Search

Intracellular tumor antigens presented on the cell surface in the context of human leukocyte antigen (HLA) molecules have been targeted by T cell-based therapies, but there has been little progress in developing small-molecule drugs or antibodies directed to these antigens. Here we describe a bispecific T-cell engager (BiTE) antibody derived from a T-cell receptor (TCR)-mimic monoclonal antibody (mAb) ESK1, which binds a peptide derived from the intracellular oncoprotein WT1 presented on HLA-A*02:01. Despite the very low density of the complexes at the cell surface, ESK1-BiTE selectively activated and induced proliferation of cytolytic human T cells that killed cells from multiple leukemias and solid tumors in vitro and in mice. We also discovered that in an autologous in vitro setting, ESK1-BiTE induced a robust secondary CD8 T-cell response specific for tumor-associated antigens other than WT1. Our study provides an approach that targets tumor-specific intracellular antigens without using cell therapy and suggests that epitope spreading could contribute to the therapeutic efficacy of this BiTE.

Wilms tumor protein 1 (WT1) is a transcription factor overexpressed in several types of leukemia and solid tumors. For this reason, WT1 is an attractive target for immunotherapy. Four peptide nonamers from WT1 have been identified by others to generate a WT1-specific cytotoxic response in the context of human leukocyte antigen (HLA)-A0201 and A2402. However, as WT1 is a self-antigen, breaking tolerance is a potential obstacle to vaccination. Here, we use a strategy to circumvent tolerance by designing synthetic immunogenic analog peptides that could crossreact to the native peptides (a heteroclitic response). A number of synthetic peptides derived from nonamer sequences of the WT1 protein were designed in which single amino-acid substitutions were introduced at HLA-A0201 major histocompatibility complex (MHC)-binding positions. Several of new peptides could stabilize MHC class I A0201 molecules better than native sequences. Some analogs were also able to elicit WT1-specific T-cell recognition and cytotoxic T-cell lymphocytes more effectively than native sequences. Importantly, T cells stimulated with the new analogs crossreacted with the native WT1 peptide sequence and were able to kill HLA-matched chronic myeloid leukemia cell lines. In conclusion, analog heteroclitic WT1 peptides with increased immunogenicity can be synthesized and are potential cancer vaccine candidates.

A tumor-specific, bcr-abl-derived fusion peptide vaccine can be safely administered to patients with chronic myelogenous leukemia (CML) and can elicit a bcr-abl peptide-specific T-cell immune response. In the present phase 2 trial, 14 patients with CML in chronic phase were vaccinated with 6 fusion peptides mixed with Quillaja saponaria (QS-21). No significant toxic effects were observed. In 14 of 14 patients, delayed-type hypersensitivity (DTH) and/or CD4 proliferative responses developed after beginning vaccinations, and 11 of 14 patients showed interferon-gamma (IFN-gamma) release by CD4 enzyme-linked immunospot (ELISPOT) at one or more time points. These responses were CD4+CD45RO+. A peptide-specific CD8+ interferon-gamma ELISPOT was found in 4 patients. Four patients in hematologic remission had a decrease in Philadelphia chromosome (Ph) percentage (3 concurrently receiving interferon-alpha and 1 on imatinib mesylate), and 3 patients in molecular relapse after allogenic transplantation became transiently polymerase chain reaction (PCR) negative after vaccination; 2 of these patients received concurrent donor lymphocyte infusion (DLI). All 5 patients on IFN-alpha ultimately reached a complete cytogenetic remission. In conclusion, a tumor-specific bcr-abl breakpoint peptide-derived vaccine can be safely administered and can reliably elicit measurable peptide-specific CD4 immune responses, including in patients after bone marrow transplantation, on interferon, or on imatinib mesylate. A relationship between the clinical responses and vaccination cannot be determined from this trial. (Blood. 2004;103:1037-1042)

