Antisense-mediated Suppression of Human Heparanase Gene Expression Inhibits Pleural Dissemination of Human Cancer Cells¹

HSPG³ is a major component of the cell surface and ECM, and provide a physical barrier to movement of cells into tissues (1). The ability of HS to interact with ECM macromolecules, such as collagen, laminin, and fibronectin, and with different attachment sites on plasma membrane indicates that this proteoglycan is essential in the self-assembly and insolubility of ECM components as well as in cell adhesion and locomotion (2). Furthermore, HSPG binds various cytokines and growth factors such as fibroblast growth factor, hepatocyte growth factor, insulin-like growth factor, and transforming growth factor-β and functions as a store of these factors (3). Malignant tumor cells can modulate the expression of HSPG by releasing degradative enzymes, such as heparanase, or cytokines that influence the turnover and biosynthesis of host-cell HSPG (4).

Heparanase is an endoglucuronidase that cleaves HS at specific interchain sites, thus facilitating the release of several protein modulators of cell function, including migration, adhesion, and angiogenesis (5). Heparanase activity correlates with the metastatic potential of tumor cells in animal models (6) and, moreover, is increased in the sera and urine of human patients with metastatic cancer. Although the importance of heparanase function has long been recognized, the mammalian heparanase has remained an elusive target. However, several laboratories recently purified human heparanase and have isolated the cDNA encoding this activity. The human heparanase enzyme is expressed as a 543-amino acid precursor protein that is modified by proteolytic cleavage to remove the first 157 amino acids (7, 8, 9, 10). Treatment with heparanase inhibitors considerably reduced the incidence of lung metastasis in animal models (11, 12, 13).

These observations led us to hypothesize that abrogation of heparanase expression could reduce pleural dissemination by inhibiting the tumor cell invasion to the basement membrane. Here we demonstrate that an adenoviral vector expressing an antisense to a full-length human heparanase gene specifically inhibits heparanase expression as well as its invasive potential in vitro, leading to the reduction of pleural dissemination in an orthotopic murine model in vivo.

The human esophageal squamous cell carcinoma cell line T.Tn and the human non-small cell lung cancer cell line A549 were propagated in monolayer culture in RPMI 1640 with 25 mm HEPES medium and DMEM:Ham’s F-12 supplemented with 10% FCS, 100 units/ml penicillin and 100 mg/ml streptomycin. The transformed embryonic kidney cell line 293 was grown in DMEM with high glucose (4.5 g/l), supplemented with 10% FCS, 100 units/ml penicillin, and 100 mg/ml streptomycin. The 293 cells were used for the production of adenovirus vectors.

Replication-deficient, E1- and E3-deleted recombinant Ad5 was used as the viral backbone. A human heparanase cDNA of 1758 bp was inserted, as a EcoRI/XhoI fragment, in sense and antisense orientation into the cosmid vector pAxCAt (Adenovirus Expression Vector Kit; TaKaRa Biomedicals, Tokyo, Japan) to generate the pAxCAt-S/hep and pAxCAt-AS/hep constructs, respectively. These constructs were cotransfected with restriction enzyme-digested DNA-TPC into 293 cells to isolate recombinant adenovirus. The resultant adenoviruses were named Ad-S/hep and Ad-AS/hep (Fig. 1 A). The E1A-deleted adenovirus vector d1312 was used as the control vector. Recombinant virus was purified by ultracentrifugation in cesium chloride step gradients, and its titer was determined by pfu on the 293 cells.

