Spatiotemporal control of vascular endothelial growth factor delivery from injectable hydrogels enhances angiogenesis


David Mooney, Division of Engineering and Applied Sciences, 29 Oxford St., 319 Pierce Hall, Harvard University, Cambridge, MA 02139, USA.
Tel.: +1 617 495 1689; fax: +1 617 495 8534; e-mail:


Therapeutic angiogenesis with vascular endothelial growth factor (VEGF) delivery may provide a new approach for the treatment of ischemic diseases, but current strategies to deliver VEGF rely on either bolus delivery or systemic administration, resulting in limited clinical utility, because of the short half-life of VEGF in vivo and its resultant low and transient levels at sites of ischemia. We hypothesize that an injectable hydrogel system can be utilized to provide temporal control and appropriate spatial biodistribution of VEGF in ischemic hindlimbs. A sustained local delivery of relatively low amounts of bioactive VEGF (3 μg) with this system led to physiologic levels of bioactive VEGF in ischemic murine (ApoE−/−) hindlimbs for 15 days after injection of the gel, as contrasted with complete VEGF deprivation after 72 h with bolus injection. The gel delivery system resulted in significantly greater angiogenesis in these limbs as compared to bolus (266 vs. 161 blood vessels mm−2). Laser Doppler perfusion imaging showed return of tissue perfusion to normal levels by day 28 with the gel system, whereas normal levels of perfusion were never achieved with saline delivery of VEGF or in control mice. The system described in this article could represent an attractive new generation of therapeutic delivery vehicle for treatment of cardiovascular diseases, as it combines long-term in vivo therapeutic benefit (localized bioactive VEGF for 1–2 weeks) with minimally invasive delivery.


Cardiovascular diseases cause over 15 million deaths in the world each year [1]; peripheral vascular disease (PVD) affects 15% of the adult population in certain Western countries, and is frequently associated with coronary artery disease [2]. Atherosclerosis is the main cause of PVD, and results in the obstruction of the blood supply to the upper and lower limbs [3]. Despite recent advances, there still remains a major need for new approaches to successfully treat this disease. One attractive approach for the treatment of ischemic diseases is therapeutic angiogenesis. Therapeutic angiogenesis involves new blood vessel formation by delivery of specific mediators [4]. The delivery of recombinant angiogenic factors to ischemic tissues is commonly pursued [5–7], and one of the most widely investigated factors is vascular endothelial growth factor (VEGF). VEGF plays a crucial role in the initial stages of angiogenesis, and VEGF and its receptors are the best-characterized signaling pathway involved in angiogenesis [8].

The typical approach to administering VEGF or other angiogenic factors involves either bolus injections into the ischemic site, or systemic administration of recombinant protein solutions. Although therapeutic benefit was reported with bolus VEGF injection in animal studies [5,9,10], no significant improvement in phase II trials was achieved [6,11,12]. Recent analysis of the pharmacokinetics of VEGF delivery in this phase II clinical trial revealed that the bioactive VEGF levels fell below critical levels in less than 8 h with this delivery approach [13]. The transient VEGF exposure is probably responsible for the limited clinical success with this molecule to date, as VEGF needs to be present for relatively long time periods to prevent regression of the newly formed vessels [14].

Polymeric delivery systems that allow localized and sustained presentation of therapeutic agents may allow one to bypass the limitations of current VEGF delivery [15,16]. In particular, hydrogels may comprise an especially appealing class of delivery vehicle, as they can be introduced into the body with minimally invasive procedures [17], and are often highly biocompatible, owing to their high water content. Hydrogels have been tested for their ability to provide localized VEGF delivery, and alginate is an attractive material for this application [18,19]. Alginate is a naturally occurring (linear unbranched) polysaccharide composed of a α-l-guluronic and β-d-mannuronic acid sugar residues. Sodium salts of alginate are water soluble, but in the presence of divalent cations (such as Ca2+), alginate chains form ionic cross-links, leading to gelation. Owing to the gentle nature of this gelation process, Ca2+ cross-linked alginate hydrogels have been widely used in several medical applications (dental impressions, drug delivery devices, and as immobilization matrices for cells) [16,20]. Despite their biocompatibility, however, these gels undergo a slow and unpredictable degradation in vivo that limits their utility. Previous studies have reported that control over alginate gel degradation can be achieved by partially oxidizing the polymer chains with sodium periodate [21], and low levels of oxidation maintain the gel biocompatibility [22]. An alternative strategy to control degradation relies on controlling the polymer molecular weight distribution, and the use of a binary molecular weight formulation allows the incorporation of low molecular weight (LMW) chains, which more readily disassociate from the gel and can be subsequently excreted from the body [21]. In contrast to non-degradable alginate hydrogels, where diffusion is the predominant mechanism of growth factor release, factor release from degradable gels could be tuned by controlling both diffusion and degradation.

