Sustained release of endothelial progenitor cell-derived extracellular vesicles from shear-thinning hydrogels improves angiogenesis and promotes function after myocardial infarction

Abstract

Aims: Previous studies have demonstrated improved cardiac function following myocardial infarction (MI) after administration of endothelial progenitor cells (EPCs) into ischaemic myocardium. A growing body of literature supports paracrine effectors, including extracellular vesicles (EVs), as the main mediators of the therapeutic benefits of EPCs. The direct use of paracrine factors is an attractive strategy that harnesses the effects of cell therapy without concerns of cell engraftment or viability. We aim to reproduce the beneficial effects of EPC treatment through delivery of EPC-derived EVs within a shear-thinning gel (STG) for precise localization and sustained delivery.

Methods and results: EVs were harvested from EPCs isolated from adult male Rattus norvegicus (Wistar) rats and characterized by electron microscopy, nanoparticle tracking analysis (NTA), and mass spectrometry. EVs were incorporated into the STG and injected at the border zone in rat models of MI. Haemodynamic function, angiogenesis, and myocardial remodelling were analyzed in five groups: phosphate buffered saline (PBS) control, STG control, EVs in PBS, EVs in STG, and EPCs in STG. Electron microscopy and NTA of EVs showed uniform particles of 50-200 nm. EV content analysis revealed several key angiogenic mediators. EV uptake by endothelial cells was confirmed and followed by robust therapeutic angiogenesis. In vivo animal experiments demonstrated that delivery of EVs within the STG resulted in increased peri-infarct vascular proliferation, preservation of ventricular geometry, and improved haemodynamic function post-MI.

Conclusions: EPC-derived EVs delivered into ischaemic myocardium via an injectable hydrogel enhanced peri-infarct angiogenesis and myocardial haemodynamics in a rat model of MI. The STG greatly increased therapeutic efficiency and efficacy of EV-mediated myocardial preservation.

Figure 1

Characterization of EPC-derived EVs. (A) Representative TEM images of EV isolate show uniform 60–120 nm round particles. (B) NTA demonstrates characteristic size and distribution of EPC-derived EVs. Red shaded areas indicate S.E.M. (C) Network maps of interactions by proteins common to EPCs and EVs, and (D) characteristic exosome markers found in EVs and angiogenic proteins identified in both EPCs and EVs (n = 3).

Figure 2

EV Uptake by HUVECs. (A) Confocal images taken at 63× show uptake of CM-DiI by perinuclear staining. Uptake was reduced after pre-incubation with CytoD at (B) 0.5 µM, (C) 1 µM, and (D) 5 µM. (E) CM-DiI signal intensity normalized to live cells. (F) CM-DiI signal intensity in the supernatant (n = 9, *P < 0.05, **P < 0.01, ***P < 0.001).

Figure 3

Tubule formation assay shows pro-angiogenic effect of EVs after 11 h. (A) HUVECs incubated in EBM alone, (B) with EVs, and (C) with 1 ng/ml VEGF. (D) Quantitative analysis of mean tubule length for each group. (E) Quantitative analysis of isolated segment, indicating disorganized tubule formation. (F) BrdU proliferation assay shows no significant difference in HUVEC proliferation after incubation with EBM alone, EVs, or VEGF (n = 3, *P < 0.05, **P < 0.01).

Figure 4

Steady release of EVs from STG. (A) EVs are added to CD-HA and Ad-HA and loaded into a syringe. CD-HA and Ad-HA interact through guest-host chemistry to form a supramolecular gel. (B) As the syringe plunger is depressed and shear stress is added, the gel shear-thins to permit flow as viscosity decreases with shear. Upon cessation of strain the gel rapidly recovers, re-assembling at the myocardial injection site with the EVs. (C) Confocal microscopy of CM-DiI-labelled EVs within the STG shows an even distribution. (D) EVs are released from the gel over time to permit cellular uptake. Cumulative EV release profile from the STG (n = 2) show steady particle release over 21 day. Error bars are hidden due to their relatively small magnitude.

Figure 5

STG + EPC and STG + EV treatment improves haemodynamics in a rat model of acute MI. Left ventricular haemodynamic function measured by pressure-volume catheter of rat hearts 4 weeks after MI and untreated (PBS Control) or treated with combinations of the STG, EVs, or EPCs (*P < 0.05, **P < 0.01). Sample sizes were as follows: PBS Control, n = 10; STG Control, n = 10; PBS + EV, n = 9; STG + EV, n = 11; STG + EPC, n = 10. Abbreviations: ESPVR, end systolic pressure-volume relationship; +dP/dt, maximum change in systolic pressure over time; −dP/dt, maximum change in diastolic pressure over time; EDPVR, end-diastolic pressure–volume relationship.

Figure 6

STG + EV increases vessel density in peri-infarct myocardium. immunohistochemistry at 20× magnification of peri-infarct myocardium to quantify vascular density in hearts treated with (A) PBS Control, (B) STG Control, (C) PBS + EV, (D) STG + EV, and (E) STG + EPC. Blue, DAPI; green, vWF; red, SAM. Quantification of the mean number of (F) arterioles (G) capillaries (H) and total vessels per HPF. Sample sizes were as follows: PBS Control, n = 8; STG Control, n = 11; PBS+ EV, n = 9; STG + EPC, n = 7, STG + EV, n = 10 (*P < 0.05, **P < 0.01).

Figure 7

Normalized CD11b signal intensity. (A) Quantification of inflammatory cell infiltration into the ischaemic border zone after treatment with PBS Control, STG Control, and STG + EV showed increased inflammation in the STG and STG + EV groups compared with PBS Control 2 days after MI and treatment injection (black data points, n = 3, *P < 0.05). The addition of EVs to the treatment did not increase inflammatory cell recruitment from the baseline amount after the injection of the empty STG. At 4 weeks, the inflammation in the STG Control and STG + EV groups was no different from that in the PBS Control group (grey data points, n = 3). (B) Representative images of the ischaemic border zone at 2 days and 4 weeks. Blue, DAPI; green, CD11b; red, WGA.

Figure 8

STG + EVs improve border zone remodelling post-infarction. representative images of sections taken at the upper, middle, and lower scar 4 weeks after infarct for (A) PBS Control, (B) STG Control, (C) PBS + EV, (D) STG + EV, and (E) STG + EPC where viable muscle is stained red, and collagen scar is stained blue. (F) Quantified scar thickness for all groups. Sample sizes were as follows: PBS Control, n = 8; STG Control, n = 9; PBS + EV, n = 9; STG + EV, n = 10; STG + EPC, n = 7 (*P < 0.05, **P < 0.01).