CD4+ helper T cells play a critical role in orchestrating host immune responses, including antitumor immunity. The limited availability of MHC class II–associated tumor antigens is still viewed as a major obstacle in the use of CD4+ T cells in cancer vaccines. Here, we describe a novel approach for the identification of MHC class II tumor-associated antigens (TAAs). By combining two-dimensional liquid chromatography and nanoelectrospray ionization tandem mass spectrometry, we developed a highly sensitive method for the detection of human leukocyte antigen (HLA)-DR–associated peptides of dendritic cells upon exposure to necrotic tumor cells. This approach led to the identification of a novel MHC class II–restricted TAA epitope derived from melanotransferrin. The epitope stimulated T cells derived from melanoma patients and healthy individuals and displayed promiscuity in HLA-DR restriction. Moreover, the same peptide was also presented by MHC class II–positive melanoma cells. This strategy may contribute to increase the number of tumor epitopes presented by MHC class II molecules and may support the development of more efficacious vaccines against cancer.

T cells play a crucial role in the induction and maintenance of antitumor immunity. This could be shown in animal models and human cancer therapy (13). The activation of antitumor T-cell immunity relies on the recognition of tumor-associated antigens (TAAs) that bear immunogenic T-cell epitopes expressed on tumor cells. The identification of the first T-cell epitope of the tumor antigen MAGE (4) paved the way for the development of cancer vaccines, which trigger cellular immune responses executed by tumor-specific T cells. Most attempts focused on the activation of tumor-specific CTLs as they can directly lyse tumor cells. Human clinical trials applying defined MHC class I–restricted tumor antigenic peptides indicated that T-cell responses are readily detectable in vaccinated tumor patients (5, 6). Accordingly, inhibition of tumor growth or tumor regression could be shown in clinical studies (79). However, the overall immune responses were often weak and transient (10, 11), and a clear association between immunologic and clinical responses has rarely been observed.

Possible reasons for the limited success of antitumor vaccinations are loss of TAAs on tumor tissue, leading to tumor escape variants (12), down-regulation of components of the antigen processing and presentation pathway (13) or mechanisms of immunosuppression exerted by the tumor (1416). Another rationale may be the use of vaccines that rely exclusively on CD8+ T-cell immunity to eradicate cancer cells. An optimal vaccine, however, might require CD8+ and CD4+ T-cell antigens to generate a strong and long-lasting antitumor response (17, 18).

Several studies have shown the essential role of CD4+ T cells in the elimination of tumors, even if MHC class II molecules are absent on the tumor tissue (19, 20). Animal models have shown the importance of TAA-specific CD4+ T cells for recruiting, priming, and maintaining of CTLs and for their ability to infiltrate tumors (21, 22). Even in the absence of CTLs, tumor regression can be mediated by CD4+ T cells through direct and indirect killing mechanisms (2325) as well as via recruitment and activation of eosinophils, macrophages, and B cells (26).

Although many tumor-specific CTL epitopes are known, only a very small number of tumor-specific helper T-cell epitopes have been characterized. Accordingly, thus far, only very few vaccination studies have included epitopes that stimulate helper T cells (2729). This is mainly due to the lack of effective methods to discover MHC class II–restricted tumor antigenic peptides although efforts are being made. Still, for most HLA haplotypes and most types of tumors, appropriate helper T-cell epitopes are elusive.

In the present study, we describe a novel strategy for the identification of MHC class II–restricted tumor antigens. Human monocyte–derived dendritic cells were pulsed with necrotic tumor cells and peptides eluted from HLA-DR molecules were sequenced by a combination of two-dimensional capillary liquid chromatography and electrospray ionization tandem mass spectrometry (LC ESI-MS/MS). Sequencing revealed several novel epitopes originating from candidate tumor antigens. One of these epitopes was derived from the melanoma-associated protein melanotransferrin (668-683). Melanotransferrin (668-683) was very potent in stimulating CD4+ T cells from healthy individuals and melanoma patients and seems a candidate epitope to be included in peptide-based immunotherapy of malignant melanoma. The described approach might offer a powerful strategy in the discovery of novel tumor-specific helper T-cell epitopes from different types of tumor.

Antibodies. The hybridomas producing HLA-DR–specific monoclonal antibody (mAb) L243 (recognizing HLA-DRαβ dimers) and mAb L235 (recognizing human melanotransferrin) were purchased from the American Type Culture Collection (Manassas, VA). Antibodies were purified from hybridoma supernatants via protein A-Sepharose. Secondary antibodies used in flow cytometry: goat anti-mouse coupled to FITC (Dianova, New York, NY). Capture and detection antibody pairs used in ELISA were purchased from BD PharMingen (San Diego, CA): IFN-γ capture antibody NIB42 and IFN-γ detection antibody 4S.B3.

Cell lines. The human melanoma cell lines UKRV-Mel-15a, Ma-Mel-18a, UKRV-Mel-17, MZ-2 (30), SK-Mel-28 (31), and Mel-Juso (32) as well as the T-cell/B-cell hybrid cell line T2 (33) stably transfected with DR4 (DRB1*0401; T2.DR4) or DR1 (DRB1*0101; T2.DR1) were maintained in complete RPMI 1640 (Life Technologies/Bethesda Research Laboratories, Rockville, MD).

Preparation of necrotic cells. Tumor and control cells were lysed by four cycles of freezing (in liquid nitrogen) and subsequent thawing (at room temperature). Lysis was monitored by light microscopy and attained an efficacy of 80% to 90%. The lysate was used for pulsing dendritic cells in a ratio of 3:1.

Generation of dendritic cells. Dendritic cells were differentiated from peripheral blood monocytes, as described (32). Briefly, we isolated monocytes from peripheral blood mononuclear cells (PBMC) of HLA-typed donors by positive selection with anti-CD14 magnetic beads (Miltenyi Biotech, Auburn, CA) and cultured them in complete RPMI supplemented with granulocyte macrophage-colony stimulating factor (50 ng/mL, Leukomax; Novartis, east Hanover, NJ) and interleukin-4 (IL-4, 3 ng/mL, R&D Systems, Minneapolis, MN). Maturation was induced on day 5 by adding tumor necrosis factor-α (TNF-α, 1 ng/mL, R&D Systems) or lipopolysaccharide (LPS, 1 μg/mL, Sigma, St. Louis, MO).

Isolation of CD4+ T cells. CD4+ T cells were isolated from PBMCs by negative selection using the CD4+ T-cell isolation kit (Milteny Biotech) consisting of a hapten antibody cocktail and anti-hapten antibodies coupled to magnetic beads. T cells were cultured in RPMI supplemented with 1% autologous human serum.

Generation of tumor antigen-specific T-cell lines. CD4+ T cells (1 × 106) were initially stimulated with autologous dendritic cells (2 × 105) that were pulsed with LPS (1 μg/mL) and antigenic peptide (20 μmol/L). After 5 days, IL-2 (1,250 units/mL, R&D Systems) was added. Responding T cells were restimulated in 10- to 14-day intervals with autologous dendritic cells pulsed with peptide (20 μmol/L) and grown in medium containing IL-2 (1,000 units/mL). After every round of restimulation, the specificity of the growing T cells was assessed by sandwich immunoassays for IFN-γ and IL-4.

Peptides. Peptides were synthesized by F-moc chemistry and purified by reversed-phase high-performance liquid chromatography (HPLC). Some peptides were biotinylated by coupling biotinyl-amino-hexanoic acid at the NH2 terminus during F-moc synthesis. HA (307-319), PKYVKQNTLKLAT, is an immunodominant epitope from influenza virus hemagglutinin that binds well to HLA-DR1, HLA-DR2, HLA-DR4, and HLA-DR5. CLIP (81-105), LPKPPKPVSKMRMATPLLMQALPMG, is derived from the MHC class II–associated invariant chain (Ii). Melanotransferrin (668-683), GQDLLFKDATVRAVPV, is derived from the long glycosyl phosphatidyl inositol (GPI)–anchored variant of melanotransferrin (splicing variant 1). CDC27 (768-782), MNFSWAMDLDFKGAN, is a known tumor antigen derived from the cell cycle protein CDC27. NY-ESO (115-132), PLPVPGVLLKEFTVSGNI, is a known tumor antigen derived from the tumor-specific protein NY-ESO. Vim (202-217), TLQSFRQDVDNASLAR, is derived from the intermediate filament protein Vimentin.

