A Novel ex Vivo Angiogenesis Assay Based on Electroporation

METHOD

doi:10.1016/j.ymthe.2003.12.002

A Novel ex Vivo Angiogenesis Assay Based on Electroporation-Mediated Delivery of Naked Plasmid DNA to Skeletal Muscle Hyung-Suk Jang, Hyun-Joong Kim, Jeong-Min Kim, Young-Sam Lee, Koung Li Kim, Jeong-A Kim, Jae-Young Lee, Wonhee Suh, Jin-Ho Choi, Eun-Seok Jeon, Jonghoe Byun,*,y and Duk-Kyung Kim*,y Cardiac and Vascular Center, Department of Medicine, and Samsung Medical Center, Sungkyunkwan University School of Medicine, 50 Ilwon-dong, Kangnam-ku, Seoul 135-710, South Korea * These authors contributed equally to this work. y To whom correspondence and reprint requests should be addressed at the Cardiac and Vascular Center, Samsung Medical Center, Department of Medicine (D.-K.K.), or the Samsung Biomedical Research Institute (J.B.), Sungkyunkwan University School of Medicine, 50 Ilwon-dong, Kangnam-ku, Seoul 135-710, South Korea. Fax: 82-2-3410-3849 (D.-K.K.) or 82-2-3410-6829 (J.B.). E-mail: [email protected] (D.-K.K.) or [email protected] (J.B.).

An angiogenesis assay based on gene transfer would be extremely useful for angiogenic gene therapy. A simple, reproducible, and quantitative assay to test angiogenic genes would provide more accurate predictions than conventional peptide-based assays. Here, we have developed a semiquantitative angiogenesis assay utilizing gene transfer into skeletal muscle, which is a target tissue for ischemic limb diseases. To facilitate quick and clean analysis, a naked plasmid DNA vector combined with an electroporation procedure was used for gene transfer. When the plasmid vector encoding vascular endothelial growth factor cDNA (pJDK-VEGF165) was injected into the tibialis anterior muscle of BALB/c mice, followed by in vivo electroporation and explant culture in growth factor-reduced Matrigel, the outward migration of sprouting cells was observed as early as day 2. The cells soon formed capillary networks, which peaked at day 7 and persisted until day 14. The capillary-like structures were positive for von Willebrand factor, platelet endothelial cell adhesion molecule, and vimentin, suggesting they were endothelial cells. There was little, if any, sprouting or formation of capillaries from the control vector (pJDK)-injected group. Consistent with the region of sprouting and network formation, the amount of secreted VEGF increased in the conditioned medium of explant cultures. The angiogenic potential of connective tissue growth factor (CTGF) was examined using the new assay. Whereas the CTGF gene alone induced weak sprouting activity, it appeared to inhibit the angiogenic activity of the VEGF165 gene during cotreatment. This attenuating activity of CTGF on VEGF was reproduced in vivo in a murine model of hindlimb ischemia. In a group of mice treated with both pJDK-CTGF and pJDK-VEGF165, the blood flow measured by laser Doppler imaging was significantly lower than that of the pJDK-VEGF165-treated group 10 days after femoral artery excision. These results are consistent with recent reports that suggest that CTGF inhibits VEGF. This confirms the usefulness of this novel ex vivo assay in assessing the angiogenic capacity of genes of interest. In summary, this new gene-based angiogenesis assay should be widely applicable in the study of angiogenic or antiangiogenic genes because it can readily predict the angiogenic potential of specific genes and their combinations. Key Words: angiogenesis, gene therapy, skeletal muscle, electroporation, connective tissue growth factor

464

MOLECULAR THERAPY Vol. 9, No. 3, March 2004 Copyright B The American Society of Gene Therapy 1525-0016/$30.00

doi:10.1016/j.ymthe.2003.12.002

INTRODUCTION Considerable progress has been made since the concept of therapeutic angiogenesis was first introduced in the treatment of ischemic diseases [1 – 3]. Much of this progress has centered on growth factors that elicit angiogenesis, a complex process that involves the migration, proliferation, and differentiation of endothelial cells; vascular tube formation; and linkage to preexisting vascular networks [4 – 6]. Although protein delivery can be successfully used for therapeutic revascularization of ischemic tissues, most studies have so far relied on the delivery of genes rather than peptides. In the development of gene therapy for therapeutic angiogenesis, an easy angiogenesis assay with which to test various genes of interest would be very useful. However, the development of such gene-based angiogenesis assays to evaluate angiogenic potential has been limited. Conventional angiogenesis assays include the rat aortic ring assay [7,8], chorioallantoic membrane (CAM) assay [9,10], corneal micropocket assay [11,12], collagen and Matrigel assay [13,14], Matrigel plug assay [15], and direct assessment of angiogenesis in animals by microangiography, Doppler ultrasonography, or immunohistological staining [16]. Although each angiogenesis assay has its own advantages and an application to which it is best suited, most of them were designed to test protein factors. Proteins or peptides, however, are often difficult to purify to complete homogeneity. They can also have such shortcomings as moderate yield, cumbersome production procedures, and high costs. Moreover, for sustained effects, repeated administrations

