Identification of Interleukin-13 Receptor A2 Peptide Analogues Capable of Inducing Improved Antiglioma CTL Responses

Research Article

Junichi Eguchi,1,4 Manabu Hatano,1,4 Fumihiko Nishimura,1,4 Xinmei Zhu,1,4 Jill E. Dusak,1,4 Hidemitsu Sato,5 Ian F. Pollack,1,4 Walter J. Storkus,3 and Hideho Okada1,2,4

Departments of 1Neurological Surgery, 2Surgery, and 3Dermatology and Immunology, University of Pittsburgh School of Medicine; 4Brain Tumor Program, University of Pittsburgh Cancer Institute, Pittsburgh, Pennsylvania;and 5Department of Neurosurgery, Kanagawa Cancer Center, Kanagawa, Japan

Abstract site (reviewed in refs. 3, 4), we and others have successfully shown Restricted and high-level expression of interleukin-13 receptor that effective anti-CNS tumor immune responses can be induced in A2 (IL-13RA2) in a majority of human malignant gliomas preclinical models using syngeneic tumor- and dendritic cell–based makes this protein an attractive vaccine target.We have vaccines (5–7). Based on our own preclinical data (8, 9), we have previously initiated phase I glioma vaccine trials using interleukin-4 previously described the identification of the IL-13RA2345-353 peptide as a human leukocyte antigen-A2 (HLA-A2)–restricted (IL-4) gene–transduced autologous glioma cells or fibroblasts (10) CTL epitope.However, as it remains unclear how efficiently and vaccination regimens using dendritic cells loaded with whole peptide-based vaccines can induce specific CTLs in patients glioma cell lysates (11). with malignant gliomas, we have examined whether analogue Although vaccinations of glioma patients with autologous epitopes could elicit heteroclitic antitumor T-cell responses dendritic cell loaded with bulk glioma antigens have not yet versus wild-type peptides.We have created three IL-13R A2 evoked a single case of autoimmune encephalopathy in treated analogue peptides by substitutions of the COOH-terminal glioma patients (12–15), glioma-associated antigen (GAA)–specific approaches would be expected to be more effective and to reduce isoleucine (I) for valine (V) and the NH2-terminal tryptophan (W) for either alanine (A), glutamic acid (E), or nonsubsti- the potential risk of autoimmune encephalitis even further. In tuted (W; designated as 1A9V, 1E9V, and 9V, respectively).In addition, these vaccines can be generated in a shorter period of comparison with the native IL-13RA2 epitope, the analogue time compared with whole glioma cell approaches, owing to the peptides 9V and 1A9V displayed higher levels of binding feasibility of obtaining an ‘‘off-the-shelf ’’ preparation of synthetic affinity and stability in HLA-A2 complexes and yielded an GAA peptides. improved stimulatory index for patient-derived, specific CTLs With regard to the human leukocyte antigen-A2.1 (HLA-A2.1)– + restricted GAA-derived epitopes available for the induction of CTL against the native epitope expressed by HLA-A2 glioma cells. + In HLA-A2-transgenic HHD mice, immunization with the responses, our report describing a CD8 T-cell epitope derived from peptides 9V and 1A9V induced enhanced levels of CTL the GAA IL-13Ra2 (16) is among the first citations in the literature. reactivity and protective immunity against an intracranial IL-13Ra2 is expressed at high levels in most human gliomas but challenge with IL13RA2-expressing syngeneic tumors when not in the normal brain (17, 18). compared with vaccines containing the native IL-13RA2 The concern for the clinical use of nonmutated, autoantigens in epitope.These findings indicate highly immunogenic IL- vaccine formulations is that high-affinity T cells recognizing such 13RA2 peptide analogues may be useful for the development epitopes have been selectively eliminated due to normal tolerance of vaccines capable of effectively expanding IL-13RA2-specific, mechanisms, leaving a low-affinity repertoire that may prove tumor-reactive CTLs in glioma patients. (Cancer Res 2006; ineffective in recognizing the naturally processed and HLA- 66(11): 5883-91) presented peptides on the surface of tumor cells in situ.To overcome this issue, it has been previously shown that by incorporating certain amino acid substitutions in the natural Introduction peptide sequence, one may enhance peptide-HLA complex stability, Despite significant advances in modern microsurgery, radiother- resulting in the induction of stronger and/or heteroclitic antitumor apy, and chemotherapy, the prognosis for patients with malignant CTL responses (19–21). In particular, analogue peptides with glioma remains poor (1). Immunotherapy for glioma is an increased HLA-binding capacity have been obtained by modifying attractive alternative treatment option because activated, anti- the NH2-terminal (22, 23) and COOH-terminal (24, 25) peptide tumor immune cells have the potential to migrate into the central positions. nervous system (CNS) and to selectively destroy malignant cells In an attempt to increase the efficiency of CTL induction against that have infiltrated normal CNS tissues (2). Although the CNS, and the natural IL-13Ra2345-353 peptide, we created three IL-13Ra2 tumors that arise therein, reside in an ‘‘immunologically privileged’’ analogue peptides (designated as 1A9V, 1E9V, and 9V) that incorporate NH2-terminal and/or COOH-terminal amino acid substitutions. Here, we show that the IL-13Ra2 analogue peptides 9V and 1A9V are superior to the natural IL-13Ra2345-353 epitope in their capacity to induce CTL responses from patient-derived HLA- Requests for reprints: Hideho Okada, Department of Neurological Surgery, + University of Pittsburgh School of Medicine, G12a The Hillman Cancer Center, 5117 A2 peripheral blood mononuclear cells (PBMC) in vitro and HLA- Center Avenue, Pittsburgh, PA 15213-1863. Phone: 412-623-1111;Fax: 412-623-4747; A2-transgenic HHD mice. This most likely reflects the enhanced E-mail: [email protected]. I2006 American Association for Cancer Research. longevity of analogue peptide presentation to specific CTL doi:10.1158/0008-5472.CAN-06-0363 precursors, allowing for increased clonotypic proliferation. www.aacrjournals.org 5883 Cancer Res 2006; 66: (11).June 1, 2006

