Oral Abstracts

Immunization of HLA-A2+ Melanoma Patients in the Adjuvant Setting with a Mutated gp100 Peptide (209-2M) Induces Specific T Cell Immune Responses

Smith II JW, Walker EB, Doran T, Wisner K, Miller W, Floyd K, Haley D, Vetto J, Wood W, Urba WJ
Franz Cancer Research Center, Chiles Research Institute, OHSU, Portland, OR

Vaccination of metastatic melanoma patients with gp100 (209-2M) peptide produced immune responses in nearly all patients but no clinical responses unless the vaccine was followed by high-dose IL-2. We reasoned that melanoma patients with a low tumor burden might have a greater immune response because of their good performance status and their ability to receive six months of vaccinations. Thirty patients with cutaneous melanomas at least 1 mm thick or metastatic to regional lymph nodes, were randomized to six months of vaccinations on an every two week (13 vaccines) or every three week (9 vaccines) schedule. Montanide ISA 51 was mixed with the gp 100 peptide. Peripheral blood mononuclear cells (PBMCs) were obtained pre and post study by leukapheresis and monthly by peripheral blood draws. PBMCs were frozen and then analyzed freshly thawed or after in vitro culture using tetramers, intracellular cytokine staining and ELISPOT assays.

Patient characteristics: median age, 51 years, range, 23-85 years, male/female 17/13. Seventeen patients (57%) had positive lymph nodes, 14 patients (47%) received interferon alfa (IFN) during their vaccines. Patient characteristics were well balanced in both arms. Of the 25 patients analyzed to date, 22 had significant increases in the number of gp100 tetramer positive CD8 T cells. Overall, the median baseline percent tetramer gp100 positive CD8 T cells was 0.08% and increased four fold to 0.32% after vaccination, the average increased 19 fold from 0.06 to 1.12%. Three patients had a striking response to vaccination with 3%, 5%, and 11% respectively, tetramer positive CD8 T cells after vaccination. The group of patients that received vaccinations every two weeks had a higher median of gp 100 tetramer positive CD8 T cells, 0.51% vs 0.26%. Patients who received IFN during vaccination had no impairment of the ability to generate a specific T cell response.

Transduction of human dendritic cells with a melanoma polyepitope simultaneously primes multiple T-cell responses but ultimately results in immunodominance

Jackson AM, Smith SG, Porte J, Patel PM
Applied Immunology Group, Imperial Cancer Research Fund, Leeds, UK

BACKGROUND: Single epitope immunistaion is limited in efficacy due to escape by mutation and modulation. Furthermore, many tumour antigens have unknown or transforming activities. Nucleic acid vaccines overcome some of the difficulties with peptide vaccines such as peptide-clearance and conformational differences between exogenous peptide and endogenously processed antigens. The polyepitope approach to immunisation represents a means to overcome these problems yet elicit cytotoxic T-cell (CTL) responses against a wide range of target antigens and epitopes.
OBJECTIVE: To examine the CTL responses to a melanoma polyepitope using human T-cells and dendritic cells (DC) from naïve donors.
METHODS: A melanoma polyepitope was generated which encoded 6 HLA-A2-restricted epitopes from human tumour antigens. Human dendritic cells were transduced in vitro with melanoma polyepitope cDNA and autologous T-cells from naïve donors were stimulated. CTL were restimulated either with polyepitope or with multiple peptides before assay for specific activity. For comparison A2Kb transgenic mice were immunised with melanoma polyepitope cDNA and CTL responses determined.
RESULTS: Polyepitope transduced human DC simultaneously primed up to 6 HLA-A2 CTL responses. However, repeated exposure of DC-primed CTL to polyepitope resulted in immunodominance to a sole epitope (MART127-35). This did not correlate with precursor frequency. Furthermore, DC-polyepitope only restimulated MART-reactive CTL lines but not CTL lines to gp100, tyrosinase or MAGE3. In contrast to the human system, repeated immunisation of A2Kb mice with polyepitope DNA resulted in multiple CTL responses.
DISCUSSION: Polyepitopes are a promising approach to cancer vaccines. However, repeated stimulation of human T-cells with a melanoma polyepitope results in immunodominance. A number of possible explanations exist however, our data demonstrates that this immunodominance may relate to preferential release of MART1 epitope by the antigen processing machinery. Studies of this kind using human T-cells may offer more informative pre-clinical information than some animal models. This may reflect differences in T-cell repertoire and in the antigen-processing machinery.

