Atsushi Ochiai

The E-cadherin gene is silenced by CpG methylation in human hepatocellular carcinomas

Our study was designed to clarify the significance of silencing the E‐cadherin gene, which is located on 16q22.1, due to CpG methylation during hepatocarcinogenesis. The CpG methylation status of primary hepatocellular carcinomas (HCCs) and corresponding liver tissues showing chronic hepatitis or cirrhosis, which are widely considered to be precancerous conditions, were assessed by digesting DNA with methylation‐sensitive and non‐sensitive restriction enzymes. CpG methylation around the promoter region of the E‐cadherin gene was detected in 46% of liver tissues showing chronic hepatitis or cirrhosis and 67% of HCCs examined. Immunohistochemical examination revealed reduced E‐cadherin expression in 59% of HCCs examined. CpG methylation around the promoter region correlated significantly with reduced E‐cadherin expression in HCCs (p < 0.05). CpG methylation around the promoter region, which increases during the progression from a precancerous condition to HCC, may participate in hepatocarcinogenesis through reduction of E‐cadherin expression, resulting in loss of intercellular adhesiveness and destruction of tissue morphology. Int. J. Cancer 71:355‐359, 1997.

Expression of E-Cadherin, α-Catenin, β-Catenin and Plakoglobin in Esophageal Carcinomas and Its Prognostic Significance

It has been suggested that inactivation of the cadherin-mediated cell-cell adhesion system plays a role in the initial steps of cancer invasion and metastasis. Expression of E-cadherin and its intracy

DNA hypermethylation at the D17S5 locus is associated with gastric carcinogenesis

The aim of this study was to elucidate the significance of aberrant DNA methylation in gastric carcinogenesis. The DNA methylation status at the D17S5 locus, at which a candidate tumor suppressor gene, HIC-1, was identified, of gastric cancers and non-cancerous gastric mucosae from 42 gastric cancer patients was examined by Southern blotting using a methylation-sensitive restriction enzyme. DNA hypermethylation was observed in 15, 13, 25 and 45% of the tissues showing no remarkable histological findings, chronic gastritis without intestinal metaplasia, intestinal metaplasia and gastric cancer, respectively. The incidence of DNA hypermethylation was significantly higher in gastric cancers than in non-cancerous gastric mucosae (P < 0.05). DNA hypermethylation was often accompanied by allelic loss at the same locus in gastric cancers. DNA hypermethylation at the D17S5 locus, which was even detected in precancerous conditions, including intestinal metaplasia, may play a role in gastric carcinogenesis.

Recurrent esophageal carcinoma after esophagectomy with three‐field lymph node dissection

To evaluate the effect of the extended lymphadenectomy for thoracic esophageal carcinoma, the pattern of recurrence in the 50 patients with pT3 tumors who underwent esophagectomy with cervical, mediastinal, and abdominal lymph node dissection (3‐F) (group A) was compared with that of 100 patients at pT3 who underwent esophagectomy without upper mediastinal and cervical lymphadenectomy (2‐F) (group B). The cumulative 5‐year survival rate for 115 patients who underwent 3‐F was 50.9%. Cumulative 5‐year survival rates for patients in groups A and B were 36.8% and 22.0%, respectively. The survival curve for group A was significantly better than group B (P = 0.02332). Lymphatic recurrence was noted less frequently in group A (8/23) than in group B (31/49) (χ2 = 5.1149), whereas the rate of hematogenous recurrence was similar. Extension of the field of lymph node dissection reduced the lymph node recurrence in patients with thoracic esophageal carcinoma, which may have positively affected patient survival.

New prognostic histological parameter of invasive ductal carcinoma of the breast: clinicopathological significance of fibrotic focus.

Immunohistochemistry, DNA ploidy analysis and molecular genetics have made it possible to predict the outcome of breast cancer more precisely than routine histological examination alone. However, in routine practice, it is difficult to incorporate these methodologies in all cases. If certain histological parameters can accurately predict the outcome of patients with breast cancer, they would be more practical for routine use. We showed that the presence of fibrotic focus (FF) in invasive ductal carcinoma (IDC) is closely associated with c‐erbB‐2 or p53 protein expression, high proliferative activity, and high angiogenesis of the tumors. Furthermore, multivariate analyses with well‐known prognostic parameters for IDC demonstrated that the presence of FF is the most useful independent parameter to predict IDC patient outcome. In addition, our data suggested that the interaction between tumor cells and stromal fibroblasts may play an important role in the formation of FF in IDC based on growth factor and growth factor receptor protein expression in the tumor cells and fibroblasts forming FF. Based on the results of our clinicopathological studies, we propose a new prognostic classification scheme for the prediction of IDC patient outcome, which consists of FF, nuclear atypia, and fat invasion. This classification has superior predicting power to existing prognostic classifications.