Materials and Methods 37jC. Thereafter, the cells were washed four times to remove free peptides and incubated at 37jC for 0, 3, or 6 hours. The cells were stained with the Peptides. a The synthetic peptides IL-13R 2 (345-353, WLPFGFILI;345-9V, BB7.2 mAb to evaluate the HLA-A2 molecule expression for each time point. WLPFGFILV;345-1A9V, ALPFGFILV;345-1E9V, ELPFGFILV), MART-1 (27-35, Peptide-induced HLA-A2 expression was evaluated by calculating the mean AAGIGILTV), Influenza M1 (58-66, GILGFVFTL), and HBVcore128 fluorescence of peptide-incubated T2 cells minus the mean fluorescence of (TPPAYRPPNAPIL) were synthesized by N-(9-fluorenyl)methoxycarbonyl T2 cells in the absence of peptide. DC50 was defined as the time required for chemistry in the University of Pittsburgh Peptide Synthesis Facility and were the loss of 50% of the HLA-A2/peptide complexes stabilized at t =0. >95% pure as indicated by analytic high-performance liquid chromatogra- In vitro induction of CTL in patient-derived PBMCs. To generate phy and mass spectrometric analysis. Peptides were dissolved in PBS/10% dendritic cells, the plastic adherent cells from PBMCs were cultured in AIM- j DMSO at a concentration of 2 mg/mL and stored at 20 C until use. V medium (Invitrogen) supplemented with 1,000 units/mL recombinant Cells and cell culture. The U251 and SNB19 glioma cell lines were human granulocyte macrophage colony-stimulating factor and 500 units/ generously provided from Drs. Martin R. Jadus (University of California, mL recombinant human IL-4 (rhIL-4;Cell Sciences) at 37j C in a humidified Irvine) and William C. Welch (University of Pittsburgh), respectively, and CO2 (5%) incubator. Six days later, the immature dendritic cells were cultured in DMEM supplemented with 10% fetal bovine serum (FBS), 100 stimulated with recombinant human tumor necrosis factor-a, IL-6, and A IU/mL penicillin, 100 g/mL streptomycin, and 10 mmol/L L-glutamine IL-1h (10 ng/mL each). Mature dendritic cells were then harvested on day 8, (all reagents from Invitrogen, Carlsbad, CA). resuspended in AIM-V medium at 1 106 per mL with peptide (10 Ag/mL), PBMCs were obtained from glioma patients and healthy donors under an and incubated for 2 hours at 37jC. Populations of autologous CD8+ T cells Institutional Review Board–approved protocol. HLA-A2 expression on the were enriched from PBMCs using magnetic microbeads (Miltenyi Biotech, PBMC was validated using the monoclonal antibodies (mAb) MA2.1 (against Auburn, CA). CD8+ T cells (2 106 per well) were cocultured with 2 105 HLA A2, B17) and BB7.2 (against HLA A2, Aw69: both from the American per well peptide-pulsed dendritic cells in 2 mL/well of AIM-V medium Type Culture Collection, Manassas, VA) in indirect immunofluorescence supplemented with 5% human AB serum, 10 units/mL rhIL-2 (R&D assays monitored by flow cytometry. Systems, Minneapolis, MN), and 10 units/mL rhIL-7 (Cell Sciences) in each b h h The murine EL4-HHD [D x 2 microglobulin ( 2M) null and transgenic well of 24-well tissue culture plates. On day 15, lymphocytes were h for modified HLA-A2.1- 2 microglobulin single chain (HHD gene)] and restimulated with autologous dendritic cells pulsed with peptide in AIM- b h control EL-4S3-Rob cells (D x 2M null) were the generous gifts of Dr. V medium supplemented with 5% human AB serum, rhIL-2, and rhIL-7 Franc¸ois A. Lemonnier (Pasteur Institute, Paris;ref. 26). The transporter (10 units/mL each). On day 20, the CD8+ cultured cells were analyzed for + associated with antigen processing–deficient, HLA-2.1 T2 cell line, EL4- CTL activity by standard 4-hour 51Cr release assay. HHD and EL4S3-Rob cells were maintained in RPMI 1640 (Invitrogen) Vaccination of HHD mice and i.c. tumor challenge. HHD mice supplemented with 10% FBS and 100 IU/mL penicillin, 100 Ag/mL received (on days 0 and 7) s.c. injections of 100 Ag of IL-13Ra2345-353 or each streptomycin, and 10 mmol/L L-glutamine (all reagents from Invitrogen). of the peptide analogues emulsified in Incomplete Freund’s adjuvant (IFA; HHD mice. HHD mice were obtained from Dr. Franc¸oisA. Lemonnier Difco, Detroit, MI) in the presence of 140 Ag of the I-Ab-restricted through Dr. Pravin T.P. Kaumaya (The Ohio State University, Columbus, OH). HBVcore128 T-helper epitope, which stimulates a CD4+ helper T-cell b h h HHD mice are D x 2 microglobulin ( 2M) null and transgenic for modi- response. Control animals received IFA containing HBV helper-peptide only. h fied HLA-A*0201- 2 microglobulin single chain (HHD gene;refs. 26, 27). On day 11 after the second immunization, the animals were sacrificed, and Creation of an HHD-syngeneic tumor cell line that expresses 5 107 splenocytes were stimulated in vitro with the same peptide that was A a IL-13R 2. The full-length IL-13R 2-cDNA (National Center for Biotech- used for in vivo stimulation (10 Amol/L). On day 6 of culture, the bulk nology Information accession no. NM_000640) fragment was generated by populations were tested for specific cytotoxicity against the EL4-HHD cells reverse transcription-PCR using the forward (AGTATGGCTTTCGTTTGC pulsed with IL-13Ra2 345-353, control unpulsed EL4-HHD, or EL4-HHD- TTGGC) and reverse (TACCGAGCTCGGATCCACTAGT) primers and U251 IL13Ra2 cells. a glioma-derived total RNA. The IL-13R 2 cDNA was then cloned into the To assess systemic protective immunity against i.c. tumor challenge, day expression plasmid pEF6/V5-His-TOPO vector (Invitrogen) to generate 7 after the second immunization, HHD mice received an i.c. inoculation of a 4 pEF6/V5-IL-13R 2. EL4-HHD cells were then transfected with the pEF6/ EL4-HHD-IL13Ra2 cells, as previously described (6). Briefly, 5 10 EL4- V5-IL-13Ra2 using Cell Line Nucleofecter kit T (Amaxa, Gaithersburg, MD), HHD-IL13Ra2 cells were stereotactically injected through an entry site at and a blasticidine-resistant clone that stably expressed the highest level of the bregma 2 mm to the right of the sagittal suture and 3 mm below the IL-13Ra2 based on flow-cytometry using anti-IL13Ra2 mAb (Cell Sciences, surface of the skull of anesthetized mice using a stereotactic frame. The Canton, MA) was selected (EL4-HHD- IL13Ra2) for further use. animals were monitored daily after treatment for the manifestation of any HLA-A2/peptide tetramer staining. Phycoerythrin-conjugated HLA- pathologic signs. All animal experiments were done under an Institutional A*0201/ALPFGFILV tetramer (IL-13Ra2-tetramer) was produced by the Animal Care and Use Committee–approved protocol. National Institute of Allergy and Infectious Disease tetramer facility within Isolation of brain-infiltrating lymphocytes. Mice bearing i.c. EL4- the Emory University Vaccine Center (Atlanta, GA) using the peptide HHD-IL13Ra2 tumors received immunizations on days 14 and 21 after the synthesized by the University of Pittsburgh Peptide Production Facility. tumor inoculation, were sacrificed by CO2 asphyxia on day 28, and perfused Cells were stained with tetramers (10 Ag/mL) in PBS containing 1% bovine through the left cardiac ventricle with PBS. Brains were enzymatically serum albumin for 15 minutes at room temperature, washed once, and digested (28, 29), and cells from each brain were resuspended in 70% stained with FITC-conjugated anti-human CD8 or anti-mouse CD8 (BD Percoll (Sigma, Saint Louis, MO), overlaid with 37% and 30% Percoll and Biosciences, San Diego, CA). Flow cytometric analyses were done using centrifuged for 20 minutes at 500 g. Enriched brain-infiltrating Coulter EPICS cytometer (Beckman Coulter, Fullerton, CA). lymphocyte (BIL) populations were recovered at the 70% to 37% Percoll Measurement of the peptide/HLA-A2 complex binding and stability. interface. T2 cells (1 106 cells/mL) were incubated with various concentrations Statistical analyses. Survival data were compared using a log-rank test. (0.1-100 nmol/L) of peptides in serum-free RPMI 1640 37jC overnight in an Comparative numbers of T-cell responses were analyzed by Student’s t test atmosphere containing 5% CO2. They were then washed twice with PBS and for two samples with unequal variances. Statistical significance was stained with the BB7.2 mAb for 30 minutes at 4jC. After cells were washed, determined at the <0.05 level. In addition, positive response was defined FITC-conjugated goat anti-mouse IgG (Caltag, Burlingame, CA) was used as as follows: the specific lysis by the responder cells against antigen-positive the secondary antibody. Surface expression levels of HLA-A2 were examined target cells is at least 15% and 2-fold higher than lytic levels by by flow cytometry. Peptide binding was evaluated by mean fluorescence corresponding control conditions in at least two effector/target (E/T) intensity (MFI). ratios. The correlation among clinical factors, including age, tumor type For assessment of stability, T2 cells (1 106 per mL) were incubated and use of temozolomide therapy, and T-cell responses, was assessed by a overnight with 100 Amol/L of each peptide in serum-free RPMI 1640 at m2 test.

