- Email: [email protected]
Transcutaneous Ultrasound Augments Naked DNA Transfection of Skeletal Muscle
ARTICLE
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
Transcutaneous Ultrasound Augments Naked DNA Transfection of Skeletal Muscle Peter Schratzberger,1 Joseph G. Krainin,1 Gabriele Schratzberger,1 Marcy Silver,1 Hong Ma,1 Marianne Kearney,1 Robert F. Zuk,2 Axel F. Brisken,2 Douglas W. Losordo,1,3,* and Jeffrey M. Isner1,3 1
Department of Cardiovascular Research, St. Elizabeth’s Medical Center, Tufts University School of Medicine, Boston, Massachusetts 02135, USA 2 PharmaSonics, Inc., Sunnyvale, California 94539, USA 3 Department of Vascular Medicine, St. Elizabeth’s Medical Center, Tufts University School of Medicine, Boston, Massachusetts 02135, USA *To whom correspondence and reprint requests should be addressed. Fax: (617) 779-6362. E-mail: [email protected].
This study was designed to test the hypothesis that transcutaneous ultrasound (US) exposure may augment the transfection efficiency and biological outcome associated with nonviral DNA gene transfer. Hindlimb muscles of New Zealand White rabbits were transfected with the reporter plasmid pCMV-, with or without US exposure. Optimization studies employed US exposure at various frequencies, mechanical indices, duty cycles, durations of exposure, and exposure time points. Based on these results, we explored the effect of US exposure on nonviral gene transfer of vascular endothelial growth factor (VEGF, phVEGF165) to promote neovascularization of ischemic hindlimbs. Ultrasound at 1 MHz, 100 W/cm2, 6% duty cycle, and 5 minutes exposure time, applied immediately following DNA injection, was found to be the most effective among the settings tested, increasing -galactosidase expression ~ 20 fold. Compared with US exposure alone, or phVEGF165 only, phVEGF165 + US exposure yielded a statistically significant improvement in revascularization, as determined by calf blood pressure ratio, angiographic score, intravascular Doppler blood flow, and capillary/myocyte ratio. These data demonstrate that ultrasound, when applied directly after intramuscular gene transfer, significantly increases transfection efficiency in vivo. The biological significance of this finding was confirmed by augmented limb perfusion in response to US exposure and naked VEGF DNA. Key Words: ultrasonics, peripheral vascular disease, gene therapy, ischemia, angiogenesis
INTRODUCTION Nonviral strategies of gene transfer have been investigated for selected disorders [1–4], in particular for therapies in which the gene product encodes a secreted protein, and in which lifelong expression of the transgene is not a prerequisite. The low transfection efficiency and limited duration of gene expression typical of nonviral gene transfer have not precluded demonstration of potentially meaningful biological outcomes [5–8]. The principal advantage of nonviral gene transfer is that it obviates concerns relating to potentially adverse consequences of viral vectors [9,10]. To preserve this advantage and at the same time optimize the efficiency of gene transfer, modifications in vector design [11,12] as well as the mode of delivery have been investigated. The use of ballistic technology, known as the “gene gun” [13], resulted in a moderate enhancement of gene transfer to skin [14] and tumors [15]. The principal limitation of this approach was poor penetration into deep tissues. Gene transfer using cationic lipids [16,17] has been widely
576
employed for in vitro gene transfer and continues to be investigated for clinical gene transfer [18]. Neutral polymers have been reported to enhance plasmid delivery to rat muscle by approximately fivefold [19]. More recently, two reports have described the successful use of electric pulses for enhanced transfer of plasmid DNA [20,21]. In addition to its well-known use as a diagnostic and therapeutic tool, ultrasound (US) has been investigated as a means to enhance gene transfer [22–24]. US exposure has been shown to permeabilize plasma membranes, hence encouraging DNA entry into cells [25,26]. Lipofection agents support DNA release from endosomes through a physicochemical transition that is known to be accelerated by US [27,28]. Based on these observations, Lawrie et al. conducted in vitro studies [29] demonstrating that adjunctive US exposure is associated with enhanced transgene expression. We report here the characteristics and mechanisms of highly efficient and reproducible reporter gene transfer
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy 1525-0016/02 $35.00
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
A
B
to skeletal muscle achieved in vivo by the local adjunctive application of US exposure following direct intramuscular (i.m.) injection of a plasmid DNA reporter gene. The biological relevance of this finding was confirmed by augmented revascularization in an ischemic hindlimb model in response to US and naked vascular endothelial growth factor (VEGF) DNA.
RESULTS Effects of Ultrasound Parameters on Reporter Gene Expression To investigate the effects of ultrasound parameters on reporter gene expression, a -galactosidase gene-encoding plasmid (pCMV) was administered by direct i.m. injection into the quadriceps muscles of New Zealand White (NZW) rabbits. Upon completion of i.m. gene transfer, transcutaneous US was carried out directly over the site of injection and surrounding tissue defined by a diameter of ~ 5 cm; the latter was accomplished by changing the position of the US transducer at 60-second intervals. Total US administration time was 5 minutes per transfection site. Transfection efficiency of pCMV without US exposure, based upon measurement of tissue activity of β-galactosidase, was 6.3 ± 1.7 relative light units (RLU)/mg. In the first set of experiments, an US frequency of 1 MHz was used. A mechanical index (MI) of 2.0 and a duty cycle (DC) of 1.5% yielded a 9.1-fold increase in transgene expression to 57.9 ± 8.9 RLU/mg (P < 0.01 versus no US exposure). A reduced MI of 0.5 with an increased DC of 25% yielded similar results (69.9 ± 5.6 RLU/mg), consistent with a reciprocal relationship between these two parameters. When the MI was maintained
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy
ARTICLE
FIG. 1. Effects of ultrasound parameters on reporter gene expression. (A) After 100 g of the plasmid pCMV was injected into the upper thigh muscles of New Zealand White rabbits, the injection site was exposed for 5 minutes to ultrasound at the conditions indicated. Ultrasound frequency was 1 MHz; MI, mechanical index; DC, duty cycle. -Galactosidase expression was analyzed 5 days later by a chemoluminescent reporter gene assay. Results in relative light units (RLU) were corrected for total protein in the tissue samples. Bars represent the mean ± SEM; n = 10 per group. *P < 0.05; **P < 0.01. (B) Histogram illustrating the fold increases in transfection efficiency after exposure at different ultrasound conditions.
in a low range (1.8), and the DC extended up to 6%, transgene expression was 20.4-fold higher in comparison with no US exposure (129.9 ± 16.8 RLU/mg, P < 0.01 versus no US exposure; n = 10 per group; Figs. 1A and 1B). Further increases in either MI or DC did not result in higher transfection efficiency (data not shown). To investigate the temporal relationship between US exposure and gene transfer, experiments were conducted in which the transfection site was pretreated with US under optimized conditions (1.8 MI, 6% DC) for 5 minutes preceding gene transfer. US exposure also augmented transgene expression under this condition (72.2 ± 11.4 RLU/mg, P < 0.01 versus no US exposure); however, the increase in transgene expression of ~ 11-fold was less than that achieved when US was applied after gene transfer (Fig. 2). Finally, we assessed the effects of US wavelength on gene transfer. US at a lower frequency of 300 kHz and 1.8 MI, 6% DC resulted in increased transgene expression (120.6 ± 15.1 RLU/mg) that did not differ significantly
FIG. 2. Characteristics of ultrasound (US)-mediated transfer of a reporter gene. A 100-g aliquot of the plasmid pCMV was injected (TX) into the upper thigh muscles of New Zealand White rabbits. Ultrasound was applied for 5 minutes either before (pre) or after (post) TX. The mechanical index was 1.8, the duty cycle was 6%, and the frequencies were either 1 MHz or 300 kHz. -Galactosidase expression was analyzed 5 days later by a chemiluminescent reporter gene assay. Results in relative light units (RLU) were corrected for total protein in the tissue samples. Bars represent the mean ± SEM; n = 10 per group. *P < 0.05; **P < 0.01.
577
ARTICLE
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
A
FIG. 3. Morphometric analysis of -galactosidase expression. Animals were transfected with pGSV plasmid encoding nuclear-localizing -galactosidase. Shown are representative sections of muscle stained with X-gal, counterstained with eosin, 5 days after transfection. In both sections, animals were transfected with the plasmid, but in (A) the tissue was not exposed to ultrasound following gene transfer, resulting in fewer transfected nuclei than in (B), where tissue was exposed to ultrasound. Magnification, ⫻100. White-boxed insets in (A) and (B) show ⫻200 magnifications of selected areas.
shown). Moreover, we also monitored tissue temperature at 4 mm depth within the sonicated muscle tissue for temperature increases that might be caused by US exposure. We observed no substantial temperature changes.
