FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4229028075> ?p ?o ?g. }

• W4229028075 abstract "Article Figures and data Abstract Editor's evaluation Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Elevation of soluble wild-type (WT) tau occurs in synaptic compartments in Alzheimer’s disease. We addressed whether tau elevation affects synaptic transmission at the calyx of Held in slices from mice brainstem. Whole-cell loading of WT human tau (h-tau) in presynaptic terminals at 10–20 µM caused microtubule (MT) assembly and activity-dependent rundown of excitatory neurotransmission. Capacitance measurements revealed that the primary target of WT h-tau is vesicle endocytosis. Blocking MT assembly using nocodazole prevented tau-induced impairments of endocytosis and neurotransmission. Immunofluorescence imaging analyses revealed that MT assembly by WT h-tau loading was associated with an increased MT-bound fraction of the endocytic protein dynamin. A synthetic dodecapeptide corresponding to dynamin 1-pleckstrin-homology domain inhibited MT-dynamin interaction and rescued tau-induced impairments of endocytosis and neurotransmission. We conclude that elevation of presynaptic WT tau induces de novo assembly of MTs, thereby sequestering free dynamins. As a result, endocytosis and subsequent vesicle replenishment are impaired, causing activity-dependent rundown of neurotransmission. Editor's evaluation This study provides an interesting new insight into the synaptic disease mechanisms of tauopathies. The paper is based on a technically very rigorous dataset indicating that increased levels of soluble Tau impair pre-synaptic endocytosis and, consequently, neurotransmission by sequestering Dynamin-1 on microtubules. The findings are of major relevance for basic neuronal cell biology and translational neuroscience alike. https://doi.org/10.7554/eLife.73542.sa0 Decision letter Reviews on Sciety eLife's review process Introduction The microtubule (MT)-binding protein tau assembles and stabilizes MTs (Brunden et al., 2009; Lee et al., 2011) mainly in axonal compartments (Binder et al., 1985; Kubo et al., 2019). Phosphorylation of tau proteins reduces their binding affinity (Biernat et al., 1993; Drechsel et al., 1992), thereby shifting the equilibrium from MT-bound form to soluble free form (Ballatore et al., 2007). Soluble tau proteins also exist in dynamic equilibrium between phosphorylated and dephosphorylated forms (Spillantini and Goedert, 2013) as well as between soluble and aggregated forms. When the cytosolic tau concentration is elevated, monomeric tau undergoes oligomerization and eventually precipitates into neurofibrillary tangles (NFT) (Greenberg and Davies, 1990; Grundke-Iqbal et al., 1986; Yoshida and Ihara, 1993), which is a hallmark of tauopathies, including Alzheimer’s disease (AD), frontotemporal dementia with Parkinsonism-17 (FTDP-17), and progressive supranuclear palsy (Lee et al., 2011; Ballatore et al., 2007; Spillantini and Goedert, 2013). Although the NFT density can correlate with the degree of AD progression (Lee et al., 2011; Ballatore et al., 2007; Morris et al., 2011), soluble tau protein levels are more closely linked to disease progression and cognitive decline (Götz et al., 2008; Koss et al., 2016). Genetic ablation of tau shows little abnormal phenotype (Harada et al., 1994; Vossel et al., 2010; Yuan et al., 2008), presumably due to compensation by other MT-associated proteins (Harada et al., 1994). Instead, tau ablation can prevent amyloid β-induced impairments of mitochondrial transport (Vossel et al., 2010) or memory defects (Ittner et al., 2010; Roberson et al., 2007). Thus, loss of tau function due to its dissociation from MTs is unlikely to be an important cause of neuronal dysfunction in AD (Spillantini and Goedert, 2013; Morris et al., 2011). In postmortem brains of both AD patients and intact humans, tau is present in synaptosomes (Fein et al., 2008; Tai et al., 2012). In a transgenic mice AD model, soluble tau is accumulated in the hippocampal nerve terminal zone (de Calignon et al., 2012; Liu et al., 2012). Both in in vivo and in culture models of tauopathy, tau is released from axon terminals upon KCl stimulation in a