Nonviral in vivo gene delivery into tumors using a novel low ... - Nature

Gene Therapy (2001) 8, 173–180  2001 Nature Publishing Group All rights reserved 0969-7128/01 $15.00 www.nature.com/gt

RESEARCH ARTICLE

Nonviral in vivo gene delivery into tumors using a novel low volume jet-injection technology W Walther1, U Stein1, I Fichtner1, L Malcherek1, M Lemm1 and PM Schlag2 1

Max-Delbru¨ck-Center for Molecular Medicine, and 2Robert-Ro¨ssle-Clinic, Universita¨tsklinikum Charite´, Berlin, Germany

The jet-injection technology has developed as an applicable alternative to viral or liposomal gene delivery systems. In this study a novel, low-volume, ‘high-speed jet injector’ handheld system was used for the direct gene transfer of naked DNA into tumors. Lewis-lung carcinoma bearing mice were jet-injected with the ␤-galactosidase (LacZ), the green fluorescence (GFP) or the human tumor necrosis factor alpha (TNF-␣) gene carrying vector plasmids. The animals received five jet injections into the tumor at a pressure of 3.0 bar, delivering 3–5 ␮l plasmid DNA (1 ␮g DNA/␮l in water) per single jet injection. The jet injection of DNA leads to a widespread expression pattern within tumor tissues with

penetration depths of 5–10 mm. Analysis of tumor cryosections revealed moderate LacZ or GFP expression at 48 h and strong reporter gene expression 72 h and 96 h after jet injection. The simultaneous jet injection of the TNF-␣ and LacZ carrying vectors demonstrated efficient expression and secretion of both the cytokine, as well as LacZ expression within the tumor 24 h, 48 h, 72 h, 96 h and 120 h after jet injection. These studies demonstrate the applicability of jet injection for the efficient in vivo gene transfer into tumors for nonviral gene therapy of cancer using minimal amounts of naked DNA. Gene Therapy (2001) 8, 173–180.

Keywords: jet injection; nonviral; gene transfer; gene therapy; cancer

Introduction Gene delivery is still one major issue in gene therapy since efficiency of gene transfer can be decisive for the desired therapeutic effect. Many attempts are being made to develop efficient viral and nonviral gene delivery systems. Particularly viral vectors are of high efficiency, however are still associated with great preparative efforts, potential risks, such as induction of immune reponses towards viral vectors and concerns for their clinical use.1 Furthermore, viral and to some extent also liposomal gene transfer systems require comparatively high preparative efforts for their generation. The use of naked DNA for gene therapy has developed into a feasible alternative to viral and liposomal technologies. Therefore, transfer of naked DNA for genetic immunization and other gene therapy application is receiving increasing attention.2,3 Different procedures are used to deliver naked DNA into the desired cells or tissues in vitro and more importantly in vivo, such as by simple needle injection, particle bombardment, in vivo electroporation or jet injection.4–7 The advantage of these nonviral techniques is the circumvention of administration of recombinant viral particles, minimal or no immune response towards the DNA applied and low toxicity. The majority of these nonviral gene transfer technologies is used for vaccination studies to introduce DNA coding for proteins or peptides which induce immune

Correspondence: W Walther, Max-Delbru¨ck-Center for Molecular Medicine, Robert-Ro¨ssle-Strasse 10, 13092 Berlin, Germany Received 22 June 2000; accepted 14 September 2000

reponses resulting in antibody production in the host. For these applications utilization of naked DNA for intradermal or intramuscular application has proven to be an efficient vaccine as shown in numerous animal models for the development of DNA vaccines against different viral infections or cancer.8–11 Although needle injection can transduce DNA into muscle tissues, this technique is inefficient for other tissues including tumors. The particle bombardment using DNA-coated microparticles is effective in gene transfer for different tissues. However, in ballistic gene transfer DNA penetration is limited to dermal applications and cannot reach deeper tissue areas. Among the various nonviral gene delivery technologies jet injection is gaining increasing attention, since this technique allows gene transfer into different tissues with deeper penetration of the desired DNA into the targeted tissue and at comparable transfer efficiencies achieved by particle bombardment.7,12 Using the jet-injection technology, the DNA-containing solution is a jet of high velocity possessing the force to penetrate skin and underlying tissues leading to transfection of the affected areas.12 In vivo application of this technology does not induce tissue damage or inflammatory reactions at jet injection sites that could have an impact on the safety of jet injection in vivo.13 Most of the aforementioned technologies including the jet-injection technique have been employed for DNA vaccination studies to induce an immune response of the host toward the gene product encoded and expressed by the plasmid applied.10,14–16 In the majority of these studies, muscle or epidermis are targeted for gene transfer, whereas only a few studies are aimed at the direct in vivo gene transfer into tumor tissue,17,18 and none of the jet injection studies are aimed

