In a significant proportion of melanoma patients, CTL specific for the melan-A26/7–35 epitope can be detected in peripheral blood using HLA-A2/peptide tetramers. However, the functional capacity of these CTL has been controversial, since although they prove to be effective killers after in vitro expansion, in some patients they have blunted activation responses ex vivo. We used phenotypic markers to characterize melan-A tetramer+ cells in both normal individuals and melanoma patients, and correlated these markers with ex vivo assays of CTL function. Melanoma patients with detectable melan-A tetramer+ cells in peripheral blood fell into two groups. Seven of thirteen patients had a CCR7+ CD45R0 CD45RA+ phenotype, the same as that found in some healthy controls, and this phenotype was associated with a lack of response to melan-A peptide ex vivo. In the remaining six patients, melan-A tetramer+ cells were shifted toward a CCR7 CD45R0+ CD45RA phenotype, and responses to melan-A peptide could be readily demonstrated ex vivo. When lymph nodes infiltrated by melan-A-expressing melanoma cells were examined, a similar dichotomy emerged. These findings demonstrate that activation of melan-A-specific CTL occurs in only some patients with malignant melanoma, and that only patients with such active immune responses are capable of responding to Ag in ex vivo assays.

The discovery that melanomas can be recognized and killed by human CTL has allowed a number of CTL epitopes to be defined (1). These epitopes are important targets for immunotherapy, and since many of them are expressed in other tumors, there are hopes that immunotherapeutic strategies initially tested in melanoma may prove broadly applicable to other tumors.

Among the known melanoma Ags recognized by CTL, melan-A or MART-1 is probably the best studied, and it has become an extremely useful target for immunotherapy. Melan-A was originally defined in a patient with a favorable disease course after vaccination with an autologous tumor cell line, by cloning from her blood HLA-A2-restricted CTL capable of lysing this tumor cell line (2). Expression libraries of tumor cDNAs showed that the target of these CTL was melan-A, a lineage-specific molecule also expressed on normal melanocytes (2, 3). The dominant HLA-A2-restricted epitope of melan-A originally appeared to be melan- A27–35 (AAGIGILTV) (4), but subsequent experiments showed melan-A26–35 (EAAGIGILTV) is also recognized by CTL, and binds better to HLA-A2 (5). Many laboratories have since found it relatively easy to generate CTL specific for this epitope from HLA-A2+ melanoma patients (6, 7), and it therefore seemed that melan-A26/7–35 might be the most commonly recognized epitope among all the known targets of melanoma Ag-specific CTL (6). Many studies have also demonstrated that similar CTL could be generated from healthy individuals (6), although this usually required repeated priming of PBL, and was possible in fewer healthy individuals than melanoma patients (8). Nevertheless, the comparative ease with which melan-A26/7–35-specific CTL could be obtained in many healthy individuals suggested that precursor frequencies are elevated in a large percentage of the general population. One explanation proposed was that melan-A26/7–35-specific CTL may merely be cross-reactive with this epitope, having been originally primed against a microbial epitope (9). However, peptide stimulation experiments suggested that melan-A26/7–35-specific CTL in healthy subjects did not derive from the memory compartment, whereas they did in melanoma patients (10). Unfortunately, the technical difficulties of characterizing CTL activity ex vivo have imposed severe limitations on such comparisons between patients and healthy controls. Similarly, any correlation between CTL responses in different patients and their clinical features has proven technically daunting.

The development of tetrameric MHC class I/peptide complexes (11) has greatly aided the analysis of tumor-specific CTL responses. In keeping with the relative ease with which melan-A-specific CTL can be derived in vitro, tetramers based on this epitope have to date proven more useful in this work than those based on any other known melanoma epitope. After initial studies confirmed that melan-A tetramer+ CD8+ cells represented true tumoricidal CTL (12), we showed that some lymph nodes infiltrated with melan-A-expressing melanoma cells contain high frequencies of melan-A tetramer+ cells, and that these cells have phenotypic markers consistent with previous Ag exposure (13). Small populations of melan-A tetramer+ CD8+ cells could also be visualized in the peripheral blood of melanoma patients (13, 14, 15), and in some patients these cells responded to peptide Ag in vitro by proliferating, forming highly tumoricidal cell lines (12, 15). Melan-A-specific CTL clones could also be generated by directly sorting single melan-A tetramer+ CD8+ cells from melanoma patient peripheral blood (14). As expected from earlier work, melan-A tetramer+ CD8+ cells could also be detected in the PBL of some healthy individuals, and these cells were CD45RA+ CD45R0 CD28+, a phenotype suggestive of a naive rather than a memory status (16). Several melanoma patients also had tetramer+ CD8+ cells exclusively of this phenotype, while others showed mixed populations, in which CD45RA expression had been lost (13, 17), and CD45R0 had been gained on a proportion of tetramer+ cells (16). These data suggested that only a minority of circulating melan-A-specific CTL in a minority of melanoma patients had a history of Ag exposure. However, CD45 isoforms are not in themselves adequate markers of memory/naivety, since memory cells can change expression patterns in vitro (18, 19), and subpopulations of memory CTL specific for viral Ags can express converse CD45 isoforms in vivo (20, 21).

Recently, expression of the chemokine receptor CCR7, which is involved in homing to lymphoid tissue, has emerged as a useful phenotypic marker in determining the memory status of T lymphocytes, particularly when used in conjunction with CD45 isoforms (22). CCR7 is expressed only by naive cells and central memory cells, while effector memory cells and terminally differentiated effectors are CCR7 (22). CCR7 and CD45 isoform expression also seems to distinguish between the functional capacity of different populations of T lymphocytes (22). The functional capacity of circulating tumor-specific lymphocytes has become an issue of some importance, since data published recently suggested melan-A-specific CTL circulating in melanoma patients are anergic (23).

We analyzed CCR7 and CD45 isoform expression on melan-A tetramer+ cells in the peripheral blood of melanoma patients and normal controls, taking advantage of a modified tetramer-staining protocol, which improves the specificity and intensity of tetramer staining (24), and minimizes sample manipulation before assay. The results of these phenotyping experiments were then compared with assays of CTL function ex vivo, and further extended by analysis of tumor-infiltrated lymph nodes (TILN)3 using similar techniques.

The HLA-A2+ lymphoblastoid cell line T2 was maintained in RPMI 1640 medium with 10% FCS. The melanoma cell line SK-mel-29 was maintained in Dulbecco’s modified medium with 10% FCS (D10). FACS buffer was PBS with 1% FCS. Cell culture media for patient samples were Iscove’s medium plus 5% human serum (I5) or I5 plus 100 U/ml human rIL-2 (Chiron, Emeryville, CA).