Pleiotropic phenotypic expression in cybrids derived from mouse teratocarcinoma cells fused with rat myoblast cytoplasts

Cybrid clones were isolated by fusing mouse embryonal carcinoma (PCC4) cells with cytoplasts of rat myoblastic cells (L6TG X CAPr). Although some clones were similar to PCC4 (Type II), a high proportion (88%) were differentiated; the differentiated cells had a mesh-like arrangement (Type I) or were flat with many projections (Type III). Protein patterns of both Type I and Type III cells changed markedly from that of PCC4 cells. Type III cells lacked alkaline phosphatase and expressed endo A and B proteins predominantly. One Type III clone produced alpha-fetoprotein and plasminogen activator (visceral endoderm-like), while another clone consisted of trophectodermal cell-like giant cells. Therefore it was shown that introduction of the somatic cell cytoplasm induces differentiation of teratocarcinoma stem cells, suggesting a cytoplasmic element (or elements) regulating gene expression.

Expression of Cadherin-Catenin Cell Adhesion Molecules, Phosphorylated Tyrosine Residues and Growth Factor Receptor-tyrosine Kinases in Gastric Cancers

Tyrosine phosphorylation of β‐catenin, an intracytoplasmic E‐cadherin‐binding protein, has been shown to disrupt the cadherin‐mediated cell adhesion system in vitro. In order to investigate the relationships of expression and tyrosine phosphorylation of cadherin‐catenin molecules and expression of growth factor receptor‐tyrosine kinase with loose cell‐to‐cell adhesion, immunohistochemical staining for E‐cadherin, α‐ and β‐catenin, phosphorylated tyrosine residues and tyrosine kinase receptors, including c‐erbB‐2, epidermal growth factor‐receptor (EGF‐R), c‐met and K‐sam, in 17 undifferentiated‐ and 10 differentiated‐type human gastric cancers was performed. Loss or reduced expressions of E‐cadherin and α‐ and β‐catenin (11, 11, 10 cancers, respectively) were observed in the former, but not the latter. Diffuse cytoplasmic staining of E‐cadherin, α‐ and β‐catenin and phosphotyrosine residues was observed frequently in the undifferentiated‐type cancers. The cytoplasmic localization of phosphotyrosine residues in undifferentiated‐type cancers was correlated significantly with K‐sam expression (P<0.01) and diffuse cytoplasmic staining of E‐cadherin (P<0.05) and β‐catenin (P<0.05). Expression of K‐sam protein was detected significantly more frequently in undifferentiated‐ (6/17; P<0.05) than differentiated‐type adenocarcinomas whereas the converse applied to c‐erbB‐2 expression (8/10 of the latter, P<0.05). Tyrosine phosphorylation of β‐catenin was directly confirmed in the protein extracts of one undifferentiated‐type gastric cancer. These data indicate that alteration of tyrosine phosphorylation status associated with K‐sam expression may cause the cytoplasmic distribution of cadherin‐catenin molecules and loose cell‐cell adhesion in undifferentiated‐type gastric cancers.

Genetic Alterations in the Precursors of Gastric Cancer

The presence of multiple genetic alterations has been established in human cancers. The evolution of sequential histological changes from normal cells through invasive cancer provides the opportunity to elucidate the molecular basis of this progression. The paradigm for such studies involves the identification of molecular genetic events that occur during the evolution of various human carcinomas, including gastric cancer. Several studies have demonstrated that multiple genetic alterations, comprising oncogene and tumor-suppressor gene mutations and loss of heterozygosity, play roles in cancer development and progression. Two histological forms of gastric carcinoma (intestinal and diffuse) and their different histogenetic and etiological backgrounds suggest that different spectra of genetic changes are involved in different tumor types. That same genetic alterations, including oncogenes and tumorsuppressor genes, have been detected in the precursors of both intestinal- and diffuse-type gastric cancers, but no specific genetic alterations have been detected in the latter. The genetic alterations in precursor lesions of gastric cancer and chronic atrophic gastritis, a precancerous condition, are summarized here. Also presented are the results of investigations into p.53 tumor-suppressor gene mutations in nonneoplastic gastric mucosa exhibiting chronic atrophic gastritis with intestinal metaplasia, an early genetic change.

