Spontaneous CD8 T Cell Responses against the Melanocyte Differentiation Antigen RAB38/NY-MEL-1 in Melanoma Patients

This information is current as Senta M. Walton, Marco Gerlinger, Olga de la Rosa, Natko of October 2, 2021. Nuber, Ashley Knights, Asma Gati, Monika Laumer, Laura Strauss, Carolin Exner, Niklaus Schäfer, Mirjana Urosevic, Reinhard Dummer, Jean-Marie Tiercy, Andreas Mackensen, Elke Jaeger, Frédéric Lévy, Alexander Knuth, Dirk Jäger and Alfred Zippelius

J Immunol 2006; 177:8212-8218; ; Downloaded from doi: 10.4049/jimmunol.177.11.8212 http://www.jimmunol.org/content/177/11/8212 http://www.jimmunol.org/ References This article cites 42 articles, 24 of which you can access for free at: http://www.jimmunol.org/content/177/11/8212.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

by guest on October 2, 2021 • No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2006 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Spontaneous CD8 T Cell Responses against the Melanocyte Differentiation Antigen RAB38/NY-MEL-1 in Melanoma Patients1

Senta M. Walton,2* Marco Gerlinger,2* Olga de la Rosa,* Natko Nuber,* Ashley Knights,* Asma Gati,* Monika Laumer,† Laura Strauss,* Carolin Exner,* Niklaus Scha¨fer,* Mirjana Urosevic,‡ Reinhard Dummer,‡ Jean-Marie Tiercy,§ Andreas Mackensen,† Elke Jaeger,¶ Fre´de´ric Le´vy,ʈ Alexander Knuth,* Dirk Ja¨ger,3# and Alfred Zippelius3,4*

The melanocyte differentiation Ag RAB38/NY-MEL-1 was identified by serological expression cloning (SEREX) and is expressed in the vast majority of melanoma lesions. The immunogenicity of RAB38/NY-MEL-1 has been corroborated previously by the frequent occurrence of specific Ab responses in melanoma patients. To elucidate potential CD8 T cell responses, we applied in vitro Downloaded from sensitization with overlapping peptides spanning the RAB38/NY-MEL-1 protein sequence and the reverse immunology approach.

The identified peptide RAB38/NY-MEL-150–58 exhibited a marked response in ELISPOT assays after in vitro sensitization of CD8 0201؉ melanoma patients. In vitro digestion assays using purified proteasomes provided evidence of naturalءT cells from HLA-A processing of RAB38/NY-MEL-150–58 peptide. Accordingly, monoclonal RAB38/NY-MEL-150–58-specific T cell populations were /capable of specifically recognizing HLA-A2؉ melanoma cell lines expressing RAB38/NY-MEL-1. Applying fluorescent HLA-A2

RAB38/NY-MEL-150–58 multimeric constructs, we were able to document a spontaneously developed memory/effector CD8 T cell http://www.jimmunol.org/ response against this peptide in a melanoma patient. To elucidate the Ag-processing pathway, we demonstrate that RAB38/NY-

MEL-150–58 is produced efficiently by the standard proteasome and the immunoproteasome. In addition to the identification of a RAB38/NY-MEL-1-derived immunogenic CD8 T cell epitope, this study is instrumental for both the onset and monitoring of future RAB38/NY-MEL-1-based vaccination trials. The Journal of Immunology, 2006, 177: 8212–8218.

ritical advances have been made toward understanding nogenic epitopes derived from the two major groups of nonmu- the magnitude and characteristics of cellular immune re- tated shared Ags, namely cancer-testis (i.e., NY-ESO-1, MAGE, sponses in cancer patients (for review, see Refs. 1 and 2). and SSX) and differentiation Ags (i.e., Melan-A, tyrosinase, and C by guest on October 2, 2021 In particular, the discovery of tumor Ags and the precise charac- NY-BR-1). Melanocyte differentiation Ags are expressed in nor- terization of their immunogenic sequences involved in T cell me- mal melanocytes and melanomas, and their function in melano- diated anti-tumor immunity (3) have allowed, for the first time, to genesis is frequently associated with pigment production. systematically analyze anti-tumor T cell responses and have sub- There is considerable evidence that melanocyte differentiation stantially progressed the development of specific cancer vaccines Ags presented to the immune system of melanoma patients can (4–6). Vaccination trials are focused mainly on the use of immu- overcome tolerance and induce T cell-based immunity (7–9). The relative immunogenicity of Melan-A/MART-1 is thought to be the highest of this group of Ags. We could demonstrate previously that *Medical Oncology, Department of Internal Medicine, University Hospital Zurich, Zurich, Switzerland; †Department of Hematology/Oncology, University of Regens- Melan-A/MART-1-specific CD8 T cells with potential tumoricidal burg, Regensburg, Germany; ‡Dermatologische Klinik, University Hospital Zurich, capacity are selectively primed in vivo and are directly detectable Zurich, Switzerland; §Transplantation Immunology Unit, University Hospital, Geneva, Switzerland; ¶II Medizinische Klinik, Krankenhaus Nordwest, Frankfurt, by fluorescent HLA-A2/peptide multimeric constructs (thereafter Germany; ʈLudwig Institute for Cancer Research, Lausanne Branch, University of multimers) even in circulating lymphocytes from melanoma pa- Lausanne, Epalinges, Switzerland; and #Medizinische Onkologie, National Center for tients (10). Apart from the Melan-A/MART-1-specific CD8 T cell Tumor Diseases, University Hospital Heidelberg, Germany repertoire, but similarly to the apparent paucity of ex vivo detect- Received for publication December 15, 2005. Accepted for publication September 22, 2006. able CD8 T cell responses to cancer testis Ags, multimers can only rarely be used to readily identify CD8 T cells specific for other 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 differentiation Ags, e.g., tyrosinase (11, 12). Notably, the applica- with 18 U.S.C. Section 1734 solely to indicate this fact. tion of CD8 T cell epitopes derived from melanocyte differentia- 1 This work was supported in part by a Swiss National Science Foundation special tion Ags in cancer vaccination trials has demonstrated favorable program and a grant from the Cancer Research Institute/Ludwig Institute for Cancer effects on disease progression in individual patients with only very Research Cancer Vaccine Collaborative; the Terry-Fox, Hanne-Liebermann, and Claudia-von-Schilling Foundation; and the UBS Wealth Management. A.Z., A.G., limited toxicity (13–15). Recently, we reported on the melanocyte and S.W. were supported in part by the Emmy-Noether Program (Zi685-2/3) of the differentiation Ag RAB38/NY-MEL-1 (thereafter RAB38) (16), Deutsche Forschungsgemeinschaft. F.L. was supported by the Swiss National Foun- dation and the Cancer Research Institute. that is highly expressed in normal melanocytes and melanoma tis- sues but not in other normal tissues or cancer types. Importantly, 2 S.M.W. and M.G. contributed equally to this work. RAB38 is immunogenic leading to spontaneous Ab responses in a 3 D.J. and A.Z. contributed equally to this work. large proportion of melanoma patients (17). In this study, we dem- 4 Address correspondence and reprint requests to Dr. Alfred Zippelius, Medical On- cology, Department of Internal Medicine, University Hospital Zurich, Raemistrasse onstrate for the first time the identification of a naturally occurring 100, CH-8091 Zurich, Switzerland. E-mail address: [email protected] RAB38-specific CD8 T cell response through the association of in

