Pigment epithelium-derived factor overexpression inhibits ... - Nature

Cancer Gene Therapy (2007) 14, 616–626 r

2007 Nature Publishing Group All rights reserved 0929-1903/07 $30.00

www.nature.com/cgt

ORIGINAL ARTICLE

Pigment epithelium-derived factor overexpression inhibits orthotopic osteosarcoma growth, angiogenesis and metastasis ETH Ek, CR Dass, KG Contreras and PFM Choong Department of Orthopaedics, University of Melbourne, St Vincent’s Hospital, Melbourne, Victoria, Australia Despite significant improvements, the current management of primary osteosarcoma is still limited by the development of metastatic disease, which occurs in approximately 30% of patients despite aggressive multiagent chemotherapy and tumor-ablative surgery. Therefore, there is a need for the development of novel agents to improve the outcome of these patients. Pigment epithelium-derived factor (PEDF) has been shown to be one of the most potent inhibitors of angiogenesis, and more recently has demonstrated a functional role in tumor growth, invasion and metastasis. In this study we report, for the first time, the multitargeted role of PEDF in the inhibition of growth, angiogenesis and metastasis of two orthotopic models of osteosarcoma (rat UMR 106-01 and human SaOS-2). Through stable plasmid-mediated gene transfer of full-length human PEDF, we show that PEDF overexpression significantly reduced tumor cell proliferation (Po0.05) and Matrigel invasion (UMRPEDF, Po0.001; SaOSPEDF, Po0.05) and increased adhesion to collagen type-1 (Po0.01), in vitro. In vivo, PEDF overexpression dramatically suppressed orthotopic osteosarcoma growth (Po0.05) and the development of spontaneous pulmonary metastases (UMRPEDF, Po0.05; SaOSPEDF, Po0.001). Furthermore, PEDF-overexpressing tumors exhibited reduced intratumoral angiogenesis, evidenced by a significant decrease in microvessel density (Po0.05). Therefore, together these results suggest that PEDF may be a new and promising approach for the treatment of osteosarcoma. Cancer Gene Therapy (2007) 14, 616–626; doi:10.1038/sj.cgt.7701044; published online 4 May 2007 Keywords: PEDF; osteosarcoma; angiogenesis; metastasis; antitumor

Introduction

Osteosarcoma is the most common primary bone tumor in childhood and adolescence and represents the second highest cause of cancer-related death in this age group.1 Up to three decades ago, first-line treatment of this tumor consisted of limb amputation, and despite this, the overall survival was 10%, with most patients succumbing within 1 year from presentation.2 It has been since the introduction of effective multiagent chemotherapeutic regimes and aggressive limb-sparing surgery, that the long-term survival for such patients has dramatically improved to approximately 60–70%.3 However, in spite of this improvement, 30–40% of patients with localized osteosarcoma of the extremities relapse, most commonly to the lungs.4 Therefore, it is the inability to eradicate the micrometastases that are invariably present at initial diagnosis that remains the greatest challenge in the management of this disease. Moreover, of the 20% of patients that present with primary metastatic osteoCorrespondence: Dr CR Dass, Department of Orthopaedics, St Vincent’s Hospital, PO Box 2900, Fitzroy 3065, Melbourne, Victoria 3065, Australia. E-mail: [email protected] Received 1 October 2006; revised 7 December 2006; accepted 11 February 2007; published online 4 May 2007

sarcoma, the prognosis, even with aggressive treatment, remains poor, with a 5-year survival of B30%.5,6 Hence, it is clear that novel therapies are indeed needed to improve the outcome of these patients. Pigment epithelium-derived factor (PEDF) is a 50 kDa secreted glycoprotein that was first identified and isolated from the conditioned media of primary human fetal retinal pigment epithelial cells.7 Although initially shown to be an effective neurotrophic factor,8 PEDF was later discovered to have potent anti-angiogenic activity, far greater than any other known endogenously produced factor.9 Furthermore, PEDF has been implicated in the pathogenesis of various conditions, including chronic inflammatory disease, atherosclerosis, diabetic complications and cancer.10 In cancer, PEDF has recently been reported to be a key prognostic marker in the development and progression of several tumors, including breast,11 lung,12 prostate13 and pancreatic carcinoma,14 as well as in gliomas15,16 and lymphangiomas.12 Decreased intratumoral expression of PEDF is associated with a higher microvessel density (MVD), a more metastatic phenotype and poorer clinical outcome. The mechanisms through which PEDF exerts its antitumor activity seems to be multiple, with studies showing widespread activities against tumor proliferation, cellular invasion, migration, differentiation and tumor angiogenesis.17 To date, a few studies have tested the

PEDF overexpression inhibits osteosarcoma growth and metastasis ETH Ek et al

therapeutic potential of PEDF against a number of tumor models, both in vitro and in vivo, and have demonstrated inhibition of growth and the suppression of metastasis in cancers, such as pancreatic carcinoma,18 prostate carcinoma,19 melanoma,20 neuroblastoma21 and ovarian cancer.22 In these studies, however, all have used models that have developed following ectopic implantation of cancer cell lines. Hence, no paper has reported the role and antitumor effect of PEDF in an orthotopic model of cancer (i.e., a tumor grown within its organ of origin). In this study, we investigated both the in vitro and in vivo growth characteristics of two osteosarcoma cell lines, rat UMR 106-01 and human SaOS-2, that were stably transfected to overexpress human PEDF. Here we demonstrate for the first time the ability of plasmidmediated gene transfer of PEDF to directly inhibit tumor growth and metastasis, and also suppress angiogenesis in two clinically relevant orthotopic models of osteosarcoma.

