Tuck BJ, Miller LV, Katsinelos T, Smith AE, Wilson EL, Keeling S, Cheng S, Vaysburd MJ, Knox C, Tredgett L, Metzakopian E, James LC, McEwan WA. Cholesterol determines the cytosolic entry and seeded aggregation of tau. Cell Rep. 2022 May 3;39(5):110776. PubMed.

Recommends

Please login to recommend the paper.

Comments

  1. Tuck et al. make an important contribution to the problem of tau uptake as it transits between cells. This problem has become more interesting with the recent publication of many tau cryoEM structures that link specific disease phenotypes to tau conformations. Thus, the power of the prion hypothesis: Tau not only templates misfolding in neighboring cells, but does so with very high shape fidelity.

    The starting observation by the McEwan group in this report is that depletion of membrane cholesterol enhances tau uptake. They proceed to show that depletion of the Niemann-Pick C1 protein, which causes Niemann-Pick type C disease and is associated with both mis-sorted cholesterol and tau inclusions, increases tau entry into cultured neurons. Depletion of cholesterol also decreased the concentration of assembled tau needed to induce seeding. Furthermore, cholesterol supplementation was able to protect the cells in the seeding assay. Similar results were obtained in organotypic slice cultures, with the additional observation that the low amounts of tau found in the extracellular space were insufficient to induce seeded aggregation, but did so when membrane cholesterol was reduced even at low nanomolar concentrations.

    The data in this paper will catalyze a more in-depth understanding of the relationship between tau uptake and tau seeding, a relationship linked through the shared effects of cholesterol depletion on both. The authors point to distinct mechanisms of tau uptake in neurons versus non-neural cells which utilize canonical endocytic pathways. In contrast, tau endocytosis in neurons is clathrin- and dynamin-independent but curiously remained dependent on LRP1 and HSPGs.

    A summary of the literature suggests that multiple tau uptake pathways are operational and their relationship to the physical properties conferred by cholesterol on membranes, such as membrane rigidity and thickness, lateral diffusion within membranes, and receptor density is an important next direction. The cell biological direction forward will need to find the intracellular path that tau travels as it ultimately escapes from an endosomal compartment to the cytosol, all the while retaining its folded structure in a form capable of templating endogenous tau to misfold.

    View all comments by Kenneth Kosik

  2. Growing evidence indicates that tau pathology spreads in the brains of AD patients and potentially other tauopathies. One of the proposed mechanisms through which spreading occurs is tau secretion/release followed by uptake into other neurons, where internalized tau aggregates can seed further aggregation in the cytosol of recipient neurons. The latter implies tau seeds have to gain access to the cytosol, although the precise mechanism underlying this process is unclear. In this elegant study, Tuck et al. developed a highly sensitive, split luciferase-based method to measure cytosolic entry of seeding-competent tau assemblies, a critical and yet understudied step in cell-to-cell transmission of tau pathology. This method will be a valuable tool for the field to further understand mechanisms involved in the uptake and cytosolic entry of tau seeds.

    In this new assay, luciferase is reconstituted by binding of two components, LgBiT and HiBiT (Dixon et al., 2016), which are restricted to the cell interior and coupled to exogenous recombinant tau, respectively. To prevent leakage of cytosolic LgBiT into the medium, the authors used a low-activity promoter coupled to a nuclear localization signal to maintain a very low concentration of cytosolic LgBiT. With the optimized protocol, luciferase signals obtained after adding tau-HiBiT were resistant to trypsin treatment, confirming intracellular reconstitution of luciferase activity derived from interactions between tau-HiBiT and LgBiT in the cytosol.

    Using this method, along with seeded tau aggregation as an alternative readout in many experiments, the authors confirmed some earlier findings on tau seeding, including distinct uptake mechanisms exploited by monomeric and aggregated tau, the importance of clathrin-mediated endocytosis and endosome sorting machinery in modulating cytosolic escape of tau seeds in HEK293 cells, and the critical role of HSPGs and LRP1 in the uptake of tau seeds in neurons.