Copyright © 2006 by The American Association of Immunologists, Inc. 0022-1767/06/$02.00 The Journal of Immunology 8213 vitro sensitization with overlapping peptides and in silico predic- cells were then cocultured overnight with the target cells in the presence of tion of potential HLA class I ligands. Tumor cell recognition and 10 ␮g/ml brefeldin A (GolgiPlug; BD Pharmingen) and then analyzed ␥ in vitro proteasomal digestion assays document natural processing using the intracellular IFN- assay. and presentation of the HLA-A2 restricted RAB3850–58 peptide. RNA extraction, quantitative real-time PCR, and Western blot We show that, in the T cell repertoire of melanoma patients, analysis RAB38 -specific CD8 T cells exist that can even be detected 50–58 RNA extraction, RAB38-specific quantitative real-time PCR, and Western spontaneously applying multimer technology and are endowed blot assays to detect circulating RAB38-specific Ab responses were per- with sufficient avidity to lyse RAB38-expressing tumor cells. Its in formed as described previously (17). To analyze the proteasome subunits, vivo immunogenicity reflected by the presence of cellular and hu- rabbit anti-hLMP-2 (b1i; PW8345), rabbit anti-hLMP-7 (b5i; PW8355), moral anti-RAB38 immune responses in cancer patients opens the rabbit anti-hMECL-2 (b2i; PW8350), rabbit anti-hb5 (PW8895), mouse avenue for developing Ag-specific immunotherapy targeting anti-hb1 (Y; PW8140), and mouse anti-hb2 (Z; PW8145) were purchased from Affiniti Research Products. Secondary HRP-labeled goat anti-rabbit RAB38. Abs or anti-mouse (Jackson ImmunoResearch Laboratories) were used, respectively. Detection of specific binding was performed with the ECL Materials and Methods detection kit (Amersham Biosciences). PBMCs, tumor cell lines, and EBV-transformed B-LCL cells Results Peripheral blood samples were obtained from HLA-A2ϩ patients with stage III/IV malignant melanoma and from 11 HLA-A2ϩ and 10 HLA- RAB38-specific CD8 T cell responses in circulating lymphocytes A2Ϫ healthy donors after written informed consent. Mononuclear cells from HLA-A2 melanoma patients were purified and immediately frozen as described (18). All used HLA- ϩ melanoma patient (ZH-48) with RAB38- 0201ءϩ Initially, a HLA-A A2 tumor cell lines, and EBV-transformed B-LCL cells exhibited the Downloaded from -subtype. expressing tumor lesions and a spontaneous anti-RAB38 Ab re 0201ءA sponse was selected to elucidate CD8 T cell responses against Flow cytometry immunofluorescence analysis and peptides RAB38. Total CD8 T cells were stimulated with 18-aa-long

PE-labeled HLA-A2/peptide multimers were synthesized around RAB3850–58 peptides spanning the entire RAB38 protein sequence and over- (VLHWDPETV) or Flu-MA58–66 (GILGFVFTL), a peptide from the Influ- lapping by 10 amino acids, in the presence of autologous APCs. Ͼ enza matrix protein. Enriched CD8 T cells ( 80%) were stained with the Cultures were tested for the presence of specific, IFN-␥-secreting appropriate PE-labeled multimers and mAbs as described previously (18). http://www.jimmunol.org/ Cells were immediately analyzed on a FACSCalibur using CellQuestPro CD8 T cells by ELISPOT assays using those peptides. A spe- software, both from BD Biosciences. mAbs also were from BD Bio- cific response to the RAB3841–58 peptide corresponding to a sciences. Peptides were synthesized by standard solid-phase chemistry frequency of 2.1% of IFN-␥-secreting CD8 T cells was clearly on a multiple peptide synthesizer (Biosynthesis). All peptides were detected (Fig. 1a). In addition, this patient exhibited a weaker, Ͼ 95% pure as indicated by analytical HPLC. Lyophilized peptides were still significant response (0.2%) against RAB peptide diluted in DMSO and stored at Ϫ20°C. 105–122 (data not shown). No clear response was detectable against all Epitope prediction and peptide digestion by proteasomes the other peptides, although for some of the tested peptides, the ␥ Epitope prediction (ͳwww.syfpeithi.deʹ) was performed as described (19). proportion of IFN- -secreting CD8 T cells was 2- to 3-fold