In vitro peptide binding assay. Purified detergent-solubilized HLA-DR1 (DRB1*0101 purified from T2.DR1), HLA-DR2 (DRB5*0101 from T2.DR2a), HLA-DR4 (DRB4*0401 from T2.DR4), or HLA-DR5 (DRB1*1101 from T2.DR5) molecules (20 nmol/L) were coincubated with biotinylated HA(307-319) peptide (200 nmol/L) and different concentrations of the corresponding competitor peptide for 24 hours at 37°C in binding buffer [50 mmol/L sodium phosphate, 50 mmol/L sodium citrate (pH 5.0), 0.1% Zwittergent 3-12]. Hence, samples were diluted 10-fold in PBS containing 0.05% Tween 20 and 1% bovine serum albumin (BSA) and incubated in a microtiterplate (Nunc, Naperville, IL), coated with the anti-HLA-DR mAb L243 (3 μg/mL) for 3 hours. Plates were developed by incubation for 45 minutes with 0.1 μg/mL streptavidin-Europium (Wallac Oy, Turku, Finland) according to the manufacturer's protocol. Quantification of binding of biotinylated HA(307-319) peptide to HLA-DR molecules was done using time-resolved Europium fluorescence and the VICTOR multilabel counter (Wallac Oy).

Flow cytometry. Dendritic cells were stained with the indicated antibodies (5 μg/mL) followed by goat anti-mouse-FITC. Analysis was done on a Becton Dickinson FACScalibur flow cytometer with the CellQuest software package (Becton Dickinson, Mountain View, CA). Background fluorescence was evaluated using irrelevant isotype-matched mAbs.

Sandwich ELISA. Supernatants were diluted 1:5 in PBS/Tween (0.05%) + BSA (1%) and incubated for 2 hours in microtiter plates (Wallac Oy) that were previously coated with the corresponding capture antibody. After intensive washing with PBS/Tween (0.05%), samples were incubated for 1 hour with biotinylated detection antibody. Quantitation was done as described for the in vitro peptide binding assay.

Analysis of T-cell responses in melanoma patients. Blood donations from melanoma patients were approved by the Institutional Review Board and an informed consent was given by all donors. Frozen PBMC from HLA-DRB1*0401-typed melanoma patients were thawed and seeded at 6 × 106 cells per well of a six-well plate in 3 mL Iscove's modified Dulbecco's medium/HEPES/glutamine (PAA Laboratories, Cölbe, Germany) supplemented with 10% human AB serum (PAA Laboratories). Peptide melanotransferrin (668-683), dissolved in DMSO, was added at a concentration of 10 μg/mL; control cells were incubated with DMSO only (day 0). One day later, cytokines IL-2 (20 units/mL) and IL-7 (10 ng/mL) were added to the cultures. After 7 to 14 days, cells were harvested and screened for their peptide reactivity by IFN-γ enzyme-linked immunospot (ELISPOT) assay using 105 PBMCs and 2 × 104 T2.DR4 target cells.

Isolation of HLA-DR–restricted peptides. Dendritic cells (4-6 × 106) were lysed in hypotonic lysis buffer containing 1% Triton X-100 and precipitated with mAb L243 conjugated to Sepharose beads. After several washing steps with double-distilled water (Merck, Darmstadt, Germany) peptides were eluted from the HLA-DR binding groove with 0.1% trifluoracetic acid at 37°C for 35 minutes and immediately lyophilized.

Mass spectrometry. Peptide identification was achieved by the multidimensional protein identification technology, which is based on a two-dimensional liquid chromatographic fractionation followed by mass spectrometric sequencing (LC-ESI-MS/MS). Briefly, lyophilized peptides were resuspended in 5% acetonitrile, 0.5% acetic acid, 0.012% heptafluoro butyric acid, and 1% formic acid. Peptide fractionation was achieved on a fused-silica microcapillary column (100-μm inner diameter) packed with C18 reverse-phase material C18-ACE 3 μm (ProntoSIL 120-3-C18 ACE-EPS; Bischoff Chromatography, Atlanta, GA) followed by cation exchange material (Partishere SCX; 5-μm particle size, Whatman, Hillsboro, OR). A fully automated 10-step gradient separation was carried out on an ULTIMATE nanoflow HPLC (Dionex, Sunnyvale, CA). The first six steps consisted of a short 5-minute salt elution step with increasing concentrations of ammonium acetate (0-225 mmol/L) followed by a nonlinear acetonitrile gradient (5-64%); the last four steps consisted each of a 20-minute salt elution step (250-1,500 mmol/L) followed by a nonlinear acetonitrile gradient. The HPLC column was directly coupled to a Finnigan LCQ ion trap mass spectrometer (Finnigan, San Jose, CA) equipped with a nano-LC ESI source. MS in the MS/MS mode was done according to the manufacturer's protocol. The identification of peptides was done by the SEQUEST algorithm against the Swiss-Prot database (http://www.expasy.org/sprot/sprot-top.html). Peptide evaluation was done according to strict quality requirements based on the sequence variables cross-correlation (Xcorr), delta cross-correlation (dCn), preliminary score (Sp), and ranking of the peptide (Rsp). That is, only those peptides were considered that showed a Xcorr value of 1.8 for singly charged ions, 2.0 for doubly charged ions, and 2.8 for triply charged ions. Furthermore, peptides had to exhibit a dCn value of >0.1, a Sp of >500, and an ion coverage of >50% to be considered. In addition, each spectrum was evaluated manually with respect to its plausibility.

Single target expression profiling. Single target expression profiling (STEP) analysis was done on a plate containing cDNA of cancer and normal tissues from a variety of sources. Data was generated by quantitative real-time PCR to the gene of interest and glycerine aldehyde-3-phosphate dehydrogenase (GAPDH; for standardization) in the same tube at the same time on a Taqman (Applied Biosystems, Foster City, CA), running 40 cycles. A primer pair was generated, specifically picking up a sequence from exon 9 of the melanotransferrin gene therefore being restricted to the long transcript of melanotransferrin from which the antigenic epitope was derived: 5′-Mtf1-I (5′-CAGTGCGTGTCAGCCAAGTC) and 3′-Mtf1-I (5′-TTCCCCGCCGTGTAAATGT). A site-specific probe sequence labeled with a fluorescent reporter dye and a fluorescent quencher dye was used for detection, P-Mtf1-I (5′-AGCGTCGACCTGCTCAGCCTGG). The relative expression of the gene of interested was E = 2ΔCT. ΔCT is the difference in the thermocycles of the GAPDH gene versus gene of interest after which the fluorescent signal pierces the threshold. The expression of GAPDH in each tissue was adjusted to the expression level of a panel of eight housekeeping genes.

Identification of HLA-DR–restricted tumor-associated peptides on dendritic cells. In the necrotic zones of tumors, dead tumor cells or tumor cell debris may be ingested by immature dendritic cells and delivered to the draining lymph node. In the T cell–rich areas, dendritic cells might initiate a T-cell response against tumor-specific epitopes presented by MHC molecules and induce tumor-infiltrating lymphocytes (34, 35). To mimic in vitro tumor cell uptake, processing and tumor antigen presentation by dendritic cells and to exploit this scenario for the identification of MHC class II–restricted tumor antigens, we established a strategy, denoted as MHC class II–associated peptide proteomics (MAPPs), which allows the identification of self and foreign HLA-restricted peptides on as little as 1 to 5 × 106 dendritic cells (ref. 36; Fig. 1A). That is, we stimulated dendritic cells with TNF-α and concomitantly exposed them to necrotic MHC class II–negative tumor cells. Dendritic cells were allowed to ingest necrotic tumor cells for 24 hours, whereupon dendritic cells were lysed, HLA-DR-associated peptides isolated via affinity beads and analyzed by two-dimensional LC ESI-MS/MS. Through analysis of dendritic cells that were exposed to necrotic tumor cells, compared with dendritic cells that were not, peptides that relied on the presence of tumor cells could be identified. Table 1 shows a list of peptides that were presented on HLA-DR molecules of dendritic cells, haplotype HLA-DRB1* 0401/1301, after encounter of necrotic Ma-Mel-18a melanoma cells and absent on unpulsed dendritic cells. Only those among ∼600 identified peptides are enlisted that were present in two independent measurements of tumor-pulsed dendritic cells and absent in both corresponding measurements of the autologous unpulsed dendritic cells. Peptides were further evaluated and selected according to several quality criteria (see Materials and Methods). By this means, 40 peptides derived from 13 different proteins were identified. Many of these peptides were derived from proteins that are ubiquitously expressed (e.g., from cytoskeletal or matrix proteins like tubulin or collagen, or constitutively expressed enzymes, like Lysyl hydroxylase or Ribophorin I). Others were derived from proteins that have been described to be overexpressed in certain cancers, like the regulator of G-protein signaling proteins (37), Cathepsin D (38), Sirp α1 (39), or Vimentin (40). These proteins, however, are also ubiquitously expressed on healthy tissues. One of the peptides that were selectively found only in the HLA-DR-associated peptide repertoire of tumor-pulsed dendritic cells was derived from the protein melanotransferrin, shown to be strongly expressed on melanoma cells (41). The melanotransferrin peptide melanotransferrin (668-683) as well as its length variant melanotransferrin (668-684) could be identified. The MS/MS spectrum of the identified epitope melanotransferrin (668-683) is depicted in Fig. 1B. The very same peptide could also be detected in an independent experiment in which dendritic cells of a different donor of the haplotype DRB1*0101/04011 were pulsed with necrotic Ma-Mel-18a cells (data not shown), suggesting that the epitope is HLA-DR4 restricted and very abundant on MHC class II molecules of dendritic cells after uptake of necrotic Ma-Mel-18a melanoma cells.