Comparative Analysis of Necrotic and Apoptotic Tumor Cells as a Source of Antigen(s) in Dendritic Cell-Based Immunization

Yoshihito Kotera and James J. Mulé
Department of Surgery, University of Michigan Medical Center, Ann Arbor, Michigan

Introduction: There is considerable controversy as to whether necrotic (lysate) or apoptotic tumor cells (Apo) serve as the superior source of multiple tumor-associated antigens (TAAs) to pulse dendritic cells (DCs) for immunotherapeutic applications. Here, we used highly enriched apoptotic cells vs. lysate of B16 melanoma to examine whether important distinctions exist between these two forms of TAAs.

Methods: Apo melanoma cells generated by UVB light were highly enriched by FACS sorting. Lysate was generated by four cycles of rapid freeze/thaw. Levels of apoptotic and necrotic cells were determined by the Annexin V-PI FACS assay and their content of heat shock proteins (HSPs) was examined by immunoblotting. DCs and killed tumor cells were co-cultured and analyzed for uptake by two-color FACS analysis, electron microscopy, and confocal microscopy. Also, IL-12 production by and changes in phenotypic markers as a measure of maturation of the tumor-loaded DCs was determined by ELISA and FACS, respectively. Apo- and lysate-pulsed DCs were compared for immune priming and antitumor therapeutic efficacy in vivo. For immune priming, naive B6 mice were immunized s.c. twice at 7 day intervals and were then challenged with a lethal dose of viable B16 melanoma cells. For therapy, B6 mice received a lethal dose of viable melanoma cells 7 days before immunization. These mice were then immunized s.c. with either Apo- or lysate-pulsed DCs twice, 7 days apart. Tumor growth was measured.

Results: After FACS sorting, UVB-irradiated B16 melanoma cells were about 90% apoptotic. The lysate contained essentially 100% necrotic cells. DCs could efficiently engulf both forms of killed tumor cells within 18 hr. For IL-12 production, DCs that had engulfed sorted Apo produced 528+/- 20 pg/ml vs. 255 +/- 65 pg/ml when exposed to lysate. DCs pulsed with either Apo or lysate could equally promote immune priming and inhibit tumor growth in vivo.
Conclusions: Although some differences existed in HSP content and IL-12 production with the two forms of TAAs, their respective capacities to mature DCs phenotypically and to elicit both effective immune priming and antitumor therapy in vivo when presented by DCs were equivalent.

Pharmacogenomic Applied to Drug Development

Joseph A. Rininger, Ph.D., Senior Research Scientist, Pharmacogenomics, CuraGen Corporation

The industrialization of tools for rapid DNA sequencing and global gene expression analysis has been a major technological scientific advancement that is being applied from target discovery to surveying of clinical populations (Pharmacogenomics). The ultimate goal is to improve the efficiency of drug development by prioritization of lead compounds to the clinic with a decrease in attrition due to lack of efficacy and unforeseen toxicity. This presentation will focus on specific case-study examples providing experimental support to the utility of genomics to from target discovery, lead prioritization and association of genetic polymorphisms with disease.

Directed Molecular Evolution of Vaccines and Immunotherapeutics

Juha Punnonen Maxygen, Inc.

Directed molecular evolution by DNA shuffling followed by screening is a technology that allows for rapid evolution of genes and genomes in vitro. DNA shuffling of natural diversity is used to create large libraries of chimeric sequences of genes of interest. The libraries are subjected to a multi-tiered high throughput screening process using cell-based robotic assays, FACS selection and/or in vivo assays. If desired, a pool of the best sequences from one round of screening is shuffled and rescreened to obtain further optimization and the best hits are subjected to detailed biological characterization.

We have used DNA shuffling and screening to generate libraries of vaccine antigens, cytokines and viruses. We have generated libraries of chimeric antigens that provide crossreactive immune responses in contrast to the parental antigens. In addition, we have used mammalian cDNAs encoding cytokine genes as the starting sequences to generate cytokine libraries and several chimeras with improved expression and specific activity have been identified. We have also evolved immunomodulatory molecules with altered ligand binding specificities, and hence, with altered biological properties. Moreover, DNA shuffling and screening of MLV viruses resulted in variants with improved resistance to ultracentrifugation. Directed molecular evolution by DNA shuffling and screening is a powerful platform technology to generate novel and improved vaccines and immunotherapeutics.

Approaches for Identifying and Accelerating the Prioritization of all the Potential Target Genes and Therapeutic Proteins in the Genome

Benjamin G. Cocks, Ph.D.
Senior Director, Immunobiology
Incyte Genomics Inc.

The LifeSeq® Gold database is the best available resource for the preliminary identification of genes associated with disease processes. The selection of potential antibody targets, drug targets, diagnostics, and therapeutic proteins for further investigation can be facilitated by additional complementary knowledge such as protein structure analysis and gene expression network information obtained by microarray analysis. We will describe the process by which all the potential targets can be rapidly characterized at multiple levels, stratified, and functionally assessed, by incorporating specialized and integrated technological and analysis capabilities.