METHOD

of peptides are usually required. Therefore, a gene in the form of cDNA, which is easily isolated by reverse transcription-polymerase chain reaction (RT-PCR), offers advantages in terms of simple purification, availability, and sustained expression. Among the many gene transfer vectors, naked plasmid DNA provides the simplest means of expressing target genes. It avoids infectious agents and the injected DNA remains episomal, greatly reducing the potential risks associated with integration of DNA into the host genome. Direct injection of naked plasmid DNA was first studied in skeletal muscle [17], with subsequent extended use of this methodology by several laboratories [18 – 20]. The major drawback to this approach is its relative inefficiency. Although this drawback does not affect the utility of this method in testing the functional roles of certain genes, its low efficiency has precluded its widespread use. Recently, however, physicochemical developments, including electroporation, have provided tools that increase transfer efficiency. Electroporation is now a leading gene transfer procedure for skeletal muscle tissues because it provides drastically increased, localized, and sustained expression of target genes in the target muscle [21 – 25]. Gene delivery to skeletal muscle has many applications, including the correction of myopathies, vaccination, and the local expression of angiogenic growth factors. Another interesting application is the use of highly vascularized muscle as an endocrine factory for the systemic secretion of therapeutic proteins such as erythropoietin and hormones [17,26]. Because most therapeutic angiogenesis studies have been directed toward occlusive diseases in skeletal muscle or myocar-

FIG. 1. Procedures of the ex vivo skeletal muscle angiogenesis assay. (A) Injection of 10 Ag of DNA in 0.4% NaCl into the tibialis anterior muscle with a 30-gauge insulin syringe. The shaved and depilated leg is shown in gray. (B) Transcutaneous electrical stimulus was applied using an electroporator (ECM 830), through two electrodes placed on the surface of the injection site. (C) Preparation of the tibialis anterior muscle 2 days after injection. The tibialis anterior muscle was detached from the bone and cut in the middle of the injection area to expose an even surface for embedding. (D) Embedding of the explanted muscle in a 24-well plate. The cut muscle was placed on growth factor-reduced Matrigel with the section surface facing the Matrigel. The explant was then covered with 1.5 ml of M199 medium containing 2% FBS.

MOLECULAR THERAPY Vol. 9, No. 3, March 2004 Copyright B The American Society of Gene Therapy

465

METHOD

dium, a naked plasmid DNA vector coupled to electroporation can be considered one of the best gene transfer strategies for skeletal muscle. Here, we tested the feasibility of a novel angiogenesis assay based on explant cultures of skeletal muscle tissue that had been transfected with naked plasmid DNA vectors encoding angiogenic genes. An optimized electroporation procedure was used to boost the low

doi:10.1016/j.ymthe.2003.12.002

transfection efficiency of the naked DNA vector. Surprisingly, a simple explant culture of the transfected muscle in growth factor-reduced Matrigel gave a reproducible measure of the angiogenic activity of the transferred vascular endothelial growth factor (VEGF165) gene. With a low level of basal angiogenesis, the new method also simplified the analysis of the interaction between different angiogenic genes,

FIG. 2. Expression of the transfected VEGF165 gene and correlation between the amount of VEGF in the medium and the sprouting area around the explanted muscle. The tibialis anterior muscle was injected with 10 Ag of DNA in 30 Al of half-saline solution, followed by electroporation and embedding as described for Fig. 1. (A) Detection of pJDK-vector-specific expression of VEGF165 mRNA. Total RNA from the transfected tibialis anterior muscle was subjected to RT-PCR analysis. () RT indicates the aliquot of RNA from the pJDK-VEGF165-injected muscle that was not subjected to reverse transcription. (B) Level of VEGF in the conditioned medium, as determined by ELISA (R&D) at different time points. The bars represent means F SEM (n = 4). *P < 0.0001 (one-way ANOVA). (C) Calculation of the sprouting area. A total of 15 magnified images were used to calculate the mean area using ImageLab imaging software. (D) Measurement of the sprouting area at the indicated time points. The pJDK-VEGF165 group showed significantly larger areas of capillary networks than the pJDK group. The bars represent means F SEM (n = 4). *P < 0.001 (one-way ANOVA). (E) The x – y plot of the amount of secreted VEGF against the sprouting area on day 7 shows a positive correlation.

466


doi:10.1016/j.ymthe.2003.12.002

VEGF165 and connective tissue growth factor (CTGF), which was confirmed in vivo in a murine model of hindlimb ischemia. This novel assay should have wide application in angiogenesis research because it easily provides valuable information related to the angiogenic capacity of given genes, together with analysis of any synergism or antagonism among multiple factors in the target tissue.

RESULTS Efficient Naked Plasmid/Electroporation-Mediated Gene Transfer into Skeletal Muscle In the development of a gene-based assay, highly efficient gene transfer is a prerequisite for any tangible effect. To optimize electroporation-mediated plasmid delivery into skeletal muscle, we cloned a bacterial CAT reporter gene into the pJDK plasmid vector [28] and evaluated different

METHOD

electroporation parameters in the tibialis anterior muscles of BALB/c mice that received injections of pJDK-CAT (Fig. 1A and B, data not shown). Maximum transfection efficiency with least tissue damage was achieved with the following parameters: voltage-to-distance ratio of 125 V/ cm, pulse duration of 50 ms, 8 pulses, frequency of 1 Hz. The optimized solvent for plasmid DNA was half-saline solution. These optimal conditions were independently identified by two different groups and are supported by our former studies [29,30]. Expression and Secretion of Transfected Gene Product in Explant Cultures We tested whether the transfected skeletal muscle expressed and secreted VEGF165 during explant culture on growth factor-reduced Matrigel. Two days after gene transfer, we carried out RT-PCR analysis of total RNAs from the transfected tibialis anterior muscle. As shown in FIG. 3. Formation of networks of vessel-like structures over time. Sprouting cells from the explanted muscles were observed on the days indicated. In the pJDK-VEGF165-injected group, migration of sprouted cells and formation of vessel-like structures are evident on day 3, with a maximum on day 7. Little endothelial migration or network formation was seen in the control (pJDK) group. Original magnification, 40.