In this study, we present an injectable alginate hydrogel system that can provide sustained delivery and spatial control of the presentation of bioactive VEGF in ischemic hindlimbs. The role of alginate gel biodegradation in controlling the rate of growth factor delivery was first analyzed in vitro, and the utility of this system was subsequently evaluated in a mouse model of PVD (ApoE−/−) [23]. This system may be useful in therapeutic angiogenesis, as it sustains and localizes the delivery of relatively low amounts of VEGF and stimulates angiogenesis, returns limb perfusion to normal levels, and prevents limb necrosis.

Materials and methods

Binary molecular weight alginate gel formulation and modifications

Ultrapure alginates were purchased from ProNova Biomedical (Norway). MVG alginate, a high-G-containing alginate (M/G ratio of 40/60 as specified by the manufacturer) was used as the high molecular weight (molecular mass = 250 000 Da) component to prepare gels. LMW alginate (molecular mass = 50 000 Da) was obtained by γ-irradiating high molecular weight alginate with a cobalt-60 source for 4 h at a γ-dose of 5.0 Mrad (Phoenix Lab, University of Michigan, Ann Arbor, MI, USA), as specified by Kong et al. [24]. The alginate used to form gels was a combination of the two different molecular weight polymers at a ratio of 7.5:2.5. Both alginate polymers were diluted to 1% w/v in double-distilled H2O, and 1% of the sugar residues in the polymer chains were oxidized with sodium periodate (Aldrich, St Louis, MO, USA) by maintaining solutions in the dark for 17 h at room temperature, as previously described [21]. An equimolar amount of ethylene glycol (Fisher, Pittsburgh, PA, USA) was added to stop the reaction, and the solution was subsequently dialyzed (MWCO 1000, Spectra/Por®) over 3 days. The solution was sterile filtered, frozen (−20 °C overnight), lyophilized and stored at −20 °C. To prepare gels, modified alginates were reconstituted in EBM-2 (Cambrex Corporation, Walkersville, MD, USA) to obtain a 2% w/v solution (75% LMW, 25% MVG used in all experiments) prior to gelation. The 2% w/v alginate solutions were cross-linked with aqueous slurries of a calcium sulfate solution (0.21 g CaSO4 mL−1 distilled H2O) at a ratio of 25:1 (40 μL of CaSO4 per 1 mL of 2% w/v alginate solution) using a 1-mL syringe. Reconstituted alginate was stored at 4 °C.

Growth factor incorporation and release kinetics

Alginates were first mixed with recombinant human VEGF165 protein (Biological Resources Branch of National Cancer Institute) by using two syringes coupled by a syringe connector, and the calcium slurry (Sigma, St Louis, MO, USA) was then mixed with the resulting alginate/VEGF solution using two syringes coupled by a syringe connector to facilitate the mixing process and prevent entrapment of air bubbles during mixing. The mixture was allowed to gel for 30 min, and then was maintained at 4 °C prior to animal injections. [125I]VEGF165 was purchased from PerkinElmer Life Sciences (Wellesley, MA, USA). Alginates were mixed with the radiolabeled growth factor as described above in certain experiments to quantify VEGF release in vitro. The resulting mixture was cast between two glass plates separated with 1-mm spacers and allowed to gel for 30 min. The gels were divided into four samples, and subsequently incubated in 3 mL of phosphate-buffered saline (PBS) ([Invitrogen, Carlsbad, CA, USA] with 0.1 g L−1 of MgCl2.6H2O and 0.132 g L−1 of CaCl2.2H2O [Sigma]) at 37 °C. At each experimental time point, the radiolabeled growth factor present in the buffer solution was measured using a gamma counter (1470 WIZARD; PerkinElmer), and compared with the initial total [125I]VEGF165 incorporated into the sample.

