identiﬁcationofpromiscuouskif20alongpeptidesbearing both...

Cancer Therapy: ClinicalSee related commentary by Melief, p. 4295

Identification of Promiscuous KIF20A Long Peptides BearingBoth CD4þ and CD8þ T-cell Epitopes: KIF20A-Specific CD4þ

T-cell Immunity in Patients with Malignant Tumor

Yusuke Tomita1,2, Akira Yuno1,3, Hirotake Tsukamoto1, Satoru Senju1, Yasuhiro Kuroda1,Masatoshi Hirayama1,3, Atsushi Irie1, Kenta Kawahara3, Junji Yatsuda1, Akinobu Hamada4,Hirofumi Jono4, Koji Yoshida5,6, Takuya Tsunoda5,6, Hirotsugu Kohrogi2, Yoshihiro Yoshitake3,Yusuke Nakamura5,7, Masanori Shinohara3, and Yasuharu Nishimura1

AbstractPurpose: To identify long peptides (LP) derived from a novel tumor-associated antigen (TAA), kinesin

family member 20A (KIF20A), which induce tumor-specific T-helper type 1 (TH1) cells and CTLs.

Experimental Design: We combined information from a recently developed computer algorithm

predicting HLA class II–binding peptides with KIF20A-derived CTL-epitope sequences presented by

HLA-A2 (A�02:01) or HLA-A24 (A�24:02) to select candidate promiscuous TH1-cell epitopes containing

CTL epitopes. Peripheral blood mononuclear cells (PBMC) derived from healthy donors or patients with

head-and-neckmalignant tumor (HNMT)were used to study the immunogenicity of KIF20A-LPs, and the in

vitro cross-priming potential of KIF20A-LPs bearing CTL epitopes. We used HLA-A24 transgenic mice to

address whether vaccination with KIF20A-LP induces efficient cross-priming of CTLs in vivo. The TH1-cell

response to KIF20A-LPs in HNMT patients receiving immunotherapy with TAA-derived CTL-epitope

peptides was analyzed using IFN-g enzyme-linked immunospot assays.

Results: We identified promiscuous KIF20A-LPs bearing naturally processed epitopes recognized by

CD4þ T cells and CTLs. KIF20A-specific CTLs were induced by vaccination with a KIF20A-LP in vivo. KIF20A

expressionwasdetected in55%ofHNMTby immunohistochemistry, and significant frequencies of KIF20A-

specific TH1 cell responses were detected after short-term in vitro stimulation of PBMCs with KIF20A-LPs in

50% of HNMT patients, but not in healthy donors. Furthermore, these responses were associated with

KIF20A expression in HNMT tissues.

Conclusions: These are the first results showing the presence of KIF20A-specific TH1 cell responses in

HNMT patients and underline the possible utility of KIF20A-LPs for propagation of TH1 cells andCTLs.Clin

Cancer Res; 19(16); 4508–20. �2013 AACR.

IntroductionWe recently used genome-wide cDNA microarray analy-

sis to identify a novel tumor-associated antigen (TAA),

kinesin family member 20A (KIF20A), which is frequent-ly overexpressed in lung cancer, pancreatic cancer, gastriccancer, bladder cancer, breast cancer, melanoma, andvarious other malignancies (1–3). KIF20A has beenreported to be a promising immunotherapeutic target forcancers. Studies have reported that 2 short peptides (SP)derived from KIF20A, KIF20A-A2809–917, and KIF20A-A2466–75 stimulated CTLs, which kill cancer cells endog-enously expressing KIF20A antigen and that KIF20A-spe-cific CTLs were present in peripheral blood mononuclearcells (PBMC) obtained from patients with pancreaticcancer (1, 4). Phase I/II clinical trials of cancer immu-notherapy for lung cancer, pancreatic cancer, and cho-langiocellular carcinoma using KIF20A-derived SP areunderway. In these trials, we observed that vaccinationwith a KIF20A-derived SP induced a KIF20A-specificCTL response and yielded promising results in patientswith advanced cancer (manuscript in preparation). There-fore, we propose that KIF20A is an attractive target mol-ecule for cancer immunotherapy. We have attempted to

Authors' Affiliations: Departments of 1Immunogenetics, 2RespiratoryMedicine, and 3Oral andMaxillofacial Surgery, Graduate School ofMedicalSciences; 4Department of Clinical Pharmaceutical Sciences, GraduateSchool of Pharmaceutical Sciences, Kumamoto University, Kumamoto;5Laboratory of Molecular Medicine, Human Genome Center, Institute ofMedical Science, University of Tokyo, Tokyo; 6OncoTherapy ScienceIncorporation, Research and Development Division, Kanagawa, Japan;and 7Department of Medicine, University of Chicago, Chicago, Illinois

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

Y. Tomita and A. Yuno contributed equally to this article.

Corresponding Author: Yasuharu Nishimura, Department of Immunoge-netics,Graduate School ofMedical Sciences, KumamotoUniversity, Honjo1-1-1, Chuo-ku, Kumamoto 860-8556, Japan. Phone: 81-96-373-5310;Fax: 81-96-373-5314; E-mail: [email protected]

doi: 10.1158/1078-0432.CCR-13-0197

�2013 American Association for Cancer Research.

ClinicalCancer

Research

Clin Cancer Res; 19(16) August 15, 20134508

on August 12, 2019. © 2013 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst May 28, 2013; DOI: 10.1158/1078-0432.CCR-13-0197

http://clincancerres.aacrjournals.org/

identify LP that induce both antigen-specific CD4þ helperT (TH) cells and CTLs to further develop peptide vaccineimmunotherapy.Tumor-specific TH cells, particularly T-helper type 1

(TH1) cells, play a critical role in efficient induction ofCTL-mediated antitumor immunity (5). IFN-g producedby TH1 cells is critical for induction and maintenance oflong-lived CTL responses through multiple interactions(6, 7). IFN-g secreted by TH1 cells also mediates directantitumor or antiangiogenic effects (8). Furthermore,TH cells pave the way for entry of CTLs at the tumor site(9). Therefore, identification of TH-cell epitopes thatcan activate tumor-specific TH1 cells is important forinduction of effective tumor immunity in tumor-bearinghosts.Melief and colleagues recently reported a synthetic LP

naturally bearing a CTL epitope as an attractive vaccinecompound. Following injection of LP, a patient’s den-dritic cells take up the LP, process it, and present allpossible CTL epitopes and TH-cell epitopes in the contextof various HLA class I and class II molecules (10). Inaddition, recent clinical studies using a promiscuoustelomerase–derived helper-epitope vaccine calledGV1001 bearing CTL epitopes, increased survival of can-cer patients when combined with radiotherapy and che-motherapy (11, 12). Thus, we proposed that an idealpeptide vaccine for cancer immunotherapy may be asingle polypeptide containing epitopes for both TH1 cellsand CTLs, which are naturally proximal to each other andcan be induced simultaneously (10, 13).In this study, a recently developed computer algorithm

predicting HLA class II–binding peptides and known CTL-epitope sequences recognized by HLA-A2 or HLA-A24–restricted CTLs were used to select candidate promiscuousTH-cell epitopes containing the CTL epitopes. Our results

show that the predicted LPs triggered TH1 responses inindividuals expressing several common HLA-DR or HLA-DP alleles, and that the KIF20A-LPs bearing CTL epitopesefficiently stimulated KIF20A-specific CTLs.

Materials and MethodsPatients

Blood samples were collected from 16 patients with ahead-and-neck malignant tumor (HNMT). The immuneresponses of TH cells reactive to KIF20A-LPs were investi-gated. The patients were receiving immunotherapy withTAA-derived CTL-epitope peptides, and were enrolled in2 peptide vaccine trials. These phase I/II clinical trials ofcancer immunotherapy using 3HLA-A24-binding SPs (clin-ical-grade 9-10-amino acid-SPs) derived from 3 cancer-testis antigens, LY6K (LY6K-A24177–186), IMP-3 (IMP-3-A24508–516), andCDCA1 (CDCA1-A2456–64)were reviewedand approved by the Institutional Review Board of Kuma-moto University (Kumamoto, Japan; ref. 14–16). This vac-cine cocktail did not include KIF20A-derived SPs. Allpatients with HNMT were selected on the basis of HLA-A24 presence after providing written informed consent. Thepatients suffered from inoperable advanced HNMT withrecurrent or metastatic tumors and were resistant to stan-dard therapy; they were enrolled in the trial under Univer-sity Hospital Medical Information Network Clinical TrialsRegistry (UMIN-CTR) number 000008379 (CTR-8379).The patients with radical resection were enrolled in the trialunder UMIN-CTR number 000008380 (CTR-8380). In thelatter trial, patients were treated with postoperative peptidevaccine combined with S-1, ifosfamide, or doxorubicin.These clinical trials and analyses are ongoing.

Cell linesMouse fibroblast cell lines (L-cells), genetically engi-

neered to express DR4 (DRB1�04:05), L-DR4; DR8(DRB1�08:03), L-DR8; DR15 (DRB1�15:02), L-DR15; orDR53 (DRB4�01:03), L-DR53were used as antigen-present-ing cells (APC). The HLA-A24-positive C1R-A2402 cellswere a gift from Dr. M. Takiguchi (Kumamoto University;ref. 17).

