This story was updated June 10, 2019.

The side effects that emerged in AD trials have given γ-secretase inhibitors a bad rap. According to a paper published May 29 in Science Translational Medicine, it might be time to take a second look. Researchers led by Bart De Strooper and Jan Cools of KU Leuven, Belgium, reported that a small-molecule inhibitor that selectively targets presenilin-1 stifled the proliferation of leukemic T cells expressing a mutant form of the γ-secretase substrate Notch. The molecule—Merck’s MRK-560—prolonged survival in mice transplanted with human leukemic T cells. What’s more, it did so without any of the immune or gastrointestinal side effects caused by broad spectrum γ-secretase inhibitors.

  • Deleting or blocking PS1 in T cells suppressed leukemia in mice.
  • The inhibitor caused neither gastrointestinal nor immune side effects.
  • Could specific γ-secretase inhibitors work for AD?

While the relevance of these findings to Alzheimer’s disease is unclear, they suggest that selectively targeting specific γ-secretase complexes could avoid some side effects of the drugs, De Strooper said.

De Strooper and colleagues reasoned that the side effects of GSIs hinge on their inhibition of all γ-secretase complexes. Four different types exist, based on which subunits they comprise. While all complexes contain one nicastrin and one presenilin enhancer-2 (PEN-2) protein, they can incorporate either APH-1A or APH-1B, and presenilin-1 (PS1) or presenilin-2 (PS2). PS1 and PS2 are expressed at different ratios across cell types and even across different compartments within the cell. The researchers proposed that selectively targeting PS1 complexes might reduce systemic inhibition of Notch processing, thought to cause the gastrointestinal upset, skin lesions, and worsened cognition seen in Phase 3 trials of GSIs in AD (Aug 2010 news; Dec 2012 news). 

First author Roger Habets and colleagues decided to test this idea in models of T cell acute lymphoblastic leukemia (T-ALL). Sixty percent of people with T-ALL have Notch1 mutations, which free this transmembrane receptor from its usual dependence on ligand binding for activation and expose it to γ-secretase for cleavage. The cleaved Notch intracellular domain (NICD) then enters the nucleus, where it triggers unchecked proliferation of T cells. However, even though γ-secretase is needed for T cells to keep proliferating in this disease, cancer trials of GSIs, too, have failed due to systemic side effects. Interestingly, while normal T cells express four times more PS1 than PS2, cancerous T-ALL cells make 30 times more. Could targeting only PS1-containing γ-secretase complexes stifle Notch processing in mutant T cells?

The researchers first tested this hypothesis genetically. They transplanted hematopoietic progenitor cells into normal adult mice. Progenitors expressing mutant Notch1 under an inducible promoter differentiated into cancerous T cells, causing leukemia, and death within around 72 days. However, when the researchers turned off the PS1 gene in the Notch 1 mutant progenitor cells three weeks after transplant, T cell proliferation slowed down. Only nine out of 17 mice transplanted with the PS1-deficient cells developed leukemia, and those that did survived for an average of 120 days.

Might selective PS1 inhibitors have the same effect? Mice treated with MRK-560, a GSI that selectively inhibits PS1-bearing γ-secretase, survived for 30 days after injection of mutant Notch1 lymphoblasts, compared with 18 days for control mice. The scientists delivered MRK-560 after they spotted leukemic clones in the mice, and continued for 14 days or until death.

The drug also stifled leukemia in immunodeficient mice transplanted with human cells from three different cases of T-ALL, nudging up their survival 1.5-fold over that of transplanted control mice. MRK-560 also prolonged survival when the researchers held off dosing until the burden of leukemia was high.

Presenilin1 Blocker Slows Cancer. Lymphocytes from patients with T-ALL ran amok after transplant into five mice, as measured by in vivo bioluminence (top). In mice co-administered MRK-560, leukemia was delayed (bottom). [Courtesy of Habets et al., Science Translational Medicine, 2019]

The researchers checked for side effects commonly associated with nonspecific GSIs. In normal mice treated for four weeks with the broad-spectrum GSI dibenzazepine, the number of intestinal goblet cells increased fourfold, a known effect of Notch inhibition. In contrast, the intestinal cells of mice taking MRK-560 looked normal. The researchers reasoned that the near-equal expression in the small intestine of PS1 and PS2 allowed the latter to compensate for inhibition of the former. Similarly, development of wild-type T cells, which have much less PS1 than Notch1 mutants, was curtailed when mice were dosed with the broad-spectrum GSI, but not on MRK-560. This, too, the authors attributed to PS2 compensation in healthy T cells.

Todd Golde of the University of Florida, Gainesville, had previously compared the activity of several GSIs in clinical development for cancer, reporting vastly different substrate cleavage properties (Ran et al., 2017). He called this new work a complete and elegant preclinical study. Golde’s group has also found that MRK-560 preferentially inhibits PS1 over PS2, though he questioned whether the benefit would translate to humans or extend to other selective GSIs.

Still, Golde believes the strategy is worth investigating. “I do think, with warts emerging on BACE1 inhibitors, that the field should revive efforts around γ-secretase modulators (GSMs) and possibly this kind of GSI approach, but we should do so knowing that success is uncertain,” he wrote. “Either approach needs to be used in primary prevention or at the earliest sign of conversion to amyloid positivity.”