Figure 1.

MAPPs strategy for the differential analysis of MHC class II–restricted peptides of dendritic cells (DC). A, dendritic cells were stimulated with TNF-α in the presence or absence of necrotic tumor cells. HLA-DR molecules were purified and peptides eluted. Peptide mixtures were fractionated by two-dimensional nano-HPLC and peptides concomitantly sequenced online by ESI-MS/MS. Identified peptides of pulsed and unpulsed dendritic cells were comparatively analyzed and those peptides presented only after encounter of necrotic tumor cells were further evaluated. B, fragment spectrum of [M+H+] of peptide GQDLLFKDATVRAVPV (m/zobserved = 1,728.7) derived from melanotransferrin (668-683) identified only in the HLA-DR-associated peptide mixture of dendritic cells after exposure to necrotic Ma-Mel-18a cells. List of theoretical masses of the b- and y-ion series of the corresponding peptide. The fragments actually detected by MS/MS are in bold letters and are annotated in the fragment spectrum.

Figure 1.

MAPPs strategy for the differential analysis of MHC class II–restricted peptides of dendritic cells (DC). A, dendritic cells were stimulated with TNF-α in the presence or absence of necrotic tumor cells. HLA-DR molecules were purified and peptides eluted. Peptide mixtures were fractionated by two-dimensional nano-HPLC and peptides concomitantly sequenced online by ESI-MS/MS. Identified peptides of pulsed and unpulsed dendritic cells were comparatively analyzed and those peptides presented only after encounter of necrotic tumor cells were further evaluated. B, fragment spectrum of [M+H+] of peptide GQDLLFKDATVRAVPV (m/zobserved = 1,728.7) derived from melanotransferrin (668-683) identified only in the HLA-DR-associated peptide mixture of dendritic cells after exposure to necrotic Ma-Mel-18a cells. List of theoretical masses of the b- and y-ion series of the corresponding peptide. The fragments actually detected by MS/MS are in bold letters and are annotated in the fragment spectrum.

Close modal
Table 1.

Neoantigens on dendritic cells (DRB1*0401/1301) after pulse with necrotic Ma-Mel-18a melanoma cells

ProteinPeptide sequenceEpitope
Melanotransferrin (melanoma-associated antigen p97)* GQDLLFKDATVRAVPVG 668-684 
 GQDLLFKDATVRAVPV 668-683 
Regulator of G-protein signaling 11 (RGS11) PALLPTPVEPTAACGPGGGD 445-464 
Cathepsin D IHHKYNSDKSSTYVK 119-133 
 VDQNIFSFYLSRDPDAQPGGE 224-244 
 NIFSFYLSRDPDAQPGGE 227-244 
 NIFSFYLSRDPDAQPGGEL 227-245 
 IFSFYLSRDPDAQPG 228-242 
 NIFSFYLSRDPDAQPGG 227-243 
 NIFSFYLSRDPDAQPG 227-242 
Signal regulatory protein α 1 (Sirp α 1) EPNNHTEYASIQTSPQPA 445-462 
 NHTEYASIQTSPQPA 448-462 
Vimentin LTNDKARVEVERDNLAEDIM 163-182 
 TNDKARVEVERDNLAEDIM 164-182 
 DKARVEVERDNLAEDIM 166-182 
 NDKARVEVERDNLAEDI 165-181 
 NDKARVEVERDNLAEDIM 165-182 
 LQEEIAFLKKLHEEEIQ 226-242 
 EEIAFLKKLHEEEIQ 228-242 
 LQEEIAFLKKLHE 226-238 
 LQEEIAFLKKLHEEE 226-240 
Ribophorin I GAKNIEIDSPYEISRAPD 377-394 
 LPEGAKNIEIDSPYEISRAPD 374-394 
Lysyl hydroxylase 3 NVPTVDIHM*KQVGYEDQ 606-622 
Lysyl hydroxylase 1 NVPTIDIHMNQIGFERE 595-611 
 NVPTIDIHM*NQIGFER 595-610 
 VPTIDIHM*NQIGFER 596-610 
 NVPTIDIHM*NQIGFERE 595-611 
Putative adenosylhomocysteinase 2 KRTTDVMFGGKQVVVCG 272-288 
Protein pM5 precursor  NAM*TFTFDNVLPGKYK 541-556 
Collagen α 1(II) chain KSGDYWIDPNQGCTL 1225-1239 
 KSGDYWIDPNQGCT 1225-1238 
Tubulin β-5 chain GAKFWEVISDEHGIDPT 17-33 
 IGAKFWEVISDEHGIDPT 16-33 
 KFWEVISDEHGIDPT 19-33 
 AKFWEVISDEHGIDPT 18-33 
HUMAN 40S ribosomal protein S13 DSHGVAQVRFVTGNKIL 55-71 
 SDDVKEQIYKLAKKGLTPSQ 29-48 
 DDVKEQIYKLAKKGLTPS 30-47 
 VKEQIYKLAKKGLTPS 32-47 
ProteinPeptide sequenceEpitope
Melanotransferrin (melanoma-associated antigen p97)* GQDLLFKDATVRAVPVG 668-684 
 GQDLLFKDATVRAVPV 668-683 
Regulator of G-protein signaling 11 (RGS11) PALLPTPVEPTAACGPGGGD 445-464 
Cathepsin D IHHKYNSDKSSTYVK 119-133 
 VDQNIFSFYLSRDPDAQPGGE 224-244 
 NIFSFYLSRDPDAQPGGE 227-244 
 NIFSFYLSRDPDAQPGGEL 227-245 
 IFSFYLSRDPDAQPG 228-242 
 NIFSFYLSRDPDAQPGG 227-243 
 NIFSFYLSRDPDAQPG 227-242 
Signal regulatory protein α 1 (Sirp α 1) EPNNHTEYASIQTSPQPA 445-462 
 NHTEYASIQTSPQPA 448-462 
Vimentin LTNDKARVEVERDNLAEDIM 163-182 
 TNDKARVEVERDNLAEDIM 164-182 
 DKARVEVERDNLAEDIM 166-182 
 NDKARVEVERDNLAEDI 165-181 
 NDKARVEVERDNLAEDIM 165-182 
 LQEEIAFLKKLHEEEIQ 226-242 
 EEIAFLKKLHEEEIQ 228-242 
 LQEEIAFLKKLHE 226-238 
 LQEEIAFLKKLHEEE 226-240 
Ribophorin I GAKNIEIDSPYEISRAPD 377-394 
 LPEGAKNIEIDSPYEISRAPD 374-394 
Lysyl hydroxylase 3 NVPTVDIHM*KQVGYEDQ 606-622 
Lysyl hydroxylase 1 NVPTIDIHMNQIGFERE 595-611 
 NVPTIDIHM*NQIGFER 595-610 
 VPTIDIHM*NQIGFER 596-610 
 NVPTIDIHM*NQIGFERE 595-611 
Putative adenosylhomocysteinase 2 KRTTDVMFGGKQVVVCG 272-288 
Protein pM5 precursor  NAM*TFTFDNVLPGKYK 541-556 
Collagen α 1(II) chain KSGDYWIDPNQGCTL 1225-1239 
 KSGDYWIDPNQGCT 1225-1238 
Tubulin β-5 chain GAKFWEVISDEHGIDPT 17-33 
 IGAKFWEVISDEHGIDPT 16-33 
 KFWEVISDEHGIDPT 19-33 
 AKFWEVISDEHGIDPT 18-33 
HUMAN 40S ribosomal protein S13 DSHGVAQVRFVTGNKIL 55-71 
 SDDVKEQIYKLAKKGLTPSQ 29-48 
 DDVKEQIYKLAKKGLTPS 30-47 
 VKEQIYKLAKKGLTPS 32-47 

NOTE: HLA-DR-associated peptides of dendritic cells pulsed with Ma-Mel-18a and unpulsed autologous control dendritic cells were analyzed twice by two-dimensional LC ESI-MS/MS. Only peptides that were present in the peptide repertoire of both pulsed dendritic cell samples and absent in both unpulsed dendritic cell samples are listed.

*

Protein for which melanoma-associated expression has been described. No ubiquitous expression.

Ubiquitously expressed proteins that have been shown overexpressed in cancer tissues.

Ubiquitously expressed proteins without apparent association to cancer.

To verify the identity of the discovered peptide, the synthetic analogue of melanotransferrin (668-683) was applied to MS/MS fragmentation on the same Finnigan LCQ Ion Trap MS instrument. The fragmentation pattern of the synthetic peptide was almost identical to the naturally processed peptide thereby confirming the identity of melanotransferrin (668-683) (Fig. 2A). To our knowledge, neither CTL nor helper T-cell epitopes of melanotransferrin have been identified thus far.

Figure 2.