Ca2+-dependent manner, like neurotransmitters (Pooler et al., 2013; Yamada et al., 2014). Tau oligomers produced by released tau triggers endogenous tau seeding in neighboring neurons, thereby causing trans-synaptic propagations (de Calignon et al., 2012; Guo and Lee, 2011). FTDP tauopathy model mice that overexpressed with mutant tau are widely used to examine tau toxicities on synaptic plasticity (Polydoro et al., 2014; Sydow et al., 2011; Yoshiyama et al., 2007), memory formation (Sydow et al., 2011; Santacruz et al., 2005), as well as on synaptic vesicle (SV) transport (McInnes et al., 2018; Zhou et al., 2017). In contrast to FTDP, which is a rare familial disease associated with tau mutation, AD is a widespread sporadic disease unassociated with tau mutation, but the expression level of WT tau being crucial. As AD models, the effects of WT tau overexpression have been examined in culture cells (Ebneth et al., 1998; Shahpasand et al., 2012; Stamer et al., 2002; Thies and Mandelkow, 2007) or in Drosophila (Mudher et al., 2004), where impaired axonal transports associated with increased MT density were found. These observations suggest that WT tau can be detrimental when its levels are elevated (Stamer et al., 2002; Thies and Mandelkow, 2007). However, unlike FTDP tau mutant, it is unknown whether elevated soluble WT tau can affect mammalian central synaptic transmission. We addressed this question using the giant nerve terminal calyx of Held visualized in slices from mice brainstem, where axonal MTs extended into the depth of terminals (Piriya Ananda Babu et al., 2020). In this presynaptic terminal, we loaded recombinant WT h-tau from a whole-cell patch pipette at fixed concentrations to model the elevation of WT tau associated with AD and found that WT h-tau newly assembled MTs and strongly impaired synaptic transmission. Capacitance measurements indicated that the primary target of WT h-tau is vesicle endocytosis. Immunocytochemical image analysis after cell permeabilization revealed an increase in the MT-bound fraction of the endocytic GTPase dynamin in WT h-tau-loaded terminals. Since the endocytic key protein dynamin is an MT-binding protein (Shpetner and Vallee, 1989), dynamin is likely sequestered by newly assembled MTs. Out of screening, we found that a synthetic dodecapeptide corresponding to amino acids 560–571 of dynamin 1 inhibited MT-dynamin interaction. When we co-loaded this peptide ‘PHDP5’ with WT h-tau, its toxicities on vesicle endocytosis as well as on synaptic transmission were rescued. Thus, we propose a novel synaptic dysfunction mechanism underlying AD, in which WT tau-induced over-assembly of MTs depletes dynamins, thereby impairing vesicle endocytosis and synaptic transmission. Results Intra-terminal loading of WT h-tau impairs excitatory synaptic transmission To address whether elevation of soluble h-tau in presynaptic terminals can affect synaptic transmission, we purified WT recombinant h-tau (0N4R) and its deletion mutant (del-MTBD) lacking the MT-binding site (244Gln–367Gly) (Figure 1—figure supplement 1A), obtained using an Escherichia coli expression system (Xie et al., 2014). These recombinant h-tau proteins are highly soluble at room temperature (RT) without any sign of granulation (Maeda et al., 2007). In simultaneous pre- and postsynaptic recording at the calyx of Held in mouse brainstem slices, we recorded EPSCs evoked at 1 Hz by presynaptic action potentials (Figure 1). After confirming stable EPSC amplitude for 10 min, we injected a large volume of internal solution containing WT h-tau (20 µM) from an installed fine tube to a presynaptic whole-cell pipette to replace most of pipette solution and allow h-tau to diffuse into a presynaptic terminal (illustration in Figure 1A; Hori et al., 1999; Takahashi et al., 2012). After loading h-tau (20 µM), the amplitude of glutamatergic EPSCs gradually declined and reached 23% ± 9% in 30 min (Figure 1A, p < 0.01, paired t-test, n = 6 synapses in 6 slices). WT h-tau loaded at a lower concentration (10 µM) caused a slower EPSC rundown to 65% ± 5% in 30 min (p < 0.01, n = 5 synapses in 5 slices). Del-MTBD (20 µM), lacking tubulin polymerization capability (Figure 1—figure supplement 1B), likewise