Preferentially expressed antigen in melanoma (PRAME) is a cancer-testis antigen that is expressed in many cancers and leukemias. In healthy tissue, PRAME expression is limited to the testes and ovaries, making it a highly attractive cancer target. PRAME is an intracellular protein that cannot currently be drugged. After proteasomal processing, the PRAME300-309 peptide ALYVDSLFFL (ALY) is presented in the context of human leukocyte antigen HLA-A*02:01 molecules for recognition by the T cell receptor (TCR) of cytotoxic T cells. Here, we have described Pr20, a TCR mimic (TCRm) human IgG1 antibody that recognizes the cell-surface ALY peptide/HLA-A2 complex. Pr20 is an immunological tool and potential therapeutic agent. Pr20 bound to PRAME+HLA-A2+ cancers. An afucosylated Fc form (Pr20M) directed antibody-dependent cellular cytotoxicity against PRAME+HLA-A2+ leukemia cells and was therapeutically effective against mouse xenograft models of human leukemia. In some tumors, Pr20 binding markedly increased upon IFN-γ treatment, mediated by induction of the immunoproteasome catalytic subunit β5i. The immunoproteasome reduced internal destructive cleavages within the ALY epitope compared with the constitutive proteasome. The data provide rationale for developing TCRm antibodies as therapeutic agents for cancer, offer mechanistic insight on proteasomal regulation of tumor-associated peptide/HLA antigen complexes, and yield possible therapeutic solutions to target antigens with ultra-low surface presentation.

Bi-specific T cell engager antibody (BiTE) therapy has recently emerged as an effective immunotherapy by redirecting polyclonal T cell cytotoxicity against cell surface protein on tumor cells. We generated the first BiTE construct derived from a TCR-mimic monoclonal antibody (mAb), ESK1, specific for a peptide from an intracellular oncoprotein, WT1, in the context of HLA-A*02:01 molecules. Despite the low density peptide/HLA-A2 complex on the cell surface, ESK-BiTE was able to selectively activate and induce proliferation of cytolytic human T cells to kill multiple leukemias and cancers in vitro and in mice. Surprisingly, we also discovered that in an autologous setting, ESK-BiTE induced a robust secondary CD8 T cell response specific for antigens other than WT1, including HLA-A2-restricted her2-neu-derived peptide 369-377, in patients with her2- positive ovarian cancer. Therefore, the study demonstrated a new vaccinal mechanism for BiTE mAb action that could contribute to more effective long-term therapeutic activity of BiTE's and further broaden their reach to other tumor antigens not previously known or originally targeted. Citation Format: Tao Dao, Dmitry Pankov, Andrew Scott, Tatyana Korontsvit, Victoriya Zakhaleva, Manuel Direito de Morais Guerrerio, Yiyang Xu, Jingyi Xiang, Su Yan, Nicholas Veomett, Nicholas Veomett, Leonid Dubrovsky, Michael Curcio, Ekaterina Doubrovina, Cheng Liu, Richard J. O'Reilly, David A. Scheinberg. Potent therapeutic and immunological effects of the first T-BiTE derived from a TCR-mimic antibody targeting intracellular oncoprotein WT1. [abstract]. In: Proceedings of the CRI-CIMT-EATI-AACR Inaugural International Cancer Immunotherapy Conference: Translating Science into Survival; September 16-19, 2015; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2016;4(1 Suppl):Abstract nr A055.

We studied the feasibility of using single-wall carbon nanotubes (SWNTs) as antigen carriers to improve immune responses to peptides that are weak immunogens, a characteristic typical of human tumor antigens. Binding and presentation of peptide antigens by the MHC molecules of antigen presenting cells (APCs) is essential to mounting an effective immune response. The Wilm's tumor protein (WT1) is upregulated in many human leukemias and cancers and several vaccines directed at this protein are in human clinical trials. WT1 peptide 427 induces human CD4 T cell responses in the context of multiple human HLA-DR.B1 molecules, but the peptide has a poor binding affinity to BALB/c mouse MHC class II molecules. We used novel, spectrally quantifiable chemical approaches to covalently append large numbers of peptide ligands (0.4 mmol/g) onto solubilized SWNT scaffolds. Peptide-SWNT constructs were rapidly internalized into professional APCs (dendritic cells and macrophages) within minutes in vitro, in a dose dependent manner. Immunization of BALB/c mice with the SWNT-peptide constructs mixed with immunological adjuvant induced specific IgG responses against the peptide, while the peptide alone or peptide mixed with the adjuvant did not induce such a response. The conjugation of the peptide to SWNT did not enhance the peptide-specific CD4 T cell response in human and mouse cells, in vitro. The solubilized SWNTs alone were nontoxic in vitro, and we did not detect antibody responses to SWNT in vivo. These results demonstrated that SWNTs are able to serve as antigen carriers for delivery into APCs to induce humoral immune responses against weak tumor antigens.