B
from that observed with 1 MHz (Fig. 2). Because the design of the 1-MHz US probes allowed for the generation of wide US beams with a long field of view, all subsequent experiments were done using US at the conditions of 1 MHz, an MI of 1.8, and a DC of 6%. Morphometric analysis of myocyte nuclei staining positive for -galactosidase expression after gene transfer with the nuclear-targeted plasmid pGSVLacZ showed that US exposure, when applied after gene transfer, increased the number of positive nuclei from 9 ± 4 (without US exposure) of total nuclei to 68 ± 11 (1 MHz, 1.8 MI, 6% DC, 5 minutes; P < 0.01 versus. no US exposure) of total nuclei per low-power field (Fig. 3). Evaluation of Ultrasound Biosafety To investigate whether US (1 MHz, 1.8 MI, 6% DC, 5 minutes) interfered with tissue integrity in our in vivo experiments, we conducted light-microscopic analyses of muscle samples harvested from treatment sites, and we found that no alteration in tissue integrity occurred (data not
578
Effects of Ultrasound Exposure on Therapeutic Gene Expression Based upon results of the reporter gene transfer experiments, we studied the biological relevance of US-mediated gene transfer using the plasmid encoding for human VEGF165 (phVEGF165) to promote therapeutic angiogenesis in a rabbit model of hindlimb ischemia. Previous studies using this animal model have demonstrated that an i.m. dose of 500 g of phVEGF165 is sufficient in young (~ 6 months old) NZW rabbits to induce a level of therapeutic angiogenesis that constitutes a near-optimal response [5]. In the current study, we used old (5–6 years) NZW rabbits. The rationale for this approach is that angiogenesis is impaired in old animals [30]; consequently, a greater dynamic range in the old animals permits comparison of strategies such as US exposure designed to augment gene transfer and thereby facilitate hindlimb neovascularization. Anatomic Assessment Representative angiograms recorded from the four treatment groups (saline injection only = control; saline injection + US exposure; phVEGF165 transfection only; phVEGF165 transfection + US exposure; n = 5 per group) at day 30 after treatment are shown in Fig. 4. In control animals, collateral development in the medial thigh typically appeared unchanged in serial angiograms recorded at days 0 (data not shown) and 30 (Fig. 4A). When US was applied after the injection of saline, no differences in collateral growth could be detected as compared with the control group (Fig. 4B). In contrast, in the phVEGF165transfected group, marked progression of collateral artery development was observed in comparison with the control animals (Fig. 4C). When VEGF gene transfer was followed by US exposure, however, collateral development was even more abundant (Fig. 4D). Quantitative analysis of collateral vessel development in the medial thigh was done by calculating an angiographic score (Fig. 5). At day 30, the angiographic score for rabbits undergoing US exposure alone (not transfected with naked VEGF DNA) was similar to that recorded for control animals (saline injection only, 0.47 ± 0.02 versus 0.48 ± 0.02,
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy
ARTICLE
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
FIG. 4. Representative images of selective internal iliac angiography at day 30 after treatment. (A) Control rabbit, saline injection, no ultrasound. (B) Saline-injected rabbit with US exposure shows no obvious increase in collateral development, as compared with (A). (C) This hindlimb was transfected with 500 g i.m. of phVEGF165 and shows extensive collateral artery formation. (D) When VEGF transfection was followed by US exposure, collateral formation was most abundant.
A
respectively, P = not significant). The phVEGF165-transfected group achieved a significantly higher score (0.68 ± 0.04; P < 0.01 versus control and P < 0.01 versus US exposure only). Consistent with C results illustrated in Fig. 5, the highest angiographic score was recorded for the group in which phVEGF165 gene transfer was followed by US exposure (0.8 ± 0.01; P < 0.01 versus VEGF alone). A favorable effect of US exposure on gene transfer–induced revascularization was also apparent at the capillary level (Fig. 6). The adductor muscle of ischemic limbs was histologically examined at day 30. Analysis of capillary/myocyte ratio was higher in the US-mediated phVEGF165 gene transfer group (0.77 ± 0.05) than in animals in which gene transfer was carried out by standard i.m. injection (0.59 ± 0.02; P < 0.01). The outcomes were superior for both treatment groups versus the control groups (0.39 ± 0.03 for saline only, P < 0.01 versus phVEGF165 only; and 0.43 ± 0.03 for saline + US exposure, P < 0.01 versus phVEGF165 only; n = 5 per group). Physiological Assessment Changes in the hemodynamic deficit in the ischemic limb after gene transfer were confirmed by measurement of the calf blood pressure (Fig. 7). By day 30 after gene transfer, blood pressure ratio for the phVEGF165-transfected group (0.74 ± 0.04) was significantly higher (P < 0.01) than that for either the saline-injected (0.49 ± 0.03) or US-only control groups (0.47 ± 0.03), respectively. In the rabbits that underwent US-mediated phVEGF165 gene transfer, an additional statistically significant increase in blood pressure ratio was achieved (0.89 ± 0.04; P < 0.05 versus VEGF only; n = 5 per group). Resting blood flow and papaverine-stimulated (maximum) blood flow in the ischemic limb were assessed by intravascular Doppler flow measurements (Fig. 8). Ultrasound exposure itself had no effect on either resting (13.7 ± 1.4 ml/minute for saline, 13.7 ± 1.4 ml/minute for saline + US exposure), or stimulated (22.8 ± 2.8 ml/minute for saline, 22.7 ± 1.5 ml/minute for
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy
B
D
saline + US exposure) blood flow. VEGF gene transfer significantly augmented blood flow to the ischemic limb through the internal iliac artery, both at rest (19.0 ± 0.8 ml/minute; P < 0.05 versus control) and following papaverine stimulation (30.8 ± 2.6 ml/minute; P < 0.05 versus control). Ultrasound-mediated VEGF gene transfer enhanced flow to the ischemic limb after papaverine stimulation (41.0 ± 2.8 ml/minute; P < 0.05 versus phVEGF165 only at maximum; n = 5 per group); although there was a trend toward higher resting flow in rabbits undergoing US-mediated phVEGF165 gene transfer (20.7 ± 0.9 ml/minute), this did not achieve statistical significance.
DISCUSSION Intramuscular injection of naked plasmid DNA is associated with little evidence of toxicity and/or toxicity and immunogenicity [31]. The disadvantage of this gene transfer strategy is relatively low transfection efficiency, particularly in comparison with viral vectors. Nevertheless, preclinical studies done in rabbit [5,32,33] and murine [30,34] models of hindlimb ischemia have demonstrated proof of concept that collateral vessel development may be induced by i.m. gene transfer of naked DNA encoding
579
ARTICLE
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
FIG. 5. Effects of i.m. phVEGF165 gene transfer and ultrasound exposure on revascularization in the ischemic hindlimb model. As indicated by angiographic score, saline-injected rabbits and rabbits treated with saline + ultrasound had significantly fewer collaterals than the phVEGF165-treated animals at day 30 after treatment. Those treated with i.m. phVEGF165 together with ultrasound had the most prominent angiogenic response. Bars show the mean ± SEM; n = 5 per group. *P < 0.05; **P < 0.01.
FIG. 6. Capillary/myocyte ratio as a measure for angiogenesis. Adductor muscles of ischemic limbs were histologically examined at day 30 to determine capillary/myocyte ratios. Each bar of the graph represents the mean ± SEM; n = 5 per group. *P < 0.01.
for VEGF. Ultimately, this has been reproduced in patients with critical limb ischemia [35–37], and more recently in patients with myocardial ischemia [38]. Numerous attempts have been made so far to overcome the relatively low transfection efficiency of naked DNA gene transfer, including the use of cationic lipids or polymers [16,17], or the application of electric pulses to DNA injection sites [20,21]. We report here a method for efficiently increasing naked DNA transfer in skeletal muscle, consisting of standard i.m. injection of naked plasmid DNA, followed by delivery of US by an external probe. Initial experiments employed the use of naked DNA encoding for a reporter gene (β-galactosidase), injected in a standardized manner into normal rabbit hindlimb muscle. Analysis of transgene expression, which was assessed by a chemiluminescent enzyme assay, demonstrated the anticipated low transfection efficiency, which was substantially and consistently increased when US was applied to the area surrounding the injection site. In this context, we found that US exposure applied to skeletal muscle after i.m. gene transfer was superior to sonication applied directly before gene transfer; this finding may be interpreted to indicate that US permeabilizes cell membranes for an increased uptake of naked plasmid DNA [23,39]. The US dose/response studies that we conducted in this in vivo assay revealed that US at a frequency of 1 MHz, 1.8 MI, 6% duty cycle, and 5 minutes exposure time yielded the highest -galactosidase expression from 100 g of pCMV. The applied energy of this
condition was higher than what others have used for cell culture transfection experiments [23,29]. To assess morphometrically the effects of US exposure on gene transfer, we used naked plasmid DNA encoding for nuclear-targeted GSVLacZ. Results from these experiments support the possibility that US exposure significantly increased the overall number of transfected skeletal myocyte nuclei; the methods used, however, do not allow assessment of the absolute amount of total transgene expression per cell. We monitored the safety of US exposure by light-microscopic analyses of tissue samples and close control of the core temperature of the target tissue. Considering that we detected no adverse effects of US exposure on the target tissue, our findings contrast with reports of several in vitro transfection techniques, which were associated with cell death [40]. Similarly, although the use of electrical pulses has been demonstrated to increase transfection efficiency [20,21], side effects included pain, passive involuntary muscle contractions, and tissue damage associated with thermal changes. Furthermore, when Mir et al. [21] applied electric pulses to rabbit rectus femoris muscle, they found > 50% variability in transgene expression, whereas USmediated gene transfer in our experiments showed a variability of ~ 7.5%. No evidence of muscle contractions or pain was seen with US-mediated gene transfer. To determine the biological relevance of the reporter gene experiments, we investigated US-mediated gene transfer for revascularization of ischemic hindlimbs. In previous reports, young (~ 6 months old) NZW rabbits
580
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy
ARTICLE
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
FIG. 7. Revascularization assessed by calf blood pressure ratio. Differences in hemodynamic outcomes among treatment groups were assessed by measuring calf blood pressure in ischemic and nonischemic limbs at day 30 after treatment. Results are shown as ratio of ischemic to nonischemic limbs. Ultrasound alone did not have an effect on calf blood pressure ratio, whereas i.m. phVEGF165 resulted in a significant increase. When ultrasound was applied for 5 minutes after the DNA injection, blood pressure ratio was significantly higher than in the group treated with VEGF only. Bars represent the mean ± SEM; n = 5 per group. *P < 0.05; **P < 0.01.
have been used to establish the concept of therapeutic angiogenesis with the i.m. injection of naked plasmid DNA encoding VEGF165 [5]. In these studies, VEGF gene transfer to ischemic hindlimbs was shown to improve revascularization parameters to almost normal levels in these animals. In the current experiments we therefore used old (~ 5–6 years old) NZW rabbits, in which angiogenesis has been shown to be naturally retarded [30]. We reasoned that this impaired response might provide a more dynamic range to assess whether the increase in transfection efficiency achieved with US-mediated gene transfer might yield superior revascularization compared with direct injection of naked plasmid DNA alone. Indeed, we found that US exposure yielded a statistically significant improvement in revascularization over direct injection alone, as determined by calf blood pressure ratio, angiographic score, intravascular Doppler blood flow, and capillary/myocyte ratio. It is interesting to note that direct injection of phVEGF165 alone induced significant revascularization of the ischemic hindlimb in these old rabbits, but to a lesser extent than found in young NZW rabbits. When phVEGF165 gene transfer was followed by US exposure, however, the improvement in hindlimb neovascularization was comparable to that seen with phVEGF165 alone in young NZW rabbits [5].
