Purpose: Merkel cell polyomavirus (MCPyV) is prevalent in the general population, integrates into most Merkel cell carcinomas (MCC), and encodes oncoproteins required for MCC tumor growth. We sought to characterize T-cell responses directed against viral proteins that drive this cancer as a step toward immunotherapy.

Experimental Design: Intracellular cytokine cytometry, IFN-γ enzyme-linked immunospot (ELISPOT) assay, and a novel HLA-A*2402–restricted MCPyV tetramer were used to identify and characterize T-cell responses against MCPyV oncoproteins in tumors and blood of MCC patients and control subjects.

Results: We isolated virus-reactive CD8 or CD4 T cells from MCPyV-positive MCC tumors (2 of 6) but not from virus-negative tumors (0 of 4). MCPyV-specific T-cell responses were also detected in the blood of MCC patients (14 of 27) and control subjects (5 of 13). These T cells recognized a broad range of peptides derived from capsid proteins (2 epitopes) and oncoproteins (24 epitopes). HLA-A*2402–restricted MCPyV oncoprotein processing and presentation by mammalian cells led to CD8-mediated cytotoxicity. Virus-specific CD8 T cells were markedly enriched among tumor infiltrating lymphocytes as compared with blood, implying intact T-cell trafficking into the tumor. Although tetramer-positive CD8 T cells were detected in the blood of 2 of 5 HLA-matched MCC patients, these cells failed to produce IFN-γ when challenged ex vivo with peptide.

Conclusions: Our findings suggest that MCC tumors often develop despite the presence of T cells specific for MCPyV T-Ag oncoproteins. The identified epitopes may be candidates for peptide-specific vaccines and tumor- or virus-specific adoptive immunotherapies to overcome immune evasion mechanisms in MCC patients. Clin Cancer Res; 17(21); 6671–80. ©2011 AACR.

This article is featured in Highlights of This Issue, p. 6603

Translational Relevance

Merkel cell carcinoma (MCC) is an aggressive neuroendocrine skin cancer (5-year disease-associated mortality rate of 46%) with limited treatment options for progressive disease. Current evidence suggests that the recently discovered Merkel cell polyomavirus (MPCyV) is causally associated with most MCCs. Reported MCC dependence on viral oncoproteins and our findings that virus-specific CD8 and CD4 T cells are present in MCC patients suggest that immunotherapeutic targeting of MCPyV and reversal of immune evasion mechanisms are promising options for patients with this disease.

The MCPyV epitopes in this report provide tools (i) to isolate both antigen- and tumor-specific T lymphocytes from blood and tumors of MCC patients, (ii) to characterize immune evasion mechanisms, (iii) to develop tumor-specific therapies such as peptide vaccines or adoptive immunotherapy, and (iv) to track T-cell responses during tumor progression or clinical trials.

Merkel cell carcinoma (MCC) is an aggressive neuroendocrine skin cancer. Its reported incidence has quadrupled in the past 20 years to approximately 1,600 cases per year in the United States (1). In 2008, the Merkel cell polyomavirus (MCPyV) was discovered and found to be integrated into the host genome in approximately 80% of MCC tumors (2). This association of MCPyV with MCC has been confirmed by multiple groups worldwide (3). MCPyV infection, defined by serology or detection of viral DNA, is prevalent in both healthy persons and MCC patients (4). However, in MCC patients, MCPyV acquires oncogenic potential via rare integration and T-antigen (T-Ag) truncation mutations (5, 6). These large T-Ag (LT) truncation mutants bind and inactivate the retinoblastoma tumor suppressor but no longer induce lytic viral replication that would be lethal to a cancer cell (7). The small T-Ag (ST) shares the N-terminus with LT and plays an important role in activating the AKT–mTOR signaling pathway (8, 9). Importantly, MCPyV T-Ag oncoproteins are persistently expressed in MCCs and are required for growth (6, 9).

The cellular immune system appears to be critical in preventing and controlling MCC. Patients who are chronically immunosuppressed by HIV infection, chronic lymphocytic leukemia, or medications after solid organ transplant have a 3- to 30-fold increased risk of MCC and represent approximately 10% of MCC patients (10). There are numerous cases of spontaneous MCC regression, suggesting immune-mediated cancer control (11, 12). Furthermore, intratumoral infiltration of CD8 lymphocytes is an independent predictor of improved survival in MCC (13). The high prevalence of antibodies to MCPyV T-Ag (14) and VP1 capsid proteins (15–17) in MCC patients is suggestive of virus-specific helper CD4 T-cell responses.

Despite the virus dependence of most MCC tumors and the high seroprevalence of MCPyV in the general population (15–17), there have been no reports of MCPyV-specific T-cell reactivity in MCC patients or control subjects. We hypothesized that MCPyV proteins are immunogenic and are recognized by T lymphocytes in MCC tumors and peripheral blood of patients and control subjects. Here, we show that MCPyV-specific CD8 and CD4 T cells can localize to MCC tumors and report 26 novel MCPyV T-cell epitopes, some of which may serve as targets for immunotherapy in the future.

Human subjects and clinical samples

This study was approved by the Fred Hutchinson Cancer Research Center (FHCRC) Institutional Review Board and conducted in accordance with Helsinki principles. Subjects (Supplementary Table S1) gave informed consent. Peripheral blood mononuclear cells (PBMC) obtained from heparinized blood samples were cryopreserved. Tumor samples obtained from medically necessary biopsies or surgeries were transported in T-cell medium (TCM; ref. 18).

Synthetic viral peptides

A total of 428 peptides (13-mers, overlapping by 9 amino acids) were synthesized (Sigma-Aldrich; Supplementary Table S2) from MCPyV reference sequence GenBank #EU375803 and strain variant #EU375804. Peptide pools (∼25 peptides per pool) were organized by viral domains (Fig. 1). Additional truncated MCPyV peptides and polyomavirus-derived homologous peptides (predicted by CLUSTAL 2.0.12 sequence alignment software) were synthesized (Supplementary Table S3).

Figure 1.

MCPyV proteome and synthetic peptide pools. MCPyV proteins (gray boxes) are annotated with nucleotide (nuc) and amino acid (a.a.) numbers. Peptides spanned the entire MCPyV proteome and were grouped into pools (boxes with a.a. range indicated) on the basis of protein domains. Double outlined boxes represent persistently expressed LT, ST, and CT (shared T-Ag N-terminus, light-gray slashed lines) domains. Jagged line at nucleotide 429 represents the genomic splice site. c(#) indicates the corresponding complementary DNA nucleotide position number.

Figure 1.

