Role of Wnt5a in the proliferation of human glioblastoma cells

Available online at www.sciencedirect.com

Cancer Letters 257 (2007) 172–181 www.elsevier.com/locate/canlet

Role of Wnt5a in the proliferation of human glioblastoma cells Ji Min Yu

b

a,b

, Eun Sook Jun c, Jin Suk Jung d, Su Young Suh e, Jin Young Han f, Jee Yeon Kim g, Ki Wook Kim h, Jin Sup Jung a,b,*

a Department of Physiology, School of Medicine, Pusan National University, Pusan 602-739, Republic of Korea Medical Research Center for Ischemic Tissue Engineering, Pusan National University, Pusan 602-739, Republic of Korea c Medical Research Institute, Pusan National University Hospital, Pusan 602-739, Republic of Korea d Department of Pathology, College of Medicine, Dong-A University, Pusan 602-715, Republic of Korea e Department of Microbiology, College of Medicine, Dong-A University, Pusan 602-715, Republic of Korea f Department of Laboratory medicine, College of Medicine, Dong-A University, Pusan 602-715, Republic of Korea g Department of Pathology, School of Medicine, Pusan National University, Pusan 602-739, Republic of Korea h Department of Neurosurgery, College of Medicine, Dong-A University, Pusan 602-715, Republic of Korea

Received 20 April 2007; received in revised form 11 July 2007; accepted 11 July 2007

Abstract Wnt5a operates as either a tumor suppressor or a tumor stimulator, according to tumor type. The functions of Wnt5a in human glioblastoma (GBM) have yet to be determined. We initially evaluated the expression of Wnt5a in human glioma. The results of immunohistochemical analyses have revealed that Wnt5a expression was higher in human GBM than in normal brain tissue and low-grade astrocytoma. In order to assess the role of Wnt5a on proliferation in human glioblastoma cells, we employed U87MG and GBM-05, a newly established GBM cell line. GBM-05 was established from a patient diagnosed with GBM. GBM-05 cells were shown to express Nestin, but did not express GFAP and Map2ab. GBM-05 cells formed infiltrating brain tumors after being intracerebrally transplanted into nude mice, and xenotransplanted GBM-05 cells were observed to differentiate into neuronal and astrocyte lineages. Wnt5a expression in the xenotransplanted tumors was higher than that detected in the surrounding brain tissues. The overexpression of Wnt5a increased the proliferation of GBM-05 and U87MG in vitro. By way of contrast, the downregulation of Wnt5a expression as the result of RNA interference reduced proliferation from GBM-05 and U87MG cells in vitro, and reduced tumorigenicity in vivo. Our data indicate that Wnt5a signaling is an important regulator in the proliferation of human glioma cells.  2007 Elsevier Ireland Ltd. All rights reserved. Keywords: Wnt5a; Proliferation; Glioblastoma; Human

1. Introduction *

Corresponding author. Address: Department of Physiology, School of Medicine, Pusan National University, 1 Ga, Ami-Dong, Suh-Gu, Pusan 602-739, Republic of Korea. Tel.: +82 51 240 7734; fax: +82 51 246 6001. E-mail address: [email protected] (J.S. Jung).

Glioblastoma multiforme (GBM) is the most common type of primary brain tumor, and accounts for approximately 50% of all gliomas and for 12–15% of all intracranial tumors [1]. Glioblastomas are comprised of morphologically diverse cells that

0304-3835/$ - see front matter  2007 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.canlet.2007.07.011

J.M. Yu et al. / Cancer Letters 257 (2007) 172–181

express a broad variety of neural lineage markers [2]. In recent years, an understanding of the numerous genetic changes occurring in gliomas has begun to emerge. As demonstrated in conjunction with many malignancies, a stepwise progression of molecular genetic events involving the overexpression of proto-oncogenes and the loss of tumor suppressor genes underlies the development of malignant gliomas [3]. A number of Wnt genes, including Wnt2, Wnt7b, and Wnt5a, have been associated with a variety of tumors [4,5]. The role of Wnt5a in tumorigenesis remains a matter of some controversy. Wnt5a has been identified as a tumor suppressor in thyroid carcinoma [6], breast cancer [7], and hematopoietic tumors [8]. By way of contrast, Wnt5a has been shown to stimulate cell migration and invasion in breast carcinoma cells [9], gastric cancer cells [10], and melanoma [11]. Wnt5a overexpression was associated with the proliferation of tumor cells in non-small cell lung cancer [12]. These reports show that Wnt5a functions as either a tumor suppressor or simulator according to the tumor type. The functions of Wnt5a in glioma have yet to be determined. Wnt signals have been shown to be involved in a variety of developmental processes, including neurogenesis in the nervous system [13]. Canonical Wnt/b-catenin signal transduction pathways have been suggested to variably influence proliferation and lineage decisions of neural progenitor cells (NPCs), as well as their progeny [14]. The noncanonical Wnt molecule, Wnt5a, is expressed in proliferating neural stem cells, and the expressional levels increase during differentiation [15]. Wnt5a is an important component of dopaminergic inductive activity in the ventral midbrain glia [16]. In our previous study, it was shown that Wnt5a increased neuronal differentiation, as well as postnatal NPC differentiation [17]. The results of recent studies have suggested that glioma harbor cells with stem cell properties, including neurosphere formation and self-renewal [18], and that glioma originate from the oncogenic transformation of stem or progenitor cells. The objective of this study was to ascertain whether Wnt5a is involved in glioma progression. In order to study this, we evaluated Wnt5a expression in human glioma tissues and assessed the role of Wnt5a in a newly established cell line (GBM05) derived from the GBM tissues of a 71-year-old male and U87MG.