Jet injection for nonviral tumor gene therapy W Walther et al

174

at the intratumoral in vivo application for cancer gene therapy. The present study was carried out for the establishment of the low volume versatile hand-held ‘high-speed jet injector’ for efficient gene transfer into tumors. The novel low volume ‘high speed jet injection system’ employed in our recent studies combines efficiency in transfer of naked DNA with the simplicity in DNA formulations (DNA solution in water) for the in vivo gene transfer, independently from the size of DNA used. The new jet injection system requires only minimal amounts of DNA to achieve efficient transgene expression within the tissue and makes simultaneous application of two or more different DNA constructs possible. More importantly, transgene expression is not restricted to narrow circumscribed areas (as seen with ballistic gene transfer or needle injection), but generates areas of transgene expression wider than the jet track with deeper penetration depths within the tissue. The use of this jet-injection system can be defined as a safe technology for repeated application of naked DNA in vivo possessing the potential for clinical use such as cancer gene therapy.

Results Influence of jet injection on DNA integrity In order to determine the influence of jet injection on possible damage to the plasmid DNA by shearing forces, we determined the integrity of the fired DNA. The reporter plasmid pCMV␤ (size 7.2 kb) solution of 1 ␮gDNA/␮l water was loaded into the jet injector and fired into an Eppendorf tube at 2.0 or 3.0 bar and constant distances. This DNA was then linearized by restriction enzyme digest and evaluated by gel electrophoresis. As shown in Figure 1, jet injection of the plasmid DNA has

Figure 1 Influence of jet injection on DNA integrity. The pCMV␤ reporter plasmid-solution of 1 ␮g DNA/␮l water was loaded into the ‘high-speed jet injector’. The DNA was fired into an Eppendorf tube at 2.0 or 3.0 bar. An DNA aliquot was then linearized with HindIII. The digested (lanes 2, 4 and 6) or undigested (lanes 1, 3 or 5) plasmid aliquots were loaded on to 1% agarose gel and the video image was analyzed. M, DNA size marker II (Boehringer); lanes 1, 2, non-jet-injected plasmid DNA; lanes 3, 4, plasmid DNA fired at 2.0 bar; lanes 5, 6: plasmid DNA fired at 3.0 bar. Gene Therapy

only a minor destructive effect at air pressures of 2.0 or 3.0 bar; conditions that are used for the jet-injection experiments into tumor tissues in this study. The semiquantitative analyses of video images revealed a 5 to 10% shearing of plasmid DNA at a pressure of 2.0 bar and a 15 to 20% shearing of plasmid DNA at 3.0 bar. Furthermore, comparison of the undigested plasmid DNAs (Figure 1, lanes 1, 3 and 5) revealed that the jet injection reduces the amount of supercoiled DNA which seems to be dependent on the pressure used. Although the DNA shearing can be associated with loss of transducible DNA, the remaining nucleic acid is sufficient in yield and quality to efficiently transduce the tumor tissue.

Evaluation of possible tumor cell dissemination by jet injection Before we started the in vivo experiments with the jet injection of reporter plasmids, the potential risk of tumor cell dissemination by the jet injection was determined. For these studies we used the B16 malignant melanoma model which possesses metastasizing potential to the lung. It was of interest to determine whether jet injection with saline into the primary tumor tissue would lead to increased tumor cell spreading associated with increase in lung metastases of treated animals. For this tumor model, eight animals per treatment group were either jetinjected at 2.0 bar, 3.0 bar or remained non-treated in the control group. During the observation time of 22 days after the jet injection, tumor growth was observerd (relative tumor volume) and at day 22 all animals were killed to count the number of lung metastases in the treated and the control groups. Table 1 lists the results for the B16 malignant melanoma model indicating that intratumoral jet injection does not provoke an increased risk for tumor cell spread which could potentially lead to elevated numbers of lung metastases. The frequence in lung metastases of the treated versus non-treated animals was not significantly different for the animals jetinjected at 2.0 bar (P = 0.4418) or at 3.0 bar (P = 0.9591). Furthermore, no differences have been determined regarding the relative tumor volume and the lung weights in the control versus treated animal groups (data not shown). Jet injection of reporter plasmids into tumors To evaluate the jet-injection efficiency for gene transfer into tumor tissue in vivo we used the low volume, ‘highspeed jet injector’ prototype for transduction of the LacZ or GFP reporter gene carrying pCMV␤ or pEGFP-N1 plasmids in Lewis-lung carcinoma. In all animal experiments, solutions of 1 ␮g DNA/␮l sterile water were used for gene transfer at a pressure of 3.0 bar. In earlier experiments we determined that the jet-injection force at 3.0 bar was most efficient for gene transfer in vivo. Therefore, this instrumental setting was used in our studies for the direct gene transfer into tumors through the overlying skin. Each animal received five jet injections of the respective plasmid DNA, whereas the control animals received 0.9% saline. Since the jet injector possesses a small ejection volume of 3 to 5 ␮l, the maximum amount administered per jet injection was equal to 3 to 5 ␮g plasmid DNA keeping the required amount of DNA at a minimum. Therefore, each animal received a maximum of only 25 ␮g plasmid DNA by the five jet injections applied in the study. The animals were killed 48, 72 or 96 h after jet injection