loaded had no effect on EPSC amplitude (Figure 1A). Since h-tau concentrations in presynaptic terminals are equilibrated with those in a presynaptic whole-cell pipette with a much greater volume than terminals (Pusch and Neher, 1988), these results suggest that WT h-tau >10 µM can significantly impair excitatory synaptic transmission. Figure 1 with 1 supplement see all Download asset Open asset Wild-type (WT) human tau (h-tau) loaded in presynaptic terminals inhibited excitatory synaptic transmission. (A) In simultaneous pre- and postsynaptic whole-cell recordings, intra-terminal infusion of WT h-tau at 10 µM (blue filled circles) or 20 µM (black filled circles), from a tube in a presynaptic patch pipette (top illustration), caused a concentration-dependent rundown of EPSCs evoked by presynaptic action potentials at 1 Hz. In the time plots, EPSC amplitudes averaged from 60 events are sampled for data points and normalized to the mean amplitude of baseline EPSCs before h-tau infusion. Sample records of EPSCs 5 min before (i) and 30 min after (ii) tau infusion are superimposed and shown on the left panels. The EPSC amplitude remaining 30 min after infusion was 23% ± 9% and 65% ± 5%, respectively, for 10 and 20 µM h-tau (means and SEMs, 6 synapses from 6 slices, p < 0.01 in paired t-test between before and after h-tau infusion). Infusion of microtubule (MT)-binding site-deleted h-tau mutant (del-MTBD, 20 µM, Figure 1—figure supplement 1A) had no effect on the EPSC amplitude (open circles, sample EPSC traces shown on the left bottom panel). (B) The amplitude of EPSCs evoked at 0.1 Hz remained unchanged after h-tau infusion (85% ± 12%, 5 synapses from 5 slices, p = 0.22 in paired t-test). Sample records of EPSCs before (i) and 30 min after (ii) h-tau infusion at 0.1 Hz are superimposed on the left panel. (C) Taxol (1 µM) caused activity-dependent rundown of EPSC amplitude to 41.4% ± 12% at 1 Hz (p < 0.01, 5 synapses from 5 slices), but remained unchanged when stimulated at 0.1 Hz (105% ± 3.0%, open circles, 4 synapses from 4 slices). Sample records of EPSCS at 0.1 and 1 Hz are superimposed on the left panels. Figure 1—source data 1 Wild-type (WT) h-tau loaded in presynaptic terminals inhibited excitatory synaptic transmission. https://cdn.elifesciences.org/articles/73542/elife-73542-fig1-data1-v2.xlsx Download elife-73542-fig1-data1-v2.xlsx The inhibitory effect of WT h-tau on EPSCs was apparently frequency-dependent. When evoked at 0.1 Hz, WT h-tau (20 µM) caused only a minor reduction of EPSC amplitude (to 85% ± 12%, 30 min after loading, p = 0.21, n = 5; Figure 1B). Since taxol shares a common binding site of MTs with tau (Kar et al., 2003) and assembles tubulins into MTs (Figure 1—figure supplement 1B), we tested the effect of taxol (1 µM) on EPSCs (Figure 1C). Like h-tau, taxol caused a significant rundown of EPSCs evoked at 1 Hz (to 41 ± 12 at 30 min, n = 5, p < 0.05), but not those evoked at 0.1 Hz (104% ± 3.0%, n = 4, p = 0.60). These results together suggest that MTs newly assembled in presynaptic terminals by WT h-tau or taxol cause activity-dependent rundown of excitatory synaptic transmission. WT h-tau primarily inhibits SV endocytosis and secondarily exocytosis To determine the primary target of h-tau causing synaptic dysfunction, we performed membrane capacitance measurements at the calyx of Held (Eguchi et al., 2017; Sun and Wu, 2001; Wang et al., 2020; Yamashita et al., 2005). Since stray capacitance of perfusion pipettes prevents capacitance measurements, we backfilled h-tau into a conventional patch pipette after preloading normal internal solution only at its tip to secure GΩ seal formation. This caused substantial and variable delays of the intra-terminal diffusion, so no clear effect could be seen more than 10 min after whole-cell patch membrane was ruptured. Twenty minutes after whole-cell patch loading of WT h-tau (20 µM), endocytic capacitance showed a significant slowing (Figure 2), whereas exocytic capacitance magnitude (ΔCm) or charge of Ca2+ currents (QCa) induced by a depolarizing pulse was not different from controls without h-tau loading. Thirty minutes after loading h-tau, the endocytic capacitance change became further slowed (p < 0.01), and