Nonviral Jet-Injection Gene Transfer for Efficient in Vivo Cytosine ...

ARTICLE

doi:10.1016/j.ymthe.2005.07.700

Nonviral Jet-Injection Gene Transfer for Efficient in Vivo Cytosine Deaminase Suicide Gene Therapy of Colon Carcinoma Wolfgang Walther,1,* Ulrike Stein,1 Iduna Fichtner,1 Dennis Kobelt,2 Jutta Aumann,2 Franziska Arlt,2 and Peter M. Schlag2 1

Max-Delbru¨ck-Center for Molecular Medicine, Robert-Ro¨ssle-Strasse 10, 13092 Berlin, Germany 2 Robert-Ro¨ssle-Clinic, Charite´, Lindenberger Weg 80, 13125 Berlin, Germany

*To whom correspondence and reprint requests should be addressed. Fax: +49 30 9406 2780. E-mail: [email protected].

Available online 3 October 2005

Jet-injection technology has developed into an efficient gene delivery system for nonviral in vivo gene transfer. In this study the jet-injector system was used for the intratumoral gene transfer of small volumes of naked DNA encoding the Escherichia coli cytosine deaminase (CD) suicide gene. In our in vivo studies human colon carcinoma (patient-derived tumor model Colo5734 and SW480 colon carcinoma)-bearing NMRI-nu/nu male mice received four jet injections (10 Ml per injection) of the CD-gene-carrying plasmid, representing 40 Mg plasmid DNA per animal. Forty-eight hours after jet injection, treatment of tumors with 5-fluorocytosine (5-FC; 500 mg/kg ip) was started and during treatment tumor volumes were measured. Starting from day 5 of 5-FC treatment inhibition of tumor growth was seen in the CD-gene-transduced tumors compared to the respective control groups, which lasted for the entire observation time. Expression analysis at the mRNA and protein levels revealed efficient expression of the CD gene in the jet-injected tumors. Therefore, in this in vivo study jet-injection gene transfer of 40 Mg CD-expressing naked plasmid DNA leads to a significant tumor growth inhibition. This study demonstrates the applicability of the jet-injection technology for in vivo gene transfer into tumors to achieve efficient tumor gene therapy. Key Words: suicide gene, cytosine deaminase, jet-injection, nonviral gene therapy, colon cancer

INTRODUCTION The transfer of naked DNA for gene therapy is gaining increasing attention for various gene therapy applications and has advantages in terms of ease of DNA preparation, simplicity of use, and lack of immunogenicity [1–3]. Based on this, different physical methods have been developed to deliver naked DNA into the desired cells or tissues in vitro and in vivo, such as needle and syringe injection, ultrasound and hydrodynamics pressure, particle bombardment (gene gun), in vivo electroporation, or jet-injection [4–9]. Among the various nonviral gene delivery technologies jet-injection has developed into an applicable method, allowing gene transfer into different tissues [10,11]. The jet-injection technology is based on jets of high velocity (N300 m/s) possessing the required energy to penetrate skin and underlying tissues, leading to transfection of the targeted areas [12]. In previous studies a jet-injector prototype (Swiss Injector) was tested for efficient and reliable in vivo gene transfer using a hgalactosidase (LacZ) reporter gene construct, green fluo-

1176

rescent protein-expressing plasmid, or tumor necrosis factor a gene-expressing vector [12,13]. The key parameters of in vivo jet-injection such as jet-injection volume, pressure, depth of jet penetration in the tumor tissue, and DNA stability have been defined for optimized nonviral gene therapy. A high percentage of experimental and clinical cancer gene therapy studies is based on the suicide/enzymeprodrug concept. Among the various suicide systems employed, Escherichia coli cytosine deaminase (CD) is of particular interest if treatment of colon cancer is envisaged, since the CD gene converts the nontoxic 5-fluorocytosine (5-FC) into the toxic metabolite 5-fluorouracil (5FU). 5-FU is known as one of the most effective cytostatic drugs for the treatment of colorectal cancer; however, its systemic use is often limited by gastrointestinal and hematological toxicity [14]. CD suicide gene therapy could circumvent such systemic toxicity by the local conversion of 5-FC into 5-FU. Numerous in vitro and in vivo studies have described CD gene transfer with varying therapeutic efficacy in different tumor models [15,16]. In

MOLECULAR THERAPY Vol. 12, No. 6, December 2005 Copyright C The American Society of Gene Therapy 1525-0016/$30.00

doi:10.1016/j.ymthe.2005.07.700

ARTICLE

FIG. 1. Schematic representation of the (A) application schedule of jet-injection of cytosine deaminase (CD)expressing pCMV-CD plasmid and the treatment with 5-fluorocytosine (5-FC) and (B) the direction of intratumoral jet-injection. (A) Tumors were jet-injected with the pCMV-h plasmid, indicated as day 0 of treatment. At the termination of the in vivo experiment the tumors were removed and series of cryosections were generated for detailed analysis of distribution of CD expression at mRNA and protein levels (see Materials and Methods). (B) The jetinjection was applied parallel to the longitudinal axis of the tumor. The arrows indicate the orientation of the consecutive cryosections made from the tumor for the analyses of pCMV-CD DNA distribution and CD expression in the tumor tissue.