173

2. Materials and methods 2.1. Patient Normal brain samples were obtained from five autopsy cases, with the prior informed consent of all participants. Astrocytoma and GBM tissues were obtained from biopsy samples with the patient’s consent, as well as the approval of the Institutional Review Board. The tumors obtained from 10 patients with gliomas (five diffuse astrocytomas and five GBM; patient age range, 18–71 years; 6 men and 4 women) were classified in accordance with the most recent World Health Organization (WHO) classification criteria [19]. For the GBM primary culture, the surgical specimen was obtained from a 71-year-old male patient. This patient had no previous medical history, with the exception of a peptic ulcer diagnosed more than 10 years ago.

2.2. Primary tumor cell culture GBM tissues were digested in a solution comprised of papain (2.5 U/ml; Worthington), DNase (250 U/ml; Worthington), and neutral protease (1 U/ml Dispase; Boehringer-Mannheim) dissolved in HBSS. Whole digested tissues were then suspended in DMEM/F12– 10% FBS and mixed with an equal volume of Percoll solution. The Percoll solution was prepared by mixing nine parts of Percoll (Sigma) with one part 10· PBS. The cell suspension was then fractionated via 30 min of centrifugation at 4 C at 20,000g. The cell fractions were harvested and grown at clonal density (2500–5000 cells/ cm2) as tumorospheres in neurobasal medium (Invitrogen) containing 20 ng/ml of both epidermal growth factor (EGF) and fibroblast growth factor-2 (FGF2) (R&D) and a 1:50 dilution of B27 (Invitrogen) in a petri dish, or were grown adhering to laminin-coated dishes. After 10 passages, GBM-05 cells can be propagated in DMEM-10% FBS without laminin-coating. GBM-05 cells at the 15–20th passages were utilized in most of our experiments.

2.3. Reverse transcription-polymerase chain reaction (RTPCR) PCR amplification was conducted using the following primer sets. Nestin, 5 0 -ACC AAG AGA CAT TCA GAC TCC-3 0 , 5 0 -CCT CAT CCT TAT TTT CCA CTC C-3 0 , Map2ab, 5 0 -CAG CAA AGG GAT ACT TTC AC-3 0 , 5 0 -ATG CTT TTT GTT GCT TCT TC-3 0 , GFAP, 5 0 -GCT CGA TCA ACT CAC CGC CAA CA-3 0 , 5 0 GGG CAG CAG CGT CTG TCA GGT C-3 0 , Wnt5a, 5 0 -GGA ATA TTA AGC CCA GGA GT-3 0 , 5 0 - AGT GGC ACA GTT TCT TCT GT-3 0 , GAPDH 5 0 -TCCATG ACAACTTTGGTATCG-3 0 , 5 0 -TGTAGCCAAATTCG

174

J.M. Yu et al. / Cancer Letters 257 (2007) 172–181

TTGTCA-3 0 . Duplicate PCRs were amplified with the designed primers, and GAPDH was utilized as a control for the evaluation of PCR efficiency. 2.4. Chromosome analysis

GAA TCG GAT ATC AAA TTC CGA CAT CGA AGG TGG AAC GG-3 0 . Semiquantitative analysis of transcript abundance relative to GAPDH expression was conducted using Image Analysis software (Image Gauge ver 3.0, Fuji, Japan).

GBM-05 cells were treated with medium containing 10 lg/ml of colcemid (Irvine Scientific) for 1–2 h, then resuspended for 30 min in hypotonic 1% sodium citrate at room temperature. The cells were then washed in methanol–acetic acid (3:1, v/v) fixative solution for 30 min, then spread onto clean dry slides. Trypsin–Giemsa banding staining was then conducted, and 20 metaphases were analyzed for the samples.