Prediction of HLA class II–binding peptidesTo predict possible promiscuous HLA class II–binding

human KIF20A-derived peptides, the amino acid sequenceof the human KIF20A protein was analyzed by a recentlydeveloped computer algorithm (IEDB analysis resource,consensus method, http://tools.immuneepitope.org/ana-lyze/html/mhc_II_binding.html; refs. 18 and 19). The pro-gram analyzed 15-amino acid-long sequences offset toencompass the entire protein. The 24- and 25-amino acid-LPs with overlapping high consensus percentile ranksformultipleHLA class IImolecules encoded byDRB1�04:05,DRB1�15:02, or DRB4�01:03 alleles, KIF20A60–84-LP(DSMEKVKVYLRVRPLLPSELERQED), KIF20A494–517-LP(TLHVAKFSAIASQLVHAPPMQLGF), and KIF20A809–833-LP(CIAEQYHTVLKLQGQVSAKKRLGTN), were selected (Sup-plementary Fig. S1 and Table S1).

Translational RelevanceWe recently identified a novel tumor-associated anti-

gen, kinesin family member 20A (KIF20A), overex-pressed in various malignancies. Phase I/II clinical trialsof immunotherapy for several cancers using a KIF20A-derived CTL epitope are underway. We identified pro-miscuous KIF20A LPs bearing naturally processed epi-topes recognized by CD4þ T cells and CTLs. KIF20A-LPsstimulated KIF20A-specific CTLs in vitro and in vivo. Inaddition, KIF20A-LP and a T-helper type 1 (TH1) cellclone enhanced induction of KIF20A-specific CTLs invitro. Thus, KIF20A-LPs provide a useful tool for prop-agation of both TH1 cells and CTLs. This report alsodescribes the first immunohistochemical detection ofKIF20A expression in cases of head-and-neck malignanttumor (HNMT), and the first detection of KIF20A-LPs–specific TH1 cell responses in patients withHNMT. Thesefindings will support clinical trials of KIF20A peptide–based immunotherapy for various cancers.

KIF20A-Specific TH1 Cell Immunity in Tumor Patients

www.aacrjournals.org Clin Cancer Res; 19(16) August 15, 2013 4509




Synthetic peptides and recombinant proteinsHuman KIF20A-derived SPs presented by HLA-A2

(KIF20A-A2809–817) or HLA-A24 (KIF20A-A2466–75), and 3LPs (KIF20A60–84-LP, KIF20A494–517-LP, and KIF20A809–833

-LP) were synthesized (MBL, Nagoya, Japan; purity >95%;Supplementary Fig. S1B). A human immunodeficiencyvirus (HIV) peptide that binds to HLA-A24 (HIV-A24) wasused as a negative control SP (17). A WT1-derived LPpresented by HLA-DR4 (WT1-peptide) and a promiscuousHIV-derived LP were used as negative control LPs (20, 21).Peptides were dissolved in dimethylsulfoxide at 10 mg/mL.The recombinant whole KIF20A and CDCA1 protein wereexpressed by Escherichia coli BL21 with a pET28a vector(Novagen). The CDCA1 protein was used as a control. Eachrecombinant protein was purified and assessed by SDS-PAGE.

Generation of antigen-specific CD4þ T cells fromhealthy donors

The research protocol for collecting and using PBMCsfrom healthy donors (HD1–HD11; Supplementary TableS2) was approved by the Institutional Review Board ofKumamoto University. We obtained PBMCs from 11healthy donors with written informed consent. Genotyp-ing ofHLA-A, DRB1, and DPB1 was conducted at the HLALaboratory (Kyoto, Japan; Supplementary Table S2). Withsome modifications, induction of antigen-specific CD4þ

T cells was conducted as described previously (22).Detailed methods are provided in Supplementary Materi-als and Methods. In some instances, T cells were clonedby limiting dilution for further studies as described pre-viously (23).

Assessment of T-cell responses to peptides andproteins

The immune response of TH cells to APCs pulsed withpeptides (10mg/mL)or proteins (50mg/mL)was assessedbyIFN-g enzyme-linked immunospot (ELISPOT) assays (BDBiosciences) as described previously (17). In brief, thefrequency of peptide-specific CD4þ T cells producingIFN-g per 3 � 104 bulk CD4þ T cells in response tostimulation with peptide-pulsed PBMCs (3 � 104 cells/well), or 1 � 104 bulk CD4þ T cells in response to stimu-lation with peptide-pulsed L-cells expressing HLA-DR (5 �104 cells/well) was analyzed. The frequency of KIF20A-LP–specific TH-cell clone (TH clone) producing IFN-g per 2 �104 TH clone in response to stimulation with protein-loaded dendritic cells (5 � 103 cells/well) was analyzed.To determine the HLA molecules involved in antigenpresentation, antigen-induced IFN-g production wasblocked by adding anti-HLA-DR monoclonal antibody(mAb; L243, BioLegend), anti-HLA-DP mAb (B7/21,Abcam), anti-human HLA-DQ mAb (SPV-L3, Abcam), oranti-HLA class I mAb (W6/32, Abcam; refs. 24–26). AllmAbs were used at a final concentration of 5 mg/mL. Allassessments of IFN-g ELISPOT assays were conducted intriplicate or duplicate, and results are presented as means� SD.

Immunohistochemical examinationImmunohistochemical staining of KIF20A using a rabbit

polyclonal antibody against KIF20A (A300-879A, BethylLaboratories) was conducted as described previously (1,2, 27).

Cytokine assays and CD107a mobilization assayHLA-DP2–restricted KIF20A60–84-LP–specific bulk TH

cells (3 � 104 cells/well) were cultured with autologousPBMCs (3 � 104 cells/well) in the presence of cognatepeptide in 96-well culture plates. HLA-DR53–restrictedKIF20A809–833-LP–specific bulk TH cells were cultured withL-DR53 (5 � 104 cells/well) in the presence of cognatepeptide. After 24 hours, culture supernatants were collectedand cytokine [TNF-a, IFN-g , granulocyte macrophage col-ony-stimulating factor (GM-CSF), macrophage inflamma-tory protein (MIP)1b, interleukin (IL)-2, IL-4, and IL-7]levels were measured using the Bio-Plex system (Bio-Rad)according to manufacturer’s instructions.

Peptide-stimulated degranulation of CD4þ T lympho-cytes was determined by flow cytometric analysis ofCD107a exposed on the cell surface as described previously,with some modifications (28). Detailed methods are pro-vided in Supplementary Materials and Methods.

The synergistic effect of KIF20A-LPs on induction ofKIF20A-specific CTLs

PBMCs obtained from anHLA-A2þ/DR53þ/DP2þ donor(HD2), from whom the KIF20A60–84-LP or KIF20A809–833

-LP–specific TH clones were generated, were plated in 24-well plates (3 � 106 cells/well), followed by addition ofSP alone (KIF20A-A2809–817 SP, 20 mg/mL), or SP þ LP(KIF20A60–84-LP or KIF20A809–833-LP, 20 mg/mL), or SP þLPþ Th-clone (5� 105 cells/well) in a final volumeof 2mL.After culture for 7 days, these peptides and IL-2 (20 U/mL)were added, then IL-15 (5 ng/mL) was added on day 9.On day 11, cells were stained with phycoerythrin-labeledtetramer of the HLA-A�02:01/KIF20A-A2809–817-complexand a fluorescein isothiocyanate–labeled anti-human CD8mAb. Data acquisition was conducted on a FACSCalibur(BD Biosciences), and data files were analyzed with FlowJosoftware (Tree Star).

In vitro stimulation of KIF20A-A2809–817 SP-specificCTLs with KIF20A809–833-LP

Induction of KIF20A-A2809–817 SP-reactive bulk CTLsfrom anHLA-A2þ andDR53þ donor (HD5) by stimulationwith KIF20A-A2809–817 SP was conducted as described (28,29). In brief, purified human CD8þ T cells were stimulatedwith KIF20A-A2809–817 SP-pulsed dendritic cells. After 3rounds of stimulation, we confirmed that CTLs specificallyproduced IFN-g in response to restimulation with T2 cellspulsed with KIF20A-A2809–817 SP; thus, the KIF20A-A2809–817 SP–reactive CTLs were used as effector cells.

To prove the stimulation of KIF20A-A2809–817 SP–reac-tive CTLs with KIF20A809–833-LP–loaded dendritic cells,the number of IFN-g producing KIF20A-A2809–817 SP–specific bulk CTLs in response to stimulation with the

Tomita et al.

Clin Cancer Res; 19(16) August 15, 2013 Clinical Cancer Research4510




KIF20A809–833-LP–loaded autologous dendritic cells wascounted by an ELISPOT assay. The LP-loaded mature den-dritic cells were prepared from positively isolated CD14þ

cells (day 0). CD14þ cells were cultured in the presence ofIL-4 (10 ng/mL) and GM-CSF (100 ng/mL). KIF20A809–833

-LP (50 mg/mL) and OK432 were added on day 5. The LP-loaded mature dendritic cells were harvested on day 7,washed, and used as APCs in an ELISPOT assay.

Assessment of KIF20A-LPs–specific CD4þ T-cellresponses in patients with HNMTFresh PBMCs from patients with HNMT or healthy

donors were cultured with amixture of KIF20A60–84-LP andKIF20A809–833-LP (10mg/mLeach) in afinal volumeof 2mLAIM-V supplemented with 5% human decomplementedplasma at 37�C (2 � 106 cells/well, 24-well plates); bothrecombinant human (rh) IL-2 and rhIL-7 were added onday 0 and 2. After 1 week of cell culture, the PBMCs werecollected, washed, and cultured in ELISPOT plates (1� 105

cells/well) with KIF20A60–84-LP, KIF20A809–833-LP, or con-trol LP for 18 hours. The number of KIF20A-LP–specific THcells expressed as spot-forming cells/105 cells was calculatedafter subtracting control values (background). Responseswere scored as positive when the mean number of IFN-gspots was more than 15 and more than twofold overbackground. The ELISPOT assays on patients’ cells wereconducted in single, duplicate, or triplicate wells because ofthe limited number of available cells. This study was con-ducted in a laboratory that operates under exploratoryresearch principles, and was conducted using investigativeprotocols. We acknowledge the recommendations of theMinimal Information About T-cell Assays reporting frame-work for human T-cell assays (30).