Materials and methods

Cells, culture conditions and mice Osteosarcoma cell lines, rat UMR 106-01 (from St Vincent’s Institute of Medical Research) and human SaOS-2 (American Tissue Culture Collection, Manassas, VA), were both cultured in a-MEM (Invitrogen, Melbourne, Australia) supplemented with 10% fetal calf serum (FCS, giving complete medium). Culture conditions comprised of 371C in a humidified 5% CO2 atmosphere. All animal experimentations were performed in accordance to the St Vincent’s Health Animal Ethics Committee guidelines. Five-week-old Balb/c nude mice were purchased from the Animal Resource Centre, Australia, and maintained under specific pathogen-free conditions. Male mice were used for orthotopic inoculation of UMR 106-01 cells, and female mice were used for SaOS-2 cells. Stable transfection of PEDF into osteosarcoma cells Human PEDF cDNA was isolated from muscle RNA and cloned into a pCR-II vector (Invitrogen) for transformation. A BamH1-Xba1 insert was then cloned into a pcDNA3.1-his-myc()A vector (Invitrogen) for transfection. Cells were transfected with Lipofectamine (Invitrogen) according to the manufacturer’s instructions. Osteosarcoma cells transfected with pcDNA3 empty vector plasmid DNA were used as controls. Transfected cells were selected using a G418 (neomycin analogue). The expression of PEDF mRNA in stably transfected clones was confirmed by reverse transcription-polymerase chain reaction (RT-PCR), whereas PEDF protein was examined by western blot analysis and immunohistochemistry. After confirming the efficiency of the transfection, three groups of cells were selected for further experiments, a parental cell line (UMRPAR and SaOSPAR), PEDF-overexpressing cell line (UMRPAR and SaOSPAR) and a vector-only cell line (UMRVECT and SaOSVECT).

Table 1 Oligonucleotides used as primers in RT–PCR reactions Primers

Nucleotide sequence

Rat PEDF

Forward: ATGAAGGCGACGTTACCAAC Reverse: TTTTCTGTAGCGCACTGG Forward: CATGGAGAAGGCTGGGGCTC Reverse: AACGGATACATTGGGGGTAG Forward: TGTCTCCAACTTCGGCTACGATC Reverse: CGAGTCAAACTTGGTTACCCACTG Forward: ACCACCATGGAGAAGGCTGG Reverse: CTCAGTGTAGCCCAGGATGC

Rat GAPDH Human PEDF Human GAPDH

Abbreviations: PEDF, pigment epithelium-derived factor; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; RT– PCR, reverse transcription–polymerase chain reaction.

RNA extraction and RT-PCR Cells were grown in 175 cm2 culture flasks (SigmaAldrich, Sydney, Australia) to B90% confluency and were then trypsinized and washed in 1 phosphatebuffered saline (PBS) and placed into Trizol reagent (Invitrogen) for total RNA extraction according to the manufacturer’s guidelines. The Expand Reverse Transcriptase-PCR and PCR Master (Roche Diagnostics, Australia) kits were utilized in accordance to the manufacter’s guidelines. A negative control (no cDNA template) was always included in each series of PCRs. Thrity-five cycles were used at the respective or relatively close theoretical Tann, Text ¼ 721C, Tden ¼ 941C and PCR products electrophoresed on a 1% agarose gel (for primers, refer to Table 1). Western blot analysis Briefly, cells were grown to 90–95% confluency, harvested and total cellular proteins extracted using modified radioimmuno precipitation assay buffer (150 nM NaCl, 50 mM Tris, pH 8.0, 1 mM ethylenediaminetetraacetic acid (EDTA), 0.1% sodium dodecyl sulfate (SDS) and 1% Triton X-100) with phenylmethylsulfonylfluoride (PMSF) (1 mM), leupeptin (2 mg/ml), aprotinin (2 mg/ml) and pepstatin (2 mg/ml). Approximately 15 mg of protein was electrophoresed through a 4–20% SDS–PAGE (SDS– polyacrylamide gel electrophoresis) gel (Invitrogen) and electrotransferred onto a polyvinylidene difluoride (PVDF; Invitrogen) membrane. Antibodies were visualized using the ECL-Plus chemiluminescence system (Amersham Biosciences, Sydney, Australia). After initial blotting with PEDF antibody (Upstate Biotechnologies, Charlottesville, VA, USA), the PVDF membrane was stripped with 325 mM Tris—HCl, pH 6.7/10% SDS at 701C and probed with GAPDH (glyceraldehyde-3-phosphate dehydrogenase) antibody (Santa Cruz Biotechnology, CA). Semiquantitative assessment of band intensities was made by Image Pro Plus (Media Cybernetics, MD) after normalization with GAPDH. Cell proliferation assay Cells were seeded at a density of 5 104 cells per well into six-well plates and cultured in a-MEM media containing 10% FBS at 371C/5% CO2 in quadruplicates. Cells were trypsinized and quantified using trypan blue exclusion

Cancer Gene Therapy

617


618

and a hemacytometer at day 3 post-seeding. Cells were used within 15 passages.

