In vivo electroporation enhances plasmid-based gene transfer of basic fibroblast growth factor for the treatment of ischemic limb

Journal of Surgical Research 120, 37– 46 (2004) doi:10.1016/j.jss.2003.12.016

In Vivo Electroporation Enhances Plasmid-Based Gene Transfer of Basic Fibroblast Growth Factor for the Treatment of Ischemic Limb Seiji Nishikage,* ,† Hiroyuki Koyama,* ,† ,1 Tetsuro Miyata,† Shigeyuki Ishii,* ,† Hirohumi Hamada,‡ and Hiroshi Shigematsu† *Department of Vascular Regeneration, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan; †Division of Vascular Surgery, Department of Surgery, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan; and ‡Department of Molecular Medicine, Sapporo Medical University, Hokkaido, Japan Submitted for publication August 21, 2003

and bFGF-E ⴚ groups were significantly higher than that in the LacZ-E ⴚ group. Conclusion. These data suggest that in vivo electroporation enhances bFGF gene transfer for the treatment of ischemic limb muscles. © 2004 Elsevier Inc. All rights reserved. Key Words: basic fibroblast growth factor; in vivo electroporation; chronic limb ischemia; collateral vessel; naked DNA.

Background. Angiogenic therapy for ischemic tissues using angiogenic growth factors has been reported on an experimental and a clinical level. Electroporation enhances the efficiency of plasmid-based gene transfer in a variety of tissues. The purpose of this study was to evaluate the angiogenic effects of plasmidbased gene transfer using basic fibroblast growth factor (bFGF) in combination with electroporation. Materials and methods. The transfection efficiency of in vivo electroporation in rabbit skeletal muscles was evaluated using pCAcclucⴙ encoding luciferase. To evaluate the angiogenic effects of bFGF gene in ischemic limb, we constructed a plasmid, pCAcchbFGFcs23, containing human bFGF cDNA fused with the secretory signal sequence of interleukin (IL)-2. Then, 500 ␮g of pCAcchbFGFcs23 or pCAZ3 (control plasmid) was injected into the ischemic thigh muscles in a rabbit model of hind limb ischemia with in vivo electroporation (bFGF-E ⴙ group and LacZ-E ⴙ group). Other sets of animals were injected with pCAcchbFGFcs23 (bFGF-E ⴚ group) or pCAZ3 (LacZ-E ⴚ group) without electroporation. Then 28 days later, calf blood pressure ratio, angiographic score, in vivo blood flow, and capillary density in the ischemic limb were measured. Results. Gene transfer efficiency increased markedly with the increase in voltage up to 100 V. Regarding angiogenic responses, calf blood pressure ratio, in vivo blood flow, and capillary density only in the bFGF-E ⴙ group were significantly higher than those in LacZ-E ⴚ group. Angiographic scores in the bFGF-E ⴙ

INTRODUCTION

Enhancement of collateral vessel development is expected to open a new therapeutic strategy for the treatment of vascular occlusive disease. Local delivery of angiogenic growth factors is a promising concept to enhance collaterals, and a variety of delivery procedures have been developed and tested in animal models of ischemia [1– 4]. Of these procedures, the most investigated is intramuscular injection of naked plasmid DNA encoding an angiogenic growth factor gene [3, 5]. Since striated muscle cells potentially take up and express a foreign gene in the form of naked plasmid DNA [6], the injected gene of an angiogenic growth factor was transferred into ischemic muscle cells, and then the gene-transduced muscle cells secreted the growth factor protein, inducing collateral vessel development in the ischemic muscle tissue. Previous studies using a rabbit model of hind limb ischemia showed that injection of vascular endothelial growth factor (VEGF) plasmid DNA significantly promoted collateral vessel development and improved tissue perfusion [3], and favorable results were also reported in a clinical trial of the treatment of leg and myocardial ischemia [7, 8]. However, one possible disadvantage of this procedure might be relatively low efficiency of gene transfer to

1 To whom correspondence and reprint requests should be addressed at Division of Vascular Surgery, Department of Surgery, Graduate School of Medicine, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8655, Japan. E-mail: hkoyama-tky@umin. ac.jp.

37

0022-4804/04 $30.00 © 2004 Elsevier Inc. All rights reserved.

