The novel stem cell marker SALL4 has been identified as a diagnostic marker of germ cell tumors, especially yolk sac tumors, in gonadal organs. To clarify the significance of SALL4 as an oncofetal protein, we investigated SALL4 expression by immunohistochemistry in non-neoplastic stomach and gastric carcinoma with particular emphasis on á-fetoprotein (AFP)-producing gastric carcinoma, as AFP-producing gastric carcinoma shares expression of AFP and glypican 3 (GPC3) with yolk sac tumors and hepatic neoplasms. A total of 338 gastric carcinomas, 60 hepatocellular carcinomas, and 48 cholangiocellular carcinomas were studied by immunohistochemistry on tissue microarrays. In addition, more detailed whole tissue section immunohistochemistry was performed on non-neoplastic gastric tissue from 5 adult and 8 fetal specimens, 6 hepatoblastomas, and 31 cases of AFP-producing gastric carcinomas. SALL4 expression was observed in the neofetal stomach in gestational week 9 and disappeared thereafter. It was also identified by tissue microarray study in a fraction of gastric carcinomas (51 of 338, 15%), associated with older age (P=0.0001), male sex (P=0.0033), intestinal-type histology (P=0.0001), and synchronous liver metastasis (P=0.0047). AFP and GPC3 were closely associated with SALL4 expression in gastric carcinoma (both, P<0.0001), and a full-section study indicated that SALL4 was positive in all 31 cases of AFP-producing gastric carcinoma with diffuse staining in 24 cases (78%). Diffuse SALL4 expression was observed in the histologic patterns of hepatoid (89%), glandular (57%), and clear cell (39%) AFP-producing gastric carcinoma. In addition, SALL4 expression was completely negative in hepatoblastoma (n=6) and hepatocellular carcinoma (n=60). SALL4 is an oncofetal protein similar to AFP and GPC3, but it represents fetal gut differentiation in gastric carcinoma. SALL4 is a sensitive marker for AFP-producing gastric carcinoma and is especially useful to distinguish hepatoid gastric carcinoma from hepatocellular carcinoma.

Lymph node metastasis occurs in as many as 16% of patients with submucosal invasive colorectal carcinoma. We investigated the association between histopathological factors and lymph node metastases in 322 consecutive patients with submucosal invasive colorectal carcinoma who had undergone radical colectomy with lymph node dissection to detect patients at high risk of lymph node metastasis without measuring the depth of submucosal invasion. Lymph node metastasis was found in 46 (14.3%) of 322 patients with submucosal invasive colorectal carcinoma. Univariate analysis showed that each of the following histopathological factors had a significant influence on lymph node metastasis: invasion depth, lymphatic invasion, venous invasion, tumor differentiation, growth pattern of the intramucosal tumor component, complete disruption of the muscularis mucosa due to tumor invasion, and tumor budding at the submucosal invasive front. Multivariate analysis showed that lymphatic invasion (Po0.01), tumor differentiation (Po0.01), and tumor budding (Po0.01) were significantly associated with lymph node metastasis. All 46 cases of lymph node metastasis showed at least one of the following findings: lymphatic invasion, moderately or poorly differentiated tumor grade, tumor budding, or complete disruption of the muscularis mucosa due to tumor invasion. Patients with submucosal invasive colorectal carcinoma that show at least one of three factorslymphatic invasion, moderately or poorly differentiated tumor grade, or tumor budding-are at high risk for lymph node metastasis. All of the patients with lymph node metastasis, who did not have any of these factors, showed a completely disrupted muscularis mucosa.

