Notch and Leukemia: Lessons from Mouse Models.

Mammalian Notch1 was first identified by Dr. Jeffrey Sklar (Yale University) and coworkers as a chromosomal translocation partner in a rare subset of human T-ALLs carrying a t(7;9) translocation. Notch moved from a bit player to center stage when Drs. Jon Aster (Brigham & Women’s Hospital), A. Thomas Look (Dana-Farber Cancer Institute), and colleagues identified Notch1 activating mutations in the majority of human T-ALLs. These mutations occur in the Notch extracellular domain in the heterodimerization domain, allowing for ligand-independent signaling (HD mutations), or in the C-terminal PEST domain (PEST mutations), which increases Notch protein stability. In retrospect, the frequent association of Notch mutations with human T-ALL may have been predicted given that Notch is a potent T cell oncogene in mice. Notch-induced T-ALL shares many features with human T-ALL, suggesting that oncogenic Notch signaling may have similar functions in T cell tumorigenesis in mice and people. Notch functions as a T cell oncogene in both retroviral and transgenic models, where it is introduced as the primary oncogenic event, as well as in models where it is likely acquired as a collaborating event. Notch1 PEST mutations are frequent in the latter. Acquired Notch mutations in both humans and mouse T-ALL have only been identified in Notch1, whereas ectopic expression of the intracellular domain of Notch1, 2, and 3 all give rise to T cell neoplasms in mice. These differences may be related to variation in signal strength between these different Notch family members. Although intracellular Notch1 (Notch1-IC) is a potent inducer of T-ALL in retroviral and transgenic mouse models, this form of Notch is rarely present in human T-ALL. We recently assayed more commonly mutated Notch1 alleles in the retroviral model and found that HD, PEST, and HD + PEST mutations were weak inducers of T-ALL. When assayed on a background that predisposes to T-ALL, such as oncogenic ras, these weakly leukemogenic alleles shortened disease latency and gave rise to cell lines that were Notch-dependent. These data suggest that the common Notch mutations found in human T-ALL are weak tumor initiators but still give rise to tumors that are “addicted” to Notch. This strengthens the rationale for inhibiting Notch in T-ALL. The mouse also appears to be a useful model for testing Notch-targeted therapeutics, as the murine tumors are frequently susceptible to Notch inhibition, and mice exhibit some of the same toxicities as people when exposed to continuous Notch inhibition. A challenge for the field is to identify a safe therapeutic index for Notch inhibition. As a transcriptional regulator, an area of active investigation is identifying oncogenic Notch1 transcriptional targets. Several targets have been identified in both mouse and human, such as c-myc and regulators of T cell identify. Ongoing studies will likely reveal additional transcriptional targets, which is important for understanding the function of Notch in the pathogenesis of T-ALL.