Targeted dephosphorylation of TFEB promotes its nuclear translocation

Abstract

Reversible phosphorylation of the transcription factor EB (TFEB) coordinates cellular responses to metabolic and other stresses. During nutrient replete and stressor-free conditions, phosphorylated TFEB is primarily localized to the cytoplasm. Stressor-mediated reduction of TFEB phosphorylation promotes its nuclear translocation and context-dependent transcriptional activity. In this study, we explored targeted dephosphorylation of TFEB as an approach to activate TFEB in the absence of nutrient deprivation or other cellular stress. Through an induction of proximity between TFEB and several phosphatases using the AdPhosphatase system, we demonstrate targeted dephosphorylation of TFEB in cells. Furthermore, by developing a heterobifunctional molecule BDPIC (bromoTAG-dTAG proximity-inducing chimera), we demonstrate targeted dephosphorylation of TFEB-dTAG through induced proximity to bromoTAG-PPP2CA. Targeted dephosphorylation of TFEB-dTAG by bromoTAG-PPP2CA with BDPIC at the endogenous levels is sufficient to induce nuclear translocation and some transcriptional activity of TFEB.

Keywords: Health sciences; Medical specialty; Medicine; Precision medicine.

Graphical abstract

Figure 1

aGFP_6M-AdPhosphatases dephosphorylate phospho-TFEB-GFP (A) Schematic representation of antigen-stabilized anti-GFP nanobody (aGFP_6M)-directed recruitment of either PPP1CA or PPP2CA to C-terminal GFP-tagged TFEB (TFEB-GFP) to mediate TFEB-GFP dephosphorylation. (B) TFEB^+/GFP heterozygous C2C12 myoblasts stably expressing FLAG-aGFP_6M-PPP1CA, catalytically inactive FLAG-aGFP_6M-PPP1CA^H125Q, FLAG-aGFP_6M-PPP2CA or catalytically inactive FLAG-aGFP_6M-PPP2CA^H118Q were lysed and subjected to immunoblot analysis. ○, TFEB-GFP; ●, unmodified endogenous TFEB. (C) TFEB^GFP/GFP/TFE3^−/− C2C12 myoblasts stably expressing FLAG-aGFP_6M-PPP2CA or FLAG-aGFP_6M-PPP2CA^H118Q were lysed and lysates (1 mg protein) subjected to anti-FLAG immunoprecipitation (IP) prior to immunoblot analysis on input extracts (25 μg protein), IPs and post-IP flow-through extracts (25 μg protein) as indicated. (D) TFEB^GFP/GFP/TFE3^−/− C2C12 myoblasts stably expressing FLAG-aGFP_6M-PPP2CA or FLAG-aGFP_6M-PPP2CA^H118Q were treated with vehicle (0.1% DMSO; D), MK-8722 [M] (10 μM) or Torin 1 [T] (100 nM) for 4 h prior to lysis and subjected to immunoblot analysis as indicated. (E) Immunofluorescence staining of GFP were performed in TFEB^GFP/GFP/TFE3^−/− C2C12 myoblasts stably expressing FLAG-aGFP_6M-PPP2CA or FLAG-aGFP_6M-PPP2CA^H118Q. Nucleus was stained with DAPI. Scale bars, 10 μm. Representative images included. (F) The expression of Hexa, Lamp1, Flcn and Fnip1 transcripts was measured by RT-qPCR in TFEB^GFP/GFP/TFE3^−/− C2C12 myoblasts stably expressing FLAG-aGFP_6M-PPP2CA or Flag-aGFP_6M-PPP2CA^H118Q. All data shown in (B–F) are representative of 3 independent experiments. All quantitative data are mean ± SEM from 3 independent experiments. Statistical analysis involved t-test with Mann-Whitney test.

Figure 2

Targeting TFEB-dTAG dephosphorylation through bromoTAG-dTAG proximity-inducing chimera (BDPIC)-mediated recruitment of bromoTAG-PPP2CA (A) Structure of BDPIC with FKBP12^F36V (dTAG) and Brd4BD2^L387A (bromoTAG) binding warheads connected with the indicated PEG linker. (B) Schematic representation of BDPIC-mediated induction of proximity between TFEB-dATG-FLAG and HA-bromoTAG-PPP2CA. (C–F) U2OS cells stably co-expressing TFEB-dTAG-FLAG and HA-bromoTAG-empty, HA-bromoTAG-PPP2CA or HA-bromoTAG-PPP2CA^H118Q were generated. (C) Cells were treated with BDPIC at indicated concentrations for 24 h before lysis and extracts (10 μg protein) were subjected to immunoblot analysis as indicated. (D and E) Cells were treated with 100 nM BDPIC for indicated times prior to lysis. TFEB and PPP2CA protein abundance was analyzed by immunoblotting using the indicated antibodies. (F) Cells were treated with 100 nM BDPIC for 2 h. BDPIC was washed off by PBS and fresh culture media without BDPIC was added to cells. Cells were subsequently cultured for indicated times prior to lysis and immunoblot analysis was performed as indicated. All data (C–F) are representative of 3 independent experiments.

