Overexpression of TFEB Drives a Pleiotropic Neurotrophic Effect and Prevents Parkinson's Disease-Related Neurodegeneration

Abstract

The possible implication of transcription factor EB (TFEB) as a therapeutic target in Parkinson's disease has gained momentum since it was discovered that TFEB controls lysosomal biogenesis and autophagy and that its activation might counteract lysosomal impairment and protein aggregation. However, the majority of putative direct targets of TFEB described to date is linked to a range of biological processes that are not related to the lysosomal-autophagic system. Here, we assessed the effect of overexpressing TFEB with an adeno-associated viral vector in mouse substantia nigra dopaminergic neurons. We demonstrate that TFEB overexpression drives a previously unknown bona fide neurotrophic effect, giving rise to cell growth, higher tyrosine hydroxylase levels, and increased dopamine release in the striatum. TFEB overexpression induces the activation of the mitogen-activated protein kinase 1/3 (MAPK1/3) and AKT pro-survival pathways, phosphorylation of mTORC1 effectors 4E-binding protein 1 (4E-BP1) and S6 kinase B1 (S6K1), and increased protein synthesis. We show that TFEB overexpression prevents dopaminergic cell loss and counteracts atrophy and the associated protein synthesis decline in the MPTP mouse model of Parkinson's disease. Our results suggest that increasing TFEB activity might prevent neuronal death and restore neuronal function in Parkinson's disease and other neurodegenerative diseases through different mechanisms.

Keywords: MPTP; Parkinson's disease; TFEB; dopamine; neuronal atrophy; neuroprotection; neurorescue; neurotrophic; prosurvival pathways.

Graphical abstract

Figure 1

Overexpressed TFEB Translocates to the Nucleus and Increases Autophagy-Lysosomal Markers (A) Tfeb cDNA levels measured by qPCR in ventral midbrain homogenates of mice after vehicle (n = 6) or AAV-TFEB nigral injection (n = 5); Mann-Whitney test. *p < 0.05 compared to control. (B) Immunofluorescence for tyrosine hydroxylase (red), Flag (green), and nucleus (blue) in substantia nigra sections showing 3xFlag-TFEB translocation to the nucleus after AAV-TFEB nigral injection. Scale bar, 25 μm. (C–E) Representative western blots and protein levels in ventral midbrain homogenates from mice overexpressing TFEB (n = 6) compared to vehicle-injected mice (n = 5–6) of (C) LAMP1, (D) CTSD, and (E) LC3-I and -II. Mann-Whitney test, *p < 0.05 compared to control. In all panels, samples were collected 5 weeks after AAV-TFEB or vehicle injections. All data are represented as mean ± SEM.

Figure 2

TFEB Neurotrophic and Functional Effects in Substantia Nigra Dopaminergic Neurons (A) Representative photomicrograph of a tyrosine-hydroxylase-immunostained substantia nigra section of a mouse injected with AAV-TFEB into the right substantia nigra. Scale bar, 25 μm. (B) Stereological volumes of both SNpc of unilateral vehicle-injected (n = 7) and AAV-TFEB-injected mice (n = 8); two-way ANOVA, post hoc Tukey’s. *p < 0.05 compared to control injected side. ^#p < 0.05 compared to AAV-TFEB non-injected side. (C) Average cell area of SNpc dopaminergic neurons of vehicle-injected (n = 7) and AAV-TFEB-injected mice (n = 8); two-way ANOVA, post hoc Tukey’s. *p < 0.05 compared to control injected side. ^#p < 0.05 compared to AAV-TFEB non-injected side. (D) Stereological cell counts of dopaminergic neurons in substantia nigra pars compacta of mice injected with vehicle (n = 7) or AAV-TFEB (n = 9); two-way ANOVA, post hoc Tukey’s. (E) Intraneuronal optical densitometry of tyrosine hydroxylase immunoreactivity; two-way ANOVA, post hoc Tukey’s. *p < 0.05 compared to control injected side. ^#p < 0.05 compared to AAV-TFEB non-injected side. (F) Content of dopamine and its metabolites DOPAC and HVA and DOPAC/DA and HVA/DA ratios in striatal homogenates measured by HPLC of vehicle-injected (n = 6) and AAV-TFEB-injected mice (n = 6); Mann-Whitney test, *p < 0.05 compared to control. (G) Striatal DA relative amounts in AAV-TFEB and vehicle-injected mice (n = 4–6 per group) measured by microdialysis in the ipsilateral striatum following local amphetamine administration by reverse dialysis; ANOVA for repeated measures, post hoc Tukey’s. *p < 0.05 compared to control. In all panels, samples were collected 5 weeks after AAV-TFEB or vehicle injections. All data are represented as mean ± SEM.