Invasion assay Tumor cell invasiveness was examined using chamber inserts with 8 mm filter membranes (BD Biosciences, Sydney, Australia) coated with Matrigel (1:4 dilution; BD Biosciences). Each insert was filled with 5 104 cells suspended in 200 ml of a-MEM (serum-free) and placed into wells filled with 800 ml of a-MEM þ 10% FCS, and incubated for 48 h (UMR 106-01) and 72 h (SaOS-2) at 371C in a 5% CO2 incubator. Cells that had invaded through the Matrigel-coated pores were quantitated after Quick-Dip (Froline, Sydney, Australia) staining, and images were acquired using SPOT Advanced software (SciTech, Melbourne, Australia). Collagen adhesion assay Twenty-four-well plates were coated with 0.2% collagen I (BD Biosciences) at 371C for 60 min. Cells (1 105) were seeded and incubated at 371C for 1 h. Wells were washed twice in 1 PBS to remove nonadherent cells, and the remaining cells were enumerated at 200 magnification under three random fields using a Nikon Eclipse TE2000U microscope (Nikon, Sydney, Australia) and photographed with SPOT Advanced software (SciTech). Orthotopic implantation of tumor cells and monitoring of growth A total of 2 104 cells (UMR 106-01) and 2 105 cells (SaOS-2) were injected into the proximal tibia of nude mice as described previously.23,24 Tumors were measured in the anteroposterior (AP) and lateral (L) planes) twice weekly using digital calipers and volume calculated using 4/3p (1/4(AP þ L))2. Legs were X-rayed at 35 kV for 30 s using a cabinet system (Faxitron Corp., Wheeling, IL, USA). At week 5, the mice were killed and the primary tumors and lungs were harvested and fixed in 4% paraformaldehyde. All tissues were embedded in paraffin for histological analysis. Lung metastases were analyzed using Q500MC Qwin software (Leica, Sydney, Australia) following hematoxylin and eosin (H&E) staining on 4-mm sections. Immunohistochemistry Tumor microvessels. Immunohistochemical staining of tumor microvessel was carried out by the streptavidinbiotin-peroxidase technique. Briefly, sections were deparaffinized with Solv21 (United Biosciences, Brisbane, QLD, Australia) and rehydrated through a series of decreasing ethanol concentrations. Antigen retrieval was performed by microwave heating in a high pH buffer consisting of 10 mM Tris and 1 mM EDTA (pH 9.0) for 12 min at low heat. Sections were then treated for 30 min with 10% hydrogen peroxide in dH2O to block endogenous peroxidase activity. Nonspecific binding was blocked using 10% normal rabbit serum for 30 min at room temperature. The primary antibody (CD34, monoclonal mouse-antihuman; DakoCytomation, Glostrup, Denmark) was diluted to 1:25 and added to sections and incubated overnight at 41C. Sections were then

Cancer Gene Therapy

washed twice in 1 PBS and the secondary antibody (biotinylated rabbit-anti-mouse secondary antibody, 1:300 dilution; DakoCytomation, Denmark) was added for 30 min at room temperature. Following this, the slides were washed in 1 PBS, treated with peroxidaseconjugated streptavidin for 20 min and developed in 3,3’-diaminobenzidine tetrahydrochloride for 10 min. Light counterstaining with Harris hematoxylin was performed before mounting on coverslips.

PEDF expression. PEDF overexpression was assessed by fluorescent immunohistochemistry. Paraffin sections were prepared as described above with the same antigenretrieval technique. Nonspecific binding was blocked using 10% goat serum for 30 min at room temperature. The primary antibody (PEDF, polyclonal rabbit-antihuman; Santa Cruz Biotechnology, CA) was diluted to 1:25 and added to sections and incubated overnight at 41C. Sections were washed twice in 1 PBS for 10 min each and the secondary antibody (Alexa Fluor 488-conjugated goat anti-rabbit IgG antibody, 1:2000 dilution; Invitrogen, Australia) was applied for 30 min at room temperature in the dark. For nuclear staining, sections were rinsed twice for 5 min each in 1 PBS and then incubated for 5 min with 4,6-diamidino-2-phenyl-indole dihydrochloride. Subsequently, the slides were rinsed in 1 PBS and then mounted with DakoCytomation Fluorescent mounting medium, and clear nail polish was applied to the edges of the cover slips. The slides were then stored in the dark at 41C. To confirm specificity for the primary antibody, negative controls were used in which the primary antibody was substituted with 1 PBS. Assessment of microvessel density The degree of angiogenesis was determined by calculating the MVD. Immunohistochemical assessment was performed using an inverted fluorescent microscope (Nikon, Eclipse TE2000-U). The MVD was assessed according to a modified version of the International Consensus Report,25 and this was conducted by two blinded independent investigators (EE & KC). Briefly, the entire section was systematically scanned at low magnification and the areas of most intense vascularization, that is, greatest number of CD34-antigen-positive cells was classified as a ‘hot spot’. A hot spot was defined as an area where the density of antigen-positive cells and cell clusters was greater relative to adjacent areas. The selected areas were viewed at 400 magnification and the slide was repositioned over the area of the hot spot with the most vessels per field. A positive vessel count was defined as a single endothelial cell, endothelial cell cluster or microvessel that is clearly separated from adjacent microvessels. The mean value of the number of microvessels was calculated from three vascular hotspots within a representative section of tumor. Statistical analysis Results were analyzed using MedCalc for Windows v8.1. (Medcalc Software, Mariakerke, Belgium). In vitro and


in vivo data were analyzed for statistical significance using the Mann–Whitney U test (two tailed). A P-value p0.05 was considered significant.

