Engineering high-affinity T cell receptors for treatment of brain cancerA recent
Wall Street Journal article described the painful dilemma that faces brain tumor patients and parents of children with brain tumors: should they undergo the standard therapy of radiation and chemotherapy, when the survival benefit is marginal, the side effects are debilitating, and the long-term consequence of the therapy is often dramatic loss of cognitive function. For children the damage is especially problematic as it continues to worsen the longer the child lives, with a progressive loss of approximately two IQ points per year, eventually resulting in mental retardation if the child lives long enough. The alternative being offered parents in 2005 is to reduce the radiation intensity below the optimal level for damaging the tumor, perhaps risking recurrence, with no assurance that the brain damaging effects will be avoided.
There is no disagreement that medical research is needed to develop better alternatives. The McDonnell Foundation has already recognized the great potential of immunotherapies, and T cell mediated immunotherapy in particular, as an alternative strategy. T cells are white blood cells that can be exquisitely selective in recognizing tumor cells and killing them, while sparing neighboring normal cells. They are especially suited for brain tumor therapy because, unlike antibodies, they are able to actively cross the blood brain barrier. In recent years the McDonnell Foundation has funded several projects to develop T cell-based brain tumor therapies, one designed to enhance the role of glial cells in activating T cells, one to study ways that tumors suppress T cells, and one to identify glioma-specific antigens that could be recognized by T cells. Our project seeks to overcome a major hurdle that will be encountered in these approaches: that the T cells best suited to attack tumors have been eliminated in a patient because of a process called "tolerance".
What is tolerance? T cells recognize tumor cells or virus-infected cells by interacting with ("binding to") fragments of proteins ("peptides") that are displayed on the surface of the tumor cells. T cells accomplish this through surface receptors, called T cell receptors, that bind to peptides. T cell receptors vary in how tightly they bind to peptides, and in many cases the ones that bind the best have been eliminated, by a process called tolerance, because they react with one's own tissue (called self). Unfortunately most of the known targets on tumors are these "self" antigens. This has been shown in humans, where tumor specific T cells on average have a ten-fold lower binding strength for their peptide than T cells that are specific for a virus. In cancer models, we and others have shown that animals with naturally arising tumors lack the strongest binding T cells. The remaining, weaker binding T cells, can be activated and they are able to restrain tumor growth for a while, but eventually the rapidly growing tumor wins out. These animal studies are consistent with the encouraging but ultimately disappointing results of current tumor vaccines in clinical trials.
We hypothesize that protein engineering techniques can be applied to this problem in efforts to overcome tolerance. We have developed technologies to engineer strongly binding T cell receptors (also called high-affinity T cell receptors). In some cases the engineered T cell receptor has 1,000-fold stronger binding to the peptide than the naturally occurring T cell receptor. The technique is called "yeast display," because we take the gene for a T cell receptor, make millions of variations in the T cell receptor by mutating the gene, then insert these genes into yeast cells. Individual yeast cells display copies of a particular T cell receptor on their surface, and then using automated cell sorting we can select the few mutants that bind more strongly to the peptide. We take this process through several rounds, each time selecting the best binding T cell receptor. We have already shown that we can produce T cell receptors with very strong binding to a peptide that is on the surface of a tumor used in mouse models of brain tumors. The studies that we are proposing to the McDonnell Foundation would determine whether T cells that express these strongly binding T cell receptors can effectively target and eliminate brain tumors in mice. Among the questions to be addressed are: Can we provide an animal with better T cells for attacking brain tumors? Will such T cells result in eradication of brain tumors? What are the binding strengths of the T cell receptors that provide the most optimal benefit?
We also propose to initiate the engineering of human T cell receptors that may some day be used directly for the treatment of brain cancers. There already exist peptides on human glioblastomas that have been identified as potential targets for T cell therapies. Yet other potential target peptides are being discovered by various research laboratories. In our proposal, we will isolate human T cells that recognize one of these peptides. The genes that encode the T cell receptor will be cloned into our "yeast display" system and higher affinity T cell receptors will be engineered. These T cell receptors will be re-introduced back into T cells and assayed for potency against various human glioblastoma cell lines. The clinical application of this strategy would be to (1) isolate T cells from a patient and expand these cells in culture (outside the patient, in an incubator) (2) transfer the stronger binding engineered T cell receptors into the T cells; (3) reintroduce large numbers of the engineered T cells back into the patient. The goal of the study is to use these T cells as an effective treatment for patients, with minimal side effects because only the tumors are targeted by the T cell receptors.