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy
FIG. 8. Intra-arterial blood flow measurements. Resting blood flow and papaverine-stimulated blood flow in the ischemic limb were measured at day 30 after treatment by intra-arterial Doppler guide wire measurements. Bars show the mean ± SEM; n = 5 per group. *P < 0.05; **P < 0.01.
Moreover, when ischemic hindlimbs of young NZW rabbits were injected with suboptimal doses of phVEGF165, with or without ultrasound (unpublished data), US treatment following i.m. injection was able to rescue an otherwise insufficient angiogenic response to low-dose phVEGF165 injection. It is important to note that in both young and old rabbits, US exposure itself did not augment therapeutic neovascularization when compared to saline-injected controls. This is consistent with results from the reporter gene transfection experiments, in which we found that US exposure alone did not increase expression of endogenous -galactosidase. Reher et al. [41] previously reported that US exposure of fibroblasts, osteoblasts, or mononuclear blood cells increased concentrations of VEGF protein in cell culture media. The principal tissue target in our in vivo experiments consisted of skeletal myocytes; even if US exposure did induce the release of endogenous VEGF protein, the absolute amount may have been too limited to induce revascularization in this animal model.
MATERIALS
AND
METHODS
Animal models. All protocols were approved by St. Elizabeth’s Institutional Animal Care and Use Committee. The effects of ultrasound on transfection efficiency and transgene expression were investigated in two rabbit models. In all experiments, investigators performing the in vivo and in vitro follow-up examinations were blinded to the identity of the treatment administered.
581
ARTICLE
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
Reporter gene expression in normal rabbit hindlimbs. For analysis of -galactosidase expression in skeletal muscle, 70 male NZW rabbits (Pine Acre Rabbitry, Norton, MA) (weight 3.5–4 kg) were anesthetized with a mixture of ketamine (50 mg/kg) and acepromazine (0.8 mg/kg) after premedication with xylazine (2 mg/kg). The quadriceps muscles of both thighs were exposed by skin incisions, and the sites of the plasmid injections into these muscles were each marked by single sutures. Per hindlimb, one injection of 100 g of pCMV in 500 l of saline was done, which was then followed by US exposure in a subgroup of animals. After completion of gene transfer and US exposure, the wounds were closed with 4-0 nylon sutures and the animals were kept alive for 5 days. At death, the single sutures facilitated identification of injection sites, following which muscle samples were then harvested, snap-frozen in liquid nitrogen, and stored at –80⬚C until further analysis. Ultrasound exposures. Wide-field ultrasonic transducers (PharmaSonics, Sunnyvale, CA) were placed in direct contact with the exposed muscles with acoustic coupling provided by ultrasound transmission gel (Ultraphonic; Pharmaceutical Innovations, Newark, NJ). Possible thermal changes within the target tissue were monitored by a temperature probe embedded at a depth of 4 mm in the muscle that was subject to sonication. Transducer drive signals were generated by a 16-MHz function/arbitrary waveform generator (33120A; Hewlett Packard, Palo Alto, CA), amplified by a custom power amplifier (PharmaSonics), and monitored by a 200-MHz digital oscilloscope (TDS360; Tektronix, Beaverton, OR). Transducers were calibrated in a water tank against a calibrated PVDF bilaminar membrane hydrophone (Series 805 with Series 806 preamp.; Sonic Industries, Hatboro, PA). Transducers were typically driven with 30-cycle bursts of 950 kHz ultrasound, at a burst repetition rate of 1890 Hz (6% DC). Drive amplitudes were adjusted to achieve the reported output MIs, where the MI is defined as the maximum negative peak pressure in MPa divided by the square root of the ultrasound frequency in MHz. Over the depth of field of interest for these experiments, the transducers typically displayed –6-dB beam profiles with beamwidths on the order of 9–10 mm. Total US exposure per DNA injection site was 5 minutes, during which the US probe was shifted in 1-cm increments once a minute to cover fully the spread of the injectate in the tissue by the US beam. Plasmids and DNA preparation. To study the effects of US on VEGF overexpression in hindlimb ischemia, we administered to rabbits a eukaryotic expression vector encoding the human VEGF165 gene [42] transcriptionally regulated by the cytomegalovirus (CMV) promoter/enhancer (phVEGF165) [5,37]. To study the influence of US exposure on reporter gene expression, pCMV (Clontech, Palo Alto, CA) was used as promotermatched reporter plasmid. This plasmid encodes -galactosidase under the control of a CMV promoter/enhancer. The plasmid pGSVLacZ (courtesy of Claire Bonnerot) containing a nuclear-targeted -galactosidase sequence coupled to the simian virus-40 (SV40) early promoter was used for the control transfection experiments. Preparation and purification of all plasmids from cultures of phVEGF165-, pCMV-, or pGSVLacZ-transformed Escherichia coli were carried out by the column method (Qiagen Mega Kit; Qiagen, Valencia, CA). Measurement of -galactosidase activity. Muscles were analyzed for -galactosidase expression using the Galacto-Light Plus chemoluminescent reporter gene assay system (Tropix, Bedford, MA) according to the manufacturer’s directions. Briefly, muscles were homogenized in lysis buffer (100 mM potassium phosphate, 0.2% Triton X-100) supplemented with protease inhibitors (aprotinin, 2 g/ml; leupeptin, 5 g/ml; and pepstatin, 2 g/ml, all from Sigma, St. Louis, MO) and dithiothreitol (1 mM, Sigma). After adding the Galacton Plus substrate, samples were incubated for 45 minutes at room temperature before light emission was read in a luminometer (LB 9507; Berthold, Germany). Total protein concentration per muscle sample was measured using the BCA protein assay kit (Pierce, Rockford, IL), according to the manufacturer’s protocol. All results are expressed as relative light units (RLU) per mg total protein. Endogenous -galactosidase produced a mean of 54.4 RLU/mg, as averaged from nontransfected quadriceps muscles of five normal NZW rabbits. All data from -galactosidase transfection experiments were therefore corrected for this background.
582
The efficiency of transgene expression after US exposure was also evaluated morphometrically in 10 additional rabbits following i.m. injection of 100 g of the plasmid pGSVLacZ, containing a nuclear-targeted -galactosidase sequence. This obviated nonspecific cytoplasmic background staining due to endogenous -galactosidase activity in nontransfected control muscles. The transfected muscles were harvested 5 days later, fixed in 1% paraformaldehyde, and incubated with X-Gal (5-bromo-4-chloro-3-indolyl-D-galactosidase chromogen; Sigma) overnight at 37⬚C [43]. Harvested tissues were then paraffin-embedded, sectioned, and counterstained with eosin. Five sections from each sample were randomly selected, and the numbers of positive nuclei in five low-power fields distributed among an area of skeletal muscle section measuring 10 mm in diameter were counted manually for each specimen by a blinded observer. To assess possible tissue damage caused by US exposure, muscle samples from select rabbits were stained with H&E, followed by light-microscopic analysis. Rabbit ischemic hindlimb model. To investigate the biological relevance of US-enhanced transfection efficiency, the rabbit ischemic hindlimb model was used in 32 old (5–6 years) animals according to previously published methods [30]. Briefly, after anesthesia and unilateral excision of the entire superficial femoral artery, rabbits were allowed to recover for 10 days. After reanesthetizing the rabbits as outlined above, the skin was opened and five i.m. injections, each containing 100 g of naked plasmid DNA encoding for human VEGF165 (phVEGF165) in 500 l, were administered to three major muscles of the thigh [5]. A subgroup of rabbits underwent US treatment immediately after VEGF DNA injections. Upon completion of treatment, the skin was closed with 4-0 nylon sutures. Blood pressure (BP) was measured as described [44]. The ratio of ischemic to normal hindlimb BP (BPR) was defined for each rabbit as the ratio of systolic pressure measured in the ischemic limb to systolic pressure measured in the normal limb. Selective angiography of the ischemic hindlimb was conducted on day 30 after surgery as described [32,33,44]. To assess collateral vessel development quantitatively, we used an acetate overlay with an imprinted grid composed of 2.5-mm-diameter circles arranged in rows spaced 5 mm apart to yield an angiographic score as described [44]. Blood flow to the ischemic hindlimb was quantified in vivo before selective internal iliac angiography on day 30 with a 0.018-in Doppler guide wire (Cardiometrics, Mountain View, CA) as described [5]. APV (time average of the spectral peak velocity) was recorded at rest, and maximum APV was obtained after bolus injection of 2 mg of papaverine. Dopplerderived flow was calculated as QD = (pd2/4) (0.5 ⫻ APV), where QD is Doppler-derived time-averaged flow, and d is vessel diameter, which was determined from an angiogram with an automated edge-detection system as described [45]. Tissue specimens obtained as transverse sections from the adductor and semimembranous muscle groups of both limbs of each rabbit at the time of sacrifice (day 30) were embedded in OCT compound (Miles, Elkhart, IN) and snap-frozen in liquid nitrogen. Tissue sections were stained for alkaline phosphatase by an indoxyl-tetrazolium method to detect capillary endothelial cells and were then counterstained with eosin. A total of 20 different fields from the two muscles were randomly selected, and capillary density was evaluated by a single observer blinded to the treatment group under a 20⫻ objective. Finally, the number of myofibers in each field were also counted to calculate the capillary/myocyte ratio. Statistics. All results are expressed as the mean ± SEM (m ± SEM). Statistical comparisons between groups were done by ANOVA, which was followed by an unpaired, two-tailed Student’s t-test when a significant difference was detected. P values < 0.05 were considered to denote statistical significance.
