Synaptotoxic effects of extracellular tau are mediated by its microtubule-binding region

Abstract

Immunotherapies targeting extracellular tau share the premise that interrupting cell-to-cell spread of tau pathology in Alzheimer's disease (AD) will slow dementia pathogenesis. Whether these interventions affect the actions of synaptotoxic, extracellular tau species that may contribute to cognitive impairment is relatively unknown. Here, we assayed synaptic plasticity disruption in anaesthetised live rats caused by intracerebral injection of synaptotoxic tau present either in (a) secretomes of induced pluripotent stem cell-derived neurons (iNs) from people with Trisomy 21, the most common genetic cause of AD, or (b) aqueous extracts of human AD brain. Extracellular tau in iN secretomes was found to include fragments that contain the extended microtubule-binding regions of tau, MTBR/R' and adjacent C-terminal sequences. Immunodepletion or co-injection with antibodies targeting epitopes within these fragments prevented the acute disruption of synaptic plasticity by these patient-derived synaptotoxic tau preparations. Moreover, a recombinant human tau fragment encompassing the core MTBR/R'-region present in tau fibrils, tau_297-391, potently mimicked the deleterious action of patient-derived tau. MTBR/R'-directed antibodies also rapidly reversed a very persistent synaptotoxic effect of soluble brain tau. Our findings reveal a hitherto relatively unexplored potential benefit of targeting extracellular MTBR/R' tau on correcting synaptic dysfunction.

Keywords: Hippocampus; In vivo; Long-term potentiation; Mass spectrometry; Tauopathies.

Fig. 1

MTBR/R’-containing extracellular tau fragments secreted by Ts21 iNs are synaptotoxic. a Antibodies used to capture/detect MR-MTBR/R’ tau fragments in a custom Meso Scale Diagnostics assay (c, e, and g). b High-frequency stimulation (HFS, arrow) triggered robust LTP of hippocampal synaptic transmission in anaesthetised (urethane, 1.6 g/kg, i.p.) rats after intracerebroventricular (i.c.v, asterisk) injection of 15–20 µL of either vehicle or the secretome from non-demented control (NDC) iPSC-derived iNs. In contrast, LTP decayed to baseline 3 h after application of HFS in animals injected with secretomes from one of two Ts21 lines. c Immunodepletion (ID) with the monoclonal MR-directed antibody Tau5 lowered MTBR/R′ tau level by ~ 55%. d After ID with Tau5, Ts21 secretomes no longer inhibited LTP. e–h ID either with the monoclonal antibody Gen2B, directed at R’, (e and f), or Gen2A, directed at R’ and the adjacent CT, (g and h), caused a > 85% reduction of MR-MTBR/R’ fragment concentration, and prevented the inhibition of LTP by Ts21 secretomes. In b, d, f and h, left-hand panels show the time course of LTP. Summary bar charts of LTP magnitude during the last 10 min are displayed in the right-hand panels. Insets in b show representative field EPSP traces at the times indicated. Calibration bars: vertical, 1 mV; horizontal, 10 ms. In b, d, f and h, values are mean ± SEM. ^#p < 0.05, ^##p < 0.01, ^###p < 0.001, ^####p < 0.0001 compared with pre-HFS, paired t-test; *p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA followed by Bonferroni’s multiple-comparison tests

