γ-Secretase inhibitors (GSIs)
γ-Secretase is a key enzyme in the activation process of NOTCH pathway. Inhibiting its activity can induce G0/G1 cell cycle arrest and inhibit cell proliferation.
The original GSI trial drug was MK-0752. Seven patients with T-ALL received MK-0752 for 56 days. One of the patients showed a certain anti-leukemia effect to the drug, but the response was transient. Most patients developed gastrointestinal toxicity (mainly diarrhea) at a dose of 300 mg/m
2, which lead to the termination of the trial [
17]. The GSI may block NOTCH1 and NOTCH2 at the same time and disturbs the normal function of physiological NOTCH signals. As a result, serious imbalance of digestive tract homeostasis is considered as the main reason of diarrhea [
2,
18].
PF-03084014 is an oral, non-competitive, reversible, selective γ-secretase inhibitor. In a HPB-ALL (a subset of human T-ALL) cell line, cells in S-G2-M phase were suppressed after 7 days of PF-03084014 treatment, while cells in G0-G1 phase were accumulated. The ICN1 protein was completely inhibited, and the expression of
Notch target genes
Hes-1 and
cMyc were down-regulated in these cells [
19]. PF-03084014 was well tolerated in mice at dose levels below 100 mg/kg. HPB-ALL xenotransplanted mice were tested at 150 mg/kg with different intermittent treatment schedules: continuous treatment, 3 days treatment/4 days withdrawal, 7 days treatment/7 days withdrawal. The PF-03084014 at 7-day treatment/7-day withdrawal schedule exerted the best anti-cancer efficacy and the least toxicity in HPB-ALL xenotransplanted mice [
19]. In a phase I study, eight patients with T-ALL or T-cell lymphoblastic lymphoma (5 T-cell lymphoblastic lymphoma and 3 T-ALL) received PF-03084014 at 150 mg twice daily [
20]. All eight patients had previously received systemic treatment and relapsed. One of the T-ALL patients achieved complete remission lasting about 3 months, and then relapsed. The
HES4 gene expression level was suppressed during remission and was higher than the baseline level when the disease relapsed in this case. The most common adverse effects were nausea and vomiting [
20].
BMS-906024, another small molecule γ-secretase inhibitor was screened by efficacy/tolerance profile. The efficacy of BMS-906024 was dose-dependent, and the lowest effective dose was 1 mg/kg/day in xenotransplantation mice [
21]. At higher doses, a significant delay in tumor growth can be observed even after the administration was complete. And no significant toxicity such as weight loss of mice was observed [
21]. A clinical trial of the agent allowed a 53-year-old male patient with relapsed and refractory early T-cell progenitor acute lymphoblastic leukemia to reach complete remission with a deep molecular response [
22]. In a Phase I clinical trial, BMS-906024 was administered in 25 children with T-ALL or T-cell lymphoblastic lymphoma at 4–6 mg weekly. One-third of cases (32%) showed at least a 50% reduction in bone marrow blasts and had tolerable side effects [
23].
MRK-560, a precise selective γ-secretase inhibitor, mainly targets PSEN1 in the γ-secretase complex in T-ALL cell lines [
24]. MRK-560 was used to treat T-ALL patient-derived xenograft mouse models with different
NOTCH1 mutations. The anti-tumor activity and improved survival were observed in all mice with various response. And the mice administered with MRK-560 did not show obvious gastrointestinal pathological changes and thymus T cell development defects [
24].
