Notch Inhibitors as a New Tool in the War on Cancer: A Pathway to Watch

Abstract

Notch was first recognized as an important developmental pathway in Drosophila in the first half of the 20^th century. Many decades later, this pathway has been found to play central roles in humans in stem cell maintenance, cell fate decisions, and in cancer as well. Notch family members are being revealed as oncogenes in an ever-increasing number of cancers. Though significant progress has been made in dissecting the complex workings of this signaling pathway, there are very limited options available for Notch inhibitors. However, the pioneering class of Notch inhibitors is already in clinical trials for two cancer types. This review will address the current state-of-the-art, agents in the pipeline, and potential strategies for future Notch inhibitors. Successful development of Notch inhibitors in the clinic holds great promise as a new anti-cancer strategy.

INTRODUCTION

Importance of the Notch Pathway in Normal and Cancerous Cells

Notch is among the most central pathways in self-maintenance of stem cells, along with Hedgehog, Wnt, and perhaps TGF-β. Its necessity has been particularly well-established for stem cells in the nervous system, hematopoietic system, and gut [1–5]. Interestingly, the Notch pathway also determines cell fate at numerous decision points. For example, it drives toward a glial cell fate in the central nervous system [6], away from secretory goblet cell fate in the gastrointestinal tract [5, 7], and regulates the T helper 1 versus T helper 2 decision in the immune system (8). Notch has been found to be critical in development of the brain, heart, vasculature, fat, gut, and immune system [1, 5, 7, 9–14]. It interacts with several essential pathways in development and tumorigenesis, and is a mediator of the oncogenic function of Ras and a driver of the Akt/mTOR and c-myc pathways [15–18].

Given its powerful roles in stem cell maintenance and differentiation and its interactions with key oncogenic pathways, it is not surprising that Notch has been implicated in numerous cancers. This was first and most clearly demonstrated for T cell leukemia/lymphoma; it was noted nearly two decades ago that chromosomal translocations in the NOTCH1 gene occur in T cell leukemia [19]. Additionally, recent work has shown that a majority of T cell leukemias harbor either activating mutations in Notch-1 or mutations/deletions in a ubiquitin ligase that normally curbs Notch activity [20, 21]. More recently, Notch has been shown to contribute to tumorigenesis and/or tumor cell survival in cancers including breast, pancreatic, brain, melanoma, and subtypes of lung [22–27]. Therefore, a safe and effective Notch inhibitor would have potential utility against a host of human cancers.

The recent tumor stem cell hypothesis may make Notch a particularly exciting target in oncology. This hypothesis states that cancers harbor a small therapy-resistant subpopulation, perhaps as little as a few percent of the total, that act as “tumor stem cells” (TSCs) [28]. Such cells have now been isolated and cultured from leukemias, breast cancers, glioblastomas, and many other cancers [29–36]. TSCs are postulated to be the only tumorigenic cells, capable of self-renewal but also of generating the other cells in the tumor. They retain other similarities to normal stem cells in their tissue of origin, such as the ability to differentiate into cells resembling normal cell types in that tissue [32, 37]. Also similar to normal stem cells, TSCs are resistant to chemotherapy and radiation. This may be secondary to overexpression of ABC export pumps and cell cycle checkpoint proteins [38, 39]. It is possible then that standard therapies kill most of the cells comprising a tumor, but that the TSCs survive and eventually re-constitute the tumors. We must therefore identify and exploit the vulnerabilities of these cells, to develop novel targeted strategies or to sensitize them to standard therapies. Given the similarity of TSCs to normal stem cells, they may also depend on classic stem cell pathways such as Hedgehog, Wnt, and Notch. Supporting this, recent reports show a role for Hedgehog in glioblastoma brain tumor TSC survival [40, 41]. Additionally, it was demonstrated in medulloblastomas, an embryonal brain tumor, that the TSC-like side population is particularly vulnerable to cell death from a Notch inhibitor [42]. TSCs may also be driven toward differentiation by Notch inhibition, since Notch activity has been shown to maintain precursor cells in some tissues [1,43]. Thus Notch inhibitors may be a means to target this critical and resistant sub-population of tumor cells.

It should be noted, however, that Notch may have the opposite effect in some cancers, acting as a tumor suppressor. This has been suggested by reports in non-small cell lung cancer, certain skin cancers, and possibly in B cell malignancies [44–46]. This dichotomy illustrates a key feature of Notch signaling—that its effects are particularly context-dependent. It also raises the concern that Notch inhibitors might even increase the risk of certain cancers.