38

JOURNAL OF SURGICAL RESEARCH: VOL. 120, NO. 1, JULY 2004

muscle cells. Indeed, Tsurumi et al. showed that only 2.2% of myocytes were transfected with injected plasmid DNA under optimal conditions [3]. This finding indicates a limitation of the efficacy of this procedure. Although animal experiments have demonstrated adequate effects after plasmid DNA injection, there is a possibility that these effects are not sufficient for the treatment of severe ischemic conditions in humans. Electroporation is known to introduce DNA into several kinds of cells in vitro [9]. Recent studies showed that in vivo electroporation markedly enhanced the gene transfer of injected plasmid DNA to several tissues [10 –12]. Aihara and Miyazaki injected an expression plasmid vector containing the IL-5 gene into the bupivacaine-treated muscle of the mouse hind limb followed by electroporation and detected more than 20-fold higher expression of IL-5 protein in serum, as compared with the sole injection of the plasmid [13]. Basic fibroblast growth factor (bFGF) protein is known to stimulate the development of collaterals in ischemic tissues [14, 15], so the objective of the present study is to evaluate the efficacy of in vivo electroporation method to improve the effect of sole injection of plasmid DNA encoding bFGF on collateral vessel development. As an angiogenic growth factor for gene transfer, we used modified human bFGF gene fused with the secretory signal sequence of IL-2 (pCAcchbFGFcs23). Because of the signal sequence, the muscle cells transfected with pCAcchbFGFcs23 could secrete bFGF protein to the outside of the cells [16].

described previously [19]. The DNA synthetic activity of secreted bFGF was assessed by incorporation of 3H-thymidine, as described previously [19].

MATERIALS AND METHODS

Animal Model of Hind Limb Ischemia and in Vivo Gene Transfer by Electroporation

Gene Transfer Efficiency into Muscle by Electroporation in Vivo Male Japanese White rabbits (Saitama Rabbitry, Saitama, Japan) were used to determine the optimal conditions of electroporation in vivo. All of the protocols were approved by the institutional review board, and animal care complied with the “Guide for the Care and Use of Laboratory Animals,” Institute of Laboratory Animal Resources, Commission of Life Sciences, National Research Council. Under anesthesia, the surface of the semimembranous muscle was exposed, and the central part of the exposed muscle received direct injection of 100 ␮g of pCAccluc⫹ suspended in 500 ␮l PBS. A 27gauge needle was used for these injections, and its tip was positioned at a depth of 5 mm below the muscle surface. Subsequently, a pair of tungsten steel electrode needles 10 mm apart was inserted into the muscle where the DNA was injected. The electrodes were positioned transverse to muscle fibers, and the depth of the electrode tip was 7 mm below the muscle surface. An electric pulse generator ECM 830 (BTX, San Diego, CA, USA) delivered three square-wave pulses of the indicated voltage followed by three pulses of opposite polarity at a rate of 1 pulse/s, with a pulse duration of 50 ms. The animals were killed 7 days later, and luciferase activity of the muscle was measured. The excised semimembranous muscle was homogenized, and 1 g of the homogenate was lysed in 1 ml luciferase lysis reagent (Promega, Madison, WI, USA). The luciferase activity of the lysate was measured using a Luciferase Assay System (Promega). Some rabbits were injected with pCAZ3 instead of pCAccluc⫹, and X-gal staining of the semimembranous muscle visualized the gene transfer in vivo. To evaluate muscle damage caused by electroporation, the muscle tissue between the electrodes was fixed and embedded in paraffin. Sections (4 ␮m) were stained with hematoxylin/eosin and evaluated.

Plasmids Plasmid pCAcchbFGFcs23 was constructed by inserting a modified human bFGF cDNA with the secretory signal sequence of IL-2 into the pCAGGS expression vector [17]. The cDNA of modified human bFGF fused with the signal sequence of IL-2 was obtained from plasmid pTB1079 (donated by Dr. K. Igarashi, Biotechnology Research Laboratories, Takeda Chemical Industries, Osaka, Japan) [16]. The control plasmid for the animal experiment was pCAZ3, which contains Escherichia coli LacZ gene in the same expression plasmid. pCAccluc⫹ containing the recombinant luciferase gene (luc⫹) was used to quantify gene transfer efficiency. Plasmids were grown in E. coli JM109 and prepared using Qiagen EndoFree Mega Kits (Qiagen, Hilden, Germany).