MOLECULAR THERAPY Vol. 12, No. 6, December 2005 Copyright C The American Society of Gene Therapy

1177

ARTICLE

doi:10.1016/j.ymthe.2005.07.700

the majority of these studies either transduced cell lines or adenovirus-mediated CD-gene transfer was used and only one report describes the transfer of naked CD-geneexpressing vector DNA for in vivo treatment of hepatic metastases of colon cancer [17–20]. By contrast, intratumoral jet-injection gene transfer for in vivo CD suicide gene therapy has not been described. The present novel protocol was performed for testing the Swiss Injector prototype for efficient in vivo gene transfer of the cytosine deaminase suicide gene into xenotransplanted human colon carcinoma. It was of major interest to evaluate whether a single application of naked plasmid DNA is sufficient to express the transgene efficiently in vivo and if this level and duration of CD expression will cause a reduction in tumor growth in vivo. The in vivo data presented here reveal for the first time the feasibility of jet-injection technology for efficient suicide tumor gene therapy.

RESULTS Analysis of CD Expression in pCMV-CD Jet-Injected Tumors In this study we used the Swiss Injector for gene transfer of naked pCMV-CD DNA into the tumors. To evaluate if this method can efficiently transduce the targeted tumor tissue, we analyzed the CD protein expression in xenotransplanted Colo5734 tumors 28 days after the jetinjection gene transfer (see Fig. 1A). These tumors originated from the animals that are also shown in Fig. 5B for the therapeutic in vivo experiment. The Western blot shown in Fig. 2 depicts the results obtained with the lysates from pCMV-CD jet-injected Colo5734 tumors treated with either saline (Fig. 2A) or 5FC (Fig. 2B). As shown, a specific band for the bacterial CD protein is detectable in both animal groups, but with varying levels of CD expression. This is a clear demonstration that the single intratumoral jet-injection of 40 Ag of naked pCMV-CD DNA leads to the persisting and efficient expression of this suicide gene in the tumor tissues. To determine the distribution of CD expression within the individual tumors, we analyzed eight different fractions (see Figs. 1A and 1B) of each tumor for CD protein in the Western blot. We prepared these fractions following the longitudinal axis of the respective tumor (Fig. 1B). Figs. 2C and 2D show the distribution of CD protein in the eight representative tumor fractions of tumor B2 (saline treated) and tumor C5 (5-FC treated). The analysis revealed that throughout the entire tumor tissue CD protein is detectable, although at varying extents regarding the expression level. In fact, this is an indication that jet-injection-mediated CD-gene transfer leads to a widespread distribution of CD expression. Moreover, it was apparent that, particularly in the lysates of the 5-FC-treated tumors, CD expression was lower compared to the saline-treated group. This could be due to

1178

FIG. 2. Western blot analysis of CD expression in patient-derived xenotransplanted human colon carcinoma Colo5734 jet-injected with pCMV-CD. (A) CD expression detected in tumor lysates jet-injected with pCMV-CD and treated with saline. (B) CD expression in tumor lysates jet-injected with pCMVCD and treated with 5-FC. (C and D) The actual distribution of CD expression from those representative tumors marked with a rectangle, tumor B2 of the saline-treated group in (A) and tumor C5 of the 5-FC-treated group in (B). Lanes 1 to 8 represent the series of eight cryosections of the respective tumor (see Fig. 1). The control tumor lysates were prepared from nontransduced tumor tissues.

the 5-FC treatment-mediated elimination of CD-expressing portions of the respective tumor tissues. We saw similar CD-expression characteristics in the Western blot analysis of the second SW480 tumor model, in which 5FC-treated tumors showed lower CD expression compared to the untreated group (Figs. 6A and 6B). Examination of the location of CD expression within the jet-injected tumor tissue by immunohistochemistry supported the results of Western blot analysis. In all pCMV-CD jetinjected tissue samples we found scattered expression of the CD protein using a CD-specific antibody (Fig. 3). Interestingly, particularly in the 5-FC-treated animal