    Importantly, the authors provided convincing data that cytosolic access of tau seeds occurs via different mechanisms in neurons versus HEK293 cells, highlighting the necessity to study this pathological process in the most relevant CNS cell types.

    The central finding of this study is that cholesterol levels in the membranes of mouse and human neurons inversely correlated with cytosolic entry of tau seeds. Treatments that deplete membrane content of cholesterol (e.g., methyl-beta-cyclodextrin treatment or activation of CYP46A1, an enzyme that promotes hydroxylation of cholesterol to 24(S)hydroxycholesterol) invariably promoted tau entry and thus seeding. Conversely, treatments that increase membrane cholesterol (e.g., TopFluor cholesterol loading) substantially reduced tau seeding.

    Other treatments were used and shown to modulate cytosolic entry of tau seeds, but the data interpretation is a little more challenging. For instance, the authors knocked down the endolysosomal cholesterol transporter Niemann-Pick type C 1 (NPC1), which causes an accumulation of free cholesterol in endolysosomes. While this treatment promotes tau seed entry, it is unclear whether it is a depletion of plasma membrane cholesterol or an accumulation of endolysosomal cholesterol that underlies the increased seeding. Treatment with 25-hydroxycholesterol increased tau seed entry, but the reason why treatment with the isomer 24(S)hydroxycholesterol caused the opposite phenotype is unclear, in the absence of assessments of membrane cholesterol levels. Generally, the findings obtained from cultured neurons were replicated in organotypic slice cultures.

    This intriguing mechanistic link between membrane cholesterol and tau seeding offers a potential explanation for the early development of tau pathology in NPC patients carrying loss-of-function mutations in NPC1 or NPC2, the protein that presents endolysosomal cholesterol to NPC1. Prior to this study, one may have speculated that endolysosomal accumulation of free cholesterol in NPC1 patients is the primary contributor to enhanced tau aggregation. However, new data from this study raises the interesting possibility that it may rather be the depletion of plasma membrane cholesterol resulting from the redistribution of free cholesterol into endolysosomes that promotes the cytosolic escape of tau seeds in NPC1 patients, culminating in greatly accelerated tau pathology.

    This interesting study also raises important new questions:

    1. The observation that cholesterol content in the membrane does not modulate cytosolic entry of control protein GFP argues against a general breaching of membrane with cholesterol depletion. Given this, what is the exact mechanism underlying cholesterol-sensitive tau entry? Which specific membrane compartment(s) do tau seeds penetrate to gain cytosolic access? Specifically, is it the plasma membrane or endocytic/lysosomal membranes?
    1. Would pathological tau found in AD brain utilize the same uptake/cytosolic entry mechanism(s) as recombinant tau aggregates? It was previously reported that AD brain-purified tau fibrils show structural and seeding properties that are distinct from recombinant tau fibrils (Guo et al., 2016; Zhang et al., 2019). To answer this question using the split luciferase method, one may need to generate HiBiT-tagged recombinant tau fibrils seeded by AD brain-purified tau fibrils or employ a newly published protocol to assemble recombinant tau fibrils with identical ultrastructure as AD tau fibrils (Lovestam et al., 2022). 
    1. Would membrane cholesterol affect cytosolic entry of α-synuclein aggregates, which also demonstrate prion-like transmission similar to tau? It has been shown that NPC1 patients frequently present α-synuclein pathology in addition to tau pathology and the two types of protein lesions are often found in the same neuron (Saito et al., 2004). 
    1. Rik van der Kant et al. (2019). showed a beneficial effect of CYP46A1 activation on reducing p-tau accumulation in iPSC-differentiated human neurons. AAV-mediated overexpression of CYP46A1 was also found to rescue cognitive deficits in a tauopathy mouse model without modulating hyperphosphorylation of tau (Burlot et al., 2015). In this study, however, CYP46A1 activation with Efavirenz increased cytosolic entry of tau and was thus deemed detrimental. What would be the recommended therapeutic approach on CYP46A1 modulation given the conflicting outcome of CYP46A1 activation?
    1. Several studies demonstrated that ApoE isoforms differentially mediate efflux of cholesterol and other lipids from cultured cells in the order of ApoE2 > ApoE3 >= ApoE4, which inversely correlates with ApoE isoform-dependent risk for AD (Michikawa et al., 2000; Minagawa  et al., 2009) and thus suggests that higher cellular content of cholesterol may increase the risk for AD. These findings appear to be at odds with this new study, which suggests higher cellular cholesterol, at least in the membrane, is protective against tau seeding. What could explain this discrepancy on the risk vs. benefit of cellular cholesterol? Answering this question likely requires quantitative assessments of cholesterol levels in the membrane.
    1. Membrane cholesterol levels have been identified as a key parameter affecting the processing of APP, with cholesterol-rich “raft-like” membranes enhancing the amyloidogenic pathway via BACE1 and presenilins (reviewed in Di Paolo and Kim, 2011). Therefore, membrane cholesterol reduction would facilitate the non-amyloidogenic pathway, while facilitating tau seed uptake and spreading based on this new study. If modulation of membrane cholesterol levels is considered a potential therapeutic approach to limit tau spreading, what effect would it have on APP processing? Does the fact that APP itself binds to membrane cholesterol (Barrett et al., 2012) have any impact on tau uptake and seeding?   