ACKNOWLEDGMENTS We thank Mickey Neely for assistance in the preparation of this manuscript. This study was supported in part by NIH grants HL53354, HL57516, HL60911, and HL63414, and by the Peter Lewis Foundation (Cleveland, OH). P.S. is the recipient of an Erwin Schroedinger fellowship grant, provided by the Austrian Funds for the Advancement of Science (FWF).
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
RECEIVED FOR PUBLICATION AUGUST 24, 2001; ACCEPTED SEPTEMBER 10, 2002.
REFERENCES 1. Blau, H. M., and Springer, M. L. (1995). Gene therapy—a novel form of drug delivery. N. Engl. J. Med. 333: 1204–1207. 2. Ledley, F. D. (1995). Nonviral gene therapy: the promise of genes as pharmaceutical products . Hum. Gene Ther. 6: 1129–1144. 3. Rolland, A. P. (1998). From genes to gene medicines: recent advances in nonviral gene delivery. Crit. Rev. Ther. Drug Carrier Syst. 15: 143–198. 4. Lee, R. J., and Huang, L. (1997). Lipid vector systems for gene transfer. Crit. Rev. Ther. Drug Carrier Syst. 14: 173–206. 5. Tsurumi, Y., et al. (1996). Direct intramuscular gene transfer of naked DNA encoding vascular endothelial growth factor augments collateral development and tissue perfusion. Circulation 94: 3281–3290. 6. Bauters, C., et al. (1994). Physiologic assessment of augmented vascularity induced by VEGF in ischemic rabbit hindlimb. Am. J. Physiol. 267: H1263–H1271. 7. Tabata, H., Silver, M., and Isner, J. M. (1997). Arterial gene transfer of acidic fibroblast growth factor for therapeutic angiogenesis in vivo: critical role of secretion signal in use of naked DNA. Cardiovasc. Res. 35: 470–479. 8. Alila, H., et al. (1997). Expression of biologically active human insulin-like growth factor-I following intramuscular injection of a formulated plasmid in rats. Hum. Gene Ther. 8: 1785–1795. 9. Balter, M. (2000). Gene therapy on trial. Science 288: 951–957. 10. Kielian, T., and Hickey, W. F. (2000). Inflammatory thoughts about glioma gene therapy. Nat. Med. 5: 1237–1238. 11. Hartikka, J., et al. (1996). An improved plasmid DNA expression vector for direct injection into skeletal muscle. Hum. Gene Ther. 7: 1205–1217. 12. Berthou-Soulie, L., et al. (2000). PCOR-mediated gene transfer in ischemic tissues: application of electrotransfer to peripheral ischemia (Abstract). Mol. Ther. 1: S209. 13. Leclerc, G., and Isner, J. M. (1993). Percutaneous gene therapy for cardiovascular disease. In Interventional Cardiology (E. Topol, Ed.), pp. 1019–1029. W. B. Saunders Company, Philadelphia, PA. 14. Fynan, E. F., et al. (1993). DNA vaccines: protective immunizations by parenteral, mucosal, and gene-gun inoculations. Proc. Natl. Acad. Sci. USA 90: 11478–11482. 15. Yang, N.-S., and Sun, W. H. (1995). Gene gun and other non-viral approaches for cancer gene therapy. Nat. Med. 1: 481–483. 16. Wolff, J. A., et al. (1992). Expression of naked plasmids by cultured myotubes and entry of plasmids into T tubules and caveolae of mammalian skeletal muscle. J. Cell Sci. 103: 1249–1259. 17. Schwartz, B., et al. (1996). Gene transfer by naked DNA into adult mouse brain. Gene Ther. 3: 405–411. 18. Yla-Herttuala, S., and Martin, J. F. (2000). Cardiovascular gene therapy. Lancet 355: 213–222. 19. Mumper, R. J., et al. (1998). Protective interactive noncondensing (PINC) polymers for enhanced plasmid distribution and expression in rat skeletal muscle. J. Controlled Release 52: 259–272. 20. Aihara, H., and Miyazaki, J.-I. (1998). Gene transfer into muscle by electroporation in vivo. Nat. Biotechnol. 16: 867–870. 21. Mir, L. M., et al. (1999). High-efficiency gene transfer into skeletal muscle mediated by electric pulses. Proc. Natl. Acad. Sci. USA 96: 4262–4267. 22. Fechheimer, M., et al. (1987). Transfection of mammalian cells with plasmid DNA by scrape loading and sonication loading. Proc. Natl. Acad. Sci. USA 84: 8463–8467. 23. Kim, H. J., Greenleaf, J. F., Kinnick, R. R., Bronk, J. T., and Bolander, M. E. (1996). Ultrasound-mediated transfection of mammalian cells. Hum. Gene Ther. 7: 1339–1346.
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy
ARTICLE
24. Zhang, L. J., et al. (1991). Efficient transformation of tobacco by ultrasonification. Biotechnology 9: 996–997. 25. Gambihler, S., Delius, M., and Ellwart, J. W. (1994). Permeabilization of the plasma membrane of L1210 mouse leukemia cells using lithotripter shock waves. J. Membr. Biol. 141: 267–275. 26. Pohl, P., Rosenfeld, E., and Millner, R. (1993). Effects of ultrasound on the steady-state transmembrane pH gradient and the permeability of acetic acid through bilayer lipid membranes. Biophys. Acta 1145: 279–283. 27. Gao, X., and Huang, L. (1995). Cationic liposome-mediated gene transfer. Gene Ther. 2: 710–722. 28. Tata, D. B., and Dunn, F. (1992). Interaction of ultrasound and model membrane systems: analyses and predictions. J. Phys. Chem. 96: 3548–3555. 29. Lawrie, A., et al. (1999). Ultrasound enhances reporter gene expression after transfection of vascular cells in vitro. Circulation 99: 2617–2620. 30. Rivard, A., et al. (1999). Age-dependent impairment of angiogenesis. Circulation 99: 111–120. 31. Wolff, J. A., Ludtke, J. J., Acsadi, G., Williams, P., and Jani, A. (1992). Long-term persistence of plasmid DNA and foreign gene expression in mouse muscle. Hum. Mol. Genet. 1: 363–369. 32. Takeshita, S., et al. (1996). Gene transfer of naked DNA encoding for three isoforms of vascular endothelial growth factor stimulates collateral development in vivo. Lab. Invest. 75: 487–502. 33. Takeshita, S., et al. (1995). Time course of increased cellular proliferation in collateral arteries following administration of vascular endothelial growth factor in a rabbit model of lower limb vascular insufficiency. Am. J. Pathol. 147: 1649–1660. 34. Murohara, T., et al. (1998). Nitric oxide synthase modulates angiogenesis in response to tissue ischemia. J. Clin. Invest. 101: 2567–2578. 35. Baumgartner, I., et al. (1998). Constitutive expression of phVEGF165 following intramuscular gene transfer promotes collateral vessel development in patients with critical limb ischemia. Circulation 97: 1114–1123. 36. Isner, J. M., et al. (1998). Treatment of thromboangiitis obliterans (Buerger’s disease) by intramuscular gene transfer of vascular endothelial growth factor: preliminary clinical results . J. Vasc. Surg. 28: 964–975. 37. Isner, J. M., et al. (1996). Clinical evidence of angiogenesis following arterial gene transfer of phVEGF165. Lancet 348: 370–374. 38. Vale, P. R., et al. (2000). Left ventricular electromechanical mapping to assess efficacy of phVEGF165 gene transfer for therapeutic angiogenesis in chronic myocardial ischemia. Circulation 102: 965–974. 39. Tata, D. B., Dunn, F., and Tindall, D. J. (1997). Selective clinical ultrasound signals mediate differential gene transfer and expression in two human prostate cancer cell lines: LnCap and PC-3. Biochem. Biophys. Res. Commun. 234: 64–67. 40. Bao, S., Thrall, B. D., and Miller, D. L. (1997). Transfection of a reporter plasmid into cultured cells by sonoporation in vitro. Ultrasound Med. Biol. 23: 953–959. 41. Reher, P., Doan, N., Bradnock, B., Meghji, S., and Harris, M. (1999). Effect of ultrasound on the production of IL-8, basic FGF and VEGF. Cytokine 11: 416–423. 42. Tischer, E., et al. (1991). The human gene for vascular endothelial growth factor: multiple protein forms are encoded through alternative exon splicing. J. Biol. Chem. 266: 11947–11954. 43. Feldman, L. J., et al. (1995). Low efficiency of percutaneous adenovirus-mediated arterial gene transfer in the atherosclerotic rabbit. J. Clin. Invest. 95: 2662–2671. 44. Takeshita, S., et al. (1994). Therapeutic angiogenesis: a single intra-arterial bolus of vascular endothelial growth factor augments revascularization in a rabbit ischemic hindlimb model. J. Clin. Invest. 93: 662–670. 45. Ku, D. D., Zaleski, J. K., Liu, S., and Brock, T. A. (1993). Vascular endothelial growth factor induces EDRF-dependent relaxation in coronary arteries. Am. J. Physiol. 265: H586–H592.