ARTICLE

doi:10.1016/j.ymthe.2005.07.700

after the jet-injection, CD mRNA expression was clearly measurable in both pCMV-CD jet-injected saline-treated and 5-FC-treated tumors. Time Course of Biodistribution and Clearance of Jet-Injected pCMV-CD Plasmid DNA and CD Expression To evaluate the time course of pCMV-CD plasmid DNA distribution and CD expression after jet-injection in more detail, we analyzed tumor, skin, liver, and kidney at different time points (1, 2, 4, 12, 23 days) after jet-injection. Fig. 4A shows the presence of the jet-injected pCMV-CD DNA at all time points. The quantitative analysis by realtime PCR demonstrated the time-dependent decline in pCMV-CD DNA amounts in the tumor from 1.46 ng plasmid DNA 1 day after jet-injection to 0.004 ng 23 days after jet-injection (Fig. 4A). Analysis of the time course of CD expression by Western blot, however, revealed the continuous expression of the CD protein in the tumors for the entire observation time (Fig. 4B). The analysis of biodistribution of the jet-injected plasmid DNA demonstrated minimal appearance in skin, liver, and kidney, which ranged from only 0.19 to 0.01 pg (Table 2). However, despite the presence of plasmid DNA we detected no expression in these organs by real-time RTPCR (not shown). Jet-Injection of pCMV-CD DNA for Inhibition of Tumor Growth To test whether the jet-injection gene transfer of naked pCMV-CD DNA would have an influence the tumor

FIG. 3. Immunohistochemistry of pCMV-CD jet-injected and uninjected patient-derived xenotransplanted human colon carcinoma Colo5734. CD expression in the jet-injected tumor tissues is detected by brown staining of the respective areas. (A and B) The immunohistochemistry in saline jetinjected tumor tissue of a control animal. (C and D) The immunohistochemistry of a pCMV-CD jet-injected, but not 5-FC-treated, tumor. (E and F) The immunohistochemistry of a pCMV-CD jet-injected and 5-FC-treated tumor. The arrows indicate the CD-specific staining. Original magnification: (A, C, and E) 40-fold, (B, D, and F) 100-fold.

group, the areas of positive CD protein staining were often in close vicinity to those parts of the tissue that were affected by necrotic processes. To support the results obtained at the protein level, we also performed CD-expression analysis at the mRNA level using quantitative real-time RT-PCR. The results of the quantitative analysis shown in Table 1 support the fact of persisting CD expression in the jet-injected tumor tissues. Furthermore, we also observed a lower CD-expression level in the 5-FC-treated tumors, which supports our assumption of treatment-related loss of CD-expression areas in the respective tumors. It is noteworthy that, although the RNA was prepared from the tumors 28 days


TABLE 1: Quantitative analysis of CD mRNA expression in pCMV-CD jet-injected tumors by real-time RT-PCR (LightCycler) Animal group pCMV-CD + saline

Mean of B tumors (n = 8) pCMV-CD + 5FC

Mean of C tumors (n = 8)

Tumor

Relative CD expression

B1 B2 B3 B4 B5 B6 B7 B8

16.5 0.73 0.54 7.6 7.8 0.37 0.51 0.24 4.28 0.25 1.1 0.22 0.27 0.88 1.17 0.35 0.36 0.57

C1 C2 C3 C4 C5 C6 C7 C8

The ratio of CD (mean of the duplicates) to G6PDH (mean of the duplicates) for each tumor sample was calculated and expressed as relative CD mRNA expression (CD/G6PDH). SD of all values shown is b10%.

1179

ARTICLE

doi:10.1016/j.ymthe.2005.07.700

FIG. 4. Analysis of time-dependent presence of jet-injected pCMV-CD plasmid-DNA and CD protein expression in tumors. (A) Top: Detection of pCMV-CD plasmid DNA in tumors 1, 2, 4, 12, and 23 days after jet-injection by agarose gel electrophoresis. Bottom: Quantitation of pCMV-CD plasmid DNA by real-time PCR. The amount of plasmid DNA was determined in 250 ng tumor DNA using a "spiked" standard curve (see Materials and Methods). (B) Corresponding Western blot analysis of CD expression in the tumors 1, 2, 4, 12, and 23 days after jet-injection of pCMV-CD. For each time point tumors and tissues of two animals were analyzed.