Healthy subjects were blood donors, registered with the U.K. National Blood Service. HLA-A2-negative healthy donors were used to define background levels of nonspecific binding by HLA-A2 tetramers. All other subjects discussed in the text, including healthy donors and melanoma patients, were HLA-A2+ by PCR. Clinical characteristics of each melanoma patient appear in Table I. All PBL obtained were cryopreserved immediately after separation, and analyzed after thawing and culturing briefly in I5 to allow metabolic recovery. TILN were surgically resected, and were mechanically disrupted before cryopreservation as single cell suspensions. On thawing, the numbers of tumor cells were counted as well as the number of lymphocytes, to provide an index of the extent of tumor infiltration, expressed as a percentage of the total cell count accounted for by tumor cells. Expression of melan-A was examined in TILN by RT-PCR (2), and protein expression confirmed in one sample by Western blotting using the primary Ab A103 (Novocastra, Newcastle, U.K.) (25). When tumor samples were available from patients, these were examined by immunohistochemistry, including staining for melan-A with the primary Ab A103 (Novocastra) (25). A tumor cell line was established from patient M2 by mechanically disrupting a surgically excised skin metastasis and culturing in D10 medium. This cell line was subjected to FACS analysis after incubation for 72 h with or without 100 U/ml IFN-γ and 1 ng/ml TNF-α.

Table I.

Melanoma patient characteristics

PatientSexAgeYears Since PrimaryaStagebSites of MetastasisPrior TherapyPrior Immunotherapy
M1 64 IV Lymph nodes, skin Surgery, chemotherapy – 
M2 65 IV Lymph nodes, adrenal Surgery, chemotherapy – 
M3 47 0.5 IV Lymph nodes, skin, liver, lungs, bone Surgery, chemotherapy – 
M4 68 IV Lymph nodes, adrenal Surgery, chemotherapy IFN-γ 
M5 58 IV Lymph nodes, skin, lungs Surgery, chemotherapy IFN-γ 
M6 58 0.7 III Lymph nodes Surgery, chemotherapy – 
M7 26 IIA – – – 
M8 59 IA – – – 
M9 43 III Lymph nodes Surgery, chemotherapy – 
M10 67 IIA – – – 
M11 59 IIB – – – 
M12 49 1.5 III Lymph nodes Surgery, chemotherapy – 
M13 55 1.7 IV Lymph nodes, lungs Surgery, chemotherapy – 
M14 78 0.5 IIB – Surgery – 
M15 83 III Lymph nodes, local Surgery, chemotherapy, laser therapy IFN-γ 
M16 48 14 IV Lymph nodes, lungs Surgery, chemotherapy – 
M17 59 15 IV Lymph nodes, skin Surgery, chemotherapy, radiotherapy IFN-α 
M18 53 IV Lymph nodes, chest wall Surgery, chemotherapy – 
M19 54 III Lymph nodes Surgery, chemotherapy – 
M20 56 10 IV Lymph nodes, skin Surgery, chemotherapy, radiotherapy – 
M21 65 0.3 IV Liver Surgery – 
M22 40 0.7 III Lymph node Surgery IFN-α 
M23 46 IB – – – 
M24 29 IB – – – 
M25 51 IB – – – 
M26 49 III Lymph node – – 
M27 47 0.5 III Lymph nodes Surgery, chemotherapyc – 
M28 51 III Lymph nodes Surgery – 
M29 55 0.6 III Lymph nodes Surgery – 
M30 43 1.6 III Lymph nodes Surgery – 
PatientSexAgeYears Since PrimaryaStagebSites of MetastasisPrior TherapyPrior Immunotherapy
M1 64 IV Lymph nodes, skin Surgery, chemotherapy – 
M2 65 IV Lymph nodes, adrenal Surgery, chemotherapy – 
M3 47 0.5 IV Lymph nodes, skin, liver, lungs, bone Surgery, chemotherapy – 
M4 68 IV Lymph nodes, adrenal Surgery, chemotherapy IFN-γ 
M5 58 IV Lymph nodes, skin, lungs Surgery, chemotherapy IFN-γ 
M6 58 0.7 III Lymph nodes Surgery, chemotherapy – 
M7 26 IIA – – – 
M8 59 IA – – – 
M9 43 III Lymph nodes Surgery, chemotherapy – 
M10 67 IIA – – – 
M11 59 IIB – – – 
M12 49 1.5 III Lymph nodes Surgery, chemotherapy – 
M13 55 1.7 IV Lymph nodes, lungs Surgery, chemotherapy – 
M14 78 0.5 IIB – Surgery – 
M15 83 III Lymph nodes, local Surgery, chemotherapy, laser therapy IFN-γ 
M16 48 14 IV Lymph nodes, lungs Surgery, chemotherapy – 
M17 59 15 IV Lymph nodes, skin Surgery, chemotherapy, radiotherapy IFN-α 
M18 53 IV Lymph nodes, chest wall Surgery, chemotherapy – 
M19 54 III Lymph nodes Surgery, chemotherapy – 
M20 56 10 IV Lymph nodes, skin Surgery, chemotherapy, radiotherapy – 
M21 65 0.3 IV Liver Surgery – 
M22 40 0.7 III Lymph node Surgery IFN-α 
M23 46 IB – – – 
M24 29 IB – – – 
M25 51 IB – – – 
M26 49 III Lymph node – – 
M27 47 0.5 III Lymph nodes Surgery, chemotherapyc – 
M28 51 III Lymph nodes Surgery – 
M29 55 0.6 III Lymph nodes Surgery – 
M30 43 1.6 III Lymph nodes Surgery – 
a

Years elapsed between excision of primary lesion and sample studied.

b

American Joint Committee on Cancer staging system.

c

, Two TILN were obtained from M27, one pre- and one post-chemotherapy.

Peptides synthesized by FMOC chemistry were: melan-A, ELAGIGILTV, a variant of the 26–35 epitope, which binds better to HLA-A2 than the natural peptide due to an altered anchor residue, but which is recognized by the same CTL as the natural peptide (26); influenza, GILGFVFTL, the influenza matrix protein 58–66 epitope (27); EBV, GLCTLVAML, the 280–288 epitope from the lytic protein BMLF1 of EBV (28, 29).

Tetramers were all made from HLA-A2.1 heavy chain, and the peptides above, and are referred to in the text simply as melan-A, influenza, and EBV tetramers. HLA-A2/peptide complexes were synthesized as previously described (11, 12, 30). Briefly, the HLA-A2.1 heavy chain cDNA was modified by substitution of the trans-membrane and cytosolic regions with a sequence encoding the BirA biotinylation enzyme recognition site. This modified HLA-A2.1 and β2-microglobulin were synthesized in a prokaryotic expression system (pET; R&D Systems, Minneapolis, MN), purified from bacterial inclusion bodies, and allowed to refold with the relevant peptide by dilution. Refolded complexes were purified by FPLC and biotinylated using BirA (Avidity, Denver, CO), then combined with PE-labeled streptavidin (Sigma, St. Louis, MO) at a 4:1 molar ratio to form tetramers. Tetramers were titrated against appropriate CTL clones to determine the dose that induced maximal staining (12, 14).

