- Email: [email protected]
VEGF induces ascites in ovarian cancer patients via increasing peritoneal permeability by downregulation of Claudin 5
Gynecologic Oncology 127 (2012) 210–216
Contents lists available at SciVerse ScienceDirect
Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno
VEGF induces ascites in ovarian cancer patients via increasing peritoneal permeability by downregulation of Claudin 5 Daniel Herr, Alexandra Sallmann, Inga Bekes, Regina Konrad, Iris Holzheu, Rolf Kreienberg, Christine Wulff ⁎ Department of Obstetrics and Gynecology, Ulm University Medical Centre, Ulm, Germany
a r t i c l e
i n f o
Article history: Received 3 April 2012 Accepted 1 May 2012 Available online 8 May 2012 Keywords: Ovarian cancer Ascites Claudin 5 Permeability
a b s t r a c t Objective. To evaluate the role of VEGF-dependent Claudin 5 production for the development of ascites via influencing endothelial permeability in peritoneal tissue of ovarian cancer patients. Methods. This study investigates the mechanisms of formation of ascites in ovarian cancer patients, performing RT-PCR, VEGF-ELISA and immunohistochemical dual staining for CD31 and Claudin 5. In addition, in order to analyze the connectivity of VEGF, Claudin 5, and endothelial permeability, an endothelial cell/ ovarian cancer cell-co-culture-system was established and evaluated performing Western blot analysis and a permeability assay. Results. Firstly, VEGF-gene expression was demonstrated for all ovarian cancer and peritoneal biopsies. In addition, quantification of VEGF in the serum and ascites of ovarian cancer patients revealed significantly increased values. We subsequently demonstrated Claudin 5 production in the peritoneal vessels, which was weaker than in the vessels of the controls. Evaluation of endothelial permeability finally showed a VEGF-dependent regulation via Claudin 5 suggesting a mechanism for the development of ascites in ovarian cancer patients. Conclusion. VEGF induces ascites in ovarian cancer patients. This instance happens due to increased peritoneal permeability, caused by downregulation of the tight junction protein Claudin 5 in the peritoneal endothelium. © 2012 Elsevier Inc. All rights reserved.
Introduction Worldwide, ovarian cancer occurs in about 200,000 women per year. Concerning mortality, ovarian cancer ranges around place 5 of malign disorders. The five-year overall survival rate is only about 40%. Unfortunately, in most cases ovarian cancer is diagnosed in advanced tumor stages. Thereby, one of the most impressing clinical symptoms is an increase of the abdominal circumference due to ascites [1-3]. Development of ascites is a consequence of dysregulated endothelial function leading to increased vascular permeability of peritoneal vessels [4]. Under physiological conditions, vascular permeability is mediated by strictly regulated opening and closure of cell–cell-junctions [5-7]. For that reason, any disturbance of junctional organization might result in dysregulated vascular function leading to pathological conditions by modifying the regular structure of the vessel wall [8]. Under malign circumstances, the typical example with clinical importance caused by increased vascular permeability is ascites. In the endothelium, paracellular permeability involves at least two different types of intercellular junctions; adherens junctions (AJ) and tight junctions (TJ). Those are formed by different transmembrane ⁎ Corresponding author at: Department of Obstetrics and Gynecology, University of Ulm, Prittwitzstrasse 43, 89075 Ulm, Germany. Fax: + 49 73150058502. E-mail address: [email protected] (C. Wulff). 0090-8258/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2012.05.002
proteins that promote homophilic cell–cell-interactions and transfer of intracellular signals [9]. Many reports support the concept, that intercellular junctions are dynamically remodeled not only during embryogenesis but also in resting cells [10]. Adhesive membrane proteins of AJ and TJ form adhesive complexes which act as zipper-like structures between interacting cells [11-15]. Endothelial cells express tissue-specific transmembrane adhesion proteins: the AJ VE-(vascular endothelial)-Cadherin and the TJ Claudin 5 [15]. Knockdown of Claudin 5 in mice is associated with a normal embryological development but, due to a defective blood brain barrier function, the Claudin 5 deficient mice die shortly after birth [15]. For the junctional protein Claudin 5 paracrine regulatory mechanisms have been shown in the human ovary. We demonstrated recently that Claudin 5 is expressed in the luteal vasculature and that the expression was decreased after in vivo treatment with hCG [16]. This effect may be secondary to vascular endothelial growth factor (VEGF) as hCG induces VEGF expression in luteal cells in vitro [17,18] and in vivo [19]. Indeed VEGF has been shown to be able to enhance vascular permeability [20,21] and influence endothelial adherens junctions [22]. In addition, it has been shown in an in vitro corpus luteum model, that downregulation of Claudin 5 in endothelial cells is associated with increased vascular permeability [23] demonstrating the importance of this junctional protein for regulation of vascular permeability.
D. Herr et al. / Gynecologic Oncology 127 (2012) 210–216
Concerning malign tumors VEGF, which is produced by cancer cells itself, induces angiogenesis in order to insure supply with oxygen and nutritive substances [24], which has also been shown for ovarian cancer cells [25,26]. Malign ovarian cysts induce VEGF-level 24-fold higher than in benign ovarian cysts [27]. Apparently, these high levels of VEGF increase local permeability causing extravasation of fluids, thus ascites. Here, we address the question of whether or not VEGF-dependent production of Claudin 5 acts as a regulator of vascular permeability in ovarian cancer patients. We hypothesized that VEGF produced by ovarian cancer cells decreases the amount of Claudin 5 in peritoneal vessels causing ascites by increasing endothelial permeability. We therefore went on to study Claudin 5 in peritoneal tissue as well as VEGF in serum and ascites of ovarian cancer patients. In addition, in order to elucidate the causality of a possible impact of VEGF on Claudin 5 and vascular permeability, we established a co-culturesystem of ovarian cancer- and endothelial-cells. Material and methods Patients Tumor- and peritoneal-tissue, serum and ascites were collected from patients undergoing laparatomy for advanced serous papillary ovarian cancer (FIGO stage III). As controls, we used tissue of patients with leiomyoma and uterine prolapse. In summary, 10 patients with ovarian cancer and 10 controls have been analyzed. This was institutionally approved after favorable ethical review and written consent of the patients. Morphological characterization of ovarian cancer Consecutive sections stained for hematoxylin and eosin were used to classify the ovarian cancer and peritoneal tissue. The embedded ovaries were serially sectioned, and tissue sections (4 μm) were placed onto BDH SuperFrost slides (BDH, Merck & Co., Inc., Poole, UK). Tissue sections were dewaxed in xylene, rehydrated in descending concentrations of ethanol, washed in distilled water, and stained with hematoxylin (Richard-Allan, Richland, MI) for 5 min, followed by a wash in water and acetic alcohol before staining with eosin (Richard-Allan) for 20 s. After dehydrating in ascending concentrations of ethanol and xylene, sections were mounted. Quantification of VEGF by ELISA (serum and ascites) In order to quantify the secreted amount of VEGF, a quantitative VEGF immunoassay was performed according to the manufacturer's protocol (R&D Systems, Minneapolis, USA. The optical density was measured at 450 nm (Dynatech Laboratory, Bath, UK). Immunohistochemistry dual staining Immunofluorescence double-staining was performed using the TSA-Kit (Perkin Elmer, Boston, USA) according to the manufacturer's instructions. The slides were incubated with mouse anti-human Claudin 5 (Zymed 18–7364, South San Francisco, USA, 1:100 dilution) overnight at 4 °C. CD 31 staining was performed to prove the colocalization to the endothelial compartment using a mouse antihuman CD 31 antibody (Dako M0823, Dako, Hamburg, Germany,1:30 dilution). Cell-culture for HUVEC and OvCar-3 Endothelial cell isolation and culture HUVEC (human umbilical vein endothelial cells) were isolated from multiple segments of normal term umbilical cords, pooled and
211