Rheologic properties of pregelled hydrogel solutions

The viscosities of pregelled alginate solutions were investigated by using a controlled-stress rheometer (CS-50, Bohlin Instrument) at 25 °C. Preceding the measurement, all samples were presheared at a high shear rate, and then allowed to rest for 5 min. While the shear stress was increased from 0.008 to 10 Pa, the resulting strains were measured, and the corresponding viscosity (η) was calculated.

Bioactivity assay

Recombinant human VEGF protein was incorporated into alginate solutions before gelling (250 ng mL−1), and gels were cast as previously described for analysis of release kinetics. After gels had polymerized, they were transferred to 24-well plates, and 3 mL of EGM-2MV medium without growth factors (Cambrex Corporation) was added to each well. At certain time points, medium was collected and added to cultured endothelial cells, and fresh medium was introduced to the alginate gels. Human microvascular dermal endothelial cells (HMVEC) (Cambrex Corporation) (passage 6) were seeded into 24-well plates (5000 cells cm−2 cell seeding density) overnight with EGM-2MV. Endothelial cells were then washed twice with PBS, and cultured with EGM-2MV without growth factors, EGM-2MV with control VEGF, and EGM-2MV with VEGF released from alginate gels (in both of the latter cases, the VEGF concentration in the medium was 25 ng mL−1 at day 1, and 4 ng mL−1 from days 2–4). After 24 h and 4 days, the endothelial cells were detached via trypsinization, and counted in a Coulter Counter (Beckman Coulter, Fullerton, CA, USA). The biological activity of VEGF released from the gels was determined by comparison with the stimulatory effect observed in the culture wells containing control VEGF.

Sprouting assay

Cytodex 3 microcarriers (Amersham Biosciences, Piscataway, NJ, USA) were hydrated in PBS at room temperature (0.2 mL mg−1 of dry Cytodex 3), and after 3 h, the supernatant was decanted and replaced with fresh PBS; this was followed by sterilization by autoclaving. HMVECs (passage 4) in EGM-2MV were combined with 50 mg of microcarriers at a 7:1 (cell/microcarrier) ratio in a spinner vessel (Bellco Glass Inc.,Vineland, NJ, USA). After 3 h, microcarriers with cells were incubated for 20 h with continuous stirring. The microcarriers with cells were subsequently transferred to tissue culture flasks, and cultured for 1–2 days. To perform the sprouting assay, the beads in suspension (57 μL) were combined with 170.5 μL of fibrinogen (Sigma) solution (4 mg mL−1) and 22.7 μL of aprotinin (Sigma) (500 μg mL−1). This solution was then added to 200 μL of thrombin (Sigma) (22.72 U mL−1), and incubated at 37 °C for 20 min, allowing gel formation. Cultures were fed every day with 0.8 mL of EGM-2MV without growth factors, EGM-2MV with control VEGF, and EGM-2MV with VEGF released from alginate gels (in both of the latter cases, the VEGF concentration in the medium was 50 ng mL−1 at day 1, and 10 ng mL−1 from days 2 to 3). After 3 days, gels were washed twice with PBS, and incubated with 4% formaldehyde overnight at 4 °C. The formaldehyde solution was then aspirated, the gels were washed twice with PBS, and sprouts per bead were quantified from microscopic images (average of 100 beads analyzed per condition).