2.8. Fluorescent activated cell sorting

2.5. Transfection of Wnt5a cDNA

The sorted GFP-positive glioma cells were plated at a density of 1000/cm2. Cells were determined via direct counting at 24, 48, and 72 h after plating. For the colony forming unit (CFU) assay, 60 cells were transferred to 10 cm2 plates. The colonies of GFP-positive cells in each colony were counted 7 days after plating.

All transient transfections were conducted using Lipofectamine Plus Reagent (Invitrogen). Transient transfections were conducted with pUSEamp(+)Wnt5a and pUSEamp(+) vectors (Upstate). The transfected cells were then cultured for 24 h in DMEM–10% FBS.

The lentivirus-infected glioma cells were passed through a 40–70 lm filter in order to eliminate clumps (BD Falcon), and the GFP-positive cells were sorted via FACS (FACSaria, BD Biosciences). 2.9. Proliferation assay

2.10. In vivo engraftment mode using Balb/c nude mice 2.6. Western blot of Wnt5a expression The cultured cells were trypsinized and washed in cold PBS, and the cell pellets were resuspended in ice-cold lysis buffer (210 mM mannitol, 70 mM sucrose 5 mM Tris, pH 7.5, 1 mM EDTA) supplemented with protease inhibitors. After 15 min of incubation on ice, the cells were homogenized (15 strokes). The total cell lysates were then centrifuged (15,000g, 10 min, 4 C) and the supernatants were employed for further processing. After the protein concentration was determined using a QuantiT Protein assay kit (Invitrogen), the proteins were loaded on 10% SDS–polyacrylamide gels, electrotransferred to nitrocellulose membranes (Hybond-ECL, Amersham Pharmacia Biotech, Piscataway, NJ) and probed with primary antibody (Wnt5a, R&D; GAPDH, BD Biosciences). Immunoreactive bands were detected with anti-mouse and anti-goat peroxidase-conjugated secondary antibodies (Amersham Pharmacia Biotech, Piscataway, NJ), and visualized via enhanced chemiluminescence (ECL Detection kit, Amersham Pharmacia Biotech, Piscataway, NJ). TM

2.7. Construction of lentivirus siRNA vector and transduction of letivirus

Balb/c nude mice were utilized for the xenotransplantation of glioma cells. The animals were anesthetized with an intraperitoneal injection of pentothal sodium (0.5 mg/ g) and positioned in a small animal stereotaxic apparatus to which a microinjector unit (WPI) was affixed. In order to measure the tumorigenesis effects of GBM-05 after the establishment of the cell line, 102–106 cells in 10 ll PBS per mouse were stereotactically injected into the area adjacent to the bregma site, and the mice were killed after 5– 8 weeks according to the number of injected cells and the development of neurological deficits. For the Wnt5a experiment, each mouse received 2 · 105 U87MG or GBM-05 cells in 10 ll PBS. The mice were killed after 5 weeks for the Wnt5a experiment, and were perfused with 4% (w/v) paraformaldehyde. Images of the H&E-stained sections containing tumors were analyzed with Metamorpho image analysis software. The tumors in each section were outlined manually using a freehand selection tool in order to measure the tumor area in mm2. The area was then multiplied by the section thickness to generate the section volume measurements. The volumes of all of the sections were added to determine the total volume for each tumor. 2.11. Histopathology and immunohistochemistry

Lentiviral vector construction was carried out as previously described [21]. The oligonucleotides used for the generation of human Wnt5a shRNA were as follows: F 5 0 -GTT ACC CGT TCC ACC TTC GAT GTC GGA ATT TGA TAT CCG ATT CCG ACA TCG AAG GTG GAA TTT TTT CCA AGC-3 0 , 5 0 -GGC CGC TTG GAA AAA ATT CCA CCT TCG ATG TGC

Patient tumor tissues and mouse brains were immediately removed and fixed in 4% paraformaldehyde for 24 h. The tissue samples were processed with an auto tissue processing machine (Leica) and embedded in paraffin. The brains were sectioned at a thickness of 5 lm and H&E stained as a standard histopathological tech-

J.M. Yu et al. / Cancer Letters 257 (2007) 172–181

nique. The antibodies were diluted to 1:1000 for rabbit GFAP (Dako), 1:200 for mouse Nestin (Pharmingen), and 1:400 for mouse Map2ab (Sigma). Wnt5a antibody (R&D) was employed at a concentration of 15 lg/ml. Immunodetection was conducted using the Elite Vector Stain ABC System (Vector Laboratories). Haematoxylin was the preferred counterstain for nuclear detail. After the incubation of primary antibodies, the slides were washed in PBS and incubated for 1 h with TRITC-conjugated anti-mouse, TRITC-conjugated anti-rabbit, and FITC-conjugated anti-goat antibodies (Jackson Immunoresearch Laboratories). Fluorescent antibody-labeled tissues were visualized via confocal microscopy (Leica). The quantitative analysis of expression intensity was evaluated relative to Wnt5a expression in normal tissues using Image Analysis software (Image Gauge ver 3.0, Fuji).