cultured in the endothelial cell growth medium (Promocell GmbH, Heidelberg, Germany). Cell culture for OvCar-3 cells According to the instructions of the American Type Culture Collection/ The Global Bioresource Center (ATCC), the OvCar-3 cells were cultivated in RPMI 1640 (PAA Laboratories, 3-Pasching, Austria) with 1% penicillin/ streptomycin and 10% FCS (PAA Laboratories, 3-Pasching, Austria). Media were changed every 48 h. RNA-isolation Total RNA from HUVEC, OvCar-3-cells, and ovarian cancer tissue was extracted from cells with the RNeasy kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. The RNA product was quantified by absorbance at 260 nm and total RNA (2.5 μg) was reverse transcribed into cDNA using random hexamer primers (Applied Biosystems, Foster City, USA). RT-PCR-analysis PCR amplification was carried out with 0.2 mM dNTP, 10 mM forward primer, 10 mM reverse primer, 1.5 mM MgCl2 and 1 U Taq polymerase (Qiagen) in the recommended buffer. The cycling conditions were 45 cycles at 94 °C for 45 s, 59 °C for 45 s and 72 °C for 60 s after the initial denaturation step. The primers had the following sequences: VEGF (367 bp) (forward: 5′-CGG GCC TCC GAA ACC ATG AAC TTT-3′; primer: 5′-CTA TGT GCT GGC CCT GGT GAG GTT GAT3′), Claudin-5 (462 bp) (forward: 5′-ACC GGC GAC TAC GAC AAG AAG A-3′; reverse: 5′-GCC CTG CCG ATG GAG TAA AGA-3′), and GAPDH (657 bp) (forward: 5′-CTG GCG CTG AGT ACG TCG-3′; reverse: 5′-TTG ACA AAG TGG TCG TTG A-3′. Bands were visualized after electrophoresis on a 2% agarose gel (Invitrogen) by staining with ethidium bromide. Pictures were taken with a camera system of Alpha Corporation (San Leandro, CA, USA). Co-culture of OvCar-3-cells and HUVEC HUVECs (2 × 10 5 per well) were seeded onto 6-well plates and grown to 90–100% confluency. On day 2, OvCar-3 cells were transferred to cell culture inserts (0.4 μm pore size, Becton Dickinson Nr. 353090). On day 3, cell culture inserts with OvCar-3-cells were inserted into the 6-well plates with HUVECs in culture. Endothelial cells were treated with 200 ng and OvCar-3 cells with 300 ng Flt1Fc. In this setting, molecules secreted by OvCar-3-cells follow the concentration gradient between the two different media such that the HUVEC can be stimulated. Extraction of membrane protein To obtain membrane fractions, cells were harvested into PBS. After centrifugation (5 min, 1200 rpm), the pellet was resuspended in 200 μl detergent-free lysis buffer (20 mM TrisHCl, pH 7.5, 0.5 mM EDTA, 10 mM BME, proteinase- and phosphatase-inhibitor cocktail (Sigma, Aldrich, Germany)). Cells were disrupted in a glass dounce homogenizer, incubated on ice, and ultracentrifuged at 35,000 rpm for 60 min at 4 C. The supernatant was referred to as the cytosolic fraction. The remaining pellet was resuspended in lysis buffer. Lysates were centrifuged at 15,000 rpm for 15 min and the supernatant was reserved as the membrane fraction. Protein concentration was determined using the BCA Protein Assay (Pierce, Rockford, IL, USA). Protein extraction from peritoneal tissue For protein extraction, peritoneal tissue was homogenized in 2 ml lysis buffer on ice using Ultraturrax (Janke & Kunkel GmbH, Staufen, Germany). Afterwards, the tissue was incubated on ice for 1 h followed by centrifugation at 1200 rpm for 10 min at 4 °C. The supernatant was extracted and centrifuged at 35,000 rpm at 4 °C for 1 h. The supernatant was referred to as the cytosolic fraction. The pellet containing the membranous fraction was resuspended and lysed in
D. Herr et al. / Gynecologic Oncology 127 (2012) 210–216
Permeability assay To test the permeability of endothelial cells for macromolecules, the HUVEC were seeded in the insert with the OvCar-3-cells underneath (supplemental data) to allow the flow of the dye through the membrane. HUVEC were seeded on micropore inserts (pore size 0.4 μm) at a density of 2 × 10 5 cells per insert and cultured for 2 days. On day 2, 1.75 × 10 5 OvCar-3-cells were seeded into 4 wells and cultured in medium for 1 day. The inserts were then moved into the wells containing the OvCar-3-cells . In this co-culturesystem, OvCar-3-cells and HUVEC were treated with Flt-1Fc (OvCar3: 300 ng/ml; HUVEC: 200 ng/ml). Bovine serum albumin (BSA) was labeled with Trypan Blue (66.7 mg Trypan Blue to 1.6 g BSA and 40 ml PBS), mixed with culture medium 1:1, and added to the insert, whereas unlabeled albumin was added to the lower compartment. Samples of the lower wells were taken after 10 and 30 min and protein concentration was measured at 607 nm. Each experiment was repeated on three to six separate occasions. Statistics Statistical analysis was performed using SPSS for Windows version 6.2. Data obtained from experiments was compared to controls and was tested for significant differences using a two tailed, unpaired t-test. Data are shown as mean ± SEM. Statistical analyses for the permeability assay were tested using ANOVA, followed by Duncan's multiple range test. Differences were considered to be significant at p b 0.05.
1000 800
VEGF
Western blotting Proteins were separated by SDS-PAGE in 4–20% Tris-glycine gels (Lonza, Rockland, USA) and transferred onto 0.45 μm nitrocellulose membranes (Schleicher Schuell, Dassel, Germany). Membranes were incubated with the primary antibody at 4 °C overnight (anti-Claudin 5 antibody at 1:500 dilution, Zytomed, Berlin, Germany; or GAPDH antibody at 1:500 dilution, Abcam, Cambridge, UK). Detection involved incubation with peroxidase-conjugated secondary antibody (Claudin 5: 1:1000, and GAPDH: 1:1000; HRP-labled sheep-anti mouse, GE Healthcare UK Limited, Little Chalfont, Buckinghamshire, UK) at room temperature for 1 h. Protein-antibody complexes were visualized using the ECL Western blot detection reagents (GE Healthcare UK Limited, Little Chalfont, Buckinghamshire, UK). Optical band density was quantified (Imager of Alpha Corporation, San Leandro, Ca., USA) and the results obtained for Claudin 5 were normalized to the values for GAPDH.
a 1200
600 400 200 0 control
b
ovarian cancer
2000 1500
VEGF
TX-100 buffer and again ultracentrifuged at 15,000 rpm at 4 °C for 15 min. The protein concentration was calculated using the BCA Protein Assay (Pierce, Rockford, IL, USA) according to the instructions of the manufacturer.
1000 500 0
c
3500 3000 2500
VEGF
212
2000 1500 1000 500 0
Fig. 1. VEGF level in pg/ml (± SEM) in serum and ascites of ovarian cancer patients. (a) VEGF in serum of ovarian cancer patients and controls at day 0 after surgery; the difference is significant (p b 0.05; n = 10). (b) VEGF in serum of ovarian cancer patients at days 0, 2, and 4 after surgery; the difference between days 2 and 4 is significant (p b 0.05; n = 10). (c) VEGF in ascites of ovarian cancer patients at days 0, 2, and 4 after surgery; the difference between days 0 and 2 is significant (p b 0.05; n = 10).
VEGF-concentration in ascites Since we supposed, that the tumor itself might be the main source of VEGF, the concentration of VEGF was measured in the ascites at days 0, 2, and 4 after surgery. As expected, we revealed a decrease of VEGF, which was significant between days 0 and 2 (Fig. 1c).
Results Ovarian cancer biopsies
Claudin 5 in the peritoneal vessels
In order to verify the histological diagnosis of ovarian cancer, hematoxylin-eosin staining was performed for all ovarian and peritoneal tissue samples, proving serous papillary ovarian cancer tissue in all cases. In addition, VEGF-gene expression was shown for all tissue samples (supplemental data).
In order to analyze the existence of Claudin 5 in the vessels of the peritoneal tissue, double-staining of Claudin 5 with CD 31 was performed, revealing co-localization in the endothelial cells of both, controls and ovarian cancer patients. In addition, it has been shown, that the proportion of Claudin 5 in the vessels of ovarian cancer patients is considerably weaker as in the vessels of the controls (Fig. 2). In order to further investigate the amount of Claudin 5 in the endothelial cell compartment, Western blot analysis confirmed the production of the Claudin 5 protein. Quantification of Claudin 5 thereby revealed a significant reduction of Claudin 5 in the peritoneal vessels of ovarian cancer patients as compared to the controls (Fig. 3).
VEGF-concentration in serum Measurement of VEGF in the serum of ovarian cancer patients preoperatively revealed significant higher values as compared with healthy controls (Fig. 1a). The analysis of VEGF in the serum at day 2 revealed no significant changes. However, at day 4 after surgery a significant increase of VEGF could be observed (Fig. 1b).
D. Herr et al. / Gynecologic Oncology 127 (2012) 210–216
213
control
ovarian cancer
CD 31
Claudin 5
Co-localisation
Fig. 2. Representative immunocytochemical staining of Claudin 5, CD 31, and their co-localization in peritoneal tissue (a–c control, d–f ovarian cancer) proving the occurrence of Claudin 5 in the endothelial cell compartment. Note the light Claudin 5-staining in f suggesting a reduction of Claudin 5 in the peritoneum of ovarian cancer patients as compared to controls.
