Screening And Identification Of COX-2 Regulated Molecules In Gastric Carcinogenesis And Functional Study

【Background】Gastric cancer is one of the most common malignant tumors in the world, especialy in Asia. But the pathogenesy is still unclear. COX-2, frequently detected over-expressed in gastric cancer tissues and cell lines, was believed to play a crucial role in promotion of proliferation and angiogenesis in gastric cancer. Up-regulation of COX-2 might facilitate invasion of gastric cancer and was significantly related to the low survival rate of gastric cancer patients. Reduced COX-2 activity could result in decreased angiogenesis, increased apoptosis and suppressed synthesis of prostaglandins. COX-2 was closed related to gastric carcinogenesis and progression, however, the underlying mechanisms are not fully understood. The proteomics approach provides a new tool to comparative studies of protein expression levels. Two-dimensional gel electrophoresis (2-DE) and matrix-assisted laser desorption/ionization-time of flight mass spectrometry analysis (MALDI-TOF MS) are principal techniques which could efficiently find differential expressing proteins and screen out the disease related markers. In this study, proteomic strategy was used to compare the differential expressing proteins between COX-2 siRNA stably transfected gastric cancer cells SGC7901 and the mock vector transfected SGC7901 cells. COX-2 modulated proteins were screened out and follow-up verification and functional studies were carried out to elucidate the mechanism of COX-2 related gastric carcinogenesis and progression, and provide solid experimental basis for the development of novel targets for the gastric cancer vaccine and gene therapy.【Objectives】(1) To screen out the COX-2 related molecules in gastric carcinogenesis and progression. (2) To investigate the relation between 15-PGDH and COX-2 in gastric cancer. (3) To study the expression status of 15-PGDH in gastric cancer and correlate it with clinicopathological parameters. (4) To explore the role of 15-PGDH in the malignant behavior of gastric cancer.【Methods】(1) 2-DE and the PDQuest software analysis were applied to compare the differential expression proteins of SGC7901 transfected with COX-2 siRNA vector. The differential protein dots were excised and further analyzed by MALDI-TOF-MS. Corresponding peptide mass figureprints (PMFs) were analyzed by bioinformatics analysis. (2) COX-2 and 15-PGDH expression in COX-2 siRNA vector transfectants and full-length cDNA vector transfectants were determined by Western blot and immunocytochemistry assay. (3) Immunohistochemistry and Western blot were used to determine the expression status of both COX-2 and 15-PGDH in gastric cancer tissues. (4) Statistic analyses were explored to value the correlation of COX-2 and 15-PGDH expression and their relations with the clinicopathological parameters. (5) 15-PGDHsiRNA vector and sense expression plasmid was constructed and transfected into the gastric cancer cell line MKN45 and SGC7901 with the mock vector as the control. The stable transfectants were screened out successfully. (6) The effects of 15-PGDH on the proliferation and in vivo tumor growth of gastric cancer cells were respectively investigated by MTT assay, plan plate colony formation assay, soft agar colony formation assay, flow cytometry and nude mice tumoration. (7) Screening candidate molecules regulated by COX-2 with oligonucleotide array-based transcription factor assay (OTAFA) and gene expression profiling microarray.【Results】1. 14 differential expressed proteins were identified by proteomics.Western blotting analysis confirmed that the stable SGC7901 clones transfected with COX-2siRNA showed lower COX-2 expression compared to empty vector-transfected SGC7901-pSilencer cells. Among them, the COX-2 expression of the first stable clone (siRNA1) was the lowest in all clones, so the first stable clone was used to do the further study. 2-DE was used to identify the differentially expressed proteins in the two cell lines. Twenty-two differentially expressed protein spots have been identified and subjected to MALDI-TOF MS. Fourteen PMFs were obtained and submitted to database query. Compared with SGC7901-pSilencer, the expressional levels of 7 proteins, namely Cell division cycle protein 16 homolog (CDC16), Actin cytoplasmic 2 (ACTG), 14-3-3 protein sigma (1433S), Vimentin (VIME), Heat shock cognate 71kDa protein (HSP7C), Histamine N-methyltransferase (HNMT) and Ras-related protein Rab-3B (RAB3B) decreased in SGC7901-COX-2/siRNA, while the expression levels of the other 7 proteins, namely Lactoylglutathione lyase (LGUL), 15-hydroxyprostaglandin dehydrogenase [NAD+] (15-PGDH), Complement component 1 Q subcomponent-binding protein precursor (C1QBP), Growth differentiation factor 2 precursor (GDF2), Heat shock 70kDa protein 5 (GRP78), Phosphatidylinositol transfer protein beta isoform (PIPNB) and Stress-70 protein (GRP75) increased in SGC7901-COX-2/siRNA. Further Western blot and immunocytochemical assay verified that the protein level of 15-PGDH was much higher in SGC7901-COX-2/siRNA with the lower expression of COX-2 than that in SGC7901-pSilencer.2. 