583
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
Transcutaneous Ultrasound Augments Naked DNA Transfection of Skeletal Muscle Peter Schratzberger,1 Joseph G. Krainin,1 Gabriele Schratzberger,1 Marcy Silver,1 Hong Ma,1 Marianne Kearney,1 Robert F. Zuk,2 Axel F. Brisken,2 Douglas W. Losordo,1,3,* and Jeffrey M. Isner1,3 1
Department of Cardiovascular Research, St. Elizabeth’s Medical Center, Tufts University School of Medicine, Boston, Massachusetts 02135, USA 2 PharmaSonics, Inc., Sunnyvale, California 94539, USA 3 Department of Vascular Medicine, St. Elizabeth’s Medical Center, Tufts University School of Medicine, Boston, Massachusetts 02135, USA *To whom correspondence and reprint requests should be addressed. Fax: (617) 779-6362. E-mail: [email protected].
This study was designed to test the hypothesis that transcutaneous ultrasound (US) exposure may augment the transfection efficiency and biological outcome associated with nonviral DNA gene transfer. Hindlimb muscles of New Zealand White rabbits were transfected with the reporter plasmid pCMV-, with or without US exposure. Optimization studies employed US exposure at various frequencies, mechanical indices, duty cycles, durations of exposure, and exposure time points. Based on these results, we explored the effect of US exposure on nonviral gene transfer of vascular endothelial growth factor (VEGF, phVEGF165) to promote neovascularization of ischemic hindlimbs. Ultrasound at 1 MHz, 100 W/cm2, 6% duty cycle, and 5 minutes exposure time, applied immediately following DNA injection, was found to be the most effective among the settings tested, increasing -galactosidase expression ~ 20 fold. Compared with US exposure alone, or phVEGF165 only, phVEGF165 + US exposure yielded a statistically significant improvement in revascularization, as determined by calf blood pressure ratio, angiographic score, intravascular Doppler blood flow, and capillary/myocyte ratio. These data demonstrate that ultrasound, when applied directly after intramuscular gene transfer, significantly increases transfection efficiency in vivo. The biological significance of this finding was confirmed by augmented limb perfusion in response to US exposure and naked VEGF DNA. Key Words: ultrasonics, peripheral vascular disease, gene therapy, ischemia, angiogenesis
INTRODUCTION Nonviral strategies of gene transfer have been investigated for selected disorders [1–4], in particular for therapies in which the gene product encodes a secreted protein, and in which lifelong expression of the transgene is not a prerequisite. The low transfection efficiency and limited duration of gene expression typical of nonviral gene transfer have not precluded demonstration of potentially meaningful biological outcomes [5–8]. The principal advantage of nonviral gene transfer is that it obviates concerns relating to potentially adverse consequences of viral vectors [9,10]. To preserve this advantage and at the same time optimize the efficiency of gene transfer, modifications in vector design [11,12] as well as the mode of delivery have been investigated. The use of ballistic technology, known as the “gene gun” [13], resulted in a moderate enhancement of gene transfer to skin [14] and tumors [15]. The principal limitation of this approach was poor penetration into deep tissues. Gene transfer using cationic lipids [16,17] has been widely
576
employed for in vitro gene transfer and continues to be investigated for clinical gene transfer [18]. Neutral polymers have been reported to enhance plasmid delivery to rat muscle by approximately fivefold [19]. More recently, two reports have described the successful use of electric pulses for enhanced transfer of plasmid DNA [20,21]. In addition to its well-known use as a diagnostic and therapeutic tool, ultrasound (US) has been investigated as a means to enhance gene transfer [22–24]. US exposure has been shown to permeabilize plasma membranes, hence encouraging DNA entry into cells [25,26]. Lipofection agents support DNA release from endosomes through a physicochemical transition that is known to be accelerated by US [27,28]. Based on these observations, Lawrie et al. conducted in vitro studies [29] demonstrating that adjunctive US exposure is associated with enhanced transgene expression. We report here the characteristics and mechanisms of highly efficient and reproducible reporter gene transfer
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy 1525-0016/02 $35.00
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
A
B
to skeletal muscle achieved in vivo by the local adjunctive application of US exposure following direct intramuscular (i.m.) injection of a plasmid DNA reporter gene. The biological relevance of this finding was confirmed by augmented revascularization in an ischemic hindlimb model in response to US and naked vascular endothelial growth factor (VEGF) DNA.
RESULTS Effects of Ultrasound Parameters on Reporter Gene Expression To investigate the effects of ultrasound parameters on reporter gene expression, a -galactosidase gene-encoding plasmid (pCMV) was administered by direct i.m. injection into the quadriceps muscles of New Zealand White (NZW) rabbits. Upon completion of i.m. gene transfer, transcutaneous US was carried out directly over the site of injection and surrounding tissue defined by a diameter of ~ 5 cm; the latter was accomplished by changing the position of the US transducer at 60-second intervals. Total US administration time was 5 minutes per transfection site. Transfection efficiency of pCMV without US exposure, based upon measurement of tissue activity of β-galactosidase, was 6.3 ± 1.7 relative light units (RLU)/mg. In the first set of experiments, an US frequency of 1 MHz was used. A mechanical index (MI) of 2.0 and a duty cycle (DC) of 1.5% yielded a 9.1-fold increase in transgene expression to 57.9 ± 8.9 RLU/mg (P < 0.01 versus no US exposure). A reduced MI of 0.5 with an increased DC of 25% yielded similar results (69.9 ± 5.6 RLU/mg), consistent with a reciprocal relationship between these two parameters. When the MI was maintained
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy
ARTICLE
FIG. 1. Effects of ultrasound parameters on reporter gene expression. (A) After 100 g of the plasmid pCMV was injected into the upper thigh muscles of New Zealand White rabbits, the injection site was exposed for 5 minutes to ultrasound at the conditions indicated. Ultrasound frequency was 1 MHz; MI, mechanical index; DC, duty cycle. -Galactosidase expression was analyzed 5 days later by a chemoluminescent reporter gene assay. Results in relative light units (RLU) were corrected for total protein in the tissue samples. Bars represent the mean ± SEM; n = 10 per group. *P < 0.05; **P < 0.01. (B) Histogram illustrating the fold increases in transfection efficiency after exposure at different ultrasound conditions.
in a low range (1.8), and the DC extended up to 6%, transgene expression was 20.4-fold higher in comparison with no US exposure (129.9 ± 16.8 RLU/mg, P < 0.01 versus no US exposure; n = 10 per group; Figs. 1A and 1B). Further increases in either MI or DC did not result in higher transfection efficiency (data not shown). To investigate the temporal relationship between US exposure and gene transfer, experiments were conducted in which the transfection site was pretreated with US under optimized conditions (1.8 MI, 6% DC) for 5 minutes preceding gene transfer. US exposure also augmented transgene expression under this condition (72.2 ± 11.4 RLU/mg, P < 0.01 versus no US exposure); however, the increase in transgene expression of ~ 11-fold was less than that achieved when US was applied after gene transfer (Fig. 2). Finally, we assessed the effects of US wavelength on gene transfer. US at a lower frequency of 300 kHz and 1.8 MI, 6% DC resulted in increased transgene expression (120.6 ± 15.1 RLU/mg) that did not differ significantly
FIG. 2. Characteristics of ultrasound (US)-mediated transfer of a reporter gene. A 100-g aliquot of the plasmid pCMV was injected (TX) into the upper thigh muscles of New Zealand White rabbits. Ultrasound was applied for 5 minutes either before (pre) or after (post) TX. The mechanical index was 1.8, the duty cycle was 6%, and the frequencies were either 1 MHz or 300 kHz. -Galactosidase expression was analyzed 5 days later by a chemiluminescent reporter gene assay. Results in relative light units (RLU) were corrected for total protein in the tissue samples. Bars represent the mean ± SEM; n = 10 per group. *P < 0.05; **P < 0.01.
577
ARTICLE
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
A
FIG. 3. Morphometric analysis of -galactosidase expression. Animals were transfected with pGSV plasmid encoding nuclear-localizing -galactosidase. Shown are representative sections of muscle stained with X-gal, counterstained with eosin, 5 days after transfection. In both sections, animals were transfected with the plasmid, but in (A) the tissue was not exposed to ultrasound following gene transfer, resulting in fewer transfected nuclei than in (B), where tissue was exposed to ultrasound. Magnification, ⫻100. White-boxed insets in (A) and (B) show ⫻200 magnifications of selected areas.
shown). Moreover, we also monitored tissue temperature at 4 mm depth within the sonicated muscle tissue for temperature increases that might be caused by US exposure. We observed no substantial temperature changes.