Figure 3

BDPIC-mediated targeted dephosphorylation of endogenous TFEB-dTAG in TFEB^dTAG/dTAG knock-in U2OS cells through overexpressed bromoTAG-PPP2CA (A–E) TFEB^dTAG/dTAG U2OS cells stably expressing HA-bromoTAG-empty, HA-bromoTAG-PPP2CA or HA-bromoTAG-PPP2CA^H118Q were generated. (A) Cells were treated with DMSO [D], MK-8722 [M] (10 μM), Torin 1 [T] (100 nM) or BDPIC [B] (100 nM) for 2 h before lysis and extracts (25 μg protein) were subjected to immunoblot analysis as indicated. (B) Cells were treated with DMSO [D], 100 nM Torin 1 [T] or BDPIC [B] for 2 h prior to lysis. Extracts (25 μg protein) were subjected to immunoblot analysis as indicated. (C) Cells were treated with 100 nM BDPIC for 2 h. DSP-crosslinking was performed prior to lysis. Cell extracts were subjected to anti-HA pull down. TFEB-dTAG and HA-bromoTAG-PPP2CA proteins were analyzed by immunoblotting. Cells were treated with excess amount (10 μM) of dTAG^V-1-NEG PROTAC compound (D) or (1 μM) cis-AGB1 (E) for 30 min prior to treating cells with 100 nM BDPIC. After 2 h, cells were lysed and subjected to immunoblot with the indicated antibodies. All data (A–E) are representative of 3 independent experiments.

Figure 4

BDPIC-mediated targeted dephosphorylation of TFEB-dTAG promotes its nuclear translocation and transcriptional activation of some reported target genes (A) Immunostaining of TFEB was performed in TFEB^dTAG/dTAG knock-in U2OS cells stably expressing HA-bromoTAG, HA-bromoTAG-PPP2CA or HA-bromoTAG-PPP2CA^H118Q treated with DMSO, Torin 1 (100 nM) or BDPIC (100 nM) for 2 h prior to fixation. Nucleus was stained with DAPI. Scale bars, 10 μm. Data are representative of 3 independent experiments. (B) TFEB^dTAG/dTAG knock-in U2OS cells stably expressing HA-bromoTAG, HA-bromoTAG-PPP2CA or HA-bromoTAG-PPP2CA^H118Q were treated with DMSO or 100 nM BDPIC for 2 h prior to lysis. Cytoplasmic and nuclear fractions were isolated and subjected to immunoblot analysis with the indicated antibodies. Data are representative of 3 independent experiments. (C) TFEB^dTAG/dTAG knock-in U2OS cells were treated with DMSO or 100 nM BDPIC for 8 h prior to lysis. The expression of Hexa, Fnip1, Flcn, and Gpnmb transcripts was examined by RT-qPCR. All quantitative data are mean ± SEM from 3 independent experiments. ∗p < 0.05 compared with DMSO treatment. Statistical analysis involved t-test with Mann-Whitney test.

Figure 5

BDPIC-mediated targeted dephosphorylation of TFEB-dTAG via bromoTAG-PPP2CA at the endogenous level (A–E) TFEB^dTAG/dTAG/^bromoTAG/+PPP2CA U2OS cells were generated by CRISPR/Cas9 technology. (A) Cells treated with DMSO, MK-8722 (10 μM), Torin 1 (100 nM) or BDPIC (100 nM) for 2 h prior to lysis and extracts (25 μg protein) subjected to immunoblot analysis. (B) Cells were treated with either DMSO, 100 nM Torin 1 or 100 nM BDPIC for 2 h prior to lysis. Extracts (25 μg protein) were subjected to immunoblot analysis as indicated. (C) Cytoplasmic and nuclear fractions from cells treated with DMSO or 100 nM BDPIC for 2 h were extracted and subjected to immunoblot with the indicated antibodies. (D) Cells were treated with DMSO, MK-8722 (10 μM) or BDPIC (100 nM) for 2 h before fixation and immunostaining for TFEB. Nucleus was stained with DAPI. Scale bars, 10 μm. (A–D) All data representative of 3 independent experiments. (E) Cells were treated with DMSO, Torin 1 (100 nM), MK-8722 (10 μM) or BDPIC (100 nM) for 4 h prior to lysis. The expression of Hexa, Flcn, Gpnmb, and Fnip1 transcripts was examined by RT-qPCR. All quantitative data are mean ± SD from 3 independent experiments, with statistical analysis involving paired t-tests to compare to DMSO treatment. ∗p < 0.05 and ∗∗p < 0.01 compared with DMSO treatment.

Figure 6

Global quantitative proteomic and phospho-proteomic analyses of BDPIC-mediated targeted dephosphorylation of TFEB-dTAG (A) TFEB^dTAG/dTAG/^bromoTAG/+PPP2CA U2OS cells were treated with either DMSO or 100 nM BDPIC for 2 h prior to lysis. Extracts (25 μg protein) were subjected to immunoblot analysis as indicated. (B) Volcano plot showing quantitative changes in the identified proteins. Data plotted represents log₂ of the fold change in protein abundance in BDPIC-treated extracts normalized to DMSO-treated controls against log₁₀ of the p value for each identified protein. The only protein showing a significant change of >2-fold is indicated. (C) Volcano plot showing global phospho-proteome alterations in BDPIC-treated cells compared to DMSO-treated controls. Data plotted represents log₂ of the fold change of phospho-peptides identified in BDPIC-treated cell extracts normalized to DMSO-treated controls against log₁₀ of the p value for each phospho-peptide. Significant alterations of >1.5-fold are indicated. (D and E) Violin plots of individual TFEB phospho-peptides identified by quantitative phospho-proteomic analysis.