Cells were stained with the appropriate PE-labeled tetramer at 37°C for 20 min before washing in FACS buffer at 37°C, and incubating for 30 min on ice with Abs, including Tricolor anti-CD8α (Caltag, Burlingame, CA) or peridinin chlorophyl protein anti-CD8α (Becton Dickinson, Mountain View, CA), FITC anti-CD45RA (Dako, Glostrup, Denmark), and APC anti-CD45R0 (Becton Dickinson). For CCR7 staining, the rat mAb 3D12 (22) was incubated for 30 min before two washes in FACS buffers and incubation with FITC mouse anti-rat IgG1/2a (PharMingen, San Diego, CA) along with the other Abs. After extensive washes in ice-cold FACS buffer, cells were kept on ice without fixation and analyzed on a Becton Dickinson FACScalibur using CellQuest software. For analysis, small lymphocytes were gated according to forward/side scatter profiles. Tetramer+ CD8+ were defined as CD8high cells with a PE fluorescence of at least 100 fluorescence units (while CD8+ cells in the same sample unstained with tetramer had a maximum fluorescence of 10 fluorescence units). CD8low cells were excluded from analysis because most CD8low cells stained with anti-CD8α Abs were CD8α+ CD8β CD3 CD45RA+ (data not shown). Frequencies of melan-A tetramer-positive cells were averaged from between two and six replicate stainings. Minimal melan-A tetramer binding was observed on HLA-A2 CD8+ PBL (=5 cells/105 CD8+ cells), but in view of this slight background staining, HLA-A2+ samples containing less than 20 melan-A tetramer+ cells per 105 CD8+ cells were considered negative for melan-A tetramer staining, and excluded from further analysis.

FACS analysis of MHC class I and HLA-A2 expression on cell lines was performed using the mAbs W6/32 and BB7.2, respectively (American Type Culture Collection, Manassas, VA).

Where sufficient PBL were available, ELISPOT analysis for IFN-γ secretion was performed according to the protocol provided by the manufacturer (Mabtech, Stockholm, Sweden). PBL were thawed and cultured overnight in I5 medium to ensure good viability before assay, and plated in duplicate at up to 5 × 105 cells/well. In some experiments, CD8+ cells were enriched from PBL using immunomagnetic separation (MACS; Miltenyi Biotec, Bergsich Gladbach, Germany) before assay. Peptide was added at 10 μM to appropriate wells, and cells incubated for 40 h before development. Results are presented normalized for CD8 counts in each PBL sample (determined by FACS), and represent mean values for each peptide after background (no peptide) has been subtracted. ELISPOT analysis of TILN was conducted in a similar fashion, except that CD8+ cells were separated away from tumor cells by immunomagnetic selection (MACS). CD8+ cells were cultured in I5 medium for 48 h before assay, then 4000 CD8+ cells were plated in duplicate ELISPOT wells containing T2 cells pulsed or not pulsed with 10 μM melan-A peptide.

A total of 106 PBL from healthy donors and melanoma patients with detectable melan-A tetramer+ CD8+ cells was cultured at 37°C in I5 plus 100 U/ml human rIL-2 for 7 days after addition of 10 μM melan-A peptide or no peptide, before harvesting and staining with tetramer and Abs, as described above.

Melan-A tetramer+ cells could be detected in the peripheral blood of 5 of 10 healthy blood donors (Fig. 1). These cells were CCR7+ CD45R0 (Fig. 2, a and d). In contrast, influenza and EBV tetramer+ (memory) cells in healthy blood donors were never CCR7+ CD45R0 (data not shown). Melan-A tetramer+ CD45R0 cells were also CD45RA+ (Fig. 2,g), so that the full phenotype of these cells was CCR7+ CD45RA+ CD45R0, consistent with the naive phenotype proposed in recent work (22). ELISPOT analysis revealed no detectable secretion of IFN-γ from the PBL of these healthy donors in response to melan-A peptide (Fig. 3), even when 5 × 105 enriched CD8+ cells were plated per well (Fig. 4), and viral recall responses could readily be visualized (Fig. 4). Hence, the CCR7+ CD45R0 phenotype in melan-A tetramer+ cells in normal donors is associated with a lack of recall response to peptide Ag.

FIGURE 1.

Frequencies of melan-A tetramer+ CD8+ cells in peripheral blood. PBL from melanoma patients (▪ and ▨) and healthy controls (□) were stained directly with melan-A tetramer and anti-CD8, and frequencies of tetramer-high cells calculated as a proportion of total CD8 cells. Subsequent phenotyping of melan-A tetramer+ cells showed the melanoma patients could be divided into two groups: group B (▨) had the same phenotype as healthy controls (CCR7+ CD45R0), while group A (▪) had circulating melan-A tetramer+ cells that were CCR7 CD45R0+.

FIGURE 1.

Frequencies of melan-A tetramer+ CD8+ cells in peripheral blood. PBL from melanoma patients (▪ and ▨) and healthy controls (□) were stained directly with melan-A tetramer and anti-CD8, and frequencies of tetramer-high cells calculated as a proportion of total CD8 cells. Subsequent phenotyping of melan-A tetramer+ cells showed the melanoma patients could be divided into two groups: group B (▨) had the same phenotype as healthy controls (CCR7+ CD45R0), while group A (▪) had circulating melan-A tetramer+ cells that were CCR7 CD45R0+.

Close modal
FIGURE 2.

Phenotypes of melan-A tetramer+ CD8+ cells in peripheral blood. PBL from melanoma patients and healthy controls were quadruple stained with melan-A tetramer and anti-CD8, along with either anti-CD45R0 and anti-CCR7 (d–f) or anti-CD45R0 and anti-CD45RA (g and h). Three representative samples are shown, gated on lymphocytes to show the size of the tetramer+ CD8+ population (a–c), or gated on tetramer+ CD8+ cells to show the phenotype of these cells only (d–h).

FIGURE 2.

Phenotypes of melan-A tetramer+ CD8+ cells in peripheral blood. PBL from melanoma patients and healthy controls were quadruple stained with melan-A tetramer and anti-CD8, along with either anti-CD45R0 and anti-CCR7 (d–f) or anti-CD45R0 and anti-CD45RA (g and h). Three representative samples are shown, gated on lymphocytes to show the size of the tetramer+ CD8+ population (a–c), or gated on tetramer+ CD8+ cells to show the phenotype of these cells only (d–h).

Close modal
FIGURE 3.

Frequencies of IFN-γ-secreting CD8+ cells compared with tetramer+ CD8+ cells. PBL from melanoma patients (filled symbols) and healthy controls (○) were stained with melan-A tetramer and anti-CD8, and frequencies of tetramer+ cells as a proportion of total CD8 cells were calculated (x-axis). The same PBL were tested in ELISPOT assay for secretion of IFN-γ in response to melan-A peptide, and spot-forming cells calculated as a proportion of total CD8 cells (y-axis). a, PBL from group A melanoma patients (▪) showed a correlation between spot-forming cells and tetramer+ cells. b, PBL from healthy controls (○) and group B melanoma patients (▴) showed no spot-forming cells, despite detectable melan-A tetramer+ cells. Two group A melanoma patients (also shown on the larger scale in a) are shown for comparison.

FIGURE 3.

