Previous studies from our laboratory 1 6 and others 7 13 have

Transcription

1 Basic Science Reports Endothelial Progenitor Cell Vascular Endothelial Growth Factor Gene Transfer for Vascular Regeneration Hideki Iwaguro, MD; Jun-ichi Yamaguchi, MD, PhD; Christoph Kalka, MD; Satoshi Murasawa, MD, PhD; Haruchika Masuda, MD, PhD; Shin-ichiro Hayashi, MD, PhD; Marcy Silver, BS; Tong Li, MD; Jeffrey M. Isner, MD; Takayuki Asahara, MD, PhD Background Previous studies have established that bone marrow derived endothelial progenitor cells (EPCs) are present in the systemic circulation. In the current study, we investigated the hypothesis that gene transfer can be used to achieve phenotypic modulation of EPCs. Methods and Results In vitro, ex vivo murine vascular endothelial growth factor (VEGF) 164 gene transfer augmented EPC proliferative activity and enhanced adhesion and incorporation of EPCs into quiescent as well as activated endothelial cell monolayers. To determine if such phenotypic modulation may facilitate therapeutic neovascularization, heterologous EPCs transduced with adenovirus encoding VEGF were administered to athymic nude mice with hindlimb ischemia; neovascularization and blood flow recovery were both improved, and limb necrosis/autoamputation were reduced by 63.7% in comparison with control animals. The dose of EPCs used for the in vivo experiments was 30 times less than that required in previous trials of EPC transplantation to improve ischemic limb salvage. Necropsy analysis of animals that received DiI-labeled VEGF-transduced EPCs confirmed that enhanced EPC incorporation demonstrated in vitro contributed to in vivo neovascularization as well. Conclusions In vitro, VEGF EPC gene transfer enhances EPC proliferation, adhesion, and incorporation into endothelial cell monolayers. In vivo, gene-modified EPCs facilitate the strategy of cell transplantation to augment naturally impaired neovascularization in an animal model of experimentally induced limb ischemia. (Circulation. 2002;105: ) Key Words: gene therapy endothelium angiogenesis ischemia Previous studies from our laboratory 1 6 and others 7 13 have established that bone marrow derived endothelial progenitor cells (EPCs) are present in the systemic circulation, are augmented in response to certain cytokines and/or tissue ischemia, and home to as well as incorporate into sites of neovascularization. More recently, EPCs have been investigated as therapeutic agents; in these studies of supply-side angiogenesis, EPCs harvested from the peripheral circulation have been expanded ex vivo and then administered to animals with limb 14 or myocardial 15 ischemia to successfully enhance neovascularization. Physiological evidence of neovascular function in these preclinical animal models includes a high rate of limb salvage and improvement in myocardial function. See p 672 EPC transplantation thus constitutes a novel therapeutic strategy that could provide a robust source of viable endothelial cells (ECs) to supplement the contribution of ECs resident in the adult vasculature that migrate, proliferate, and remodel in response to angiogenic cues, according to the classic paradigm of angiogenesis developed by Folkman and colleagues. 16 Just as classic angiogenesis may be impaired in certain pathological phenotypes, however, aging, diabetes, hypercholesterolemia, and hyperhomocysteinemia may likewise impair EPC function, including mobilization from the bone marrow and incorporation into neovascular foci. Gene transfer of EPCs during ex vivo expansion constitutes a potential means of addressing such putative liabilities in EPC function. Moreover, phenotypic modulation of EPCs ex vivo may also reduce the number of EPCs required for optimal transplantation after ex vivo expansion and thus serve to address a practical limitation of EPC transplantation, namely the volume of blood required to extract an optimal number of EPCs for autologous transplantation. Accordingly, we investigated the hypothesis that gene transfer can be used to achieve phenotypic modulation of EPCs. In particular, we sought to determine the impact of vascular endothelial growth factor (VEGF) gene transfer on certain properties of EPCs in vitro and the consequences of VEGF EPC-gene transfer on neovascularization in vivo. Received October 12, 2001; revision received November 29, 2001; accepted December 14, From the Division of Cardiovascular Research and Medicine (H.I., J.Y., C.K., S.M., H.M., S.H., M.S., T.L., J.M.I., T.A.), St Elizabeth s Medical Center, Tufts University School of Medicine, Boston, Mass; and the Department of Physiology (H.I.) and Institute of Medical Science (T.A.), Tokai University School of Medicine, Japan. Correspondence to Takayuki Asahara, MD, PhD, St Elizabeth s Medical Center, 736 Cambridge St, Boston, MA, asa777@aol.com 2002 American Heart Association, Inc. Circulation is available at DOI: /hc

