Vascular endothelial growth factor and fibroblast growth factor 2 delivery from spinal cord bridges to enhance angiogenesis following injury

Vascular endothelial growth factor and fibroblast growth factor 2delivery from spinal cord bridges to enhance angiogenesisfollowing injury

Laura De Laporte,1 Anne des Rieux,2 Hannah M. Tuinstra,1 Marina L. Zelivyanskaya,1

Nora M. De Clerck,3,4 Andrei A. Postnov,3,4 Veronique Preat,2 Lonnie D. Shea1

1Department of Chemical and Biological Engineering, Northwestern University, 2145 Sheridan Road/E156 Evanston,

Illinois 60208-31202Universite Catholique de Louvain, LDRI, Unite de Pharmacie Galenique, UCL, 1200 Brussels, Belgium3Department of Biomedical Sciences, University of Antwerp, Belgium4Department of Physics, University of Antwerp, Belgium

Received 16 September 2010; revised 18 December 2010; accepted 13 January 2011

Published online 31 May 2011 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.a.33112

Abstract: The host response to spinal cord injury can lead to

an ischemic environment that can induce cell death and

limits cell transplantation approaches to promote spinal cord

regeneration. Spinal cord bridges that provide a localized and

sustained release of vascular endothelial growth factor (VEGF)

and fibroblast growth factor 2 (FGF-2) were investigated for

their ability to promote angiogenesis and nerve growth

within the injury. Bridges were fabricated by fusion of poly-

(lactide-co-glycolide) microspheres using a gas foaming/par-

ticulate leaching technique, and proteins were incorporated

by encapsulation into the microspheres and/or mixing with

the microspheres before foaming. Compared to the mixing

method, encapsulation reduced the losses during leaching

and had a slower protein release, while VEGF was released

more rapidly than FGF-2. In vivo implantation of bridges

loaded with VEGF enhanced the levels of VEGF within

the injury at 1 week, and bridges releasing VEGF and

FGF-2 increased the infiltration of endothelial cells and

the formation of blood vessel at 6 weeks postimplanta-

tion. Additionally, substantial neurofilament staining was

observed within the bridge; however, no significant differ-

ence was observed between bridges with or without protein.

Bridges releasing angiogenic factors may provide an

approach to overcome an ischemic environment that

limits regeneration and cell transplantation-based ap-

proaches. VC 2011 Wiley Periodicals, Inc. J Biomed Mater Res Part A:

98A: 372–382, 2011.

Key Words: spinal cord injury, angiogenesis, spinal cord

bridges, PLG, protein delivery

How to cite this article: Laporte LD, des Rieux A, Tuinstra HM, Zelivyanskaya ML, De Clerck NM, Postnov AA, Preat V, Shea LD.2011. Vascular endothelial growth factor and fibroblast growth factor 2 delivery from spinal cord bridges to enhance angiogenesisfollowing injury. J Biomed Mater Res Part A 2011:98A:372–382.

INTRODUCTION

Drug-delivery strategies are increasingly being investigatedto address the limited regeneration that occurs followingspinal cord injury. Spinal cord injury is characterized by anacute phase, which includes hemorrhage, the destruction ofthe blood brain barrier, and infiltration of inflammatorycells, followed by a subacute and chronic phase, which con-sists of secondary injury distinguished by the Wallerian axo-nal degeneration, and the formation of a cavity and glialscar surrounding this cavity.1 The lack of neurotrophic fac-tors, ischemia resulting from insufficient blood flow, and thepresence of inhibiting factors, all contribute to an unfavora-ble microenvironment for spinal cord regeneration. Severalneurotrophic factors have been investigated to promoteneuronal survival and axonal regrowth through the injury

site.2–4 The delivery of these factors has typically beenachieved using direct injection, osmotic pumps, or trans-plantation of genetically engineered cells. More recently,bridges have been implanted into spinal cord injuries to sta-bilize the injury site, promote cell infiltration, and limit cystor cavity formation. These bridges have the potential to beused as vehicles for the sustained release of factors to tar-get one or more processes that limit regeneration.5–11

Angiogenic factors, such as vascular endothelial growthfactor (VEGF) and fibroblast growth factor 2 (FGF-2), havebeen proposed to address ischemia following injury andwere implicated in neuroprotection1,12 and/or nerve regen-eration2,4,13,14 after spinal cord injury. The ischemic environ-ment after spinal cord injury leads to limited neuron sur-vival and has complicated the transplantation of cells

Correspondence to: L. D. Shea; e-mail: [email protected]

Contract grant sponsor: NIH; contract grant numbers: RO1 EB005678, R21 EB006520, RO1 EB 003806

Contract grant sponsors: Belgium American Education Foundation, IBBT TIRPA project, IWT SBO Quantiviam project

372 VC 2011 WILEY PERIODICALS, INC.

designed to promote a permissive environment at theimplant site.15,16 The delivery of exogeneous VEGF can sparespinal cord tissue, reduce retrograde degeneration, increaseblood vessel density, and reduce the number of apoptoticcells.12 Transplantation of neural stem cells after spinalcord contusion that were genetically modified to produceVEGF enhanced stem cell differentiation into mature oligo-dendrocytes and improved endogenous gliogenesis, angio-genesis, and tissue sparing.17 However, a significant chal-lenge is to control the release of these factors, as there isthe potential for increased microvascular permeability, asso-ciated with leukocyte infiltration, and an increased lesionvolume. In addition to enhancing vessel ingrowth, angio-genic factors may be able to promote axonal ingrowth. Boththe vascular and neural networks are structurally similarcomplex branched systems that are wired by multiple regu-latory factors.18,19 Several axon guidance cues are involvedin blood vessel guidance,19,20 while VEGF and FGF-2 havebeen suggested to play a role in nerve regeneration.2,4,21

VEGF can pattern small arteries along peripheral nerves,22

stimulate neurogenesis of neural stem cells in the adult cen-tral nervous system,23 have a mitogenic effect on neural celltypes, such as astrocytes24 and Schwann cells,25,26 andinduce limited regeneration of the corticospinal tract acrossthe injury site.13 Although the delivery of angiogenic factorshas many potential advantages after spinal cord injury, theircontrolled release from bridges has, to the best of ourknowledge, not yet been investigated.

