. Activation of PPARα enhances astroglial uptake and degradation of β-amyloid. Sci Signal. 2021 Oct 26;14(706):eabg4747. PubMed.


Please login to recommend the paper.


  1. Raha et al. investigated the therapeutic potential of gemfibrozil and retinoic acid combination in Alzheimer’s disease, which is exciting since GFB is already FDA-approved for hyperlipidemia and RA is a derivative of vitamin A. The authors reported that the combination therapy reduces Aβ plaques and improves cognitive functions in 5XFAD mice. They also performed comprehensive mechanistic studies both in vitro and in vivo to delineate the mechanism underlying these effects. GFB-RA induced Aβ uptake and degradation in astrocytes. Raha et al. further identified the specific pathways involved in these processes. In their previous study, they had demonstrated that GFB-RA upregulates lysosomal biogenesis by inducing TFEB and that this mechanism is dependent on PPARα (Ghosh et al., 2015). In this new study, the authors investigated whether this PPARα pathway could be responsible for the Aβ uptake and degradation.

    To address this important question, Raha et al. generated an astrocyte-specific PPARα knockout mouse model. Using this new mouse model, they reported that the TFEB-PPARα pathway was responsible for Aβ degradation in astrocytes. Although Figure S7 clearly demonstrated the reduction of PPARα in astrocytes, it is unclear whether there was any reduction in other cell types. Many commonly used Gfap-Cre lines are known to have issues with leaky expression. Because the exact mouse line information was not provided in this article, it will strengthen this study if astrocyte-specific PPARα deletion can be corroborated by co-staining other cell types.

    Raha et al. also demonstrated that GFB-RA induces the expression of low-density lipoprotein receptor (Ldlr) in a PPARα-dependent manner in astrocytes and this mechanism underlies the Aβ uptake, strengthening the role of LDLR in astrocytic Aβ uptake as demonstrated earlier (Basak et al., 2012). We and David Holtzman’s group also demonstrated that overexpression of Ldlr reduces Aβ accumulation and tau-associated neurodegeneration, pointing to the therapeutic potential of LDLR in AD (Castellano et al., 2012; Kim et al., 2009; Shi et al., 2021). In addition, we and Peter Tontonoz’s group identified a potentially druggable target, an inducible degrader of the LDL receptor (IDOL) (Choi et al., 2015). It will be interesting to determine whether the GFB-RA combination therapy also affects IDOL or any other regulators of LDLR.

    Given the beneficial effect of LDLR overexpression on tau-associated neurodegeneration, it would be important to determine in future studies if this combination therapy could alleviate tau pathology, as well. Not only Ldlr but Tfeb overexpression has beneficial effects on tauopathy (Polito et al., 2014). Furthermore, RA was reported to reduce tau hyperphosphorylation and improve learning and memory functions in a tauopathy mouse model (Ding et al., 2008). 

    Although the effects of GFB-RA on astrocytes were clearly demonstrated, it is possible that these drugs can alter other cell types and pathways, which may contribute to the observed phenotypes in 5XFAD mice. Indeed, GFB-mediated PPARα activation can reduce Aβ production by inducing non-amyloidogenic processing of amyloid precursor protein (Corbett et al., 2015). RA was also shown to reduce Aβ plaques and the activation of microglia and astrocytes (Ding et al., 2008). Therefore, it appears that GFB-RA combination therapy may induce multiple beneficial pathways in several cell types.

    Taken together, a GFB-RA combination therapy may be a promising strategy for the treatment of AD. However, it remains to be determined whether combining GFB and RA has any advantage over monotherapy on AD-related pathologies. Based on the authors’ earlier study, there was no significant difference between combination vs. individual therapy on Tfeb expression in astrocytes (Ghosh et al., 2015). They also recently demonstrated the beneficial effects of GFB alone, even with a slightly lower dose, on AD-related pathologies in 5XFAD mice at the same age used in this study (Chandra and Pahan, 2019). GFB alone significantly reduced Aβ plaques, microgliosis, and astrogliosis, and improved cognitive functions in 5XFAD mice.

    Therefore, it is necessary to determine whether the GFB-RA combination therapy has any synergistic or additive effects on alleviating the disease pathology or decreasing side effects. For example, the authors found a reduction in astrocyte number with GFB alone treatment in the earlier study, whereas they did not observe this reduction with GFB-RA treatment in 5XFAD mice. Therefore, RA might have a compensatory effect on astrogliosis, but this warrants further studies.