Figure 3

Activation of EIF4E/4E-BP1 and S6K1 Protein Synthesis Pathways (A–C) Representative western blots and protein levels in ventral midbrain homogenates from mice overexpressing TFEB (n = 6) compared to control mice (n = 5 to 6) of (A) EIF4E, (B) phosphorylated and non-phosphorylated 4E-BP1, p4E-BP1^S65, expressed as a ratio, and (C) endogenous ribosomal protein S6K1, its activated form (pS6K1^T389), and the ratio. Mann-Whitney test, *p < 0.05 compared to control. (D) Representative photomicrograph of a substantia nigra section immunostained for the phosphorylated/activated RPS6, pRPS6^S235/236, of a mouse injected with AAV-TFEB into the right substantia nigra. Scale bar, 50 μm. In all panels, samples were collected 5 weeks after AAV-TFEB or vehicle injections. All data are represented as mean ± SEM.

Figure 4

Activation of MAPK1/3 and AKT Pro-survival Pathways (A–E) Representative western blots and protein levels in ventral midbrain homogenates from mice overexpressing TFEB (n = 6) compared to control mice (n = 5 to 6) of (A) MAPK 1 and 3, their activated forms (pMAPK1/3^{Thr202/Tyr204}), and the ratio; (B) 90-kDa ribosomal protein S6 kinase (P90RSK) and its phosphorylated/activated form (pP90RSK^T359/S363) and the ratio; (C) total AKT and phosphorylated/activated AKT (pAKT^S473) and the ratio; (D) total NDRG1, its phosphorylated/activated form (pNDRG1^T346), and the ratio; and (E) TBK1, its phosphorylated/activated form (pTBK1^S172), and the ratio. Mann-Whitney test, *p < 0.05 compared to control. In all panels, samples were collected 5 weeks after AAV-TFEB or vehicle injections. All data are represented as mean ± SEM.

Figure 5

Neuroprotective Effect of TFEB Overexpression in the MPTP Mouse Model of Parkinson’s Disease (A) Top: representative photomicrographs of tyrosine-hydroxylase-immunostained substantia nigra of MPTP-intoxicated mice previously injected with AAV-TFEB (n = 14) or vehicle (n = 10), and control mice treated with saline and previously injected with vehicle (n = 7). Scale bar, 200 μm. Bottom: stereological cell counts of dopaminergic neurons in substantia nigra pars compacta of all groups of animals; two-way ANOVA, post hoc Tukey’s. *p < 0.05 compared to control non-injected side. ^#p < 0.05 compared to control injected side. ^$p < 0.05 compared to MPTP+sham injected side. ˆp < 0.05 compared to MPTP+AAV-TFEB non-injected side. (B) Left: representative photomicrographs of tyrosine-hydroxylase-immunostained striata of MPTP-intoxicated mice previously injected with AAV-TFEB (n = 14) or vehicle (n = 10) and control mice treated with saline and previously injected with vehicle (n = 7). Scale bar, 500 μm. Right: optical densitometry of striatal TH-positive dopaminergic fibers of all groups of animals; two-way ANOVA, post hoc Tukey’s. *p < 0.05 compared to control non-injected side. ^#p < 0.05 compared to control injected side. ⁺p < 0.05 compared to MPTP+sham non-injected side. ^$p < 0.05 compared to MPTP+sham injected side. ˆp < 0.05 compared to MPTP+AAV-TFEB non-injected side. (C) Average cell area of dopaminergic neurons of all groups of animals (n = 6–10 mice per group); two-way ANOVA, post hoc Tukey’s. *p < 0.05 compared to control non-injected side. ^#p < 0.05 compared to control injected side. ^$p < 0.05 compared to MPTP+sham injected side. ˆp < 0.05 compared to MPTP+AAV-TFEB non-injected side. (D) Cumulative rotations induced by systemic administration of amphetamine during 10-min intervals 14 days after the last administration of MPTP. The net contralateral rotations were obtained as follows: total left − total right 360° turns. The test was performed with MPTP-intoxicated mice previously injected with AAV-TFEB (n = 9) or vehicle (n = 13) and control mice treated with saline and previously stereotaxically injected with vehicle (n = 10); two-way ANOVA for repeated-measures, post hoc Tukey’s. *p < 0.05 compared to MPTP+sham. ^#p < 0.05 compared to control. (E) Striatal DA amount measured by microdialysis in the ipsilateral striatum following local amphetamine administration by reverse dialysis in MPTP-intoxicated mice (21 days after last MPTP injection) previously injected with AAV-TFEB (n = 5) or vehicle (n = 5) and control mice treated with saline and previously injected with vehicle (n = 7); two-way ANOVA for repeated-measures, post hoc Tukey’s. *p < 0.05 compared to control. ^#p < 0.05 compared to MPTP+sham. All data are represented as mean ± SEM.