Fragmentation and HLA-DR binding of synthetic melanotransferrin (668-683). A, fragment spectrum [M+H+] of synthetic GQDLLFKDATVRAVPV. Observed peptide fragments are indicated in bold. B, peptide binding assay. The concentration of each peptide required to reduce binding of biotinylated HA (307-319) by 50% (IC50) through competition was determined by capture ELISA. The reciprocal 1/IC50 is given which directly correlates with peptide binding affinity. The IC50 values were determined in the context of HLA-DR4, HLA-DR1, HLA-DR2, and HLA-DR5 (left to right). Columns, means of three independent experiments for each haplotype; bars, ±SD.

Figure 2.

Fragmentation and HLA-DR binding of synthetic melanotransferrin (668-683). A, fragment spectrum [M+H+] of synthetic GQDLLFKDATVRAVPV. Observed peptide fragments are indicated in bold. B, peptide binding assay. The concentration of each peptide required to reduce binding of biotinylated HA (307-319) by 50% (IC50) through competition was determined by capture ELISA. The reciprocal 1/IC50 is given which directly correlates with peptide binding affinity. The IC50 values were determined in the context of HLA-DR4, HLA-DR1, HLA-DR2, and HLA-DR5 (left to right). Columns, means of three independent experiments for each haplotype; bars, ±SD.

Close modal

HLA-DR binding of melanotransferrin (668-683). The identified melanotransferrin peptide possessed two overlapping HLA-DR binding motives, which may confer binding to a diverse set of HLA-DR molecules. The first motif comprises anchor residues L-672, D-675, T-677, and A-680, whereas residues F-673, A-676, V-678, and V-681 constitute the second motif. As melanotransferrin (668-683) was identified in the peptide repertoire of two donors sharing the HLA-DRB1*0401 molecules, binding of the synthetic peptide to HLA-DR4 was assessed by an in vitro peptide binding assay (Fig. 2B). As expected, melanotransferrin (668-683) showed high binding affinity to HLA-DR4, even exceeding peptide HA (307-319), which is a strong binder in the context of HLA-DR4, HLA-DR1, HLA-DR2, and HLA-DR5. Melanotransferrin (668-683) was also superior to CDC27 (768-782) in binding to HLA-DR4. In contrast to CDC27 (768-782), which was identified as a tumor antigen in the context of HLA-DRB1*0401 (35), melanotransferrin (668-683) displayed mediocre binding affinities to additional HLA-DR alleles, such as HLA-DR1, HLA-DR2, and HLA-DR5, accounting for almost 50% of the HLA-DR genotypes among Caucasians. Haplotype promiscuity of melanotransferrin (668-683) was also higher than that of the tumor antigen NY-ESO (115-132), recognized by CD4+ T cells of patients with NY-ESO-expressing tumors (42). Thus, melanotransferrin (668-683) apparently exhibits a reasonable degree of promiscuity with regard to binding to allelic variants of HLA-DR.

Expression of melanotransferrin in melanoma cell lines. Melanotransferrin is a cell membrane protein that is attached to the cell surface through a GPI anchor. We compared the expression of melanotransferrin on melanoma cells and dendritic cells by flow cytometry using anti–melanotransferrin-specific mAb L235 (Fig. 3A). As expected from our experiments described above, dendritic cells apparently do not express melanotransferrin neither in the immature nor in the mature state. The melanoma cell line Ma-Mel-18a, however, which was used as antigen source, exhibited a relatively strong expression of melanotransferrin compared with a panel of other melanoma cell lines and gave rise to the presentation of melanotransferrin (668-683) by dendritic cells. Interestingly, when necrotic UKRV-Mel-15a cells that express substantially less melanotransferrin than Ma-Mel-18a cells were subjected to the same dendritic cells, no presentation of any melanotransferrin peptide could be observed (data not shown). Both UKRV-Mel-15a and Ma-Mel-18a melanoma cell lines were negative for MHC class II (Fig. 3B).

Figure 3.

Expression of melanotransferrin (MTf) and HLA-DR on melanoma cells. A, flow cytometric analysis of a panel of melanoma cells, immature (IM) and mature dendritic cells (DC; Mat, stimulated with 1 μg/mL LPS for 24 hours) stained with the melanotransferrin-specific mAb L235 (5 μg/mL). Columns, mean fluorescence intensity of the specific staining for each cell type from three experiments; bars, ±SD. Isotype controls for each cell type were set to a mean fluorescence intensity of 1. B, expression of HLA-DR on melanoma cell lines. Flow cytometry of melanoma cell lines Ma-Mel-18a, UKRV-Mel-15a, and UKRV-Mel-17 with mAb L243 (5 μg/mL) specific for HLA-DR. Solid lines, isotype control. One of three experiments with similar results.

Figure 3.

Expression of melanotransferrin (MTf) and HLA-DR on melanoma cells. A, flow cytometric analysis of a panel of melanoma cells, immature (IM) and mature dendritic cells (DC; Mat, stimulated with 1 μg/mL LPS for 24 hours) stained with the melanotransferrin-specific mAb L235 (5 μg/mL). Columns, mean fluorescence intensity of the specific staining for each cell type from three experiments; bars, ±SD. Isotype controls for each cell type were set to a mean fluorescence intensity of 1. B, expression of HLA-DR on melanoma cell lines. Flow cytometry of melanoma cell lines Ma-Mel-18a, UKRV-Mel-15a, and UKRV-Mel-17 with mAb L243 (5 μg/mL) specific for HLA-DR. Solid lines, isotype control. One of three experiments with similar results.

Close modal

MHC class II–positive UKRV-Mel-17 cells present melanotransferrin (668-683). We asked whether melanoma cells that coexpress both melanotransferrin and HLA-DR molecules would themselves present the melanotransferrin peptide. We therefore analyzed the HLA-DR-restricted self-peptide repertoire of the melanoma cell line UKRV-Mel-17, which strongly expresses both melanotransferrin and HLA-DRB1* 0401/03011 (Fig. 3A and B). As melanotransferrin (668-683) was identified in the context of HLA-DRB1*0401 (Table 1) and showed strong binding affinity for HLA-DR4 (Fig. 2B), we expected melanotransferrin (668-683) to be also present in the HLA-DR-restricted peptide repertoire of UKRV-Mel-17 cells.

HLA-DR-molecules of ca. 5 × 107 UKRV-Mel-17 cells were isolated, peptides eluted and subjected to two-dimensional LC ESI-MS/MS. 971 HLA-DR-associated peptides could be sequenced, by this approach. Table 2 displays a subset of these peptides. We listed only those peptides that were derived from proteins described to be associated with cancer. Indeed, melanotransferrin (668-683) and its elongation variant melanotransferrin (668-684) were among the set of HLA-DR-associated self-peptides presented by this melanoma cell line.

Table 2.

HLA-DR-restricted peptides of tumor-associated proteins eluted from UKRV-Mel-17 melanoma cells (DRB1*0401/03011)

ProteinPeptide sequenceEpitope
Melanotransferrin (melanoma-associated antigen p97) GQDLLFKDATVRAVPV 668-683 
 GQDLLFKDATVRAVPVG 668-684 
Gp100 (melanoma-associated ME20 antigen) NRQLYPEWTEAQR 45-57 
 NRQLYPEWTEAQRLD 45-59 
 RQLYPEWTEAQR 46-57 
 WNRQLYPEWTEAQR 44-57 
 WNRQLYPEWTEAQRLD 44-59 
MART-1 (Melan-A) APPAYEKLSAEQSPPPY 100-116 
 APPAYEKLSAEQSPPP 100-115 
Melanoma-associated chondroitinsulfate-proteoglycan NG2 (HSN tumor-specific antigen) GPWPQGATLRLDPTVLDAGEL 2053-2073 
 GPWPQGATLRLDPTVLDAGE 2053-2072 
 GATLRLDPTVLDAGEL 2058-2073 
 GPWPQGATLRLDPTVLDAGELA 2053-2074 
MMP-14 (MT-MMP 1, membrane-type-1 MMP) GDKHWVFDEASLEPG 384-398 
 GDKHWVFDEASLEPGYPK  
Syntenin 1 (melanoma differentiation–associated protein-9) ITSIVKDSSAARNGLL 218-234 
 ITSIVKDSSAARNGL 218-233 
 ITSIVKDSSAARN 218-231 
 ITSIVKDSSAARNGLLT 218-235 
ProteinPeptide sequenceEpitope
Melanotransferrin (melanoma-associated antigen p97) GQDLLFKDATVRAVPV 668-683 
 GQDLLFKDATVRAVPVG 668-684 
Gp100 (melanoma-associated ME20 antigen) NRQLYPEWTEAQR 45-57 
 NRQLYPEWTEAQRLD 45-59 
 RQLYPEWTEAQR 46-57 
 WNRQLYPEWTEAQR 44-57 
 WNRQLYPEWTEAQRLD 44-59 
MART-1 (Melan-A) APPAYEKLSAEQSPPPY 100-116 
 APPAYEKLSAEQSPPP 100-115 
Melanoma-associated chondroitinsulfate-proteoglycan NG2 (HSN tumor-specific antigen) GPWPQGATLRLDPTVLDAGEL 2053-2073 
 GPWPQGATLRLDPTVLDAGE 2053-2072 
 GATLRLDPTVLDAGEL 2058-2073 
 GPWPQGATLRLDPTVLDAGELA 2053-2074 
MMP-14 (MT-MMP 1, membrane-type-1 MMP) GDKHWVFDEASLEPG 384-398 
 GDKHWVFDEASLEPGYPK  
Syntenin 1 (melanoma differentiation–associated protein-9) ITSIVKDSSAARNGLL 218-234 
 ITSIVKDSSAARNGL 218-233 
 ITSIVKDSSAARN 218-231 
 ITSIVKDSSAARNGLLT 218-235 

NOTE: HLA-DR molecules were isolated from 5 × 107 cells, peptides eluted and subjected to two-dimensional LC ESI-MS/MS. Enlisted peptides are derived from proteins known to be tumor associated.