ERK1/ERK2 activation

Tissue culture dishes, 60 × 15 mm, of HMVECs (passage 5) were cultured in EGM2-MV to subconfluency. The medium was then removed, and cells were washed twice in PBS; subsequently, starvation medium was added, and cells were cultured for 18 h (starvation medium was composed of EBM). The starvation medium was then removed, and cells were placed in starvation medium supplemented with control VEGF or VEGF released from alginate gels (for both conditions, the VEGF concentration was 40 ng mL−1). After 5, 15 and 45 min, medium was removed from dishes, and cells were washed twice with ice-cold PBS. Levels of phosphorylated intracellular ERK1/ERK2 in cell lysates were then quantified via a two-site sandwich enzyme-linked immunosorbent assay (ELISA) (R & D Systems, Minneapolis, MN, USA).

Animals and surgical procedures

All animal work was performed in compliance with NIH and institutional guidelines. Female ApoE−/− mice aged 6 weeks, and 6–7-week-old female C57BL/6J mice (Jackson Laboratories, Bar Harbour, ME, USA), were used for these studies. ApoE−/− mice were fed a high-fat diet (21% fat, 0.15% cholesterol; Harlan Teklad, Madison, WI, USA) for at least 6 weeks prior to enrollment in the study. Mice were anesthetized with an intraperitoneal injection of a mixture of ketamine 80 mg kg−1 and xylazine 5 mg kg−1 prior to all surgical procedures. Hindlimb ischemia was induced by unilateral external iliac and femoral artery and vein ligation, as previously described [23].

After the vessel ligation, mice were injected with a total volume of 50 μL of alginate gel containing 3 μg of VEGF165, gel with no VEGF165, or PBS solution containing 3 μg of VEGF165. Injections were performed using a 25 G needle (Becton Dickinson, Franklin Lakes, NJ, USA), directly into the area where the vessels were ligated. The incisions were subsequently surgically closed, and animals monitored over time.

VEGF localization in vivo

At various time points, three distinct hindlimb muscle regions were dissected from the ischemic hindlimbs (±5 mm apart from each other), weighed, and digested with T-Per Reagent (Pierce, Rockford, IL, USA). Tissue samples were subsequently sonicated for 5 s and centrifuged at 17 949 g for 5 min (4 °C). Supernatant was collected and stored at −20 °C for future VEGF quantification. Blood samples were collected retro-orbitally, using hematocrit heparinized capillaries (Fisher Scientific), and maintained at room temperature for 30 min to allow coagulation. After coagulation, blood samples were centrifuged at 17 949 g for 10 min (4 °C). Blood serum was collected and stored at −20 °C. Total recombinant human VEGF levels present in samples were measured via quantitative human ELISA (Quantikine, R&D Systems).

Immunohistochemistry and blood vessel quantification

Hindlimb muscle tissues (n = 6 per time point per experimental condition) were retrieved, fixed, paraffin embedded, and stained for mouse CD31 (BD Biosciences Pharmingen, San Diego, CA, USA). For measurement of capillary densities, 30 randomly chosen high-power fields of the tissue were analyzed. The number of positively stained blood vessels was manually counted and normalized to the tissue area. Sections from each sample were visualized at 200× and 400× with an Olympus-IX81 light microscope (Japan) connected to an Olympus DP70 digital image capture system (Japan), and analyzed using IPLab 3.7 software (Scanalytics, Rockville, MD, USA). Vessel size was determined using iplab 3.7 software.

Laser Doppler perfusion imaging (LDPI) and limb integrity

Before surgery, and 0, 1, 3 and 7 days and 2, 4 and 6 weeks postsurgery, measurements of the ischemic/normal limb blood flow ratio were performed on anesthetized animals (n = 6 per time point per experimental condition) using Periscan system blood perfusion monitor laser Doppler equipment (Perimed, Stockholm, Sweden). Perfusion measurements were obtained from the right (ischemic) and left (non-ischemic) limb. To minimize variability because of ambient light and temperature, the index was expressed as a ratio of ischemic to non-ischemic limb blood flow.

Ischemic hindlimbs were visually observed at 1, 2, 4 and 6 weeks postoperatively in order to grade limb integrity, and categorized according to degree of necrosis: normal compared to non-surgical limb, one necrotic toe, multiple necrotic toes, and necrotic foot.

