Chaperonin CCT controls extracellular vesicle production and cell metabolism through kinesin dynamics

Abstract

Cell proteostasis includes gene transcription, protein translation, folding of de novo proteins, post-translational modifications, secretion, degradation and recycling. By profiling the proteome of extracellular vesicles (EVs) from T cells, we have found the chaperonin complex CCT, involved in the correct folding of particular proteins. By limiting CCT cell-content by siRNA, cells undergo altered lipid composition and metabolic rewiring towards a lipid-dependent metabolism, with increased activity of peroxisomes and mitochondria. This is due to dysregulation of the dynamics of interorganelle contacts between lipid droplets, mitochondria, peroxisomes and the endolysosomal system. This process accelerates the biogenesis of multivesicular bodies leading to higher EV production through the dynamic regulation of microtubule-based kinesin motors. These findings connect proteostasis with lipid metabolism through an unexpected role of CCT.

Keywords: CCT; chaperonin; extracellular vesicle; lipid droplet; lipidomic; peroxisome.

FIGURE 1

CCT regulates the proteomic profile of cells and extracellular vesicles. (a) Spectral counts of the 8 CCT subunits and the internal control beta actin recognized by mass spectrometry analysis in siCtrl and siCCT cells. Graphs, means ± SEM. (n = 3). (b) Volcano plot of the proteomic profile identified in siCtrl and siCCT cells (left); and pie chart representing the relative amount of proteins that are increased (dark orange), unchanged (light orange) and decreased (grey) in siCCT versus siCtrl cells (right). A total of 1652 proteins were identified. (c) Volcano plot of the proteomic profile identified in siCtrl and siCCT EVs (left); and pie chart representing the relative amount of proteins that are increased (dark orange), unchanged (light orange) and decreased (grey) in siCCT versus siCtrl‐derived EVs (right). A total of 514 proteins were identified. (d) Canonical pathways predicted by IPA related to proteins increased in siCCT versus siCtrl cells. (e) Venn diagram showing the criteria used to select the proteins that are simultaneously reduced in siCCT cells and increased in siCCT‐derived versus to siCtrl‐derived EVs (dark green). (f) Canonical pathways predicted by IPA for the proteins that are simultaneously reduced in cells and increased in EVs from siCCT condition. (g) Lipid pathways that are altered in siCCT‐derived EVs predicted by IPA. (a–f) A threshold of at least 1 peptide detected was established for this analysis. Fisher exact test was used to calculate the relative p‐value and fold change of the proteomic analysis (n = 3); cells used for this study were Jurkat E6‐1. See also Figure S1.

FIGURE 2

CCT is required for the maintenance of lipid profile in cells. (a) Heatmap of lipid abundance in siCtrl and siCCT cells. The figure shows the clustering results in the form of a dendrogram. Values are measured by Euclidean distance with a Ward clustering algorithm (n = 5 per group); cells used for this study were Jurkat E6‐1. 1–5: sample number. (b) Scheme of lipids differently present in siCCT cells, which are related to EVs biogenesis or structure. Arrows show the increase (red) or decrease (blue) of the lipid amount in the cell. Cer, ceramide; DG, diacylglycerol; MG, monoacylglycerol; PC, phosphatidylcholine; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; TG, triacylglycerol; SM, sphingomyelin.

FIGURE 3

CCT is required for correct lipid droplet internal structures. (a) CryoSXT virtual slices of LDs observed as black vesicles. Bar, 200 nm. (b) 3D segmentation of the reconstructed cryoSXT volume showing MVB (white), LDs (olive green), mitochondria (red), centrioles (light green) and nucleus (blue). Field of view, 100 μm². (c) Quantification of cryoSXT imaging data from (a) and (b) (mean ± interquartile range; Kruskal‐Wallis test; siCtrl n = 108 [8 tomograms], siCCT n = 48 [6 tomograms]). (d) Neutral lipids were analysed by FACS with the Bodipy 493/503 lipid probe (mean ± SEM; two‐tailed unpaired t‐student test; n = 6, representative of 3 independent experiments). (e) Images of LDs (green; Bodipy 493/503) and mitochondria (magenta; Mitotracker Orange) in live cells. Bar, 10 μm. Left graph, LDs per cell. Right graph, LDs contacting mitochondria per cell (estimated distance 0.5 μm; mean ± SD; Mann‐Whitney test; siCtrl n = 63, siCCT n = 62). (f) Images of LDs (green) and lysosomes (magenta; Lysotracker Red) in live cells. Bar, 10 μm. Upper graphs show the number of LD or lysosomes per cell, lower graph shows the ratio of lysosomes contacting LDs and LDs contacting lysosomes per cell (estimated distance 1.2 μm; estimated average diameter of LDs and lysosomes were 0.5 and 0.750 μm, respectively; mean ± SD; Welch's t‐Test; siCtrl n = 47, siCCT n = 48); cells used for this study were Jurkat E6‐1. See also Figure S2.