growth in vivo, in a first experimental setting we jetinjected Colo5734-bearing mice with a total dose of 40 Ag per animal of this DNA in two independent experiments (Figs. 5A and 5B). Forty-eight hours after gene transfer we started 5-FC treatment as described under Materials and Methods. As shown in Fig. 5A, the nontransduced Colo5734 tumor per se is sensitive toward the cytostatic drug 5-FU, the toxic metabolite of 5-FC. Therefore, this control group demonstrated the usefulness of the Colo5734 colon carcinoma model for in vivo testing of our suicide approach. In both therapeutic experiments jet-injection of control animals with PBS had no inhibitory effect on the tumor growth, showing growth behavior similar to that of uninjected tumors. Animals jet-injected with the empty vector construct also showed no growth inhibition, even if they were treated with 5-FC (Fig. 5B). However, jet injection of the pCMV-CD plasmid exerted slight,

but not significant ( P N 0.05), reduction of tumor growth in both experiments, in which these animals received PBS treatment after the jet-injection. By contrast, treatment of pCMV-CD jet-injected tumors with the prodrug 5-FC led to the significant ( P = 0.0412 and P = 0.0156) suppression of tumor growth in the two experiments. This growth inhibitory effect started as early as 4 to 5 days after the initiation of 5-FC treatment and lasted for the entire observation time in both animal experiments. The observed suppression in tumor growth led to a 70–80% reduction of the tumor volumes in the jet-injected, 5-FC-treated animals compared to the untreated, nontransduced control tumors. In a second tumor model of xenotransplanted SW480 human colon carcinoma, we tested the nonviral approach of jet-injection suicide gene transfer using the same treatment schedule as for the Colo5734 tumors. Paralleling the results obtained with the Colo5734 tumor model, SW480 tumors jet-injected with the CD-expressing vector also showed significant reduction in their growth after 5-FC treatment (Fig. 6C, P = 0.026). Similar to the previous tumor model, jet-injected but nontransduced tumors did not show reduction in tumor growth. More interestingly, also in this tumor model the CD-expressing 5-FC-treated tumors showed a better reduction in tumor growth compared to the 5-FUtreated control tumors. These observations in the SW480 tumor model support the applicability and reliability of the nonviral jet-injection for efficient suicide gene therapy. Determination of animal body weight revealed no changes in this parameter during the entire treatment in both tumor models (data not shown). Taken together, these in vivo experiments indicate the reproducible antitumor efficacy of a single in vivo application of the CDexpressing naked pCMV-CD DNA by jet-injection to achieve a therapeutic effect.

DISCUSSION Several in vivo studies have shown that the needle-less jetinjection is efficient for gene transfer of naked DNA into

TABLE 2: Quantitative real-time PCR (LightCycler) analysis of pCMV-CD plasmid DNA distribution in skin, liver, and kidney 1, 2, 4, 12, and 23 days after intratumoral jet-injection Time after jet-injection

Skin

1 day 2 days 4 days 12 days 23 days

0.19 0.10 n.d. 0.40 0.06

Amount of pCMV-CD plasmid DNA in pg/250 ng tissue DNA (mean) Liver Kidney 0.05 0.03 0.19 0.24 0.07

0.08 0.04 0.01 0.03 0.01

The amount of plasmid DNA was determined in 250 ng tissue DNA and quantified using a "spiked" standard plasmid DNA curve (see Materials and Methods).

1180


doi:10.1016/j.ymthe.2005.07.700

ARTICLE

FIG. 5. Influence of pCMV-CD jet-injection on tumor growth. (A and B) Xenotransplanted patient-derived human colon carcinoma Colo5734 was jet-injected once at day 0 of treatment with naked pCMV-CD vector DNA in two independent experiments. 48 h after jet-injection the animals were treated daily either with saline or with 5-FC. Nontransduced tumor-bearing animals served as controls. In addition, nontransduced control animals that were treated with 5-FU are shown (see A). Each group of the experiment shown in (A) consists of n = 4 animals; in (B) each group represents n = 8 animals. P values are indicated by asterisks: for (A) *P N 0.05, **P = 0.0412, and for (B) *P N 0.05, **P = 0.0156.

somatic tissues, is useful for genetic immunization/DNA vaccination, and is effective for direct gene transfer into tumors [10,12,21,24,25]. In the majority of the earlier studies reporter gene constructs were delivered for intratumoral gene transfer [12,13]. In this report, for the first time in vivo jet-injection was employed to transduce the cytosine deaminase suicide gene-expressing pCMV-CD construct to demonstrate the therapeutic efficiency of this technology. This report is of particular value, since our jetinjection-based gene transfer diminishes the amount of DNA needed for the efficient expression of the transgene. We showed that a single intratumoral in vivo application of 40 Ag of naked pCMV-CD DNA leads to long-lasting expression of the CD mRNA and protein in two colon carcinoma models. In our earlier reports we have shown that 120 h after jet-injection expression of the transgene is detectable [12]. In this report, however, for a period of 16 to 28 days after jet-injection CD expression was seen in the Colo5734 and SW480 tumors. The level of CD expression in the jet-injected tumor tissues was shown to be sufficient