175

3. Results 3.1. Immunohistochemistry of Wnt5a protein normal and glioma samples In order to evaluate Wnt5a expression in normal tissues and glioma samples, we obtained five samples from autopsy cases, five samples of grade II astrocytoma, and five samples of GBM (grade IV glioma). The measurement of the percentage of Wnt5a(+) cells and the signal intensity of Wnt5a(+) cells as determined via the immunohistochemistry analysis of five different samples indicated that Wnt5a expression was the lowest in normal tissues, intermediate in grade II astrocytoma, and the highest in the GBM samples (Fig. 1a–c). In order to verify the increased expression of Wnt5a in GBM, Wnt5a expression in the normal brain and GBM tissues was determined

Fig. 1. Immunohistochemistry of Wnt5a in human normal brains and glioma tissues (a) Normal and glioma tissues were immunohistochemically stained for Wnt5a expression. Wnt5a expression was higher in WHO grade II astrocytoma (middle) and WHO grade IV glioblastoma (lower) as compared to normal brain tissue (upper). Wnt5a (green), Nuclei (red). (b) Measurement of the percentage of Wnt5a(+) cells relative to total cells. Total cell numbers were counted via propidium iodide staining (PI+). (c) Comparison of the signal intensity of Wnt5a on the immunohistochemical images. The signal intensity of Wnt5a(+) cells was quantitated with an image analyzer. (d and e) RT-PCR analysis of Wnt5a expression in normal cells and GBM. The quantification of Wnt5a expression level via RTPCR was determined using an image analyzer. The data are expressed as means ± SEM (n = 5). *p < 0.05, significant difference from control.

176

J.M. Yu et al. / Cancer Letters 257 (2007) 172–181

via RT-PCR analysis. Wnt5a gene expression was higher in the GBM tissues than in the normal brain tissues (Fig. 1d and e). 3.2. Patient tumor tissues expressed Nestin, Map2ab and GFAP We established a new tumor cell line (GBM-05) from the GBM of a 71-year-old patient. The histological features of the patient tumors, upon hematoxylin and eosin staining, evidenced increased cellularity, elevated mitotic activity, microvascular proliferation, and necrosis (Fig. 2a). The expression of neural markers in patient tumor tissues was determined via immunohistochemistry. The patient tumor tissues evidenced immunoreactivity for Nestin, GFAP, and Map2ab (Fig. 2b). 3.3. GBM-05 cells were able to form spheres, similarly to neural progenitor cells GBM-05 derived from tumorospheres of GBM grew adherent to culture dishes. These cells evidenced a round shape with processes, as well as a high nuclear-to-cytoplasmic ratio (Fig. 2c). GBM-05 cells were subcultured as far as passage 40 (about 12 months) without any special treatment applied to the culture dish during the collection of data described herein. The results of RT-PCR analysis indicated that GBM05 cells expressed Nestin, but did not express markers of differentiated neural cell types, such as GFAP for astrocytes and Map2ab for neurons, in contrast with the patient tumor tissue samples (Fig. 2d). Karyotype analysis was conducted using metaphase preparations obtained from GBM-05 cells at the seventh passage. No numerical aberrations were detected, but a host of structural aberrations were observed in GBM-05. The structural aberrations 46 XY, der(1p), der(6q), der(9p+), del(9)(p13), del(10)(q23), der(11q+), del(12)(q21), del(13)(q14), der(16p+), der(20p+), del(22)(q13) [20] were present in all of the metaphase cells examined (Fig. 2e), although variations were observed in some cells, including the loss of the y chromosome. The long-term cultivation of established cells had no effect on their karyotypic signatures (data not shown). 3.4. In vivo tumorigenic potential of GBM-05 cells In order to evaluate the tumorigenic potential of established GBM-05 cells, we injected 102–106 cells into the brains of Balb/c nude mice. Five to eight weeks after transplantation according to the number of injected cells, all of the transplanted mice developed brain tumors. As was indicated by H&E staining, the transplanted tumor was composed of highly mitotic, large cell nuclei with irregular shapes. The transplanted tumor tissues were found to be immunoreactive for GFAP, Map2ab, and

