Expressions Of VEGF-C And VEGFR-3 In Esophageal Squamous Cell Carcinoma And Effect Of VEGF-C SiRNA On Anti-lymphangiogenesis In Vitro And In Vivo

Lymphatic vessel is a major pathway for the metastasis of malignant solid tumors,especially epithelial carcinomas. Such metastasis accounts for an important reason forpostoperative recovery of tumor patients. Thus, it is important to understand the associationof lymphangiogenesis with tumor metastasis, and to explore the new potential therapeutictarget in the treatment of cancer with anti-lymphangiogenic approach.Esophageal carcinoma (ES) is one of the most common malignancies of thegastrointestinal tract and greatly threatens human health and life. The mechanism oftumorigenesis remains unclear for ES. Most ES patients are already in the late stage at thetime of diagnosis due to lack of specific symptoms, often with lymphatic metastasis, andcannot be treated effectively. Therefore, it is very important clinically to understand the roleof lymphangiogenesis in the metastasis of ES and to search for the anti-lymphangiogenesisagents to block ES metastasis.Vascular endothelial growth factor C (VEGF-C) and its receptor VEGF receptor-3(VEGFR-3, also designated Flt-4) are recently identified lymphangiogenic stimulators thatcan selectively induce growth of the lymphatic vasculature. VEGFR-3 was cloned byPajusola and his colleagues in 1992 from the cDNA library of human HEL erythroleukemiacells. It is a transmembrane glycoprotein with a molecular weight of 180 kD and acts as atyrosine kinase receptor. Its gene is located on chromosome 5q33-q35. VEGFR-3 is requiredfor vascular formation in the embryo and expressed in both venous and lymphatic endothelia.However, it is present only in lymphatic endothelia in adults. VEGF-C was cloned from thelibrary of human prostate carcinoma (PC-3) cells by Joukov et al in 1996. It is a 46.9 kDsecretory glycoprotein with homology to VEGF-A and functions as a ligand for VEGFR-3.VEGF-C gene is localized to chromosomes 4q34 and generates a protein of 419 amino acids. VEGF-C expresses at low level in heart, skeletal muscle, lung and kidney.Recent studies have showed over-expression of VEGF-C in cancer cells. VEGF-C bindsits receptor VEGFR-3 via autocrine/paracrine pathway and activates VEGFR-3. ActivatedVEGFR-3 undergoes autophosphorylation, and then initiates its signal transduction pathwayto induce mitosis and proliferation of lymphatic endothelial cells and finally formation ofnew lymphatics. During lymphangiogenesis, the chance for tumor to metastasize is enhanced.In 2001, Mandriota et al found that tumors in double-transgenic (Rip VEGF-C) mice oftenmetastasized to the lymph nodes in the pancreas and peritumoral tissue were rich inlymphatics. These findings demonstrated that VEGF-C could induce lymphangiogenesis,which may mediate tumor metastases. Skobe et al observed within human breast cancersafter orthotopic transplantation into nude mice the intratumoral lymphangiogenesis, whichcould be promoted by over-expression of VEGF-C in breast cancer cells and significantlyenhanced tumor metastasis to regional lymph nodes and lungs. Kinoshita et al found that theover-expression of VEGF-C showed a significant positive correlation with lymphatic vesselinvasion and negative correlation with the patients'5-year disease free survival rate.However, there are few reports showing that breast carcinomas contained nowell-developed lymphatics; the poorly-developed lymphatics were mostly present in theparatumoral tissue and not associated with tumor size, grade, or nodal metastasis; and thedensity of lymphatics was inversely correlated to tumor invasiveness. These reportssuggested breast tumor lymphangiogenesis may not always occur