Analysis in OvCar-3 cells and HUVEC in the co-culture-system First of all gene expression as well as protein production of VEGF in the ovarian cancer cell-line Ovcar-3 cells was demonstrated. We found a time-dependent increase of the VEGF-protein amount in the period from 5 to 120 h (Fig. 4 a,b). In addition, analysis of Claudin 5 in HUVEC revealed its gene expression and protein production (Fig. 4 c,d). Claudin 5 and permeability in the endothelial-ovarian-cancer-cell co-culture-system
production of Claudin 5 in HUVEC grown in the co-culture-system in vitro. We found a significant decrease of Claudin 5 in HUVEC cocultured with OvCar-3 cells, which was prevented by simultaneous treatment with the VEGF-inhibitor Flt1-Fc (Fig. 5). In addition, the possible influence of changed Claudin 5-production on endothelial permeability was investigated in the co-culture-system. We observed a significant increase of permeability in the co-cultured HUVEC as compared to the controls. Furthermore, additional treatment with Flt1-Fc again antagonized this effect and caused a significant decrease of permeability reaching again the level of the control (Fig. 6). Discussion
Since it has been shown in vivo, that Claudin 5 is reduced in the peritoneal vessels of ovarian cancer patients, we measured the
a GAPDH
Claudin 5 control
cancer
b
control
cancer
Fig. 3. a) Representative Western blot of Claudin 5 (19 kDa) in peritoneal tissue of controls and ovarian cancer patients with corresponding production of GAPDH (37 kDa). b) Densitometric quantification of Claudin 5 normalized to GAPDH. The observed difference is significant (p b 0.05; n = 6).
The current study was performed in order to investigate the molecular mechanisms of formation of ascites in patients with ovarian cancer. In this regard, we showed increased levels of VEGF in the serum and ascites of preoperative ovarian cancer patients. In addition, a decreased amount of the tight junction protein Claudin 5 was detected in the peritoneal vessels of ovarian cancer patients compared to healthy controls. We therefore went on to study a possible functional association between Claudin 5 and increased peritoneal permeability as morphological correlation of ascites and established a co-culture-system of ovarian cancer- and endothelial-cells. In this way, we revealed a VEGF-dependent decrease of Claudin 5 in endothelial cells co-cultured with ovarian cancer cells. The observed changes of this junctional protein did not only remain on a structural level, but could also be translated into a significant increase of endothelial permeability. It is likely that this novel paradigm of interaction of VEGF and Claudin 5 in the endothelial compartment will facilitate the explanation of the molecular regulation of altered peritoneal permeability causing ascites in ovarian cancer patients. In order to investigate this connectivity, we firstly focused on protein production of VEGF proving its presence in serous papillary ovarian cancer as well as in peritoneal tissue. This result is in line with that of others [25-27]. In addition, quantification of VEGF-levels in the serum of ovarian cancer patients revealed significantly increased values as compared to healthy controls, indicating a possible functional role for VEGF concerning formation of ascites by increasing endothelial permeability. In contrast to the majority of other tumor entities, metastases of ovarian cancer mainly occur intra‐abdominally and metastasize only exceptionally via blood vessels [28]. Therefore, VEGF facilitates cancer cells to disseminate in the abdominal cavity, which might provoke these cells to accumulate, thus again increasing
214
D. Herr et al. / Gynecologic Oncology 127 (2012) 210–216
b
h
a
1
2
3
140 120 100 80 60 40 20 0 0 1000 2000 3000 4000 5000 6000 7000 8000 9000
VEGF in pg/ml
d
c
GAPDH
1
2
3 Claudin 5
Fig. 4. a) RT-PCR amplification of (1) GAPDH (657 bp), (2) negative control, and (3) VEGF (367 bp) in OvCar-3 cells. b) Production of the VEGF-protein (ELISA) in Ovcar-3 cells revealing a significant time-dependent increase (+ standard deviation) in the period from 5 to 120 h (n = 3). c) RT-PCR amplification of (1) GAPDH (657 bp), (2) negative control, and (3) Claudin 5 (462 bp) in HUVEC. d) Western blot of Claudin 5 (19 kDa) and GAPDH (37 kDa) in HUVEC.
the production of VEGF, again with its monitored effects on peritoneal permeability. Such interrelation has been already supposed for different tumor entities such as breast cancer, colorectal cancer and ovarian cancer [28,29]. In addition, the VEGF-levels in the serum depend on the cumulative amount of tumor tissue. Therefore, they were measured on days 2 and 4 after surgery. At day 2 no changes had been observed, however at day 4 a significant increase of VEGF was found. This increase might be explained by the beginning process of wound healing tissue regeneration, since several authors described increased levels of VEGF in this context [30–32]. It is conspicuous, that the systemically increased serum-levels of VEGF only influence the peritoneal
a GAPDH
Claudin 5
control
Claudin-5/GAPDH
b
co-culture
co-culture + Flt1-Fc
12000 10000 8000 6000 4000 2000 0 control
co-culture
co-culture + Flt1-Fc
Fig. 5. Representative Western blot (a) of Claudin 5 and GAPDH in HUVEC co-cultured with OvCar-3 cells ± Flt1-Fc. Densitometric quantification (b) reveals a significant reduced amount of Claudin 5 in co-cultured cells. Treatment with Flt1-Fc antagonizes this effect increasing Claudin 5 again significantly. Values are normalized to GAPDH (p b 0.05; n = 6).
vessels but obviously not the complete periphery. It has been shown by Monsky et al., that there are huge differences of the amount of VEGF needed to increase endothelial permeability in different areas of the human body [33]. For example, increase of intracranial permeability requires a bigger amount of VEGF compared to vessels of the subcutaneous tissue [33]. Therefore, it can be supposed, that the observed VEGF-concentration in the ovarian cancer patients is too small in order to influence the periphery but is sufficient to alter the integrity of the peritoneal vessels. According to the consideration of tumor-amount dependent levels of VEGF, we also investigated VEGF-levels in the ascites, proving a significant postoperative decline at day 2 after surgery. Since its has been shown, that ascites with high concentrations of VEGF which has been injected into mice resulted in increased rates of peritoneal permeability [34], our result suggests, that ovarian cancer cells produce VEGF in order to increase the peritoneal permeability as a base for the formation of ascites which allows cancer cells to grow and to disperse in the visceral cavity. Since VEGF levels in ascites are much higher as compared to serum levels it is likely to hypothesize that cancer cell derived VEFG secreted into the ascites acts locally on the peritoneal vessels increasing their permeability. This also explains the observation that hyperpermeability in ovarian cancer patients is mostly restricted to peritoneal cavity. Initially, we had hypothesized that VEGF, produced by ovarian cancer cells, influences peritoneal permeability via regulation of the tight junction protein Claudin 5. Therefore, we first demonstrated production of Claudin 5 in the peritoneal vessels. In addition, quantification of the Claudin 5 proportion in the peritoneum of ovarian cancer patient revealed significant lower values as compared to healthy patients as controls. It has been shown in a physiologic in vitro model of the human corpus luteum, that VEGF-dependent downregulated levels of Claudin 5 cause increased endothelial permeability [23]. Therefore, we now assumed a functional role of VEGF-dependent production of Claudin 5 concerning regulation of endothelial permeability in the peritoneal tissue of ovarian cancer patients. In order to investigate this presumption, a new co-culture-system of ovarian cancer- and endothelial cells was established using OvCar3 cells and HUVEC which produce VEGF and Claudin 5, respectively. As expected, due to VEGF produced by OvCar-3 cells, endothelial cells co-cultured with ovarian cancer cells presented with significantly lower levels of Claudin 5. This result validated the functionality of
D. Herr et al. / Gynecologic Oncology 127 (2012) 210–216
permeability/permeability of control
a
References 1,8 1,6 1,4 1,2 1 0,8 0,6 0,4 0,2 0
permeability/permeability of control
b
215
1,6 1,4 1,2 1 0,8 0,6 0,4 0,2 0
Fig. 6. Relative permeability in the HUVEC-Ovcar-3-cell co-culture-system after 10 min (a) and 30 min (b). The black bar represents the control (HUVEC alone), the white bar the HUVEC-Ovcar-3 cells-co-culture, and the grey bar the co-culture with simultaneous treatment of both cell compartments with the VEGF-antagonist Flt1-Fc. Note the significant increase (after 10 and 30 min) of permeability in the co-culture (white) and the significant decrease after inhibition of VEGF (grey). The values are normalised to the permeability of the control (p b 0.05; n = 12).
our in vitro model-system, since in vivo we had observed the same effects as described above. We now finally focused on endothelial permeability and found significantly increased rates of permeability after VEGF-induced downregulation of Claudin 5. These findings make it possible that ovarian cancer cells produce VEGF in order to increase endothelial permeability via Claudin 5 which finally induces the formation of ascites and thereby dissemination of cancer cells in the abdominal cavity. Taken together, our results indicate an important role of VEGF as regulator of endothelial permeability via Claudin 5. In addition, the data provides an explanation of the mechanisms of the intraperitoneal treatment of ascites with the VEGF-antagonist bevacizumab in ovarian cancer patients, being a promising approach for symptomatic therapy in clinical studies [35]. For this reason, our data suggest a new pathophysiological model, explaining the molecular connectivity of the development of ascites in ovarian cancer patients, having possible clinical implications on the symptomatic treatment of patients in palliation.
