Outsmarting a Mastermind

Abstract

Moellering et al. recently reported that a cell-permeable “stapled” synthetic peptide of the Notch coactivator Mastermind is a potent dominant-negative inhibitor of oncogenic Notch signaling in T cell acute lymphoblastic leukemia. This new class of inhibitor may find broad utility in blocking protein-protein interactions that underlie many human diseases.

The Notch signaling pathway is transiently induced during development to control cell proliferation and differentiation (Aster et al., 2008). In this pathway, binding of extracellular ligands to the Notch1 receptor causes the proteolytic cleavage and release of the Notch1 intracellular domain (NICD). The short-lived NICD protein translocates to the nucleus, where it binds to Suppressor of Hairless [Su(H)] CSL DNA-binding proteins to activate downstream target genes (Figure 1).

Figure 1. Diagram of MAML1 versus SAHM1 Binding.

(A) The Mastermind-like 1 (MAML1) coactivator binds to the preformed NICD:CSL DNA-binding protein complex to activate Notch target genes.

(B) A synthetic stapled 16aa MAML1 peptide (SAHM1, shown in brown) binds to the preformed NICD:CSL complex and blocks recruitment of MAML1 to the Notch enhancer complex. This prevents the expression of MAML1-dependent oncogenic Notch target genes, such as HES1 and MYC, which are necessary for growth of T cell acute lymphoblastic leukemia cells.

Loss of the normal cellular restraints on Notch receptor processing and NICD stability in the nucleus leads to an inappropriate reactivation in Notch signaling in T cell acute lymphoblastic leukemia (T-ALL) and several other cancers. Notch signaling can be inhibited by chemical inhibitors of the gamma-secretase complex (GSIs), which cleave the Notch1 receptor; however, chronic disruption of Notch signaling combined with off-target activities have limited the potential therapeutic use of these compounds and prompted a search for more selective inhibitors of the pathway.

A different approach was suggested from high-resolution structures of the Notch enhancer complex, which revealed a novel shallow groove that is created upon binding of NICD to CSL (Nam et al., 2006). This shared interface creates a cavity that recruits Mastermind-like (MAML) coactivators (Saint Just Ribeiro and Wallberg, 2009) to the enhancer complex. A 62 amino acid (aa) α-helical fragment of MAML1 (aa 13–74) precisely fits into this groove. The Bradner lab and colleagues reasoned that a hydrocarbon-stapled (stabilized) peptide mimetic might serve as a dominant-negative inhibitor of Notch signaling by binding to the NICD: CSL enhancer complex and blocking recruitment of endogenous MAML1. After probing this surface with multiple synthetic peptides, Moellering et al. (2009) selected a 16aa peptide (designated SAHM1; aa 21–36) that bound tightly (Kd, 0.12 ± 0.02 μM) to the NICD:CSL complex in vitro. Exposure of cells to 20 μM SAHM1 peptide significantly reduced transcription of HES1, MYC, and other Notch target genes, whereas various mutant peptides were relatively inactive.

Transcription profiling studies established that known Notch targets were among those most severely downregulated by SAHM1, which also included downstream target genes of two Notch-induced activators, c-Myc and E2F1. A noninvasive bioluminescent cell imaging approach was used to measure the ability of SAHM1 to impede the progression of established leukemia in a mouse model of the disease. Twice-daily intraperitoneal SAHM1 injections led to a dose-dependent regression of tumors in these mice, accompanied by reduced transcription of the HES1 and MYC genes, indicating that the Notch signaling pathway is a major target of SAHM1 in T-ALL leukemic cells in vivo.

Taken together, these new findings provide exciting evidence that stabilized, cell-permeable peptides can effectively disrupt complex protein assemblies in vivo. Unlike inhibitors that inactivate kinases or other enzymes, this strategy aims for greater specificity by interfering with a single domain of a regulatory protein. But how specific is SAHM1 inhibition? Although SAHM1 was found to affect a similar set of genes to GSIs, it may ultimately prove to be more selective because MAML1 is not universally required for Notch transactivation. Previous studies have shown that MAML1 acts preferentially at promoters containing tandem inverted repeat Su(H)/CSL DNA-binding sites (so-called “paired” CSL sequences; Figure 1), and can act synergistically with nearby basic Helix-loop-helix activators. Even within this subset of Notch target genes, MAML1 has been suggested to function in a promoter-specific manner (Cave and Caudy, 2008). Consequently, the SAHM1 peptide might target only a subset of Notch target genes, which fortuitously includes some of the most oncogenic of the Notch-induced genes. Viewed in this context, SAHM1 may function more selectively than GSIs, which should downregulate all Notch-dependent genes.

On the other hand, the SAHM1 peptide might have unintended consequences on other signaling pathways, because MAML1 does not function exclusively in Notch signaling. Emerging studies indicate that MAML1 is also an important coactivator for beta-catenin, p53, and the muscle-specific MEF2C protein (Saint Just Ribeiro and Wallberg, 2009). Mastermind proteins bind extensively to Drosophila polytene chromosomes, including the ecdysone-induced puffs, further suggesting a widespread role in gene expression. Mastermind binds tightly to the p300 histone acetyltransferase (HAT) and can recruit it to active genes. However, the order of Notch enhancer complex assembly in vivo is not firmly established, and histone acetylation has also been shown to promote or stabilize binding of Notch enhancer complexes to their target genes in vivo (Krejčí and Bray, 2007). Similarly, the binding of Mastermind to polytene chromosomes is lost in mutants of the Drosophila TRRAP/Tra1 protein, a critical component of several HAT complexes (Gause et al., 2006). Together, these data indicate that histone acetylation may play roles in both recruiting and activating Notch enhancer complexes on chromatin.

Interestingly, expression of MAML1 strongly upregulates p300 catalytic activity in vivo, and forms large nuclear foci containing p300 and acetylated histones (Saint Just Ribeiro and Wallberg, 2009). MAML1 also interacts directly with CDK8, which, like MAML1, is required for transactivation by p53 (Donner et al., 2007) and beta-catenin (Firestein and Hahn, 2009). Coexpression of MAML1 with NICD and CSL proteins in cells induces proteolytic destruction of NICD, indicating that it might also help in the disassembly of the enhancer complex following signaling. Thus MAML1 may serve as a scaffold to integrate the activities of multiple enzymatic complexes in transcription.

The various activities of the Mastermind-like coactivators illustrate the great utility of the approach used by Moellering et al. (2009). Whereas transcription would be affected broadly by targeted elimination of MAML1 or chemical inhibiton of p300 or CDK8, the more selective SAHM1 peptide may simply uncouple MAML1 from the Notch pathway while leaving its other activities intact. A central question then is whether the SAHM1 peptide might also block binding of MAML1 to p53, beta-catenin, or other activators. SAHM1 might be expected to also block growth of colon cancer cells, which depends upon Notch as well as Wnt/beta-catenin signaling (van Es et al., 2005). However, the utility of this peptide could be compromised if it is found to disrupt transcription by p53 or other tumor suppressors. Therefore to better appreciate the potential therapeutic utility of SAHM1 peptides, it will be important to define how MAML1 functions in other signaling pathways.

These new findings from Bradner and colleagues provide an important “proof of principle” that effective domain-selective disruption of regulatory protein interactions is achievable both in cells and in animals. Looking forward, this elegant new technology could be used to block other critical developmental and cancer signaling pathways, or to inhibit viral proteins that selectively engage host cell factors to promote virus replication in infectious diseases or cancer.