The focus of our work is on using mouse models of retinoblastoma to address important questions in tumor biology. Retinoblastoma is a malignant pediatric eye cancer caused by inactivating mutations in both alleles of the RB gene. Some cells in the developing human retina are particularly sensitive to tumorigenesis upon RB loss and children who inherit a mutant copy of the RB gene develop retinoblastoma at extremely high frequency.  Individuals with a germline RB mutation are also prone to the development of other tumors later in life and RB is widely mutated in many human cancers. Also, in most tumors that do not exhibit direct RB mutation, the RB pathway is deregulated through mutations in genes that control the activity of RB as well its related family members p107 and p130.

We have generated the first accurate breedable mouse model of retinoblastoma and are now using mouse models to better understand tumor origins and progression. These models make use of tissue specific inactivation of the RB gene in concert with additional mutation of either p107 or p130. These models mimic the histological progression of advanced human retinoblastoma, including late stage tumor invasion of the brain.

One of our aims is to better understand mutational events that cooperate with inactivation of RB in cancer. Using genomic approaches to identify regions of chromosomal deletion or amplification in the tumors, we have recently found that a subset of murine retinoblastomas exhibit gene amplification at the N-myc locus. N-myc amplification is also found in a subset of human retinoblastomas, suggesting that the mouse models accurately recapitulate the human disease. We would like to understand when and why specific secondary genetic alterations occur in the course of tumor progression.

We are also using these mouse models to identify the cell type in the retina that gives rise to retinoblastoma and to elucidate the normal functions of the RB family members in the developing retina.