Release of basic fibroblast growth factor from acoustically-responsive scaffolds promotes therapeutic angiogenesis in the hind limb ischemia model

Abstract

Pro-angiogenic growth factors have been studied as potential therapeutics for cardiovascular diseases like critical limb ischemia (CLI). However, the translation of these factors has remained a challenge, in part, due to problems associated with safe and effective delivery. Here, we describe a hydrogel-based delivery system for growth factors where release is modulated by focused ultrasound (FUS), specifically a mechanism termed acoustic droplet vaporization. With these fibrin-based, acoustically-responsive scaffolds (ARSs), release of a growth factor is non-invasively and spatiotemporally-controlled in an on-demand manner using non-thermal FUS. In vitro studies demonstrated sustained release of basic fibroblast growth factor (bFGF) from the ARSs using repeated applications of FUS. In in vivo studies, ARSs containing bFGF were implanted in mice following induction of hind limb ischemia, a preclinical model of CLI. During the 4-week study, mice in the ARS + FUS group longitudinally exhibited significantly more perfusion and less visible necrosis compared to other experimental groups. Additionally, significantly greater angiogenesis and less fibrosis were observed for the ARS + FUS group. Overall, these results highlight a promising, FUS-based method of delivering a pro-angiogenic growth factor for stimulating angiogenesis and reperfusion in a cardiovascular disease model. More broadly, these results could be used to personalize the delivery of therapeutics in different regenerative applications by actively controlling the release of a growth factor.

Keywords: Acoustic droplet vaporization; Angiogenesis; Basic fibroblast growth factor; Drug delivery; Fibrin; Hind limb ischemia; Phase-shift emulsion; Ultrasound.

Figure 1.

Focused ultrasound (FUS) was used to control release of basic fibroblast growth factor (bFGF) from an acoustically-responsive scaffold (ARS). The ARS consists of a fibrin matrix doped with a phase-shift double emulsion containing bFGF. The FUS-based release mechanism is termed acoustic droplet vaporization (ADV). A) At subthreshold acoustic pressures (i.e., P < P_ADV), the perfluorocarbon (PFC) phase within the emulsion remains a liquid. At suprathreshold pressures (i.e., P > P_ADV), the PFC phase is vaporized into a bubble, thereby causing release of the encapsulated bFGF. Due to its high bulk boiling point, the PFC used in this study ultimately recondenses [28]. B) An ARS was implanted at the site of femoral artery resection in mice. Suprathreshold FUS was transcutaneously applied to release bFGF, which subsequently stimulated angiogenesis and reperfusion. The figure was created with BioRender.com.

Figure 2.

In vitro release of bFGF from ARSs correlated with the number of suprathreshold FUS exposures. A) On day 1, ARSs containing 3.3% (v/v) bFGF-loaded, phase-shift emulsion were exposed to FUS either once or every three days (denoted by red arrows). Data are represented as mean ± standard deviation (N=5 per group). Statistically significant differences (p < 0.01) are denoted as follows. α: ARS + FUS (day 1) vs. ARS + FUS (every 3 days). B) ARSs were immunohistochemically stained for bFGF. Brightfield images reveal morphological differences caused by FUS-induced vaporization of the phase-shift emulsion. Scale bar: 200 μm.

Figure 3.

The greatest therapeutic efficacy, as evaluated visually based on the level of discoloration and/or necrosis in the left leg, was observed in mice that received ARSs in conjunction with FUS applied every three days (i.e., ARS + FUS). A) Representative, longitudinal images of mice from the five experimental groups are shown. The ischemia score for each image is displayed as an inset. B) The level of ischemia was quantitatively scored across the 28 day study. The ischemia score is inversely related to the amount of ischemia, with a score of 13 assigned to a healthy leg prior to surgery. Data are represented as mean ± standard error of the mean (N=12–13 per group for days 0–14 and N=6 for days 21–28). Statistically significant differences (p < 0.05) are denoted as follows. α: no intervention vs. ARS; β: no intervention vs. ARS + FUS; γ: FUS vs. ARS; δ: FUS vs. ARS + FUS; ε: fibrin + bFGF vs. ARS + FUS; and ζ: ARS vs. ARS + FUS. C) Kaplan-Meier curves are shown for the five experimental groups. Mice that displayed necrosis exceeding the approved endpoint were euthanized, which occurred between days 6 and 11 of the study.

Figure 4.

Perfusion was longitudinally tracked using laser speckle contrast analysis (LASCA) imaging. A) Representative, longitudinal images of mice from the five experimental groups are shown. B) Perfusion was assessed in four different regions of interest (ROIs) within the leg as well as in the foot (not shown). This image shows a leg prior to surgery. C) For each ROI, a perfusion ratio was calculated by normalizing the average relative perfusion in the ischemic limb by the normal limb. The longitudinal profiles for the calf muscle ROI are shown. Data are represented as mean ± standard error of the mean (N=12–13 per group for days 0–14 and N=6 for days 21–28). Statistically significant differences (p < 0.05) are denoted as follows. α: no intervention vs. ARS +FUS; β: FUS vs. ARS + FUS; γ: fibrin + bFGF vs. ARS + FUS; and δ: ARS vs. ARS + FUS.

Figure 5.

The density of blood vessels was assessed using CD31 immunohistochemistry on day 14 (A, B) and day 28 (C, D) within the left (i.e., ischemic) as well as right (i.e., normal) legs. Four different regions of muscle were analyzed: left lower leg (LLL), left upper leg (LUL), right lower leg (RLL), and right upper leg (RUL). All groups underwent HLI surgery except for the ‘no surgery’ group. Data are represented as mean ± standard error of the mean (N=6–7 per group for day 14, N=6 for day 28, and N=4 for the ‘no surgery’ group). Statistically significant differences (p < 0.05) are denoted by brackets as well as the following symbols. α: LLL at day 14 vs. RLL at day 14; β: LLL at day 28 vs. RLL at day 28; γ: vs. ARS + FUS (LLL at day 14); δ: vs. ARS + FUS (LUL at day 14); ε: vs. ARS + FUS (LLL at day 28); and ζ: vs. ARS + FUS (LUL at day 28).

Figure 6.

Fibrosis was evaluated based on Masson’s trichrome staining. A) Representative images are shown from the left lower leg (LLL) and left upper leg (LUL) from the experimental groups on day 28 as well as control mice (i.e., ‘no surgery’). B) The area percent of each image that stained positive for collagen (i.e., blue) was quantified. Data are represented as mean ± standard error of the mean (N=6 for each experimental group and N=4 for the ‘no surgery’ group). Statistically significant differences (p < 0.05) are denoted as follows. α: vs. no surgery.

Figure 7.

The density of macrophages was assessed using CD68 immunohistochemistry on day 28 within the left (A) and right legs (B). Four different regions of muscle were analyzed: left lower leg (LLL), left upper leg (LUL), right lower leg (RLL), and right upper leg (RUL). Data are represented as mean ± standard error of the mean (N=6 per group). An asterisk (*) denotes each group where zero CD68+ cells were counted. Statistically significant differences (p < 0.05) are denoted by brackets as well as the following symbols. α: LLL vs RLL; β: LUL vs. RUL; γ: vs. ARS + FUS (LLL); and δ: vs. ARS + FUS (LUL).