    We look forward to seeing these questions addressed by the field in future studies.

    References:

    Dixon AS, Schwinn MK, Hall MP, Zimmerman K, Otto P, Lubben TH, Butler BL, Binkowski BF, Machleidt T, Kirkland TA, Wood MG, Eggers CT, Encell LP, Wood KV. NanoLuc Complementation Reporter Optimized for Accurate Measurement of Protein Interactions in Cells. ACS Chem Biol. 2016 Feb 19;11(2):400-8. Epub 2015 Dec 10 PubMed.

    Guo JL, Narasimhan S, Changolkar L, He Z, Stieber A, Zhang B, Gathagan RJ, Iba M, McBride JD, Trojanowski JQ, Lee VM. Unique pathological tau conformers from Alzheimer's brains transmit tau pathology in nontransgenic mice. J Exp Med. 2016 Nov 14;213(12):2635-2654. Epub 2016 Oct 17 PubMed.

    Zhang W, Falcon B, Murzin AG, Fan J, Crowther RA, Goedert M, Scheres SH. Heparin-induced tau filaments are polymorphic and differ from those in Alzheimer's and Pick's diseases. Elife. 2019 Feb 5;8 PubMed.

    Lövestam S, Koh FA, van Knippenberg B, Kotecha A, Murzin AG, Goedert M, Scheres SH. Assembly of recombinant tau into filaments identical to those of Alzheimer's disease and chronic traumatic encephalopathy. Elife. 2022 Mar 4;11 PubMed.

    Saito Y, Suzuki K, Hulette CM, Murayama S. Aberrant phosphorylation of alpha-synuclein in human Niemann-Pick type C1 disease. J Neuropathol Exp Neurol. 2004 Apr;63(4):323-8. PubMed.

    van der Kant R, Langness VF, Herrera CM, Williams DA, Fong LK, Leestemaker Y, Steenvoorden E, Rynearson KD, Brouwers JF, Helms JB, Ovaa H, Giera M, Wagner SL, Bang AG, Goldstein LS. Cholesterol Metabolism Is a Druggable Axis that Independently Regulates Tau and Amyloid-β in iPSC-Derived Alzheimer's Disease Neurons. Cell Stem Cell. 2019 Mar 7;24(3):363-375.e9. Epub 2019 Jan 24 PubMed.