Frequencies of IFN-γ-secreting CD8+ cells compared with tetramer+ CD8+ cells. PBL from melanoma patients (filled symbols) and healthy controls (○) were stained with melan-A tetramer and anti-CD8, and frequencies of tetramer+ cells as a proportion of total CD8 cells were calculated (x-axis). The same PBL were tested in ELISPOT assay for secretion of IFN-γ in response to melan-A peptide, and spot-forming cells calculated as a proportion of total CD8 cells (y-axis). a, PBL from group A melanoma patients (▪) showed a correlation between spot-forming cells and tetramer+ cells. b, PBL from healthy controls (○) and group B melanoma patients (▴) showed no spot-forming cells, despite detectable melan-A tetramer+ cells. Two group A melanoma patients (also shown on the larger scale in a) are shown for comparison.

Close modal
FIGURE 4.

ELISPOT analysis of IFN-γ secretion by enriched CD8+ cells. PBL from a melanoma patient (M1) and a healthy control with detectable melan-A tetramer+ CD8+ cells (N1) were enriched for CD8+ cells by immunomagnetic sorting. A total of 5 × 105 CD8+ cells/well was then plated in duplicate into an ELISPOT assay for secretion of IFN-γ, with or without melan-A peptide or influenza matrix peptide at 10 μM. The assay was developed after a 40-h incubation.

FIGURE 4.

ELISPOT analysis of IFN-γ secretion by enriched CD8+ cells. PBL from a melanoma patient (M1) and a healthy control with detectable melan-A tetramer+ CD8+ cells (N1) were enriched for CD8+ cells by immunomagnetic sorting. A total of 5 × 105 CD8+ cells/well was then plated in duplicate into an ELISPOT assay for secretion of IFN-γ, with or without melan-A peptide or influenza matrix peptide at 10 μM. The assay was developed after a 40-h incubation.

Close modal

Peripheral blood from 26 patients with malignant melanoma (whose clinical characteristics are provided in Table I) was examined with melan-A tetramers. Melan-A tetramer+ cells were repeatably detected on ex vivo staining in 13 of the 26 patients (Fig. 1). In contrast, tetramers made from other melanoma epitopes only rarely detected tumor-specific CTL in peripheral blood without Ag stimulation (data not shown). The frequency of melan-A tetramer+ cells in melanoma patients was often similar to those found in normal subjects (Fig. 1), with only four patients showing more frequent melan-A-specific CTL than the normal controls. When the phenotype of these cells was studied, the patients fell into two distinct groups. In one group (group B), comprising 7 of the 13 patients with detectable CTL, melan-A tetramer+ cells were CCR7+ CD45R0 (Fig. 2, b and e), the same phenotype as in the healthy controls. In contrast, influenza and EBV tetramer+ cells, where detected in these patients, were never CCR7+ CD45R0 (data not shown). In the other group (group A), comprising 6 of 13 patients with detectable CTL, a substantial proportion of melan-A tetramer+ cells was CCR7 CD45R0+ (Fig. 2, c and f). Tetramer+ CD45R0+ cells in group A patients were also CD45RA (Fig. 2,h), so that these patients had substantial populations of CCR7 CD45RA CD45R0+ cells, the converse phenotype of those found in group B patients and healthy controls. However, in all group A patients, some melan-A tetramer+ cells still displayed the phenotype typical of group B patients and healthy controls (Fig. 2, c, f, and h), while CCR7+ CD45R0+ cells were rare. Hence, the distinguishing characteristic of group A is the appearance of CCR7 CD45RA CD45R0+ cells, among the melan-A tetramer+ population. The percentage of CCR7 CD45R0+ cells among the melan-A tetramer+ cells in group A patients ranged from 24 to 89%, with the highest phenotypic shifts seen in those with the highest frequencies of melan-A tetramer+ cells.

In ELISPOT assay, only patients from group A were capable of producing IFN-γ in response to melan-A peptide (Fig. 3). CD8 enrichment allowed substantial numbers of melan-A-reactive cells to be visualized in patients from group A, but did not result in any melan-A-specific signal from subjects with melan-A tetramer+ cells bearing an exclusively CCR7+ CD45RA+ CD45R0 phenotype (Fig. 4). Comparison of the frequencies of ELISPOT-reactive cells with melan-A tetramer+ cells in group A revealed that on average, one-third of tetramer+ cells were detected in ELISPOT (Fig. 3). This discrepancy is very similar to that reported for influenza-specific CTL (16). To confirm that melan-A tetramer+ cells in group A patients were capable of functional responses to Ag, PBL were cultured in the presence of IL-2 for 7 days after addition of melan-A peptide. Melan-A tetramer+ cells from group A patients expanded rapidly under these experimental conditions (Fig. 5).

FIGURE 5.

Tetramer analysis of in vitro expansion in response to melan-A peptide. PBL from a melanoma patient (M2) were plated at 1 × 106 cells/well, with or without 10 μM melan-A peptide, then cultured in the presence of IL-2 for 7 days. Cells were then harvested and stained with melan-A tetramer and anti-CD8, and frequencies of tetramer+ CD8+ cells in each culture were calculated as a percentage of the total CD8+ cells.

FIGURE 5.

Tetramer analysis of in vitro expansion in response to melan-A peptide. PBL from a melanoma patient (M2) were plated at 1 × 106 cells/well, with or without 10 μM melan-A peptide, then cultured in the presence of IL-2 for 7 days. Cells were then harvested and stained with melan-A tetramer and anti-CD8, and frequencies of tetramer+ CD8+ cells in each culture were calculated as a percentage of the total CD8+ cells.

Close modal

TILN from four patients (described in Table I) were examined, including two TILN from the same patient. The TILN had variable percentages of infiltrating tumor cells, as shown in Fig. 6 (M30 was 9%). Melan-A expression was confirmed in all TILN samples by RT-PCR, and in TILN from patient M29 by Western blot. Melan-A tetramer+ CD8+ cells were also detectable in all samples. Two TILN, from patients M29 and M30, showed frequencies of melan-A tetramer+ cells similar to those found in healthy controls and group B patient PBL, less than 100 cells per 105 CD8+ cells (Fig. 6,d, and data not shown). Melan-A tetramer+ CD8+ cells in these TILN had a CCR7+ CD45R0 phenotype (Fig. 6,h), identical with PBL from healthy controls and group B melanoma patients. Patients M27 and M28 had melan-A tetramer+ cells that were not of this phenotype. Patient M28 had a melan-A tetramer+ population of 0.27% of CD8+ cells (Fig. 6,c), which were largely CCR7 CD45R0+ (Fig. 6,g), although some CCR7+ CD45R0+ cells were also present (Fig. 6,g). Patient M27, in the first TILN sample, had a massively expanded population of melan-A tetramer+ CD8+ cells compared with all other patients, equivalent to 13% of the CD8+ cells (Fig. 6,a), and the vast majority of these cells were CCR7 CD45R0+ (Fig. 6,e), although 6% were CCR7+ CD45R0+ (Fig. 6,e). A second TILN, obtained from patient M27 8 mo later, had a similar expansion of melan-A tetramer+ CD8+ cells (Fig. 6,b), although these cells were split evenly between the CCR7 CD45R0+ and CCR7+ CD45R0+ phenotypic patterns (Fig. 6,f). Perhaps most strikingly, in neither of these TILN, in which expansion of melan-A-specific CTL had clearly been stimulated, did significant numbers of these CTL bear the CCR7+ CD45R0 phenotype seen in healthy PBL (Fig. 6, e and f). Hence, in the TILN, as in the PBL, melan-A tetramer+ cells had strikingly different phenotypes in different patients, with some patients showing exclusively a CCR7+ CD45R0 phenotype, and some patients showing a shift to the converse phenotype. This shift in phenotype was associated with a slightly greater level of tumor infiltration in the small number of samples investigated in this study (Fig. 6).

