FN matrix stimulates sEV secretion by VSMCs in vitro.

A, Immobilised FN but not a soluble FN promotes sEV secretion. Cells were cultured for 24h and sEVs in conditioned media were measured by CD63-beads assay. N=3 biological replicates, n=4 technical replicates, ANOVA, ****p<0.0001 B, Micrograph showing VSMC plated onto plastic or 3D matrix. C, VSMCs were plated on the 3D matrix for 24h, fixed and labelled for Syndecan-4 (green), F-actin (phalloidin, red) and fibronectin (blue) Size bar, 10µm. D, 3D matrix promote sEV secretion. Cells were cultured for 24h and conditioned media was collected and sEV secretion was measured by CD63-beads assay. N=5, biological replicates, n=4 technical replicates, t-test E, FN matrix does not affect sEV mode size. VSMCs were plated on non-coated or FN-coated flasks and incubated for 24h. Isolated sEVs were analysed by Nanoparticle Tracking Analysis. Representative from N=3 biological replicates. F, sEV and EV markers distribution is not altered by FN matrix. Cells were plated on non-coated or FN-coated flasks and apoptotic bodies (AB, 1.2K pellet), extracellular vesicles (EV, 10K pellet) and small extracellular vesicles (sEVs, 100K pellet) were isolated by differential ultracentrifugation and analysed by western blotting. Representative image from N=3 biological replicates. G, FN induces secretion of sEVs by activating β1 integrin. VSMCs were plated on non-coated or FN-coated plates in the absence or presence of integrin activating (12G10) or inhibiting (4B4) antibodies for 24h and conditioned media was analysed by CD63-bead assay. N=3 biological replicates, n=4 technical replicates, ANOVA, ****p<0.0001 H, Src is required for the sEV secretion. Cells were plated and sEV secretion was measured as in 2A. N=3, biological replicates, n=4, technical replicates, ANOVA, **p<0.0001 I, Inhibition of FAK blocks FN-induced sEV secretion. Cells were plated and sEV secretion was measured as in 1A. N=3 biological replicates, n=4 technical replicates, ANOVA.

sEV secretion is regulated by Arp2/3 and formin-dependent actin cytoskeleton remodelling.

A, Arp2/3 inhibition with CK666 reduces sEV secretion in VSMC. Cells were plated onto non-coated or FN-coated plates for 24h and conditioned media was analyzed by CD63-bead assay. N=4 biological replicates, n=4 technical replicates, ANOVA. B, Formin inhibitor, SMIFH2 blocking filopodia formation reduces sEV secretion from FN-plated cells only. Cells were plated onto non-coated or FN-coated plates for 24h and conditioned media was analyzed by CD63-bead assay. N=4-5 biological replicates, n=2-4 technical replicates, ANOVA. C, Still images of a time-lapse showing that transported MVBs are attached to F-actin tails. VSMCs were co-transfected with CD63-GFP and F-tractin-RFP and time-lapse was captured using confocal spinning-disk microscopy. Arrow head – position of CD63 MVB across time. Time, min. Size bar, 10µm. D, VSMC were plated for 24h in the absence (top, -ve) or presence of sEV secretion inhibitor (3-OMS, 10μM). Cells were fixed and stained for cortactin, F-actin and CD63. Note an overlap between CD63 endosomes and branched actin proteins (F-actin and cortactin), which is enhanced by 3-OMS treatment (arrow). Size bar, 10µm. E, CD63 MVBs (arrows) are detected in filopodia-like structures. VSMCs were plated on FN-coated plates for 24h and cells were stained for Myo10 (green), CD63 (blue), F-actin (phalloidin, red). Size bar, 10µm. C, Still images of a time-lapse showing that MVBs are transported to the filopodia tip in the live VSMC. Cells were co-transfected by CD63-RFP and Myo10-GFP, cultured for 24h and time-lapse video was captured using confocal spinning disk microscopy. Snapshots were taken at T=3s and T=4s after start of the video.

Endogenous and exogenous sEVs are trapped by ECM in vitro.

