Vasohibin-2 Expressed in Human Serous Ovarian Adenocarcinoma Accelerates Tumor Growth by Promoting Angiogenesis

Ovarian cancer is the second most common malignant tumor in gynecology and is the leading cause of cancer-related death for women worldwide (1). Cytotoxic therapy with platinum and taxanes is initially effective in many cases of ovarian cancer, but there is a considerable risk of recurrence and resistance to such cytotoxic therapy (2). Thus, it is critical to develop alternative options that target pathways responsible for the progression of ovarian cancer. Angiogenesis is recognized as one of the principal hallmarks of various cancers (3). Indeed, tumor angiogenesis is thought to be a key process that enables ovarian cancer growth as well as dissemination in the peritoneal space, and thus one of the promising options for treating ovarian cancer is considered to be antiangiogenic therapy (4).

Ovarian cancer cells express various angiogenesis stimulators (1). Among them, VEGF plays the most important role, as it stimulates the migration and proliferation of, and tube formation by, endothelial cells. VEGF is the prototype of the VEGF family, and its proangiogenic signals are mainly transmitted via its type 2 receptor (VEGFR2) on endothelial cells (5). High levels of VEGF have been found in ovarian cancers, which is associated with poor survival of patients in both early and advanced stages of the disease (6, 7). In animal models of ovarian cancer, inhibition of VEGF reduces tumor growth and inhibits ascites accumulation (8). It is suggested that VEGF may also promote tumor growth by direct action on VEGF receptors expressed on ovarian cancer cells (9). Angiopoietins and their receptor, TIE2, are another ligand receptor system that regulates angiogenesis. Angiopoietin-1 (Ang-1) is an agonistic ligand of TIE2, and it facilitates pericyte covering of vessels for vascular stabilization, whereas Ang-2 is an antagonistic of ligand of TIE2 and induces detachment of pericytes from vessels for vascular destabilization (10). An increase in the Ang-2 level in cancers may be related to the immature phenotype of tumor vessels. Regarding these 2 angiopoietins, the serum Ang-2 level is significantly high in patients with ovarian cancer and is proposed to be a biomarker of malignant potential and poor prognosis in ovarian cancer (11).

The local balance between angiogenesis stimulators and inhibitors determines the occurrence and progress of angiogenesis. We recently isolated vasohibin-1 (VASH1) as a negative feedback regulator of angiogenesis that is induced in endothelial cells by angiogenesis stimulators such as VEGF and fibroblast growth factor 2 (FGF-2; refs. 12, 13). By conducting a database search, we found one gene homologous to VASH1 and named it VASH2 (14). The amino acid sequence of the human VASH2 protein is 52.5% homologous to that of human VASH1, and both VASH1 and VASH2 are highly conserved among species. VASH1 and VASH2 lack the classical signal sequence; but they bind to the small intracellular vasohibin-binding protein (SVBP), and this binding with SVBP facilitates their secretion (15). Because of the similarity between VASH1 and VASH2, we examined their expression and function by the use of hypoxia-induced subcutaneous angiogenesis in mice. Our analysis revealed that VASH1 is mainly expressed in endothelial cells in the termination zone to halt angiogenesis, whereas VASH2 is mainly expressed in mononuclear cells mobilized from the bone marrow in the sprouting front to stimulate angiogenesis (16). Thus, these 2 VASH family members regulate angiogenesis perhaps in a seemingly contradictory manner.

Previously, we investigated the expression of VASH1 under conditions accompanied by conducive to pathologic angiogenesis and showed its presence in endothelial cells of various cancers (17–21), atherosclerotic lesions (22), age-dependent macular degeneration (23), and diabetic retinopathy (24). However, the expression of VASH2 is ill-defined. Here, we conducted immunohistochemical analysis of VASH2 in ovarian cancers and showed for the first time the expression of VASH2 in them. This expression seems to be restricted to serous adenocarcinoma of the ovary. Our subsequent analysis indicated that VASH2 in ovarian cancer cells promoted tumor growth and peritoneal dissemination via the stimulation of angiogenesis.