exocytic ΔCm eventually showed a significant reduction (p < 0.05, n = 5) without a change in QCa. These results suggest that the primary target of h-tau toxicity is SV endocytosis. Endocytic block inhibits recycling replenishment of SVs via recycling, thereby reducing the exocytic release of neurotransmitter as a secondary effect. Figure 2 Download asset Open asset Inhibition of synaptic vesicle (SV) endocytosis is the primary effect of wild-type (WT) human tau (h-tau) loading. (A) Exo-endocytic membrane capacitance changes in presynaptic terminals without (Control) or after direct loading of WT h-tau (20 µM; Tau). WT h-tau was directly loaded by diffusion into a terminal from a whole-cell patch pipette (illustration). Capacitance traces were sampled from (i) 10, (ii) 20, and (iii) 30 min after patch membrane rupture (superimposed). Left panel, non-loading control. Right panel, WT h-tau-loaded terminal. Capacitance changes were evoked every 2 min by Ca2+ currents induced by a 20 ms depolarizing pulse (not shown). (B) Time plots of endocytic rate (left panel), exocytic magnitude (middle panel), and presynaptic Ca2+ current charge (right panel). Data points represent averaged values from five events from 4 min before and 4 min after the time points. In calyceal terminals, 20 min after patch membrane rupture with a pipette containing WT h-tau (filled circles; Tau), endocytic rate was significantly prolonged (*p < 0.05 compared to controls, open circles, repeated-measures two-way ANOVA with post hoc Scheffe test, n = 5 from 5 slices), whereas exocytic magnitude remained similar to controls (p = 0.45). Thirty minutes after rupture, endocytic rate was further prolonged (**p < 0.01) and exocytic magnitude became significantly less than controls (*p < 0.05). Ca2+ current charge (QCa) remained unchanged throughout recording. Figure 2—source data 1 Inhibition of synaptic vesicle (SV) endocytosis is the primary effect of wild-type (WT) human tau (h-tau) loading. https://cdn.elifesciences.org/articles/73542/elife-73542-fig2-data1-v2.xlsx Download elife-73542-fig2-data1-v2.xlsx Inhibition of SV endocytosis and synaptic transmission by WT h-tau requires de novo MT assembly Since new MT assembly might take place after h-tau loading (Figure 1, Figure 1—figure supplement 1), we tested whether the tubulin polymerization blocker nocodazole might reverse the toxic effects of h-tau on SV endocytosis and synaptic transmission. In tubulin polymerization assays, nocodazole inhibited h-tau-dependent MT assembly in a concentration-dependent manner, with a maximal inhibition reached at 20 µM (Figure 3A). In presynaptic capacitance measurements, nocodazole (20 µM) co-loaded with h-tau (20 µM) fully prevented the h-tau toxicities on endocytosis (Figure 3B) and synaptic transmission (Figure 3C). Nocodazole alone (20 µM) had no effect on exo-endocytosis (Figure 3B) or EPSC amplitude (Figure 3C). It is highly likely that WT h-tau loaded in calyceal terminals newly assembled MTs, thereby impairing SV endocytosis and synaptic transmission. Figure 3 Download asset Open asset The microtubule (MT) assembly blocker nocodazole prevented tau-induced block of synaptic vesicle (SV) endocytosis and EPSC rundown. (A) Concentration-dependent inhibitory effects of nocodazole on MT assembly in tubulin polymerization assay. MT assembly by 0N4R human tau (h-tau) (20 µM) in the absence (pink symbols and a fitting line) or presence of nocodazole at 1 µM (blue), 10 µM (red), 20 µM (purple), and 50 µM (orange). Data points and error bars in all graphs represent means and SEMs (n = 3). (B) Nocodazole prevented h-tau-induced block of SV endocytosis. Presynaptic membrane capacitance changes (superimposed traces) 25 min after loading h-tau alone (20 µM, red trace), h-tau and nocodazole (20 µM, blue), nocodazole alone (20 µM, green), and controls with no loading (black). Bar graphs indicate endocytic rates in non-loading controls (Ctr, black, 6 terminals from 6 slices), h-tau-loaded terminals (Tau, red, 8 terminals from 8 slices), co-loading of nocodazole with h-tau (N + T, blue, 7 terminals from 7 slices), and nocodazole alone (Noc, green, 8 terminals from 8 slices). Nocodazole co-loading fully prevented endocytic block