FIGURE 6.

Phenotypes of melan-A tetramer+ CD8+ cells in TILN. TILN from three melanoma patients (including two different nodes from patient M27) were quadruple stained with melan-A tetramer, anti-CD8, anti-CD45R0, and anti-CCR7. a–d, Scatter plots of tetramer and CD8 stains (gated on small lymphocytes by forward/side scatter profile) show the size of the tetramer+ CD8+ population, as indicated. e–h, Scatter plots gated on tetramer+ CD8+ cells show the phenotype of these cells with respect to expression of CCR7 and CD45R0. The proportion of the cells in each sample accounted for by tumor cells is shown above each pair of plots, as an index of tumor infiltration.

FIGURE 6.

Phenotypes of melan-A tetramer+ CD8+ cells in TILN. TILN from three melanoma patients (including two different nodes from patient M27) were quadruple stained with melan-A tetramer, anti-CD8, anti-CD45R0, and anti-CCR7. a–d, Scatter plots of tetramer and CD8 stains (gated on small lymphocytes by forward/side scatter profile) show the size of the tetramer+ CD8+ population, as indicated. e–h, Scatter plots gated on tetramer+ CD8+ cells show the phenotype of these cells with respect to expression of CCR7 and CD45R0. The proportion of the cells in each sample accounted for by tumor cells is shown above each pair of plots, as an index of tumor infiltration.

Close modal

ELISPOT assay was performed on TILN1 from patient M27 to test whether the expanded population of melan-A tetramer+ cells (13% of total CD8+ cells) was able to release IFN-γ in response to peptide Ag. To minimize contamination of the assay with melan-A-expressing tumor cells, CD8+ cells were separated from tumor cells, and rested in tissue culture for 2 days. In subsequent ELISPOT assay, secretion of IFN-γ in response to melan-A peptide was detected in 4.5% of the CD8+ cells plated (Fig. 7), a similar proportion of tetramer+ cells to that detected in PBL.

FIGURE 7.

ELISPOT analysis of IFN-γ secretion by enriched CD8+ cells from melanoma-infiltrated lymph nodes. CD8+ lymphocytes from a melanoma-infiltrated lymph node (TILN1 from patient M27) were enriched by immunomagnetic sorting, and rested for 48 h in tissue culture. Four thousand cells/well were then plated in duplicate into an ELISPOT assay for secretion of IFN-γ, with T2 cells pulsed or not pulsed with 10 μM melan-A peptide. The assay was developed after 40-h incubation.

FIGURE 7.

ELISPOT analysis of IFN-γ secretion by enriched CD8+ cells from melanoma-infiltrated lymph nodes. CD8+ lymphocytes from a melanoma-infiltrated lymph node (TILN1 from patient M27) were enriched by immunomagnetic sorting, and rested for 48 h in tissue culture. Four thousand cells/well were then plated in duplicate into an ELISPOT assay for secretion of IFN-γ, with T2 cells pulsed or not pulsed with 10 μM melan-A peptide. The assay was developed after 40-h incubation.

Close modal

Patient M1, who had the highest level of melan-A-specific CTL measured by both tetramer staining and ELISPOT, had tissue available from a sternal metastasis for examination by immunohistochemistry. Although the tumor cells expressed melan-A strongly, and a CD8+ lymphocytic infiltrate was present, MHC class I expression was not detected on tumor cells (data not shown). From patient M2, who had the second strongest melan-A-specific CTL response, a resected skin metastasis was cultured in vitro, and a tumor cell line was established. This cell line proved to have a selective loss of HLA-A2 (Fig. 8). Interestingly, this cell line exhibited substantial expression of total MHC class I, indicating that at least one allele other than HLA-A2 was still expressed (Fig. 8). Treatment with IFN-γ up-regulated total expression of MHC class I slightly, but did not restore HLA-A2 expression (Fig. 8).

FIGURE 8.

Expression of HLA-A2 and total MHC class I on melanoma lines. The melanoma cell line M2, derived from patient M2, was stained for surface expression of HLA-A2 (mAb BB7.2) and total MHC class I (mAb W6/32) and analyzed by FACS, with and without prior treatment with IFN-γ and TNF-α. The HLA-A2+ melanoma cell line SK-mel-29 served as a positive control.

FIGURE 8.

Expression of HLA-A2 and total MHC class I on melanoma lines. The melanoma cell line M2, derived from patient M2, was stained for surface expression of HLA-A2 (mAb BB7.2) and total MHC class I (mAb W6/32) and analyzed by FACS, with and without prior treatment with IFN-γ and TNF-α. The HLA-A2+ melanoma cell line SK-mel-29 served as a positive control.

Close modal

In this study, we used optimized tetramer-staining techniques to study melan-A-specific CTL directly ex vivo, to correlate phenotypic markers with functional parameters.

Circulating CTL specific for melan-A were detectable ex vivo in 13 of 26 patients with malignant melanoma and 5 of 10 healthy controls. These percentages of tetramer positivity in PBL are far higher than for other tumor epitopes studied to date with tetramers (31, 32, 33), in which only rare patients have circulating tumor-specific CTL detectable ex vivo. Given that so many normal subjects also had melan-A tetramer+ cells, the high percentage of positivity in melanoma patients is not surprising, and clearly may relate to a higher precursor frequency in many HLA-A2+ individuals rather than a response to the melanoma-associated Ag. Four patients with melanoma (M1–4) had a higher frequency of melan-A tetramer+ cells than any normal control, suggesting they might have some degree of Ag-specific CTL expansion. However, it is important to note that even the highest tetramer+ frequency seen in this patient series (0.3% of CD8+ cells) is relatively low compared with those seen in acute viral infections (21, 30).

We used expression of the chemokine receptor CCR7 alongside CD45 isoforms to analyze the phenotype of melan-A tetramer+ CD8+ cells, to resolve questions about the nature of these cells that could not be answered using CD45 isoform expression pattern alone. The CD45R0 CD45RA+ phenotype classically ascribed to naive cells suffers from the ability of memory cells such as EBV-specific memory and effector CTL to revert to this phenotype (20, 21). Hence, new markers of memory status are needed, and CCR7 is clearly one such candidate (22). In control experiments, we found no EBV- or influenza-specific CTL with a CCR7+ CD45R0 phenotype, suggesting that memory CTL do not bear this combination of markers. This CCR7+ CD45R0 phenotype was found on melan-A tetramer+ cells from both healthy controls and some melanoma patients (group B), confirming that the melan-A-specific CTL precursors detected in these individuals are unlikely to be memory CTL.