Purpose: Wilms' tumor 1 protein (WT1), a transcription factor overexpressed in malignant mesothelioma, leukemias, and other solid tumors, is an ideal target for immunotherapy. WT1 class I peptide epitopes that were identified and shown to stimulate CD8+ T cells are being tested as vaccine candidates in several clinical trials. The induction and maintenance of a robust memory CD8+ cytotoxic T-cell response requires CD4+ T-cell help.Experimental Design: Three HLA class II peptide epitopes of WT1 with high predictive affinities to multiple HLA-DRB1 molecules were identified using the SYFPEITHI algorithm. Due to the highly polymorphic nature of the HLA class II alleles, such reactivity is critical in the development of a broadly useful therapeutic. One of the WT1 CD4+ peptide epitopes, 122-140, comprises a previously identified CD8+ peptide epitope (126-134). By mutating residue 126 from an arginine to a tyrosine, we embedded a synthetic immunogenic analogue CD8+ epitope (126-134) inside the longer peptide (122-140). This analogue was previously designed to improve immunogenicity and induce a potent CD8+ response.Results: WT1 peptides 328-349 and 423-441 are able to stimulate a peptide-specific CD4+ response that can recognize WT1+ tumor cells in multiple HLA-DRB1 settings as determined by IFN-γ enzyme-linked immunospot assays. The mutated WT1 peptide epitope 122-140 is able to induce CD4+ and cytotoxic CD8+ WT1-specific T-cell responses that can recognize the native WT1 epitopes on the surface of human WT1+ cancer cells. Cross-priming experiments showed that antigen-presenting cells pulsed with either mesothelioma or leukemia tumor lysates can process and present each of the CD4+ peptides identified.Conclusions: These studies provide the rationale for using the WT1 CD4+ peptides in conjunction with CD8+ peptide epitopes to vaccinate patients with WT1-expressing cancers.

Purpose: Presentation of exogenous antigen by MHC class I molecules, or cross-presentation, is a property of dendritic cells, which is considered crucial for the priming of cytotoxic T-cell response to tumor antigens. However, the precise mechanisms of this process are not fully understood.Experimental Design and Results: We show here in a human in vitro system, using B lymphoma cells as a tumor model, that the cross-presentation of cell-associated antigens to T cells by dendritic cells requires “help” from natural killer cells. When autologous dendritic cells that had taken up apoptotic B lymphoma cells and induced to a fully mature state were used to stimulate nonadherent cells of peripheral blood mononuclear cells from healthy donors, they induced strong cytotoxicity against B lymphoma cells in a HLA-A0201-restricted manner. The cells failed to induce cytotoxicity, however, when purified T cells were used as effector cells. Depletion of CD56+ cells, but not CD14+ or CD19+ cells, abrogated the cytotoxicity of nonadherent cells, showing that the help was provided by natural killer cells. Further, when natural killer cells were present in the cultures, a strong and persistent production of interleukin-18, but not interleukin-12 and interleukin-15, was observed. Blocking interleukin-18 significantly reduced the cytotoxicity of nonadherent cells against B lymphoma cells.Conclusions: These results suggest that capture of tumor cells and a full maturation status of dendritic cells are not sufficient to cross-prime CD8 T cells. Effective cross-priming requires further activation of dendritic cells by natural killer cells and an abundant production of interleukin-18, which, along with other yet undefined mechanisms, contribute to the generation of CTL response against B-cell lymphoma.