Other epitopes present in the self-peptide repertoire of UKRV-Mel-17 cells were derived from gp100, MART-1, melanoma-associated chrondoitinsulfate-proteoglycan, matrix metalloproteinase-14 (MMP-14), and melanoma differentiation–associated protein-9 (syntenin). From each parent protein, one epitope and several of the respective length variants could be identified (Table 2). The peptide derived from gp100 sequenced here has been described as helper T-cell epitope before (43). A helper T-cell epitope of MART-1 has already been discovered in the context of HLA-DRB1*0401 (44), The peptide identified here, however, has not been described before. The other tumor-associated MHC class II peptides listed here, to our knowledge, have not been identified, as yet.

Melanotransferrin (668-683) is a helper T-cell epitope. To assess whether melanotransferrin (668-683) may sensitize helper T cells, we isolated CD4+ T cells from PBMCs of an HLA-DRB1*0401-positive healthy donor and repeatedly stimulated them with autologous dendritic cells that had been pulsed with melanotransferrin (668-683). Melanotransferrin (668-683)–pulsed dendritic cells induced a helper T-cell type 1 (TH1) response, as indicated by the specific release of IFN-γ, whereas dendritic cells pulsed with control peptide did not (Fig. 4A). Release of TH2 cytokines, such as IL-4, could not be observed in response to melanotransferrin (668-683)–pulsed dendritic cells (data not shown). To verify that the melanotransferrin (668-683)–specific T-cell line recognized melanotransferrin (668-683) in the context of HLA-DR4, we did an MHC restriction analysis. The T-cell line was stimulated with peptide-pulsed T2.DR4 or T2.DR1 transfectants. T2.DR4 cells were capable of inducing a substantial T-cell response, whereas peptide-pulsed T2.DR1 cells provoked only a weak release of IFN-γ (Fig. 4B). T-cell recognition of T2.DR4 cells could be substantially inhibited by 10 μg/mL anti-HLA-DR mAb L243, whereas no effect was observed at the same concentration of MHC class I–specific control mAb W6/32.

Figure 4.

Stimulation of melanotransferrin [MTf (668-683)]–specific T cells. A, CD4+ T cells of a HLA-DRB1* 0401/0701 donor were isolated and probed for its specific recognition of melanotransferrin (668-683) after five rounds of restimulation with autologous dendritic cells pulsed with peptide. Dendritic cells were activated with LPS (1 μg/mL) and pulsed with 20 μmol/L of melanotransferrin (668-683) or control peptide CLIP (81-104) for 24 hours before T-cell stimulation. T cells were cocultured in a ratio of 10:1 with pulsed autologous dendritic cells for 16 hours. Supernatants were probed for secreted IFN-γ by sandwich ELISA. B, 5 × 104 T2.DR4 or T2.DR1 cells were used as antigen-presenting cells (APC) and incubated with 1 × 105 T cells for 24 hours in the presence of 20 μmol/L peptide or peptide + 10 μg/mL of the blocking antibodies W6/32 (anti MHC class I) or L243 (anti-HLA-DR). Secreted IFN-γ was determined as in (A). C, melanoma cells were either pretreated with 100 units/mL IFN-γ for 48 hours or left untreated. UKRV-Mel-17 or UKRV-Mel-15a melanoma cells (5 × 104) were then incubated with (+TCL) or without 1 × 105 melanotransferrin-specific T-cell lines for 24 hours in the presence of 10 μg/mL mAb W6/32 or L243 or peptide melanotransferrin (668-683). D, IFN-γ release of melanotransferrin (668-683)–reactive T-cell line in response to dendritic cells pulsed for 24 hours with LPS and various concentrations of melanotransferrin (668-683) or control CLIP (81-104). Columns (points), mean concentration in ng/mL (n = 3); bars, ±SD. *, P < 0.05 (significant). E, detection of melanotransferrin (668-683)–specific T cells in the peripheral blood of HLA-DRB1*0401+ melanoma patients. PBMCs (6 × 106) from melanoma patient EM (HLA-DRB1*0301 and HLA-DRB1*0401), MD (HLA-DRB1*0401 and HLA-DRB1*1101), and ASn (HLA-DRB1*0401 and HLA-DRB1*1302) were seeded in the absence or presence of peptide melanotransferrin (668-683). After an incubation period of 7 days (for EM and MD) and 14 days (for ASn), T cells were analyzed for their specificity in an IFN-γ ELISPOT assay by incubation with T2.DR4 cells as antigen presenters in the presence of melanotransferrin (668-683), Vim (202-217), or without (w/o) peptide. Columns, means of three-well (EM, MD) and two-well (Asn) determinations; bars, ±SD. *, P < 0.05, (significant).

Figure 4.

Stimulation of melanotransferrin [MTf (668-683)]–specific T cells. A, CD4+ T cells of a HLA-DRB1* 0401/0701 donor were isolated and probed for its specific recognition of melanotransferrin (668-683) after five rounds of restimulation with autologous dendritic cells pulsed with peptide. Dendritic cells were activated with LPS (1 μg/mL) and pulsed with 20 μmol/L of melanotransferrin (668-683) or control peptide CLIP (81-104) for 24 hours before T-cell stimulation. T cells were cocultured in a ratio of 10:1 with pulsed autologous dendritic cells for 16 hours. Supernatants were probed for secreted IFN-γ by sandwich ELISA. B, 5 × 104 T2.DR4 or T2.DR1 cells were used as antigen-presenting cells (APC) and incubated with 1 × 105 T cells for 24 hours in the presence of 20 μmol/L peptide or peptide + 10 μg/mL of the blocking antibodies W6/32 (anti MHC class I) or L243 (anti-HLA-DR). Secreted IFN-γ was determined as in (A). C, melanoma cells were either pretreated with 100 units/mL IFN-γ for 48 hours or left untreated. UKRV-Mel-17 or UKRV-Mel-15a melanoma cells (5 × 104) were then incubated with (+TCL) or without 1 × 105 melanotransferrin-specific T-cell lines for 24 hours in the presence of 10 μg/mL mAb W6/32 or L243 or peptide melanotransferrin (668-683). D, IFN-γ release of melanotransferrin (668-683)–reactive T-cell line in response to dendritic cells pulsed for 24 hours with LPS and various concentrations of melanotransferrin (668-683) or control CLIP (81-104). Columns (points), mean concentration in ng/mL (n = 3); bars, ±SD. *, P < 0.05 (significant). E, detection of melanotransferrin (668-683)–specific T cells in the peripheral blood of HLA-DRB1*0401+ melanoma patients. PBMCs (6 × 106) from melanoma patient EM (HLA-DRB1*0301 and HLA-DRB1*0401), MD (HLA-DRB1*0401 and HLA-DRB1*1101), and ASn (HLA-DRB1*0401 and HLA-DRB1*1302) were seeded in the absence or presence of peptide melanotransferrin (668-683). After an incubation period of 7 days (for EM and MD) and 14 days (for ASn), T cells were analyzed for their specificity in an IFN-γ ELISPOT assay by incubation with T2.DR4 cells as antigen presenters in the presence of melanotransferrin (668-683), Vim (202-217), or without (w/o) peptide. Columns, means of three-well (EM, MD) and two-well (Asn) determinations; bars, ±SD. *, P < 0.05, (significant).

Close modal

When T-cell lines were stimulated by melanoma cell line UKRV-Mel-17, which is presenting the melanotransferrin (668-683) epitope in the context of HLA-DRB1*0401, as identified by MS (Table 2), only a very weak T-cell response could be observed (Fig. 4C). This weak T-cell stimulation could be inhibited by coincubation with the anti-HLA-DR mAb L243. When UKRV-Mel-17 cells were pretreated with IFN-γ for 48 hours, carefully washed, and then cocultured with T cells, a stronger T-cell stimulation occurred which was abolished by mAb L243. IFN-γ pretreatment of UKRV-Mel-17 cells led to elevated expression of HLA-DR, melanotransferrin, and intercellular adhesion molecule-1 (data not shown), aspects that may all contribute to the stronger recognition of UKRV-Mel-17 cells by melanotransferrin (668-683)–specific T cells. Pulsing of UKRV-Mel17 cells with exogenous melanotransferrin (668-683) peptide to further increase the number of melanotransferrin (668-683)/HLA-DR complexes led to a roughly 25-fold elevation of IFN-γ release. When the same set of experiments was done with the MHC class II–negative control melanoma cell line UKRV-Mel-15a, no T-cell stimulation could be observed (Fig. 4C).

