Unilateral HLI was generated via surgical ligation and excision of the femoral artery to create an experimental PAD model, as described previously (16). In the present study, 32 male BALB/c mice (age, 14 weeks; weight, 20–25 g) were anesthetized with 3% isoflurane. Immediately after HLI, the mice were randomized into two groups (n=16 in each): In the control group, the mice received 300 µl olive oil only, and in the curcumin group, the mice received 1,000 mg/kg curcumin (Sigma-Aldrich; Merck KGaA, Darmstadt, Germany) in 300 µl olive oil. All mice received treatment by gavage, once per day for two weeks.

All procedures of the present study followed the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (publication no. 85–23, revised 1996). The experimental protocol was approved by the Committee on Animal Experiments of Wuhan University School of Medicine (Wuhan, China). The BALB/c mice were obtained from the Experimental Animal Center of Wuhan University (Wuhan, China). Mice were housed in a specific pathogen-free laboratory environment at a temperature of 25°C and a constant humidity of 50±10% with free access to food and water under a 12-h light/dark cycle.

Mice were anesthetized and subjected to a non-invasive assessment of ischemic and non-ischemic limb perfusion using a laser Doppler perfusion imaging system (LDPI; Perimed Instruments AB, Stockholm, Sweden) at 0, 7, 14, 21 and 28 days after HLI, as described previously (17). Perfusion of the ischemic limb was quantified and normalized to the non-surgical limb, and the results are presented as a percentage of the values in the non-ischemic side.

Mice were sacrificed in a CO₂ chamber at 28 days after HLI, and the gastrocnemius anterior muscles from the ischemic side were cryo-sectioned in 6-µm sections. Anti-CD31 antibody (rat anti-mouse CD31; cat. no. 550274; 1:100 dilution; BD Pharmingen, San Jose, CA, USA) was applied to acetone-fixed sections (fixed for −20°C for 10 min) of ischemic gastrocnemius muscle tissue, followed by incubation overnight at 4°C with an Alexa Fluor 555 anti-rabbit secondary antibody (1:400 dilution; cat. no. BM2004; Boster Biological Technology, Wuhan, China). Images were acquired using an Olympus IX71 high-magnification microscope (Olympus, Tokyo, Japan). Capillary densities were analyzed by counting in four randomly selected high-power fields (magnification, ×100) and expressed as the number of CD31⁺ cells per field.

Total RNA was isolated from tissue or cells using a PureLink^® RNA Mini kit (cat. no. 12183018A; Thermo Fisher Scientific, Inc., Waltham, MA, USA) according to the manufacturer's protocol. Real-time qPCR for microRNA (miR) quantification and a miR assay (assay no. 001090; cat. no: 4427975; Thermo Fisher Scientific, Inc.) were used for RT-qPCR according to the manufacturer's protocols. Small nucleolar RNA MBII-202 (assay no. 001095; Thermo Fisher Scientific, Inc.) served as an internal control for miR quantification. The quantification cycle (Cq) value obtained for each gene was normalized to that of the respective internal control (ΔCq). Each gene was then further normalized to the average ΔCq value of its control group (ΔΔCq). The final fold expression changes were calculated using the 2^−ΔΔCq equation (18).

Human umbilical vein endothelial cells (HUVECs) were isolated from a donor umbilical cord, as described previously (19), and then cultured in endothelial cell growth medium (Cell Applications, Inc., San Diego, CA, USA) supplemented with 10% fetal bovine serum (Wuhan Boster Biological Technology). To mimic endothelial cells under ischemic conditions as a model for HLI, HUVECs were subjected to hypoxia (2% oxygen; BioSpherix, Lacona, NY, USA) and serum starvation (HSS). The use of HUVECs was approved by the Institutional Review Boards of Wuhan University (Wuhan, China).

In vitro transfection of miRNA inhibitors was used to knock down miR-93 expression in HUVECs, as described previously (20). In brief, a reverse transfection protocol using neofx transfection agent (Ambion, Austin, TX, USA) was used to transfect miR-93 inhibitor, or miRNA inhibitor negative control (cat. no. 4464084; Thermo Fisher Scientific, Inc) into HUVECs for 48 h.

For assessment of cellular viability, HUVECs transfected with an miR-93 inhibitor or control miR were seeded into a 96-well plate at a density of 1×10⁴ cells/well (n=8/group), and then cultured under HSS conditions with/without curcumin (10 µM) for 48 h. Subsequently, the cell viability was assessed using tetrazolium dye incorporation (BioVision, Milpitas, CA, USA).

In vitro angiogenesis assays were performed as previously described (20), under HSS conditions. In brief, HUVECs transfected with miR-93 inhibitor or control were seeded in 96-well dishes coated with growth factor-reduced Matrigel (BD Biosciences, Franklin Lakes, NJ, USA) at a density of 1×10⁴ cells/well, and then exposed to HSS conditions in the presence of curcumin (10 µM) or vehicle alone for 12 h to assess tube formation. Each set of conditions was replicated in 6 wells from the 96-well dish. The degree of tube formation was determined by measuring the length of the tubes and the number of loops in each well under a magnification, ×40 using ImageJ software 1.15K (National Institutes of Health, Bethesda, MD, USA). Each experiment was repeated using at least two different batches of HUVECs in total.

Statistical analysis was performed with GraphPad Prism 7.0 software (GraphPad Inc., La Jolla, CA, USA). An unpaired t-test was used for comparisons between two groups; comparisons between ≥3 groups were performed with one-way analysis of variance and Tukey's post-hoc test. P<0.05 was considered to indicate a statistically significant difference.

LDPI imaging revealed that BALB/c mice receiving curcumin experienced better perfusion recovery at 14, 21 and 28 days after HLI compared with those receiving olive oil (Fig. 1). Immunostaining with CD31 to visualize capillaries indicated that after 28 days of post-HLI treatment, the gastrocnemius anterior muscles from the ischemic side of the mice receiving curcumin exhibited a higher capillary density compared with those receiving olive oil (81.9±4.3 vs. 44.3±2.7 capillaries/field, n=10/group; P<0.01; Fig. 2). A previous study indicated that miR-93 is a potent regulator of angiogenesis in the ischemic limbs of patients with PAD and in animal models (20). Given that curcumin induced angiogenesis as a therapeutic benefit after HLI, the level of miR-93 was assessed, revealing that curcumin treatment increased miR-93 expression by ~5 fold (n=5/group) in the ischemic muscle at 7 days after HLI (Fig. 3A).

In HUVECs cultured under HSS conditions (mimicking in vivo ischemia), curcumin treatment for 12 h significantly increased miR-93 expression (Fig. 3B), consistent with the in vivo results obtained with ischemic muscle tissue. In addition, curcumin increased endothelial cell viability (Fig. 4A) and tube formation (Fig. 4B) in vitro under HSS conditions. It was also revealed that miR-93 knockdown using a miR-93 inhibitor reduced angiogenesis and curcumin-induced angiogenesis and endothelial cell survival were attenuated when miR-93 was knocked down by an miR-93 inhibitor in vitro (Fig. 4A and B).