In this report, we implanted poly(lactide-co-glycolide)(PLG) bridges with the ability to locally deliver VEGF andFGF-2 to a spinal cord injury and analyzed blood vessel for-mation, endothelial cell infiltration, and nerve growth.Bridges were fabricated using a gas foaming/particulateleaching process and have previously demonstrated good tis-sue apposition without cavity formation, cell ingrowth intothe pores of the bridge, and the ingrowth of neurofilament(NF)-positive fibers,6,27 which contrasts with the insertion ofgelfoam that resulted in cavity formation. VEGF and FGF-2were loaded into the bridge through either encapsulation intothe microspheres and/or direct mixing of microspheres withprotein before gas foaming. The encapsulation efficiency,release, and bioactivity of the released protein were charac-terized in vitro, while the presence of exogeneous VEGF at theinjury site in vivo was analyzed 1-week postimplantation in arat spinal cord hemisection model. Noninvasive high resolu-tion X-ray microcomputed tomography (micro-CT) was usedto visualize the 3D position and formation of functional bloodvessels for multiple protein doses in vivo.28 Subsequent stud-ies investigated two VEGF/FGF-2 combinations for the pro-motion of angiogenesis and neurotrophic effects, such as axo-nal extension into the lesion. Angiogenic factor delivery maybe used to combat the ischemic environment that can inducecell death and limits cell transplantation approaches to pro-mote spinal cord regeneration.

MATERIALS AND METHODS

Fabrication of protein loaded multiple channel bridgesProtein-loaded multiple channel bridges were made of PLGand fabricated using a combination of a cryogenic double

emulsion technique and a gas foaming/particulate leachingtechnique, which have been previously described.29,30 Thetechnique in these reports was adapted to the fabricationtechniques for multiple channel bridges, which have a morecomplex architecture based on its application to spinal cordregeneration.5,6,8,27 A mixture of protein-loaded micro-spheres (2.6 mg), lyophilized proteins, and salt as a porogen(10 mg, 63–106 lm) was mixed with 0.5 lL water usingthe wet granulation method31 and deposited in a custommade mold layer-by-layer before gas foaming. This resultedin a porous 3D bridge after 1 h leaching in water, which issufficient to leach out all the salt, based on the dry weightof the bridges.6 The bridges contained seven channels andwere �90% porous based on their volume, weight, and thedensity of PLG.6 Their dimensions were 4 mm in length, 2.6mm in width, and 1.5 mm in height, which matched thedimensions of the lateral hemisection created in the rat spi-nal cord.6,27 Images of the bridges can be found in these pre-vious reports, and a schematic is provided in Figure 4(A).

Proteins were encapsulated into the microspheres byemulsifying a protein solution (17 lL), containing VEGF(recombinant mouse VEGF-165, Prospec, Rehovot, Israel) orFGF-2 (recombinant human FGF-2, Chemicon, Billerica, MA),700 lg bovine serum albumin (BSA), 50 mg/mL sucrose,and MgCO3 (3% wt of BSA), with 500 lL 3 wt % PLG indichloromethane using sonication at 40 W for 15 s on ice.This first emulsion was sequentially frozen in liquid nitro-gen for 4 s in order to freeze the aqueous protein solutionpart without freezing the polymer solution. In a secondemulsion, the first protein/polymer emulsion was added to25 mL of 5% poly(vinyl alcohol) (PVA, 88% hydrolyzed, av-erage MW 22,000, Acros Organics, Morris Plains, NJ) con-taining 50 mg/mL sucrose, and homogenized at 7000 rpmfor 45 s. The resulting solution was diluted in 15 mL 1%PVA with 50 mg/mL sucrose and stirred at room tempera-ture for 3 h. Microspheres were collected by centrifugation,washed three times, and lyophilized overnight.

For the mixing method, protein was lyophilized with 15lL 1% alginate and mixed with the microspheres (5 wt %alginate to PLG32) and salt before loading into the mold forbridge fabrication. The initial amount of protein added tothe protein solutions was based on the encapsulation effi-ciency and losses during leaching measured by radioactiveI125-labeled VEGF and FGF-2.

Protein encapsulation efficiency and release frombridgeThe release kinetics for VEGF and FGF-2 from multiplechannel bridges was determined using radiolabeled protein.A total of �0.1 lCu I125-labeled VEGF and FGF-2 (PerkinElmer, Waltham, MA) and 0.2 lg nonradiolabeled proteinwas either mixed with and/or encapsulated within themicrospheres per bridge to quantify the encapsulation effi-ciency and losses during leaching. The growth factor wasmixed with an excess of BSA (100 lg) for encapsulation;thus, small changes in growth factor loading were notexpected to substantially alter the bioactivity or release pro-file.33 To measure the encapsulation efficiency, a known

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | 1 SEP 2011 VOL 98A, ISSUE 3 373

amount of microspheres was dissolved in 5M NaOH to mea-sure its radioactivity using a Gamma counter (Micromedic4/600 Plus, Micromedic, Horsham, PA). The efficiency wascalculated as the ratio of protein loading per milligrammicrospheres and the maximum protein loading, which wascalculated as the ratio of the initial amount of protein addedto the PLG solution and the PLG weight in this solution. Thelosses during leaching were calculated by measuring theleaching fluid. After leaching, the bridges were dried, incu-bated in 1 mL phosphate-buffered saline (PBS) at 37�C, andtransferred to fresh PBS at specific time points. The activityof the PBS that contained the bridge for that specific timeperiod was measured to quantify the amount of protein thatwas released. The cumulative release of VEGF and FGF-2from the bridges was calculated by dividing the cumulativeradioactivity released into the buffer by the total amount ofradioactivity released over the course of the study plus theradioactivity that remained inside the bridge at the end ofthe study. The latter was analyzed by dissolving the bridgesin 5M NaOH to measure its activity.

Bioactivity of protein released from bridgeThe activity of released VEGF and FGF-2 was measured bytheir ability to phosphorylate the VEGF receptor 2 (R2) andFGF-2 receptor, respectively, of human umbilical vein endo-thelial cells (prescreened HUVECs, ECACC, # S200-05n).Bridges containing 1 lg of VEGF or FGF-2 (encapsulated ormixed, n ¼ 3) were incubated in 500 lL of PBS at 37�C. Atday 1, 5, 10, 20, and 42, release solutions were collectedand replaced by fresh PBS. Because of the low proteinamount in release solutions, the release buffer of the tripli-cates was combined before incubation with HUVECs.HUVECs were first incubated 4 h in serum-free medium andthen incubated with the release buffer for 5 min. Cells werewashed with cold HBSS and lysed for 15 min in 50 lL ofice-cold RIPA buffer [50 mM Tris, pH 7.4, 150 mM NaCl, 1%Triton, 0.05% sodium deoxycholate, 1 mM EDTA, 0.1% SDS,and protease inhibitors (Protease Inhibitor Cocktail, Sigma)].Cell debris was removed by centrifugation at 14,000 � g for10 min at 4�C, and protein concentrations were determinedwith the BCA protein assay (Pierce, Rockford, IL). Cellextracts (70 lL) were then resolved by 6% SDS–PAGE andtransferred to PVDF membranes (Millipore, Billerica, MA).VEGF and FGF-2 that were not encapsulated within thebridge were used as positive controls.