Total RNA was isolated from cells using RNAzol (Cinna/BioTecx, Friendswood, TX) in a single-step phenol-extraction method and used as templates. Reverse transcription was performed at 22°C for 10 min and then at 42°C for 20 min, and PCR was performed with specific primers in volumes of 50 μl according to the protocol provided by the manufacturer (PCR kit; Perkin-Elmer/Cetus, Norwalk, CT). The following specific primers were used: for heparanase, sense (Hep-S: 5′-TTC GAT CCC AAG AAG GAA TCA AC-3′) and antisense (Hep-AS: 5′-GTA GTG ATG CCA TGT AAC TGA ATC-3′); for the sequence just after the inserted fragment in the vector (Ad5 pax: 5′-ATC GAT TCT AGA CTA GTT TAA TTA ATT T-3′); for MMP-2, sense (5′-TTT TCT CGA ATC CAT GAT GG-3′) and antisense (5′-CTG GTG CAG CTC TCA TAT TT-3′); and glyceraldehyde-3-phosphate dehydrogenase, sense (5′-CAG CCG AGC CAC ATC-3′) and antisense (5′-TGA GGC TGT TGT CAT ACT TCT-3′). The amplification reaction involved denaturation at 95°C for 45 s, annealing at 52°C for 1 min, and 72°C for 1 min using a thermal cycler (Perkin-Elmer, Foster City, CA). The PCR products were resolved on 1% agarose gels and visualized by ethidium bromide staining.

Cells were collected by trypsinization and washed twice in cold PBS. Cells then were dissolved in lysis buffer containing 50 mm Tris-HCl (pH 7.5), 150 mm NaCl, 0.5% Triton X-100, and protease inhibitors (0.2 mm phenylmethylsulfonyl fluoride, 0.2 mm 4-(2-aminoethyl) benzenesulfonylfluoride, 10 μg/ml leupeptin, 10 μg/ml pepstatin, and 1 μg/ml aprotinin). The lysis was carried out at 4°C for 30 min and centrifuged at 15,000 rpm. The protein concentration of the supernatant was determined using the Bio-Rad protein determination method (Bio-Rad, Hercules, CA). Equal amounts (60 μg) of proteins were electrophoresed under reducing conditions on 12% (w/v) polyacrylamide gels. Proteins were electrophoretically transferred to a Hybond- polyvinylidene difluoride transfer membranes (Amersham, Arlington Heights, IL) and incubated with primary antibodies against heparanase and then peroxidase-linked secondary antibody. An Amersham enhanced chemiluminescence chemiluminescent Western system (Amersham, Tokyo, Japan) was used to detect secondary probes.

Immunohistochemical staining was performed using a HISTOFINE SAB PO kit (Nichirei, Tokyo, Japan) according to the instructions provided by the manufacturer. Formalin-fixed, paraffin-embedded tissue sections were mounted on silanized slides, deparaffinized in xylene for 20 min, and rehydrated in graded ethanol solutions. Endogenous peroxidase was blocked by incubating the sections in 3.0% H₂O₂ in methanol for 15 min. Antigen retrieval on paraffin sections was performed by heating three times in 10 mm citrate buffer solution (pH 6.0) in a microwave. After the blocking of nonspecific reactivity with rabbit serum for 10 min at room temperature, sections were incubated overnight at 4°C with the antihuman heparanase mouse monoclonal antibody raised against recombinant human heparanase. After rinsing, the slides were incubated with biotinylated rabbit antimouse IgG and then with avidin-biotin-peroxidase complex. Peroxidase activity was determined using DAB/H₂O₂ solution (Histofine DAB substrate kit; Nichirei). The slides were counterstained with methyl green with Mayer’s hematoxylin. As a negative control, some sections were subjected to normal serum blocking and omission of the primary antibody.

Heparanase mRNA copy number in surgical specimens and cell lines were determined by real-time quantitative RT-PCR using a LightCycler instrument and a LightCycler DNA Master SYBR Green I kit (Roche Molecular Biochemicals, Indianapolis, IN). Amplifications were done in glass capillary tubes using 20-μl reaction containing 3 mm MgCl₂, 0.5 μm each primer, and 2 μl of 10× LightCycler FastStart DNA Master SYBR Green I. PCR amplification began with a 60-s denaturation step at 95°C and then 40 cycles of denaturation at 95°C for 15 s, annealing at 58°C for 10 s, and extension at 72°C for 9 s. The oligonucleotides used as specific primers were 5′-GCG GTT ACC CTA TCC TTT TT-3′ and 5′-GCA GCA ACT TTG GCA TTT C-3′. Copy numbers of mRNA were calculated from serially diluted standard curves (Fig. 3 B) generated from purified cDNA template, which consisted of a human heparanase cDNA of 1758 bp inserted into the expression vector. Data analysis was performed using LightCycler Software (Roche Molecular Biochemicals).

