MAGE-6 Encodes HLA-DRβ1*0401-presented Epitopes Recognized by CD4+ T Cells from Patients with Melanoma or Renal Cell Carcinoma¹

MAGE-6 is a member of the cancer-testis family of tumor-associated antigens and is broadly expressed by tumors of diverse histologies, including melanoma and RCC⁴ (1, 2, 3, 4, 5). Although the MAGE family of proteins all bear strong homologies with each other, MAGE-6 is most highly homologous (98%) to MAGE-3 (1). Interestingly, the expression of MAGE-6/-3 by tumor tissue has been reported to be inversely correlated with clinical stage (6), and these two markers may be indicative of early premalignant lesions in situ (3, 7). This suggests that cellular immunity against MAGE-6 may modulate the incidence or time to incidence in individuals prone to develop certain types of cancer, the progression status of MAGE-6+ tumors and protection against recurrence of MAGE-6+ disease in the adjuvant setting.

Although RCC and melanoma are considered among the most responsive cancers to immunotherapy, many recent such approaches have been focused solely on the induction of CD8+ antitumor T cells in vivo (8, 9, 10). CD4+ T-cell recognition of these tumors is also clearly possible, with some RCC and melanoma in situ expressing MHC class II molecules (HLA-DR, HLA-DP, and HLA-DQ; Refs. 11, 12, 13). Cross-presentation of tumor-associated epitopes by HLA-DR+ patient DCs is also anticipated in vivo (14). The ability or inability of Th1-type CD4+ T cells to recognize MHC class II-presented tumor epitopes in situ may play a dominant role in determining whether resistance or susceptibility to disease progression occurs, respectively, and whether therapeutic approaches are both effective and durable (15, 16).

The current study was undertaken to define MAGE-6-derived peptides recognized in a HLA-DR-restricted fashion by CD4+ T cells and to use these probes to survey Th1-type CD4+ T-cell responses in patients with RCC or melanoma. These epitopes have promise as vaccine candidates for the treatment of patients with MAGE-6+ tumors and for the monitoring of evolving Th1-type CD4+ T-cell responses in patients receiving therapy.

The T2.DR4 (DRB1*0401+) cell line [kindly provided by Dr. Janice Blum (Indiana University School of Medicine, Indianapolis, IN] was used as the peptide-presenting cell in these studies (17). Melanoma cell line MEL-SLM2 (Storkus Lab Melanoma no. 2; HLA-DR4+, MAGE-6+) was generated from primary tumor cultures at the University of Pittsburgh. All cell lines were maintained in RPMI 1640 supplemented with 10% heat-inactivated fetal bovine serum, 100 IU/ml penicillin, 100 μg/ml streptomycin, and 10 mm l-glutamine (all reagents Life Technologies, Inc.).

The protein sequence of the MAGE-6 protein was obtained from GenBank (accession no. JC2360) and analyzed for HLA-DRB1*0401 binding peptides using a neural network algorithm (18, 19, 20). High-scoring 9 amino acid long sequences were typically extended by three amino acids on either flank using the genomic corresponding sequences. Alternatively, if high-scoring 9 amino acid long sequences were found to overlap, the longer overlapping sequences were synthesized, with amino acid extensions added to the end(s) of putative DR4 binding sequence(s). Overall, peptides were 15–23 amino acids in length (Table 1) and were synthesized by N-(9-fluorenyl)methoxycarbonyl chemistry by the University of Pittsburgh Cancer Institute’s Peptide Synthesis Facility (Shared Resource). Peptides were >90% pure based on high-performance liquid chromatography profile and tandem mass spectrometry mass spectrometric analysis performed by the UPCI Protein Sequencing Facility (Shared Resource).

The peptide binding determinations were performed using a solid-state competitive-binding assay as described previously (21).

Forty to 100 ml of patient or normal donor heparinized blood were obtained with informed consent under Institutional Review Board-approved protocols and diluted 1:2 with HBSS, applied to Ficoll-Hypaque gradients (LSM; Organon-Teknika, Durham, NC) per the manufacturer’s instructions, and centrifuged at 550 × g for 25 min at room temperature. Patient and normal donor information is provided in Table 2. Lymphocytes at the buoyant interface were recovered and washed twice with HBSS to remove residual platelets and Ficoll-Hypaque. HLA-DR4+ status was confirmed by flow cytometry using the anti-HLA-DR4-reactive mAb clone 359-13F10 [IgG, kindly provided by Dr. Janice Blum (Indiana University School of Medicine)] in indirect immunofluorescence assays. Cells were frozen in 90% FCS containing 10% DMSO (Sigma Chemical Co., St. Louis, MO) at 10⁷ lymphocytes/vial using controlled rate freezing technique. On the day of establishing DC-T-cell cultures, nonadherent cells were thawed and washed twice with HBSS. CD4+ T cells were then isolated using MACS (Miltenyi Biotec, Auburn, CA) antihuman CD4 beads and MiniMACS columns per the manufacturer’s protocol. CD4+ T-cell yields were typically 25–35% of starting PBMC numbers loaded, with purity exceeding 97% as assessed by flow cytometry.