Nestin, as in the original patient tumors (Fig. 2f). In order to determine whether the transplanted cells express neural markers, we labeled glioma cells with the GFP lentivirus. GFP-labeled cells also form tumors in nude mice. The immunohistochemical results indicated that the xenotransplanted GFP cells expressed Nestin, Map2 and GFAP (Fig. 2g). 3.5. Wnt5a expression in glioma tissue We utilized GBM-05 and U87MG in further experiments. In order to further verify the role of Wnt5a in glioma formation, we assessed Wnt5a expression in xenotransplanted tumors via immunohistochemistry. As the Wnt5a antibody to be used in this experiment was derived from mouse Wnt-5a peptide harboring amino acid residues Gln 254–Cys 334 of mouse Wnt-5a and the amino acid sequences of human Wnt5a in this region are identical, the antibody is capable of recognizing Wnt5a protein in both mice and humans. The results of immunohistochemical analyses indicate that Wnt5a protein is expressed at higher levels in the xenotransplanted tumor region than was observed in the adjacent normal area (Fig. 2h). 3.6. Wnt5a overexpression promotes glioma cell proliferation In order to determine the role of Wnt5a in glioma cell proliferation, Wnt5a was overexpressed in GBM-05 and U87MG cells via the transient transfection of the Wnt5a plasmid. We determined the efficiency of plasmid transfection with a plasmid encoding for green fluorescent protein (GFP) under the control of the CMV promoter. By the second day after transfection, approximately 75% of the total glioma cells were GFP-positive. The transfected cells were replated at 1000 cells/cm2. Wnt5a-overexpressed cells evidenced higher numbers of cells 3 days after plating than was observed with the vector-transfected cells (Fig. 3). 3.7. Effect of downregulation of Wnt5a expression via the transduction of siGFP lentivirus on the proliferation and tumorigenesis of human glioma cells In order to further confirm the role of Wnt5a in the proliferation of glioma cells, the expression of Wnt5a was downregulated via RNA interference. GBM-05 and U87MG were infected with LV-GFP and LV-GFPsiWnt5a, respectively. The GFP-positive cells were sorted by FACS. The RT-PCR analysis of sorted cells showed that Wnt5a expression was downregulated significantly in LV-GFP-siWnt5a-transduced cells as compared with the LV-GFP-transduced cells (Fig. 4a). Western blot analysis indicated that Wnt5a cDNA transfection increased Wnt5a protein levels and Wnt5a siRNA transduction

J.M. Yu et al. / Cancer Letters 257 (2007) 172–181

177

Fig. 2. Histopathologic features of patient tumor tissue and transplanted tumor. (a) Patient’s tumor tissues were stained with haematoxylin and eosin. High vascularity and necrotic areas were observed. (b) Immunohistochemical staining was conducted using the Elite Vector Stain ABC System. Paraffin-embedded sections were labeled with Nestin, GFAP, and Map2ab. (c) Morphology of cultured GBM-05 cells. (d) RT-PCR of gene expression of GBM-05 cells and patient tumor tissues. (e) Karyotypic analysis of GBM-05 cells. (f) GBM-05 cells were transplanted into the brains of Balb/c nude mice. Paraffin-embedded tissues were stained with H&E and neural markers. (g) Immunofluorescent staining of the tumors generated from LV-GFP-infected GBM-05. GBM-05 cells (green), Nestin (red), GFAP (red) and Map2ab (red). (h) Wnt5a expression in xenotransplanted tumors. Xenotransplanted tumor sections were stained with Wnt5a antibody (green).

178

J.M. Yu et al. / Cancer Letters 257 (2007) 172–181

Fig. 3. Overexpression of Wnt5a promotes the proliferation of glioma cells. The proliferation of GBM-05 and U87MG cells were determined via direct cell counting for 4 days after cell plating. The data are expressed as means ± SEM (n = 3). *p < 0.05, significant difference from control.

reduced it (Fig. 4b). In order to determine the effects of Wnt5a siRNA on the proliferation of glioma cells, the lentivirally infected GFP-positive glioma cells were sorted via FACS. For the colony forming unit (CFU) assay, 60 GFP-positive GBM-05 or U87MG were plated on 100 cm culture plates. The numbers of CFU in the LVGFP-siWnt5a-infected glioma cells at 3 days to 1 week after plating were significantly lower than those observed in the LV-GFP-infected glioma cells. Furthermore, the numbers of cells per LV-GFP-siWnt5a expressing colony were found to be significantly lower than in the colonies formed from the LV-GFP-infected glioma cells (Fig. 4c–e) We attempted to determine whether Wnt5a promotes tumorgenesis in vivo. LV-GFP-siWnt5a or LV-GFPinfected cells (2 · 105 cells per animal) was injected stereotactically into the brains of Balb/c nude mice. The brains in the LV-GFP cell-transplanted group evidenced the formation of extruding tumors after 5 weeks, but the LVsiWnt5a cell-transplanted group mice evidenced no changes in the external countours of the brain. Tumor volume was measured in the histologic sections of xenotransplanted tumors. LV-GFP-siWnt5a cells-derived tumors were significantly smaller than the tumors derived from the control LV-GFP cells (GBM-5, 60 mm3 vs 12 mm3; U87MG, 31.5 mm3 vs 2.75 mm3; p < 0.05) (Fig. 4f and g).