MCPyV proteome and synthetic peptide pools. MCPyV proteins (gray boxes) are annotated with nucleotide (nuc) and amino acid (a.a.) numbers. Peptides spanned the entire MCPyV proteome and were grouped into pools (boxes with a.a. range indicated) on the basis of protein domains. Double outlined boxes represent persistently expressed LT, ST, and CT (shared T-Ag N-terminus, light-gray slashed lines) domains. Jagged line at nucleotide 429 represents the genomic splice site. c(#) indicates the corresponding complementary DNA nucleotide position number.

Close modal

Tumor-infiltrating lymphocytes culture

Minced MCC tissue was plated in 48-well plates in TCM plus 1.6 μg/mL phytohaemagglutinin (PHA; Remel), 32 units/mL natural human interleukin-2 (nIL-2; Hemagen Diagnostics), and allogeneic-irradiated PBMCs. nIL-2 in fresh TCM was added every second day for 14 to 20 days (18). When conducted, a second tumor-infiltrating lymphocyte (TIL) expansion used anti-CD3 monoclonal antibody (mAb) OKT3 (Ortho), recombinant IL-2 (Chiron Corporation), and feeder cells (19). Cultures expanded with anti-CD3 mAb are comprised entirely of CD3-Qdot655 (Invitrogen)-positive cells (Supplementary Fig. S1).

T-cell cloning

Subject w347 primary tumor TILs were cloned by limiting dilution and subsequently expanded (20). Subject w347 metastatic tumor was collagenase (Sigma)-digested and plated as single-cell suspensions in TCM plus PHA and nIL-2 for limiting dilution cloning. Subject w447 bulk TILs were tetramer-sorted and expanded (19).

ELISPOT

Cultured assays.

PBMCs plated in TCM at 3 × 105 cells per well in 96-well flat bottom plates were stimulated on day 1 with MCPyV peptide pools (1 μg/mL of each peptide), individual peptides (1 μg/mL), cytomegalovirus (CMV), Epstein–Barr virus (EBV), influenza peptides (CEF, 2 μg/mL; Cellular Technology Ltd.) as positive control or dimethyl sulfoxide (DMSO; <0.2%) and/or TCM as negative control and cultured for 10 days. Fresh TCM with 20 U/mL IL-2 (Chiron Corporation) and 20 ng/mL IL-7 (R&D Systems) were added on days 3, 6, and 9. On day 11, cells were plated in 96-well multiscreen IP plate (Millipore) precoated with anti-IFN-γ capture antibody (1-D1K; Mabtech). Mitogens corresponding to the prior stimulation cycle were added on day 12 and the plates developed on day 13 (21).

Direct assays.

PBMCs plated at 5 × 105 cells per well in TCM were stimulated with relevant peptides. After 24 hours, plates were developed (21), scanned with an enzyme-linked immunospot (ELISPOT) reader (C.T.L.), and counted using ImmunoSpot 5.0 Software (C.T.L.). Stringent ELISPOT interpretation criteria were used to determine T-cell reactivity (Supplementary Methods).

Intracellular cytokine and flow cytometry

Responder cells stimulated with peptide in the presence of anti-CD28, anti-CD49d mAb, and brefeldin A (22) were incubated with carboxyfluorescein succinimidyl ester (CFSE; Invitrogen)-labeled antigen-presenting cells (APC) for 12 to 18 hours (23). Controls included DMSO and PMA-ionomycin. Cells subsequently stained with LIVE/DEAD-Violet (Invitrogen), followed by mAbs CD4-PerCP and CD8-APC (BD Biosciences), were permeabilized and stained with anti-IFN-γ-PE (BD Biosciences; ref. 23). Data acquired with FacsCanto-II cytometer (Becton Dickinson) and BD-FACSDiva software (v6.1.1) were analyzed with FlowJo (v9.1). Data are reported as percentage of viable cells of phenotypic interest identified as CFSE-negative responder cells in the lymphocyte forward/side scatter region.

Cytokine analysis

EBV transformed lymphoblastoid cell line (LCL; 2 × 105 cells per well) and responders (2 × 105 cells per well) were cocultured with peptides (1 μg/mL) in a 48-well plate in 1 mL TCM. Supernatants collected at 24 and 48 hours were analyzed for IFN-γ, IL-10, or IL-5 by fluorescent microbead-based flow cytometric ELISA (FHCRC, Seattle).

Plasmids, transfection, and imumunoblot

Cos7 cells were transfected with plasmid-encoding MCPyV LT and/or HLA-A*2402 cDNA using FuGENE6 transfection reagent (Roche) following manufacturer-recommended protocol (Supplementary Methods). Protein product was confirmed by immunoblotting for MCPyV LT-specific mAb CM2B4 (Santa Cruz; ref. 24). Expression of HLA-A*2402 was confirmed by flow cytometry with A23/A24 mAb (One Lambda) followed by phycoerithrin (PE)-labeled IgG secondary antibody (Invitrogen) at 48 hours. Controls included HLA-A*2402–positive and negative LCLs. Transfected Cos7 or control cells were used as APCs in intracellular cytokine (ICC) cytometry.

Cytotoxicity assay

Target cell lysis was determined by a 4-hour 51Cr release assay in quadruplicates (25). Negative controls included LT-transfected Cos7 plus effectors and Cos7 transfected with LT and HLA-A*2402 without effectors. Positive control included detergent-lysed target cells.

Tetramer staining

PE- and APC-conjugated complexes of HLA-A*2402 and peptide LT-92-101 (EWWRSGGFSF; CPC Scientific) or irrelevant peptide (NY-ESO-158-166; ref. 26), were synthesized at the Immune Monitoring Laboratory (FHCRC, Seattle). Cells were stained with tetramer for 60 minutes at 37°C, followed by anti-CD8-APC mAb (BD Biosciences) for 30 minutes at 4°C, then washed with fluorescence-activated cell-sorting (FACS) buffer, fixed, and analyzed.

MCPyV DNA sequencing and detection

DNA extraction, amplification, and sequencing are detailed in Supplementary Methods (3, 27). Sequence data were submitted to GenBank for subjects w447 (accession number JF912157) and w347 (accession number JF912158).

HLA typing

HLA-typing genotypes from blood-derived DNA were determined by sequence-based typing of exons 2 and 3 (ref. 28; FHCRC, Seattle) or with commercially available (OneLambda; Qiagen) sequence-specific primer PCR kits (Puget Sound Blood Center, Seattle; Hematologisk Laboratorium, Denmark).

Immunohistochemistry

MCPyV T-Ag staining was done using CM2B4 antibody (Santa Cruz; ref. 14, 24). Anti-CD8 (Novocastra) and anti-CD4 (Cell Marque) antibodies were used at 1:200 and 1:25 dilution, respectively.

Statistical analyses

Analyses were conducted using 2-tailed Fisher exact test in Stata v11.0 (StataCorp). A value of P < 0.05 was considered significant.