Conflict of interest The authors declare that there are no conflicts of interest.
Appendix A. Supplementary data Supplementary data to this article can be found online at http:// dx.doi.org/10.1016/j.ygyno.2012.05.002.
[1] Ayantunde AA, Parsons SL. Pattern and prognostic factors in patients with malignant ascites: a retrospective study. Ann Oncol 2007;18:945–9. [2] Ayhan A, Guven S, Guven ES, Kucukali T. Is there a correlation between tumor marker panel and tumor size and histopathology in well staged patients with borderline ovarian tumors? Acta Obstet Gynecol Scand 2007;86:484–90. [3] Cheng MH, Yen MS, Chao KC, Sheu BC, Wang PH. Differential diagnosis of gynecologic organ-related diseases in women presenting with ascites. Taiwan J Obstet Gynecol 2008;47:384–90. [4] Garrison RN, Kaelin LD, Galloway RH, Heuser LS. Malignant ascites. Clinical and experimental observations. Ann Surg 1986;203:644–51. [5] Dejana E. Endothelial cell–cell junctions: happy together. Nat Rev Mol Cell Biol 2004;5:261–70. [6] Bazzoni G, Dejana E. Endothelial cell-to-cell junctions: molecular organization and role in vascular homeostasis. Physiol Rev 2004;84:869–901. [7] Walz A, Keck C, Weber H, Kissel C, Pietrowski D. Effects of luteinizing hormone and human chorionic gonadotropin on corpus luteum cells in a spheroid cell culture system. Mol Reprod Dev 2005;72:98–104. [8] Herr D, Fraser HM, Konrad R, Holzheu I, Kreienberg R, Wulff C. HCG Controls luteal Vascular Permeability via VEGF by Down Regulation of a Cascade of Adhesion Proteins. Fert Ster submitted for publication. [9] Dejana E, Orsenigo F, Molendini C, Baluk P, McDonald DM. Organization and signaling of endothelial cell-to-cell junctions in various regions of the blood and lymphatic vascular trees. Cell Tissue Res 2009;335:17–25. [10] Dejana E, Tournier-Lasserve E, Weinstein BM. The control of vascular integrity by endothelial cell junctions: molecular basis and pathological implications. Dev Cell 2009;16:209–21. [11] Cavey M, Rauzi M, Lenne PF, Lecuit T. A two-tiered mechanism for stabilization and immobilization of E-cadherin. Nature 2008;453:751–6. [12] Chitaev NA, Troyanovsky SM. Adhesive but not lateral E-cadherin complexes require calcium and catenins for their formation. J Cell Biol 1998;142:837–46. [13] Nelson WJ, Veshnock PJ. Ankyrin binding to (Na + + K +)ATPase and implications for the organization of membrane domains in polarized cells. Nature 1987;328:533–6. [14] Yap AS, Brieher WM, Gumbiner BM. Molecular and functional analysis of cadherin-based adherens junctions. Annu Rev Cell Dev Biol 1997;13:119–46. [15] Nitta T, Hata M, Gotoh S, Seo Y, Sasaki H, Hashimoto N, et al. Size-selective loosening of the blood–brain barrier in claudin-5-deficient mice. J Cell Biol 2003;161:653–60. [16] Groten T, Fraser HM, Duncan WC, Konrad R, Kreienberg R, Wulff C. Cell junctional proteins in the human corpus luteum: changes during the normal cycle and after HCG treatment. Hum Reprod 2006;21:3096–102. [17] Neulen J, Yan Z, Raczek S, Weindel K, Keck C, Weich HA, et al. Human chorionic gonadotropin-dependent expression of vascular endothelial growth factor/vascular permeability factor in human granulosa cells: importance in ovarian hyperstimulation syndrome. J Clin Endocrinol Metab 1995;80:1967–71. [18] Neulen J, Raczek S, Pogorzelski M, Grunwald K, Yeo TK, Dvorak HF, et al. Secretion of vascular endothelial growth factor/vascular permeability factor from human luteinized granulosa cells is human chorionic gonadotrophin dependent. Mol Hum Reprod 1998;4:203–6. [19] Wulff C, Wilson H, Largue P, Duncan WC, Armstrong DG, Fraser HM. Angiogenesis in the human corpus luteum: localization and changes in angiopoietins, Tie-2 and vascular endothelial growth factor messenger ribonucleic acid. J Clin Endocrinol Metab 2000;85:4302–9. [20] Dvorak HF, Nagy JA, Feng D, Brown LF, Dvorak AM. Vascular permeability factor/vascular endothelial growth factor and the significance of microvascular hyperpermeability in angiogenesis. Curr Top Microbiol Immunol 1999;237: 97–132. [21] Dvorak HF. Vascular permeability factor/vascular endothelial growth factor: a critical cytokine in tumor angiogenesis and a potential target for diagnosis and therapy. J Clin Oncol 2002;20:4368–80. [22] Esser S, Lampugnani MG, Corada M, Dejana E, Risau W. Vascular endothelial growth factor induces VE-cadherin tyrosine phosphorylation in endothelial cells. J Cell Sci 1998;111:1853–65. [23] Rodewald M, Herr D, Duncan WC, Fraser HM, Hack G, Konrad R, et al. Molecular mechanisms of ovarian hyperstimulation syndrome: paracrine reduction of endothelial claudin 5 by hCG in vitro is associated with increased endothelial permeability. Hum Reprod 2009;24:1191–9. [24] Senger DR, Van de Water L, Brown LF, et al. Vascular permeability factor (VPF, VEGF) in tumor biology. Cancer Metastasis Rev 1993;12:303–24. [25] Olson TA, Mohanraj D, Carson LF, Ramakrishnan S. Vascular permeability factor gene expression in normal and neoplastic human ovaries. Cancer Res 1994;54: 276–80. [26] Yamamoto S, Konishi I, Mandai M, Kuroda H, Komatsu T, Nanbu K, et al. Expression of vascular endothelial growth factor (VEGF) in epithelial ovarian neoplasms: correlation with clinicopathology and patient survival, and analysis of serum VEGF levels. Br J Cancer 1997;76:1221–7. [27] Hazelton D, Nicosia RF, Nicosia SV. Vascular endothelial growth factor levels in ovarian cyst fluid correlate with malignancy. Clin Cancer Res 1999;5:823–9. [28] Yamamoto S, Konishi I, Mandai M, Kuroda H, Komatsu T, Nanbu K, et al. Expression of vascular endothelial growth factor (VEGF) in epithelial ovarian neoplasms: correlation with clinicopathology and patient survival, and analysis of serum VEGF levels. Br J Cancer 1997;76(9):1221–7. [29] Cooper BC, Ritchie JM, Broghammer CL, Coffin J, Sorosky JI, Buller RE, et al. Preoperative serum vascular endothelial growth factor levels: significance in ovarian cancer. Clin Cancer Res 2002 Oct;8(10):3193–7.
216
D. Herr et al. / Gynecologic Oncology 127 (2012) 210–216
[30] Abell RG. The permeability of blood capillary sprouts and newly formed blood capillaries as compared to that of older blood capillaries. Am J Physiol 1946;147:237–41. [31] Futami R, Miyashita M, Nomura T, Makino H, Matsutani T, Sasajima K, et al. Increased serum vascular endothelial growth factor following major surgical injury. J Nippon Med Sch 2007;74:223–9. [32] Justinger C, Schluter C, Oliviera-Frick V, Kopp B, Rubie C, Schilling MK. Increased growth factor expression after hepatic and pancreatic resection. Oncol Rep 2008;20:1527–31.
[33] Monsky WL, Fukumura D, Gohongi T, Ancukiewcz M, Weich HA, Torchilin VP, et al. Augmentation of transvascular transport of macromolecules and nanoparticles in tumors using vascular endothelial growth factor. Cancer Res 1999 Aug 15;59(16):4129–35. [34] Nagy JA, Masse EM, Herzberg KT, Meyers MS, Yeo KT, Yeo TK, et al. Pathogenesis of ascites tumor growth: vascular permeability factor, vascular hyperpermeability, and ascites fluid accumulation. Cancer Res 1995;55:360–8. [35] Hamilton CA, Maxwell GL, Chernofsky MR, Bernstein SA, Farley JH, Rose GS. Intraperitoneal bevacizumab for the palliation of malignant ascites in refractory ovarian cancer. Gynecol Oncol 2008 Dec;111(3):530–2 Epub 2008 Jun 18.
