UBE4B, a microRNA-9 target gene, promotes autophagy-mediated Tau degradation

Abstract

The formation of hyperphosphorylated intracellular Tau tangles in the brain is a hallmark of Alzheimer's disease (AD). Tau hyperphosphorylation destabilizes microtubules, promoting neurodegeneration in AD patients. To identify suppressors of tau-mediated AD, we perform a screen using a microRNA (miR) library in Drosophila and identify the miR-9 family as suppressors of human tau overexpression phenotypes. CG11070, a miR-9a target gene, and its mammalian orthologue UBE4B, an E3/E4 ubiquitin ligase, alleviate eye neurodegeneration, synaptic bouton defects, and crawling phenotypes in Drosophila human tau overexpression models. Total and phosphorylated Tau levels also decrease upon CG11070 or UBE4B overexpression. In mammalian neuroblastoma cells, overexpression of UBE4B and STUB1, which encodes the E3 ligase CHIP, increases the ubiquitination and degradation of Tau. In the Tau-BiFC mouse model, UBE4B and STUB1 overexpression also increase oligomeric Tau degradation. Inhibitor assays of the autophagy and proteasome systems reveal that the autophagy-lysosome system is the major pathway for Tau degradation in this context. These results demonstrate that UBE4B, a miR-9 target gene, promotes autophagy-mediated Tau degradation together with STUB1, and is thus an innovative therapeutic approach for AD.

Fig. 1. Genome-wide Drosophila miRNA library screening identified miR-9 family miRNAs as modifiers of hTau in Drosophila eyes.

a Screening of the miRNA library in GMR > hTau Drosophila eyes revealed significant phenotypic enhancement, as indicated by decreased eye size, in flies overexpressing miR-9 family miRNAs compared with GMR > hTau control flies. Eye sizes were arranged from smallest to largest. b Volcano plot of mean eye sizes in flies expressing various miRNAs in the GMR > hTau background versus their respective p-values derived from a one-way analysis of variance followed by pairwise t-tests and a Bonferroni correction for multiple comparisons. All points above the dotted line can be considered significant. This analysis identified that flies overexpressing miR-9 family miRNAs exhibited a severe Tau toxicity phenotype, as indicated by decreased eye size. N = 3 biologically independent experiments. c, d The overexpression of miR-9a, miR-9b, or miR-9c in GMR > hTau Drosophila eyes significantly reduced eye sizes relative to GMR > hTau Drosophila. N = 5 biologically independent experiments. Data are presented as the mean ± s.e.m. Statistical significance was determined with a two-tailed Student’s t-test. In the box plots the whiskers represent the 5th to 95th percentile range. e Alignment of mature Drosophila miR-9a, miR-9b, and miR-9c sequences with human and murine miR-9 sequences identified that miR-9a had 100% identity with mammalian miR-9 sequences. Statistical source data.

Fig. 2. CG11070, a miR-9a target identified from secondary screening, strongly modulated hTau in Drosophila eyes.

a Screening of flies with RNAi knockdown of miR-9a targets in GMR > hTau Drosophila eyes identified a significant reduction in eye size in CG11070-RNAi flies relative to the control. b Volcano plot of mean eye sizes of flies expressing various miR-9a target gene RNAis in GMR > hTau flies versus their respective p-values derived from a one-way analysis of variance followed by pairwise t-tests and a Bonferroni correction for multiple comparisons. All points above the dotted line can be considered significant. CG11070-RNAi flies exhibited more severe ocular Tau toxicity, as demonstrated by decreased eye size. N = 3 biologically independent experiments. c, d The knockdown of CG11070 (CG11070-RNAi) in GMR > hTau flies significantly decreased eye size relative to GMR > hTau controls. N = 5 biologically independent experiments. In the box plots the whiskers represent the 5th to 95th percentile range. Data are presented as the mean ± s.e.m. Statistical significance was determined with a two-tailed Student’s t-test. e miRNA–mRNA–RISC pull-down assays in Drosophila S2 cells revealed that miR-9a bound to CG11070 mRNA. Transfection of miR-9a enriched CG11070 mRNA levels, similar to the known miR-9a targets senseless and sNPFR1, as demonstrated by qRT-PCR. N = 4 biologically independent experiments. Data are presented as the mean ± s.e.m. Statistical significance was determined with a two-tailed Student’s t-test. f Expression of miR-9a using GMR-GAL4 significantly decreased CG11070 expression. N = 3 biologically independent experiments. Data are presented as the mean ± s.e.m. Statistical significance was determined with a two-tailed Student’s t-test. Statistical source data.

