Funded Grants

Identification and characterization of glioma-CTL epitopes

Malignant gliomas are intractable despite advances in modern microsurgery, radio-therapy and chemotherapy. These tumors grow aggressively and infiltrate surrounding normal brain. Even after microsurgical resection, they recur from these infiltrating elements of residual tumors.

Immunotherapy for cancers has been developed on the notion that a patient’s immune system may be able to distinguish cancer cells as “abnormal” tissues and attack microscopic small elements of cancers. Indeed, cancer cells produce many proteins that are not usually produced by normal tissues. Investigators for other types of cancers have developed vaccine strategies in which they induce specific immune responses against proteins called tumor-associated antigens. Early reports from these studies have demonstrated safety, feasibility and some level of efficacy of these approaches.

The central nervous system (CNS) and CNS tumors have long been regarded as immunologically privileged due to their distinctive anatomical characteristics such as the absence of conventional draining lymphatics and the presence of the blood-brain-barrier. Nonetheless, the immunologic privilege of the brain is not absolute, and our understanding of neuro-immunology has actually been extensively revised in the last several years. For instance, in experimental allergic encephalomyelitis (EAE), which is an animal disease model simulating multiple sclerosis (MS) in humans, peripheral immunization with brain specific proteins elicits antigen specific, T-lymphocyte-mediated auto-immune brain disease. This clearly indicates that specific immunoreactivity in the CNS can result from the appropriate presentation of target antigen and the efficient induction of a systemic effector response.

This observation is further strengthened by recent reports that effective anti-CNS tumor immune responses can be generated through the use of cytokine (a type of immuno-stimulatory substances) -gene modified tumor cell vaccines. We have developed clinical trials of vaccine therapies using patients’ own glioma cells that are genetically modified to produce a cytokine, interleukin-4 (IL-4). We have treated two patients with this strategy, and have observed some level of clinical response in each patient.

Although these whole glioma cell vaccines appear to be promising, there is a theoretical concern for inducing immune responses against normal brain compartments because glioma cells and normal brain cells express many common proteins. One way to limit this attack on healthy brain tissues is to prime immunity against tumor specific antigens. To date, few human gliomaspecific antigens have been identified. However, we recently reported identification of one of the first glioma-derived T-cell epitopes, which are small pieces of proteins (peptides) capable of inducing cytotoxic T cell-immune responses specifically against brain tumors. We have incorporated cutting-edge bioinformatics and immunobiology to establish an efficient system to identify these immunogenic molecules. It is thought to be very important to expand the number of available CTL epitopes for therapy of human gliomas because immunotherapy against a single antigen epitope is likely to result in outgrowth of “antigen-loss variant” tumors because the antigen-profile of human glioma cells are different each other even in the same individual. In this proposal, we will extend our approach in order to identify more glioma antigen-epitopes. These peptides will be identified within the proteins that are strongly expressed in brain tumors, and ones that have altered amino-acid sequences in brain tumor cells. We already have one candidate antigen called EphA2. We have recently shown it to be expressed in many human malignant gliomas. Therefore, identification of CTL epitopes in EphA2 may allow us to treat malignant gliomas with specific vaccines. In order to identify glioma associated antigen-derived epitopes, we will use computer programs that will help us to predict epitopes that can bind to major histocompatibility complex (MHC) molecules that present epitopes to T-cells. Synthetic peptides for these predicted sequences will be then loaded on donor-derived dendritic cells (DCs) that are sometimes called “quarterbacks” in the immune system because of their important roles in presenting antigens to T-cells and activating them. We will co-culture the peptide-loaded DCs and T-cells in order to find whether these peptides are able to elicit specific T-cell responses. If we find any response, these T-cells will be tested for their killing ability against human glioma cells that express the antigens.

Some immunologists have also reported that introduction of alterations in amino-acid sequences of the antigen peptides may even enhance the efficiency of immune-priming ability; and T-cell responses induced by such altered peptides may still recognize the original form of antigen-peptides on the target cells. In order to develop even more effective vaccine strategies, we will also investigate whether peptides that have artificial alterations in amino-acid sequences can elicit even more efficient immune responses against human glioma-derived antigens.

We believe it is also important to demonstrate that anti-glioma T-cell responses can be elicited with these peptide-epitopes in glioma patient-derived T-cells in vitro because we eventually plan to develop peptide-based vaccine protocols for these patients. We will therefore test in vitro whether these peptides can elicit immuno-reactivity in T-cells obtained from glioma patients. In addition, we plan to use these peptides to measure anti-glioma immunity elicited in patients who participated in our IL-4 glioma vaccine trials. We expect to find that specific T-cell immuno-reactivity may increase following glioma-cell vaccines, and such responses may correlate with clinical responses.

Our ultimate goal of this proposal, although not directly a part of the proposal, is to design and conduct peptide-based vaccines against malignant gliomas using the epitope-peptides identified in the proposed study. We envision the use of DCs loaded with the peptide-epitopes as vaccines. Peptide-based vaccines are expected to have the following advantages over our ongoing whole glioma cell vaccines: 1) because of its specificity, there will be less concern of inducing auto-immune reaction to normal brain; 2) in contrast to whole cell vaccines that require resection of significant amount of tumor tissue and cumbersome laboratory work for cellprocessing, large amounts of synthetic peptides will be available in a timely fashion; and 3) if we find modified peptides that are more efficient in inducing responses, vaccines with these epitopes may be more efficient than whole natural glioma cells. We are confident in proposing these projects because of the extensive experience in this field at the University of Pittsburgh.