467

METHOD

Fig. 2A, we detected a band corresponding to pJDK-vectorderived VEGF165 mRNA only in pJDK-VEGF165-injected muscle, whereas neither the same RNA sample that had not been reverse transcribed nor the control RNAs (phosphatebuffered saline (PBS), pJDK) produced a band, indicating that transfected VEGF165 was expressed in the muscle. Northern blot analysis also confirmed this result (data not shown). However, VEGF165 mRNA was not detected in the cultured explant on days 3, 5, or 7, suggesting rapid degradation in the culture medium. To confirm the secretion of VEGF, we collected conditioned medium from the explant culture plates on days 3, 5, and 7 and measured the VEGF concentrations by enzyme-linked immunosorbent assay (ELISA). As shown in Fig. 2B, the total amount of VEGF was maintained at 2 – 4.7 ng/ml in the pJDK-VEGF165injected groups for 7 days. These values are well within the known effective concentrations of VEGF, 1 – 10 ng/ml. There was no detectable level of VEGF in the medium of the pJDK-injected group (less than 15.6 pg/ml). This indicates that the transfected muscle secreted enough VEGF for functional activity, whereas there was little expression of VEGF in the mock-transfected samples. In addition to VEGF levels, we monitored any potential angiogenic response in and around the explant (Fig. 2C). Consistent with the increased and sustained expression of VEGF165 in the pJDK-VEGF165-transfected group, there was a corresponding increase in the area of sprouted vessel-like structures (Fig. 2D). However, there was minimal sprouting from the control vector (pJDK)-injected group. We measured the sprouting area at different time points by taking a picture of the whole explant and digitally integrating the sprouted area (Fig. 2C). The mean areas of vessel-like structures from the pJDK-VEGF165 group were 1.5 and 4.7 mm2 on days 3 and 7, respectively. However, the pJDK groups displayed negligible sprouting and the sprouted areas were less than 0.02 mm2, even on day 7 (Fig. 2D). When the amount of VEGF and the extent of sprouting were compared, a positive correlation curve was obtained (Fig. 2E).

Capillary-like Structures and Formation of Networks over Time We next looked closely at the sprouting areas over time. Sprouting and the outward migration of cells from the muscle started on day 2. By day 3, some of the migrating cells began to take the shape of linear tube-like structures, which later connected to neighboring cells to form networks (Fig. 3). The extent of capillary network formation peaked at day 7 and persisted for 2 weeks in the absence of additional growth factors or added medium. We observed little outward sprouting in the control vector (pJDK)-injected group. Although the degree of sprouting was not consistent along the perimeter of the explant, this did not affect the overall assay significantly.

468

doi:10.1016/j.ymthe.2003.12.002

Immunostaining to Identify Endothelial Cells To identify the tube-like structures and sprouted cells, we used immunostaining of the explanted muscle with antibodies that are specific for markers of vessel cells. As shown in Fig. 4, the capillary-like structures from the pJDK-VEGF165-transfected muscle stained positive (dark brown color) for von Willebrand factor (vWF), platelet endothelial cell adhesion molecule (PECAM), and vimentin, indicating that they are endothelial cells. However, the negative staining with a-SMA antibody indicated that no smooth muscle cells were present. In

FIG. 4. Identification of cell types in the vessel-like structures by immunostaining. Sprouting cells were examined for the presence of vWF, vimentin, a-SMA, and PECAM (CD31) markers. In the controls, normal goat serum was used for primary antibody staining. Positive staining (dark brown color) for vWF, PECAM, and vimentin indicated the presence of endothelial cells, whereas negative staining for a-SMA indicated the absence of smooth muscle cells. The cells that were negative for vWF (white arrow in vWF image) and PECAM were positive for vimentin, indicating that these nonendothelial cells were fibroblasts. Original magnification: left, 40; right, 200.


doi:10.1016/j.ymthe.2003.12.002

all controls, normal goat serum was used for primary antibody staining. In some areas, we found cells negative for vWF and PECAM markers around the vWF and PECAM-positive cells (white arrow in the vWF image of Fig. 4). These sprouting cells, however, stained positive for vimentin. Based on the fact that vimentin is present in cells of mesenchymal origin, such as fibroblasts and endothelial cells, we identified these nonendothelial cells as fibroblasts. Interestingly, these putative fibroblasts, which had