To study the T-cell response to MCPyV proteins, we isolated lymphocytes from MCC tumor tissues and blood donated by MCC patients and control subjects (Supplementary Table S1). Fresh tumor specimens were obtained from 10 MCC patients requiring diagnostic or therapeutic surgery. TILs were successfully cultured from each tumor in sufficient numbers to test T-cell reactivity to regions of T-Ag that are persistently expressed in MCC tumors (peptide pools 1–3 and 10; Fig. 1 and Supplementary Table S2). MCPyV T-Ag oncoproteins were present in 24 of 33 tumors (72%) as evaluated by immunohistochemistry (IHC). T-cell responses to MCPyV T-Ag as assessed by IFN-γ secretion were found in 2 of 6 T-Ag protein-expressing tumors. T-Ag–specific T cells were not detected among the 4 tumors that did not express viral oncoproteins.

Identification of the MCPyV large T-Ag epitope recognized by CD8+ TILs

Subject w447 presented at age 67 with a 1.5-cm primary MCC lesion on his left hip. The patient underwent wide local excision of the tumor, a portion of which was used for obtaining TILs. IHC revealed CD8+ cells within an MCPyV T-Ag protein–positive tumor that had low HLA class I expression (Fig. 2A). After 14 days of expansion in the presence of PHA and nIL-2, TILs were screened by ICC assay for IFN-γ production to peptides from regions of T-Ag that are persistently expressed in MCC tumors (pools 1, 2, 3, and 10). A discrete population of IFN-γ–secreting CD8 cells was only reactive to peptide pool 2 (Fig. 2B). Upon challenge with single peptides from pool 2, these TILs responded to a single immunogenic peptide, LT-89-101 (KFKEWWRSGGFSF; Fig. 2B). Truncation analysis identified a 10-amino acid section, LT-92-101 (EWWRSGGFSF), that generated strong CD8 IFN-γ responses that persisted in peptide titrations to as low as 1 ng/mL (Fig. 2B and C). Notably, engagement of TCR signals may downregulate CD8 expression (29), explaining slightly low CD8 signal among IFN-γ–producing cells. Sequencing of DNA amplified from this patient's tumor confirmed the predicted MCPyV large T-Ag amino acid sequence (amino acids 92–101, data not shown).

Figure 2.

CD8+ TIL isolated from an MCC patient specifically recognize an MCPyV-derived peptide and endogenously processed large T-Ag (LT). A, serial sections from w447 MCC tumor. Left, hematoxylin and eosin (H&E) stain; T-Ag expression (CM2B4 antibody); CD8 lymphocyte staining; HLA-I staining (EMR8-5 clone). Scale bar, 200 μm. B, CD8 IFN-γ response to LT-derived peptides in an ICC assay. Left, TILs stimulated with DMSO; pool 2; 13-mer peptide LT-89-101 from pool 2; and minimal 10-mer epitope LT-92-101. Autologous PBMCs were used as APCs. C, TILs have high functional avidity for peptide LT-92-101. D, TILs recognize MCPyV LT-89-101 but not homologous peptides from indicated polyomaviruses. Y-axis represents flow cytometry data gated as in B. E, Western blotting of primary bands after probe with anti-MCPyV LT (CM2B4) or anti-α-actin loading control in MKL-1 (MCPyV-positive MCC cell line) and Cos7 cells transfected as indicated. * and **, distinct truncation isoforms. F, flow cytometric analysis of HLA-A24 expression on Cos7 cells transfected as indicated. Data represent percentage of cells above PE fluorescence intensity threshold. G, TIL IFN-γ responses to Cos7 cells transfected as indicated. Numbers shown in plots are the percentage of CD8-gated TILs.

Figure 2.

CD8+ TIL isolated from an MCC patient specifically recognize an MCPyV-derived peptide and endogenously processed large T-Ag (LT). A, serial sections from w447 MCC tumor. Left, hematoxylin and eosin (H&E) stain; T-Ag expression (CM2B4 antibody); CD8 lymphocyte staining; HLA-I staining (EMR8-5 clone). Scale bar, 200 μm. B, CD8 IFN-γ response to LT-derived peptides in an ICC assay. Left, TILs stimulated with DMSO; pool 2; 13-mer peptide LT-89-101 from pool 2; and minimal 10-mer epitope LT-92-101. Autologous PBMCs were used as APCs. C, TILs have high functional avidity for peptide LT-92-101. D, TILs recognize MCPyV LT-89-101 but not homologous peptides from indicated polyomaviruses. Y-axis represents flow cytometry data gated as in B. E, Western blotting of primary bands after probe with anti-MCPyV LT (CM2B4) or anti-α-actin loading control in MKL-1 (MCPyV-positive MCC cell line) and Cos7 cells transfected as indicated. * and **, distinct truncation isoforms. F, flow cytometric analysis of HLA-A24 expression on Cos7 cells transfected as indicated. Data represent percentage of cells above PE fluorescence intensity threshold. G, TIL IFN-γ responses to Cos7 cells transfected as indicated. Numbers shown in plots are the percentage of CD8-gated TILs.

Close modal

We next determined whether TILs from this subject cross-reacted to peptide homologues of MCPyV LT-92-101 derived from other human polyomaviruses. Cross-reactivity with other polyomaviruses could indicate T-cell priming by other pathogens prior to tumor infiltration. Conversely, if the TIL response was MCPyV specific, this would suggest that MCPyV was the relevant immunogen. To address this, we tested 13-mer peptides homologous to MCPyV LT-92-101 (Supplementary Table S3) and did not detect cross-reactivity among TILs to these peptides (Fig. 2D).

HLA restriction and endogenous processing of MCPyV large T-Ag

We sought to determine the restricting HLA allele from among this patient's genotype (A*03G1, A*24G1, B*1501, B*3502, Cw*0304, and Cw*04G1). Two published HLA-peptide binding algorithms (30) predicted strong binding of the immunogenic peptide (MCPyV LT 92-101) to HLA-A*2402. This in silico prediction was confirmed using cells matched to subject w447 only at HLA-A*2402. Although allogeneic partially matched PBMCs induced IFN-γ secretion in CD8 TIL from subject w447, peptide-loaded PBMCs that were not matched at any HLA locus induced minimal IFN-γ secretion (Supplementary Fig. S2).

