Shared MHC Class II–Dependent Melanoma Ribosomal Protein L8 Identified by Phage Display

Antigens recognized by T helper (Th) cells in the context of MHC class II molecules have vaccine potential against cancer and infectious agents. We have described previously a melanoma patient's HLA-DR7–restricted Th cell clone recognizing an antigen, which is shared among melanoma and glioma cells derived from various patients. Here, this antigen was cloned using a novel antigen phage display approach. The antigen was identified as the ribosomal protein L8 (RPL8). A peptide of RPL8 significantly stimulated proliferation and/or cytokine expression of the Th cell clone and lymphocytes in four of nine HLA-DR7⁺ melanoma patients but not in healthy volunteers. The RPL8 antigen may represent a relevant vaccine target for patients with melanoma, glioma, and breast carcinoma whose tumors express this protein. [Cancer Res 2007;67(8):3555–9]

CD4⁺ T helper (Th) lymphocytes play a central role in the development of protective immunity against tumors. Adoptively transferred CD4⁺ Th cells, in the absence of CD8⁺ CTL, inhibit tumor growth in mice. In melanoma patients, spontaneous tumor regression was associated with CD4⁺ lymphocyte infiltrates (reviewed in ref. 1, 2).

A few human CD4⁺ Th cell lines and clones directed against various tumors have been described (3–7). Th antigens are usually recognized by MHC class II–restricted CD4⁺ Th cells after processing by antigen-presenting cells (APC) through the exogenous pathway.

Although peptides derived from tumor antigens originally defined by MHC class I–restricted CTL have been shown to stimulate Th cells (reviewed in ref. 8), peptide vaccines have not held promise in clinical trials (reviewed in ref. 9). Antigen vaccines expressing multiple potentially immunogenic epitopes may be superior to peptide vaccines expressing only a single epitope (10). Antigens recognized by noncytolytic MHC class II–restricted CD4⁺ T cells have been cloned using the Ii-cDNA fusion approach (11–16).

In the present study, we have cloned, using a novel antigen phage display approach, a Th antigen shared between melanomas and gliomas, each derived from various patients. A peptide derived from this antigen is recognized by the lymphocytes of several melanoma patients but not healthy donors.

Melanoma patients. Patient 35 has no evidence of disease after excision of primary lesion ∼23 years ago. All other melanoma patients included in this study had metastatic lesions excised between 2002 and 2005. Peripheral blood mononuclear cells (PBMC) were obtained from the patients' peripheral blood on the day of surgery (3522, 3523, 3626, and 3704) or up to 24 months after surgery (all other patients) with informed consent and under approved protocols.

Cell lines. Melanoma cell line WM35, EBV-B35 cell line, and Th35-1A Th cell clone were established and maintained in culture as described (7). COS-7L cells (Life Technologies-Invitrogen, Carlsbad, CA) were maintained in DMEM (Life Technologies-Invitrogen) supplemented with 10% fetal bovine serum.

Antibodies. Anti-HLA class II antibody B33.1 directed to HLA-DR was obtained from B. Perussia (Thomas Jefferson University, Philadelphia, PA) and normal mouse IgG from Cappel-ICN (Costa Mesa, CA). FITC-conjugated antihuman CD4 antibody (clone RPA-T4) and APC-conjugated antihuman interleukin-2 (IL-2) antibody (clone MQ1-17h12) were obtained from eBioscience (San Diego, CA).

Th antigen cloning approach using antigen phage display—overview. The currently used molecular cloning approach of HLA class II–dependent human melanoma and colon carcinoma antigens is based on fusing cDNA tumor libraries to MHC invariant chain (Ii) fragments with the aim of targeting the fusion proteins to the endosomal and lysosomal compartments (6). Fused libraries were transfected into 293 cells genetically engineered to express DRα, DRβ, DMA, DMB, and Ii and screened for reactivity with CD4⁺ T cells (Fig. 1, left).

We have expressed a melanoma cDNA library in bacteriophages followed by phage-library protein presentation to Th lymphocytes by APCs and identification of the relevant Th antigen by its capacity to induce proliferation and cytokine release in Th cells (Fig. 1, right). This strategy, in contrast to the cDNA library-Ii fusion approach, does not require prior knowledge of the MHC class II restriction element and it involves natural processing of the antigen by APC.

cDNA library construction and screening. mRNA was isolated from cultured WM35 cells using the FastTrack 2.0 kit (Invitrogen, Carlsbad, CA). PolyA⁺ RNA was converted to cDNA using the OrientExpress system (EMD Biosciences Novagen, La Jolla, CA) and ligated into T7Select10-3b vector (EMD Biosciences Novagen) according to the manufacturer's instructions. The ligated DNA was packed in vitro using T7 packing extract. The library was 1.2 × 10⁶ phages and was amplified on plates once and divided into 100 phages per pool. For screening, each pool was amplified once in liquid culture, and released phages were purified twice by polyethylene glycol/NaCl precipitation. The phage titers were determined, and ∼3 × 10³ phages were used per well of 96-well microtiter plate (392 wells screened) in lymphocyte proliferation and IFN-γ release assays (see below). Phages from one pool stimulated proliferation and IFN-γ release in Th35-1A cells. Stimulatory pools were subdivided until a single positive phage was identified.

