Genome-wide analysis of DNA copy number alterations and gene expression in gastric cancer

Genomic copy number aberrations (CNAs) are believed to play a major role in the development and progression of human cancers. Although many CNAs have been reported in gastric cancer, their genome-wide transcriptional consequences are poorly understood. In this study, to reveal the impact of CNAs on genome-wide expression in gastric cancer, we analysed 30 cases of gastric cancers for their CNAs by array comparative genomic hybridization (array CGH) and 24 of these 30 cases for their expression profiles by oligonucleotide-expression microarray. We found that with the application of laser microdissection, most CNAs were detected at higher frequency than in previous studies. Notably, gain at 20q13 was detected in almost all cases (97%), suggesting that this may play an important role in the pathogenesis of gastric cancer. By comparing the array CGH data with expression profiles of the same samples, we showed that both genomic amplification and deletion strongly influence the expression of genes in altered genomic regions. Furthermore, we identified 125 candidate genes, consisting of 114 up-regulated genes located in recurrent regions (>10%) of amplification and 11 down-regulated genes located in recurrent regions of deletion. Up-regulation of several candidate genes, such as CDC6, SEC61G, ANP32E, BYSL and FDFT1, was confirmed by immunohistochemistry. Interestingly, some candidate genes were localized at genomic loci adjacent to well-known genes such as EGFR, ERBB2 and SMAD4, and concordantly deregulated by genomic alterations. Based on these results, we propose that our list of candidate genes may contain novel genes involved in the pathogenesis of advanced gastric cancer.

Genome-wide microRNA expression profiling in renal cell carcinoma: significant down-regulation of miR-141 and miR-200c

We investigated expression profiles of microRNA (miRNA) in renal cell carcinoma [clear cell carcinomas (CCC) and chromophobe renal cell carcinomas (ChCC)] and in normal kidneys by using a miRNA microarray platform which covers a total of 470 human miRNAs (Sanger miRBase release 9.1). Unsupervised hierarchical cluster analysis revealed that CCC and ChCC were separable and that no subgroups were identified in CCCs. We found that 43 miRNAs were differentially expressed between CCC and normal kidney, of which 37 were significantly down-regulated in CCC and the other 6 were up-regulated. We also found that 57 miRNAs were differentially expressed between ChCC and normal kidney, of which 51 were significantly down-regulated in ChCC and the other 6 were up-regulated. Together, these observations indicate that expression of miRNAs tends to be down-regulated in both CCC and ChCC compared with normal kidney. We observed that miR-141 and miR-200c were the most significantly down-regulated miRNAs in CCCs. Indeed, in all cases of CCC analysed, both miR-141 and miR-200c were down-regulated in comparison with normal kidney. Microarray data and quantitative RT-PCR showed that these two miRNAs were expressed concordantly. TargetScan algorithm revealed that ZFHX1B mRNA is a hypothetical target of both miR-141 and -200c. We established by quantitative RT-PCR that, in CCCs in which miR-141 and miR-200c were down-regulated, ZFHX1B, a transcriptional repressor for CDH1/E-cadherin, tended to be up-regulated. Furthermore, we found that overexpression of miR-141 and miR-200c caused down-regulation of ZFHX1B and up-regulation of E-cadherin in two renal carcinoma cell lines, ACHN and 786-O. On the basis of these findings, we suggest that down-regulation of miR-141 and miR-200c in CCCs might be involved in suppression of CDH1/E-cadherin transcription via up-regulation of ZFHX1B.

High-resolution analysis of DNA copy number alterations and gene expression in renal clear cell carcinoma

We analysed chromosomal copy number aberrations (CNAs) in renal cell carcinomas by array-based comparative genomic hybridization, using a genome-wide scanning array with 2304 BAC and PAC clones covering the whole human genome at a resolution of roughly 1.3 Mb. A total of 30 samples of renal cell carcinoma were analysed, including 26 cases of clear cell carcinoma (CCC) and four cases of chromophobe renal cell carcinoma (ChCC). In CCCs, gains of chromosomes 5q33.1-qter (58%), 7q11.22-q35 (35%) and 16p12.3-p13.12 (19%), and losses of chromosomes 3p25.1-p25.3 (77%), 3p21.31-p22.3 (81%), 3p14.1-p14.2 (77%), 8p23.3 (31%), 9q21.13-qter (19%) and 14q32.32-qter (38%) were detected. On the other hand, the patterns of CNAs differed markedly between CCCs and ChCCs. Next, we examined the correlation of CNAs with expression profiles in the same tumour samples in 22/26 cases of CCC, using oligonucleotide microarray. We extracted genes that were differentially expressed between cases with and without CNAs, and found that significantly more up-regulated genes were localized on chromosomes 5 and 7, where recurrent genomic gains have been detected. Conversely, significantly more down-regulated genes were localized on chromosomes 14 and 3, where recurrent genomic losses have been detected. These results revealed that CNAs were correlated with deregulation of gene expression in CCCs. Furthermore, we compared the patterns of genomic imbalance with histopathological features, and found that loss of 14q appeared to be a specific and additional genetic abnormality in high-grade CCC. When we compared the expression profiles of low-grade CCCs with those of high-grade CCCs, differentially down-regulated genes tended to be localized on chromosomes 14 and 9. Thus, it is suggested that copy number loss at 14q in high-grade CCC may be involved in the down-regulation of genes located in this region.

