Drosophila appear resistant to trans-synaptic tau propagation

Abstract

Alzheimer's disease is the most common cause of dementia in the elderly, prompting extensive efforts to pinpoint novel therapeutic targets for effective intervention. Among the hallmark features of Alzheimer's disease is the development of neurofibrillary tangles comprised of hyperphosphorylated tau protein, whose progressive spread throughout the brain is associated with neuronal death. Trans-synaptic propagation of tau has been observed in mouse models, and indirect evidence for tau spread via synapses has been observed in human Alzheimer's disease. Halting tau propagation is a promising therapeutic target for Alzheimer's disease; thus, a scalable model system to screen for modifiers of tau spread would be very useful for the field. To this end, we sought to emulate the trans-synaptic spread of human tau in Drosophila melanogaster. Employing the trans-Tango circuit mapping technique, we investigated whether tau spreads between synaptically connected neurons. Immunohistochemistry and confocal imaging were used to look for tau propagation. Examination of hundreds of flies expressing four different human tau constructs in two distinct neuronal populations reveals a robust resistance in Drosophila to the trans-synaptic spread of human tau. This resistance persisted in lines with concurrent expression of amyloid-β, in lines with global human tau knock-in to provide a template for human tau in downstream neurons, and with manipulations of temperature. These negative data are important for the field as we establish that Drosophila expressing human tau in subsets of neurons are unlikely to be useful to perform screens to find mechanisms to reduce the trans-synaptic spread of tau. The inherent resistance observed in Drosophila may serve as a valuable clue, offering insights into strategies for impeding tau spread in future studies.

Keywords: Alzheimer’s disease; Drosophila melanogaster; neurodegeneration; tau; trans-synaptic.

Graphical Abstract

Figure 1

hTau stays restricted to expressing neurons, even in flies kept at 29°C for 4 weeks. (A) hTau staining was restricted to within the PDF neuron membrane—marked by membrane-bound GFP (mCD8:GFP). Control flies without hTau exhibited no Tau 5A6 staining, unlike Tau13-stained controls in Supplementary Fig. 1. (B) In flies kept at 29°C for 4 weeks, Tau 5A6 staining remained restricted to within the PDF neuron membrane—marked by membrane-bound red fluorescent protein (mCD8:RFP). Control flies without hTau that expressed both GFP and RFP [to control for two upstream activation sequence (UAS) lines] exhibited no Tau 5A6 staining. Haemagglutinin (HA) was also imaged and was found to overlap with RFP. (C) Fluorescence intensity of co-localized Tau 5A6-RFP was measured and normalized to the mean RFP intensity of each brain. Tau 5A6 detected significant hTau overexpression in PDF neurons compared to GFP controls (Kruskal–Wallis rank sum test, χ² = 25.2, df = 2, P < 0.0001; N = 10–11 brains). There was no significant difference between hTau^WT and hTau^E14 expression (P = 0.12, Dunn post hoc test, P-value adjusted with the Bonferroni method). (D) Tau 5A6 staining outside RFP-labelled PDF neurons was not observed. Images were blinded and scored, and no difference between genotypes was observed (χ² test, χ² = 0.063, df = 2, P = 0.97; N = 10–11 brains). Scale bar is 20 μm. Genotypes: (A) PDF-GAL4 > UAS-mCD8:GFP, PDF-GAL4 > UAS-mCD8:GFP + UAS-hTau^WT(0N4R). (B) PDF-GAL4 > UAS-mCD8:GFP + UAS-mCD8:RFP, PDF-GAL4 > UAS-mCD8:RFP + UAS-hTau^WT(0N4R):HA, PDF-GAL4 > UAS-mCD8:RFP + UAS-hTau^E14(0N4R):HA.