to generate tumor growth suppression in 5-FC-treated animals. The analysis of distribution of CD expression showed an inhomogeneous pattern in the jet-injected tumor due to the physical properties of the jet-injection procedure, which is in agreement with our previous findings regarding the scattered distribution of green fluorescence or h-galactosidase gene expression in different in vivo tumor models [12,13,22]. The Western blot analyses of the consecutive tumor fractions, however, indicated that the CD expression affects the entire tumor, since almost every tumor fraction showed CD expression. The analysis of the time course of the CD expression revealed the early appearance of the CD protein 24 h after jet-injection and the persistence of gene expression during the entire observation time. Although jet-injection permits inhomogeneous scattered CD gene transfer, the known bystander effect could contribute to an increase in therapeutic efficacy [23]. Particularly, our observation by immunohisto-

1181

ARTICLE

doi:10.1016/j.ymthe.2005.07.700

The major finding of this study, however, is the demonstration in two different tumor models of the therapeutic effectiveness of jet-injection gene transfer by a single application of the pCMV-CD DNA. Although in various approaches adenovirus vectors have shown their effectiveness for suicide gene transfer, our nonviral approach has demonstrated its applicability with comparable therapeutic effects in vivo [17–19]. In addition, adenovirus-associated problems, such as human pathogenicity or unwanted host immune reaction toward viral proteins, could be circumvented by using nonviral gene transfer such as jet-injection. With respect to the potential clinical application of gene therapy for cancer patients it is of importance to deliver the transgene safely into the target tissue with minimal DNA load. We therefore favor direct local intratumoral jet-injection gene delivery as an effective and safe option. We conclude that the jet-injection gene transfer technology can be performed very efficiently by the single administration of the CD-expressing construct and possesses the potential for clinical application.

MATERIALS

FIG. 6. (A and B) Analysis of CD expression by Western blot and (C) determination of tumor growth in pCMV-CD jet-injected SW480 tumors. The xenotransplanted human colon carcinoma SW480 was jet-injected once at day 0 of treatment with naked pCMV-CD vector DNA. 48 h after jet-injection the animals were treated daily either with saline or with 5-FC. Nontransduced (jetinjection with saline) tumor-bearing animals served as controls. In addition, nontransduced control animals that were treated with 5-FU are shown. Each group of the experiment consisted of n = 6 animals. Lanes marked by asterisks in (A) represent lysates from large necrotic tumors, in which no CD expression could be detected. P values in (C) are indicated by asterisks: *P N 0.05, **P = 0.026.

chemistry that in the vicinity of CD-positive stained areas of jet-injected and 5-FC-treated tumors necrotic lesions were found supports this assumption. In addition, we observed a lower CD-expression level in the 5FC-treated tumors (Figs. 2 and 6) compared to the jetinjected, but saline-treated, tumors in the Western blot, which is supported by the quantitative real-time RTPCR CD-expression analysis (see Table 1). These observations might account for the fact that the 5-FC treatment led to the successful elimination of the CDexpressing cells within the tumors, leading to the decrease in CD expression. The observation that not all CD-expressing tumor areas were eliminated by our treatment schedule suggests the therapeutic situation of a bstable diseaseQ in the two colon carcinoma models, in which tumor growth was stopped but the tumor was not eradicated.