In Vitro Study The plasmid pCAcchbFGFcs23 (15 ␮g) was transfected into 1.5 ⫻ 10 6 cultured rabbit fibroblasts using the modified calcium phosphate method followed by glycerol shock [18]. After transfection, the fibroblasts were cultured in 4 ml Dulbecco’s modified Eagle’s minimum essential medium (DMEM, Life Technologies, Inc., Grand Island, NY, USA) containing 10% FBS. The culture medium was changed daily and harvested at 4 days after transfection. The culture medium of nontransfected fibroblasts was used as the control. Each culture medium (50 ␮l) was subjected to Western blotting analysis using mouse monoclonal antibody against bovine bFGF (1:500, Upstate Biotechnology, Lake Placid, NY, USA) according to the protocol

We used a rabbit model of hind limb ischemia to evaluate in vivo angiogenic responses. To induce an ischemic state of the hind limb, male Japanese White rabbits weighing 3.0 –3.5 kg underwent complete excision of the left femoral artery as described previously [3]. Ten days after the operation, the left semimembranous muscle and adductor muscle were exposed, and 125 ␮g pCAZ3 or pCAcchbFGFcs23 suspended in 500 ␮l PBS was injected at two sites in each muscle; a total of 500 ␮g of the plasmid DNA was administered in each animal. After the injection, electric pulses of 75 V were delivered to the four muscle regions in the same manner as described above. As a control for animals with electroporation, another set of plasmid DNA-treated rabbits was not subjected to electroporation (Fig. 1). That is, we prepared four groups of animals, as follows: pCAZ3 injection without electroporation (LacZ-E ⫺ group, n ⫽ 7); pCAcchbFGFcs23 injection without electroporation (bFGF-E ⫺ group, n ⫽ 8); pCAZ3 injection with electroporation (LacZ-E ⫹ group, n ⫽ 7); and pCAcchbFGFcs23 injection with electroporation (bFGF-E ⫹ group, n ⫽ 8).

Calf Blood Pressure Ratio, Angiographic Score, Arterial Diameter, in Vivo Blood Flow, and Capillary Density The systolic pressure of the calf was measured in both hind limbs immediately before and 28 days after plasmid DNA injection, and the calf blood pressure ratio was calculated as the ratio of the left pressure to the right. Also, at 28 days after the injection, a 3-French end-hole catheter

NISHIKAGE ET AL.: GENE THERAPY BY IN VIVO ELECTROPORATION

FIG. 1.

39

Experimental protocol.

was introduced into the left internal iliac artery and selective left internal iliac arteriography was carried out. Then the angiographic score and the diameter of the left caudal gluteal artery were determined as described previously [3]. Before angiography, a 0.014-in. Doppler guide wire (EndoSonics, Rancho Cordva, CA, USA) was introduced through the catheter to the proximal part of the left internal iliac artery. Average peak velocity (APV) was measured at rest, and the maximum APV was quantified immediately after bolus intraarterial injection of 2 mg of papaverine (Dainippon Pharmaceutical, Osaka, Japan) was measured. Blood flow was calculated as described previously [3]. After the above evaluation, the rabbits were killed, and the left semimembranous muscle was resected. Transverse sections of the muscle tissue were prepared, and capillary density was determined as described previously [3].

Expression of bFGF in Vivo To assess the expression of bFGF protein, rabbits were killed at 1, 4, 7, 14, and 28 days after pCAcchbFGFcs23 injection with electroporation (n ⫽ 5 at each time point), and another group of normal rabbits (n ⫽ 5) was used as the control. Serum bFGF concentration was measured by ELISA, and bFGF level in the left adductor muscle was analyzed by Western blotting after concentration with heparinsepharose, as described previously [20].

Statistical Analysis Results were expressed as mean ⫾ SD. Statistical significance was evaluated using ANOVA followed by Dunnett’s test for comparisons

between the control group (Lac Z-E ⫺) and other groups. If two groups showed any significant difference from the control group, unpaired t test was used between the two groups. ANOVA was used for analysis of systemic bFGF level. All data were considered significant at P ⬍ 0.05.