Membranes were blocked for 1 h at room temperaturein 5% nonfat dry milk powder in 1M Tris (pH 7.4), 5MNaCl, and 0.05% Tween 20 (TBST) and incubated overnightat 4�C with the primary antibody rabbit antiphospho-VEGFR2 (TYR1175) or rabbit antiphospho-FGF R (TYR653)(Cell Signaling Technology, Beverly, MA) diluted 1/1000 in1% nonfat dry milk powder in TBST. After washing in TBST(3 � 15 min), membranes were incubated for 1 h at roomtemperature in peroxidase-conjugated anti-rabbit secondaryantibody (0.1 lg/mL, Jackson ImmunoResearch Laborato-ries, Suffolk, UK). After washing in TBST, enhanced chemilu-minescence was done according to the manufacturer’s rec-

ommendation (ImmobilonTM; Western, Millipore). Proteinexpression was analyzed via Geliance 600 (Perkin Elmer).

Rat spinal cord hemisection modelTo analyze the performance of the bridges in vivo, bridgeswere implanted into a rat spinal cord hemisectionmodel.5,6,27 Thirty-four female Long-Evans rats (CharlesRiver, 180–200 g) were treated according to IACUC guide-lines at Northwestern University and prehandled for 2weeks presurgery. Rats were anaesthetized using a RC2Rodent Anesthesia System (Colonial Medical Supply, Franco-nia, NH) with vaporized isoflurane (Baxter, Deerfield, IL) toperform a laminectomy at T9-10 and expose the spinalcord. A lateral hemisection of 4 mm long up to the midlinewas created for bridge implantation, after which the injurysite was covered with gelfoam (Henry Schein, Melville, NY),the muscles were sutured together, and the skin stapled.Postoperative care included administering Baytril (enroflox-acin 2.5 mg/kg s.c., once a day for 2 weeks), buprenorphine(0.01 mg/kg s.c., twice a day for 2 days), and lactate ringersolution (5 mL/100 g, once a day for 5 days), and bladderexpression twice a day until bladder function recovered.

ELISA to quantify VEGF at the injury siteAn enzyme-linked immunosorbent assay (ELISA) was per-formed to analyze the presence of delivered VEGF at theinjury site and its adjacent spinal cord segments 1 weekpost injury. The bridges (n ¼ 4) were fabricated with 0.5 lgrat VEGF (kindly provided by Prof. Carmeliet, KatholiekeUniversiteit Leuven, Belgium) encapsulated inside themicrospheres plus 0.5 lg rat VEGF mixed with the micro-spheres. This loading of VEGF was sufficient to determine ifVEGF delivery was increasing levels above background. Con-trol implants (n ¼ 4) were done with bridges without VEGF.Upon the removal of the spinal cord, the gelfoam wasremoved from the cord, and the cord was cut into three seg-ments: the injury site and two 0.5 cm long segments rostraland caudal of the injury site [Fig. 4(A)]. All segments werestored on dry ice until being thawed and cut into smallpieces. Lysis buffer (100 lL) (Cell Culture Lysis Reagent 1�,Promega, Madison, WI) was added, and the lysate was vor-texed for 15 s, rotated for 30 min using a rotamix (Appro-priate Technical Resources, Laurel MD), and centrifuged at14,000 RPM for 10 min at 4�C to collect the supernatant.The supernatant was diluted (1/200) to perform the VEGFELISA, according to supplier instructions (Rat VEGF ELISAkit for cell and tissue lysate, RayBioVR , RayBiotech, NorcrossGA). Note that this processing of the spinal cord does notsolubilize the polymer; thus, the procedure measures theprotein within the tissue and not within the polymer.

Microcomputed tomography to analyze functionalblood vesselsMicrocomputed tomography (MicroCT) was performed asan initial qualitative screening for five different VEGF/FGF-2doses ranging from 1 to 2 lg of one or both proteins (n ¼3) to visualize the presence of functional blood vessels inthe bridge after spinal cord injury and determine the doses

374 DE LAPORTE ET AL. VEGF AND FGF-2 DELIVERY FROM SPINAL CORD BRIDGES TO ENHANCE ANGIOGENESIS

for a more quantitative analysis of cell infiltration.34–36 Toanalyze this functional response of protein delivery in vivo,greater protein amounts were used compared to the releasestudies. Fourteen rats (one rat died) were euthanized at 6weeks postimplantation using a transcardiac perfusion toinject Microfil compounds (Flow Tech, Carver, MA) that cureto form a three-dimensional cast of the animal’s vasculature,which can then be imaged by micro-CT. Animals weredeeply anesthetized using Euthasol to open the abdominalcavity, cut the diaphragm, and expose the heart. A perfusionneedle was inserted into the left ventricle oriented to theascending aorta, while the right atrium was opened usingsurgical microscissors. Animals were perfused at 20 mL perminute with PBS (50 mL), followed by 4% phosphate-buf-fered paraformaldehyde (300–400 mL) to fix the tissue, and50 mL of PBS to rinse out paraformaldehyde.

The Microfil injection compound (MV-122) was preparedby adding 500 lL of curing agent per 10 mL MV-122 andthorough mixing. It was sequentially manually injected at aslow rate into the animal’s vasculature, using a 10-mL sy-ringe and a 21-gauge needle. This procedure fills blood ves-sels that connect to the animal vasculature and are referredto as functional vessels. After curing, the spinal cords wereremoved for image analysis with a high-resolution in vitromicro-CT system, SkyScan 1072 (SkyScan, Kontich, Belgium).The scanner was equipped with a point X-ray source (8-lmfocal spot) operating at 80 kV and 100 lA, and the studywas performed with 10-lm pixel size without filter. Theseconditions allow for imaging the smallest vessels with opti-mal contrast provided by the soft part of X-ray spectra.From the virtual cross-sections, 3D models were createdusing ANT software (Skyscan, Kontich, Belgium) in order tovisualize the position of the blood vessels.

ImmunohistochemistryThe effect of VEGF and FGF-2 delivery from the bridges onthe presence of angiogenic markers, such as endothelial cells,and neurotrophic effects, such as nerve extension into thelesion, was sequentially investigated using immunohisto-chemistry for three different conditions: bridges containing4 lg of VEGF encapsulated within their microspheresand 2 lg of FGF-2 and VEGF mixed (high protein dose, n ¼ 4),4 lg of VEGF encapsulated within their microspheres and1 lg of FGF-2 and VEGF mixed (medium protein dose, n ¼ 3),and bridges without protein loading (no protein, n ¼ 4).