Long synthetic peptide RAB3844–63 was digested with purified protea- higher in the presence than in the absence of the corresponding by guest on October 2, 2021 somes as described previously (20). After lyophilization, the samples were peptide (data not shown). further tested for recognition by RAB3850–58-specifc T cell clones in chro- mium release cytotoxicity assays. Peptide stimulation and isolation of RAB38-specific CD8 T cells CD8 T cells enriched (Ͼ80%) from PBMCs using a miniMACS device (Miltenyi Biotec) were stimulated with peptide-pulsed irradiated autolo- gous APCs in the presence of recombinant human IL-2 (25 IU/ml) and recombinant human IL-7 (10 ng/ml). CD8 T cell depleted PBMCs were used as APCs. T cells secreting cytokines in response to peptide stimula- tion in a cytokine secretion detection kit (Miltenyi Biotec) were isolated by single-cell, cytokine-guided flow cytometry using a FACSAria (BD Bio- sciences) and further cloned as described previously (10). Transfection of cells, intracellular IFN-␥ secretion, IFN-␥ ELISPOT and chromium release assay, and proteasome inhibition COS-7 and HEK293 EBNA cells were transfected with pcDNA3.1 con- taining the RAB38 cDNA, Melan-A cDNA or HLA-A2 cDNA by Fu- FIGURE 1. Enumeration of RAB38 -specific CD8 T cells in mel- GENE 6 transfection reagent (Roche). Recognition of COS-7 transfectants 50–58 anoma patients and isolation of specific monoclonal populations. a and b, by RAB38 -specific T cells was tested in an MTT colorimetric assay 50–58 IFN-␥ ELISPOT analysis of PBMCs from patient ZH-48 and ZH-104 stim- (21). Recognition of HEK293 EBNA cells by RAB3850–58-specific T cells was tested using intracellular IFN-␥ staining as described previously (10). ulated with RAB3841–58 and RAB3850–58 peptides, respectively, in the IFN-␥ ELISPOT assays were performed in nitrocellulose lined 96-wells absence or presence of relevant peptides. Numbers correspond to the per- microplates (MAHA S45; Millipore) using IFN-␥ ELISPOTs (Diaclone) as centage of specifically IFN-␥ secreting CD8 T cells. c, IFN-␥ secretion described previously (22). Spots were counted using Bioreader 3000 (Bio- assay of PBMCs from patient ZH-104. After in vitro sensitization, PBMCs sys). Lytic activity of clones was measured against peptide-pulsed T2 cells were stimulated in the absence or presence of RAB38 peptide and ϩ Ϫ/Ϫ 50–58 (HLA-A2 , TAP ), different EBV-transformed B-LCL cells, and mel- stained with a cytokine secretion detection kit. FACS single-cell sorting anoma cell lines at the indicated E:T ratios in the presence or absence of ϩ was used to isolate IFN-␥ CD8 T cells and consequently to generate peptide in 4-h chromium release cytotoxicity assays (18). For proteolysis inhibition assays, EBV-transformed B-LCL cells were pretreated for2hat clonal populations. Numbers correspond to the percentage of specifically ␥ 37°C with and without 20 ␮M lactacystin (Calbiochem) and electroporated IFN- -secreting CD8 T cells. d, Clone ZH-104/7 was stained with PE- with 10 ␮g mRNA encoding for full-length RAB38 sequence (RNAactive labeled A2/RAB3850–58 (left) or A2/FLU-MA58–66 (right) multimers, re- Rab38; CureVac). After electroporation, target cells were incubated with spectively, anti-CD8, and anti-CD3 Abs, and analyzed by flow cytometry. and without 20 ␮M lactacystin during 4 h. RAB38-specific monoclonal T Dot plots are gated on CD3ϩ lymphocytes. 8214 RAB38-SPECIFIC CD8 T CELLS IN MELANOMA PATIENTS

To further restrict the number of candidate epitopes, RAB3841–58 sequence was screened for peptides predicted to bind to HLA-A2 molecules with high affinity (i.e., with a score of Ͼ20) by the SYFPEITHI algorithm. As a result, we chose the following two peptides for analysis: RAB3841–50 (TIGVDFALKV; score 21) and RAB3850–58 (VLHWDPETV; score 23). Using those two pep- tides, CD8 T cells of four melanoma patients were stimulated with one round of in vitro sensitization, and cultures were tested in ELISPOT assays. Unfortunately, PBMCs from patient ZH-48 were no longer available as the patient died from disease progression.

Although no reactivity was detected against the RAB3841–50 pep- tide, RAB3850–58 peptide induced significant responses with val- ues of IFN-␥-secreting CD8 T cells being at least 3-fold higher than baseline (irrelevant peptide) in three (ZH-104, ZH-169, ZH- 224) of four patients. In these patients, percentages of IFN-␥-se- creting CD8 T cells varied between 0.35 and 0.9% (data not shown). Of these patients, one patient (ZH-104) exhibited a de- tectable anti-RAB38 Ab response (data not shown). PBMCs of patient ZH-104 with a frequency of 0.35% of IFN-␥ secreting CD8 T cells (Fig. 1b) were further used to perform an additional round Downloaded from of restimulation. After reaching a frequency of 2.1%, IFN-␥-se- creting CD8 T cells were isolated from the cultures using a cyto- kine secretion detection kit and FACS sorting (Fig. 1c). This pro- FIGURE 2. Functional characterization of RAB3850–58-specific clones. cedure resulted in the isolation of monoclonal T cell populations 51 multimers folded a, Lytic activity of clone ZH104/7 against T2 cells was tested in a Cr 0201ءthat were specifically stained by HLA-A

release assay at an E:T ratio of 10:1 in the presence of exogenously added http://www.jimmunol.org/ around the RAB3850–58 peptide, but not by irrelevant HLA-A2/ f F Œ Ⅺ RAB3850–58 ( ), RAB3850–59 ( ), RAB3849–58 ( ), RAB3251–59 ( ), and Flu58–66 multimers (Fig. 1d). Clone ZH-104/7 is shown as a rep- E FLU-MA58–66 peptide ( ). b, Lytic activity of the clone was tested in a resentative example. 51Cr release assay against EBV-transformed B-LCL cells expressing or not HLA-A2 in the presence of RAB38 or FLU-MA peptide Functional analysis of RAB38 -specific CD8 T cell clones 50–58 58–66 50–58 (data not shown). Concentrations of peptides corresponded to 1 ␮M, First, we confirmed the specificity of the generated CTL clones. In E:T ratio ϭ 10:1. a chromium release cytotoxicity assay, the clone ZH-104/7 lysed

T2 cells incubated with peptide RAB3850–58 (with a functional avidity of 10Ϫ10 M) but not T2 cells incubated with the irrelevant by guest on October 2, 2021

Flu58–66 peptide (Fig. 2a). To further document HLA-A2-re- for different time periods with standard proteasomes purified from stricted recognition, we assessed the ability of RAB3850–58-spe- human erythrocytes as reported (24). Applying the CTL clone ZH- cific CTL clones to lyse RAB3850–58 peptide-pulsed EBV-trans- 104/7 that specifically recognizes both RAB3850–58 and formed B-LCL cells expressing or not HLA-A2. As depicted in RAB3849–58 peptides, the digested products were tested for rec- ,0201ϩ EBV-transformed B-LCL cells were ognition when pulsed onto T2 cells. As demonstrated in Fig. 3aءFig. 2b, only HLA-A recognized. Fine specificity of Ag recognition was assessed using the 15-, 30-, and 60-min digests obtained with the proteasomes

RAB3850–58 nonameric peptide, and, additionally, the N- and C- were recognized by the CTL, whereas no recognition was observed terminally extended RAB3849–58 and RAB3850–59 decameric pep- for the RAB3844–63 polypeptide before digestion (0 min). This tides, respectively. Peptide titration experiments showed clearly suggests that, indeed, the CTL clone could recognize a product that the RAB3849–58 decameric peptide was as efficiently recog- generated by the proteasome in vitro, most likely due to an efficient nized as the nonamer, while the relative antigenic activity of the cleavage at the C terminus. To assess whether RAB3850–58-spe-