4. Discussion GBM is the most frequently identified malignant brain tumor. Despite recent advances in surgery and adjuvant therapy, GBM patients continue to have generally poor prognoses due to incomplete resection and resistance to radiotherapy and chemotherapy [22]. Recent studies have suggested that GBM patients may benefit from molecularly targeted therapies [23]. The concept of targeted cancer therapies is the blockage of the growth and spreading of cancer via interference with specific molecules that perform crucial functions in the cancer development

and progression. Growth factors, growth factor receptors (usually receptor tyrosine kinases), and receptor-mediated signaling pathways are known to be critical in gliomagenesis due to their ability to affect fundamental cellular processes, including differentiation, proliferation, survival, migration, and metabolism [24]. In this study, we determined that the expression of Wnt5a is increased significantly in GBM as compared with normal tissue or astrocytoma. We suggest that Wnt5a may be associated with tumor progression or aggressiveness, and also attempted to determine whether the up- and downregulation of Wnt5a expression affects the proliferation of tumor cells. In service of this goal, we utilized a newly established cell line and U87MG as a wellknown glioblastoma cell line. The new cell line, GBM-05, was established from tumorospheres obtained from the GBM tissues of a 71-year-old male patient. The patient had no clinical problems associated with brain tumor prior to admission, thereby indicating that the tumor arose de novo (primary). The patient tumor tissue used for the establishment of the cell line evidenced Nestin, GFAP, and Map2ab expression, as shown by the results of immunohistochemistry and RT-PCR. The cultured GBM-05 cells displayed multiple and stable chromosomal changes. GBM-05 cells transplanted into the brains of nude mice retained the potential to differentiate into neuronal or glial lineages, although they did not express Map2ab or GFAP during the in vitro culture. Therefore, GBM-05 cells may provide an advantageous resource, in addition to the available glioma cell lines, for therapeutic and diagnostic purposes, and to develop patient-tailored pharmacological approaches to a cure for GBM. Wnt5a overexpression by the transfection of cDNA expression plasmids increased the prolifera-

J.M. Yu et al. / Cancer Letters 257 (2007) 172–181

179

Fig. 4. Effect of Wnt5a downregulation in vitro and in vivo. (a) Wnt5a expression of GBM-05 and U87MG infected with LV-GFP or LVGFP-si-Wnt5a was analyzed via RT-PCR. (b) Immunoblot analysis of control, Wnt5a-overexpressed or Wnt5a-downregulated U87MG or GBM-05 cells. (c) The CFU assay was performed by plating 60 cells onto a 100 cm2 culture dish. A representative fluorescent photograph of a lentivirus-transduced glioma cell colony is shown. (d and e) The colony numbers and cell numbers/colony of GFPpositive cells were counted 7 days after plating. The data are expressed as means ± SEM (n = 5). *p < 0.05, significant difference from control. (f) Morphological analysis of siWnt5a effects on tumor growth. Cells (2 · 105) of tumor cells were injected into nude mouse brains and allowed to develop for 5 weeks. Mice were sacrificed, and the brains were removed, fixed, paraffin embedded, and sectioned (7 lm). Serial sections were stained with H&E. (g) siWnt5a-mediated changes in tumor volume. Brains were sectioned and H&E stained, and total tumor volume was calculated by adding the volumes of individual tumor sections. The data are expressed as means ± SEM (n = 4). *p < 0.05, significant difference from control data.

