All animal procedures were performed in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals and were approved by the Animal Ethics Committee of Guangdong Academy of Medical Sciences. FVB male mice were acquired from Harlan Laboratories Inc. (Frederick, MD), weighing 25–30 g and aging 6–7 weeks. All surgical processes and imaging scan session were performed under anesthesia with 1.0–2.0% isoflurane in oxygen delivered at a flow of 1.0 L/min.

An electric shaver was used to remove the hair on both hindlimbs of each mouse with the help of depilatory cream. The skin on the top of thighs was then incised to expose the arteries, veins, and nerves. The mouse hindlimbs mainly include the thigh muscles in the top layer and the gastrocnemius in the deep layer. For the induction of right hindlimb ischemia, we ligated and excised a section of femoral artery and all branches derived from this section of femoral artery. The proximal and distal ends of the superficial femoral artery and vein, as well as the origins of all branches were ligated with surgical silk, size 6-0. Each of the vessels surrounded by three ligated points was excised. Femoral nerve was carefully separated from the contaminant vessels and preserved. The sham operated left hindlimb served as a control with an incision in the skin of the thigh (16).

As stimulus for ischemic hindlimb, a single dose of VEGF (0.5 μg) in 100 μl PBS was injected into inflamed gastrocnemius muscles below the site of occlusion at three different sites at day 3 post surgery and daily thereafter for a total of 3 days The same amount of PBS (100 μl) alone was injected intramuscularly in the same way after the induction of ischemia, which served as the non-treatment group.

Laser Doppler imaging (LDI) was used to serially monitor the blood flow of the ischemic hindlimbs. Briefly, anesthetized mice were placed in supine position on a 37.4–38.0°C heating pad and then imaged using an analyzer (PeriScan-PIM3 Perimed AB, Sweden). The blood perfusion was quantitatively evaluated by perfusion ratio (ratio of average LDI index of ischemic to non-ischemic hindlimb) by LDPI win 3.1.3 (Perimed AB,Sweden).

Before and 1 day post-surgery, Power Doppler Imaging (PDI) and Color Doppler Imaging (CDI) were performed using a Vevo 2100 Imaging Systems (VisualSonics, Inc., Toronto, ON, Canada). After FVB mice were anesthetized, flow velocity and spatial vascular profile in right hindlimb were tested by Power Doppler with a linear transducer in three-dimensional mode. Meantime, Color Doppler mode scan in three-dimension was also performed to provide a visual overview of flow within the normal and ischemic skeletal muscle tissue as well as delineate the flow direction and velocity by red and blue color spectra.

The synthesis of ¹⁸F-NEB was performed according to the procedure reported previously (14). In brief, 0.13 ml of acetonitrile and 0.05 ml of aqueous ¹⁸F-fluoride (0.3–0.9 GBq) were added to a 1-ml plastic tube containing 3 ml of 2 mM aluminum chloride in 0.5 M, pH 4.0, sodium acetate buffer and 6 ml of 3 mM NEB in 0.5 M, pH 4.0, sodium acetate buffer. The mixture was stirred in a vortex mixer and heated in a 105°C heating block for 10 min. The vial was cooled, and the solution was diluted with 10 ml of water and trapped on a Varian Bond Elut C₁₈ column (100 mg). The radioactivity trapped on the C₁₈ column was eluted with 0.3 ml of 80% ethanol/water containing 1 mM HCl. The ethanol solution was evaporated with argon flow, and the final product was dissolved in phosphate-buffered saline (PBS) and analyzed by HPLC (15).

PET scans were performed using an INVEON small animal PET scanner (Siemens Preclinical Solutions). As designed, PET scan was done at 1, 3, 5, 7, 10, and 14 days post-surgery respectively. At each day mentioned, mice were anesthetized and approximately 3.75 MBq (100 μCi) of ¹⁸F-NEB tracer was administered via tail vein injection. Then 10-min static PET images were acquired at 15 min and 2 h after tracer injection respectively. The images were reconstructed by a three-dimensional ordered subsets expectation maximization (3D-OSEM) algorithm and analyzed using ASI Pro VMTM software. The mean (%ID) uptake of the ischemic hindlimb tissue was determined by drawing three-dimensional regions of interest (ROIs) surrounding an entire limb on the coronal images. The radioactivity contained in the ROI divided by the dose administered to the animal gave the %ID. Each group contained 5–6 mice.