2 Iwaguro et al EPC Gene Transfer 733 Figure 1. Profile of transfection efficiencies for Ad/ -gal in ex vivo expanded human EPCs. Four different multiplicities of infection (MOI, 250, 500, 1000, and 2000) were tested in 2 different media conditions (1% or 5% serum EBM-2) for 1 or 3 hours of incubation. Error bars represent SEM of triplicate experiments. After these preliminary experiments, human EPCs were transduced with 1000 MOI Ad/VEGF or Ad/ gal for 3 hours in 1% serum media. Methods EPC Culture Ex vivo expansion of EPC was performed as recently described. 14 In brief, peripheral blood mononuclear cells from human volunteers were plated on human fibronectin coated (Sigma) culture dishes and maintained in EC basal medium-2 (EBM-2) (Clonetics) supplemented with 5% fetal bovine serum, human VEGF-A, human fibroblast growth factor-2, human epidermal growth factor, insulinlike growth factor-1, and ascorbic acid. After 4 days in culture, nonadherent cells were removed by washing, new media was applied, and the culture maintained through day 7. EPC Gene Transfer After 7 days in culture, cells were transduced with an adenovirus encoding the murine VEGF 164 gene (Ad/VEGF) or lacz gene (Ad/ gal) (generously provided by Kevin Peters). 21 To establish optimum conditions for EPC adenovirus gene transfer serum concentration, virus incubation time and virus concentration were evaluated (Figure 1). After preliminary experiments were performed, human EPCs were transduced with 1000 MOI Ad/VEGF or Ad/ gal for 3 hours in 1% serum media. After transduction, cells were washed with PBS and incubated with EPC media for 24 hours before transplantation. Proliferative Activity Assay At 24 hours after gene transfer, EPCs transduced with Ad/VEGF (Td/V-EPCs), Ad/ -gal (Td/ -EPCs), or nontransduced EPCs (non- Td/EPCs) were reseeded on 96-well plates coated with human fibronectin for assay of proliferative activity with the use of the MTS [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4- sulfophenyl)-2h-tetrazolium] Assay (Promega). After 48 hours in culture, MTS/PMS (phenazine methosulfate) solution was added to each well for 3 hours, and light absorbance at 490 nm was detected by ELISA plate reader (Bionetics Laboratory). In Vitro Incorporation of Td-EPCs Into Human Umbilical Vein Endothelial Cell Monolayer At 24 hours after gene transfer, Td/V-EPCs and Td/ -EPCs were stained with fluorescent carbocyanine DiI (Biomedical Technologies). DiI-labeled EPCs were incubated on a monolayer of human umbilical vein endothelial cells (HUVECs) in 4-well culture slides with or without pretreatment of tumor necrosis factor (TNF)- (1ng/mL) for 12 hours. 22 Three hours after incubation, nonadherent cells were removed by washing with PBS, new media was applied, and the culture was maintained for an additional 24 hours. The total number of adhesive EPCs in each well was counted in a blinded manner under a 200 magnification field of a fluorescent microscope. Td-EPCs Transplantation Animal Model All procedures were performed in accordance with the St Elizabeth s Institutional Animal Care and Use Committee. Female athymic nude mice (Jackson Laboratory, Bar Harbor, Maine), 8 to 9 weeks old and 17 to 20 g weight, were anesthetized with 160 mg/kg IP pentobarbital for operative resection of one femoral artery 23 and subsequently for laser Doppler perfusion imaging. As a preliminary experiment, we performed dose-dependent EPC transplantation to determine the minimum number of VEGFtransduced EPCs that was required to achieve a magnitude of therapeutic neovascularization similar to that which could be achieved with nontransduced EPCs. According to this result, VEGF-transduced EPCs, 30 times less than the number required for cell therapy alone, were used in the current series of in vivo experiments. One day after unilateral femoral artery excision, Td/V-EPCs (n 11), Td/ -EPCs (n 11), or non-td/epcs (n 5) in 100 L EBM-2 media without growth factors were administered through the tail vein. To track the fate of transplanted EPCs, 4 mice in