Results

Expression of PEDF in osteosarcoma cells after plasmidmediated gene transfer After transfection of two clonal osteosarcoma cell lines, rat UMR 106-01 and human SaOS-2, with the human PEDF plasmid and subsequent neomycin (G418) anti-

a

Parental

biotic selection, individual clones were grown in 175 cm2 flasks to confluency. The expression of PEDF mRNA and protein in stably transfected osteosarcoma cells (UMRPEDF and SaOSPEDF) was confirmed by RT-PCR, western blot analysis and immunohistochemistry, and compared to parental (UMRPAR and SaOSPAR) and vector-only (UMRVECT and SaOSVECT) controls (Figure 1). Evaluation of endogenous levels of PEDF expression revealed that in UMR 106-01 cells, PEDF was easily detectable on western blot analysis (Figure 1b) and immunohistochemistry (Figure 1a). In contrast, SaOS-2 cells demonstrated only minimal levels of PEDF protein (Figure 1b) and had

b

PEDF

Parental

PEDF

Vector

PEDF

GAPDH

UMR 106-01

UMR 106-01

Parental

PEDF

Vector

PEDF

SaOS-2 GAPDH

U

600bp 600bp

400bp

Sa Ve OS ct -2 or

01 610 tor R M Vec

Sa PE OS D -2 F

U

01 610 F R D M PE

S Pa aO re S-2 nt al

c

U M Pa R 1 re 06 nt -0 al 1

SaOS-2

Human PEDF

400bp Human PEDF 200bp

300bp

Figure 1 Expression of PEDF in UMR 106-01 and SaOS-2 cell lines. (a) Immunofluorescence staining for PEDF (green) demonstrating endogenous levels of PEDF within parental cells and the level of PEDF after plasmid-mediated transfection. Cell nuclei were counterstained with DAPI (blue) (magnification 400). (b) Western blot analysis showing expression levels of PEDF protein within the cell lysates of parental, PEDFtransfected and vector-only cells. Levels were normalized against GAPDH. (c) RT-PCR demonstrating the detection of human PEDF mRNA within each subclone.

Cancer Gene Therapy

619


620

undetectable levels of PEDF mRNA on RT-PCR (Figure 1c). Upon transfection, overexpression of PEDF was confirmed in the UMRPEDF cells, with an approximately fourfold increase seen on Western blotting and increased levels of human PEDF mRNA expression on RT-PCR. Similarly in SaOSPEDF cells, there was an approximately fivefold increase in PEDF protein levels and an obvious increase in PEDF mRNA. Both UMRVECT and SaOSVECT demonstrated levels similar to that of endogenous PEDF.

PEDF overexpression inhibits proliferation in vitro Having determined that PEDF was overexpressed in both UMR 106-01 and SaOS-2 cells, we then investigated whether this influenced tumor cell proliferation in vitro (Figure 2a). After enumerating cells grown in culture over 3 days, we observed that, in both cell lines, the growth rate of PEDF-transfected cells was significantly reduced compared to controls (Po0.05). As shown in Figure 2a, PEDF overexpression resulted in an approximately 45 and 50% reduction in cell numbers in UMRPEDF and SaOSPEDF, respectively, compared to vector controls (UMR, P ¼ 0.02; SaOS, P ¼ 0.03). Effect of PEDF overexpression on invasion and adhesion in vitro Previous studies have reported that decreased intratumoral levels of PEDF is associated with an increased incidence of metastasis and a poorer clinical outcome.10 Given that metastasis to the lung remains the major cause of death from osteosarcoma, we next investigated the role of PEDF in the inhibition of osteosarcoma metastasis by studying its effect on tumor cell invasion in a Matrigel assay. We found that both UMRPEDF and SaOSPEDF cells showed reduced ability to invade through Matrigel by 96 and 40%, respectively (UMRPEDF, Po0.05; SaOSPEDF, Po0.001; Figure 2b). Furthermore, PEDF has also been shown to have high affinity to collagen type-1, the most abundant extracellular matrix protein present within bone.26 Hence, we studied the effect of PEDF overexpression on cellular adhesion to wells coated with collagen type-1. We observed that both UMRPEDF and SaOSPEDF demonstrated significantly increased adhesion compared to parental and vector controls (Po0.01, Figure 3). Interestingly, closer inspection of the adherent cells revealed that, in the presence of PEDF overexpression, the morphology of the osteosarcoma cells changed to appear flatter and there was a tendency to form clusters with neighboring cells (data not shown). Together these results suggest that PEDF may play role in regulating tumor cell invasion and adhesion to the surrounding extracellular matrix, which are important aspects in determining a malignant cell’s metastatic potential. PEDF overexpression suppresses primary tumor growth and prevents spontaneous metastasis Having demonstrated the antitumor effects of PEDF overexpression in vitro, we then examined whether PEDF overexpression influences in vivo primary tumor growth and the development of pulmonary metastasis in two

Cancer Gene Therapy

orthotopic models of osteosarcoma. As shown in Figure 4, UMRPEDF tumors demonstrated slower growth kinetics compared to the parental and vector groups, with a twoto threefold reduction in tumor volumes seen by day 28 (Po0.05; Figure 4a). Similarly, SaOSPEDF tumors were 59 and 51% smaller at days 28 and 32, respectively (Po0.05; Figure 4b). While all mice in the control groups developed lung metastases, only 20% (1/5) of mice with UMRPEDF tumors (Po0.05) and no mice with SaOSPEDF tumors developed pulmonary metastases (Po0.001; Figure 4a and b). In addition, mice that were orthotopically inoculated with tumors that overexpressed PEDF did not show any difference in body weight (Figure 4c), and there was no overt evidence of lethargy or soft stools.

Inhibition of tumor angiogenesis in vivo by PEDF The ability of PEDF overexpression to inhibit tumor angiogenesis in vivo was evaluated by immunohistochemical staining of the tumor microvascular endothelial cells with an antibody against CD34 and determining the MVD. As demonstrated in Figure 5, we observed a significant reduction in MVD in both PEDF-overexpressing tumor groups with an 85 and 74% decrease seen in the UMRPEDF and SaOSPEDF groups, respectively (Po0.05).