Statistical analysisData were compared by the two-tailed Student t test (bar

graphs), or by the nonparametric Mann–Whitney U test(scatter-dot graph). Differences with a P value < 0.05 wereconsidered statistically significant for all tests.

ResultsPrediction and selection of possible promiscuous HLAclass II–binding KIF20A-LPsTo identify possible promiscuous HLA class II–binding

TH-cell epitopes of KIF20A, we first examined the aminoacid sequence of KIF20A using a recently developed com-puter algorithm (Supplementary Fig. S1A; refs. 18 and 19).Two LPs, KIF20A60–84-LP and KIF20A809–833-LP, predictedby the computer algorithm to be potent promiscuous HLAclass II–binding peptides, were identified proximal toknown 9- or 10-mer CTL epitopes recognized by HLA-A2- or A24-restricted CTLs (Supplementary Fig. S1B andTable S1). Another peptide, KIF20A494–517-LP, was alsopredicted to be a potent promiscuous HLA class II–bindingpeptide, although it did not include a known CTL-epitopesequence. Therefore, 3 candidate LPs, KIF20A60–84-LP,KIF20A494–517-LP, and KIF20A809–833-LP, predicted to havestrong binding affinity to HLA class II molecules HLA-DR4,

HLA-DR15, or HLA-DR53 were synthesized for subsequentanalyses.

Identification of promiscuous KIF20A-derived TH-cellepitopes

CD4þT cells isolated fromPBMCsofhealthydonorswerestimulated at weekly intervals with autologous dendriticcells or PBMCs pulsed with KIF20A60–84-LP as described intheMaterials andMethods section. After at least 3 rounds ofstimulation, KIF20A60–84-LP–specific responses of CD4þ Tcells were examined by IFN-g ELISPOT assays. In an HLA-DR15þ HD1, the generated TH cells produced a significantamount of IFN-g in response to KIF20A60–84-LP–pulsedPBMCs in an HLA-DR–dependent manner. The bulk THcells specifically recognized L-DR15 cells pulsed withKIF20A60–84-LP in anHLA-DR–dependentmanner, but notKIF20A60–84-LP–pulsed L-DR8 cells (Fig. 1A). These resultsindicated that KIF20A60–84-LPwaspresented byHLA-DR15.

To investigatewhether KIF20A60–84-LP induces responsesin TH cells restrictedbyotherHLAclass IImolecules, CD4þTcells from HLA-DR15-negative healthy donors were tested.The TH cells generated from HLA-DP2þ HD2 produced asignificant amount of IFN-g in response to KIF20A60–84-LP–pulsed PBMCs in an HLA-DP–dependent manner. L cellstransduced withHLA-DP2 were unavailable; therefore, weestablished a KIF20A60–84-LP–reactive TH-cell clone (THclone) and used allogeneic PBMCs from 4 differentdonors as APCs to determine restriction by sharedHLA-DP molecules. We confirmed that KIF20A60–84-LPgenerates HLA-DP2–restricted TH cells (Fig. 1B). Thus,KIF20A60–84-LP binds to HLA-DR15 and HLA-DP2, whichsuggests that KIF20A60–84-LP is a promiscuous TH-cellepitope presented by frequent HLA class II molecules(31, 32). In addition, KIF20A60–84-LP could induce anti-gen-specific and HLA-DR–restricted bulk TH cells fromHD3, who was HLA-DR4þ and DR53þ, but DR15- andDP2-negative (Supplementary Fig. S2A). Although wewere unable to define the HLA class II molecule present-ing the LP to the TH cells, this result supports the pro-miscuous nature of KIF20A60–84-LP.

Next, we assessed whether KIF20A809–833-LP could gen-erate antigen-specific TH cells. The TH cells generated fromHLA-DR15þ healthy donors (HD1 and HD4) produced asignificant amount of IFN-g in response to KIF20A809–833

-LP–pulsed PBMCs in anHLA-DR–dependentmanner (Fig.1C and Supplementary Fig. S2B). The bulk TH cells specif-ically recognized L-DR15 cells pulsed with KIF20A809–833

-LP in an HLA-DR–dependent manner (Fig. 1C, HD1).CD4þ T cells from an HLA-DR15-negative healthy donor(HD2) were tested to investigate whether KIF20A809–833-LPinduces responses in TH cells restricted by other HLA class IImolecules. KIF20A809–833-LP could generate HLA-DR53–restricted TH cells in this donor (Fig. 1D). These resultsshow that KIF20A809–833-LP is also a promiscuous TH-cellepitope.

The immunogenicity of KIF20A494–517-LP, which did notinclude a known CTL epitope was evaluated. KIF20A494–517

-LP could generate HLA-DR4–restricted bulk TH cells in






HLA-DR4þ HD2 (Supplementary Fig. S2C). Although theimmunogenicity of KIF20A494–517-LP was assessed inother donors, it was less immunogenic. Thus, we focusedon promiscuous LPs bearing known CTL epitopes, such asKIF20A60–84-LP and KIF20A809–833-LP, for further studies.

KIF20A60–84-LP and KIF20A809–833-LP are naturallyprocessed and presented by dendritic cells

We proceeded to assess whether dendritic cells take upandprocess theKIF20Aprotein to stimulate KIF20A-specificTH clones. Dendritic cells loaded with recombinant KIF20Aprotein were prepared and used as APCs in IFN-g ELISPOTassays (16, 28). An HLA-DP2–restricted KIF20A60–84-LP–specific TH clone efficiently recognized dendritic cells load-ed with KIF20A protein in an HLA-DP–dependent manner,but did not recognize control protein–loaded dendriticcells, indicating that this epitope was naturally processedand presented by HLA-DP2 molecules (Fig. 2A). An HLA-DR53–restricted KIF20A809–833-LP–specific TH clone effi-ciently recognized dendritic cells loaded with KIF20A pro-tein in an HLA-DR–dependent manner, but did not recog-nize control protein–loaded dendritic cells, indicating thatthis epitope was also naturally processed and presented byHLA-DR53 molecules (Fig. 2B).

KIF20A-LPs stimulate TH1-type CD4þ T cellsTo further characterize KIF20A-LP–reactive TH cells, we

used the Bio-Plex system to measure the levels of severalcytokines released in response to stimulation by the cognatepeptide. KIF20A60–84-LP– or KIF20A809–833-LP–specificbulk TH cells from HD2 produced a large amount ofTNF-a, IFN-g , GM-CSF, MIP-1b, and IL-2, but less IL-4 andIL-17 after restimulation with cognate peptide, indicatingTH1-polarized characteristics (Fig. 3A). The cytotoxicitymarker CD107a was detected on the KIF20A-LPs–specificbulk TH cells stimulated with cognate peptide (Fig. 3B), aswas previously showed for antiviral CD4þ effectors andtumor-infiltrating lymphocytes (33–36). These data suggestthat KIF20A-specific TH cells provide a helper function anddirect cytotoxic activity, both of which are advantageous forcancer immunotherapy.

Enhanced induction of KIF20A-specific CTLs byKIF20A-LP–specific CD4þ T cells

Next, we tested whether KIF20A-LPs could enhanceinduction of KIF20A-A2809–817 SP–specific CTLs as shownin Fig. 4A. When PBMCs from an HLA-A2þ/DP2þ/DR53þ

donor (HD2) were stimulated with KIF20A-A2809–817 SPalone (SP), the frequency of KIF20A-A2809–817 SP-specific

0

200

400

600

AIF

N-γ

sp

ots

/

3×

10

4 C

D4

+ T

cells

KIF20A60–84-LP

Anti-DRIF

N-γ

sp

ots

/

10

4 C

D4

+ T

cells

Anti-class I

APC :

HD1 : DRB1*08:03/15:02

Effector cells: KIF20A60–84-LP–specific bulk TH cells

PBMCAPC: Autologous

KIF20A60–84-LP

Anti-DR

Anti-DP

++

+ –

+

+

–

–

–

–

––

0

100

200

300

400

B

IFN

-γ s

po

ts/5

×10

3

CD

4+

T c

ell c

lon

e

HD2 : DPB1*02:01/04:02

IFN

-γ s

po

ts/

3×

10

4 C

D4

+ T

cells

APC: Autologous PBMC

KIF20A60–84-LP

Anti-DR

++

+ –

Anti-DP +

+

–

–

–

–

––

0

200

400

600

800

KIF20A60–84-LP

Anti-DR

Anti-DP

HLA-DPB1 genotypes of allo-PBMCs

DP*02:01/ -DP*02:01/09:01 DP*05:01/ - DP*05:01/09:01

APC: allo-PBMC

+ +

+ –

+

+

–

–

–

–

––

+ +

+ –

+

+

–

–

–

–

––

+ +

+ –

+

+

–

–

–

–

––

+ +

+ –

+

+

–

–

–

–

––

0

200

400

600

800

1,000

L-DR8L-DR15

+ +

+ ––

+

+

–

–

–

–

– ––

–

–

–

+

C D HD2 : DRB1*04:05/09:01, DR53

IFN

-γ s

po

ts/

3×

10

4 C

D4

+ T

cells


KIF20A809–833-LP

Anti-DR

Anti-DP

IFN

-γ s

po

ts/

10

4 C

D4

+ T

cells

Anti-class I

L-DR4L-DR53APC :