Contents lists available at SciVerse ScienceDirect
Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno
VEGF induces ascites in ovarian cancer patients via increasing peritoneal permeability by downregulation of Claudin 5 Daniel Herr, Alexandra Sallmann, Inga Bekes, Regina Konrad, Iris Holzheu, Rolf Kreienberg, Christine Wulff ⁎ Department of Obstetrics and Gynecology, Ulm University Medical Centre, Ulm, Germany
a r t i c l e
i n f o
Article history: Received 3 April 2012 Accepted 1 May 2012 Available online 8 May 2012 Keywords: Ovarian cancer Ascites Claudin 5 Permeability
a b s t r a c t Objective. To evaluate the role of VEGF-dependent Claudin 5 production for the development of ascites via influencing endothelial permeability in peritoneal tissue of ovarian cancer patients. Methods. This study investigates the mechanisms of formation of ascites in ovarian cancer patients, performing RT-PCR, VEGF-ELISA and immunohistochemical dual staining for CD31 and Claudin 5. In addition, in order to analyze the connectivity of VEGF, Claudin 5, and endothelial permeability, an endothelial cell/ ovarian cancer cell-co-culture-system was established and evaluated performing Western blot analysis and a permeability assay. Results. Firstly, VEGF-gene expression was demonstrated for all ovarian cancer and peritoneal biopsies. In addition, quantification of VEGF in the serum and ascites of ovarian cancer patients revealed significantly increased values. We subsequently demonstrated Claudin 5 production in the peritoneal vessels, which was weaker than in the vessels of the controls. Evaluation of endothelial permeability finally showed a VEGF-dependent regulation via Claudin 5 suggesting a mechanism for the development of ascites in ovarian cancer patients. Conclusion. VEGF induces ascites in ovarian cancer patients. This instance happens due to increased peritoneal permeability, caused by downregulation of the tight junction protein Claudin 5 in the peritoneal endothelium. © 2012 Elsevier Inc. All rights reserved.
Introduction Worldwide, ovarian cancer occurs in about 200,000 women per year. Concerning mortality, ovarian cancer ranges around place 5 of malign disorders. The five-year overall survival rate is only about 40%. Unfortunately, in most cases ovarian cancer is diagnosed in advanced tumor stages. Thereby, one of the most impressing clinical symptoms is an increase of the abdominal circumference due to ascites [1-3]. Development of ascites is a consequence of dysregulated endothelial function leading to increased vascular permeability of peritoneal vessels [4]. Under physiological conditions, vascular permeability is mediated by strictly regulated opening and closure of cell–cell-junctions [5-7]. For that reason, any disturbance of junctional organization might result in dysregulated vascular function leading to pathological conditions by modifying the regular structure of the vessel wall [8]. Under malign circumstances, the typical example with clinical importance caused by increased vascular permeability is ascites. In the endothelium, paracellular permeability involves at least two different types of intercellular junctions; adherens junctions (AJ) and tight junctions (TJ). Those are formed by different transmembrane ⁎ Corresponding author at: Department of Obstetrics and Gynecology, University of Ulm, Prittwitzstrasse 43, 89075 Ulm, Germany. Fax: + 49 73150058502. E-mail address: [email protected] (C. Wulff). 0090-8258/$ – see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2012.05.002
proteins that promote homophilic cell–cell-interactions and transfer of intracellular signals [9]. Many reports support the concept, that intercellular junctions are dynamically remodeled not only during embryogenesis but also in resting cells [10]. Adhesive membrane proteins of AJ and TJ form adhesive complexes which act as zipper-like structures between interacting cells [11-15]. Endothelial cells express tissue-specific transmembrane adhesion proteins: the AJ VE-(vascular endothelial)-Cadherin and the TJ Claudin 5 [15]. Knockdown of Claudin 5 in mice is associated with a normal embryological development but, due to a defective blood brain barrier function, the Claudin 5 deficient mice die shortly after birth [15]. For the junctional protein Claudin 5 paracrine regulatory mechanisms have been shown in the human ovary. We demonstrated recently that Claudin 5 is expressed in the luteal vasculature and that the expression was decreased after in vivo treatment with hCG [16]. This effect may be secondary to vascular endothelial growth factor (VEGF) as hCG induces VEGF expression in luteal cells in vitro [17,18] and in vivo [19]. Indeed VEGF has been shown to be able to enhance vascular permeability [20,21] and influence endothelial adherens junctions [22]. In addition, it has been shown in an in vitro corpus luteum model, that downregulation of Claudin 5 in endothelial cells is associated with increased vascular permeability [23] demonstrating the importance of this junctional protein for regulation of vascular permeability.
D. Herr et al. / Gynecologic Oncology 127 (2012) 210–216
Concerning malign tumors VEGF, which is produced by cancer cells itself, induces angiogenesis in order to insure supply with oxygen and nutritive substances [24], which has also been shown for ovarian cancer cells [25,26]. Malign ovarian cysts induce VEGF-level 24-fold higher than in benign ovarian cysts [27]. Apparently, these high levels of VEGF increase local permeability causing extravasation of fluids, thus ascites. Here, we address the question of whether or not VEGF-dependent production of Claudin 5 acts as a regulator of vascular permeability in ovarian cancer patients. We hypothesized that VEGF produced by ovarian cancer cells decreases the amount of Claudin 5 in peritoneal vessels causing ascites by increasing endothelial permeability. We therefore went on to study Claudin 5 in peritoneal tissue as well as VEGF in serum and ascites of ovarian cancer patients. In addition, in order to elucidate the causality of a possible impact of VEGF on Claudin 5 and vascular permeability, we established a co-culturesystem of ovarian cancer- and endothelial-cells. Material and methods Patients Tumor- and peritoneal-tissue, serum and ascites were collected from patients undergoing laparatomy for advanced serous papillary ovarian cancer (FIGO stage III). As controls, we used tissue of patients with leiomyoma and uterine prolapse. In summary, 10 patients with ovarian cancer and 10 controls have been analyzed. This was institutionally approved after favorable ethical review and written consent of the patients. Morphological characterization of ovarian cancer Consecutive sections stained for hematoxylin and eosin were used to classify the ovarian cancer and peritoneal tissue. The embedded ovaries were serially sectioned, and tissue sections (4 μm) were placed onto BDH SuperFrost slides (BDH, Merck & Co., Inc., Poole, UK). Tissue sections were dewaxed in xylene, rehydrated in descending concentrations of ethanol, washed in distilled water, and stained with hematoxylin (Richard-Allan, Richland, MI) for 5 min, followed by a wash in water and acetic alcohol before staining with eosin (Richard-Allan) for 20 s. After dehydrating in ascending concentrations of ethanol and xylene, sections were mounted. Quantification of VEGF by ELISA (serum and ascites) In order to quantify the secreted amount of VEGF, a quantitative VEGF immunoassay was performed according to the manufacturer's protocol (R&D Systems, Minneapolis, USA. The optical density was measured at 450 nm (Dynatech Laboratory, Bath, UK). Immunohistochemistry dual staining Immunofluorescence double-staining was performed using the TSA-Kit (Perkin Elmer, Boston, USA) according to the manufacturer's instructions. The slides were incubated with mouse anti-human Claudin 5 (Zymed 18–7364, South San Francisco, USA, 1:100 dilution) overnight at 4 °C. CD 31 staining was performed to prove the colocalization to the endothelial compartment using a mouse antihuman CD 31 antibody (Dako M0823, Dako, Hamburg, Germany,1:30 dilution). Cell-culture for HUVEC and OvCar-3 Endothelial cell isolation and culture HUVEC (human umbilical vein endothelial cells) were isolated from multiple segments of normal term umbilical cords, pooled and
211
