Intraperitoneal Injection of an hpRNA-expressing Plasmid Targeting uPAR and uPA Retards Angiogenesis and Inhibits Intracranial Tumor Growth in Nude Mice

Abstract

Purpose

The purpose of this study was to evaluate the therapeutic potential of using plasmid expressed RNAi targeting uPAR and uPA to treat human glioma.

Experimental Design

In the present study, we have used plasmid based RNAi to simultaneously downregulate the expression of uPAR and uPA in SNB19 glioma cell lines and EGFR overexpressing 4910 human glioma xenografts in vitro and in vivo, and evaluate the intraperitoneal route for RNAi expressing plasmid administered to target intracranial glioma.

Results

Plasmid mediated RNAi targeting uPAR and uPA did not induce OAS1 expression as seen from RT-PCR analysis. In 4910 EGFR-over expressing cells, downregulation of uPAR and uPA induced the downregulation of EGFR and VEGF and inhibited angiogenesis in both in vitro and in vivo angiogenic assays. In addition, invasion and migration were inhibited as indicated by in vitro spheroid cell migration, matrigel invasion and spheroid invasion assays. We did not observe OAS1 expression in mice with pre-established intracranial tumors, which were given intraperitoneal injections of plasmid expressing siRNA targeting uPAR and uPA. Furthermore, the siRNA plasmid targeting uPAR and uPA caused regression of pre-established intracranial tumors when compared to the control mice.

Conclusion

In conclusion the plasmid expressed RNAi targeting uPAR and uPA via the intraperitoneal route has potential clinical applications for the treatment of glioma.

INTRODUCTION

Nucleic acid based methods have been used to target the transcriptional regulation of gene expression. One of these methods, RNA interference (RNAi), is now routinely used for the transient knockdown of gene expression in a wide range of organisms (1). Unlike antisense and triple helix approaches, the RNAi-mediated method of gene regulation utilizes normal cellular responses to double stranded RNA (dsRNA), which lead to the highly specific degradation of the target mRNA (1, 2). In addition and perhaps more importantly, this process triggers a cell-to-cell spreading of gene silencing as seen in several RNAi models (3–5). One potential obstacle for the therapeutic use of RNAi in mammalian cells is the activation of dsRNA-dependent protein kinase (PKR) by long dsRNA (6, 7). The ability to express intracellular siRNA has been demonstrated by several groups, mainly through the use of plasmids containing RNA pol III promoter (8–10).

Tumor cell invasion is a multistep process involving tumor cell attachment to the extracellular matrix (ECM) followed by the degradation of host barriers by proteolysis and tumor colony formation at distinct sites (11). Urokinase plasminogen activator (uPA) plays a key role in tumor progression and invasion by virtue of its ability to initiate a cascade of proteases that can degrade most matrix and basement membrane components and interfere with cell-cell and cell-matrix interactions (12, 13). uPA and its high affinity receptor, uPAR, which is a glycosyl phosphatidylinositol-anchored membrane protein (CD87), are believed to be critical elements in tumor biology since they control cell motility, tissue remodeling and are involved in the bioavailability of angiogenic factors (14). Formation of the uPA-uPAR complex at the cell surface is required for efficient activation of plasmin, a protease that can degrade ECM components and release various growth factors (15, 16). Although many studies have documented that uPAR has a central role in uPA-mediated cell surface plasminogen activation, further studies with uPAR-deficient mice have demonstrated the existence of additional pathways of uPA-mediated plasminogen activation that are independent of uPAR (17).

Glioblastomas are characterized by their invasive infiltration and destruction of surrounding normal brain tissue, making complete surgical resection of these tumors virtually impossible. Their invasive behavior seems to depend in part on a variety of proteolytic enzymes including serine, metalloproteases and cysteine proteases. Our previous work and that of others have suggested a direct correlation between the expression of uPA and uPAR and the invasiveness of human gliomas (18–23). Our studies have also demonstrated that antisense clones for uPAR and uPA are unable to form tumors in nude mice intracerebrally (24–26). Further, adenoviral vectors carrying antisense uPAR inhibited the invasion of glioma cells in vitro and inhibition of ex vivo tumor formation (27). Taken together, these studies indicate the biological significance of uPA and uPAR in glioma invasion and tumor growth.

The delivery approach of Ad-vectors and antisense technology to intracellularly target RNA seems to be a crucial limiting factor in exerting its inhibitory effect on the targeted molecule. The siRNA duplex is significantly more stable in cells than the cognate single stranded sense or antisense RNA, with transcription, under the control of the identical promoter in each case (28). Here, we have produced a single construct driven by a cytomegalovirus promoter (CMV) to deliver hpRNA molecules for both uPAR and uPA. We demonstrate the simplicity and ease of using vectors expressing hpRNA molecules for more than one target molecule using a single promoter and the subsequent, effective inhibition of glioma cell invasion, angiogenesis and tumor growth both in vitro and in vivo.

MATERIALS & METHODS

Construction of hpRNA expressing plasmid

A pcDNA3 plasmid with a CMV promoter was used in the construction of the hpRNA-expressing vector. The uPA sequence agcttGagagccctgctggcgcgccatatataatggcgcgccagcagggctctca and uPAR sequence gatccTacagcagtggagagcgattatatataataatcgctctccactgctgtag were used for the siRNA sequence. Inverted repeat sequences were synthesized for both uPA and uPAR. The inverted repeats were laterally symmetrical making them self-complimentary with a five-base pair mismatch in the loop region. This five-base pair mismatch would aid in the loop formation of the hpRNA. Oligos were heated in a boiling water bath in 6xSSC for 5 min and self-annealed by slow cooling to room temperature. The resulting annealed oligos were ligated to pcDNA 3 at the Hind III site for uPA and BamHI site for uPAR sequentially and the plasmid was named pU2. Single constructs were also made: puPAR targeting uPAR and puPA targeting uPA. An inverted-repeat sequence targeting GFP mRNA was synthesized and cloned into the pcDNA 3 Hind III site as described above (Figure 1).

Figure 1.

Schematic representation showing the possible mechanism involved in the formation of siRNA molecules from hpRNA to induce RNAi targeting of uPAR and uPA.