There are currently a variety of GSIs in the experimental stage [
5]. But overall, GSI has a limited anti-tumor effect, mostly manifested as transient growth arrest rather than cell death, and a single agent seems to be insufficient to eradicate T-ALL blasts [
25]. Combining GSI with other agents that can trigger cell death may be an alternative option to treatment of
NOTCH1-mutant T-ALL [
26,
27]. Chloroquine (CQ) can increase the accumulation of reactive oxygen species in T-ALL, activate DNA damage, enhance GSI-induced cell cycle arrest in T-ALL, and interfere with ligand-independent NOTCH1 transportation and localization [
28]. In addition, it can also reduce the concentration of GSI with less side effects [
28]. The combination therapy of GSIs and CQ showed excellent synergistic effect in vitro on T-ALL cell line [
28]. Because there is PI3K/mTOR pathway activation in T-ALL [
29], some researchers have conducted experiments using GSI and mTOR inhibitor rapamycin in mouse model [
30]. T-ALL xenograft mice were divided into 4 treatment groups: control group, GSI group, rapamycin group and combination group. The average survival time of the combination group was significantly longer than that of the single agent group and the control group. The combined treatment increased the mean survival time of mice by more than 50 days [
30]. These results suggest that the combination of GSI and mTOR inhibitors have synergistic inhibitory effect on T-ALL blasts. The addition of glucocorticoid to GSI can overcome the severe gastric-intestinal toxicity of GSI in a xenograft mice model [
19]. The xenograft mice were treated with PF-03084014 (150 mg/kg, twice a day) or PF-03084014 plus dexamethasone (15 mg/kg, once a day) [
19]. After euthanizing of mice at day 14, the duodenum in combined treatment mice had similar number of goblet cells to that in untreated animals assayed by histology, which demonstrated that dexamethasone had a protective effect on gastrointestinal toxicity caused by PF-03084014 [
19]. Furthermore, the application of GSI can reverse glucocorticoid resistance of T-ALL [
31]. In the highly resistant human T-ALL cell line with glucocorticoids, when treated with dexamethasone concentrations as high as 1 × 10
–5 M, cells only showed the minimum loss of cell viability. However, the addition of GSI to dexamethasone showed a synergistic dose-dependent response and could effectively reduce cell viability, with a 50% inhibitory concentration value of 7.7 × 10
–8 M for dexamethasone in the presence of GSI. The addition of glucocorticoids antagonists can reverse the synergistic effect of combination therapy [
32]. There are also other combination options tested. For example, protein kinase CK2 inhibitor CX-4945 combined with GSI can destroy the stability of NOTCH1, reduce the growth and survival of human T-ALL cells [
33,
34]. The combined application of cyclin D-dependent kinase CD4 inhibitor and GSI made ALL cells more sensitive to GSI, with a synergistic effects on cell cycle arrest and cell death [
35,
36]. Therefore, it is promising to apply GSI with other agents combining with chemotherapy in the management of T-ALL cases.
Currently, no GSI has been approved for the clinical use in treatment of T-ALL. In addition to gastrointestinal toxicity and limited anti-leukemia effects, researchers also found that there are differences in the degree of inhibition of NOTCH1 cleavage by various GSIs and a lack of bioequivalence [
37]. Further study found that some T-ALL cells are resistant to GSIs, which may be related to PTEN inactivating mutations [
26,
38]. Therefore, further researches on GSIs with different agents are necessary to guarantee the effective management of T-ALL cases with
NOTCH1 mutation.
Monoclonal antibody targeting NOTCH1
OMP-52M51 is a monoclonal antibody produced by mice immunized by human NOTCH1 protein fragments [
40]. It can bind to the negative regulatory region of NOTCH1 and prevent ligand-free activation of NOTCH1 receptors. OMP-52M51 reduces the level of NOTCH1 intracellular domain in T-ALL cell line in vitro and significantly inhibits T-ALL cell growth in xenograft models [
40]. Agnusdai and colleagues [
40] reported seven T-ALL xenograft mice (4
NOTCH1 mutants and 3
NOTCH1 wild types) treated with OMP-52M51. All of the T-ALL xenograft mice carrying
NOTCH1-mutant had reduced circulating leukemia cells, while the animals carrying
NOTCH1 wild-type had no response to OMP-52M51 treatment. The survival time of
NOTCH1-mutant T-ALL mice were significantly prolonged in comparison to the animals carrying
NOTCH1 wild-type after treatment with OMP-52M51(mean 44 days vs. 15 days). Leukemia burden were reduced by 90% in animals with
NOTCH1-mutant by administration of OMP-52M51. At necropsy, the spleen size was significantly reduced in mice treated with anti-NOTCH1 antibody compared to that in the untreated group. Ki67 staining showed a significant reduction after anti-NOTCH1 antibody administration. Similar to GSIs, the combined application of OMP-52M51 and dexamethasone enhances the anti-NOTCH1 efficacy [
40]. However, some T-ALL xenografts developed resistance to OMP-52M51. And gene sequencing analysis suggested that OMP-52M51 resistant mice mostly carried two activation mutations of
NOTCH1 gene: p.Q1584H and p.L1585P [
41]. Therefore, T-ALL patients with p.Q1584H and p.L1585P mutations of
NOTCH1 gene may not benefit from OMP-52M51. However, there are no further clinical trials to support this hypothesis, and further research is needed.