METHOD

a more densely packed appearance, did not form capillary-like structures. Effect of Connective Tissue Growth Factor on Angiogenesis To validate the usefulness of the new assay, we tested the angiogenic activity of the CTGF gene, the potential of which as an angiogenic factor is controversial. As shown in Fig. 5A, the Ctgf gene alone appeared to be weakly angiogenic, producing more sprouting cells than the FIG. 5. Attenuation of the angiogenic activity of VEGF165 by CTGF in the ex vivo explant assay. (A) Photograph of muscle explant culture on day 7. (1) pJDK, (2) pJDK-VEGF165, (3) pJDKCTGF, (4) pJDK-VEGF165 + pJDK-CTGF. Tibialis anterior muscles were injected with 20 Ag of DNA in 30 Al of half-saline solution (0.45% NaCl), followed by electroporation and embedding in 0.2 ml of growth-factor-reduced Matrigel. Samples were then covered with 1.5 ml of M199 medium containing 2% FBS in a 24-well plate 2 days after injection. In the pJDKVEGF165 + pJDK-CTGF group, 10 Ag of each vector was combined before injection. In the other groups, 10 Ag of each plasmid was injected together with 10 Ag of the pJDK vector to a final 20 Ag. The areas of sprouting vessels from 15 magnified images were used for quantitative comparisons. Few sprouting vessels were seen in the pJDK and pJDK-CTGF groups, whereas extensive sprouting was observed in the pJDK-VEGF165 group. Modest sprouting was seen in the group treated with both VEGF165 and CTGF. (B) Comparison of the integrated areas of sprouting vessels in each group. The data are means F SEM (n = 4). *P < 0.05 (one-way ANOVA).


469

METHOD

doi:10.1016/j.ymthe.2003.12.002

FIG. 6. Time course of hindlimb blood flow after electroporation of the plasmid-injected muscle. (A) Representative LDI recorded on the days indicated. One day after femoral artery excision, a marked reduction in blood flow was apparent in the left leg, as blue color. Gradual recovery was seen in the group injected with pJDK-VEGF165. The perfusion signal is displayed in color codes ranging from dark blue (0) to white (1000). The far right column, with a green background, shows the gross findings on day 10. The LDI flux ratio correlates with the overall limb loss score. (B) Summary of LDI flux ratios from eight different animals on day 10. The data are means F SEM (n = 8). The total amount of DNA injected was 20 Ag. In the single-gene group, 10 Ag of the control vector (pJDK) was included to a final total of 20 Ag. *P < 0.05 (one-way ANOVA).

470


doi:10.1016/j.ymthe.2003.12.002

control but far less than the VEGF165 gene. Interestingly, however, we observed a significant reduction in the degree of sprouting and capillary network formation in the explant treated with both pJDK-VEGF165 and pJDKCTGF plasmids (Fig. 5). We observed little sprouting in the control vector (pJDK)-injected group, whereas we saw extensive sprouting and network formation in the pJDKVEGF165-injected group, confirming the accuracy of the assay. To confirm this unexpected attenuating activity of CTGF on VEGF, we examined the effect of CTGF on VEGF in a murine model of hindlimb ischemia (Fig. 6). On the day of femoral artery excision, we injected naked plasmids into four different sites on the skeletal muscle, followed by electroporation and daily monitoring. Consistent with the above observation, the LDI flux ratio of the group treated with both pJDK-CTGF and pJDK-VEGF165 was significantly lower than that of the pJDK-VEGF165-treated group on day 10. The LDI flux ratio correlated well with the overall limb loss score. Five of eight mice treated with the control or pJDK-CTGF vector suffered limb necrosis or autoamputation. In the pJDK-VEGF165-injected group, six of eight mice appeared normal, whereas all the mice showed necrosis or autoamputation in the group that received both pJDK-VEGF165 and pJDK-CTGF (Fig. 6).

DISCUSSION In this study, we developed a novel angiogenesis assay that utilizes explant cultures of skeletal muscle transfected with naked plasmid vectors encoding angiogenic genes. This new assay offers, for the first time, a simple means of testing the angiogenic capacity of a particular gene(s). Despite the recent avalanche of studies on therapeutic angiogenesis, methods for evaluating the angiogenic potential of genes of interest in the target muscle have been limited. Therefore, it is timely and proper that a new genebased assay is developed for angiogenic gene therapy studies. Our new assay uses a single injection of naked plasmid and an optimized electroporation procedure to minimize tissue damage, followed by explant culture on growth factor-reduced Matrigel. The assay is validated by the strong correlation between the level of secreted VEGF and the sprouting area (Fig. 2E) and by accurately predicting the behavior in vivo of the controversial angiogenic factor CTGF (Figs. 5 and 6). One of the advantages of the new assay is that quantification is easy, using digital images captured in situ, which allows real-time analysis. Another advantage is that the assay is relatively inexpensive in that it uses a simple naked DNA vector and multiple samples can be prepared from one animal. Because our gene transfer-based assay is performed on skeletal muscle, the results are pertinent to the development of angiogenic gene therapies for ischemic limb diseases. Indeed, there is a growing need for functionally significant angiogenesis in the treatment of peripheral