Fig. 3. The overexpression of Drosophila CG11070 and its mammalian orthologue UBE4B alleviated hTau phenotypes in Drosophila.

a, b Eye-specific overexpression of Drosophila CG11070 or its mammalian orthologue UBE4B in GMR > hTau flies increased eye size relative to GMR > hTau controls. N = 5 biologically independent experiments. In the box plots the whiskers represent the 5th to 95th percentile range. c, d Neuronal overexpression of Drosophila CG11070 or its mammalian orthologue UBE4B using Elav-Gal4 in Elav > hTau flies significantly increased larval crawling. N = 10 biologically independent experiments. In the box plots the whiskers represent the 5th to 95th percentile range. e, f Neuronal overexpression of Drosophila CG11070 or its mammalian orthologue UBE4B using Elav-Gal4 in Elav > hTau flies significantly increased synaptic bouton numbers relative to Elav > hTau controls. Scale bar 100 µm. N = 10 biologically independent experiments. In the box plots the whiskers represent the 5th to 95th percentile range. g–i Western blotting revealed that ocular overexpression of Drosophila CG11070 or its mammalian orthologue UBE4B in GMR > hTau flies significantly reduced total and phosphorylated Tau protein levels relative to GMR > hTau controls. N = 4 biologically independent experiments. Data are presented as the mean ± s.e.m. Statistical significance was determined with a two-tailed Student’s t-test. Statistical source data.

Fig. 4. Tau was ubiquitinated and degraded by UBE4B and STUB1 in mammalian neuroblastoma cells.

a Tau was co-expressed with His-Ubiquitin, UBE4B, and STUB1 WT or dominant-negative mutant STUB1^H260Q in SH-SY5Y neuroblastoma cells. Ubiquitinated Tau was significantly increased by co-expression of UBE4B and STUB1 (lane 5), but not by expression of UBE4B (lane 4) or STUB1 (lane 3) alone. Tau ubiquitination by co-expression of UBE4B and STUB1 required the ligase activity of STUB1 (lanes 5 and 6). b, c Tau protein degradation was enhanced by co-expression of UBE4B and STUB1 (lane 6–9) compared with expression of UBE4B (lane 1–4) or STUB1 (lane f–i) alone. Quantification of Tau levels was normalized to the amount of β-actin protein in each case. Data represent the mean ± s.e.m of three independent experiments (*P < 0.05, **P < 0.005, ***P < 0.001 two-tailed Student’s t-test). d UBE4B was co-expressed with HA-STUB1 and Tau in SH-SY5Y cells and immunoprecipitated on anti-HA-agarose beads. UBE4B did not directly interact with STUB1 in the absence of Tau, but indirectly interacted with STUB1 in the presence of Tau. e HA-UBE4B was co-expressed with Tau in SH-SY5Y cells and immunoprecipitated on anti-HA-agarose beads. Co-precipitated Tau was detected by Western blot, revealing that Tau directly interacted with UBE4B. All western blots were performed more than three times. Statistical source data.

Fig. 5. STUB1 knockdown reduces UBE4B-mediated alleviated hTau phenotypes in Drosophila.

a, b Eye-specific knockdown of Drosophila STUB1 significantly reduced eye size relative to GMR > hTau + hUBE4B flies. N = 5 biologically independent experiments. In the box plots the whiskers represent the 5th to 95th percentile range. c, d Neuronal knockdown of Drosophila STUB1 using Elav-Gal4 in Elav > hTau + hUBE4B flies significantly reduced larval crawling. N = 10 biologically independent experiments. In the box plots the whiskers represent the 5th to 95th percentile range. e, f Neuronal knock down of Drosophila STUB1 using Elav-Gal4 in Elav > hTau + hUBE4B flies significantly reduced synaptic bouton numbers relative to Elav > hTau controls. Scale bar 100 µm. N = 10 biologically independent experiments. In the box plots the whiskers represent the 5th to 95th percentile range. Data are presented as the mean ± s.e.m. Statistical significance was determined with two-tailed Student’s t-test. Statistical source data.