each EPC cohort received EPCs that were marked with the fluorescent carbocyanine DiI dye (Molecular Probes). In brief, before cellular transplantation, EPCs in suspension were washed with PBS and incubated with DiI at a concentration of 2.5 g/ml PBS for 5 minutes at 37 C and 15 minutes at 4 C. After two washing steps in PBS, the cells were resuspended in EBM-2 medium. At 30 minutes before the animals were killed, a subgroup (n 4 each group) of mice received an intravenous injection of 50 g of Bandeiraea simplicifolia lectin 1 (BS-1) conjugated with FITC (Vector Laboratories). Plasma VEGF Levels To confirm that the Ad/VEGF could mediate successful gene transfer at the protein level, an enzyme-linked immunoassay (ELISA, R&D System) was used to quantify VEGF levels in plasma from animals 1, 4, 7, and 28 days after intravenous injections of Td/V-EPCs or Td/ -EPCs. The results were compared with a standard curve constructed with murine VEGF (each assay carried out in duplicate for each animal). Absorbance was measured at 450 nm by means of a microplate reader. Physiological Assessment of Animals Laser Doppler perfusion imaging (LDPI) (Moor Instruments) was used to record blood flow measurements on day 0 and day 28 after surgery, as previously described. In these digital color-coded images, red hue indicates regions with maximum perfusion, medium perfusion values are shown in yellow, and lowest perfusion values are represented as blue. The resulting images display absolute values in readable units. For quantification, the ratio of readable units in ischemic to nonischemic hindlimb is determined. Histological Assessment of Animals Vascular density was evaluated at the microvascular level through the use of light microscopic sections harvested from the ischemic hindlimbs at necropsy. Tissue sections from the lower calf muscles of ischemic limbs were harvested on days 7 and 28. Muscle samples were embedded in OCT compound (Miles), snap-frozen in liquid nitrogen, and cut into 5- m-thick sections. Tissue sections were stained for alkaline phosphatase with an indoxyltetrazolium method to detect capillary endothelial cells as previously described 23 and then were counterstained with eosin. A total of 20 different fields were randomly selected, and the number of capillaries and myofibers were counted ( 40 magnification for 20 fields).

3 734 Circulation February 12, 2002 Figure 2. Proliferative activity assay. Proliferative activity of EPCs transduced in 5% serum was measured by MTS assay after 48 hours in culture. Increase in mitogenic response of EPCs transduced with Ad/VEGF (Td/V-EPCs) was statistically significant in comparison with EPCs transduced with Ad/ gal (Td/ -EPCs) and nontransduced EPCs (non-td). *P 0.01 vs Td/ gal and non-td. Statistical Analysis All results are expressed as mean SEM. Statistical significance was evaluated by means of a paired Scheffé t test or ANOVA. A value of P 0.05 was considered to denote statistical significance. Results Proliferative Activity Assay MTS assay was used to determine proliferative activity of transduced EPCs. With the use of 5% serum-conditioned media, proliferative activity of Ad/VEGF-transduced EPCs exceeded proliferative activity of Ad/ -gal ( versus corrected absorbance at 490 nm, P 0.01) and nontransduced EPCs (non-td , P 0.05) in vitro (Figure 2). In Vitro Incorporation of Td-EPCs Into HUVEC Monolayer At 24 hours after transduction, EPCs were labeled with the fluorescent marker DiI for cell tracking. DiI-labeled, VEGFtransduced EPCs were incubated on a HUVEC monolayer with or without TNF- (1 ng/ml) pretreatment for 12 hours (Figure 3A). After 3 hours of incubation, nonadherent cells were removed by washing with PBS, and DiI-marked cells adherent to the HUVEC monolayer were manually counted. In the quiescent HUVEC monolayer, adhesion of DiI-labeled EPCs was not significantly different between Td/V-EPCs and Td/ -EPC ( versus , P NS) (Figure 3B). In activated HUVECs, however, adhesion of DiI-labeled Td/V- EPCs exceeded Td/ -EPCs ( versus , P 0.01). Alternatively, the same cells were incubated in new media and maintained for 24 hours on the HUVEC monolayer to confirm incorporation in vitro. In the quiescent HUVEC monolayer, incorporation of DiI-labeled Td/V-EPCs exceeded Td/ -EPCs ( versus , P 0.01) (Figure 3B). In activated HUVECs, incorporation of DiI-labeled Td/V-EPCs also exceeded Td/ -EPCs ( versus , P 0.001). Transgene Expression After Td/V-EPC Transplantation Ad/VEGF-mediated