sophageal squamous cell carcinoma (SCC) is one of the most aggressive cancers. Esophageal SCC is relatively common in countries in Eastern Asia, such as China and Japan, and is characterized by poor prognosis and rapid clinical progression, with a high frequency of lymph node metastasis and recurrence. The 5-year survival rate of patients with esophageal SCC showing submucosal invasion is low (40-75%) 1-5) compared with that for patients with colon cancer (over 80%). 4,5) In addition, lymph node metastasis is commonly found in esophageal SCC, even when the tumor invades only the submucosa. Lymph node metastasis is the main cause of the poor prognosis for the patients with esophageal SCC.Consequently, the identification of the genes associated with metastasis of esophageal SCC is very important. Microarray analysis has been used to investigate the gene expression profiles of esophageal SCC tissues and esophageal cancer cell lines with various characteristics.6-10) When clinical materials are used for microarray analysis, it might be necessary to look at the overall expression profiles of genes (if possible) by clustering analysis in the different pathological stages, such as dysplasia, carcinoma in situ, and invasive cancer with or without metastasis. However, to isolate gene(s) related to metastasis, it would be convenient to use cell lines with different metastatic potentials derived from the same parental cells, and a reliable model system is therefore required for examining the metastatic potential of cancer cells.In our previous experiments, we established in vitro and in vivo model systems for studying invasion and metastasis of esophageal SCC cells. 11,12) We first clarified in detail the molecular and genetic characteristics of a human non-metastasizing esophageal SCC cell line, T.Tn, 12) and then developed an orthotopic inoculation model for esophageal cancer cells in nude mice.11) In the present study, we isolated a metastasizing subclone from the parental non-metastasizing T.Tn cell line by in vitro selection and by the use of a nude mouse orthotopic inoculation model. Then, we compared the expression profiles of 9206 genes in the parental T.Tn cells and the metastasizing subclone by cDNA microarray analysis, and identified several genes differentially expressed in the metastasizing subclone. Materials and MethodsCell line. A human esophageal SCC cell line, T.Tn 13) was obtained from JCRB (Japanese Collection of Research Bioresources, Osaka). T.Tn cells were grown in 1:1 mixture of Dulbecco's modified Eagle's medium (Nissui, Tokyo) and F-12 (Life Technologies, Inc., Gaithersburg, MD) supplemented with 10% fetal calf serum (FCS; Sigma, St. Louis, MO), 100 µg/ml streptomycin, 100 units/ml penicillin (Life Technologies, Inc.), and 0.25 µg/ml amphotericin B (Life Technologies, Inc.) in a humidified atmosphere of 95% air and 5% CO 2 at 37°C. HT1080 cells are derived from a human fibrosarcoma cell line known to secrete a large amount of several matrix-degrading enzymes (purchased from Dai-Nippon Seiyaku, Osa...

beta-2-Glycoprotein 1, an abundant plasma glycoprotein, binds anionic cell surfaces and functions as a regulator of thrombosis. Here, we show that cleavage of the kringle domain at Lys317/Thr318 switches its function to a regulator of angiogenesis. In vitro, the cleaved protein specifically inhibited the proliferation and migration of endothelial cells. The protein was without effect on preformed endothelial cell tubes. In vivo, the cleaved protein inhibited neovascularization into subcutaneously implanted Matrigel and Gelfoam sponge implants and the growth of orthotopically injected tumors. Collectively, these data indicate that plasmin-cleaved beta-2-glycoprotein 1 is a potent antiangiogenic and antitumor molecule of potential therapeutic significance.

Maximizing the area under the receiver operating characteristic curve (AUC) is a standard approach to imbalanced classification. So far, various supervised AUC optimization methods have been developed and they are also extended to semi-supervised scenarios to cope with small sample problems. However, existing semi-supervised AUC optimization methods rely on strong distributional assumptions, which are rarely satisfied in real-world problems. In this paper, we propose a novel semi-supervised AUC optimization method that does not require such restrictive assumptions. We first develop an AUC optimization method based only on positive and unlabeled data (PU-AUC) and then extend it to semi-supervised learning by combining it with a supervised AUC optimization method. We theoretically prove that, without the restrictive distributional assumptions, unlabeled data contribute to improving the generalization performance in PU and semi-supervised AUC optimization methods. Finally, we demonstrate the practical usefulness of the proposed methods through experiments.

SUMMARY:Although regenerating gene (Reg) protein is reported to have a trophic effect on gastric epithelial cells, its involvement in human gastric diseases is not clear. We have recently shown that both gastrin and gastric mucosal inflammation enhance Reg gene expression in the fundic mucosa in rats. This study was designed to clarify whether Reg protein is involved in Helicobacter pylori-induced gastritis and whether Reg gene expression is linked to serum gastrin levels in this condition. Mongolian gerbils were inoculated with an H. pylori strain isolated from a gastric cancer patient. Four weeks later, some of the gerbils with H. pylori infection were eradicated by lansoprazole, amoxicillin, and clarithromycin. The time courses of changes in Reg gene expression, serum gastrin levels, gastric acidity, and histopathologic factors were examined. Four weeks after H. pylori infection, gastritis started spreading to the fundic mucosa, and gastric acidity started reducing. Serum gastrin levels and Reg mRNA expression in the fundus were significantly increased 6 weeks after infection. Reg mRNA expression in the fundus correlated significantly with both serum gastrin levels and the severity of fundic mucosal inflammation. After H. pylori eradication, serum gastrin levels and fundic mucosal inflammation were normalized, and the increase in Reg mRNA expression was abolished. The Reg gene is associated with hypergastrinemia and fundic mucosal inflammation and may be involved in H. pylori-induced gastritis. (Lab Invest 2003, 83:1777-1786.