Fig. 6. Tau oligomers were degraded by UBE4B and STUB1 in the Tau-BiFC mouse model.

a Schematic representation of AAV-CMV-mCherry and AAV-UBE4B-mCherry virus constructs. b Schematic illustration of AAV-CMV-mCherry or AAV-UBE4B-mCherry virus delivery into the dentate gyrus of Tau-BiFC mice. c A fluorescence staining image indicating the foci (red) of AAV-CMV-mCherry or AAV-UBE4B + STUB1-mCherry virus delivery in the dentate gyrus and hippocampus of Tau-BiFC mice. Scale bar (white), 1 mm. d Schematic illustration of the work flow for virus injection and pathological examination in Tau-BiFC mice. e AAV-UBE4B and AAV-STUB1 decreased oligomer Tau-BiFC and pTau (S202/T205) levels in the dentate gyrus compared with the control. GCL, granular cell layer; POL, polymorphic layer. Scale bars (white), 80 μm. These experiments were performed four times. f, g Densitometry analysis revealed that AAV-UBE4B and STUB1 significantly decreased both Tau-BiFC and pTau (S202/T205) levels in the dentate gyrus relative to the control (AAV-Cont N = 4; AAV-UBE4B N = 4; AAV-STUB1 N = 4; AAV-UBE4B + AAV-STUB1, N = 4; N = 4 biologically independent animals), respectively. In the box plots the whiskers represent the 5th to 95th percentile range. Data are presented as means as ± s.e.m. Statistical significance was determined with two-tailed Student’s t-test. Statistical source data.

Fig. 7. Tau was degraded by UBE4B and STUB1 primarily via autophagy.

a Treatment of SH-SY5Y cells with the proteasome inhibitor MG132 did not affect Tau degradation mediated by UBE4B/STUB1. b Treatment with chloroquine (CQ), an autophagy inhibitor, affected Tau degradation by UBE4B/STUB1. c Pepstatin A (PEPA) and E64D, autophagy inhibitors, inhibited Tau degradation by UBE4B/STUB1. All western blots were performed three times. d Schematic illustration of autophagy inhibitor injection to the dentate gyrus of Tau-BiFC mice. e Schematic illustration of the work flow for autophagy inhibitor injection and pathological examination in Tau-BiFC mice. f CQ and E64D plus PEPA (E64D + PEPA) increased pTau (S202/T205) levels in the dentate gyrus relative to saline-injected controls. GCL granular cell layer, POL polymorphic layer. g, h Densitometry analysis revealed that chloroquine and ED64 + PEPA significantly increased both pTau (S202/T205) and Tau-BiFC levels in the dentate gyrus relative to saline-injected controls (AAV-UBE4B + AAV-STUB1 (saline control), N = 4; CQ + AAV-UBE4B + AAV-STUB1, N = 4; E64D/PEPEA + AAV-UBE4B + AAV-STUB1, N = 4; N = 4 biologically independent animals). In the box plots the whiskers represent the 5th to 95th percentile range. i AAV-UBE4B and AAV-STUB1 decreased LC3 levels in the dentate gyrus relative to control. Scale bars (white): 40 μm. j AAV-UBE4B and AAV-STUB1 decreased p62 levels in the dentate gyrus relative to control. Scale bars (white): 40 μm. k Densitometry analysis revealed that autophagy inhibitors significantly increased LC3 levels in the dentate gyrus relative to control (AAV-Control, N = 4; AAV-UBE4B + AAV-STUB1, N = 4; CQ + AAV-UBE4B + AAV-STUB1, N = 4; E64D/PEPA + AAV-UBE4B + AAV-STUB1, N = 4; N = 4 biologically independent animals). In the box plots the whiskers represent the 5th to 95th percentile range. l Densitometry analysis revealed that autophagy inhibitors significantly increased p62 levels in the dentate gyrus relative to control (AAV-Control, N = 4; AAV-UBE4B + AAV-STUB1, N = 4; CQ + AAV-UBE4B + AAV-STUB1, N = 4; E64D/PEPA + AAV-UBE4B + AAV-STUB1, N = 4; N = 4 biologically independent animals). In the box plots the whiskers represent the 5th to 95th percentile range. Data are presented as means ± s.e.m. Statistical significance was determined with a two-tailed Student’s t-test. Statistical source data. Each exact p value was listed in Statistical source data.