    Burlot MA, Braudeau J, Michaelsen-Preusse K, Potier B, Ayciriex S, Varin J, Gautier B, Djelti F, Audrain M, Dauphinot L, Fernandez-Gomez FJ, Caillierez R, Laprévote O, Bièche I, Auzeil N, Potier MC, Dutar P, Korte M, Buée L, Blum D, Cartier N. Cholesterol 24-hydroxylase defect is implicated in memory impairments associated with Alzheimer-like Tau pathology. Hum Mol Genet. 2015 Nov 1;24(21):5965-76. Epub 2015 Sep 10 PubMed.

    Michikawa M, Fan QW, Isobe I, Yanagisawa K. Apolipoprotein E exhibits isoform-specific promotion of lipid efflux from astrocytes and neurons in culture. J Neurochem. 2000 Mar;74(3):1008-16. PubMed.

    Minagawa H, Gong JS, Jung CG, Watanabe A, Lund-Katz S, Phillips MC, Saito H, Michikawa M. Mechanism underlying apolipoprotein E (ApoE) isoform-dependent lipid efflux from neural cells in culture. J Neurosci Res. 2009 Aug 15;87(11):2498-508. PubMed.

    Di Paolo G, Kim TW. Linking lipids to Alzheimer's disease: cholesterol and beyond. Nat Rev Neurosci. 2011 May;12(5):284-96. Epub 2011 Mar 30 PubMed.

    Barrett PJ, Song Y, Van Horn WD, Hustedt EJ, Schafer JM, Hadziselimovic A, Beel AJ, Sanders CR. The amyloid precursor protein has a flexible transmembrane domain and binds cholesterol. Science. 2012 Jun 1;336(6085):1168-71. PubMed.

    View all comments by Jing Guo

  3. I totally agree with the comments of Gilbert Di Paolo and Jing Guo from Denali Therapeutics pointing out that the findings by Tuck and colleagues (namely that the content of plasma membrane cholesterol determines the cytosolic entry and seeded aggregation of tau) do not agree well with some previous studies in the field. Specifically, Rik van der Kant and colleagues (van der Kant et al., 2019) showed that CYP46A1 activation by efavirenz reduces p-tau accumulation in iPSC-differentiated human neurons. Burlot and colleagues (Burlot et al., 2015) demonstrated that CYP46A1 overexpression by gene therapy in the THY-Tau22 mouse model of Alzheimer’s disease-like tau pathology rescues cognitive deficits without affecting tau hyperphosphorylation and associated gliosis. Also, enrichment with cholesterol of plasma membranes was shown to enhance the amyloidogenic pathway (reviewed in Maulik et al., 2013), i.e., it has an opposite effect as compared to tau pathology.

    In their studies, Tuck et al. used cell cultures, a convenient system to conduct some proof-of-principle experiments. Yet, these are isolated systems, which do not reflect the complexity of cholesterol maintenance in the brain and the strong homeostatic mechanisms that are operative in this organ to prevent significant fluctuations of cholesterol in different cell types. The cholesterol content in the cytosol and plasma membranes of cultured neurons was not measured after the treatments that modulated the neuronal cholesterol levels. Hence, it is not clear if the resulting modulations of cellular cholesterol levels were of physiological relevance and if they occur in the Alzheimer’s disease brain. In addition, it would be helpful in the future to generate dependence curves for tau entry versus the plasma membrane cholesterol (or exogenous oxysterol) concentrations, both below and above the sterol content.

    Despite the caveats, the study by Tuck et al. is important because it links plasma membrane cholesterol content with cellular tau entry followed by seeded aggregation. Accordingly, this finding should be kept in mind in future in vivo studies that use treatments affecting brain cholesterol homeostasis. The small-dose, anti-HIV drug efavirenz activates CYP46A1, the major cholesterol-eliminating enzyme in the brain, in a mouse model of Alzheimer’s disease (5XFAD mice) (Mast et al., 2017; Petrov et al., 2019), as well as in a Phase 1 clinical trial in subjects with mild cognitive impairment due to Alzheimer’s disease. In mouse brain, efavirenz was shown to increase both cholesterol elimination by CYP46A1 and brain cholesterol biosynthesis, i.e., it increased brain cholesterol turnover without affecting the whole brain cholesterol levels. Cholesterol turnover determines the rate of sterol flux through plasma membranes, and it was also shown that the brain sterol flux mediated by CYP46A1 affects membrane properties, e.g., cholesterol accessibility, ordering, osmotic resistance, and thickness, and membrane-dependent processes, including glutamate release and protein phosphorylation (Petrov et al., 2020). Accordingly, it would be interesting to study whether the rate of sterol flux affects cellular tau entry in any way.