KIF20A809–833-LP

Anti-DR

Effector cells: KIF20A809–833-LP–specific bulk TH cells

++

+ –

+

+

–

–

–

–

––

0

100

200

300

400

IFN

-γ s

po

ts/

3×

10

4 C

D4

+ T

cells


KIF20A809–833-LP

Anti-DR

+ ++–

+ –––

Anti-DP +–––

IFN

-γ s

po

ts/

10

4 C

D4

+ T

cells

HD1 : DRB1*08:03/15:02

Anti-class I

APC : L-DR8L-DR15

+ +

+ ––

+

+

–

–

–

–

– ––

–

–

–

+KIF20A809–833-LP

Anti-DR

0

50

100

150

200

0

200

400

600

800

0

100

200

300

400

500

+ +

+ ––

+

+

–

–

–

–

– ––

–

–

–

+

Figure 1. Induction of KIF20A-specific TH cells from healthy donors. A, KIF20A-specific TH cells were generated from a DR15þ healthy donor (HD1) bystimulation of purifiedCD4þ T cells with KIF20A60–84-LP. The generated TH cells were restimulatedwith autologous PBMCs or L cells pulsedwith KIF20A60–84

-LP. The number of IFN-g–producing TH cells was analyzed by ELISPOT assay. Data are presented as themean�SDof triplicate assays. Representative datafrom at least 3 independent experiments with similar results obtained from HD1 are shown. The HLA class II genotype of HD1 is indicated above thepanels. The underlined HLA class II alleles encode the HLA class II molecule presenting the peptides to TH cells. B, HLA-DP2–restricted KIF20A60–84-LP–specific bulk TH cells (left) or a TH clone (right) derived from HD2 was established. An HLA-DP–restricted TH clone was cocultured with allogeneicPBMCs derived from HLA-DP2-positive or HLA-DP2-negative donors pulsed/unpulsed with KIF20A60–84-LP. C, KIF20A809–833-LP–specific HLA-DR15–restricted TH cells were generated from a healthy donor (HD1). D, KIF20A809–833-LP–specific HLA-DR53–restricted TH cells were generated from a healthydonor (HD2).

Tomita et al.





tetramerþ cells was 0.24% of CD8þ T cells (Fig. 4B). Addi-tion of KIF20A60–84-LP (LP) into the SP culture (SP þ LP)induced a slight increase in the frequency of tetramerþ cells.In contrast, when the PBMCs were costimulated withKIF20A-A2809–817 SP, KIF20A60–84-LP, and KIF20A60–84

-LP–specific TH clone (SP þ LP þ TH clone), the frequencyof KIF20A-A2809–817 SP-specificCTLs increased significantlyto 0.87%. As shown in Supplementary Fig. S3A and S3B,KIF20A809–833-LP alone (LP), which encompasses KIF20A-A2809–817 SP, or addition of KIF20A809–833-LP into the SPculture (SPþ LP) induced a slight increase in the frequencyof KIF20A-A2809–817 SP-specific CTLs. However, theKIF20A809–833-LP–specific TH clone rapidly increased inresponse to the KIF20A809–833-LP when we added both LPand TH clone into the PBMCs without KIF20A-A2809–817 SP(LPþ TH clone) and thenwe could not detect the increase offrequency of KIF20A-A2809–817 SP–specific CTLs. We alsoobserved that the stimulation of PBMCs with KIF20A-A2809–817 SP, KIF20A809–833-LP, and KIF20A809–833-LP–specific TH clone (SP þ LP þ TH clone) most stronglyenhanced induction of KIF20A-A2809–817 SP–specifictetramerþ T cells. These results indicate that the activatedTH1 cells enhanced induction of KIF20A-specificCTLs in thepresence of KIF20A-A2809–817 SP.Next, we examined that the CD107a expression of

KIF20A-A2466–75 SP–specific CTLs cultured in the presenceof KIF20A809–833-LP–specific TH cells stimulated with therelevant LP for 1 week to assess the function of KIF20A-specific CTLs expanded by activated KIF20A-specific THcells. KIF20A809–833-LP–specific bulk CD4þ T cells andKIF20A-A2466–75 SP–specific bulk CD8þ T cells derivedfrom HLA-A24þ and HLA-DR15þ HD4 were cultured withautologous dendritic cells in the presence of KIF20A-A2466–75 SP (SP alone), KIF20A-A2466–75 SP þ Control LP (Con-trol LP þ SP), or KIF20A-A2466–75 SP þ KIF20A809–833-LP(KIF20A809–833-LPþ SP) without addition of any cytokine.After 1-week in vitro culture with peptides, the cultured cellswere stained with tetramer of the HLA-A�24:02/KIF20A-A2467–75 complex, anti-human CD107a mAb, and anti-

human CD8mAb as described in the Supplementary Mate-rials and Methods section. As shown in Supplementary Fig.S3C, the addition of KIF20A-A2466–75 SP þ KIF20A809–833

-LP (KIF20A809–833-LP þ SP) significantly increased theabsolute number of KIF20A-A2466–75 SP–specific CD8þ Tcells compared with the addition of KIF20A-A2466–75 SPalone (SP) or KIF20A-A2466–75 SPþControl LP (Control LPþ SP). The absolute number of KIF20A-A2466–75 SP–spe-cific CTLs expressing CD107a on the cell surface afterrestimulationwithKIF20A-A2466–75 SPwas also significant-ly augmented by the addition of KIF20A-A2466–75 SP þKIF20A809–833-LP (KIF20A809–833-LP þ SP; SupplementaryFig. S3D). These results suggest that activated KIF20A-LP–specific TH cells enhance induction of KIF20A-A2466–75 SP–specific CTLs expressing CD107a.

KIF20A-LPs stimulate KIF20A-specific CD8þ T-cells invitro and in vivo

We assessed whether the KIF20A-LPs bearing CTL epi-topes could stimulate KIF20A-SP–specific CTLs. The capac-ity of KIF20A809–833-LP to stimulate KIF20A-A2809–817 SP–specific CTLs was examined by IFN-g ELISPOT assay asdescribed in the Materials and Methods section. As shownin Fig. 4C, KIF20A-A2809–817 SP–reactive bulk CTLs from anHLA-A2þ and -DR53þ donor (HD5) specifically producedIFN-g in response to stimulation with KIF20A809–833-LP–loaded dendritic cell, but not with control LP-loadeddendritic cell. The specific IFN-g production was signifi-cantly inhibited by addition of the anti–HLA class I mAb,but not by the anti–HLA-DR mAb, thus suggesting thatKIF20A-A2809–817 SP–reactive CTLs were stimulatedthrough the cross-presentation of KIF20A809–833-LP by den-dritic cells in vitro.

Subsequently, the capacity of KIF20A60–84-LP to induceKIF20A-A2466–75 SP–specific CTLs was examined by IFN-gELISPOT assay. After in vitro stimulation of PBMCswith KIF20A60–84-LP for 2 weeks, the cells were harvestedand the number of IFN-g–producing T cells in responseto stimulation with KIF20A-A2466–75 SP–pulsed C1R-

APC: DC

BA HD2 : DRB1*04:05/09:01, DR53

IFN

-γ s

po

ts/

2×

10

4 C

D4

+ T

cell c

lon

e

0

100

200

300

400

KIF20A809–833-LP–specific TH-clone

KIF20A protein

Control protein

Anti-DR

Anti-class I

+ ++–

– –

+ –––

+–––

+ – –

–

–

–

–

–

–

–

+––– ––KIF20A809–833-LP

>400

HD2 : DPB1*02:01/04:02

IFN

-γ s

po

ts/

2×

10

4 C

D4

+ T

cell c

lon

e

KIF20A60–84-LP–specific TH-cloneEffector cells

KIF20A protein

Control protein

+ ++–

– –

Anti-DP

Anti-class I

+ –––

+–––

+ – –

–

–

–

>400

0

100

200

300

400

–

–

–

–

+–––– –KIF20A60–84-LP

Figure 2. Natural processing andpresentation of KIF20A-LPs bydendritic cells. A, an HLA-DP2–restricted KIF20A60–84-LP–specific TH clone established fromHD2 recognized autologousdendritic cells loaded with arecombinant KIF20A protein.B, an HLA-DR53–restrictedKIF20A809–833-LP–specific THclone established from HD2recognized autologous dendriticcells loaded with a recombinantKIF20A protein. The relevantKIF20A-derived LP-pulseddendritic cells were used aspositive controls. Representativedata from 3 independentexperiments with similar resultsare shown.






A2402 cells was counted by ELISPOT assay. Details areprovided in Supplementary Materials and Methods. Asshown in Fig. 4D, KIF20A-A2466–75 SP–specific CTLs wereinduced in PBMCs cultured with KIF20A60–84-LP. In addi-tion, KIF20A809–833-LP also induced KIF20A-A2809–817SP–specific CTLs from PBMCs derived from HLA-A2þ

donors (HD2 and HD3; data not shown).The capacity of KIF20A60–84-LP to prime KIF20A-

A2466–75 SP–specific CTLs in vivo was examined by

an ex vivo IFN-g ELISPOT assay. HLA-A24 Tgm wereimmunized three times with KIF20A60–84-LP. TheCD8þ T cells of HLA-A24 Tgm vaccinated withKIF20A60–84-LP produced IFN-g in response to stimula-tion with bone marrow–dendritic cells pulsed with theKIF20A-A2466–75 SP (Supplementary Fig. S4). Theseresults suggest that after uptake of KIF20A60–84-LP, APCscross-prime KIF20A-A2466–75 SP–specific CTLs in vitroand in vivo.