Invasion of tumor cells was assessed by counting the number of cells that migrated through transwell inserts with 8-μm pores (Becton Dickinson, Mountain View, CA), according to the protocol recommended by the manufacturer. Briefly, transwell membranes were coated with 100 μl of Matrigel (Collaborative Research Inc., Boston, MA) at a final concentration of 0.1 mg/ml and dried. Tumor cells (4 × 10⁴) in 200 μl of medium from each treatment were added in triplicate wells and allowed to invade through Matrigel for 12 h at 37°C in a 5% CO₂ atmosphere. The cells that migrated through the membranes into the lower wells were assessed by trypan blue uptake. Then the number of cells was counted within a field at ×200 under a light microscope. For each membrane, a total of five fields were selected at random, and the numbers were averaged.

The human non-small cell lung cancer cell line A549 was infected with Ad-S/hep at a MOI of 5, harvested 24 h after infection, and resuspended in HBSS at a density of 4.0 × 10⁶ cells/200 μl. Female BALB/c nu/nu mice were intrathoracically injected with 200 μl of cell suspension through a 27-gauge needle. After 24 h, either 3.0 × 10⁸ pfu/100 μl of dl312 or 3.0 × 10⁸ pfu/100 μl of Ad-AS/hep was injected into the thoracic space by the same technique. The procedure was repeated over 3 consecutive days. Thirty days after cell inoculation, mice killed, and their thoracic spaces were examined macroscopically for any growths. Tumors in the thoracic spaces were removed and weighed. The experimental protocol was approved by the Ethics Review Committee for Animal Experimentation of Okayama University School of Medicine, Okayama, Japan.

To examine the relevance of human heparanase in tumor cell invasion, we constructed replication-defective recombinant adenoviruses Ad-S/hep and Ad-AS/hep expressing, respectively, sense and antisense RNA to a full-length human heparanase sequence as described in “Materials and Methods” (Fig. 1,A). First we examined, by strand-specific RT-PCR assay, the expression of the sense and antisense constructs in human esophageal and lung cancer cell lines T.Tn and A549, respectively (Fig. 1 B). The antisense transcript was detected with specific primers (Hep-AS and Ad5 pax) in Ad-AS/hep-infected cells, whereas RT-PCR with another set of primers (Hep-S and Ad5 pax) detected heparanase mRNA expression in cells transduced with the Ad-S/hep vector. Both sense and antisense transcripts could be seen when Hep-AS and Hep-S primers were used. This analysis also confirmed the dose-dependent expression of the transcripts in T.Tn cells.

Western blot analysis was carried out to examine the effect of Ad-S/hep and Ad-AS/hep infection on heparanase protein levels in T.Tn and A549 cells. As shown in Fig. 2,A, proheparanase (M_r 65,000) as well as cleaved, active heparanase (M_r 50,000) protein expression was detected in both cell lines as early as 24 h after infection, but not in parental and dl312- or Ad-AS/hep-infected cells. Experiments using various concentrations of Ad-S/hep showed that infection at a MOI of 5 was sufficient to exhibit a detectable heparanase protein expression in T.Tn and A549 cells, although the protein level was much higher in A549 cells compared with that in T.Tn cells (Fig. 2,B). In addition, infection with dl312, Ad-S/hep, or Ad-AS/hep vectors at a MOI of 30 had no apparent effect on the in vitro growth of T.Tn and A549 cells (Fig. 2 C).

Heparanase has been reported to be expressed in many types of neoplastic tissues (6). We examined heparanase protein expression by immunohistochemistry in surgical specimens of human esophageal squamous cell carcinoma and lung adenocarcinoma. Sections of human tumor tissues were obtained from the Okayama University Medical School. As shown in Fig. 3 A, heparanase protein expression was present in the tumor cells in all of tumor specimens examined. Although some cells in the stromal regions adjacent to the tumor cells exhibited heparanase immunoreactivity, the staining was less intense than the tumor cells. These results suggest that heparanase expression in the tumor tissues is derived primarily from the tumor cells.