This study was approved by the ethics committee of Jichi Medical University Hospital (Tochigi, Japan). Twenty-one patients with epithelial ovarian carcinoma who underwent surgery at Jichi Medical University Hospital between 2007 and 2009 were included in this study. Histologic types were assigned according to the criteria of the World Health Organization (WHO) classification.

Paraffin-embedded blocks of cancer tissue were prepared, and thin sections were cut and placed on glass slides. After deparafinization and hydration, endogenous peroxidase activity was quenched by a 5-minutes incubation in 3% hydrogen peroxide. The sections were then incubated for 1 hour at room temperature with anti-VASH2 mAb (14) at the concentration of 2 μg/mL. The samples were washed with Tris-buffered saline, and the color was developed using the EnVision⁺System (Dako), according to the manufacturer's instructions. Sections were counterstained with hematoxylin and mounted. A sample from which the primary antibody was omitted served as the negative control.

The human ovarian serous adenocarcinoma cell line SKOV-3 (25) was purchased from American Type Culture Collection (ATCC), and the DISS one was described previously (26). The murine lymphoma cell line EL-4 and murine Lewis lung carcinoma (LLC) were provided from the Cell Resource Center for Biomedical Research, Institute of Development, Aging, and Cancer, Tohoku University (Sendai, Japan). The murine malignant melanoma cell line B16F1 and murine Leydig tumor cell line MLTC-1 were purchased from ATCC. The murine fibrosarcoma cell line 505-05-01 and the murine ovarian carcinoma cell line OV2944-HM-1 were purchased from RIKEN Cell Bank (Tsukuba, Japan). These cell lines were maintained in Dulbecco's Modified Eagle Medium (DMEM; Wako) supplemented with 10% heat-inactivated fetal calf serum (FCS; JRH Biosciences and 100 μg/mL kanamycin (Meiji Seika Kaisha Ltd.). Human umbilical vein endothelial cells (HUVEC) and human microvascular endothelial cells (HMVEC) were obtained from Kurabo Industries, Ltd. and were cultured on type I collagen-coated dishes (IWAKI) in endothelial basal medium EBM-2 (Lonza) supplemented with EGM-2-MV-SingleQuots (Lonza) containing VEGF, FGF-2, insulin-like growth factor-I, EGF, and 5% FBS. MS1, an immortalized cell line with a SV40 large T antigen from mouse pancreatic endothelial cells, was purchased from ATCC and was cultured in αMEM (Wako) supplemented with 10% FCS. Human ovarian epithelial cells (HOEC) were purchased from ScienCell Research Laboratories, and were cultured on poly-l-lysine–coated dishes (IWAKI) in ovarian epithelial cell medium (ScienCell) supplemented with ovarian epithelial cell growth supplement (ScienCell).

All the cells were cultured at 37°C in a humidified atmosphere with 5% CO₂.

The short hairpin RNA (shRNA) sequence that suppresses VASH2 expression is 5′-CACCAGGTGATCTAGAATTGCATACGTGTGCTGTCCGTATGTAATTCTGGATCGCCTTTTTT. This sequence was inserted into the piGENE hU6 Vector (iGENE) to make the VASH2 shRNA expression vector. Cells were transfected with this vector or control vector by using Effectene transfection reagent. After the transfection, the cells were selected in puromycin (Calbiochem)-containing medium; and the bulk cells were obtained. Next, the bulk cells were seeded at a density of 0.3 cells per well in 96-well plates with 100 μL medium per well, and visible clones were picked and expanded in 24-well plates. These clones were finally transferred to regular cell culture flasks, and the VASH2 knockdown clones were thus established.

Human VASH2 cDNA was cloned into the pCALL2-pcDNA3.1/Hygro vector at multiple cloning sites (XhoI and NotI). To improve the activity of transcription, we replaced the cytomegalovirus (CMV) promoter of the pcDNA3.1/Hygro plasmid (Invitrogen) with the chicken β-actin promoter derived from pCALL2 (27). Cells were transfected with the VASH2 expression vector or control vector by using Effectene transfection reagent (QIAGEN) according to the manufacturer's protocol. After the transfection, the cells were selected in hygromycin (Invitrogen)-containing medium, and the clones having high VASH2 expression were established according to the procedure similar to that for the selection of the VASH2 knockdown clones.