Table 1 Influence of jet injection on lung metastases formation in jet-injected B16 malignant melanoma in BDF1 mice Animal group

Non-injected animals

Mouse

Jet injection at 2.0 bar

0 5 4 0 10 0 3 0 2.75 3.6 1.5

1.14 1.09 0.86 0.80 2.46 0.97 1.2 1.49 1.25 0.53

1 2 3 4 5 6 7 8

12 5 4 3 7 0 3 0 4.25 3.9 3.5

1.43 0.85 0.74 1.43 1.32 1.69 1.17 0.47 1.17 0.36

18 5 0 NE 2 3 0 0 4 6.5 1.0

1.17 1.32 0.91 NE 1.32 1.32 0.69 1.11 1.16 0.44

Mean s.d. Median Jet injection at 3.0 bar

Mean s.d. Median

1 2 3 4 5 6 7 8

b

c

d

e

f

g

h

i

j

175

Lung nodules Lung weight (% of body weight)

1 2 3 4 5 6 7 8

Mean s.d. Median

a

NE, not evaluable.

to remove the tumors and to evaluate the expression pattern of the LacZ or GFP reporter genes within the tumor tissue. Figure 2 shows the results obtained with the jet injection of the LacZ expressing pCMV␤ plasmid at the different time points. Neither the untreated (Figure 2b), nor the saline jet-injected control tumors (Figure 2d) showed background staining for ␤-galactosidase-like activity in the tumor tissue. The application of the LacZ reporter plasmid however leads to ␤-galactosidase expression 48 h (Figure 2e,f), 72 h (Figure 2g, h) and 96 h (Figure 2i, j) after jet injection. As shown in Figure 2, the reporter expression is strongest 72 and 96 h after DNA application, whereas reporter expression at 48 h is somewhat weaker and almost not detectable 24 h after jet injection (not shown). Analysis of the tumor cryosections revealed that the jet path of plasmid solution penetrates the skin and travels 5 to 10 mm through the tumor tissue with only minor DNA transfer efficiency within this jet path (Figure 3a). However, reaching the end of the jetinjection path the jet is spreading into an area of high transfer efficiency encompassing an area of 3 × 3 mm2 to a maximum of 5 × 5 mm2 (Figure 3b). Such gene expression pattern has been observed in all jet-injected tumors independently of the time after jet injection.

Figure 2 Expression of the LacZ reporter gene in pCMV␤ jet-injected Lewis-lung tumors. The tumors were removed 48, 72 or 96 h after jet injection and formalin fixed cryosections were used for X-gal staining (see Materials and methods) and visualized through a light microscope at 60fold or 100-fold magnification. (a): Hematoxylin/eosin stained cryosection of untreated tumor; (b): X-gal staining of the same untreated tumor (magnification 60-fold). (c) Hematoxylin/eosin stained cryosection of saline control tumor 96 h after jet injection; (d) X-gal staining of the same saline jet-injected tumor (magnification 60-fold). (e and f) X-gal staining of tumor cryosection 48 h after jet injection with the pCMV␤ plasmid DNA at 3.0 bar (magnifications: (e), 60-fold; (f), 100-fold). (g and h) Xgal staining of tumor cryosection 72 h after jet injection with the pCMV␤ plasmid DNA at 3.0 bar (magnifications: (g), 60-fold; (h) 100-fold). (i and j), X-gal staining of tumor cryosection 96 h after jet injection with the pCMV␤ plasmid DNA at 3.0 bar (magnifications: (i), 60-fold; (j), 100fold).