Rats (n ¼ 11) were sacrificed at 6 weeks postimplanta-tion to retrieve the spinal cords and prepared for immuno-histochemistry as described previously.27 The tissue sectioncontaining the injury site (T8–T11) was frozen in isopentane(Fisher Scientific, Pittsburgh, PA), stored in siliconizedeppendorf tubes at �80�C, cryopreserved in optimum cut-ting temperature (O.C.T.) compound, and sliced longitudi-nally in 10-lm thick sections using a cryostat (Micron,Microm HM 505 N). Every other section was collected onpoly(L-lysine)-coated glass slides (Fisher Scientific), post-fixed, and stained.

Each 10th collected section was stained for a specificcell stain with primary IgG1 antibodies against endothelial

cells (mouse-anti-rat RECA-1, AbD Serotec, Raleigh, NC,1:150 dilution) and NF (rabbit-anti-NF 200; Sigma-Aldrich;1:5000 dilution) in combination with a secondary immuno-peroxidase stain [biotinylated anti-mouse IgG (Vector Labo-ratories, Burlingame, CA, 1:200 dilution) and anti-rabbit IgG(Jackson Lab, West Grove, PA, 1:200 dilution), respectively].Negative controls were performed by eliminating the pri-mary antibodies. Images were taken of the entire bridgesection in phase (20�) and six sections per rat were ana-lyzed for each stain in Photoshop. The results were reportedas a percentage of the surface area of the entire bridge thatwas stained positive and normalized to the results obtainedfor bridges without protein. The surface areas of the posi-tive stain and the entire bridge were measured by selectingthese areas and measuring their total amount of pixelsusing the histogram tool in Photoshop.

Statistical analysisStatistical analyses were done using statistical package JMP(SAS, Cary, NC). For multiple-pair comparison, an ANOVAwith post hoc Tukey test was performed with a p-level of0.05. A t-test was performed to analyze differences betweenindividual pairs. Error bars represent standard deviations inall figures.

RESULTS

VEGF and FGF-2 encapsulation and release from bridgesThe encapsulation efficiency of the proteins within thebridge, losses during leaching, and release profile either byencapsulation within microspheres or mixing with micro-spheres was initially investigated. High molecular weight75:25 PLG was used to encapsulate VEGF and FGF-2 intoPLG microspheres, resulting in encapsulation efficiencies of44% 6 16% and 56% 6 14%, respectively, which were notstatistically different [Fig. 1(A)]. Varying the polymer solventagent (ethylacetate and dichloromethane), the lactide to gly-colide ratio (85:15, 75:25, and 65:35 PLG), or the PLG con-centration (3 and 5%) did not enhance the encapsulation ef-ficiency (data not shown). Protein incorporated inside thebridges through initial encapsulation into microspheresresulted in minimal protein losses during the leaching pro-cess, with protein retention of 93% 6 0% and 95% 6 2%,respectively, inside the bridge. Bridges fabricated by mixingthe protein with the microspheres before gas foaming had aretention of 41% 6 5% and 52% 6 1%, respectively, afterleaching [Fig. 1(B)], which was statistically different forboth proteins. Importantly, the total amount of protein thatcan be loaded inside the microspheres is limited due to thefirst emulsion volume and the initial protein concentration,which is not a limiting factor in the case of protein mixingwith the microspheres.

The release study was subsequently investigated, andrelease was a strong function of the incorporation methodand the protein incorporated. All conditions had an initialrapid release that persisted for 4 days, followed by a moregradual release. Bridges loaded with VEGF demonstrated afaster release compared to FGF-2 for both incorporationmethods with a greater percentage released after 23 weeks.

ORIGINAL ARTICLE


For bridges with mixed VEGF or FGF-2, respectively, the ini-tial phase led to a release of 61% 6 9% and 27% 6 6%,while the second phase released 26% 6 6% and 36% 63% during the subsequent 157 days. For encapsulatedVEGF or FGF-2, respectively, the initial phase had a releaseof 21% 6 4% and 12% 6 1%, and a sustained release dur-ing the subsequent 157 days with a maximal release of57% 6 4% and 40% 6 2% of the encapsulated protein.Bridges that were loaded with protein by both mixing andencapsulation inside the microspheres had an intermediaterelease profile (Fig. 2).

Bioactivity of protein released from bridgeThe bioactivity of VEGF and FGF-2 was subsequently deter-mined for protein that was released from bridges in which

the protein was encapsulated within the microspheres ormixed with the microspheres. Phosphorylation of VEGF-R2and FGF-R was observed for protein obtained followingrelease for 24 h, 5, 10, 20, and 42 days (Fig. 3), with activityobserved for protein that was mixed with or encapsulatedinto the microspheres. Variations of intensity between sam-ples are due to the variability in protein content for therelease solutions. This activity of VEGF and FGF-2 was con-sistent with multiple reports involving the release of theseproteins from PLG.35,37

VEGF localization at the injury siteVEGF-loaded bridges were subsequently implanted into thelateral hemisection. We have previously reported that thesebridges prevent the formation of a cystlike cavity typical ofmany spinal cord injuries, likely due to stabilizing the injuryfollowing implantation.6,27 Cells infiltrate into the pores andchannels of the bridge within a week, with the channelsaligning cells along the major axis of the channel. The pres-ence of the delivered protein at the injury site was analyzedby quantifying the amount of VEGF in the injury site and inadjacent spinal cord segments [Fig. 4(A)]. Bridges were

FIGURE 1. Protein encapsulation efficiency and protein retention after leaching. (A) VEGF and FGF-2 encapsulation efficiency after microsphere

encapsulation and collection. (B) VEGF and FGF-2 retention after leaching bridges for 1 h. Proteins were incorporated using either the encapsula-

tion only or mixing only approaches. Significant differences between protein conditions are denoted by an asterisk (p < 0.05).

FIGURE 2. Protein release from bridge. Bridges were fabricated with

VEGF (filled line) or FGF-2 (dotted line) in three different manners:

mixing only (squares), encapsulation only (triangles), or a combina-

tion of both encapsulation and mixing (circles). The amount of protein

left in the bridge after leaching was set as 100%.

FIGURE 3. VEGF and FGF-2 activity during in vitro release. Phospho-

rylation of (A) VEGF and (B) FGF-2 receptors was assessed by western

blot. Legend: (1) control (100 ng/mL VEGF or 50 ng/mL FGF-2), (2)

mixing 24 h, (3) encapsulation 24 h, (4) mixing 5 days, (5) encapsula-

tion 5 days, (6) mixing 10 days, (7) encapsulation 10 days, (8) mixing

20 days, (9) encapsulation 20 days, (10) mixing 42 days, (11) encapsu-

lation 42 days, and (12) negative control (PBS).


loaded with 0.5 lg VEGF inside microspheres and 0.5 lgVEGF mixed with microspheres, a loading that was sufficientfor quantification by ELISA. VEGF levels at the injury sitewere 20-fold greater than for bridges without VEGF [Fig.4(B)]. Additionally, for VEGF-loaded bridges, VEGF levels atthe injury site (4300 6 1924 pg/mL) were significantlygreater than the levels in the gelfoam covering the implantor in the segments rostral and caudal to the injury site.These results suggest that the delivered protein remainedlocal at the injury site 1-week postimplantation, with VEGFlevel that were 20-fold greater than the endogenous VEGFlevels after spinal cord injury.

