Journal of the American College of Cardiology Vol. 59, No. 14, 2012© 2012 by the American College of Cardiology Foundation ISSN 0735-1097/$36.00

PRE-CLINICAL RESEARCH

Sustained Improvement in Perfusion andFlow Reserve After Temporally SeparatedDelivery of Vascular Endothelial Growth Factorand Angiopoietin-1 Plasmid Deoxyribonucleic Acid

Alexandra H. Smith, MSC,* Michael A. Kuliszewski, BSC,* Christine Liao, BSC,*Dmitriy Rudenko, BSC,* Duncan J. Stewart, MD,† Howard Leong-Poi, MD*

Toronto, Ontario, Canada

Objectives The aim of this study was to compare temporally separated vascular endothelial growth factor (VEGF) and angio-poietin (Ang)-1 delivery with concomitant delivery or single VEGF delivery, for therapeutic angiogenesis in chronicischemia.

Background Single gene delivery of VEGF results in immature neovessels that ultimately regress. Endogenously, VEGF actsearly to initiate angiogenesis, whereas Ang-1 acts later to induce vessel maturation. Timing VEGF and Ang-1gene delivery to mimic endogenous angiogenesis might be more effective for sustained neovascularization.

Methods Unilateral hindlimb ischemia was induced in 170 rats. Ultrasound-mediated gene delivery was performed withcationic microbubbles and plasmid deoxyribonucleic acid. Groups included VEGF at 2 weeks, VEGF/Ang-1 at 2weeks, VEGF at 2 weeks with Ang-1 at 4 weeks, and untreated control subjects. At 2, 4, and 8 weeks after liga-tion, blood flow and flow reserve (FR) were assessed by contrast-enhanced ultrasound. Vascular density, organiza-tion, and supporting cell coverage were assessed by fluorescent microangiography and immunohistochemistry.

Results In untreated control subjects, blood flow, FR, and vessel density remained reduced. The VEGF delivery improvedflow and vessel density at 4 weeks; however, FR remained low, supporting cell coverage was poor, and flow andvessel density regressed by 8 weeks. The VEGF/Ang-1 co-delivery marginally increased flow and vessel density;however, FR and supporting cell coverage improved. After temporally separated VEGF and Ang-1 delivery, bloodflow, vessel density, and FR increased and were sustained, with improved pericyte coverage at 8 weeks.

Conclusions In conclusion, temporally separated VEGF and Ang-1 gene therapy results in sustained and functionalneovascularization. (J Am Coll Cardiol 2012;59:1320–8) © 2012 by the American College of CardiologyFoundation

Published by Elsevier Inc. doi:10.1016/j.jacc.2011.12.025

b

Cardiovascular gene therapy aimed at promoting new vesselformation or angiogenesis for the treatment of refractorycoronary artery disease (CAD) and peripheral arterial dis-ease (PAD) has been extensively studied (1). Althoughadvances in our understanding of the biology of new vessel

From the *Division of Cardiology, Keenan Research Centre in the Li Ka ShingKnowledge Institute, St. Michael’s Hospital, University of Toronto, Toronto,Ontario, Canada; and the †Ottawa Hospital Research Institute, University of Ottawa,Ottawa, Ontario, Canada. This work was supported by an Operating Grant (MOP62763) from the Canadian Institutes of Health Research, Ottawa, Ontario, Canada,and an Equipment Grant from the Canadian Foundation for Innovation, Ottawa,Ontario, Canada. Dr. Leong-Poi is supported by a Clinician Scientist Phase II Awardfrom the Heart and Stroke Foundation of Ontario, Ottawa, Ontario, Canada, and anEarly Researcher Award from the Ministry of Research and Innovation, Ontario,Canada. All other authors have reported that they have no relationships relevant tothe contents of this paper to disclose.

(Manuscript received August 11, 2011; revised manuscript received November 26,

2011, accepted December 15, 2011.

growth has led to promising results of early open-labeledhuman trials, the translation of the basic science of angio-genesis into a clinically useful and widely applicable strategyfor therapeutic angiogenesis in patients has not yet occurred.The largely disappointing results of randomized clinicaltrials of gene therapy for ischemic CAD (2,3) and PAD (4)can be attributed to several factors, including the choice ofthe optimal biological agent, a limited time course oftreatment (usually single fixed dose), and suboptimal deliv-ery techniques (5).