by h-tau (**p < 0.01, between Tau and N-T) to control level (one-way ANOVA with Scheffe post hoc test). (C) Nocodazole prevented EPSC rundown caused by wild-type (WT) h-tau. Nocodazole (20 µM) co-loaded with WT h-tau (20 µM) prevented EPSC rundown (filled circles, 4 synapses from 4 slices, **p < 0.01, unpaired t-test). Data of WT h-tau effect on EPSCs (Figure 1A) is shown as a red dashed line for comparison. Nocodazole alone (20 µM) had no effect on EPSC amplitude throughout (open circles, 4 synapses from 4 slices). Figure 3—source data 1 The microtubule (MT) assembly blocker nocodazole prevented tau-induced block of synaptic vesicle (SV) endocytosis and EPSC rundown. https://cdn.elifesciences.org/articles/73542/elife-73542-fig3-data1-v2.xlsx Download elife-73542-fig3-data1-v2.xlsx WT h-tau assembles MTs and sequesters dynamins in calyceal terminals The monomeric GTPases dynamins 1 and 3 play critical roles in the endocytic fission of SVs (Hinshaw and Schmid, 1995; Raimondi et al., 2011; Takei et al., 1995). Since dynamin is originally discovered as an MT-binding protein (Shpetner and Vallee, 1989), we hypothesized that newly assembled MTs might trap free dynamins in cytosol. If this is the case, MT-bound form of dynamin would be increased. To test this hypothesis, we performed immunofluorescence microscopy and image analysis to quantify MTs and dynamin. After whole-cell infusion of h-tau into calyceal terminals, slices were chemically fixed and permeabilized to allow cytosolic-free molecules such as tubulin monomers to be washed out of the terminal, thereby enhancing the signals from large structures such as MTs or MT-bound molecules. Fluorescent h-tau antibody identified calyceal terminals loaded with WT h-tau (20 µM, Figure 4A). Double staining with mouse β3-tubulin antibody revealed a 2.1-fold increase in MT signals in h-tau-loaded terminals, compared with those without h-tau loading (p = 0.01, n = 5, two-tailed unpaired t-test with Welch’s correction, Figure 4B). Triple labeling with dynamin antibodies further revealed a 2.6-fold increase in dynamin signal (p = 0.01, n = 5, two-tailed t-test with Welch’s correction, Figure 4B). In super-resolution imaging, dynamins are shown in clusters along MTs in tau-loaded calyceal terminal (Figure 4—figure supplement 1). These results suggest that soluble WT h-tau can assemble MTs in presynaptic terminals, thereby sequestering cytosolic dynamins that are indispensable for SV endocytosis. Figure 4 with 2 supplements see all Download asset Open asset Wild-type (WT) human tau (h-tau) assembled microtubules (MTs) and increased bound-form dynamins in calyceal terminals. (A) Immunofluorescence images of brainstem slices showing loaded WT h-tau (green, left, arrowhead) labeled with anti h-tau/Alexa Fluor 488 antibodies (green, left panel), newly assembled MTs labeled with anti-β3-tubulin/Alexa Fluor 647 antibodies (magenta, middle), and increased bound-form dynamin labeled with anti-dynamin 1/Alexa Fluor 568 antibodies (red, right panel). (B) Bar graphs showing immunofluorescence intensities of h-tau (green), β3-tubulin (magenta), and dynamin (red) relative to controls with no loading (black bars). WT h-tau loading significantly increased β3-tubulin (p = 0.0105) and dynamin 1 (p = 0.0109) intensity in terminals compared to control terminals without WT h-tau loading (n = 5 terminals from 5 slices for each data set, two-tailed unpaired t-test with Welch’s correction; *p < 0.05, ***p < 0.001). Figure 4—source data 1 Raw immunofluorescence images from Figure 4A. https://cdn.elifesciences.org/articles/73542/elife-73542-fig4-data1-v2.pptx Download elife-73542-fig4-data1-v2.pptx Figure 4—source data 2 Data from Figure 4B. https://cdn.elifesciences.org/articles/73542/elife-73542-fig4-data2-v2.xlsx Download elife-73542-fig4-data2-v2.xlsx Besides dynamins, MTs can bind to various other proteins. Among them, formin mDia can bind to MTs (Bartolini and Gundersen, 2010) and involved in the endocytic scaffold functions together with F-actin, intersectin, and endophilin. Although acute depolymerization of F-actin (Piriya Ananda Babu et al., 2020; Eguchi et al., 2017) or genetic ablation of intersectin (Sakaba et al., 2013) has no effect on SV