Autologous DCs were prepared as previously described by 7-day culture of plastic-adherent PBMCs in granulocyte macrophage colony-stimulating factor and interleukin 4 (22). Harvested, nonadherent DCs were then loaded with antigen in one of three ways. DCs were either pulsed with 10 μm synthetic peptides for 3–4 h at 37°C or infected with empty recombinant adenovirus (Ad-ψ5) or adenovirus encoding the MAGE-6 (Ad-MAGE6) at a multiplicity of infection of 250 for 24 h at 37°C. The resulting antigen-loaded DCs were washed and used to stimulate purified CD4+ T cells at a 10–50:1 responder-to-stimulator ratio. Primary in vitro cultures were performed in AIM-V medium containing 5% human serum AB serum and 1 ng/ml of both rhIL-1 and rhIL-7 (Genzyme). In some cases, cultures were restimulated with the residual one-half of the DC-tumor stimulator preparation (cryopreserved for 1 week) on day 7 in AIM-V medium containing 5% HuAB serum and 10 IU/ml rhIL-2 (kind gift of Chiron Corporation, Emeryville, CA). In vitro-stimulated T cells were harvested either on day 7 and/or on day 14–17 and analyzed for MAGE-6 peptide/tumor specificity in ELISPOT assays.

ELISPOT assays were performed essentially as described previously (23, 24). HLA restriction of CD4+ T-cell responses was demonstrated by addition of the blocking anti-HLA-DR4 mAb 359-13F10 (5 μg/well).

PCR analyses were performed to determine patient HLA-DR4 genotype using a commercial PCR panel according to the manufacturer’s instructions (Dynal, Oslo, Norway) and peripheral blood lymphocyte as a source DNA. Reverse transcription-PCR analysis was also used to determine tumor expression of MAGE-6 mRNA. The following primer set was used: MAGE-6 (forward: TGGAGGACCAGAGGCCCCC, reverse: CAGGATGATTATCAGGAAGCCTGT, product size 728 bp with cycles: melting 94°C for 1 min, annealing 68°C for 1 min, extension 72°C for 1 min).

To identify a series of candidate peptides epitopes, we subjected the MAGE-6 cDNA sequence to a computer algorithm analysis designed to predict HLA-DR*0401 binding peptides (18, 19, 20). Nine amino acid-long core sequences were evaluated and scored from 0–10, with the highest scoring sequences taken to represent peptides most likely to bind to HLA-DRβ1*0401. Genomically encoded 3 amino acid long flanking extensions were added to core epitopes (i.e., 9-mers receiving an algorithm score ≥ 3) in syntheses. If overlapping 9-mers were identified, the cumulative sequence was produced, again with the flanking extensions added in the synthesis. A total of 15 peptides was produced for analysis (Table 1).

These selected synthetic peptides were then assessed for their comparative abilities to bind to purified HLA-DR4 molecules using a solid-state competitive-binding assay (21). The data reported as the dose of peptide capable of inhibiting 50% of the binding of a radiolabeled reference DR4 binding peptide (IC₅₀ in nm) to HLA-DR4w4 (-DRβ1*0401) are listed in Table 1. The strongest HLA-DR4 binding peptides tested were the MAGE-6_102–116 and MAGE-6_121–144 peptides, although the degree of affinity of these peptides for HLA-DR4 would be considered only moderate-to-low when compared with an extensive array of previously analyzed HLA-DR4/peptide binding events that used the same reference peptide (18, 19, 20).

In a preliminary screen of the immunogenicity of these peptides, we evaluated the ability of CD4+ T cells isolated directly from the peripheral blood of 14 HLA-DR4 (-DRβ1*0401)+ patients treated for melanoma (Table 2) to recognize these putative peptide epitopes using the IFN-γ ELISPOT assay. Six of these individuals (SLM1, SLM2, SLM5, SLM6, SLM9, and SLM10) were disease-free after surgery and/or immunotherapy, and we hypothesized that the current disease status might, in part, be attributed to circulating Th1-type antimelanoma CD4+ T cells. Reciprocally, the absence of disease (and chronic antigenic stimulation) may have been permissive of a biasing toward Th1-type immunity. As indicated in Table 1 and Fig. 1,A, a number of these disease-free patients displayed detectable frequencies of circulating Th0/Th1-type (i.e., IFN-γ secreting) CD4+ T cells that recognized a subset of the peptides selected for analysis (i.e., MAGE-6_102–116, MAGE-6_121–144, MAGE-6_{140–170 (140L)}, MAGE-6_145–160, MAGE-6_150–165, and MAGE-6_246–263). Interestingly, 7 HLA-DRB1*0401+ melanoma patients with active disease, either ocular melanoma (SLM12) or stage III or IV disease (SLM14, SLM16–19, and SLM21) exhibited minimal or no reactivity to any of the peptides analyzed in IFN-γ ELISPOT assays (Fig. 1,B). Among the 6 HLA-DRβ1*0401+ RCC patients evaluated, a pattern of reactivity similar to that of the melanoma patients was observed. The highest frequency response was observed for a patient that was successfully treated by surgery (i.e., SLR2) and had no evidence of disease at the time of testing, whereas those RCC patients with active disease (SLR3–SLR7) were poorly responsive to MAGE-6 peptides (Fig. 1 C).