Lymphocyte proliferation assay. This assay was done as described (17). For screening of Th35-1A cell reactivity with phage libraries, Th cells (1 × 10⁴ to 2 × 10⁴ per well of 96-well round-bottomed microtiter plates; Corning, Corning, NY) were cultured with irradiated autologous EBV-B cells (10⁴ per well) prepulsed with 1 × 10³ to 3 × 10³ phages. To determine Th or PBMC reactivity with peptide, adherent monocytes (5 × 10⁴ per well, obtained from PBMC) prepulsed with various concentrations (3.1–50 μmol/L) of peptide were incubated with Th35-1A cells or PBMC (5 × 10⁴ per well). Th cells or PBMC were stimulated with peptide-pulsed monocytes once or twice. Proliferative responses of lymphocytes were determined by standard [³H]thymidine ([³H]Tdr) incorporation assay. All determinations were done in triplicate. The lymphocyte proliferation inhibition assay with anti-HLA class II antibody B33.1 was done as described (17).

IFN-γ release assay. Supernatants obtained 48 h after Th cell stimulation with phage-pulsed EBV-B cells were tested for the presence of IFN-γ using an ELISA kit (Endogen, Rockford, IL).

Intracellular staining of lymphocytes for IL-2. Peptide-pulsed monocytes (see above) were incubated with lymphocytes at 37°C and cultures were treated with Golgi Stop (2 μmol/L; BD Biosciences, San Diego, CA) after 1 h. Five hours later, nonadherent cells were stained for CD4 and intracellular IL-2 using a kit (BD Biosciences). Samples were analyzed by CyAn ADP cytometer (DakoCytomation, Fort Collins, CO) and data were analyzed using FlowJo software (Ashland, OR).

DNA sequencing. DNA sequencing was done by The Wistar Institute's DNA sequencing facility. DNA and deduced amino acid sequence (ExPASy)⁸

comparisons were done with the BLAST program.⁹

Peptide design. Potentially DRB1*070101 binding epitopes were determined from the deduced amino acid sequence by using the Rammensee epitope prediction model¹⁰

and were limited to epitopes with a binding score of >20. Selected peptides were synthesized and high-pressure liquid chromatography purified by Invitrogen. The following peptides were used: VGLIAARRTGRLRGT, with a score of 24 [peptide 1, ribosomal protein L8 (RPL8) position 235–249]; TGRLRGTKTVQEKEN, with a score of 24 (peptide 2, RPL8 position 243–257); and RPGLLGASVLGLDDI, with a score of 22 (control peptide, telomerase reverse transcriptase).

Full-length RPL8 cloning. The GeneRacer kit (Invitrogen) and oligonucleotides based on the cDNA sequence of the phage that stimulated Th35-1A cell proliferation were used to determine the 5′ and 3′ end of RPL8 mRNA in WM35 cells. Both fragments (5′ and 3′ end) were sequenced and oligonucleotides were designed to clone full-length RPL8 cDNA by reverse transcription-PCR (RT-PCR; SuperScript III One-Step RT-PCR with Platinum Taq; Invitrogen).

Northern blot analysis. Northern blot analysis was done according to standard procedures. RNA levels were compared using a Storm PhosphorImager system (GE Healthcare, Piscataway, NJ).

Real-time PCR. Single-strand cDNA derived from tumor cells was subjected to real-time PCR using Power SYBR Green PCR Master Mix (Applied Biosystems, Foster City, CA) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH), actin, or RPL8-specific primers on an ABI Prism 7000 Sequencing Detection System (Applied Biosystems). The data were analyzed using 7000 System SDS software (Applied Biosystems).

Statistical analyses. Differences between experimental and control values were analyzed for significance by Student's two-sided t test or nonparametric Wilcoxon test using Proc NPAR1WAY in SAS/STAT software (SAS Institute, Inc., Cary, NC).

Library screening and clone identification. A single phage clone of the cDNA library from WM35 melanoma cells induced proliferation and IFN-γ release in Th35-1A cells. The clone had an insert of 189 bp, encoding an open reading frame of 58 amino acids, and represented the COOH-terminal part of RPL8 (18). The nucleotide sequence of full-length RPL8 subsequently cloned from WM35 melanoma cells was 100% identical with the published RPL8 sequence (Genbank GI:15431305; data not shown). Thus, the WM35 antigen recognized by Th35-1A cells was not mutated, which is in contrast to the other cloned mutated Th antigens or Th antigens using alternative reading frames (12–16).