A

38.3%

KIF20A60–84-LP

1.5%

CD4

CD

10

7a

B

0.5% 45.0%

KIF20A809–833-LPIrrelevant peptideStimulation : Irrelevant peptide

Cyto

kin

e levels

(p

g/m

L)

KIF20A809–833-LP–specific bulk TH cells

Cognate

peptide

HD2: DPB1*02:01/ 04:02


IL-2

+ –

0

100

200

300

IL-4

0

100

200

300

+ –

IL-17

0

100

200

300

+ –

MIP1β

+ –

0

20,000

40,000

60,000

GM-CSF

+ –

0

200

400

TNF-α

+ –

0

2,000

4,000

6,000

IFN-γ

+ –

0

5,000

10,000

Cognate

peptide

TNF-α

+ –

0

1,000

2,000

3,000

IFN-γ

+ –

0

2,000

4,000

6,000

8,000

GM-CSF

+ –

0

200

400

600

IL-2

+ –

0

100

200

MIP1β

+ –

0

1,000

2,000

3,000

4,000

IL-4

+ –

0

100

200

IL17

+ –

0

100

200

HD2 : DRB1*04:05/09:01, DR53


HD2 : DRB1*04:05/09:01, DR53HD2: DPB1*02:01/ 04:02


Gated on CD4+ T cells

Figure 3. Cytokine profile produced by KIF20A-LPs–specific bulk TH cells. A, after 24 hours incubation of TH cells cocultured with autologous PBMCs(for KIF20A60–84-LP–specific bulk TH cells) or L-DR53 (for KIF20A809–833-LP–specific bulk TH cells) in the presence of cognate peptides, the culturesupernatant was collected and the concentration of cytokines (TNF-a, IFN-g , GM-CSF, MIP1b, IL-2, IL-4, and IL-7) was measured using the Bio-Plexassay system. Data are presented as the mean � SD of triplicate assays. B, detection of CD107a on the cell surface of bulk KIF20A-specific CD4þ

T cells after antigenic stimulation. Cells were restimulated with cognate LPs or a control peptide. Events shown are gated for CD4þ T cells. The numbersinside the plots indicate the percentage of the cell population with the quadrant characteristic (CD4þ CD107aþ T cells).

Tomita et al.





Presence of KIF20A-specifc TH1 cells in patients withHNMT receiving immunotherapy with TAA-derivedCTL-epitope peptidesTo the best of our knowledge, no studies have examined

KIF20A expression in HNMT. Immunohistochemical anal-

ysis of KIF20A expression was done on 56 cases of HNMT(39 squamous cell carcinoma, 14 adenoid cystic carcinoma,2 osteosarcoma, and 1 angiosarcoma tissue specimens).Twenty-six of the 39 head and neck squamous cell carci-nomas (67%), 4of the 14 adenoid cystic carcinomas (29%),

D

A

B

Day 0Day 7

Peptides (SP ± LP) ± TH-clone Peptides (SP ± LP)

IL-2

PBMCs +

Flow cytometry

SP : KIF-A2809–817

LP : KIF20A60–84-LP

Th-clone : KIF20A60–84-LP–specific TH cell clone

IL-15

Day 9Day 11

SP + LP SP + LP + TH-cloneSP

0.330.24 0.87

CD8

KIF

20

A-A

28

09

–8

17

tetr

am

er

HD2 : HLA-A2+, DP2+, DR53+

Gated on CD8+ T cells

HD5 : HLA-A2+, DR53+C

APC: DC

IFN

-γ s

po

ts/

10

5 C

D8

+ T

cells

KIF20A-A2809–817SP–specific bulk CTLs

Effector cell:

KIF20A809–833-LP

Anti-DR

Anti-class I

+ +

+ ––

+ ––

Control-LP –– –

–

KIF20A-A2809–817SP +

+

–

–

–

–

–

–

–

–

––

–

–

+

–

––

*

**

0

200

400

600*

IFN

-γ s

po

ts/1

05 c

ells

APC: C1R-A2402 cell

KIF20A

A2466–75

HIV

A24Peptide (SP)

HD8 : HLA-A24+, DP2+

In vitro stimulation

with KIF20A60–84-LP

0

40

80

120

160

200 P < 0.005

Figure 4. KIF20A-LPs induce efficient expansion of KIF20A-specific CTLs in vitro and in vivo. A, enhanced induction of KIF20A-A2809–817 SP–specific CTLs inresponse to stimulation with KIF20A60–84-LP and KIF20A60–84-LP–specific TH clones. PBMCs obtained from HLA-A2þ/DP2þ HD2, from which anHLA-DP2–restricted KIF20A60–84-LP–specific CD4

þ TH clone was generated, were cultured for 11 days with KIF20A-A2809–817 SP (SP), SPþKIF20A60–84-LP(LP), or SP þ LP þ KIF20A60–84-LP–specific TH clone (TH clone). On day 11, the cells were stained with KIF20A-A2809–817 SP–specific tetramer withan anti-humanCD8mAb and analyzed by flow cytometry. B, representative KIF20A-A2809–817 SP–specific tetramer staining (gated onCD8þ T cells) obtainedfrom 3 independent experiments with similar results has been shown. C, the KIF20A809–833-LP stimulated KIF20A-A2809–817 SP–specific CD8þ T cellsin vitro. The number of IFN-g–producing KIF20A-A2809–817 SP–specific bulk CTL in response to stimulation with the KIF20A809–833-LP–loaded orcontrol LP–loaded autologous dendritic cells was counted by ELISPOT assay. Statistically significant differences (P < 0.05) are indicated with asterisks.Representative data from 3 independent experiments with similar results are shown. D, PBMCs from HLA-A24þ donors (HD4 and HD8) were incubated withKIF20A60–84-LP for 2 weeks. On days 0 and 7, KIF20A60–84-LP was added. On day 14, the cells were harvested and the number of IFN-g-producing T cells(1 � 105 cells/well) in response to stimulation with KIF20A-A2466–75 SP–pulsed C1R-A2402 cells (2 � 104 cells/well) was counted by ELISPOT assay.Representative data from 5 independent experiments with similar results are shown.






Tab

le1.

Clinical

charac

teris

ticsof

HNMTpatients

KIF20

A-spec

ificCD4þ

T-cellres

pons

esa

Patient

IDAge/se

xKIF20

A60-84-LP

KIF20

A809-833-LP

Histologic

subtype

IHC

analys

isofKIF20

AHLA

-DRB1

HLA

-DRB4

HLA

-DPB1

Positive

/total

Staining

2/16

(13%

)7/16

(44%

)

CTR

-837

9HNMT1

061

/M�

�Squa

mou

sce

llca

rcinom

an.t.

01:01/04

:05

DR53

05:01/�

HNMT2

057

/F�

þSqua

mou

sce

llca

rcinom

an.t.

01:01/09

:01

DR53

02:01/05

:01

HNMT2

670

/M�

�Bas

aloidsq

uamou

sce

llca

rcinom

an.t.

04:05/15

:02

DR53

05:01/09

:01

HNMT2

964

/F�

�Squa

mou

sce

llca

rcinom

an.t.

09:01/14

:54

DR53

03:01/05

:01

HNMT3

169

/F�

þAden

oidcy

stic

carcinom

aPos

itive

01:01/11

:01

-02

:01/04

:02

HNMT3

465

/M�

þSqua

mou

sce

llca

rcinom

an.t.

08:03/15

:02

-02

:01/05

:01

HNMT3

585

/F�

þSqua

mou

sce

llca

rcinom

an.t.

04:05/08

:02

DR53

05:01/�

HNMT3

977

/M�

�Aden

oidcy

stic

carcinom

an.t.

04:06/14

:54

DR53

05:01/19

:01

HNMT4

076

/M�

�Squa

mou

sce

llca

rcinom

an.t.

01:01/09

:01

DR53

04:02/05

:01

HNMT4

151

/Fþ

þAden

oidcy

stic

carcinom

an.t.

01:01/04

:05

DR53

04:02/05

:01

HNMT4

236

/F�

þUnk

nown

n.t.

01:01/08

:02

-04

:02/05

:01

HNMT4

350

/M�

þSqua

mou

sce

llca

rcinom

an.t.

08:02/09

:01

DR53

05:01/�

CTR

-838

0HNMT1

0280

/F�

�Squa

mou

sce

llca

rcinom

aNeg

ative

15:02/�

–02

:01/09

:01

HNMT1

0565

/M�

�Ang

iosa

rcom

aNeg

ative

04:05/13

:02

DR53

03:01/04

:01

HNMT1

0720

/M�

�Osteo

sarcom

aNeg

ative

09:01/�

DR53

02:01/02

:02

HNMT1

0841

/Mþ

�Osteo

sarcom

aPos

itive

04:05/09

:01

DR53

05:01/�

Abbreviations

:CTR

,clinical

trials

registry;D

R53

,DRB4�01

:03;

IHC,immun

ohistoch

emistry;

M/F,m

ale/female;

n.t.,n

ottested

.aKIF20

A-spec

ificT-ce

llresp

onse

smea

suredby

IFN-g

ELISPOTas

sayas

detailedintheMaterialsan

dMetho

dsse

ction.