As the response of the T-cell line to naturally processed melanotransferrin (668-683) epitope presented by UKRV-Mel-17 was rather weak but could be increased by IFN-γ pretreatment or exogenous peptide administration, we supposed that the T-cell line that we generated from a healthy donor was a low-avidity melanotransferrin-specific T-cell line that required rather high numbers of surface peptide/MHC class II complexes. Peptide titration experiments using mature autologous dendritic cells as antigen-presenting cells revealed that T-cell responses could only be observed with peptide concentrations higher than 0.1 μmol/L (Fig. 4D). Our results show that the mass spectrometrically identified HLA-DR-restricted melanotransferrin peptide presented by dendritic cells after uptake of necrotic tumor cells is a true helper T-cell epitope as it proved to be immunogenic.

Recognition of melanotransferrin (668-683) by T cells of melanoma patients. We further wanted to explore whether melanotransferrin (668-683) would also be suitable to induce a T-cell response in late-stage melanoma patients, as in these patients potentially reactive T cells might have been anergized or deleted due to the induction of antigen-specific tolerance exerted by the tumor (45). PBMCs of three HLA-DRB1*0401-positive melanoma patients were isolated and incubated with or without melanotransferrin (668-683). After 7 to 14 days, these PBMCs were analyzed for their melanotransferrin (668-683)–specific reactivity in the presence of DRB1*0401-expressing T2 transfectants. Reactive T cells were enumerated by ELISPOT analysis. Melanotransferrin (668-683)–specific T-cell responses could be detected in all three patients when PBMCs were presensitized with melanotransferrin (668-683) as indicated by the significantly increased number of IFN-γ-producing cells (Fig. 4E). We conclude that late-stage melanoma patients do bear helper T cells that respond to the melanotransferrin (668-683) peptide antigen.

Expression of melanotransferrin in normal and cancer tissues. Melanotransferrin has been described to be melanoma associated and absent on most other normal tissues (41). To extend this knowledge and to verify melanotransferrin (668-683) as a tumor antigen that could be employed in immunotherapy against cancer, we assessed the expression of melanotransferrin (transcription variant 1, of which the epitope was derived) by real-time PCR in a panel of 189 normal and 119 cancer tissues. A selection of several normal versus cancer tissues is shown in Fig. 5A. Interestingly, not only melanoma cells but also lung cancer and colon cancer primary tissues exhibited elevated expression of melanotransferrin. In the case of lung cancer tissues, 4 of 14 assessed samples (∼30%) showed an expression that was 12- to 45-fold increased over the average expression of normal lung tissues. The average expression of all lung cancer samples was about eight times higher than the average expression of all normal lung tissues. With respect to colon cancer and colon cancer metastasis, elevation of melanotransferrin expression was not that distinct, however, in 9 of 40 samples (∼25%) expression levels were increased by >5-fold compared with normal tissues. No increased expression could be observed in prostate cancer (Fig. 5A), bladder, or kidney cancer (data not shown). Importantly, melanotransferrin was hardly expressed in normal tissues tested (Fig. 5A,, dotted line). Slightly elevated expression could only be observed in normal breast and small intestine tissues (data not shown). In depth analysis of mRNA expression in different types of lung cancer further revealed that melanotransferrin is overexpressed in different neoplastic lung cell types but most pronounced in large cell lung cancer cells (Fig. 5B). Based on these results, melanotransferrin seems specifically overexpressed in certain cancers, whereas it is absent on most normal tissues.

Figure 5.

Single target expression profiling for melanotransferrin. cDNA was generated from a panel of normal versus cancer tissues. Quantitative real time PCR was done with primers specific for melanotransferrin (MTf) and the reference gene GAPDH. Expression levels are given as arbitrary units based on the relative expression of melanotransferrin mRNA versus normalized GAPDH in the corresponding tissue. A, each filled box is indicative of the expression of melanotransferrin mRNA per sample. The numbers above the panels for each tissue show the average expression of all samples of the corresponding tissue (horizontal line). The following numbers of samples per tissue were analyzed: testis, n = 8; brain, n = 6; spleen, n = 5; muscle, n = 18; lymph node, n = 3; adipose, n = 19; lung, n = 21; lung cancer, n = 14; melanoma cell lines, n = 3; colon, n = 16; colon cancer, n = 25; colon cancer metastases, n = 15; prostate, n = 8; prostate cancer, n = 11. Dotted line, average expression of melanotransferrin in all normal tissues that were assessed. B, expression of melanotransferrin in different cell types of normal lung and lung cancer tissues. Expression is given in arbitrary units as in (A). Abbreviation: NSCLC, non–small cell lung cancer.

Figure 5.

Single target expression profiling for melanotransferrin. cDNA was generated from a panel of normal versus cancer tissues. Quantitative real time PCR was done with primers specific for melanotransferrin (MTf) and the reference gene GAPDH. Expression levels are given as arbitrary units based on the relative expression of melanotransferrin mRNA versus normalized GAPDH in the corresponding tissue. A, each filled box is indicative of the expression of melanotransferrin mRNA per sample. The numbers above the panels for each tissue show the average expression of all samples of the corresponding tissue (horizontal line). The following numbers of samples per tissue were analyzed: testis, n = 8; brain, n = 6; spleen, n = 5; muscle, n = 18; lymph node, n = 3; adipose, n = 19; lung, n = 21; lung cancer, n = 14; melanoma cell lines, n = 3; colon, n = 16; colon cancer, n = 25; colon cancer metastases, n = 15; prostate, n = 8; prostate cancer, n = 11. Dotted line, average expression of melanotransferrin in all normal tissues that were assessed. B, expression of melanotransferrin in different cell types of normal lung and lung cancer tissues. Expression is given in arbitrary units as in (A). Abbreviation: NSCLC, non–small cell lung cancer.

Close modal

The importance of tumor-reactive CD4+ helper T cells in the development of antitumor immunity has become increasingly clear over the past decade (17, 18, 46). The development of broadly applicable methods to expand the number of MHC class II–restricted tumor antigens, therefore, remains an important task in tumor immunotherapy. In tumor vaccination, the use of synthetic peptides compared with whole cell preparations or adoptive transfer approaches has the advantage that peptides can easily be manufactured and pharmaceutically formulated.

We developed a procedure that allows us to identify HLA-DR ligands of as little as 106 dendritic cells (36). Dendritic cells are likely to be most critical in mounting antitumor immunity, as they take up antigens from damaged tumor cells in tumor lesions, traffic to draining lymph nodes, and prime antigen-specific naive T cells (47, 48). To mimic this process in vitro, we incubated dendritic cells with necrotic tumor cells and identified HLA-DR-restricted antigenic peptides derived from tumor cells through a highly sensitive two-dimensional LC ESI-MS/MS method. We believe that this approach has several advantages over other current strategies:

(a) Approaches using T cells as screening tools have the limitation that T-cell clones from individual cancer patients need to be generated, which can be particularly difficult for to nonmelanoma tumors. (b) In case only low-affinity T-cell clones are available, T-cell screenings might give negative results, although the antigenic peptide is presented by the tumor. (c) Screening by T cells is laborious and time consuming as it depends on testing of overlapping peptides or peptides that have been predicted in silico by algorithms that are still far less reliable in the context of MHC class II than MHC class I molecules. (d) Epitopes identified by T-cell screening need to be confirmed by testing of the respective recombinant protein in a T-cell assay, because cryptic epitopes might have been identified. (e) Our approach circumvents the limitation that most tumors are MHC class II negative, as it identifies tumor antigens that are presented by MHC class II molecules of dendritic cells. (f) As whole tumor cell preparations are used in the present approach, it is independent of purification or cloning steps of preselected tumor antigens pulsed onto antigen presenting cells or cloned as invariant chain fusion proteins. This approach, however, bears the imminent danger that dendritic cells not only take up tumor-specific proteins but also housekeeping proteins expressed by any type of cell. To solve this issue, we subtract the broad panel of HLA-DR–associated self-peptides derived from dendritic cell–resident proteins (36) from those presented after ingestion of necrotic tumor material. Hence, we focus on epitopes that are either exclusively presented after ingestion of tumor cells and absent among the naturally occurring self-peptides on dendritic cells or peptides that are strongly up-regulated in the presence of tumor cells. Following this strategy, several epitopes could be identified, which were exclusively present on tumor-pulsed dendritic cells. However, we also found peptides of ubiquitously expressed proteins (e.g., vimentin, collagen, tubulin, and ribosomal protein S13), which were not present in the self-peptide repertoire of corresponding unpulsed dendritic cells. These proteins might particularly be set free in large quantities after tumor cell necrosis, thereby giving rise to neoepitopes. The LC-MS method applied here does not identify peptides by de novo sequencing but through comparison of real mass spectra with theoretical spectra of a protein database. To date, our approach therefore has the limitation that epitopes containing sequence mutations that are not covered by the protein database can not be detected. However, as such mutations often only occur in distinct individuals, mutated epitopes may be less broadly applicable for immunotherapy compared with those derived from nonmutated antigens, such as cancer testis and tissue-specific antigens expressed in a majority of cancer patients.