    Importantly, an earlier clinical trial with GFB alone failed in the prevention of early stage treatment of AD although there was a trend in reduction of Aβ42 and phospho-tau in the cerebrospinal fluid of GFB-treated patients (Gemfibrozil). It would be important to determine if GFB-RA combination therapy has more beneficial effects for the treatment of AD. However, more rigorous animal studies are warranted given the limitations of this first combination trial with a 5xFAD mouse model.


    . Low-density lipoprotein receptor represents an apolipoprotein E-independent pathway of Aβ uptake and degradation by astrocytes. J Biol Chem. 2012 Apr 20;287(17):13959-71. PubMed.

    . Low-density lipoprotein receptor overexpression enhances the rate of brain-to-blood Aβ clearance in a mouse model of β-amyloidosis. Proc Natl Acad Sci U S A. 2012 Sep 18;109(38):15502-7. Epub 2012 Aug 27 PubMed.

    . Gemfibrozil, a Lipid-Lowering Drug, Lowers Amyloid Plaque Pathology and Enhances Memory in a Mouse Model of Alzheimer's Disease via Peroxisome Proliferator-Activated Receptor α. J Alzheimers Dis Rep. 2019 May 18;3(1):149-168. PubMed.

    . The E3 ubiquitin ligase Idol controls brain LDL receptor expression, ApoE clearance, and Aβ amyloidosis. Sci Transl Med. 2015 Nov 18;7(314):314ra184. PubMed.

    . Activation of peroxisome proliferator-activated receptor α stimulates ADAM10-mediated proteolysis of APP. Proc Natl Acad Sci U S A. 2015 Jul 7;112(27):8445-50. Epub 2015 Jun 15 PubMed.

    . Retinoic acid attenuates beta-amyloid deposition and rescues memory deficits in an Alzheimer's disease transgenic mouse model. J Neurosci. 2008 Nov 5;28(45):11622-34. PubMed.

    . Activation of peroxisome proliferator-activated receptor α induces lysosomal biogenesis in brain cells: implications for lysosomal storage disorders. J Biol Chem. 2015 Apr 17;290(16):10309-24. Epub 2015 Mar 6 PubMed.

    . Overexpression of low-density lipoprotein receptor in the brain markedly inhibits amyloid deposition and increases extracellular A beta clearance. Neuron. 2009 Dec 10;64(5):632-44. PubMed.

    . Selective clearance of aberrant tau proteins and rescue of neurotoxicity by transcription factor EB. EMBO Mol Med. 2014 Jul 28;6(9):1142-60. PubMed.

    . Overexpressing low-density lipoprotein receptor reduces tau-associated neurodegeneration in relation to apoE-linked mechanisms. Neuron. 2021 Aug 4;109(15):2413-2426.e7. Epub 2021 Jun 21 PubMed.

    View all comments by Hande Karahan
  2. This interesting paper shows the therapeutic potentials of gemfibrozil (GFB) and retinoic acid (RA) to treat Alzheimer’s disease (AD). The authors' finding indicates the significant contribution of astrocytic autophagy and lysosomal degradation to Aβ clearance in the disease.

    While both GFB and RA are small compounds that modulate lipid metabolism, GFB-RA has been shown in this paper to mediate its effect through PPARα-LDLR. Since PPARα and LDLR play a critical role in maintaining lipid homeostasis, future studies should define how lipids are involved in the therapeutic mechanisms.

    Particularly, it would be important to dissect the roles of apolipoprotein E, as GFB and RA might increase ApoE production. Whereas ApoE receptors including LDLR and LRP1 mediate the cellular uptake of Aβ, our knowledge is still limited regarding if ApoE is predominantly involved in the Aβ clearance through those receptors in astrocytes. It is also important to explore how APOE genotype influences the effect of GFB-RA on ameliorating AD-related phenotypes in future studies.

    In addition, their findings may implicate a link between astrocytic lipids and AD pathogenesis. Since astrocytes possibly communicate with other brain cell types including neurons and microglia through lipid metabolism, the therapeutic effects of GFB-RA may be induced through multiple mechanisms beyond astrocytic Aβ degradation.

    View all comments by Takahisa Kanekiyo
  3. This group previously published (Ghosh et al, 2015) that a combination of the PPARα agonist gemfibrozil combined with retinoic acid (GFB/RA) allowed low doses of these drugs to be used to increase lysosomal abundance. This new study shows the action of this drug combination as a potential therapeutic in an Alzheimer’s disease (AD) model.