gene transfer and expression were confirmed by ELISA assay of plasma samples obtained from Figure 3. In vitro incorporation of Td-EPCs on HUVEC monolayer. A, Representative macroscopic photographs of Td/V- EPCs and Td/ -EPCs on HUVEC monolayer 24 hours after transduction with Ad/VEGF or Ad/ gal, respectively. Left panel shows Td/ -EPCs; right panel, Td/V-EPCs on HUVEC monolayer, in each case pretreated with TNF- stimulation. White bars indicate 50- m length. B, Quantitative analysis of EPC adhesion observed 3 hours and incorporation observed 24 hours after transduction with ( ) and without ( ) pretreatment of TNF-. *P 0.01 vs Td/ -EPCs. mice after Td/V-EPC transplantation. Mice transplanted with Td/V-EPCs disclosed significantly higher VEGF protein levels (day 1, versus pg/ml; day 4, versus pg/ml; day 7, versus pg/ml; day 28, versus pg/ml; P 0.01) at each time point than did mice transplanted with Td/ -EPCs (Figure 4). Figure 4. VEGF plasma levels after administration of Td/V-EPCs and Td/ -EPCs. A, Quantification of VEGF expression was measured by ELISA assay at each time point (days 1, 4, 7, and 28). B, table of results. *P 0.01 vs Td/ -EPCs.

4 Iwaguro et al EPC Gene Transfer 735 Figure 5. Histological identification of EPC incorporation in vivo. Transplanted human DiI-labeled, EPC-derived cells were identified by red fluorescence in histological sections retrieved from ischemic muscle. Host mouse vasculature was identified by green fluorescence in the same tissue sections. Td/V-EPC transplanted animals had increased numbers of EPC-derived vasculature as well as mouse vasculature compared with control animals (Td/ -EPC and non-td groups). White bars indicate 100- m length. A, Transplanted EPCderived vasculature and mouse vasculature were quantified for the same light microscopic fields. Density of DiI-labeled EPCs in tissue sections of skeletal muscle removed from ischemic limb was greater in the Td/V-EPCs group than either of the other two groups (B). Density of BS-1 lectin positive vasculature in tissue sections of skeletal muscles removed from the ischemic limb was also greater in the Td/V-EPCs group than either of the other 2 groups (C). *P vs non-td and Td/ -EPCs, **P 0.01 vs non-td and Td/ -EPCs. EPC Incorporation Into Ischemic Hindlimb Transplanted human EPC derived cells marked with DiI were identified in tissue sections by red fluorescence. In contrast, the mouse vasculature, stained by premortem administration of BS-1 lectin, was identified by green fluorescence in the same tissue sections. Td/V-EPC transplanted animals disclosed increased numbers of DiI-labeled, red fluorescent EPC derived vasculature as well as mouse vasculature versus control groups (Td/ -EPC and non-td group) (Figure 5A). Both transplanted EPC-derived vasculature and mouse vasculature were analyzed quantitatively in the same microscopic field. The density of DiI-labeled EPCs in tissue sections of hindlimb muscles was greater in the Td/V-EPC group ( mm 2 ) than either of the other two groups (non-td/epcs mm 2 ; Td/ -EPCs mm 2, P 0.001) at day 7 (Figure 5 B). The density of BS-1 lectin positive vasculature in sections of skeletal muscle removed from the ischemic limb was also greater in the Td/V-EPC group ( mm 2 ) than the other two groups (non-td/epcs mm 2 ; Td/ -EPCs mm 2, P 0.01) at day 7 (Figure 5C). Physiological Assessment of Animals The impact of human gene modified EPC administration on neovascularization was investigated in a murine model of hindlimb ischemia. One day after operative excision of one femoral artery, athymic nude mice (n 27), in which angiogenesis is characteristically impaired, 14,21 received an intravenous injection of transduced EPCs (Td/V-EPCs, n 11) or Td/ -EPCs (n 11). As additional control animals, 5 mice with hindlimb ischemia were identically injected with non-td/epcs. Enhanced neovascularization in mice transplanted with Td/V-EPCs led to important biological consequences, compared with control animals. After administration of Td/ -EPCs to 11 mice, 3 (27.2%) had extensive toe necrosis, and the remaining 8 (72.7%) underwent autoamputation of the ischemic limb (Figure 6A). In contrast, Td/V-EPC transplantation was associated with successful limb salvage in 7 (63.6%) of 11 animals; toe necrosis was limited to 3 (27.2%) mice, and only 1 (9%) had spontaneous limb amputation (Figure 6B). Serial examination of hindlimb perfusion by LDPI was performed at days 0 and 28 (Figure 7A). The ratio of ischemic/normal blood flow in mice transplanted with Td/V- EPCs indicated significantly greater hindlimb perfusion compared with those mice transplanted with Td/ -EPCs and nontransduced EPCs at day 28 ( versus versus , P 0.05) (Figure 7B). Histological Assessment of Animals To further evaluate the impact of EPC gene transfer on revascularization of the ischemic hindlimb, histological examination of skeletal muscle sections retrieved from the ischemic hindlimbs of mice killed at day 28 was performed as described above. Capillary density observed in the mice transplanted with Td/V-EPCs was significantly higher than in