To validate the findings obtained with the LacZ reporter system we repeated the experiments under the same jet-injection conditions using the GFP expressing pEGFP-N1 plasmid for transfection of Lewis-lung tumors. Figure 4 clearly indicates that again 48 h, 72 h and 96 h after jet injection, expression of the GFP can be Gene Therapy


176

a

b

a

b

c

d

e

f

g

h

Figure 3 Schematic representation of jet injection pattern in the tumor tissue jet-injected with plasmid DNA. Characteristic patterns of reporter gene expression within the jet-path (a) and the area of highest transfection efficiency (b) are shown in the images taken from Lewis-lung carcinomas jet-injected with the LacZ-expressing pCMV␤ plasmid DNA. The arrows indicate the tumor area where LacZ-expression was detected by X-gal staining.

detected within the targeted tumor tissue, whereas no fluorescence is seen in the respective controls (Figure 4a, b). In the transduced tumor cryosections we observed a similar pattern of GFP expression as seen with the LacZ plasmid DNA. This is an indication that comparable gene transfer efficiencies can be achieved independently of the reporter system or size of the DNA construct used (7.2 kb pCMV␤ versus 4.7 kb pEGFP-N1). In extended studies using alternative tumor models, jet injection of pCMV␤ plasmid DNA into B16 melanoma or into a xenotransplant model of a human mammary carcinoma was performed under similar conditions. In parallel to our observations in the Lewis-lung tumor model we were able to achieve gene transfer at comparable efficiency. Figures 5 and 6 show strong LacZ gene expression in B16 tumors 72 h and 96 h after jet injection or 48 h after jet injection in the human mammary xenotransplants with a similar pattern of LacZ distribution as observed in the Lewis-lung tumors.

Simultaneous intratumoral jet injection of LacZ reporter plasmid and TNF-␣ expressing vector To test whether the jet-injection technology is applicable for the transfer of a therapeutically relevant gene and whether two DNA constructs can be jet-injected simultaneously, we mixed the LacZ reporter expressing pCMV␤ plasmid with the human TNF-␣ expressing pM3CMV-hTNF vector. This mix was jet-injected into Gene Therapy

Figure 4 Expression of the GFP in pEGFP-1 jet-injected Lewis-lung tumors. The tumors were removed 48, 72 or 96 h after jet injection and formalin fixed cryosections were visualized by fluorescence or light microscopy at 60-fold magnification. Panels a, c, e and g are hematoxylin/eosin stained cryosections; b, d, f and h are the corresponding fluorescence images indicating GFP expression within the tumor tissue. (a and b), tumor cryosection of a saline control animal 96 h after jet injection; (c and d), tumor cryosection 48 h after jet injection with the pEGFP-N1 plasmid DNA; (e and f), tumor cryosection 72 h after jet injection with the pEGFP-N1 plasmid DNA; (g and h), tumor cryosection 96 h after jet injection with the pEGFP-N1 plasmid DNA.

Lewis-lung tumors at 3.0 bar and subsequently both, LacZ and TNF-␣ expression was determined in these tumors. The analysis of LacZ expression by X-gal staining and in parallel by TNF-␣-specific immunohistochemistry revealed that both genes are efficiently expressed at the same site of jet injection. Figure 7 depicts a representative expression analysis in a Lewis-lung tumor 48 h (Figure 7a) and 72 h (Figure 7c) after the jet injection showing the blue staining for LacZ expression. At the same site in the corresponding cryosection TNF-␣ expression (Figure 7b, d) has been observed by the TNF␣ specific immunostaining using a FITC-labeled monoclonal anti-TNF-␣ antibody. It is noteworthy, that TNF␣ is detectable in a more widespread area compared with the area where LacZ expression is present. This is an indication that TNF-␣ diffuses as a secretory cytokine from its site of transduction and expression into the vicinity


a

b

c

d

177

a

b Figure 5 LacZ expression in B16 tumors. The X-gal-stained control tumors which were not jet injected (a) or jet injected with 0.9% saline (b) showed no LacZ expression. The X-gal stained B16 tumors jet-injected with the LacZ expressing plasmid pCMV␤ showed strong LacZ expression 72 h (c) and 96 h (d) after jet injection.

of the transduced cell populations, therefore affecting a greater than only transduced area within the tumor. To determine the quantitative level of TNF-␣ expression and duration of expression we used cryosections of those tumors simultaneously jet-injected with the two plasmid DNAs for TNF-specific ELISA. Table 2 shows the amounts of TNF-␣ secreted 24 h, 48 h, 72 h, 96 h and 120 h after jet injection with the plasmid mix. The results clearly indicate an appropriate expression of the cytokine ranging from 1300 to 2600 pg TNF-␣/mg protein. Interestingly, the cytokine expression remained at this level over the entire observation time. However, it has to be determined what the maximum expression duration after jet injection of naked DNA will be in in vivo applications, since no significant drop in TNF-␣ expression has been observed even after 120 h.