Strikingly, in TILN, in which expansions of melan-A-specific CTL had clearly taken place (Fig. 6), this CCR7+ CD45R0 phenotype was entirely absent from melan-A tetramer+ cells. Hence, conversion away from this phenotype is associated with strong in vivo proliferative CTL responses against the tumor, and Ag-specific effector function that is measurable ex vivo. In the PBL of six patients, melan-A tetramer+ cells that no longer expressed the CCR7+ CD45RA+ CD45R0 phenotype could be found, suggesting an active immune response against melan-A had occurred in these patients. Indeed, in functional assays of recall responses, these were the patients who responded to melan-A peptide, demonstrating that these CTL were not anergic. Therefore, in both PBL and TILN, loss of the CCR7+ CD45RA+ CD45R0 phenotype was associated with a functional immune response against melan-A.

During such a response, the phenotype of melan-A tetramer+ cells in the PBL typically shifted to CCR7 CD45RA CD45R0+. Cells with this phenotype have been termed effector memory cells (22) and are expected to have effector function, including the ability to secrete IFN-γ (22), as detected in our assays. Populations of melan-A-specific CTL that fit the proposed central memory population (CCR7+ CD45RA CD45R0+) were only rarely seen in peripheral blood (Fig. 2, and data not shown). However, in TILN, CCR7+ CD45R0+ cells with a putative central memory phenotype were detected (Fig. 6), albeit as smaller populations than the effector memory cells (CCR7 CD45R0+). Interestingly, of two TILN taken from the same patient 8 mo apart, it was at the later time point that the most central memory cells were found among the melan-A-specific CTL.

Our data suggest that many melan-A-specific CTL that appear anergic ex vivo may in fact be naive or unprimed, because they are phenotypically identical with those found in many normal subjects (CCR7+ CD45RA+ CD45R0). Functional assays that depend on rapid response to peptide Ag do not detect these cells, because they probably require prolonged stimulation to become activated (34). Hence, we failed to detect secretion of IFN-γ in response to peptide when samples contained such naive phenotype CTL precursors, while this assay readily detected the presence of Ag-experienced (CCR7 CD45RA CD45R0+) CTL. Our data suggest that brisk effector function can be readily detected ex vivo provided there is evidence of CTL priming.

In some patients with malignant melanoma (group B), melan-A tetramer+ cells were readily detected, but there was no phenotypic evidence that these cells had ever responded to Ag. This phenomenon was observed even in TILN, in which melan-A-expressing tumor cells were effectively adjacent to melan-A CTL precursors (patients M29 and M30). It therefore seems likely that these patients are in a state analogous to the peripheral ignorance proposed from some animal tumor models (35, 36), in which tumor-specific CTL are present both in the circulation and the lymphoid tissue, but have not been primed by tumor Ag. As noted above, unprimed CTL were seen even in the circulation of patients who had managed to prime at least a proportion of their melan-A-specific CTL. Hence, even in the circulation of melanoma patients with active immunity against melan-A, priming of CTL may not be complete. In fact, only in those TILN in which expansion of melan-A-specific CTL had occurred, had all melan-A tetramer+ cells converted away from the naive CCR7+ CD45R0 phenotype (Fig. 6). All these data suggest priming of melan-A-specific CTL by malignant melanoma is weak. This should perhaps not be unduly surprising, because tumor cells usually lack costimulatory molecules (37, 38), and tumor Ags have no direct route into the MHC class I pathway of professional APCs.

The fact that only some patients primed their melan-A CTL precursors begs the question as to whether any factors can be distinguished that might correlate with CTL priming. It is tempting to suggest that priming of melanoma Ag-specific CTL is a late phenomenon, because only stage III or stage IV patients (patients M1-6, M27-8) had the group A phenotype indicating an active immune response. Conversely, four patients with no clinical evidence of metastatic disease (M7, M8, M10, M11) had the group B unprimed phenotype, suggesting substantial metastatic disease may be required for CTL priming. As noted above, however, many patients with metastases also showed no evidence of priming melan-A-specific CTL, so while metastatic disease may be required for priming CTL, it may not in itself be sufficient. Indeed, the current data suggest that the cross-presentation events postulated to explain the priming of tumor-specific CTL in humans (39, 40) may be quite rare.

When clinical parameters in group A and group B patients were compared, no striking correlation with disease progression was evident. Indeed, further study will be necessary to determine whether the character of the melan-A-specific CTL response has any impact on clinical course, because we did not follow patients longitudinally or study patients at equivalent disease stages. Nevertheless, the clinical features of group A patients offer important clues as to the possible impact of an active immune response against melan-A.

Most importantly, the two patients from group A with the strongest melan-A-specific CTL responses both died within 1 yr of developing metastatic disease. Hence, active immunity to melan-A, comprising activated Ag-experienced melan-A-specific CTL circulating at numbers of up to 400 cells/ml of blood, did not protect these patients from succumbing to their tumors. If priming of melanoma Ag-specific CTL is a relatively late event, then it is possible that these patients simply generated a melan-A-specific CTL response too late to show substantial clinical benefit. The loss of HLA-A2 expression observed in some melanoma cells from these patients would have rendered these cells impervious to the effects of the HLA-A2-restricted melan-A-specific CTL, even in the presence of IFN-γ and TNF-α (in the case of patient M2). The maintenance of total MHC class I expression in some of these cells (Fig. 8), presumably due to continued expression of at least one other MHC class I allele, may also have allowed these cells to escape attack by NK cells. Although similar antigenic loss variants have previously been reported in melanoma (41, 42), in this study we confirm that this phenomenon can occur in patients with functional Ag-specific CTL detectable ex vivo. It is possible that the pressure of melan-A CTL attack in patient M2 may have played a role in selecting the tumor variant we propagated in tissue culture. However, it is important to note that MHC class I loss variants are also likely to be present in patients without an active immune response against melan-A, so our data do not necessarily support the concept of CTL pressure selecting tumor variants in vivo.

These findings have important implications for immunotherapy of melanoma. If many melanoma patients are failing to prime their melan-A-specific CTL, then clearly immunogens that are capable of priming these CTL may be clinically helpful. The fact that CTL precursors for the melan-A26/7–35 epitope seem to be present at relatively high frequencies in such a large proportion of HLA-A2+ individuals is encouraging, because it may prove easier to generate a sizeable CTL response against this epitope than for epitopes in which precursor CTL are rare. In contrast, CTL responses against melan-A26/7–35 are clearly not protective in late disease, at least in part due to the development of tumor escape loss variants. Even so, it is clear that a sound immunotherapeutic strategy in melanoma may be to prime melan-A26/7–35-specific CTL early in disease, when tumor burden and genetic variation are at a minimum. Using polyvalent agents to induce CTL against other melanoma Ags would also seem advisable, to minimize the risk of Ag loss variants emerging. Powerful immunogens capable of reliably priming tumor-specific CTL are needed, and developing agents that are also simple to store and administer will greatly assist their wider use in early disease.

We thank Pru Bahl for expert technical assistance.

