Learning about retinoblastoma from mouse models that missed

Abstract

Retinoblastoma has been an attractive cancer for mouse modeling, as it is thought to develop in response to the inactivation of a single gene (RB) and thus provides an extraordinary opportunity to dissect mechanisms of tumor initiation. However, difficulties have repeatedly arisen in efforts to model this malignancy. For example, mice with inactivating Rb mutations fail to develop retinoblastoma, and instead develop pituitary and thyroid cancers. Moreover, mice with combined mutations in Rb and Rb-related genes develop retinal tumors that differ from human retinoblastomas in important aspects, including the mutations necessary to form the tumors, the retinal cell phenotypes of the tumors, and potentially the tumor cell of origin. Despite these discrepancies, mouse models may provide insight into human retinoblastoma by revealing features that contribute to retinal tumors in both species. Nevertheless, the mouse's inexact representation of retinoblastoma challenges us to develop more accurate models and to decipher how human-specific circuitry sensitizes to the loss of Rb and enables retinoblastoma initiation. © 2013 Springer Science+Business Media New York.