Ischemic skeletomuscular tissues at different time points post HI surgery were harvested and fixed in 4% paraformaldehyde solution, dehydrated through graded solutions of ethanol, and embedded in paraffin. Serial sections (5 μm thick) were cut and mounted on glass slides (Fisher, Pittsburgh, PA). After dewaxing and microwave antigen retrieval, slides were incubated with 10% normal goat serum for 1 h and then overnight at 4°C with mouse monoclonal antibody to CD31 (1:100) or PBS as a control. Biotinylated anti-mouse IgG (Sigma) was applied and detected with a streptavidin-peroxidase complex and 0.1% of 3,3′-diaminobenzidine (Sigma) in PBS with 0.05% H₂O₂ for 5 min at room temperature. In addition, slides were stained with hematoxylin and eosin (H & E) for tissue morphology (17).

Results were expressed as mean and standard deviation (mean ± SD). Statistical analysis was performed using one-way ANOVA followed by the Bonferroni multiple comparison test or Student t-test. P < 0.05 was considered as statistically significant.

We used a well-established mouse hindlimb ischemia (HI) model to mimic PAD. Ligation and excision of a section of femoral artery and all branches originated from this section were performed in the right hindlimb of FVB mice. The sham operated left hindlimb served as the control. Laser Doppler Imaging (LDI) showed significantly diminished blood flow in the right hindlimb at day 1 post surgery (Figure 1). Power Doppler Imaging (PDI) (18) and Color Doppler Imaging (CDI) were also utilized to assess the blood perfusion within ischemic area. Both three-dimensional PDI and CDI scans displayed the abrupt blockade of blood flow and the absence of multiple branches in femoral artery at the site of occlusion, as compared to the clear and intact vasculature before the surgery (Figure 1). All these results indicated successful establishment of mouse hindlimb ischemia model.

As the chemical structure in Figure 2A showed, ¹⁸F-AIF-NOTA-EB (¹⁸F-NEB) was developed by labeling a NOTA (1,4,7-triazacyclononane-N,N′,N″-triaceticacid)- conjugated truncated EB (NEB) with ¹⁸F through the formation of ¹⁸F-aluminum- fluoride complex. Due to the high concentration of albumin in the blood circulation, only a trace amount of NEB is needed for NEB-albumin complex formation and equilibrium, the process of which takes just a few minutes (Figure 2B). In the early phase post injection (p.i.), most of the ¹⁸F-NEB-albumin complex is restrained in the blood circulation (Supplementary Figure S1), allowing assessment of blood volume (Figure 2C). In response to inadequate blood flow and oxygenation, acute inflammation is initiated and characterized by significant immunovascular changes (19), involving vasodilation, increased vascular permeability and the movement of plasma fluid containing plenty of serum proteins (20). Gradually, ¹⁸F-NEB-albumin complex is extravasated into interstitial space. Therefore, PET imaging with radiolabeled NEB in the late phase p.i. permits non-invasive evaluation of vascular permeability in HI model (Figure 2C).

Fifteen minutes after intravenous administration of ¹⁸F-NEB probe (100 μCi per mouse), static PET scans were performed to assess blood perfusion at 1, 3, 5, 7, 10, and 14 days post-surgery, respectively. As shown in Figure 3A, ¹⁸F-NEB signal intensity in surgical hindlimb (area within white dotted line) was significantly diminished in the first day post-surgery (0.154±0.029) and gradually increased over time, while the non-surgical hindlimb showed no changes. At 14 days post-surgery (0.812 ± 0.068), local signal of the surgical hindlimb was significantly augmented, suggesting recovery of blood perfusion within ischemic muscle. Serial Laser Doppler imaging (LDI) was also performed to monitor the changes of blood flow, which was in consistent with the PET images (Figure 3B).