Functional blood vessels formed following proteindeliveryThe presence of functional blood vessels in the bridge afterimplantation in the spinal cord hemisection was character-ized by micro-CT. As a control, the vascular network of anuninjured spinal cord was filled with contrast agent to vali-date the technique [Fig. 5(A)]. The protein loadings thatwere used were selected from literature reports in whichPLG scaffolds were used to deliver VEGF in vivo.34–36 Thecombinatorial delivery of VEGF and FGF-2 was investigatedbased on the potential for synergistic activity between thetwo factors to promote blood vessel development.38–41 Fol-lowing bridge implantation for 6 weeks, a greater extent ofblood vessels was visualized inside bridges when a highdose of VEGF (2 lg) was encapsulated within the micro-

spheres and when FGF-2 (1 lg) was mixed with the micro-spheres [Fig. 5(B)]. Bridges with an intermediate VEGF dose(1 lg) or without the proteins did not demonstrate thesame extent of blood vessel growth after spinal cord injury.Mixing FGF-2 with the microspheres in addition to VEGFencapsulation (1 lg) enhanced blood vessel ingrowth com-pared to VEGF mixed with the microspheres [Fig. 5(B)],while additional encapsulation of FGF-2 in the microspheresdid not enhance blood vessel growth (data not shown).

Infiltration of endothelial cells in bridgeThe presence of endothelial cells in the bridge was subse-quently evaluated at 6 weeks postimplantation by immuno-histochemistry. Based on the qualitative micro-CT results,which demonstrated that bridges with intermediate dosesof protein did not result in the same extent of blood vesselingrowth as bridges with the maximal doses and thatbridges with FGF-2 mixed with the microspheres in additionto VEGF enhanced the presence of blood vessels at theinjury site, the protein doses were increased to ensure thatthe protein loading was sufficient to promote endothelialcell infiltration and axonal extension into the bridge. Thedose increase was also done to account for the flow of cere-brospinal fluid that may enable convective transport of theproteins from the implant.5 Therefore, one dose of VEGF (4lg) was encapsulated in microspheres, and two differentdoses of VEGF and FGF-2 were mixed within the bridge.

For VEGF and FGF-2 releasing bridges, substantial RECAstaining was observed within the bridge and adjacent tissue[Fig. 6(A)], while intact circular-shaped blood vessels weredetected adjacent to the bridge [Fig. 6(B)]. Bridges loadedwith medium or high protein dosages had significantlygreater RECA staining relative to bridges without protein[Fig. 6(C)]. Bridges loaded with 4 lg of VEGF encapsulatedwithin their microspheres and 2 lg of FGF-2 and VEGFmixed (high protein) had a 3.3 6 1.0-fold greater RECAstaining relative to bridges without protein. Bridges contain-ing 4 lg of VEGF encapsulated within their microspheresand 1 lg of FGF-2 and VEGF mixed (medium protein) had a2.5 6 0.7-fold greater RECA staining. Differences betweenbridges with medium and high protein doses were not stat-istically significant.

Neurite growth into the bridgeThe extent of neurite growth within the bridge at 6 weekspostimplantation was subsequently characterized in orderto investigate synergy between the endothelial cells andneurite outgrowth. Substantial NF staining was present inall conditions [Fig. 7(A)]. For bridges without protein orwith the intermediate levels of VEGF/FGF-2, the mean levelof staining was similar [Fig. 7(B)]. For bridges loaded withthe highest dose of VEGF and FGF-2, the mean staining levelwas greater [1.7-fold relative to no protein, Fig. 7(B)]; how-ever, this difference was not statistically significant.

DISCUSSION

In this report, VEGF and FGF-2 were delivered to theinjured spinal cord from a bridge to enhance blood vessel

FIGURE 4. ELISA for VEGF at the injury and in adjacent tissue at 1-

week postimplantation. (A) Schematic of the spinal cord injury and

segments that were analyzed for ELISA, adapted from De Laporte L,

Yang Y, Zelivyanskaya ML, Cummings BJ, Anderson AJ, and Shea

LD. Plasmid releasing multiple channel bridges for transgene expres-

sion after spinal cord injury. Originally published in Ref. 6 with per-

mission of the Nature Publishing Group. (B) VEGF levels in different

segments for bridges loaded with and without VEGF. Significant dif-

ferences between conditions are denoted by an asterisk (p < 0.05).

ORIGINAL ARTICLE


formation and reduce ischemia, and to investigate if thedelivery of these factors impacts neurite outgrowth. Angio-genic factors, such as VEGF and FGF-2, have been deliveredfrom PLG scaffolds formed by the gas foaming process inorder to increase blood vessel formation locally29,32,42–44;yet, these strategies have not been applied to the spinalcord. The gas foaming process has been effective in previousreports at maintaining the bioactivity of both VEGF andFGF-2.29,42,45 Proteins were loaded into the bridge through(i) encapsulation into the microspheres and/or (ii) directmixing with the microspheres, which were subsequentlyprocessed by gas foaming to create the bridge structure.The method of protein incorporation impacts the releaseprofile,29 with encapsulation within the microspheres reduc-ing the release rate relative to the mixing procedure. FGF-2demonstrated a greater encapsulation efficiency and de-creased release relative to VEGF. FGF-2 (17.2 kDa) has asmaller molecular weight than VEGF (38.2 kDa), yet hasa greater isoelectric point that would impart a more positivecharge than VEGF and could enhance electrostatic interac-tions with the polymer. Both proteins retained their bioac-tivity after their release from the bridge up to 42 days.

Delivery of angiogenic factors from a bridge was hypothe-sized to provide the combination of structural support andbiochemical factors to promote revascularization of theinjury. In the normal spinal cord, VEGF mRNA is not detecta-ble; however, following spinal cord injury, VEGF expression is

upregulated after 6 h with a peak at 24 h.46 Expression is re-stricted to the border of the injury, with new vessels present24-h postinjury that eventually disappears with cavity forma-tion and glial scar deposition.46 These previous studies sug-gested that delivery should maintain VEGF at the injury andavoid release to the uninjured tissue and that limiting cavityformation may be essential for long-term vessel stability.Bridge implantation provides mechanical support to theinjury and supports cell infiltration, whose combination pre-vents cavity formation and decreases the extent of glial scarformation.27 Delivery of 1 lg of VEGF increased VEGF levels20-fold at the injury site 1-week postimplantation relative tobridges without VEGF. The delivered VEGF remained at theinjury, as segments rostral and caudal to the injury and thegelfoam had VEGF levels comparable to levels observed withbridges without VEGF. With a VEGF half-life of 3 min in thecirculation47 and 6 min when bound to heparin,48 a 1-weektime point suggests a local and sustained delivery of exogene-ous VEGF from the bridge in vivo after implantation, whichmay have the potential to support the formation of matureblood vessels at the injury site, while avoiding blood vesselretraction.

