Gene expression analysis in colorectal cancer using practical DNA array filter

Siva‑1 regulates multidrug resistance of gastric cancer by targeting MDR1 and MRP1 via the NF‑κB pathway

Siva-1 is a well-known anti-apoptosis protein that serves a role in multiple types of cancer cells. However, whether Siva-1 affects multidrug resistance via the NF-κB pathway in gastric cancer is currently unknown. The present study aimed to determine the possible involvement of Siva-1 in gastric cancer anticancer drug resistance in vitro. A vincristine (VCR)-resistant KATO III/VCR gastric cancer cell line with stable Siva-1 overexpression was established. The protein expression levels of Siva-1, NF-κB, multidrug resistance 1 (MDR1) and multidrug resistance protein 1 (MRP1) were detected via western blotting. The effect of Siva-1 overexpression on anticancer drug resistance was assessed by measuring the 50% inhibitory concentration of KATO III/VCR cells to VCR, 5-fluorouracil and doxorubicin. The rate of doxorubicin efflux and apoptosis were detected by flow cytometry. Additionally, colony formation, wound healing and Transwell assays were used to detect the proliferation, migration and invasion of cells, respectively. The results of the current study revealed that the Siva-1-overexpressed KATO III/VCR gastric cancer cells exhibited a significantly decreased sensitivity to VCR, 5-fluorouracil and doxorubicin. The results of flow cytometry revealed that the percentage of apoptotic cells decreased following overexpression of Siva-1. The colony formation assay demonstrated that cell growth and proliferation were significantly promoted by Siva-1 overexpression. Additionally, Siva-1 overexpression increased the migration and invasion of KATO III/VCR cells in vitro. Western blot analysis determined that Siva-1 overexpression increased NF-κB, MDR1 and MRP1 levels. The current study demonstrated that overexpression of Siva-1, which functions as a regulator of MDR1 and MRP1 gene expression in gastric cancer cells via promotion of NF-κB expression, inhibited the sensitivity of gastric cancer cells to certain chemotherapies. These data provided novel insight into the molecular mechanisms of gastric cancer, and may be of significance for the clinical diagnosis and therapy of patients with gastric cancer.

Transcriptomic profiles differentiate normal rectal epithelium and adenocarcinoma

Adenocarcinoma is a histologic diagnosis based on subjective findings. Transcriptional profiles have been used to differentiate normal tissue from disease and could provide a means of identifying malignancy. The goal of this study was to generate and test transcriptomic profiles that differentiate normal from adenocarcinomatous rectum. Comparisons were made between cDNA microarrays derived from normal epithelium and rectal adenocarcinoma. Results were filtered according to standard deviation to retain only highly dysregulated genes. Genes differentially expressed between cancer and normal tissue on two-groups t test (P < 0.05, Bonferroni P value adjustment) were further analyzed. Genes were rank ordered in terms of descending fold change. For each comparison (tumor versus normal epithelium), those 5 genes with the greatest positive fold change were grouped in a classifier. Five separate tests were applied to evaluate the discriminatory capacity of each classifier. Genetic classifiers derived comparing normal epithelium with malignant rectal epithelium from pooled stages had a mean sensitivity and specificity of 99.6% and 98.2%, respectively. The classifiers derived from comparing normal and stage I cancer had comparable mean sensitivities and specificities (97% and 98%, respectively). Areas under the summary receiver-operator characteristic curves for each classifier were 0.981 and 0.972, respectively. One gene was common to both classifiers. Classifiers were tested in an independent Gene Expression Omnibus-derived dataset. Both classifiers retained their predictive properties. Transcriptomic profiles comprising as few as 5 genes are highly accurate in differentiating normal from adenocarcinomatous rectal epithelium, including early-stage disease.

Discovery of genes from feces correlated with colorectal cancer progression

Colorectal cancer (CRC) is considered to develop slowly via a progressive accumulation of genetic mutations. Markers of CRC may serve to provide the basis for decision-making, and may assist in cancer prevention, detection and prognostic prediction. DNA and messenger (m)RNA molecules that are present in human feces faithfully represent CRC manifestations. In the present study, exogenous mouse cells verified the feasibility of total fecal RNA as a marker of CRC. Furthermore, five significant genes encoding solute carrier family 15, member 4 (SLC15A4), cluster of differentiation (CD)44, 3-oxoacid CoA-transferase 1 (OXCT1), placenta-specific 8 (PLAC8) and growth arrest-specific 2 (GAS2), which are differentially expressed in the feces of CRC patients, were verified in different CRC cell lines using quantitative polymerase chain reaction. The present study demonstrated that the mRNA level of SLC15A4 was increased in the majority of CRC cell lines evaluated (SW1116, LS123, Caco-2 and T84). An increased level of CD44 mRNA was only detected in an early-stage CRC cell line, SW1116, whereas OXCT1 was expressed at higher levels in the metastatic CRC cell line CC-M3. In addition, two genes, PLAC8 and GAS2, were highly expressed in the recurrent CRC cell line SW620. Genes identified in the feces of CRC patients differed according to their clinical characteristics, and this differential expression was also detected in the corresponding CRC cell lines. In conclusion, feces represent a good marker of CRC and can be interpreted through the appropriate CRC cell lines.