Both the quantification data of PET and LDI demonstrated the similar trend of perfusion recovery in ischemic hindlimb. There has been significant increase of blood perfusion in ischemic hindlimb since day 5 post surgery (0.334 ± 0.047) based on PET data (Figure 3C). However, LDI did not recognize this remarkable change until day 7 post surgery (0.476±0.034) (Figure 3D), which may be due to the relatively low sensitivity of LDI.

Following the first PET scan at 15 min post ¹⁸F-NEB injection, the same batch of surgical mice were further scanned at 2 h p.i. to monitor the inflammatory leakage of albumin from the vessel within the ischemic region of occlusive hindlimb. Figure 4A showed the representative coronal PET images of mice hindlimbs at multiple designated time points (day 1, 3, 5, 7, 10, and 14) post-surgery. According to the PET imaging results, we observed the highest radioactivity level of ¹⁸F-NEB probe within ischemic right hindlimb in the first day (8.303±2.1), indicating the increased vascular permeability during the acute inflammation phase. Then, lessened leakage of serum albumin was revealed by the decreased radioactivity within 2 weeks (day 14: 1.953 ± 0.69). The probe uptake within ischemic right limb and non-ischemic left limb was quantified based on PET images and presented as Figure 4B. Based on the quantitative data, the high uptake ratio (ischemic limb to non-ischemic) of ¹⁸F-NEB was maintained until day 7 post surgery.

Ischemic skeletomuscular of surgical mice were also collected for histology analysis. Immunohistochemical (IHC) staining with anti-CD31 (marker for endothelial cell) antibody were performed (Figure 5A). IHC staining results showed the irregular vessel lumen (highlighted by brown color), tangled endothelial cells arrangement and swelling interstitial space at the first day post-surgery, compared with the pre-surgery one (Supplementary Figure S2). Meanwhile, vast neutrophils (nucleus in blue or purple) surrounding the damaged lumen indicated the initial leakage of plasma fluid from the vessel. The irregular vessel lumen and intense neutrophil infiltration in the ischemic skeleton-muscular tissue reflected initial enhanced vascular permeability at the acute stage. From day 5 (117.6 ± 12.93), newly regenerated capillaries were prevalent and reached the peak at day 7 (159.0 ± 13.95), as displayed by CD31 staining (Figure 5B).

Given the fact that angiogenesis development in HI can lead to both blood perfusion and vascular permeability enhancement, we next used vascular endothelial growth factor (VEGF), the most widely studied angiogenic stimuli, to treat ischemic limbs in our study. Based on previous reports and our histological data as shown in Figure 5, we chose day 7 post surgery, the peak time of angiogenesis, to assess the response to VEGF treatment. As expected, a significant increase of the capillary number at 7 days post-surgery was evidenced by CD31 staining in VEGF-treated group (Figure 6A: upper row). Quantitative data showed the capillary density of 185.7±16.27 at VEGF-treated group and 162.3 ± 10.65 at PBS-treated group (Figure 6B).

Next, we evaluated whether ¹⁸F-NEB was capable of assessing the alteration of blood perfusion and vascular permeability in response to VEGF treatment in HI model. Figure 6A showed representative coronal PET images of hindlimbs at 15 min (Figure 6A: middle row) and 2 h (Figure 6A: lower row) p.i, of ¹⁸F-NEB, respectively. Based on PET quantification, the ¹⁸F-NEB uptake ratio of ischemic hindlimb to control one in the early phase (Figure 6C) in VEGF-treated group (0.6 ± 0.1) was significantly higher than PBS-treated group (0.48 ± 0.087; P < 0.05). In the late phase (Figure 6D), the ¹⁸F-NEB uptake ratio was similarly elevated by VEGF treatment (6.23 ± 0.83) as compared to PBS treatment (4.85 ± 0.66; P < 0.05). These imaging results were consistent with CD31 staining. All these data demonstrated that the resulting blood perfusion and vascular permeability enhancement of VGEF-induced angiogenesis in ischemic tissues could be non-invasively detected by ¹⁸F-NEB probe.