5 736 Circulation February 12, 2002 Figure 6. Administration of hepcs leads to reduced limb loss and increased limb salvage. A, Representative macroscopic photographs of mice show two different outcomes observed in the study. Left panel, Td/ -EPC treated animals; right panel, Td/V-EPC treated animals at day 28. B, Percent distribution of above outcomes among mice receiving Td/ -EPCs and Td/V- EPCs. *P 0.01 vs Td/ -EPCs. mice receiving Td/ -EPCs ( versus /mm 2, P 0.05) or nontransduced mice EPCs (183 16/mm 2, P 0.05) (Figure 8). Similarly, the capillary/muscle fiber ratios in the Ad/VEGF transplanted mice were significantly higher than in Td/ -EPCs or non-td/epcs transplanted mice ( versus versus , P 0.01). Figure 7. LDPI performed at day 28. A, LDPI was used to record serial blood flow measurements immediately before and 28 days after administration of transduced EPCs. In these digital, color-coded images, red hue indicates regions with maximum perfusion; medium perfusion values are shown in yellow; lowest perfusion values are represented as blue. Upper panel, Animal that received Td/ -EPCs; lower panel, animal that received Td/V-EPCs. Panel on right displays absolute values in readable units (RU). B, Color-coded recordings were analyzed by calculating the average perfusion for each foot (ischemic and nonischemic). To account for variations, including ambient light and temperature, perfusion is expressed as ratio of left (ischemic) and right (normal) hindlimbs. *P 0.05 vs Td/ -EPCs and non-td/epcs. Figure 8. Histological evidence of neovascularization in ischemic hindlimb. Extent of neovascularization was assessed by measuring capillary density in light microscopic sections prepared from muscles of ischemic hindlimbs 28 days after administration of Td/ -EPCs, Td/V-EPCs and non-td/epcs. *P 0.05 vs Td/ -EPCs and non-td/epcs. Similar findings were documented for capillary/myocyte ratio (see text). Discussion The finding that circulating EPCs may home to sites of neovascularization and differentiate into ECs in situ is consistent with vasculogenesis, 24 a critical paradigm for establishment of the primordial vascular network in the embryo. Although the proportional contributions of angiogenesis and vasculogenesis to postnatal neovascularization remain to be clarified, our findings together with the recent reports from other investigators 7 13 suggest that growth and development of new blood vessels in the adult are not restricted to angiogenesis but encompass both embryonic mechanisms. As a corollary, augmented or retarded neovascularization whether endogenous or iatrogenic probably includes enhancement or impairment of vasculogenesis. Therapeutic Vasculogenesis We therefore considered a novel strategy of EPC transplantation to provide a source of robust ECs that might supplement fully differentiated ECs thought to migrate and proliferate from preexisting blood vessels according to the classic paradigm of angiogenesis developed by Folkman and colleagues. 16 Our studies indicated that ex vivo cell therapy, consisting of culture-expanded EPC transplantation, successfully promotes neovascularization of ischemic tissues, even when administered as sole therapy, for example, in the absence of angiogenic growth factors. Such a supply side version of therapeutic neovascularization in which the substrate rather than ligand comprises the therapeutic agent was first demonstrated in a immune deficient murine model of hindlimb ischemia, with the use of donor cells from human volunteers. 14 These findings provided novel evidence that exogenously administered EPCs augment naturally impaired neovascularization in an animal model of experimentally induced critical limb ischemia. Not only did heterologous cell transplantation improve neovascularization and blood flow recovery, but important biological consequences notably limb necrosis and autoamputation were reduced by 50% in comparison with mice receiving differentiated ECs or control mice receiving media in which harvested cells were expanded ex vivo before transplantation. More recently, this same strategy has been used successfully to enhance myocardial function after myocardial infarc-