Statistical analysis

All statistical comparisons were performed using Student's t-test (two-tail comparisons), and analyzed using instat 2.01 (GraphPad Software, Inc., San Diego, CA, USA) software. Differences between conditions were considered significant if P < 0.05.


In vitro validation of injectable polymeric system

Alginate hydrogels with well-controlled degradation rates [22] can be generated by combining control over molecular weight distribution and partial oxidation to make the polymer chains susceptible to hydrolysis, and the ability of these gels to serve as sustained VEGF release vehicles was first assessed in vitro. The degradation of hydrogels formed from a combination of high and LMW polymers that were partially oxidized (1%), and control gels formed from both high molecular weight alginate and a mixture of high and LMW polymer that were not oxidized, was evaluated by dry weight loss as a function of degradation time (Fig. 1A). As expected, the degradation rate of the gels formed from partially oxidized alginate was relatively rapid, with significant mass loss within 7 days, and almost complete mass loss by 40 days. In contrast, neither of the control gels degraded significantly over this time frame. The capabilities of alginate hydrogels to be used as an injectable delivery system were evaluated by characterizing the rheologic properties of pregelled solutions (Fig. 1B). The combination of low and high molecular weight alginates exhibited an intermediate low-shear viscosity (η), as compared to the high or LMW polymers alone.

Figure 1.

 Biophysical properties and vascular endothelial growth factor (VEGF) release from alginate gels. The dry masses of gels formed from high molecular weight non-oxidized alginate (Δ), binary molecular weight non-oxidized alginate (bsl00000) and binary molecular weight partially oxidized alginate (bsl00001) were analyzed as a function of time following incubation in phosphate-buffered saline (A). The viscosities of solutions of high (bsl00000), low (bsl00001) and binary (inline image) molecular weight partially oxidized alginate solutions were assessed (B). The release kinetics of [125I]VEGF165 from gels formed from binary molecular weight alginate, either partially oxidized (bsl00001) or non-oxidized (bsl00000), was monitored over time (C). Greater endothelial cell proliferation was observed when cells were cultured with VEGF released from binary molecular weight partially oxidized alginate gels (inline image), as compared to medium without VEGF (bsl00001) and control VEGF added to medium (bsl00000), as determined from the cell number counts from day 0 to day 4 (D). VEGF delivered from gels induced a greater number of endothelial cell sprouts, as compared to control VEGF, or no VEGF added to culture medium. Control VEGF (bsl00000) and VEGF released from gels (inline image) activated ERK1/2 signaling with a comparable profile over time, as opposed to control medium containing no VEGF (bsl00001) (F). In (A), (B) and (C), values represent mean and SD (n = 5) (A, B, C) n = 5. (D) n = 6. (E) n = 4.

To investigate the relationship between gel degradation rate and VEGF165 (referred to as VEGF) release, 125I-labelled VEGF was incorporated into the gels, and its release following incubation in medium was monitored over time (Fig. 1C). After an initial burst, VEGF was released in a sustained manner over time, and within 7 days, approximately 60% of the total VEGF loaded was released. In contrast, VEGF delivered from non-oxidized alginates showed slower release, and after 30 days, only approximately 40% was released. VEGF is susceptible to rapid degradation in vitro and in vivo [8], and VEGF incorporated and released from the hydrogels was examined to determine whether it maintained its bioactivity. This was first monitored by quantifying its ability to stimulate endothelial cell proliferation in vitro. The VEGF released for the first 4 days from the gels was not only bioactive, but was even more potent than the same VEGF concentration added directly to the culture medium (Fig. 1D). Although striking, these data are consistent with previous studies of growth factor release from alginate gels [19]. We also investigated whether the higher bioactivity of VEGF released from the gels correlated with an increased ability to promote endothelial cell sprouting in vitro. After 3 days, VEGF delivered from alginate gels led to an approximately 2.5-fold increase in the number of sprouts per bead, as compared to the VEGF control (Fig. 1E). Finally, we considered the ability of VEGF released from gels to activate the ERK1/ERK2 intracellular pathway, in terms of duration of activation. The VEGF released from alginate gels was able to activate and maintain ERK1/ERK2 intracellular signal in a similar manner to control VEGF (Fig. 1F).