Notch Signaling

In discussing inhibitors of Notch, it is important to consider the complex workings of this pathway (summarized in a simplified fashion in Fig. 1). There are four members of the Notch family (Notch-1, -2, -3, and -4), each a single-pass transmembrane protein with an extracellular domain containing epidermal growth factor (EGF)-like repeats and an intracellular domain containing RAM, ankyrin, transactivation, and PEST regions. The ligands for Notch are similar but smaller single-pass transmembrane proteins, consisting of three Delta-like proteins (DLL-1, -3, and -4) and two Jagged proteins (JAG-1 and -2). Other ligands have been proposed, such as the NOV (nephroblastoma-overexpressed) and F3/contactin genes [47, 48], but they have not been as well-established as Delta-like and Jagged. Binding to Notch by ligand on an adjacent cell triggers two enzymatic cleavages of Notch, extracellularly by an alpha-secretase (a disintegrin and metalloprotease ADAM-10 and ADAM-17) [49, 50] and intracellularly by the gamma-secretase/presenilin complex. The liberated Notch intracellular domain (NICD) then travels to the nucleus, displaces corepressors from CSL (CBF1/Suppressor of hairless/LAG-1) transcription factors, and recruits coactivators such as histone acetyl-transferase and Mastermind-like. In humans, CBF1 is the only CSL factor identified to date. Its targets include the Hes and Hey family of transcription factors, c-myc, NF-κB2, p21, and many others yet to be discovered [51–55]. Up-regulating the transcription of numerous transcription factors enables Notch activation to trigger a cascade of actions and potent effects in the cell.

Fig. (1).

Overview of Notch pathway.

Other aspects of the Notch pathway are also noteworthy. The Notch ligands Delta-like and Jagged are processed in a similar fashion to Notch following binding; they too are cleaved by alpha- and gamma-secretase, liberating an intracellular domain that is believed to localize to the nucleus [56, 57]. The functions of their intracellular domains remain mysteries. One report suggested that the intracellular domain of Jagged-1 up-regulates activity of the AP-1 transcription factor important in cancer and inflammation, but no mechanism was proposed [57]. The likelihood of independent functions for Delta-like and Jagged proteins should be considered in the development of Notch inhibitors, as most strategies for Notch inhibition are also likely to inhibit Delta-like and Jagged.

Non-canonical Notch signaling has also been reported to occur through the Deltex protein [58, 59], though mechanisms for this or specific downstream effects are as yet unknown. Curiously, Deltex has also been reported to be involved in ubiquitylation and degradation of Notch, leading to its proteasomal degradation [60,61]. It is therefore difficult at this time to know how to factor in non-canonical Notch signaling in a discussion of Notch inhibitors. Further complicating Notch signaling, evidence is mounting that different Notch family members and different Notch ligands may yield quite divergent signaling. One study has proposed that Notch-1 and Notch-2 play opposing roles in the embryonal brain tumor medulloblastoma [62]. A dramatic example of differing Notch outputs from different ligands has come from the immune system, in which differentiation of T helper precursors into T helper 1 or T helper 2 cells was determined by whether the adjacent antigen-presenting cell expressed Delta-like-1 or Jagged-1 (8). Evidence is accumulating that diverse cancer types depend on different Notch family members and ligands. Notch-1 has been found most frequently to be up-regulated in cancer, but two recent studies have implicated Notch-3 in a subtype of lung cancer and in choroid plexus carcinoma [26, 63]. A steadily increasing number of reports have indicated high expression of Jagged- 1 in various cancers [22, 64–68], but over-expression of other Notch ligands has also been shown [22, 64, 69].

CURRENT AGENTS AND STRATEGIES FOR NOTCH INHIBITION

Gamma-Secretase Inhibitors

Development of Notch inhibitors is not a straightforward enterprise, as the Notch pathway consists primarily of protein- protein interactions without enzymatic activity by pathway members. However, one point of leverage against the pathway has come from the gamma-secretase enzymatic cleavage essential to Notch processing. Gamma-secretase inhibitors (GSIs) were first developed as potential therapies for Alzheimers disease, as the Amyloid Precursor Protein (APP) implicated in this illness is also cleaved by gammasecretase [70]. GSIs were since adapted for use as Notch inhibitors, and have been widely used in research for this purpose. A Merck GSI, MK-0752, continues in clinical trials for T-cell leukemia/lymphoma and breast cancer.