B
from that observed with 1 MHz (Fig. 2). Because the design of the 1-MHz US probes allowed for the generation of wide US beams with a long field of view, all subsequent experiments were done using US at the conditions of 1 MHz, an MI of 1.8, and a DC of 6%. Morphometric analysis of myocyte nuclei staining positive for -galactosidase expression after gene transfer with the nuclear-targeted plasmid pGSVLacZ showed that US exposure, when applied after gene transfer, increased the number of positive nuclei from 9 ± 4 (without US exposure) of total nuclei to 68 ± 11 (1 MHz, 1.8 MI, 6% DC, 5 minutes; P < 0.01 versus. no US exposure) of total nuclei per low-power field (Fig. 3). Evaluation of Ultrasound Biosafety To investigate whether US (1 MHz, 1.8 MI, 6% DC, 5 minutes) interfered with tissue integrity in our in vivo experiments, we conducted light-microscopic analyses of muscle samples harvested from treatment sites, and we found that no alteration in tissue integrity occurred (data not
578
Effects of Ultrasound Exposure on Therapeutic Gene Expression Based upon results of the reporter gene transfer experiments, we studied the biological relevance of US-mediated gene transfer using the plasmid encoding for human VEGF165 (phVEGF165) to promote therapeutic angiogenesis in a rabbit model of hindlimb ischemia. Previous studies using this animal model have demonstrated that an i.m. dose of 500 g of phVEGF165 is sufficient in young (~ 6 months old) NZW rabbits to induce a level of therapeutic angiogenesis that constitutes a near-optimal response [5]. In the current study, we used old (5–6 years) NZW rabbits. The rationale for this approach is that angiogenesis is impaired in old animals [30]; consequently, a greater dynamic range in the old animals permits comparison of strategies such as US exposure designed to augment gene transfer and thereby facilitate hindlimb neovascularization. Anatomic Assessment Representative angiograms recorded from the four treatment groups (saline injection only = control; saline injection + US exposure; phVEGF165 transfection only; phVEGF165 transfection + US exposure; n = 5 per group) at day 30 after treatment are shown in Fig. 4. In control animals, collateral development in the medial thigh typically appeared unchanged in serial angiograms recorded at days 0 (data not shown) and 30 (Fig. 4A). When US was applied after the injection of saline, no differences in collateral growth could be detected as compared with the control group (Fig. 4B). In contrast, in the phVEGF165transfected group, marked progression of collateral artery development was observed in comparison with the control animals (Fig. 4C). When VEGF gene transfer was followed by US exposure, however, collateral development was even more abundant (Fig. 4D). Quantitative analysis of collateral vessel development in the medial thigh was done by calculating an angiographic score (Fig. 5). At day 30, the angiographic score for rabbits undergoing US exposure alone (not transfected with naked VEGF DNA) was similar to that recorded for control animals (saline injection only, 0.47 ± 0.02 versus 0.48 ± 0.02,
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy
ARTICLE
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
FIG. 4. Representative images of selective internal iliac angiography at day 30 after treatment. (A) Control rabbit, saline injection, no ultrasound. (B) Saline-injected rabbit with US exposure shows no obvious increase in collateral development, as compared with (A). (C) This hindlimb was transfected with 500 g i.m. of phVEGF165 and shows extensive collateral artery formation. (D) When VEGF transfection was followed by US exposure, collateral formation was most abundant.
A
respectively, P = not significant). The phVEGF165-transfected group achieved a significantly higher score (0.68 ± 0.04; P < 0.01 versus control and P < 0.01 versus US exposure only). Consistent with C results illustrated in Fig. 5, the highest angiographic score was recorded for the group in which phVEGF165 gene transfer was followed by US exposure (0.8 ± 0.01; P < 0.01 versus VEGF alone). A favorable effect of US exposure on gene transfer–induced revascularization was also apparent at the capillary level (Fig. 6). The adductor muscle of ischemic limbs was histologically examined at day 30. Analysis of capillary/myocyte ratio was higher in the US-mediated phVEGF165 gene transfer group (0.77 ± 0.05) than in animals in which gene transfer was carried out by standard i.m. injection (0.59 ± 0.02; P < 0.01). The outcomes were superior for both treatment groups versus the control groups (0.39 ± 0.03 for saline only, P < 0.01 versus phVEGF165 only; and 0.43 ± 0.03 for saline + US exposure, P < 0.01 versus phVEGF165 only; n = 5 per group). Physiological Assessment Changes in the hemodynamic deficit in the ischemic limb after gene transfer were confirmed by measurement of the calf blood pressure (Fig. 7). By day 30 after gene transfer, blood pressure ratio for the phVEGF165-transfected group (0.74 ± 0.04) was significantly higher (P < 0.01) than that for either the saline-injected (0.49 ± 0.03) or US-only control groups (0.47 ± 0.03), respectively. In the rabbits that underwent US-mediated phVEGF165 gene transfer, an additional statistically significant increase in blood pressure ratio was achieved (0.89 ± 0.04; P < 0.05 versus VEGF only; n = 5 per group). Resting blood flow and papaverine-stimulated (maximum) blood flow in the ischemic limb were assessed by intravascular Doppler flow measurements (Fig. 8). Ultrasound exposure itself had no effect on either resting (13.7 ± 1.4 ml/minute for saline, 13.7 ± 1.4 ml/minute for saline + US exposure), or stimulated (22.8 ± 2.8 ml/minute for saline, 22.7 ± 1.5 ml/minute for
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy
B
D
saline + US exposure) blood flow. VEGF gene transfer significantly augmented blood flow to the ischemic limb through the internal iliac artery, both at rest (19.0 ± 0.8 ml/minute; P < 0.05 versus control) and following papaverine stimulation (30.8 ± 2.6 ml/minute; P < 0.05 versus control). Ultrasound-mediated VEGF gene transfer enhanced flow to the ischemic limb after papaverine stimulation (41.0 ± 2.8 ml/minute; P < 0.05 versus phVEGF165 only at maximum; n = 5 per group); although there was a trend toward higher resting flow in rabbits undergoing US-mediated phVEGF165 gene transfer (20.7 ± 0.9 ml/minute), this did not achieve statistical significance.
DISCUSSION Intramuscular injection of naked plasmid DNA is associated with little evidence of toxicity and/or toxicity and immunogenicity [31]. The disadvantage of this gene transfer strategy is relatively low transfection efficiency, particularly in comparison with viral vectors. Nevertheless, preclinical studies done in rabbit [5,32,33] and murine [30,34] models of hindlimb ischemia have demonstrated proof of concept that collateral vessel development may be induced by i.m. gene transfer of naked DNA encoding
579
ARTICLE
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
FIG. 5. Effects of i.m. phVEGF165 gene transfer and ultrasound exposure on revascularization in the ischemic hindlimb model. As indicated by angiographic score, saline-injected rabbits and rabbits treated with saline + ultrasound had significantly fewer collaterals than the phVEGF165-treated animals at day 30 after treatment. Those treated with i.m. phVEGF165 together with ultrasound had the most prominent angiogenic response. Bars show the mean ± SEM; n = 5 per group. *P < 0.05; **P < 0.01.
FIG. 6. Capillary/myocyte ratio as a measure for angiogenesis. Adductor muscles of ischemic limbs were histologically examined at day 30 to determine capillary/myocyte ratios. Each bar of the graph represents the mean ± SEM; n = 5 per group. *P < 0.01.
for VEGF. Ultimately, this has been reproduced in patients with critical limb ischemia [35–37], and more recently in patients with myocardial ischemia [38]. Numerous attempts have been made so far to overcome the relatively low transfection efficiency of naked DNA gene transfer, including the use of cationic lipids or polymers [16,17], or the application of electric pulses to DNA injection sites [20,21]. We report here a method for efficiently increasing naked DNA transfer in skeletal muscle, consisting of standard i.m. injection of naked plasmid DNA, followed by delivery of US by an external probe. Initial experiments employed the use of naked DNA encoding for a reporter gene (β-galactosidase), injected in a standardized manner into normal rabbit hindlimb muscle. Analysis of transgene expression, which was assessed by a chemiluminescent enzyme assay, demonstrated the anticipated low transfection efficiency, which was substantially and consistently increased when US was applied to the area surrounding the injection site. In this context, we found that US exposure applied to skeletal muscle after i.m. gene transfer was superior to sonication applied directly before gene transfer; this finding may be interpreted to indicate that US permeabilizes cell membranes for an increased uptake of naked plasmid DNA [23,39]. The US dose/response studies that we conducted in this in vivo assay revealed that US at a frequency of 1 MHz, 1.8 MI, 6% duty cycle, and 5 minutes exposure time yielded the highest -galactosidase expression from 100 g of pCMV. The applied energy of this
condition was higher than what others have used for cell culture transfection experiments [23,29]. To assess morphometrically the effects of US exposure on gene transfer, we used naked plasmid DNA encoding for nuclear-targeted GSVLacZ. Results from these experiments support the possibility that US exposure significantly increased the overall number of transfected skeletal myocyte nuclei; the methods used, however, do not allow assessment of the absolute amount of total transgene expression per cell. We monitored the safety of US exposure by light-microscopic analyses of tissue samples and close control of the core temperature of the target tissue. Considering that we detected no adverse effects of US exposure on the target tissue, our findings contrast with reports of several in vitro transfection techniques, which were associated with cell death [40]. Similarly, although the use of electrical pulses has been demonstrated to increase transfection efficiency [20,21], side effects included pain, passive involuntary muscle contractions, and tissue damage associated with thermal changes. Furthermore, when Mir et al. [21] applied electric pulses to rabbit rectus femoris muscle, they found > 50% variability in transgene expression, whereas USmediated gene transfer in our experiments showed a variability of ~ 7.5%. No evidence of muscle contractions or pain was seen with US-mediated gene transfer. To determine the biological relevance of the reporter gene experiments, we investigated US-mediated gene transfer for revascularization of ischemic hindlimbs. In previous reports, young (~ 6 months old) NZW rabbits
580
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy
ARTICLE
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
FIG. 7. Revascularization assessed by calf blood pressure ratio. Differences in hemodynamic outcomes among treatment groups were assessed by measuring calf blood pressure in ischemic and nonischemic limbs at day 30 after treatment. Results are shown as ratio of ischemic to nonischemic limbs. Ultrasound alone did not have an effect on calf blood pressure ratio, whereas i.m. phVEGF165 resulted in a significant increase. When ultrasound was applied for 5 minutes after the DNA injection, blood pressure ratio was significantly higher than in the group treated with VEGF only. Bars represent the mean ± SEM; n = 5 per group. *P < 0.05; **P < 0.01.
have been used to establish the concept of therapeutic angiogenesis with the i.m. injection of naked plasmid DNA encoding VEGF165 [5]. In these studies, VEGF gene transfer to ischemic hindlimbs was shown to improve revascularization parameters to almost normal levels in these animals. In the current experiments we therefore used old (~ 5–6 years old) NZW rabbits, in which angiogenesis has been shown to be naturally retarded [30]. We reasoned that this impaired response might provide a more dynamic range to assess whether the increase in transfection efficiency achieved with US-mediated gene transfer might yield superior revascularization compared with direct injection of naked plasmid DNA alone. Indeed, we found that US exposure yielded a statistically significant improvement in revascularization over direct injection alone, as determined by calf blood pressure ratio, angiographic score, intravascular Doppler blood flow, and capillary/myocyte ratio. It is interesting to note that direct injection of phVEGF165 alone induced significant revascularization of the ischemic hindlimb in these old rabbits, but to a lesser extent than found in young NZW rabbits. When phVEGF165 gene transfer was followed by US exposure, however, the improvement in hindlimb neovascularization was comparable to that seen with phVEGF165 alone in young NZW rabbits [5].