In vivo localization of VEGF delivered with gels

We next examined the potential of the degradable hydrogels to serve as an injectable system to deliver VEGF in a sustained and localized manner in vivo. C57BL6 mice were submitted to ischemic hindlimb surgery, and immediately injected with either VEGF in PBS (bolus delivery) or hydrogels containing VEGF. The total VEGF dose and volume of injection were the same for both conditions. At different time points, the levels of VEGF in the blood and in discrete tissue regions of the hindlimb relative to the injection site (Fig. 2A) were assessed via ELISA. During the first 24 h after injection, higher concentrations of VEGF in the injection region were found when VEGF was delivered from the alginate, as compared with bolus (Fig. 2B). At 12 h after injection, 95% of the total VEGF dose in the gels was still localized in the region (A) closest to the injection site, whereas VEGF delivered via bolus injection was much more widely dispersed (43% of VEGF located in regions B and C) (Fig. 2C). Bolus VEGF injection led to small amounts of bioactive VEGF at the injection site at 24 h, and complete VEGF deprivation was observed 72 h after injection (Fig. 2D). In contrast, physiologic levels of VEGF were still present in the ischemic hindlimbs 15 days after delivery with the alginate (Fig. 2E). Bolus delivery did lead to very high VEGF concentrations in peripheral serum within hours after injection, whereas the VEGF levels in serum were low in the first 12 h when gels were used for VEGF delivery, and physiologically insignificant or zero levels were observed after 12 h (Fig. 2F).

Figure 2.

 Photograph of ischemic hindlimb tissue, showing regions of tissue used for analysis (A); A, B and C represent regions of the mice hindlimb at increasing distances from the injection site. Higher concentrations of bioactive vascular endothelial growth factor (VEGF) were present in the injection region (region A) when VEGF was released from binary alginate gels (bsl00045), as compared to bolus delivery (bsl00000) (B). The relative VEGF distribution in distinct hindlimb regions, 12 h postinjection for bolus delivery and delivery from alginate hydrogels (C). VEGF levels in ischemic tissue resulting from bolus injection (D) were initially low, and rapidly fell to insignificant levels, as contrasted to tissue concentrations of bioactive VEGF when delivered using gels (E). In both (D) and (E), tissues were analyzed after 12 h (inline image), 1 day (bsl00066), 3 days (○), 7 days (•), and 15 days (bsl00001). Low levels of VEGF were observed in peripheral serum when delivered from gels (bsl00045), as compared to bolus delivery (□) (F). Mean values are presented with SD (n = 5) in all graphs.

Hydrogel VEGF delivery promotes angiogenesis, alleviates ischemia, and prevents necrosis

The effects of VEGF delivery from hydrogels on vascularization and perfusion of ischemic tissues was next analyzed in ApoE−/− mice that were subjected to femoral artery and vein ligation, a standard model of peripheral ischemia that mimics some aspects of human atherosclerosis [23]. Retrieval of muscle tissue 42 days postoperatively, and immunohistochemical analysis, revealed that VEGF-delivering gels increased blood vessel densities, as compared with bolus delivery of VEGF or injection of a gel with no VEGF (Fig. 3A). Quantification of these data revealed that bolus delivery of VEGF had no significant effect on vascularization, as expected, and delivery of blank gels resulted in a small increase in blood vessel densities (Fig. 3B). VEGF delivery from the gels resulted in an approximately 2-fold increase in vessel density, as compared to the condition in which there was no intervention.

Figure 3.

 Photomicrographs of tissue sections from ischemic hindlimbs of apoE−/− mice at postoperative day 42, immunostained for the endothelial marker CD-31. (A) Quantification of blood vessel density after 6 weeks with no injection of gel or vascular endothelial growth factor (VEGF) (−−), bolus delivery of VEGF (− +), injection of a gel with no VEGF (+ −) or delivery of VEGF from a gel (+ +) in ApoE −/− mice (B) revealed a statistically significant enhancement with VEGF-delivering alginate gels. *P < 0.02.