The use of a gamma-secretase Notch inhibitor in clinical trials in humans has already yielded valuable information. The initial trial in patients with T cell leukemia/lymphoma showed diarrhea to be the dose-limiting toxicity [71]. This is likely an on-target side effect of Notch inhibition, as Notch inhibition has been shown to convert intestinal cells from a nutrient-absorbing to a mucus-secreting type (5). Thus other classes of Notch inhibitor introduced in the future may also have to contend with this side effect. Fortunately, early indications from patients suggest that the diarrhea is lessened by a dosing schedule that does not require continuous daily dosing by patients, e.g. giving a week off each month. There has also been concern about other potential on-target effects of Notch inhibition, which might include impairing or ablating the adult stem cell populations that have been found in various organs.

Clinical application of gamma-secretase inhibitors may suffer not only from potential on-target effects of Notch inhibition but also from off-target effects, as a number of other proteins besides Notch family members and APP are also cleaved by gamma-secretase. Other gamma-secretase targets include CD44, ErbB4, LRP, syndecan-3, p75 NTR, Apo ER2, DCC, Nectin-1α, E-cadherin, and possibly N-cadherin [72–82]. There has not yet been a systematic effort to dissect out effects of GSIs not secondary to Notch inhibition. It should be noted, however, that the lack of specificity of GSIs does not necessarily make them poor therapeutic candidates. Some of the most promising targeted agents available or being developed block several targets, such as poly-tyrosine kinase inhibitors.

There are numerous gamma-secretase inhibitors commercially available for research; Calbiochem (a division of EMD Biosciences, San Diego, CA) alone sells over twenty varieties. There is some diversity in the chemical structures used to generate these GSIs. The most common type consists of a modified di- or tri-peptide, usually with one to two aromatic hydrocarbon rings. These inhibitors are generally cell-permeable, reversible inhibitors of gamma-secretase. The most commonly used in the literature is DAPT (N-[N-(3,5-Difluorophenylacetyl-L-alanyl)]-S-phenylglycine t-Butyl ester), possibly followed by L685,458 which shares a similar structure. Another which has seen common usage and possessing a different structural base is compound E ((s,s)-2-(3,5-Difluorophenyl)-acetylamino]-N-(1-methyl-2-oxo-5- phenyl-2,3-dihydro-1H-benzo[e][1, 4]diazepin-3-yl)-propionamide). Others such as DBZ (dibenzazepine) contain diazepine- type structures. Still another type is the irreversible inhibitor JLK6, containing an isocoumarin foundation (7-amino-4-chloro-3-methoxyisocoumarin). Potent gamma-secretase inhibitors with a sulfonamide core have been developed, such as Compound 18 ([11-endo]-N-(5,6,7,8,9,10-hexahydro-6,9-methanobenzo[9][8]annulen-11-yl)-thiophene-2-sulfonamide) [42,83]. Other chemical classes of gamma-secretase inhibitors have also been explored [84].

Established medications of other types may also demonstrate gamma-secretase inhibitor activity. Non-steroidal anti-inflammatory drugs (NSAIDs) were reported several years ago to demonstrate GSI activity [85,86], though one report described them as sparing Notch cleavage and activation [87]. However, in our hands the NSAID flurbiprofen was able to reduce Notch activity40–50% in a CBF1-luciferase reporter plasmid assay in glioma cells (unpublished data). A50% decrease in this assay actually compares favorably to the results we have obtained with a number of commercially available dedicated GSIs.

Notch Inhibitors in Pre-Clinical Development

There is little in the literature relative to preclinical development of other inhibitors of the Notch pathway. However, one recent development promises to have very high clinical impact. A few references suggested in years past that Notch signaling is important for the angiogenesis, or new blood vessel formation, required by tumors to sustain their growth [88–90]. Then in 2007, new reports demonstrated the Notch ligand Delta-like-4 (DLL4) to be vital in tumor angiogenesis, and blocking antibodies to DLL4 were shown in animal models to dramatically inhibit tumor growth [91,92]. Interestingly, DLL4 was shown to suppress endothelial cell growth, so blocking DLL4/Notch signaling led to an increase in tumor blood vessels. However, this occurred in a disorganized fashion, so the tumor blood vessels resulting from DLL4 blockade were chaotic and dysfunctional. DLL4 antibodies therefore led to marked inhibition of tumor growth. Also important from a translational perspective, DLL4 antibodies were effective against tumors resistant to the current anti-angiogenic standard, antibody to vascular endothelial growth factor (VEGF) [92]. Thus these targeted inhibitors of Notch/DLL4 signaling offer tremendous promise as novel anti-tumor agents.