15-PGDH expression was down-regulated in gastric cancer tissues and was negatively correlated to COX-2 expression.Expression levels of COX-2 and 15-PGDH were examined by Western blotting in gastric cancer tissues and adjacent non-tumor tissues taken from 8 patients. 15-PGDH expression was found decreased or even absent while COX-2 was over-expressed in gastric cancer tissues compared with the corresponding non-tumor tissues. Meanwhile, immunohistochemistry was conducted in 55 cases of gastric cancer specimens and normal mucosa. 15-PGDH expression was also found decreased or even absent while COX-2 was over-expressed in gastric cancer tissues compared with the corresponding non-tumor tissues. Further analysis about their relationship with clinicopathological parameters revealed that the level of 15-PGDH was significantly lower in patients of III-IV stage than in I-II stage (p<0.05). Its expression had significant difference among differentiation grade (p<0.05) and was related with lymph node metastasis (p<0.01). However, COX-2 expression just had the opposite pattern. Spearman analysis showed that significant negative correlation between COX-2 and 15-PGDH immunoreactivity with rs=-0.564 (p<0.01). Western blotting analysis confirmed that the stable SGC7901clones transfected with pcDNA3.1-COX-2 showed higher COX-2 expression and much lower 15-PGDH compared to empty vector-transfected SGC7901 cells. All these results demonstrated that 15-PGDH was down-regulated in gastric cancer tissues, especially decreased in those cases of lowly differentiation, late TNM stage and with lymph node metastasis. Expression of 15-PGDH has a negative correlation with that of COX-2. Up-regulation of COX-2 could suppress the expression of 15-PGDH in vitro, which was consistent with above findings by 2-DE.3. Exogenous upregulation of 15-PGDH could partially reverse the malignant phenotype of gastric cancer while knock-down of its expression by siRNA stimulate proliferation in gastric cancer cells.15-PGDH expression was detected in 5 kinds of human gastric cancer cell lines (SGC7901,AGS,BGC823,MKN45,MKN28) and human normal gastric epithelial cell line GES. Its expression decreased in all the gastric cancer cell lines compared to GES, especially with low level in SGC7901 cells. cDNA eukaryotic expression vector from Prof. Ido Wolf (Israel) was stably transfected into SGC7901 in order to up-regulate 15-PGDH expression. SGC7901 cells stably transfected with the 15-PGDH cDNA were found to exhibit significantly lower rate of proliferation than empty vector transfectants. 15-PGDH suppresses tumor cell growth in soft agar and tumor formation in athymic nude mice. Flow cytometry detected more cells blocked in G1 stage. Meanwhile, 15-PGDH siRNA vector was constructed and stably transfected into MKN45 cells which expressed relatively high level of 15-PGDH. MKN45 cells stably transfected with the 15-PGDHsiRNA were found to exhibit significantly proliferation and stimulate growth of MKN45 cells in soft agar and tumor formation in athymic nude mice than empty vector transfectants. In conclusion, 15-PGDH participates in many gastric cancer malignant behaviors such as proliferation. Exogenous upregulation of 15-PGDH could partially reverse the gastric cancer malignant phenotype.4. The candidate molecules regulated by COX-2 were screened with oligonucleotide array-based transcription factor assay (OTAFA) and gene expression profiling microarray.Nucleoprotein was extracted from the SGC7901 cells transfected with COX-2siRNA plasmid or empty vector with the kit of Pierce biotechnology. OATFA was then performed to detect the activity variation of transcription factors. Spot intensity 1.5 fold increased or decreased was used as the criteria for differential expression of transcription factors. Compared with the control cells, 4 up-regulated transcription factors and 2 down-regulated transcription factors were detected in SGC7901-COX-2/siRNA cells.Total RNA of SGC7901-COX-2/siRNA and SGC7901-pSilence cells was prepared using Trizol according to the manufacturer's instructions. After array hybridization, image scanning and data analysis, candidates with the filtering criteria of intensity 2 fold increased or decreased were selected out. Compared to the control cells, 1040 up-regulated genes and 234 down-regulated genes were found in SGC7901-COX-2/siRNA cells.【Conclusions】(1) Differential expression proteins related to COX-2 mediated gastric carcinogenesis and progression were screened out. These proteins function in the process of cell cycle regulation, motility, differentiation, apoptosis, proliferation as well as protein transportation, and may play critical roles in gastric carcinogenesis, progression and metastasis. (2) Follow-up verification and functional study revealed that 15-PGDH is negatively regulated by COX-2, and may play its role in gastric carcinogenesis and progression as an onco-suppressor. Up-regulation of 15-PGDH expression by sense transfection could partially reverse the malignant phenotype of gastric cancer. (3) Differential expression genes and transcription factors were successfully screened out by gene expression profiling microarray and OTAFA. Further studies may yield novel clues for elucidating the mechanisms of COX-2 mediated gastric carcinogenesis.