All large double-blind, randomized, placebo-controlledclinical trials of gene therapy for angiogenesis have used a1-time administration of a single therapeutic gene and havefailed to conclusively show significant benefit. Most recently,oth the Euroinject One Trial (2) and the NORTHERN

NOGA angiogenesis Revascularization Therapy: assess-

mpcn

giiiinciAtasolftWh

M

DAtHhDlMesdp

CwS(s4IwvCam(csRdlwpVwdEamgaAau

1321JACC Vol. 59, No. 14, 2012 Smith et al.April 3, 2012:1320–8 Temporally Separated Multi-Gene Delivery

ment by RadioNuclide imaging) trial (3) used a catheterdelivery system to percutaneously administer a single dose ofVEGF165 plasmid deoxyribonucleic acid (DNA) to ischemic

yocardium in “no-option” patients, showing no difference inrimary efficacy endpoints between active treatment and pla-ebo. The results of these studies suggest that VEGF mightot be the ideal candidate for single gene therapy.Angiopoietins (Ang) are another important class of

rowth factors specific for vascular endothelial cells thatnteract closely with VEGF (6). Although VEGF seems to bemportant during the initiation phase of angiogenesis, increas-ng permeability and degrading the extracellular matrix, Ang-1s important late in the angiogenic process, contributing toeovessel maturation (6). Accordingly, Ang-1 is a promisingandidate to compliment VEGF gene delivery. Given themportance of the temporal relationship between VEGF andng-1 during endogenous angiogenesis, the timeline of

ransfection of these growth factors during therapeuticngiogenesis might be important. Therefore, we hypothe-ized that a strategy of temporally separated transfectionf VEGF and Ang-1, delivering VEGF early and Ang-1ater, would result in a greater, more sustained, andunctional angiogenic response, as compared with simul-aneous delivery of VEGF and Ang-1 or VEGF alone.

e tested our hypothesis with a model of chronicindlimb ischemia in rats.

ethods

etailed methods are available in the Online Appendix.nimal preparation. The study protocol was approved by

he Animal Care and Use Committee at St. Michael’sospital Research Centre, University of Toronto. Proximal

indlimb ischemia was induced in 170 male Sprague-awley rats (Charles River, Wilmington, Massachusetts) by

igation of the left common iliac artery (7,8).icrobubble and DNA preparation. For contrast-

nhanced ultrasound (CEU) perfusion imaging, lipid-helled perfluoropropane microbubbles were used. For geneelivery, cationic lipid microbubbles were charge-coupled tolasmid complementary DNA (500 �g) (7,8). Plasmids

were constructed by incorporating the human VEGF165gene into the vector pcDNA3.1(�) and the human Ang-1gene into the vector pFLAG-CMV-1 (Addgene, Cam-bridge, Massachusetts).Gene delivery. Gene delivery to the proximal ischemichindlimb was performed by ultrasound-targeted micro-bubble destruction (UTMD), as previously described (7–9).For VEGF/Ang-1 early animals, Ang-1 was delivered 5min after VEGF.Perfusion imaging: CEU. Blood flow was measured in theproximal hindlimb adductor muscles with CEU duringcontinuous intravenous infusion of perfusion microbubbles(1 � 107 min�1) (10). In brief, after background subtrac-tion, contrast-enhanced images from multiple increasing

pulsing intervals (200 ms to 40 s) are used to generate w

replenishment curves, which areanalyzed to obtain parameters ofmicrovascular perfusion (10). Forthe measurement of blood flowreserve (FR), the ratio of exerciseblood flow to resting flow wasmeasured by CEU during elec-trically stimulated muscle con-traction (3 mA, pulse frequency 2Hz) induced by a pulse generator(AV Sequential Demand PulseGenerator, Medtronic 5330,Medtronic, Mississauga, On-tario, Canada).Fluorescent microangiography-. Fluorescent microangiography(FMA), a technique of postmor-tem vascular casting, was per-formed as previously described(7–9). Vessel density was calcu-lated from the total length ofvessels/tissue volume, and totallength of a vascular tree in 0.036cm3 of tissue was quantified.

apillary and arteriolar densityas calculated by branch order analysis based on thetrahler Method (11) with Neurolucida Software packageMBF Bioscience, Williston, Vermont), as previously de-cribed (9). Ischemic limbs were compared between groups atand 8 weeks after ligation.

mmunohistochemistry. Immunohistochemical stainingas performed on hindlimb muscle, with antibodies againston Willebrand Factor (Santa Cruz Biotechnology, Santaruz, California) (endothelial cells), alpha-smooth muscle

ctin (SMA) (Abbiotec, San Diego, California) (smoothuscle cells), platelet-derived growth factor receptor