Chronic myelogenous leukemia (CML) presents a unique opportunity to develop therapeutic strategies using vaccination against a truly tumor-specific antigen that is also the oncogenic protein required for neoplasia. We have shown in phase I and II trials that a tumor-specific, bcr-abl-derived peptide vaccine can be safely administered to patients with chronic phase CML and can elicit a reliable specific CD4 immune response. However, variable CD8 responses and no HLA A0201-restricted responses were found. One strategy to circumvent this poor immunogenicity is to design synthetic immunogenic analog peptides that cross-react with the native peptides (a heteroclitic response). The aim of this study was to design such peptides.By using computer prediction analysis. We designed a number of synthetic peptides derived from the junctional sequences of CML (p210/b3a2 or p210/b2a2) in which single and double amino acid substitutions were introduced at key HLA A0201 binding positions. The binding of these peptides was tested by a thermostabilization assay using a T2 cell line.We found three peptides that predicted good binding to HLA A0201 molecules and stabilized MHC class I A0201 molecules on the surface of T2 cell lines. These peptides were screened for eliciting HLA restricted, peptide-specific cytotoxic T lymphocyte responses using CD3+ T cells from several A0201 donors and CML patients. The CD8+ cytotoxic T lymphocytes lines were assessed by either interferon-g ELISPOT or a chromium release assay using pulsed, HLA-matched leukemic cell lines. The analog peptides generated larger immune responses (increased CD8 T-cell precursor frequency) than did the native peptides. Importantly, CD8+ T cells stimulated with the new synthetic peptides cross-reacted with the native bcr-abl peptides.In conclusion, analog CML fusion peptides with increased immunogenicity and heteroclitic properties can be synthesized and may be useful in vaccination strategies.

Abstract 1677 The Wilms' tumor oncogene protein (WT1) is an attractive target for immunotherapy for a wide range of leukemias and cancers. WT1 is an intracellular transcription factor that is widely expressed in acute and chronic leukemias, but not appreciably in normal adult tissues. WT1 has been implicated in oncogenesis and appears to be expressed in the CML stem cell. Thus far, the intracellular location of WT1 has limited its use as a cancer target to T cell-based vaccine therapies. Several such vaccines directed to WT1 peptides are in clinical trials worldwide. However, such approaches is likely to be weaker and slower acting than typical antibody therapies that act directly to kill cells, especially for a leukemia that is not in remission. The WT1 peptide (RMFPNAPYL) is presented by HLA-A0201 and induces cytotoxic CD8 T cells capable of killing WT1+ leukemia cells. Therefore, we hypothesized that a monoclonal antibody (mAb) specific for WT1 peptide/HLA-A2 complexes on the cell surface (a “TCR-like antibody”) could be an effective direct therapeutic agent alone, or armed with potent anti-cancer agents. In addition, such an antibody would provide a potential tool for identifying appropriate patients for clinical trials of both WT1 vaccine and WT1 mAb therapies, or as a prognostic tool. Using phage-display technology, we selected a number of human scFv phage mAbs specific for WT1 peptide RMFPNAPYL, when bound in the context of HLA-A0201 molecules. Bivalent, full-length human IgG1 mAbs were next constructed and tested for biochemical and immunological functions. The scFv phage mAbs bound specifically to T2 cells (TAP deficient) pulsed with WT1 peptide RMFPNAPYL, but not other HLA-A0201-binding peptides, demonstrating specific recognition of the WT1peptide/HLA-A0201 complex. Flow cytometry and radio-immunoassay was used to quantify sites, determine avidity, and show specificity. The IgG1 mAb showed high binding affinity (Kd= 0.2 nM) for the WT1 peptide/A0201 complex on live cells. The IgG1 mAb also recognized leukemia and tumor cell lines (n= 8 of 12) that expressed WT1, in an HLA-A0201, WT1 peptide-restricted manner, but not WT1-negative or HLA-A02-negative cells (n=13). The mAb was capable of killing WT1 peptide-pulsed T2 cells, as well as un-pulsed WT1+, A0201 + cancer cell lines, demonstrating presentation of the epitope in sufficient numbers for therapeutic attack. Complement mediated killing of cells was not seen. Alanine scanning of the epitope pointed to peptide position #1 as critical. In conclusion, cytotoxic human monoclonal antibodies targeting WT1 peptide/HLA complexes represent a novel therapeutic approach to target leukemias and solid tumors that over-express the intracellular WT1 oncoprotein. Disclosures: Yan: eureka therapeutics: Employment. Zhou:eureka therapeutics: Employment. Liu:eureka therapeutics: Employment.