C-terminal extended RAB3850–59 decameric peptide was at least 2 cific T cells also were able to recognize endogenously produced orders of magnitude lower than the RAB3850–58 nonameric pep- HLA-A2/RAB3850–58 complexes, COS-7 cells were transfected tide (Fig. 2a). Among the human RAB family, RAB38 is closest transiently with plasmids encoding HLA-A2 and either RAB38 or related to RAB32 both at the DNA and protein levels. RAB32 has Melan-A as a control. Transfected COS-7 cells were tested for ␣ been shown to be ubiquitously expressed in multiple normal tis- their capacity to induce TNF- secretion by RAB3850–58-specific sues (23). To exclude reactivity against peptides derived from the clone ZH-104/7. As shown in Fig. 3b, the CTL clone produced RAB32 sequence, recognition of clone ZH-104/7 was tested in a high levels of TNF-␣ in the presence of cells cotransfected with chromium release assay using T2 cells pulsed with the plasmids encoding HLA-A2 and RAB38 but not HLA-A2 and

RAB3251–59 peptide exhibiting the highest sequence homology to Melan-A. Untransfected cells or cells expressing either RAB38 or ϳ RAB3850–58 ( 60%). Importantly, no cross-recognition to this HLA-A2 were not recognized (data not shown). Melan-A26–35- peptide was observed (Fig. 2a). specific clone ZH-7/5 served as a positive control. Similarly,

RAB3850–58-specific clone ZH-104/7 was able to specifically se- RAB3850–58 peptide is naturally processed and presented crete IFN-␥ when coincubated with HLA-A2-expressing HEK293 In addition to sufficient binding capacities to HLA class I mole- EBNA cells that were transiently transfected with RAB38 plasmid cules, an important requirement for a peptide to be naturally pre- (Fig. 3c). To document T cell recognition of tumor cells endog- sented as CD8 T cell epitope is proper excision from the protein by enously expressing RAB38, two HLA-A2ϩ melanoma cell lines, the proteolytic machinery. Therefore, we analyzed in vitro protea- NW-Mel-16 and NW-Mel-449, were tested for recognition by the some-mediated digestions of a 20-mer polypeptide (RAB3844–63) RAB3850–58-specific T cell clone ZH-104/7. As already demon- encompassing the RAB3850–58 peptide. The 20-mer was incubated strated previously by Northern blotting (16), we confirmed a The Journal of Immunology 8215

FIGURE 4. RAB3850–58 peptide is processed by the proteasome and the immunoproteasome. a, Western bot analysis of ␤5, MECL-1, and LMP-7 subunits with lysates from melanoma cell line NW-Mel-16 treated (b)ornot treated (a) with 100 U/ml IFN-␥, and with lysate from an EBV-transformed

B-LCL cells (c). b, Lytic activity of the RAB3850–58-specific clone ZH-104/7 was tested in a chromium release assay against tumor cell lines NW-Mel-16 (HLA-A2ϩ, RAB38ϩ) treated with IFN-␥ for 2, 4, 6, 8, and 10 days at an E:T ␮ f Ⅺ ratio of 10:1. RAB3850–58 peptide (1 M) was added ( )ornot( ).

shown). Similar results, in terms of peptide recognition as well as

recognition of transfected cells and tumor cells, were obtained with Downloaded from six other CTL clones (ZH-104/1, ZH-104/19, ZH-104/23, ZH-104/ 24, ZH-104/28, and ZH-104/30) from patient ZH-104. Altogether,

these data identify RAB3850–58 as an HLA-A2-restricted CD8 T cell epitope that is correctly processed by the intracellular machin- ery of the tumor cells and presented on the cell surface of mela-

noma cell lines. http://www.jimmunol.org/

RAB3850–58 peptide is processed by the immunoproteasome FIGURE 3. RAB3850–58 is naturally processed and presented by mela- noma cells. a, Lytic activity of the clone ZH104/7 was tested in a chro- Proteasomes play a critical role in the generation of antigenic pep- mium release assay against T2 cells pulsed with RAB3850–58 peptide and tides for MHC class I presentation. As shown above, in vitro di- RAB3844–63 peptide incubated with standard proteasomes for 0, 15, 30, gestion analysis and specific T cell recognition of NW-MEL-16 and 60 min, respectively (see Materials and Methods). b, Recognition of and NW-MEL-449 cells expressing both RAB38 and the standard transfected COS-7 cells. Cells transfected with plasmids encoding RAB38, proteasome strongly indicate efficient processing of RAB3850–58 HLA-A2, and Melan-A (as indicated) were cocultured with RAB3850–58- peptide by the multicatalytic protease complex of the tumor cells. specific clone ZH-104/7 and Melan-A -specific clone ZH-7/5. The pro- by guest on October 2, 2021 26–35 However, mature dendritic cells (DCs)5 and tumor cells located in duction of TNF-␣ was measured by the MTT colorimetric assay (see Ma- an IFN-␥-rich environment constitutively express the immunopro- terials and Methods). c, Intracellular IFN-␥ assay was performed using teasome that may significantly differ in the catalytic activities and, RAB3850–58-specific clone ZH-104/7 coincubated with HEK293 EBNA cells that were nontransfected with RAB38 plasmid (left), pulsed with 1 consequently, in the cleavage specificities (25, 26). Thus, we ␮ aimed to determine the relative efficiency of the two proteasome M RAB3850–58 peptide (middle left), and transfected with RAB38 plas- mid (middle right). As control (right), clones were stimulated with PMA/ types to process the RAB3850–58 peptide. To this end, the NW- ionomycin. d, RAB38 cDNA-specific quantitative RT-PCR analysis of Mel-16 tumor cell line was treated with 100 U/ml IFN-␥ for 2, 4, ϩ HLA-A2 tumor cell lines NW-Mel-16 (b), NW-Mel-449 (c), and NTC 6, 8, and 10 days before being tested for recognition by the (Non Template Control, d). The data are expressed relative to the mRNA RAB3850–58-specific T cell clone ZH-104/7. We confirmed by level in normal melanocytes (a), arbitrarily set to 1. The value of each Western blot analysis that LMP-7, MECL-1, and LMP-2 were de- sample was determined in triplicate reactions. e, Lytic activity of the clone tected similarly in NW-Mel-16 cells upon IFN-␥ treatment and in was tested in a chromium release assay against tumor cells NW-Mel-16 ϩ ϩ ϩ Ϫ EBV-transformed B-LCL cells (Fig. 4a). It was shown previously (HLA-A2 , RAB38 ; Ⅺ, f), MCF-7 (HLA-A2 , RAB38 ; ‚, Œ), and NW-Mel-1453 (HLA-A2Ϫ, RAB38Ϫ; छ, ࡗ) at the indicated E:T ratios that the latter express the immunoproteasome constitutively (27). in the presence (filled symbols) or in the absence (white symbols) of As depicted in Fig. 4b, clone ZH-104/7 efficiently recognized tu- ␮ mor cells both treated or not with IFN-␥, suggesting that process- RAB3850–58 peptide (1 M). ing of RAB3850–58 peptide is not affected by the replacement of the standard proteasome. The IFN-␥ treatment did not affect the weaker expression of RAB38 in those melanoma cell lines, com- level of RAB38 expression, as measured by real-time PCR and pared with melanocytes using quantitative real-time PCR (Fig. 3d Western blot analysis (data not shown). Thus, in contrast with and Ref. 17). Protein expression was further confirmed by Western other tumor-specific antigenic peptides, RAB3850–58 peptide is ef- blot analysis (data not shown). As shown in Fig. 3e, RAB38ϩHLA- ficiently processed by the two proteasome types. Additionally, to A2ϩ NW-Mel-16 tumor cell line was efficiently lysed by the CTL document that the proteasome is indeed responsible for the gen- clone, both in the absence and presence of exogenously added eration of RAB3850–58 peptide, EBV-transformed B-LCL cells peptide. Similar results were obtained for tumor cell line NW-Mel- were electroporated with mRNA encoding for full-length RAB38 449 (data not shown). RAB38Ϫ HLA-A2ϩ MCF-7 cells were not and treated with lactycystin, which affects Ag processing by in- hibiting the action of the proteasome (28). After lactacystin treat- lysed unless peptide RAB3850–58 was added exogenously. Finally, ϩ Ϫ ␥ RAB38 HLA-A2 NW-Mel-1453 was not lysed by the CTL ment, IFN- secretion of RAB3850–58-specific T cell clone ZH- clone, neither in the presence nor in the absence of synthetic pep- 104/7 was reduced by 50%, further indicating that its processing ϩ ϩ tide. Lysis of RAB38 HLA-A2 NW-Mel-16 cells by RAB3850–58- specific CTL could be blocked using w6/32 mAb (data not 5 Abbreviation used in this paper: DC, dendritic cell. 8216 RAB38-SPECIFIC CD8 T CELLS IN MELANOMA PATIENTS