PDGFR)-beta, Abcam, Cambridge, Massachusetts) (peri-ytes) and NG-2 (Abcam) (pericytes). Nuclei were counter-tained with Topro-3 (blue).eal-time polymerase chain reaction. At 3 days afterelivery (2 and 4 weeks after ligation) and 8 weeks after

igation, animals were sacrificed, and adductor muscle tissueas isolated, ribonucleic acid was extracted, and real-timeolymerase chain reaction for exogenous Ang-1 and totalEGF expression was performed on complementary DNAith the housekeeping gene Cyclophylin, as previouslyescribed (7–9).xperimental protocol. Iliac artery ligation was performed

t Day 0. After 2 weeks, baseline perfusion and FR wereeasured. Gene delivery was performed with UTMD as per

roup assignment—VEGF at 2 weeks after ligation (VEGFlone), VEGF and Ang-1 co-delivery at 2 weeks (VEGF/ng-1 early), or VEGF at 2 weeks and Ang-1 at 4 weeks

fter ligation (VEGF � Ang-1 late). Control animals werentreated. Blood flow and FR were reassessed at 4 and 8

Abbreviationsand Acronyms

Ang-1 � angiopoietin-1

Ang-2 � angiopoietin-2

CAD � coronary arterydisease

CEU � contrast-enhancedultrasound

DNA � deoxyribonucleicacid

FMA � fluorescentmicroangiography

FR � flow reserve

PAD � peripheral arterialdisease

PDGFR � platelet-derivedgrowth factor receptor

SMA � smooth muscleactin

UTMD � ultrasound-targeted microbubbledestruction

VEGF � vascularendothelial growth factor

eeks after ligation. Animals were

sacrificed at 3 days after

CaGmD

R

VAtitiCpFrDw

l8idV

d(0cAaiaptoc0a

1322 Smith et al. JACC Vol. 59, No. 14, 2012Temporally Separated Multi-Gene Delivery April 3, 2012:1320–8

delivery and at 4 and 8 weeks after ligation to collect tissuefor postmortem analysis.Statistical methods. Data are expressed as mean � SEM.

omparisons between groups and time points were made with2-way analysis of variance (GraphPad Prism5, version 5.0a;raphPad, La Jolla, California). When differences betweeneans were found, Bonferroni correction was performed.ifferences were considered significant when p � 0.05.

esults

EGF and Ang-1 expression after gene delivery. Humanng-1 messenger ribonucleic acid expression in Ang-1-

reated animals and total VEGF messenger ribonucleic acidn VEGF-treated animals peaked 3 days after delivery inheir respective groups and decreased by 8 weeks, confirm-ng transient transgene expression (Fig. 1).

EU resting perfusion and FR. Representative CEUerfusion images from each treatment group are shown inigure 2A. In control untreated ischemic muscle, resting flow

emained reduced at all time-points after ligation (Fig. 2B).elivery of VEGF at 2 weeks after ligation increased flow at 4eeks (0.58 � 0.26 to 0.86 � 0.23, p � 0.001), but flow

returned to baseline levels by 8 weeks. Blood flow increasedmarginally after simultaneous VEGF/Ang-1 co-delivery(0.69 � 0.13 at 4 weeks, p � NS vs. 2 weeks, p � 0.05 vs.Control at 4 weeks). After VEGF with Ang-1 late delivery,resting blood flow increased 2 weeks after VEGF delivery andremained elevated after Ang-1 delivery at 4 weeks (0.90 �0.21, p � 0.001 vs. 2 weeks and vs. all other groups at 8 weeks).Similar to changes in blood flow, blood volume increased afterVEGF single delivery in the VEGF alone (0.94 � 0.33 at 2weeks to 1.21 � 0.33 at 4 weeks, p � 0.01) and VEGF withAng-1 late groups (0.82 � 0.20 at 2 weeks to 1.30 � 0.43 at4 weeks, p � 0.001), but increases in blood volume persisted at8 weeks only in the VEGF with late Ang-1 treated animals.

Figure 1 Gene Expression by Real-Time Polymerase Chain Rea

(Left) Expression of human angiopoietin (Ang)-1, compared with cyclophilin (Cyclo)sion in all groups. Data from after 2-week delivery in white, after 4-week delivery iweeks. #p � 0.05, ##p � 0.01 versus Control. mRNA � messenger ribonucleic a

The FR in normal limbs was 2.41 � 0.06 (Fig. 2C). Incontrol and VEGF alone animals, FR remained reduced atall time-points. By comparison, FR improved after VEGF/Ang-1 co-delivery and after temporally separated delivery ofVEGF and Ang-1.Vessel density and vascular architecture by FMA. Rep-resentative FMA images from each treatment group at 4and 8 weeks are shown in Figure 3A. After delivery ofVEGF alone, vascular density (Fig. 3C) improved com-pared with control at 4 weeks (p � 0.001) but was notsustained to 8 weeks. After VEGF/Ang-1 co-delivery,vessel density was not significantly different from controlsubjects. After VEGF with Ang-1 late delivery, vesseldensity was elevated compared with all other treatmentgroups (21,817 � 1,036 vessels/mm3 in VEGF with Ang-1ate at 8 weeks, p � 0.001 vs. Control and VEGF alone at