Total RNA was extracted from cell cultures by using an RNeasy mini kit (QIAGEN) according to the manufacturer's instructions. Total tissue RNA was extracted from tumors with ISOGEN (Nippon Gene) according to the manufacturer's instructions. First-strand cDNA was generated by using ReverTra Ace (TOYOBO). The reverse transcriptase PCR (RT-PCR) procedure was carried out in a DNA thermal cycler (Takara). PCR conditions consisted of an initial denaturation step at 95°C for 5 minutes followed by 25 to 35 cycles consisting of 15 seconds at 95°C, 15 seconds at the appropriate annealing temperature, and 30 seconds at 72°C. PCR products were separated on a 1.5% agarose gel and visualized under ultraviolet by ethidium bromide staining. The primer pairs used were as follow: mouse β-actin forward, 5′-TCGTGCGTGACATCAAAGAG, and reverse, 5′-TGGACAGTGAGGCCAGGATG (annealing temperature: 58°C); human β-actin forward, 5′-ACAATGAGCTGCGTGTGGCT, and reverse, 5′-TCGTGCGTGACATTAAGGAGA (annealing temperature: 58°C); mouse VASH2 forward, 5′-TGGAGACAGCGAAGGAGATG, and reverse, 5′-GAAGCAACTTGTCCTCAACG (annealing temperature: 56°C); human VASH2 forward, 5′-AGCTGATGGACAAGCCATTG, and reverse, 5′-CTCTGAATGAAGTGGGCTATC (annealing temperature: 56°C).

SKOV-3 cells were grown to 60% to 70% confluence and then transfected with 50 nmol/L anti-miR-200b (Ambion), pre-miR-200b (Ambion), or their negative control oligonucleotides (Ambion) by using Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer's instructions. The medium was changed after 12 hours of transfection, and the cells cultured for an additional 48 hours.

Total RNA was extracted from HOECs, ovarian cancer cells, and ovarian cancer tissues using RNeasy Mini Kit. First-strand cDNA was generated using ReverTra Ace for RT-PCR. Quantitative real-time RT-PCR was carried out using the CFX96 real-time PCR detection system (Bio-Rad Laboratories) according to the manufacturer's instructions. PCR conditions consisted of an initial denaturation step at 95°C for 3 minutes, followed by 40 cycles of 10 seconds at 95°C, 10 seconds at 56°C, and 30 seconds at 72°C. Each mRNA level was measured as a fluorescent signal corrected according to the signal for β-actin. The primer pairs used were as follows: human VASH2 forward, 5′-TGCACACAGTCAAGAAGGTC-3′, and reverse, 5′-TTCTCACTTGGGTCGGAGAG-3′; human β-actin forward, 5′-ACAATGAGCTGCGTGTGGCT-3′, and reverse, 5′-TCTCCTTAATGTCACGCACGA-3′. miR-200b levels were analyzed by the TaqMan real-time PCR method. Ten nanograms of total RNA was reverse transcribed by using a TaqMan MicroRNA Reverse Transcription kit (Applied Biosystems) according to the manufacturer's instructions. The specific primer for miR200b was designed and produced by Applied Biosystems. Real-time PCR was carried out with the CFX96 real-time PCR detection system. PCR conditions consisted of an initial denaturation step at 95°C for 10 minutes, followed by 40 cycles of 15 seconds at 95°C and 60 seconds at 60°C. Relative expression levels were calculated by using the comparative C_t method.

Proliferation of tumor cells was measured by conducting the Tetra Color ONE cell proliferation assay (28). Briefly, the cells were seeded at a density of 3 × 10³ cells per well in a 96-well plate and incubated at 37°C. After 48 hours, 5 μL of Tetra Color ONE (Seikagaku Co.) was added to each well; and mixture was then incubated for an additional 2 hours. Absorbance at 450 nm was monitored.

