Comparative proteomics of exosomes secreted by tumoral Jurkat T cells and normal human T cell blasts unravels a potential tumorigenic role for valosin-containing protein

Abstract

We have previously characterized that FasL and Apo2L/TRAIL are stored in their bioactive form inside human T cell blasts in intraluminal vesicles present in multivesicular bodies. These vesicles are rapidly released to the supernatant in the form of exosomes upon re-activation of T cells. In this study we have compared for the first time proteomics of exosomes produced by normal human T cell blasts with those produced by tumoral Jurkat cells, with the objective of identify proteins associated with tumoral exosomes that could have a previously unrecognized role in malignancy. We have identified 359 and 418 proteins in exosomes from T cell blasts and Jurkat cells, respectively. Interestingly, only 145 (around a 40%) are common. The major proteins in both cases are actin and tubulin isoforms and the common interaction nodes correspond to these cytoskeleton and related proteins, as well as to ribosomal and mRNA granule proteins. We detected 14 membrane proteins that were especially enriched in exosomes from Jurkat cells as compared with T cell blasts. The most abundant of these proteins was valosin-containing protein (VCP), a membrane ATPase involved in ER homeostasis and ubiquitination. In this work, we also show that leukemic cells are more sensitive to cell death induced by the VCP inhibitor DBeQ than normal T cells. Furthermore, VCP inhibition prevents functional exosome secretion only in Jurkat cells, but not in T cell blasts. These results suggest VCP targeting as a new selective pathway to exploit in cancer treatment to prevent tumoral exosome secretion.

Keywords: T cells; apoptosis; exosomes; leukemia; proteomics.

Figure 1. Localization of FasL and Apo2L/TRAIL in exosomes derived from human T cell blasts

Human T cell blasts were stimulated with PHA for 5 minutes, PHA washed out, and supernatants collected 1 hour later A. or they were stimulated with immobilized anti-CD59 mAb VJ1/12.2 for 3h and supernatants collected B.. Exosomes present in supernatants were placed on bacitracin-treated grids as indicated in Material and Methods and observed by TEM. In both cases, FasL (15 nm dots) and APO2L/TRAIL (5 nm dots, marked with arrows in A) were located using, respectively, N20 rabbit pAb and 5C2 mouse mAb, plus secondary anti-rabbit and anti-mouse IgGs.

Figure 2. Lipid composition of Jurkat cells and of derived exosomes

Cells were first metabolically labeled with [¹⁴C]-acetate during 72h. Then, cells were stimulated with PHA as indicated in the legend of Figure 1, cells separated from supernatants by centrifugation, and exosomes isolated from the supernatants as indicated in Materials and Methods. Cell or exosome lipids were extracted and subjected to thin-layer chromatography, and radiolabelled bands revealed by autoradiography. Bands were marked in the thin layers, scraped, and radioactivity determined in a b-counter by liquid scintillation. Results are shown as percentage of the total labeling (left graphic) or of the labeling corresponding to neutral lipids (right graphic). PI, phosphatidylinositol; PS, phosphatidylserine; SM, sphingomyelin; PC, phosphatidylcholine; PE, phosphatidylethanolamine; CL, cardiolipin; NL, neutral lipids; TAG, triacylglycerol; FFA, free fatty acids; Chol, cholesterol; DAG, diacylglycerols; MAG, monoacylglycerol. The results are expressed as the mean±SD of two different analysis. *, P<0.05.

Figure 3. Initial immunoblot analysis of proteins expressed in cells and in exosomes

Protein extracts were obtained from cells or from exosomes, proteins separated by electrophoresis in Nu-PAGE Bis-Tris 12% gels, proteins transferred to PVDF membranes and proteins revealed by immunoblot using specific antibodies. In all cases, extracts correspond to 5 μg of total protein. A. proteins from whole Jurkat cell extracts (1) or from exosomes secreted by Jurkat cells after pulse-stimulation with PHA (2). B. proteins from whole Jurkat cell extracts (1), from exosomes secreted by Jurkat cells after pulse-stimulation with PHA (2) or from exosomes secreted by human T cell blasts pulse-stimulated with PHA (3). In the case of CD63, proteins from non-stimulated Jurkat cells (1), from Jurkat cells pulse-stimulated with PHA (2), from exosomes secreted by Jurkat cells after pulse-stimulation with PHA (3) or from exosomes secreted by human T cell blasts pulse-stimulated with PHA (4). Molecular weights are indicted on the left.