Cancer Res 2006; 66: (11).June 1, 2006 5884 www.aacrjournals.org

Results of HLA-A2+ human glioma cells that endogenously expressed and presented IL-13Ra2-derived epitopes. Human glioma cell lines IL-13RA2345-353 analogue peptides 9V and 1A9V exhibit enhanced binding affinities for HLA-A2. We synthesized three U251 and SNB19 express both the HLA-A2 and IL-13Ra2 proteins, whereas the human glioma cell line A172 expresses IL-13Ra2 but IL-13Ra2345-353 analogue peptides that incorporated a COOH- terminal I for V substitution based on previous studies (24, 25); not HLA-A2 (16). Therefore, we used U251 and SNB19 as relevant target glioma cells, whereas A172 served as a negative control line and either no NH2-terminal substitution or a W for either A or E in 4-hour 51Cr release assays. As illustrated in Fig. 1B, the U-251 and NH2-terminal substitution designated as 9V, 1A9V, and 1E9V, SNB19 cell lines were highly susceptible to cytotoxic activity respectively. The NH2-terminal substitutions were designed to improve steric fit (via replacement of W with the small A residue) mediated by CTL lines that had been induced with IL-13Ra2345-353 and solubility (by replacing W with E). or with analogue peptides. A172 cells, in contrast, were not lysed V Because stable binding of HLA-A2 with peptide epitopes further above a background level (<15%) at E/T ratio of 40, by any of the stabilizes the surface expression of HLA-A2 (30, 31), quantitative CTL lines tested, suggesting that the IL-13Ra2345-353 or modified expression levels of HLA-A2, which is indicated by MFI in Table 1, peptide-induced CTL lines lysed SNB19 and U-251 glioma cells in correlate with the binding affinity of the peptide epitopes that are an HLA-A2-restricted and antigen-dependent manner. In contrast, coincubated with the T2 cells. The MFI values for T2 cells with no the T cells stimulated with no peptide showed only background peptide indicate the baseline HLA-A2 expression level. The level lysis (<20%) against the SNB19 and U-251 glioma cell lines at analogue peptides 9V and 1A9V but not 1E9V possess a higher E/T ratios of V40. Although there was a trend for an E/T ratio– binding affinity to HLA-A2 and higher stabilization capacity as dependent increase of lysis in these control effector cells, we shown by the prolonged DC50 when compared with the native believe it was non–antigen-specific killer activity of the T cells IL-13Ra2345-353 epitope. This suggested the possibility that 9V and based on experiments discussed in Fig. 1D, in which neither 1A9V peptides might prove more immunogenic than the wild-type addition of cold peptide-loaded T2 cells or neutralizing antibody peptide when incorporated into vaccines. against MHC-class I (W6/32) significantly reduced the cytolysis of Peptide analogues 9V and 1A9V were more immunogenic the control effector cells. than wild-type IL-13RA2345-353 peptide in promoting the As shown in Fig. 1C, peptide analogues 9V and 1A9V induced in vitro activation of antitumor CTL. We initially sought to higher levels of CTL reactivity against the SNB19 and U-251 show that the CD8+ T cells that were stimulated with the peptide target cells at all E/T ratios when compared with CTLs primed analogues could still recognize the native IL-13Ra2345-353 peptide against the wild-type IL-13Ra2345-353 peptide. These data further presented in the context of HLA-A2. CD8+ T cells that had been confirm that these peptide analogues consistently induce anti-IL- stimulated with either the wild-type or analogue peptides 13Ra2 CTL more efficiently than does the native IL-13Ra2345-353 efficiently lysed T2 target cells loaded with escalating concen- peptide. trations of the natural IL-13Ra2 peptide (Fig. 1A). T cells that had To corroborate the specificity of CTL activity, cold target been stimulated with the 9V or 1A9V peptides showed higher levels competition experiments were done by addition of nonradiolabeled of lytic activity against T2 cells loaded with low concentrations (cold) T2 cells pulsed with IL-13Ra2345-353 peptide in the 4-hour 51 (0.1-10 nmol/L) of IL-13Ra2345-353 wild-type peptide compared with Cr release assay (Fig. 1D). The anti-SNB19 glioma cell lytic T cells stimulated with other peptides. Only low background lysis of activities mediated by CTL lines induced by the wild-type IL- T2 cells was observed for any of the CTL lines in the absence of 13Ra2345-353 or its peptide analogues were almost completely specific peptide. T cells that had been stimulated with the control inhibited by the addition of the cold T2 cells pulsed with IL- Influenza M1 peptide or no peptide did not show any specific lytic 13Ra2345-353 wild-type peptide. The CTL activities, however, were activity against the IL-13Ra2345-353-pulsed T2 cells over background not inhibited by the addition of non–peptide-pulsed cold T2 cells, control levels. showing that the lytic ability and inhibition were specifically linked We next examined whether the CTLs developed in response to to the IL-13Ra2345-353 epitope. Furthermore, addition of an anti- the 9V and 1A9V analogue peptides exhibited improved recognition MHC class I antibody (W6/32) inhibited the CTL-mediated lysis,

Table 1. Enhanced binding ability and stability of peptide analogues 9V and 1A9V to HLA-A2

c Peptide Sequence T2 binding (MFI*) DC50 (h)

1 nmol/L 0.1 nmol/L

IL13Ra2 345-353 WLPFGFILI 28.4 23.7 3 IL13Ra2 345-9V WLPFGFILV 38.1 29.1 6 IL13Ra2 345-1A9V ALPFGFILV 44.6 33.2 >6 IL13Ra2 345-1E9V ELPFGFILV 25.2 23.4 3-6 Influenza M1 GILGFVFTL 53.4 38.9 >6 b b No peptide (control) 9.4 8.9

