| BackgroundOvarian cancer, one of the fetal malignancies of the female reproductive system, has the highest mortality rate among all gynecological oncology. It has been well known oxygen and nutrition provided by blood vessels are mainly responsible for tumor growth, development and metastasis. Jain et al. have proposed a theory about suppressing tumor growth via inhibiting its blood vessel formation. However, tumor neovascularization only composing of vascular endothelial cell has been challenged by vasculogenic mimicry (VM). VM is an alternative way to provide sufficient blood perfusion for highly malignant solid tumors, such as ovarian cancer, breast cancer and hepatocellular cancer. Accumulating studies have indicated that VM is correlated with five-year survival, tumor staging and metastasis in ovarian cancer patients, suggesting VM may be a potential therapeutic target for ovarian cancer. Therefore, it has significant implication to investigate the molecular mechanism of VM formation and seek for its related inhibitors thereby improving treatment outcome of ovarian cancer.The unusual channel pattern of VM is composed of highly malignant tumor cells and extracellular matrix (ECM). Tumor cells involved in VM formation exhibit endothelial phenotypes of mesenchymal cells which is similar to process of Epithelial-mesenchymal transition (EMT). EMT is a dynamic biological process characterized by loss of epithelial feature and the acquisition of mesenchymal feature as well as a change in cellular morphology, and EMT subsequently improves the motility of tumor cells. More importantly, reports suggest that EMT-like tumor cells more easily form VM structure and regulators contributing to EMT have been verified to involve in VM formation. However, EMT-like tumor cells require to dissolve and pass through ECM before VM formation, which is not an ignored process of VM formation.uPA, a serine protease with multiple function, is secreted by many cancer cells including ovarian cancer cells, and uPA enables to stimulate tumor cell proliferation and adhesion. It has been reported the high level of uPA indicates poor prognosis, and American society clinical oncology have appealed that uPA should act as risk assessment and a possible treatment target. In particular, uPA can accelerate tumor metastasis and promote tumor angiogenesis by degrading ECM and basement membranes such as laminin, fibronectin and collagen, allowing cells to migrate. However, it is unknown whether uPA enables to promote VM formation and possibly serves as a therapeutic target of VM based on uPA function to ECM degradation.Integrin avβ3 with high expression in neovascular endothelial cells and cancer cells can regulate cellular activities involved in angiogenesis. It’s reported that RGD domain enables to inhibit tumor neovascularization dependent on endothelial cells by specific and competitive binding with integrin avβ3. Compared with linear RGD, Cyclic RGD (cRGD) has much more binding sites of integrin avP3, thus cRGD presents promising therapeutic efficiency in angiogenesis induced by endothelial cells. Also, it is worth to note that endogenic RGD domain also enables to down-regulate uPA expression through high affinity with integrin avβ3. Moreover, intergrin avβ3 enables to promote interactions between cells and cells or ECM and plays pivotal role in invasion, migration and EMT process involving VM formation. Accordingly, it is provoked our interest to explore whether exogenous cRGD can act as VM antagonist to suppress its formation by down-regulating uPA expression and attenuating EMT.Herein, we first examine the correlation of uPA with VM in the level of ovarian cancer tissues and cells, and then attempt to validate its underlying molecular mechanism and estimate therapeutic potential of uPA. Most importantly, it is necessary to investigate whether exogenous cRGD has capacity to inhibit VM formation via down-regulating uPA expression and reducing EMT at protein level, gene level and biological characteristics. In this study, we aim to seek effective inhibitors for VM formation thereby further enriching therapeutic strategies for ovarian cancer.Chapter 1 The influence of uPA on VM formation in ovarian cancerObjectiveIn order to investigate the the influence of uPA on VM formation in ovarian cancer and reavel its relevant molecular mechanism, we conducted a serious expriments to study it from the level of tumor tissues and cell. It is important to further enrich and perfect the theory of angiogenesis mimicry, and it enable to provide a new therapeutic target for