We concentrated our efforts on the peptide derived from melanotransferrin, because melanotransferrin has already been described to be a melanoma-associated protein (41). Melanotransferrin was not expressed in dendritic cells, but it was strongly expressed by the MHC class II–negative tumor cell line Ma-Mel-18a, which was used as an antigen source in the presence of dendritic cells. This provided evidence that melanotransferrin was transferred from necrotic Ma-Mel-18a melanoma cells to dendritic cells, where it was processed and presented. Interestingly, the very same melanotransferrin epitope could also be detected as a natural self-peptide on the melanoma cell line UKRV-Mel-17, indicating that the peptide can indeed be generated by natural processing and was unlikely to be created exogenously by proteases liberated during necrosis.

Among the naturally presented self-peptides of the tumor cell line UKRV-Mel-17, further epitopes from tumor-specific proteins could be identified, one of which, gp100 (44–59), has already been described by others and proved to be immunogenic (43). We are therefore confident that the other novel HLA-DR-restricted peptides derived from MART-1, melanoma-associated chondroitinsulfate-proteoglycan, MMP-14, and Syntenin 1 may also qualify as MHC class II–restricted tumor antigens, although T-cell assays will have to confirm the immunogenicity of these peptides.

For melanotransferrin (668-683), the capacity to activate T cells could be shown by in vitro T-cell sensitization experiments. Reactive T cells showed a high level of melanotransferrin (668-683)–specific IFN-γ release, indicative of a TH1 response. This result proves that nontolerant melanotransferrin (668-683)–specific naive T cells exist in the naturally occurring T-cell repertoire, which could be stimulated by peptide vaccination.

The induction of antigen-specific unresponsiveness by cancer cells seems one of the predominant means by which tumors evade the attack of the immune system. It is thus likely that antigen-specific tolerance among T cells is of paramount importance for tumor survival. Tolerance induction by tumors has been shown for both CD4+ and CD8+ T cells (49, 50), and it is conceivable that in tumor patients, T-cell stimulation against certain antigens may fail due to such tolerization mechanisms. Hence, it was interesting to observe that peptide sensitization experiments carried out with PBMCs of HLA-DRB1*0401-positive late-stage melanoma patients revealed that after a single round of stimulation with melanotransferrin (668-683), substantial T-cell responses could be detected. The fact that a single round of restimulation already induced a significant T-cell response suggests that either the precursor frequency of melanotransferrin (668-683)–reactive T cells is high or that melanotransferrin (668-683)–specific T cells proliferate well. This finding supports our contention that under favorable conditions epitope melanotransferrin (668–683) may be capable of inducing an anti–melanotransferrin-specific immune response also in melanoma patients.

Ideally, tumor peptides to be included in vaccination trials should be applicable to a broad range of tumor patients bearing a variety of HLA haplotypes. A certain degree of promiscuity in haplotype restriction is therefore desirable. Our peptide binding assays showed that melanotransferrin (668-683) exhibited a very pronounced binding affinity for HLA-DR4 but also moderate affinities for HLA-DR1, HLA-DR2, and HLA-DR5. These four HLA-DR alleles together cover ∼50% of the HLA-DR genotypes of the Caucasian population, implying that melanotransferrin (668-683) may be a broadly applicable tumor peptide vaccine.

More importantly, tumor antigens to be included in cancer vaccination trials have to exhibit tumor-specific expression or at least substantial overexpression in tumors compared with healthy tissues, to prevent the induction of potentially life-threatening autoreactive immune responses against tumor-unrelated tissues. Our expression profiling extended already existing melanotransferrin expression data (41, 51, 52) and showed that melanotransferrin mRNA is present only in trace amounts in the assessed normal tissues. However, it was strongly overexpressed in tumor tissues (e.g., melanoma) but also certain other cancer types such as lung and colon cancer, supporting our contention that melanotransferrin may qualify as a shared tumor antigen in immunotherapy.

Early studies using recombinant vaccinia virus vaccines against melanotransferrin in mice and Macaca fascicularis monkeys have shown that cell-mediated and humoral immune responses could be induced against melanotransferrin-expressing xenografts and transfected syngeneic tumor cells resulting in in vivo rejection of tumor cells (53, 54). Despite the expression of trace amounts of cross-reactive melanotransferrin in normal tissues of primates, no adverse effects or normal tissue damage could be observed after eliciting the immune response against melanotransferrin (53). This suggests that vaccination with melanotransferrin (668-683) against melanotransferrin-expressing tumors may also be safely applicable in man.

In conclusion, we have shown that our MAPPs strategy based on human dendritic cells and LC/MS-MS is suitable for the identification of tumor-associated MHC class II peptide antigens, which may be of value for vaccination against cancer. We believe that MAPPs provides a powerful complementation to already existing methods that aim at expanding the number of tumor-specific helper T-cell epitopes. Future vaccination with appropriate combinations of cytotoxic CD8+ and CD4+ helper T-cell epitopes may contribute to increase the success rate of cancer immunotherapy.

Grant support: Wilhelm-Sander-Stiftung.

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.

We thank Nadine Daniel (Roche Pharmaceuticals, Basel, Switzerland) for expert technical assistance, Bernd Müller (Roche Pharmaceuticals) for help with MS, and Silke Schnell for helpful discussions during her diploma thesis.