A, VSMC were plated onto non-coated (top) or FN-coated (bottom) plastic for 24h and stained for the membrane sEV marker, CD63 and F-actin. Note accumulation of CD63 puncta in close proximity to filopodia-like projections. Size bar, 10µm. B, VSMCs were plated onto gelatin-coated plates and treated with control or SMPD3 siRNA for 72h. 3D matrices were generated and stained for CD63 and FN. Images were acquired using Nikon AX inverted confocal microscope. Note the decrease in CD63-positive sEVs associated with the FN fibrils. C, FN is presented on the surface of the VSMC-derived sEV along with α5β1 integrin. VSMC sEVs were immobilised on the 4µm beads. sEV-beads and VSMCs were stained with the antibodies (filled graphs) in non-permeabilised conditions and analysed by flow cytometry. D, VSMC were plated on FN-coated dishes and Alexa568-labelled sEV were added to the cell media for 3h. Cells were fixed and stained for filopodia marker Myo10 (green) and vinculin (blue). Note perinuclear localisation of internalised sEVs. Size bar, 10µm. Representative image from N=3 biological replicates. E, VSMC were plated on FN-coated dishes pre-coated with Alexa568-labelled sEV and incubated for 24h. Cell staining as in Fig 3D. Note even distribution of sEVs across the matrix and cell area. F, VSMC were plated on the FN-coated dishes in the absence of Alexa568-labelled sEV and incubated for 24h. Cell staining as in Fig 3D. Note the absence of signal in sEV channel.

Fibronectin deposition in the atherosclerotic plaque is spatially associated with sEV markers.

A, Proteomic profiling of pairwise collected carotid atherosclerotic plaques and adjacent intact arterial segments. Venn diagram shows the number of plaque-specific (n = 213) and intact-specific (n = 111) proteins as well as the number of proteins which are common for both vascular regions (1,368). B. Partial least-squares discriminant analysis indicates clear classification pattern between the carotid plaques (indicated by red triangles) and adjacent intact arterial segments (indicated by blue circles). N = 14 patients. C. Volcano plot illustrates that 46 proteins are significantly overexpressed in plaques whilst 13 proteins are significantly upregulated in adjacent intact arterial segments. D, E. Atherosclerotic plaques were co-stained for fibronectin (FN) and sEV marker, CD81. Cell nuclei were counterstained with DAPI. Main figure: x200 magnification, size bar, 50 µm. Box: x400 magnification, size bar, 15µm. Note an accumulation of FN in the neointima. F, Spatial distribution of FN and CD81 in the neointima. Note high overlap between FN and CD81 in the extracellular matrix. x200 magnification, size bar, 50 µm. Manders’ split colocalization coefficient for the overlap of FN with CD81 (M1) and CD81 with FN (M2). Neointima region as in Fig. 4E. G, Co-localisation of endothelial cells (stained for CD31/PECAM1, red colour) and sEVs (stained for CD81, green colour). Nuclei are counterstained with 4′,6-diamidino-2-phenylindole (DAPI, blue colour). Magnification: ×200, scale bar: 50 µm. Note endothelialised (i.e., those covered with CD31-positive cells) segments at the luminal side and abundant CD81 expression both in the neointima and in the tunica media. Endothelial cells (CD31, red colour) are largely co-localised with CD81-positive extracellular vesicles (green colour). Note CD31-positive capillaries in the neointima. H, I, Quantification of FN content in atherosclerotic plaques. Samples were analysed by western blot and bands intensity was quantified in ImageJ. Fold change was calculated as ratio of band intensity in the atherosclerotic plaque to band intensity in the adjacent intact arterial segments normalised to GAPDH. Note that FN content is elevated in atherosclerotic plaques relative to the adjacent intact arterial segments. Paired t-test.

sEV induces directional VSMC invasion.

A, B, VSMC migration in 2D assay. VSMC were plated onto FN in the absence or presence of SMPD3 inhibitor (3-OMS) and/or sEV (25ug). Cells were tracked for 8h. Kruskal-Wallis with Dunn’s multiple comparison test, N=2-3 biological replicates, n indicates the total number of tracked cells per condition, **, p<0.01, ****, p<0.0001. C, Chemotaxis µ-slide diagram. Yellow, cell chamber, blue chemoattractant-free medium chamber, dark blue – chemoattractant medium chamber. D-H, sEV promote directional VSMC invasion. Cells were seeded to the FN-enriched Matrigel matrix in μ-Slide Chemotaxis assay and stained with Draq5. Cell tracking was conducted by OperaPhenix microscope for 12h and cell invasion parameters were quantified using Columbus. Kruskal-Wallis with Dunn’s multiple comparison test, N=4, biological replicates, n indicates the total number of tracked cells per condition, **, p<0.01, ***, p<0.001, ****, p<0.0001.

sEVs induce formation of peripheral FAs.