Specific antibodies or peptides broadly offer the potential to modulate individual Notch family members and ligands, and this may ultimately be extremely useful in fine-tuning Notch inhibition and avoiding adverse effects. A commercially available blocking antibody to Notch-1 has been available for many years. In one important recent publication, Li and colleagues derived antibodies that up- or down-regulate activity of Notch-3 only [93]. The authors noted that the most effective of these regulatory antibodies bound to a region of Notch-3 that in the resting, inactive state blocks a site of proteolytic cleavage. This interesting observation, one that makes intuitive sense, may be of particular use in designing future agents to modulate activity of Notch family members.

Soluble forms of the extracellular domains of Notch family members and ligands represent another means of blocking Notch/ligand interactions [94]. In the past, this has been done through attaching the extracellular domain of DLL1 to an antibody Fc region [95]. Such agents present challenges, however, as it has been shown that cross-linking them or changing their concentration can lead them to act as Notch agonists instead of antagonists [96]. Their use may therefore remain restricted to the laboratory.

SOME FUTURE APPROACHES TO INHIBITING NOTCH

Though the limited number of enzymes involved in Notch processing and activity increases the difficulty of deriving inhibitors, there are still many potential vulnerabilities that may be exploited by future inhibitors of the pathway. These will likely require the use of blockers of protein-protein interactions such as antibodies, peptides binding with high affinity, or with technological advances even small-molecule agents. Newer agents such as hydrocarbon-stapled peptide helices and RNA or protein aptamers may also be applicable to blocking these protein-protein interactions.

Some of the points that may be targeted in the Notch pathway include the interactions between NICD, MAML, and CBF1. However, it should be kept in mind that there is likely more to Notch signaling than that through CBF1. It has been reported to act through the Deltex proteins, as mentioned above, and a recent paper from the Gaiano laboratory further supports that Notch signaling and CBF1 signaling are not equivalent (43). Recent work has indicated that NICD proteins dimerize on the promoters of certain target genes, cross-linking NICD/MAML/CBF1 complexes and increasing their effects [97]; if a blocker of this dimerization were developed, it might dampen transcription only of especially sensitive target genes. Deriving new blockers of these interactions should become more achievable as structural insights into the Notch pathway continue to accrue from the laboratories of Steven Blacklow and others.

Up-regulation of endogenous Notch antagonists represents another conceivable strategy for inhibiting Notch. Genes such as Numb and Numb-like have been shown to diminish Notch activity [98,99], and therapies that increase their transcription may therefore block Notch—either alone or as an adjunct to other forms of Notch inhibition. The ubiquitin ligase FBXW7 has been shown to trigger Notch degradation [100], and an intriguing recent report indicates that FBXW7 mutation or deletion represents a mechanism by which some T cell leukemias up-regulate Notch activity (20). FBXW7 is a target gene for the tumor suppressor p53 [101,102], so it is possible that therapeutic approaches for restoring p53 function to cancer cells will lead to up-regulation of FBXW7 transcription and thus Notch down-regulation. There are doubtless other endogenous inhibitors of Notch activity yet to be discovered, with the newly-described microRNAs—small non-coding RNAs thought to regulate translation of much of the human genome—likely to include some that down-regulate Notch activity. Delivery of such Notch-inhibiting microRNAs to cancer cells may soon be another strategy for Notch inhibition.

SUMMARY

Notch is a critical pathway in stem cell maintenance, development, and cancer. It has been shown to be important in survival and tumorigenesis in numerous cancer types, and its potential utility against “tumor stem cells” makes it a particularly high-value target. Notch signaling is complex, with activation of Notch family members on one cell triggered by ligands expressed on the surface of an adjacent cell. Notch/ligand binding triggers enzymatic cleavage of Notch, liberating an intracellular domain that travels to the nucleus and complexes with the CBF1 transcription factor to up-regulate target genes. Despite the attractiveness of this pathway as a therapeutic target, very few Notch inhibitors have been developed, likely because pathway members do not themselves have enzymatic activities that lend themselves to blockade. However, inhibitors of the gamma-secretase enzyme complex essential in Notch processing have long been used in the laboratory as Notch inhibitors, and they have recently begun clinical trials in patients with cancer. These trials have revealed issues likely to be faced by other Notch inhibitors, such as gastrointestinal toxicity. Other more selective inhibitors of Notch and Notch ligands, such as antibodies to Delta-like-4, are in preclinical development and may show great promise against not only cancer cells but also tumor angiogenesis. In the near future, other points of protein/protein interaction in the Notch pathway may be vulnerable to inhibition through recent technological advances, and it may also be possible to exploit endogenous inhibitors of Notch. After many years of basic and pre-clinical investigation, the rapidly-progressing field of Notch research now seems poised to have a major impact on the treatment of cancer in patients.