6 Iwaguro et al EPC Gene Transfer 737 tion in experimental animal models as well. 15 Peripheral blood mononuclear cells obtained from healthy human adults were cultured in EPC medium, harvested 7 days later, and administered intravenously to Hsd:RH-rnu (athymic nude) rats with myocardial ischemia induced by ligation of the left anterior descending coronary artery. Fluorescent microscopy of DiI-labeled EPCs revealed that transplanted EPCs accumulated to the ischemic area and incorporated into foci of myocardial neovascularization. Echocardiography disclosed ventricular dimensions that were significantly smaller and fractional shortening that was significantly greater in the EPC versus control animals. Necropsy examination disclosed that capillary density was significantly greater and the extent of left ventricular scarring was significantly less in rats receiving EPCs versus control animals. Logistics of Primary EPC Transplantation Despite promising potential for regenerative applications, the fundamental scarcity of EPC populations in the hematopoietic system, combined with possible functional impairment of EPCs associated with a variety of phenotypes such as aging, diabetes, hypercholesterolemia, and homocysteinemia, constitute important limitations of primary EPC transplantation. Ex vivo expansion of EPCs cultured from the peripheral blood of healthy human volunteers typically yields cells per 100 ml of blood. Our animal studies 14 suggest that heterologous transplantation requires human EPCs/grams of weight (of the recipient mouse) to achieve satisfactory reperfusion of the ischemic hindlimb. Rough extrapolation of this experience to humans suggests that a volume of as much as 12 L of peripheral blood may be necessary to harvest the EPCs required to treat critical limb ischemia. Even with the integration of certain technical improvements, the adjustment of species compatibility by autologous transplantation, and adjunctive strategies (eg, cytokine supplements) to promote EPC mobilization, 3,4 the primary scarcity of a viable and functional EPC population constitutes a potential liability of therapeutic vasculogenesis based on the use of ex vivo expansion alone. Advantage of VEGF EPC Gene Transfer Our current findings provide the first evidence that exogenously administered, gene-modified EPCs augment naturally impaired neovascularization in an animal model of experimentally induced limb ischemia. Previous studies from our laboratory 14,21 established that neovascularization is impaired in nude rodents; failure of a satisfactory endogenous response to limb ischemia leads to extensive necrosis, including limb autoamputation, in nearly all animals. Transplantation of heterologous EPCs transduced with adenovirus encoding VEGF improved not only neovascularization and blood flow recovery but also had meaningful biological consequences: Limb necrosis and autoamputation were reduced by 63.7% in comparison with control animals. The dose of EPCs used in the current in vivo experiments was subtherapeutic; that is, this dose of EPCs was 30 times less than that required in previous experiments to improve the rate of limb salvage above that seen in untreated control animals. Adenoviral VEGF EPC gene transfer, however, accomplished a therapeutic effect, as evidenced by a functional outcome, despite a subtherapeutic dose of EPCs. Thus, VEGF EPC gene transfer constitutes one option to address the limited number of EPCs that can be isolated from peripheral blood before ex vivo expansion and subsequent autologous readministration. The results of our in vitro studies provide potential insights into the mechanisms responsible for the in vivo outcomes. First, VEGF gene transfer augmented EPC proliferative activity. Second, VEGF gene transfer enhanced adhesion and incorporation of EPCs into a quiescent endothelial cell monolayer as well as ECs activated by pretreatment with TNF-. These findings suggest that modulation of adhesion molecule expression after VEGF gene transfer may promote homing of EPCs to foci of ischemia and are consistent with previous studies demonstrating VEGF-induced upregulation of certain EC integrins and matrix proteins. 