endocytosis at the calyx of Held, the formin mDia inhibitor SMFH2 reportedly inhibits endocytosis at the calyx terminals in pre-hearing rats (postnatal days [P] 8–12) (Soykan et al., 2017). We re-examined whether the drug might inhibit SV endocytosis at calyceal terminals in slices from post-hearing mice (P13–14). SMFH2 slightly prolonged SV endocytosis, but this effect was statistically insignificant (Figure 4—figure supplement 2A). Thus, formin unlikely makes substantial contribution to the marked endocytic slowing observed after intra-terminal tau loading (Figure 2). It may also be argued that binding of endophilin to MTs (Schuske et al., 2003) might cause EPSC rundown since endophilin is involved in clathrin uncoating (Watanabe et al., 2018), which is required for SV refilling with glutamate. If SV refilling during recycling is impaired, miniature EPSCs are decreased in amplitude and frequency (Takami et al., 2017). However, neither amplitude nor frequency was affected by intra-terminal loading of tau (20 µM) (Figure 4—figure supplement 2B). Thus, endophilin-MT binding unlikely underlies EPSC rundown by intra-terminal tau infusion (Figure 1). An MT-dynamin-binding inhibitor peptide attenuates h-tau toxicities on SV endocytosis and synaptic transmission To prevent toxic effects of h-tau on endocytosis and transmission, we searched for a dominant-negative (DN) peptide blocking MT-dynamin binding. Since the MT-binding domain of dynamins is unknown, we synthesized 11 peptides from the pleckstrin-homology (PH) domain and 11 peptides from the proline-rich domain of dynamin 1 (Figure 4—figure supplement 1A) and submitted them to the MT-dynamin 1-binding assay. Out of 22 peptides, one peptide corresponding to the amino acid sequence 560–571 of PH domain, which we named ‘PHDP5’, significantly inhibited the MT-dynamin 1 interaction (Figure 5, Figure 5—figure supplement 1B and C). By SYPRO orange staining, dynamin 1 is found as an ~100 kDa band, 1.7% ± 0.4% in precipitates (ppts). In the presence of MT, dynamin 1 in ppts increased to 22.6% ± 2.4%, indicating sequestration of dynamin 1 by MTs. When PHDP5 was added to MT and dynamin 1, dynamin 1 in the ppt fraction decreased to 6.3% ± 2.4%, indicating that PHDP5 works as a DN peptide for inhibiting MT-dynamin interactions (Figure 5A). Figure 5 with 2 supplements see all Download asset Open asset Dynamin 1 pleckstrin-homology (PH) domain peptide inhibited microtubule (MT)-dynamin 1 binding and prevented endocytic slowing and EPSC rundown caused by wild-type (WT) human tau (h-tau). (A) Top, partial amino acid sequence of PH domain of mouse dynamin 1 indicating the sequence of the synthetic dodecapeptide PHDP5 (560–571). Left, SDS-PAGE of MT-dynamin 1-binding assay. S, supernatant; P, precipitates. Dyn1, dynamin 1. Right, quantification of MT-dynamin 1 interaction. The bars indicate the percentage of dynamin 1 found in precipitates relative to total amount. PHDP5 significantly inhibited MT-dynamin 1 interaction (**p < 0.01, n = 3). (B) Presynaptic membrane capacitance records (superimposed) after loading h-tau alone (20 µM, red trace, taken from Figure 4B), h-tau co-loaded with DPHP5 (0.25 mM, blue) or scrambled DPHP5 (SDPHP5, green). DPHP5 alone (0.25 mM, black trace, 7 terminals from 7 slices) had no effect on capacitance changes compared to non-loading terminal controls (taken from Figure 3B). Bar graphs of endocytic rates (middle panel) indicate significant difference (*p < 0.05, 7 terminals from 7 slices) between tau (red bar, 8 terminals from 8 slices) and DPHP5 + tau (blue, 7 terminals from 7 slices) as well as between SDPHP5 + tau (blue) and DPHP5 + tau (green, n = 6 terminals from 6 slices). The magnitudes of exocytic capacitance changes were not significantly different between the groups, recorded 25 min after rupture. (C) DPHP5 attenuated h-tau-induced EPSC rundown. The EPSC rundown after h-tau infusion (20 µM, red dashed line; data taken from Figure 1A) was attenuated by DPHP5 (1 mM) co-loaded with h-tau (filled circles) but not by scrambled DPHP5 peptide (SDPHP5, open triangles, 1 mM). DPHP5 alone (1 mM, open circles) had no effect on EPSC amplitude throughout. Bar graphs indicate EPSC amplitude (normalized to that before