In vitro reconstitution of human respiratory epithelium

Dear Editor: Data concerning the reconstruction of respiratory epithelium by use of the human cells cultivated in vitro are still scanty. In view of the recent increasingly clear indication that human and animal cells are different in various respects (Emura et al., 1996; Gray et al., 1996; Jacob et al., 1996) more effort should be made than before to carry out experiments in which human cells are used in parallel with animal cells. In this letter to the Editor we would like to discuss the results of our preliminary experiments to reconstitute in vitro a human respiratory epithelium in terms of three-dimensional cell-cell contact, formation of polarized configurations, and subsequent cell differentiation. The normal bronchial epithelial explants were obtained from a surgical specimen of a 49-yr-old male patient. They were trimmed in a size of 2-4 mm square in an ice-cold Ca2+-free phosphatebuffeted saline. Thereafter, each piece was transferred in a plastic tissue culture grade Petri dish with the epithelial side downwards. Ten to 20 rain later a small amount of special growth medium (GM) was given to the pieces of explants as described previously (Emura et al., 1985) and the explants were cultivated for 3M0 wk. During this period they were able to be transferred consecutively when the dish became 60-80% confluent, each time yielding a batch of rapidly growing undifferentiated cells. The cells of a later transfer appeared more undifferentiated under a phase-contrast microscope. The medimn (GM) used for collecting and passaging these undifferentiated cells consisted of 250 ml of a low Ca 2+ human foreskin keratinocyte culture medium modified on the basis of MCDB 153 (KGM) (Nissui Pharmaceutical Co., Tokyo), 2.5 gg EGF (Sigma Chemical Co., Tokyo), 2 ml bovine pituitm 7 extract attached to each purchased bottle of KGM, 1.25 ml antibiotics mixture (GIBCO BRL, Tokyo). and 2.5 ml chelatefiltered, heat-inactivated fetal bovine sermn (FBS, approx. 0.005 mM Ca2+). For the tissue reconstruction experiment, the bronchial epithelial cells were harvested front the second explant transfer, showing a considerably undifferentiated phenotype (Fig. 1). The medium (3DM) used for the tissue reconstruction experiment consisted of a mixture of Dulbeecos MEM and RPMI 1640 (1:1) with a standard Ca 2+ concentration 240 ml, an additional 0.6 mg L-glutamine (Nissui)/ml, 10 pg insulin (Sigma)/ml, 0.1 gg hydroeortisone (Sigma)/ml, 0.1 gg cholera toxin (Sigma)/ml, 5 pg transferrin (Sigma) ml, 25 ng EGF (Sigma)/ml, 5 gg ethanolamine (Sigma)/mL 7.2 mg HEPES (Sigma)/ml, and 9.4 gg additional Ca(NOz)~4H~O (Kanto Chemical, Tokyo)/ml, an additional 0.22 mg Na-pyruvate (Sigma)/ml, and 2% heat-inactivated FBS with a standard Ca 2 + concentration. The cells at two passages (not cloned) were suspended at a density of 6.2 x 104/ml in a 100-ml plastic conical flask (Coming, Tokyo) containing 10 ml 3-DM: pH was balanced in a CO2 incubator for 0.5 h; and then the flask with its cap tightly closed was shaken for 4 d on a 37 ° C water bath at 80 cycles/rain. Thereafter, the aggregates together with the medium were transferred into a 25-ml centrifuge tube that was subsequently left to stand for 10 rain, and divided into freely sedimenting coarse aggregate and the remaining fine aggregate fractions. Each fraction was further divided in two identical groups. One group from each fraction received 5 × 105 MRC-5 fibroblasts, which have been known to release a hepatocyte growth factor (HGF) stimulating morphogenesis in various organs (Birchmeier et al., 1995). After thorough mixing, the suspension was centrifuged at 800 ~3?m/ min for 5 rain to remove most of the medium. Then 3 ml collagen (type I) solution mixture (see below) was added to each pellet: the aggregates and free fibroblasts were well suspended allowing incidental contact between the two components; and the suspension was laid on top of a double-layered collagen gel complex, which was prepared 4 d earlier. For the preparation of the collagen gel, an icecold collagen solution (Nitta Gelatine, Osaka), a 10-times concentrated RPMI 1640 (without sodium bicarbonate; GIBCO-BRL, Tokyo) and a mixture of 25 ml 0.08 N NaOH and 1.2 g HEPES plus 0.55 g NaHCO:~ were mixed in a ratio of 8:1:1 in an ice-cold water bath. For the lowermost layer, 1 ml of this collagen mixture was pipetted in a 25-ram Falcon Cell Culture Insert with a Cyclopore PC membrane (3.0 ~m pore size) fitted in a well of a Coming 6-well flat plate. After gelling at 37 ° C in a humidified CO2 incubator, 2 ml of the same collagen mixture with 5 X 105 esophagus-derived stromal fi-