Fig. 2

Tau fragments in certain Ts21 iN secretome size exclusion chromatography (SEC) fractions potently inhibit LTP in an MTBR/R’-CT-dependent manner. a Mass spectrometry (MS) analysis of MR and MTBR tau in SEC-fractionated Ts21 iN secretomes. Tau fragments containing MTBR appear to be most abundant in fractions 12–14 and 16–17. Note that insufficient quantity of fraction 15 was available following the functional LTP experiments to analyse by MS. The elution of globular protein standards is indicated by downward pointing arrows. An increase in red colour intensity denotes an increase in peptide abundance. Quality control (QC) sample comprises all fractions analysed (see Fig. 5S). Unique 2N4R tau peptide sequences used in the targeted assay are shown on the left (see also Table 2S). Data from adjacent fragments containing unique tryptic peptides were combined (2). MR and MTBR tau are specified on the right. b In anaesthetised rats, i.c.v. injection of 20 µL of Ts21 Mock ID SEC fractions 15 (Fr 15) and 16 (Fr 16) significantly inhibited hippocampal LTP. Circles represent individual animal values. *p < 0.05, ****p < 0.0001, compared with Vehicle group, one-way ANOVA followed by Bonferroni’s multiple-comparison tests. c Co-injection of Gen2A mAb (2.5 µg per injection) prevented the inhibition of LTP by Fr 15 (10 µL of 2X concentrated Fr 15). d The recombinant human tau fragment tau_297-391 (also known as dGAE fragment [39]) (80 pg/injection in 5 µL i.c.v.) potently inhibited LTP. This dose did not affect baseline synaptic transmission in the absence of HFS conditioning stimulation (pre- vs 210 min post-injection, n = 3, p = 0.3778, data not shown). Importantly, the amount of tau_297-391 used was similar to MR-MTBR/R’-containing fragments in the Ts21 unfractionated secretomes (34–60 pg/injection i.c.v., Fig. 1c,e,g). Left-hand panels in c, d show the time course of LTP. Summary bar charts of LTP magnitude during the last 10 min are displayed in the right-hand panels. Values are mean ± SEM. ^##p < 0.01, compared with pre-HFS, paired t-test; ****p < 0.0001, unpaired t-test tests

Fig. 3

MTBR/R’ is required for acute plasticity disruption by synaptotoxic tau in AD brain extracts. a Aqueous extracts from the brains of two people with AD (AD1 and AD6) were analysed by Western blotting in the presence or absence of lambda protein phosphatase (± LPP) using the monoclonal anti-tau antibodies Gen2A and Gen2B. Migration of SDS-PAGE molecular weight standards (in kDa) is indicated on the left. The bracket on the right highlights the position of full-length tau whereas arrows indicate likely tau fragments detected by both mAbs. It should be noted that AD6 extract had much higher total tau concentration as measured by immunoassay [41]. b Unlike isotype control IgG, anaesthetised rats co-injected with AD1 extract (10 µL) and either Gen2A or Gen2B antibodies (2.5 µg i.c.v.) maintained normal hippocampal LTP. c Similarly, robust LTP was induced when Gen2B mAb (2.5 µg i.c.v.) was co-administered with the extract AD6 (10 µL). Left-hand panels in b, c show the time course of LTP. Summary bar charts of LTP magnitude during the last 10 min are displayed in the right-hand panels. Values are mean ± SEM. ^##p < 0.01, ^###p < 0.001 compared with pre-HFS, paired t-test; ***p < 0.001, ****p < 0.0001, one-way ANOVA followed by Bonferroni’s multiple-comparison tests in b and unpaired t-test in c

Fig. 4

Rapid reversal of the persistent inhibition of LTP by synaptotoxic tau in AD aqueous brain extract by acute injection of anti-tau antibodies. a Study design. Animals received an i.c.v. injection of AD1 extract under recovery anaesthesia (ketamine and medetomidine, 60 and 0.4 mg/kg, respectively, i.p.), followed 3 weeks later by a single injection with anti-tau or control antibodies under non-recovery urethane anaesthesia. b Human tau was detected in rat hippocampus 3 weeks (3 wk, n = 18, after i.c.v. injection of 10 µL of AD1). LLoQ = 0.03125 ng/ml (see Fig. 4S). For comparison, the levels of human tau detected after acute (30 min) injection with the same volume of AD1 (pooled sample from 3 rats) or a sham injection (n = 1) are also shown. c Whereas LTP was strongly inhibited in animals receiving control mAb IgG (AD1 + IgG), robust LTP was induced in rats injected with the mid-region anti-tau mAb Tau5 (AD1 + Tau5) (2.5 μg i.c.v., 15 min prior to HFS). d The same protocol was followed using anti-tau mAbs targeting the MTBR/R’ or MTBR/R’ and the adjacent CT region of tau with Gen2B (AD1 + Gen2B) or Gen2A (AD1 + Gen2A), respectively. HFS triggered stable LTP in these animals. Left-hand panels in c, d show the time course of LTP. Summary bar charts of LTP magnitude during the last 10 min are displayed in the right-hand panels. Values are mean ± SEM. ^#p < 0.05, ^##p < 0.01 compared with pre-HFS, paired t-test; **p < 0.01, ***p < 0.001 unpaired t-test in c and one-way ANOVA followed by Bonferroni’s multiple-comparison tests in d