*Peptide binding was evaluated by MFI. c DC50 was defined as the time required for the loss of 50% of the HLA-A*0201 peptide complexes stabilized at t =0. bMFI values for control, non–peptide-loaded cells indicate baseline expression levels when T2 cells were treated with solvent (DMSO in PBS) only.

www.aacrjournals.org 5885 Cancer Res 2006; 66: (11).June 1, 2006

+ Figure 1. IL-13Ra2345-353 analogue peptides 9V and 1A9V induced enhanced levels of anti-IL-13Ra2345-353 CTL activity in vitro. A, CD8 T cells isolated from an + HLA-A2 glioma patient were stimulated with autologous dendritic cells loaded with either wild-type IL-13Ra2345-353 (n), 9V (o), 1A9V (E), 1E9V (4), Influenza M158-66 (5), or no ( ) peptide for 10 days. Responder T cells were then tested for their lytic activity against T2 cells loaded with indicated concentrations of IL-13Ra2345-353 51 peptide in 4-hour Cr release assays at an E/T ratio of 12.5. Points, mean % specific lysis; bars, SD. P < 0.01 for 9V versus wild-type IL-13Ra2345-353 peptide and 1A9V versus wild-type IL-13Ra2345-353 peptide at 0.1 and 1 nmol/L by two-tailed Student’s t test. From one representative experiments of three done for donors + + exhibiting a positive CTL response against the wild-type IL-13Ra2345-353 epitope. B and C, CD8 T cells derived from an HLA-A2 glioma patient were stimulated with autologous dendritic cells loaded with no peptide, the wild-type IL-13Ra2345-353 peptide, or the 9V, 1A9V or 1E9V analogue peptides. On day 19 after the primary stimulation, responder T cells were tested for their lytic ability against human glioma cells SNB19, U-251 (both are IL-13Ra+/HLA-A2+), or A172 (IL13Ra2+/HLA-A2) using 4-hour 51Cr release assays. Points, mean % specific lysis; bars, SD. Some SD values were smaller than the point marks. B, P < 0.05 at all E/T ratios for SNB19 versus A172 and for U251 versus A172 by two-tailed Student’s t tests (except for the control group). C, against SNB19 glioma cells (P < 0.05) at all E/T ratios for 9V versus wild-type IL-13Ra2345-353 peptide and 1A9V versus wild-type IL-13Ra2345-353 by two-tailed Student’s t tests. Against U251 glioma cells (P < 0.05) at E/T ratios of 10 and 40 for 9V versus wild-type IL-13Ra2345-353 and 1A9V versus wild-type IL-13Ra2345-353 peptide by two-tailed Student’s t tests. Stimulation with the 1E9V analogue peptide did not improve the CTL reactivity when compared to stimulation using the wild-type peptide. D, addition of ‘‘cold’’ T2 cells pulsed with the IL-13Ra2345-353 peptide or anti-HLA-class I antibody inhibited CTL activity, supporting the antigen specificity and the HLA-class I restriction of the CTL lines, respectively. The CTL lines induced with each peptide were incubated for 4 hours with 51Cr -labeled human glioma SNB19 cells at the indicated E/T ratios for evaluation of specific lytic ability (n). For the cold target inhibition assay, 51Cr-labeled target SNB19 cells (1 103 per well) and cold T2 cells (1 104 per well) pulsed with (E) or without (.) the IL-13Ra2345-353 peptide were incubated with the CTLs. To block the HLA class I–mediated recognition by the T cells, the pan-class I mAb W6/32 (10 Ag/mL) was added before standard 4-hour 51Cr release assay (). Points, mean % specific lysis; bars, SD. In IL-13Ra2 peptide groups, P < 0.05 at all E/T ratios for SNB19 versus cold target or mAb-treated target by two-tailed Student’s t tests. From one of the three experiments done yielding similar results using PMBCs isolated from three independent donors that displayed a positive response against the IL-13Ra2345-353 wild-type epitope.

Cancer Res 2006; 66: (11).June 1, 2006 5886 www.aacrjournals.org

confirming that the antiglioma CTL reactivity induced by these evaluate CTLs that recognize the native IL-13Ra2345-353 peptide in peptides was HLA class I restricted (Fig. 1D). the context of HLA-A2, as shown in Fig. 2B, we next showed that the In vitro stimulation with the peptide analogues 9V and 1A9V CTLs induced by the native peptide recognized T2 cells loaded with + promotes the enhanced development of antigen-specific CD8 the 1A9V or the wild-type IL-13Ra2345-353 peptide, and reciprocally, T cells. To determine whether the enhanced CTL activities induced that CTLs induced by the peptide analogues recognized both native by the 9V and 1A9V analogue peptides are attributable to the and analogue peptides, suggesting a mutual cross-reactivity of the efficient induction of the CTL proliferation (quantity), CD8+ T cells modified and the native peptides. Therefore, we consider that an from a HLA-A2+ healthy donor were stimulated with dendritic cells evaluation using the HLA-A2/IL-13Ra2 (1A9V) tetramer allows us to loaded with either the native or peptide analogues. Control groups evaluate CTL recognition of both native and analogue peptides. were stimulated with negative control MART-127-35 peptide or with In vivo immunization of HHD mice with the 9V or 1A9V no peptide. The cells were then evaluated for the number of CD8+ peptides results in enhanced frequencies of specific CTL react- T cells capable of binding to HLA-A2/IL-13Ra2 tetramer by flow ive against target cells naturally expressing IL-13Ra2345-353. cytometry (Fig. 2A). T cells stimulated with the 9V or 1A9V peptides To examine whether immunization with IL-13Ra2345-353 and/or its induced higher numbers of tetramer-positive cells (6.86% and peptide analogues can elicit CTL responses in vivo, we immunized 6.80%, respectively) when compared with the naive peptide (4.60%). HLA-A2 transgenic HHD mice twice with the wild-type, 9V, or 1A9V T cells that had been stimulated with the MART-1 peptide or no peptides emulsified in IFA in the presence of a heterologous peptide had only background levels of tetramer-positive events. HBVcore128 T-helper peptide. After a 6-day in vitro stimulation, the Although we initially synthesized an HLA-A2-tetramer using the splenocytes derived from vaccinated HHD mice were able to lyse native IL-13Ra2345-353 peptide, we were unable to optimize the EL4-HHD cells loaded with the native IL-13Ra2345-353 peptide staining conditions, most likely due to the poor stability of the HLA- (Fig. 3A) and EL4-HHD cells transfected to express the IL-13Ra2 A2/IL-13Ra2345-353 peptide complex (data not shown). To examine (EL4-HHD-IL-13Ra2) gene product (Fig. 3B). Control nonpulsed whether a tetramer made using the 1A9V peptide would allow us to EL4-HHD cells were not lysed by the splenocytes beyond