    Gemfibrozil’s mechanism of action is not fully understood—as well as a PPARα agonist, it is a CYP2C8 inhibitor and also acts to inhibit some membrane transporters. However, the previous study, as well as this new study, used a PPARα knockout mouse to show that PPARα plays a key role in this action. Particularly impressive in this new paper is that they knock out astrocytic PPARα in the 5xFAD mouse and show this blocks the action of GFB/RA to reduce Aβ plaques and improve cognition in this AD model.

    Retinoic acid (and related molecules), when used by itself, has a number of the actions described in this new paper, such as reduction of Aβ and improvement of cognition, so it is surprising that this is dependent on PPARα. Although retinoic acid can, in fact, bind to PPAR, it is not a ligand for PPARα, only PPAR β/δ. Perhaps the dependence on PPARα for the GFB/RA combination is because, as the authors describe, only low doses of the drugs are used. However it is difficult to know the dose of retinoic acid used because, in the paper, it is expressed in International Units (IU). Retinoic acid RA doses are seldom expressed as IU, and the amount per body weight is not given. Personally I don’t understand how it can be expressed in IU, as vitamin A is, because retinoic acid cannot completely replace the action of vitamin A.

    That gemfibrozil and retinoic acid are already in clinical use is an advantage with this approach, but more on the mechanism would be useful to refine this. For instance, could an RXR agonist replace retinoic acid (such as bexarotene, already proposed as an AD treatment)? The earlier idea was that retinoic acid would synergize with a PPARα agonist because retinoic acid (specifically its 9-cis isomer) is a ligand for the RXR partner to which PPARα binds as a heterodimer in its action to regulate transcription. Hence it was assumed that the addition of retinoid acid, some of it isomerizing to 9-cis retinoic acid, would further promote the action of gemfibrozil on PPARα.

    In a similar line of thinking, would a drug other than gemfibrozil, which was more specific for PPARα, be active? 

    Although not discussed to any extent, what might be the relationship of the action of these drugs on lipid metabolism? Gemfibrozil will reduce triglyceride levels and decrease the risk of hyperlipidemia, whereas retinoids can have the opposite effect and elevate triglycerides. What will be the net effect in combination?


    . Activation of peroxisome proliferator-activated receptor α induces lysosomal biogenesis in brain cells: implications for lysosomal storage disorders. J Biol Chem. 2015 Apr 17;290(16):10309-24. Epub 2015 Mar 6 PubMed.

    View all comments by Peter McCaffery
  4. Raha et al. provide new fuel for the Aβ hypothesis in that regimens that increase Aβ clearance benefit brain pathology and cognition. Using a combination of low-dose gemfibrozil and retinoic acid, which the authors had previously shown to induce lysosomal biogenesis and autophagy through PPARα-mediated transcription of TFEB, the role of this type of astrocytic PPARα activation was investigated in the context of AD. The study is a tour de force, with eight multiple-pane figures in the main text and nine supplementary figures.

    While this study did not clarify which species of Aβ clearance is requisite—monomer and oligomer, mainly oligomer, or oligomer and fibrillar—for the pathological and cognitive outcomes, its strength is the relatively benign treatment regimen using a vitamin supplement with an FDA-approved cholesterol drug that achieved in vivo results relatively quickly and initiated during disease for this model.

    The in vitro and ex vivo data tell us that astrocytes can be stimulated to clear Aβ, both soluble and associated with plaque deposits. The resulting multidimensional cognitive benefits offer optimism that human studies may improve quality of life.

    Using a semi-quantitative imaging assay for fluorophore-tagged Aβ uptake and degradation, it was shown that mouse primary astrocytes exhibit enhanced FAM- and HF-647-tagged Aβ1-42 uptake and degradation with GFB-RA treatment compared to vehicle control. Appropriate controls determined that GFB-RA treatment increased not only lysosomal biogenesis but also activity and autophagic flux. In addition, Aβ uptake was specific to astrocytes rather than contaminating microglia. Uptake resulted in Aβ localization to lysosomes/endosomes (LysoTracker), likely occurred via heparan sulfate proteoglycans (heparin competition) on the surface of astrocytes, and required both PPARα and PPARβ using GFAP-cre-driven knockout astrocyte cultures for these PPAR isotypes.