25 Given that EPCs by definition express VEGF receptors, the potential for autocrine effects demonstrated previously by our laboratory for ECs 26 on proliferation, migration, and possibly survival of EPCs probably contributed to the reduced requirement of harvested EPCs. Given the ELISA results demonstrating increased levels of VEGF protein in animals that received VEGF-transduced EPCs, it is also likely that indirect effects of VEGF transduction contributed to improved limb neovascularity. Because ex vivo expanded EPCs are preferentially recruited to ischemic foci for neovascularization, 14,15 EPCs may operate as a Trojan horse, promoting local overexpression of secreted VEGF that may in turn promote migration, proliferation, and remodeling of differentiated ECs resident in the target ischemic tissue. The extent to which augmented neovascularity derives from phenotypically modified EPCs versus enhanced proliferation and migration of native ECs in response to VEGF secreted from transduced EPCs is difficult to discern because of the lack of valid methods available to quantify local VEGF overexpression. Furthermore, it must be acknowledged that the possibility of similar outcomes achieved with non-epc circulating cells that overexpress VEGF (or a combination of VEGFexpressing non-epcs administered together with nontransduced EPCs) has not been excluded by the current studies. Testing such alternative approaches, however, is currently precluded by the lack of non-epc cells that can (a) be ex vivo transduced with equal efficiency; (b) circulate in vivo for some reasonable time period; and (c) be recruited to as well as incorporate into foci of neovascularization. Vector-Specific Issues Transient gene expression is characteristic of adenoviral vectors. 27 For purposes of therapeutic neovascularization, this feature is unlikely to constitute a liability, given the plethora of previous preclinical and clinical studies suggesting that VEGF overexpression for 4 weeks or less, whether achieved by transfer of naked plasmid DNA or use of an adenoviral vector, 32,33 is sufficient to augment angiogenesis. Indeed, previous work by others has demonstrated that protracted VEGF overexpression may result in hemangioma formation in normal skeletal muscle 34 as well as myocardi-

7 738 Circulation February 12, 2002 um. 35 This potential complication was not observed in the experiments described herein. The potential for immunologic complications remains a concern attached to the use of adenoviral vectors, despite certain reports to the contrary. 33 In the current application, however, the ex vivo transduction strategy used may preclude exposure of the adenovirus to the immune system of the transplant recipients. Thus, administration of the transgene by ex vivo viral gene transfer may not detract from the safety of this application for clinical gene therapy. References 1. Asahara T, Murohara T, Sullivan A, et al. Isolation of putative progenitor endothelial cells for angiogenesis. Science. 1997;275: Asahara T, Masuda H, Takahashi T, et al. Bone marrow origin of endothelial progenitor cells responsible for postnatal vasculogenesis in physiological and pathological neovascularization. Circ Res. 1999;85: Takahashi T, Kalka C, Masuda H, et al. Ischemia-, and cytokine-induced mobilization of bone marrow-derived endothelial progenitor cells for neovascularization. Nat Med. 1999;5: Asahara T, Takahashi T, Masuda H, et al. VEGF contributes to postnatal neovascularization by mobilizing bone marrow-derived endothelial progenitor cells. EMBO J. 1999;18: Kalka C, Masuda H, Takahashi T, et al. Vascular endothelial growth factor165 gene transfer augments circulating endothelial progenitor cells in human subjects. Circ Res. 2000;86: Kalka C, Tehrani H, Laudenberg B, et al. Mobilization of endothelial progenitor cells following gene therapy with VEGF 165 in patients with inoperable coronary disease. Ann Thorac Surg. 2000;70: Shi Q, Rafii S, Wu MH-D, et al. Evidence for circulating bone marrowderived endothelial cells. Blood. 1998;92: Hatzopoulos AK, Folkman J, Vasile E, et al. Isolation and characterization of endothelial progenitor cells from mouse embryos. Development. 1998;125: Gunsilius E, Duba HC, Petzer AL, et al. Evidence from a leukaemia model for maintenance of vascular endothelium by bone-marrow-derived endothelial cells. Lancet. 2000;355: Gehling UM, Ergun S, Schumacher U, et al. In vivo differentiation of endothelial cells from AC133-positive progenitor cells. Blood. 