Carotid Plaque‐Derived Small Extracellular Vesicles Mediate Atherosclerosis and Correlate With Plaque Vulnerability

Figures

• 

  FIGURE 1. Isolation and characterization of plEVs and psEVs from aCAS patients in the sequencing cohort. (A) Representative TEM images of carotid plaque ultrathin sections showed the presence of Ti‐EVs within the extracellular spaces (red arrows). Scale bar = 100 or 1000 nm. (B) A schematic diagram illustrated the isolation of plEVs and psEVs. (C) Representative TEM images of the isolated plEVs and psEVs. Scale bar = 100 nm. (D–E) The size distribution of the isolated plEVs and psEVs was detected by NTA (D), and the particle size and concentration were obtained (E; n = 18/group). (F) The protein contents derived from the isolated plEVs and psEVs were quantified using a BCA kit (n = 18/group). G Representative western blotting bands of typical EV markers TSG101, HSP70, and CD63, and a negative marker calnexin (n = 5/group). (H) Non‐EV contaminants, including ApoA1, ApoB100, collagen I, and collagen III were measured using ELISA kits. Technical replicates = 3. Data were presented as the mean ± SD, and analyzed by t‐test or one‐way ANOVA followed by Bonferroni's multiple comparison test. **p < 0.01 and ***p < 0.001.

• 

  FIGURE 2. Effects of psEVs on endothelial inflammation in vitro and in vivo. (A) Representative 3D horizontal and vertical microphotograph showing PKH26‐labeled psEVs (red) taken up by HUVECs (F‐actin, green). Scale bar = 40 µm. (B) Representative western blotting bands and densitometric quantifications of NLRP3, CL‐caspase‐1, GSDMD‐NT, and mature IL‐1β in the treated HUVECs (n = 5/group). (C) Representative microphotographs of immunostaining and quantitative analysis of Hoechst 33342/PI‐positive cells (n = 5/group). Scale bar = 200 µm. (D) Representative western blotting bands and densitometric quantifications of VE‐cadherin, ICAM‐1, and ICAM‐1 in the treated HUVECs (n = 5/group). (E) Endothelial permeability was measured by FITC‐dextran (70 kDa) transendothelial leakage (n = 5/group). (F) Representative microphotographs and quantitative analysis of the adhesion of THP‐1 monocytes to treated HUVECs (n = 5/group). Scale bar = 200 µm. (G) Schematic of the experimental procedure to test the effect of psEVs on endothelial inflammation and atherogenesis using a high‐fat diet ApoE−/− mouse model. i.v.: intravenous injection. (H) Representative ex vivo in vivo imaging system images of excised…

• 

  FIGURE 3. Bioinformatic analysis of miRNA sequencing of 13 vulnerable psEVs (5 IPH and 8 FCR) compared with five stable psEVs in the sequencing cohort. (A) Volcano plot of the DEmiRNAs between vulnerable and stable psEVs. Red and blue dots represent significantly upregulated and downregulated DEmiRNAs in vulnerable psEVs, respectively. (B) Heatmap of the normalized expression profiles of these DEmiRNAs. (B–F) GO and KEGG enrichment analyses for the target genes of the upregulated (C, D) and downregulated (E, F) DEmiRNAs.

• 

  FIGURE 4. Bioinformatic analysis of miRNA sequencing of psEVs isolated from 8 FCR or 5 IPH plaques in the sequencing cohort. (A) Volcano plot of the DEmiRNAs between psEVs from IPH and FCR plaques. Red and blue dots represent significantly upregulated and downregulated DEmiRNAs in the FCR group, respectively. (B) Heatmap of the normalized expression profiles of these DEmiRNAs. (C–F) GO and KEGG enrichment analyses for the target genes of the upregulated (C, D) and downregulated (E, F) DEmiRNAs.