Wilms tumor protein (WT1) is a transcription factor selectively over expressed in several types of leukemia and solid tumors, making it a promising potential target antigen for immunotherapy. Several open clinical trials use native or altered peptide sequences derived from the WT1 protein in order to overcome the weak immunogenicity of the self-antigen. Here we report a new strategy to circumvent tolerance by designing peptides that incorporate non-natural amino acids into the native sequence of WT1 peptides. Starting from the nonamer sequences WT1 187–195 and WT1 235–243, eight peptides containing natural amino acids and nine peptides in which different chemical modifications (fluorination, photo-reactive azido groups or benzophenone groups) were introduced at major histocompatibility complex (MHC) and T cell receptor binding positions, were synthesized. The new non-natural peptides could stabilize MHC class I molecules better than the native sequences and were also able to elicit strong specific T-cell responses. Photo-reactive peptides were additionally modified with biotin handles to allow streptavidin-biotin pull down and western blot analysis of kinetics of binding and catabolism. Upon UV irradiation, these peptides covalently bound to MHC molecules on the live cells; clearance of the peptide-MHC covalent complex occurred over 24 hours, consistent with the T2 thermo-stabilization data for the same peptide. Further catabolic studies may elucidate the important or novel cellular proteins involved in antigenic peptide processing and cross presentation and should aid in vaccine development. We are investigating whether covalent interaction with the MHC may lead to alterations in immune responses as well. T cells stimulated with one of the synthetic peptides (WT1J-W4WF) cross-reacted with the native WT1J sequence and were able to kill WT1 positive HLA-A0201 matched acute lymphoblastic leukemia cell lines. In conclusion, this study shows that peptides with non-natural amino acids can be successfully incorporated into T cell epitopes to provide increased immunogenicity and novel biological information.

Wilms tumor protein 1 (WT1) is a transcription factor over-expressed in several types of leukemias and solid tumors, making it an ideal target for immunotherapy. A number of class I binding WT1 peptides have been identified and shown to stimulate CD8+ T cells. These peptides are being tested as potential cancer vaccine candidates in a variety of clinical trials. However, the induction and maintenance of a robust memory CD8+ cytotoxic T cell response requires CD4+ T cell help. Herein we report the identification of three HLA Class II peptide epitopes of WT1 using the SYFPEITHI and RANKPEP predictive algorithms. Peptides 328–349 and 423–441 are able to stimulate a peptide specific CD4+ response that can recognize WT1 positive tumor cells in multiple HLA-DRB*1 settings, as determined by IFN-gamma ELISPOT assays. Due to the highly polymorphic nature of the HLA class II alleles, such broad reactivity is critical in the development of a universal therapeutic. In addition, we identified a WT1 CD4+ peptide epitope (122–140) that lies within close proximity to a previously identified CD8+ peptide epitope (126–134). Residue 126 was mutated from an Arginine (R) to a Tyrosine (Y) thereby embedding a synthetic immunogenic analog CD8+ peptide that was previously designed to improve immunogenicity and induce a potent CD8+ response. Mutated peptide 122–140 is able to induce a CD4+ and cytotoxic CD8+ WT1 specific T cell response that can recognize the native WT1 epitopes on the surface of human CML and solid tumor cells. Cross-priming experiments demonstrated that APCs pulsed with either CML or mesothelioma tumor lysates can process and present each of the CD4+ peptides identified. These studies provide the rationale for using the three WT1 CD4+ peptides in conjunction with CD8+ peptide epitopes to vaccinate patients with WT1 expressing cancers.