Discussion The present study has been performed to evaluate the applicability of the jet-injection technology using a novel low volume air propulsion-based jet injector for gene transfer into tumors. The jet-injection technology is being tested over several years for its feasibility for transfer of pharmacological substances and of genetic material in vivo and it has been demonstrated that it can be successfully utilized for pharmacologic applications and for gene transfer.7,19,20 However, most of the air propulsion and also spring-powered jet injectors required relatively large amounts of plasmid DNA for transfection due to the large ejection volumes ranging from 100 ␮l7,14 up to 500 ␮l21 per jet injection. In contrast, recent particle bombardment technology has been shown to be feasible with low quantities of plasmid DNA when employed for vaccination studies.22 However, if intratumoral and not dermis-directed gene transfer is attempted, the particle bombardment technology does not reach the required penetration to transduce deeper tissue areas. To obtain a more applicable jet injection-based technology, recent

Figure 6 LacZ expression in xenotransplanted human mammary carcinoma in nude mice (magnification, 60-fold). The X-gal-stained tumor tissue (a) shows spread LacZ expression within the jet-injected areas. The structure of the tissue is shown by the hematoxylin/eosin stained corresponding section of the same tumor area (b).

development led to a new prototype of jet injector which requires only minimal amounts of naked plasmid DNA. The ejection volumes have been significantly reduced down to 3–5 ␮l per jet injection associated with higher accuracy of DNA amounts applied. The use of naked DNA has gained growing acceptance for gene therapy, since numerous studies demonstrated uptake of the genetic material leading to efficient transgene expression and also showed advantages over liposomal or adenoviral vectors.23–25 The new ‘high speed jet injector’ device used in the in vivo study possesses these properties of low volume jets for repeated naked DNA application into the targeted tissue. Similar to earlier experiences with other types of jet injectors, we and others observed no serious sideeffects in treated animals associated with the use of the injector.13 More importantly, particularly for cancer gene therapy applications we evaluated the potential risk of tumor cell dissemination by jet injection, which could then lead to increased tumor metastases. The data indicate that the utilization of the jet injector does not lead to an increased risk for tumor metastases in the B16 Gene Therapy


178

a

b

c

d

Figure 7 Representative cryosections of simultaneous pCMV␤ plasmid and pM3CMV-hTNF vector jet-injected Lewis-lung tumors. The tumors were removed 48 h (a, b) and 72 h (c, d) after jet injection and subjected to X-gal staining (a, c) or TNF␣-specific immunostaining (b, d), respectively.

Table 2 TNF-␣ secretion determined by TNF-ELISA in pM3CMVhTNF jet-injected Lewis-lung tumors in BDF1 mice Hours after jet injection 24 48 72 96 120 pCMV␤ control Saline control Non-jet-injected

TNF-␣ in pg/mg proteina 1318 ± 130 2629 ± 182 2510 ± 160 2448 ± 175 1890 ± 105 0 0 0

a

TNF-␣ values are given as the mean of at least two independent determinations.

malignant melanoma model, which points to the safety of this technology. Furthermore, regarding the DNA damage by shearing forces associated with jet injection, our analysis revealed no dramatic loss of intact plasmid DNA indicating that a sufficient amount of functionally active DNA can enter the target tissue. This is in agreement with findings by others, who also observed only small destructive effects on DNA after jet injection.15 Additional analyses using plasmids of different sizes revealed that even vectors of bigger size (up to 10 kb) underwent a maximum degradation of only 20–40% depending on the pressure used for the jet injections (data not shown). Vahlsing et al8 report that jet injection by the Med-E-Jet has an influence on the plasmid DNA when fired through a PVC membrane or skin, reflected by slight reduction of the supercoiled form of the plasmid and the appearance of linear plasmid molecules. However, these effects have possibly only minor impact on expression efficiency of the jetinjected DNA. These and our own findings suggest that jet injection-associated DNA damage is not a limiting factor for efficient in vivo gene transfer. Gene Therapy