To study MCPyV oncoprotein processing into short peptides, we cotransfected full-length MCPyV LT and HLA-A*2402 into the primate cell line Cos7 (Fig. 2E–F). Transfected cells were exposed to subject w447 TILs, and T-cell activation was measured by IFN-γ production. Mock-transfected cells, or cells transfected with A*2402 alone, did not activate subject w447 TIL. However, cotransfection of LT and A*2402 led to readily detectable activation of CD8 TIL by IFN-γ ICC (Fig. 2G). Furthermore, we observed specific lysis of Cos7 cells that expressed MCPyV LT and HLA-A*2402 by a TIL-derived CD8 T-cell clone (Supplementary Fig. S3). These MCPyV-specific CD8 cells also have excellent cytolytic potential against peptide-pulsed autologous cells (91% and 100% lysis of PBMCs pulsed with peptide LT.92-101 at 1 μg/mL and 5 μg/mL, respectively; Supplementary Fig. S3). We conclude that (i) MCPyV LT can be processed into appropriate peptides, (ii) these peptides can access the HLA class-I antigen presentation pathway for proper loading, and (iii) MCPyV-specific CD8 TILs can recognize and kill cells expressing LT and the appropriate HLA.

MCPyV-specific CD8 T cells are enriched in the tumor as compared with blood

We synthesized a fluorescent HLA-peptide tetramer (A24:MCPyV.LT.92-101). This MCPyV-specific tetramer and a control HLA-A*2402 tetramer were used to stain subject w447 PBMCs and his TILs that had been expanded in an unbiased fashion with PHA and IL-2. Our data are consistent with marked enrichment of MCPyV-specific CD8 T cells among TILs (4.0%) as compared with blood (0.14%; Fig. 3A, top and middle).

Figure 3.

Detection and functional status of MCPyV-specific CD8 T cells in tumor and blood of MCC patients. A, subject w447 TILs (top), w447 PBMCs (middle), and w350 PBMCs (bottom) specifically stain with an MCPyV-specific tetramer but not with control tetramer. Specific gating for cell clusters was used in the bottom 4 panels as indicated. B, IFN-γ secretion by TILs and PBMCs in response to LT-92-101 and positive controls as assessed by direct ELISPOT assay.

Figure 3.

Detection and functional status of MCPyV-specific CD8 T cells in tumor and blood of MCC patients. A, subject w447 TILs (top), w447 PBMCs (middle), and w350 PBMCs (bottom) specifically stain with an MCPyV-specific tetramer but not with control tetramer. Specific gating for cell clusters was used in the bottom 4 panels as indicated. B, IFN-γ secretion by TILs and PBMCs in response to LT-92-101 and positive controls as assessed by direct ELISPOT assay.

Close modal

MCPyV-specific T cells identified in PBMCs in two MCC patients fail to produce IFN-γ

PBMCs from 5 MCC patients (all of whom were positive for an HLA-A*24xx genotype) were stained with the MCPyV tetramer. We observed a well-defined population of tetramer-bright, CD8-positive cells in PBMCs from both the index patient w447 and a second patient, w350 (Fig. 3A, middle and bottom), but not in other HLA-A24–positive or HLA-A24–negative subjects (Supplementary Fig. S4). Both of these patients had T-Ag–expressing tumors. In contrast to expanded TILs from subject w447, physiologically relevant direct ELISPOT functional analyses of PBMCs from subjects w447 and w350 revealed no IFN-γ production in response to antigen (Fig. 3B). These observations were further supported by ICC assays examining IFN-γ production in response to peptide among tetramer-specific cells in subject w447. Although 20% of tetramer-positive TILs secreted IFN-γ, only 2% of tetramer-positive PBMCs produced IFN-γ (Supplementary Fig. S5). The PBMCs analyzed for IFN-γ were drawn during an early asymptomatic time point after recurrence (subject w447) or 1 month post–tumor removal (w350).

CD4+ T cells with the same MCPyV specificity isolated from an MCC primary tumor and a subsequent metastasis

Subject w347 was 78 years old when she presented with a 3.5-cm MCC on the right forearm. IHC revealed CD4+ and FoxP3+ cells within an MCPyV T-Ag protein–positive tumor (Fig. 4A). TILs were expanded in the absence of exogenous antigen and later screened for IFN-γ production to peptide pools 1, 2, 3, and 10 by the ICC assay. CD4 reactivity was observed only to peptide pool 1 (Fig. 4B). Pool breakdown showed reactivity only to peptide CT-57-69 (TLWSKFQQNIHKL; Fig. 4B). Of note, strong antigenic triggers may downregulate CD4 cell surface expression (31) among IFN-γ–producing cells. Sequencing of MCPyV DNA amplified from tumor w347 confirmed the predicted T-Ag sequence (amino acids 57–69, data not shown).

Figure 4.

Characterization and cloning of MCPyV-specific CD4 TIL from an MCC patient. A, MCC tumor from subject w347. Left, primary tumor arising on forearm; MCPyV T-Ag expression (CM2B4); CD4 and FoxP3 (236A/E7 clone) staining. Scale bar, 200 μm. B, identification of IFN-γ CD4 T-cell responses to CT domain peptides by ICC assay. From left, TILs stimulated with DMSO; CT pool 1; representative nonreactive pool 1 peptide; reactive pool 1 peptide CT-57-69. PBMCs were used as APCs. Gating as in Fig. 2. C, TIL cytokine response to MCPyV peptide or its homologues as measured by ELISA. Left, bulk CD4 TIL IFN-γ response to CT-57-69 or homologous peptides from the indicated polyomaviruses. Right, Bulk CD4 TIL secretes IL-5 and IL-10. Error bars represent standard error between triplicates. D, CD4 clone isolated from primary tumor TIL secretes IFN-γ and IL-10 in response to CT-57-69. Data represent single measurements at 2 time points. E, CD4 TILs from recurrent lesion recognize the same peptide as TILs from primary lesion. Left, recurrent w347 MCC (red outline); representative nonreactive T-cell clone; T-cell clone reactive to peptide CT-57-69.

Figure 4.

Characterization and cloning of MCPyV-specific CD4 TIL from an MCC patient. A, MCC tumor from subject w347. Left, primary tumor arising on forearm; MCPyV T-Ag expression (CM2B4); CD4 and FoxP3 (236A/E7 clone) staining. Scale bar, 200 μm. B, identification of IFN-γ CD4 T-cell responses to CT domain peptides by ICC assay. From left, TILs stimulated with DMSO; CT pool 1; representative nonreactive pool 1 peptide; reactive pool 1 peptide CT-57-69. PBMCs were used as APCs. Gating as in Fig. 2. C, TIL cytokine response to MCPyV peptide or its homologues as measured by ELISA. Left, bulk CD4 TIL IFN-γ response to CT-57-69 or homologous peptides from the indicated polyomaviruses. Right, Bulk CD4 TIL secretes IL-5 and IL-10. Error bars represent standard error between triplicates. D, CD4 clone isolated from primary tumor TIL secretes IFN-γ and IL-10 in response to CT-57-69. Data represent single measurements at 2 time points. E, CD4 TILs from recurrent lesion recognize the same peptide as TILs from primary lesion. Left, recurrent w347 MCC (red outline); representative nonreactive T-cell clone; T-cell clone reactive to peptide CT-57-69.