Figure 2

Absence of hTau propagation to post-synaptic neurons with different hTau isoforms and different post-synaptic Aβ isoforms. (A) PDF neurons project to the mushroom body calyces. Trans-Tango labels pre-synaptic PDF neurons with GFP and their post-synaptic partners with haemagglutinin (HA)-tagged tdTomato (tdTom). Additional experiments using GAL4 and QF drivers (from the trans-Tango system) can also express hTau and human Aβ (hAβ) in pre- and post-synaptic compartments. Created using BioRender. (B) Representative axon terminal region from PDF neurons highlighting GFP and hTau expression in PDF neurons, and tdTom:HA expression in post-synaptic partners. This fly also expressed Aβ₄₂ in post-synaptic neurons. Staining with Tau 5A6 antibody (full-length hTau) indicated hTau expression in the PDF neuron but not outside. Scale bar = 20 μm. (C) Quantification of total hTau fluorescence intensity showed an expected significant increase in hTau-expressing brains compared to controls without hTau (this difference is indicated by *** above the conditions without hTau; main effect of hTau isoform: one-way ANOVA on Tukey-transformed data: F(2,77) = 58.73, P < 0.0001; N = 7–10 brains). (D) Quantification of hTau co-localization signal outside GFP and inside tdTom signals showed no difference between genotypes (main effect of genotype: one-way ANOVA on Tukey-transformed data: F(8,71) = 0.78, P = 0.62; N = 7–10 brains). n.s. = not significant. Genotypes (left to right): PDF-GAL4 > trans-Tango; PDF-GAL4 > trans-Tango + QUAS-Aβ₄₀; PDF-GAL4 > trans-Tango + QUAS-Aβ₄₂; PDF-GAL4 > trans-Tango + UAS-hTau^WT(0N4R); PDF-GAL4 > trans-Tango + UAS-hTau^WT(0N4R)+QUAS-Aβ₄₀; PDF-GAL4 > trans-Tango + UAS-hTau^WT(0N4R)+QUAS-Aβ₄₂; PDF-GAL4 > trans-Tango + UAS-hTau^WT(2N4R); PDF-GAL4 > trans-Tango + UAS-hTau^WT(2N4R)+QUAS-Aβ₄₀; PDF-GAL4 > trans-Tango + UAS-hTau^WT(2N4R)+QUAS-Aβ₄₂.

Figure 3

Absence of hTau propagation in post-synaptic neurons. (A) Schematic shows the brain regions involved in the experiment. Pre-synaptic olfactory neurons are labelled by Orco-GAL4, while PNs connect the antennal lobes to the mushroom body, labelled by GH146-QF. These flies contain a global hTau^WT(0N4R) knock-in mutation. Created using BioRender. (B) Schematic of hTau adapted from Chi et al. Small bars indicate the locations of the 14 phospho-epitope changes of hTau^E14. AT8 cannot detect hTau^E14 due to these mutations. (C) In a global hTau^WT(0N4R) heterozygous background, GH146-QF labels post-synaptic neurons with membrane-bound red fluorescent protein (mCD8:RFP) while pre-synaptic Orco-GAL4 expresses hTau^WT(0N4R) or control. AT8 did not detectably spread trans-synaptically. Arrowheads and dashed areas label clusters of post-synaptic cell bodies and endogenous hTau accumulation, respectively. (D) Fluorescence intensity of AT8 was measured and normalized to the maximum area of each brain. AT8 detected significant hTau^WT overexpression in the antennal lobes compared to control and hTau^E14 flies (this difference is indicated by ** above the hTau^WT condition, Kruskal–Wallis rank sum test, χ² = 13.01, df = 2, P = 0.0015; N = 6–7 brains), but there was no difference in AT8 staining between control and hTau^E14 flies. (E) AT8 staining within RFP-labelled post-synaptic neurons was not observed. Images were blinded and scored, and no difference between genotypes was observed (χ² test, χ² = 0.1, df = 2, P = 0.95; N = 6–7 brains). Scale bar is 50 μm. Genotypes: Orco-GAL4; GH146-QF > QUAS-mCD8:RFP; hTau^WT(0N4R), Orco-GAL4 > UAS-hTau^WT(0N4R); GH146-QF > QUAS-mCD8:RFP; hTau^WT(0N4R), Orco-GAL4 > UAS-hTau^E14(0N4R); GH146-QF > QUAS-mCD8:RFP; hTau^WT(0N4R).