A, VSMC were plated on FN matrix for 30min and adhered cells were counted by crystal violet staining. N=6 biological replicates, n=4-10 technical replicates, ANOVA, ***p<0.001, ****, p<0.0001 B, C, VSMC spreading onto FN was tracked by using ACEA’s xCELLigence Real-Time Cell Analysis. Note that FN matrix promoted VSMC adhesion but addition of sEVs inhibited cell spreading. N=3 biological replicates, n=2-3 technical replicates, ANOVA. D, VSMCs were spread onto FN for 30 min and cells were fixed and stained with CellMask (magenta) and vinculin (green). Size bar, 10µm. E, F, G, I Quantification of FA number, distance from plasma membrane, cell size and mean FA size per cell, respectively. FA were stained as in 5D and quantified. Representative data from N=3 biological replicates, n=80-82 total number of tracked cells per condition, ANOVA, *p<0.05, **p<0.01. J, Focal adhesion turnover is not affected by sEVs. VSMC were transfected with Paxillin-RFP and plated on the FN in the absence or presence of immobilised sEVs. Images were captured for 30min using confocal spinning disk microscopy and FA turnover was quantified using extracted images analysis, N=4, total n=53-61 cells per condition, Unpaired T-test. K, sEV induces formation of strong-pulling FAs. VSMC transfected with Paxillin-RFP were plated on the PDMS pillars which were covered with FN and sEVs and pillar displacements were quantified. **p<0.01, Unpaired t-test, Data represent 2,170-2322 measurements per condition pooled from two independent biological replicates.

sEVs block focal adhesion formation by presenting collagen VI.

A, Proteomic analysis of VSMC-derived sEVs and EVs. Venn diagram. N=3 biological replicates. B, Protein enrichment in the EV and sEV proteome. Heat Map. N=3 biological replicates. C, Western blot validation of sEV cargos. EVs and sEVs were isolated from VSMC conditioned media by differential ultracentrifugation and analysed by western blotting. Representative image from N=3 biological replicates. D, Collagen VI was presented in sEVs from conditioned media and in ECM-associated sEVs. sEVs were isolated from conditioned media and 0.5M NaCl ECM fraction by differential ultracentrifugation and analysed by western blotting. E, Collagen VI is presented on the surface of sEVs. sEVs were analysed by dot-blot in the non-permeabilised or permeabilised (PBS-0.2%-tween20) conditions. Staining intensity was quantified for non-permeabilised (outside) and permeabilised (inside) conditions. Representative from N=2 biological replicates. F, VSMC adhesion is regulated by collagen VI loaded on sEV. FN matrices were incubated with sEV and anti-collagen VI antibody (COLVI IgG) or control IgG. Cell adhesion was tracked by using ACEA’s xCELLigence Real-Time Cell Analysis. ANOVA, N=3 biological replicates, ***, p<0.001, ****, p<0.0001. G, H, I, sEVs promote directional VSMC invasion. VSMCs were treated with control siRNA (Scramble) or collagen VI-specific siRNA pools for 24h and were seeded to the FN-enriched Matrigel matrix in a μ-Slide Chemotaxis assay and stained with Draq5. Cell tracking was conducted by OperaPhenix microscope for 12h and cell invasion parameters were quantified using Columbus, n indicates the total number of tracked cells per condition, Kolmogorov-Smirnov test, *, p<0.05. I, Real-time PCR analysis of expression of CD9, CD63, CD81, COL6A3, EDIL3 and TGFBI in atherosclerotic plaque. *, p<0.05, Paired t-test, N=5 patients.

FN matrix stimulates sEV secretion by VSMCs.