De Strooper noted that PS1 complexes produce the lion’s share of Aβ42 in the brain. While MRK-560 has not been tested in people, it prevented Aβ plaques in mouse models of amyloidosis without causing Notch-related side effects (Best et al., 2007; Borgegård et al., 2012). The inhibitor may be effective at inhibiting Aβ production while avoiding the gastrointestinal and immune side effects of broader GSIs, De Strooper said. Despite this promising preclinical data, Merck dropped γ-secretase inhibitor research. “Merck decided not to pursue development of GSI inhibitors based on results from other programs indicating lack of efficacy and adverse clinical effects,” wrote Mike Egan, Merck Research Laboratories, North Wales, Pennsylvania. 

Would selective GSIs avoid the cognitive decline that nonselective GSIs caused in AD trials run by Eli Lilly and Bristol-Myers Squibb? Eric Siemers, of Siemers Integration LLC, headed the Phase 3 semagacestat GSI trial while at Lilly. He was skeptical. “In my view, inhibition of Notch cleavage in the gut and elsewhere was not the problem with the Lilly and BMS GSIs,” Seimers wrote. “I suspect one of the other multiple substrates of GS was the source of the cognitive worsening. So while this looks like a promising therapy in T-ALL, it would require a great deal of speculation to make the case that it would be useful in AD.”

De Strooper acknowledged that γ-secretase substrates other than Notch could have played a role in the cognitive decline. However, he thinks that because PS2 is expressed in neurons, it has a good chance of processing a variety of substrates to compensate for the loss of PS1. That remains to be seen. A previous study led by Wim Annaert of KU Leuven, on which De Strooper is a co-author, reported that PS2 activity is restricted primarily to endocytic compartments, where it generates intraneuronal Aβ (Jun 2016 news). Some researchers believe that the intracellular pool of Aβ fuels amyloidosis (Gouras et al., 2005; Pensalfini et al., 2014). 

Many other groups have attempted to develop GSMs, which selectively reduce the production of longer Aβ peptides while sparing γ-secretase’s cleavage of other important substrates (for review, see Xia, 2019). However, several candidates have not moved beyond early trials due to poor pharmacological properties, toxicity, or because whole drug development programs were dropped (Jun 2013 news; Aug 2008 conference newsPF-06648671). De Strooper and colleagues are investigating γ-secretase stabilizing compounds, given that these allow the enzyme to more fully process amyloid precursor protein C-terminal fragments into shorter, less-amyloidgenic Aβ peptides (Szaruga et al., 2017). He sees modulators and stabilizers as parallel approaches to selective GSIs like MRK-560.

Despite the failure of broad GSIs, De Strooper maintains that γ-secretase remains the best-validated target for treating AD, if the enzyme can be targeted safely and patients start treatment early. Beyond AD, De Strooper thinks MRK-560 holds promise for the treatment of T-ALL, and potentially other cancers where mutant Notch unleashes proliferation of cells.—Jessica Shugart


No Available Comments

Make a Comment

To make a comment you must login or register.


News Citations

  1. Lilly Halts IDENTITY Trials as Patients Worsen on Secretase Inhibitor
  2. Drug Company Halts Development of γ-Secretase Inhibitor Avagacestat
  3. Lodged in Late Endosomes, Presenilin 2 Churns Out Intraneuronal Aβ
  4. Satori Pharmaceuticals Shuts Down, Abandons γ-Secretase Modulators
  5. Chicago: Flurizan Postmortem

Therapeutics Citations

  1. PF-06648671

Paper Citations

  1. . γ-Secretase inhibitors in cancer clinical trials are pharmacologically and functionally distinct. EMBO Mol Med. 2017 Jul;9(7):950-966. PubMed.
  2. . The novel gamma secretase inhibitor N-[cis-4-[(4-chlorophenyl)sulfonyl]-4-(2,5-difluorophenyl)cyclohexyl]-1,1,1-trifluoromethanesulfonamide (MRK-560) reduces amyloid plaque deposition without evidence of notch-related pathology in the Tg2576 mouse. J Pharmacol Exp Ther. 2007 Feb;320(2):552-8. PubMed.
  3. . Alzheimer's disease: presenilin 2-sparing γ-secretase inhibition is a tolerable Aβ peptide-lowering strategy. J Neurosci. 2012 Nov 28;32(48):17297-305. PubMed.
  4. . Intraneuronal Abeta accumulation and origin of plaques in Alzheimer's disease. Neurobiol Aging. 2005 Oct;26(9):1235-44. PubMed.
  5. . Intracellular amyloid and the neuronal origin of Alzheimer neuritic plaques. Neurobiol Dis. 2014 Nov;71:53-61. Epub 2014 Aug 1 PubMed.
  6. . γ-Secretase and its modulators: Twenty years and beyond. Neurosci Lett. 2019 May 14;701:162-169. Epub 2019 Feb 11 PubMed.
  7. . Alzheimer's-Causing Mutations Shift Aβ Length by Destabilizing γ-Secretase-Aβn Interactions. Cell. 2017 Jul 27;170(3):443-456.e14. PubMed. Correction.

Further Reading

Primary Papers

  1. . Safe targeting of T cell acute lymphoblastic leukemia by pathology-specific NOTCH inhibition. Sci Transl Med. 2019 May 29;11(494) PubMed.