Figure 2. In vitro stimulation of PBMC with the 9V or 1A9V analogue peptides promotes enhanced proliferation of antigen-specific CD8+ CTLs. CD8+ T cells isolated from a healthy HLA-A2+ donor were stimulated with autologous dendritic cells loaded with either no peptide, MART-127-35 peptide, wild-type IL-13Ra2345-353 peptide, or the analogue peptides 9V, 1A9V, or 1E9V for 2 weeks. A, responder cells were then incubated with FITC-conjugated anti-human CD8 antibody and pycoerythrin-labeled HLA-A2/IL-13Ra2-tetramer for 15 minutes at room temperature. After washing, the cells were evaluated by flow cytometry for the percentage of CD8+/ tetramer+ in the lymphocyte-gated population (numbers in the parenthesis). B, the same responder cells were evaluated for their lytic activity against T2 cells loaded with either the wild-type IL-13Ra2345-353 (E), analogue 1A9V (.), or MART-1 () peptides in 4-hour 51Cr release assays. Points, mean % specific lysis; bars, SD.

www.aacrjournals.org 5887 Cancer Res 2006; 66: (11).June 1, 2006

Figure 3. Vaccination of HHD mice with the IL-13Ra2 analogue peptides 9V or 1A9V promotes enhanced CTL reactivity against EL4-HHD cells loaded with wild-type IL-13Ra2345-353 peptide and EL4-HHD-IL- 13Ra2 tumor cells. HHD mice received s.c injections of 100 Ag of either wild-type IL-13Ra2345-353 peptide, or the 9V or 1A9V analogue peptides emulsified in IFA in the presence of the 150 Ag of the heterologous I-Ab-restricted HBVcore128 T-helper epitope on days 0 and 7. Control animals received IFA containing HBV helper peptide only. Eleven days after the second immunization, the animals were sacrificed, and splenocytes were restimulated in vitro with the same peptide that was used for in vivo immunization for an additional 6 days. The effector cells were then evaluated for specific cytotoxicity using standard 4-hour 51Cr release assays against EL4-HHD target cells pulsed with or without the wild-type IL-13Ra2345-353 peptide (A); or EL4-HHD-IL-13Ra2 cells that constitutively express the IL-13Ra2 versus control EL4-HHD cells (B). Points, mean % specific lysis; bars, SD. C, the CTL activities against the peptide-pulsed (left) or IL-13Ra2-expressing (right) EL4-HHD cells are combined for the comparison of relative CTL activities induced by either the wild-type IL-13Ra2345-353 peptide or its analogues. Against peptide-pulsed EL4-HHD (P < 0.05) at all E/T ratios for 9V versus wild-type peptide and 1A9V versus wild-type peptide by two-tailed Student’s t tests. Against IL-13Ra2-expressing EL4-HHD (P < 0.05) at E/T ratio of 10 and 40 for 9V versus wild-type and 1A9V versus wild-type peptide by two-tailed Student’s t tests. D, freshly isolated cells isolated from draining lymph nodes were stained with FITC-conjugated anti-mouse CD8 antibody and pycoerythrin-labeled HLA-A2/IL-13Ra2 tetramers. Numbers in each histogram indicate the percentage of CD8+/tetramer+ T cells within the lymphocyte-gated population.

background levels in both settings. Furthermore, immunization based vaccination, HHD mice were preimmunized twice with the with the 9V or 1A9V peptides induced higher levels of CTL IL-13Ra2345-353 wild-type or analogue peptides before being reactivity against the IL-13Ra2 345-353-loaded EL4-HHD and EL4- challenged i.c. with EL4-HHD-IL-13Ra2 tumor cells, which typically HHD-IL13Ra2 cells when compared with the CTL induced by the form solid and lethal tumor lesions in the brains of nonimmunized wild-type peptide (Fig. 3C). syngenic animals (Fig. 4A). As shown in Fig. 4B, all HHD mice To determine whether immunizations with the 9V or 1A9V treated with control vaccines succumbed due to progressive tumor + peptides induced the proliferation of IL-13Ra2345-353-reactive CD8 growth by day 70. In contrast, preimmunizations with the 9V or T cells more efficiently than the wild-type peptide, freshly isolated 1A9V analogue peptides resulted in extended, long-term survival draining lymph node cells isolated from immunized HHD mice (longer than 100 days) in three of six mice treated in each group were stained using the HLA-A2/IL13Ra2 tetramer. As shown in and showed statistically significant prolongations of survival when Fig. 3D, the analogue peptides induced higher percentages of compared with the control group (P = 0.0031 and 0.0016, tetramer-positive/CD8+ T cells (2.00% and 2.19%, respectively) respectively). Although one of the six mice preimmunized with when compared with the wild-type peptide (1.19%), suggesting that the wild-type IL-13Ra2345-353 peptide survived for longer than 100 improved CTL reactivity induced by peptide analogues may be, at days, the mean difference with the control group did not reach least in part, attributable to the enhanced proliferation of antigen- statistical significance (P = 0.0533). These results indicate that specific CD8+ T cells. We also confirmed that the CTLs induced by preimmunization with the 9V or 1A9V peptide induces effective the wild-type peptide also recognize the 1A9V peptide loaded on anti-brain tumor responses in HHD mice that are superior to those EL4-HHD cells, providing a validation for the use of the HLA-A2/IL- promoted by the wild-type IL-13Ra2-derived epitope. 13Ra2 tetramer in the evaluation of CTLs induced by both wild- Furthermore, splenocytes isolated from long-term survivor type and analogue peptides (data not shown). animals following immunizations with the 9V or 1A9V peptides 9V and 1A9V analogue peptide-based vaccines elicit superior exhibited specific lytic reactivity against both EL4-HHD cells pulsed protective immunity against lethal i.c. challenge dose of EL4- with the wild-type IL-13Ra2345-353 peptide and EL4-HHD-IL-13Ra2 HHD-IL-13RA2 tumor cells in HHD mice. To examine the cells in vitro (Fig. 4C). In contrast, control splenocytes isolated from comparative in vivo anti-intracranial tumor efficacy of peptide- nonimmunized mice did not show antigen-specific lytic ability

Cancer Res 2006; 66: (11).June 1, 2006 5888 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 2006 American Association for Cancer Research. IL-13Ra2-Antigenic Peptide Analogues following 6-day in vitro restimulation with the 9V or 1A9V peptides, suggesting that the observed lytic activities in immunized mice reflect memory T-cell responses in these long-term survivors. Peripheral immunizations with the IL-13Ra2-derived peptides also induced the infiltration of antigen-specific CD8+ T cells into the i.c. EL4-HHD-IL-13Ra2 tumor lesions, as shown by elevated numbers of IL-13Ra2-tetramer+/CD8+ T cells in BILs isolated from immunized versus control nonimmunized mice (Fig. 4D). To determine whether any autoimmune status is induced, we evaluated lungs, livers, and brains from the immunized mice for any excessive lymphocyte infiltration or tissue damage using H&E staining. In addition, brain sections were evaluated by Luxol Fast Blue staining (32) for possible demyelination. Although we are aware that vaccine-induced atypical autoimmune encephalitis may present differently than typical demyelination, we found no evidence of autoimmune response against these normal organs. In vitro induction of anti-IL-13Ra2345-353 CTL responses from PBMCs isolated from HLA-A2+ glioma patients. We have evaluated the ability of wild-type or analogue IL-13Ra2 peptides to stimulate specific CTL responses in vitro from 42 HLA-A2+ glioma patients. When using the 1A9V peptide for stimulations, 18 of 42 (42.9%) patients displayed positive cross-reactivity against the wild- + type IL-13Ra2345-353 peptide loaded onto T2 cells or HLA-A2 glioma cell lines that naturally process and present this epitope (Table 2). As we found that the 1A9V was more immunogenic than the wild type in the first several donors (Fig. 1), we did not test the wild-type peptide in all subsequent donors. Among 19 cases that were stimulated with wild-type peptide, 8 (42.1%) cases displayed positive reactivity. Although the positive response rate was not significantly different than that was observed with 1A9V, the magnitude of positive response in each case was constantly improved by the use of the 1A9V peptide compared with the wild- type peptide in accordance with the data in Fig. 1 (data not shown). We also evaluated whether positive responsiveness was related to clinical variables, such as patient age (>50 or <50), type of glioma (glioblastoma multiforme versus others), and prior treatment (i.e., chemotherapy). We did not discern any statistically meaningful correlations among any of these indices.