weeks, p � 0.01 vs. VEGF/Ang-1 at 8 weeks). Changesn total vascular length (Fig. 3B) were similar to vascularensity. Significant improvement was observed only afterEGF with Ang-1 late delivery, similar to vessel density.After VEGF delivery, capillary density and arteriole

ensity increased, compared with control muscle at 4 weekscapillary density p � 0.001 vs. control; arteriole density p �.01 vs. control) (Figs. 4A and 4B). By 8 weeks, bothapillary and arteriole densities returned to baseline levels.fter VEGF with Ang-1 late delivery, however, capillary

nd arteriole densities remained increased at 8 weeks (cap-llary density p � 0.05 vs. control and p � 0.01 vs. VEGFlone at 8 weeks; arteriole density p � 0.001 vs. control and� 0.01 vs. VEGF alone at 8 weeks). In addition, after

emporally separated delivery of VEGF and Ang-1, arteri-le/capillary ratio was significantly higher compared withontrol and VEGF alone-treated muscle at 8 weeks (0.81 �.13 after VEGF with Ang-1 late vs. 0.68 � 0.03 in controlnd vs. 0.67 � 0.10 after VEGF alone, both p � 0.05).

-1-treated groups. (Right) Total vascular endothelial growth factor (VEGF) expres-, and at 8 weeks after ligation in black. Mean � SEM. **p � 0.01 versus 8A1 � vascular endothelial growth factor/angiopoietin 1; wk � week.

ction

in Angn greycid; V/

1323JACC Vol. 59, No. 14, 2012 Smith et al.April 3, 2012:1320–8 Temporally Separated Multi-Gene Delivery

Arteriolar density and mural cell coverage by immuno-fluorescence. Late (8 weeks after ligation) after gene de-livery, increased alpha-SMA-positive arterioles were foundin VEGF with Ang-1 late-treated ischemic muscle(Fig. 4C). The PDGFR-beta and NG-2 expressing peri-cytes were increased in the VEGF/Ang-1 early and VEGF �Ang-1 late groups, compared with control and VEGFgroups, with the greatest co-localization of von Willebrandfactor-positive endothelial cells and pericytes expressingPDGFR-beta and NG-2 seen after temporally separated

Figure 2 Contrast-Enhanced Ultrasound Perfusion Data

(A) Representative contrast-enhanced ultrasound images at three pulsing intervalsment groups at 8 weeks after ligation. Contrast signal increases at a faster rate ablood flow. (B) Resting blood flow in the ischemic hindlimb adductor muscles norm0.001 versus 2 weeks and control. #p � 0.05 versus control. ###p � 0.001 versand 8 weeks after ligation. Dotted line highlights normal flow reserve. Mean � SEAbbreviations as in Figure 1.

VEGF � Ang-1 delivery (Fig. 5).

Discussion

Clinical trials of VEGF gene therapy for therapeutic angio-genesis in CAD and PAD have yielded disappointingresults. One of the major limitations of “first generation”angiogenic factors like VEGF is that resultant neovessels areimmature and poorly formed (12,13), and might regress.Our previous studies using UTMD to deliver VEGFplasmid DNA to ischemic skeletal muscle (7,8) and theresults of our present study support this, with an increase inblood flow and vessel density 2 weeks after VEGF delivery

, 7 s, and 20 s (top to bottom) of the proximal ischemic hindlimb for all 4 treat-s a higher plateau in the VEGF�Ang-1 late-treated muscle, signifying greater

to contralateral limb at 2, 4, and 8 weeks after ligation. Mean � SEM. ***p �

other groups at 8 weeks. (C) Flow reserve in the ischemic hindlimbs at 2, 4,p � 0.01, ***p � 0.001 versus 2 weeks. #p � 0.05 versus control and VEGF.

—1 snd haalizedus allM. **

that does not persist at later time points. Delivery of Ang-1

1324 Smith et al. JACC Vol. 59, No. 14, 2012Temporally Separated Multi-Gene Delivery April 3, 2012:1320–8

plasmid DNA 2 weeks after VEGF delivery results in asustained improvement in blood flow and vessel density andaugmentation of blood FR. Our study demonstrates that astrategy of temporally separated multi-gene delivery ofVEGF and Ang-1, based on the temporal pattern of growthfactor expression during endogenous angiogenesis, results ina more sustained and functional angiogenic response.Limitations of VEGF single delivery. Although pre-clinical studies of VEGF for therapeutic angiogenesis haveshown consistent increases in blood flow and vessel density(7,8,14), studies that measured angiogenic outcomes at �4weeks after delivery often found late regression (7,8,11,15).Disorganized, leaky neovasculature after increased VEGFexpression has been noted in several studies of therapeuticangiogenesis (12,13,16,17). Some trimming of excessiveneovascularization normally accompanies angiogenesis (18),likely related to endothelial cell apoptosis upon withdrawalof pro-angiogenic factors (19) and reduced shear stress dueto disorganized branching (20). In our study, FR did not