180

J.M. Yu et al. / Cancer Letters 257 (2007) 172–181

tion of human glioma cells in vitro. Although the efficiency of transient transfection in this experiment was shown to be approximately 75%, the Wnt5a generated in the transfected cells could exert its effects on nontransfected cells, because Wnt5a is a soluble factor. The results of the direct cell counting and colony forming unit assay indicated that the downregulation of Wnt5a as the consequence of the lentiviral transduction of Wnt5a shRNA effected a reduction in the proliferation of GBM-05 and U87MG cells in vitro. The effects of Wnt5a on the glioblastoma cell proliferation were confirmed further by the observation of reduced tumorigenesis in nude mice into which si-Wnt5a-transduced cells were xenotransplanted. Collectively, these data indicated that Wnt5a signaling performs an important function in the proliferation of glioblastoma. The function of Wnt5a in the central nervous system has yet to be fully elucidated. Wnt5a has been reported to be associated with patterning decisions in the embryonic nervous system [25], and to be involved in dopaminergic differentiation and midbrain development [26]. Recently, reduced Wnt5a gene expression has been observed in cases of high-risk neuroblastoma, as well as in cultured metastatic neuroblasts. The retinoic acid-induced differentiation of neuroblastoma cells increased Wnt5a expression [27]. As neuroblastomas arise from embryonal cells committed to the development of the sympathetic nervous system [28] and Wnt5a evidences differential activity according to neural lineage [26], the differential effects of Wnt5a in the proliferation of neuroblastoma and glioblastoma may result from the fact that they originate from different neural lineages. In our previous study, it was demonstrated that Wnt5a increased neural stem cell proliferation [17]. The similarities between the self-renewal mechanisms of stem cells and cancer stem cells have culminated in the recently proposed cancer stem cell model hypothesis, which asserts that brain tumor stem cells arise from intrinsic mutations in normal stem/progenitor cells, thereby leading to their uncontrolled proliferation [29]. The finding that Wnt5a increases the proliferation of human glioblastoma cells in this study may provide supportive evidence regarding this hypothesis. Acknowledgement This work was supported by MRC program of MOST/KOSEF (R13-2005-009).

References [1] P.L. Lantos, S.R. Vanderberg, P. Kleihues, Tumours of the nervous system, in: D.I. Graham, P.L. Lantos, (Eds.), Greenfield’s neuropatholology, sixth ed., Arnold 2, London, 1997, pp. 600–626. [2] A. Salmaggi, A. Boiardi, M. Gelati, A. Russo, C. Calatozzolo, E. Ciusani, F.L. Sciacca, A. Ottolina, E.A. Parati, C. La Porta, G. Alessandri, C. Marras, D. Croci, M. De Rossi, Glioblastoma-derived tumorospheres identify a population of tumor stem-like cells with angiogenic potential and enhanced multidrug resistance phenotype, Glia 54 (2006) 850–860. [3] A.D. Norden, P.Y. Wen, Glioma therapy in adults, Neurologist 12 (2006) 279–292. [4] M.J. Smalley, T.C. Dale, Wnt signalling in mammalian development and cancer, Cancer Metastasis Rev. 18 (1999) 215–230. [5] M. Jonsson, K. Smith, A.L. Harris, Wnt and fibroblast growth factor gene expression during development of the mammary gland and role of wnts in human cancer, in: R.B. Dickson, D.S. Salomon (Eds.), Hormones and Growth Factors in Development and Neoplasia, Wiley & Liss, New York, 1998, pp. 361–382. [6] N. Kremenevskaja, R. von Wasielewski, A.S. Rao, C. Schofl, T. Andersson, G. Brabant, Wnt5a has tumor suppressor activity in thyroid carcinoma, Oncogene 24 (2005) 2144– 2154. [7] M. Jonsson, J. Dejmek, P.O. Bendahl, T. Andersson, Loss of Wnt5a protein is associated with early relapse in invasive ductal breast carcinomas, Cancer Res. 62 (2002) 409–416. [8] H. Liang, Q. Chen, A.H. Coles, S.J. Anderson, G. Pian, A. Bradley, R. Gerstein, R. Jurecic, S.N. Jones, Wnt5a inhibits B cell proliferation and functions as a tumor suppressor in hematopoietic tissue, Cancer Cell 4 (2003) 349–360. [9] T. Pukrop, F. Klemm, T. Hagemann, D. Gradl, M. Schulz, S. Siemes, L. Trumper, C. Binder, Wnt5a signaling is critical for macrophage-induced invasion of breast cancer cell lines, Proc. Natl. Acad. Sci. 103 (2006) 5454–5459. [10] M. Kurayoshi, N. Oue, H. Yamamoto, M. Kishida, A. Inoue, T.W. Yasui, A. Kikuchi, Expression of Wnt5a is correlated with aggressiveness of gastric cancer by stimulating cell migration and invasion, Cancer Res. 66 (2006) 10439–10448. [11] T.B. Lewis, J.E. Robison, R. Bastien, B. Milash, K. Boucher, W.E. Samlowski, S.A. Leachman, R. Dirk Noyes, C.T. Wittwer, L. Perreard, P.S. Bernard, Molecular classification of melanoma using real-time quantitative reverse transcriptase-polymerase chain reaction, Cancer 104 (2005) 1678– 1686. [12] C.L. Huang, D. Liu, J. Nakano, S. Ishikawa, K. Kontani, H. Yokomise, M. Ueno, Wnt5a expression is associated with the tumor proliferation and the stromal vascular endothelial growth factor—an expression in non-small-cell lung cancer, J. Clin. Oncol. 23 (2005) 8765–8773. [13] J. Huelsken, W. Birchmeier, New aspects of Wnt signaling pathways in higher vertebrates, Curr. Opin. Genet. Dev. 11 (2001) 547–553. [14] Y. Hirabayashi, Y. Gotoh, Stage-dependent fate determination of neural precursor cells in mouse forebrain, Neurosci. Res. 51 (2005) 331–336.