Figure 6

Mitochondrial-Lysosomal Cell Death Axis Is Blocked by TFEB Overexpression in the MPTP Model of Parkinson’s Disease (A–C) Representative western blots and protein levels of ventral midbrain homogenates in MPTP-intoxicated mice previously injected with AAV-TFEB (n = 7 to 8) or vehicle (n = 7) and control mice treated with saline and previously injected with vehicle (n = 6) of (A) pCASP9^T125, (B) LAMP1, and (C) LC3-I and -II; one-way ANOVA, post hoc Tukey’s. *p < 0.05 compared to control. ^#p < 0.05 compared to MPTP+sham. (D) Confocal images showing immunofluorescence for tyrosine hydroxylase (red), LAMP1 (green), and nuclear staining (blue). Scale bar, 25 μm. (E) mRNA levels in ventral midbrain extracts in MPTP-intoxicated mice previously injected with AAV-TFEB (n = 6 to 7) or vehicle (n = 6) and control mice treated with saline and previously injected with vehicle (n = 6) of mitochondrial genes Ppargc1a and Tfam and lysosomal genes Lamp1 and Ctsd; one-way ANOVA, post hoc Tukey’s. *p < 0.05 compared to MPTP+sham. ^#p < 0.05 compared to control. All samples were collected 1 day after the last administration of MPTP prior to neuronal cell death taking place. All data are represented as mean ± SEM.

Figure 7

TFEB Overexpression Prevents the Decline of Protein Synthesis that Occurs in the MPTP Model of Parkinson’s Disease (A, B, D, and E) Representative western blot and protein levels of ventral midbrain homogenates in MPTP-intoxicated mice previously injected with AAV-TFEB (n = 7 to 8) or vehicle (n = 7) and control mice treated with saline and previously injected with vehicle (n = 5 to 6) of (A) Ser65-phosphorylated and total 4E-BP1, expressed as a ratio; (B) Thr-37/Thr-46 phosphorylated-4E-BP1/non-Thr-37/Thr-46 phosphorylated-4E-BP1 ratio; (D) phosphorylated eukaryotic translation initiation factor 4E, pEIF4E^S209; and (E) endogenous ribosomal protein S6K1, its activated form (pS6K1^T389), and the ratio; one-way ANOVA, post hoc Tukey’s. *p < 0.05 compared to control. ^#p < 0.05 compared to MPTP+sham group. (C) Eif4e mRNA levels of ventral midbrain homogenates in MPTP-intoxicated mice previously injected with AAV-TFEB (n = 7) or vehicle (n = 6) and control mice treated with saline and previously injected with vehicle (n = 6); one-way ANOVA, post hoc Tukey’s. *p < 0.05 compared to control. ^#p < 0.05 compared to MPTP+sham group. All samples were collected 1 day after the last administration of MPTP prior to neuronal cell death taking place. All data are represented as mean ± SEM.