Cell culture and transfection conditions

SNB19 (or SNB19 GFP) cell lines, established from a human high-grade glioma, and EGFR-overexpressing human glioma xenograft tumors, designated as 4910 (Xeno), (kindly provided by Dr. David James at UCSF) were used for this study. Cells were grown in Dulbecco’s modified Eagle medium/F12 media (1:1, v/v) supplemented with 10% fetal calf serum in a humidified atmosphere containing 5% CO₂ at 37°C. SNB19 and 4910 cells were transfected with plasmid constructs (EV, SV, puPAR, puPA or pU2) using lipofectamine as per the manufacturer’s instructions (Life Technologies, Rockville, MD).

PCR and RT-PCR

Total RNA was isolated from SNB19 and 4910 control cells and the cells were transfected with empty vector (EV), scrambled vector (SV), puPAR, puPA or pU2. RNA was also isolated from cells transfected with antisense expression vectors for uPAR and uPA, and from cells transfected with a plasmid vector expressing siRNA-targeting GFP. RT-PCR was performed as per standard protocol for uPAR and uPA. To determine whether these siRNA-expressing plasmids induce an interferon response, RT-PCR for OAS1 was performed. Cells were also treated with interferon α (0.5ng/ml) to visualize OAS1 mRNA expression as positive control. Total RNA was isolated from fresh or paraffin-embedded brain tissue (Ambion, Catalog #47000) from control or mice injected intraperitoneally with EV, SV, puPAR, puPA or pU2, or interferon α (0.5ng). RT-PCR was performed to determine OAS1 expression. PCR was performed using primers specific for PCDNA3 vector amplifying CMV to BGH sequence regions using deparaffinized and protease-treated intracranial tumors. These primers had no significant homology to mouse or human sequences.

Primers used for PCR and RT-PCR

CMV to BGH	Forward CTGGTGTCGACCTGCTTCCGCGATGTACGGGC, Reverse CTGGTGTCGACATCCCCAGCATGCCTGCTAT
uPAR	Forward CATGCAGTGTAAGACCCAACGGGGA Reverse AATAGGTGACAGCCCGGCCAGAGT
uPA	Forward TGCGTCCTGGTCGTGAGCGA Reverse CAAGCGTGTCAGCGCTGTAG
GAPDH	Forward CGGAGTCAACGGATTTGGTCGTAT Reverse AGCCTTCTCCATGGTGGTGAAGAC
OAS1	Forward AGGTGGTAAAGGGTGGCTCC Reverse ACAACCAGGTCAGCGTCAGAT

Western blotting

SNB19 cells were transfected with mock, empty vector (EV)/scrambled vector (SV), puPA, puPAR or pU2. After 48 h, cells were collected and total cell lysates were prepared in extraction buffer containing Tris [0.1 M (pH 7.5)], Triton-X114 (1.0%), EDTA (10 mΜ), aprotinin, and phenylmethylsulfonyl fluoride as described previously (27). The extracts were incubated at 37°C for 5 min and centrifuged to separate the lower (detergent) phase that contains mainly hydrophobic membrane proteins including the glycosylphosphatidylinositol-anchored uPAR. Subsequently, 20μg of protein from these samples were separated under nonreducing conditions by 12% SDS-PAGE and transferred to nitrocellulose membranes (Schleicher & Schuell, Keene, NH). The membranes were probed for 2 h with antibodies against uPAR. The membranes were subsequently washed three times with PBS to remove excess of primary antibodies, incubated with secondary antibodies as required, and then developed according to enhanced chemiluminescence protocol (Amersham, Arlington Heights, IL). For loading control, the membranes were stripped and probed with monoclonal antibodies for GAPDH as per standard protocol. Similarly, 4910 EGFR-overexpressing xenograft cells were transfected with mock, empty vector (EV)/scrambled vector (SV), puPA, puPAR or pU2 and proteins extracted. Western blot analysis was performed for EGFR and VEGF per standard protocol.

Fibrin zymography

The enzymatic activity and molecular weight of electrophoretically separated forms of uPA were determined in conditioned medium of SNB19 cells transfected with mock, empty vector (EV)/scrambled vector (SV), puPA, puPAR or pU2 by SDS-PAGE as described previously (24, 29). Briefly, the SDS-PAGE gel contains acrylamide to which purified plasminogen and fibrinogen were substrates before polymerization. After polymerization, equal amounts of proteins in the samples were electrophoresed and the gel was washed and stained as described previously (24, 29).

Immunocytochemical analysis

4910 cells (1×10⁴) were seeded on vitronectin-coated 8-well chamber slides, incubated for 24 h and transfected with mock, EV, puPAR, puPA or pU2. After 72 h, cells were fixed with 3.7% formaldehyde and incubated with 1% bovine serum albumin in PBS at room temperature for 1 h for blocking. After the slides were washed with PBS, either IgG anti-VEGF (mouse) or IgG anti-EGFR (mouse) was added at a concentration of 1:200. The slides were incubated at room temperature for 1 h and washed three times with PBS to remove excess primary antibody. Cells were then incubated with anti-mouse FITC conjugated IgG (1:500 dilution) for 1 h at room temperature. The slides were then washed three times, covered with glass cover slips with DAPI containing mounting media and fluorescent photomicrographs were obtained.

In vitro angiogenic assay

4910 and SNB19 cells (2x10⁴/well) were seeded in 8-well chamber slides and transfected with mock, EV, puPAR, puPA, or pU2. After a 24 h incubation period, the conditioned medium was removed and added to a 4x10⁴ human dermal endothelial cell monolayer in 8-well chamber slides and the human dermal endothelial cells were allowed to grow for 72 h. Cells were then fixed in 3.7% formaldehyde, blocked with 2% bovine serum albumin, and the endothelial cells were incubated with factor VIII primary antibody (DAKO Corp., Carpinteria, CA). Cells were washed with PBS and incubated with a FITC-conjugated secondary antibody for 1 h. The slides were then washed and the formation of capillary-like structures was observed by fluorescent microscopy. Endothelial cells were also grown in conditioned media of either 4910 or SNB19 cells transfected with mock, EV, puPAR, puPA or pU2. Endothelial cells were allowed to grow for 72 h and H&E stained to visualize capillary network formation. The degree of angiogenesis was quantified based on the numerical value for the product of the number of branches and number of branch points as an average of 10 fields.

Dorsal skin-fold chamber model

Athymic nude mice (nu/nu; 18 male/female, 28–32 g) were bred and maintained within a specific-pathogen, germ-free environment. The implantation technique of the dorsal skin-fold chamber model has been described previously (30). Sterile small-animal surgical techniques were followed. Mice were anesthetized by intraperitoneal injection with ketamine (50mg/kg)/xylazine (10mg/kg). Once the animal was anesthetized completely, a dorsal air sac was made in the mouse by injecting 10ml of air. Diffusion chambers (Fisher) were prepared by aligning a 0.45-micron Millipore membranes (Fisher) on both sides of the rim of the “O” ring (Fisher) with sealant. Once the chambers were dry (2–3 min), they were sterilized by UV radiation for 20 min. 20 μl of PBS was used to wet the membranes. 2x10⁶ glioblastoma cells (4910 cells transfected with either EV or pU2), suspended in 100–150 μl of sterile PBS, were injected into the chamber through the opening of the “O” ring. The opening was sealed by a small amount of bone wax. A 1.5 to 2 cm superficial incision was made horizontally along the edge of the dorsal air sac and the air sac was opened. With the help of forceps, the chambers were placed underneath the skin and sutured carefully. After 10 days, the animals was anesthetized with ketamine/xylazine and sacrificed by intracardiac perfusion with saline (10 ml) followed by a 10 ml of 10% formalin/0.1 M phosphate solution and 0.001% FITC solution in PBS. The animals were carefully skinned around the implanted chambers and the implanted chambers were removed from the sc. air fascia. The skin fold covering the chambers was photographed under visible light and for FITC fluorescence. The number of blood vessels within the chamber in the area of the air sac fascia was counted and their lengths measured.

Spheroid migration assay

Migration was assayed as described previously (27) with some modifications. Spheroids of SNB19 and 4910 cells were prepared by seeding a suspension of 2x10⁶ cells in Dulbecco's modified Eagle medium on ultra low attachment 100 mm tissue culture plates and cultured until spheroid aggregates formed. Spheroids measuring ~150 μm in diameter (approximately 4x10⁴ cells/spheroid) were selected and transfected with pU2. Three days after infection, single glioma spheroids were placed in the center of each well of a vitronectin-coated (50 μg/ml) 96-well microplate and 200 μl of serum-free medium was added to each well. Spheroids were incubated at 37°C for 24 h, after which the spheroids were fixed and stained with Hema-3 and photographed. The migration of cells from spheroids to monolayers was quantified using a microscope calibrated with a stage and ocular micrometer and represented graphically.

Matrigel invasion assay

Invasion of glioma cells in vitro was measured by the invasion of cells through matrigel-coated (Collaborative Research, Inc., Boston, MA) transwell inserts (Costar, Cambridge, MA). Briefly, transwell inserts with 8 μm pores were coated with a final concentration of 1 mg/ml of Matrigel, SNB19 and 4910 cells transfected with mock, EV/SV, puPAR, puPA or pU2 were trypsinized, and 200 μl aliquots of cell suspension (1x10⁶ cells/ml) were added in triplicate wells. After a 24 h incubation period, cells that passed through the filter into the lower wells were quantified as described earlier (27, 31) and expressed as a percentage of the sum of cells in the upper and lower wells. Cells on the lower side of the membrane were fixed, stained with Hema-3 and quantified as percent invasion.

In vitro spheroid invasion assay

Multicellular SNB19 spheroids were cultured in 6-well ultra low attachment plates. Briefly, 3x10⁶ cells were suspended in 10 ml of medium, seeded onto the plates and cultured until spheroids formed. Spheroids, 100–200 μm in diameter, were selected and transfected with mock, EV/SV, puPAR, puPA or pU2. Three days after infection, tumor spheroids were stained with the fluorescent dye DiI and confronted with fetal rat brain aggregates stained with DiO. The progressive destruction of fetal rat brain aggregates and invasion of SNB19 cells were observed by confocal laser scanning microscopy and photographed as described previously (32, 33). The remaining volume of the rat brain aggregates at 24, 48 and 72 h were quantified using image analysis software as described previously (32) and graphically represented (33).

Intracranial tumor growth inhibition

For the intracerebral tumor model, 2x10⁶ 4910 xenograft tumor cells were intracerebrally injected into nude mice. Ten days after tumor implantation, the mice were treated with intraperitoneal injections of pU2 (150μg/injection/mouse) every other day three times. Control mice were either injected with PBS alone or with empty plasmid, vector (150μg/injection/mouse). Five weeks after tumor inoculation, six mice from each group were sacrificed by cardiac perfusion with 3.5% formaldehyde in PBS, their brains removed, and paraffin sections prepared. Sections were stained with hematoxylin and eosin to visualize tumor cells and to examine tumor volume (32, 33). The sections were blindly reviewed and scored semiquantitatively for tumor size. Whole mount images of brains were also taken to determine infiltrative tumor morphology. The average tumor area per section integrated to the number of sections where the tumor was visible was used to calculate tumor volume and compared between controls and treated groups. RT-PCR was performed on fresh or paraffin-embedded brain tissue for OAS1 pcDNA3 plasmid and GAPDH as previously described.

Immunohistochemical analysis

Brains of control and pU2 intraperitoneally treated mice implanted with 4910 tumors were fixed in formaldehyde and embedded in paraffin as per standard protocol. Sections were deparaffinized as per standard protocol. Sections were blocked in 1% BSA in PBS for 1 h, and the sections were subsequently transferred to primary antibody (uPAR and uPA) diluted in 1% BSA in PBS (1:500). Sections were allowed to incubate in the primary antibody solution for 2 h at 4ºC in a humidified chamber, followed by a wash in 1% BSA in PBS and placed in a solution with the appropriate (anti-mouse and anti-rabbit FITC) secondary antibody. The sections were allowed to incubate with the secondary antibody for 1 h and visualized using a confocal microscope. Images were obtained for FITC. Transmitted light images were also obtained after H & E staining as per standard protocol to visualize the morphology of the sections. A control study was performed using a normal rabbit immunoglobulin fraction as the primary antibody (control Ab) instead of uPAR and uPA.

In situ hybridization

2x10⁶ xenograft tumor cells were intracerebrally injected into nude mice. Ten days after tumor implantation, the mice were treated with intraperitoneal injections of mock, EV/SV, puPAR, puPA or pU2 (150μg/injection/mouse) every other day three times. Control mice were either injected with PBS alone or with empty plasmid vector (150μg/injection/mouse). Five weeks after tumor inoculation, six mice from each group were killed by cardiac perfusion with 3.5% formaldehyde in PBS, their brains harvested, and paraffin sections prepared. Sections were deparaffinized and probed for PCDNA3-CMV promoter using specific alkaline phosphatase labeled (Alkaphos, Amersham Cat # RPN 3680) DNA oligo (CTGGTGTCGACCTGCTTCCGCGATGTACGGGC) as per standard protocol. Hybridization was observed using NBT alkaline phosphatase substrate Western Blue® (Promega) as per manufacturer’s instructions.

Animal survival analysis

Nude mice were implanted with intracranial 4910 xenograft tumors and their survival ability was determined based on symptoms of intracranial pressure, arched back and dehydration. If the animals exhibited excessive pain, they were euthanized. Two sets of animals were used (6 mice/group). Both sets were implanted with intracranial xenograft tumors as described previously. Ten days after tumor implantation, the mice were treated with intraperitoneal injections of pU2 (150μg/injection/mouse) every other day three times. Control mice were either injected with PBS alone or EV/SV (150μg/injection/mouse). The mice were maintained in clean room conditions and monitored every day for 112 days after which the experiment was artificially terminated. Brains were collected from the control and treated mice, paraffin embedded, sectioned, and H & E stained per standard protocols. The survival curve was plotted as per standard methods and graphically represented as percent survival.

RESULTS

Effect of pU2 plasmid on uPA and uPAR mRNA levels in total cell extracts

Total RNA was isolated from control cells and cells transfected with empty vector (EV), scrambled vector (SV), puPAR, puPA and pU2. RNA was also isolated from cells transfected with antisense expression vectors for uPAR and uPA, and from cells transfected with plasmid vector expressing siRNA-targeting GFP. RT-PCR was performed as per standard protocol for uPAR and uPA. To determine whether these siRNA-expressing plasmids induce an interferon response, RT-PCR for OAS1 was performed. As a positive control, cells were treated with interferon α (0.5ng/ml) to visualize OAS1 mRNA expression. Figure 2A shows RT-PCR analysis of total RNA isolated from cells treated with interferon α, control, antisense uPAR (asuPAR), antisense uPA (asuPA) and siRNA expressing vectors for uPAR (puPAR), uPA (puPA) and both upAR and uPA (pU2). As RNAi controls, siRNA targeting GFP was used in non–GFP cells, empty vector (pEV) and scrambled vector (pSV). RT-PCR for GAPDH served as an internal control. Interferon-treated cells did not show a change in the expression levels of uPA or uPAR, OAS1 induction was observed and served as control for OAS1 RT-PCR. Control, pGFP, pEV and pSV-treated cells did not show changes in uPAR or uPA expression and no detectable levels of OAS1 were observed. Antisense for uPAR and uPA did not show a change in the mRNA levels of uPAR or uPA, whereas the protein levels did show a 20–50% decrease (not shown). Cells treated with puPAR showed a 20% decrease in uPA expression and a 50% decrease in uPAR expression. Cells treated with puPA did not show appreciable change in uPAR expression levels whereas uPA levels showed a 30–60%-decrease. Cells treated with pU2 showed a 75% decrease in uPAR mRNA levels and 60% decrease in uPA mRNA levels. GAPDH levels did not change and served as controls.

Figure 2. Determination of uPAR and uPA mRNA levels using semi-quantitative RT-PCR.

Total RNA was isolated from control cells and cells transfected with empty vector (EV), scrambled vector (SV), puPAR, puPA or pU2. RNA was also isolated from cells transfected with antisense expression vectors for uPAR and uPA, and from cells transfected with a plasmid vector expressing siRNA for GFP. RT-PCR was performed per standard protocol for uPAR and uPA. To determine whether these siRNA-expressing plasmids induce an interferon response, RT-PCR for OAS1 was performed. As a positive control, cells were also treated with interferon alpha (0.5ng/ml) to visualize OAS1 mRNA expression (Fig 2A). SNB19 cells were transfected with mock, EV/SV, puPA, puPAR or pU2. After 48 h, cells were collected and total cell lysates were prepared in extraction buffer containing Tris [0.1 M (pH 7.5)], Triton-X114 (1.0%), EDTA (10 mΜ), aprotinin, and phenylmethylsulfonyl fluoride as described previously. Subsequently, 20mu;g of protein from these samples were separated under nonreducing conditions by 12% SDS-PAGE and transferred to nitrocellulose membranes (Schleicher & Schuell, Keene, NH). The membranes were probed for 2 h with antibodies against uPAR (Fig 2B). The membranes were subsequently washed three times with PBS to remove excess primary antibody, incubated with a secondary antibody as required, and then developed per standard protocol. For loading control, the membranes were stripped and probed with monoclonal antibodies for GAPDH. The enzymatic activity and molecular weight of electrophoretically separated forms of uPA were determined from the conditioned media of SNB19 cells transfected with mock, EV/SV, puPA, puPAR or pU2 by SDS-PAGE as described previously (Fig. 2B). Western blot analysis was also performed using cell lysates of 4910 EGFR-overexpressing 4910 xenograft cells transfected with mock, EV/SV, puPA, puPAR or pU2. Western blots were immunoprobed for EGFR and VEGF per standard protocols (Fig. 2C).

Effect of pU2 on uPA and uPAR enzymatic activity in SNB19 cells and EGFR and VEGF protein levels in 4910 xenograft cells

RNAi targeted against proteolytic degradation could be an important intervention to prevent cancer cell invasion. uPA and uPAR have been shown to play significant roles in ECM degradation. As demonstrated by western blotting, transfection of SNB19 cells with a vector expressing siRNA for uPAR and uPA (pU2) strongly inhibited the protein expression of uPAR levels when compared to mock and empty vector (EV)/scrambled vector (SV)-transfected cells (Fig. 2B). Fibrin zymography revealed that uPA enzymatic activity significantly decreased in SNB19 cells transfected with puPAR, puPA and pU2 as compared to mock and EV/SV transfection (Fig. 2B). GAPDH protein levels served as a loading control (Fig. 2B). Quantitative analysis of uPAR and uPA bands by densitometry revealed a significant (P<0.001) decrease in uPAR protein (13- to 16-fold) and uPA enzymatic activity (10- to 12-fold) in pU2-transfected cells as compared to mock and EV/SV-transfected cells. Cells transfected with puPAR and puPA inhibited uPAR and uPA levels in almost the same manner as pU2, but downregulation of the target molecules was much more pronounced with the bicistronic construct as compared to either of the single constructs. 4910 xenograft cells exhibited similar downregulation of uPA and uPAR as SNB19 cells (data not shown). Since EGFR levels are overexpressed in 4910 cells, we determined the effect of pU2 transfection. Simultaneous downregulation of uPAR and uPA decreased EGFR levels in 4910 cells at least 15 fold while VEGF levels decreased 5 fold (Fig 2C). GAPDH levels did not change and served as a loading control. In situ levels of VEGF and EGFR were also determined and those results correlated with western blot analysis (Fig 3A).

Figure 3.

To visualize VEGF and EGFR expression in EGFR-overexpressing 4910 cells, 1×10⁴ cells were seeded on vitronectin-coated 8-well chamber slides, incubated for 24 h, and transfected with mock, empty vector (EV) and a vector expressing siRNA for uPAR (puPAR), uPA (puPA), or both (pU2). After 72 h, cells were fixed with 3.7% formaldehyde and incubated with 1% bovine serum albumin in PBS at room temperature for 1 h for blocking. After the slides were washed with PBS, either IgG anti-VEGF (mouse) or IgG anti-EGFR (mouse) was added at a concentration of 1:200. The slides were incubated at room temperature for 1 h and washed three times with PBS to remove excess primary antibody. Cells were then incubated with anti-mouse FITC conjugated IgG (1:500 dilution) for 1 h at room temperature. The slides were washed three times, covered with glass cover slips with DAPI-containing mounting media, and fluorescent photomicrographs were obtained. Figure 3A shows expression of EGFR and VEGF in control and EV/SV-, puPAR-, puPA- and pU2-transfected 4910 cells. To determine in vitro angiogenesis, 4910 or SNB19 cells (2x10⁴/well) were seeded in 8-well chamber slides and transfected with mock, empty vector (EV) and a vector expressing siRNA for uPAR (puPAR), uPA (puPA), or both (pU2). After a 24 h incubation period, the conditioned medium was removed and added to 4x10⁴ human dermal endothelial cell monolayer in 8-well chamber slides. The human dermal endothelial cells were allowed to grow for 72 h. Cells were then fixed in 3.7% formaldehyde, blocked with 2% bovine serum albumin, and incubated with factor VIII primary antibody (DAKO Corp., Carpinteria, CA). The cells were then washed with PBS and incubated with a FITC-conjugated secondary antibody for 1 h. The slides were washed and the formation of capillary-like structures was observed using fluorescent microscopy (Fig. 3B). Endothelial cells were also grown in conditioned media of 4910 or SNB19 cells transfected with mock, empty vector (EV) and a vector expressing siRNA for uPAR (puPAR), uPA (puPA), or both (pU2). Endothelial cells were allowed to grow for 72 h, and H&E stained for visualization of network formation (Fig. 3B). As determined by in vitro angiogenesis quantification, similar results were obtained for SNB19 and 4910 cells. The degree of angiogenic induction was quantified for both SNB19 and 4910 cells based on the numerical value for the product of the number of branches and number of branch points (*p value=0.005) (Fig. 3C). Figure 3D shows in vivo angiogenic assay using the dorsal skin fold model as described in Methods. Briefly, the animals were implanted with diffusion chambers containing control or pU2-transfected 4910 cells in a dorsal cavity. Ten days after implantation, the animals were sacrificed and vasculature flushed with FITC solution. The skin fold covering the diffusion chamber was observed for FITC fluorescence and in visible light for the presence of tumor-induced neovasculature (TN) and pre-existing vasculature (PV).

Inhibition of tumor cell-induced capillary network formation by pU2

Emerging tumors are dependent on the formation of new blood vessels to fuel tumor growth. Since uPA and uPAR have been reported to regulate angiogenesis, we next assessed the effect of pU2 on tumor cell-induced angiogenesis. We performed immunohistochemical analysis using factor VIII antigen to evaluate tumor-induced vessel formation in an in vitro co-culture system and H&E staining of endothelial cells cultured in conditioned media of SNB19 and 4910 xenograft cells after transfection with mock, EV/SV, puPA, puPAR or pU2. The results demonstrated that endothelial cells formed capillary-like structures in the presence of mock and EV/SV-transfected cells within 48 h. In contrast, the pU2 vector significantly inhibited tumor cell-induced capillary-like network formation. SNB19 and 4910 xenograft cells exhibited similar behavior although 4910 xenograft cells were slightly more aggressive in terms of angiogenesis when compared to SNB19 cells (Fig 3B). The quantitation of the branch points and number of branches was extremely low in pU2-transfected co-cultures as compared to mock and EV/SV-transfected cells. In the case of xenograft co-cultures, rudimentary network-like structures were seen in the pU2-treatment group (Fig. 3B). The effect was less than 50% to 60% in puPA- and puPAR-transfected co-cultures, respectively, when compared to mock and EV/SV-transfected co-cultures in relation to capillary-like structure formation (Fig 3C). To confirm the in vitro co-culture experiments, we examined whether pU2 could inhibit tumor angiogenesis in vivo as assessed by the dorsal window model. Implantation of a chamber containing mock and empty vector-transfected 4910 xenograft cells resulted in microvessel development with thin, curved structures and numerous tiny bleeding spots [as indicated by arrows (TN= tumor-induced neovasculature; PV= pre-existing vasculature)]. In contrast, implantation of 4910 xenograft cells transfected with pU2 did not result in the development of any additional microvessels (Fig. 3D). Similar results were obtained with SNB19 cells (data not shown).