Discussion

In the present study, we have shown that the IL-13Ra2345-353 peptide analogues 9V and 1A9V exhibit a superior capacity (versus the wild-type peptide) to induce CTLs capable of (cross-)reacting against the wild-type IL-13Ra2345-353 epitope. This was true in both HLA-A2+ patients with glioma and HLA-A2-transgenic HHD mice. HLA-A2-binding and stability assays suggest that the improved Figure 4. Vaccination with the IL-13Ra2345-353 analogue peptides 9V or 1A9V promote enhanced protective immunity in HHD mice against a normally lethal immunogenicity of the 9V and 1A9V peptides was at least partially intracranial challenge with IL13Ra2-expressing syngeneic tumors. A to C, HHD attributable to the higher binding/stability of these analogue mice were immunized twice in the same manner as described for experiments peptides in HLA-A2 complexes that are required for specific CTL in Fig. 3 (n = 6 per group). On day 7 after the last vaccination, mice received stereotactic injections of 5 104 EL4-HHD-IL13Ra2 cells per mouse into the recognition. CTL assays analyzing reactivity versus peptide dose right basal ganglia. A, H&E staining of the tumor from a control, nonimmunized titration on T2 target cells indicated that the CTLs developed using mouse. Original magnification, 4. B, disease-free survival was monitored for each vaccination group. C, splenocytes were harvested from mice that survived the 9V or 1A9V analogue peptides possessed a higher functional for 100 days after the tumor challenge and stimulated in vitro for 6 days with avidity than those primed with wild-type peptide. Alternatively or the same peptide used for in vivo immunization (right). Splenocytes isolated from naive mice were also stimulated in vitro with the 9V or 1A9V peptide as additionally, our data could indicate that these peptide analogues controls (left). Splenocytes were tested for their lytic activity against peptide- promoted superior proliferation of specific CTL precursors owing pulsed EL4-HHD cells and EL4-HHD-IL-13Ra2 in 4 hours 51Cr release 5 to their improved binding/stability in HLA-A2 complexes. assays. D, HHD mice received 1 10 EL4-HHD-IL13Ra2 cells per mouse into the right basal ganglia on day 0 and were vaccinated with 9V or 1A9V peptides As to whether the analogue-induced CTLs possess higher avidity, on days 14 and 21 (n = 5 per group). BILs were harvested on day 28, pooled it has been shown that the affinity (33) or stability (34) between within the same treatment group, stained with pycoerythrin-conjugated HLA-A2/ IL13Ra2-tetramer and FITC-conjugated anti-CD8 mAb, and analyzed by flow TCRs and peptide-MHC complexes correlates with the avidity of cytometry. Numbers in each histogram indicate the percentage of CD8+/tetramer+ T cell-target interactions and the antitumor responsiveness of T cells among the total lymphocyte-gated population. www.aacrjournals.org 5889 Cancer Res 2006; 66: (11).June 1, 2006

Table 2. Characteristics of the glioma patients evaluated trafficking of vaccine-induced T effector cells into tumor lesions in in this study concert with therapeutic efficacy. In addition to these murine responders, we have also evaluated a + Characteristic Responder Nonresponder P total of 42 HLA-A2 glioma patients for their in vitro responsive- (n = 18) (n = 24) ness against wild-type and analogue IL-13Ra2345-353 peptides. Nearly half of our patients developed specific CTLs capable of Age recognizing the wild-type IL-13Ra2 peptide after stimulation with >50 7 10 0.6517 the 1A9V analogue peptide (i.e., 42.9%, 18 of 42). These CTL <50 11 13 recognized peptide-pulsed T2 cells or HLA-A2+ glioma cell lines Unknown 0 1 that constitutively express the IL-13Ra2 protein. Interestingly, Tumor type various background factors, such as age, tumor type (glioblastoma GBM 7 10 0.6925 multiforme versus other gliomas), and prior treatment with chemo- Not GBM 10 11 therapy, did not correlate with the patient’s ability to respond Unknown 1 3 Chemotherapy preferentially to the analogue peptides. Although further analyses + 4 6 0.9772 are clearly warranted to identify patient populations most suitable 10 13 for vaccine trials implementing IL-13Ra2 wild-type and/or ana- Unknown 4 5 logue peptides, these preliminary data suggest that conditions, Chemotherapy such as prior application of chemotherapy by itself, should not TMZ, CPT11+ 2 3 0.9871 serve as an exclusion criteria for participation in such prospective TMZ, CPT11 12 16 vaccine trials. Combination strategies using chemotherapy and Unknown 4 5 vaccination may also be appropriate but require further characterization of the effects of chemotherapy on the tumor-bearing Abbreviations: GBM, glioblastoma multiforme;TMZ, temozolomide; host immune response (41). CPT11, irinotecan. As the wild-type IL-13Ra2345-353 epitope and its analogue peptides were originally characterized in the context of presentation by the HLA-A*0201 allele (16), it is possible that our flow cytometry–based assessment of HLA-A2+ PBMC may have allowed us to include T cells. In these published studies, the intensity (33), or stability patients bearing non-HLA-A*0201 A2 subtypes. This could have (34) of specific T-cell staining with HLA tetramers, and threshold influenced our overall observed response rates, as previously of positive staining using titrating doses of tetramers (35) are observed for other CTL epitope peptides (42). In this regard, we purported to be indicative of the relative avidity of specific T cells. recognize that further characterization of the analogue peptides for Using these methods, however, we were unable to detect any their ability to bind to a variety of A2 subtypes is warranted. It has differences at the single cell level, suggesting that the functional been shown that >70% of the peptides that bind HLA-A *0201 with avidity of specific CTLs was comparable regardless of the priming high affinity also bind at least two other A2 supertype molecules peptide used (data not shown). (42);thus, identification of those analogue peptides exhibiting high In contrast, based on tetramer staining results obtained from HLA-A*0201 binding affinity may portend to the use of these both human HLA-A2+ donor- and HHD mouse–derived CTLs peptides in patients bearing a variety of HLA-A2 subtypes. (splenocytes), the 9V and 1A9V analogue peptides seem to induce We have also noted that donor-derived PBMCs that did not higher frequencies of specific CTLs when compared with respond to IL-13Ra2-1A9V can often react well against other GAAs, stimulations using the wild-type peptide. Hence, we believe that such as EphA2883-891 (43) and GP100 (44). In addition, tumor cells the improved CTL-inducing capacity of the analogue peptides is may undergo ‘‘immunoediting’’ in vivo, selecting for heterogeneous likely attributable to the stabilization of specific antigen presen- progressor lesions that may lose specific antigen or HLA allele tation to T-cell precursors, as has been shown by others (26). expression (45). Given such variability, we believe that an optimal As a surrogate marker for CTL responses, cytokine responses, glioma vaccine design should include multiple GAA-derived T-cell such as IFN-g (36, 37) and IL-2 (38), are often monitored. In our epitopes (43, 44). Our current study strongly supports the inclusion current study, we also evaluated IFN-g and IL-2 secretion levels from of peptides, such as the IL-13Ra2-9V and/or -1A9V analogues, in CTL cultures stimulated with native or analogue peptides using such formulations designed to expanding GAA-specific CTL in cytokine-specific ELISA and IFN-g enzyme-linked immunospot glioma patients. assays. However, our results failed to show any correlation between the levels of specific cytolysis and cytokine secretion by responder Acknowledgments CD8+ Tcells (data not shown). These suggests that more appropriate Received 1/30/2006;revised 3/28/2006;accepted 4/5/2006. laboratory surrogate assays for CTL responses may involve Grant support: Grants 1P01 NS40923 (I.F. Pollack and H. Okada) and 1P01 measurements of perforin (39) and/or granzyme B (40) release. CA100327 (W.J. Storkus and H. Okada), Doris Duke Charitable Foundation, James S. Our HHD mouse models support the improved in vivo efficacy of McDonnell Foundation (H. Okada), and Copeland Fund of the Pittsburgh Foundation (J. Eguchi). protective vaccinations incorporating the IL-13Ra2analogue The costs of publication of this article were defrayed in part by the payment of page peptides. For BIL analyses, as we needed a large volume of tumors charges. This article must therefore be hereby marked advertisement in accordance to obtain sufficient numbers of BILs, we used established tumors with 18 U.S.C. Section 1734 solely to indicate this fact. We thank Dr. Franc¸oisA. Lemonnier and Dr. Pravin T.P. Kaumaya for generously for immunizations. This may explain the relatively low frequency of providing us with HHD mice and Drs. Ronald L. Hamilton (University of Pittsburgh), IL-13Ra2-tetramer+/CD8+ T cells observed in BILs;nevertheless, Naruo Kuwashima (Jikei University, Tokyo), Melanie Tate (Brain Tumor Program, University of Pittsburgh Cancer Institute), and Wendy K. Fellows-Mayle (University of our data suggest that vaccinations with the 9V or 1A9V peptides Pittsburgh) for providing neuropathologic expertise, valuable insights, assistance in can yield systemic antigen-specific immunity but also the effective collecting blood samples, and technical assistance, respectively.