In contrast to the majority of other studies which employ the jet-injection technology for genetic immunization approaches, our experiments are aimed at direct gene transfer into the targeted tumor tissue. In this context, the expression of the respective LacZ and GFP reporter gene and also of the therapeutic human TNF-␣ gene has been demonstrated directly within the jetinjected tissue. In fact, this approach revealed the parameters of distribution of gene expression and penetration of the transducing jet into the tumor tissue. In contrast to particle bombardment9 or other jet injection studies,15 we were able to show deeper penetration of the jet in the target tissue associated with widespread and strong gene expression. This is of essential importance, if not genetic immunization by targeting intradermal areas,14,15 but direct tumor gene therapy is attempted by introduction of cytokine or suicide genes. Furthermore, the widespread distribution of transgene expression ensures that sufficient areas of the tumor are affected for the desired therapeutic effect. This is of great advantage over results described for conventional needle injection-mediated gene transfer into the target tissue, where more localized expression is seen around and close to the injection track. Although more efficient expression can be achieved when needle injections of naked DNA are performed via vascular injection, this approach requires comparatively large volumes of 1 ml to 10 ml for successful gene transfer.26,27 The experiments in this study demonstrate that jet injection leads to significant gene expression by using only small amounts of naked DNA. In addition to the simple formulation of the plasmids applied, the study also demonstrated that the simultaneous jet injection of two different plasmids results in successful expression of both genes (LacZ and TNF-␣) at the same site of jet injection. This points to the possible application of particular ‘DNA-mixes’ to achieve maximum synergy of the therapeutic genes transduced into the desired tissue. Analysis of the TNF-␣ expression revealed that expression of the cytokine lasted for up to 120 h. Further experiments are required to determine the maximum duration of gene expression after jet injection, since jet injection of DNA leads rather to transient than stable gene expression.21

Materials and methods Determination of possible DNA degradation by jetinjection The high-speed jet injector was loaded with DNA solution in sterile water containing the pCMV␤ LacZ reporter plasmid (Clontech, Palo Alto, CA, USA). The plasmid DNA was amplified in ‘one shot’ competent E. coli bacteria (Invitrogen, Groningen, The Netherlands) and DNA was isolated by using the JETSTAR maxi preparation kit (Genomed, Bad Oeynhausen, Germany), which employs the alkaline/SDS and anion exchange method for plasmid DNA purification. The DNA solution was fired into an Eppendorf tube at pressures of 2.0 or 3.0 bar. Aliquots of this jet-injected plasmid DNA were used for linearization by digestion with HindIII and both undigested and digested DNA were then subjected to agarose gel electrophoresis. DNA integrity was evaluated from video images by densitometry using the NIH Image 1.44b11 software (obtained from Wayne Rasband, NIMH, Bethesda, MD, USA).


In vivo evaluation of potential jet injection-associated metastases 1 × 107 B16 malignant melanoma cells were transplanted into the right flank of 6–8-week-old female BDF1 mice. The animals were staged for 2 weeks to allow for the development of tumors with a size of approximately 6 × 6 mm. Tumor bearing mice were randomized for the studies: eight animals received no jet injection, eight animals received jet injections at 2.0 bar and eight animals received jet injections at 3.0 bar. The anesthetized animals received 10 jet injections at the respective pressure through the skin directly into the tumor applying 3–5 ␮l 0.9% sodium chloride solution. Over the observation time of 22 days tumor growth was evaluated and after day 22 all animals were killed and lung metastases were counted. In vivo jet injection of tumor bearing mice 1 × 107 Lewis-lung carcinoma cells or B16 malignant melanoma cells were transplanted into the right flank of 6–8-week-old female BDF1 mice. The animals were staged for 2–3 weeks to allow for the development of tumors with a size of approximately 6 × 6 mm. Tumor bearing mice were randomized for the studies. The anesthetized animals received five jet injections using the hand-held ‘high-speed jet injector’ prototype (developed and manufactured by EMS Medical, Konstanz, Germany) at a pressure of 3.0 bar through the skin directly into the tumor. Each jet injection applied 3–5 ␮l plasmid solution containing 1 ␮g/␮l naked DNA in sterile water. For LacZ reporter expression the pCMV␤ plasmid was used (Clontech, Palo Alto, CA, USA), for GFP reporter expression the pEGFP-N1 plasmid was employed (Clontech) and for the expression of the human TNF-␣ the pM3CMV-hTNF vector was used.28 In the experiments of simultaneous application of the LacZ reporter plasmid and of the pM3CMV-hTNF vector, DNAs of both constructs were mixed in sterile water to a final concentration of 1 ␮g DNA/␮l sterile water for each construct. The mix was applied in a similar manner as described for the reporter plasmids. In all experiments control animals received five jet injections of 0.9% saline. The animals were killed at 48, 72 and 96 h, tumors were excized and cryosections were prepared for subsequent expression analyses. To evaluate the efficiency of jet injection in human tumor tissue, mammary tumor xenotransplants on NMRI-nu/nu were additionally used for jet injection of the LacZ reporter plasmid. 4 × 4 mm tumor pieces of a xenograft line, originally derived from surgical material of a ductal invasive mammary carcinoma, were transplanted on to female nude mice with an age of 4–8 weeks.29 These animals were staged for 6–8 weeks for the establishment of tumor of a size of approximately 6 × 6 mm. The jet injections were performed as described and animals were killed 48 h after jet injection for tumor removal and subsequent X-gal staining. Analysis of LacZ and GFP reporter expression For detection of reporter gene expression animals were killed 48, 72 and 96 h after jet injection and tumors were removed for preparation of cryosections. The cryosections (7 ␮m) were fixed in 2% formaldehyde for 10 min at 4°C and rinsed twice with ice cold PBS. The slides were the placed in quadriPERM plus (Heraeus Instruments,