at physiological levels ofVEGF-C and tumor may metastasize via pre-existing lymphatics.Recent emergence of RNA interference (RNAi) technique has provided a new approachfor gene therapy of tumors. It uses short interfering RNAs (siRNA) to specifically andefficiently turn off the expression of the target gene, but not to affect normal genes. Thistechnique has become a powerful tool not only to study gene function and viral and inheritablediseases, but also to develop highly specific therapy for human cancers.Literature search shows a few studies on expression of VEGF-C and its receptorVEGFR-3 in esophageal squamous cell carcinoma (ESCC), but no report for effect of siRNAon expression of VEGF-C in esophageal carcinoma cell line or tissue. This study is designedto examine expressions of VEGF-C and VEGFR-3 in ESCC, to determine the effect ofsiRNA on expression of VEGF-C and/or VEGFR-3 in vitro and in vivo, and to explored thepossibility of using specific siRNA of VEGF-C or VEGFR-3 as antilymphangiogenesisagents in gene therapy of ESCC.The expression of VEGF-C and VEGFR-3 in cancer tissue was examined with immunohistochemistry (IHC), in situ hybridization (ISH), and reverse transcriptasepolymerase chain reaction (RT-PCR) in 49 cases of ESCC containing area of tumor (n=49),normal esophageal mucosa (n=49) and dysplasia (n=23). The correlation of expression ofVEGF-C and VEGFR-3 with lymphangiogenesis was also analyzed.The effect of siRNA on expression of VEGF-C in vitro was analyzed in esophagealsquamous cell carcinoma cell line EC9706. After confirmation of expression of endogenousVEGF-C mRNA in EC9706 by RT-PCR, 4 different constructs of pSINsi-U6 VEGF-CsiRNA plasmids were made and transfected into EC9706 cells with liposome. Theproliferation and cell cycle of transfected EC9706 cells were analyzed by MTT assay & flowcytometry, respectively, and also correlated with expression of VEGF-C mRNA and proteindetected by IHC, ISH and RT-PCR.The effect of siRNA on expression of VEGF-C in vivo was analyzed in nude mice modelfor human esophageal carcinoma, which was established by subcutaneous injection ofEC9706 cells transfected with pSINsi-U6 VEGF-C siRNA plasmids. The growth of tumor innude mice was recorded. The expression of VEGF-C in tumors was examined by IHC,RT-PCR, and ISH.Our aims are to understand the expression of VEGF-C and/or VEGFR-3 in ESCC cellline and tissue; to characterize the expression of VEGF-C mRNA and protein in EC9706cells transfected with pSINsi-U6 VEGF-C siRNA plasmids; and correlate the effect of siRNAon expression of VEGF-C with tumor cell proliferation, cell cycle, and tumor growth in nudemice; and to provided experimental evidences for the antilymphangiogenesis treatment ofESCC through specifically turning off VEGF-C gene.This study consists of three partsPart 1: Expressions of VEGF-C and its receptor VEGFR-3 in ESCC,dysplasia, and normal esophageal mucosa.Methods:1. Immunohistochemistry (IHC) staining was used to examine the expression ofVEGF-C and VEGFR-3 in 49 cases of ESCC containing area of tumor (n=49), normalesophageal mucosa (n=49) and dysplasia (n=23).2. ISH and RT-PCR were used to determine the expression of VEGF-C and VEGFR-3mRNA in 49 cases of ESCC containing area of tumor (n=49), normal esophageal mucosa(n=49) and dysplasia (n=23). 