    Cultures were incubated for up to eight hours at 37°C; this suggests that the peak uptake timepoint of 4 hours reflects uptake of soluble monomeric and oligomeric fluorophore conjugated Aβ (Jungbauer et al., 2009). Unfortunately, the aggregation status of the tagged Aβ was not documented in the figures provided.

    The work then migrated to in vivo studies using control and 5XFAD mice (male and female, 6 months old), given 60 days oral treatment with GFB-RA or vehicle. Multiple measures were conducted to verify that hippocampus lysosomal biogenesis, as per TFEB, LAMP2, TPP1, p62, LC3 immunoblot and immunohistostaining, occurred in treated 5XFAD mice leading to reduced Aβ plaque load; measures included Thio-S staining for plaque number, area, perimeter, as well as 82e1 and 6E10 immunoblot densitometry analysis of full-length APP as well as CTF and Aβ fragments. Immunoblot analysis of APP and CTF remained unchanged by the treatment in 5XFAD mice; whereas Aβ was significantly reduced. Reduction of plaque number, size, and perimeter along with a reduction in Aβ, further supported that GFB-RA treatment enhanced Aβ clearance, likely misfolded but possibly monomeric forms as well.

    The supplementary figures show that additional studies in these 5XFAD cohorts determined that GFB-RA treatment increased hippocampal lysosomal biogenesis, insulin-degrading enzyme level, and decreased hippocampal astroglial inflammation and activation morphology via PPARα.

    To test the role of astrocytic PPARα in vivo, 5XFAD mice crossed with PPARα astrocyte KO (PPARαastroKO ) mice were insensitive to GFB-RA treatment, in that lysosomal biogenesis markers, total Aβ, and measures of plaque number, area, and size remained unchanged compared to treated 5XFAD with intact astrocytic PPARα. Additional controls were included, e.g., nontransgenic, vehicle-treated 5XFAD, untreated 5XFAD-PPARαastroKO.

    Finally, cognitive function was tested using this same genetic cohort to demonstrate that GFB-RA improved memory in 5XFAD via astrocytic PPARα. Barnes maze, T-maze, and novel object recognition were performed to evaluate spatial, associative, working, and non-spatial memory. Open field test was used to determine that locomotion was not affected by genotype or treatment.


    . Preparation of fluorescently-labeled amyloid-beta peptide assemblies: the effect of fluorophore conjugation on structure and function. J Mol Recognit. 2009 Sep-Oct;22(5):403-13. PubMed.

    View all comments by Kelly Dineley
  5. It will be critical to characterize the contributions astrocytes make to clearance of Aβ and other aggregates, either by autophagy or related processes such as "LANDO" (Heckmann et al., 2019). It will also be important to determine the relevant mechanisms, including coordinated induction of transcriptional programs.

    Nevertheless, the focus on TFEB is somewhat surprising. Nearly all CNS TFEB is expressed in oligodendrocytes; astrocytes express very little, as indicated by several single-cell transcript databases, such as the Allen Brain Institute's Transcriptomics Explorer (Allen Brain Map) and the resource provided by the lab of the late Ben Barres (Brain RNA-Seq). This may explain why the effects Raha et al. obtained with TFEB siRNA were rather meager. The more abundant member of the TFEB family found in astrocytes is TFE3 (Clarke et al., 2018). It should be noted that the Abcam catalog entry for the anti-TFEB antibody used here contains a warning of cross-reactivity with TFE3.

    It is also important to note that ApoE4 has been reported by multiple laboratories to specifically bind the "CLEAR" DNA sequence motifs through which TFEB and related transcription factors act (Theendakara et al., 2016; Parcon et al., 2018; Lima et al., 2020). As astrocytes produce the bulk of CNS ApoE, this competitive inhibition of autophagy-related gene induction by ApoE4 provides a novel mechanism to explain the link of Alzheimer's disease to APOE genetics.


    . LC3-Associated Endocytosis Facilitates β-Amyloid Clearance and Mitigates Neurodegeneration in Murine Alzheimer's Disease. Cell. 2019 Jul 25;178(3):536-551.e14. Epub 2019 Jun 27 PubMed.

    . Normal aging induces A1-like astrocyte reactivity. Proc Natl Acad Sci U S A. 2018 Feb 20;115(8):E1896-E1905. Epub 2018 Feb 7 PubMed.

    . Direct Transcriptional Effects of Apolipoprotein E. J Neurosci. 2016 Jan 20;36(3):685-700. PubMed.