Osterode, Germany) chambers and covered with 3 ml Xgal mix (350 ␮l X-gal (5-bromo-4-chloro-3-indolyl-␤-dgalactoside) stock solution of 20 mg X-gal in 1 ml dimethylformamide; 7 ml 1.1 mm MgCl2; 0.5 ml 50 mm K3Fe(CN)6 and 0.5 ml K4Fe(CN)6) for ␤-galactosidase detection. The cryosections were incubated for at least 24 h at 37°C (blue color started to develop after 2–4 h) and then covered in Faramount aqueous mounting medium (Dako, Carpinteria, CA, USA). The tumors jet injected with pEGFP-N1 were also cryosectioned and fixed in 2% formaldehyde at 4°C for 10 min, mounted in Kaiser’s glycerol gelatin (Merck, Darmstadt, Germany) and then analyzed using a fluorescence microscope (Leica, Wetzlar, Germany). For better topology of the reporter expression within the tissue, corresponding cryosections were stained using hematoxylin-eosin solution.

179

TNF-␣ immunohistochemistry The tumors which were simultaneously transduced with the pCMV␤ and the pM3CMV-hTNF vectors were cryosectioned and the consecutive sections were subjected either to X-gal staining or TNF-␣ immunohistochemistry. For the immunohistochemistry, cryosections were fixed in 2% formaldehyde for 10 min at 4°C. The sections were washed once in PBS and the FITC-labeled monoclonal anti-human TNF-␣ specific antibody (R&D Systems, Wiesbaden, Germany) at a dilution of 1:2 in 2% BSA-PBS was added. The antibody was incubated over night at 4°C, the sections were then washed twice in PBS. TNF-␣-specific staining was evaluated using a fluorescence microscope and were checked for co-localization of the respective LacZ expression in the corresponding X-gal-stained cryosection of the same tumor. TNF-␣ ELISA For the TNF-␣ ELISA, three consecutive cryosections (7 ␮m) of one tumor (see Figure 7) were homogenized in 300 ␮l ice cold TE buffer (pH 8.0) containing the protease inhibitor aprotinin (10 mg/ml) and PMSF (phenylmethylsulfonyl fluoride, 100 ␮g/ml) by two cycles of freeze–thawing. The homogenates were centrifuged at 14 000 r.p.m., 4°C for 10 min. Then, 2 × 100 ␮l of the supernatants were subjected to the TNF-␣ ELISA (Cytoscreen ELISA, Biosource, Camarillo, CA, USA) as duplicates and the ELISA was performed according to the manufacturer’s instructions. Absorbance was measured in a microplate reader at 450 nm (SLT-Labinstruments, Crailsheim, Germany). The TNF-␣ values were calculated from the respective TNF-␣ standard curve using the EasySoftG200/Easy-Fit software (SLTLabinstruments) and normalized to the corresponding protein content of the tumor homogenates which was determined by using the Coomassie Plus Protein Assay Reagent (Pierce, Rockford, IL, USA). Statistical analysis In the metastasis study, statistical significance was evaluated by using the non-parametric Mann–Whitney rank sum test.

Acknowledgements The work was kindly supported by the EMS-Medical GmbH, Konstanz, Germany and granted by the HW and J Hector Foundation, Mannheim, Germany. Gene Therapy