RESULTS In Vitro Study

Western blot analysis showed that pCAcchbFGFcs23treated rabbit fibroblasts secreted two forms of bFGF (18 and 22 kDa) into the culture medium at 4 days after transfection (Fig. 2a). The DNA synthetic activity of the secreted bFGF was assessed by 3H-thymidine incorporation assay, and significant bFGF activity was demonstrated in cultured rabbit fibroblasts (Fig. 2b). Efficiency of in Vivo Gene Transfer by Electroporation

Low luciferase activity was detected in the muscles without electroporation, while in the samples with electroporation, gene transfer efficiency increased linearly according to the increase in voltage up to 100 V (Fig. 3a). At 150 V electroporation, luciferase activity was markedly lower than that at 100 V. Macromorphological analysis using the LacZ gene also showed abun-

40

JOURNAL OF SURGICAL RESEARCH: VOL. 120, NO. 1, JULY 2004

rounding the electrode (Fig. 4c, d, and e), and the area showing these findings expanded according to the increase in voltage. Further, 150 V electroporation induced massive fibrosis between the regenerative myocytes (Fig. 4f). Calf Blood Pressure Ratio, Angiographic Score, Arterial Diameter, in Vivo Blood Flow, and Capillary Density

Calf blood pressure ratio showed no significant difference between the LacZ-E ⫺, bFGF-E ⫺, LacZ-E ⫹, and bFGF-E ⫹ groups immediately before gene transfer. At 28 days after plasmid DNA injection (pCAZ3 or pCAcchbFGFcs23) with/without in vivo electroporation, the calf blood pressure ratio in the bFGF-E ⫹ group was significantly higher than that in the control group (LacZ-E ⫺), while there was no significant difference between the LacZ-E ⫺, bFGF-E ⫺, and LacZ-E ⫹ groups (Fig. 5). Angiographic score reflects the development of relatively large-diameter collateral arteries. The scores in the bFGF-E ⫺ and bFGF-E ⫹ groups at 28 days after plasmid DNA injection were significantly higher than that in the LacZ-E ⫺ group, but no significant difference in the score was detected between the bFGF-E ⫹ and bFGF-E ⫺ groups (Fig. 6a). Regarding blood flow at rest, maximum blood flow, capillary density, and diameter of the left caudal gluteal artery at 28 days after plasmid DNA administration, only the values in the bFGF-E ⫹ group were significantly greater than those in the control group (LacZ-E ⫺) (Fig. 6b– e). FIG. 2. (a) Western blotting probed with anti-bFGF antibody shows bFGF expression in culture medium. Note that pCAcchbFGFcs23 (expression plasmid vector containing modified bFGF gene)treated rabbit fibroblasts secreted bFGF protein in the culture medium. (b) 3H-thymidine incorporation assay shows the DNA synthetic activity of secreted bFGF. Values are shown as mean ⫾ SD. *P ⬍ 0.05. PC, recombinant human bFGF (Upstate Biotechnology) used as positive control; C, control; T, medium of pCAcchbFGFcs23treated cells.

dant LacZ-positive myofibers after 75 V electroporation, while scarce LacZ-positive myofibers were observed without electroporation (Fig. 3b and c). Muscle Damage Caused by in Vivo Electroporation

Histological analysis on day 7 after electroporation showed various findings indicating damage of the muscle tissue. The main features of electrical damage were necrosis, degeneration and regeneration of muscle cells, inflammatory cell infiltration, and fibrosis in the connective tissue. All samples subjected to electroporation showed these findings in the muscle region where the electrodes had been inserted (Fig. 4b–f). In the samples with 50, 75, and 100 V electroporation, muscle degeneration and regeneration with inflammatory infiltration were also observed in the region sur-

Systemic bFGF Level

Statistical analysis detected no significant increase in systemic bFGF level during the time course after injection of pCAcchbFGF with electroporation (Fig. 7). Local Expression of bFGF Protein

The expression of bFGF in muscle was increased from 1 day after gene transfer, and abundant bFGF was detected on day 4 and day 7 (Fig. 8). The protein expression decreased, although the protein level of bFGF on days 14 and 28 was slightly higher than that in the control muscle sample. DISCUSSION

