Direct imaging of the recruitment and phosphorylation of S6K1 in the mTORC1 pathway in living cells

Abstract

Knowledge of protein signalling pathways in the working cell is seen as a primary route to identifying and developing targeted medicines. In recent years there has been a growing awareness of the importance of the mTOR pathway, making it an attractive target for therapeutic intervention in several diseases. Within this pathway we have focused on S6 kinase 1 (S6K1), the downstream phosphorylation substrate of mTORC1, and specifically identify its juxtaposition with mTORC1. When S6K1 is co-expressed with raptor we show that S6K1 is translocated from the nucleus to the cytoplasm. By developing a novel biosensor we demonstrate in real-time, that phosphorylation and de-phosphorylation of S6K1 occurs mainly in the cytoplasm of living cells. Furthermore, we show that the scaffold protein raptor, that typically recruits mTOR substrates, is not always involved in S6K1 phosphorylation. Overall, we demonstrate how FRET-FLIM imaging technology can be used to show localisation of S6K1 phosphorylation in living cells and hence a key site of action of inhibitors targeting mTOR phosphorylation.

Figure 1

S6K1 live cell localisation and recruitment onto mTORC1 in HEK293 cells. (a) Confocal image of EGFP-S6K1 only. (b) Western blot validation of the functionality of the EGFP-S6K1 construct. (c) Graph and blot showing S6K1 phosphorylation with and without Rheb co-expression +/− rapamycin (200 nM) treatment. Data taken from Western blot of HEK293 cells and bands quantified using densitometry analysis (ImageJ software 1.48 V), ratio = Phospho-EGFP-S6K1/EGFP-S6K1 where EGFP-S6K1 band correlates to both phosphorylated and unphosphorylated. (d,e) Confocal images of EGFP-S6K1 with mCherry-raptor co-expression. (f) Graph of fraction of mean cytoplasmic/nuclear S6K1 intensities against fraction of mean cytoplasmic/nuclear raptor intensities (C/N) with linear fit. Data representative of three independent experiments with errors representative of standard deviation. Full-length blots are presented in Supplementary Fig. S10 for (b) and in Supplementary Fig. S13 for (c).

Figure 2

Live cell interaction between S6K1 and raptor using FRET-FLIM in HEK293 cells. (a) Confocal image of EGFP-S6K1. (b,c) FLIM of EGFP-S6K1 with corresponding lifetime distribution histogram showing lifetime (τ). (d–g) Confocal images of EGFP-S6K1 with mCherry-raptor co-expression with FLIM and lifetime distribution histogram. (h) Schematic showing summary of interaction. (i) Graph showing cytoplasmic against whole cell lifetimes between EGFP-S6K1 and mCherry-raptor, selected by selecting subcellular regions and obtaining the mode lifetime in SPCImage V6.0. Data representative of three independent experiments with errors representative of standard deviation.

Figure 3

mTOR mediated recruitment of S6K1 onto mTORC1 in the cytoplasm. (a–c) Confocal images of wildtype YFP-mTOR, mCherry-raptor and S6K1-mTurquoise2 triple-colour expression in live HEK293 cells. (d) Western blot for S6K1 phosphorylation (phospho-S6K1) with raptor and S6K1 expression. (e–g) Confocal images of truncated mTOR-mCherry (ΔNmTOR-mCherry), raptor-YFP and S6K1-mTurquoise2 expression. (h) Western blot for S6K1 phosphorylation with truncated mTOR-mCherry and S6K1 expression. (i) Truncated mTOR-mCherry construct showing domains and numbered amino acid sequence. (j) Schematic of possible sequence of recruitment and assembly of mTORC1. Data representative of three independent experiments with errors representative of standard deviation. Full-length blots are presented in Supplementary Fig. S13.

Figure 4

Localisation of phospho-S6K1 in living cells using SensOR. (a) Schematic diagram showing conformational changes in the biosensor to induce FRET. (b) Western blot for SensOR phosphorylation by endogenous mTOR. (c) Confocal images of biosensor alone in HEK293 cells. (d) Cytoplasmic versus nuclear lifetime distributions of EGFP-S6K1. (e,f) FLIM of SensOR with lifetime scale bar below in nanoseconds (ns) and graph showing cytoplasmic against nuclear lifetime distributions of SensOR, selected by selecting sub-cellular regions and obtaining the mode lifetime in SPCImage V6.0 software. Data representative of three independent experiments with errors representative of standard deviation. Full-length blots are presented in Supplementary Fig. S14.

Figure 5

mTORC1 activation and inhibition using SensOR. (a) FLIM of serum and amino acid starved HEK293 cells expressing SensOR. (b) FLIM at 10 minutes following serine + leucine activation. (c) FLIM collected at 40 minutes after subsequent rapamycin treatment of serine + leucine activated cells for 30 minutes. (d) Summary of lifetime changes of SensOR with serum starvation, amino acid addition and rapamycin treatment from mean lifetimes and also pixel by pixel analysis of image with increased binning. Opening and closing of sensor also shown via schematics. Data representative of three independent experiments with errors representative of standard deviation, where two experiments were treated with leucine and the third treated with the combination of leucine and serine.

Figure 6

Summary of FRET-FLIM results and revised pathway and mTORC1 structure. (a) Summary of FRET-FLIM studies for S6K1-mTORC1 interactions (error bars = SD). (b) 3D structural model of S6K1 docked onto the mTORC1 complex made in Swiss PDB Viewer V4.10 software using PDB files: 3a62, 1xts, 5h64. Weak/ transient interaction with mTOR while strong interaction with raptor shown from a monomeric prospective, although the mTOR dimer formation could influence the stoichiometry. (c) Schematic of mTORC1 signalling in live cells showing possible S6K1 recruitment onto mTORC1. Modified from Yadav et al..