None of the peptides tested was recognized or they were recognized extremely poorly by freshly isolated CD4+ T cells harvested from a series of 5 normal HLA-DRβ1*0401+ donors (Fig. 2,A). To evaluate whether normal HLA-DR4+ donors could recognize any of these sequences, if appropriately activated in vitro, donor T cells were stimulated and restimulated (1 week later) with autologous DCs that had been prepulsed with a given MAGE-6 peptide identified in Table 1. Resulting CD4+ T cells were used as responders against T2.DR4 target cells pulsed with the candidate DR4 binding peptides and against a HLA-DRβ1*0401+/MAGE-6+ melanoma cell line SLM2 in IFN-γ ELISPOT assays. Cryopreserved, freshly isolated (i.e., nonstimulated) CD4+ T cells obtained from each donor were also analyzed in these same assays to determine the basal in situ level and the impact of in vitro stimulation on the calculated frequencies of peptide-specific responder CD4+ T cells in these individuals. As shown in Fig. 2,B, high frequency responses were consistently observed from in vitro-stimulated CD4+ T cells against 6 of the MAGE-6-derived peptides in the IFN-γ ELISPOT assay. As depicted in Fig. 3, CD4+ T cells recognizing specific peptides also recognized the HLA-DRβ1*0401+, MAGE-6+ SLM2 melanoma cell line, and this recognition was inhibited by inclusion of anti-DR4 mAb in the assay.

We next investigated whether autologous DCs transfected with MAGE-6 cDNA could promote the induction of epitope-specific CD4+ T cells from normal DR4+ donors in vitro. As shown in Fig. 4, autologous DCs infected with Ad-MAGE6, but not Ad-ψ5, were able to promote the specific CD4+ Th1-type T-cell immunity. In Ad-MAGE6/DC-stimulated cultures obtained from 2 normal donors, IFN-γ-secreting CD4+ T cells preferentially recognized peptides MAGE-6_121–144, MAGE-6_140–170, and MAGE-6_150–165, with weaker responses noted against MAGE-6_102–116 and MAGE-6_246–263. These CD4+ T cells also recognized the MAGE-6+ SLM2 melanoma cell line in an HLA-DR4-restricted manner (Fig. 4). Overall, this suggests that immunogenic core epitope(s) resident within each of these biologically active, synthetic sequences is naturally processed and presented by MAGE-6+ Ghost DCs and by HLA-DR4+ tumors (i.e., SLM2) that constitutively express the relevant antigens.

We used peripheral blood T cells harvested from normal donors and patients with RCC or melanoma in conjunction with a peptide-binding algorithm and IFN-γ ELISPOT assays to initially identify a series of MAGE-6-derived peptides that are recognized by Th1-type, HLA-DR4-restricted CD4+ T cells. Of 15 peptides analyzed, 6 (encompassing 4 distinct, nonoverlapping sequences) were consistently recognized by CD4+ T cells isolated from melanoma and RCC patients (that were clinically free of disease at the time of analysis) and from normal donors after in vitro sensitization. Freshly isolated normal donor T cells and CD4+ T cells isolated from patients with active disease did not produce IFN-γ in response to any MAGE-6 peptides and typically two to three rounds of in vitro restimulation were required to expand MAGE-6-specific responders.

These data suggest that Th1-type CD4+ T cell responses to MAGE-6 may be important for immune-mediated control of melanoma/RCC tumor progression and may play a role in disease clearance and prevention of recurrence. The loss of such responses in patients with active disease may be attributable to tumor-induced immunosuppression or to the activation-induced cell death of these specific T-cell populations. Hence, therapeutic strategies targeting the potentiation of MAGE-6-reactive Th1-type CD4+ (and CD8+) T-cell responses may prove clinically beneficial.