The most important finding in the present study was that only rabbits in the bFGF-E ⫹ group showed significantly augmented development of collateral vessels, while few parameters indicated functional collateral development in the bFGF-E ⫺ group. There are some possible mechanisms to explain the additive effect of application of in vivo electroporation. One explanation is that in vivo electroporation improved gene transfer

NISHIKAGE ET AL.: GENE THERAPY BY IN VIVO ELECTROPORATION

41

FIG. 3. Luciferase assay shows gene transfer efficiency in semimembranous muscles subjected to injection of pCAccluc⫹ (expression plasmid vector containing luciferase gene) and electroporation with various voltages (a). Photographs of semimembranous muscles from animals treated by sole injection of pCAZ3 (expression plasmid vector containing LacZ gene) (b), and with 75 V electroporation (c). Muscle tissue was stained with X-gal. The LacZ-positive area is stained blue. Values are shown as mean ⫾ SD. Bar, 1 mm.

efficiency to muscle cells and then increased the expression of bFGF protein. Although the efficacy of gene transfer was low, sole injection of naked plasmid DNA into muscle tissue reliably promoted gene transfer to muscle cells, suggesting that a certain level of bFGF was secreted in the ischemic limb tissues in the bFGF-E ⫺ group. The negative data in the bFGF-E ⫺ group, therefore, indicate that the level of bFGF secretion was lower than the requirement for significant

angiogenic effects in ischemic muscle. In contrast, since analysis of luciferase activity in the present study revealed that expression of the gene product after 75 V electroporation was approximately 520 times higher than that without electroporation, bFGF secretion in the animals in the bFGF-E ⫹ group is presumed to be markedly greater than that in the bFGF-E ⫺ group, and the net secretion of bFGF exceeded the requirement for a significant angiogenic effect. Recently, Masaki et al.

42

JOURNAL OF SURGICAL RESEARCH: VOL. 120, NO. 1, JULY 2004

FIG. 4. Microphotographs of semimembranous muscle between the electrodes at 7 days after electroporation with 0 V (a), 25 V (b), 50 V (c), 75 V (d), 100 V (e), and 150 V (f). Sections were stained with hematoxylin and eosin. E, electrode-positioned area. Bar, 50 ␮m.

presented a new method of angiogenic gene therapy using recombinant Sendai virus vector containing the bFGF gene in a mouse model of limb ischemia [21]. They reported in their study that the gene expression level after sole injection of plasmid DNA was below the therapeutic level. This finding indicates that bFGF expression achieved by sole plasmid DNA injection was far below the level required to induce significant collateral development in ischemic muscle tissue, which might support our speculation. Another possible expla-

nation for the effects of electroporation is a direct angiogenic effect of electrical stimulation. Previous study showed that electric pulses stimulated skeletal muscle cells to increase blood flow in muscle tissue in a rat model of hind limb ischemia [22]. The electrical stimulation performed in this study, however, was lowvoltage, high-frequency, and continuous as compared with that in the present study, indicating that electroporation in this study was quite different from the above therapeutic electrical stimulation. Further,

NISHIKAGE ET AL.: GENE THERAPY BY IN VIVO ELECTROPORATION

43

FIG. 5. Calf blood pressure ratio measured immediately before and 28 days after plasmid injection (pCAZ3 or pCAcchbFGFcs23) with/without 75 V electroporation. Values are shown as mean ⫾ SD. *P ⬍ 0.05.

since no significant effects were detected in the animals treated with electroporation and the control plasmid (LacZ-E ⫺ group), it might be possible to exclude direct angiogenic effects caused by electroporation in the present study. Although sole administration of plasmid DNA encoding the bFGF gene induced weak angiogenic effects as described above, previous studies demonstrated that injection of plasmid DNA encoding other angiogenic growth factors promoted significant development of collateral vessels [3, 5]. Isner and co-workers especially demonstrated in their studies using the VEGF gene that intramuscular injection of naked plasmid DNA achieved favorable results in animal experiments and clinical trials [3, 7]. These findings suggested that the effectiveness of plasmid-based gene transfer varies according to the transferred gene, and that one critical factor regulating the effectiveness might be a minimum requirement for an exogenous growth factor to induce a significant angiogenic response. It might be difficult to accurately determine the minimum requirement of a growth factor, although previous studies have provided some hints for speculation on this issue [14, 21, 23]. Arras et al. established a local delivery system for bFGF protein using control-release microspheres, applied this system to pig myocardium, and detected a significant response of cell proliferation [23]. They calculated that each gram of targeted tissue was exposed to a daily dose of about 17 ng bFGF. In other animal experiments, two studies presented data indi-