METHOD

vascular diseases, the clinical outcomes of which have not been satisfactory [31]. Effective gene therapy for such diseases requires an effective screen for angiogenic factors in their natural environments, which would have more biological meaning. Our muscle-based assay should, therefore, be very helpful in evaluating the angiogenic activity of particular factors and their ideal combinations in target tissue, i.e., skeletal muscle. In this regard, it is noteworthy that CTGF behaves differently depending on its circumstances. Initially, CTGF was reported to be an angiogenic factor [32,33], but recent studies have shown that CTGF inhibits the angiogenic activity of VEGF165 [34,35]. Reflecting these controversies, our results indicate that CTGF is a weak angiogenic factor by itself, but significantly inhibits the activity of VEGF 165 when cotransfected (Fig. 5). Our data therefore suggest that the complex nature of vascular responses, which result from interactions between different factors, must be taken into account in angiogenesis research. Angiogenesis assays have been widely used to test angiogenic molecules and their inhibitors. Currently, many assays are available, including the Matrigel plug assay [15], corneal neovascularization assays [11,36], CAM assays [9,10], aortic ring assays [7,8], and chick aortic arch assay [37]. However, current methods for assessing angiogenesis vary greatly among different laboratories. Moreover, most assays frequently lack quantification and are limited in their clinical relevance. They often have high backgrounds, which cause a narrow range of responses between the control and the stimulus. In other aspects, this high basal level of angiogenesis is useful in determining antiangiogenic effects. In our case, we had to change such parameters as serum level in the medium, growth factor levels in the Matrigel, and mouse strain to optimize the assay conditions for proangiogenesis. In medium (M199) containing no serum, explanted muscle was detached from the Matrigel 1 or 2 days after embedding, whereas in medium containing 10% serum, high basal levels of sprouting were observed. This high background was also observed when we used standard Matrigel for explant cultures. As for mice, the C57BL/6 strain showed a higher level of basal angiogenesis than the BALB/c strain, as did younger mice relative to older mice (data not shown). Although our assay was primarily designed for proangiogenic genes, it can also be used for antiangiogenic genes because the results of our assay show that CTGF acts as an inhibitor of VEGF (Fig. 5). Further optimization of the assay for testing antiangiogenic genes and chemicals is currently under way. With regard to the electroporation technique, we examined the distribution and the proportion of transgenepositive cells after electroporation-coupled naked plasmid DNA injection in a previous study with the h-galactosidase reporter gene [30]. Consistent with other studies [38,39], the transfection efficiency (30 – 50%) was very high, with localized expression of the transgene and

471

METHOD

minimal tissue damage (data not shown). The fact that gene expression was observed only in the area of plasmid DNA injection argues against any concerns related to unwanted transfection. The in vivo electroporation technique has also been successful in various tissues in addition to skeletal muscle, such as liver, testis, skin, cornea, cardiovascular, and mammary tumor tissues [40 – 43]. Direct injection of DNA coupled to electroporation thus appears to offer a simple and powerful tool with which to assess the behavior of specific genes in vivo. Although plasmid vectors are the simplest way to introduce exogenous DNA sequences under study, we cannot rule out the possibility that the given plasmid vector will integrate into the host chromosome via either homologous or nonhomologous recombination. It was recently reported that naked DNA vectors are capable of integration in hepatocytes in vivo and that incorporation of adeno-associated virus inverted terminal repeats (AAV ITRs) into the vector seems to enhance the capacity for in vivo integration [44]. Because the pJDK vector used in this study contains the AAV-2 ITRs, further characterization of the genome is required and caution should be exercised in future studies. However, it is worth noting that integration of circular dsDNA was detectable only after selection, compared with the appreciable levels of linear dsDNA vector [44]. A dose – response curve plotting the amount of DNA injected against sprouting area was not achieved with our assay. One reason for this could be that VEGF, unlike reporters like luciferase or bacterial CAT, is a secretory protein and that this contributed to the reduced magnitude of the dose response. Another possible explanation is that there is a narrow range of effective VEGF concentrations, above which there is little additional sprouting despite the increased amounts of DNA injected. Moreover, measurement of the sprouting area is in itself only semiquantitative. Nevertheless, lack of dose response does not compromise the merits of this novel assay, because it can successfully differentiate the angiogenic capacity of different genes under the optimized condition used in this study. One of the caveats of this study is that we have yet to extend the general use of this muscle-based technology to assess the angiogenic capacity of a particular gene in different tissues. Although our new assay was successful in predicting the angiogenic activity of particular genes, such as CTGF, we cannot rule out the possibility that the given angiogenic gene might behave differently in other tissues. Moreover, it is still possible that the performance of a given angiogenic gene in the context of an ex vivo culture assay has little bearing on its functions in vivo. In screening potential angiogenic genes, therefore, the positive results in our assay should be seen as a prerequisite, rather than a determinant, of in vivo efficacy. Another caveat is that the explants were cultured under normoxic conditions. A muscle explant grown under hypoxic con-

472

doi:10.1016/j.ymthe.2003.12.002

ditions could give different results. Currently, the effects of hypoxia on the validity of our assay are under investigation. Despite these qualifications, our new method provides a platform on which different combinations of proangiogenic and antiangiogenic genes can be tested easily. In conclusion, we have developed a novel angiogenesis assay that is simple, quantitative, and reproducible. This new and easy method should facilitate the study of various angiogenic factors together with their inhibitors.