VASH2-expressing EL-4 clone or EL-4 cells transfected with empty vector were cultured at 1 × 10⁶ cells/mL for 24 hours and centrifuged at 1,000 rpm for 5 minutes to obtain the conditioned medium (CM). Next, the cellular components were removed from the CM by using a MILLEX-GP PES 0.22-μm filter (Millipore) and concentrated 10-fold with a VIVASPIN15; MWCO 10,000 (Sartorius Stedim Biotech). MS1 cells were plated in a 96-well plate at 2 × 10³ cells per well and cultured in medium to which the CM had been added, and the proliferation was measured by using the Tetra Color ONE.

The migratory activity of endothelial cells was measured by use of the modified Boyden chamber method. MS1 cells were preincubated in 1% FCS in α-MEM for 24 hours and then plated at 5 × 10⁵ cells/mL on the upper chamber (insert) of a Boyden chamber (8.0-μm pore size, Corning). The low chamber was filled with CM corrected from VASH2 gene or mock transfectants as described above. MS1 cells were allowed to migrate for 6 hours, the cells that had migrated across the filter were stained with Difu Quick (Sysmex), and the number of cells that had migrated was counted in 5 fields per insert in a blind manner.

Female 6- to 8-week-old BALB/c nude mice were obtained from Clea Laboratories. VASH2^−/− mice on a C57BL/6 background were previously described (16). All of the animal experiments were approved by Tohoku University Center for Gene Research and carried out under the guidelines for animal experimentation of Tohoku University.

Tumor cells were subcutaneously transplanted into the back of mice at 2 × 10⁶ cells per mouse. Two dimensions of the tumors were measured every 3 days by using a caliper. The tumor volume was calculated by the formula: volume = (short diameter)² × (long diameter) × 1/2.

Tumor cells were intraperitoneally injected into BALB/c nude mice at 2 × 10⁶ cells per mouse. Two or 3 weeks after the injection, the mice were sacrificed. Thereafter, the ascites fluid was collected, and its volume was measured. Peritoneal dissemination was evaluated by counting the number of tumor nodes on the surface of the small intestines and mesentery. The survival of the mice was monitored twice daily. The survival rate was calculated by the Kaplan–Meier method.

Orthotopic tumor cell inoculation was conducted according to the method described by Cordero and colleagues. (29). Briefly, a small incision was made at the dorsomedial position and directly above the ovarian fat pad. The ovarian fat pad was gently pulled out and cancer cells (1 × 10⁵) were inoculated between the brusa and the ovary. Thirty-two days after the inoculation, mice were sacrificed and tumors in the ovary were examined.

For immunohistochemical analysis, tumors were frozen in optimum cutting temperature (OCT) compound (Sakura), cut into 7-μm sections, fixed in methanol for 20 minutes at −20°C, and blocked with 1% bovine serum albumin in PBS for 30 minutes at room temperature. Primary antibody reactions were conducted overnight at 4°C with rat monoclonal antibody against mouse CD31 (Research Diagnostics) at a dilution of 1:500, mouse monoclonal antibody against mouse αSMA (Sigma-Aldrich) at a dilution of 1:200. Secondary antibody reactions were conducted for 1 hour at room temperature with Alexa Fluor 488–conjugated donkey anti-rat IgG, Alexa Fluor 568–conjugated goat anti-rat IgG, Alexa Fluor 555–conjugated goat anti-mouse IgG (Molecular Probes) at a dilution of 1:500. After having been washed 3 times with PBS, the sections were covered with fluorescent mounting medium. All samples were analyzed with a BZ-9000 fluorescence microscope (KEYENCE) with a ×10, ×20, ×40, ×100 objective lens at room temperature. The vascular luminal area was calculated from 5 different fields. Quantitative analyses of vessels and cells were done by using BZ-H1C software (KEYENCE) and ImageJ software (http://rsbweb.nih.gov/ij/).