FIGURE 4

CCT regulates the number of peroxisomes and energy metabolism in cells. (a) Organelle distribution in live siCtrl and siCCT cells spreading onto stimulating anti‐CD3 and anti‐CD28 antibodies coated glass‐bottom chambers; a single plane from confocal volumes is shown. Merge: yellow, Golgi apparatus; cyan, membrane; blue, nucleus; green, mitochondria; magenta, peroxisomes; red, ER. Bar, 10 μm. (b) 3D‐reconstruction of siCtrl and siCCT cells (magenta, peroxisomes; cyan, membrane). Bar, 5 μm. (c–d) Number (c) and volume (d) of peroxisomes per cell. (e–g) Graphs showing the distance from plasma membrane to the peroxisomes (e), the mitochondria (f) or the ER (g) (mean ± SEM; two‐tailed t‐Student test; siCtrl n = 45, siCCT n = 40). (h) Peroxisome distribution in fixed cells was assessed with Pex‐14 antibody in resting Jurkat E6‐1 cells. Images are the sum of slices from a confocal volume. Bar, 5 μm. MFI (0‐255); Graph, peroxisomes per cell (mean ± SD; Mann Withney Test; siCtrl n = 236, siCCT n = 226). (i–k) Lipid peroxidation by flow cytometry (BODIPY™ 581/591 C11 probe): (i) peroxidation signal, B530/30 channel; (j) basal staining, YG586/15 channel; (k) fluorometric ratio (B530/30:YG586/15). Mean ± SEM; two‐tailed t‐Student test; n = 6, representative of 3 independent experiments. (l–m) OCR (l) and ECAR (m) in siCtrl and siCCT cells fed with 25 mM glucose, sodium piruvate and glutamine (1 mM each) and subjected to mitostress test. (n–o) ECAR from siCtrl and siCCT cells fed with glucose 10 mM and subjected to glycolysis stress test (n) or glycolytic rate assay (o). (p–q) OCR (p) and ECAR (q) in siCtrl and siCCT cells fed with palmitate (16:0; 125 μM) and treated or not with 40 μM etomoxir after basal measurements. (l–q), linear mixed model; (n = 5, 4 technical replicates each; from 2 (glucose mitostress test), 1 (glycolysis stress test and glycolytic rate assay) and 3 (palmitate mitostress test) experiments; graphs show means of 3 or 5 (etomoxir) serial measures; treatments are indicated (X‐axis). Px: peroxisome. Memb: membrane. Mito: mitochondria; cells used for this figure were Jurkat E6‐1. See also Figures S3 and S4.

FIGURE 5

CCT controls the biogenesis of MVB and EVs. (a) Imaging of siCtrl and siCCT Jurkat E6‐1 cells. Projections of XY stacks for single bright‐field (BF) and fluorescence images are shown. Merge shows EEA1 (green) and CD63 (magenta). Bar, 10 μm. (b and c) Statistical analysis of EEA1+ particles. (d and e) Statistical analysis of CD63+ particles. (a–e) Mean ± SEM; Mann Whitney test; siCtrl n = 64, siCCT n = 81). (f) CryoSXT virtual slices of MVB. Bar, 200 nm. (g) Statistical analysis of MVB mean radius from the reconstructed tomograms. Median ± interquartile range; Kruskal‐Wallis Test; siCtrl n = 38 (8 tomograms), siCCT n = 26 (6 tomograms). (h) Size distribution analysis by NTA of purified EVs from siCtrl (black) and siCCT (red) cells (representative graph of n = 3 independent experiments). (i) EV production ratio per 10⁶ cells obtained from NTA quantification (Mean ± SEM; Wilcoxon t‐test; n = 7). (j) Western blot analysis of EVs purified fraction. Cells and EVs were blotted for the ER‐specific protein calnexin and for the EV markers TSG101, CD63 and CD81 (representative image of at least 3 independent experiments). Cells used for this study were Jurkat E6‐1.

FIGURE 6

CCT regulates inter‐organellar contacts through the cytoskeleton. (a and b) Tubulin cytoskeleton of siCtrl and siCCT Jurkat cells (a) and HEK 293 cells (b). A single slice from XYZ confocal stacks containing the centrosome is shown. Merge shows α‐tubulin (green) and acetylated tubulin (K40 α‐tubulin; magenta). Graphs, normalized acetylated (K40) α‐Tubulin:α‐Tubulin ratio (Mean ± SEM; Mann Whitney test; (a) siCtrl n = 36, siCCT n = 26 and (b) siCtrl n = 20, siCCT n = 31). Bar, 10 μm. (c and d) Western blot showing distribution of indicated proteins in (c) subcellular fractions (T, total extract; C, cytosol; O, organelles; I, insoluble) and in (d) isolated mitochondria (M) from siCtrl and siCCT HEK 293 cells (T, total; Ex, excluded volume from the columns) (representative image of three independent experiments). (e) Images, Peroxisomes (magenta) and α‐tubulin (green) location in siCtrl and siCCT Jurkat E6‐1 cells. A maximal projection is shown. Plots, MFI and location of the signal in siCtrl and siCCT cells. (f) Graphs, normalized peroxisome number and distance from centrosome to peroxisomes per cell volume in mock, siCtrl and siCCT cells (mock, n = 54, siCtrl n = 101, siCCT n = 102). Graph, mean ± SEM. One way Anova, KHC, kinesin heavy chain; KLC, kinesin light chain. See also Figure S5.