Pos

itive

andne

gativ

eresp

onse

sareden

oted

by(þ

)and

(�),

resp

ectiv

ely.Th

eun

derlined

HLA

classIIallelesen

codeHLA

clas

sIImolec

ules

prese

ntingKIF20

A-LPto

T Hce

llsinhe

althydon

ors(Fig.1;H

LA-D

RB1�15

:02,DR53

,and

DPB1�02

:01).

Tomita et al.





and 1 of the 2 osteosarcoma (50%) showed positive expres-sion of KIF20A. (Supplementary Table S3). No staining wasdetected in the benign tumor samples.In context of cancer immunotherapy, there is strong

evidence suggesting that vaccines using restricted epitopescan result in broad CD8þ T-cell responses to antigens notpresented in the vaccine (37–39). Thus, we considered thatKIF20A-specific TH-cell responses may be induced by vac-cination with TAA-derived CTL-epitope peptides, which donot include KIF20A-derived SPs. We assessed T-cellresponses specific for KIF20A in peripheral blood from16 patients who were receiving immunotherapy for treat-ment of HNMT. The donor characteristics are summarizedinTable 1. After 1weekof in vitro stimulationof PBMCswithKIF20A-LPs, the frequency of individual KIF20A-LP–specif-ic T cells was detected by IFN-g ELISPOT assay (Fig. 5A).PBMCs isolated from 9 healthy volunteers were used ascontrols. Responses were considered positive when thenumber of IFN-g–secreting cells was at least twofold abovethe negative control. KIF20A-specific TH-cell responses wereobserved in 8 of 16 patients (KIF20A60–84-LP, 2 of 16, 13%;KIF20A809–833-LP, 7 of 16, 44%), but no specific immuneresponses to KIF20A-LPs were detected in the 9 healthydonors (Table 1).We also found that the number of specificspots against KIF20A60–84-LP and KIF20A809–833-LP inpatients were significantly larger than in healthy donors(Fig. 5B). KIF20A809–833-LP–specific IFN-g production by Tcells in HNMT31 and HNMT43 was significantly inhibitedby addition of anti-HLA-DRmAb but not by anti–HLA classI mAb (Fig. 5C). Interestingly, specific responses toKIF20A809–833-LP were augmented in some patients(HNMT31 and HNMT42), or induced in HNMT43 duringthe course of immunotherapy (Fig. 5D). As shown in Fig.5E, KIF20A antigenwas expressed in patients withHNMT inwhom KIF20A-specific TH1 cell responses were detected(Table 1; HNMT31 and HNMT108), but was not expressedin those for whom KIF20A-specific TH1 cell responses werenot detected (Table 1; HNC102 and 107). These observa-tions suggest that APCs collected and processed a KIF20Aantigen derived from tumor cells expressing KIF20A, andthen activated KIF20A-specific TH1 cells in vivo.

DiscussionWe identified 2 promiscuous KIF20A-derived TH-cell

epitopes bearing known HLA-A2– or HLA-A24–restrictedCTL epitopes. KIF20A-specific CTLs were induced by stim-ulation with KIF20A-LPs in vitro and in vivo. This is the firstdemonstration of KIF20A expression in HNMT, and thepresence of KIF20A-specific TH1-cell responses in patientswith HNMT. Furthermore, these responses were related toKIF20Aexpression inHNMT tissues.Wepostulate that theseKIF20A-LPs–specific TH1-cell responses may occur throughtheKIF20Aantigen in apoptotic bodies released from tumorcells. When collected and processed by APCs, the KIF20Aantigen most probably activates KIF20A-specific TH1 cellsin vivo.Recent reports showed that a new wave of tumor-specific

CTL clones became detectable in the blood after vaccination

and provide convincing evidence that the phenomenon ofantigen spreading is critical to the development of effectiveantitumor immunity (37–41). The interaction betweenantivaccine CTLs and the tumor facilitates stimulation oflargenumbers of antitumorCTLs that proceed todestroy thetumor cells. However, CD4þ T-cell responses against a TAAin tumor-bearing patients vaccinated with CTL-epitopepeptides have not been investigated in detail. In this study,we analyzed CD4þ T-cell responses to a KIF20A antigen,which was not present in the vaccine cocktail, in thepatient’s blood before and after vaccination to investigatethe phenomenon of antigen spreading triggered by CTL-epitope vaccination. In patient HNMT43 with advancedcancer, the KIF20A809–833-LP–specific TH1 response wasabsent before vaccination, but after vaccination thisresponse was significantly induced (Fig. 5D, right panel).We believe that this KIF20A809–833-LP–specific TH1-cellresponse is a phenomenon of antigen spreading triggeredby vaccination with CTL epitopes.

Aarntzen and colleagues have recently reported that tar-geting TH cells with dendritic cells pulsed with both HLAclass I- and II-restricted epitopes enhances vaccine-specificimmunologic responses and improves clinical responses(42).We showed that KIF20A-LPs and TH1 clones enhancedinduction of HLA-A2–restricted KIF20A-specific CTLs invitro. These findings indicate that KIF20A-LPs can augmentthe induction of antigen-specific CTLs in combination withimmunotherapy using KIF20A CTL-epitope peptides. Wespeculate that immunotherapy with peptide vaccinationsusing KIF20A-LPs and KIF20A-SPs may improve clinicaloutcomes in KIF20A-expressing cancer.

HLA class II restriction of the KIF20A809–833-LP–specificCD4þ T-cell response in HNMT31 was confirmed by inhi-bition of IFN-g secretion in the presence of HLA-DR–block-ing antibody (Fig. 5C), although HLA-DR alleles ofHNMT31 (Table 1; DRB1�01:01/11:01) were not sharedby theHLA-DR alleles, which encodeHLA class IImoleculespresenting KIF20A809–833-LP to TH cells in healthy donors(HLA-DR15 and DR53; Fig. 1). This result supports thepromiscuous nature of KIF20A809–833-LP and indicates thatLP is naturally processed and presented by HLA class IImolecules in vivo.

In this study, significant frequencies ofKIF20A-specific TH1-cell responses were detected in patients with HNMTreceiving immunotherapy (8 of 16, 50%). Godet and col-leagues reported a possible synergistic effect of the telome-rase-specific CD4þ T-cell response with chemotherapy inlung cancer. They showed that the existence of spontaneoustelomerase-specific TH1 cells before first-line chemotherapysignificantly increased overall survival in lung cancer thatresponded to chemotherapy (43). In addition, the recentintroduction of immunotherapy in clinical practice empha-sized the influence of immune responses on cancer prog-nosis and chemotherapy effectiveness (11, 12, 44). Thesepieces of evidence support the hypothesis that induction ofKIF20A-specific TH1 cells by KIF20A-LP vaccination mayimprove the clinical outcome of cancer patients whencombined with chemotherapy or other standard therapies






A

Day 0 Day 2

PBMCs + KIF20A-LPs

IFN-γ ELISPOT assay

IL-2/ IL-7

Day 7

0

50

100

150

200

250

300

0

20

40

60

80

100

Sp

ecif

ic IF

N-γ

sp

ots

/

10

5 c

ells

Sp

ecif

ic IF

N-γ

sp

ots

/

10

5 c

ells

KIF20A809–833-LPKIF20A60–84-LP

n = 16

n = 9

B

Healthy

donors

HNMT

patients Healthy

donors

HNMT

patients

E

P < 0.001

P < 0.01

HNMT107HNMT108HNMT31 HNMT102

n = 16

n = 9

IHC staining : Positive IHC staining : Negative

Squamous cell carcinomaOsteosarcomaOsteosarcomaAdenoid cystic carcinoma

KIF20A-specific CD4+ T-cell response

NegativeNegativePositivePositive

KIF20A809–833-LP KIF60–84-LP D

Sp

ec

ific

IF

N-γ

sp

ots

/10

5 c

ell

s

Number of

vaccinations

0

100

200

300

16 20

HNMT31

0

50

100

150HNMT42

Pre-vac. 4 Pre-vac. 4

HNMT43

0

20

40

60

80

0

100

200

300

KIF20A809–833-LP

Sp

ec

ific

IF

N-γ

sp

ots

/10

5 c

ell

s

HNMT31

C

0

40

80

120HNM43

Blo

ckin

g mAb (-

)

Anti-

DQ

Anti-

clas

s I

Anti-

DR

Anti-

DP

Blo

ckin

g mAb (-

)

Anti-

clas

s I

Anti-

DR

Anti-

DP

Figure 5. Presence of KIF20A-LPs–specific TH cells in PBMCs isolated from patients with HNMT receiving immunotherapy with TAA-derived CTL-epitopepeptides. A, after in vitro stimulation of PBMCs with a mixture of KIF20A60–84-LP and KIF20A809–833-LP for 1 week, the frequency of individual KIF20A-LPs–specific T cells was detected by IFN-g ELISPOT assay. B, KIF20A-LPs–specific TH1 cell responses were assessed in 16 patients with HNMT receivingimmunotherapy and in 9 healthy donors. The results represent specific IFN-g spots after background subtraction. Each dot represents an individual donor.Horizontal lines denote median values, and P values represent statistical results from a nonparametric Mann–WhitneyU test. C, HLA class II restriction of theIFN-g–producing KIF20A809–833-LP–specific TH cells in HNMT31 and HNMT43. PBMCs stimulated with LPs for 1 week were restimulated with KIF20A809–833

-LP in the presence of mAbs specific to HLA-DR, HLA-DP, HLA-DQ, or HLA class I. D, KIF20A-LPs–specific TH1 cell responses in patients with HNMT weredetected during the course of immunotherapy. E, immunohistochemical analyses of the KIF20A protein in cancer tissues and osteosarcoma (originalmagnification � 400). Positive KIF20A immunohistochemical staining on tissue sections of adenoid cystic carcinoma in HNMT31 and osteosarcoma inHNMT108 are shown. Negative KIF20A immunohistochemical staining on tissue sections of squamous cell carcinoma in HNMT102 and osteosarcoma inHNMT107 are also shown. Malignant cells positive for KIF20A showed homogeneous cytoplasmic staining. Expression of KIF20A in HNMT tissues wasassociated with KIF20A-LPs–specific TH1-cell responses in patients with HNMT.

