Human Pericardial Fluid Contains Exosomes Enriched with Cardiovascular-Expressed MicroRNAs and Promotes Therapeutic Angiogenesis

Abstract

The pericardial fluid (PF) is contained in the pericardial sac surrounding the heart. MicroRNA (miRNA) exchange via exosomes (endogenous nanoparticles) contributes to cell-to-cell communication. We investigated the hypotheses that the PF is enriched with miRNAs secreted by the heart and that it mediates vascular responses through exosome exchange of miRNAs. The study was developed using leftover material from aortic valve surgery. We found that in comparison with peripheral plasma, the PF contains exosomes enriched with miRNAs co-expressed in patients' myocardium and vasculature. At a functional level, PF exosomes improved survival, proliferation, and networking of cultured endothelial cells (ECs) and restored the angiogenic capacity of ECs depleted (via Dicer silencing) of their endogenous miRNA content. Moreover, PF exosomes improved post-ischemic blood flow recovery and angiogenesis in mice. Mechanistically, (1) let-7b-5p is proangiogenic and inhibits its target gene, TGFBR1, in ECs; (2) PF exosomes transfer a functional let-7b-5p to ECs, thus reducing their TGFBR1 expression; and (3) let-7b-5p depletion in PF exosomes impairs the angiogenic response to these nanoparticles. Collectively, our data support the concept that PF exosomes orchestrate vascular repair via miRNA transfer.

Keywords: angiogenesis; clinical samples; exosomes; extracellular vesicles; human; ischemia; let-7b; microRNAs; pericardial fluid.

Graphical abstract

Figure 1

Human PF Is Enriched with MicroRNAs of Potential Cardiovascular Origin The expression of selected microRNAs (miRNAs) was measured (RT-qPCR) in paired PF (black column) and peripheral plasma (gray column) samples. The PF/plasma concentration ratios of individual miRNAs are plotted on the right side of the graphs. Exogenous spike-in cel-miR-39 was the normalizer. MiR-122-5p was used as a non-cardiovascular control. N.D., not detectable. All values are mean + SEM. *p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001 versus plasma (unpaired Student’s t test); n = 5–9.

Figure 2

PF Contains Nanoparticles of the Size and Antigenic Profile of Exosomes (A and B) Nanoparticle tracking analysis (NTA) was used to quantify the concentration (A) and size distribution (B) of particles in the total PF and plasma. The dotted rectangles evidence exosome-sized vesicles (30–120 nm). (C and D) PF and plasma exosome preparations were positive for the exosomal markers ALIX, FLOT1, EpCAM and CD63. Transmission electron microscopy (TEM) using a gold particle-conjugated anti-CD63 antibody further confirmed exosome identity. In the bottom right corner of (D), next to CD63-immunoreactive exosomes, is shown a larger unreactive cytoplasmic vesicle as negative control (the scale bar represents 100 nm). Unpaired two-tailed Student’s t test was applied. All values are mean + SEM; n = 5.

Figure 3

Human PF Exosomes Carry Cardiovascular miRNAs (A) miRNA level normalized to the spike-in cel-miR-39. (B) PF exosome/total PF ratios calculated for each miRNA. All values are mean + SEM. *p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001 versus plasma (unpaired Student’s t test); n = 5.

Figure 4

PF Exosomes Contain DICER and AGO-2 Protein (A) Validation by immunoblotting of AGO-2 immunoprecipitation (IP) performed on human PF samples using AGO-2 antibody. Mouse non-specific IgG antibody was used as control for the IP. ECs were used as positive control (CTRL). (B) The miRNA expression after AGO-2 IP is presented as fold enrichment relative to IgG; mean + SEM; n = 5. (C) AGO-2 in exosomes enriched from PF samples (representative western blot images). (D) AGO-2 IP was performed on exosomes enriched from PF samples. miRNA expression is expressed as fold enrichment in the AGO-2 IP relative to IgG; mean + SEM; n = 2. (E) Representative western blot images of DICER protein incorporated in the exosomes.

