BACE inhibitors, with their ability to squelch Aβ production, seemed like a promising preventative therapy for Alzheimer’s disease. Unfortunately, their trials foundered, with some of them worsening cognition in an apparent class effect (Dec 2017 conference news; Nov 2018 conference newsJul 2019 conference news). Hoping the class can yet be salvaged, some scientists suggest that a low dose could nudge down Aβ while sparing cognition (Dec 2019 conference news). 

Researchers led by Henrik Zetterberg at the University of Gothenburg, Sweden, now say this just might work. In the May 26 Alzheimer’s Research & Therapy, they report that inhibiting Aβ production below a 50 percent threshold in cultured neurons harmed synaptic transmission, while inhibiting it by less left synaptic function intact.

Joint first authors Tugce Munise Satir and Lotta Agholme treated cultures of primary rat cortical neurons with the BACE inhibitor LY2886721, BACE Inhibitor IV, or lanabecestat, at concentrations of 0.04, 0.3, or 3 μM. All three of these inhibitors were developed by Eli Lilly. LY2886721 was discontinued due to liver toxicity, while BACE Inhibitor IV was never tested in clinical trials, and lanabecestat was stopped in Phase 3 because of lack of efficacy and hints of cognitive harm.

Four days after a single application of inhibitor, the researchers measured secreted Aβ40 and Aβ42, and evaluated synaptic activity using an optical electrophysiology assay developed by biotech company Cellectricon in Mölndal, Sweden. In this assay, a calcium indicator is added to a culture of synaptically connected neurons. Neurons in one spot are stimulated with an electric field, and the researchers measure the response from connected cells.

Lanabecestat was most potent. It suppressed Aβ production by 70 percent, but also dampened synaptic activity by 15 percent, at all three doses tested. LY2886721 was milder. At the highest dose, it lowered Aβ by 57 percent and synaptic transmission by 39 percent. At the lower two doses, however, it cut Aβ levels by a bit less than 50 percent on average, and did not harm synaptic signaling. This implies that inhibition of less than 50 percent might preserve neuronal signaling, at least in this short-term paradigm, the authors note. The finding complements mouse studies that found low doses to be benign, while high doses damaged synaptic plasticity (Nov 2014 news). 

Intriguingly, BACE Inhibitor IV was the least harmful to synapses. At the highest dose, it dropped Aβ production by more than 60 percent and muted electrical transmission by 16 percent. However, the middle dose, while still suppressing Aβ by about 60 percent, did not affect synaptic signaling. At the lowest dose, Aβ production fell by about 40 percent, while synaptic signaling actually strengthened by 13 percent. BACE Inhibitor IV is reputed to be selective for BACE1 over BACE2, while the other two are not (Stachel et al., 2004). 

In addition, BACE inhibition may promote α cleavage of amyloid precursor protein, the authors noted. The sAPPα fragment has been found to protect neurons, perhaps explaining the better synaptic transmission (Mockett et al., 2017). 

The Icelandic variant of APP lowers a person’s lifetime Aβ production by 30 percent and wards off AD, suggesting that low levels of inhibition over long time periods are sufficient for protection (Jul 2012 news; Jun 2017 news). However, it is unclear how early such therapy would have to be started. “It would also be interesting to explore if combinations of low-dose BACE inhibitors could result in additive effects on Aβ reduction without causing additive synaptic dysfunction,” the authors proposed.—Madolyn Bowman Rogers


  1. Ever since the origin of Aβ was shown to be the result of proteolytic processing of the juxta-membranous and intramembranous domains of a beta-amyloid precursor protein (APP) (Kang et al., 1987), it has been apparent that this phenomenon must reflect some basic physiological function of APP and its superfamily partners (APLP1 and APLP2). Given its high neuronal expression and synaptic localization, most studies have concluded that the APP family plays a fundamental role in synaptic plasticity underlying learning and memory. Is it surprising, therefore, that high levels (more than 50 percent) of inhibition of either N-terminal (α- and β-secretases) and C-terminal (γ-secretase) processing should cause cognitive adverse effects?

    Our studies (Roberts et al., 2017) indicated that over the 20-year natural history of AD (preclinical and prodromal stages), approximately 5 mg of Aβ peptide accumulates in the average AD brain. This represents a rate of accumulation of 28 ng/hour, which is approximately 5 percent of normal hourly production (Potter et al., 2013; Patterson et al., 2015). From this we would argue that if an inhibitor of production (either β- or γ-secretase) is used early enough in the pathogenesis of AD, a rate of inhibition on the order of 5 percent to 10 percent should be sufficient to prevent the abnormal accumulation of Aβ. Of course, higher levels of inhibition would be needed for interventions commencing at later stages of AD. In these circumstances, it would be logical to use an Aβ clearing agent (immunotherapy) followed by low dose inhibition maintenance therapy, close to the 5 percent levels of abnormal rates of accumulation.

    Most compounds in the modern pharmacopoeia are toxic if given in sufficiently high doses. It’s good to see these BACE inhibitor (BACEi) in vitro results of Satir et al., which prove this point, and offer hope for re-examination of the potential benefits of these therapeutic strategies. Most of the clinical trials of BACEi have been operating at more than 70 percent to 80 percent inhibition levels. The biomarkers for Aβ-PET and p-tau changed in the right directions after relatively brief exposures. We must move forward and determine safe dosage levels, probably as a combination therapy of improved clearance followed by low dose maintenance inhibition of production.


    . The precursor of Alzheimer's disease amyloid A4 protein resembles a cell-surface receptor. Nature. 1987 Feb 19-25;325(6106):733-6. PubMed.

    . Biochemically-defined pools of amyloid-β in sporadic Alzheimer's disease: correlation with amyloid PET. Brain. 2017 May 1;140(5):1486-1498. PubMed.

    . Increased in vivo amyloid-β42 production, exchange, and loss in presenilin mutation carriers. Sci Transl Med. 2013 Jun 12;5(189):189ra77. PubMed.

    . Age and amyloid effects on human central nervous system amyloid-beta kinetics. Ann Neurol. 2015 Sep;78(3):439-53. Epub 2015 Jul 20 PubMed.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Verubecestat Negative Trial Data: What Does it Mean for BACE Inhibition?
  2. Bump in the Road or Disaster? BACE Inhibitors Worsen Cognition
  3. Cognitive Decline Trips Up API Trials of BACE Inhibitor
  4. Picking Through the Rubble, Field Tries to Salvage BACE Inhibitors
  5. At High Doses, BACE1 Inhibitors Hinder Synaptic Plasticity in Mice
  6. Protective APP Mutation Found—Supports Amyloid Hypothesis
  7. More Support for Amyloid Hypothesis: Protective APP Mutation Lowers Aβ in Blood

Therapeutics Citations

  1. LY2886721
  2. Lanabecestat

Paper Citations

  1. . Structure-based design of potent and selective cell-permeable inhibitors of human beta-secretase (BACE-1). J Med Chem. 2004 Dec 16;47(26):6447-50. PubMed.
  2. . Therapeutic Potential of Secreted Amyloid Precursor Protein APPsα. Front Mol Neurosci. 2017;10:30. Epub 2017 Feb 7 PubMed.

Further Reading

Primary Papers

  1. . Partial reduction of amyloid β production by β-secretase inhibitors does not decrease synaptic transmission. Alzheimers Res Ther. 2020 May 26;12(1):63. PubMed.