1
Houghton AN, Gold JS, Blachere NE. Immunity against cancer: lessons learned from melanoma.
Curr Opin Immunol
2001
;
13
:
134
–40.
2
Rosenberg SA. Progress in human tumor immunology and immunotherapy.
Nature
2001
;
411
:
380
–4.
3
Stevanovic S. Identification of tumor-associated T-cell epitopes for vaccine development.
Nat Rev Cancer
2002
;
2
:
514
–20.
4
Van der Bruggen P, Traversari C, Chomez P, et al. Gene encoding an antigen recognized by cytolytic T lymphocytes on a human melanoma.
Science
1991
;
254
:
1643
–7.
5
Jager E, Maeurer M, Hohn H, et al. Expansion of Melan A-specific cytotoxic T lymphocytes in a melanoma patient responding to continued immunization with melanoma-associated peptides.
Int J Cancer
2000
;
86
:
538
–47.
6
Disis ML, Gooley TA, Rinn K, et al. Generation of T-cell immunity to the HER-2/neu protein after active immunization with HER-2/neu peptide-based vaccines.
J Clin Oncol
2002
;
20
:
2624
–32.
7
Rosenberg SA, Yang JC, Schwartzentruber DJ, et al. Immunologic and therapeutic evaluation of a synthetic peptide vaccine for the treatment of patients with metastatic melanoma.
Nat Med
1998
;
4
:
321
–7.
8
Nestle FO, Alijagic S, Gilliet M, et al. Vaccination of melanoma patients with peptide- or tumor lysate-pulsed dendritic cells.
Nat Med
1998
;
4
:
328
–32.
9
Marchand M, van Baren N, Weynants P, et al. Tumor regressions observed in patients with metastatic melanoma treated with an antigenic peptide encoded by gene MAGE-3 and presented by HLA-A1.
Int J Cancer
1999
;
80
:
219
–30.
10
Lee PP, Yee C, Savage PA, et al. Characterization of circulating T cells specific for tumor-associated antigens in melanoma patients.
Nat Med
1999
;
5
:
677
–85.
11
Panelli MC, Wunderlich J, Jeffries J, et al. Phase 1 study in patients with metastatic melanoma of immunization with dendritic cells presenting epitopes derived from the melanoma-associated antigens MART-1 and gp100.
J Immunother
2000
;
23
:
487
–98.
12
Khong HT, Restifo NP. Natural selection of tumor variants in the generation of “tumor escape” phenotypes.
Nat Immunol
2002
;
3
:
999
–1005.
13
Seliger B, Ritz U, Abele R, et al. Immune escape of melanoma: first evidence of structural alterations in two distinct components of the MHC class I antigen processing pathway.
Cancer Res
2001
;
61
:
8647
–50.
14
Tada T, Ohzeki S, Utsumi K, et al. Transforming growth factor-β-induced inhibition of T cell function. Susceptibility difference in T cells of various phenotypes and functions and its relevance to immunosuppression in the tumor-bearing state.
J Immunol
1991
;
146
:
1077
–82.
15
Gabrilovich DI, Chen HL, Girgis KR, et al. Production of vascular endothelial growth factor by human tumors inhibits the functional maturation of dendritic cells.
Nat Med
1996
;
2
:
1096
–103.
16
Salazar-Onfray F. Interleukin-10: a cytokine used by tumors to escape immunosurveillance.
Med Oncol
1999
;
16
:
86
–94.
17
Topalian SL. MHC class II restricted tumor antigens and the role of CD4+ T cells in cancer immunotherapy.
Curr Opin Immunol
1994
;
6
:
741
–5.
18
Wang RF. The role of MHC class II-restricted tumor antigens and CD4+ T cells in antitumor immunity.
Trends Immunol
2001
;
22
:
269
–76.
19
Ossendorp F, Mengede E, Camps M, Filius R, Melief CJ. Specific T helper cell requirement for optimal induction of cytotoxic T lymphocytes against major histocompatibility complex class II negative tumors.
J Exp Med
1998
;
187
:
693
–702.
20
Mumberg D, Monach PA, Wanderling S, et al. CD4(+) T cells eliminate MHC class II-negative cancer cells in vivo by indirect effects of IFN-γ.
Proc Natl Acad Sci U S A
1999
;
96
:
8633
–8.
21
Casares N, Lasarte JJ, de Cerio AL, et al. Immunization with a tumor-associated CTL epitope plus a tumor-related or unrelated Th1 helper peptide elicits protective CTL immunity.
Eur J Immunol
2001
;
31
:
1780
–9.
22
Marzo AL, Kinnear BF, Lake RA, et al. Tumor-specific CD4+ T cells have a major “post-licensing” role in CTL mediated anti-tumor immunity.
J Immunol
2000
;
165
:
6047
–55.
23
Cohen PA, Peng L, Plautz GE, Kim JA, Weng DE, Shu S. CD4+ T cells in adoptive immunotherapy and the indirect mechanism of tumor rejection.
Crit Rev Immunol
2000
;
20
:
17
–56.
24
Greenberg PD. Adoptive T cell therapy of tumors: mechanisms operative in the recognition and elimination of tumor cells.
Adv Immunol
1991
;
49
:
281
–355.
25
Qin Z, Blankenstein T. CD4+ T cell-mediated tumor rejection involves inhibition of angiogenesis that is dependent on IFN γ receptor expression by nonhematopoietic cells.
Immunity
2000
;
12
:
677
–86.
26
Hung K, Hayashi R, Lafond-Walker A, Lowenstein C, Pardoll D, Levitsky H. The central role of CD4(+) T cells in the antitumor immune response.
J Exp Med
1998
;
188
:
2357
–68.
27
van Driel WJ, Ressing ME, Kenter GG, et al. Vaccination with HPV16 peptides of patients with advanced cervical carcinoma: clinical evaluation of a phase I-II trial.
Eur J Cancer
1999
;
35
:
946
–52.
28
Brossart P, Wirths S, Stuhler G, Reichardt VL, Kanz L, Brugger W. Induction of cytotoxic T-lymphocyte responses in vivo after vaccinations with peptide-pulsed dendritic cells.
Blood
2000
;
96
:
3102
–8.
29
Slingluff CL, Yamshchikov G, Neese P, et al. Phase I trial of a melanoma vaccine with gp100 (280–288) peptide and tetanus helper peptide in adjuvant: immunologic and clinical outcomes.
Clin Cancer Res
2001
;
7
:
3012
–24.
30
Eichmuller S, Usener D, Jochim A, Schadendorf D. mRNA expression of tumor-associated antigens in melanoma tissues and cell lines.
Exp Dermatol
2002
;
11
:
292
–301.
31
Carey TE, Takahashi T, Resnick LA, Oettgen HF, Old LJ. Cell surface antigens of human malignant melanoma: mixed hemadsorption assays for humoral immunity to cultured autologous melanoma cells.
Proc Natl Acad Sci U S A
1976
;
73
:
3278
–82.
32
Johnson JP, Demmer-Dieckmann M, Meo T, Hadam MR, Riethmuller G. Surface antigens of human melanoma cells defined by monoclonal antibodies. I. Biochemical characterization of two antigens found on cell lines and fresh tumors of diverse tissue origin.
Eur J Immunol
1981
;
11
:
825
–31.
33
Salter RD, Howell DN, Cresswell P. Genes regulating HLA class I antigen expression in T-B lymphoblast hybrids.
Immunogenetics
1985
;
21
:
235
–46.
34
Wang RF, Rosenberg SA. Human tumor antigens for cancer vaccine development.
Immunol Rev
1999
;
170
:
85
–100.
35
Wang RF, Wang X, Atwood AC, Topalian SL, Rosenberg SA. Cloning genes encoding MHC class II-restricted antigens: mutated CDC27 as a tumor antigen.
Science
1999
;
284
:
1351
–4.
36
Rohn TA, Boes M, Wolters D, et al. Upregulation of the CLIP self peptide on mature dendritic cells antagonizes T helper type 1 polarization.
Nat Immunol
2004
;
5
:
909
–18.
37
Furuya M, Nishiyama M, Kimura S, et al. Expression of regulator of G protein signalling protein 5 (RGS5) in the tumour vasculature of human renal cell carcinoma.
J Pathol
2004
;
203
:
551
–8.
38
Leto G, Gebbia N, Rausa L, Tumminello FM. Cathepsin D in the malignant progression of neoplastic diseases [review].
Anticancer Res
1992
;
12
:
235
–40.
39
Chen TT, Brown EJ, Huang EJ, Seaman WE. Expression and activation of signal regulatory protein α on astrocytomas.
Cancer Res
2004
;
64
:
117
–27.
40
Singh S, Sadacharan S, Su S, Belldegrun A, Persad S, Singh G. Overexpression of vimentin: role in the invasive phenotype in an androgen-independent model of prostate cancer.
Cancer Res
2003
;
63
:
2306
–11.
41
Brown JP, Woodbury RG, Hart CE, Hellstrom I, Hellstrom KE. Quantitative analysis of melanoma-associated antigen p97 in normal and neoplastic tissues.
Proc Natl Acad Sci U S A
1981
;
78
:
539
–43.
42
Jager E, Jager D, Karbach J, et al. Identification of NY-ESO-1 epitopes presented by human histocompatibility antigen (HLA)-DRB4*0101–0103 and recognized by CD4(+) T lymphocytes of patients with NY-ESO-1-expressing melanoma.
J Exp Med
2000
;
191
:
625
–30.
43
Li K, Adibzadeh M, Halder T, et al. Tumour-specific MHC-class-II-restricted responses after in vitro sensitization to synthetic peptides corresponding to gp100 and Annexin II eluted from melanoma cells.
Cancer Immunol Immunother
1998
;
47
:
32
–8.
44
Zarour HM, Kirkwood JM, Kierstead LS, et al. Melan-A/MART-1(51–73) represents an immunogenic HLA-DR4-restricted epitope recognized by melanoma-reactive CD4(+) T cells.
Proc Natl Acad Sci U S A
2000
;
97
:
400
–5.
45
Pardoll D. Does the immune system see tumors as foreign or self?
Annu Rev Immunol
2003
;
21
:
807
–39.
46
Toes RE, Ossendorp F, Offringa R, Melief CJ. CD4 T cells and their role in antitumor immune responses.
J Exp Med
1999
;
189
:
753
–6.
47
Finn OJ. Cancer vaccines: between the idea and the reality.
Nat Rev Immunol
2003
;
3
:
630
–41.
48
Knutson KL, Disis ML. Tumor antigen-specific T helper cells in cancer immunity and immunotherapy.
Cancer Immunol Immunother
2005
;
54
:
721
–8.
49
Bogen B. Peripheral T cell tolerance as a tumor escape mechanism: deletion of CD4+ T cells specific for a monoclonal immunoglobulin idiotype secreted by a plasmacytoma.
Eur J Immunol
1996
;
26
:
2671
–9.
50
Staveley-O'Carroll K, Sotomayor E, Montgomery J, et al. Induction of antigen-specific T cell anergy: an early event in the course of tumor progression.
Proc Natl Acad Sci U S A
1998
;
95
:
1178
–83.
51
Alemany R, Vila MR, Franci C, Egea G, Real FX, Thomson TM. Glycosyl phosphatidylinositol membrane anchoring of melanotransferrin (p97): apical compartmentalization in intestinal epithelial cells.
J Cell Sci
1993
;
104
:
1155
–62.
52
Rothenberger S, Food MR, Gabathuler R, et al. Coincident expression and distribution of melanotransferrin and transferrin receptor in human brain capillary endothelium.
Brain Res
1996
;
712
:
117
–21.
53
Estin CD, Stevenson US, Plowman GD, et al. Recombinant vaccinia virus vaccine against the human melanoma antigen p97 for use in immunotherapy.
Proc Natl Acad Sci U S A
1988
;
85
:
1052
–6.
54
Kahn M, Sugawara H, McGowan P, et al. CD4+ T cell clones specific for the human p97 melanoma-associated antigen can eradicate pulmonary metastases from a murine tumor expressing the p97 antigen.
J Immunol
1991
;
146
:
3235
–41.

Supplementary data