The effects of VEGF delivery from these gels on perfusion and amelioration of limb ischemia were also analyzed. LDPI was used to quantify perfusion in the entire hindlimbs, and tissue necrosis was quantified by visual observation. Regional blood flow was abruptly reduced after surgery in all conditions, as expected (Fig. 4A). No treatment led to little increase in perfusion over time, and the ischemic limbs became necrotic. Bolus VEGF delivery resulted in little difference from the no-treatment control. In contrast, sustained and localized VEGF delivery from the gels led to an increase in tissue perfusion over time, and largely spared the limbs from necrosis. Quantification confirmed the beneficial effect of gel-based VEGF delivery on return of perfusion, as compared to no treatment, bolus VEGF delivery, or delivery of gel without VEGF (Fig. 4B). After a small increase in perfusion of the ischemic hindlimb, the level of perfusion was almost constant over time for bolus VEGF delivery as well for the other control conditions. In contrast, animals treated with alginate gels delivering VEGF showed a gradual and marked increase in blood flow over time (45% by day 3, and 83% of the normal level by day 42). The surgically induced hindlimb ischemia led to severe toe or foot gangrene, and animals were observed and categorized by the level of limb integrity. Necrosis of toes or foot loss was prevented by gel-based VEGF delivery in four of the animals in that cohort by week 6 (Fig. 4C). In contrast, bolus VEGF delivery was incapable of preventing toe necrosis and foot loss, and the other control conditions led to similar results (data not shown).

Figure 4.

 Representative color-coded laser Doppler perfusion imaging (LDPI) images at various time points, and photographs of ApoE−/− mice at postoperative day 42 (A). No treatment [no gel or bolus vascular endothelial growth factor (VEGF) injection] (left), bolus VEGF injection (center) and alginate hydrogel (VEGF-releasing) (right) conditions were among those analyzed. Hindlimb normal baseline (before) perfusion is illustrated in green/yellow/red. Immediately after unilateral femoral artery ligation (after), animals were scanned at different time points. LDPI blood perfusion profile of hindlimbs with no injection of gel or VEGF (bsl00000), injection of blank control gels (Δ), bolus VEGF delivery (inline image), and VEGF delivered from alginate hydrogels (•) in ApoE −/− mice (B). Hindlimbs subjected to surgery were visually examined, and grouped as normal (displaying no discrepancy in color or limb integrity from non-ischemic hindlimbs of the same animal), or presenting one necrotic toe, multiple necrotic toes, or a complete necrotic foot (C). *Statistically significant difference (P < 0.02) as compared to controls (blank alginate, bolus, and no gel or VEGF); mean values are presented with SD (n = 6) in (B).


The results of this study indicate that VEGF delivered from injectable alginate hydrogels promotes angiogenesis, alleviates ischemia, and prevents necrosis. Alginate gels can be manipulated and designed to exhibit favorable degradation kinetics and rheologic properties, allowing their use as injectable delivery vehicles capable of presenting proteins in a spatially and temporally regulated order. This injectable system provided a sustained and localized release of VEGF in ischemic hindlimbs, resulting in high retention of the VEGF at the site of interest at physiologically relevant concentrations. The VEGF delivered from this injectable system was biologically active and available for long time periods, and led to significant increases in blood vessel formation and subsequent increases in tissue perfusion.

Alginate gels can be assigned specific degradation rates by combining partial oxidation and a controlled molecular weight distribution, resulting in an injectable delivery vehicle capable of sustained release of bioactive factors (Fig. 1). The results of this study indicate that protein release was controlled by the alginate gel degradation rate, as faster-degrading gels led to more rapid release kinetics (Fig. 1A,C). Alginate gels have a highly porous (5–200 nm) nanostructure, favorable for high diffusion rates for macromolecules. However, diffusion is also regulated by protein–polymer interactions, and alginate has been reported to reversibly bind heparin-binding growth factors such as VEGF [19,25], which slows the release and increases its dependency on gel degradation. The control over gel degradation in this study was consistent with previous reports, in which alginate gel degradation was demonstrated to be controllable by combining partial oxidation of the polymer chains and the utilization of a binary molecular weight distribution in gel formation [21,22,26]. Furthermore, this injectable delivery vehicle was capable of not only maintaining, but also increasing, the biological activity of VEGF incorporated into gels (Fig. 1D). This result was consistent with previous studies [19], suggesting that VEGF bioactivity can be enhanced through interactions with alginate, perhaps because of shielding of the VEGF from environmental conditions and protection from premature denaturation. This increase of the biological activity of VEGF included in gels was correlated with increased endothelial cell sprouting in vitro, probably resulting from activation of ERK1/2 (Fig. 1E,F).

