CYCLIN B1 AS A TUMOR ANTIGEN AND A CANCER VACCINE

Olivera J. Finn, Ph.D.
Department of Immunology, University of Pittsburgh School of Medicine
W1040 Biomedical Science Tower
Pittsburgh, PA 15261-0001
ojfinn@pitt.edu

Molecular characterization and isolation of shared tumor antigens that are immunogenic in humans has been an important goal of tumor immunologists in the last two decades.  A large number of such antigens has been described and some of them could be considered candidates for inclusion in vaccines (1).  Uncontrolled cell division is an indispensable event in tumor progression and numerous molecules involved in this process have been the subjects of intense investigation in tumor biology.   Cyclins, molecules that control different checkpoints in the cell cycle, have been shown to be abnormally overexpressed in various human cancers. 

Cyclin B1 is important in the cell cycle progression from G2 to M-phase and it is also associated with cell proliferation.   Correlation has been found between the extent of cyclin B1 overexpression and clinical outcome for patients with cancer of the lung, esophagus and tongue.  We have a particular interest in cyclin B1 since we recently showed that cyclin B1 can serve as a tumor specific antigen and that tumor-specific CTL derived from patients with breast and head and neck cancer recognize HLA-A2 presented peptides derived from cyclin B1 protein (2).   

Cyclin B1 is aberrantly expressed in the cytoplasm of tumor cells where it is found in large amounts that are never encountered in normal proliferating cells.  Tumors that aberrantly express cyclin B1 belong to many different types, solid tumors as well as lymphomas, and they all have functionally inactive p53.  There is a direct correlation between inactivation of p53 function and overexpression of cyclin B1  (3), even though a direct interaction of these two molecules at either transcriptional or translational level has not been shown.  Inactivation of p53 with consequent overexpression of cyclin B1 are very early changes in the transformation process and are characteristic of premalignant lesions as well as fully developed tumors.  This suggests the possibility of using a cyclin B1-based vaccine in patients who have been diagnosed with preneoplastic changes, in hope of eliminating these lesions and preventing further transformation to cancer.

Our data on the correlation of cyclin B1 overexpression and p53 inactivation suggested that a mouse with a homozygous deletion in the p53 gene that spontaneously develops cancer may be a good animal model for immunoprevention of cyclin B1 over expressing tumors. Theoretically, every cell in the body of a p53-/- mouse could be expected to overexpress cyclin B1.  What we found instead was that normal cells had undetectable (normal) levels of cyclin B1, while the arising tumors, lymphomas and sarcomas showed very high levels of expression.  We have designed several cyclin B1 based vaccines and are in the process of immunizing wild type and p53-/- mice to elicit anti-cyclin B1 immunity.  We are testing the ability of anti-cyclin B1 immune responses to prevent or lower tumor incidence in p53-/- mice and /or decrease the metastatic potential of the tumors that do arise.  Our studies show that mice are not tolerant to cyclin B1 and that T-cell dependent immunity can readily be elicited.  A vaccine based on a cyclin B1 DNA is able to immunize mice and protect them from cyclin B1 positive tumor challenge without inducing autoimmunity. 

ACKNOWLEDGEMENT.
This work is supported by a from the National Institutes of Health 5P50 CA90440.

REFERENCES 

1. Finn, O.J. (2003) Nature Immunol. Rev. 3: 630-641
2. Kao, H., Marto J. A., Hoffmann, T. K., Shabanowitz, J., Finkelstein,S. D., Whiteside, T. L., Hunt, D. F., and Finn O. J. (2001) J. Exp. Med., 194: 1313-1323
3. Yu, M., Zhan, Q., and Finn, O. J. (2001) Mol. Imunol. 38: 981-987