Cancer Res 2006; 66: (11).June 1, 2006 5890 www.aacrjournals.org

References Okada H. Identification of a novel HLA-A*0201 restricted antitumor CTL response without autoimmune side cytotoxic T lymphocyte epitope in a human glioma effects. J Immunol 2002;169:4840–9. 1. Wen PY, Kesari S. Malignant gliomas. Curr Neurol associated antigen, interleukin-13 receptor 2 chain. Clin 31. Alves PM, Faure O, Graff-Dubois S, et al. EphA2 as Neurosci Rep 2004;4:218–27. Cancer Res 2002;8:2851–5. target of anticancer immunotherapy: identification of 2. Holladay FP, Heitz T, Chen YL, Chiga M, Wood GW. 17. Debinski W, Gibo DM, Slagle B, Powers SK, Gillespie HLA-A*0201-restricted epitopes. Cancer Res 2003;63: Successful treatment of a malignant rat glioma with GY. Receptor for interleukin 13 is abundantly and 8476–80. cytotoxic T lymphocytes. Neurosurgery 1992;31:528–33. specifically over-expressed in patients with glioblastoma 32. al-Sabbagh AM, Goad EP, Weiner HL, Nelson PA. 3. Lampson LA. Beyond inflammation: site-directed multiforme. Int J Oncol 1999;15:481–6. Decreased CNS inflammation and absence of clinical immunotherapy. [Review] [32 refs]. Immunol Today 18. Debinski W, Gibo DM. Molecular expression ana- exacerbation of disease after six months oral adminis- 1998;19:17–22. lysis of restrictive receptor for interleukin 13, a brain tration of bovine myelin in diseased SJL/J mice with 4. Walker PR, Calzascia T, De Tribolet N, Dietrich PY. tumor-associated cancer/testis antigen. Mol Med 2000; chronic relapsing experimental autoimmune encepha- T-cell immune responses in the brain and their rele- 6:440–9. lomyelitis. J Neurosci Res 1996;45:424–9. vance for cerebral malignancies. Brain Res Brain Res Rev 19. Rosenberg SA, Yang JC, Schwartzentruber DJ, et al. 33. Yee C, Savage PA, Lee PP, Davis MM, Greenberg PD. 2003;42:97–122. Immunologic and therapeutic evaluation of a synthetic Isolation of high avidity melanoma-reactive CTL from 5. Ashley DM, Faiola B, Nair S, et al. Bone marrow- peptide vaccine for the treatment of patients with heterogeneous populations using peptide-MHC tet- generated dendritic cells pulsed with tumor extractsor metastatic melanoma. Nat Med 1998;4:321–7. ramers. J Immunol 1999;162:2227–34. tumor RNA induce antitumor immunity against central 20. Parkhurst MR, Salgaller ML, Southwood S, et al. 34. Dutoit V, Rubio-Godoy V, Doucey MA, et al. nervous system tumors. J Exp Med 1997;186:1177–82. Improved induction of melanoma-reactive CTL with Functional avidity of tumor antigen-specific CTL 6. Okada H, Tahara H, Shurin MR, et al. Bone marrow peptides from the melanoma antigen gp100 modified at recognition directly correlates with the stability of derived dendritic cells pulsed with a tumor specific HLA-A*0201-binding residues. J Immunol 1996;157: MHC/peptide multimer binding to TCR. J Immunol peptide elicit effective anti-tumor immunity against 2539–48. 2002;168:1167–71. intracranial neoplasms. Int J Cancer 1998;78:196–201. 21. Keogh E, Fikes J, Southwood S, et al. Identification of 35. Ercolini AM, Ladle BH, Manning EA, et al. Recruit- 7. Liau LM, Black KL, Prins RM, et al. Treatment of new epitopes from four different tumor-associated ment of latent pools of high-avidity CD8(+) T cells to intracranial gliomas with bone marrow-derived dendrit- antigens: recognition of naturally processed epitopes the antitumor immune response. J Exp Med 2005;201: ic cells pulsed with tumor antigens. J Neurosurg 1999;90: correlates with HLA-A*0201-binding affinity. J Immunol 1591–602. 1115–24. 2001;167:787–96. 36. Mailliard RB, Wankowicz-Kalinska A, Cai Q, et al. 8. Okada H, Giezeman-Smits KM, Tahara H, et al. 22. Rivoltini L, Squarcina P, Loftus DJ, et al. A super- alpha-type-1 polarized dendritic cells: a novel immuni- Effective cytokine gene therapy against an intracranial agonist variant of peptide MART1/Melan A27–35 elicits zation tool with optimized CTL-inducing activity. glioma using a retrovirally transduced IL-4 plus HSV-TK anti-melanoma CD8+ T cells with enhanced functional Cancer Res 2004;64:5934–7. tumor vaccine. Gene Ther 1999;6:219–26. characteristics: implication for more effective immuno- 37. Herr W, Ranieri E, Olson W, et al. Mature dendritic 9. Okada H, Villa LA, Attanucci J, et al. Cytokine gene therapy. Cancer Res 1999;59:301–6. cells pulsed with freeze-thaw cell lysates define an therapy of gliomas: effective induction of therapeutic 23. Kirksey TJ, Pogue-Caley RR, Frelinger JA, Collins EJ. effective in vitro vaccine designed to elicit EBV-specific immunity to intracranial tumors by peripheral immu- The structural basis for the increased immunogenicity CD4(+) and CD8(+) T lymphocyte responses. Blood nization with interleukin-4 transduced glioma cells. of two HIV-reverse transcriptase peptide variant/class I 2000;96:1857–64. Gene Ther 2001;8:1157–66. major histocompatibility complexes. J Biol Chem 1999; 38. Carrabba MG, Castelli C, Maeurer MJ, et al. 10. Okada H, Lieberman FS, Edington HD, et al. 