The data presented in this study indicate that sustainable and localized release of VEGF from alginate gels in ischemic hindlimbs maintains higher VEGF levels at the hypoxic regions than does bolus injection (Fig. 2). The specific biophysical properties of these alginate gels led to VEGF delivery in vivo in a regulated, temporal manner (Fig. 2E). The control over the factor spatial distribution also resulted in marginal levels of VEGF in the peripheral serum (Fig. 2F). Many lines of evidence indicate that VEGF is an attractive molecule to revascularize ischemic regions [8,27,28]; however, VEGF delivery failed in the only large clinical trial performed to date [29]. Owing to the short VEGF half-life in vivo, massive doses of VEGF were typically delivered into the circulation [5] to stimulate a therapeutic effect in past trials, probably resulting in an absence of spatial resolution (ineffective microenvironmental localization) and a lack of appropriate temporal dynamics of VEGF presentation (local environment exposed to marginal VEGF doses). Because the timing of availability of angiogenic factors is believed to be crucial to obtain a physiologically adequate angiogenic response in vivo [28], the use of polymeric constructs to deliver these types of therapeutic agent is very appealing [15,19]. The specific hydrogel system developed in the current studies is capable of maintaining bioactive VEGF in ischemic tissue over significant time periods (7–15 days) with little biodistribution outside of the ischemic regions, and this contrasts with the results of bolus delivery.

The results of this study show that therapeutic angiogenesis could probably benefit from sustained and localized VEGF delivery from the injectable and degradable alginate gels (Figs 3 and 4). The sustained VEGF tissue exposure made possible with this injectable system led to significant increases in blood vessel formation in the ischemic muscle tissue (Fig. 3A,B). Furthermore, the controlled and localized delivery of VEGF led to significant increases in tissue perfusion (Fig. 4A,B), alleviating ischemia and preventing necrosis associated with ischemia (Fig. 4A,C). Altogether, these data indicate that VEGF delivered from alginate gels regulates the formation of functional new blood vessels in ischemic hindlimbs. Previous studies have indicated that lipid disorders can directly impair or retard angiogenesis [23], and our results indicate that the low endogenous angiogenic potential of ApoE−/− mice can be bypassed with sustainable VEGF delivery. However, although it was possible to rescue and reverse limb necrosis, fully normal tissue perfusion was not achieved. These findings suggest that VEGF delivery alone may not be sufficient, and future studies may involve combining VEGF delivery with other angiogenic factors to augment and mature the angiogenic response, as previously demonstrated with polymer systems that require surgical implantation [15].

In summary, an injectable biodegradable alginate hydrogel allowing sustained and localized release of VEGF has been developed. This system is particularly appealing for therapeutic use, as it allows for minimally invasive factor deliver. We have demonstrated that this injectable gel system provides VEGF at a desirable concentration for extended time periods. The spatiotemporal factor bioavailability provided by this system led to a significant angiogenic response in ischemic hindlimbs, and this system may also be broadly useful for the manipulation or exploitation of the presentation of a wide variety of other growth factors, and examination of their roles in this and other biological processes.

Disclosure of Conflict of Interests

The authors thank the National Institutes of Health for financial support of this research (RO1 HL069957), and the Biological Resources Branch of National Cancer Institute (NCI) for generously providing the VEGF used in our studies. E. A. Silva is a student of the Gulbenkian PhD Program in Biomedicine, Portugal, and is supported by a predoctoral fellowship from Fundacao para Ciencia e Tecnologia, FCT (SFRH/BD/9613/2002), Portugal.