Close modal

To determine whether this CD4 TIL response was cross-reactive, we tested 13-mer homologue peptides (Supplementary Table S3) from other human polyomaviruses. No specific reactivity to the homologous peptides was noted (Fig. 4C, left). To investigate the nature of the CD4 response to MCPyV CT-57-69, we investigated the cytokine secretion profile of the expanded CD4 TIL population. Interestingly, we observed the secretion of both Th1 and Th2 type cytokines, namely, IFN-γ, IL-5, and IL-10 (Fig. 4C), although the latter cytokine was only modestly increased compared with background. Moreover, 2 of 58 tested T-cell clones isolated by limiting dilution from the primary TILs also secreted IFN-γ and 1 of 1 clones tested also secreted IL-10 in response to peptide (Fig. 4D).

Two months after primary tumor excision, subject w347 developed a skin recurrence of MCC (Fig. 4E, left). After clonal microculture of the digested tumor tissue, 1 of 53 clones tested produced a robust IFN-γ response to the same peptide (CT-57-69) as in the primary tumor (Fig. 4E, right).

T-cell responses to MCPyV in the peripheral blood

PBMC reactivity was screened against the MCC-associated T-Ag oncoprotein domains (pools 1, 2, 3, and 10) in MCC patients and control subjects. We observed a trend toward a higher frequency of T-cell responses in patients than in control subjects for pool 1 (37% of patient PBMC samples vs. 8% of control subject samples), pool 2 (31% vs. 17%), and pool 10 (37% vs. 17%; Fig. 5A). The magnitude of pool-level responses for each subject in Fig. 5A is included as Supplementary Fig. S6. We detected PBMC responses against T-Ag in 12 of 19 patients with MCPyV-positive tumors. In contrast, we found T-Ag–specific T-cell reactivity in blood from 1 of 6 MCC patients with MCPyV-negative tumors (P = 0.07).

Figure 5.

Prevalence of PBMC reactivity to MCPyV oncoproteins and determination of immunoreactive epitopes. A, PBMC T-cell reactivity (as described in ELISPOT interpretation criteria in Supplementary Methods) to specified MCPyV peptide pools among MCC patients (black) and control subjects (gray). Although there is a trend, data did not reach statistical significance. B, identification of immunoreactive epitopes derived from previously reactive T-Ag peptide pools by cultured IFN-γ ELISPOT assay in MCC patients (black) and control subjects (gray). In subject C4, 5 peptides (arrows) were identified to be immunogenic to CD4+ T cells.

Figure 5.

Prevalence of PBMC reactivity to MCPyV oncoproteins and determination of immunoreactive epitopes. A, PBMC T-cell reactivity (as described in ELISPOT interpretation criteria in Supplementary Methods) to specified MCPyV peptide pools among MCC patients (black) and control subjects (gray). Although there is a trend, data did not reach statistical significance. B, identification of immunoreactive epitopes derived from previously reactive T-Ag peptide pools by cultured IFN-γ ELISPOT assay in MCC patients (black) and control subjects (gray). In subject C4, 5 peptides (arrows) were identified to be immunogenic to CD4+ T cells.

Close modal

Twenty-four additional immunogenic MCPyV epitopes were identified. Breakdown of the reactive pools to single reactive peptides is shown in Fig. 5B for 2 MCC patients and 2 control subjects. A T-cell response to a single peptide (ST-125-137) was common to both subjects C7 and w147. Responder T-cell phenotype was determined for 5 immunoreactive T-Ag–derived peptides in control subject C4 by a cultured ICC assay. In each case, the reactive cells were CD4 positive (data not shown). To explore whether these CD4 T-cell responses were likely to be MCPyV specific, we compared the sequence of the reactive MCPyV 13-mers with the homologous regions from other known polyomaviruses (Supplementary Table S3). Among these 5 peptides, 3 were from an MCPyV unique protein domain and 1 was sequence divergent when compared with homologous peptides, suggesting MCPyV specificity in most cases.

T-cell responses to the entire MCPyV proteome (18 peptide pools) were also assessed in 4 control subjects. PBMCs from each subject showed strong reactivity to at least 1 portion of the MCPyV proteome (Fig. 6A). Of note, PBMCs from all 4 control subjects reacted to peptides in pool 13 (derived from the capsid protein VP1). Single peptide-level responses were deconvoluted for 2 subjects (C7 and C8). Subject C7 showed robust T-cell responses to multiple peptides within pool 13 (Fig. 6B, top). The peptide with the strongest response (VP1-225-237) was shown to be CD4 positive (Fig. 6B, top, inset). This particular peptide shares amino acid identity at 5 of 13 positions, with all of the homologous peptides derived from 5 other polyomaviruses (Supplementary Table S3). All 6 of these related peptides elicited T-cell reactivity from this subject's PBMCs (Fig. 6B, top, right). In subject C8, VP1-245-257 was the only reactive peptide within pool 13 and it stimulated CD4 cells (Fig. 6B, bottom, inset). This immunoreactivity was MCPyV specific, as homologous peptides from other polyomaviruses were nonreactive (Fig. 6B, bottom, right). Aggregate results for all 26 peptides described in this study are shown in Supplementary Table S4.

Figure 6.

MCPyV proteome-wide screen of T-cell responses and characterization of select epitopes. A, PBMCs from 4 control subjects showed reactivity to both MCPyV T-Ag and structural proteins as measured by cultured IFN-γ ELISPOT assay. B, characterization of T-cell responses to pool 13 peptides derived from VP1 capsid protein in control subject C7 (top) and C8 (bottom). Left, stimulation with CEF-positive control, pool 13, or single 13-mer peptides from pool 13. Inset, ICC/flow cytometric analysis identifies IFN-γ+CD4+ T cells upon peptide stimulation. Right, PBMCs challenged with relevant peptide or homologous peptides from indicated polyomaviruses.

Figure 6.

MCPyV proteome-wide screen of T-cell responses and characterization of select epitopes. A, PBMCs from 4 control subjects showed reactivity to both MCPyV T-Ag and structural proteins as measured by cultured IFN-γ ELISPOT assay. B, characterization of T-cell responses to pool 13 peptides derived from VP1 capsid protein in control subject C7 (top) and C8 (bottom). Left, stimulation with CEF-positive control, pool 13, or single 13-mer peptides from pool 13. Inset, ICC/flow cytometric analysis identifies IFN-γ+CD4+ T cells upon peptide stimulation. Right, PBMCs challenged with relevant peptide or homologous peptides from indicated polyomaviruses.

