A  feedback loop underlying brain degeneration in Alzheimer’s disease which may explain why so many drug trials have failed has been uncovered by researchers at King’s College London. The study also identifies a clinically approved drug, Fasudil, which breaks the vicious cycle and protects against memory-loss in animal models of Alzheimer’s.

Researchers found that when beta-amyloid destroys a synapse, the nerve cells make more beta-amyloid driving yet more synapses to be destroyed.

Overproduction of the protein beta-amyloid is strongly linked to development of Alzheimer’s disease but many drugs targeting beta-amyloid have failed in clinical trials. Beta-amyloid attacks and destroys synapses — the connections between nerve cells in the brain — resulting in memory problems, dementia and ultimately death.

“We show that a vicious positive feedback loop exists in which beta-amyloid drives its own production. We think that once this feedback loop gets out of control it is too late for drugs which target beta-amyloid to be effective, and this could explain why so many Alzheimer’s drug trials have failed,”

says senior author Dr. Richard Killick from the Institute of Psychiatry, Psychology & Neuroscience (IoPPN).

“Our work uncovers the intimate link between synapse loss and beta-amyloid in the earliest stages of Alzheimer’s disease. This is a major step forward in our understanding of the disease and highlights the importance of early therapeutic intervention,”

says lead author Dr. Christina Elliott from the IoPPN.

Positive Feedback Loop

The researchers also found that a protein called Dickkopf-1 (Dkk1), which potently stimulates production of beta-amyloid, is central to the positive feedback loop. Previous research by Dr. Killick and colleagues identified Dkk1 as a central player in Alzheimer’s, and while Dkk1 is barely detectable in the brains of young adults its production increases as we age.

Instead of targeting beta-amyloid itself, the researchers believe targeting Dkk1 could be a better way to halt the progress of Alzheimer’s disease by disrupting the vicious cycle of beta-amyloid production and synapse loss.

“Importantly, our work has shown that we may already be in a position to block the feedback loop with a drug called fasudil which is already used in Japan and China for stroke. We have convincingly shown that fasudil can protect synapses and memory in animal models of Alzheimer’s, and at the same time reduces the amount of beta-amyloid in the brain,”

says Dr. Killick.


The researchers found that in mice engineered to develop large deposits of beta-amyloid in their brains as they age, just two weeks of treatment with fasudil dramatically reduced the beta-amyloid deposits.

APP overexpression enhanced Wnt-β-catenin signalling

(a) APP overexpression enhanced Wnt-β-catenin signalling as measured by TCF/LEF transcriptional activity using SuperTOPflash luciferase reporter assay in HEK293A cells; activity was measured in relative light units (RLU). The enhancement of Wnt-β-catenin signalling was only detectable in cells also overexpressing Wnt3a (a “canonical Wnt”) and not in the presence of the “non-canonical Wnt”, Wnt5a.
(b) APP overexpression also enhanced Wnt-PCP signalling as measured by AP1 transcriptional activity (APRE-luciferase) in HEK239A cells. The enhancement of Wnt-PCP signalling was only detectable in cells also overexpressing Wnt5a and not in the presence of Wnt3a.
(c) APP co-immunoprecipitated from lysates of primary rat cortical neurons with Wnt-β-catenin co-receptor protein, LRP6 and with Wnt-PCP co-receptor protein, Vangl2. (The full blots are presented in the supplemental data.)
(d) Schematic illustrating the sites of APP-LRP6 interaction determined by peptide array hybridisation. Overlay of APP peptides with recombinant extracellular domain (ECD) of LRP6 demonstrated binding between the LRP6 and APP ECDs.
(e) Schematic illustrating the site of APP-Vangl2 interaction as elucidated by peptide array. Overlay of APP peptides with HEK293A lysates overexpressing Vangl2-HA demonstrated binding of Vangl2 to a single site in the APP-ICD.
(f) Schematic of the precise binding regions of LRP6/Vangl2 binding within the APP sequence. Binding of LRP6 is restricted to the ECD with one region lying close to the site affected by the Swedish APP mutation. In contrast, Vangl2 binding at the APP-ICD occurs at the YENPTY motif, which has been shown to control APP internalisation.
(g) Overexpression of LRP6 enhanced the activity of APP in Wnt-β-catenin signalling as measured by TCF/LEF transcriptional activity using SuperTOPflash luciferase reporter assay in HEK293A cells.
(h) Overexpression of Vangl2 enhanced the activity of APP in Wnt-PCP signalling as measured by APRE luciferase reporter assay in HEK293A cells.
Credit: Christina Elliott et al, CC-BY

