Efficient Inhibition of Notch3 and Notch4 Family Members In Vivo by a Dominant Negative Mutant of Mastermind.

The four mammalian Notch receptors play multiple roles in the hematolymphoid system. Notch1 is critical for the generation of hematopoietic stem cells (HSCs) during embryonic development and for commitment to the T cell lineage, while Notch2 regulates marginal zone B cell development. In addition, important roles for Notch are starting to be identified in peripheral T cells, in which all four Notch family members are upregulated upon T cell activation. Canonical Notch signaling involves proteolytic cleavage of Notch receptors and translocation of intracellular Notch (ICN) to the nucleus where it interacts with the transcription factor CSL/RBP-J and recruits Mastermind-like proteins (MAMLs) for transcriptional activation. By inhibiting Notch1 and Notch2-dependent developmental decisions with a dominant negative mutant of MAML1 (DNMAML1), we have previously demonstrated that MAMLs critically regulate Notch signaling in vivo (Maillard et al., Blood 2004). However, it is unclear if Notch3 and Notch4 are equally dependent on the activity of MAMLs in vivo, and if DNMAML1 can thus be used as a true pan-Notch inhibitor.

To address this question, we used a modified fetal thymic organ culture (FTOC) system and Notch-mediated T lineage commitment as a readout for Notch function. Fetal liver cells (FLCs) were transduced with retroviruses expressing ICN1, ICN3 or ICN4 and a retrovirus expressing DNMAML1. Irradiated fetal thymic lobes were reconstituted with transduced FLCs and cultured in the presence of gamma secretase inhibitors to inhibit the activation of endogenous Notch receptors. In this manner the effect of a specific ICN could be evaluated in the absence of endogenous Notch signals. Gamma secretase inhibitors resulted in a profound block in T cell development and in the generation of intrathymic B cells, consistent with Notch1 inhibition. ICN1, ICN3 and ICN4 were all able to rescue commitment to the T cell lineage and to block B cell development, indicating that each ICN can deliver appropriate signals for T lineage commitment in the absence of endogenous Notch1 signaling. In cells doubly transduced with specific ICNs and DNMAML1, the rescuing effect of ICN1, ICN3 or ICN4 was blocked, leading to the generation of B cells. In this competitive situation, DNMAML1 only partially inhibited ICN1, the most potent of all four Notch family members, but it was able to completely inhibit ICN3 and ICN4. We conclude that DNMAML1 can efficiently inhibit Notch3 and Notch4 in vivo. The results validate the use of DNMAML1 as a bona fide pan-Notch inhibitor in multiple settings, including studying the role of Notch in HSCs and peripheral T cells.