3. IHC and enzyme-histochemical methods were used to measure the lymphatic vesseldensity (LVD) in ESCC and peritumoral tissue.4. The SPSS 10.0 statistical software was used for all statistical analyses. Thedifferences in positive rates and means were analyzed by chi-square or Fisher, and thedifferences in means were analyzed by t-test (two groups) or ANVOA (more than twogroups). The relationship of two variables was analyzed by correlation analysis. P values lessthan 0.05 were defined statistically significant.Results:1. Immunohistochemistry showed positive staining of VEGF-C and VEGFR-3 as browngranules in the cytoplasm of cancer cells. VEGF-C was positive in 73.5% (36/49) of ESCC,but negative in the normal esophageal mucosa and dysplasia. VEGF-C expression correlatedsignificantly with lymphatic node metastases (P＜0.01), but not with the patient's age, genderor tumor grade (P＞0.05). VEGFR-3 was positive in 32.7% (16/49) of ESCC. The expressionof VEGFR-3 were correlated with metastasis in lymphatic nodes (P＜0.05), but not with thepatient's gender or tumor grade (P＞0.05). The results revealed that VEGF-C and VEGFR-3expression was correlated significantly with lymph node metastasis. VEGF-C and VEGFR-3could express in ESCC tissues at the same case with a good correlations.2. ISH showed expression of VEGF-C mRNA in 46.9% (23/49) ESCC, but in none ofthe normal esophageal mucosa. VEGF-C mRNA expression was correlated significantly withlymph node metastases (P＜0.01), but not with patient's gender or pathologic grade (P＞0.05).VEGFR-3 mRNA was expressed in 24.5% (12/49) of ESCC, and significantly related withlymphatic node metastasis (P＜0.01), but not with patient's gender and pathologic grade(P＞0.05). These results indicated that expressions of VEGF-C and VEGFR-3 mRNAcorrelate significantly with lymph node metastasis in ESCC.3. RT-PCR showed expression of VEGF-C mRNA was in 59.2% (29/49) of ESCC, butin none of the normal esophageal mucosa. VEGF-C mRNA expression was correlatedsignificantly with lymph node metastases (P＜0.01), but not with the patient's gender andpathologic grade (P＞0.05). VEGF-C mRNA was also expressed in 52.2% (12/23) ofdysplasia with the relative optical density (ROD) of 0.31±0.18, significantly lower than thatin ESCC (0.68±0.24, P＜0.01). Expression of VEGF-C mRNA detected by RT-PCR werecorrelated with that detected by ISH (P＜0.01) without significantly difference (P＞0.05) inESCC. VEGFR-3 mRNA was positive in 28.6% (14/49) of ESCC, but in none of the normalesophageal mucosa. Expression of VEGFR-3 mRNA was correlated significantly with lymph node metastases (P＜0.01), but no with patient's gender or pathologic grade (P＞0.05).Expression of VEGFR-3 mRNA was also observed in 44.9% (22/49) of paratumoral tissue,and correlated significantly with lymph node metastasis (P＜0.05).4. The lymphatic vessel density detected by IHC staining (LVDi) and double enzymehistochemical stains was significantly higher in the ESCC with than without lymph nodemetastasis. The same was true in paratumoral tissue for LVDi and LVD detected byenzyme-histochemical stain (LVDe) in the ESCC with than without metastasis. LVDi wassignificantly lower than LVDe in ESCC with lymph node metastasis (P＜0.01). LVD inparatumoral tissue was significantly higher than that in tumor tissue (P＜0.01).PartⅡ: Effects of VEGF-C siRNA on expression of VEGF-C mRNAand protein in esophageal carcinoma EC9706 cells in vitro.Methods:1. Three different single strand DNAs (HX1, HX2, HX3) were synthesized according tothe hair loop RNA sequence, and subcloned into eukaryotic expression vector pSINsi-U6 tomake 3 constructs of VEGF-C siRNA (pSINsi-U6-HX1, pSINsi-U6-HX2, pSINsi-U6-HX3).One additional construct of random siRNA (pSINsi-U6-HX4) not homologous to anyhuman genes were made in the same way as control.2. Above 4 constructs (pSINsi-U6-HX1, pSINsi-U6-HX2, pSINsi-U6-HX3, andpSINsi-U6-HX4) and control (vector only) were transfected into esophageal carcinomaEC9706 cells with liposome. Positive single colonies were selected with G418.3. Cell growths, cell cycles, and invasive ability of transfected EC9706 cells wereevaluated by MTT assay, FCM, and Boyden