To compare expression levels of heparanase message between tumor tissues and cell lines, we next performed a quantitative real-time RT-PCR assay using the LightCycler instrument (Fig. 3,B). The assay demonstrated that heparanase mRNA expression was easily detectable in clinical samples of human esophageal and lung cancers, which is consistent with previous reports; human esophageal and lung cancer cell lines, however, expressed relatively low levels of heparanase mRNA. The relative expression of heparanase mRNA in primary tumors obtained from four patients as well as in A549 lung cancer cells was shown in Fig. 3 C. The median expression was ∼300-fold higher in tumor tissues than in the A549 cell line.

We next evaluated the efficacy of Ad-AS/hep in inhibiting heparanase expression. However, having verified the presence of low expression levels of heparanase in T.Tn and A549 cell lines, we selected cells infected with Ad-S/hep for these experiments. T.Tn cells were infected with Ad-S/hep at a MOI of 30 and then Ad-AS/hep infection at various MOIs. RT-PCR analysis with specific primers demonstrated expression of the antisense transcript as well as a diminution of heparanase mRNA levels in a dose-dependent manner (Fig. 4,A). For Western blot analysis, T.Tn and A549 cells were infected with Ad-S/hep at MOI of 10 and 5, respectively, and then Ad-AS/hep infection at various MOIs. The assay showed significantly decreased levels of heparanase protein in a dose-dependent fashion with an increase in MOI (Fig. 4 B).

To investigate the importance of heparanase expression in tumor cell invasion, the Matrigel invasion assay was performed using A549 cells in a modified Boyden chamber system. As shown in Fig. 5 A, A549 cells infected with Ad-S/hep at 5 MOI showed aggressive invasion characteristics compared with dl312-infected cells, which could be significantly inhibited upon simultaneous infection with 100 MOI of Ad-AS/hep vector. These results suggest that the antisense construct selectively inhibited the heparanase-induced invasive properties of A549 cells.

We also evaluated the therapeutic effect of Ad-AS/hep injection against Ad-S/hep-transduced A549 cells growing intrathoracically in a thymic nu/nu mice. Preliminary experiments revealed that, when 4 × 10⁶ A549 cells transduced with Ad-S/hep at 5 MOI were inoculated into the thoracic space, dissemination of tumors appeared in all mice 30 days after tumor injection. Injection of Ad-AS/hep vector (3.0 × 10⁸ pfu/100 μl) into the thoracic cavity over 3 consecutive days reduced the incidences of tumor cell dissemination and tumor nodules compared with mice that received intrathoracic dl312 injections (Fig. 5 B). In addition, Ad-AS/hep-treated animals showed a reduced weight of tumor nodules compared with dl312-treated animals.

MMP is a neutral proteinase that mediates ECM degradation and potentiates the invasiveness and metastatic potential of malignant tumors (14). To determine whether heparanase expression affects other invasion-related enzyme expression, we carried out semiquantitative RT-PCR analysis with specific primer pairs for type IV collagenases: MMP-2 and MMP-9. Densitometric quantification revealed that Ad-S/hep infection at a MOI of 5 caused a 2-fold increase of MMP-2 mRNA expression (Fig. 5 C), whereas there were no changes in MMP-9 mRNA levels (data not shown). When A549 cells were simultaneously infected with Ad-AS/hep at a MOI of 50, MMP-2 mRNA expression decreased by 1.3-fold compared with that in cells transduced with 5 MOI of Ad-S/hep alone.

Inhibition of tumor invasion is an attractive approach for the treatment of highly malignant tumors (15). Tumor cell invasion is characterized by the secretion of enzymes that facilitate tumor cell spread by degrading the ECM surrounding tumors and solubilizing the vascular basement (16). Thus, heparanase, which degrades the HSPG in the ECM, is an attractive target for the development of new antimetastatic drugs because of considerable evidence implicating the enzyme in tumor cell invasion. In the present study, we demonstrated that overexpression of human heparanase enhanced tumor cell invasiveness in vitro, and that this invasive ability was significantly reduced by inhibiting heparanase expression using an adenovirus-mediated antisense gene-delivery strategy. Furthermore, injection of an antisense adenovirus construct against human heparanase into the pleural space inhibited the growth of human non-small cell lung cancer cells implanted in the pleural cavity of nu/nu mice.