Figure 3 Representative Fluorescent Microangiography Images

(A) Representative fluorescent microangiography images at 4 and 8 weeks after li8 weeks after ligation. (C) Vascular density/tissue volume. Mean � SEM. *p � 00.01 versus V/A1 early. Abbreviations as in Figure 1.

improve after VEGF single delivery, suggesting the new

vasculature lacked the capacity to respond to increasedoxygen demand. Our findings suggest that the lack offunctional FR in VEGF-induced vasculature might be dueto poor supporting cell coverage (21) and disorganizedcollateralization.Temporal regulation of endogenous angiogenesis. Simi-lar to VEGF, the angiopoietins, Ang-1 and Ang-2, are amajor class of angiogenic cytokines specific for vascularendothelial cells (6). Ang-1 has only a modest influence onvessel growth (6,11,22–24), in contrast to the potent angio-genic effect of VEGF; however, Ang-1 plays an integral rolein vascular remodeling and maturation (6). AlthoughVEGF causes vascular instability and increased permeabil-ity, Ang-1 stabilizes vessels, increases pericyte coverage, anddecreases vascular leakage (22,25,26). The VEGF andAng-2 initiate angiogenesis, acting synergistically to in-crease permeability and degrade the extracellular matrix.Ang-2 inhibits Tie-2 phosphorylation (27), blocking thestabilizing actions of Ang-1 and facilitating VEGF

Total Vascular Length and Density/Tissue Volume

. (B) Total vascular length (mm)/tissue volume (mm3) in ischemic limbs at 4 and*p � 0.01, ***p � 0.001 versus Control. ###p � 0.001 versus VEGF. ��p �

and

gation.05, *

action in the initiation of angiogenesis. Ang-1 seems to

IArwp

Vad

omhrIodvfawgpOd

1325JACC Vol. 59, No. 14, 2012 Smith et al.April 3, 2012:1320–8 Temporally Separated Multi-Gene Delivery

be important late in the angiogenic process, contributingto stabilization and maturation of neovessels (27), whileopposing VEGF (22).Combination VEGF and Ang-1 gene therapy for angio-genesis. Due to the different yet complementary roles ofthese 2 angiogenic growth factors in angiogenesis, theirco-administration has been previously studied. Interest-ingly, studies of simultaneous delivery of VEGF with Ang-1have shown conflicting results; some show benefit (24,28),whereas others show that dual delivery is less effectivecompared with single VEGF delivery (15,23,29). Viral genedelivery of VEGF/Ang-1 after myocardial infarction indiabetic rats increases capillary/arteriolar density after 1week (28); however, an incremental angiogenic effect aboveVEGF alone is not observed in rat hindlimb ischemia (23).n a rabbit ischemic hindlimb model, combination VEGF/ng-1 gene delivery produced only a modest effect on

esting blood flow and capillary formation, as comparedith either factor alone (24). Our study has clarified the

Figure 4 Capillary and Arteriole Densities by Fluorescent Micro

(A) Number of capillaries (order 0 vessels)/tissue volume (mm3). (B) Number of anumbered and color-coded by branch order. White arrow indicates direction of flowversus VEGF. �p � 0.05, ���p � 0.001 versus V/A1 early. (C) Immunostaining oligation—smooth muscle actin (SMC) (magenta) and Topro-3 nuclei (blue), showintions as in Figure 1.

attern emerging from previous multi-gene studies that V

EGF/Ang-1 co-delivery causes a significant yet modestnd transient increase in resting blood flow and vesselensity.The limitation of using VEGF and Ang-1 simultane-

usly might relate to the ability of Ang-1 to disrupt the idealicroenvironment of VEGF. Studies using transgenic mice

ave clearly demonstrated that Ang-1 can act as a negativeegulator of VEGF-induced angiogenesis in the heart (30).ndeed, late in the angiogenic process, Ang-1 binds to Tie-2n endothelial cells to induce vessel maturation by matrixeposition, increasing intercellular interactions and reducingascular permeability (25), counteracting VEGF-mediated ef-ects. The VEGF-induced blood-brain barrier permeability isssociated with increased matrix metalloproteinase-9 activity,ith both effects being decreased by Ang-1 therapy, sug-esting that Ang-1 attenuation of VEGF effects might inart be mediated through matrix metalloproteinase-9 (31).ur study results indicate that, when VEGF and Ang-1 are