Close modal

The strong association of MCC with both a polyomavirus and T-cell dysfunction suggests that dissection of the specificity, trafficking, and effector functions of MCPyV-specific T cells is medically significant. We show that circulating and tumor-infiltrating T cells can recognize specific peptides from MCPyV T-Ag oncoproteins and capsid structural proteins and that both CD8 and CD4 MCPyV-specific T cells are locally enriched in at least some MCC tumors. Twenty-six novel MCPyV T-cell epitopes have been identified. We confirmed that the T-Ag oncoprotein could be processed and presented to stimulate cell lysis by virus-specific CD8 cells isolated from a patient's tumor. We developed an MCPyV tetramer that should serve as the first of several tools for isolating and determining the function of MCPyV-specific CD8 T cells directly ex vivo. These findings advance our understanding of the association of immunosuppression with MCC and help enable rational therapies for MCC that augment virus-specific T-cell activity.

Although not previously described, several lines of evidence suggest that MCPyV-specific T cells should be present in patients and control subjects. Serologic data indicate that MCPyV infection is commonly acquired during childhood (16, 32, 33). The presence of MCPyV-specific antibodies in 53% to 88% of healthy persons (15–17) implies the presence of MCPyV-specific CD4 T-cell help (indeed, all MCPyV-responsive T cells identified in our PBMC-based assays were CD4+). Similar to these findings in MCPyV, T-cell responses have been described (34–37) in other cancer-associated viruses, such as human papilloma virus and hepatitis C virus HCV, but are typically of low abundance (38, 39). Cultured ELISPOT or ICC assays are usually required for their detection (35, 40), but occasionally, their abundance is sufficient to be detected ex vivo by tetramers (41).

Our data document significant enrichment of MCPyV-specific T cells in an MCC tumor compared with blood. In subject w447 (with a small, slow-growing tumor), we detected a marked enrichment of antigen-specific CD8 cells in TILs compared with blood using tetramer staining. Because primary tumor tissue was sparse, TILs were expanded using a nonspecific mitogen (PHA + IL-2) cocktail that uniformly expands the peripheral T-cell repertoire (42). We therefore believe the relative frequency of antigen-specific TILs remains stable after expansion. Regardless, this finding indicates that local immune evasion mechanisms in this tumor did not block the trafficking and enrichment of virus-specific T cells. However, the tumor may persist despite these antigen-specific CD8+ T cells due in part to low HLA-I expression. Conversely, chronic exposure to antigen may mediate an exhausted phenotype of antigen-specific CD8+ T cells with poor effector function and sustained expression of inhibitory receptors, such as PD-1 and CTLA-4 (43). Interestingly, in subject w323 (with a large, rapidly growing tumor), we detected MCPyV T-Ag–specific T cells only in the blood but not in TILs derived from her T-Ag–expressing tumor. In the latter case, there are several possible explanations about antigen-specific T cells: (i) they were present in the blood but failed to home to the tumor; (ii) they homed to the tumor but were undetectable because they failed to produce IFN-γ; or (iii) they were present in the tumor but failed to expand in culture.

It is possible that local MCPyV-specific T cells actually promoted tumor growth in some cases. Chronic antigen exposure, a common feature of MCC tumors, has been reported to induce IL-10 and IFN-γ–secreting CD4 T cells that have a regulatory function (44). In subject w347, we observed MCPyV-specific IFN-γ and IL-10 cytokine secretion by CD4 T cells isolated from the primary tumor. Notably, we also detected FoxP3+ cells within the tumor microenvironment, which may be similar to the antigen-specific, IL-10–secreting, FoxP3+ T-regulatory cell subset found to be immunosuppressive in metastatic melanoma (45). Further investigation is required to determine the role of T-regulatory cells in MCC.

There are several limitations to this study. PBMC availability prevented proteome-wide screens for some subjects; therefore, we focused on the oncogenic portions of T-Ag persistently expressed in tumors. Cultured ELISPOT assay may sometimes generate false-positive reactivity via in vitro priming (46). To eliminate such artifactual results, we used stringent ELISPOT interpretation criteria (Supplementary Methods). Finally, it was not possible to determine the extent of prior exposure to MCPyV for all studied subjects because no “gold standard” serologic or viral shedding test exists. The single most relevant indicator of MCPyV status is viral oncoprotein expression in the tumor, which was assessed for 33 of 38 patients.

In conclusion, this study is the first report of T-cell immune responses to the MCPyV. The necessary and persistent expression of T-Ags to drive MCC proliferation makes these oncoproteins especially promising targets for measuring and manipulating tumor-specific immune responses. As has been accomplished in other cancers (20, 47, 48), the identified MCPyV T-cell epitopes may be used to design peptide-specific vaccines, to generate virus-specific T cells for adoptive immunotherapy, or to track tumor-associated antigen-specific T-cell responses.

No potential conflicts of interest were disclosed.

The authors thank Stephanie Pietromonaco, Miranda Schmidt, Renee Thibodeau, Sherry Lee, Yuki Yoshimatsu, Jianhong Cao, Sine Hadrup, Julie Randolph-Habecker's laboratory, and FHCRC Specimen Processing for their contributions toward this study.