Deparaffinized and rehydrated tumor tissue sections were incubated overnight at 4°C with anti–proliferating cell nuclear antigen (PCNA) antibody (Santa Cruz Biotechnology) at a dilution of 1:200. The samples were incubated in biotin-conjugated antibody solution and streptavidin followed by staining with 3,3′-diaminobenzidine (DAB). The sections were then counterstained with Mayer hematoxylin, and the percentage of PCNA-positive cells was quantified by using HistoQuest software (NOVEL SCIENCE).

To evaluate apoptotic cancer cells, we conducted terminal deoxynucleotidyl transferase–mediated dUTP nick end labeling (TUNEL) staining. Deparaffinized and rehydrated tumor tissue sections were immersed in protease solution and were incubated with terminal deoxynucleotidyl transferase to label 3′ terminals of DNA. Then, they were incubated in peroxidase-conjugated antibody solution and were stained with DAB. The sections were counterstained with methyl green, and the percentage of TUNEL-positive cells was quantified by using HistoQuest software (NOVEL SCIENCE).

The statistical significance of differences was evaluated by unpaired ANOVAs, and probability values were calculated with the Student t test. Survival rates were analyzed by the generalized Wilcoxon and log-rank tests. P < 0.05 was considered statistically significant.

To test the possible involvement of VASH2 in the tumor, we analyzed human ovarian cancers for its presence. Immunohistochemical analysis of the pathologic sections revealed that VASH2 protein was preferentially detected in serous ovarian adenocarcinoma cells (Fig. 1A). Indeed, cancer cells in 8 of 12 cases of serous ovarian adenocarcinoma (67%) were positive for VASH2, whereas cancer cells in all of the cases of non–serous ovarian adenocarcinoma; that is, 3 cases of clear cell carcinoma, 3 cases of mucinous adenocarcinoma, and 3 cases of endometrioid carcinoma, were negative for it (Fig. 1B).

To verify the function of VASH2 in ovarian cancers, we conducted a loss-of-function experiment by knocking down the expression of VASH2. We used SKOV-3 and DISS for the following experiments, as they are representative ovarian cancer cells that are highly tumorigenic in animals. At least, 7 variant transcripts of human VASH2 are registered in the database (Supplementary Fig. S1). We designed shRNAs to knock down most of the splicing variants and established 2 VASH2 knocked-down (sh-VASH2) clonal cell lines from each of SKOV-3 and DISS (Fig. 2A). The knockdown of VASH2 did not alter the in vitro proliferation of SKOV-3 but slightly decreased that of DISS (Fig. 2B). We then inoculated either of these VASH2-knocked down cancer cells subcutaneously into nude mice and observed a significant reduction in tumor growth in terms of both tumor volume and weight (Fig. 2C and D).

Peritoneal dissemination often occurs in serous adenocarcinoma of the ovary and is a sign of poor prognosis. We injected cancer cells into the peritoneal cavity as a model of peritoneal dissemination and observed that there was a significant decrease in the number of disseminated tumors in mice injected with SKOV-3 or DISS cells that had been transfected with sh-VASH2 (Fig. 3A and C). In addition, DISS cells caused the accumulation of a bloody ascites, but it was almost completely abrogated by sh-VASH2 (Fig. 3E). As the result, the survival period was prolonged in mice that had been injected with either SKOV-3 or DISS sh-VASH2 cells (Fig. 3B and D).

We further examined an orthotopic mouse model. Again we observed a significant reduction in tumor growth of SKOV-3 or DISS cells that had been transfected with sh-VASH2 (Fig. 4).

We previously reported that VASH1 inhibits angiogenesis whereas VASH2 promotes it (16). We therefore examined the vasculature in the tumors and found a significant decrease in angiogenesis in tumors derived from either SKOV-3 or DISS sh-VASH2 cells (Fig. 5A and C). The decrease of tumor angiogenesis resulted in the significant increase of cancer cell apoptosis but no changes in cancer cell proliferation in vivo (Supplementary Fig. S2). We further investigated the vascular composition of endothelial cells and mural cells. The association of mural cells with endothelial cells was not significantly altered in either type of tumor (Fig. 5B and D). We did not observe any differences in the extent of mural cell coverage in tumor vessels (Fig. 5B).