ovarian cancer anti-angiogenesis therapy.Method1. To investigate the correlation of uPA expression and VM in ovarian cancer tissuesWe conducted immunohistochemistry and CD34-PAS for 90 ovarian cancer tissue which collected from Zhujiang Hospital, Southern Medical University. Then we analysised the correlation of uPA expression (VM) with clinicopathological data and the correlation of uPA expression and VM2.3D-culture to asses the ability of ovarian cancer cells to form VM150μL of Matrigel (BD Biosciences) was dropped on 24-well plates and incubated at 37℃ for 45 min. SKOV-3, OVCAR-3 and HUVEC cells (3.5×105 cells/well) were seeded into the surface of the gels and maintained in 1640 medium supplemented without serum to observe tubule structures formation, including the number and the completeness of the tubule.Images were captured using an inverted light microscope at 100×after incubating for 8 h.3. Western Blot test the level of uPA expression in ovarian cancer cellsSKOV3 and OVCAR-3 cells were lysed in the RIPA buffer containing protease inhibitors and phosphatase inhibitors. Proteins were quantified with BCA kits and calculated the volume of 50μg protein. Proteins were separated using 10% SDS-PAGE gels, and then transferred to polyvinylidene difluoride membrane. After blocking in 3% BSA for 1 h, the membranes were incubated with specific primary antibodies overnight at 4℃, and then incubated with the secondary antibody for 2 h at room temperature. GAPDH was used as a protein loading control.4. Transient transfection and detection transfection efficiency with RT-PCR and Western BlotuPA si-RNA and negative control si-RNA were transfected into SKOV-3 and OVCAR-3 cells using Lipofectamine RNAiMAX Reagent according to the manufacturer’s protocol.Respectively, extracted the protein from SKOV-3 and OVCAR-3 cells which transfected and then incubated with 48-72 h. And then collected to assess transfection efficiency by western blot and Real time-PCR and then were used in further experiments.5.3D-culture to asses the ability of ovarian cancer cells which treated with uPA si-RNA to form VMAfter down-regulated uPA, SKOV-3 and OVCAR-3 cells were cultured in 3D-culture to compare the changes of their ability to form VM.6. Evaluated the influence of down-regulated uPA on VM-associated protein in ovarian cancer cells with western blotRespectively, extracted the proteins from SKOV-3 and OVCAR-3 cells which transfected and then incubated with 72 h. Next assessed the expression level of uPA, AKT, p-AKT, mTOR, p-mTO, MMP-2, Laminin5y2 by western Blot. GAPDH was used as a protein loading control.Results1. VM and uPA expression are correlated with clinicopathological datauPA protein expression in ovarian cancer tissues was assessed by immunohistochemistry. We detected different expression of uPA in ovarian cancer tissues and divided them into three group uPA (-/+), uPA (++) and uPA(+++) according the proportionof positive cell and the intensity of staining. Then the correlation of uPA expression with clinicopathological data was analyzed. High uPA expression (uPA++ and uPA+++) was found in 36% of low FIGO stage and in 82.5 % of high FIGO stage. Statistical analysis show that the uPA expression between low FIGO stage and high FIGO stage,were statistically significant.(P=0.000). High uPA expression (uPA++ and uPA+++) was aslo found in 41.6% G1, in 59.4% G2 and in 77.3% G3, Statistical analysis show that the uPA expression between G1, G2 and G3, were statistically significant (P=0.028). VM was found in 26% of low FIGO stage and in 57.5% of high FIGO stage. Statistical analysis show that the VM positive proportion between low FIGO stage and high FIGO stage,were statistically significant.(P=0.003). VM positive proportion was aslo found in 14 %G1, in 53% G2 and in 64% G3, Statistical analysis show that VM positive proportion between G1, G2 and G3, were statistically significant (P=0.000).2. The correlation of uPA expression and VM in ovarian cancer tissuesTo examine the correlation of uPA expression and VM, we analyzed uPA expression in tissues (VM+) and VM phenomenon in tissues with uPA expression. According to the level of uPA expression, all tissues were divided into three groups: uPA (-/+), uPA (++) and uPA (+++). It was observed more of VM in the uPA (+++) group:VM was detected in 18.1% of uPA (-/+) tissues,41.7 of uPA (++) tissues, and 57.1% of uPA(+++) tissues (Figure 2A). Further analysis presented that uPA expression was up-regulated in VM (+) group:there were 22.2% uPA (-/+),30.5% uPA (++) and 47.3% uPA (+++), while there was only 20.3% uPA (+++) in the VM (-) group.3. The correlation of uPA expression and VM in ovarian cancer cellThree-dimensional cultures were conducted to estimate the ability of the vessel-like channels formation in SKOV-3, OVCAR-3 and A2780 ovarian cancer cells using HUVEC cells as positive control. We foud cells enable head-tail formed a large number of vessel-like channels namely VM and we observed their different ability to form channels. It was found vessel-like channels formed by SKOV-3 and OVCAR-3 cells were similar to that of HUVEC cells, while the forming ability of SKOV-3 cells was better than that of OVCAR-3 cells (p<0.05). In contrast, no channels appeared in A2780 cells. Next, we investigated their expression level of uPA protein by western blot. The expression of uPA protein in SKOV-3 cells was highest among them, and the expression of uPA protein in OVCAR-3 cells was higher than in A2780 cells.4. Down-regulated uPA inhibited VM formation in ovarian cancer cellsSpecific uPA-siRNA fragments were transfected into both cells and the interfering effects of uPA were evaluated by RT-PCR and western blot. It was detected the expression levels of uPA mRNA and protein were obvious decreased after treating with uPA-siRNA. In three-dimensional cultures, down-regulated uPA caused notable decrease for the complete channels formed by SKOV-3 and OVCAR-3 cells compared with their si-NC (p<0.001 in SKOV-3, p<0.01 inOVCAR-3).5. The realated molecular mechanism of uPA promoted VM formationThe signal pathway of AKT/mTOR/MMP-2/Laminin5y2 was detected in the interfered SKOV-3 and OVCAR-3 cells by western blot. Compared with the si-NC, the expression of AKT, mTOR, their phosphorylation and MMP-2 was significantly decreased, while the uncleaved Laminin5y2 was increased in SKOV-3 si-uPA cells. For OVCAR-3, Compared with the si-NC, there was little differences on the amounts of total AKTand mTOR, whereas significantly less phosphorylated PI3k and Akt as well as the uncleaved Laminin5y2 was increased.Conclusion1. uPA expression and VM had correlation with ovarian cancer grad and stage:the higher the tumor grade/stage, the higher the percentage of VM structures or uPA expression2. uPA promoted VM formation in ovarian cancer.3. uPA was a positive mediator for VM formation in ovarian cancer via AKT/mTOR/MMP-2/Laminin5y2 signal pathway.Chapter 2 The effect of cRGD on VM formation in ovarian cancer cells and its relevant molecular mechanism Section 1 cRGD inhibited VM formation in ovarian cancer through down-regulated uPAObjectiveDocuments disclosed that endogenic RGD domain down regulated uPA expression in breast cancer, and our previous indicated that uPA was a a new VM antagonist for ovarian cancer. So next we explored ectogenic cycle RGD (cRGD) whether could inhibit VM formation in ovarian cancer through down-regulated uPA.Method1. MTT detection the optimal dose of cRGDSKOV-3 and OVCAR-3 cells were reseeded into 96-well culture plates at a density of 5×103 cells/well and incubated at 37℃. After incubated with 24 h, Different concentrations of cRGD (PBS) was added to plates and incubated with 48 h. And then the media were detached and replaced with 20 uL of MTT solution (5 mg/mL) followed by incubation for 4 h. After incubation, the solution was replaced with DMSO and the plates were slightly shaken. The plates were detected at 540 nm using a microplate reader. This experiment was repeated three times.2. The effect of cRGD on VM formation for ovarian cancer cells in 3D-culturedSKOV-3 and OVCAR-3 cells were collected and resuspended with RPMI-1640 without FBS medium.3.5×105/cells were seeded into 24-well paltes which containing Matrigel. cRGD which dissolved in PBS or PBS were added to cell suspensions before plating the cell into Matrigel. Images were captured using an inverted light microscope at 100x after incubating for 8 h.3. The change of AKT/mTOR/MMP-2/Laminin5y2 signal pathway after treated cRGD with western Blot in ovarian cancer cells.Respectively, extracted the protein from SKOV-3 and OVCAR-3 cells which treated with cRGD or not. Next assess the expression level of uPA, AKT, p-AKT, mTOR, p-mTO, MMP-2, Laminin5y2 with western Blot. GAPDH was used as a protein loading control.Result1.40 nM cRGD was selected in subsequent experimentsThe results of MTT exhibited that the IC50 of cRGD for SKOV-3 cells was 6160 nM and OVCAR-3 was 536 nM. To eliminate cRGD itself cytotoxicity, we screened appropriate cRGD dose (40 nM) far below their IC50 in SKOV-3 and OVCAR-3 cells applying in our subsequent experiments.2. cRGD inhibited VM formation in ovarian cancerwe