1

This work was supported by funding from the Cancer Research Campaign, the U.K. Medical Research Council, the Cancer Research Institute, the Imperial Cancer Research Foundation, and the Axe Immunologie des Tumeurs de La Ligue Nationale contre le Cancer.

3

Abbreviations used in this paper: TILN, tumor-infiltrated lymph node; ELISPOT, enzyme-linked immunospot.

1
Boon, T., L. J. Old.
1997
. Cancer tumor antigens.
Curr. Opin. Immunol.
9
:
681
2
Coulie, P. G., V. Brichard, P. A. Van, T. Wolfel, J. Schneider, C. Traversari, S. Mattei, E. De Plaen, C. Lurquin, J. P. Szikora.
1994
. A new gene coding for a differentiation antigen recognized by autologous cytolytic T lymphocytes on HLA-A2 melanomas.
J. Exp. Med.
180
:
35
3
Kawakami, Y., S. Eliyahu, C. H. Delgado, P. F. Robbins, L. Rivoltini, S. L. Topalian, T. Miki, S. A. Rosenberg.
1994
. Cloning of the gene coding for a shared human melanoma antigen recognized by autologous T cells infiltrating into tumor.
Proc. Natl. Acad. Sci. USA
91
:
3515
4
Kawakami, Y., S. Eliyahu, K. Sakaguchi, P. F. Robbins, L. Rivoltini, J. R. Yannelli, E. Appella, S. A. Rosenberg.
1994
. Identification of the immunodominant peptides of the MART-1 human melanoma antigen recognized by the majority of HLA-A2-restricted tumor infiltrating lymphocytes.
J. Exp. Med.
180
:
347
5
Romero, P., N. Gervois, J. Schneider, P. Escobar, D. Valmori, C. Pannetier, A. Steinle, T. Wolfel, D. Lienard, V. Brichard, et al
1997
. Cytolytic T lymphocyte recognition of the immunodominant HLA-A*0201-restricted Melan-A/MART-1 antigenic peptide in melanoma.
J. Immunol.
159
:
2366
6
Rivoltini, L., Y. Kawakami, K. Sakaguchi, S. Southwood, A. Sette, P. F. Robbins, F. M. Marincola, M. L. Salgaller, J. R. Yannelli, E. Appella.
1995
. Induction of tumor-reactive CTL from peripheral blood and tumor-infiltrating lymphocytes of melanoma patients by in vitro stimulation with an immunodominant peptide of the human melanoma antigen MART-1.
J. Immunol.
154
:
2257
7
Stevens, E. J., L. Jacknin, P. F. Robbins, Y. Kawakami, M. el Gamil, S. A. Rosenberg, J. R. Yannelli.
1995
. Generation of tumor-specific CTLs from melanoma patients by using peripheral blood stimulated with allogeneic melanoma tumor cell lines: fine specificity and MART-1 melanoma antigen recognition.
J. Immunol.
154
:
762
8
Marincola, F. M., L. Rivoltini, M. L. Salgaller, M. Player, S. A. Rosenberg.
1996
. Differential anti-MART-1/MelanA CTL activity in peripheral blood of HLA-A2 melanoma patients in comparison to healthy donors: evidence of in vivo priming by tumor cells.
J. Immunother. Emphasis Tumor Immunol.
19
:
266
9
Loftus, D. J., C. Castelli, T. M. Clay, P. Squarcina, F. M. Marincola, M. I. Nishimura, G. Parmiani, E. Appella, L. Rivoltini.
1996
. Identification of epitope mimics recognized by CTL reactive to the melanoma/melanocyte-derived peptide MART-1(27–35).
J. Exp. Med.
184
:
647
10
D’Souza, S., D. Rimoldi, D. Lienard, F. Lejeune, J. C. Cerottini, P. Romero.
1998
. Circulating Melan-A/Mart-1 specific cytolytic T lymphocyte precursors in HLA-A2+ melanoma patients have a memory phenotype.
Int. J. Cancer
78
:
699
11
Altman, J. D., P. A. H. Moss, P. J. R. Goulder, D. H. Barouch, M. G. McHeyzer-Williams, J. I. Bell, A. J. McMichael, M. M. Davis.
1996
. Phenotypic analysis of antigen-specific T lymphocytes.
Science
274
:
94
12
Dunbar, P. R., G. S. Ogg, J. Chen, N. Rust, P. Van der Bruggen, V. Cerundolo.
1998
. Direct isolation, phenotyping, and cloning of low-frequency antigen-specific cytotoxic T lymphocytes from peripheral blood.
Curr. Biol.
8
:
413
13
Romero, P., P. R. Dunbar, D. Valmori, M. Pittet, G. S. Ogg, D. Rimoldi, J.-L. Chen, D. Lienard, J.-C. Cerottini, V. Cerundolo.
1998
. Ex-vivo staining of metastatic lymph nodes by class I MHC tetramers reveals high numbers of antigen-experienced tumor-specific cytotoxic T lymphocytes.
J. Exp. Med.
188
:
1641
14
Dunbar, P. R., J. L. Chen, D. Chao, N. Rust, H. Teisserenc, G. S. Ogg, P. Romero, P. Weynants, V. Cerundolo.
1999
. Cutting edge: rapid cloning of tumor-specific CTL suitable for adoptive immunotherapy of melanoma.
J. Immunol.
162
:
6959
15
Valmori, D., M. J. Pittet, D. Rimoldi, D. Lienard, R. Dunbar, V. Cerundolo, F. Lejeune, J. C. Cerottini, P. Romero.
1999
. An antigen-targeted approach to adoptive transfer therapy of cancer.
Cancer Res.
59
:
2167
16
Pittet, M. J., D. Valmori, P. R. Dunbar, D. E. Speiser, D. Lienard, F. Lejeune, K. Fleischhauer, V. Cerundolo, J. C. Cerottini, P. Romero.
1999
. High frequencies of naive Melan-A/MART-1-specific CD8(+) T cells in a large proportion of human histocompatibility leukocyte antigen (HLA)-A2 individuals.
J. Exp. Med.
190
:
705
17
Speiser, D. E., M. J. Pittet, D. Valmori, R. Dunbar, D. Rimoldi, D. Lienard, H. R. MacDonald, J. C. Cerottini, V. Cerundolo, P. Romero.
1999
. In vivo expression of natural killer cell inhibitory receptors by human melanoma-specific cytolytic T lymphocytes.
J. Exp. Med.
190
:
775
18
Rothstein, D. M., A. Yamada, S. F. Schlossman, C. Morimoto.
1991
. Cyclic regulation of CD45 isoform expression in a long term human CD4+CD45RA+ T cell line.
J. Immunol.
146
:
1175
19
Hargreaves, M., E. B. Bell.
1997
. Identical expression of CD45R isoforms by CD45RC+ ‘revertant’ memory and CD45RC+ naive CD4 T cells.
Immunology
91
:
323
20
Tan, L. C., N. Gudgeon, N. E. Annels, P. Hansasuta, C. A. O’Callaghan, S. Rowland-Jones, A. J. McMichael, A. B. Rickinson, M. F. Callan.
1999