cultured in the endothelial cell growth medium (Promocell GmbH, Heidelberg, Germany). Cell culture for OvCar-3 cells According to the instructions of the American Type Culture Collection/ The Global Bioresource Center (ATCC), the OvCar-3 cells were cultivated in RPMI 1640 (PAA Laboratories, 3-Pasching, Austria) with 1% penicillin/ streptomycin and 10% FCS (PAA Laboratories, 3-Pasching, Austria). Media were changed every 48 h. RNA-isolation Total RNA from HUVEC, OvCar-3-cells, and ovarian cancer tissue was extracted from cells with the RNeasy kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. The RNA product was quantified by absorbance at 260 nm and total RNA (2.5 μg) was reverse transcribed into cDNA using random hexamer primers (Applied Biosystems, Foster City, USA). RT-PCR-analysis PCR amplification was carried out with 0.2 mM dNTP, 10 mM forward primer, 10 mM reverse primer, 1.5 mM MgCl2 and 1 U Taq polymerase (Qiagen) in the recommended buffer. The cycling conditions were 45 cycles at 94 °C for 45 s, 59 °C for 45 s and 72 °C for 60 s after the initial denaturation step. The primers had the following sequences: VEGF (367 bp) (forward: 5′-CGG GCC TCC GAA ACC ATG AAC TTT-3′; primer: 5′-CTA TGT GCT GGC CCT GGT GAG GTT GAT3′), Claudin-5 (462 bp) (forward: 5′-ACC GGC GAC TAC GAC AAG AAG A-3′; reverse: 5′-GCC CTG CCG ATG GAG TAA AGA-3′), and GAPDH (657 bp) (forward: 5′-CTG GCG CTG AGT ACG TCG-3′; reverse: 5′-TTG ACA AAG TGG TCG TTG A-3′. Bands were visualized after electrophoresis on a 2% agarose gel (Invitrogen) by staining with ethidium bromide. Pictures were taken with a camera system of Alpha Corporation (San Leandro, CA, USA). Co-culture of OvCar-3-cells and HUVEC HUVECs (2 × 10 5 per well) were seeded onto 6-well plates and grown to 90–100% confluency. On day 2, OvCar-3 cells were transferred to cell culture inserts (0.4 μm pore size, Becton Dickinson Nr. 353090). On day 3, cell culture inserts with OvCar-3-cells were inserted into the 6-well plates with HUVECs in culture. Endothelial cells were treated with 200 ng and OvCar-3 cells with 300 ng Flt1Fc. In this setting, molecules secreted by OvCar-3-cells follow the concentration gradient between the two different media such that the HUVEC can be stimulated. Extraction of membrane protein To obtain membrane fractions, cells were harvested into PBS. After centrifugation (5 min, 1200 rpm), the pellet was resuspended in 200 μl detergent-free lysis buffer (20 mM TrisHCl, pH 7.5, 0.5 mM EDTA, 10 mM BME, proteinase- and phosphatase-inhibitor cocktail (Sigma, Aldrich, Germany)). Cells were disrupted in a glass dounce homogenizer, incubated on ice, and ultracentrifuged at 35,000 rpm for 60 min at 4 C. The supernatant was referred to as the cytosolic fraction. The remaining pellet was resuspended in lysis buffer. Lysates were centrifuged at 15,000 rpm for 15 min and the supernatant was reserved as the membrane fraction. Protein concentration was determined using the BCA Protein Assay (Pierce, Rockford, IL, USA). Protein extraction from peritoneal tissue For protein extraction, peritoneal tissue was homogenized in 2 ml lysis buffer on ice using Ultraturrax (Janke & Kunkel GmbH, Staufen, Germany). Afterwards, the tissue was incubated on ice for 1 h followed by centrifugation at 1200 rpm for 10 min at 4 °C. The supernatant was extracted and centrifuged at 35,000 rpm at 4 °C for 1 h. The supernatant was referred to as the cytosolic fraction. The pellet containing the membranous fraction was resuspended and lysed in
D. Herr et al. / Gynecologic Oncology 127 (2012) 210–216
Permeability assay To test the permeability of endothelial cells for macromolecules, the HUVEC were seeded in the insert with the OvCar-3-cells underneath (supplemental data) to allow the flow of the dye through the membrane. HUVEC were seeded on micropore inserts (pore size 0.4 μm) at a density of 2 × 10 5 cells per insert and cultured for 2 days. On day 2, 1.75 × 10 5 OvCar-3-cells were seeded into 4 wells and cultured in medium for 1 day. The inserts were then moved into the wells containing the OvCar-3-cells . In this co-culturesystem, OvCar-3-cells and HUVEC were treated with Flt-1Fc (OvCar3: 300 ng/ml; HUVEC: 200 ng/ml). Bovine serum albumin (BSA) was labeled with Trypan Blue (66.7 mg Trypan Blue to 1.6 g BSA and 40 ml PBS), mixed with culture medium 1:1, and added to the insert, whereas unlabeled albumin was added to the lower compartment. Samples of the lower wells were taken after 10 and 30 min and protein concentration was measured at 607 nm. Each experiment was repeated on three to six separate occasions. Statistics Statistical analysis was performed using SPSS for Windows version 6.2. Data obtained from experiments was compared to controls and was tested for significant differences using a two tailed, unpaired t-test. Data are shown as mean ± SEM. Statistical analyses for the permeability assay were tested using ANOVA, followed by Duncan's multiple range test. Differences were considered to be significant at p b 0.05.
1000 800
VEGF
Western blotting Proteins were separated by SDS-PAGE in 4–20% Tris-glycine gels (Lonza, Rockland, USA) and transferred onto 0.45 μm nitrocellulose membranes (Schleicher Schuell, Dassel, Germany). Membranes were incubated with the primary antibody at 4 °C overnight (anti-Claudin 5 antibody at 1:500 dilution, Zytomed, Berlin, Germany; or GAPDH antibody at 1:500 dilution, Abcam, Cambridge, UK). Detection involved incubation with peroxidase-conjugated secondary antibody (Claudin 5: 1:1000, and GAPDH: 1:1000; HRP-labled sheep-anti mouse, GE Healthcare UK Limited, Little Chalfont, Buckinghamshire, UK) at room temperature for 1 h. Protein-antibody complexes were visualized using the ECL Western blot detection reagents (GE Healthcare UK Limited, Little Chalfont, Buckinghamshire, UK). Optical band density was quantified (Imager of Alpha Corporation, San Leandro, Ca., USA) and the results obtained for Claudin 5 were normalized to the values for GAPDH.
a 1200
600 400 200 0 control
b
ovarian cancer
2000 1500
VEGF
TX-100 buffer and again ultracentrifuged at 15,000 rpm at 4 °C for 15 min. The protein concentration was calculated using the BCA Protein Assay (Pierce, Rockford, IL, USA) according to the instructions of the manufacturer.
1000 500 0
c
3500 3000 2500
VEGF
212
2000 1500 1000 500 0
Fig. 1. VEGF level in pg/ml (± SEM) in serum and ascites of ovarian cancer patients. (a) VEGF in serum of ovarian cancer patients and controls at day 0 after surgery; the difference is significant (p b 0.05; n = 10). (b) VEGF in serum of ovarian cancer patients at days 0, 2, and 4 after surgery; the difference between days 2 and 4 is significant (p b 0.05; n = 10). (c) VEGF in ascites of ovarian cancer patients at days 0, 2, and 4 after surgery; the difference between days 0 and 2 is significant (p b 0.05; n = 10).
VEGF-concentration in ascites Since we supposed, that the tumor itself might be the main source of VEGF, the concentration of VEGF was measured in the ascites at days 0, 2, and 4 after surgery. As expected, we revealed a decrease of VEGF, which was significant between days 0 and 2 (Fig. 1c).
Results Ovarian cancer biopsies
Claudin 5 in the peritoneal vessels
In order to verify the histological diagnosis of ovarian cancer, hematoxylin-eosin staining was performed for all ovarian and peritoneal tissue samples, proving serous papillary ovarian cancer tissue in all cases. In addition, VEGF-gene expression was shown for all tissue samples (supplemental data).
In order to analyze the existence of Claudin 5 in the vessels of the peritoneal tissue, double-staining of Claudin 5 with CD 31 was performed, revealing co-localization in the endothelial cells of both, controls and ovarian cancer patients. In addition, it has been shown, that the proportion of Claudin 5 in the vessels of ovarian cancer patients is considerably weaker as in the vessels of the controls (Fig. 2). In order to further investigate the amount of Claudin 5 in the endothelial cell compartment, Western blot analysis confirmed the production of the Claudin 5 protein. Quantification of Claudin 5 thereby revealed a significant reduction of Claudin 5 in the peritoneal vessels of ovarian cancer patients as compared to the controls (Fig. 3).
VEGF-concentration in serum Measurement of VEGF in the serum of ovarian cancer patients preoperatively revealed significant higher values as compared with healthy controls (Fig. 1a). The analysis of VEGF in the serum at day 2 revealed no significant changes. However, at day 4 after surgery a significant increase of VEGF could be observed (Fig. 1b).
D. Herr et al. / Gynecologic Oncology 127 (2012) 210–216
213
control
ovarian cancer
CD 31
Claudin 5
Co-localisation
Fig. 2. Representative immunocytochemical staining of Claudin 5, CD 31, and their co-localization in peritoneal tissue (a–c control, d–f ovarian cancer) proving the occurrence of Claudin 5 in the endothelial cell compartment. Note the light Claudin 5-staining in f suggesting a reduction of Claudin 5 in the peritoneum of ovarian cancer patients as compared to controls.