cells in patient MeM297 collected in October 2001 (before IL-2) corresponded to ϳ2 ϫ 10Ϫ4 circulating CD8 T cells (0.02%), which is clearly above background values (Ͻ0.01%). Interestingly, the frequency of specific T cells increased threefold (ϳ6 ϫ 10Ϫ4) while the patient was treated with IL-2 until January 2003 (Fig. 5).

The vast majority of those RAB3850–58-specific CD8 T cells dis- played an Ag experienced CCR7ϪCD45RAlow phenotype, corre- sponding to the recently described effector memory T cell subset (30). In addition, 10 melanoma patients were evaluated for the

response to RAB3850–58 by short-term culture of PBMCs in the presence of exogenously added peptide RAB3850–58 and cyto- kines. Specific CD8 T cells could readily be identified in 2 of10 patients (data not shown). Taken together, our data indicate that

RAB3850–58-specific CD8 T cells have been expanded previously in vivo in the response to the autologous tumor.

Discussion RAB proteins form the largest family of the Ras superfamily of small GTP-binding proteins involved in vesicular trafficking path- ways (31). The melanocyte differentiation Ag RAB38 identified Downloaded from previously by the SEREX approach (16) is characterized mainly by 1) a highly restricted expression in normal melanocytes and melanoma lesions, and 2) strong immunogenicity reflected by spontaneous humoral immune responses in a large proportion of melanoma patients (17). In this study, we demonstrate for the first time that RAB38 can elicit a spontaneous CD8 T cell response in http://www.jimmunol.org/ addition to Abs. To date, only a few tumor Ags, e.g., NY-ESO-1, have been reported that may elicit immune responses involving both humoral and cellular effectors (32, 33). FIGURE 5. Ex vivo analysis of RAB3850–58-specific CD8 T cells in mel- ϩ The short protein sequence of RAB38 made it possible to com- anoma patients and healthy donors. a and c, PBMCs from 19 HLA-A2 mel- bine in vitro sensitization using overlapping peptides with in silico anoma patients (Mel Pat) and 11 HLA-A2ϩ healthy donors (HD) were en- prediction of potential HLA-A2 ligands, coined “reverse immu- riched for CD8 T cells and then stained by multimer A2/RAB3850–58. Shown are representative dot plots (a) and an overview of all multimer stainings nology” (19). The peptides that were predicted to bind with high affinity based on appropriately positioned anchor residues were performed (c). Numbers indicate frequency of RAB3850–58-specific CD8 T by guest on October 2, 2021 cells among the CD8 T cell population. Two samples of patient MeM297 tested further for their immunogenicity. RAB3850–58-specific CD8 collected at different time points had a frequency above the detection limit T cell responses were readily detectable in the PBMCs of three of of 0.01%. b, A2/RAB3850–58 multimer-positive CD8-positive cells from four melanoma patients, who have been short-term stimulated with MeM297 were gated and further analyzed for CCR7 as well as CD45RA RAB3850–58 peptide. Interestingly, only one of those patients ex- expression. hibited an anti-RAB38 humoral response, which would implicate that CD8 T cell responses may take place in the absence of specific humoral responses. This is in line with observations with respect to is dependent on proteolytic action cellular and humoral NY-ESO-1 responses (34). However, only a 0201ءand loading on HLA-A (data not shown). larger series of seropositive and seronegative patients will give a more comprehensive picture of this issue. Although reverse im- Ex vivo detectable frequency of circulating RAB3850–58-specific munology has successfully led to the identification of numerous CD8 T cells epitopes, it revealed some peptides, however, that are unlikely to

To evaluate the frequency at which RAB3850–58-specific CD8 T be naturally processed as demonstrated by the poor recognition of cell responses are found among HLA-A2ϩ individuals, we ran- target cells endogenously expressing the Ag (20, 35, 36). Thus, we domly selected a group of 19 patients with stage III/IV malignant chose to perform in vitro digestion assays of a synthetic long oli- melanoma and 11 healthy individuals. Serum was not available gopeptide encompassing RAB3850–58 peptide, and recognition ex- from patients with melanoma. Enriched circulating CD8 T cells periments with RAB3850–58-specific T cells using COS-7 and (Ͼ80% CD3ϩCD8ϩ) from each individual were stained with flu- HEK293 cells transfected with the RAB38 sequence and, more orescent HLA-A2/RAB3850–58 multimeric complexes. To deter- importantly, melanoma cells endogenously expressing RAB38 mine the levels of nonspecific A2/multimer staining, a lower detection as targets. These assay confirmed the adequate generation of Ϫ limit for multimer staining was defined using 10 HLA-A2 blood RAB3850–58 peptide by the cellular processing machinery. As the ϭ ϩ ϫ Ͻ donors (cutoff mean 3 SD 0.01%), as described previ- N-terminal extended decapeptide RAB3849–58, however, is recog- ously (18). According to this limit, ex vivo circulating RAB3850–58- nized with the same functional avidity, either of the two peptides specific CD8 T cells were detected in one of 19 melanoma patients, may be presented by tumor cells. It should be emphasized that but in none of the healthy donors (Fig. 5). Patient MeM297 was a recognition by CD8 T cell clones was specifically directed toward 63-year-old patient with an acrolentiginous melanoma (Clark level RAB38. Undesirable cross-recognition against cells expressing IV), who developed a relapse with s.c. and lymphatic metastases. ubiquitously expressed RAB proteins that share Ͼ75% amino acid In November 2001, the patient was included in a phase I study of sequence identity, e.g., RAB32, was ruled out by TAP-deficient T2 adoptively transferred Melan-A-specific CTL together with low- cells pulsed with a RAB32 peptide exhibiting the highest sequence dose IL-2 (29). The frequency of RAB3850–58-specific CD8 T homology to RAB3850–58. The Journal of Immunology 8217