1182

AND

METHODS

Tumor model. For the in vivo experiments the patient-derived human colon carcinoma Colo5734 was xenotransplanted subcutaneously in NMRI-nu/nu male mice [26]. The xenotransplants were grown to a size of 6 6 mm before jet-injection and further treatments were started. As a second tumor model the human colon carcinoma cell line SW480 was used. For establishment of tumors 1 107 cells of this line were inoculated subcutaneously into the animals and tumors were grown to a size of 6 6 mm before jet-injection and further treatments were begun. Intratumoral jet injection of CD-expressing naked plasmid DNA, treatment of animals, and evaluation of tumor growth. For the nonviral in vivo gene transfer, nude mice xenotransplanted with the human colon carcinoma Colo5734 or the SW480 colon carcinoma were used. The tumor-bearing animals were anesthetized with Radenarkon (40 mg/kg ip; Jansen Cilag, Neus, Germany) and received four jet-injections of the pCMV-CD plasmid DNA using the Swiss Injector prototype (EMS Medical Systems SA, Nyon, Switzerland) at a pressure of 3.0 bars through the skin into the tumor (Fig. 1B). The control animals were jet-injected with saline or the empty vector construct. Each jet-injection applied 10 Al plasmid solution containing 1 Ag/Al naked DNA dissolved in phosphate buffer (150 mM Na2HPO4/NaH2PO4; pH 7.0). Thus, each animal received a total dose of 40 Ag pCMV-CD plasmid DNA. Fortyeight hours after the jet-injection animals received daily ip injections of 5-FC (500 mg/kg). The jet-injected control animals received saline and the uninjected control group received daily ip injections of either saline or 5-FU (20 mg/kg). During the entire treatment tumor volume and body weight were determined. The tumor volume was calculated by using the measurements of two perpendicular tumor diameters using the spheroid equation tumor volume = [(tumor width) 2 tumor length] 0.5. After termination of the experiment animals were sacrificed and tumors were removed and snap frozen in liquid nitrogen for further analyses (Fig. 1A). Analysis of plasmid biodistribution and time schedule of CD expression of jet-injected animals. At days 1, 2, 4, 12, and 23 after jet-injection animals were sacrificed for removal of tumor, liver, skin, and kidney. The tumors and organs were shock frozen and then used for preparation of DNA and RNA using the QIAamp DNA Mini Kit (Qiagen, Valencia, CA, USA) and Trizol according to the manufacturer’s instructions (Life


ARTICLE

doi:10.1016/j.ymthe.2005.07.700

Technologies, Karlsruhe, Germany). Portions of the tumors were used for preparation of tumor lysates, which were subjected to Western blot analysis of the time course of CD protein expression. Western blot analysis of CD expression in jet-injected tumors. For preparation of tumor lysates, tumor tissues of the jet-injected animals were cryosectioned (15 Am sections) in a consecutive series of sections for fractionated analysis of CD expression in each tumor (see Fig. 1). For each fraction 10 cryosections were collected in lysis buffer (50 mM TrisHCl, pH 7.4, 1 mM EDTA, 150 mM NaCl, 1% NP-40, 0.25% sodium deoxycholate). Precast 10% polyacrylamide gels (Biozym, Oldendorf, Germany) were loaded with tumor lysates containing 5 Ag protein and electrophoresed at 125 V for 90 min. The gel was blotted onto a nitrocellulose (NC) filter (Hybond-C Extra, Amersham, Freiburg, Germany). The filter was blocked overnight at 48C in TBS blocking buffer (50 mM Tris, 150 mM NaCl, pH 7.5, 5% fat-free dry milk) and washed in TBST (0.25% Tween 20 in TBS buffer) at room temperature. The mouse anti-CD antibody (1:2000 dilution in TBST containing 5% BSA; MTMLaboratories, Heidelberg, Germany) was added and incubated for 90 min at room temperature. The NC filter was washed in TBST; POD-labeled anti-mouse IgG antibody (1:8000 dilution in TBST, 5% BSA; BD Pharmingen, Heidelberg, Germany) was added and incubated for 60 min. After being washed in TBST, the CD protein was detected using ECL solution (Amersham, Freiburg, Germany) and exposed to Kodak XOmat AR film. Real-time PCR for detection of pCMV-CD plasmid DNA in tumors, skin, liver, and kidney after jet-injection. Each real-time PCR (958C for 30 s, 45 cycles of 958C for 10 s, 628C for 10 s, 728C for 10 s) was performed using 250 ng of cellular DNA (LightCycler DNA Master Hybridization Probes Kit; Roche Diagnostics, Mannheim, Germany) in duplicate. For the pCMV-CD plasmid a 145-bp amplicon (forward primer, 5V-TTCGGATGCAACGCTAACT-3V; FITC-labeled probe, 5V-TCGTCGCCTTCCCTCAGGAAG-3V-FITC; LCRed640-labeled probe, LCRed640-5V-GATTTTGTCGTATCCCAACGGTGAAG-3V; reverse primer, 5V-TAACGCCTCTTCCAGCAAC-3V) was generated, which was detected by the gene-specific fluorescein- and LCRed640labeled hybridization probes (synthesis of primers for pCMV-CD by BioTeZ, Berlin, Germany; synthesis of probes for LacZ by TIB MOLBIOL, Berlin, Germany). The concentration of pCMV-CD was calculated using the bspikedQ standard curve of the pCMV-CD plasmid in serial dilutions (0.0125 pg, 0.05 pg, 0.2 pg, 3.3 pg, 33 pg, 330 pg, 3.3 ng, and 16.5 ng in duplicate) in 250 ng cellular DNA (isolated from SW480 cells), generated in parallel in each real-time PCR run. Real-time RT-PCR analysis of CD expression in jet-injected tumors. The total RNA from tumor tissue cryosections was isolated using the Trizol method (Life Technologies). Reverse transcriptase reaction was performed with 50 ng of total RNA (MuLV reverse transcriptase; Perkin-Elmer, Weiterstadt, Germany). Each real-time PCR was done as described above. Expression of the CD gene and of the housekeeping gene glucose-6phosphate dehydrogenase (G6PDH) was determined in parallel from the same RT reaction, each done in duplicate per sample. For the CD gene a 145-bp amplicon (forward primer, 5V-TTCGGATGCAACGCTAACT-3V; FITC-labeled probe, 5V-TCGT CGCCTT CCCTCAGGAAG-3 V-FITC; LCRed640-labeled probe, LCRed640-5V-GATTTTGTCGTATCCCAACGGTGAAG-3V; reverse primer, 5V-TAACGCCTCTTCCAGCAAC-3V) and for G6PDH a 113-bp amplicon were produced, which were detected by gene-specific fluorescein- and LCRed640-labeled hybridization probes (BioTeZ and TIB MOLBIOL; sequences and synthesis of primers and probes for G6PDH by Roche Diagnostics). The calibrator cDNA derived from a stably transduced CD-expressing cell clone (SW480 colon carcinoma cell line) was employed in serial dilutions (in duplicate) simultaneously in each run and served as internal standard for CD and G6PDH calculations. The ratios of CD (mean of the duplicates) and G6PDH (mean of the duplicates) for each tumor sample were calculated and expressed as relative CD mRNA expression (CD/G6PDH). Immunohistochemistry of jet-injected CD-expressing tumor tissues. For immunohistochemistry of CD expression in jet-injected tumor tissues, cryosections were washed in NKH buffer (in 1000 ml dest. water: 8 g NaCl,