[Pulmonary benign metastasizing leiomyoma from uterine myoma; report of a case]

A 46-year-old female with benign metastasizing leiomyoma was reported. The patient had undergone a lobectomy of thyroid for adenomatous goiter at the age of 30, and a myomectomy for uterine myoma and a modified radical masmectomy for breast cancer at the age of 32. The chest X-ray on health screening revealed multiple nodules in the bilateral lung. Computed tomography (CT) revealed 6 nodules in the right side and 3 nodules in the left side of the lung. The tumors of the right lung were resected under thoracoscopic surgery. Pathologically these lesions were diagnosed as benign metastasizing leiomyoma. Left lung nodules have been followed-up.

Expression of RUNX3 protein in human gastric mucosa, intestinal metaplasia and carcinoma

BACKGROUND

Human runt-related transcription factor gene 3 (RUNX3) is considered as a possible candidate of tumour suppressor gene in human gastric carcinoma.

MATERIALS AND METHODS

To investigate the RUNX3 protein expression in human gastric mucosa and carcinoma, we generated a polyclonal antibody, AS251, which recognized amino acid sequences from 251 to 266 of human RUNX3. The AS251 antibody was immunoreactive with only RUNX3 protein, but not with RUNX1 and RUNX2. The AS251-antibody was available for Western blotting and immunohistochemistry using paraffin-embedded specimens.

RESULTS

Western-blot analysis revealed that three (MKN-1, -7 and -45) of six human gastric carcinoma cell lines variably expressed RUNX3 protein, consistent with the expression pattern of RUNX3 mRNA reported previously by Li et al. (Cell 2002;109:113-24). Immunohistochemistry disclosed RUNX3 protein in most chief cells and a few gastrin-containing G cells in normal mucosa, but not in intestinal metaplasia and carcinoma cells.

CONCLUSIONS

These data suggest that RUNX3 may play a physiologic role in chief cells and G cells in gastric mucosa, and that suppression of RUNX3 expression in intestinal metaplasia and carcinoma of human stomach may be implicated in gastric carcinogenesis.

Identification of a novel human ankyrin-repeated protein homologous to CARP

We cloned a novel ankyrin repeat protein, Arpp, by immunoscreening a cDNA library constructed from a human esophageal carcinoma cell line, TE1, with an antibody directed to a hypothetical protein encoded by antisense p53 mRNA. Arpp protein is composed of 333 amino acids and contains four ankyrin-like repeat motifs in the middle portion of the protein, a PEST-like sequence and a lysine-rich sequence similar to a nuclear localization signal in the N-terminal region, and a proline-rich region containing consensus phosphorylation sites in the C-terminal region. Protein sequence analysis revealed that Arpp is homologous (52.7% identity) to Carp which is shown to be involved in the regulation of the transcription of the cardiac ventricular myosin light chain 2 gene. Arpp mRNA was found to be expressed in normal skeletal and cardiac muscle. Interestingly, Arpp expression was detectable in bilateral ventricles, but undetectable in bilateral atria and large vessels, suggesting that Arpp may play a specific function in cardiac ventricles as well as skeletal muscles.

Separation and characterization of carboxyl-linked glucuronides of bile acids in incubation mixture of rat liver microsomes

The carboxyl-linked 24-glucuronides of common bile acids have been identified by means of liquid chromatography (LC)/atmospheric pressure chemical ionization (APCI)-mass spectrometry (MS) in an incubation mixture with a male Wistar rat liver microsomal fraction. The authentic specimens of bile acid 24-glucuronide acetate-methyl esters were synthesized unequivocally using the Mitsunobu reaction, and the APCI-mass spectrometric properties of these glucuronide derivatives were also characterized. After incubation of common unconjugated bile acids with hepatic microsomes, glucuronides were extracted and purified with a Sep-Pak C18 cartridge and lipophilic ion exchange gel, piperidino-hydroxypropyl Sephadex LH-20, and then derivatized into the acetate-methyl esters. Subsequent resolution into alpha- and beta-isomers at the glucuronosyl linkage was attained by LC on Cosmosil 5C8 and Sumichiral OA-2500 columns using 200 mM ammonium acetate (pH 7.0)-methanol (1:4, v/v), where 24-glucuronides were monitored with characteristic positive ions [M + NH4]+. The 24-glucuronides of lithocholic, chenodeoxycholic, deoxycholic, ursodeoxycholic and cholic acid were definitely characterized, in contrast to no formation of corresponding 3-glucuronides.