To further confirm proangiogenic function of VASH2 in tumors, we carried out a gain-of-function experiment by introducing the VASH2 gene in cancer cells. VASH2 can be expressed in bone marrow–derived CD11b⁺ mononuclear cells (16). To avoid the possible involvement of recipient VASH2, we planned to use VASH2 (−/−) mice. As our VASH2 (−/−) mice were on a C57BL6 background, we examined various tumorigenic C57BL6 murine tumor cells and found that EL-4 and MLTC-1 did not express endogenous VASH2 (Supplementary Fig. S2). We introduced the human VASH2 gene into EL-4 and established stable clones (Fig. 6A). The introduction of the VASH2 gene slightly but significantly inhibited the in vitro proliferation of these tumor cells (Fig. 6B). We then inoculated these cells subcutaneously into VASH2 (−/−) mice. Parental EL-4 cells are tumorigenic, indicating that they have sufficient angiogenic activity despite of the lack of VASH2 expression. Even though the VASH2 gene slightly decreased the proliferation of tumor cells, we observed a significant intensification of tumor growth in VASH2 transfectants (Fig. 6C). Then, we examined the vasculature in the tumors. As expected, tumors of mock transfectants contained a certain quantity of tumor vessels. Nevertheless, we observed that there was a significant increase in tumor angiogenesis, as evidenced by the increased vascular luminal area, in the VASH2 transfectants (Fig. 6D). We further observed that the association of mural cell with endothelial cells was not significantly altered in the EL-4 transfectants (Fig. 6E).

VASH1 exhibits its antiangiogenic activity by inhibiting migration and proliferation of endothelial cells (12). As a homolog of VASH1, VASH2 may exhibit its proangiogenic effect by acting on endothelial cells as well. Here, we showed that CM from VASH2 transfectants stimulated the migration but not the proliferation of endothelial cells when compared with CM from mock transfectants (Fig. 7A). Moreover, those VASH2 transfectants as well as 2 human ovarian cancer cell lines SKOV-3 and DISS expressed SVBP (15) that plays an essential role in vasohibin secretion (Supplementary Fig. S3). These results suggest that VASH2 is secreted from cancer cells and affects on endothelial cells as a paracrine manner.

We intended to verify the mechanism of expression of VASH2 in cancer cells. We first confirmed that SKOV-3 and DISS expressed more VASH2 mRNA than HOECs (Fig. 7B). This increased expression of VASH2 in serous ovarian adenocarcinoma cells was constitutive, as no stimulus, including hypoxia, induced its expression (data not shown). To understand the mechanism of this sustained increase in VASH2 expression, we examined the involvement of microRNAs because a database for microRNA targets prediction and functional annotations (http://mirdb.org/miRDB/)/indicates VASH2 to have the highest rank as a target of miR-200b. Indeed, there are multiple binding sites of miR-200bc/429 in the 3′-untranslated region of human VASH2 mRNA (Fig. 7C). As expected, the expression of miR-200b was apparently low in SKOV-3 and DISS (Fig. 1B). Pre-miR-200b decreased the expression of VASH2 in SKOV-3 cells (Fig. 7D). Moreover, we observed an inverse correlation between VASH2 expression and miR-200b expression in human ovarian cancer tissue (Fig. 7E). These results suggest that the decreased expression of miR-200b was responsible for the upregulation of VASH2 in serous ovarian adenocarcinoma cells.

Here, we showed for the first time that VASH2 was preferentially expressed in serous adenocarcinoma among human ovarian cancers. Moreover, specific knockdown of VASH2 from cell lines of human serous adenocarcinoma remarkably attenuated the growth of inoculated tumor cells and their peritoneal dissemination as well as tumor angiogenesis. In contrast, transfection of VASH2-negative tumor cells with the VASH2 gene augmented tumor angiogenesis and tumor growth when inoculated into mice. Collectively, our data suggest that VASH2 was responsible for promoting tumor angiogenesis in serous ovarian adenocarcinoma. Whereas several splicing variants of VASH2 are registered in the database, the significance and functional differences among those variants are presently obscure.