Representing self-proteins, intentional targeting of melanocyte types. Importantly, this implies that, even in an IFN-␥ rich envi- differentiation Ags, e.g., RAB38, may result in inducing autoim- ronment, e.g., present at tumor sites heavily infiltrated with T cells, munity. This is further underlined by our finding that even tumor RAB38 can be presented on melanoma cells. In addition, given cells with expression levels lower or comparable to melanocytes the high expression of immunoproteasomes in mature DCs, are efficiently recognized by RAB38-specific CD8 T cells. How- RAB3850–58 peptide may also be involved in initiating the im- ever, it needs to be emphasized that although abundantly expressed mune response to RAB38. However, as a highly diverse set of in melanoma lesions, RAB38 expression is limited strictly to me- genes representing tissue restricted self Ags are promiscuously ex- lanocytes (17). As demonstrated for one patient who received in- pressed in thymic cells (43), it is possible that weak expression of fusion of CD8 T cell clones specific for another melanocyte dif- RAB38 in the thymus could play a role in specific thymocyte se- ferentiation Ag, Melan-A/MART-1, postinfusion biopsies of the lection processes. skin showed clinical signs of vitiligo and localization of the in- Overall, in this study, we have described the first characteriza- fused Melan-A/MART-1-specific CD8 T cell clones (37). Given tion of a natural CD8 T cell response against the melanocyte dif- the autoimmune toxicity and temporary clinical response in this ferentiation Ag RAB38. Aside from important implications for patient, tumor immunity against self-Ags can be regarded as a clinical trials of vaccination with RAB38 immunogenic molecules, special case of autoimmune reaction against transformed melano- the data presented here confirm and extend our previous findings cytes. In this regard, effective anti-tumor immunity induced by supporting the immunogenicity of RAB38 in melanoma patients. It RAB38 peptide based vaccinations may be limited to variable ex- thus appears that RAB38, as shown so far only for some Ags, can tents by the process of central and peripheral tolerance that are elicit immune responses involving both humoral and cellular ef- essential to avoid the potentially devastating effects of autoimmunity. fectors. The role of CD4 T cells in generating and maintaining To assess spontaneously occurring CD8 T cell responses to anti-RAB38 immune responses is currently investigated in our Downloaded from RAB38 we used fluorescent multimers to stain circulating lym- laboratory. phocytes from HLA-A2 melanoma patients. Although the frequen- cies of RAB38-specific CD8 T cells were generally below the Acknowledgments multimer detection limit, one patient (MeM297) exhibited a di- We thank Dr. I. Luescher (Ludwig Institute for Cancer Research, Epalin- rectly ex vivo detectable frequency. As recently demonstrated for ges, Switzerland) for synthesis of multimers and Dr. K. Osanai (Depart- the vast majority of Ag experienced CD8 T cells specific for ment of Internal Medicine, Kanazawa Medical University, Ishikawa, http://www.jimmunol.org/ Melan-A/MART-1 (10), RAB38-specific CD8 T cells were Japan) for providing anti-RAB38 Ab. We thank Conny Schneider, Heidi CD45RAlow CCR7Ϫ, a phenotype proposed to characterize effec- Mattlin, Renate Wenig, Julia Karbach, Claudia Frei, Nicole Le´vy, and tor memory T cells potentially able to home into inflamed tissue Anne-Lise Peitrequin for their excellent technical assistance. (30). Importantly, this differentiation stage indicates encounter and Disclosures activation with naturally processed RAB3850–58 peptide presented on APCs in vivo in this patient. Remarkably, the high frequency of The authors have no financial conflict of interest. memory effector RAB38-specific T cells even increased under low-dose IL-2 treatment. Why those cells failed to effectively pre- References by guest on October 2, 2021 vent tumor progression in this patient is difficult to evaluate, as 1. Coulie, P. G., and T. Connerotte. 2005. Human tumor-specific T lymphocytes: does function matter more than number? Curr. Opin. Immunol. 17: 320–325. various mechanisms underlying the cancer immunoediting process 2. Blattman, J. N., and P. D. Greenberg. 2004. Cancer immunotherapy: a treatment could intervene at that stage. Encouraged by the growing appre- for the masses. Science 305: 200–205. ciation that tumors are capable to create an immunosuppressive in 3. van der Bruggen, P., C. Traversari, P. Chomez, C. Lurquin, E. De Plaen, B. Van den Eynde, A. Knuth, and T. Boon. 1991. A gene encoding an antigen situ network that shields tumor cells from T cell immune attack recognized by cytolytic T lymphocytes on a human melanoma. Science 254: (38), it is of our future interest to further dissect characteristics of 1643–1647. 4. Mahnke, Y. D., D. Speiser, I. F. Luescher, J. C. Cerottini, and P. Romero. 2005. RAB38-specific immune responses, e.g., T cell function, present at Recent advances in tumour antigen-specific therapy: in vivo veritas. Int. J. Can- tumor sites. Importantly, in sharp contrast to the large repertoire of cer 113: 173–178. Melan-A-specific CD8 T cells (18), RAB38-specific CD8 T cells 5. Jager, E., D. Jager, and A. Knuth. 2003. Antigen-specific immunotherapy and cancer vaccines. Int. J. Cancer 106: 817–820. are not identifiable ex vivo in healthy donors. However, these cells 6. Rosenberg, S. A. 2001. Progress in human tumour immunology and immuno- become detectable in some healthy individuals, but only after re- therapy. Nature 411: 380–384. petitive in vitro stimulations. Similar findings have been docu- 7. Kawakami, Y., S. Eliyahu, K. Sakaguchi, P. F. Robbins, L. Rivoltini, J. R. Yannelli, E. Appella, and S. A. Rosenberg. 1994. Identification of the im- mented previously for other epitopes, e.g., SSX-2 (39). Additional munodominant peptides of the MART-1 human melanoma antigen recognized by experiments will aim at assessing functional avidity of RAB38- the majority of HLA-A2-restricted tumor infiltrating lymphocytes. J. Exp. Med. specific CD8 T cells in those tumor-free individuals. 180: 347–352. 8. Coulie, P. G., V. Brichard, A. Van Pel, T. Wolfel, J. Schneider, C. Traversari, The proteasome is a multicatalytic protease complex that plays S. Mattei, E. De Plaen, C. Lurquin, J. P. Szikora, et al. 1994. A new gene coding a critical role in the Ag processing pathway. In contrast to the for a differentiation antigen recognized by autologous cytolytic T lymphocytes on HLA-A2 melanomas. J. Exp. Med. 180: 35–42. standard proteasome expressed in all somatic cells, lymphoid cells 9. Brichard, V., A. Van Pel, T. Wolfel, C. Wolfel, E. De Plaen, B. Lethe, P. Coulie, (including myeloid-derived DCs) or cells under the influence of and T. Boon. 1993. The tyrosinase gene codes for an antigen recognized by IFN-␥ express the immunoproteasome that exhibits an altered autologous cytolytic T lymphocytes on HLA-A2 melanomas. J. Exp. Med. 178: 489–495. cleavage site preferences (40). In contrast to other approaches 10. Zippelius, A., P. Batard, V. Rubio-Godoy, G. Bioley, D. Lienard, F. Lejeune, based on in vitro proteasomal digestion (41), we initially used D. Rimoldi, P. Guillaume, N. Meidenbauer, A. Mackensen, et al. 2004. Effector here the standard proteasome to confirm proper processing of function of human tumor-specific CD8 T cells in melanoma lesions: a state of local functional tolerance. Cancer Res. 64: 2865–2873. RAB3850–58 peptide. For some tumor Ags, however, immunopro- 11. Lee, P. P., C. Yee, P. A. Savage, L. Fong, D. Brockstedt, J. S. Weber, D. Johnson, teasomes negatively affect epitope processing (25, 42). In contrast, S. Swetter, J. Thompson, P. D. Greenberg, et al. 1999. Characterization of cir- culating T cells specific for tumor-associated antigens in melanoma patients. Nat. other antigenic peptides appear to be produced more efficiently by Med. 5: 677–685. the immunoproteasome (26). Thus, those peptides that are poorly 12. Valmori, D., C. Scheibenbogen, V. Dutoit, D. Nagorsen, A. M. Asemissen, processed by immunoproteasomes may not be generated at all or V. Rubio-Godoy, D. Rimoldi, P. Guillaume, P. Romero, D. Schadendorf, et al. 2002. Circulating tumor-reactive CD8ϩ T cells in melanoma patients contain a only poorly generated by professional APCs. We show in this CD45RAϩCCR7Ϫ effector subset exerting ex vivo tumor-specific cytolytic ac- study that RAB3850–58 peptide is processed by both proteasome tivity. Cancer Res. 62: 1743–1750. 8218 RAB38-SPECIFIC CD8 T CELLS IN MELANOMA PATIENTS