Localized delivery of angiogenic factors increased endo-thelial cell infiltration and also the presence of vessels inthe adjacent tissue. Significant differences in endothelial cellinfiltration into the bridges at 6 weeks postspinal cordinjury were observed for bridges releasing VEGF and FGF-2

FIGURE 5. (A) Cured microfil injection compound that formed a three-dimensional cast of the rat’s spinal cord vasculature. From left to right:

cross section and longitudinal view. (B) 3D reconstructions of micro-CT scans of bridge implants in spinal cord hemisection model at 6 weeks

postimplantation. From left to right: bridge implant without protein loading, bridge implant with 1 lg VEGF encapsulated and 1 lg VEGF mixed,

bridge implant with 1 lg VEGF encapsulated and 1 lg FGF-2 mixed, bridge implant with 2 lg VEGF encapsulated and 1 lg VEGF mixed. The

red-dotted line marks the contours of the bridge at the implant site. The green-dotted line marks the contour of some residual gelfoam that

appeared on micro-CT. The gelfoam was used to cover the injury site, and mostly removed upon tissue retrieval. [Color figure can be viewed in

the online issue, which is available at wileyonlinelibrary.com.]


relative to bridges without proteins. Interestingly, endothe-lial cell infiltration was not significantly different betweenmedium and high protein doses. This result suggests that amaximal dose was exceeded or that the mixing process wasineffective at greater doses due to the rapid release. Micro-CT analysis indicated the presence and position of functionalvessels within the injury site. Furthermore, histological anal-ysis indicated the presence of circular-shaped blood vesselsadjacent to the bridge in quantities that were increased for

bridges containing VEGF and FGF-2 protein relative to noprotein. Inside the bridges, however, few circular-shapedblood vessels were detected, which may suggest that thepore structure and size might need to be re-evaluated tosupport blood vessel ingrowth.49,50 In addition, the relation-ship between VEGF dosage and the morphology and func-tion of newly formed blood vessels has previously beeninvestigated, demonstrating that elevated VEGF levels canlead to abnormal blood vessels and hemangiomas.51,52 The

FIGURE 6. Endothelial cell infiltration in bridge at 6 weeks postimplantation. (A) RECA stain (brown): from left to right: bridge containing 4 lg of

VEGF encapsulated within microspheres and 2 lg of FGF-2 and VEGF mixed (high protein), and bridge without protein. The red-dotted line

marks the contours of the bridge at the implant site. Scale bar: 200 lm. (B) Blood vessels adjacent to the bridge containing 4 lg of VEGF encap-

sulated within microspheres and 1 lg of FGF-2 and VEGF mixed (medium protein). Scale bar: 200 lm. (C) Quantifcation of RECA staining for the

three conditions. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

ORIGINAL ARTICLE


studies reported here indicate the potential for localizedVEGF delivery to promote angiogenesis locally in the spinalcord. A key challenge will be forming normal mature bloodvessels that will restore the integrity of the blood brain bar-rier. To obtain more mature vessels, multiple researchgroups are currently investing a combination of multipleangiogenic factors (e.g., PDGF53), their doses, and the timeframe of delivery. More research is necessary to overcomethe challenges of controlling the time frame and doses ofprotein release, as parameters may have to be re-optimizedfor each protein or protein combination.

VEGF has been proposed as a neurotrophin,54 while,more generally, delivery of angiogenic factors may lead tothe simultaneous regeneration of blood vessels and nerves.Blood vessels and nerves may be congruent in order to pro-vide nerves with oxygen and nutrients and allow vasoregu-lation of blood vessels.19 Additionally, the blood vessels andnerve bundles may provide mechanical support for eachother and thus orienting regeneration. In this report, sub-stantial staining for NF was observed in empty bridges (i.e.,no protein release), for which there was limited staining forendothelial cells, suggesting that there is endogenous neuro-trophin production that is promoting neurite growth. Deliv-ery of the high dose of VEGF and FGF-2 increased the meanquantity of staining for NF in the bridge; however, thisincrease was not statistically significant. Additionally, pre-liminary staining for astrocytes revealed that the delivery ofVEGF and FGF-2 induced astrocytes infiltration into thepores and channels of the bridge (data not shown), which

was not observed in the absence of this delivery. Withoutprotein delivery, a dense layer of astrocytes is observed atthe edge of the bridge without infiltration inside the bridge,consistent with previous reports.27 Thus, the delivery ofangiogenic factors is impacting the local environment andstimulating cells other than endothelial cells.

The significant increase in endothelial cell staining, yetinsignificant change in NF staining, with protein deliverymay reflect that the pore structure of the scaffold or thequantity or duration of delivery may need to be altered.Although endothelial cell staining was increased within thebridge, the presence of organized vessel structures may benecessary to provide the mechanical support and orientationto promote complementary nerve growth. Altering the poresize has the potential to enhance the number of vesselswithin the bridge.49,50 Importantly, the action of VEGF as aneurotrophin is proposed to require long-term VEGF produc-tion,13 and the concentration at which VEGF is neurotrophicmay differ from the concentration at which it is angiogenic.

CONCLUSION

Bridges releasing angiogenic factors (VEGF and FGF-2)locally after spinal cord injury increased endothelial cellinfiltration and blood vessel formation at the injury, whilealso supporting neurite outgrowth, which are both neces-sary to overcome ischemia and promote functional spinalcord regeneration. VEGF and FGF-2 were delivered from thebridge in a controlled manner using a combination of twomethods of protein loading: encapsulation inside the

FIGURE 7. Neurite ingrowth with bridge at 6 weeks postimplantation. (A) Neurofilament stain (brown) of bridge implanted with the highest

VEGF/FGF-2 protein dose. The red-dotted line marks the contours of the bridge at the implant site. Scale bar: 500 lm. (B) Quantification of posi-

tive NF stain for bridge implants without VEGF/FGF-2 and bridge implants with a medium and high dose of VEGF/FGF-2 protein. [Color figure

can be viewed in the online issue, which is available at wileyonlinelibrary.com.]


microspheres and mixing with the microspheres before gasfoaming. Bridges stabilized the injury to prevent cavity for-mation and localized release promoted a local increase ofVEGF at the injury, increased endothelial cell infiltration andenhanced the number of circular-shaped blood vessels. Thisreport demonstrates that implantation of bridges releasingproteins into the spinal cord can stabilize the injury whilepromoting and supporting processes critical to regeneration.