274:37259–64. Suboptimal activation of CD8(+) T cells by melanoma- Autologous glioma cell vaccine admixed with interleu- 24. Bownds S, Tong-On P, Rosenberg SA, Parkhurst M. derived altered peptide ligands: role of Melan-A/MART-1 kin-4 gene transfected fibroblasts in the treatment of Induction of tumor-reactive cytotoxic T-lymphocytes optimized analogues. Cancer Res 2003;63:1560–7. recurrent glioblastoma: preliminary observations in a using a peptide from NY-ESO-1 modified at the carboxy- 39. Monsurro V, Nagorsen D, Wang E, et al. Functional patient with a favorable response to therapy. J Neuro- terminus to enhance HLA-A2.1 binding affinity and heterogeneity of vaccine-induced CD8(+) T cells. oncol 2003;64:13–20. stability in solution. J Immunother 2001;24:1–9. J Immunol 2002;168:5933–42. 11. Okada H, Pollack IF, Lieberman F, et al. Gene therapy 25. Chen JL, Dunbar PR, Gileadi U, et al. Identification of 40. Waterhouse NJ, Sedelies KA, Clarke CJ. Granzyme B: of malignant gliomas: a pilot study of vaccination with NY-ESO-1 peptide analogues capable of improved the chalk-mark of a cytotoxic lymphocyte. J Transl Med irradiated autologous glioma and dendritic cells stimulation of tumor-reactive CTL. J Immunol 2000; 2004;2:36. admixed with IL-4 transduced fibroblasts to elicit an 165:948–55. 41. Su YB, Sohn S, Krown SE, et al. Selective CD4+ immune response. Hum Gene Ther 2001;12:575–95. 26. Gross DA, Graff-Dubois S, Opolon P, et al. High lymphopenia in melanoma patients treated with temo- 12. Yu JS, Wheeler CJ, Zeltzer PM, et al. Vaccination vaccination efficiency of low-affinity epitopes in anti- zolomide: a toxicity with therapeutic implications. J Clin of malignant glioma patients with peptide-pulsed den- tumor immunotherapy. J Clin Invest 2004;113:425–33. Oncol 2004;22:610–6. dritic cells elicits systemic cytotoxicity and intracranial 27. Pascolo S, Bervas N, Ure JM, et al. HLA-A2.1- 42. Sidney J, Southwood S, Mann DL, et al. Majority of T-cell infiltration. Cancer Res 2001;61:842–7. restricted education and cytolytic activity of CD8(+) peptides binding HLA-A*0201 with high affinity cross- 13. Yu JS, Liu G, Ying H, et al. Vaccination with tumor T lymphocytes from beta2 microglobulin (beta2m) HLA- react with other A2-supertype molecules. Hum Immunol lysate-pulsed dendritic cells elicits antigen-specific, A2.1 monochain transgenic H-2Db beta2m double 2001;62:1200–16. cytotoxic T-cells in patients with malignant glioma. knockout mice. J Exp Med 1997;185:2043–51. 43. Hatano M, Eguchi J, Tatsumi T, et al. EphA2 as a Cancer Res 2004;64:4973–9. 28. Walker PR, Calzascia T, Schnuriger V, et al. The brain glioma-associated antigen: a novel target for glioma 14. Kikuchi T, Akasaki Y, Irie M, et al. Results of a phase I parenchyma is permissive for full antitumor CTL vaccines. Neoplasia 2005;7:717–22. clinical trial of vaccination of glioma patients with effector function, even in the absence of CD4 T cells. 44. Liu G, Ying H, Zeng G, et al. HER-2, gp100, and fusions of dendritic and glioma cells. Cancer Immunol J Immunol 2000;165:3128–35. MAGE-1 are expressed in human glioblastoma and Immunother 2001;50:337–44. 29. Calzascia T, Masson F, Di Berardino-Besson W, et al. recognized by cytotoxic T cells. Cancer Res 2004;64: 15. Liau LM, Prins RM, Kiertscher SM, et al. Dendritic Homing phenotypes of tumor-specific CD8 T cells are 4980–6. cell vaccination in glioblastoma patients induces predetermined at the tumor site by crosspresenting 45. Jager E, Ringhoffer M, Karbach J, et al. Inverse systemic and intracranial T-cell responses modulated APCs. Immunity 2005;22:175–84. relationship of melanocyte differentiation antigen by the local central nervous system tumor microenvi- 30. Francini G, Scardino A, Kosmatopoulos K, et al. High- expression in melanoma tissues and CD8+ cytotoxic- ronment. Clin Cancer Res 2005;11:5515–25. affinity HLA-A(*)02.01 peptides from parathyroid T-cell responses: evidence for immunoselection of 16. Okano F, Storkus WJ, Chambers WH, Pollack IF, hormone-related protein generate in vitro and in vivo antigen-loss variants in vivo. Int J Cancer 1996;66:470–6.

www.aacrjournals.org 5891 Cancer Res 2006; 66: (11).June 1, 2006

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 2006 American Association for Cancer Research. Identification of Interleukin-13 Receptor α2 Peptide Analogues Capable of Inducing Improved Antiglioma CTL Responses

Junichi Eguchi, Manabu Hatano, Fumihiko Nishimura, et al.

Cancer Res 2006;66:5883-5891.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/66/11/5883

Cited articles This article cites 44 articles, 23 of which you can access for free at: http://cancerres.aacrjournals.org/content/66/11/5883.full#ref-list-1

Citing articles This article has been cited by 10 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/66/11/5883.full#related-urls

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/66/11/5883. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.