Discussion

As osteosarcoma is invariably a systemic disease at the time of initial diagnosis, the goal of current treatment regimes is to not only destroy primary tumor cells, with reduction in tumor burden, but primarily to eradicate micrometastases. However, despite the effectiveness of modern chemotherapeutics, a significant proportion of patients will still suffer from metastatic relapse, hence the need for discovery and development of novel agents to combat this. In this study, we have demonstrated for the first time that plasmid-mediated PEDF overexpression significantly inhibits in vivo tumor growth and angiogenesis and dramatically suppresses the development of pulmonary metastases in two clinically relevant orthotopic models of rat and human osteosarcoma. These results are supported by the in vitro findings that PEDF overexpression decreases osteosarcoma cell proliferation and invasion and also increases adhesion to collagen type-1. The overall role of PEDF in the human body is still relatively unclear. However, it is a protein that is widely expressed throughout fetal and adult tissues including the brain, spinal cord, eye, liver, heart, lung and bone.10 Given this wide distribution, PEDF’s actions have shown to be pleiotropic and depending on the organ of expression, PEDF exhibits neuroprotective, neurotrophic, anti-angiogenic and anti-tumorigenic properties. Physiologically, we have previously demonstrated that PEDF is developmentally expressed within both cartilage and osteoblastic cells during the process of endochondral bone formation and have suggested that it plays a key regulatory role in osteoblastic differentiation, endochondral


621

a

* *

UMR 106-01

SaOS-2

b

*

**

UMR 106-01

SaOS-2

UMR 106-01

SaOS-2

Parental

PEDF

Vector

Figure 2 PEDF overexpression inhibits osteosarcoma cell proliferation and invasion in vitro. (a) Cell proliferation of UMR 106-01 and SaOS-2 parental, PEDF-transfected and vector-only cells measured 3 days post-seeding. (b) Effect of PEDF overexpression on cellular invasion through Matrigel-coated 8 mm pores. Representative sections are shown (below). Data shown are mean7s.d. of triplicates or quadruplicates. *Po0.05 vs vector control, **Po0.001 vs vector control (Mann–Whitney U-test, two-tailed).

ossification and bone remodeling.27 Furthermore, we have shown that the expression of PEDF within the avascular epiphyseal growth plate cartilage may contribute to its ability to inhibit the spatial progression of osteosarcoma.28 Hence, based on these findings, we

sought to further study the role of PEDF in osteosarcoma growth and metastasis, and determine its potential as a therapeutic anticancer target. Although previous investigators have shown inhibition of in vivo tumor growth with either PEDF gene transfer or

Cancer Gene Therapy


622

UMR 106-01 Parental

SaOS-2 PEDF

Vector

UMR 106-01

SaOS-2

Figure 3 PEDF overexpression increases adhesion to collagen type-1. Cellular adhesion to collagen type-1 lined plates as a result of PEDF overexpression in UMR 106-01 (right) and SaOS-2 (left) cells. Representative images demonstrating increased cell adhesion in cells transfected with PEDF compared to vector and parental controls (bottom). Representative photomicrographs are shown at 200 magnification. Data shown are mean7s.d. of quadruplicates. *Po0.01 vs vector control (Mann–Whitney U-test, two-tailed).

recombinant PEDF treatment,17 no studies have tested the role of PEDF within an orthotopic model of cancer. The importance of this lies firstly in the clinical relevance, but also in the fact that the tumor microenvironment plays an important paracrine role in regulating tumor growth and spread, hence the limitation of subcutaneous xenograft models. In both the syngeneic UMR 106-01 and human SaOS-2 models, primary tibial tumors become visible by approximately 2–3 weeks postinoculation, and spontaneous pulmonary metastases develop by week 5.23,24 Therefore, in our study, by using these models, we have shown, as proof of principle, that through stable PEDF overexpression, primary tumor growth is retarded by approximately 50–70%, and more impressively, there is almost complete inhibition of spontaneous pulmonary metastasis. Clinical studies have demonstrated that decreased expression of PEDF within several tumors is associated with an increased MVD, greater metastatic potential and a poorer clinical outcome.10 Although the mechanisms through which PEDF exerts its activity still remain to be

Cancer Gene Therapy

clearly elucidated, it is apparent that its antitumor activity is multifaceted, having the ability to target tumor progression at various levels of the metastatic cascade, through direct effects on cell proliferation, apoptosis, invasion and differentiation, and indirectly through potent anti-angiogenesis.10 Through the use of gene transfer, we have established, in vitro, that PEDF is effective in significantly decreasing osteosarcoma cell proliferation and also exerting anti-metastatic properties evidenced by inhibited invasion and increased collagen adhesion. The ability of PEDF to inhibit directly tumor cell proliferation has been shown to occur at two levels, through induction of cell apoptosis via the Fas/Fas ligand (FasL) death pathway,29 and also through regulation of the cell cycle leading to a decrease in the entry of cells into S phase.30 In addition, PEDF has also demonstrated the capacity to promote cell differentiation in neuroblastoma21 and prostate carcinoma cells,31 although the mechanism by which this occurs is unknown. In bone, we have previously shown that PEDF is temporally expressed


623

b

a

250

SaOS-2 (Parental) UMR 106-01 (Parental)

250

UMR 106-01 (Parental)

200

UMR 106-01 (PEDF)

200

SaOS-2 (Parental) SaOS-2 (PEDF) SaOS-2 (Vector)