    No doubt that the work of Tuck et al. will give impetus to many additional studies, which may ultimately enhance our understanding of Alzheimer’s disease and lead to disease-modifying treatment of this devastating brain disorder.

    References:

    Burlot MA, Braudeau J, Michaelsen-Preusse K, Potier B, Ayciriex S, Varin J, Gautier B, Djelti F, Audrain M, Dauphinot L, Fernandez-Gomez FJ, Caillierez R, Laprévote O, Bièche I, Auzeil N, Potier MC, Dutar P, Korte M, Buée L, Blum D, Cartier N. Cholesterol 24-hydroxylase defect is implicated in memory impairments associated with Alzheimer-like Tau pathology. Hum Mol Genet. 2015 Nov 1;24(21):5965-76. Epub 2015 Sep 10 PubMed.

    Mast N, Saadane A, Valencia-Olvera A, Constans J, Maxfield E, Arakawa H, Li Y, Landreth G, Pikuleva IA. Cholesterol-metabolizing enzyme cytochrome P450 46A1 as a pharmacologic target for Alzheimer's disease. Neuropharmacology. 2017 Sep 1;123:465-476. Epub 2017 Jun 24 PubMed.

    Maulik M, Westaway D, Jhamandas JH, Kar S. Role of Cholesterol in APP Metabolism and Its Significance in Alzheimer's Disease Pathogenesis. Mol Neurobiol. 2012 Sep 16; PubMed.

    Petrov AM, Lam M, Mast N, Moon J, Li Y, Maxfield E, Pikuleva IA. CYP46A1 Activation by Efavirenz Leads to Behavioral Improvement without Significant Changes in Amyloid Plaque Load in the Brain of 5XFAD Mice. Neurotherapeutics. 2019 Jul;16(3):710-724. PubMed.

    Petrov AM, Mast N, Li Y, Denker J, Pikuleva IA. Brain sterol flux mediated by cytochrome P450 46A1 affects membrane properties and membrane-dependent processes. Brain Commun. 2020;2(1) Epub 2020 Apr 11 PubMed.

    van der Kant R, Langness VF, Herrera CM, Williams DA, Fong LK, Leestemaker Y, Steenvoorden E, Rynearson KD, Brouwers JF, Helms JB, Ovaa H, Giera M, Wagner SL, Bang AG, Goldstein LS. Cholesterol Metabolism Is a Druggable Axis that Independently Regulates Tau and Amyloid-β in iPSC-Derived Alzheimer's Disease Neurons. Cell Stem Cell. 2019 Mar 7;24(3):363-375.e9. Epub 2019 Jan 24 PubMed.

    View all comments by Irina Pikuleva

  4. We thank the commentators above for their interesting observations on our recent paper. We agree that several important questions are raised by the finding that membrane cholesterol levels determine the membrane-crossing ability of tau. We particularly share the view that determining levels of tau entry under conditions of physiological membrane cholesterol will be critical to determining the contribution of changes in tau entry on disease progression.

    However, we think it is worth discussing comparisons with previous literature discussed above, and we broadly disagree that our findings are necessarily contradictory with some of these studies. The paper by van der Kant et al., 2019, showed that activation of Cyp46A1, the enzyme responsible for 24(s)-HC production, was able to reduce levels of phosphorylated tau in iPSC-derived human neurons. In their models, the tau phospho-epitopes were generated by cell-intrinsic signalling pathways stimulated by duplication of the APP locus. This is inherently a different metric to our study, which measured the entry of exogenously applied tau assemblies to neurons and subsequent formation of insoluble tau aggregates. The two papers’ findings—that efavirenz can increase entry of extracellular tau and reduce intracellular tau phosphorylation in the context of APP duplication—are therefore not mutually exclusive from a mechanistic perspective.