Tomita et al.





(45–47). Weide and colleagues have recently reported thatthe presence of circulating TH1 cells responding to Melan-Aor NY-ESO-1 had a strong prognostic impact on survivalamong chemotherapy-treated advanced melanomapatients (48). Thus, KIF20A-LPs–specific TH1-cell responsesin patients with HNMT receiving immunotherapy maypositively influence overall survival. The impact ofKIF20A-specific TH1-cell responses on clinical outcome willbe evaluated in a future study.Recent studies of LPs have shown that vaccines contain-

ing natural CTL epitopes were superior to those comprisingminimal CTL epitopes in antitumor CTL immunity becauseof long-lasting cross-presentation of the LPs (10, 49). Weshowed that KIF20A-LPs stimulatedKIF20A-specificCTLs invitro and in vivo. However, we did not compare the capacityto induce KIF20A-specific CTLs between KIF20A-LPs andKIF20A-SPs in human cells. In vivo assay using HLA-A24Tgm, vaccination with KIF20A60–84-LP was not superior toKIF20A-A2466–75 SP in induction of KIF20A-specific CTLs(data not shown). Therefore, we are not able to concludethat we should use KIF20A-LP encompassing CTL epitopealone or both KIF20A-SP and the KIF20A-LP to elicit stron-ger antitumor T-cell immunity at this moment. These issueswill be evaluated in a future study.In this study, we did not confirm the response of gener-

ated TH cells to the shorter, 15-mer TH epitope fromKIF20A-LPs, because we considered that these 25-mer amino acidlong KIF20A-LPs encompass both HLA class I– and class II–restricted T-cell epitopes including unconfirmed T-cell epi-topes by in vitro experiments and could cover many cancerpatients. Based on the findings from healthy donors, 2KIF20A-derived LPs, KIF20A60–84-LP andKIF20A809–833-LP,cover at least 93% of Japanese individuals. We are planningto use these 25-mer KIF20A-LPs in a future clinical trial ofpeptide-based immunotherapy.In conclusion, we identified promiscuous T-H-cell epi-

topes derived from KIF20A and present the first clinicalevaluation of KIF20A-specific TH1-cell responses inpatients with HNMT receiving immunotherapy. These

TH-cell epitopes provide a tool for propagation ofKIF20A-specific TH1 cells and CTLs. These findings sup-port a clinical trial of KIF20A peptide–based immuno-therapy for cancer treatment.

Disclosure of Potential Conflicts of InterestK. Yoshida and T. Tsunoda are employed (other than primary affiliation;

e.g., consulting) by OncoTherapy Science, Inc. Y. Nakamura is a stockholderand scientific advisor for OncoTherapy Science, Inc. Y. Nishimura has acommercial research grant from OncoTherapy Science, Inc. No potentialconflicts of interest were disclosed by the other authors.

Authors' ContributionsConception and design: Y. Tomita, H. Tsukamoto, Y. NishimuraDevelopment of methodology: Y. Tomita, A. Yuno, H. Tsukamoto, A. IrieAcquisitionofdata (provided animals, acquired andmanagedpatients,provided facilities, etc.): Y. Tomita, A. Yuno, Y. Kuroda, M. Hirayama, A.Irie, K. Kawahara, J. Yatsuda, A. Hamada, H. Jono, K. Yoshida, Y. Yoshitake,M. ShinoharaAnalysis and interpretation of data (e.g., statistical analysis, biosta-tistics, computational analysis): Y. Tomita, A. Yuno, M. Hirayama, A. Irie,K. Kawahara, J. YatsudaWriting, review, and/or revision of the manuscript: Y. Tomita, A. Yuno,Y. NishimuraAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): Y. Tomita, A. Yuno, S. Senju, Y.Kuroda, T. Tsunoda, H. Kohrogi, Y. NishimuraStudy supervision: Y. Tomita, H. Tsukamoto, S. Senju, Y. Nakamura, Y.Nishimura

AcknowledgmentsThe authors thank Mr. R. Nishikata (Department of Immunogenetics,

Kumamoto University) for his technical assistance andMs. T. Kubo (Depart-ment of Molecular Pathology, Kumamoto University) for technical assis-tance with immunohistochemical analyses.

Grant SupportThis research was supported by MEXT KAKENHI Grant No. 22133005;

JSPS KAKENHI Grant No. 24300334, a research grant from the PrincessTakamatsu Cancer Research Fund No. 10-24215, and funding fromOncoTherapy Science, Inc.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received January 22, 2013; revised April 8, 2013; accepted May 7, 2013;published OnlineFirst May 28, 2013.

References1. Imai K, Hirata S, Irie A, SenjuS, Ikuta Y, YokomineK, et al. Identification

of HLA-A2-restricted CTL epitopes of a novel tumour-associatedantigen, KIF20A, overexpressed in pancreatic cancer. Br J Cancer2011;104:300–7.

2. Yamashita J, FukushimaS, JinninM,HondaN,MakinoK,Sakai K, et al.Kinesin family member 20A is a novel melanoma-associated antigen.Acta Derm Venereol 2012;92:593–7.

3. Claerhout S, Lim JY, Choi W, Park YY, Kim K, Kim SB, et al. Geneexpression signature analysis identifies vorinostat as a candidatetherapy for gastric cancer. PLoS One 2011;6:e24662.

4. Osawa R, Tsunoda T, Yoshimura S,Watanabe T,MiyazawaM, Tani M,et al. Identification of HLA-A24-restricted novel T cell epitope peptidesderived from P-cadherin and kinesin family member 20A. J BiomedBiotechnol 2012;2012:848042.

5. Chamoto K, Tsuji T, Funamoto H, Kosaka A, Matsuzaki J, Sato T, et al.Potentiation of tumor eradication by adoptive immunotherapy with T-cell receptor gene-transduced T-helper type 1 cells. Cancer Res 2004;64:386–90.

6. Bevan MJ. Helping the CD8(þ) T-cell response. Nat Rev Immunol2004;4:595–602.

7. Shedlock DJ, Shen H. Requirement for CD4 T cell help in generatingfunctional CD8 T cell memory. Science 2003;300:337–9.

8. Street SE, Cretney E, Smyth MJ. Perforin and interferon-gammaactivities independently control tumor initiation, growth, and metas-tasis. Blood 2001;97:192–7.

9. BosR, Sherman LA.CD4þ T-cell help in the tumormilieu is required forrecruitment and cytolytic function of CD8þ T lymphocytes. CancerRes 2010;70:8368–77.

10. Melief CJ, van der Burg SH. Immunotherapy of established (pre)malignant disease by synthetic long peptide vaccines. Nat RevCancer2008;8:351–60.

11. Brunsvig PF, Kyte JA, Kersten C, Sundstrom S, Moller M, Nyakas M,et al. Telomerase peptide vaccination in NSCLC: a phase II trialin stage III patients vaccinated after chemoradiotherapy and an8-year update on a phase I/II trial. Clin Cancer Res 2011;17:6847–57.

12. Kyte JA, Gaudernack G, Dueland S, Trachsel S, Julsrud L, Aamdal S.Telomerase peptide vaccination combined with temozolomide: aclinical trial in stage IV melanoma patients. Clin Cancer Res 2011;17:4568–80.






13. Kenter GG, Welters MJ, Valentijn AR, Lowik MJ, Berends-van derMeer DM, Vloon AP, et al. Vaccination against HPV-16 oncopro-teins for vulvar intraepithelial neoplasia. N Engl J Med 2009;361:1838–47.

14. Kono K, Iinuma H, Akutsu Y, Tanaka H, Hayashi N, Uchikado Y, et al.Multicenter, phase II clinical trial of cancer vaccination for advancedesophageal cancer with three peptides derived from novel cancer-testis antigens. J Transl Med 2012;10:141.

15. Suda T, Tsunoda T, Daigo Y, Nakamura Y, Tahara H. Identificationof human leukocyte antigen-A24-restricted epitope peptidesderived from gene products upregulated in lung and esophagealcancers as novel targets for immunotherapy. Cancer Sci 2007;10:1803–8.

16. Harao M, Hirata S, Irie A, Senju S, Nakatsura T, Komori H, et al. HLA-A2-restricted CTL epitopes of a novel lung cancer-associated cancertestis antigen, cell division cycle associated 1, can induce tumor-reactive CTL. Int J Cancer 2008;123:2616–25.

17. Tomita Y, Imai K, Senju S, Irie A, Inoue M, Hayashida Y, et al. A noveltumor-associated antigen, cell division cycle 45-like can induce cyto-toxic T-lymphocytes reactive to tumor cells. Cancer Sci 2011;102:697–705.

18. Wang P, Sidney J, Kim Y, Sette A, Lund O, Nielsen M, et al. Peptidebinding predictions forHLADR,DPandDQmolecules. BMCBioinform2010;11:568.

19. Wang P, Sidney J, Dow C, Mothe B, Sette A, Peters B. A systematicassessment of MHC class II peptide binding predictions and eval-uation of a consensus approach. PLoS Comput Biol 2008;4:e1000048.