13. Rosenberg, S. A., J. C. Yang, D. J. Schwartzentruber, P. Hwu, F. M. Marincola, 27. Frisan, T., V. Levitsky, A. Polack, and M. G. Masucci. 1998. Phenotype-depen- S. L. Topalian, N. P. Restifo, M. E. Dudley, S. L. Schwarz, P. J. Spiess, et al. dent differences in proteasome subunit composition and cleavage specificity in B 1998. Immunologic and therapeutic evaluation of a synthetic peptide vaccine for cell lines. J. Immunol. 160: 3281–3289. the treatment of patients with metastatic melanoma. Nat. Med. 4: 321–327. 28. Craiu, A., M. Gaczynska, T. Akopian, C. F. Gramm, G. Fenteany, A. L. Goldberg, 14. Jager, E., M. Maeurer, H. Hohn, J. Karbach, D. Jager, Z. Zidianakis, and K. L. Rock. 1997. Lactacystin and clasto-lactacystin ␤-lactone modify multiple A. Bakhshandeh-Bath, J. Orth, C. Neukirch, A. Necker, et al. 2000. Clonal ex- proteasome ␤-subunits and inhibit intracellular protein degradation and major histo- pansion of Melan A-specific cytotoxic T lymphocytes in a melanoma patient compatibility complex class I antigen presentation. J. Biol. Chem. 272: 13437–13445. responding to continued immunization with melanoma-associated peptides. Int. J. 29. Meidenbauer, N., J. Marienhagen, M. Laumer, S. Vogl, J. Heymann, R. Andreesen, Cancer 86: 538–547. and A. Mackensen. 2003. Survival and tumor localization of adoptively transferred 15. Lienard, D., D. Rimoldi, M. Marchand, P. Y. Dietrich, N. van Baren, C. Geldhof, Melan-A-specific T cells in melanoma patients. J. Immunol. 170: 2161–2169. P. Batard, P. Guillaume, M. Ayyoub, M. J. Pittet, et al. 2004. Ex vivo detectable 30. Sallusto, F., D. Lenig, R. Forster, M. Lipp, and A. Lanzavecchia. 1999. Two activation of Melan-A-specific T cells correlating with inflammatory skin reac- subsets of memory T lymphocytes with distinct homing potentials and effector tions in melanoma patients vaccinated with peptides in IFA. Cancer Immun. 4: 4. functions. Nature 401: 708–712. 16. Jager, D., E. Stockert, E. Jager, A. O. Gure, M. J. Scanlan, A. Knuth, L. J. Old, 31. Pereira-Leal, J. B., and M. C. Seabra. 2001. Evolution of the Rab family of small and Y. T. Chen. 2000. Serological cloning of a melanocyte rab guanosine 5Ј- GTP-binding proteins. J. Mol. Biol. 313: 889–901. triphosphate-binding protein and a chromosome condensation protein from a mel- 32. Jager, E., S. Gnjatic, Y. Nagata, E. Stockert, D. Jager, J. Karbach, A. Neumann, anoma complementary DNA library. Cancer Res. 60: 3584–3591. J. Rieckenberg, Y. T. Chen, G. Ritter, et al. 2000. Induction of primary NY- ESO-1 immunity: CD8ϩ T lymphocyte and antibody responses in peptide-vac- 17. Zippelius, A., A. Gati, T. Bartnick, S. Walton, B. Odermatt, E. Jaeger, ϩ R. Dummer, M. Urosevic, V. Filonenko, K. Osanai, et al. Melanocyte differen- cinated patients with NY-ESO-1 cancers. Proc. Natl. Acad. Sci. USA 97: tiation antigen RAB38/NY-MEL-1 induces frequent antibody responses exclu- 12198–12203. sively in melanoma patients. Cancer Immunol. Immunother. In press. 33. Gnjatic, S., D. Atanackovic, E. Jager, M. Matsuo, A. Selvakumar, N. K. Altorki, R. G. Maki, B. Dupont, G. Ritter, Y. T. Chen, et al. 2003. Survey of naturally 18. Zippelius, A., M. J. Pittet, P. Batard, N. Rufer, M. de Smedt, P. Guillaume, ϩ K. Ellefsen, D. Valmori, D. Lie´nard, J. Plum, et al. 2002. Thymic selection occurring CD4 T cell responses against NY-ESO-1 in cancer patients: correla- generates a large T cell pool recognizing a self-peptide in humans. J. Exp. Med. tion with antibody responses. Proc. Natl. Acad. Sci. USA 100: 8862–8867. 195: 485–494. 34. Valmori, D., V. Dutoit, D. Lie´nard, D. Rimoldi, M. J. Pittet, P. Champagne, K. Ellefsen, U. Sahin, F. Lejeune, J. C. Cerottini, and P. Romero. 2000. Naturally 19. Rammensee, H., J. Bachmann, N. P. Emmerich, O. A. Bachor, and S. Stevanovic. Downloaded from occurring HLA-A2 restricted CD8ϩ T cell response to the cancer testis antigen 1999. SYFPEITHI: database for MHC ligands and peptide motifs. Immunoge- NY-ESO-1 in melanoma patients. Cancer Res. 60: 4499–4506. netics 50: 213–219. 35. Zaks, T. Z., and S. A. Rosenberg. 1998. Immunization with a peptide epitope 20. Valmori, D., U. Gileadi, C. Servis, P. R. Dunbar, J. C. Cerottini, P. Romero, (p369-377) from HER-2/neu leads to peptide-specific cytotoxic T lymphocytes V. Cerundolo, and F. Levy. 1999. Modulation of proteasomal activity required ϩ that fail to recognize HER-2/neu tumors. Cancer Res. 58: 4902–4908. for the generation of a cytotoxic T lymphocyte-defined peptide derived from the 36. Ayyoub, M., M. Migliaccio, P. Guillaume, D. Lienard, J. C. Cerottini, P. Romero, tumor antigen MAGE-3. J. Exp. Med. 189: 895–906. F. Levy, D. E. Speiser, and D. Valmori. 2001. Lack of tumor recognition by 21. Traversari, C., P. van der Bruggen, B. Van den Eynde, P. Hainaut, C. Lemoine, hTERT peptide 540–548-specific CD8ϩ T cells from melanoma patients reveals N. Ohta, L. Old, and T. Boon. 1992. Transfection and expression of a gene coding inefficient antigen processing. Eur. J. Immunol. 31: 2642–2651. http://www.jimmunol.org/ for a human melanoma antigen recognized by autologous cytolytic T lympho- 37. Yee, C., J. A. Thompson, P. Roche, D. R. Byrd, P. P. Lee, M. Piepkorn, cytes. Immunogenetics 35: 145–152. K. Kenyon, M. M. Davis, S. R. Riddell, and P. D. Greenberg. 2000. Melanocyte 22. Pittet, M. J., A. Zippelius, D. E. Speiser, M. Assenmacher, P. Guillaume, destruction after antigen-specific immunotherapy of melanoma: direct evidence D. Valmori, D. Lienard, F. Lejeune, J. C. Cerottini, and P. Romero. 2001. Ex vivo of T cell-mediated vitiligo. J. Exp. Med. 192: 1637–1644. ␥ IFN- secretion by circulating CD8 T lymphocytes: implications of a novel ap- 38. Zou, W. 2005. Immunosuppressive networks in the tumour environment and their proach for T cell monitoring in infectious and malignant diseases. J. Immunol. therapeutic relevance. Nat. Rev. Cancer 5: 263–274. 166: 7634–7640. 39. Ayyoub, M., D. Rimoldi, P. Guillaume, P. Romero, J. C. Cerottini, D. Valmori, 23. Bao, X., A. E. Faris, E. K. Jang, and R. J. Haslam. 2002. Molecular cloning, and D. Speiser. 2003. Tumor-reactive, SSX-2-specific CD8ϩ T cells are selec- bacterial expression and properties of Rab31 and Rab32. Eur J. Biochem. 269: tively expanded during immune responses to antigen-expressing tumors in mel- 259–271. anoma patients. Cancer Res. 63: 5601–5606. 24. Ayyoub, M., S. Stevanovic, U. Sahin, P. Guillaume, C. Servis, D. Rimoldi, 40. Kloetzel, P. M. 2004. The proteasome and MHC class I antigen processing.