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy
FIG. 8. Intra-arterial blood flow measurements. Resting blood flow and papaverine-stimulated blood flow in the ischemic limb were measured at day 30 after treatment by intra-arterial Doppler guide wire measurements. Bars show the mean ± SEM; n = 5 per group. *P < 0.05; **P < 0.01.
Moreover, when ischemic hindlimbs of young NZW rabbits were injected with suboptimal doses of phVEGF165, with or without ultrasound (unpublished data), US treatment following i.m. injection was able to rescue an otherwise insufficient angiogenic response to low-dose phVEGF165 injection. It is important to note that in both young and old rabbits, US exposure itself did not augment therapeutic neovascularization when compared to saline-injected controls. This is consistent with results from the reporter gene transfection experiments, in which we found that US exposure alone did not increase expression of endogenous -galactosidase. Reher et al. [41] previously reported that US exposure of fibroblasts, osteoblasts, or mononuclear blood cells increased concentrations of VEGF protein in cell culture media. The principal tissue target in our in vivo experiments consisted of skeletal myocytes; even if US exposure did induce the release of endogenous VEGF protein, the absolute amount may have been too limited to induce revascularization in this animal model.
MATERIALS
AND
METHODS
Animal models. All protocols were approved by St. Elizabeth’s Institutional Animal Care and Use Committee. The effects of ultrasound on transfection efficiency and transgene expression were investigated in two rabbit models. In all experiments, investigators performing the in vivo and in vitro follow-up examinations were blinded to the identity of the treatment administered.
581
ARTICLE
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
Reporter gene expression in normal rabbit hindlimbs. For analysis of -galactosidase expression in skeletal muscle, 70 male NZW rabbits (Pine Acre Rabbitry, Norton, MA) (weight 3.5–4 kg) were anesthetized with a mixture of ketamine (50 mg/kg) and acepromazine (0.8 mg/kg) after premedication with xylazine (2 mg/kg). The quadriceps muscles of both thighs were exposed by skin incisions, and the sites of the plasmid injections into these muscles were each marked by single sutures. Per hindlimb, one injection of 100 g of pCMV in 500 l of saline was done, which was then followed by US exposure in a subgroup of animals. After completion of gene transfer and US exposure, the wounds were closed with 4-0 nylon sutures and the animals were kept alive for 5 days. At death, the single sutures facilitated identification of injection sites, following which muscle samples were then harvested, snap-frozen in liquid nitrogen, and stored at –80⬚C until further analysis. Ultrasound exposures. Wide-field ultrasonic transducers (PharmaSonics, Sunnyvale, CA) were placed in direct contact with the exposed muscles with acoustic coupling provided by ultrasound transmission gel (Ultraphonic; Pharmaceutical Innovations, Newark, NJ). Possible thermal changes within the target tissue were monitored by a temperature probe embedded at a depth of 4 mm in the muscle that was subject to sonication. Transducer drive signals were generated by a 16-MHz function/arbitrary waveform generator (33120A; Hewlett Packard, Palo Alto, CA), amplified by a custom power amplifier (PharmaSonics), and monitored by a 200-MHz digital oscilloscope (TDS360; Tektronix, Beaverton, OR). Transducers were calibrated in a water tank against a calibrated PVDF bilaminar membrane hydrophone (Series 805 with Series 806 preamp.; Sonic Industries, Hatboro, PA). Transducers were typically driven with 30-cycle bursts of 950 kHz ultrasound, at a burst repetition rate of 1890 Hz (6% DC). Drive amplitudes were adjusted to achieve the reported output MIs, where the MI is defined as the maximum negative peak pressure in MPa divided by the square root of the ultrasound frequency in MHz. Over the depth of field of interest for these experiments, the transducers typically displayed –6-dB beam profiles with beamwidths on the order of 9–10 mm. Total US exposure per DNA injection site was 5 minutes, during which the US probe was shifted in 1-cm increments once a minute to cover fully the spread of the injectate in the tissue by the US beam. Plasmids and DNA preparation. To study the effects of US on VEGF overexpression in hindlimb ischemia, we administered to rabbits a eukaryotic expression vector encoding the human VEGF165 gene [42] transcriptionally regulated by the cytomegalovirus (CMV) promoter/enhancer (phVEGF165) [5,37]. To study the influence of US exposure on reporter gene expression, pCMV (Clontech, Palo Alto, CA) was used as promotermatched reporter plasmid. This plasmid encodes -galactosidase under the control of a CMV promoter/enhancer. The plasmid pGSVLacZ (courtesy of Claire Bonnerot) containing a nuclear-targeted -galactosidase sequence coupled to the simian virus-40 (SV40) early promoter was used for the control transfection experiments. Preparation and purification of all plasmids from cultures of phVEGF165-, pCMV-, or pGSVLacZ-transformed Escherichia coli were carried out by the column method (Qiagen Mega Kit; Qiagen, Valencia, CA). Measurement of -galactosidase activity. Muscles were analyzed for -galactosidase expression using the Galacto-Light Plus chemoluminescent reporter gene assay system (Tropix, Bedford, MA) according to the manufacturer’s directions. Briefly, muscles were homogenized in lysis buffer (100 mM potassium phosphate, 0.2% Triton X-100) supplemented with protease inhibitors (aprotinin, 2 g/ml; leupeptin, 5 g/ml; and pepstatin, 2 g/ml, all from Sigma, St. Louis, MO) and dithiothreitol (1 mM, Sigma). After adding the Galacton Plus substrate, samples were incubated for 45 minutes at room temperature before light emission was read in a luminometer (LB 9507; Berthold, Germany). Total protein concentration per muscle sample was measured using the BCA protein assay kit (Pierce, Rockford, IL), according to the manufacturer’s protocol. All results are expressed as relative light units (RLU) per mg total protein. Endogenous -galactosidase produced a mean of 54.4 RLU/mg, as averaged from nontransfected quadriceps muscles of five normal NZW rabbits. All data from -galactosidase transfection experiments were therefore corrected for this background.
582
The efficiency of transgene expression after US exposure was also evaluated morphometrically in 10 additional rabbits following i.m. injection of 100 g of the plasmid pGSVLacZ, containing a nuclear-targeted -galactosidase sequence. This obviated nonspecific cytoplasmic background staining due to endogenous -galactosidase activity in nontransfected control muscles. The transfected muscles were harvested 5 days later, fixed in 1% paraformaldehyde, and incubated with X-Gal (5-bromo-4-chloro-3-indolyl-D-galactosidase chromogen; Sigma) overnight at 37⬚C [43]. Harvested tissues were then paraffin-embedded, sectioned, and counterstained with eosin. Five sections from each sample were randomly selected, and the numbers of positive nuclei in five low-power fields distributed among an area of skeletal muscle section measuring 10 mm in diameter were counted manually for each specimen by a blinded observer. To assess possible tissue damage caused by US exposure, muscle samples from select rabbits were stained with H&E, followed by light-microscopic analysis. Rabbit ischemic hindlimb model. To investigate the biological relevance of US-enhanced transfection efficiency, the rabbit ischemic hindlimb model was used in 32 old (5–6 years) animals according to previously published methods [30]. Briefly, after anesthesia and unilateral excision of the entire superficial femoral artery, rabbits were allowed to recover for 10 days. After reanesthetizing the rabbits as outlined above, the skin was opened and five i.m. injections, each containing 100 g of naked plasmid DNA encoding for human VEGF165 (phVEGF165) in 500 l, were administered to three major muscles of the thigh [5]. A subgroup of rabbits underwent US treatment immediately after VEGF DNA injections. Upon completion of treatment, the skin was closed with 4-0 nylon sutures. Blood pressure (BP) was measured as described [44]. The ratio of ischemic to normal hindlimb BP (BPR) was defined for each rabbit as the ratio of systolic pressure measured in the ischemic limb to systolic pressure measured in the normal limb. Selective angiography of the ischemic hindlimb was conducted on day 30 after surgery as described [32,33,44]. To assess collateral vessel development quantitatively, we used an acetate overlay with an imprinted grid composed of 2.5-mm-diameter circles arranged in rows spaced 5 mm apart to yield an angiographic score as described [44]. Blood flow to the ischemic hindlimb was quantified in vivo before selective internal iliac angiography on day 30 with a 0.018-in Doppler guide wire (Cardiometrics, Mountain View, CA) as described [5]. APV (time average of the spectral peak velocity) was recorded at rest, and maximum APV was obtained after bolus injection of 2 mg of papaverine. Dopplerderived flow was calculated as QD = (pd2/4) (0.5 ⫻ APV), where QD is Doppler-derived time-averaged flow, and d is vessel diameter, which was determined from an angiogram with an automated edge-detection system as described [45]. Tissue specimens obtained as transverse sections from the adductor and semimembranous muscle groups of both limbs of each rabbit at the time of sacrifice (day 30) were embedded in OCT compound (Miles, Elkhart, IN) and snap-frozen in liquid nitrogen. Tissue sections were stained for alkaline phosphatase by an indoxyl-tetrazolium method to detect capillary endothelial cells and were then counterstained with eosin. A total of 20 different fields from the two muscles were randomly selected, and capillary density was evaluated by a single observer blinded to the treatment group under a 20⫻ objective. Finally, the number of myofibers in each field were also counted to calculate the capillary/myocyte ratio. Statistics. All results are expressed as the mean ± SEM (m ± SEM). Statistical comparisons between groups were done by ANOVA, which was followed by an unpaired, two-tailed Student’s t-test when a significant difference was detected. P values < 0.05 were considered to denote statistical significance.