cating the bFGF concentration required for significant collateral development in ischemic limb muscle and heart, which was 12 and 4.4 ng/g of muscle, respectively [14, 21]. Thus, bFGF concentration on the order of nanograms per gram was suggested to be necessary for a therapeutic angiogenic response in ischemic muscle. On the contrary, a recent study demonstrated in a mouse ischemic limb model that plasmid-based gene transfer of VEGF achieved a concentration of 102 pg/g of muscle tissue, and the VEGF expression promoted a mild limb salvaging effect, showing that a VEGF concentration on the order of picograms per gram was appropriate for an angiogenic effect [21]. Therefore, it is possible that higher expression is needed to achieve an effect as compared with VEGF, and this possibility could explain the discrepancy in effectiveness between bFGF and VEGF. The present study showed some distinct features of the in vivo electroporation method for gene therapy. One was that the expression level of gene product correlated linearly with the voltage of electrical pulses up to 100 V. A similar finding of a linear correlation was reported in a previous study that transferred another gene [13]. This property indicates that the expression level of the gene product could be predictably regulated within a certain range by controlling the voltage of electroporation, which might ensure appropriate therapeutic effects. Further, increasing the electrode voltage could possibly reduce the injection volume of plasmid DNA. Since bacterial DNA has an

44

JOURNAL OF SURGICAL RESEARCH: VOL. 120, NO. 1, JULY 2004

FIG. 6. Angiographic score analyzed from selective angiograms of left internal iliac artery (a). In vivo blood flow of the left internal iliac artery at rest (b) and after injection of papaverine (maximum blood flow) (c). Capillary density of the left semimembranous muscle was calculated in tissue sections stained by indoxy-tetrazolium method (d). The diameter of the left caudal gluteal artery was measured (e). All data were measured 28 days after plasmid injection (pCAZ3 or pCAcchbFGFcs23) with/without 75 V electroporation. Values are shown as mean ⫾ SD. *P ⬍ 0.05.

expected frequency of CpG motifs, injection of plasmid DNA grown in E. coli potently activates innate immune cells of the host, and this immune response occasionally results in severe complications [24]. Reducing the dose of plasmid DNA by electroporation might decrease the risk of a host innate immune reaction and improve the safety of gene therapy. There might be some possible disadvantage using electrical stimulation in angiogenic gene therapy. One is that electrical stimulation must be administered under anes-

thesia. However, since some other techniques such as cardioversion or electroconvulsive therapy are in clinical use, in vivo electroporation would be tolerated in a clinical setting. Electrical damage was also an important feature of in vivo electroporation. Histological study revealed inflammatory and regenerative changes in the treated muscle, but these changes were detected in a limited area just around the electrodes when treated with 75 V electroporation or lower. So it is highly possible that it would not cause any functional disorders or chronic

NISHIKAGE ET AL.: GENE THERAPY BY IN VIVO ELECTROPORATION

45

FIG. 7. ELISA-quantified time course of serum bFGF concentration after injection of pCAcchbFGFcs23 with 75 V electroporation. No significant change was detected throughout the time course. Values are shown as mean ⫾ SD. C, control serum from normal rabbit; d, day(s) after plasmid injection and electroporation.