. A re-evaluation of the frequency of CD8+ T cells specific for EBV in healthy virus carriers.
J. Immunol.
162
:
1827
21
Callan, M. F., L. Tan, N. Annels, G. S. Ogg, J. D. Wilson, C. A. O’Callaghan, N. Steven, A. J. McMichael, A. B. Rickinson.
1998
. Direct visualization of antigen-specific CD8+ T cells during the primary immune response to Epstein-Barr virus in vivo.
J. Exp. Med.
187
:
1395
22
Sallusto, F., D. Lenig, R. Forster, M. Lipp, A. Lanzavecchia.
1999
. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions.
Nature
401
:
708
23
Lee, P. P., C. Yee, P. A. Savage, L. Fong, D. Brockstedt, J. S. Weber, D. Johnson, S. Swetter, J. Thompson, P. D. Greenberg, et al
1999
. Characterization of circulating T cells specific for tumor-associated antigens in melanoma patients.
Nat. Med.
5
:
677
24
Whelan, J. A., P. R. Dunbar, D. A. Price, M. A. Purbhoo, F. Lechner, G. S. Ogg, G. Griffiths, R. E. Phillips, V. Cerundolo, A. K. Sewell.
1999
. Specificity of CTL interactions with peptide-MHC class I tetrameric complexes is temperature dependent.
J. Immunol.
163
:
4342
25
Chen, Y. T., E. Stockert, A. Jungbluth, S. Tsang, K. A. Coplan, M. J. Scanlan, L. J. Old.
1996
. Serological analysis of Melan-A (MART-1), a melanocyte-specific protein homogeneously expressed in human melanomas.
Proc. Natl. Acad. Sci. USA
93
:
5915
26
Valmori, D., J. F. Fonteneau, C. M. Lizana, N. Gervois, D. Lienard, D. Rimoldi, V. Jongeneel, F. Jotereau, J. C. Cerottini, P. Romero.
1998
. Enhanced generation of specific tumor-reactive CTL in vitro by selected Melan-A/MART-1 immunodominant peptide analogues.
J. Immunol.
160
:
1750
27
Gotch, F., J. Rothbard, K. Howland, A. Townsend, A. McMichael.
1987
. Cytotoxic T lymphocytes recognize a fragment of influenza virus matrix protein in association with HLA-A2.
Nature
326
:
881
28
Steven, N. M., N. E. Annels, A. Kumar, A. M. Leese, M. G. Kurilla, A. B. Rickinson.
1997
. Immediate early and early lytic cycle proteins are frequent targets of the Epstein-Barr virus-induced cytotoxic T cell response.
J. Exp. Med.
185
:
1605
29
Bogedain, C., H. Wolf, S. Modrow, G. Stuber, W. Jilg.
1995
. Specific cytotoxic T lymphocytes recognize the immediate-early transactivator Zta of Epstein-Barr virus.
J. Virol.
69
:
4872
30
Ogg, G. S., X. Jin, S. Bonhoeffer, P. R. Dunbar, M. A. Nowak, S. Monard, J. P. Segal, Y. Cao, S. L. Rowland-Jones, V. Cerundolo, et al
1998
. Quantitation of HIV-1 specific cytotoxic T lymphocytes and plasma load of viral RNA.
Science
279
:
2103
31
Valmori, D., M. J. Pittet, C. Vonarbourg, D. Rimoldi, D. Lienard, D. Speiser, R. Dunbar, V. Cerundolo, J. C. Cerottini, P. Romero.
1999
. Analysis of the cytolytic T lymphocyte response of melanoma patients to the naturally HLA-A*0201-associated tyrosinase peptide 368–376.
Cancer Res.
59
:
4050
32
Youde, S. J., P. R. Dunbar, E. M. Evans, A. N. Fiander, L. K. Borysiewicz, V. Cerundolo, S. Man.
2000
. Use of fluorogenic histocompatibility leukocyte antigen-A*0201/HPV 16 E7 peptide complexes to isolate rare human cytotoxic T-lymphocytes recognizing endogenous human papillomavirus antigens.
Cancer Res.
60
:
365
33
Jager, E., M. Ringhoffer, M. Altmannsberger, M. Arand, J. Karbach, D. Jager, F. Oesch, A. Knuth.
1997
. Immunoselection in vivo: independent loss of MHC class I and melanocyte differentiation antigen expression in metastatic melanoma.
Int. J. Cancer
71
:
142
34
Anichini, A., A. Molla, R. Mortarini, G. Tragni, I. Bersani, M. Di Nicola, A. M. Gianni, S. Pilotti, R. Dunbar, V. Cerundolo, G. Parmiani.
1999
. An expanded peripheral T cell population to a cytotoxic T lymphocyte (CTL)-defined, melanocyte-specific antigen in metastatic melanoma patients impacts on generation of peptide-specific CTLs but does not overcome tumor escape from immune surveillance in metastatic lesions.
J. Exp. Med.
190
:
651
35
Speiser, D. E., R. Miranda, A. Zakarian, M. F. Bachmann, K. McKall-Faienza, B. Odermatt, D. Hanahan, R. M. Zinkernagel, P. S. Ohashi.
1997
. Self antigens expressed by solid tumors do not efficiently stimulate naive or activated T cells: implications for immunotherapy.
J. Exp. Med.
186
:
645
36
Hermans, I. F., A. Daish, J. Yang, D. S. Ritchie, F. Ronchese.
1998
. Antigen expressed on tumor cells fails to elicit an immune response, even in the presence of increased numbers of tumor-specific cytotoxic T lymphocyte precursors.
Cancer Res.
58
:
3909
37
Allison, J. P., A. A. Hurwitz, D. R. Leach.
1995
. Manipulation of costimulatory signals to enhance antitumor T-cell responses.
Curr. Opin. Immunol.
7
:
682
38
Townsend, S. E., J. P. Allison.
1993
. Tumor rejection after direct costimulation of CD8+ T cells by B7-transfected melanoma cells.
Science
259
:
368
39
Albert, M. L., B. Sauter, N. Bhardwaj.
1998
. Dendritic cells acquire antigen from apoptotic cells and induce class I-restricted CTLs.
Nature
392
:
86
40
Sauter, B., M. L. Albert, L. Francisco, M. Larsson, S. Somersan, N. Bhardwaj.
2000
. Consequences of cell death: exposure to necrotic tumor cells, but not primary tissue cells or apoptotic cells, induces the maturation of immunostimulatory dendritic cells.
J. Exp. Med.
191
:
423
41
Wang, Z., B. Seliger, N. Mike, F. Momburg, A. Knuth, S. Ferrone.
1998
. Molecular analysis of the HLA-A2 antigen loss by melanoma cells SK-MEL-29.1.22 and SK-MEL-29.1.29.
Cancer Res.
58
:
2149
42
Hicklin, D. J., Z. Wang, F. Arienti, L. Rivoltini, G. Parmiani, S. Ferrone.
1998
. β2-Microglobulin mutations, HLA class I antigen loss, and tumor progression in melanoma.
J. Clin. Invest.
101
:
2720