Researchers hope they can treat or prevent Alzheimer's by blocking β-secretase (BACE), the enzyme that kick-starts production of Aβ from amyloid precursor protein. It is commonly believed that the non-amyloidogenic α-secretase takes over processing of APP when BACE activity falls, limiting Aβ production. However, a study in the June 11 Journal of Neuroscience suggests it is not so simple. Scientists led by Randall Bateman, Washington University School of Medicine, St. Louis, found that in rhesus monkeys, a BACE inhibitor blocked β-secretase as expected, but did not boost α-secretase processing of APP to the same degree. Rather, APP was processed in other ways. “The finding opens the door to the idea that APP metabolism may be more complex than simple α and β processing,” Bateman told Alzforum.

Previous evidence from cell models suggests that BACE inhibitors shunt APP toward processing by α-secretase (see Fukumoto et al., 2010). In a Phase 1 trial, Eli Lilly’s LY2811376 reportedly hiked the concentration of soluble APPα (sAPPα), the N-terminal fragment of the precursor cleaved by α-secretase (see May et al., 2011). However, other evidence suggests that the pathways compete neither in cell models nor in humans (see Kim et al., 2008; Dobrowolska et al., 2014). Bateman’s group aimed to find out how the two pathways interact by using their stable isotope labeling kinetics (SILK) method (see Jun 2006 news story). This distinguishes newly produced proteins from ones that are already present, and allows the researchers to get a sensitive read on dynamic changes in protein metabolism.

Bateman's group, including first author Justyna Dobrowolska, who is now at Nationwide Children’s Hospital, Columbus, Ohio, chose the rhesus monkey as a model because its APP shares 91 percent similarity with the human protein. The researchers first took baseline samples of the blood and cerebrospinal fluid of five monkeys, then gave them each four successive treatments, separated by a washout period. In random order, the monkeys received vehicle control and three doses of Merck’s BACE inhibitor MBI-5. An hour after treatment, the scientists gave the animals an infusion of 13C-labeled leucine, an amino acid that incorporates into new proteins. Every few hours for the next day, and then periodically over the following five, the researchers sampled CSF and blood again. They drew out sAPPα, sAPPβ, or Aβ by immunoprecipitation and analyzed what fraction incorporated the heavy leucine. ELISA measurements gave the absolute concentration of each protein in the CSF. By combining the two measures, the authors calculated and compared how much of each new peptide was made at the end of 57 hours.

As expected, the newly formed sAPPβ and Aβ fell in a dose-dependent manner. Compared with monkeys given vehicle control, the fraction of sAPPβ labelled with 13C-labeled leucine fell by 30 percent in those given the highest dose of MBI-5.  The monkeys made 90 percent less Aβ and 83 percent less sAPPβ, according to the combined measurements of SILK and ELISA. While the researchers anticipated an opposite and equal rise in newly made sAPPα, they saw no change in the fraction of the peptide labelled with 13C-labeled leucine in the SILK test. Nevertheless, by taking their ELISA measurements into account, they calculated that newly made sAPPα rose by 35 percent compared with vehicle-treated controls. That suggested that as APP accumulated due to BACE inhibition, only some was shunted toward the α-secretase pathway. "The reasons for the differing magnitude of the effect of BACE inhibition on sAPPα levels detected by either ELISA or LC-MS are not understood," wrote co-author Mary Savage, Merck Research Laboratories, West Point, Pennsylvania (see full comment below).

What happened to the rest of the APP? Bateman suggested that α-secretase cleaves the protein at alternative sites (see Brinkmalm et al., 2013) to create fragments that neither ELISA nor SILK detect. How this might affect the monkeys is unclear. In vitro evidence suggests some alternative APP fragments could be toxic to neurons, said Bateman, but the clinical relevance of this is unknown. Extra APP could also undergo lysosomal degradation, the authors wrote.