On the basis of data provided in Table 1, peptides that bound HLA-DR4w4 (-DRB1*0401) with an IC₅₀ > 10,000 nm failed to consistently elicit IFN-γ production from CD4+ T cells, and those peptides that promoted CD4+ T-cell responses would be considered to be moderate-to-low affinity (i.e., IC₅₀s between 100-5000 nm) HLA-DR4 binders. It is hypothesized (although not investigated in this study) that variant analogues of these sequences designed to improve binding of these sequences to HLA-DR4 will promote enhanced immunoreactivity.

On the basis of the strong sequence homology between the MAGE-6 and MAGE-3 gene products (1), the recently defined MAGE-3_141–155 and MAGE-3_281–295 sequences (25) and the MAGE-3_119–134 sequence (26) that are also found unaltered in the MAGE-6 protein should represent MAGE-6 epitopes recognized by CD4+ T-cell lines in the context of HLA-DR11 and HLA-DR13, respectively. In the HLA-DR4 system evaluated in the current study, the MAGE-6_140–155 (which contains the MAGE-6_141–155 sequence) and MAGE-6_280–296 (containing the MAGE-6_281–295 sequence) were analyzed and found to be poorly or nonimmunogenic, suggesting that these epitopes do not represent global pan-DR-presented peptides. Interestingly, the MAGE-3_146–160 peptide, which differs from the MAGE-6 sequence only at position 156 [serine (S) in MAGE-3 versus aspartate (D) in MAGE-6], has recently been reported to serve as both an HLA-DR4 and HLA-DR7-restricted epitope for CD4+ T-cell reactivity (27). Our data suggests that the MAGE-6_145–160 peptide is also recognized in the context of HLA-DR4 (albeit weakly).

Data provided in Figs. 3 and 4 suggest that CD4+ T cells recognize naturally processed epitopes constitutively presented by MAGE-6+, HLA-DR4+ tumor cells, and/or cross-presented by autologous DCs infected with rAd-MAGE-6 (but not control rAd-ψ5). CD4+ T cells stimulated with MAGE-6-expressing DCs were able to recognize not only peptide-pulsed T2.DR4 cells but also the HLA-DR4-matched MAGE-6+ SLM2 melanoma cell line in a manner that was HLA-DR4 restricted. The range of MAGE-6 peptides recognized by Th1-type CD4+ T cells generated by priming with the complete MAGE-6 protein was essentially identical to that derived by in vitro sensitization with peptides (Figs. 2 and 4), suggesting that these sequences are not cryptic in vivo.

Interestingly, three of the MAGE-6 sequences recognized by CD4+ T cells in this study contain the amino acid asparagine (N), which may be posttranslationally modified into aspartate (D), potentially resulting in the HLA-DR4-presentation of the D-variant on cell surface of MAGE-6+/HLA-DR4+ DCs (via rAd-MAGE-6 infection) or MAGE-6+/HLA-DR4+ tumor cells. Skipper et al. (28) have previously demonstrated that this modification yields the naturally processed and HLA-A2-presented tyrosinase_368–376 (YMDGTMSQV) epitope, and we have also recently shown that a posttranslationally modified tyrosinase_365–381D, but not the genomically encoded tyrosinase_365–381N peptide, serves as a strong HLA-DR4-presented immunogen recognized by melanoma-reactive CD4+ T cells. The observed enhanced immunoreactivity of the tyrosinase_365–381D peptide could be attributed, in part, to its ∼100-fold better binding affinity to HLA-DRβ1*0401 (19). We will prospectively evaluate the immunogenicity of the D-variants of MAGE-6_121–144D, MAGE-6_140–170D, and MAGE-6_246–263D in promoting CD4+ T-cell responses.

The immunogenic MAGE-6 peptides identified in this study are presented by HLA-DR4, which is expressed by 15–20% of the North American population afflicted with RCC or melanoma (29). The clinical use of these sequences would be enhanced greatly if they could be presented in the context of additional HLA-DR alleles to tumor-reactive CD4+ T cells. Our preliminary analysis suggests that a number of these epitopes bind to a broad range of MHC class II alleles (Table 3). For instance, the MAGE-6_102–116 peptide binds to (at least) the HLA-DR1, HLA-DR4, HLA-DR7, HLA-DR8, HLA-DR9, HLA-DR11, HLA-DR13, HLA-DR15, and HLA-DRw53 (i.e., -DRβ4*0401) alleles that cover in excess of one-half of patients with RCC or melanoma (29). Prospective studies will assess the ability of Th1- and Th2-type CD4+ T cells restricted by non-DR4 class II alleles to recognize such pan-DR epitopes to justify the use of these peptides in vaccines and immune monitoring protocols treating patients with melanoma or RCC.

We thank Dr. Walter Olson and William Knapp for their excellent technical support and to Drs. Jan Mueller-Berghaus, Eva Pizzoferrato, Anna Kalinska, and Cynthia Brissette-Storkus for their careful review and comments during the generation of this manuscript.