A, FN and collagen I but not laminin stimulate secretion of CD63-enriched sEVs. VSMCs were plated on the various matrices for 24h and sEV secretion was measured by CD63-beads assay. N=3, biological replicates, n=6 technical replicates. ANOVA **p<0.01 B, Inhibition of SMPD3 blocks sEV secretion by VSMC plated onto FN matrix. VSMCs were plated on non-coated or FN-coated plates for 24h and conditioned media was analysed by CD63-bead assay. N=2 biological replicates, n=4 technical replicates, ANOVA, ***, p<0.001, ****p<0.0001

A, Still images of a time-lapse showing that Arp2/3 and F-actin form tails in VSMC cytosol. VSMCs were co-transfected by ARPC2–GFP and F-tractin-RFP and cultured for 24h. Time-lapse video was captured using confocal spinning disk microscopy. Note, that Arp2/3 and F-actin are observed in lamellipodia but also detected in the cytosol with the unknown activity (arrow). Size bar, 10µm B, Analysis of F-actin/CD63 overlap using ImageJ in a fixed VSMCs. F-actin and CD63 images (Fig 2D) were converted to 8 bits and thresholded. Actin stress fiber structures were extracted from the image by using ParticleAnalysis plug-in (Size (pixel^2) 10-infinity; Circularity 0.5-1) and overlap image created (merged). CD63 endosomes (pseudo coloured in green) were overlapped with F-actin spot-like structures (pseudo coloured in red). Note that endosomes partially overlapped with F-actin tail-like structures (arrows) though some F-actin spots and endosomes were not colocalised (arrowheads). C, sEV secretion detected with CD63-pHluorin. VSMCs were co-transfected with CD63-pHluorin and incubated for 24h. Time-lapse was captured using confocal spinning-disk microscopy. Arrows, typical “burst”-like appearance of sEV secretion at the cell-ECM interface. Arrows, an intense CD63-pHluorin staining along filopodia-like structures indicating that sEV release can occur in filopodia. Size bar, 10µm

SMPD3-dependent sEVs are trapped in ECM.

A, Characterisation of sEVs extracted from VSMC-derived media and ECMs fractions. sEVs were sequentially extracted and fractions were collected and analysed by using ExoView. X-axis – capturing spot antibodies (CD63, CD81, CD9 and control MigG1). B, CD63-bead capturing assay shows similar enrichment in CD63+/CD81+ sEVs in the conditioned media and ECM-associated sEVs extracted using 0.5M NaCl. VSMCs were plated onto gelatin-covered plates for 10 days and media was collected every 3d. sEVs were extracted in a mild (0.5M NaCl) conditions and media and 0.5M NaCl fractions were analysed by using CD63-beads capturing assay. C, D, Exogenously added FN-Alexa568 can be detected in the early endosomes and MVBs. VSMC were serum-deprived for 24h and then incubated with FN-Alexa568 for 30min (C) or 3h (D) in RPMI-0.5% BSA, thoroughly washed and fixed and stained for EEA-1 (C, green) or CD63 (D, blue). Size bar, 10um, E, Exogenously added FN-Alexa568 can be detected in MVBs. VSMC were incubated with FN-Alexa568 for 3h and stained for CD63 which was visualised with anti-mouse-Alexa488 antibody. Size bar, 10um, Note the partial colocalization of FN and CD63. F, G, FN can be detected in the CD63+ MVBs in a sparsely growing VSMCs. VSMC were plated at low (5,000 cells per well, F) or high density (20,000 cells per well, G) and were cultured for 24h in RMPI supplemented with 2.5% EV-free FBS, fixed and stained for F-actin, FN and CD63. H, VSMC plating density influences secretion of CD63+/CD81+ sEVs. Cells were plated at the different density and cultured for 24h and sEVs in conditioned media were measured by CD63-beads assay. N=1 biological replicates, n=8 technical replicates. I, FN is presented in sEV along with β1 integrin. J, VSMC were plated on the FN-coated dishes and Alexa568-labelled sEV were added to the cell media for 3h. Cells were fixed and stained for filopodia marker Myo10 (green) and vinculin (blue). Note perinuclear localisation of internalised sEVs. Size bar, 10µm. 3D projection. Myo10 staining channel is removed. K, VSMC were plated on the FN-coated dishes pre-coated with Alexa568-labelled sEV and incubated for 24h. Cell staining as in Fig S3J. Note even distribution of sEVs across the extracellular matrix and cell area.