2000;95: Crosby JR, Kaminski WE, Schatteman G, et al. Endothelial cells of hematopoietic origin make a significant contribution to adult blood vessel formation. Circ Res. 2000;87: Moldovan NI, Goldschmidt-Clermont PJ, et al. Contribution of monocytes/macrophages to compensatory neovascularization: the drilling of metalloelastase-positive tunnels in ischemic myocardium. Circ Res. 2000; 87: Murohara T, Ikeda H, Duan J, et al. Transplanted cord blood-derived endothelial precursor cells augment postnatal neovascularization. J Clin Invest. 2000;105: Kalka C, Masuda H, Takahashi T, et al. Transplantation of ex vivo expanded endothelial progenitor cells for therapeutic neovascularization. Proc Natl Acad Sci U S A. 2000;97: Kawamoto A, Gwon H-C, Iwaguro H, et al. Therapeutic potential of ex vivo expanded endothelial progenitor cells for myocardial ischemia. Circulation. 2001;103: Folkman J. Tumor angiogenesis. In: Holland JF, Frei E III, Bast RC Jr, et al, eds. Cancer Medicine. 3rd ed. Philadelphia, Pa: Lea & Febiger; 1993: Gerhard M, Roddy M-A, Creager SJ, et al. Aging progressively impairs endothelium-dependent vasodilation in forearm resistance vessels of humans. Hypertension. 1996;27: Cosentino F, Luscher TF. Endothelial dysfunction in diabetes mellitus. J Cardiovasc Pharmacol. 1998;32:S54 S Drexler H, Zeiher AM, Meinzer K, et al. Correction of endothelial dysfunction in coronary microcirculation of hypercholesterolemic patients by L-arginine. Lancet. 1991;338: Luscher TF, Tshuci MR. Endothelial dysfunction in coronary artery disease. Annu Rev Med. 1993;44: Couffinhal T, Silver M, Kearney M, et al. Impaired collateral vessel development associated with reduced expression of vascular endothelial growth factor in ApoE -1- Mice. Circulation. 1999;99: Simmons PJ, Masinovdky B, Longenecker BM, et al. Vascular cell adhesion molecule-1 expressed by bone marrow stromal cells mediates the binding of hematopoietic progenitor cells. Blood. 1992;80: Couffinhal T, Silver M, Zheng LP, et al. A mouse model of angiogenesis. Am J Pathol. 1998;152: Risau W, Sariola H, Zerwes H-G, et al. Vasculogenesis and angiogenesis in embryonic stem cell-derived embryoid bodies. Development. 1988; 102: Senger DR, Ledbetter SR, Claffey KP, et al. Stimulation of endothelial cell migration by vascular permeability factor/vascular endothelial growth factor through cooperative mechanisms involving the v 3 integrin, osteopontin, and thrombin. Am J Pathol. 1996;149: Namiki A, Brogi E, Kearney M, et al. Hypoxia induces vascular endothelial growth factor in cultured human endothelial cells. J Biol Chem. 1995;270: Watanabe T, Kusznyski C, Ino K, et al. Gene transfer into human bone marrow hematopoietic cells mediated by adenovirus vectors. Blood. 1996;87: Losordo DW, Pickering JG, Takeshita S, et al. Use of the rabbit ear artery to serially assess foreign protein secretion after site specific arterial gene transfer in vivo: evidence that anatomic identification of successful gene transfer may underestimate the potential magnitude of transgene expression. Circulation. 1994;89: Isner JM, Pieczek A, Schainfeld R, et al. Clinical evidence of angiogenesis following arterial gene transfer of phvegf 165. Lancet. 1996;348: Baumgartner I, Pieczek A, Manor O, et al. Constitutive expression of phvegf 165 following intramuscular gene transfer promotes collateral vessel development in patients with critical limb ischemia. Circulation. 1998;97: Losordo DW, Vale PR, Symes J, et al. Gene therapy for myocardial angiogenesis: initial clinical results with direct myocardial injection of phvegf 165 as sole therapy for myocardial ischemia. Circulation. 1998; 98: Mack CA, Patel SR, Schwarz EA, et al. Biologic bypass with the use of adenovirus-mediated gene transfer of the complementary deoxyribonucleic acid for vascular endothelial growth factor 121 improves myocardial perfusion and function in the ischemic porcine heart. J Thorac Cardiovasc Surg. 1998;115: Giordano FJ, Ping P, McKirnan D, et al. Intracoronary gene transfer of fibroblast growth factor-5 increases blood flow and contractile function in an ischemic region of the heart. Nat Med. 1996;2: Springer ML, Chen AS, Kraft PE, et al. VEGF gene delivery to muscle: potential role of vasculogenesis in adults. Mol Cell. 1998;2: Lee RJ, Springer ML, Blanco-Bose WE, et al. VEGF gene delivery to myocardium: deleterious effects of unregulated expression. Circulation. 2000;102: