Inhibition of cathepsin B and MMP-9 gene expression in glioblastoma cell line via RNA interference reduces tumor cell invasion, tumor growth and angiogenesis

Article metrics


Extracellular proteases have been shown to cooperatively influence matrix degradation and tumor cell invasion through proteolytic cascades, with individual proteases having distinct roles in tumor growth, invasion, migration and angiogenesis. Matrix metalloproteases (MMP)-9 and cathepsin B have been shown to participate in the processes of tumor growth, vascularization and invasion of gliomas. In the present study, we used a cytomegalovirus promoter-driven DNA template approach to induce hairpin RNA (hpRNA)-triggered RNA interference (RNAi) to block MMP-9 and cathepsin B gene expression with a single construct. Transfection of a plasmid vector-expressing double-stranded RNA (dsRNA) for MMP-9 and cathepsin B significantly inhibited MMP-9 and cathepsin B expression and reduced the invasive behavior of SNB19, glioblastoma cell line in Matrigel and spheroid invasion models. Downregulation of MMP-9 and cathepsin B using RNAi in SNB19 cells reduced cell–cell interaction of human microvascular endothelial cells, resulting in the disruption of capillary network formation in both in vitro and in vivo models. Direct intratumoral injections of plasmid DNA expressing hpRNA for MMP-9 and cathepsin B significantly inhibited established glioma tumor growth and invasion in intracranial tumors in vivo. Further intraperitoneal (ip) injections of plasmid DNA expressing hpRNA for MMP-9 and cathepsin B completely regressed pre-established tumors for a long time (4 months) without any indication of these tumor cells. For the first time, these observations demonstrate that the simultaneous RNAi-mediated targeting of MMP-9 and cathepsin B has potential application for the treatment of human gliomas.


RNA interference (RNAi) is the process of sequence-specific, post-transcriptional gene silencing in animals and plants, initiated by double-stranded RNA (dsRNA) that is homologous in sequence to the silenced gene. The mechanism of RNAi is not yet fully understood, but recent genetic and biochemical studies have revealed some details at the molecular level. dsRNA are processed into small 21–23 bp dsRNAs (siRNAs) by a dsRNA-specific RNase DICER. Subsequently, these small fragments, known as small interfering RNA (siRNA), are incorporated into the RNA-induced silencing complex (RISC) (Hammond et al., 2001). The active complexes containing the guide RNA recognize and to cleave the target RNA. RNA hydrolysis occurs within the region of homology directed by the original siRNA, thereby selectively inhibiting expression of the target gene. Elbashir et al. (2001) showed that synthetic 21–23 nucleotide siRNA could induce efficient RNAi in mammalian cells. Since then, in order to circumvent the high cost of synthetic siRNA and establish stable gene knockdown cell lines by siRNA, several plasmid vector systems have been designed to produce siRNA inside cells driven by RNA polymerase III-dependent promoters such as U6 and H1-RNA. In addition, RNAi has been shown to be potent since only a few molecules of dsRNA per cell were necessary to trigger gene silencing throughout the treated animal (Zamore, 2001).

Tumor progression involves modulation of tumor cell adhesion during migration and extracellular matrix (ECM) degradation during invasion. An intricate balance of proteases, their activators and inhibitors regulate both these processes during tumor invasion. ECM-degrading proteinases can be divided into three classes: cysteine proteinases, matrix metalloproteases (MMPs) and serine proteinases. Various studies have clearly demonstrated a close correlation between the mRNA and protein levels of the cysteine proteinase cathepsin B and the invasive potential of tumors (Sloane et al., 1986). The proteolytic activity of cathepsin B has been suggested to facilitate direct degradation of ECM proteins, including fibronectin, types I and IV collagen and Laminin (Guinec et al., 1993). In addition, cathepsin B has an indirect role that involves the activation of other enzymes, including MMPs and both the soluble and receptor-bound forms of the serine protease urokinase plasminogen activator (uPA) (Kobayashi et al., 1992, 1993; Guinec et al., 1993). MMPs and uPA have been shown to modulate the proteolytic cascade that mediates ECM degradation. Our previous studies demonstrated that the stable clones of antisense cathepsin B were less invasive in in vitro models and formed undetectable, small tumors in nude mice (Mohanam et al., 2001).

MMPs are implicated in cancer invasion through ECM degradation and the processing of a range of molecules, including growth factors and cytokines. These processes are involved in promoting all aspects of tumor growth, such as cell proliferation, adhesion and dispersion, migration, differentiation, angiogenesis, apoptosis and host defense evasion (Egeblad and Werb, 2002). Glioblastoma (GBM) cells secrete MMP-9 and MMP-2, and their mRNA and protein levels are found to be elevated in patient biopsy tissue (Rooprai et al., 1998; Kachra et al., 1999; Raithatha et al., 2000). MMP-9 is frequently upregulated in cancer cells as well as the adjacent host tissues (Himelstein et al., 1994). We have previously reported that MMP-9 levels were significantly correlated with the histological grade of malignancy (Rao et al., 1993). We have also shown that MMP-9 is important in endothelial cell morphogenesis and capillary formation in glial/endothelial cocultures in vitro (Chandrasekar et al., 2000). Furthermore, antisense oligonucleotides that blocked MMP-9 gene expression dramatically reduced the invasive phenotype of SNB19 GBM cells. These results show that MMP-9 expression facilitates glioma invasion in vivo (Kondraganti et al., 2000; Lakka et al., 2002).

Malignant GBM is typically unresponsive to conventional treatments. Glioma cells infiltrate widely into the surrounding normal white matter making complete surgical resection of these tumors virtually impossible. Some of the key molecules involved in glioma progression are proteases such as cathepsins, uPA and various metalloproteases, and thus, constitute important therapeutic targets. Previous methodologies that have been used include antisense technology and catalytic ribozymes. However, the efficacy of these approaches has been mixed, particularly the in vivo treatments. The stability/specificity of the oligonucleotides and/or problems in target sequence selection hamper the widespread use of these approaches (Jen and Gewirtz, 2000). More promising is the selective degradation of the corresponding mRNAs by RNAi (Elbashir et al., 2001). Given the effectiveness of dsRNA-mediated gene silencing, we wanted to assess whether this mechanism could be used to specifically downregulate two different proteases involved in glioma invasion. Our results demonstrate efficient and specific knockdown of MMP-9 and cathepsin B in a glioma cell line with the concomitant decrease in pre-established tumor growth, invasion and angiogenesis. These results indicate the potential applicability of RNAi for cancer gene therapy.


Gene-specific siRNAs lower expression of MMP-9 and cathepsin B protein in a glioma cell line

To test the effectiveness of simultaneously inhibiting two endogenous genes with a hairpin siRNA expression vector, we constructed a vector-expressing siRNA for cathepsin B (732–753 bases of human cathepsin B mRNA) and MMP-9 (360–381 bases of human MMP-9 mRNA) under the control of the human cytomegalovirus (CMV) promoter (pCM) (Scheme 1). Figure 1a demonstrates that transfection with pC and pCM vector specifically inhibited cathepsin B levels compared to mock, empty and pM vector controls. β-Actin levels assessed in the same blot indicated that the inhibition of cathepsin B was specific and confirmed equal sample loading. MMP-2 and MMP-9 levels were determined in the conditioned medium in the transfected cells. The amount of MMP-9 released from the mock and empty vector (EV) transfected cells were the same. Cells transfected with pM and pCM vector expressed very low levels of MMP-9 compared to the mock, EV and pC controls. There was no change in the expression of MMP-2 demonstrating the sequence-specific inhibition of the pM and pCM vector (Figure 1b). To confirm that the decrease in MMP-9 activity was due to a decrease in protein expression, the conditioned medium was analysed using immunoblotting with an MMP-9-specific antibody. MMP-9 protein band was decreased dramatically by immunoblotting of the conditioned medium from cells transfected with pM and pCM vector, but bands were significantly much higher in the conditioned medium from the cells infected with the EV or with pC vector (Figure 1c).

Scheme 1

Schematic representation of siRNA expression for cathepsin B and MMP-9 from pCM vector. pCDNA 3 plasmid constructs were developed having two complimentary inverted repeats driven by a CMV promoter directed against cathepsin B and MMP-9. The CMV promoter would drive the formation of a dual hairpin structure which, in turn, was processed by the double-strand RNA recognizing enzyme DICER to form viable SiRNA-molecules. Stability of the dual hairpin molecule was ensured because of the secondary structure of the molecule which is reminiscent of an mRNA molecule having a poly-A tail driven by a bovine growth hormone (BGH) poly-a-signal sequence

Figure 1

RNA interference decreased cathepsin B and MMP-9 levels in SNB19 cells. Total cell lysates and serum-free medium were collected from SNB19 cells transfected with mock, EV or a vector-encoding siRNA for MMP-9 (pM) and cathepsin B (pC) and together (pCM). Subsequently, 30 μg of protein from these samples were separated under nonreducing conditions on 8–12% SDS–PAGE and transferred onto nitrocellulose membranes. The membranes were probed with antibodies for Cathepsin B (a) and MMP-9 (c) and with appropriate secondary antibody (horseradish peroxidase conjugate) and developed according to the manufacturer's protocol (Amersham, Arlington Heights, IL, USA). β-Actin was simultaneously immunodetected to verify the loading of similar amounts of cell lysates. MMP-9 activity of SNB19 cells infected with EV, pC, pM or pCM vector for 3 days in serum-free medium and were determined by gelatin zymography (b)

Inhibition of tumor cell-induced capillary network formation by pCM vector

The growth of a glial tumor depends on the induction of new capillary blood vessels as this is necessary to support the developing tumor mass. We used a coculture system in which microvascular endothelial cells were induced by glial cells to form capillary-like structures in order to examine the RNAi-mediated suppression of cathepsin B and MMP-9. SNB19 cells induced endothelial cells to differentiate into capillary-like structures within 72 h. In contrast, transfection of SNB19 cells with the vector expressing siRNA for cathepsin B and MMP-9 completely inhibited tumor cell-induced microvessel morphogenesis (Figure 2a). Further quantification of the branch points and number of branches were undetectable in pCM transfected cocultures compared to mock and EV (Figure 2b). Further, the effect was only 50% in pC or pM-treated cocultures when compared to pCM vector in relation to capillary-like structures. To confirm the in vitro coculture experiments, we examined whether the pCM vector could inhibit tumor angiogenesis in vivo as assessed by the dorsal window model. Implantation of a chamber containing mock and EV-transfected SNB19 cells resulted in the development of microvessels (as indicated by arrows) with curved thin structures and many tiny bleeding spots. In contrast, implantation of SNB19 cells transfected with the pCM vector did not result in the development of any additional microvessels (Figure 2c).

Figure 2

RNAi inhibits tumor cell-induced capillary network formation. SNB19 cells were transfected with mock, EV, pM, pC and pCM for 24 h. Then, cells were cocultured with human dermal endothelial cells for 48 h. After incubation, cells were fixed and blocked with 2% bovine serum albumin for 1 h and endothelial cells were probed with antibody for factor VIII antigen (Factor VIII antibody, DAKO Corporation, Carpinteria, CA, USA) and examined under a fluorescent microscope after probing with an appropriate FITC-congugated secondary antibody. Endothelial cells grown in the presence of SNB19 (control, empty vector, pM, pC or pCM transfected) conditioned media were H and E stained and photographed (see Materials and methods) (a). Quantification of angiogenesis in cocultures infected with mock, EV or pCM vector as described in Materials and methods (b). Inhibition of tumor angiogenesis in SNB19 cells infected with pCM vector by mouse dorsal skin-fold assay as described in Materials and methods (c). PV-pre-existing vasculature, TN-tumor-induced vasculature. Photographs were taken using light microscopy (upper panel) and for FITC fluorescence (lower panel) to determine the presence of newly developed vasculture

Suppressive effects of pCM vector on glioma migration and invasion

Cell migration requires the coordinated regulation of cell–cell attachments, cell–matrix attachment and matrix remodeling. We studied the influence of suppressing cathepsin B and MMP-9 on the capacity of the cells to migrate on vitronectin in a spheroid migration assay. Multicellular glioma spheroids were grown from SNB19-green fluorescent protein (GFP) cells in six-well plates coated with agarose. After checking for viability using morphology and trypan blue exclusion, spheroids of similar diameter (100–200 μm) were transfected with mock, EV or the pCM vector expressing siRNA for cathepsin B and MMP-9. After 3 days, single spheroids were placed on vitronectin-coated plates and allowed to migrate. Figure 3a indicates that cells from the control spheroids and spheroids infected with the EV showed a significantly higher capability of cells to migrate as compared to the pCM vector-infected cells. Proteolytic degradation of ECM components is critical for tumor cell invasion. To investigate whether expression of siRNA for cathepsin B and MMP-9 plays a role in glioma invasiveness, we compared the invasive ability of SNB19 cells transfected with the pCM vector to those cells infected with mock and EV. SNB19 cells transfected with mock and EV invaded through Matrigel more extensively compared to the pCM vector-transfected cells penetrated through the matrigel (Figure 3b).

Figure 3

RNAi inhibits glioma cell migration and invasion. SNB19 GFP spheroids were infected with mock, EV and a vector encoding siRNA for cathepsin B and MMP-9 (pCM). After 3 days, single glioma spheroids were placed in the center of a vitronectin-coated well in a 96-well plate and cultured for 48 h. At the end of the migration assay, spheroids were fixed and photographed (a). SNB19 cells were trypsinized 3 days after transfection with mock, EV and a vector encoding siRNA for cathepsin B and MMP-9 (pCM), washed with PBS and resuspended in serum-free medium. Invasion assays were carried out in a 12-well transwell unit (Costar, Cambridge, MA, USA) on polycarbonate filters with 8 μm pores coated with Matrigel. After a 24 h incubation period, the cells that had passed through the filter into the lower wells were stained, counted and photographed under a light microscope (b). Spheroids of SNB19 cells were transfected with mock, EV and a vector-encoding siRNA for cathepsin B and MMP-9 (pCM) and stained with DiI and cocultured with DiO-stained fetal rat brain aggregates. Progressive destruction of fetal brain aggregates by tumor spheroids was observed (c).

We further examined the extent of suppressive effects of the siRNA in a spheroid invasion assay. A potential advantage of using glioma spheroids is that tumor cells grown in three-dimensional cultures have been shown to exhibit properties that more closely resemble those of tumors in vivo. Mock- and EV-transfected spheroids invaded 25% of the normal brain aggregates within 1 day, 50% within 2 days and by 3 days, 95% of the tumor spheroid and brain aggregate had combined into a single entity (Figure 3c). In contrast, glioma spheroids transfected with the pCM vector-expressing siRNA for cathepsin B and MMP-9 remained separate from the normal brain aggregates.

RNAi induces complete regression of GBM tumors in nude mice

The capacity of the siRNA for MMP-9 and cathepsin B to inhibit regression of intracranial SNB19 tumors was tested in nude mice. Mice with pre-established glioma growth were stereotactically injected with phosphate-buffered saline (PBS) (mock), empty vector (EV), pC, pM and pCM vector. Brain sections of mice treated with mock and EV showed rapid tumor growth whereas mice injected with the pCM vector using mini osmotic pumps into a pre-established tumor resulted in complete inhibition of tumor growth over a 5-week time period (Figures 4a, b). Quantification of tumor size showed a total regression of tumor in the pCM vector-treated group compared to the mock or EV (Figure 4c). Brain sections of mice treated with pC or pM vector-treated group resulted in around 50% tumor regression compared to control groups. Intraperitoneal (ip) injections of the vector also resulted in complete regression of pre-established intracranial tumor growth with no indication of tumor cells for long period of several months (Figure 4d). Thus, RNAi was able to completely eradicate malignant glioma tumor growth in this nude mouse model.

Figure 4

RNAi-mediated regression of pre-established tumor growth. SNB19 GFP tumor cells were injected intracerebrally with the help of a stereotactic frame into nude mice. After 1 week, either an EV or a vector expressing siRNA for cathepsin B and MMP-9 (pCM), cathepsin B (pC) or MMP-9 (pM) was injected into the brain using an Alzet mini osmotic pump. Photographs of tumor sections were observed for GFP fluorescence (a) and subsequently, stained with hematoxylin and eosin (b). Semiquantification of tumor volume in mock/EV, pM, pC and pCM vector-treated groups after 5 weeks was done as described in Materials and methods. Data shown are the ±s.d. values from six animals from each group (*P<0.001) (c). In another experiment, 10 days after intracerebral injection, pCM vector was injected intraperitoneally twice and the animals killed after 4 months (d)


Tumor cell invasion involves attachment of tumor cells to the underlying basement membrane, local proteolysis and migration of tumor cells through the proteolytically modified region (Liotta et al., 1986). ECM degradation by proteases such as cysteine proteases and MMPs is critical to malignant progression. Antisense DNA and oligonucleotides are a convenient approach for the selective downregulation of protein expression. In a recent study, siRNA was quantitatively more efficient than antisense oligonucleotides at suppressing cotransfected GFP expression both in vitro and in vivo (Bertrand et al., 2002). Limitations to the use of RNAi in mammalian cells derive from sequence-nonspecific responses to long dsRNA, such as induction of interferon synthesis (Stark et al., 1998) and apoptosis (Zamore, 2001). The recent development of 21-nucleotide siRNA duplexes has circumvented these problems and allowed successful RNAi in cultures of several types of mammalian cells (Elbashir et al., 2001). An important potential application for RNAi in mammals is the simultaneous inhibition of multiple genes in somatic cells. Previous observations demonstrated that inhibition of a synthetic reporter by a transfected hairpin siRNA vector was constant even in the presence of another hairpin siRNA vector, indicating that there is no significant competition between cotransfected hairpin siRNA vectors. In the present study, we demonstrated that effective simultaneous inhibition of two genes using an expression vector encoding for two-hairpin siRNA was feasible in glioma cells and had potential application for glioma therapy.

The invasive potential of cathepsin B has been shown in several kinds of carcinoma cells in vitro. Intracellular and extracellular cathepsin B activity contributes to the in vitro invasiveness of MCF10AT cells (Premzl et al., 2003). Synthetic cysteine proteinase inhibitors, selective for cathepsin B, have been shown to significantly reduce the invasiveness of MCF10AT cells (Bervar et al., 2003). Increased cathepsin B activity was observed in the invasive tumor regions of human colon cancer samples (Emmert-Buck et al., 1994). Plasma membrane binding can also be demonstrated for both cathepsin B and gelatinase B in bone metastasis (Arkona and Wiederanders, 1996). MMP-9 inhibition in transformed keratinocytes and squamous-cell carcinoma cells was accompanied by reduced invasion through collagen and reconstituted basement membranes (Simon et al., 1998). In addition, there was a reduction in the number of lung colonies formed in an experimental metastasis system in MMP-9 null mice (Itoh et al., 1999) as well as through MMP-9 downregulation (Hua and Muschel, 1996). Our present study shows that the CMV promoter-driven expression of siRNA against cathepsin B and MMP-9 (pCM) can successfully silence cathepsin B and MMP-9 expression in the SNB19 GBM cell line, as analysed by Western blotting and gelatin zymography. Our results also demonstrated that the invasive potential of glioma cells treated with the pCM vector was significantly inhibited. Together, these studies establish the significance of MMP-9 and cathepsin B on tumor invasion.

Cancer cells must detach from the neighboring cells and ECM components to migrate and invade. Matrix proteolysis can directly modulate cell–matrix adhesion either by removal of adhesion sites or by exposing a binding site, which in turn may effect cell migration. MMP-9 was shown to bind to CD44, a receptor for extracellular components (e.g. hyaluronic acid, fibronectin, collagen, etc.), thereby localizing the enzyme to the cell surface, which is required for tumor invasion and angiogenesis (Yu and Stamenkovic, 1999). In the case of EJ human bladder-carcinoma cells, E-64, a cathepsin B/cysteine inhibitor proteases abrogates metastasis in mice by abolishing cell migration (Redwood et al., 1992) as well as blocking locomotion of W256 rat carcinosarcoma cells (Boike et al., 1992). In the present study, RNAi-mediated inhibition of cathepsin B and MMP-9 significantly blocked the migration of SNB19 glioma cells as shown in a spheroid migration assay.

The therapeutic effect of suppressing cathepsin B and MMP-9 described in vitro was then achieved in athymic mice defective in immune response. Intratumoral injection of a vector expressing siRNA for cathepsin B and MMP-9 completely inhibited pre-established tumor growth in the SNB19 intracranial model. The sustained suppression of glioma growth could be due to siRNA amplification. It has been previously reported that the siRNA-directed suppression of a cotransfected gene in vivo in mouse liver was maintained for several days (Lewis et al., 2002). Recent work has reported the possibility of knocking-down gene expression in adult mice by injecting synthetic siRNA oligos or a vector system encoding siRNA (McCaffrey et al., 2002). siRNA against cathepsin B and MMP-9 suppressed glioma growth more efficiently than antisense oligodeoxynucleotide for MMP-9 and cathepsin B as described in our previous reports (Mohanam et al., 2001; Lakka et al., 2002). Thus, the control of both cathepsin B and MMP-9 expression has considerable significance for regulation of tumor progression.

Growth maintenance of malignant tumors is closely related with development of the vascular network that supplies the tumor with nutrients. We next assessed the potential antiangiogenic action of siRNA expression of MMP-9 and cathepsin B on microvessel formation. We did not observe the formation of a vascular network characterized by closed polygons and complex mesh-like structures in cells treated with the pCM vector. This network is typically observed when glioma cells are co-cultured with endothelial cells. Proteolysis of ECM components allows endothelial cells to migrate and releases stored angiogenic signaling molecules from the ECM (Bergers et al., 2000). Immunohistochemical analysis demonstrated that cathepsin B was strongly expressed in malignant anaplastic astrocytomas and GBMs as compared to normal brain tissue. Cathepsin B expression has also been detected in association with new vessel formation in malignant gliomas (Mai et al., 2002). A direct role for MMP-9 in the regulation of angiogenesis via modulation of growth factor availability was initially demonstrated in homozygous mice with a null mutation in MMP-9 (Vu et al., 1998). MMP-9 deficiency and/or an MMPI (batimastat) inhibits tumor development and angiogenesis in RIP-TAg islet cell carcinoma model (Bergers et al., 2000). MMP-9-null mice demonstrate reduced development of HPV-induced squamous cell carcinoma (Coussens et al., 2000). Altogether, these results clearly suggest that MMPs can promote angiogenesis, and that absolute lack of MMP activity can prevent new blood vessel formation. However, recent reports have shown that MMP-9 plasma levels control endothelial cell proliferation and tumor vascularization, and that in vitro, this activity is dependent on angiostatin generation from plasminogen, while in vivo there is a correlation between angiostatin and MMP-9 levels (Pozzi et al., 2002). This provides an explanation for the recently observed failures of anti-MMP therapy in tumor treatment. The tumor regression achieved by the combined treatment in the present study could be most likely attributable to the complementary actions of cathepsin B and MMP-9. Targeting expression of cathepsin B and MMP-9 in tumor cells may therefore be an effective approach to control angiogenesis and tumor growth.

In conclusion, our study has demonstrated the anticancer efficacy of RNAi-mediated inhibition of cathepsin B and MMP-9. Comparison of the suppressive effects of antisense oligonucleotides and siRNAs directed against the same targets in mammalian cells revealed that the IC50 value for the siRNA was about 100-fold lower than that of the antisense oligonucleotides (Miyagishi et al., 2003). The ability of siRNA to silence sequence-specific target genes and the lower concentrations required to inhibit gene expression make RNAi a powerful tool for gene therapy.

Materials and methods

Construction of hpRNA-expressing vector

To develop vector capable of producing hairpin siRNA molecules for cathepsin B and MMP-9, we used the mammalian expression plasmid vector pCDNA 3. Self-complimentary inverted repeat sequences spaced by a nine base GC-deficient region targeted to cathepsin B (732–753) and MMP-9 (360–381) were synthesized. Oligos for cathepsin B were terminated with XhoI sites and the oligos for MMP-9 were terminated with EcoRI and self-annealed by heating to 100°C for 5 min and cooled to room temperature in 6 × SSC which would result in the formation of dsDNA molecules with the respective sticky restriction site ends. These dsDNA molecules were ligated to the XhoI and EcoRI sites of the pCDNA plasmid vector, resulting in the formation of a plasmid containing inverted repeats for cathepsin B and MMP-9 downstream of the CMV promoter and terminated by a SV40 terminator. The resultant plasmid termed pCM transfected to mammalian cells would result in the production of a dual hairpin siRNA molecule targeted both to cathepsin B and MMP-9 which would be further processed by a dsRNA recognizing enzyme (DICER) to produce individual siRNA molecules to induce RNAi (Scheme 1).

Cell culture and transfections

High-grade human glioma cells (SNB19) were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 1% glutamine, 100 μg/ml streptomycin, 100 U/ml penicillin, and 10% fetal bovine serum (pH 7.2–7.4) in a humidified atmosphere containing 5% CO2 at 37°C. Transfections were performed with Lipofectamine™ 2000 reagent (Life Technologies, Rockville, MD, USA) using 1–2 μg of expression vector/ml serum-free medium as described by the manufacturer. After 5 h of transfection, the medium was replaced by serum-containing medium and incubated for a further 48 h.

Gelatin zymography

Gelatin-substrate gel electrophoresis was accomplished as described previously (Lakka et al., 2000). SNB19 cells were transfected with mock, an EV or a vector expressing siRNA for cathepsin B (PC), MMP-9 (PM) and cathepsin B and MMP-9 (pCM). To prepare conditioned medium, cells were washed with serum-free medium and resupplied with fresh serum-free medium. After 24 h, the conditioned medium was harvested, centrifuged to remove cellular debris and protein concentrations were determined. Equal amounts of protein were electrophoresed under nonreducing conditions through 8% SDS–PAGE containing 0.1% gelatin. Gels were washed in 2.5% Triton X-100 and incubated overnight in Tris-CaCl2 buffer. The gels were then stained with 0.2% Coomassie blue for 1 h and destained in 20% methanol and 10% acetic acid. The clear bands represented gelatinase activity.

Western blotting

SNB19 cells were transfected with mock, EV, pC, pM and pCM vector and cultured for 48 h. At the end of incubation cells were harvested, washed twice with cold PBS and lysed in buffer (150 mM NaCl, 50 mM Tris-HCl, 2 mM EDTA, 1% NP-40, pH 7.4) containing protease inhibitors.

Western blot analysis of MMP-9 protein expression in conditioned medium in mock-, EV- pC- pM- and pCM-infected SNB19 cells was performed as described above. Total proteins (30 μg/lane) were resolved on a 10% SDS-polyacrylamide gel and transferred onto a nylon membrane and incubated with anti-cathepsin B (Pharmingen, San Diego, CA, USA) anti-MMP-9 (Santa Cruz Biotechnology, Santa Cruz, CA, USA) anti-actin (Sigma, St Louis, MO, USA) monoclonal antibodies, followed by incubation with horseradish-peroxidase conjugated anti-mouse IgG secondary antibody (Sigma St Louis, MO, USA). The bands were visualized using the enhanced chemiluminescence system (Pierce, Rockford, IL, USA) according to the manufacturer's protocol and exposed to radiographic film (Eastman Kodak, Rochester, NY, USA).

Cell migration assay

SNB19-GFP cells (2 × 105/ml) were grown as multicellular tumor spheroids on 100 mm tissue culture plates coated with 0.75%. agar. Spheroids measuring 100–150 μm in diameter (4 × 104 cells/spheroid) were selected and infected with either mock or EV or vector expressing siRNA cathepsin B and MMP-9 (pCM). After 2 days, a single glioma spheroid was placed in the center of each well of vitronectin-coated 96-well microplate and 200 μl of serum-free medium was added to each well. Spheroids were cultured at 37°C for 72 h, after which the spheroids were fixed and stained with Hema-3. Cell migration from the spheroids was assessed under light microscopy. The migration of cells from spheroids to monolayers was used as an index of cell migration and was measured using a microscope calibrated with a stage and ocular micrometer.

Matrigel invasion assay

The invasiveness of the transfected SNB19 cells was tested in vitro with the Boyden chamber invasion assay after transfection with either the EV or the vector expressing siRNA for cathepsin B and MMP-9 (pCM). Briefly, transwell inserts with 8 μm pores were coated with Matrigel (0.7 mg/ml) (Collaborative Research, Inc., Boston, MA, USA). SNB19 cells were trypsinized and 500 μl of the cell suspension (1 × 106 cells/ml) was added to the wells in triplicate. After incubation for 24 h at 37°C, cells that passed through the filters into the lower wells were quantified and expressed as a percentage of the sum of the cells in the upper and lower wells (Mohanam et al., 2001). Cells on the lower side of the membrane were fixed, stained with Hema-3 and photographed.

Spheroid invasion assay

SNB19 cells (3 × 105/ml) and fetal rat brain cells (2 × 106/ml) were cultured in low-attachment 35 mm Petri dishes with constant shaking at 60 r.p.m. until multicellular spheroids were formed. Spheroids with a diameter of 100–200 μm were transfected with mock, EV and vector expressing siRNA for cathepsin B and MMP-9 (1 μg/ml). After 3 days, tumor spheroids were stained with the fluorescent dye 1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanineperchlorate (DiI) and confronted with fetal rat brain aggregates stained with 3,3′-dioctadecyloxacarbocyanine perchlorate (DiO). Progressive destruction of fetal rat brain aggregates and invasion of SNB19 cells were observed with confocal laser scanning microscope and photographed as described previously (Go et al., 1997). The remaining volume of brain aggregates or tumor spheroids during the cocultures in the presence of these constructs was determined as described previously (Go et al., 1997).

In vitro angiogenic assay

SNB19 cells (2 × 104) were seeded in eight-well chamber slides and transfected with mock, an EV or a vector expressing siRNA for cathepsin B and MMP-9 (pCM). After a 48 h incubation period, human dermal endothelial cells (4 × 104) (Center for Disease Control and Prevention, Atlanta, GA, USA) were seeded and cocultured for another 72 h. Endothelial cells were stained with factor VIII antigen (Dako Corporation, Carpinteria, CA, USA) for 1 h after cells were fixed in 3.7% formaldehyde and blocked with 2% bovine serum albumin. The cells were washed with PBS and incubated with fluoresceine-5-isothiocyanate (FITC)-conjugated secondary antibody for 1 h. The specimens were then washed, examined and photographed with fluorescence microscopy. Further, endothelial cells were cultured as a monolayer on eight-well chamber slides and incubated with conditioned media from SNB19 cells infected with the appropriate vectors. After 72 h of incubation these cell cultures were stained. Following the 72 h seeding and coculturing process, Image Pro software was used for quantification of angiogenesis. The degree of angiogenesis was measured by the following method: number of branch points and the total number of branches per point were counted, with the product indicating the degree of angiogenesis.

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 (Leunig et al., 1992). Sterile small-animal surgical techniques were followed. Mice were anesthetized by i.p. injection with ketamine (50 mg/kg) zylazine (10 mg/kg). Once the animal was anesthetized completely, a dorsal air sac was made in the mouse by injecting 10 ml of air. Diffusion chambers (Fisher) were prepared by aligning a 0.45-μm 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. PBS (20 μl) was used to wet the membranes. SNB (2 × 106) 19 cells (mock, EV or the pCM transfected), 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 –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 were anesthetized with ketamine/xylazine and killed by intracardiac perfusion with saline (10 ml) followed by a 10 ml of 10% formalin/0.1 M phosphate solution and followed by 0.001% FITC solution in PBS. The animals were carefully skinned around the implanted chambers, and the implanted chambers were removed from the s.c. air fascia. The skin fold covering the chambers were photographed under visible light and for FITC fluorescence. The number of blood vessels within the chamber in the area of the air sac fascia were counted and their lengths measured.

Intracranial injections

SNB19-GFP cells were trypsinized and resuspended in serum-free medium at a concentration of 2 × 105 × cells/μl. Mice were anesthetized with an i.p. injection of a 0.35–0.45 μl solution consisting of 0.06 M 2,2,2-tribromethanol (Aldrich Chemical Co., Milwaukee, WI, USA), 1.25 isoamyl alcohol, and 98.5% bacteriostatic saline. Then, the mice were injected intracerebrally with a 10 μl aliquot (2 × 105 cells/μl) with the aid of a stereotactic frame as described elsewhere (Konduri et al., 2001). After 10–12 days, mice were separated into three groups and mock (PBS), EV (150 μg) or pCM vector (150 μg) were injected into the brain using Alzet mini pumps at the rate of 0.25 μl/h (six mice/group). After 5 weeks, the mice were killed by intracardiac perfusion, first with PBS and then with 4% paraformaldehyde in normal saline. The brains were removed, placed in 4% paraformaldehyde for 4 h, and then incubated for 2 days in 30% sucrose in PBS at 4°C. The following day, the brains were sectioned, embedded in microscopic slides, and frozen at −20°C. Cryostat sections were screened for GFP fluorescence to examine tumor growth. The sections were evaluated by a neuropathologist who was blinded as to the treatment group and scored semiquantitatively for tumor size, as described previously (Konduri et al., 2001). In another set of experiments, after 10 days of GBM cells injected intracerebrally, the animals received pCM vector intraperitoneally (two injections at 10 and 13 days). The animals were then sacrificed at 4 months and the sections were analysed as described above.



matrix metalloproteases -9


RNA interference vector expressing siRNA for cathepsin B and MMP-9 (pCM)




simian virus type 40


polymerase chain reaction


phosphate-buffered saline






3,3′-dioctadecyloxacarbocyanine perchlorate


green fluorescent protein


extracellular matrix


  1. Arkona C and Wiederanders B . (1996). Biol. Chem., 377, 695–702.

  2. Bergers G, Brekken R, McMahon G, Vu TH, Itoh T, Tamaki K, Tanzawa K, Thorpe P, Itohara S, Werb Z and Hanahan D . (2000). Nat. Cell Biol., 2, 737–744.

  3. Bertrand JR, Pottier M, Vekris A, Opolon P, Maksimenko A and Malvy C . (2002). Biochem. Biophys. Res. Commun., 296, 1000–1004.

  4. Bervar A, Zajc I, Sever N, Katunuma N, Sloane BF and Lah TT . (2003). Biol. Chem., 384, 447–455.

  5. Boike G, Lah T, Sloane BF, Rozhin J, Honn K, Guirguis R, Stracke ML, Liotta LA and Schiffmann E . (1992). Melanoma Res., 1, 333–340.

  6. Chandrasekar N, Jasti S, Alfred-Yung WK, Ali-Osman F, Dinh DH, Olivero WC, Gujrati M, Kyritsis AP, Nicolson GL, Rao JS and Mohanam S . (2000). Clin. Exp. Metast., 18, 337–342.

  7. Coussens LM, Tinkle CL, Hanahan D and Werb Z . (2000). Cell, 103, 481–490.

  8. Egeblad M and Werb Z . (2002). Nat. Rev. Cancer, 2, 161–174.

  9. Elbashir SM, Harborth J, Lendeckel W, Yalcin A, Weber K and Tuschl T . (2001). Nature, 411, 494–498.

  10. Emmert-Buck MR, Roth MJ, Zhuang Z, Campo E, Rozhin J, Sloane BF, Liotta LA and Stetler-Stevenson WG . (1994). Am. J. Pathol., 145, 1285–1290.

  11. Go Y, Chintala SK, Mohanam S, Gokaslan Z, Venkaiah B, Bjerkvig R, Oka K, Nicolson GL, Sawaya R and Rao JS . (1997). Clin. Exp. Metast., 15, 440–446.

  12. Guinec N, Dalet-Fumeron V and Pagano M . (1993). Biol. Chem. Hoppe Seyler, 374, 1135–1146.

  13. Hammond SM, Boettcher S, Caudy AA, Kobayashi R and Hannon GJ . (2001). Science, 293, 1146–1150.

  14. Himelstein BP, Canete-Soler R, Bernhard EJ, Dilks DW and Muschel RJ . (1994). Invas. Metast., 14, 246–258.

  15. Hua J and Muschel RJ . (1996). Cancer Res., 56, 5279–5284.

  16. Itoh T, Tanioka M, Matsuda H, Nishimoto H, Yoshioka T, Suzuki R and Uehira M . (1999). Clin. Exp. Metast., 17, 177–181.

  17. Jen KY and Gewirtz AM . (2000). Stem Cells, 18, 307–319.

  18. Kachra Z, Beaulieu E, Delbecchi L, Mousseau N, Berthelet F, Moumdjian R, Del Maestro R and Beliveau R . (1999). Clin. Exp. Metast., 17, 555–566.

  19. Kobayashi H, Moniwa N, Sugimura M, Shinohara H, Ohi H and Terao T . (1993). Biochim. Biophys. Acta, 1178, 55–62.

  20. Kobayashi H, Ohi H, Sugimura M, Shinohara H, Fujii T and Terao T . (1992). Cancer Res., 52, 3610–3614.

  21. Kondraganti S, Mohanam S, Chintala SK, Kin Y, Jasti SL, Nirmala C, Lakka SS, Adachi Y, Kyritsis AP, Ali-Osman F, Sawaya R, Fuller GN and Rao JS . (2000). Cancer Res., 60, 6851–6855.

  22. Konduri SD, Rao CN, Chandrasekar N, Tasiou A, Mohanam S, Kin Y, Lakka SS, Dinh D, Olivero WC, Gujrati M, Foster DC, Kisiel W and Rao JS . (2001). Oncogene, 20, 6938–6945.

  23. Lakka SS, Jasti SL, Kyritsis AP, Yung WK, Ali-Osman F, Nicolson GL and Rao JS . (2000). Clin. Exp. Metast., 18, 245–252.

  24. Lakka SS, Rajan M, Gondi C, Yanamandra N, Chandrasekar N, Jasti SL, Adachi Y, Siddique K, Gujrati M, Olivero W, Dinh DH, Kouraklis G, Kyritsis AP and Rao JS . (2002). Oncogene, 21, 8011–8019.

  25. Leunig M, Yuan F, Menger MD, Boucher Y, Goetz AE, Messmer K and Jain RK . (1992). Cancer Res., 52, 6553–6560.

  26. Lewis DL, Hagstrom JE, Loomis AG, Wolff JA and Herweijer H . (2002). Nat. Genet., 32, 107–108.

  27. Liotta LA, Rao CN and Wewer UM . (1986). Annu. Rev. Biochem., 55, 1037–1057.

  28. Mai J, Sameni M, Mikkelsen T and Sloane BF . (2002). Biol. Chem., 383, 1407–1413.

  29. McCaffrey AP, Meuse L, Pham TT, Conklin DS, Hannon GJ and Kay MA . (2002). Nature, 418, 38–39.

  30. Miyagishi M, Hayashi M and Taira K . (2003). Antisense Nucleic Acid Drug Dev., 13, 1–7.

  31. Mohanam S, Jasti SL, Kondraganti SR, Chandrasekar N, Lakka SS, Kin Y, Fuller GN, Yung AW, Kyritsis AP, Dinh DH, Olivero WC, Gujrati M, Ali-Osman F and Rao JS . (2001). Oncogene, 20, 3665–3673.

  32. Pozzi A, LeVine WF and Gardner HA . (2002). Oncogene, 21, 272–281.

  33. Premzl A, Zavasnik-Bergant V, Turk V and Kos J . (2003). Exp. Cell Res., 283, 206–214.

  34. Raithatha SA, Muzik H, Muzik H, Rewcastle NB, Johnston RN, Edwards DR and Forsyth PA . (2000). Neuro-oncology, 2, 145–150.

  35. Rao JS, Steck PA, Mohanam S, Stetler-Stevenson WG, Liotta LA and Sawaya R . (1993). Cancer Res., 53, 2208–2211.

  36. Redwood SM, Liu BC, Weiss RE, Hodge DE and Droller MJ . (1992). Cancer, 69, 1212–1219.

  37. Rooprai HK, Van Meter T, Rucklidge GJ, Hudson L, Everall IP and Pilkington GJ . (1998). Int. J. Oncol., 13, 1153–1157.

  38. Simon C, Goepfert H and Boyd D . (1998). Cancer Res., 58, 1135–1139.

  39. Sloane BF, Rozhin J, Johnson K, Taylor H, Crissman JD and Honn KV . (1986). Proc. Natl. Acad. Sci. USA, 83, 2483–2487.

  40. Stark GR, Kerr IM, Williams BR, Silverman RH and Schreiber RD . (1998). Annu. Rev. Biochem., 67, 227–264.

  41. Vu TH, Shipley JM, Bergers G, Berger JE, Helms JA, Hanahan D, Shapiro SD, Senior RM and Werb Z . (1998). Cell, 93, 411–422.

  42. Yu Q and Stamenkovic I . (1999). Genes Dev., 13, 35–48.

  43. Zamore PD . (2001). Nat. Struct. Biol., 8, 746–750.

Download references


We thank Karen Minter for preparing the manuscript, and Sushma Jasti and Diana Meister for manuscript review. This research was supported by the National Cancer Institute CA 85216 and CA 75557 and the National Institute of Neurological Disorders and Stroke Grant NS47699 (to JSR).

Author information

Correspondence to Jasti S Rao.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Lakka, S., Gondi, C., Yanamandra, N. et al. Inhibition of cathepsin B and MMP-9 gene expression in glioblastoma cell line via RNA interference reduces tumor cell invasion, tumor growth and angiogenesis. Oncogene 23, 4681–4689 (2004) doi:10.1038/sj.onc.1207616

Download citation


  • siRNA
  • proteases
  • glioma
  • invasion
  • angiogenesis

Further reading