pain. Severe damage of muscle cells also results in a decrease in gene production. In the present study, luciferase activity at 7 days after gene transfer showed that the maximum expression of gene product was detected in the samples with 100 V electroporation, and the gene product at 150 V electroporation was less than one-half the maximum expression. Further, histological analysis revealed massive fibrosis between the myofibers in muscle samples after 150 V electroporation, which was not detected in the samples with an electrode voltage of 100 V and lower. Focusing on the gene transfer efficacy, 100 V was the optimum voltage in the present study. However, care must be taken regarding the chronic effects induced by electroporation. There is a possibility that some cell damage may appear after day 7 and then re-

duce gene production in the late phase. Indeed, a previous study using mouse muscle demonstrated that 80 V electroporation promoted maximum gene expression at 3 weeks after gene transfer, whereas maximum expression on day 5 was achieved with 100 V [13]. We have to address the safety of the present protocol of gene transfer. Although bFGF has weak oncogenic activity, there is a risk of accelerating the growth of preexisting tumor or occult cancer [25]. Further, excessive bFGF possibly worsens diabetic retinopathy [26]. The influences on other tissues and organs might depend upon the systemic bFGF level. The present study analyzed the time course of serum bFGF concentration and showed no significant increase in bFGF level through all time points, suggesting a low possibility of

FIG. 8. Time course of bFGF expression in left adductor muscle after injection of pCAcchbFGFcs23 with 75 V electroporation. To concentrate bFGF, each muscle lysate was treated with heparin-sepharose and was then analyzed by Western blotting using anti-bFGF antibody. PC, recombinant human bFGF used as positive control; C, control sample; d, day(s) after plasmid injection and electroporation.

46

JOURNAL OF SURGICAL RESEARCH: VOL. 120, NO. 1, JULY 2004

influences on other tissues and organs. bFGF is known to bind with heparan sulfate proteoglycans of the extracellular matrix [27]. Then, secreted bFGF is captured in the gene-transduced muscle tissue, and a small amount of bFGF might flow to outside of the muscle. Because of the nature of bFGF, we believe that bFGF is a suitable growth factor for safe local delivery. In summary, the present study demonstrated the combination procedure of plasmid-based gene transfer and in vivo electroporation for angiogenic gene therapy. We injected an expression plasmid DNA encoding bFGF fused with a signal sequence into the ischemic muscles of the rabbit hind limb and delivered six pulses of electroporation with 75 V to the same sites. Gene transfer efficiency after 75 V electroporation was markedly greater than that without electroporation. In muscle tissue of the ischemic limb, bFGF protein was highly expressed until 21 days after gene transfer, and on day 28, rabbits treated with plasmid injection and electroporation showed significantly greater development of collateral vessels as compared to those with sole plasmid injection and other control animals. ACKNOWLEDGMENTS

9.

10.

11. 12.

13. 14.

15.

16.

This study was supported by a Grant-in Aid for Scientific Research (B) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan (13557098). We thank Dr. Yasuo Ohashi (Department of Biostatistics, School of Health Science and Nursing, the University of Tokyo) for statistical advice.

17.

REFERENCES

19.

1.

2.

3.

4.

5.

6. 7.

8.

Bauters, C., Asahara, T., Zheng, L. P., et al. Physiological assessment of augmented vascularity induced by VEGF in ischemic rabbit hindlimb. Am. J. Physiol. 267: H1263, 1994. Ohara, N., Koyama, H., Miyata, T., et al. Adenovirus-mediated ex vivo gene transfer of basic fibroblast growth factor promotes collateral development in a rabbit model of hind limb ischemia. Gene Ther. 8: 837, 2001. Tsurumi, Y., Takeshita, S., Chen, D., et al. Direct intramuscular gene transfer of naked DNA encoding vascular endothelial growth factor augments collateral development and tissue perfusion. Circulation 94: 3281, 1996. Takeshita, S., Weir, L., Chen, D., et al. Therapeutic angiogenesis following arterial gene transfer of vascular endothelial growth factor in a rabbit model of hindlimb ischemia. Biochem. Biophys. Res. Commun. 227: 628, 1996. Taniyama, Y., Morishita, R., Aoki, M., et al. Therapeutic angiogenesis induced by human hepatocyte growth factor gene in rat and rabbit hindlimb ischemia models: Preclinical study for treatment of peripheral arterial disease. Gene Ther. 8: 181, 2001. Wolff, J. A., Malone, R. W., Williams, P., et al. Direct gene transfer into mouse muscle in vivo. Science 247: 1465, 1990. Baumgartner, I., Pieczek, A., Manor, O., et al. Constitutive expression of phVEGF165 after intramuscular gene transfer promotes collateral vessel development in patients with critical limb ischemia. Circulation 97: 1114, 1998. Losordo, D. W., Vale, P. R., Symes, J. F., et al. Gene therapy for

18.