Ankur and colleagues generated a monoclonal antibody mAb604.107 against the negative regulatory region of NOTCH1 [
42]. At low concentrations (1–2 μg/mL), it can distinguish the conformation of the NRR region between mutated
NOTCH1 and wild-type
NOTCH1. T-ALL leukemia-initiating cells were inhibited by mAb604.107. The mAb604.107 can increase the sensitivity of the chemotherapy drug doxorubicin in an animal model [
42]. Therefore, mAb604.107 may be a promising monoclonal antibody targeting
NOTCH1-mutant in T-ALL.
SERCA inhibitors
The formation of the NOTCH1 receptor protein HD region and the transport process of NOTCH1 require the participation of calcium ions [
17]. The NOTCH1 signal is related to the intracellular Ca
2+ homeostasis in T-ALL [
44]. In a NOTCH1-dependent model of T-ALL, the absence of Ca
2+ channel activators STIM1 and STIM2 can significantly prolong the survival of these animals [
45]. Ca
2+ ATPase (SERCA) regulates the calcium ion concentration. Therefore, some studies explore the role of SERCA inhibitors in activation of NOTCH1 signaling pathway.
The compound thapsigargicin was screened from a series of SERCA inhibitors. The study in vitro showed that thapsigargicin induced cellular apoptosis by depletion of endoplasmic reticulum calcium ions and oxidative stress. The inhibitory effect of thapsigargicin on T-ALL was evaluated in cell lines and mouse model [
46]. Thapsigargicin inhibits
NOTCH1 mutated T-ALL cells more specifically than
NOTCH1 wild-type T-ALL cells. However, severe dose-limited cardiotoxicity was observed, which limited the clinical application [
47]. The folate conjugate named as JQ-FT combines folic acid with thapsigargicin through a cleavable bond to achieve leukemia-specific delivery of thapsigargicin. JQ-FT is a
NOTCH1 inhibitor with a dual selectivity, targeting both
NOTCH1 mutations and leukemia cells. In the xenograft model, JQ-FT inhibits
NOTCH1-mutated T-ALL growth in vivo. And the sensitivity of JQ-FT to cells in
NOTCH1-mutated T-ALL is higher than that of
NOTCH1 wild-type cells [
47].
Clerodane diterpene casearin J (CJ), a natural SERCA inhibitor, mainly targets
NOTCH1 HD domain mutant in T-ALL. Experiments in cell lines show that CJ can induce T-ALL cell death in the low molar concentration [
48]. CJ reduces the level of ICN1 in T-ALL cells carrying
NOTCH1 HD domain mutations, and this effect was significantly reduced in the cells with normal
NOTCH alleles and juxtamembrane expansion mutations [
49]. CJ can also cause slight activation of
NF-κB. The combined treatment of CJ and NF-κB inhibitor, parthenolide, resulted in significant synergistic death of T-ALL cells [
48].
Recently, Marchesini and colleagues identified an oral SERCA inhibitor, CAD204520 [
50]. The toxicity of CAD204520 to off-target calcium ions is significantly reduced. T-ALL xenograft mice treated with CAD204520 by oral gavage (dose 45 mg/kg, twice a day, 8 h apart, for 4 days) showed that the percentages of circulating leukemia cells were 56-fold reduced compared to the control group. Leukemia infiltration in the spleen also decreased significantly. There were no weight loss, no adverse effects on behavior, and no signs of acute cardiotoxicity or gastrointestinal metaplasia observed in mice treated with CAD204520. The complete blood count of the treatment group had no significant difference from control group. There were no general pathological abnormalities in internal organs, including heart, lung, liver, brain and kidney at 21 days after the treatment. These data demonstrated that CAD204520 is a promising option in the management of T-ALL with NOTCH mutation due to its high efficacy and tolerable toxicities.
Other agents
In addition to the above studies, there are other experimental protocols targeting
NOTCH1 [
51,
52], such as the proteasome inhibitor (bortezomib) [
53], histone deacetylase inhibitor (panobinostat) [
54], HSP90 inhibitor [
55,
56], insecticide (mebendazole) [
57], geranylgeranyl diphosphate synthase inhibition (digeranyl bisphosphonate, DGBP) [
58], and the antibody Rova-T against its ligand DLL3
[59]. Other natural anti-
NOTCH compounds have been shown to inhibit
NOTCH1 mutant T-ALL cells, such as plant polyphenol flavonoids [
60], artemisinin [
61], etc. However, all of the currently investigated agents targeting on Notch signaling pathway are waiting to be approved for clinical application in the management of T-ALL patients [
2].