Current views on the role of Notch signaling and the pathogenesis of human leukemia

The Notch gene was first described following the observation of Notches on the wings of fruit flies (Drosophila melanogaster) caused by partial loss of function of the Notch gene. Notch signaling is involved in many biological processes, ranging from embryonic development to cell proliferation and survival. It has been demonstrated that the Notch signaling pathway is involved in vascular formation and morphogenesis during vascular development. Notch1, Notch2 and Notch4 and its ligands (Jagged1, Jagged2, Dll1 and Dll4) are expressed in vascular endothelium, whereas Notch3 is expressed in vascular smooth muscle cells. Mutations in Notch3 are associated with CADASIL syndrome (cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy), the human degenerative vascular disease.

The human Notch family includes four receptors and five ligands [1, 2]. All four Notch receptors are synthesized as a single transmembrane polypeptide in the endoplasmic reticulum and transported to the cell surface trough the trans-Golgi network. Notch receptors are expressed as heterodimeric proteins with extracellular, transmembrane and intracellular domains (Figure 1). When a ligand of the Delta/Serrate/LAG-2 family (located on the surface of neighboring cells) binds to the extracellular domain of the Notch receptor, it triggers proteolytic cleavage by a metalloprotease (a disintegrin and metalloprotease (ADAM)). ADAM cleavage produces a substrate for a second cleavage by the presenilin-containing γ-secretase complex, releasing the Notch intracellular domain (NICD) [2, 3] (Figure 2). NCID corresponds to the activated form of Notch, which translocates to the nucleus and forms complexes with specific DNA-binding proteins (CBF1/Suppressor of Hairless/LAG-1 and Mastermind/SEL-8) and transcriptionally activates target genes [4] (Figure 2). In the absence of receptor activation and NICD, CBF1 acts as a transcriptional repressor through interactions with the corepressors SMRT (silencing mediator of retinoid and thyroid receptors), KyoT2, CIR (CBF1-interacting corepressor) and SHARP (SMRT/HDAC1 (histone deacetylase 1)-associated repressor protein) [5]. In addition to canonical intracellular signaling pathways, there are other types of noncanonical Notch signaling (Figure 3). The first one involves Notch ligation and translocation of activation signals independent of CBF1 (NICD-dependent), the second involves activation of Notch target genes that are independent of γ-secretase cleavage (NICD- and CBF1-independent) and the third involves CBF1-dependent gene activation without receptor cleavage and NICD release. Termination of Notch signaling can occur at or downstream of the Notch receptor. The Notch receptor can be degraded through the lysosomes by the ubiquitin ligase Itch/AIP4 [6] or another ubiquitin ligase, Nedd4 [7], which act together with Numb [8] and Itch/AIP4 to stimulate endocytosis and lysosomal degradation of the Notch receptor [9]. Finally, NICD1 phosphorylation by GSK3 regulates its interaction with the E3 ubiquitin ligase CDC4/FBW7, thereby controlling NICD1 ubiquitination and proteasome-mediated degradation [10]. This multifaceted control of Notch expression underscores its critical functions in cellular homeostasis.

The role of the microenvironment in the activation of Notch in leukemia is increasingly recognized. Recently, cis-inhibition of Notch signaling by the DLL1 ligand has been described in Drosophila and mice [11]. These investigations have suggested that while expression of ligands on neighboring cells stimulates Notch activation, expression on the same cell as the Notch receptor may have an inhibitory effect [11]. Along these lines, activation of Notch signaling in B-cell malignancies might result from interactions between tumor cells as well as between the tumor cell and the microenvironment. There is evidence suggesting the importance of Notch signaling in the cross-talk between multiple myeloma (MM) cells and their environment. Bone marrow stromal cells express both Notch ligands, Jagged and δ, and are able to activate Notch signaling in MM cells [12, 13].

Notch1 was discovered in humans through a t(7;9)(q34;q34.3) chromosomal translocation observed in some patients with T-cell acute lymphoblastic leukemia (T-ALL) [14, 15]. However, a direct role of Notch activation in T-ALL remained obscure, since only 1% to 3% of patients with T-ALL were found to carry this translocation. It was only after the discovery of a high rate of activating mutations that it became clear that Notch1 expression is linked to the development of T-ALL [16]. Activating mutations identified in the T-ALL cluster at the heterodimerization domain (HD) and the proline, glutamine, serine and threonine (PEST) domain led to ligand-independent cleavage of the Notch receptor and a reduced degradation of NICD1, respectively. Recently, activating mutations in Notch were identified in more than 30% of human T-lymphotropic virus type I (HTLV-I)-associated adult T-cell leukemia (ATL) patients, suggesting an important role for Notch signaling in HTLV-I-associated ATL [17]. Activating mutations found in ATL patients are different from those previously reported in patients with T-ALL and mostly involve single-substitution mutations in the PEST domain that do not create an early stop codon, but rather lead to reduced CDC4/Fbw7-mediated degradation and stabilization of NICD1 [17].

Additional mechanisms have also been reported to lead to increased Notch expression in cancer cells. Mutations and internal duplication insertions in exon 28 of NICD [18], as well as mutations in CDC4/Fbw7 [10, 19], have been reported, but they seem to occur at a very low frequency. The Wnt/β-catenin and Notch1 signaling pathways play an important role in a variety of biological processes, including cell proliferation and survival. Studies have shown that β-catenin can regulate the level and transcriptional activity of Notch1 [20]. β-catenin can prevent NICD degradation, possibly by competing with CDC4/Fbw7-dependent degradation [20]. Moreover, β-catenin increases the transcriptional activity of NICD, and the effects of β-catenin on Notch1 are noticeably reduced by overexpression of the lymphocyte enhancer-binding factor 1, LEF1.

The molecular pathogenesis of Notch has recently been reviewed [24]. Targeting Notch receptor cleavage through γ-secretase inhibitors (GSIs) is an attractive approach, as GSI treatment inhibits proliferation of T-ALL and ATL tumor cells in vitro and in vivo [42, 43]. The use of GSIs, however, poses several challenges. Current GSIs have been shown to have significant intestinal toxicity in patients because of the dual inhibition of Notch1 and Notch2. In addition, high levels of IL-6 and IL-8 have been reported to abrogate or significantly reduce the efficacy of the GSI (RO4929097) [44]. GSIs may not be useful in patients with a Notch1 mutation in the HD, as this results in weakened association or complete dissociation of the receptor subunits and thus leads to ligand-independent activation. Finally, recent studies have also shown that T-ALL patients with loss of PTEN are resistant to GSI effects [45]. Together these results highlight the need for new targeted therapies.

Recently, the use of therapeutic antibodies that selectively block Notch1 receptor signaling has been reported to inhibit tumor growth in mouse models [46]. Rather than targeting receptor processing, alternative strategies may also focus on blocking NICD functions and transcriptional activities. Along these lines, the use of a stapled peptide to inhibit the Notch transcription factor complex has been reported to result in Notch-specific antiproliferative effects in cultured cells and in a mouse model of Notch1-driven T-ALL [47].