UMR 106-01 (Vector)

150

150

100 100

17 21 Days

25

28

SaOS-2 (PEDF)

50 0

11

6

4 SaOS-2 (Vector)

2

*

0

14

28

32

4 3 2 1

**

0

UMR 106-01 (Parental) UMR 106-01 (PEDF) UMR 106-01 (Vector) SaOS-2 (Parental) SaOS-2 (PEDF) SaOS-2 (Vector)

22 Mice weight (g)

21 25 Days

5

(P O ar S en -2 ta l)

24

17

*

6

Sa

c

*

*

S (V aO ec Sto 2 r)

14

*

No. microscopic metastastases

No. mice with metastastases

11

*

U M (P R ar 10 en 6ta 01 l) U M R (P 1 0 ED 6 F) -01 U M R (V 1 ec 06 to -0 r) 1

UMR 106-01 (Vector)

0

*

S (P aOS ED -2 F)

50 UMR 106-01 (PEDF)

20 18 16 14 12 10

7

11

14 18 Days

21

25

Figure 4 PEDF overexpression inhibits osteosarcoma growth and metastasis in two orthotopic models. (a) Effect of PEDF overexpression on UMR 106-01 cells grown in vivo. Representative primary tumors of UMR 106-01 parental, PEDF-transfected and vector groups arising from the proximal tibia of nude Balb/c mice 28 days post-implantation (left). Radiographs were taken to examine extent of bony destruction (inset) and the development of pulmonary metastases was assessed by H&E staining (arrows). Tumor volumes (right top) and the number of mice containing lung metastases (right bottom) are shown. (b) Effect of PEDF overexpression on SaOS-2 cells grown in vivo. Representative primary tumors of SaOS-2 parental, PEDF-transfected and vector groups are shown at 5 weeks (left). Radiographs demonstrate bony tumor involvement within the proximal tibia (inserts). Tumor volume (middle) and the development of macroscopic metastases (arrows; right) are displayed. (c) Body weights of mice bearing orthotopic tumors. Data shown are mean7s.e.m. of at least five mice in each group. *Po0.05 vs vector control, **Po0.001 vs vector controls (Mann–Whitney U-test, two-tailed).

during endochondral bone formation, which suggests that PEDF has a key regulatory role in osteoblastic differentiation.27 In osteosarcoma, this may contribute to the inhibition in cell proliferation seen, as the degree of differentiation is often inversely related to the rate of proliferation, with more differentiated and mature cells generally exhibiting slower growth kinetics.32 In recent times, the role of angiogenesis in the pathogenesis of malignant tumors has been the subject of intense research. It is widely believed that in order for tumors to grow, invade and eventually metastasize, they must be able to continuously stimulate the ingrowth of new capillary blood vessels from the surroundings.33 This is the result of an imbalance between pro-angiogenic factors (e.g., vascular endothelial growth factor (VEGF), basic fibroblast growth factor and transforming growth factor-beta) and anti-angiogenic factors (e.g., thrombospondin-1 (TSP-1), angiostatin and endostatin). Folkman

postulated that as a result of the induction of the ‘angiogenic switch’ during tumor development, there is a threshold change in the balance between stimulatory and inhibitory influences in favor of angiogenesis.34 Therefore, interruption of this process could theoretically halt the progression of many tumors that are dependent on angiogenesis for further growth. As previously mentioned, PEDF has shown to be the most potent of the known endogenous inhibitors of angiogenesis, being twice as potent as angiostatin and seven times as potent as endostatin.9 It has shown to act by inducing apoptosis in endothelial cells through the Fas/FasL pathway,29 and also by downregulating the expression of potent proangiogenic factors, such as VEGF.35 As a result, in PEDF-deficient mice, several organs, including the pancreas, kidney and prostate, exhibit increased stromal vascularity and MVD. Furthermore, clinical studies in several tumors, including osteosarcoma, have shown that

Cancer Gene Therapy


624

a a

b

c

UMRPAR d

UMRPEDF e

f

SaOSPAR

SaOSPEDF

45

50

40

45

35

40 Microvessels per HPF

Microvessels per HPF

b

30 25 20 15

**

10

UMRVECT

SaOSVECT

35 30 25 20

**

15 10

5

5 0

0 UMR (PAR)

UMR (PEDF)

UMR (VECT)

SaOS (PAR)

SaOS (PEDF)

SaOS (VECT)

Figure 5 Inhibition of tumor angiogenesis by PEDF overexpression. Intratumoral MVD was determined by anti-CD34 staining. (a) Representative sections showing tumor microvessels (brown) in parental, PEDF-transfected and vector-only cell lines. (b) Quantitative assessment of the MVD was made by counting the number of positively stained cells or vessels within a high power field (0.26 mm2; magnification 400). Data shown are mean (7s.d.) microvessels per 400 magnification of three different hot spots. *Po0.05 compared to vector controls (Mann–Whitney U-test, two-tailed).

decreased expression of PEDF correlates with a higher MVD and a more-metastatic phenotype.17,36 In this present study, the anti-angiogenic activity of PEDF in osteosarcoma is demonstrated by its ability to inhibit in vivo angiogenesis, as reflected by the decreased MVD seen in tumors overexpressing PEDF. As osteosarcoma has shown to be a highly vascularized tumor,36 this inhibition is likely to contribute to the reduction in primary tumor volume as well as preventing the development of metastatic disease. In osteosarcoma, as with all malignant tumors, the establishment of metastatic disease remains the major challenge in management, as local recurrence rates after primary resections are relatively low.3 Although it is possible that PEDF overexpression could inhibit metastases through anti-angiogenesis, inhibition of tumor cell proliferation and increased apoptosis, and pro-differentiation, we have also demonstrated the profound effects of