Epitope determination. To confirm that RPL8 was recognized by Th35-1A, we determined the peptide epitope recognized by Th35-1A. This epitope was predicted to associate with HLA-DR7 (7). The deduced amino acid sequence of the cloned cDNA contains two potential DR7 (DRB1*070101) binding sites (see Materials and Methods). Only peptide 2 (TGRLRGTKTVQEKEN) was recognized by Th35-1A after presentation by autologous monocytes, and peptide recognition was HLA class II dependent (Fig. 2A). Th35-1A proliferation was peptide concentration dependent (Fig. 2B). Allogeneic DR7⁺ monocytes presented peptide 2 to Th35-1A cells (Fig. 2C), whereas DR7⁻ monocytes did not (Fig. 2D).

RPL8 protein is a component of the 60S subunit of ribosomes and is involved in protein synthesis (18). It is expressed by all normal cell.¹¹

RPL8 RNA is overexpressed in metastatic versus primary carcinomas (19). In light of the ubiquitous expression of RPL8 RNA (Table 1), it was surprising that some, but not all, tumor cell lysates derived from different patients stimulated proliferation of Th35-1A (Table 1). Because an antibody to RPL8 is not available, RPL8 protein levels and possible post-translational modifications in tumors of various tissue origins could not be investigated.

Recognition of peptide 2 by lymphocytes of HLA-DR7⁺ melanoma patients. PBMC from two of four HLA-DR7⁺ melanoma patients tested significantly proliferated after stimulation with RPL8 peptide 2 (Table 2). PBMC from four of seven HLA-DR7⁺ patients tested showed peptide reactivity as determined by intracellular staining for IL-2 (Table 2). Not all nine HLA-DR7⁺ patients' PBMC could be tested in both assays because we lacked sufficient numbers of PBMC. However, two patients (patients 35 and 3533) tested in both assays showed lymphocyte responses in both assays. Lymphocytes from two HLA-DR7⁻ patients and four healthy donors were nonresponsive in both assays (Table 2). Tumor RPL8 RNA levels were similar in all patients tested (Table 2).

We have successfully cloned the antigen, RPL8, recognized by a CD4⁺ HLA class II–restricted Th cell clone (Th35-1A) directed against melanoma. Until now, only the Ii-cDNA library fusion approach developed by Wang et al. (15) has been available for the molecular cloning of antigens recognized by MHC class II–restricted Th cells. The major limitation of the cDNA library-Ii fusion approach is the need for prior knowledge and isolation of the MHC class II restriction element used by the Th cell. We have developed a novel approach to the cloning of MHC class II–restricted antigens. This strategy does not require prior knowledge of the MHC class II restriction element used by Th cells for antigen recognition as phage-expressed cDNA libraries are presented to Th cells by MHC class II–positive autologous EBV-transformed B cells. Natural processing of the antigen by APC is another potential advantage of the phage protein fusion approach. Instead of the lysogenic filamentous phages commonly used in phage display libraries, we used the lytic phage T7. The use of T7 phage has two major advantages: (a) the cDNA is located at the 3′ end of protein 10B, whereas, in filamentous phage, the cDNA is located in the middle of pIII, thus requiring two in-frame fusions, and (b) the lytic life cycle of T7 phage avoids negative selection of proteins during protein transport through the bacterial membrane, which is necessary for assembling filamentous phage.

Antigen phage display is a novel and powerful approach to the cloning of MHC class II–dependent antigens recognized by CD4⁺ MHC class II–restricted T cells. Using this approach, we have cloned RPL8, which is shared by melanomas (this study and ref. 7), gliomas (7), colorectal carcinomas (19), and ovarian carcinomas (20).

Antigen recognition by Th35 is HLA-DR7 restricted (7), and a putatively HLA-DR7 binding peptide of RPL8 stimulated Th35 cells in an HLA class II–restricted manner. Moreover, RPL8 peptide-specific lymphocytes could be detected in four of nine HLA-DR7⁺ melanoma patients but not in HLA-DR7⁻ patients or healthy donors. There was no statistically significant correlation between melanoma status (stable versus progressive disease) and lymphocyte responses or RPL8 RNA levels of tumor cells and lymphocyte responses. Furthermore, recognition of allogeneic tumor cells by Th35 was independent of tumor RNA level. Interestingly, RPL8 may contain additional epitopes presented by a broad range of HLA class I/II alleles (i.e. A1, A2, A3, A24, B7, B44, B51, DR1, DR4, and DR11), based on preliminary algorithm screens. This suggests that this antigen may represent an important vaccine target in patients bearing RPL8⁺ tumors.

Grant support: NIH grants CA60975, CA88193, CA25874, and CA10815 and the Commonwealth Universal Research Enhancement Program, Pennsylvania Department of Health.

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 Jeffrey Faust and Matthew Farabaugh for flow cytometry analysis and Andrea Raymond for her help with real-time PCR.