Researchers at King’s College London are now seeking funding to run a clinical trial in early stage sufferers of Alzheimer’s to determine if fasudil improves brain health and prevents cognitive decline.

“As well as being a safe drug, fasudil appears to enter the brain in sufficient quantity to potentially be an effective treatment against beta-amyloid. We now need to move this forward to a clinical trial in people with early stage Alzheimer’s disease as soon as possible,”

Professor Dag Aarsland from the IoPPN said.

Fasudil (INN) is a potent Rho-kinase inhibitor and vasodilator. Since it was discovered, it has been used for the treatment of cerebral vasospasm, which is often due to subarachnoid hemorrhage, as well as to improve the cognitive decline seen in stroke patients. It has been found to be effective for the treatment of pulmonary hypertension.

It was demonstrated in February 2009 that fasudil could improve memory in normal mice, identifying the drug as a possible treatment for age related or neurodegenerative memory loss.

It is approved for use in Japan and China, but at the time of this writing has not been approved by the United States Food and Drug Administration or by the European Medicines Agency.


In Alzheimer’s disease (AD), the canonical Wnt inhibitor Dickkopf-1 (Dkk1) is induced by β-amyloid (Aβ) and shifts the balance from canonical towards non-canonical Wnt signalling. Canonical (Wnt-β-catenin) signalling promotes synapse stability, while non-canonical (Wnt-PCP) signalling favours synapse retraction; thus Aβ-driven synapse loss is mediated by Dkk1. Here we show that the Amyloid Precursor Protein (APP) co-activates both arms of Wnt signalling through physical interactions with Wnt co-receptors LRP6 and Vangl2, to bi-directionally modulate synapse stability. Furthermore, activation of non-canonical Wnt signalling enhances Aβ production, while activation of canonical signalling suppresses Aβ production.

Together, these findings identify a pathogenic-positive feedback loop in which Aβ induces Dkk1 expression, thereby activating non-canonical Wnt signalling to promote synapse loss and drive further Aβ production. The Swedish familial AD variant of APP (APPSwe) more readily co-activates non-canonical, at the expense of canonical Wnt activity, indicating that its pathogenicity likely involves direct effects on synapses, in addition to increased Aβ production. Finally, we report that pharmacological inhibition of the Aβ-Dkk1-Aβ positive feedback loop with the drug fasudil can restore the balance between Wnt pathways, prevent dendritic spine withdrawal in vitro, and reduce Aβ load in vivo in mice with advanced amyloid pathology. These results clarify a relationship between Aβ accumulation and synapse loss and provide direction for the development of potential disease-modifying treatments.

The research was supported by a Medical Research Council grant. The Wolfson Drug Discovery Unit receives core support from the UK National Institute for Health Research Biomedical Research Centre and Unit Funding Scheme. The CIBERNED received funding from the Spanish Ministry of Economy and Competitiveness, and an MRC Centres of Excellence in Neurodegeneration (COEN) Pathfinder award, a part of the EU-Joint Program for Neurodegenerative diseases.

Christina Elliott, Ana I. Rojo, Elena Ribe, Martin Broadstock, Weiming Xia, Peter Morin, Mikhail Semenov, George Baillie, Antonio Cuadrado, Raya Al-Shawi, Clive G. Ballard, Paul Simons & Richard Killick
A role for APP in Wnt signalling links synapse loss with β-amyloid production
Translational Psychiatry volume 8, Article number: 179 (2018)