elivered simultaneously, Ang-1 stabilization impedes

graphy and Immunostaining

es (order 1 to order 3 vessels)/tissue volume. (Right, top) Vascular tracingn � SEM. *p � 0.05, **p � 0.01, ***p � 0.001 versus Control. ##p � 0.01rioles from ischemic muscle from each treatment group at 8 weeks aftertest arteriolar density in VEGF�Ang1 late-treated ischemic muscle. Abbrevia-

angio

rteriol. Meaf arteg grea

EGF vessel formation, leading to more modest increases

1326 Smith et al. JACC Vol. 59, No. 14, 2012Temporally Separated Multi-Gene Delivery April 3, 2012:1320–8

in resting perfusion as compared with VEGF alone. How-ever, we noted an increase in blood FR in animals co-treatedwith VEGF and Ang-1 plasmid DNA, corresponding toincreased vascular network maturation with improved sup-porting cell coverage, supporting a predominant Ang-1effect. This supports some of the previously reported bene-ficial effects of co-administration (29), benefits that areenhanced even further when delivered separately in a tem-poral fashion.

This study revealed that the key to taking maximaladvantage of each growth factor is to mimic the endogenoustimeline of gene expression (27,32). Only 2 studies to datehave attempted to test this hypothesis. Yamauchi et al. (33)studied temporal combinations of VEGF and Ang-1 genedelivery in a rabbit ischemic hindlimb model. Although they

Figure 5 Representative Images From Normal Adductor MuscleEach Treatment Group at 8 Weeks After Ligation, Dem

von Willebrand factor (vWF) (Santa Cruz Biotechnology, Santa Cruz, California)-endbar � 100 �m. (B) Pericyte coverage, platelet-derived growth factor receptor (PDGeither (A) NG2 (Abcam, Cambridge, Massachusetts) or (B) PDGFR-beta. Abbreviat

concluded that pre-administration of Ang-1 followed 5 days

later by VEGF induced the greatest functional vascularformation, the results do not fully support their conclusions.Angiographic scores were highest in the VEGF-alonegroup, compared with other groups, whereas vessel densityand laser Doppler perfusion measurements varied widely.Ang-1 followed by VEGF was significantly greater thancontrol, yet there were no significant differences betweendual-treatment groups, including Ang-1 followed byVEGF, VEGF/Ang-1 co-administration, and VEGF fol-lowed by Ang-1. Differences in gene delivery methods(intramuscular vs. ultrasound-mediated [8]), perfusion as-sessment technique, and timing of sequential gene delivery(5 vs. 14 days) might have accounted for differences in studyfindings. Importantly, we chose to deliver Ang-1 at Day 14after VEGF, a time-point after delivery when exogenous

Ischemic Muscle Fromtrating Supporting Cell Coverage

l cells (red) and Topro-3-nuclei (blue). (A) Pericyte coverage, NG2 (green). Scaleta (green). Scale bar � 20 �m. Yellow indicates co-localization of vWF ands in Figure 1.

andons

otheliaFR)-beions a

VEGF expression becomes undetectable (7,8), thus avoid-

1

1

1

1

1

1

1

1

1

1

2

2

1327JACC Vol. 59, No. 14, 2012 Smith et al.April 3, 2012:1320–8 Temporally Separated Multi-Gene Delivery

ing any competing effects of VEGF on Ang-1 therapy. The5-day period between delivery of Ang-1 and VEGF in thestudy by Yamauchi et al. (33) likely results in simultaneouslyactive VEGF and Ang-1 gene expression in all theirdual-treatment groups, possibly accounting for the lack ofdifferences between groups.

Peirce et al. (11) used a dorsal skinfold backpack windowchamber to evaluate vascular pattern changes in response tofocally delivered recombinant VEGF and Ang-1 proteinwith intravital microscopy. The VEGF delivery resulted inincreased density of vessels out to 14 days but was reducedto control levels by Day 21. Ang-1 protein delivery resultedin lesser increases in vessel density when compared withVEGF alone. The addition of recombinant Ang-1 on Day7 after VEGF delivery resulted in sustained increases out toDay 21, vessel branch order ratios comparable to controllevels, and increased alpha-SMA-positive vessels (11). Ourstudy results confirm and extend these findings in a clinicallyrelevant model of chronic hindlimb ischemia. Early VEGFdelivery increased resting flow by increasing vessel density.Late (14 days) Ang-1 delivery stabilized vessels by providingperi-endothelial support (pericytes and smooth muscle cells)and improving arteriole/capillary ratio. The organized vas-culature, improved supporting cell coverage, and the recog-nized ability of Ang-1 to reduce vascular permeability (26)generate a sustained improvement in both resting bloodflow and FR. In this respect, Ang-1 mimics the function ofthe direct pericyte chemo-attractant PDGF-beta, whichalso results in maturation of VEGF-induced neovascular-ization (34).