ACKNOWLEDGMENTS

The authors are grateful for the rat VEGF that was kindly pro-vided by Prof. Carmeliet from the Katholieke UniversiteitLeuven, Belgium. The authors thank Prof. O. Feron and L. Petitfor their help with the protein bioactivity test and SamanthaHolland for help with immunohistochemistry.

REFERENCES1. Storkebaum E, Lambrechts D, Carmeliet P. VEGF: Once regarded

as a specific angiogenic factor, now implicated in neuroprotec-

tion. Bioessays 2004;26:943–954.

2. Jimenez Hamann MC, Tator CH, Shoichet MS. Injectable intrathe-

cal delivery system for localized administration of EGF and FGF-2

to the injured rat spinal cord. Exp. Neurol 2005;194:106–119.

3. Jones LL, Oudega M, Bunge MB, Tuszynski MH. Neurotrophic fac-

tors, cellular bridges and gene therapy for spinal cord injury.

J Physiol 2001;533(Pt 1):83–89.

4. Nakahara Y, Gage FH, Tuszynski MH. Grafts of fibroblasts geneti-

cally modified to secrete NGF, BDNF, NT-3, or basic FGF elicit dif-

ferential responses in the adult spinal cord. Cell Transpl 1996;5:

191–204.

5. De Laporte L, Yan AL, Shea LD. Local gene delivery from ECM-

coated poly(lactide-co-glycolide) multiple channel bridges after

spinal cord injury. Biomaterials 2009;30:2361–2368.

6. De Laporte L, Yang Y, Zelivyanskaya ML, Cummings BJ, Ander-

son AJ, Shea LD. Plasmid releasing multiple channel bridges for

transgene expression after spinal cord injury. Mol Ther 2009;17:

318–326.

7. Stokols S, Tuszynski MH. Freeze-dried agarose scaffolds with uni-

axial channels stimulate and guide linear axonal growth following

spinal cord injury. Biomaterials 2006;27:443–451.

8. De Laporte L, Huang A, Ducommun MM, Zelivyanska ML, Aviles

MO, Adler AF, Shea LD. Patterned transgene expression in multi-

ple-channel bridges after spinal cord injury. Acta Biomater 6:

2889–2897.

9. Biondi M, Ungaro F, Quaglia F, Netti PA. Controlled drug delivery

in tissue engineering. Adv Drug Deliv Rev 2008;60:229–242.

10. Tessmar JK, Gopferich AM. Matrices and scaffolds for protein

delivery in tissue engineering. Adv Drug Deliv Rev 2007;59:

274–291.

11. Willerth SM, Sakiyama-Elbert SE Approaches to neural tissue en-

gineering using scaffolds for drug delivery. Adv Drug Deliv Rev

2007;59:325–338.

12. Widenfalk J, Lipson A, Jubran M, Hofstetter C, Ebendal T, Cao Y,

Olson L. Vascular endothelial growth factor improves functional

outcome and decreases secondary degeneration in experimental

spinal cord contusion injury. Neuroscience 2003;120:951–960.

13. Facchiano F, Fernandez E, Mancarella S, Maira G, Miscusi M,

D’Arcangelo D, Cimino-Reale G, Falchetti ML, Capogrossi MC,

Pallini R. Promotion of regeneration of corticospinal tract axons

in rats with recombinant vascular endothelial growth factor alone

and combined with adenovirus coding for this factor. J Neuro-

surg 2002;97:161–168.

14. Ruiz de Almodovar C, Lambrechts D, Mazzone M, Carmeliet P.

Role and therapeutic potential of VEGF in the nervous system.

Physiol Rev 2009;89:607–648.

15. Sakakibara Y, Nishimura K, Tambara K, Yamamoto M, Lu F,

Tabata Y, Komeda M. Prevascularization with gelatin micro-

spheres containing basic fibroblast growth factor enhances the

benefits of cardiomyocyte transplantation. J Thorac Cardiovasc

Surg 2002;124:50–56.

16. Yau TM, Fung K, Weisel RD, Fujii T, Mickle DA, Li RK. Enhanced

myocardial angiogenesis by gene transfer with transplanted cells.

Circulation 2001;104(12, Suppl 1):I218–I222.

17. Kim HM, Hwang DH, Lee JE, Kim SU, Kim BG. Ex vivo VEGF

delivery by neural stem cells enhances proliferation of glial pro-

genitors, angiogenesis, and tissue sparing after spinal cord injury.

PLoS ONE 2009;4:e4987.

18. Carmeliet P, Tessier-Lavigne M. Common mechanisms of nerve

and blood vessel wiring. Nature 2005;436:193–200.

19. Eichmann A, Le Noble F, Autiero M, Carmeliet P. Guidance of vas-

cular and neural network formation. Curr Opin Neurobiol 2005;15:

108–115.

20. Autiero M, De Smet F, Claes F, Carmeliet P. Role of neural guid-

ance signals in blood vessel navigation. Cardiovasc Res 2005;65:

629–638.

21. Sondell M, Lundborg G, Kanje M. Vascular endothelial growth

factor has neurotrophic activity and stimulates axonal outgrowth,

enhancing cell survival and Schwann cell proliferation in the pe-

ripheral nervous system. J Neurosci 1999;19:5731–5740.

22. Mukouyama YS, Shin D, Britsch S, Taniguchi M, Anderson DJ.

Sensory nerves determine the pattern of arterial differentiation

and blood vessel branching in the skin. Cell 2002;109:693–705.

23. Shen Q, Goderie SK, Jin L, Karanth N, Sun Y, Abramova N, Vin-

cent P, Pumiglia K, Temple S. Endothelial cells stimulate self-

renewal and expand neurogenesis of neural stem cells. Science

2004;304:1338–1340.

24. Krum JM, Mani N, Rosenstein JM. Angiogenic and astroglial

responses to vascular endothelial growth factor administration in

adult rat brain. Neuroscience 2002;110:589–604.

25. Sondell M, Lundborg G, Kanje M. Vascular endothelial growth

factor stimulates Schwann cell invasion and neovascularization of

acellular nerve grafts. Brain Res 1999;846:219–228.

26. Hobson MI, Green CJ, Terenghi G. VEGF enhances intraneural

angiogenesis and improves nerve regeneration after axotomy.

J Anat 2000;197(Pt 4):591–605.