MATERIALS AND METHODS Chemicals. M199 medium (Earle’s salt), penicillin/streptomycin, Fungizone, trypsin/EDTA, and Hanks’ balanced salt solution were purchased from Gibco (Grand Island, NY, USA). PECAM (CD 31) antibody was from B & D Pharmingen (Palo Alto, CA, USA). Vimentin antibody was from Abcam (Cambridge, UK). Antibodies for vWF and a-smooth muscle actin (a-SMA) were purchased from Dako Corp. (Carpinteria, CA, USA). Plasmid DNA. Plasmid DNA was amplified in TOP 10 competent cells (Invitrogen, Carlsbad, CA, USA) and purified with a Mega Plasmid Purification Kit from Qiagen (Valencia, CA, USA) according to the manufacturer’s instructions. The plasmid pellets were washed twice with 70% (v/v) ethanol and resuspended in water. The plasmid was stored at 20jC until use. The pJDK plasmid was derived from the ACP plasmid [27] by replacing the backbone containing the AmpR gene (PvuII fragment) with one containing the KanR gene (HincII/XcmI) from the pVAX1 plasmid (Invitrogen) using blunt ligation. The cDNAs for CTGF and VEGF165 were cloned into the EcoRI/XhoI and EcoRI sites of the pJDK plasmid, respectively. Intramuscular DNA injection and electroporation. All animal experiments conform to the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH Publication No. 85-23, revised 1996). Twelve-week-old male BALB/c mice were anesthetized by intraperitoneal injection of 100 Al of solution containing 2.215 mg ketamine (Ketalar 50 mg/ml; Yuhan Co., Korea) and 0.175 mg xylazine (Rompun 23.32 mg/ml; Bayer, Korea). The leg was shaved and depilated to expose the tibialis anterior muscle. As shown in Fig. 1A, 10 lg of DNA in 30 ll of half-saline solution (0.4% NaCl or 75 mM) was injected into the tibialis anterior muscle of each mouse with a 30-gauge insulin syringe (Becton – Dickinson, Franklin Lakes, NJ, USA). The syringe needle was surrounded with P10 tubing to ensure an even depth of injection of 3.0 mm. Thirty seconds after DNA injection, transcutaneous electric pulses were applied to the surface of the injection site (Fig. 1B) using an ECM 830 electroporator (BTX Division of Genetronics, Inc., San Diego, CA, USA). Eight electric pulses of 125 V/cm for 50 ms at 1 Hz were applied through two electrodes (Tweezertrodes, BTX Division of Genetronics, Inc.). The mice were humanely killed 2 days after electroporation to retrieve the muscle. Explant culture of skeletal muscle on Matrigel. The muscle was cut in the middle to expose the injection area and washed three times in 3 ml of PBS to remove debris and blood (Fig. 1C). The washed muscle was placed in a 24-well plate containing 200 Al of growth factor-reduced Matrigel (Collaborative Biomedical Products, Becton – Dickinson, Bedford, MA, USA) and incubated at 37jC for 30 min to solidify the gel (Fig. 1D). The plate was then covered with 1.5 ml of M199 medium containing 2% heat-inactivated fetal bovine serum, 100 U/ml penicillin, and 100 Ag/ml streptomycin. The plate was cultured at 37jC under 5% CO2. Outgrowth of capillary-like structures was observed in specimens with an inverted microscope (Zeiss, Oberkochen, Germany) equipped with a digital photography system. The mean area of microvessels was


METHOD

doi:10.1016/j.ymthe.2003.12.002

measured by a 1300 1030 pixel image captured with a digital CCD camera (AxioCAM, Zeiss) and quantified using ImageLab imaging software (MCM Design, Birkeroed, Denmark). Hindlimb ischemia model and laser Doppler imaging. Twelve-week-old male Balb/c mice were anesthetized, shaved, and depilated as described above. The left femoral artery was completely excised at day 0, and naked plasmid DNA (a total of 20 Ag in 40 Al of half-saline solution) was injected into four different skeletal muscle sites (vastus medialis, adductor magnus). After suturing, the injected muscle was subjected to electroporation. Blood flow in both ischemic and nonischemic hindlimbs was measured using an LDI system (Moor Instruments, Axminster, Devon, England) until day 10. The perfusion signal was displayed in color codes ranging from dark blue (0) through red to white (1000). Blood flow in the ischemic limb was normalized to that of the contralateral nonischemic hindlimb. The mice were also observed for gross changes, including limb necrosis or autoamputation. Immunostaining. The explants were observed daily by inverted microscopy. For immunohistochemical analyses, the explants were fixed with 4% (w/v) paraformaldehyde (Sigma, St. Louis, MO, USA) at room temperature for 10 min. To quench endogenous peroxidase activity, the explants were immersed in PBS containing 0.1% H2O2 for 15 min at room temperature and then washed three times in PBS. The explants were then incubated in PBS containing 10% (v/v) normal goat serum in 0.5% Triton X-100 for 30 min at room temperature. Specific antibody (rat anti-mouse PECAM antibody, rabbit anti-human von Willebrand factor antibody, rabbit anti-mouse vimentin antibody, mouse anti-a-SMA antibody) that had been diluted to a ratio of 1:200 in PBS containing 1% normal goat serum and 0.5% (v/v) Triton X-100 was added and incubated overnight at 4C. For the secondary antibody, biotinylated anti-rabbit IgG (1:200; Jackson Laboratories, West Grove, PA, USA), anti-rat IgG (1:200; Dako, Copenhagen, Denmark), or anti-mouse IgG (1:200; Jackson Laboratories) was applied for 30 min at room temperature, after which the samples were washed three times in PBS. Samples were stained according to the manufacturer’s protocols (Elite Kit, Vector Laboratories, Burlingame, CA, USA). ELISA. The amount of secreted VEGF in the conditioned medium of explant cultures was measured with the human VEGF ELISA kit (R & D, Minneapolis, MN, USA), according to the manufacturer’s instructions. The conditioned medium was sampled at each time point (days 3, 5, and 7) and stored at 20C in microtubes until use. Good linear correlation was observed with standards in the range between 5.6 and 1000 pg/ml. RT-PCR. Total RNA was extracted from the tibialis anterior muscle using the Qiagen RNeasy Mini Kit. RNA (2 lg) was first treated with DNase I and then reverse transcribed using oligo(dT)15 primer and SuperScript II reverse transcriptase (Invitrogen) at 42C for 50 min. The cDNA was amplified with 35 cycles of PCR at an annealing temperature of 53C using primers specific for human VEGF165 mRNA that was transcribed from pJDK-hVEGF165: sense, 5V-TCGCCCTTATGAACTTTCTG-3V; antisense, 5V-ACAACAGATGGCTGGCAAC-3V. The sense primer contained the start codon of VEGF165 (bold characters) and the antisense primer hybridized to the noncoding region downstream of the poly(A) signal sequence of the vector. A second aliquot of the RT reaction was amplified for 25 cycles at an annealing temperature of 60jC using primers specific for the h-actin gene: sense, 5V-CATGTTTGAGACCTTCAACA-3V; antisense, 5V-ATCTCCTTCAGCATCCTGTC-3V. Statistical analysis. The data are expressed as means F SEM. Results were analyzed with GraphPad Prism statistics software (GraphPad Software, Inc., San Diego, CA, USA). One-way analysis of variance followed by Bonferroni’s post hoc multiple comparison tests were used to evaluate statistical differences between the groups. A P value of less than 0.05 was considered statistically significant.


