Tier 4 — mechanistic

Fisetin stimulates autophagic degradation of phosphorylated tau via the activation of TFEB and Nrf2 transcription factors

Kim S, Choi KJ, Cho SJ, Yun SM, Jeon JP, Koh YH, Song J, Johnson GVW, Jo C
Scientific Reports 2016 Volume 6, article 24933

Bibliography

PubMed
PMID 27112200
PubMed Central
PMC4844953
DOI
10.1038/srep24933
Funding
Grant from the Ministry of Health and Welfare and Korea National Institute of Health Intramural Research Grant (4845-302-210-13).
Competing interests
The authors declare no competing financial interests.

Study snapshot

DesignIn vitro mechanistic study combining immunoblotting, qRT-PCR, siRNA knockdown, immunocytochemistry, electron microscopy, tandem fluorescent LC3 reporter assay, and pharmacological autophagy-lysosome inhibition to establish the role of TFEB and Nrf2 in fisetin-induced tau clearance.
ModelDoxycycline-inducible mouse cortical neuronal cell line (T4 cells) expressing wild-type human tau; rat primary cortical neurons (embryonic, DIV 10-12); HEK293 cells co-transfected with tau and constitutively active GSK-3β to model sarkosyl-insoluble tau aggregation; human neuronal H4 cells stably expressing mRFP-GFP-LC3 for autophagic flux imaging.
SampleCell-based experiments performed in triplicate (n=3 independent experiments) unless otherwise noted.
InterventionFisetin 2.5-20 µM for 12-36 hours depending on cell type; TFEB and Nrf2 siRNA knockdown; autophagy-lysosome inhibitors bafilomycin A1 (100 nM) and chloroquine (50 µM).
DurationTreatment durations of 12 hours (T4 cortical cells) to 36 hours (primary cortical neurons); longer 24-hour co-treatment windows for autophagy-inhibitor experiments.
EndpointsPhosphorylated tau levels (Ser262 and Ser396/404, by 12E8 and PHF1 antibodies); Sarkosyl-insoluble tau accumulation; Tau kinase (GSK-3β) and phosphatase (PP2A) activity; TFEB and Nrf2 nuclear translocation and downstream target gene expression (ATG9b, LAMP1, NDP52, p62/SQSTM1, HO-1); Autophagy markers (LC3-II, beclin-1, ATG5, ATG7) and autophagic flux; mTORC1 pathway activity (phosphorylated p70S6K and 4E-BP1)

What the study showed, in plain terms

This South Korean neuroscience study investigated whether fisetin can help clear away a toxic, sticky form of the tau protein that builds up inside neurons in Alzheimer's disease — using mouse cortical neuron cell lines and rat primary neurons rather than living animals or people.

Fisetin significantly reduced levels of phosphorylated (chemically modified, aggregation-prone) tau in both cell types, and also reduced a particularly stubborn 'sarkosyl-insoluble' form of tau in a lab model designed to mimic tau aggregation. Crucially, this was not because fisetin blocked the enzymes that add phosphate tags to tau, or boosted the enzymes that remove them — those tau kinase and phosphatase activities were unchanged.

Instead, fisetin worked by revving up the cell's own waste-disposal system: autophagy. It activated two master regulator proteins, TFEB and Nrf2, which together switch on genes for building more lysosomes (the cell's recycling centres) and more autophagy machinery. When the researchers blocked autophagy with drugs, or silenced TFEB or Nrf2 with gene-silencing techniques, fisetin's tau-clearing effect disappeared — confirming autophagy through TFEB and Nrf2 was the actual mechanism.

The paper also found a plausible upstream trigger: fisetin partially inhibited a nutrient-sensing protein complex called mTORC1, which normally keeps TFEB locked out of the nucleus; by dialling down mTORC1, fisetin let TFEB move in and do its job. This is an important piece of mechanistic groundwork for fisetin's neuroprotective reputation, though the authors themselves frame it as evidence fisetin 'should be evaluated further' as a candidate — not as proof it clears tau or improves memory in a living brain.

Key findings

  • Fisetin (5-20 µM, 24-36 hours) significantly and dose-dependently reduced levels of tau phosphorylated at Ser262 and Ser396/404 in both inducible mouse cortical neuronal cells and rat primary cortical neurons, without reducing cell viability at the concentrations used.
  • Fisetin (10 µM, 24 hours) significantly reduced sarkosyl-insoluble tau accumulation in a HEK293 cell model co-expressing tau and constitutively active GSK-3β, a model of pathological tau aggregation.
  • In vitro kinase and phosphatase assays showed fisetin did not alter the activity of tau kinases (GSK-3β, ERK1/2) or the phosphatase PP2A, indicating the reduction in phosphorylated tau is not due to altered phosphorylation/dephosphorylation chemistry.
  • Fisetin increased nuclear localisation of both TFEB and Nrf2 in cortical cells and primary neurons, and increased mRNA of their downstream target genes (ATG9b, LAMP1, NDP52, p62/SQSTM1, heme oxygenase-1).
  • Fisetin increased autophagy markers (LC3-II, beclin-1, ATG5, ATG7), the number of autophagic vesicles by electron microscopy, and autophagic flux (mRFP-GFP-LC3 reporter); autophagy-lysosome inhibitors (bafilomycin A1, chloroquine) blocked fisetin-induced phosphorylated-tau clearance.
  • siRNA knockdown of either TFEB or Nrf2 attenuated fisetin-induced degradation of phosphorylated tau, confirming both transcription factors are required for the effect.
  • Fisetin reduced phosphorylation of mTORC1 downstream targets p70S6 kinase and 4E-BP1, consistent with fisetin-mediated mTORC1 inhibition as a likely trigger for TFEB nuclear translocation and activation.

What this study can and cannot tell us

This study used only cell-culture models (an inducible mouse neuronal cell line, rat primary neurons, and HEK293 cells); there is no in vivo animal model of tauopathy or Alzheimer's disease, and no assessment of whether fisetin improves cognitive or behavioural outcomes.

The tau-aggregation model used (HEK293 cells co-expressing constitutively active GSK-3β) is an artificial overexpression system rather than a disease-relevant neuronal tauopathy model, which may not fully capture how tau aggregates and clears in real neurodegenerative disease.

Fisetin concentrations used (5-20 µM) are, as with most in vitro fisetin work in this Data Center, above what oral fisetin supplementation is expected to achieve in human brain tissue given its poor bioavailability and blood-brain barrier penetration.

Editorial review

Reviewed by the Biohack Blueprint research team

Last verified

Dr. Khaled Mahmoud, MBBChCritical Care Medicine, Longevity Science & Editorial Direction Dr. Amr Yasser, PhDCardiometabolic, Musculoskeletal, Gut & Bioavailability Dr. Abdulla Mostafa, PhDCellular Senescence, Neurological, Mitochondrial & Brain-Ageing Biology