6514 Background: CML is characterized by a tumor-specific fusion oncoprotein, BCR-ABL. Peptides spanning the B3A2 breakpoint of BCR-ABL can elicit MHC restricted T cell responses and clinical responses. No such data have previously been reported for vaccines targeting B2A2. Methods: We initiated a clinical trial administering synthetic analog breakpoint specific peptide vaccines for either B3A2 or B2A2 breakpoints to CML patients with major or complete cytogenetic remission. Measurable minimal disease was determined by either quantitative polymerase chain reaction (RQ-PCR) or nested PCR for BCR-ABL. Vaccine was administered with GM-CSF and Montanide ISA 51 subcutaneously. Eleven vaccinations were planned over the course of 1 year with the first 5 doses administered bi-weekly. Results: Eleven of 20 planned patients have been accrued to the study. Eight had documented cytogenetic remission and 3 were BCR-ABL positive via FISH analysis. Seven of the 11 patients had measurable BCR-ABL transcript levels as determined by RQ-PCR and all 11 were positive using the nested technique. Of the 5 patients positive for BCR-ABL (by RQ-PCR) before vaccination, 3 converted from RQ-PCR positive to negative while the transcript levels in the other patients decreased by approximately 1 log after 5 doses. All 5 of these patients remained positive using the nested PCR technique. Vaccination was well tolerated with local skin reactions at the injection sites. Immunologic reactivity was assessed ex vivo by CD4 autologous proliferation assay and T cell interferon (IFN) secretion (ELISPOT) assay. Ten of 11 patients were immunologically unreactive to the peptides prior to the vaccinations. Following the 5th dose of vaccine, 8/8 patients (including 2 with B2A2) tested showed significant immunologic responses by one of the above assays. Conclusions: These preliminary results suggest vaccination with synthetic analog peptides derived from CML proteins results in immunologic responses and may be associated with molecular improvement. Clinical efficacy of these vaccines in reducing/ eliminating minimal disease has yet to be established. [Table: see text]

The recent identification of the JAK2 V617F mutation in myeloproliferative diseases including polycythemia vera (PV), essential thrombocythemia (ET), and myelofibrosis has the potential to serve as a new therapeutic target for these disorders. We hypothesized that the mutated JAK2 kinase sequence could generate new immunogenic peptide-epitopes that could be recognized by the human immune system. Therefore, such sequences could be potential vaccine candidates. The myeloproliferative diseases, especially PV and ET, are ideal diseases for a vaccine approach as they are indolent disorders and also not suitable for use of possibly toxic chemotherapies. Using computer based predictive algorithms, potentially immunogenic nonamer peptides spanning the V617F region of JAK2 were identified. Since JAK2 is a normally expressed self-protein, the goal is to break immune tolerance to the mutated JAK2 sequence without allowing significant recognition of the native sequence. Therefore, synthetic analogs of the V617F mutated peptides with stronger predictive binding to MHC were designed by introducing single amino-acid substitutions into the HLA-A0201 and A0301 binding positions. Corresponding non-mutated native sequences showed little predicted binding. The binding of the mutated and synthetic analog peptides was confirmed in vitro using an MHC class I stabilization assay on TAP deficient cells. Native sequences bound poorly as expected. JAK2 peptides derived from V617F sequences as well as their analogs were used to elicit MHC-restricted, peptide-specific CTL responses in an in vitro model using T-cells from either HLA-A0201 or A0301 healthy donors after stimulation with autologous monocyte-derived dendritic cells. The analog peptides were shown to be immunogenic in vitro, and activated CD8+ T cells that recognized the synthetic analog peptide pulsed targets as determined by an IFN-gamma ELISPOT assay. In addition, and most importantly, these T cells cross-reacted with cellular targets that were pulsed with the mutated V617F peptide associated with the myeloproliferative diseases, but not with the native JAK2 peptide. These analog peptides thus satisfied our initial goals for a vaccine candidate. Further studies with these peptides and others eliciting CD4+ responses will determine their potential for successful use as vaccine candidates

Major Histocompatibility Complex class I (MHC-I) molecules present antigenic peptides to T cells on the cell surface as a prerequisite for stimulating cytotoxic T cell response. Thus, the ability to reliably identify the peptides that can bind to MHC molecules is of practical importance for rapid vaccine development. Several computer-based prediction methods have been applied to study the interaction of MHC class I/peptide binding. Here we have compared three of the most commonly used predictive algorithms BIMAS, SYFPEITHI and Rankpep with actual binding of HLA-A*0201 peptides in vitro. Forty six HLA-A*0201 peptides were selected from several target oncoproteins: Wilms’ tumor (WT1), native and imatinib- mutated bcr-abl p210 and JAK2 protein. Experimental peptide binding to HLA-A*0201 was assessed using a MHC stabilization assay on T2, TAP deficient cells. Peptides were considered to show positive in vitro binding if the mean fluorescence was at least 50 % of the binding of a high affinity reference peptide. Peptides qualified as positive in vitro if the BIMAS score was ≥ 100, the SYFPEITHI score ranked ≥ 24 or the Rankpep was ≥ 50. Results are summarized below: BIMAS SYFPEITHI RANKPEP Sensitivity 84 % 72 % 60 % Specificity 76 % 71 % 81 % Positive Predictive Value 84 % 72 % 60 % Negative Predictive Value 80 % 68 % 63 % Combining two or more computer methods did not appear to improve the predictive value. In conclusion, of the three predictive algorithms, the best correspondence with the actual MHC binding was demonstrated with the BIMAS algorithm. Predictive computer algorithms are important for preselection of potential T-cell epitope candidates for the application in vaccine design.