chamber assay, respectively.4. Expression of VEGF-C protein and mRNA in the transfected EC9706 cells wasexamined by IHC, and ISH and RT-PCR, respectively.5. The SPSS 10.0 statistical software was used for all statistical analyses. The differencesin positive rates and means were analyzed by chi-square or Fisher, and the differences inmeans were analyzed by t-test (two groups) or ANVOA (more than two groups). Therelationship of two variables was analyzed by correlation analysis. P values less than 0.05were defined statistically significant.Results:1. After selection with G418, stable clones were established for pSINsi-U6-HX1, pSINsi-U6-HX2, pSINsi-U6-HX3, and pSINsi-U6-HX4.2. MTT assay showed no significant difference in cell proliferation betweenuntransfected and pSINsi-U6-siRNA transfected cell groups (P＞0.05), suggesting noinfluence of VEGF-C siRNA on cell proliferation.3. FCM assay revealed the percentage of cells in different cell cycle as follows:pSINsi-U6-HX1 pSINsi-U6-HX2 pSINsi-U6-HX3 pSINsi-U6-HX4 EC9706G0/G1 phase: 78.3±6.5% 84.9±7.1% 77.6±6.9%, 79.2±7.2% 79.4±7.4%S phase: 12.9±1.2% 12.5±1.0% 12.8±1.1% 13.3±1.2% 13.5±1.3%The cell population in pSINsi-U6-HX2 group was significantly increased in G0/G1phase but significantly decreased in S phase as compared with EC9706 group (P＜0.05).There were no significant differences among other groups in all the cell cycles. These resultsindicated that VEGF-C siRNA may have some impact on cell cycle.4. Boyden chamber assay showed no significant differences in cell number penetratedthe membrane among five groups (P＞0.05), suggesting no effects of VEGF-C siRNA on cellinvasive ability.5. Immunohistochemical stain showed significant less expression of VEGF-C protein inVEGF-C siRNA transfected cells than in random siRNA transfected and untransfected cells(P＜0.01). The result demonstrated that VEGF-C siRNA could inhibit VEGF-C expression inEC9706 cells.6. In situ hybridization results revealed significant less expression of VEGF-C mRNA inVEGF-C siRNA transfected cells than that in random siRNA transfected and untransfectedcells (P＜0.01). The result showed VEGF-C siRNA inhibited expression of VEGF-C mRNAin cells.7. Electrophoresis of RT-PCR products showed specific band at 203bp for VEFG-C andband at 480bp forβ-actin in all the groups. The relative optical density (ROD) in threeVEGF-C siRNA transfected groups was 0.35±0.03, 0.34±0.02 and 0.34±0.03,respectively without significant differences among them. The ROD of random siRNAtransfected and untransfected groups were 0.71±0.14 and 0.69±0.18, respectively, andsignificantly higher than those of 3 VEGF-C siRNA transfected groups (P＜0.01). The resultsuggested that VEGF-C siRNA inhibit VEGF-C mRNA expression in EC9706 cells. PartⅢEffects of VEGF-C siRNA on expression of VEGF-C proteinand mRNA in xenografts tumor tissue of nude mice.Methods:1. Four-week-old female athymic nude mice were housed in specific pathogen freeenvironment and randomly divided into five groups. Each group received flank subcutaneousinjection of one of 4 siRNA transfected or untransfected EC9706 cells.2. The growth of tumor in nude mice was recorded. The weight and volume of tumorwere measured after termination of mice.3. Expressions of VEGF-C in tumor tissues were examined by IHC, ISH and RT-PCR.4. The SPSS 10.0 statistical software was used for all statistical analyses. Thedifferences in positive rates and means were analyzed by chi-square or Fisher, and thedifferences in means were analyzed by t-test (two groups) or ANVOA (more than twogroups). The relationship of two variables was analyzed by correlation analysis. P values lessthan 0.05 were defined statistically significant.Results:1. Tumor grew in every mouse of all the