Conclusions

We have demonstrated that amplifying growth factor ex-pression according to endogenous timelines allows for anearly VEGF-induced increase in blood vessel growth and alate Ang-1–induced improvement in vascular stabilizationand supporting cell coverage, resulting in a long-lastingimprovement in resting blood flow and FR. This suggests anew paradigm for angiogenic gene therapy, whereby notonly the right combinations of therapeutic genes but alsothe timing of delivery and action need to be considered toachieve an optimal therapeutic response.

AcknowledgmentThe authors would like to thank Quiwang Zhang, PhD, forproviding the angiopoietin-1 plasmid used in the study.

Reprint requests and correspondence: Dr. Howard Leong-Poi,6-044 Queen Wing, St. Michael’s Hospital, 30 Bond Street, Toronto,Ontario M5B 1W8, Canada. E-mail: leong-poih@smh.ca.

REFERENCES

1. Gupta R, Tongers J, Losordo DW. Human studies of angiogenic genetherapy. Circ Res 2009;105:724–36.

2. Kastrup J, Jorgensen E, Ruck A, et al. Direct intramyocardial plasmid

vascular endothelial growth factor-A165 gene therapy in patients with

stable severe angina pectoris A randomized double-blindplacebo-controlled study: the Euroinject One trial. J Am Coll Cardiol2005;45:982–8.

3. Stewart DJ, Kutryk MJ, Fitchett D, et al. VEGF gene therapy fails toimprove perfusion of ischemic myocardium in patients with advancedcoronary disease: results of the NORTHERN trial. Mol Ther 2009;17:1109–15.

4. Rajagopalan S, Mohler ER III, Lederman RJ, et al. Regional angio-genesis with vascular endothelial growth factor in peripheral arterialdisease: a phase II randomized, double-blind, controlled study ofadenoviral delivery of vascular endothelial growth factor 121 inpatients with disabling intermittent claudication. Circulation 2003;108:1933–8.

5. Simons M, Bonow RO, Chronos NA, et al. Clinical trials in coronaryangiogenesis: issues, problems, consensus: an expert panel summary.Circulation 2000;102:E73–86.

6. Asahara T, Chen D, Takahashi T, et al. Tie2 receptor ligands,angiopoietin-1 and angiopoietin-2, modulate VEGF-induced postna-tal neovascularization. Circ Res 1998;83:233–40.

7. Leong-Poi H, Kuliszewski MA, Lekas M, et al. Therapeutic arterio-genesis by ultrasound-mediated VEGF165 plasmid gene delivery tochronically ischemic skeletal muscle. Circ Res 2007;101:295–303.

8. Kobulnik J, Kuliszewski MA, Stewart DJ, Lindner JR, Leong-Poi H.Comparison of gene delivery techniques for therapeutic angiogenesisultrasound-mediated destruction of carrier microbubbles versus directintramuscular injection. J Am Coll Cardiol 2009;54:1735–42.

9. Kuliszewski MA, Kobulnik J, Lindner JR, Stewart DJ, Leong-Poi H.Vascular gene transfer of SDF-1 promotes endothelial progenitor cellengraftment and enhances angiogenesis in ischemic muscle. Mol Ther2011;19:895–902.

0. Wei K, Jayaweera AR, Firoozan S, Linka A, Skyba DM, Kaul S.Quantification of myocardial blood flow with ultrasound-induceddestruction of microbubbles administered as a constant venous infu-sion. Circulation 1998;97:473–83.

1. Peirce SM, Price RJ, Skalak TC. Spatial and temporal control ofangiogenesis and arterialization using focal applications of VEGF164and Ang-1. Am J Physiol Heart Circ Physiol 2004;286:H918–25.

2. Lee RJ, Springer ML, Blanco-Bose WE, Shaw R, Ursell PC, BlauHM. VEGF gene delivery to myocardium: deleterious effects ofunregulated expression. Circulation 2000;102:898–901.

3. Schwarz ER, Speakman MT, Patterson M, et al. Evaluation of theeffects of intramyocardial injection of DNA expressing vascular endo-thelial growth factor (VEGF) in a myocardial infarction model in therat—angiogenesis and angioma formation. J Am Coll Cardiol 2000;35:1323–30.

4. Tsurumi Y, Takeshita S, Chen D, et al. Direct intramuscular genetransfer of naked DNA encoding vascular endothelial growth factoraugments collateral development and tissue perfusion. Circulation1996;94:3281–90.

5. Zhou YF, Stabile E, Walker J, et al. Effects of gene delivery oncollateral development in chronic hypoperfusion: diverse effects ofangiopoietin-1 versus vascular endothelial growth factor. J Am CollCardiol 2004;44:897–903.

