Nucleocytoplasmic Shuttling of FTO Does Not Affect Starvation-Induced Autophagy

Abstract

Polymorphic variants of the FTO (fat mass and obesity) gene associate with body mass index in humans, but the underlying molecular mechanisms have not been firmly determined. FTO is linked to energy homeostasis via amino acid sensing and is thought to activate the mammalian target of rapamycin complex 1, a negative regulator of autophagy. FTO localises both to the nucleus and the cytoplasm, and in this study we identify a functional nuclear localisation signal (NLS) in the N-terminus of FTO, as well as nuclear localization information in its very C-terminus. Inhibition of FTO nuclear transport has no effect on autophagy and in contrast to a previously proposed role of FTO in autophagy, we find no difference in starvation-induced autophagy in control cells compared to a panel of cell types depleted of FTO. Future studies that further characterise the cellular functions of FTO will be important to understand why variants in FTO are associated with body weight.

Fig 1. FTO is ubiquitously expressed.

(A) Western blot analysis of the indicated mouse tissues probing for Fto using antibody from Epitomics (see Table 1). BAT; brown adipose tissue, WAT; white adipose tissue. (B) Protein extracts from the indicated human cell lines were analysed by immunoblotting using mouse anti-β-actin and rabbit anti-FTO (Epitomics) antibodies. (C) Various subcellular fractions of U2OS cells were immunoblotted using rabbit anti-FTO (Epitomics), anti-HSP90, anti-EGFR, anti-PML, anti-Histone-3 and anti-Vimentin. One representable experiment (out of two) is shown.

Fig 2. FTO contains a functional NLS.

(A) Graphical description of the constructs used. Red denotes the N-terminal Fe(II) and 2-Oxoglutarate dependent Dioxygenase domain. Blue denotes the C-terminal domain without any known function. (B) Confocal microscopy images of HeLa cells transiently transfected with the denoted GFP-tagged constructs. (C) Graphical description of the constructs used. KKLR-> AALR denotes mutation of lysine residues to alanine residues in amino acid number 15 and 16 in the canonical sequence. (D) Confocal images of U2OS cells transiently transfected with the denoted constructs. Scale bar 20 μm.

Fig 3. The extreme C-terminal part of FTO facilitates nuclear shuttling.

(A) Graphical description of the constructs used. The numbers denotes the amino acids the constructs contain. (B) Confocal microscopy images of U2OS transiently transfected with the denoted GFP-tagged constructs. Scale bar 20 μm. (C) Confocal microscopy images of U2OS transiently transfected with the denoted GFP-tagged constructs. Scale bar 20 μm. (D) Diagram of the construct FTO 32–495 highlighting the 10 amino acid necessary for nuclear shuttling and an alignment of amino acid 486–495 in the denoted species.

Fig 4. FTO has no significant effect on autophagy.

(A) Quantification of LC3B puncta in cells transfected with GFP, full length GFP-FTO or GFP-FTO 32–475, treated either with complete media (CM) or EBSS starvation media (Starved). >1000 cells counted in each condition. Data shown is mean ± SEM. (B) Representative images of cells used for quantification of LC3B puncta, shown in red. Scale bar 20 μm. (C) Western blot analysis of protein lysates from Fto^+/+ MEFs and Fto^-/- MEFs treated either with complete media (CM), complete media and BafA1 (BafA1), EBSS starvation media (Starved) or EBSS starvation media and BafA1 (Starved + BafA1). All treatments were for 4 hours. The blot is representative for five experiments. (D) Quantification of LC3B-II normalised to beta-actin and to CM Fto^+/+ from five experiments as shown in A. The graphs show mean ± SEM (E) Quantification of p62 normalised to beta-actin and to CM Fto^+/+ from three experiments as shown in A. The graphs show mean ± SEM. (F) Long-lived protein degradation in HeLa cells treated as in A. The graph shows results from five experiments, presented as mean ± SEM.