Analysis in OvCar-3 cells and HUVEC in the co-culture-system First of all gene expression as well as protein production of VEGF in the ovarian cancer cell-line Ovcar-3 cells was demonstrated. We found a time-dependent increase of the VEGF-protein amount in the period from 5 to 120 h (Fig. 4 a,b). In addition, analysis of Claudin 5 in HUVEC revealed its gene expression and protein production (Fig. 4 c,d). Claudin 5 and permeability in the endothelial-ovarian-cancer-cell co-culture-system
production of Claudin 5 in HUVEC grown in the co-culture-system in vitro. We found a significant decrease of Claudin 5 in HUVEC cocultured with OvCar-3 cells, which was prevented by simultaneous treatment with the VEGF-inhibitor Flt1-Fc (Fig. 5). In addition, the possible influence of changed Claudin 5-production on endothelial permeability was investigated in the co-culture-system. We observed a significant increase of permeability in the co-cultured HUVEC as compared to the controls. Furthermore, additional treatment with Flt1-Fc again antagonized this effect and caused a significant decrease of permeability reaching again the level of the control (Fig. 6). Discussion
Since it has been shown in vivo, that Claudin 5 is reduced in the peritoneal vessels of ovarian cancer patients, we measured the
a GAPDH
Claudin 5 control
cancer
b
control
cancer
Fig. 3. a) Representative Western blot of Claudin 5 (19 kDa) in peritoneal tissue of controls and ovarian cancer patients with corresponding production of GAPDH (37 kDa). b) Densitometric quantification of Claudin 5 normalized to GAPDH. The observed difference is significant (p b 0.05; n = 6).
The current study was performed in order to investigate the molecular mechanisms of formation of ascites in patients with ovarian cancer. In this regard, we showed increased levels of VEGF in the serum and ascites of preoperative ovarian cancer patients. In addition, a decreased amount of the tight junction protein Claudin 5 was detected in the peritoneal vessels of ovarian cancer patients compared to healthy controls. We therefore went on to study a possible functional association between Claudin 5 and increased peritoneal permeability as morphological correlation of ascites and established a co-culture-system of ovarian cancer- and endothelial-cells. In this way, we revealed a VEGF-dependent decrease of Claudin 5 in endothelial cells co-cultured with ovarian cancer cells. The observed changes of this junctional protein did not only remain on a structural level, but could also be translated into a significant increase of endothelial permeability. It is likely that this novel paradigm of interaction of VEGF and Claudin 5 in the endothelial compartment will facilitate the explanation of the molecular regulation of altered peritoneal permeability causing ascites in ovarian cancer patients. In order to investigate this connectivity, we firstly focused on protein production of VEGF proving its presence in serous papillary ovarian cancer as well as in peritoneal tissue. This result is in line with that of others [25-27]. In addition, quantification of VEGF-levels in the serum of ovarian cancer patients revealed significantly increased values as compared to healthy controls, indicating a possible functional role for VEGF concerning formation of ascites by increasing endothelial permeability. In contrast to the majority of other tumor entities, metastases of ovarian cancer mainly occur intra‐abdominally and metastasize only exceptionally via blood vessels [28]. Therefore, VEGF facilitates cancer cells to disseminate in the abdominal cavity, which might provoke these cells to accumulate, thus again increasing
214
D. Herr et al. / Gynecologic Oncology 127 (2012) 210–216
b
h
a
1
2
3
140 120 100 80 60 40 20 0 0 1000 2000 3000 4000 5000 6000 7000 8000 9000
VEGF in pg/ml
d
c
GAPDH
1
2
3 Claudin 5
Fig. 4. a) RT-PCR amplification of (1) GAPDH (657 bp), (2) negative control, and (3) VEGF (367 bp) in OvCar-3 cells. b) Production of the VEGF-protein (ELISA) in Ovcar-3 cells revealing a significant time-dependent increase (+ standard deviation) in the period from 5 to 120 h (n = 3). c) RT-PCR amplification of (1) GAPDH (657 bp), (2) negative control, and (3) Claudin 5 (462 bp) in HUVEC. d) Western blot of Claudin 5 (19 kDa) and GAPDH (37 kDa) in HUVEC.
the production of VEGF, again with its monitored effects on peritoneal permeability. Such interrelation has been already supposed for different tumor entities such as breast cancer, colorectal cancer and ovarian cancer [28,29]. In addition, the VEGF-levels in the serum depend on the cumulative amount of tumor tissue. Therefore, they were measured on days 2 and 4 after surgery. At day 2 no changes had been observed, however at day 4 a significant increase of VEGF was found. This increase might be explained by the beginning process of wound healing tissue regeneration, since several authors described increased levels of VEGF in this context [30–32]. It is conspicuous, that the systemically increased serum-levels of VEGF only influence the peritoneal
a GAPDH
Claudin 5
control
Claudin-5/GAPDH
b
co-culture
co-culture + Flt1-Fc
12000 10000 8000 6000 4000 2000 0 control
co-culture
co-culture + Flt1-Fc
Fig. 5. Representative Western blot (a) of Claudin 5 and GAPDH in HUVEC co-cultured with OvCar-3 cells ± Flt1-Fc. Densitometric quantification (b) reveals a significant reduced amount of Claudin 5 in co-cultured cells. Treatment with Flt1-Fc antagonizes this effect increasing Claudin 5 again significantly. Values are normalized to GAPDH (p b 0.05; n = 6).
vessels but obviously not the complete periphery. It has been shown by Monsky et al., that there are huge differences of the amount of VEGF needed to increase endothelial permeability in different areas of the human body [33]. For example, increase of intracranial permeability requires a bigger amount of VEGF compared to vessels of the subcutaneous tissue [33]. Therefore, it can be supposed, that the observed VEGF-concentration in the ovarian cancer patients is too small in order to influence the periphery but is sufficient to alter the integrity of the peritoneal vessels. According to the consideration of tumor-amount dependent levels of VEGF, we also investigated VEGF-levels in the ascites, proving a significant postoperative decline at day 2 after surgery. Since its has been shown, that ascites with high concentrations of VEGF which has been injected into mice resulted in increased rates of peritoneal permeability [34], our result suggests, that ovarian cancer cells produce VEGF in order to increase the peritoneal permeability as a base for the formation of ascites which allows cancer cells to grow and to disperse in the visceral cavity. Since VEGF levels in ascites are much higher as compared to serum levels it is likely to hypothesize that cancer cell derived VEFG secreted into the ascites acts locally on the peritoneal vessels increasing their permeability. This also explains the observation that hyperpermeability in ovarian cancer patients is mostly restricted to peritoneal cavity. Initially, we had hypothesized that VEGF, produced by ovarian cancer cells, influences peritoneal permeability via regulation of the tight junction protein Claudin 5. Therefore, we first demonstrated production of Claudin 5 in the peritoneal vessels. In addition, quantification of the Claudin 5 proportion in the peritoneum of ovarian cancer patient revealed significant lower values as compared to healthy patients as controls. It has been shown in a physiologic in vitro model of the human corpus luteum, that VEGF-dependent downregulated levels of Claudin 5 cause increased endothelial permeability [23]. Therefore, we now assumed a functional role of VEGF-dependent production of Claudin 5 concerning regulation of endothelial permeability in the peritoneal tissue of ovarian cancer patients. In order to investigate this presumption, a new co-culture-system of ovarian cancer- and endothelial cells was established using OvCar3 cells and HUVEC which produce VEGF and Claudin 5, respectively. As expected, due to VEGF produced by OvCar-3 cells, endothelial cells co-cultured with ovarian cancer cells presented with significantly lower levels of Claudin 5. This result validated the functionality of
D. Herr et al. / Gynecologic Oncology 127 (2012) 210–216
permeability/permeability of control
a
References 1,8 1,6 1,4 1,2 1 0,8 0,6 0,4 0,2 0
permeability/permeability of control
b
215
1,6 1,4 1,2 1 0,8 0,6 0,4 0,2 0
Fig. 6. Relative permeability in the HUVEC-Ovcar-3-cell co-culture-system after 10 min (a) and 30 min (b). The black bar represents the control (HUVEC alone), the white bar the HUVEC-Ovcar-3 cells-co-culture, and the grey bar the co-culture with simultaneous treatment of both cell compartments with the VEGF-antagonist Flt1-Fc. Note the significant increase (after 10 and 30 min) of permeability in the co-culture (white) and the significant decrease after inhibition of VEGF (grey). The values are normalised to the permeability of the control (p b 0.05; n = 12).
our in vitro model-system, since in vivo we had observed the same effects as described above. We now finally focused on endothelial permeability and found significantly increased rates of permeability after VEGF-induced downregulation of Claudin 5. These findings make it possible that ovarian cancer cells produce VEGF in order to increase endothelial permeability via Claudin 5 which finally induces the formation of ascites and thereby dissemination of cancer cells in the abdominal cavity. Taken together, our results indicate an important role of VEGF as regulator of endothelial permeability via Claudin 5. In addition, the data provides an explanation of the mechanisms of the intraperitoneal treatment of ascites with the VEGF-antagonist bevacizumab in ovarian cancer patients, being a promising approach for symptomatic therapy in clinical studies [35]. For this reason, our data suggest a new pathophysiological model, explaining the molecular connectivity of the development of ascites in ovarian cancer patients, having possible clinical implications on the symptomatic treatment of patients in palliation.