    Whether efavirenz may be beneficial as a treatment will depend on several factors, including the contribution of the mechanisms in question to pathological progression. Also worth considering here is the relatively small effect size of 24(s)-HC and efavirenz treatment on tau entry in our assays. The effects of 24(s)-HC manipulation were of a much smaller magnitude than when cholesterol itself was altered. Targeting Cyp46A1 would therefore seem an inefficient means by which to alter tau entry or its seeded aggregation.

    The study by Burlot and colleagues expressed the Cyp46A1 enzyme in the hippocampi of tau-transgenic mice using AAV vectors. The authors found no effect of Cyp46A1 over-expression on tau pathology. In the light of our data showing that exogenous supply of 24(s)-HC or treatment with efavirenz slightly increased tau entry, it might be expected that tau pathology would be exacerbated by this treatment. In hippocampal slice cultures, we observed large changes in seeded aggregation in response to cholesterol manipulation but comparatively minor effects when 24(s)-HC was altered by its exogenous supply or by efavirenz treatment (Figures 6 and 7). Burlot et al. showed that total cholesterol was unchanged by the AAV-Cyp46A1 treatment, potentially consistent with this interpretation that cholesterol levels are likely more important that 24-HC. These issues again speak to the need for further studies where in vivo sterol levels and spatial localization are correlated to tau entry.

    Finally, the review by Maulik et al. draws on several studies to suggest that high membrane cholesterol is associated with increased activity of the amyloidogenic pathway, likely by promoting clustering of APP with its proteolytic enzyme BACE1 within cholesterol-rich lipid rafts. This again does not conflict with our findings on tau entry, though it may have implications for therapeutic targeting of brain cholesterol levels if cholesterol levels are found to have opposing effects on progression of amyloid and tau pathologies. Potentially consistent with such opposing roles, APOE genotypes differentially confer risk in Alzheimer’s disease versus other tauopathies that do not feature Aβ pathology. For instance, the APOE2 allele, while protective in Alzheimer’s disease, is associated with more severe tau pathology in progressive supranuclear palsy (Zhao et al., 2018). This speaks to the need for a deeper understanding of cholesterol’s role in regulating tau pathology at the cellular level as well as in the whole-organism context. We hope the observation that membrane cholesterol is a determinant of tau entry to the cytosol will help stimulate inquiry here.

    We note recent findings from the Diamond group that also suggest tau assemblies gain access to the cytosol via direct translocation across the plasma membrane (Dodd et al., 2022). We concur that direct transit is currently the best fit of the available data given that tau entry to neurons in our study was independent of classical endocytic machinery.

    References:

    Zhao N, Liu CC, Van Ingelgom AJ, Linares C, Kurti A, Knight JA, Heckman MG, Diehl NN, Shinohara M, Martens YA, Attrebi ON, Petrucelli L, Fryer JD, Wszolek ZK, Graff-Radford NR, Caselli RJ, Sanchez-Contreras MY, Rademakers R, Murray ME, Koga S, Dickson DW, Ross OA, Bu G. APOE ε2 is associated with increased tau pathology in primary tauopathy. Nat Commun. 2018 Oct 22;9(1):4388. PubMed.

    Dodd DA, LaCroix M, Valdez C, Knox GM, Vega AR, Kumar A, Xing C, White CL, Diamond MI. Tau seeds translocate across the cell membrane to initiate aggregation. medRxiv, May 12, 2022

    View all comments by Benjamin Tuck

Make a Comment

To make a comment you must login or register.

This paper appears in the following:

News

  1. Membrane Border Patrol: Cholesterol Stymies Tau Uptake, Aggregation

Therapeutics

  1. Efavirenz