Figure 8

Knocking Down the Master Transcriptional Repressor of Autophagy ZKSCAN3 Does Not Prevent MPTP-Induced Atrophy or Neuronal Death (A) Left: representative images showing immunofluorescence for tyrosine hydroxylase (green), the transcriptional repressor of autophagy ZKSCAN3 (red), and nuclear staining (blue) in substantia nigra sections in mice after vehicle or AAV-shZKSCAN3 nigral injection. Scale bar, 50 μm. Right: quantification of intracellular ZKSCAN3 immunolabeling intensity of both groups of animals (n = 5 to 6 per group); Mann-Whitney test, *p < 0.05 compared to control. (B) Left: representative photomicrographs of LAMP1-immunostained substantia nigra in mice after vehicle or AAV-shZKSCAN3 nigral injection. Scale bar, 200 μm. Right: quantification of intracellular LAMP-1 immunolabeling intensity of both groups of animals (n = 4 per group); Mann-Whitney test, *p < 0.05 compared to control. (C) Left: representative photomicrographs of S235/236-phosphorylated RPS6-immunostained substantia nigra in mice after vehicle or AAV-shZKSCAN3 nigral injection. Scale bar, 150 μm. Right: quantification of intracellular S235/236-phosphorylated RPS6-immunolabeling intensity of both groups of animals (n = 4 per group); Mann-Whitney test, *p < 0.05 compared to control. (D) Stereological cell counts of dopaminergic neurons in substantia nigra pars compacta in MPTP-intoxicated mice previously injected with AAV-shZKSCAN3 (n = 5) or vehicle (n = 5) and mice treated with saline and previously injected with vehicle (n = 6) or AAV-shZKSCAN3 (n = 5); two-way ANOVA, post hoc Tukey’s. *p < 0.05 compared to control non-injected side. ^#p < 0.05 compared to control injected side. ^$p < 0.05 compared to saline AAV-shZKSCAN3 non-injected side. ˆp < 0.05 compared to saline AAV-shZKSCAN3 injected side. (E) Optical densitometry of striatal TH-positive dopaminergic fibers in MPTP-intoxicated mice previously injected with AAV-shZKSCAN3 (n = 8) or vehicle (n = 7) and mice treated with saline and previously injected with vehicle (n = 7) or AAV-shZKSCAN3 (n = 7); two-way ANOVA, post hoc Tukey’s. *p < 0.05 compared to control non-injected side. ^#p < 0.05 compared to control injected side. ^$p < 0.05 compared to saline AAV-shZKSCAN3 non-injected side. ˆp < 0.05 compared to saline AAV-shZKSCAN3 injected side. (F) Average cell area of dopaminergic neurons in MPTP-intoxicated mice previously injected with AAV-shZKSCAN3 (n = 5) or vehicle (n = 5) and mice treated with saline and previously injected with vehicle (n = 6) or AAV-shZKSCAN3 (n = 6); two-way ANOVA, post hoc Tukey’s. *p < 0.05 compared to control non-injected side. ^#p < 0.05 compared to control injected side. ^$p < 0.05 compared to saline AAV-shZKSCAN3 non-injected side. ˆp < 0.05 compared to saline AAV-shZKSCAN3 injected side. (G) Left: representative photomicrographs of Thr202/Tyr204-phosphorylated MAPK1/3-immunostained substantia nigra in mice after vehicle or AAV-shZKSCAN3 nigral injection. Scale bar, 125 μm. Right: quantification of intracellular Thr202/Tyr204-phosphorylated MAPK1/3-immunolabeling intensity of both groups of animals (n = 4 per group); Mann-Whitney test. (H) Left: representative photomicrographs of S473-phosphorylated AKT-immunostained substantia nigra in mice after vehicle or AAV-shZKSCAN3 nigral injection. Scale bar, 125 μm. Right: quantification of intracellular S473-phosphorylated AKT-immunolabeling intensity of both groups of animals (n = 4 per group); Mann-Whitney test. In all panels, quantifications were performed 21 days after the last MPTP or saline administration. All data are represented as mean ± SEM.