An AMPK-dependent regulatory pathway in tau-mediated toxicity

Abstract

Neurodegenerative tauopathies are characterised by accumulation of hyperphosphorylated tau aggregates primarily degraded by autophagy. The 5'AMP-activated protein kinase (AMPK) is expressed in most cells, including neurons. Alongside its metabolic functions, it is also known to be activated in Alzheimer's brains, phosphorylate tau, and be a critical autophagy activator. Whether it plays a neurotoxic or neuroprotective role remains unclear. In tauopathies stress conditions can result in AMPK activation, enhancing tau-mediated toxicity. Paradoxically, in these cases AMPK activation does not always lead to protective autophagic responses. Using a Drosophila in vivo quantitative approach, we have analysed the impact of AMPK and autophagy on tau-mediated toxicity, recapitulating the AMPK-mediated tauopathy condition: increased tau phosphorylation, without corresponding autophagy activation. We have demonstrated that AMPK binding to and phosphorylating tau at Ser-262, a site reported to facilitate soluble tau accumulation, affects its degradation. This phosphorylation results in exacerbation of tau toxicity and is ameliorated via rapamycin-induced autophagy stimulation. Our findings support the development of combinatorial therapies effective at reducing tau toxicity targeting tau phosphorylation and AMPK-independent autophagic induction. The proposed in vivo tool represents an ideal readout to perform preliminary screening for drugs promoting this process.

Keywords: AMPK; Autophagy; Drosophila; Neurodegeneration; Tau.

Fig. 1.

Immunoprecipitation reveals a physical association between tau and AMPKα. (A) Lysates from HEK293T cells transfected with a pCMV–Flag vector (negative control) and a pCMV–Flag-tau were immunoprecipitated with ANTI-FLAG M2 Affinity Gel. Immunoprecipitates were analysed by western blotting, probing against AMPKα. Uncropped blots and input controls are shown in Fig. S1A,B. (B) Identical lysates were immunoprecipitated with anti-AMPKα antibody and then incubated with protein A-Sepharose. Immunoprecipitates were analysed by western blotting, probing against total tau. Control blots are shown in Fig. S1C. (C) Interaction of tau with active AMPK (pAMPKα). Identical lysates were immunoprecipitated with anti-phospho-AMPKα (Thr172) antibody and then incubated with protein A-Sepharose. Immunoprecipitates were analysed by western blotting, probing against total tau. Control blots are shown in Fig. S1D.

Fig. 2.

AMPKα1 affects exogenous tau stability. (A) Extracts from HEK293T transfected as indicated were used for detection of flagged-tau and phospho-tau, probing with anti-Flag and anti-p262-tau antibodies respectively, and then normalised with respect to tubulin. (B) Tau stability quantification expressed as a ratio between tau/tubulin is shown in the graph. Data are shown as mean±s.d.; 8.25±2.2 for tau wt+AMPKα1-CA, 0.77±0.24 for tau wt+AMPKα1-KD. P values were calculated by one-way ANOVA. Asterisks indicate significant differences (P<0.05). **P<0.01, n≥3 independent HEK293T transfections. (C) Quantification of phospho-tau species as a ratio between p262 tau/tubulin is shown in the graph. Data are shown as mean±s.d.; 0.37±0.03 for tau wt+AMPKα1-CA, 0.12±0.005 for tau S262A+AMPKα1-CA, 0.11±0.01 for tau wt+AMPKα1-KD. P values were calculated by one-way ANOVA. Asterisks indicate significant differences (P<0.05). **P<0.01, n≥3 independent HEK293T transfections.

Fig. 3.

Downregulation of Drosophila AMPKα (dAMPKα) partially rescued the tau-induced degenerative eye phenotype. (A) Scanning electron microscope (SEM) images of the external eye of control flies (GMR/+), flies expressing tau (GMR/+; tau/+) and flies co-expressing tau and RNAi dAMPKα (GMR/+; tau/ RNAi dAMPKα) under the control of the GMR-GAL4 driver. Scale bar: 100 µm. (B) Quantitative analysis of tau-induced toxicity displayed as the cumulative distribution of ommatidial distortion coefficients (DC) obtained via QED analysis for control flies (GMR, green line), flies expressing tau (GMR tau, black line) and flies co-expressing tau and RNAi dAMPKα (GMR tau RNAi dAMPKα, orange line) under the control of the GMR-GAL4 driver. Co-expression of RNAi dAMPKα with tau reduced tau-induced ommatidial distortion, as indicated by the leftward shift of the DC distribution, compared to the distribution of ommatidial distortion coefficients for flies expressing tau alone. The SEM images are representative of the median DC values for each genotype. The phenotypes associated with the lower and upper DC values can be seen in Fig. S5. GMR n=14, GMR tau n=14, GMR tau RNAi dAMPKα n=25. Levels of significance for Mann–Whitney directional tests are <0.01 for both the compared groups: GMR versus GMR tau and GMR tau versus GMR tau RNAi dAMPKα. Validation of dAMPK downregulation upon expression RNAi dAMPKα under the control of the GMR-GAL4 driver can be seen in Fig. S3B.

Fig. 4.