NIH-RC2-CA147820-01, NIH-R01CA162522, NIH-F30ES019463-01, FHCRC Molecular Diagnostics Pilot Award, NIH-K02-AR50993, ACS-RSG-08-115-01-CCE, NIH-K24-CA139052-0, Poncin scholarship, David & Rosalind Bloom Endowment for MCC Research, and UW MCC Patient Gift Fund.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Albores-Saavedra
J
,
Batich
K
,
Chable-Montero
F
,
Sagy
N
,
Schwartz
AM
,
Henson
DE
. 
Merkel cell carcinoma demographics, morphology, and survival based on 3870 cases: a population based study
.
J Cutan Pathol
2009
;
1
:
20
7
.
2.
Feng
H
,
Shuda
M
,
Chang
Y
,
Moore
PS
. 
Clonal integration of a polyomavirus in human Merkel cell carcinoma
.
Science
2008
;
319
:
1096
100
.
3.
Becker
JC
,
Houben
R
,
Ugurel
S
,
Trefzer
U
,
Pföhler
C
,
Schrama
D
. 
MC polyomavirus is frequently present in Merkel cell carcinoma of European patients
.
J Invest Dermatol
2009
;
129
:
248
50
.
4.
Foulongne
V
,
Kluger
N
,
Dereure
O
,
Mercier
G
,
Molès
JP
,
Guillot
B
, et al
Merkel cell polyomavirus in cutaneous swabs
.
Emerging Infect Dis
2010
;
16
:
685
7
.
5.
Shuda
M
,
Feng
H
,
Kwun
HJ
,
Rosen
ST
,
Gjoerup
O
,
Moore
PS
, et al
T antigen mutations are a human tumor-specific signature for Merkel cell polyomavirus
.
Proc Natl Acad Sci U S A
2008
;
105
:
16272
77
.
6.
Houben
R
,
Schrama
D
,
Alb
M
,
Pföhler
C
,
Trefzer
U
,
Ugurel
S
, et al
Comparable expression and phosphorylation of the retinoblastoma protein in Merkel cell polyoma virus-positive and negative Merkel cell carcinoma
.
Int J Cancer
2010
;
126
:
796
8
.
7.
Houben
R
,
Adam
C
,
Baeurle
A
,
Hesbacher
S
,
Grimm
J
,
Angermeyer
S
, et al
An intact retinoblastoma protein binding site in Merkel cell polyomavirus large T antigen is required for promoting growth of Merkel cell carcinoma cells
.
Int J Cancer
2011 Mar 16
[Epub ahead of print]
.
8.
Sablina
AA
,
Hahn
WC
. 
SV40 small T antigen and PP2A phosphatase in cell transformation
.
Cancer Metastasis Rev
2008
;
27
:
137
46
.
9.
Shuda
M
,
Kwun
HJ
,
Feng
H
,
Chang
Y
,
Moore
PS
. 
Human Merkel cell polyomavirus small T antigen is an oncoprotein targeting the 4E-BP1 translation regulator
.
J Clin Invest
2011
;
9
:
3623
34
.
10.
Heath
M
,
Jaimes
N
,
Lemos
B
,
Mostaghimi
A
,
Wang
LC
,
Penas
PF
, et al
Clinical characteristics of Merkel cell carcinoma at diagnosis in 195 patients: the AEIOU features
.
J Am Acad Dermatol
2008
;
58
:
375
81
.
11.
Kubo
H
,
Matsushita
S
,
Fukushige
T
,
Kanzaki
T
,
Kanekura
T
. 
Spontaneous regression of recurrent and metastatic Merkel cell carcinoma
.
J Dermatol
2007
;
34
:
773
7
.
12.
Wooff
JC
,
Trites
JR
,
Walsh
NMG
,
Bullock
MJ
. 
Complete spontaneous regression of metastatic Merkel cell carcinoma: a case report and review of the literature
.
Am J Dermatopathol
2010
;
6
:
614
7
.
13.
Paulson
KG
,
Iyer
JG
,
Tegeder
AR
,
Thibodeau
R
,
Schelter
J
,
Koba
S
, et al
Transcriptome-wide studies of Merkel cell carcinoma and validation of intratumoral CD8+ lymphocyte invasion as an independent predictor of survival
.
J Clin Oncol
2011
;
12
:
1539
46
.
14.
Paulson
KG
,
Carter
JJ
,
Johnson
LG
,
Cahill
KW
,
Iyer
JG
,
Schrama
D
, et al
Antibodies to Merkel cell polyomavirus T antigen oncoproteins reflect tumor burden in Merkel cell carcinoma patients
.
Cancer Res
2010
;
70
:
8388
97
.
15.
Pastrana
DV
,
Tolstov
YL
,
Becker
JC
,
Moore
PS
,
Chang
Y
,
Buck
CB
. 
Quantitation of human seroresponsiveness to Merkel cell polyomavirus
.
PLoS Pathog
2009
;
5
:
e1000578
.
16.
Tolstov
YL
,
Pastrana
DV
,
Feng
H
,
Becker
JC
,
Jenkins
FJ
,
Moschos
S
, et al
Human Merkel cell polyomavirus infection II. MCV is a common human infection that can be detected by conformational capsid epitope immunoassays
.
Int J Cancer
2009
;
125
:
1250
6
.
17.
Carter
JJ
,
Paulson
KG
,
Wipf
GC
,
Miranda
D
,
Madeleine
MM
,
Johnson
LG
, et al
Association of Merkel cell polyomavirus-specific antibodies with Merkel cell carcinoma
.
J Natl Cancer Inst
2009
;
101
:
1510
22
.
18.
Koelle
DM
,
Corey
L
,
Burke
RL
,
Eisenberg
RJ
,
Cohen
GH
,
Pichyangkura
R
, et al
Antigenic specificity of human CD4+ T cell clones recovered from recurrent genital HSV-2 lesions
.
J Virol
1994
;
68
:
2803
10
.
19.
Koelle
DM
,
Chen
H
,
Gavin
MA
,
Wald
A
,
Kwok
WW
,
Corey
L
. 
CD8 CTL from genital herpes simplex lesions: recognition of viral tegument and immediate early proteins and lysis of infected cutaneous cells
.
J Immunol
2001
;
166
:
4049
58
.
20.
Yee
C
,
Gilbert
MJ
,
Riddell
SR
,
Brichard
VG
,
Fefer
A
,
Thompson
JA
, et al
Isolation of tyrosinase-specific CD8+ and CD4+ T cell clones from the peripheral blood of melanoma patients following in vitro stimulation with recombinant vaccinia virus
.
J Immunol
1996
;
157
:
4079
86
.
21.
Hosken
N
,
McGowan
P
,
Meier
A
,
Koelle
DM
,
Sleath
P
,
Wagener
F
, et al
Diversity of the CD8+ T-cell response to herpes simplex virus type 2 proteins among persons with genital herpes
.
J Virol
2006
;
80
:
5509
15
.
22.
Posavad
CM
,
Wald
A
,
Hosken
N
,
Huang
ML
,
Koelle
DM
,
Ashley
RL
, et al
T cell immunity to herpes simplex viruses in seronegative subjects: silent infection or acquired immunity?
J Immunol
2003
;
170
:
4380
8
.
23.
Jing
L
,
Chong
TM
,
McClurkan
CL
,
Huang
J
,
Story
BT
,
Koelle
DM
. 
Diversity in the acute CD8 T cell response to vaccinia virus in humans