Inositol 1,4,5-trisphosphate (IP 3 ) 1 is a second messenger generated by the phosphatidylinositol signaling cascade response to hormones, neurotransmitters, and growth factors (1) and causes Ca 2ϩ release from intracellular stores by binding to the IP 3 receptor (IP 3 R), which is an IP 3 -gated intracellular Ca 2ϩ release channel. Ca 2ϩ release in the cytoplasm occurs in complex spatial and temporal patterns, such as Ca 2ϩ waves and Ca 2ϩ oscillations, and regulates many cellular responses, including fertilization, muscle contraction, secretion, cell growth, differentiation, apoptosis, and synaptic plasticity (1).The IP 3 R family consists of three isoforms (IP 3 R1, IP 3 R2, and IP 3 R3) (2), and the primary structure of all three IP 3 R types has been determined in rat (3-5) and human (6 -9). Only a single type of IP 3 R was identified in frog (10), fly (11), starfish (12), lobster (13), and nematode (14). Three isoforms have been found to exist in the mouse (15, 16), but the primary structure of mouse IP 3 R2 and IP 3 R3 has not yet been determined.IP 3 -gated Ca 2ϩ release channels are composed of four subunits (17). The expression levels of the three IP 3 R isoforms are different in each tissue, but most tissues contain multiple isoforms (18). Heterotetrameric channels were detected as well as homotetrameric channels (19), indicating that the structural diversity of the IP 3 -gated channels is greater than the number of genes. There is evidence of functional differences among the three types of IP 3 R in terms of their IP 3 sensitivity (20, 21) and the modulatory effects on them from cytoplasmic Ca 2ϩ (22), ATP (22), calmodulin (23), and cAMP-dependent protein kinase (24). Additional diversity of IP 3 -gated channels is produced by alternative splicing, and alternative splicing segments, designated SI, SII, and SIII, have been found in the mouse (25), rat (26), and human (9) IP 3 R1. These findings suggest that the channels composed of different isoforms possess distinctive functions, but little is known about the physiological significance and exact functional differences arising from the heterotetrameric assembly of IP 3 R.In many cells, the Ca 2ϩ increase triggered by IP 3 is propagated throughout the cytoplasm; however, the diffusion of free Ca 2ϩ is spatially restricted because many immobile or slowly diffusing Ca 2ϩ -binding proteins are present in the cytoplasm (27). Propagation of Ca 2ϩ is thought to be mediated by regenerative Ca 2ϩ release through IP 3 Rs, which are regulated by cytoplasmic Ca 2ϩ (28). Positive feedback regulation of IP 3 R by Ca 2ϩ enables the Ca 2ϩ released by one receptor to excite its

The AT-rich interactive domain 1A gene (ARID1A), which encodes one of the subunits in the Switch/Sucrose Nonfermentable chromatin remodeling complex, carries mutations and is responsible for loss of protein expression in gastric carcinoma, particularly with Epstein-Barr virus (EBV) infection and a microsatellite instability-high phenotype. We used immunohistochemistry to investigate the significance of ARID1A loss in 857 gastric carcinoma cases, including 67 EBV(+) and 136 MLH1-lost gastric carcinomas (corresponding to a microsatellite instability-high phenotype). Loss of ARID1A expression was significantly more frequent in EBV(+) (23/67; 34 %) and MLH1-lost (40/136; 29 %) gastric carcinomas than in EBV(-)MLH1-preserved (32/657; 5 %) gastric carcinomas (P < 0.01). Loss of ARID1A correlated with larger tumor size, advanced invasion depth, lymph node metastasis, and poor prognosis in EBV(-)MLH1-preserved gastric carcinoma. A correlation was found only with tumor size and diffuse-type histology in MLH1-lost gastric carcinoma, but no correlation was observed in EBV(+) gastric carcinoma. Loss of ARID1A expression in EBV(+) gastric carcinoma was highly frequent in the early stage of gastric carcinoma, although EBV infection did not cause downregulation of ARID1A: EBV-positive nasopharyngeal carcinomas (n = 8) and lymphomas (n = 15) failed to show loss of ARID1A, and EBV infection did not cause loss of ARID1A in gastric carcinoma cell lines. Taken together, loss of ARID1A may be an early change in carcinogenesis and may precede EBV infection in gastric epithelial cells, while loss of ARID1A promotes cancer progression in gastric cancer cells without EBV infection or loss of MLH1 expression. Loss of ARID1A has different and pathway-dependent roles in gastric carcinoma.

Recent studies have revealed that pregnane X receptor (PXR) can function as a master regulator to control the expression of phase I and phase II drug-metabolizing enzymes, as well as members of the drug transporter family, including multiple drug resistance (MDR) 1, which has a major role in multidrug resistance. Previously, we have demonstrated that steroid/xenobiotics metabolism by tumor tissue through the PXR-cytochrome P-450 3A (CYP3A) pathway might play an important role in endometrial cancer. In this study, we examined which endocrine-disrupting chemicals (EDCs) and anticancer agents might be ligands for PXR and whether these chemicals enhanced PXR-mediated transcription through two different PXR-responsive elements (PXREs), CYP3A4 and MDR1, in endometrial cancer cell lines. Some steroids/EDCs strongly activated PXR-mediated transcription through the CYP3A4-responsive element compared with the MDR1-responsive element, whereas these steroids/EDCs also enhanced the CYP3A4 expression compared with the MDR1 expression. In contrast, the anticancer agents, cisplatin and paclitaxel, strongly activated PXR-mediated transcription through the MDR1-responsive element compared with the CYP3A4-responsive element, whereas these drugs also enhanced the MDR1 expression compared with the CYP3A4 expression. We also analyzed how these ligands regulated PXR-mediated transcription through two different PXREs. In the presence of PXR ligands, there was no difference in the DNA binding affinity of the PXR/retinoid X receptor heterodimer to each PXRE, but there were different interactions of the coactivator to each PXR/PXRE complex. These data suggested that PXR ligands enhanced PXR-mediated transcription in a ligand- and promoter-dependent fashion, which in turn differentially regulated the expression of individual PXR targets, especially CYP3A4 and MDR1.