Figure 4. 2D separation of proteins in Jurkat exosomes

A 2D gel was performed on a sample of 30 μg total protein extracted from Jurkat exosomes and protein spots silver-stained. Isoelectric point is indicated above and molecular weight on the left. The image is representative of at least 5 different gels performed. Numbered spots were identified by MALDI-TOF, see text of Results for details.

Figure 5. Exosomal proteins detected by LCMSMS

4 different samples obtained from Jurkat cells and samples obtained from 4 different donors were analyzed by LCMSMS. Upper panel, the number of proteins with an statistically significant identification identified in exosomes from Jurkat cells or from human T cell blasts, indicating the number of proteins that were common are represented in a Venn diagram; Lower panels, protein interaction networks obtained using the STRING software between proteins identified in human T cell blast exosomes (middle panel) or in Jurkat cells (lower panel).

Figure 6. Analysis of VCP expression in cells and exosomes

A. Levels of VCP protein both in cell extracts and in exosomes were analyzed by Wetern blot. Both T cell blastas and Jurkat cells were unstimulated or stimulated with PHA for 5 min. Then, PHA was removed and cells were suspenden in fresh medium. Finally, supernatants were collected 1h later and proteins extracted. Cell and exosome proteins were separated by electrophoresis and VCP immunoblotted with an specific antibody. The levels of β-actin were also determined in the same gels as a control of protein loading. B. Upper graphs show the relative intensity of the VCP bands as a ratio with cellular β-actin expression, analyzed using the ImageJ software. Lower graph shows the VCP enrichment in exosomes, expressed as the ratio between exosome and celular VCP in each experimental condition. The results are expressed as the mean±SD of two different experiments.

Figure 7. Effect of the VCP inhibitor DBeQ on leukemic and normal lymphocytes

Dose-response assays using the indicated doses of DBeQ were performed on freshly isolated PBMC, derived T cell blasts and Jurkat cells A., or on leukemic U937, Mec1 and Jurkat cells B.. The decrease in cell viability was measured using the MTT assay after 24h of incubation. The results are expresses as the mean±SD of at least two different experiments. **, P < 0.01.

Figure 8. Effect of the VCP inhibitor DBeQ on exosomes release from T cell blasts or from tumoral Jurkat cells

A. Jurkat cells were either left untreated (control) or they were treated overnight with 1 μg/ml of soluble TRAIL or with 50 ng/ml of the anti-Fas mAb CH11, in the presence or absence of 30 μM of the caspase inhibitor Z-VAD-fmk, as indicated (white bars). The possible effect of DBeQ was tested using two incubation protocols. In the first one (black bars), 3 μM DBeQ was present during the overnight assay, and in the second (grey bars), cells were pre-incubated with 3 μM for 16h before the incubation with anti-Fas of with TRAIL and the assay performed in the absence of DBeQ. Cell death was determined by flow cytometry using 7-amino-actinomycin D (7-AAD) staining. The results are expressed as the mean±SD of at least two different experiments. B. Jurkat cells were incubated in the presence or absence, as indicated, of 3 μM DBeQ for 16h, cell extracts were obtained, and the expression of TRAIL or FasL was determined by immunoblot. C, D. T cell blasts and Jurkat cells were cultured in the presence or in the absence of 3 μM DBeQ for 16h. Then, DBeQ was removed and cells were stimulated with 10 ng/ml phorbol myristate-acetate (PMA) plus 600 nM ionomycin during 2h. After that, cell supernatants were collected by centrifugation, and their cytotoxicity was tested overnight against non-activated Jurkat cells. Finally, cell death was determined by flow cytometry using 7-amino-actinomycin D (7-AAD) staining. The results are expressed as the mean±SD of at least two different experiments. **, P < 0.01. C. Cytotoxic effect of supernatants from T cell blasts. D. Cytotoxic effect of supernatants from Jurkat cells.