groups. After 4 weeks, the mice wereterminated. The weight and volume of tumors were measured immediately. Grossexamination showed that tumors were solid and circumscribed without ascites or metastasis.2. Tumor masses in 3 VEGF-C siRNA (HX1, HX2, and HX3) transfected, untransfected,and random siRNA transfected (HX4) groups weighted 863±90mg, 784±73mg, 875±93mg,912±97mg and 8901±95mg respectively. Tumor volumes in 3 VEGF-C siRNA (HX1, HX2,and HX3) transfected, untransfected, and random siRNA transfected (HX4) groups were528±44mm~3, 483±33mm~3, 513±42mm~3, 582±59mm~3 and 561±57mm~3, respectively. Tumorweight in pSINsi-U6-HX2 group was significantly lower than that in untransfected group(P＜0.05). Tumor volume of pSINsi-U6-HX2 group was significantly lower than thant inuntransfected (P＜0.01) and random siRNA (pSINsi-U6-HX4) transfected groups (P＜0.05).There were no significant differences among other groups in both tumor weight and volume.3. Immunohistochemical stain showed that expression of VEGF-C protein wassignificantly less in VEGF-C siRNA transfected groups than that in untransfected (P＜0.01)and random siRNA transfected groups (P＜0.05). There was no significant difference inexpression of VEGF-C protein between random siRNA transfected and untransfected groups.The result indicated that VEGF-C siRNA could inhibit VEGF-C protein expression in tumor. 4. ISH and RT-PCR showed significantly lower expression of VEGF-C mRNA inVEGF-C siRNA transfected groups than that in untransfected and random siRNA transfectedgroups (P＜0.01). There was no significant difference in expression of VEGF-C proteinbetween random siRNA transfected and untransfected groups. These results suggested thatVEGF-C siRNA inhibited expression ofVEGF-C mRNA in tumor.Conclusions:1. Both frequency and level of expression of VEGF-C and VEGFR-3 protein and mRNA aresignificantly higher in esophageal squamous cell carcinoma than those in normal esophagealmucosa. Such expressions are correlated with lymph node metastasis. The results suggestVEGF-C and VEGFR-3 may serve as new targets for anti-lymphangiogenesis therapy inESCC.2. It is the first study to apply IHC and immuno-enzyme technique of 5'-NeucleiotidaseAlkaline Phosphatase (5'-Nase-ALPase) double enzyme histochemical staining inmeasurement of lymphatic vessel density (LVD) in ESCC. Our results prove that there islymphangiogenesis in ESCC tissues. The LVD in paratumoral tissue is significantly higherthan that in tumor, and correlates with lymph node metastasis.3. It is the first study to apply VEGF-C siRNA in esophageal squamous cell carcinomaEC9706 cells. Our results show that expressions of VEGF-C mRNA and protein aresignificantly lower in VEGF-C siRNA transfected cells than that in untransfected or randomsiRNA transfected cells, suggesting inhibition of expression of VEGF-C mRNA and proteinby VEGF-C siRNA in vitro. Inhibition of VEGF-C protein expression in EC9706 cells andimplanted tumor may be due to simultaneously down-regulation of VEGF-C mRNAexpression.4. Animal models for implanted tumor show that expression of VEGF-C mRNA and proteinin implanted tumors are significantly lower in VEGF-C siRNA transfected groups than thatin untransfected or random siRNA transfected groups. Results indicate that VEGF-C siRNAcan inhibit expression of VEGF-C in vivo.5. Both in vitro and in vivo experiments show pSINsi-U6-HX2 has the significantly higherinhibition than other two VEGF-C siRNAs (pSINsi-U6-HX1 and pSINsi-U6-HX3) onexpression of VEGF-C mRNA and protein, tumor growth (weight and volume), and cellentry into S phase in cell cycle. These results suggest that the specific fragment of VEGF-CsiRNA is required for more efficient turn off of VEGF-C gene in gene therapy of ESCC.