infusion) 30 min after infusion. Significant difference (**p < 0.01) between tau and tau + DPHP5, between tau + DPHP5 and tau + SDPHP5. The difference between DPHP5 alone and DPHP5 + tau was not significant (p = 0.09), indicating the partial antagonistic effect of DPHP5 against h-tau-induced EPSC rundown. Figure 5—source data 1 Data from Figure 5A. https://cdn.elifesciences.org/articles/73542/elife-73542-fig5-data1-v2.xlsx Download elife-73542-fig5-data1-v2.xlsx Figure 5—source data 2 Data from Figure 5B and C. https://cdn.elifesciences.org/articles/73542/elife-73542-fig5-data2-v2.xlsx Download elife-73542-fig5-data2-v2.xlsx Figure 5—source data 3 Images from Figure 5A. https://cdn.elifesciences.org/articles/73542/elife-73542-fig5-data3-v2.pptx Download elife-73542-fig5-data3-v2.pptx A cryo-electron microscope study on dynamin 1 assembled on lipid membrane has revealed that PH domain is tucked up into dynamin structure in apo state, but upon GTP binding, exposed toward membrane by a conformational change (Kong et al., 2018). In negatively stained electron micrographs, dynamin 1 is periodically arranged on the surface of MTs (La et al., 2020), suggesting a helical polymerization like in dynamin-membrane interaction (Zhang and Hinshaw, 2001). Therefore, PH domain including the putative-binding site PHDP5 is likely exposed toward MT surface. To examine whether PH domain of dynamin 1 can directly bind to MTs, immunofluorescence labeled MTs and glutathione transferase-tagged PH domain (GST-PH) were mixed and observed by confocal and electron microscopy (Figure 5—figure supplement 2). In confocal microscopic imaging, GST-PH co-localized with MTs, in contrast to controls, where MTs were mixed with GST alone (Figure 5—figure supplement 2A). These results were further confirmed in electron microscopic imaging, showing co-localizations of MTs and GST-PH (Figure 5—figure supplement 2B). Thus, dynamin 1 PH domain can associate with MTs, although it remains to be determined whether PHDP5 can directly bind to MTs. Loading of PHDP5 (0.25 mM) alone in calyceal terminals had no effect on exo-endocytic capacitance changes, but when co-loaded with WT h-tau (20 µM), it significantly attenuated the h-tau-induced endocytic slowing (p < 0.05, Figure 5B). Scrambled PHDP5 peptide (0.25 mM) loaded as a control had no effect on h-tau-induced endocytic slowing. Like its effect on capacitance changes, intra-terminal infusion of PHDP5 alone (1 mM) did not affect EPSC amplitude, but when co-loaded with WT h-tau (20 µM), significantly attenuated the inhibitory effect of h-tau on EPSC amplitude (p < 0.01, Figure 5C). Co-infusion of scrambled PHDP5 (1 mM) with h-tau (20 µM) did not affect the h-tau-induced EPSC rundown (p = 0.46). These results further support that WT h-tau causes dynamin deficiency via new assembly of MTs thereby impairing SV endocytosis and synaptic transmission. These results also highlight PHDP5 as a potential therapeutic tool for rescuing synaptic dysfunctions associated with AD or Parkinson’s disease (PD). Discussion Using the calyx of Held in brainstem slices as an AD model for dissecting mammalian central excitatory synaptic transmission, we demonstrated that intra-terminal loading of WT h-tau impairs vesicle endocytosis and synaptic transmission via de novo MT assembly. Previous overexpression studies in cultured cells reported MT assembly by injection or overexpression of WT tau (Thies and Mandelkow, 2007; Drubin and Kirschner, 1986; Shemesh et al., 2008) or phosphorylated tau (Shahpasand et al., 2012; Liu et al., 2007). Compared with overexpression, our whole-cell method allows targeted loading of molecules in presynaptic terminals at defined concentrations because of a large pipette-to-cell volume ratio (Pusch and Neher, 1988). In postmortem brain tissue homogenates from AD patients, soluble tau content is estimated as 6 ng/μg of protein, which is eight times higher than controls (Khatoon et al., 1992). Assuming protein contents in brain homogenate as 10%, 60 kDa tau concentration in AD patients’ brain is estimated as 10 µM. Since elevation of soluble tau concentration likely occurs mainly in axons and axon terminal compartments of neurons, soluble tau conc" @default.
