Cancer Immune Therapy Edited by G. Stuhler and P. Walden ...

6 1 Search for Universal Tumor-Associated T Cell Epitopes
1.4
Search for Universal Tumor Antigens
The biggest practical problem of currently described tumor-associated antigens is the
lack of wide expression. Clinical studies have been limited as strategies have been
tested one malignancy at a time and, in some cases (such as the immunoglobulin
idiotypic antigen in B cell malignancies), patient by patient [10]. To circumvent these
obstacles, we proposed a method of identifying tumor antigens with universal or
near universal expression in cancer, such that CTL responses against a broad range
of tumor types could be triggered [18]. Rather than analyzing patient-derived T cells
or antibody responses to tumor, an alternative strategy can be used in which tumor
antigens and their CTL epitopes are deduced from genes known to be selectively expressed
in tumors [18]. This method, often called ªepitope deductionº or ªreverse immunologyº,
does not rely on the presence of an innate anti-tumor T cell response,
which is then dissected by molecular immunology. This is important because the T
cell response against the most common types of cancers is felt to be weak or absent.
Given the hallmark features of T cell recognition outlined above, we began a search
for antigens that met the following criteria:
. Expression by the vast majority of human cancers.
. Expression of peptide sequences that bind to MHC molecules.
. Adequate processing by tumor cells such that antigen-derived peptides are available
for binding to MHC molecules.
. Recognition by the T cell repertoire in an MHC-restricted fashion, permitting the
expansion of naive CTL precursors bearing specific T cell receptors.
1.5
Epitope Deduction
The method of epitope deduction postulates that candidate T cell peptide epitopes
can be chosen based on predicted binding affinities of peptide to MHC and then
scrutinized for immunogenicity based on the capacity of experimentally generated
peptide-specific T lymphocytes to kill tumors in vitro or in vivo. Table 1.1 outlines several
algorithms publicly available for predicting the MHC class I-binding affinities of
peptides. Candidate peptides are screened systematically against the criteria described
above. Any gene product can be subjected to this analysis without the need
to dissect anti-tumor immune responses from cancer patients. Such dissection is the
cornerstone of the classical discovery approach, but becomes limited as efforts move
beyond melanoma to the majority of common cancers in which patient immunoreactivity
is weak. To be sure, the method is not restricted to the field of tumor antigens
and has been richly exploited elsewhere, particularly in the search for immunogenic
targets of infectious agents.
Epitope deduction has been used to characterize numerous epitopes from candidate
tumor-associated antigens (Tab. 1.2) [19]. Most commonly, MHC class I-restricted