Multifaceted functions of Siva-1: more than an Indian God of Destruction

Siva-1, as a p53-inducible gene, has been shown to induce extensive apoptosis in a number of different cell lines. Recent evidence suggests that Siva-1 functions as a part of the auto-regulatory feedback loop that restrains p53 through facilitating Mdm2-mediated p53 degradation. Also, Siva-1 plays an important role in suppressing tumor metastasis. Here we review the current understanding of Siva-1-mediated apoptotic signaling pathway. We also add comments on the p53-Siva-1 feedback loop, the novel function of Siva-1 in suppressing tumor metastasis, and their potential implications.

Downregulation of metallothionein 1F, a putative oncosuppressor, by loss of heterozygosity in colon cancer tissue

PURPOSE

Downregulation of metallothionein (MT) genes has been reported in several tumors with discrepant results. This study is to investigate molecular mechanism of MT gene regulation in colon cancer which is characterized by tumor suppressor gene alterations.

EXPERIMENTAL DESIGN

Integral analysis of microarray data with loss of heterozygosity (LOH) information was employed. Quantitative real-time PCR and immunohistochemistry were used to validate MT isoform expression in colon cancer tissues and cell lines. The effects of MT1F expression on RKO cell survival and tumorigenesis was analyzed. Bisulphite sequencing PCR (BSP) and methylation-specific PCR were employed to detect the methylation status of the MT1F gene in colon cancer tissues and cell lines. DNA sequencing was used to examine the LOH at the MT1F locus.

RESULTS

MT1F, MT1G, MT1X, and MT2A gene expression was significantly downregulated in colon cancer tissue (p<0.05). Exogenous MT1F expression increased RKO cell apoptosis and inhibited RKO cell migration, invasion and adhesion as well as in vivo tumorigenicity. Downregulation of MT1F gene in majority of human colon tumor tissues is mainly through mechanism by loss of heterozygosity (p=0.001) while CpG island methylation of MT1F gene promoter region was only observed in poorly differentiated, MSI-positive RKO and LoVo colon cancer cell lines.

CONCLUSIONS

MT1F is a putative tumor suppressor gene in colon carcinogenesis that is downregulated mainly by LOH in colon cancer tissue. Further studies are required to elucidate a possible role for MT1F downregulation in colon cancer initiation and/or progression.

Mdm2 inhibition induces apoptosis in p53 deficient human colon cancer cells by activating p73- and E2F1-mediated expression of PUMA and Siva-1

Camptothecin (CPT) and Nutlin-3 caused apoptosis by increasing p53 protein and its activation in intestinal epithelial cells (IEC-6). We studied the effectiveness of these inducers on apoptosis in human colon cancer cells (Caco2) lacking p53 expression. CPT failed to activate caspase-3 and cause apoptosis in these cells. The absence of p53 expression, higher basal Bcl-xL and lower Bax proteins prevented CPT-induced apoptosis. However, the Mdm2 antagonist Nutlin-3 induced apoptosis in a dose dependent manner by activating caspases-9 and -3. Nutlin-3 prevented the activation of AKT via PTEN-mediated inhibition of the PI3K pathway. Nutlin-3 increased the phosphorylation of retinoblastoma protein causing E2F1 release leading to induction of Siva-1. Nutlin-3-mediated degradation of Mdm2 caused the accumulation of p73, which induced the expression of p53 up-regulated modulator of apoptosis (PUMA). E2F1 and p73 knockdown decreased the expression of Siva and PUMA, respectively and abolished Nutlin-3-induced caspase-3 activation. Cycloheximide (CHX) inhibited Nutlin-3-induced Siva, Noxa, and PUMA expression and inhibited apoptosis in IEC-6 and Caco2 cells. These results indicate that translation of mRNAs induced by Nutlin-3 is critical for apoptosis. In summary, apoptosis in Caco2 cells lacking functional p53 occurred following the disruption of Mdm2 binding with p73 and Rb leading to the expression of pro-apoptotic proteins, PUMA, Noxa, and Siva-1.