Figure 5

PF-Derived Exosomes Are Incorporated by Cultured ECs and Enhance Their Angiogenic Capacity (A) PF-derived exosomes (10 μg/ml) stained using carboxyfluorescein succinimidyl ester (CFSE, in green) were incubated with ECs for 24 hr. Next, cells were stained with phalloidin (red) and DAPI (blue) (the scale bar represents 25 μm). (B and C) Column graphs show (B) EC apoptosis (measured by caspase-3 activity assay) and (C) EC proliferation (measured by BrdU incorporation) after treatment with 10 μg/ml of either PF- exosomes (black columns) or exosome-depleted PF (gray columns). PBS was used as an additional control (open columns); n = 7. (D) Photomicrographs show endothelial network formation on Matrigel (the scale bar represents 200 μm); 2.5× magnification. (E) Bar graphs show Matrigel assay quantification (total length of EC tube-like structures). All values are mean + SEM. *p ≤ 0.05 and ***p ≤ 0.001 versus PBS, ^{# # #}p ≤ 0.001 versus PF exosome-free (one-way ANOVA, Dunnett’s post hoc test); n = 6.

Figure 6

The PF-Enriched let-7b-5p Is a Proangiogenic MicroRNA (A) ECs were transfected with either a let-7b-5p mimic (right) or a let-7b-5p inhibitor (left). The random sequence anti-miR miRNA inhibitor (scramble) was used as control. Efficiency of let-7b-5p transfection was assessed by qPCR using U6 as a normalizer. (B) Matrigel assay photomicrograph (the scale bar represents 100 μm) and quantification (total length of tube-like structures), 5× magnification. (C and D) Caspase activity (C) and BrdU incorporation (D) of ECs transfected as previously described. (E) Relative expression of the let-7b-5p direct target gene TGFBR1 (measured by PCR using ubiquitin C [UBC] as normalizer). All values are mean + SEM. *p ≤ 0.05, **p ≤ 0.01, and ***p ≤ 0.001 versus the respective scramble control (unpaired Student’s t test); n = 4.

Figure 7

Let-7b-5p Underpins the Angiogenic Action of PF Exosomes ECs were transfected with either a sequence that exhibits no homology to the human genome (scramble) or DICER silencing RNA (siRNA) to knock down (KD) DICER for 24 hr, before being treated with either naive PF exosomes or PF exosomes previously depleted of their let-7b-5p content (exosome let-7b-5p KD). (A and B) Columns graphs show the relative expression of (A) let-7b-5p and (B) TGFBR1 in the exosome recipient ECs and untreated cell controls; n = 4. (C) Photomicrographs (the scale bar represents 100 μm) and (D) quantification of Matrigel assays performed using ECs treated as previously indicated; n = 5; 2.5× magnification. All values are mean + SEM. ^§p ≤ 0.05 and ^§§§p ≤ 0.001 versus scramble, *p ≤ 0.05 and ***p ≤ 0.001 versus DICER KD, ^#p ≤ 0.05 and ^##p ≤ 0.01 DICER KD treated with naive PF exosomes (one-way ANOVA, Tukey’s post hoc test).

Figure 8

PF-Derived Exosomes Promote Reparative Angiogenesis and Blood Flow Recovery in a Mouse Limb Ischemia Model Unilateral limb ischemia was surgically induced in CD1 mice. Next, mice were injected in their ischemic adductor muscle with 100 μg of exosomes from either the PF or plasma. A control group received PBS. (A) Representative color laser Doppler images of lower limb perfusion at baseline and at day 7 post-ischemia. (B) The blood flow recovery to the ischemic foot was calculated as percentage versus the contralateral foot (n = 20 mice/group). (C) Percentage of necrotic toes in animals treated with PBS, PF, or plasma-derived exosomes. **p < 0.01 versus PBS. (D) Representative immunofluorescent images of ischemic adductor muscle sections after staining with green fluorescent isolectin-B4. (E) Capillary densities in the ischemic adductors at 7 days post-ischemia induction (n = 8 per group). Average capillary density was determined from eight randomly selected high-power fields (magnification 20×). ^##p < 0.01 versus PBS. (F and G) Let-7b-5p (F) and Tgfbr1 (G) expression in ischemic mouse adductors injected with PF exosomes or PBS (control); n = 5. *p ≤ 0.05 versus PBS. All values are mean + SEM.