Cancer Gene Therapy

PEDF on in vitro tumor cell invasion and adhesion and more impressively on the formation of metastases in vivo. The complete inhibition of metastases with PEDF overexpression strongly indicates that it plays a key antimetastatic role, possibly through regulation of the microenvironment or direct changes in tumor cell phenotype. In a study by Guan et al.,37 decreased cellular invasion of malignant U251 cells was also seen in cells overexpressing PEDF. They showed that this was associated with a significant downregulation in matrix metalloproteinase-9 (MMP-9) activity, which plays an important role in extracellular matrix degradation. In osteosarcoma, overexpression of MMP-9 has been shown to strongly correlate with the acquisition of metastatic potential and the development of pulmonary metastases.38 Therefore, together with our findings that PEDF increases osteosarcoma cell adhesion to collagen type-1, which is abundant in bone, this indicates that PEDF may


be central to regulating tumor cell invasion and subsequently metastasis. In conclusion, our study provides strong evidence for the multifunctional role of PEDF in the inhibition of osteosarcoma growth, angiogenesis and metastasis. However, before the commencement of preclinical trials, further work is indeed needed to further understand the various mechanisms through which PEDF exerts its activity in osteosarcoma. In this study, through the overexpression of PEDF via plasmid-mediated gene transfer, we have demonstrated, both in vitro and in vivo, that PEDF may be a new and promising approach for the treatment of osteosarcoma. Acknowledgements

We thank Dr Joseline Ojaimi for construction of the human PEDF plasmid used throughout this study. Dr Eugene TH Ek is supported by scholarships awarded by the University of Melbourne, the Royal Australasian College of Surgeons and the National Health and Medical Research Council of Australia (NHMRC). This study was generously supported by grants from the Australian Orthopaedic Association, the Victorian Orthopaedic Research Trust Grant and the Cancer Council of Victoria. Competing interests statement The authors declare that they have no competing financial interests.

References 1 Campanacci M. Bone Tumors. 2nd edn. Lippincott-Verlag: Austria, 1999. 2 Huvos A. Bone tumors: Diagnosis, Treatment and Prognosis. 2nd edn. WB Saimders: Philadelphia, PA, 1991. 3 Ferrari S, Smeland S, Mercuri M, Bertoni F, Longhi A, Ruggieri P et al. Neoadjuvant chemotherapy with high-dose Ifosfamide, high-dose methotrexate, cisplatin, and doxorubicin for patients with localized osteosarcoma of the extremity: a joint study by the Italian and Scandinavian Sarcoma Groups. J Clin Oncol 2005; 23: 8845–8852. 4 Longhi A, Errani C, De Paolis M, Mercuri M, Bacci G. Primary bone osteosarcoma in the pediatric age: state of the art. Cancer Treat Rev 2006; 32: 423–436. 5 Voute PA, Souhami RL, Nooij M, Somers R, Cortes-Funes H, van der Eijken JW et al. A phase II study of cisplatin, ifosfamide and doxorubicin in operable primary, axial skeletal and metastatic osteosarcoma. European osteosarcoma intergroup (EOI). Ann Oncol 1999; 10: 1211–1218. 6 Harris MB, Gieser P, Goorin AM, Ayala A, Shochat SJ, Ferguson WS et al. Treatment of metastatic osteosarcoma at diagnosis: a pediatric oncology group study. J Clin Oncol 1998; 16: 3641–3648. 7 Tombran-Tink J, Johnson LV. Neuronal differentiation of retinoblastoma cells induced by medium conditioned by human RPE cells. Invest Ophthalmol Vis Sci 1989; 30: 1700–1707. 8 Tombran-Tink J, Chader GG, Johnson LV. PEDF: a pigment epithelium-derived factor with potent neuronal differentiative activity. Exp Eye Res 1991; 53: 411–414.