• W4229028075 created "2022-05-08" @default.
• W4229028075 creator A5001925265 @default.
• W4229028075 creator A5003029675 @default.
• W4229028075 creator A5023921886 @default.
• W4229028075 creator A5037301114 @default.
• W4229028075 creator A5056421274 @default.
• W4229028075 creator A5063750053 @default.
• W4229028075 creator A5067433430 @default.
• W4229028075 creator A5077954170 @default.
• W4229028075 creator A5088174279 @default.
• W4229028075 creator A5091097806 @default.
• W4229028075 date "2022-04-06" @default.
• W4229028075 modified "2023-09-26" @default.
• W4229028075 title "Author response: Microtubule assembly by tau impairs endocytosis and neurotransmission via dynamin sequestration in Alzheimer’s disease synapse model" @default.
• W4229028075 doi "https://doi.org/10.7554/elife.73542.sa2" @default.
• W4229028075 hasPublicationYear "2022" @default.
• W4229028075 type Work @default.
• W4229028075 citedByCount "1" @default.
• W4229028075 countsByYear W42290280752023 @default.
• W4229028075 crossrefType "peer-review" @default.
• W4229028075 hasAuthorship W4229028075A5001925265 @default.
• W4229028075 hasAuthorship W4229028075A5003029675 @default.
• W4229028075 hasAuthorship W4229028075A5023921886 @default.
• W4229028075 hasAuthorship W4229028075A5037301114 @default.
• W4229028075 hasAuthorship W4229028075A5056421274 @default.
• W4229028075 hasAuthorship W4229028075A5063750053 @default.
• W4229028075 hasAuthorship W4229028075A5067433430 @default.
• W4229028075 hasAuthorship W4229028075A5077954170 @default.
• W4229028075 hasAuthorship W4229028075A5088174279 @default.
• W4229028075 hasAuthorship W4229028075A5091097806 @default.
• W4229028075 hasBestOaLocation W42290280751 @default.
• W4229028075 hasConcept C127445978 @default.
• W4229028075 hasConcept C1491633281 @default.
• W4229028075 hasConcept C156407911 @default.
• W4229028075 hasConcept C169760540 @default.
• W4229028075 hasConcept C170493617 @default.
• W4229028075 hasConcept C185592680 @default.
• W4229028075 hasConcept C200170125 @default.
• W4229028075 hasConcept C20418707 @default.
• W4229028075 hasConcept C28005876 @default.
• W4229028075 hasConcept C55493867 @default.
• W4229028075 hasConcept C86803240 @default.
• W4229028075 hasConcept C95444343 @default.
• W4229028075 hasConceptScore W4229028075C127445978 @default.
• W4229028075 hasConceptScore W4229028075C1491633281 @default.
• W4229028075 hasConceptScore W4229028075C156407911 @default.
• W4229028075 hasConceptScore W4229028075C169760540 @default.
• W4229028075 hasConceptScore W4229028075C170493617 @default.
• W4229028075 hasConceptScore W4229028075C185592680 @default.
• W4229028075 hasConceptScore W4229028075C200170125 @default.
• W4229028075 hasConceptScore W4229028075C20418707 @default.
• W4229028075 hasConceptScore W4229028075C28005876 @default.
• W4229028075 hasConceptScore W4229028075C55493867 @default.
• W4229028075 hasConceptScore W4229028075C86803240 @default.
• W4229028075 hasConceptScore W4229028075C95444343 @default.
• W4229028075 hasLocation W42290280751 @default.
• W4229028075 hasOpenAccess W4229028075 @default.
• W4229028075 hasPrimaryLocation W42290280751 @default.
• W4229028075 hasRelatedWork W1984628987 @default.
• W4229028075 hasRelatedWork W1985953812 @default.
• W4229028075 hasRelatedWork W1990950863 @default.
• W4229028075 hasRelatedWork W2020336867 @default.
• W4229028075 hasRelatedWork W2025383253 @default.
• W4229028075 hasRelatedWork W2264730682 @default.
• W4229028075 hasRelatedWork W4229028075 @default.
• W4229028075 hasRelatedWork W4229333130 @default.
• W4229028075 hasRelatedWork W4285346917 @default.
• W4229028075 hasRelatedWork W4293168447 @default.
• W4229028075 isParatext "false" @default.
• W4229028075 isRetracted "false" @default.
• W4229028075 workType "peer-review" @default.