ACKNOWLEDGMENTS We thank Mickey Neely for assistance in the preparation of this manuscript. This study was supported in part by NIH grants HL53354, HL57516, HL60911, and HL63414, and by the Peter Lewis Foundation (Cleveland, OH). P.S. is the recipient of an Erwin Schroedinger fellowship grant, provided by the Austrian Funds for the Advancement of Science (FWF).
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy
doi:10.1006/mthe.2002.0715, available online at http://www.idealibrary.com on IDEAL
RECEIVED FOR PUBLICATION AUGUST 24, 2001; ACCEPTED SEPTEMBER 10, 2002.
REFERENCES 1. Blau, H. M., and Springer, M. L. (1995). Gene therapy—a novel form of drug delivery. N. Engl. J. Med. 333: 1204–1207. 2. Ledley, F. D. (1995). Nonviral gene therapy: the promise of genes as pharmaceutical products . Hum. Gene Ther. 6: 1129–1144. 3. Rolland, A. P. (1998). From genes to gene medicines: recent advances in nonviral gene delivery. Crit. Rev. Ther. Drug Carrier Syst. 15: 143–198. 4. Lee, R. J., and Huang, L. (1997). Lipid vector systems for gene transfer. Crit. Rev. Ther. Drug Carrier Syst. 14: 173–206. 5. Tsurumi, Y., et al. (1996). Direct intramuscular gene transfer of naked DNA encoding vascular endothelial growth factor augments collateral development and tissue perfusion. Circulation 94: 3281–3290. 6. Bauters, C., et al. (1994). Physiologic assessment of augmented vascularity induced by VEGF in ischemic rabbit hindlimb. Am. J. Physiol. 267: H1263–H1271. 7. Tabata, H., Silver, M., and Isner, J. M. (1997). Arterial gene transfer of acidic fibroblast growth factor for therapeutic angiogenesis in vivo: critical role of secretion signal in use of naked DNA. Cardiovasc. Res. 35: 470–479. 8. Alila, H., et al. (1997). Expression of biologically active human insulin-like growth factor-I following intramuscular injection of a formulated plasmid in rats. Hum. Gene Ther. 8: 1785–1795. 9. Balter, M. (2000). Gene therapy on trial. Science 288: 951–957. 10. Kielian, T., and Hickey, W. F. (2000). Inflammatory thoughts about glioma gene therapy. Nat. Med. 5: 1237–1238. 11. Hartikka, J., et al. (1996). An improved plasmid DNA expression vector for direct injection into skeletal muscle. Hum. Gene Ther. 7: 1205–1217. 12. Berthou-Soulie, L., et al. (2000). PCOR-mediated gene transfer in ischemic tissues: application of electrotransfer to peripheral ischemia (Abstract). Mol. Ther. 1: S209. 13. Leclerc, G., and Isner, J. M. (1993). Percutaneous gene therapy for cardiovascular disease. In Interventional Cardiology (E. Topol, Ed.), pp. 1019–1029. W. B. Saunders Company, Philadelphia, PA. 14. Fynan, E. F., et al. (1993). DNA vaccines: protective immunizations by parenteral, mucosal, and gene-gun inoculations. Proc. Natl. Acad. Sci. USA 90: 11478–11482. 15. Yang, N.-S., and Sun, W. H. (1995). Gene gun and other non-viral approaches for cancer gene therapy. Nat. Med. 1: 481–483. 16. Wolff, J. A., et al. (1992). Expression of naked plasmids by cultured myotubes and entry of plasmids into T tubules and caveolae of mammalian skeletal muscle. J. Cell Sci. 103: 1249–1259. 17. Schwartz, B., et al. (1996). Gene transfer by naked DNA into adult mouse brain. Gene Ther. 3: 405–411. 18. Yla-Herttuala, S., and Martin, J. F. (2000). Cardiovascular gene therapy. Lancet 355: 213–222. 19. Mumper, R. J., et al. (1998). Protective interactive noncondensing (PINC) polymers for enhanced plasmid distribution and expression in rat skeletal muscle. J. Controlled Release 52: 259–272. 20. Aihara, H., and Miyazaki, J.-I. (1998). Gene transfer into muscle by electroporation in vivo. Nat. Biotechnol. 16: 867–870. 21. Mir, L. M., et al. (1999). High-efficiency gene transfer into skeletal muscle mediated by electric pulses. Proc. Natl. Acad. Sci. USA 96: 4262–4267. 22. Fechheimer, M., et al. (1987). Transfection of mammalian cells with plasmid DNA by scrape loading and sonication loading. Proc. Natl. Acad. Sci. USA 84: 8463–8467. 23. Kim, H. J., Greenleaf, J. F., Kinnick, R. R., Bronk, J. T., and Bolander, M. E. (1996). Ultrasound-mediated transfection of mammalian cells. Hum. Gene Ther. 7: 1339–1346.
MOLECULAR THERAPY Vol. 6, No. 5, November 2002 Copyright © The American Society of Gene Therapy
ARTICLE
24. Zhang, L. J., et al. (1991). Efficient transformation of tobacco by ultrasonification. Biotechnology 9: 996–997. 25. Gambihler, S., Delius, M., and Ellwart, J. W. (1994). Permeabilization of the plasma membrane of L1210 mouse leukemia cells using lithotripter shock waves. J. Membr. Biol. 141: 267–275. 26. Pohl, P., Rosenfeld, E., and Millner, R. (1993). Effects of ultrasound on the steady-state transmembrane pH gradient and the permeability of acetic acid through bilayer lipid membranes. Biophys. Acta 1145: 279–283. 27. Gao, X., and Huang, L. (1995). Cationic liposome-mediated gene transfer. Gene Ther. 2: 710–722. 28. Tata, D. B., and Dunn, F. (1992). Interaction of ultrasound and model membrane systems: analyses and predictions. J. Phys. Chem. 96: 3548–3555. 29. Lawrie, A., et al. (1999). Ultrasound enhances reporter gene expression after transfection of vascular cells in vitro. Circulation 99: 2617–2620. 30. Rivard, A., et al. (1999). Age-dependent impairment of angiogenesis. Circulation 99: 111–120. 31. Wolff, J. A., Ludtke, J. J., Acsadi, G., Williams, P., and Jani, A. (1992). Long-term persistence of plasmid DNA and foreign gene expression in mouse muscle. Hum. Mol. Genet. 1: 363–369. 32. Takeshita, S., et al. (1996). Gene transfer of naked DNA encoding for three isoforms of vascular endothelial growth factor stimulates collateral development in vivo. Lab. Invest. 75: 487–502. 33. Takeshita, S., et al. (1995). Time course of increased cellular proliferation in collateral arteries following administration of vascular endothelial growth factor in a rabbit model of lower limb vascular insufficiency. Am. J. Pathol. 147: 1649–1660. 34. Murohara, T., et al. (1998). Nitric oxide synthase modulates angiogenesis in response to tissue ischemia. J. Clin. Invest. 101: 2567–2578. 35. Baumgartner, I., et al. (1998). Constitutive expression of phVEGF165 following intramuscular gene transfer promotes collateral vessel development in patients with critical limb ischemia. Circulation 97: 1114–1123. 36. Isner, J. M., et al. (1998). Treatment of thromboangiitis obliterans (Buerger’s disease) by intramuscular gene transfer of vascular endothelial growth factor: preliminary clinical results . J. Vasc. Surg. 28: 964–975. 37. Isner, J. M., et al. (1996). Clinical evidence of angiogenesis following arterial gene transfer of phVEGF165. Lancet 348: 370–374. 38. Vale, P. R., et al. (2000). Left ventricular electromechanical mapping to assess efficacy of phVEGF165 gene transfer for therapeutic angiogenesis in chronic myocardial ischemia. Circulation 102: 965–974. 39. Tata, D. B., Dunn, F., and Tindall, D. J. (1997). Selective clinical ultrasound signals mediate differential gene transfer and expression in two human prostate cancer cell lines: LnCap and PC-3. Biochem. Biophys. Res. Commun. 234: 64–67. 40. Bao, S., Thrall, B. D., and Miller, D. L. (1997). Transfection of a reporter plasmid into cultured cells by sonoporation in vitro. Ultrasound Med. Biol. 23: 953–959. 41. Reher, P., Doan, N., Bradnock, B., Meghji, S., and Harris, M. (1999). Effect of ultrasound on the production of IL-8, basic FGF and VEGF. Cytokine 11: 416–423. 42. Tischer, E., et al. (1991). The human gene for vascular endothelial growth factor: multiple protein forms are encoded through alternative exon splicing. J. Biol. Chem. 266: 11947–11954. 43. Feldman, L. J., et al. (1995). Low efficiency of percutaneous adenovirus-mediated arterial gene transfer in the atherosclerotic rabbit. J. Clin. Invest. 95: 2662–2671. 44. Takeshita, S., et al. (1994). Therapeutic angiogenesis: a single intra-arterial bolus of vascular endothelial growth factor augments revascularization in a rabbit ischemic hindlimb model. J. Clin. Invest. 93: 662–670. 45. Ku, D. D., Zaleski, J. K., Liu, S., and Brock, T. A. (1993). Vascular endothelial growth factor induces EDRF-dependent relaxation in coronary arteries. Am. J. Physiol. 265: H586–H592.
583