Fibronectin deposition in the atherosclerotic plaque is spatially associated with sEV marker CD81.

A. Staining of atherosclerotic plaques with haematoxylin and eosin. Unstable atheroma with an extracellular lipid or necrotic core and calcification (type IV according to the American Heart Association). Below are representative fields of view focusing on the lumen (blue contour), neointima (green contour), and tunica media (yellow contour). Red contours demarcate representative fields of view within these tissue compartments. Note multiple cells at the luminal side, numerous leukocytes in the neointima, and abundant cells in the tunica media. Overview images: magnification: ×1.8, scale bar: 500 µm; close-ups demarcated by blue, green, and gold contour: magnification: ×200, scale bar, 150 µm; close-ups demarcated by red contour: magnification: ×400, scale bar: 50 µm. B. Staining of atherosclerotic plaques with Masson’s trichrome. Note the superficial erosions at the luminal side, abundant haemorrhages in the neointima, and sporadic haemorrhages in the tunica media. Overview images: magnification: ×1.8, scale bar: 500 µm; close-ups demarcated by blue, green, and gold contour: magnification: ×200, scale bar: 150 µm; close-ups demarcated by red contour: magnification: ×400, scale bar, 50 µm. C. Staining of atherosclerotic plaques with orcein. Note the moderate connective tissue staining at the luminal side, weak connective tissue staining in the atheroma with a confluent extracellular lipid core, and strong specific staining at the tunica media containing elastic fibres. Overview images: magnification: ×1.8, scale bar: 500 µm; close-ups demarcated by blue, green, and gold contour: magnification: ×200, scale bar: 150 µm; close-ups demarcated by red contour: magnification: ×400, scale bar, 50 µm. D. Demarcation of tunica media from the neointima using specific antibodies to elastin (red colour) and non-specific autofluorescence of elastic fibers (green colour). Nuclei are counterstained with 4′,6-diamidino-2-phenylindole (DAPI, blue colour). Two representative images per each vascular compartment (luminal side, neointima, and tunica media) are provided. Negative controls (i.e., sections stained with species-specific pre-adsorbed fluorescent-labeled secondary antibodies and DAPI but without the respective antigen-specific primary antibody) are provided to the right from representative images. Magnification: ×200, scale bar, 50 µm. Note that elastin is expressed exclusively in the tunica media and is associated with elastic fibers. E. Expression of FN in atherosclerotic plaques. Atherosclerotic plaques (A) and adjacent intact arterial segments (I) were homogenised and analysed by Western blotting. FN is abundantly presented in both regions and 50 kDa FN fragment (Homandberg et al., 1992) intensity was used to quantify FN amount.

sEVs regulate VSMC motility and invasion.

A, Centripetal FAs are linked to actin stress fibers. VSMCs were plated on FN-coated plate for 24h and cells were stained for Myo10 (green), CD63 (blue) and F-actin (phalloidin, red). Note the dot-like focal complexes in lamellipodium which are not associated with the contractile actin bundles (arrow) and an appearance of elongated FAs associated with the mature actin bundles (arrowhead). Dotted line, approximate position of the lamellipodium boundaries. Size bar, 10µm. B, Mature focal adhesion turnover is not affected by sEVs. VSMC were transfected with Paxillin-RFP and plated on the FN in the absence or presence of immobilised sEVs. Images were captured for 30min using confocal spinning disk microscopy. Note the appearance of the mature FAs in the lamellipodium (Arrowhead). Arrow, direction of the VSMC movement. Representative from N=4 biological replicates. C, sEV induces formation of FAs with the enhanced pulling force. VSMC transfected with Paxillin-RFP were plated on the PDMS pillars which were covered with FN and sEVs and pillar displacements were quantified. Representative image from N=2 biological replicates.

Proteomic analysis of VSMC-derived sEVs and EV.

VSMC-derived EVs and sEVs were isolated from cells and analyzed by protein mass-spectrometry. N=3 biological replicates. A. Clustered proteomic heatmap for EV and sEV. B, The GO term analysis was performed using the differentially expressed genes between the sEV and EV groups.