6. Masaki I, Yonemitsu Y, Yamashita A, et al. Angiogenic gene therapyfor experimental critical limb ischemia: acceleration of limb loss byoverexpression of vascular endothelial growth factor 165 but not offibroblast growth factor-2. Circ Res 2002;90:966–73.

7. Drake CJ, Little CD. Exogenous vascular endothelial growth factorinduces malformed and hyperfused vessels during embryonic neovas-cularization. Proc Natl Acad Sci U S A 1995;92:7657–61.

8. Oses P, Renault MA, Chauvel R, et al. Mapping 3-dimensionalneovessel organization steps using micro-computed tomography in amurine model of hindlimb ischemia-brief report. Arterioscler ThrombVasc Biol 2009;29:2090–2.

9. Benjamin LE, Golijanin D, Itin A, Pode D, Keshet E. Selectiveablation of immature blood vessels in established human tumorsfollows vascular endothelial growth factor withdrawal. J Clin Invest1999;103:159–65.

0. Dimmeler S, Zeiher AM. Endothelial cell apoptosis in angiogenesisand vessel regression. Circ Res 2000;87:434–9.

1. Korpisalo P, Karvinen H, Rissanen TT, et al. Vascular endothelialgrowth factor-A and platelet-derived growth factor-B combination

gene therapy prolongs angiogenic effects via recruitment of interstitial

2

2

2

2

2

2

2

2

3

3

3

3

3

g

1328 Smith et al. JACC Vol. 59, No. 14, 2012Temporally Separated Multi-Gene Delivery April 3, 2012:1320–8

mononuclear cells and paracrine effects rather than improved pericytecoverage of angiogenic vessels. Circ Res 2008;103:1092–9.

2. Thurston G, Suri C, Smith K, et al. Leakage-resistant blood vessels inmice transgenically overexpressing angiopoietin-1. Science 1999;286:2511–4.

3. Arsic N, Zentilin L, Zacchigna S, et al. Induction of functionalneovascularization by combined VEGF and angiopoietin-1 genetransfer using AAV vectors. Mol Ther 2003;7:450–9.

4. Chae JK, Kim I, Lim ST, et al. Coadministration of angiopoietin-1and vascular endothelial growth factor enhances collateral vasculariza-tion. Arterioscler Thromb Vasc Biol 2000;20:2573–8.

5. Suri C, Jones PF, Patan S, et al. Requisite role of angiopoietin-1, aligand for the TIE2 receptor, during embryonic angiogenesis. Cell1996;87:1171–80.

6. Thurston G, Rudge JS, Ioffe E, et al. Angiopoietin-1 protects the adultvasculature against plasma leakage. Nat Med 2000;6:460–3.

7. Matsunaga T, Warltier DC, Tessmer J, Weihrauch D, Simons M,Chilian WM. Expression of VEGF and angiopoietins-1 and -2 duringischemia-induced coronary angiogenesis. Am J Physiol Heart CircPhysiol 2003;285:H352–8.

8. Samuel SM, Akita Y, Paul D, et al. Coadministration of adenoviralvascular endothelial growth factor and angiopoietin-1 enhances vas-cularization and reduces ventricular remodeling in the infarcted myo-cardium of type 1 diabetic rats. Diabetes 2010;59:51–60.

9. Benest AV, Salmon AH, Wang W, et al. VEGF and angiopoietin-1stimulate different angiogenic phenotypes that combine to enhance

functional neovascularization in adult tissue. Microcirculation 2006;13:423–37.

0. Visconti RP, Richardson CD, Sato TN. Orchestration of angiogenesisand arteriovenous contribution by angiopoietins and vascular endothe-lial growth factor (VEGF). Proc Natl Acad Sci U S A 2002;99:8219–24.

1. Valable S, Montaner J, Bellail A, et al. VEGF-induced BBB perme-ability is associated with an MMP-9 activity increase in cerebralischemia: both effects decreased by Ang-1. J Cereb Blood Flow Metab2005;25:1491–504.

2. Wakui S, Yokoo K, Muto T, et al. Localization of Ang-1, -2, Tie-2,and VEGF expression at endothelial-pericyte interdigitation in ratangiogenesis. Lab Invest 2006;86:1172–84.

3. Yamauchi A, Ito Y, Morikawa M, et al. Pre-administration ofangiopoietin-1 followed by VEGF induces functional and maturevascular formation in a rabbit ischemic model. J Gene Med 2003;5:994–1004.

4. Kupatt C, Hinkel R, Pfosser A, et al. Cotransfection of vascularendothelial growth factor-A and platelet-derived growth factor-B viarecombinant adeno-associated virus resolves chronic ischemic malper-fusion role of vessel maturation. J Am Coll Cardiol;56:414–22.

Key Words: angiogenesis y cationic microbubbles y chronic ischemia yene therapy y ultrasound.

APPENDIX

For detailed Methods, please see the online version of this article.