ACKNOWLEDGMENTS This work was supported by National Research Laboratory grants from the Korea Institute of Science and Technology Evaluation and Planning (M1-020300-0048), the Korean Ministry of Health and Welfare (01-PJ1-PG1-01CH060003), and Science Research Center grants from the Korea Science and Engineering Foundation to D.-K.K. RECEIVED FOR PUBLICATION MAY 12, 2003; ACCEPTED DECEMBER 4, 2003.

REFERENCES 1. Rissanen, T. T., et al. (2003). Fibroblast growth factor 4 induces vascular permeability, angiogenesis and arteriogenesis in a rabbit hindlimb ischemia model. FASEB J. 17: 100 – 102. 2. Sasaki, K., et al. (2002). Evidence for the importance of angiotensin II type 1 receptor in ischemia-induced angiogenesis. J. Clin. Invest. 109: 603 – 611. 3. Taniyama, Y., et al. (2001). Therapeutic angiogenesis induced by human hepatocyte growth factor gene in rat diabetic hind limb ischemia model: molecular mechanisms of delayed angiogenesis in diabetes. Circulation 104: 2344 – 2350. 4. Carmeliet, P. (2000). Mechanisms of angiogenesis and arteriogenesis. Nat. Med. 6: 389 – 395. 5. Folkman, J., and Haudenschild, C. (1980). Angiogenesis in vitro. Nature 288: 551 – 556. 6. Risau, W. (1997). Mechanisms of angiogenesis. Nature 386: 671 – 674. 7. Klagsbrun, M., and D’Amore, P. A. (1991). Regulators of angiogenesis. Annu. Rev. Physiol. 53: 216 – 239. 8. Jain, R. K., Schlenger, K., Hockel, M., and Yuan, F. (1997). Quantitative angiogenesis assays: progress and problems. Nat. Med. 3: 1203 – 1208. 9. Ribatti, D., et al. (1997). New model for the study of angiogenesis and antiangiogenesis in the chick embryo chorioallantoic membrane: the gelatin sponge/chorioallantoic membrane assay. J. Vasc. Res. 34: 455 – 463. 10. Nguyen, M., Shing, Y., and Folkman, J. (1994). Quantitation of angiogenesis and antiangiogenesis in the chick embryo chorioallantoic membrane. Microvasc. Res. 47: 31 – 40. 11. Muthukkaruppan, V. R., and Auerbach, R. (1979). Angiogenesis in the mouse cornea. Science 205: 1416 – 1418. 12. Gimbrone, M. A., Jr., Leapman, S. B., Cotran, R. S., and Folkman, J. (1994). Tumor dormancy in vivo by prevention of neovascularization. J. Exp. Med. 136: 261 – 276. 13. Schor, S. L., Schor, A. M., Winn, B., and Rushton, G. (1982). The use of 3D collagen gels for the study of tumour cell invasion in vitro: experimental parameters influencing cell migration into the gel matrix. Int. J. Cancer 29: 57 – 62. 14. Madri, J. A., Pratt, B. M., and Tucker, A. M. (1988). Phenotypic modulation of endothelial cell by transforming growth factor-h depends upon the composition and organization of the extracellular matrix. J. Cell Biol. 106: 1375 – 1384. 15. Passaniti, A., et al. (1982). A simple, quantitative method for assessing angiogenesis and antiangiogenic agents using reconstituted basement membrane, heparin and fibroblast growth factor. Lab. Invest. 67: 519 – 528. 16. Taylor, P. C. (2002). VEGF and imaging of vessels in rheumatoid arthritis. Arthritis Res. 4(Suppl. 3): S99 – S107. 17. Wolff, J. A., et al. (1990). Direct-gene transfer into mouse muscle in vivo. Science 247: 1465 – 1468. 18. Li, S., and Benninger, M. (2002). Applications of muscle electroporation gene therapy. Curr. Gene Ther. 2: 101 – 105. 19. Fewell, J. G., et al. (2001). Gene therapy for the treatment of hemophilia B using PINCformulated plasmid delivered to muscle with electroporation. Mol. Ther. 3: 574 – 583. 20. Hartikka, J., et al. (2001). Electroporation-facilitated delivery of plasmid DNA in skeletal muscle: plasmid dependence of muscle damage and effect of poloxamer 188. Mol. Ther. 4: 407 – 415. 21. Nicosia, R. F., and Ottinetti, A. (1990). Growth of microvessels in serum-free matrix culture of rat aorta. A quantitative assay of angiogenesis in vitro. Lab. Invest. 63: 115 – 122. 22. Aihara, H., and Miyazaki, J. I. (1999). Gene transfer into muscle by electroporation. Nat. Biotechnol. 16: 867 – 870. 23. Lucas, M. L., and Heller, R. (2001). Immunomodulation by electrically enhanced delivery of plasmid DNA encoding IL-12 to murine skeletal muscle. Mol. Ther. 3: 47 – 53. 24. 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. 25. Muramatsu, T., et al. (2001). In vivo gene electroporation in skeletal muscle with special reference to the duration of gene expression. Int. J. Mol. Med. 7: 37 – 42. 26. MacColl, G. S., Goldspink, G., and Bouloux, P. M. G. (1999). Using skeletal muscle as an artificial endocrine tissue. J. Endocrinol. 162: 1 – 9. 27. Byun, J., et al. (2001). Efficient expression of the vascular endothelial growth factor gene in vitro and in vivo, using an adeno-associated virus vector. J. Mol. Cell. Cardiol. 33: 295 – 305. 28. Kim, D.K., and Byun, J. (2002). The pJDK vector contains the human cytomegalovirus enhancer/promoter-driven expression cassette within the inverted terminal repeats of AAV-2. Patent pending in Korea, 10-2002-0008232.