Intracellular processing of the breakpoint products of the bcr-abl fusion gene may generate novel peptides that, if capable of binding to HLA class I molecules, would be potential targets for a cytotoxic T lymphocyte (CTL) response. In humans, peptides derived from the e1a2 p190 breakpoint have generated peptide-specific MHC class II proliferative responses. Short peptides have shown binding to MHC class I molecules, although processing and presentation of endogenous p190 protein has not been shown. In a Kd Balb/c mouse model, we tested the hypothesis that single, key amino acids near the breakpoint could be the reason of lack of immunogenicity of the p190 breakpoint peptides. We synthesized a native peptide from the 190 breakpoint (AFHGDAEAL) as well as a synthetic mutated peptide (AYHGDAEAL), which showed excellent predicted binding on the BIMAS algorithm (1152 and 2880 respectively), although in vivo experiments did not show any specific CTL response. In order to assess if the lack of immunogenicity in vivo was due to the absence of binding to the MHC class I molecule rather than to poor TCR interactions, we designed a series of peptides where neutral amino acid, alanine, substitutions were introduced at different potential binding sites in the synthetic peptide: at position three (AYAGDAEAL), position four (AYHADAEAL), position five (AYHGAAEAL) and position seven (AYHGDAAAL). The binding of these altered peptides to H2 class I was assessed using a MHC stabilization assay on T2-Kd cells (TAP deficient cells). In spite of the good computer prediction for binding the MHC stabilization assay did not show evidence of binding of the native and synthetic peptides (

Imatinib mesylate is now standard treatment for chronic myelogenous leukemia. In spite of its high efficacy in controlling the disease, residual bcr-abl cells measured by RT-PCR are still detectable in more than 95% of the patients treated with this drug. In addition, recent reports describe the emergence of resistance to Imatinib mesylate. A major mechanism for the appearance of resistance, are point mutations changing amino acids in the bcr-abl tyrosine kinase that make imatinib ineffective. We hypothesized that some bcr-abl kinase mutations selected by imatinib would generate new immunogenic peptide-epitopes that could be recognized by the human immune system. For this purpose, we selected with predictive computer algorithms potentially immunogenic mutated amino acid sequences and synthetic analogues of mutated bcr-abl kinase sequences to study for immunogenicity and anti-leukemia activity in an in vitro human system. Initially we designed a number of synthetic peptides derived from the most frequent bcr-abl kinase mutations induced by imatinib in which single amino acid substitution was introduced at different MHC-binding positions. We investigated the most common mutations: E255V, E255K, Y253H, Y253F, F311L, T315I, M351T and H396P. Specific single and combined mutations improved the predicted binding of peptides to HLA A0201. Further additional single amino-acid substitutions on mutated peptides generated high predicted binding. Selected sequences were evaluated for eliciting MHC-restricted, peptide-specific CTL responses in an in vitro model using T-cells from different HLA A0201 healthy donors after stimulation with monocyte-derived DC. CTL lines were assessed by either IFN gamma ELISPOT or by a chromium51 release assay. Mutated derived peptides from E255K, E255V, Y253F and T315I were shown to be immunogenic in vitro by the specific release of IFN gamma. Furthermore, single amino-acid substitutions in key residues improved the immunogenicity of these peptides. Finally, T cell lines generated with optimized peptides derived from E255K and T315I (the most common imatinib-induced mutations) were able to kill peptide pulsed targets; more important, these T cell lines cross-reacted with the mutated peptides (a heteroclitic response). In conclusion, Imatinib-selected mutations of peptides from bcr-abl kinase have a better predicted binding than native kinase sequences. Selected peptides from the most common mutations are immunogenic in vitro in a HLA A0201 restricted manner. T cells generated with optimized peptides from these mutations are able to elicit heteroclitic peptide-specific cytotoxicity.