20.

21.

22.

23.

24. 25.

26.

27.

myocardial angiogenesis: Initial clinical results with direct myocardial injection of phVEGF165 as sole therapy for myocardial ischemia. Circulation 98: 2800, 1998. Baum, C., Forster, P., Hegewisch-Becker, S., and Harbers, K. An optimized electroporation protocol applicable to a wide range of cell lines. Biotechniques 17: 1058, 1994. Muramatsu, T., Mizutani, Y., Ohmori, Y., and Okumura, J. Comparison of three nonviral transfection methods for foreign gene expression in early chicken embryos in ovo. Biochem. Biophys. Res. Commun. 230: 376, 1997. Heller, R., Jaroszeski, M., Atkin, A., et al. In vivo gene electroinjection and expression in rat liver. FEBS Lett. 389: 225, 1996. Titomirov, A. V., Sukharev, S., and Kistanova, E. In vivo electroporation and stable transformation of skin cells of newborn mice by plasmid DNA. Biochim. Biophys. Acta 1088: 131, 1991. Aihara, H., and Miyazaki, J. Gene transfer into muscle by electroporation in vivo. Nat. Biotechnol. 16: 867, 1998. Unger, E. F., Banai, S., Shou, M., et al. Basic fibroblast growth factor enhances myocardial collateral flow in a canine model. Am. J. Physiol. 266: H1588, 1994. Baffour, R., Berman, J., Garb, J. L., Rhee, S. W., Kaufman, J., and Friedmann, P. Enhanced angiogenesis and growth of collaterals by in vivo administration of recombinant basic fibroblast growth factor in a rabbit model of acute lower limb ischemia: Dose-response effect of basic fibroblast growth factor. J. Vasc. Surg. 16: 181, 1992. Sasada, R., Seno, M., Watanabe, T., and Igarashi, K. Expression of modified bFGF cDNAs in mammalian cells. Ann. NY Acad. Sci. 638: 149, 1991. Niwa, H., Yamamura, K., and Miyazaki, J. Efficient selection for high-expression transfectants with a novel eukaryotic vector. Gene 108: 193, 1991. Parker, B. A., and Stark, G. R. Regulation of simian virus 40 transcription: Sensitive analysis of the RNA species present early in infections by virus or viral DNA. J. Virol. 31: 360, 1979. Koyama, H., and Reidy, M. A. Reinjury of arterial lesions induces intimal smooth muscle cell replication that is not controlled by fibroblast growth factor 2. Circ. Res. 80: 408, 1997. Kardami, E., and Fandrich, R. R. Basic fibroblast growth factor in atria and ventricles of the vertebrate heart. J. Cell Biol. 109: 1865, 1989. Masaki, I., Yonemitsu, Y., Yamashita, A., et al. Angiogenic gene therapy for experimental critical limb ischemia: Acceleration of limb loss by overexpression of vascular endothelial growth factor 165 but not of fibroblast growth factor-2. Circ. Res. 90: 966, 2002. Kanno, S., Oda, N., Abe, M., et al. Establishment of a simple and practical procedure applicable to therapeutic angiogenesis. Circulation 99: 2682, 1999. Arras, M., Mollnau, H., Strasser, R., et al. The delivery of angiogenic factors to the heart by microsphere therapy. Nat. Biotechnol. 16: 159, 1998. Hemmi, H., Takeuchi, O., Kawai, T., et al. A Toll-like receptor recognizes bacterial DNA. Nature 408: 740, 2000. Gross, J. L., Herblin, W. F., Dusak, B. A., et al. Effects of modulation of basic fibroblast growth factor on tumor growth in vivo. J. Natl. Cancer Inst. 85: 121, 1993. Bensaid, M., Malecaze, F., Prats, H., Bayard, F., and Tauber, J. P. Autocrine regulation of bovine retinal capillary endothelial cell (BREC) proliferation by BREC-derived basic fibroblast growth factor. Exp. Eye Res. 48: 801, 1989. Friesel, R. E., and Maciag, T. Molecular mechanisms of angiogenesis: Fibroblast growth factor signal transduction. FASEB J. 9: 919, 1995.