Recapitulation of the human AMPK/tau pathway in the Drosophila eye. (A) Scanning electron microscope (SEM) images of the external eye of flies expressing tau (GMR/+; tau/+) and flies co-expressing tau and human AMPKα1-CA (GMR/+; tau/ AMPKα1-CA) or tau and human AMPKα1-KD (GMR/+; tau/ AMPKα1-KD) under the control of the GMR-GAL4 driver. Scale bar: 100 µm. (B) Quantitative analysis of tau-induced toxicity as distribution of ommatidial distortion coefficients (DC) for flies expressing tau (GMR tau, black line), flies co-expressing tau and AMPKα1-CA (GMR tau+ AMPKα1-CA, blue line), tau and AMPKα1-KD (GMR tau+ AMPKα1-KD, red line) under the control of the GMR-GAL4 driver. The SEM images are representative of the median DC values for each genotype. The phenotypes associated with the lower and upper DC values, can be seen in Fig. S6. GMR tau n=14, GMR tau+ AMPKα1-CA n=9, GMR tau+ AMPKα1-KD n=12. Levels of significance for Mann–Whitney directional tests are <0.01 for both the compared groups: GMR tau versus GMR tau+ AMPKα1-CA and GMR tau versus GMR tau+ AMPKα1-KD. Validation of AMPKα1 overexpression in transgenic lines AMPKα1-CA and AMPKα1-KD under the control of the GMR-GAL4 driver can be seen in Fig. S3A,B.

Fig. 5.

AMPKα1 modulation affects tau phosphorylation and stability. AMPKα1 overactivation promotes tau phosphorylation. (A) Extracts of fly heads from flies expressing tau alone and in combination with AMPKα1-CA probed with total tau or pS262 antibodies and normalised against actin. (B) Quantification of tau phosphorylation as a ratio between ptau/total tau is shown in the graph. Data are shown as mean±s.d.; 0.94±0.14 for tau wt and 1.97±0.21 for tau+ AMPKα1-CA. P values were calculated by one-way ANOVA. Asterisks indicate significant differences (P<0.05). ***P<0.001, n≥3 independent preparations. (C) Quantification of phospho-tau species as a ratio between the ptau slower mobility upper band (upper band ptau)/total tau is shown in the graph. Data are shown as mean±s.d.; 1.11±0.26 for tau and 3.26±0.59 for tau+AMPKα1-CA. P values were calculated by one-way ANOVA. Asterisks indicate significant differences (P<0.05). **P<0.01, n≥3 independent preparations. AMPKα1 downregulation destabilises pathological tau species and changes its phosphorylation pattern. (D) Extracts of fly heads from flies expressing tau alone and in combination with RNAi dAMPKα or AMPKα1-KD, probed with total tau antibody and normalised against actin. (E) Quantification of pathological tau species as a ratio between the slower mobility upper band/actin is shown in the graph. Data are shown as mean±s.d.; 0.90±0.13 for tau, 0.78±0.05 for tau+RNAi dAMPKα, 0.55±0.06 for tau+AMPKα1-KD. P values were calculated by one-way ANOVA. Asterisks indicate significant differences. **P<0.01, n≥3 independent preparations. (F) Extracts of fly heads from flies expressing tau alone and in combination with RNAi dAMPKα or AMPKα1-KD, probed with total phospho-tau (p262) antibody and normalised against actin. (G) Quantification of pathological phospho tau species as a ratio between the slower mobility upper band/actin is shown in the graph. Data are shown as mean±s.d.; 0.99±0.14 for tau, 0.65±0.09 for tau+RNAi dAMPKα, 0.67±0.09 for tau+AMPKα1-KD. P values were calculated by one-way ANOVA. Asterisks indicate significant differences. *P<0.05, n≥3 independent preparations.

Fig. 6.

Modulation of AMPK, autophagy and AMPKα1-CA tau-mediated toxicity. Modulation of AMPK and Ref(2)P accumulation in tau flies. (A) Western blot analysis of fly head lysates of tau, tau/AMPKα1-KD and tau/AMPKα1-CA flies probed with anti-Ref(2)P antibody and normalised for actin. Lysates from tau (control), tau/AMPKα1-KD and tau/AMPKα1-CA (experiments) were loaded on the same gel and probed on the same membrane. (B) Quantification of Ref(2)P accumulation as a ratio between Ref(2)P/actin is shown in the graph. Data are shown as mean±s.d.; 1.09±0.18 for tau, 2.98±0.48 for tau/AMPKα1KD and 0.95±0.57 for tau/AMPKα1-CA. P values were calculated by one-way ANOVA. Asterisks indicate significant differences. ***P<0.001, n≥3 independent preparations. Rapamycin treatment ameliorates AMPKα1-CA mediated tau phenotype. (C) SEM images of the external eye of control (DMSO) and rapamycin-treated (Rapamycin) flies co-expressing tau and AMPKα1-CA (GMR/+; tau/AMPK CA) under the control of the GMR-GAL4 driver. Scale bar: 100 µm. (D) Quantitative analysis of tau-induced toxicity as a distribution of ommatidial distortion coefficients (DC) for flies co-expressing tau and AMPKα1-CA (GMR/+; tau/AMPK CA) fed with DMSO (black line) and rapamycin (red line). The SEM images are representative of the median DC values for each genotype. The phenotypes associated with the lower and upper DC values can be seen in Fig. S7. DMSO n=9, Rapamycin n=7. Level of significance for a Mann–Whitney directional test is <0.01.

Fig. 7.

AMPK-mediated tau phosphorylation and autophagy activation: model of action. Downregulation of AMPK via AMPKα1-KD, although causing a reduction in the activation of autophagy, resulted in less tau toxicity, likely due to the reduction of tau phosphorylation and hence less toxic tau species. In contrast, overexpression of AMPKα1-CA results in increased tau toxicity due to AMPKα1-CA's ability to phosphorylate tau, thereby promoting the appearance of toxic tau species, without affecting autophagy, which would otherwise degrade such toxic species. Finally, overexpression of AMPKα1-CA concomitant with pharmacological induction of autophagy, via rapamycin, resulted in rescued AMPKα1-CA-mediated tau toxicity.