Many compounds, such as laminarin sulfate, phosphomannopentose sulfate (17), and maltohexaose sulfate (18), are potent inhibitors of heparanase activity; these molecules are, however, difficult to target to specific tissue sites, and their pleiotropic interactions with HSPG might produce undesired effects. Therefore, to selectively block heparanase expression, we used an antisense strategy. Expression of antisense RNA can directly influence the expression of a given gene of interest, although the precise mechanism of inhibition remains unclear (19). In addition, to ensure efficient transfer and high expression of antisense heparanase, we generated an adenovirus construct carrying human heparanase cDNA in an antisense orientation (Ad-AS/hep; Fig. 1). First, we examined whether Ad-AS/hep infection could suppress endogenous heparanase expression in human esophageal and lung cancer cell lines. However, unexpectedly, our quantitative real-time RT-PCR assay demonstrated that a specific heparanase transcript amplification was undetectable or extremely low in the cell lines tested in our study, although it was readily detected in surgical specimens (Fig. 3). We examined at least 10 clinical tissues of esophageal and lung cancer and detected heparanase expression in all of them by either RT-PCR assay or immunohistochemistry. Possible explanation for such a discrepancy is attributable to the differences of the microenvironment. Neoplastic cells may interact with surrounding cells in vivo to sustain their malignant potentials (e.g., metastatic melanoma cells and astrocytes; Ref. 20), which supports the concept that lack of heparanase expression in established cell lines results form the absence of microenvironmental stimuli.

Several groups have reported that tumor cells transfected with heparanase cDNA acquire a highly metastatic phenotype in vivo (21). Consistent with these findings, tumor cells adenovirally transduced with the human heparanase gene showed heparanase protein expression (Fig. 2, A and B) as well as heparanase activity (data not shown), which contributed to the highly invasive nature of the infected cells (Fig. 5,A). The finding that overexpression of heparanase had no apparent effect on the proliferation of tumor cells in vitro indicates its important roles for invasive properties in the microenvironment (Fig. 2,C). With the use of tumor cells expressing the ectopic heparanase gene, we showed successful transduction of an antisense construct with Ad-AS/hep, resulting in decreases of mRNA transcripts (Fig. 4,A) and protein contents (Fig. 4,B). In addition, we have noted complete inhibition of enhanced invasiveness in cells expressing heparanase by Ad-AS/hep infection (Fig. 5 A), suggesting that Ad-AS/hep has significant biological consequences.

Our data also indicate the therapeutic potential of molecular manipulation of heparanase expression in an orthotopic model of pleural dissemination. Intrathoracic administration of Ad-AS/hep vector reduced the incidence of tumor formation in the thoracic cavity (Fig. 5,B). Our preliminary experiments revealed that injection of adenovirus expressing the wild-type p53 gene, which has been known to induce apoptosis in certain tumor cells, had no significant antitumor effect in this animal model (data not shown). Thus, approaches that selectively block expression of molecules implicated in cellular invasion may be clinically more relevant for the prevention of tumor cell dissemination than directly inducing apoptosis of tumor cells. In fact, selective inhibition of other degradative enzymes such as MMP-9 (22) and urokinase-type plasminogen activator receptor (23) by antisense gene transfer has been reported to impair tumor invasion as well as metastasis. We also found that Ad-AS/hep infection suppressed heparanase-induced up-regulation of MMP-2 expression (Fig. 5 C), suggesting that Ad-AS/hep may have direct and indirect effects on the expression of heparanase and MMP-2, respectively, thereby being more clinically beneficial.

Although additional experiments using freshly isolated primary human cancer cells that consistently express endogenous heparanase are required, to the best of our knowledge, this is the first demonstration of antisense-mediated inhibition of heparanase expression as well as invasive properties by using an adenovirus construct. Our results suggest that heparanase is a promising target for the prevention of pleural dissemination of human cancer.

We thank Dr. Izumu Saito for the cosmid vectors and Dr. Jun-ichi Miyazaki for the CAG promoter.