J.M. Yu et al. / Cancer Letters 257 (2007) 172–181 [15] C. Lange, E. Mix, K. Rateitschak, A. Rolfs, Wnt signal pathways and neural stem cell differentiation, Neurodegener Dis. 3 (2006) 76–86. [16] G. Castelo-Branco, K.M. Sousa, V. Bryja, L. Pinto, J. Wagner, E. Arenas, Ventral midbrain glia express regionspecific transcription factors and regulate dopaminergic neurogenesis through Wnt5a secretion, Mol. Cell Neurosci. 31 (2006) 251–262. [17] J.M. Yu, J.H. Kim, G.S. Song, J.S. Jung, Increase in proliferation and differentiation of neural progenitor cells isolated from postnatal and adult mice brain by Wnt-3a and Wnt5a, Mol. Cell Biochem. 288 (2006) 17–28. [18] R. Galli, E. Binda, U. Orfanelli, B. Cipelletti, A. Gritti, S. De Vitis, R. Fiocco, F. Foroni, C.h. Dimeco, A. Vescovi, Isolation and characterization of tumorigenic, stem-like neural precursors from human glioblastoma, Cancer Res. 64 (2004) 7011–7021. [19] P. Kleihues, D.N. Louis, B.W. Scheithauer, L.B. Rorke, G. Reifenberger, P.C. Burger, W.K. Cavenee, The WHO classification of tumors of the nervous system, J. Neuropathol. Exp. Neurol. 61 (2002) 215–225. [20] V. Tropepe, M. Sibilia, B.G. Ciruna, J. Rossant, E.F. Wagner, D. van der Kooy, Distinct neural stem cells proliferate in response to EGF and FGF in the developing mouse telencephalon, Dev. Biol. 208 (1999) 166–188. [21] H.H. Cho, Y.J. Kim, S.J. Kim, J.H. Kim, Y.C. Bae, B.A. Bunnell, J.S. Jung, Endogenous Wnt signaling promotes proliferation and suppresses osteogenic differentiation in human adipose derived stromal cells, Tissue Eng. 12 (2006) 111–121.

181

[22] R. Stupp, W.P. Mason, M.J. van den Bent, M. Weller, B. Fisher, M.J. Taphoorn, K. Belanger, A.A. Brandes, C. Marosi, U. Bogdahn, J. Curschmann, R.C. Janzer, S.K. Ludwin, T. Gorlia, A. Allgeier, D. Lacombe, J.G. Cairncross, E. Eisenhauer, Mirimanoff, Radiotherapy plus concomitant and adjuvant temozolomide for glioblastoma, N. Engl. J. Med. 352 (2005) 987–996. [23] C. Sawyers, Targeted cancer therapy, Nature 432 (2004) 294–297. [24] S. Horvath, B. Zhang, M. Carlson, K.V. Lu, S. Zhu, R.M. Felciano, M.F. Laurance, W. Zhao, S. Qi, Z. Chen, Y. Lee, A.C. Scheck, L.M. Liau, H. Wu, D.H. Geschwind, P.G. Febbo, H.I. Kornblum, T.F. Cloughesy, S.F. Nelson, P.S. Mischel, Analysis of oncogenic signaling networks in glioblastoma identifies ASPM as a molecular target, Proc. Natl. Acad. Sci. 103 (2006) 17402–17407. [25] E. Blanc, D. Goldschneider, S. Douc-Rasy, J. Benard, G. Raguenez, Wnt5a gene expression in malignant human neuroblasts, Cancer Lett. 228 (2005) 117–123. [26] G. Castelo-Branco, J. Wagner, F.J. Rodriguez, J. Kele, K. Sousa, N. Rawal, Differential regulation of midbrain dopaminergic neuron development by Wnt-1, Wnt-3a, and Wnt5a, Proc. Natl. Acad. Sci. 100 (2003) 12747–12752. [27] E. Blanc, G.L. Roux, J. Be´nard, G. Rague´nez, Low expression of Wnt-5a gene is associated with high-risk neuroblastoma, Oncogene 24 (2005) 1277–1283. [28] J. Mora, W.L. Gerald, Origin of neuroblastic tumors: clues for future therapeutics, Expert. Rev. Mol. Diagn. 4 (2004) 293–302. [29] S.K. Nicolis, Cancer stem cells and ‘‘stemness’’ genes in neuro-oncology, Neurobiol Dis. 25 (2007) 217–229.