D. Valmori, P. Romero, J. C. Cerottini, H. G. Rammensee, et al. 2002. Protea- Biochim. Biophys. Acta 1695: 225–233. by guest on October 2, 2021 some-assisted identification of a SSX-2-derived epitope recognized by tumor- 41. Kessler, J. H., N. J. Beekman, S. A. Bres-Vloemans, P. Verdijk, P. A. van Veelen, reactive CTL infiltrating metastatic melanoma. J. Immunol. 168: 1717–1722. A. M. Kloosterman-Joosten, D. C. Vissers, G. J. ten Bosch, M. G. Kester, presented-0201ءMorel, S., F. Levy, O. Burlet-Schiltz, F. Brasseur, M. Probst-Kepper, A. L. A. Sijts, et al. 2001. Efficient identification of novel HLA-A .25 Peitrequin, B. Monsarrat, R. Van Velthoven, J. C. Cerottini, T. Boon, et al. 2000. cytotoxic T lymphocyte epitopes in the widely expressed tumor antigen PRAME Processing of some antigens by the standard proteasome but not by the immu- by proteasome-mediated digestion analysis. J. Exp. Med. 193: 73–88. noproteasome results in poor presentation by dendritic cells. Immunity 12: 42. Chapatte, L., M. Ayyoub, S. Morel, A. L. Peitrequin, N. Levy, C. Servis, 107–117. B. J. Van den Eynde, D. Valmori, and F. Levy. 2006. Processing of tumor- 26. Schultz, E. S., J. Chapiro, C. Lurquin, S. Claverol, O. Burlet-Schiltz, G. Warnier, associated antigen by the proteasomes of dendritic cells controls in vivo T-cell V. Russo, S. Morel, F. Levy, T. Boon, et al. 2002. The production of a new responses. Cancer Res. 66: 5461–5468. MAGE-3 peptide presented to cytolytic T lymphocytes by HLA-B40 requires the 43. Kyewski, B., and J. Derbinski. 2004. Self-representation in the thymus: an ex- immunoproteasome. J. Exp. Med. 195: 391–399. tended view. Nat. Rev. Immunol. 4: 688–698.