Erik Portelius, University of Gothenburg, Sweden, was surprised by the results. “Previous clinical trials in humans suggest that BACE inhibitors have strong effects on sAPPα,” he said. “I would have expected that here as well.” Bateman pointed out that those measures come not from SILK but from ELISA. Portelius suggested that perhaps BACE inhibitors affect monkeys differently, or maybe this particular BACE inhibitor has distinct effects on the levels of sAPPα. MBI-5 is distinct from Merck’s BACE inhibitor MK-8931, currently in Phase 3 clinical trials. Nevertheless, Portelius found the drop in Aβ and sAPPβ encouraging, suggesting that BACE inhibition could be an effective means of lowering the production of Aβ in people.

Bateman said the next steps would be to apply these methods to people being treated with BACE inhibitors to see if similar changes occurred. He also suggested applying the SILK method to people with AD to see if their BACE function rises above controls, as some have hypothesized.—Gwyneth Dickey Zakaib


  1. The primary expected therapeutic mechanism of action of BACE inhibition is to lower toxic Aβ burden in the CNS via inhibition of its synthesis from APP. This is founded on the strongest human genetic evidence possible, including identification of both risk-enhancing and protective mutations. The relative therapeutic impact of increased CNS sAPPα in response to CNS BACE inhibition is unknown. Several physiological roles of sAPPα have been proposed (e.g. a neuroprotective role), but none are yet widely accepted and understanding of the biology of sAPPα is still evolving. It is important to note that sAPPα levels were not decreased by the BACE inhibitor.

    The reasons for the differing magnitude of the effect of BACE inhibition on sAPPα levels detected by either ELISA or LC-MS are not understood.  However, the modest increase in ELISA sAPPα, a measure of total sAPPα, demonstrates that this pathway contributes to alternate APP clearance when BACE is considerably inhibited.  As noted in the manuscript, we continue to advance a comprehensive model of CNS APP processing in the non-human primate, including data from both ELISA and SILK/LC-MS during steady state conditions and also following both BACE and gamma-secretase inhibition.  This more comprehensive model is likely to contribute to further understanding of APP processing kinetics.

  2. A very interesting finding, indeed!

  3. Inhibition of α-secretase does not affect BACE1 processing in neuroblastoma cells expressing endogenous APP (Gandhi et al., 2004). Thus, the two pathways don't appear to compensate for each other, but appear to occur in independent compartments.

Make a Comment

To make a comment you must login or register.


News Citations

  1. CSF Aβ—New Approach Shows Rapid Flux, May Help Evaluate Therapeutics

Therapeutics Citations

  1. Verubecestat

Paper Citations

  1. . A noncompetitive BACE1 inhibitor TAK-070 ameliorates Abeta pathology and behavioral deficits in a mouse model of Alzheimer's disease. J Neurosci. 2010 Aug 18;30(33):11157-66. PubMed.
  2. . Robust central reduction of amyloid-β in humans with an orally available, non-peptidic β-secretase inhibitor. J Neurosci. 2011 Nov 16;31(46):16507-16. PubMed.
  3. . Effects of TNFalpha-converting enzyme inhibition on amyloid beta production and APP processing in vitro and in vivo. J Neurosci. 2008 Nov 12;28(46):12052-61. PubMed.
  4. . Diurnal patterns of soluble amyloid precursor protein metabolites in the human central nervous system. PLoS One. 2014;9(3):e89998. Epub 2014 Mar 19 PubMed.
  5. . Soluble amyloid precursor protein α and β in CSF in Alzheimer's disease. Brain Res. 2013 Jun 4;1513:117-26. PubMed.

Further Reading


  1. . Targeting the β secretase BACE1 for Alzheimer's disease therapy. Lancet Neurol. 2014 Mar;13(3):319-29. Epub 2014 Feb 17 PubMed.

Primary Papers

  1. . CNS amyloid-β, soluble APP-α and -β kinetics during BACE inhibition. J Neurosci. 2014 Jun 11;34(24):8336-46. PubMed.