Conflict of interest The authors declare that there are no conflicts of interest.
Appendix A. Supplementary data Supplementary data to this article can be found online at http:// dx.doi.org/10.1016/j.ygyno.2012.05.002.
[1] Ayantunde AA, Parsons SL. Pattern and prognostic factors in patients with malignant ascites: a retrospective study. Ann Oncol 2007;18:945–9. [2] Ayhan A, Guven S, Guven ES, Kucukali T. Is there a correlation between tumor marker panel and tumor size and histopathology in well staged patients with borderline ovarian tumors? Acta Obstet Gynecol Scand 2007;86:484–90. [3] Cheng MH, Yen MS, Chao KC, Sheu BC, Wang PH. Differential diagnosis of gynecologic organ-related diseases in women presenting with ascites. Taiwan J Obstet Gynecol 2008;47:384–90. [4] Garrison RN, Kaelin LD, Galloway RH, Heuser LS. Malignant ascites. Clinical and experimental observations. Ann Surg 1986;203:644–51. [5] Dejana E. Endothelial cell–cell junctions: happy together. Nat Rev Mol Cell Biol 2004;5:261–70. [6] Bazzoni G, Dejana E. Endothelial cell-to-cell junctions: molecular organization and role in vascular homeostasis. Physiol Rev 2004;84:869–901. [7] Walz A, Keck C, Weber H, Kissel C, Pietrowski D. Effects of luteinizing hormone and human chorionic gonadotropin on corpus luteum cells in a spheroid cell culture system. Mol Reprod Dev 2005;72:98–104. [8] Herr D, Fraser HM, Konrad R, Holzheu I, Kreienberg R, Wulff C. HCG Controls luteal Vascular Permeability via VEGF by Down Regulation of a Cascade of Adhesion Proteins. Fert Ster submitted for publication. [9] Dejana E, Orsenigo F, Molendini C, Baluk P, McDonald DM. Organization and signaling of endothelial cell-to-cell junctions in various regions of the blood and lymphatic vascular trees. Cell Tissue Res 2009;335:17–25. [10] Dejana E, Tournier-Lasserve E, Weinstein BM. The control of vascular integrity by endothelial cell junctions: molecular basis and pathological implications. Dev Cell 2009;16:209–21. [11] Cavey M, Rauzi M, Lenne PF, Lecuit T. A two-tiered mechanism for stabilization and immobilization of E-cadherin. Nature 2008;453:751–6. [12] Chitaev NA, Troyanovsky SM. Adhesive but not lateral E-cadherin complexes require calcium and catenins for their formation. J Cell Biol 1998;142:837–46. [13] Nelson WJ, Veshnock PJ. Ankyrin binding to (Na + + K +)ATPase and implications for the organization of membrane domains in polarized cells. Nature 1987;328:533–6. [14] Yap AS, Brieher WM, Gumbiner BM. Molecular and functional analysis of cadherin-based adherens junctions. Annu Rev Cell Dev Biol 1997;13:119–46. [15] Nitta T, Hata M, Gotoh S, Seo Y, Sasaki H, Hashimoto N, et al. Size-selective loosening of the blood–brain barrier in claudin-5-deficient mice. J Cell Biol 2003;161:653–60. [16] Groten T, Fraser HM, Duncan WC, Konrad R, Kreienberg R, Wulff C. Cell junctional proteins in the human corpus luteum: changes during the normal cycle and after HCG treatment. Hum Reprod 2006;21:3096–102. [17] Neulen J, Yan Z, Raczek S, Weindel K, Keck C, Weich HA, et al. Human chorionic gonadotropin-dependent expression of vascular endothelial growth factor/vascular permeability factor in human granulosa cells: importance in ovarian hyperstimulation syndrome. J Clin Endocrinol Metab 1995;80:1967–71. [18] Neulen J, Raczek S, Pogorzelski M, Grunwald K, Yeo TK, Dvorak HF, et al. Secretion of vascular endothelial growth factor/vascular permeability factor from human luteinized granulosa cells is human chorionic gonadotrophin dependent. Mol Hum Reprod 1998;4:203–6. [19] Wulff C, Wilson H, Largue P, Duncan WC, Armstrong DG, Fraser HM. Angiogenesis in the human corpus luteum: localization and changes in angiopoietins, Tie-2 and vascular endothelial growth factor messenger ribonucleic acid. J Clin Endocrinol Metab 2000;85:4302–9. [20] Dvorak HF, Nagy JA, Feng D, Brown LF, Dvorak AM. Vascular permeability factor/vascular endothelial growth factor and the significance of microvascular hyperpermeability in angiogenesis. Curr Top Microbiol Immunol 1999;237: 97–132. [21] Dvorak HF. Vascular permeability factor/vascular endothelial growth factor: a critical cytokine in tumor angiogenesis and a potential target for diagnosis and therapy. J Clin Oncol 2002;20:4368–80. [22] Esser S, Lampugnani MG, Corada M, Dejana E, Risau W. Vascular endothelial growth factor induces VE-cadherin tyrosine phosphorylation in endothelial cells. J Cell Sci 1998;111:1853–65. [23] Rodewald M, Herr D, Duncan WC, Fraser HM, Hack G, Konrad R, et al. Molecular mechanisms of ovarian hyperstimulation syndrome: paracrine reduction of endothelial claudin 5 by hCG in vitro is associated with increased endothelial permeability. Hum Reprod 2009;24:1191–9. [24] Senger DR, Van de Water L, Brown LF, et al. Vascular permeability factor (VPF, VEGF) in tumor biology. Cancer Metastasis Rev 1993;12:303–24. [25] Olson TA, Mohanraj D, Carson LF, Ramakrishnan S. Vascular permeability factor gene expression in normal and neoplastic human ovaries. Cancer Res 1994;54: 276–80. [26] Yamamoto S, Konishi I, Mandai M, Kuroda H, Komatsu T, Nanbu K, et al. Expression of vascular endothelial growth factor (VEGF) in epithelial ovarian neoplasms: correlation with clinicopathology and patient survival, and analysis of serum VEGF levels. Br J Cancer 1997;76:1221–7. [27] Hazelton D, Nicosia RF, Nicosia SV. Vascular endothelial growth factor levels in ovarian cyst fluid correlate with malignancy. Clin Cancer Res 1999;5:823–9. [28] Yamamoto S, Konishi I, Mandai M, Kuroda H, Komatsu T, Nanbu K, et al. Expression of vascular endothelial growth factor (VEGF) in epithelial ovarian neoplasms: correlation with clinicopathology and patient survival, and analysis of serum VEGF levels. Br J Cancer 1997;76(9):1221–7. [29] Cooper BC, Ritchie JM, Broghammer CL, Coffin J, Sorosky JI, Buller RE, et al. Preoperative serum vascular endothelial growth factor levels: significance in ovarian cancer. Clin Cancer Res 2002 Oct;8(10):3193–7.
216
D. Herr et al. / Gynecologic Oncology 127 (2012) 210–216
[30] Abell RG. The permeability of blood capillary sprouts and newly formed blood capillaries as compared to that of older blood capillaries. Am J Physiol 1946;147:237–41. [31] Futami R, Miyashita M, Nomura T, Makino H, Matsutani T, Sasajima K, et al. Increased serum vascular endothelial growth factor following major surgical injury. J Nippon Med Sch 2007;74:223–9. [32] Justinger C, Schluter C, Oliviera-Frick V, Kopp B, Rubie C, Schilling MK. Increased growth factor expression after hepatic and pancreatic resection. Oncol Rep 2008;20:1527–31.
[33] Monsky WL, Fukumura D, Gohongi T, Ancukiewcz M, Weich HA, Torchilin VP, et al. Augmentation of transvascular transport of macromolecules and nanoparticles in tumors using vascular endothelial growth factor. Cancer Res 1999 Aug 15;59(16):4129–35. [34] Nagy JA, Masse EM, Herzberg KT, Meyers MS, Yeo KT, Yeo TK, et al. Pathogenesis of ascites tumor growth: vascular permeability factor, vascular hyperpermeability, and ascites fluid accumulation. Cancer Res 1995;55:360–8. [35] Hamilton CA, Maxwell GL, Chernofsky MR, Bernstein SA, Farley JH, Rose GS. Intraperitoneal bevacizumab for the palliation of malignant ascites in refractory ovarian cancer. Gynecol Oncol 2008 Dec;111(3):530–2 Epub 2008 Jun 18.