180

References 1 Dewey RA et al. Chronic brain inflammation and persistent herpes simplex virus 1 thymidine kinase expression in survivors of syngeneic glioma treated by adenovirus-mediated gene therapy: implications for clinical trials. Nat Med 1999; 5: 1256–1263. 2 Ledley FD. Nonviral gene therapy: the promise of genes as pharmaceutical products. Hum Gene Ther 1995; 6: 1129–1144. 3 Klinman DM et al. Safe and effective regulation of hematocrit by gene gun administration of an erythropoietin-encoding DNA plasmid. Hum Gene Ther 1999; 10: 659–665. 4 Sikes ML, O’Malley BW, Finegold MJ, Ledley FD. In vivo gene transfer into rabbit thyroid follicular cells by direct DNA injection. Hum Gene Ther 1994; 6: 837–844. 5 Yang N-S et al. In vivo and in vitro gene transfer to mammalian cells by particle bombardment. Proc Natl Acad Sci USA 1990; 87: 9568–9572. 6 Aihara H, Miyazaki J-I. Gene transfer into muscle by electroporation in vivo. Nat Biotechnol 1998; 16: 867–870. 7 Furth PA, Shamay A, Wall RJ, Henninghausen L. Gene transfer into somatic tissue by jet injection. Anal Biochem 1992; 205: 365–368. 8 Vahlsing HL et al. Immunization with plasmid DNA using a pneumatic gun. J Immunol Meth 1994; 175: 11–22. 9 Rakhmilevich AL et al. Gene gun-mediated skin transfection with interleukin 12 gene results in regression of established primary and metastatic murine tumors. Proc Natl Acad Sci USA 1996; 93: 6291–6296. 10 Macklin MD et al. Immunization of pigs with a particlemediated DNA vaccine to influenza A virus protects against challenge with homologous virus. J Virol 1998; 72: 1491–1496. 11 Turner JG et al. Broadened clinical utility of gene gun-mediated granulocyte–macrophage colony-stimulating factor cDNAbased tumor cell vaccines as demonstrated with a mouse myeloma model. Hum Gene Ther 1998; 9: 1121–1130. 12 Furth PA, Shamay A, Henninghausen L. Gene transfer into mammalian cells by jet injection. Hybridoma 1995; 14: 149–152. 13 Cartier R et al. In vivo gene transfer by low-volume jet injection. Anal Biochem 2000; 282: 262–265. 14 Heansler J et al. Intradermal DNA immunization by using jetinjectors in mice and monkeys. Vaccine 1999; 17: 628–638. 15 Sawamura D et al. In vivo gene introduction into keratinocytes using jet injection. Gene Therapy 1999; 6: 1785–1787.

Gene Therapy

16 Seigne J et al. Feasibility study of gene gun mediated immunotherapy for renal cell carcinoma. J Urol 1999; 162: 1259–1263. 17 Nishi T et al. High-efficiency in vivo gene transfer using intraarterial plasmid DNA injection following in vivo electroporation. Cancer Res 1996; 56: 1050–1055. 18 Hui KM, Chia TF. Eradication of tumor growth via biolistic transformation with allogeneic MHC genes. Gene Therapy 1997; 4: 762–767. 19 Soyka P, Zellweger GA. Betamethasone in the treatment of hypertrophic scars. Helv Chir Acta 1980; 47: 137–140. 20 Bennett J, Nichols F, Rosenblum M, Condry J. Subcutaneous administration of midazolam: a comparison of the bioject jet injector with the conventional syringe and needle. J Oral Maxillofac Surg 1998; 56: 1249–1254. 21 Kerr DE, Furh PA, Powell AM, Wall RJ. Expression of genegun injected plasmid DNA in the ovine mammary gland and in lymph nodes draining the injection site. Animal Biotechnol 1996; 7: 33–45. 22 Pertmer TM et al. Gene gun-based nucleic acid immunization: elicitation of humoral and cytotoxic T-lymphocyte responses following epidermal delivery of nanogram quantities of DNA. Vaccine 1995; 13: 1427–1430. 23 Wolff JA et al. Direct gene transfer into mouse muscle in vivo. Science 1990; 247: 1465–1458. 24 Davis HL et al. Plasmid DNA is superior to viral vectors for direct gene transfer into adult mouse skeletal muscle. Hum Gene Ther 1993; 4: 733–740. 25 Yang JP, Huang L. Direct gene transfer to mouse melanoma by intratumor injection of free DNA. Gene Therapy 1996; 3: 542–548. 26 Zhang G et al. Expression of naked plasmid DNA injected into afferent and efferent vessels of rodent and dog livers. Hum Gene Ther 1997; 8: 1763–1772. 27 Budker V et al. The efficient expression of intravascularly delivered DNA in rat muscle. Gene Therapy 1997; 5: 272–276. 28 Stein U, Walther W, Shoemaker RH. Reversal of multidrug resistance by transduction of cytokine genes into human colon carcinoma cells. J Natl Cancer Inst 1996; 88: 1383–1392. 29 Naundorf H, Fichtner I, Bu¨ttner B, Frege J. Establishment and characterization of a new human estradiol- and progesteronereceptor-positive mammary carcinoma serially transplantable in nude mice. J Cancer Res Clin Oncol 1992; 119: 35–40.