27. Yang Y, De Laporte L, Zelivyanskaya M, Whittlesey KJ, Anderson

A, Cummings BJ, Shea LD. Multiple channel bridges for spinal

cord injury: Cellular characterization of host response. Tissue Eng

Part A 2009;15:3283–3295.

28. De Clerck N, Postnov A. High Resolution X-Ray Microtomography:

Applications in Biomedical Research. London: Wiley; 2007. p 57–77.

29. Ennett AB, Kaigler D, Mooney DJ. Temporally regulated delivery

of VEGF in vitro and in vivo. J Biomed Mater Res A 2006;79:

176–184.

30. Yang Y, De Laporte L, Rives CB, Jang JH, Lin WC, Shull KR, Shea

LD. Neurotrophin releasing single and multiple lumen nerve con-

duits. J Control Release 2005;104:433–446.

31. Jang JH, Rives CB, Shea LD. Plasmid delivery in vivo from porous

tissue-engineering scaffolds: Transgene expression and cellular

transfection. Mol Ther 2005;12:475–483.

32. Peters MC, Polverini PJ, Mooney DJ. Engineering vascular net-

works in porous polymer matrices. J Biomed Mater Res 2002;60:

668–678.

33. Cleland JL, Duenas ET, Park A, Daugherty A, Kahn J, Kowalski J,

Cuthbertson A. Development of poly-(D,L-lactide-co-glycolide)

microsphere formulations containing recombinant human vascu-

lar endothelial growth factor to promote local angiogenesis.

J Control Release 2001;72:13–24.

34. Chen RR, Silva EA, Yuen WW, Mooney DJ. Spatio-temporal VEGF

and PDGF delivery patterns blood vessel formation and matura-

tion. Pharm Res 2007;24:258–264.

35. Kaigler D, Wang Z, Horger K, Mooney DJ, Krebsbach PH. VEGF

scaffolds enhance angiogenesis and bone regeneration in irradi-

ated osseous defects. J Bone Miner Res 2006;21:735–744.

36. Sun Q, Chen RR, Shen Y, Mooney DJ, Rajagopalan S, Grossman

PM. Sustained vascular endothelial growth factor delivery enhan-

ces angiogenesis and perfusion in ischemic hind limb. Pharm Res

2005;22:1110–1116.

37. Golub JS, Kim YT, Duvall CL, Bellamkonda RV, Gupta D, Lin AS,

Weiss D, Robert Taylor W, Guldberg RE. Sustained VEGF delivery

ORIGINAL ARTICLE


via PLGA nanoparticles promotes vascular growth. Am J Physiol

Heart Circ Physiol 2010;298:H1959–H1965.

38. Przybylski M. A review of the current research on the role of

bFGF and VEGF in angiogenesis. J Wound Care 2009;18:516–519.

39. Pepper MS, Ferrara N, Orci L, Montesano R. Potent synergism

between vascular endothelial growth factor and basic fibroblast

growth factor in the induction of angiogenesis in vitro. Biochem

Biophys Res Commun 1992;189:824–831.

40. Asahara T, Bauters C, Zheng LP, Takeshita S, Bunting S, Ferrara

N, Symes JF, Isner JM. Synergistic effect of vascular endothelial

growth factor and basic fibroblast growth factor on angiogenesis

in vivo. Circulation 1995;92(9 Suppl):II365–II371.

41. Ferrara N, Alitalo K. Clinical applications of angiogenic growth

factors and their inhibitors. Nat Med 1999;5:1359–1364.

42. Sheridan MH, Shea LD, Peters MC, Mooney DJ. Bioabsorbable

polymer scaffolds for tissue engineering capable of sustained

growth factor delivery. J Control Release 2000;64:91–102.

43. Rives CB, des Rieux A, Zelivyanskaya M, Stock SR, Lowe WL Jr,

Shea LD. Layered PLG scaffolds for in vivo plasmid delivery. Bio-

materials 2009;30:394–401.

44. Murphy WL, Peters MC, Kohn DH, Mooney DJ. Sustained release

of vascular endothelial growth factor from mineralized poly(lac-

tide-co-glycolide) scaffolds for tissue engineering. Biomaterials

2000;21:2521–2527.

45. Riddle KW, Kong HJ, Leach JK, Fischbach C, Cheung C, Anseth

KS, Mooney DJ. Modifying the proliferative state of target cells to

control DNA expression and identifying cell types transfected in

vivo. Mol Ther 2007;15:361–368.

46. Bartholdi D, Rubin BP, Schwab ME. VEGF mRNA induction corre-

lates with changes in the vascular architecture upon spinal cord

damage in the rat. Eur J Neurosci 1997;9:2549–2560.

47. George ML, Eccles SA, Tutton MG, Abulafi AM, Swift RI. Correla-

tion of plasma and serum vascular endothelial growth factor lev-

els with platelet count in colorectal cancer: Clinical evidence of

platelet scavenging? Clin Cancer Res 2000;6:3147–3152.

48. Pantely GA, Porter JM. Therapeutic angiogenesis: Time for the

next phase. J Am Coll Cardiol 2000;36:1245–1247.

49. Brauker JH, Carr-Brendel VE, Martinson LA, Crudele J, Johnston

WD, Johnson RC. Neovascularization of synthetic membranes

directed by membrane microarchitecture. J Biomed Mater Res

1995;29:1517–1524.

50. Lim TC, Chian KS, Leong KF. Cryogenic prototyping of chitosan

scaffolds with controlled micro and macro architecture and their

effect on in vivo neo-vascularization and cellular infiltration.

J Biomed Mater Res A 2010;94:1303–1311.

51. Ozawa CR, Banfi A, Glazer NL, Thurston G, Springer ML, Kraft PE,

McDonald DM, Blau HM. Microenvironmental VEGF concentra-

tion, not total dose, determines a threshold between normal and

aberrant angiogenesis. J Clin Invest 2004;113:516–527.

52. von Degenfeld G, Banfi A, Springer ML, Wagner RA, Jacobi J,

Ozawa CR, Merchant MJ, Cooke JP, Blau HM. Microenvironmen-

tal VEGF distribution is critical for stable and functional vessel

growth in ischemia. FASEB J 2006;20:2657–2659.

53. Richardson TP, Peters MC, Ennett AB, Mooney DJ. Polymeric sys-

tem for dual growth factor delivery. Nat Biotechnol 2001;19:

1029–1034.

54. Van Den Bosch L, Storkebaum E, Vleminckx V, Moons L, Vanop-

denbosch L, Scheveneels W, Carmeliet P, Robberecht W. Effects

of vascular endothelial growth factor (VEGF) on motor neuron

degeneration. Neurobiol Dis. 2004;17:21–28.


Vascular endothelial growth factor and fibroblast growth factor 2 delivery from spinal cord bridges to enhance angiogenesis following injury

Documents