siRNA against uPA and uPAR inhibits glioma migration and invasion

To determine whether uPAR and uPA siRNA expression is capable of influencing tumor cell migration, we transfected SNB19 and 4910 xenograft spheroids with the puPAR, puPA and pU2. As shown in Figure 4A, there was much higher cell migration in the control and EV/SV-transfected spheroids than in the pU2-transfected spheroids. As demonstrated by the quantitation of the distance of migration out from the spheroids, cell migration of the control and EV/SV-transfected spheroids was significantly higher (P<0.05) as compared to pU2-transfected spheroids. Moderate cell migration from tumor spheroids transfected with puPAR and puPA was observed. Further, 4910 xenograft cells appeared to be more migratory than SNB19 cells. The effect of pU2 was similar in both SNB19 and 4910 xenograft cells (Fig 4A).

Figure 4.

Migration was assayed as previously described. Spheroids of SNB19 or 4910 cells were prepared by seeding a suspension of 2x10⁶ cells in Dulbecco's modified Eagle medium on ultra low attachment 100 mm tissue culture plates and cultured until spheroid aggregates formed. Spheroids measuring ~150 μm in diameter (about 4x10⁴ cells/spheroid) were selected and transfected with mock, empty vector (EV) and a vector expressing siRNA for uPAR (puPAR), uPA (puPA) or both (pU2). Three days after infection, single glioma spheroids were placed in the center of each well in vitronectin-coated (50 μg/ml) 96-well microplates and 200 μl of serum-free medium was added to each well. Spheroids were incubated at 37°C for 24 h, after which the spheroids were fixed and stained with Hema-3 and photographed. Cell migration from spheroids to monolayers was quantified using a microscope calibrated with a stage and ocular micrometer and represented graphically (Fig. 4A). In vitro invasion of SNB19 and 4910 cells was determined by measuring the cells that invaded through matrigel-coated (Collaborative Research, Inc., Boston, MA) transwell inserts (Costar, Cambridge, MA). Briefly, transwell inserts with 8 μm pores were coated with a final concentration of 1 mg/ml of matrigel, SNB19 cells transfected with mock, EV/SV, puPAR, puPA or pU2 were trypsinized, and 200 μl aliquots of cell suspension (1x10⁶ cells/ml) were added to the wells in triplicate. After a 24 h incubation period, cells that passed through the filter into the lower wells were quantified as described earlier and expressed as a percentage of cells in the lower wells. Cells on the lower side of the membrane were fixed, stained with Hema-3 and quantified as percent invasion (Fig. 4B). SNB19 and 4910 spheroids were cultured in 6-well ultra low attachment plates. Briefly, 3x10⁶ cells were suspended in 10 ml of medium, seeded onto the plates and cultured until spheroids formed. Spheroids, 100–200 μm in diameter, were selected and transfected with mock, EV/SV, puPAR, puPA or pU2. Three days after infection, tumor spheroids were stained with the fluorescent dye DiI and confronted with fetal rat brain aggregates stained with DiO. The progressive destruction of fetal rat brain aggregates and invasion of SNB19 cells were observed by confocal laser scanning microscopy and photographed as described previously (32, 33). The remaining volume of the rat brain aggregates at 24, 48 and 72 h were quantified using image analysis software as described previously and graphically represented (Fig 4C). X= 4910 xeno

To evaluate the effect of siRNA-mediated inhibition of uPAR and uPA on glioma invasiveness, we utilized a two-model system: the standard matrigel invasion assay and the spheroid invasion assay. SNB19 or 4910 xenograft cells transfected with mock and EV/SV extensively invaded the matrigel-coated transwell insert. In contrast, the pU2-transfected cultures had markedly much less invasiveness through the matrigel as compared to control and EV/SV-transfected cells. Quantitative determination of invasion confirmed that pU2 transfection demonstrated only 6% invasion in SNB19 cells and 8% invasion in 4910 xenograft cells as compared to mock and EV/SV-transfected cells (Fig. 4B). Further, quantitative analysis results of invasion of SNB19 and 4910 xenograft cells transfected with puPAR and puPA was 22% (SNB19), 30% (4910) and 35% (SNB19), 40% (4910), respectively (Fig 4B).

We further examined the extent of the pU2 effect in a spheroid invasion assay. The EV/SV-transfected spheroids progressively invaded fetal rat brain aggregates; and quantitation revealed that the glioma spheroids invaded the fetal rat brain aggregates by 25% within 24 h, >70% within 48 h, and >90% at 72 h. In contrast, the tumor spheroids transfected with the pU2 vector invaded the fetal rat brain aggregates by only 2% (SNB19) and 2.8% (4910 xeno) after 72 h. Invasion of fetal rat brain aggregates by the single constructs were: 20% (SNB19) and 23% (4910 xeno) with puPAR, and 40% (SNB19) and 60% (4910 xeno) with puPA (Fig. 4C). Taken together, these findings provide strong evidence that RNAi-mediated silencing of uPAR and uPA greatly inhibits glioma cell invasion in both in vitro models in comparison to the single siRNA constructs for uPAR and uPA.