9 Dawson DW, Volpert OV, Gillis P, Crawford SE, Xu H, Benedict W et al. Pigment epithelium-derived factor: a potent inhibitor of angiogenesis. Science 1999; 285: 245–248. 10 Ek ET, Dass CR, Choong PF. PEDF: a potential molecular therapeutic target with multiple anti-cancer activities. Trends Mol Med 2006; 12: 497–502. 11 Cai J, Parr C, Watkins G, Jiang WG, Boulton M. Decreased pigment epithelium-derived factor expression in human breast cancer progression. Clin Cancer Res 2006; 12: 3510–3517. 12 Zhang L, Chen J, Ke Y, Mansel RE, Jiang WG. Expression of pigment epithelial derived factor is reduced in non-small cell lung cancer and is linked to clinical outcome. Int J Mol Med 2006; 17: 937–944. 13 Halin S, Wikstrom P, Rudolfsson SH, Stattin P, Doll JA, Crawford SE et al. Decreased pigment epithelium-derived factor is associated with metastatic phenotype in human and rat prostate tumors. Cancer Res 2004; 64: 5664–5671. 14 Uehara H, Miyamoto M, Kato K, Ebihara Y, Kaneko H, Hashimoto H et al. Expression of pigment epitheliumderived factor decreases liver metastasis and correlates with favorable prognosis for patients with ductal pancreatic adenocarcinoma. Cancer Res 2004; 64: 3533–3537. 15 Guan M, Yam HF, Su B, Chan KP, Pang CP, Liu WW et al. Loss of pigment epithelium derived factor expression in glioma progression. J Clin Pathol 2003; 56: 277–282. 16 Sidle DM, Maddalozzo J, Meier JD, Cornwell M, Stellmach V, Crawford SE. Altered pigment epithelium-derived factor and vascular endothelial growth factor levels in lymphangioma pathogenesis and clinical recurrence. Arch Otolaryngol Head Neck Surg 2005; 131: 990–995. 17 Ek ET, Dass CR, Choong PF. Pigment epithelium-derived factor: a multimodal tumor inhibitor. Mol Cancer Ther 2006; 5: 1641–1646. 18 Hase R, Miyamoto M, Uehara H, Kadoya M, Ebihara Y, Murakami Y et al. Pigment epithelium-derived factor gene therapy inhibits human pancreatic cancer in mice. Clin Cancer Res 2005; 11: 8737–8744. 19 Doll JA, Stellmach VM, Bouck NP, Bergh AR, Lee C, Abramson LP et al. Pigment epithelium-derived factor regulates the vasculature and mass of the prostate and pancreas. Nat Med 2003; 9: 774–780. 20 Abe R, Shimizu T, Yamagishi S, Shibaki A, Amano S, Inagaki Y et al. Overexpression of pigment epitheliumderived factor decreases angiogenesis and inhibits the growth of human malignant melanoma cells in vivo. Am J Pathol 2004; 164: 1225–1232. 21 Crawford SE, Stellmach V, Ranalli M, Huang X, Huang L, Volpert O et al. Pigment epithelium-derived factor (PEDF) in neuroblastoma: a multifunctional mediator of Schwann cell antitumor activity. J Cell Sci 2001; 114: 4421–4428. 22 Cheung LW, Au SC, Cheung AN, Ngan HY, Tombran-Tink J, Auersperg N et al. Pigment epithelium-derived factor is estrogen sensitive and inhibits the growth of human ovarian cancer and ovarian surface epithelial cells. Endocrinology 2006; 147: 4179–4191. 23 Fisher JL, Mackie PS, Howard ML, Zhou H, Choong PF. The expression of the urokinase plasminogen activator system in metastatic murine osteosarcoma: an in vivo mouse model. Clin Cancer Res 2001; 7: 1654–1660. 24 Dass CR, Ek ET, Contreras KG, Choong PF. A novel orthotopic murine model provides insights into cellular and molecular characteristics contributing to human osteosarcoma. Clin Exp Metastasis 2006; 23: 367–380. 25 Vermeulen PB, Gasparini G, Fox SB, Toi M, Martin L, McCulloch P et al. Quantification of angiogenesis in solid

Cancer Gene Therapy

625


626

26

27

28

29

30

31

human tumours: an international consensus on the methodology and criteria of evaluation. Eur J Cancer 1996; 32A: 2474–2484. Meyer C, Notari L, Becerra SP. Mapping the type I collagenbinding site on pigment epithelium-derived factor. Implications for its antiangiogenic activity. J Biol Chem 2002; 277: 45400–45407. Quan GM, Ojaimi J, Li Y, Kartsogiannis V, Zhou H, Choong PF. Localization of pigment epithelium-derived factor in growing mouse bone. Calcif Tissue Int 2005; 76: 146–153. Quan GM, Ojaimi J, Nadesapillai AP, Zhou H, Choong PF. Resistance of epiphyseal cartilage to invasion by osteosarcoma is likely to be due to expression of antiangiogenic factors. Pathobiology 2002; 70: 361–367. Volpert OV, Zaichuk T, Zhou W, Reiher F, Ferguson TA, Stuart PM et al. Inducer-stimulated Fas targets activated endothelium for destruction by anti-angiogenic thrombospondin-1 and pigment epithelium-derived factor. Nat Med 2002; 8: 349–357. Pignolo RJ, Rotenberg MO, Cristofalo VJ. Analysis of EPC-1 growth state-dependent expression, specificity, and conservation of related sequences. J Cell Physiol 1995; 162: 110–118. Filleur S, Volz K, Nelius T, Mirochnik Y, Huang H, Zaichuk TA et al. Two functional epitopes of pigment epithelial-

Cancer Gene Therapy

32 33 34 35

36

37

38

derived factor block angiogenesis and induce differentiation in prostate cancer. Cancer Res 2005; 65: 5144–5152. Compagni A, Christofori G. Recent advances in research on multistage tumorigenesis. Br J Cancer 2000; 83: 1–5. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000; 100: 57–70. Folkman J, Shing Y. Angiogenesis. J Biol Chem 1992; 267: 10931–10934. Cai J, Jiang WG, Grant MB, Boulton M. Pigment epithelium-derived factor inhibits angiogenesis via regulated intracellular proteolysis of vascular endothelial growth factor receptor 1. J Biol Chem 2006; 281: 3604–3613. Ek ET, Ojaimi J, Kitagawa Y, Choong PF. Does the degree of intratumoural microvessel density and VEGF expression have prognostic significance in osteosarcoma? Oncol Rep 2006; 16: 17–23. Guan M, Pang CP, Yam HF, Cheung KF, Liu WW, Lu Y. Inhibition of glioma invasion by overexpression of pigment epithelium-derived factor. Cancer Gene Ther 2004; 11: 325–332. Kido A, Tsutsumi M, Iki K, Takahama M, Tsujiuchi T, Morishita T et al. Overexpression of matrix metalloproteinase (MMP)-9 correlates with metastatic potency of spontaneous and 4-hydroxyaminoquinoline 1-oxide (4-HAQO)induced transplantable osteosarcomas in rats. Cancer Lett 1999; 137: 209–216.