0.4 g KCl, 4 ml Hepes, pH 7.4), fixed in 0.04% glutaraldehyde, and washed with NKH buffer. Endogenous peroxidase was blocked with 3% H2O2 for 5 min at room temperature. Sections were washed with NKH buffer and blocked for 30 min at room temperature in blocking buffer (in 100 ml NKH buffer: 1 ml 10% NaN3, 4 ml 1 M Hepes, pH 8.0, 5 ml 5% gelatin, 1 ml 22% BSA). The primary mouse anti-CD antibody (1:40; MTM-Laboratories) was incubated for 30 min at room temperature. After a wash, biotinylated antimouse IgG antibody (1:40; DAKO Diagnostica, Hamburg, Germany) was added and incubated for 15 min. The sections were washed and streptavidin-POD conjugate (DAKO) was added for 15 min and then washed. Thereafter, 5 ml of substrate stock solution (in 10 ml NKH buffer: 5 mg 3V,3-diaminobenzidine hydrochloride, 0.5 ml Hepes, pH 8.0, 0.2 ml 5% gelatin solution) and 5 Al 30% H2O2 was added and incubated for 10 min at room temperature. After being washed, sections were counterstained with hematoxylin/eosin and covered with Glycergel. Statistical Analysis. The levels of statistical significance for in vivo tumor growth were evaluated by using the nonparametric U test of MannWhitney. The statistical significance was set at P V 0.05.

ACKNOWLEDGMENTS We thank W. Uckert for kindly providing the pCMV-CD plasmid. We thank M. Lemm for excellent technical assistance. The work was supported by EMS Medical Systems SA, Nyon, Switzerland, by the H.W. & J. Hector Foundation, Mannheim, Germany, and by the Deutsche Forschungsgemeinschaft. RECEIVED FOR PUBLICATION NOVEMBER 26, 2004; REVISED JULY 19, 2005; ACCEPTED JULY 22, 2005.