473

METHOD

29. Muramatsu, T., Nakamura, A., and Park, H. M. (1998). In vivo electroporation: a powerful and convenient means of nonviral gene transfer to tissues of living animals. Int. J. Mol. Med. 1: 55 – 62. 30. Lee, M. J., et al. (2002). Optimal salt concentration of vehicle for plasmid DNA enhances gene transfer mediated by electroporation. Exp. Mol. Med. 34: 265 – 272. 31. Khan, T. A., Sellke, F. W., and Laham, R. J. (2003). Gene therapy progress and prospects: therapeutic angiogenesis for limb and myocardial ischemia. Gene Ther. 10: 285 – 291. 32. Shimo, T., et al. (2001). Connective tissue growth factor as a major angiogenic agent that is induced by hypoxia in a human breast cancer cell line. Cancer Lett. 174: 57 – 64. 33. Shimo, T., et al. (1999). Connective tissue growth factor induces the proliferation, migration, and tube formation of vascular endothelial cells in vitro, and angiogenesis in vivo. J. Biochem. (Tokyo) 126: 137 – 145. 34. Inoki, I., et al. (2002). Connective tissue growth factor binds vascular endothelial growth factor (VEGF) and inhibits VEGF-induced angiogenesis. FASEB J. 16: 219 – 222. 35. Hashimoto, G., Inoki, I., Fujii, Y., Aoki, T., Ikeda, E., and Okada, Y. (2002). Matrix metalloproteinases cleave connective tissue growth factor and reactivate angiogenic activity of vascular endothelial growth factor 165. J. Biol. Chem. 277: 36288 – 36295. 36. Muthukkaruppan, V. R., Kubai, L., Auerbach, R. (1982). Tumor induced neovascularization in the mouse eye. J. Natl. Cancer Inst. 69: 699 – 708.

474

doi:10.1016/j.ymthe.2003.12.002

37. Muthukkaruppan, V. R., Shinners, B. L., Lewis, R. L., and Auerbach, R. (2000). The chick embryo aortic arch assay: a new, rapid, quantifiable in vitro method for testing the efficacy of angiogenic and anti-angiogenic factors in a three-dimensional, serum free organ culture system. Proc. Am. Assoc. Cancer Res. 41: 65. 38. Draghia-Akli, R., Ellis, K. M., Hill, L. A., Malone, P. B., and Fiorotto, M. L. (2003). Highefficiency growth hormone-releasing hormone plasmid vector administration into skeletal muscle mediated by electroporation in pigs. FASEB J. 17: 526 – 528. 39. Martinenghi, S., et al. (2002). Human insulin production and amelioration of diabetes in mice by electrotransfer-enhanced plasmid DNA gene transfer to the skeletal muscle. Gene Ther. 9: 1429 – 1437. 40. Blair-Parks, K., Weston, B. C., and Dean, D. A. (2002). High-level gene transfer to the cornea using electroporation. J. Gene Med. 4: 92 – 100. 41. Wells, J. M., Li, L. H., Sen, A., Jahreis, G. P., and Hui, S. W. (2002). Electroporationenhanced gene delivery in mammary tumors. Gene Ther. 7: 541 – 547. 42. Maruyama, H., Ataka, K., Higuchi, N., Sakamoto, F., Gejyo, F., and Miyazaki, J. (2001). Skin-targeted gene transfer using in vivo electroporation. Gene Ther. 8: 1808 – 1812. 43. Suzuki, T., Shin, B. C., Fujikura, K., Matsuzaki, T., and Takata, K. (1998). Direct gene transfer into rat liver cells by in vivo electroporation. FEBS Lett. 425: 436 – 440. 44. Nakai, H., et al. (2003). Helper-independent and AAV-ITR-independent chromosomal integration of double-stranded linear DNA vector in mice. Mol. Ther. 7: 101 – 111.