.
J Immunol
2005
;
175
:
7550
9
.
24.
Shuda
M
,
Arora
R
,
Kwun
HJ
,
Feng
H
,
Sarid
R
,
Fernandez-Figueras
MT
, et al
Human Merkel cell polyomavirus infection I. MCV T antigen expression in Merkel cell carcinoma, lymphoid tissues and lymphoid tumors
.
Int J Cancer
2009
;
125
:
1243
9
.
25.
Koelle
DM
,
Posavad
CM
,
Barnum
GR
,
Johnson
ML
,
Frank
JM
,
Corey
L
. 
Clearance of HSV-2 from recurrent genital lesions correlates with infiltration of HSV-specific cytotoxic T lymphocytes
.
J Clin Invest
1998
;
101
:
1500
8
.
26.
Yamaguchi
H
. 
Identification of HLA-A24-restricted CTL epitope from cancer-testis antigen, NY-ESO-1, and induction of a specific antitumor immune response
.
Clin Cancer Res
2004
;
10
:
890
6
.
27.
Garneski
KM
,
Warcola
AH
,
Feng
Q
,
Kiviat
NB
,
Leonard
JH
,
Nghiem
P
. 
Merkel cell polyomavirus is more frequently present in North American than Australian Merkel cell carcinoma tumors
.
J Invest Dermatol
2009
;
129
:
246
8
.
28.
Eiz-Vesper
B
,
Deluca
DS
,
Blasczyk
R
,
Horn
PA
. 
The nature of recombination in HLA-B*4207
.
Tissue Antigens
2007
;
70
:
164
8
.
29.
Park
J-H
,
Adoro
S
,
Lucas
PJ
,
Sarafova
SD
,
Alag
AS
,
Doan
LL
et al 
“Coreceptor tuning”: cytokine signals transcriptionally tailor CD8 coreceptor expression to the self-specificity of the TCR
.
Nat Immunol
2007
;
8
:
1049
59
.
30.
Zhang
Q
,
Wang
P
,
Kim
Y
,
Haste-Andersen
P
,
Beaver
J
,
Bourne
PE
, et al
Immune epitope database analysis resource (IEDB-AR)
.
Nucleic Acids Res
2008
;
36
:
W513
8
.
31.
Weyand
CM
,
Goronzy
J
,
Fathman
CG
. 
Modulation of CD4 by antigenic activation
.
J Immunol
1987
;
138
:
1351
4
.
32.
Chen
T
,
Hedman
L
,
Mattila
PS
,
Jartti
T
,
Ruuskanen
O
,
Soderlund-Venermo
M
, et al
Serological evidence of Merkel cell polyomavirus primary infections in childhood
.
J Clin Virol
2010
;
50
:
125
9
.
33.
Kean
J M
,
Rao
S
,
Wang
M
,
Garcea
RL
. 
Seroepidemiology of human polyomaviruses
.
PLoS Pathog
2009
;
5
:
e1000363
.
34.
Gheuens
S
,
Bord
E
,
Kesari
S
,
Simpson
DM
,
Gandhi
RT
,
Clifford
DB
, et al
Role of CD4+ and CD8+ T-cell responses against JC virus in the outcome of patients with progressive multifocal leukoencephalopathy (PML) and PML with immune reconstitution inflammatory syndrome
.
J Virol
2011
;
85
:
7256
63
.
35.
Schneidawind
D
,
Schmitt
A
,
Wiesneth
M
,
Mertens
T
,
Bunjes
D
,
Freund
M
, et al
Polyomavirus BK-specific CD8+ T cell responses in patients after allogeneic stem cell transplant
.
Leuk Lymphoma
2010
;
51
:
1055
62
.
36.
Chan
PKS
,
Liu
S-J
,
Cheung
JLK
,
Cheung
TH
,
Yeo
W
,
Chong
P
, et al
T-cell response to human papillomavirus type 52 L1, E6, and E7 peptides in women with transient infection, cervical intraepithelial neoplasia, and invasive cancer
.
J Med Virol
2011
;
83
:
1023
30
.
37.
Scott
M
,
Nakagawa
M
,
Moscicki
AB
. 
Cell-mediated immune response to human papillomavirus infection
.
Clin Diagn Lab Immunol
2001
;
8
:
209
20
.
38.
de Jong
A
,
van Poelgeest
MIE
,
van der Hulst
JM
,
Drijfhout
JW
,
Fleuren
GJ
,
Melief
CJM
, et al
Human papillomavirus type 16-positive cervical cancer is associated with impaired CD4+ T-cell immunity against early antigens E2 and E6
.
Cancer Res
2004
;
64
:
5449
55
.
39.
Rehermann
B
,
Nascimbeni
M
. 
Immunology of hepatitis B virus and hepatitis C virus infection
.
Nat Rev Immunol
2005
;
5
:
215
29
.
40.
Ramaswami
B
,
Popescu
I
,
Macedo
C
,
Metes
D
,
Bueno
M
,
Zeevi
A
, et al
HLA-A01-, -A03-, and -A024-binding nanomeric epitopes in polyomavirus BK large T antigen
.
Hum Immunol
2009
;
70
:
722
8
.
41.
Lima
MA
,
Marzocchetti
A
,
Autissier
P
,
Tompkins
T
,
Chen
Y
,
Gordon
J
, et al
Frequency and phenotype of JC virus-specific CD8+ T lymphocytes in the peripheral blood of patients with progressive multifocal leukoencephalopathy
.
J Virol
2007
;
81
:
3361
8
.
42.
Moretta
A
,
Pantaleo
G
,
Moretta
L
,
Cerottini
JC
,
Mingari
MC
. 
Direct demonstration of the clonogenic potential of every human peripheral blood T cell. Clonal analysis of HLA-DR expression and cytolytic activity
.
J Exp Med
1983
;
157
:
743
54
.
43.
Frebel
H
,
Richter
K
,
Oxenius
A
. 
How chronic viral infections impact on antigen-specific T-cell responses
.
Eur J Immunol
2010
;
40
:
654
63
.
44.
Chen
J
,
Liu
XS
. 
Development and function of IL-10 IFN-gamma-secreting CD4+ T cells
.
J Leukoc Biol
2009
;
86
:
1305
10
.
45.
Vence
L
,
Palucka
AK
,
Fay
JW
,
Ito
T
,
Liu
Y-J
,
Banchereau
J
, et al
Circulating tumor antigen-specific regulatory T cells in patients with metastatic melanoma
.
Proc Natl Acad Sci U S A
2007
;
104
:
20884
9
.
46.
Janetzki
S
,
Price
L
,
Britten
CM
,
van der Burg
SH
,
Caterini
J
,
Currier
JR
, et al
Performance of serum-supplemented and serum-free media in IFNgamma ELISpot assays for human T cells
.
Cancer Immunol Immunother
2010
;
59
:
609
18
.
47.
Hunder
NN
,
Wallen
H
,
Cao
J
,
Hendricks
DW
,
Reilly
JZ
,
Rodmyre
R
, et al
Treatment of metastatic melanoma with autologous CD4+ T cells against NY-ESO-1
.
N Engl J Med
2008
;
358
:
2698
703
.
48.
Burch
PA
,
Croghan
GA
,
Gastineau
DA
,
Jones
LA
,
Kaur
JS
,
Kylstra
JW
, et al
Immunotherapy (APC8015, Provenge) targeting prostatic acid phosphatase can induce durable remission of metastatic androgen-independent prostate cancer: a phase 2 trial
.
Prostate
2004
;
60
:
197
204
.