• 

  FIGURE 5. The expression levels of identified DEmiRNA candidates in plasma‐derived sEVs as detected by qRT‐PCR. (A–D) The relative expression of plasma‐derived sEV miR‐497‐5p, miR‐152‐3p, miR‐204‐5p, miR‐23a‐3p, and miR‐143‐5p in aCAS patients with stable (blue) or vulnerable (IPH: red; FCR: green) plaques in both the discovery cohort (n = 178, 64 stable, 55 IPH, 59 FCR; A, B) and the sequencing cohort (n = 18, 5 stable, 5 IPH, 8 FCR; C, D). (E) Pearson correlation coefficient analyses assessing correlations between DEmiRNAs from psEVs and those from plasma‐derived sEVs in the sequencing cohort (n = 18). Technical replicates = 3. Data were presented as the mean ± SD and were analyzed by t‐test or one‐way ANOVA followed by Bonferroni's multiple comparison test. **p < 0.05 and ***p < 0.001. ns: not significant.

• 

  FIGURE 6. The expression levels of identified DEmiRNA candidates in plasma‐derived sEVs as detected by the HNCIB system. (A, B) The relative expression of plasma‐derived sEV miR‐497‐5p, miR‐152‐3p, miR‐204‐5p, miR‐23a‐3p, and miR‐143‐5p in aCAS patients with stable (blue, n = 64) or vulnerable (IPH: red, n = 55; FCR: green, n = 59) plaques in the discovery cohort. (C) ROC analyses for the plasma‐derived sEV DEmiRNAs to predict plaque vulnerability in the discovery cohort (n = 178). (D, E) The relative expression of plasma‐derived sEV miR‐497‐5p, miR‐152‐3p, miR‐204‐5p, miR‐23a‐3p, and miR‐143‐5p in aCAS patients with stable (blue, n = 42) or vulnerable (IPH: red, n = 18; FCR: green, n = 22) plaques in the external cohort. (F) ROC analyses for the plasma‐derived sEV DEmiRNAs to predict plaque vulnerability in the external cohort (n = 82). Technical replicates = 3. Data were presented as the mean ± SD and were analyzed by t‐test or one‐way ANOVA followed by Bonferroni's multiple comparison test. ***p < 0.001. ns: not significant.

• 

  FIGURE 7. Effects of miR‐497‐5p on endothelial inflammation in ox‐LDL‐treated HUVECs. (A) Representative western blotting bands and densitometric quantifications of NLRP3, CL‐caspase‐1, GSDMD‐NT, mature IL‐1β, VE‐cadherin, ICAM‐1, and VCAM‐1 in the treated HUVECs (n = 5/group). (B) Endothelial permeability was measured by FITC‐dextran (70 kDa) transendothelial leakage (n = 5/group). (C) Quantitative analysis of the adhesion of THP‐1 monocytes to treated HUVECs (n = 5/group). (D) The treated HUVECs were collected and subjected to intracellular reactive ROS measurement (n = 5/group). (E) Representative western blotting bands and densitometric quantifications of NLRP3, CL‐caspase‐1, GSDMD‐NT, mature IL‐1β, VE‐cadherin, ICAM‐1, and VCAM‐1 in treated HUVECs (n = 5/group). Technical replicates = 3. Data were presented as the mean ± SD and were analyzed by one‐way ANOVA followed by Bonferroni's multiple comparison test. *p < 0.05, **p < 0.01, and ***p < 0.001. ns: not significant.

• 

  FIGURE 8. Vulnerable psEVs promoted endothelial inflammation partially through miR‐497‐5p delivery and UCP2 suppression. (A) The binding sites between miR‐497‐5p and the 3′UTR of UCP2 were predicted by an online tool TargetScan, and their binding relationships were verified using a dual‐luciferase reporter assay. (E) The mRNA and protein expression levels of UCP2 in vehicle or psEV‐treated HUVECs (n = 5/group). (D) Representative western blotting bands and densitometric quantifications of NLRP3, CL‐caspase‐1, GSDMD‐NT, mature IL‐1β, VE‐cadherin, ICAM‐1, and VCAM‐1 in the treated HUVECs (n = 5/group). (E) Endothelial permeability measured by FITC‐dextran (70 kDa) transendothelial leakage (n = 5/group). (F) Quantitative analysis of the adhesion of THP‐1 monocytes to treated HUVECs (n = 5/group). (G) The treated HUVECs were collected and subjected to intracellular reactive ROS measurement. Technical replicates = 3. Data were presented as the mean ± SD and were analyzed by one‐way ANOVA followed by Bonferroni's multiple comparison test. *p < 0.05, **p < 0.01, and ***p < 0.001. ns: not significant.