estimated the inhibiting ability of cRGD to VM formation in three-dimensional cultures after SKOV-3 and OVCAR-3 cells were treated with cRGD, respectively. Compared with control groups, it was observed that cRGD could obviously suppress VM formation for both ovarian cancer cells (P<0.001 in SKOV-3 cells, P<0.01 in OVCAR-3 cells)3. cRGD inhibited VM formation in ovarian cancer by regulating AKT/mTOR/MMP-2/Laminin5γ2 signal pathwayThe change of AKT/mTOR/MMP-2/Laminin5y2 were detected in SKOV-3 and OVCAR-3 cells treated with cRGD or not by western blot. Compared with the control group, the expression of AKT, mTOR, their phosphorylation and MMP-2 was significantly decreased, while the uncleaved Laminin5γ2 was increased in cRGD group.Conclusion1. cRGD inhibited VM formation in ovarian cancer via down-regulating uPA thereby inhibiting AKT/mTOR/MMP-2/Laminin5γ2 signal pathway.2. Reconfirmed the signal pathway of uPA involving in VM formation in ovarian cancer.Section 2. cRGD inhibited VM formation in ovarian cancer through reducing EMTObjectiveEpithelial-mesenchymal transition (EMT) is a important key to promote VM formation. In this section, we want to explore whether cRGD enables to inhibited VM formation in ovarian cancer through reducing EMT.Method1. The influence of cRGD on ovarian cancer cells migration was assessed by transwellSKOV-3 and OVCAR-3 cells which growth in logarithmic phase were collected. Dissociated cells (1×104/insert) in serum free medium were seeded on inserts and medium 10% fetal calf serum was added to the lower chambers. After incubation for 10 h, the insert was collected. The migration cells adhered to the membrane lower surface were counted under a microscope in five random optical fields.2. The influence of cRGD on ovarian cancer cells invasion was assessed by borderMatrigel was dropped into inserts, and then collected SKOV-3 and OVCAR-3 cells which growth in logarithmic phase. Dissociated cells (1×104/insert) in serum free medium were seeded on inserts and medium 10% fetal calf serum was added to the lower chambers. After incubation for 12 h, the insert was collected.The migration cells adhered to the membrane lower surface were counted under a microscope in five random optical fields.3. The influence of cRGD on EMT-associated protein were assessed by Immunofluorescence in ovarian cancer cellsSKOV-3 and OVCAR-3 cells were seeded in the chamber slides in medium and incubated overnight at 37℃. And then both cells were treated with cRGD for 48 h. The cells were were incubated with E-cadherin antibody, N-cadherin antibody and Vimentin antibody overnight at 4℃. Anti-rabbit IgG Alexa Fluor 488 were treated with cells for 1 h followed by Hochest33342 staining. The fluorescent images were viewed by confocal laser scanning microscope.4. The influence of cRGD and down-regulated uPA on EMT-associated protein were assessed by Western Blot in ovarian cancer cellsRespectively, extracted the protein from SKOV-3 and OVCAR-3 cells which transfected and then incubated with 72 h or treated with cRGD for 48h. Next assess the expression level of E-cadherin, N-cadherin and Vimentin with western Blot. GAPDH was used as a protein loading control.5. Observe the influence of cRGD and down-regulated uPA on ovarian cancer cellular morphologyAfter SKOV-3 and OVCAR-3 cells were transfected for 72 h or treated with cRGD for 48h, the change of cellular morphology were viewed by Inverted Microscope.Results1. cRGD inhibited the ability of migration and invasion in ovarian cancer cellsAfter treated with cRGD for 10 h (12 h), we found the ability of migration and invasion for SKOV-3 and OVCAR-3 cells was distinctly weakened compared with control groups (all p<0.01).2. cRGD up-regulated E-cadherin expression and down-regulated N-cadherin, VimentinFrom the Immunofluorescence and Western Blot results, we foud that cRGD up-regulated E-cadherin expression and down-regulated N-cadherin, Vimentin. We still found that there was no change in these protein after down-regulated uPA expression for SKOV-3 cells, while for OVCAR-3cells, no change in E-cadherin, decreased N-cadherin and increased Vimentin.3. cRGD reverted ovarian cancer cellular morphologyAfter treated with cRGD for 48 h, compared with control group, we observed that SKOV-3 and OVCAR-3 cells relatively acquired cell-to-cell adhesions and losing spindle-like phenotype. While,there was no change in cellular morphology after down-regulated uPA.Conclusion1. cRGD inhibited VM formation in ovarian cancer via reducing EMT2. cRGD reduced EMT wasn’t caused by the decreased uPA. |
PDF Full Text Request