20. Demachi-Okamura A, Ito Y, Akatsuka Y, Tsujimura K, Morishima Y,Takahashi T, et al. Epstein-Barr virus nuclear antigen 1-specific CD4þT cells directly kill Epstein-Barr virus-carrying natural killer and T cells.Cancer Sci 2008;99:1633–42.

21. Ramduth D, Day CL, Thobakgale CF, Mkhwanazi NP, de Pierres C,ReddyS, et al. ImmunodominantHIV-1Cd4þT cell epitopes in chronicuntreated clade C HIV-1 infection. PLoS One 2009;4:e5013.

22. Zarour HM, StorkusWJ, Brusic V,Williams E, Kirkwood JM.NY-ESO-1encodes DRB1�0401-restricted epitopes recognized by melanoma-reactive CD4þ T cells. Cancer Res 2000;60:4946–52.

23. Tabata H, Kanai T, Yoshizumi H, Nishiyama S, Fujimoto S, MatsudaI, et al. Characterization of self-glutamic acid decarboxylase 65-reactive CD4þ T-cell clones established from Japanese patientswith insulin-dependent diabetes mellitus. Hum Immunol 1998;59:549–60.

24. Monji M, Nakatsura T, Senju S, Yoshitake Y, Sawatsubashi M, Shi-nohara M, et al. Identification of a novel human cancer/testis antigen,KM-HN-1, recognized by cellular andhumoral immune responses.ClinCancer Res 2004;10:6047–57.

25. Yokomine K, Senju S, Nakatsura T, Irie A, Hayashida Y, Ikuta Y, et al.The forkhead box M1 transcription factor as a candidate of target foranti-cancer immunotherapy. Int J Cancer 2010;126:2153–63.

26. InoueM,SenjuS,HirataS, Irie A,BabaH,Nishimura Y.An in vivomodelof priming of antigen-specific human CTL by Mo-DC in NOD/Shi-scidIL2rgamma(null) (NOG) mice. Immunol Lett 2009;126:67–72.

27. Yoshitake Y, Nakatsura T, Monji M, Senju S, Matsuyoshi H, Tsu-kamoto H, et al. Proliferation potential-related protein, an idealesophageal cancer antigen for immunotherapy, identified usingcomplementary DNA microarray analysis. Clin Cancer Res 2004;10:6437–48.

28. Tomita Y, Harao M, Senju S, Imai K, Hirata S, Irie A, et al. Peptidesderived from human insulin-like growth factor-II mRNA binding protein3 can induce human leukocyte antigen-A2-restricted cytotoxic Tlymphocytes reactive to cancer cells. Cancer Sci 2011;102:71–8.

29. Imai K, Hirata S, Irie A, Senju S, Ikuta Y, YokomineK, et al. Identificationof a novel tumor-associated antigen, cadherin 3/P-cadherin, as apossible target for immunotherapy of pancreatic, gastric, and colo-rectal cancers. Clin Cancer Res 2008;14:6487–95.

30. Britten CM, Janetzki S, Butterfield LH, Ferrari G, Gouttefangeas C,Huber C, et al. T cell assays and MIATA: the essential minimum formaximum impact. Immunity 2012;37:1–2.

31. Saito S, Ota S, Yamada E, Inoko H, Ota M. Allele frequencies andhaplotypic associations defined by allelic DNA typing at HLA class Iand class II loci in the Japanese population. Tissue Antigens2000;56:522–9.

32. Mack SJ, Bugawan TL, Moonsamy PV, Erlich JA, Trachtenberg EA,Paik YK, et al. Evolution of Pacific/Asian populations inferred fromHLAclass II allele frequency distributions. Tissue Antigens 2000;55:383–400.

33. Casazza JP, Betts MR, Price DA, Precopio ML, Ruff LE, Brenchley JM,et al. Acquisition of direct antiviral effector functions by CMV-specificCD4þ T lymphocytes with cellular maturation. J Exp Med 2006;203:2865–77.

34. Attig S, Hennenlotter J, Pawelec G, Klein G, Koch SD, Pircher H,et al. Simultaneous infiltration of polyfunctional effector and sup-pressor T cells into renal cell carcinomas. Cancer Res 2009;69:8412–9.

35. Widenmeyer M, Griesemann H, Stevanovic S, Feyerabend S, Klein R,Attig S, et al. Promiscuous survivin peptide induces robust CD4þ T-cell responses in the majority of vaccinated cancer patients. Int JCancer 2012;131:140–9.

36. Martorelli D, Muraro E, Merlo A, Turrini R, Rosato A, Dolcetti R. Role ofCD4 þcytotoxic T lymphocytes in the control of viral diseases andcancer. Int Rev Immunol 2010;29:371–402.

37. Corbi�ere V, Chapiro J, Stroobant V, Ma W, Lurquin C, Leth�e B, et al.Antigen spreading contributes to MAGE vaccination-induced regres-sion of melanoma metastases. Cancer Res 2011;71:1253–62.

38. Ribas A, Timmerman JM, Butterfield LH, Economou JS. Determinantspreading and tumor responses after peptide-based cancer immuno-therapy. Trends Immunol 2003;24:58–61.

39. Hunder NN, Wallen H, Cao J, Hendricks DW, Reilly JZ, Rodmyre R,et al. Treatment ofmetastaticmelanomawith autologousCD4þ T cellsagainst NY-ESO-1. N Engl J Med 2008;358:2698–703.

40. Germeau C, MaW, Schiavetti F, Lurquin C, Henry E, Vigneron N, et al.High frequency of antitumor T cells in the blood of melanoma patientsbefore and after vaccination with tumor antigens. J Exp Med 2005;201:241–8.

41. Lurquin C, Lethe B, DePlaen E, Corbiere V, Theate I, van BarenN, et al.Contrasting frequencies of antitumor and anti-vaccine T cells inmetastases of a melanoma patient vaccinated with a MAGE tumorantigen. J Exp Med 2005;201:249–57.

42. Aarntzen EH, De Vries IJ, LesterhuisWJ, Schuurhuis D, Jacobs JF, BolK, et al. Targeting CD4þ T-helper cells improves the induction ofantitumor responses in dendritic cell-based vaccination. Cancer Res2013;73:19–29.

43. Godet Y, Fabre E, DossetM, LamuragliaM, Levionnois E, Ravel P, et al.Analysis of spontaneous tumor-specific CD4 T-cell immunity in lungcancer using promiscuous HLA-DR telomerase-derived epitopes:potential synergistic effect with chemotherapy response. Clin CancerRes 2012;18:2943–53.

44. Robert C, Thomas L, Bondarenko I, O'Day S, Weber J, Garbe C, et al.Ipilimumab plus dacarbazine for previously untreated metastatic mel-anoma. N Engl J Med 2011;364:2517–26.

45. Ding ZC, Zhou G. Cytotoxic chemotherapy and CD4þ effector T cells:an emerging alliance for durable antitumor effects. Clin Dev Immunol2012;2012:890178.

46. Lesterhuis WJ, Haanen JB, Punt CJ. Cancer immunotherapy—revis-ited. Nat Rev Drug Discov 2011;10:591–600.

47. Walter S, Weinschenk T, Stenzl A, Zdrojowy R, Pluzanska A, SzczylikC, et al.Multipeptide immune response to cancer vaccine IMA901 aftersingle-dose cyclophosphamide associates with longer patient surviv-al. Nat Med 2012;18:1254–61.

48. Weide B, Zelba H, Derhovanessian E, Pflugfelder A, Eigentler TK, DiGiacomo AM, et al. Functional T cells targeting NY-ESO-1 or Melan-Aare predictive for survival of patients with distant melanoma metas-tasis. J Clin Oncol 2012;30:1835–41.

49. Bijker MS, van den Eeden SJ, Franken KL, Melief CJ, van der Burg SH,OffringaR.Superior inductionof anti-tumorCTL immunity byextendedpeptide vaccines involves prolonged, DC-focused antigen presenta-tion. Eur J Immunol 2008;38:1033–42.


Tomita et al.




2013;19:4508-4520. Published OnlineFirst May 28, 2013.Clin Cancer Res Yusuke Tomita, Akira Yuno, Hirotake Tsukamoto, et al. Immunity in Patients with Malignant Tumor

T-cell+ T-cell Epitopes: KIF20A-Specific CD4+ and CD8+CD4Identification of Promiscuous KIF20A Long Peptides Bearing Both

Updated version

10.1158/1078-0432.CCR-13-0197doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2013/08/14/1078-0432.CCR-13-0197.DC2

http://clincancerres.aacrjournals.org/content/suppl/2013/05/29/1078-0432.CCR-13-0197.DC1Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/19/16/4508.full#ref-list-1

This article cites 49 articles, 18 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/19/16/4508.full#related-urls

This article has been cited by 5 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/19/16/4508To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-13-0197

http://clincancerres.aacrjournals.org/content/suppl/2013/05/29/1078-0432.CCR-13-0197.DC1 http://clincancerres.aacrjournals.org/content/suppl/2013/08/14/1078-0432.CCR-13-0197.DC2

http://clincancerres.aacrjournals.org/content/suppl/2013/05/29/1078-0432.CCR-13-0197.DC1 http://clincancerres.aacrjournals.org/content/suppl/2013/08/14/1078-0432.CCR-13-0197.DC2

http://clincancerres.aacrjournals.org/content/19/16/4508.full#ref-list-1

http://clincancerres.aacrjournals.org/content/19/16/4508.full#related-urls

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/19/16/4508


identiﬁcationofpromiscuouskif20alongpeptidesbearing both...

Documents