    . Apolipoprotein E4 inhibits autophagy gene products through direct, specific binding to CLEAR motifs. Alzheimers Dement. 2018 Feb;14(2):230-242. Epub 2017 Sep 22 PubMed.

    . Electrochemical detection of specific interactions between apolipoprotein E isoforms and DNA sequences related to Alzheimer's disease. Bioelectrochemistry. 2020 Jun;133:107447. Epub 2019 Dec 23 PubMed.

    View all comments by Steve Barger
  6. I thank Dr. Barger for his important comment about TFEB and TFE3. The immunogen for the Abcam antibody is a synthetic peptide corresponding to Human TFEB aa 2-14 (N terminal). Sequence: ASRIGLRMQLMRE.

    To understand this issue in a better way, we used the blast alignment tool from NCBI to do a protein sequence alignment using the mouse TFE3 sequence from Uniprot (ID Q 64092) and aligned it with the immunogen from Abcam (Ab2636) consisting of 13 amino acid synthetic peptide taken from the N terminal end. We found a 61.53 percent homology match which is quite low.

    Query  110  SSRVLLRQQLMR  121
    +SR  + LR   QLMR

    This Abcam antibody has been successfully used by many investigators. However, there is only one warning about TFEB antibody cross-reacting with mouse TFE3 and MITF in overexpressing cell lines. Perhaps, if cells are made to overexpress TFE3, there is a chance of cross-reactivity at high antibody concentration. This is also not published.

    Moreover, we have seen that the combination of gemfibrozil and retinoic acid at low doses stimulates the transcription of TFEB gene in pure cultured astrocytes via increasing the recruitment of PPARα and RXRα (Ghosh et al., 2015). We have also observed a similar finding with aspirin (Chandra et al., 2018). 

    Therefore, although the level of TFEB is low in astrocytes, it can be stimulated via activation of the PPARα/RXRα pathway.


    . Activation of peroxisome proliferator-activated receptor α induces lysosomal biogenesis in brain cells: implications for lysosomal storage disorders. J Biol Chem. 2015 Apr 17;290(16):10309-24. Epub 2015 Mar 6 PubMed.

    . Aspirin Induces Lysosomal Biogenesis and Attenuates Amyloid Plaque Pathology in a Mouse Model of Alzheimer's Disease via PPARα. J Neurosci. 2018 Jul 25;38(30):6682-6699. Epub 2018 Jul 2 PubMed.

    View all comments by Kalipada Pahan
  7. Amidst these provocative and sophisticated findings, no author was certain about the mechanism. Yet, not one scientist mentioned, even as a possibility, that viruses or other microorganisms might be playing some role.

    Thus, I reluctantly add this material to the "shunning" of microbes pile. Yet, numerous papers report on the possible antimicrobial role of retinoic acid and gemfibrozil (see below).

    For example: I predicted earlier this year that the salutary effect of some common food supplements on cognition may be due to their inhibitory effect on herpes virus influences in the brain. This was confirmed at Tufts by D. Cairns et al. (personal communication).


    . Statins and fibrates for the treatment of hyperlipidaemia in HIV-infected patients receiving HAART. AIDS. 2003 Apr 11;17(6):851-9. PubMed.

    . The p150 Isoform of ADAR1 Blocks Sustained RLR signaling and Apoptosis during Influenza Virus Infection. PLoS Pathog. 2020 Sep;16(9):e1008842. Epub 2020 Sep 8 PubMed.

    . Retinoic Acid Improves the Recovery of Replication-Competent Virus from Latent SIV Infected Cells. Cells. 2020 Sep 11;9(9) PubMed.

    . Retinoic Acid Modulates Hyperactive T Cell Responses and Protects Vitamin A-Deficient Mice against Persistent Lymphocytic Choriomeningitis Virus Infection. J Immunol. 2020 Jun 1;204(11):2984-2994. Epub 2020 Apr 13 PubMed.

    . Inhibition of herpes simplex virus replication by retinoic acid. Antiviral Res. 1997 Jan;33(2):117-27. PubMed.

    . Cholesterol reducing agents inhibit assembly of type I parainfluenza viruses. Virology. 2017 Jan 15;501:127-135. Epub 2016 Dec 2 PubMed.

    View all comments by Leslie Norins

Make a Comment

To make a comment you must login or register.

This paper appears in the following:


  1. Aβ Clearance—The Untapped Potential of Astrocytes?