uPA and uPAR siRNA suppresses intracranial tumor growth

We used an intracranial tumor model with 4910 xenograft cells to assess the effects of RNAi-mediated inhibition on pre-established tumor growth in vivo. Paraffin brain sections of the untreated (mock) and EV/SV-treated control groups were characterized by large-spread tumor growth after a 5-week follow-up period as visualized by H&E staining of similar sections (Fig. 5A). However, we could not detect tumors in mice treated with the pU2 vector (Fig. 5A). Further quantification of H&E-stained brain sections by a neuropathologist (Dr. Gujrati) blinded to treatment condition revealed no difference in tumor size between the control and EV/SV treatment groups. However, total regression of pre-established tumors was revealed in the pU2-treated group (Fig. 5A & 5B). Pre-established intracranial tumor growth was inhibited by 70% in puPAR-treated mice and 55% in puPA-treated mice. These results demonstrated that RNAi-mediated suppression of uPAR and uPA inhibited pre-established intracranial tumor growth. Paraffin sections were probed for uPA and uPAR protein as previously described (see Materials and Methods). pU2-treated mice did not exhibit uPAR or uPA protein expression above expected background levels (not shown) whereas the control and EV/SV-treated mice showed increased levels of uPAR and uPA expression localized at the tumor region (Fig 5C). To determine whether this downregulation of protein expression was caused by pU2 treatment, paraffin brain sections of 4910 cells implanted in untreated mice (mock) and mice treated with SV/EV, puPAR, puPA and pU2 were probed with an alkaline phosphatase labeled oligo for the CMV promoter region of the pcDNA3 plasmid vector. In the mock sections, no alkaline phosphatase activity was detected. In contrast, we observed alkaline phosphatase activity in the EV/SV-, puPAR-, puPA- and pU2-treated mice, indicating the presence of a pcDNA3 vector in the tissue (Fig 5D). As a positive control, mice were injected intracranially with interferon α to determine OAS1 expression. Total RNA from fresh or paraffin-embedded brain sections was used for RT-PCR to determine in vivo induction of interferon response plasmids inducing RNAi (see Materials and Methods). From the RT-PCR analysis, it is clear that no OAS1 induction was observed in the mock, EV/SV-, puPAR-, puPA- or pU2-treated mice. RT-PCR for GAPDH served as an internal control (Fig 5E). The presence of a pcDNA3 vector was confirmed by regular PCR analysis (Fig 5E). To determine the survival rate of mice implanted with 4910 xenograft intracranial tumors, we performed a long-term survival study, Thirty days after implantation, all of the control mice died and paraffin sections of these brains revealed the presence of large intracranial tumors (Fig 5A). In contrast, pU2-treated mice survived much longer. One mouse from the pU2-treated group died after 43 days, but paraffin sections revealed the presence of a hemorrhage and very few xenograft cells indicating that the death of this animal was not due to an intracranial tumor (not shown). The remaining 5 animals in the pU2-treated group were sacrificed after 112 days. Paraffin sections of these mice revealed no intracranial tumors, normal brain morphology, and no visible 4910-xenograft tumor cells. The survival rate is represented as percent survival in Figure 5F.

Figure 5.

For the intracerebral tumor model, 2x10⁶ 4910 xenograft tumor cells were intracerebrally injected into nude mice. Ten days after tumor implantation, the mice were treated with intraperitoneal injections of pU2 (150μg/injection/mouse) every other day three times. Control mice were either injected with PBS alone or with an empty plasmid vector (150μg/injection/mouse). Five weeks after tumor inoculation, six mice from each group were sacrificed via cardiac perfusion with 3.5% formaldehyde in PBS, their brains removed, and paraffin sections prepared. Sections were stained with hematoxylin and eosin to visualize tumor cells and to examine tumor volume as described in Methods (arrows point to approximate site of intracranial implantation site) (Fig. 5A & 5B). To visualize the expression levels of uPAR and uPA in intracranial tumors, mouse brains were fixed in formaldehyde and embedded in paraffin per standard protocols. Sections were deparaffinized, blocked in 1% BSA in PBS for 1 h, and subsequently transferred to primary antibody (uPAR and uPA) diluted in 1% BSA in PBS (1:500). Sections were allowed to incubate in the primary antibody solution for 2 h at 4ºC in a humidified chamber, followed by a wash in 1% BSA in PBS and placed in a solution with the appropriate (anti-mouse and anti-rabbit FITC) secondary antibody. The sections were allowed to incubate with the secondary antibody for 1 h and visualized using a confocal microscope. Images were obtained for FITC. Transmitted light images were also obtained after H&E staining to visualize the morphology of the sections. A control study was performed using a normal rabbit immunoglobulin fraction as the primary antibody (control Ab) instead of uPAR or uPA (Fig. 5C). In situ hybridization was performed to determine the presence of transfected plasmid intracranially after intraperitoneal injections. Briefly, ten days after tumor implantation the mice were treated with intraperitoneal injections of control, EV/SV, puPAR, puPA or pU2 (150μg/injection/mouse) every other day three times. Control mice were injected with PBS alone. Five weeks after tumor inoculation, six mice from each group were sacrificed and brains processed as described in Methods. Sections were deparaffinized and probed for pcDNA3-CMV promoter using specific alkaline phosphatase-labeled DNA oligo (CTGGTGTCGACCTGCTTCCGCGATGTACGGGC) per standard protocols. The presence of CMV promoter was determined by the development of a blue precipitate of NBT alkaline phosphatase substrate. Arrows point to region of localization (Fig. 5D). The presence of plasmids intracranially was also determined by PCR amplification of CMV to BGH construct region of the plasmid using deparaffinized intracranial sections of control, EV/SV-, puPAR-, puPA- or pU2-intraperitoneally injected mice. To determine if interferon induction was present intracranially, total RNA was isolated from fresh or paraffin-embedded brain tissue from mice injected with control, EV/SV, puPAR, puPA, pU2, or interferon (0.5ng) intracranially and RT-PCR was performed using primers specific for OAS1 (Fig. 5E). Nude mice were implanted with intracranial xenograft tumors and their survival ability was determined. Two sets of animals were used (6 mice/group). Both sets of mice were implanted with intracranial xenograft tumors as described previously. Ten days after tumor implantation, the mice were treated with intraperitoneal injections of pU2 plasmid (150μg/injection/mouse) three times every other day. Control mice were either injected with PBS alone or with empty plasmid vector (150μg/injection/mouse). The mice were maintained in clean room conditions and monitored every day for 112 days after which the experiment was artificially terminated. Brains were harvested, paraffin embedded, sectioned, and H&E stained as per standard protocols. Survival curve was plotted per standard methods and graphically represented (Fig 5F).