REFERENCES 1. Herweijer, H., and Wolff, J. A. (2003). Progress and prospects: naked DNA gene transfer and therapy. Gene Ther. 10: 453 – 458. 2. Niidome, T., and Huang, L. (2002). Gene therapy progress and prospects: nonviral vectors. Gene Ther. 9: 1647 – 1652. 3. Wolff, J. A., et al. (1990). Direct gene transfer into mouse muscle in vivo. Science 247: 1458 – 1465. 4. Lawrie, A., et al. (2000). Microbubble-enhanced ultrasound for vascular gene delivery. Gene Ther. 7: 2023 – 2027. 5. Zhang, G., Song, Y. K., and Liu, D. (2000). Long-term expression of human alpha1antitrypsin gene in mouse liver achieved by intravenous administration of plasmid DNA using hydrodynamics-based procedure. Gene Ther. 7: 1344 – 1349. 6. Sikes, M. L., O’Malley, B. W., Finegold, M. J., and Ledley, F. D. (1994). In vivo gene transfer into rabbit thyroid follicular cells by direct DNA injection. Hum. Gene Ther. 6: 837 – 844. 7. Yang, N.-S., Burkholder, J., Roberts, B., Martinell, B., and McCabe, D. (1990). In vivo and in vitro gene transfer to mammalian cells by particle bombardment. Proc. Natl. Acad. Sci. USA 87: 9568 – 9572. 8. Lin, M. T., Pulkkinen, L., Uito, J., and Yoon, K. (2000). The gene gun: current application in cutaneous gene therapy. Int. J. Dermatol. 39: 161 – 170. 9. Aihara, H., and Miyazaki, J.-I. (1998). Gene transfer into muscle by electroporation in vivo. Nat. Biotechnol. 16: 867 – 870. 10. Furth, P. A., Kerr, D., and Wall, R. (1995). Gene transfer by jet injection into differentiated tissues of living animals and in organ culture. Mol. Biotechnol. 4: 121 – 127. 11. Cartier, R., et al. (2000). In vivo gene transfer by low volume jet injection. Anal. Biochem. 282: 262 – 265. 12. Walther, W., Stein, U., Fichtner, I., Malcherek, L., Lemm, M., and Schlag, P. M. (2001). Non-viral in vivo gene delivery into tumors using a novel low volume jet-injection technology. Gene Ther. 8: 173 – 180. 13. Walther, W., et al. (2002). Intratumoral low volume jet-injection for efficient nonviral gene transfer. Mol. Biotechnol. 21: 105 – 115. 14. Vauthey, J. N., de Marsh, R., Cendan, J. C., Chu, N. M., and Copeland, E. M. (1996). Arterial therapy of hepatic colorectal metastases. Br. J. Surg. 83: 447 – 455. 15. Huber, B. E., Austin, E. A., Good, S. S., Knick, V. C., Tibbels, S., and Richards, C. A. (1993). In vivo antitumor activity of 5-fluorocytosine on human colorectal carcinoma cells genetically modified to express cytosine deaminase. Cancer Res. 53: 4619 – 4626. 16. Ge, K., et al. (1997). Transduction of cytosine deaminase gene makes rat glioma cells highly sensitive to 5-fluorocytosine. Int. J. Cancer 71: 675 – 679. 17. Kievit, E., et al. (1999). Superiority of yeast over bacterial cytosine deaminase for enzyme/prodrug gene therapy in colon cancer xenografts. Cancer Res. 59: 1417 – 1421.

1183

ARTICLE

18. Hirschowitz, E. A., Ohwada, A., Pascal, W. R., Russi, T. J., and Crystal, R. G. (1995). In vivo adenovirus-mediated gene transfer of the Escherichia coli cytosine deaminase gene in colon carcinoma-derived tumors induced chemosensitivity to 5-fluorocytosine. Hum. Gene Ther. 6: 1055 – 1063. 19. Li, Z., et al. (1997). Enzyme/prodrug gene therapy approach for breast cancer using recombinant adenovirus expressing Escherichia coli cytosine deaminase. Cancer Gene Ther. 4: 113 – 117. 20. Baque, P., et al. (2002). Naked DNA injection for liver metastases treatment in rats. Hepatology 35: 1144 – 1152. 21. Ren, S., Li, M., Smith, J. M., DeTolla, L. J., and Furth, P. A. (2002). Low-volume jet injection for intradermal immunization in rabbits. BMC Biotechnol. 2: 10 – 15. 22. Schramm-Baxter, J., Katrencik, J., and Mitragotri, S. (2004). Jet injection into

1184

doi:10.1016/j.ymthe.2005.07.700

23. 24. 25.

26.

polyacrylamide gels: investigation of jet injection mechanics. J. Biomech. 37: 1181 – 1188. Pope, I. M., and Kinsella, A. R. (1997). The role of the bystander effect in suicide gene therapy. Eur. J. Cancer 33: 1005 – 1016. Babiuk, S., et al. (2003). Needle-free topical electroporation improves gene expression from plasmids administered in porcine skin. Mol. Ther. 8: 992 – 998. Horiki, M., Yamato, E., Ikegami, H., Ogihara, T., and Miyazaki, J.-I. (2004). Needleless in vivo gene transfer into muscles by jet injection in combination with electroporation. J. Gene Med. 6: 1134 – 1138. Fichtner, I., et al. (2004). Anticancer drug response and expression of molecular markers in early-passage xenotransplanted colon carcinomas. Eur. J. Cancer 40: 298 – 307.