Serous adenocarcinoma is the most common histologic subtype of ovarian cancers. As mentioned earlier, because recurrence after the first-line chemotherapy is frequent, several targets have been considered for the treatment of ovarian cancers and one of them is angiogenesis (4). Antiangiogenic therapy is now approved for several cancers ranging from colon, lung, breast, and kidney; and drugs targeting VEGF signals are in clinical use (30). However, the benefit of such drugs may not last long, as patients will encounter progression of cancers because of the compensatory production of angiogenic factors other than VEGF or recruitment of bone marrow–derived angiogenic cells. Therefore, alternative targets for antiangiogenic therapy are now extensively being investigated (31). Here, we propose that VASH2 can be a candidate target for the treatment of serous adenocarcinoma in light of its stimulating effect on angiogenesis.

We have reported the proangiogenic activity of VASH2 (16), but its precise function has been unclear. Here, we showed that CM from human VASH2 transfectants stimulated the migration of endothelial cells. Ovarian cancer cells express SVBP, a secretory chaperone of vasohibins (15). These results suggest that VASH2 is secreted from the cancer cells and acts on neighboring endothelial cells to stimulate angiogenesis in a paracrine manner. VASH1 inhibits angiogenesis, whereas VASH2 stimulates it. Most plausible mechanism is that these 2 factors share a putative vasohibin receptor and one acts as an agonist whereas the other acts as an antagonist. This hypothesis is currently under investigation.

In terms of the expression of VASH2, we showed that miR-200b repressed the expression of VASH2 in ovarian cancer cells. microRNAs represent a category of small noncoding RNAs that are involved in the regulation of gene expression by translational repression and/or degradation of target mRNAs (32). The expression of VASH2 could not be induced by any of the stimuli tested (data not shown). Moreover, when murine embryonic stem cells form embryoid bodies, the expression of VASH1 is initially low but steeply induced when vessels are formed, whereas that of VASH2 is initially high in embryonic stem cells and gradually decreases during their differentiation (Abe and Sato, unpublished observations). We think that this pattern of VASH2 expression fits with the microRNA-mediated gene repression. miR-200b belongs to the miR-200 family of microRNAs. This family comprises 5 members (miR-200a, 200b, 200c, 141, and 429), and they are downregulated in cancer cells due to aberrant epigenetic gene silencing and play a critical role in the suppression of epithelial-to-mesenchymal transition (EMT) by targeting and repressing the expression of key molecules such as ZEB1 and ZEB that are involved in EMT (33). It is reported that miR-200 family members are frequently dysregulated in human ovarian cancers (34–37). In this context, it would be interesting to see whether or not VASH2 is involved in the EMT of cancer cells besides its role in angiogenesis.

In summary, VASH2 was expressed in certain ovarian cancers, and it promoted tumor growth and peritoneal dissemination of tumor cells by stimulating tumor angiogenesis. Knockdown of VASH2 significantly attenuated the tumor growth and peritoneal dissemination, which indicates that VASH2 would be a molecular target for the treatment of ovarian cancer. It is important to see whether this role of VASH2 is played in cancers from other organs.

No potential conflicts of interest were disclosed.

Conception and design: Y. Takahashi, Y. Sato

Development of methodology: Y. Takahashi, T. Moriya

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Takahashi, T. Koyanagi, N. Kanomata, T. Moriya

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y. Takahashi, Y. Suzuki, T. Moriya

Writing, review, and/or revision of the manuscript: Y. Takahashi, T. Koyanagi, Y. Sato

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y. Saga, M. Suzuki

Study supervision: Y. Sato

This work was supported by Grant-in-Aid for Scientific Research on Innovative Areas “Integrative Research on Cancer Microenvironment Network” (22112006) and Grants-in-Aid for Scientific Research (C) (22590821) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan and a Health and Labour Sciences research grant, Third Term Comprehensive Control Research for Cancer, from the Ministry of Health, Labour, and Welfare of Japan (Y. Sato), and by the research Award to JMU Student (Y. Takahashi).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.