DISCUSSION

In the present study, we provide evidence to suggest that the direct intraperitoneal injections of plasmids expressing siRNA targeting uPAR and uPA inhibit intracranial tumor growth in nude mice. Extracellular matrix (ECM) destruction is dependent on the expression of proteases, which are known to be overexpressed in gliomas (34–37). Using an adenovirus-mediated antisense construct for uPAR and uPA, we have already demonstrated that the simultaneous downregulation of uPAR and uPA has a synergistic effect rather than an additive one (38). However, using an adenovirus to deliver therapeutic genes presents significant problems, including accurate targeting and toxicity (39). In recent years, several groups have experimented with various modifications, but so far, none of these altered methods have been successful in eliminating the aforementioned problems (40). Here, we demonstrate that a simple cytomegalovirus plasmid vector driving the production of hairpin-like RNA molecules can be utilized therapeutically. As indicated by the RT-PCR results, the use of an mRNA-like molecule possessing a poly A tail and having 21 bp inverted repeats, which target uPAR and uPA, did not induce an interferon-like response. Other groups have used lentiviral vectors to successfully target specific proteases, such as PAI-1, but still induced OAS1 (41). Here, we did not observe the induction of OAS1. This might have been due to the presence of a poly A tail mimicking cellular mRNA and appearing as “self” to the cell. As expected, the downregulation of the target mRNA molecule was observed with siRNA, whereas no downregulation was observed with the use of an antisense sequence. With an antisense approach, equimolar quantities of the antisense molecule are required to silence the target gene. This is not required with RNAi where the RISC behaves like a catalyst and is reused (42). In essence, a small amount of RNAi inducing molecules, such as siRNA or hpRNA, is sufficient to induce silencing of target genes. Our results demonstrate that future therapeutic use of RNAi to treat gliomas is very probable. Urokinase plasminogen activator and its receptor play important roles in invasion and migration of metastatic gliomas.(35–37) Using these proteases as targets is very attractive for the treatment of gliomas.

In terms of angiogenesis, the use of anti-angiogenic drugs has been met with mixed success (43). The targeting of uPAR and uPA, which are also indirectly involved in angiogenic pathways (44, 45) has revealed the importance of their targeting in cancer therapy. Our results show that the simultaneous downregulation of uPAR and uPA causes the downregulation of EGFR and VEGF in EGFR-overexpressing glioma xenograft cells (4910). In situ studies confirm western blot analysis results in also demonstrating the downregulation of EGFR and VEGF. In situ angiogenic assays have shown that endothelial cells co-cultured with EGFR-overexpressing 4910 cells induce the endothelial cells to form a network-like pattern mimicking tumor angiogenesis. Progressive reduction in the network formation was seen in puPAR- and puPA-transfected cells. In cells transfected with pU2, we observed complete regression of network formation, indicating that the simultaneous downregulation of uPAR and uPA causes the tumor cells to retard or stop secreting factors necessary for the induction of angiogenesis. The dorsal skin fold assay, an in vivo angiogenic assay, revealed complete inhibition of angiogenesis by pU2-transfected 4910 cells. Although our previous work using an adenovirus-mediated strategy had similar results (38), the major difference is that 100 MOI of virus particles were required to achieve the same effect as 6 μg of plasmid. We also observed similar results with the other assays. For example, spheroid migration was significantly inhibited in pU2-treated SNB19 and 4910 xenograft cells. Our invasion studies demonstrated that after transfection with pU2, both SNB19 and 4910 cells exhibited a significant reduction in their invasive ability, only 5%–8% invasion when compared to the controls. From these spheroid invasion assay results, it is clear that the simultaneous downregulation of uPAR and uPA retards the invasion of fetal rat brain aggregates.

uPAR is associated with integrins and is known to mediate cellular motility via ECM components such as vitronectin (46). In addition, uPAR’s association with integrins mediates ras signaling pathways (47, 48). uPAR and uPA have also been observed at the leading edge of invading tumors (48). Hence, uPAR is a logical target to inhibit tumor invasion and migration. Our animal studies demonstrate that the simultaneous downregulation of uPAR and uPA causes the regression of intracranial tumors. Nude mice implanted with 4910 xenograft cells intracranially usually die in 4 weeks due to tumor invasion. In contrast, mice injected with pU2 intraperitoneally do not exhibit tumor establishment and survived for over 112 days after implantation. It would be of academic interest to see if similar results can be obtained using antisense vector constructs. Nevertheless, as the in situ hybridization studies indicated, there was translocation of the intraperitoneally injected plasmid to the brain. The plasmids may pass through the blood brain barrier (BBB) probably due to the already compromised BBB at the tumor site. The presence of the plasmid intracranially in control was seen primarily surrounding vessels (not shown). As such, the potential for using siRNA vectors for future therapy is promising. Agrawal and Iyer (49) have reviewed the advantages and disadvantages of using an antisense strategy for therapeutic purposes. Our results demonstrate that, in spite of being injected intraperitoneally, siRNA-expressing plasmids localize intracranially and effectively downregulate uPAR and uPA. RT-PCR of the brain tissue showed that even though plasmid localization was observed in the brain, no OAS1 induction was detected. This indicates that the presence of a poly A tail probably prevented the induction of an interferon-like response. In conclusion, the RNAi-mediated downregulation of uPAR and uPA has clear clinical implications for the treatment of gliomas as well as other cancers.

Abbreviations

uPA[R]

urokinase-type plasminogen activator [receptor]

CMV

cytomegalovirus

BGH

bovine growth hormone

hpRNA

hairpin RNA

PCR

polymerase chain reaction

PBS

phosphate-buffered saline

FITC

fluoresceine-5-isothiocyanate

DiI

1'-dioctadecyl-3,3,3',3'-tetramethylindocarbocyanineperchlorate

DiO

3,3'-dioctadecyloxacarbocyanine perchlorate

GFP

green fluorescent protein

ECM

extracellular matrix

FRBA

fetal rat brain aggregrate

EGFR

epidermal growth factor receptor

VEGF

vascular endothelial growth factor