[Functional mRNA expression analysis and classification of colonic biopsy samples using overall cDNA microarray technique]

BACKGROUND

Overall mRNA expression analysis of colon biopsies can contribute to the understanding of molecular background of the local alterations and gene ontology-based functional classification of colonic biopsies into inflammatory and neoplastic diseases.

METHODS

Total RNA was extracted from frozen biopsies and amplified by T7-method. Expression profile was evaluated by Atlas Glass 1K microarrays. After microarray quality control, applicable data were available from 10 adenomas, 6 colorectal adenocarcinoma (CRC), and inflammatory bowel diseases (IBDs: 3-3 CD and UC). Multivariate statistical and cell functional analyses were performed. Real-time RT-PCR and immunohistochemistry were used for validation.

RESULTS

Discriminant analysis of selected genes could correctly reclassify all 22 samples using 4 parameters (heat shock transcription factor-1, bystin-like, calgranulin-A, TRAIL receptor 3). IBD samples were characterized by overregulated chemokine (C-X-C motif) ligand 13, replication protein A1, E74-like factor 2 and downregulated TNF receptor-associated factor 6, BCL2-interacting killer genes. In adenomas upregulation of TNF receptor-associated factor 6, replication protein A1, E74-like factor 2 and underexpression of BCL2-associated X protein, calgranulin-A genes were found. CRC cases had significantly increased epidermal growth factor receptor, topoisomerase-1, v-jun, TNF receptor-associated factor 6 and TRAIL receptor 3, and decreased RAD51 and RAD52 DNA repair gene, protein phosphatase-2A and BCL2-interacting killer mRNA levels. Epidermal growth factor receptor RT-PCR and immunohistochemistry, topoisomerase-1 RT-PCR confirmed the chip results.

CONCLUSIONS

Different histological alterations can be objectively classified by functional, multivariate analysis using cDNA microarrays. Disease-specific gene expression patterns can assist to avoid the intermediate and nondescript cases in diagnostics, which do not belong to any of the conventional diagnostic groups.

cDNA array analysis for prediction of hepatic metastasis of colorectal carcinoma

PURPOSE

Distant metastasis is a significant prognostic factor of colon carcinoma. Adjuvant chemotherapy has been shown to decrease its recurrence. However, there are no definitive methods for the diagnosis of hepatic recurrence after potentially curative surgery. The aim of this study was to evaluate the accuracy of mRNA expression profiling using samples obtained from primary tumors to predict hepatic recurrence.

METHODS

Patients with stage III colorectal carcinoma without any recurrence for at least 5 years (group A: n = 9) and patients with stage IV carcinoma with hepatic metastasis (group B: n = 10) were included in this study. Tissue samples were collected from the primary tumor and adjacent normal colonic mucosa at the time of surgery in each patient. Total RNA was extracted and the mRNA expression profile was examined using a cDNA macroarray.

RESULTS

A hierarchical clustering analysis revealed a dendrogram in which the patients were divided into two clusters. One cluster consisted of seven patients in group A and two in group B. The other consisted of two patients in group A and eight in group B. Therefore, the positive and negative predictive value of hierarchical clustering analysis for hepatic metastasis was 80.0% and 78.8%, respectively. Fifteen genes were revealed to be upregulated and 12 were downregulated in group B. The upregulated genes included CCNA2, TP53, and MDM2, while the downregulated genes included CDH1, GADD45A, and BCL2L2.

CONCLUSIONS

mRNA expression profiling by a cDNA array analysis of specimens obtained from primary tumors was found to be useful for distinguishing patients with and without hepatic metastasis. This method is expected to contribute to the identification of patients at high risk for hepatic recurrence, while also helping in the administration of intensive adjuvant chemotherapy for such high risk patients.

Adequate design of customized cDNA macroarray for convenient multiple gene expression analysis

To establish a convenient, cost-effective, and reasonably reliable method for monitoring multiple gene expression using customized membrane-based macroarray, we constructed a cDNA macroarray with multiple probes for 13 human vascular endothelial genes and assessed the accuracy of the macroarray measurements. For each gene, two cDNA probes (450-550 bp) were designed from different regions (coding region and 3'-untranslated region [3'-UTR], respectively) on the basis of simple criteria concerning length and sequence specificity and spotted on the macroarray. In addition, unmodified oligonucleotide probes (80 mer) targeted to a unique sequence from the coding region of each gene were spotted on the same macroarray. Using this macroarray, shear stress-induced mRNA expression changes were analyzed in human coronary artery endothelial cells. Comparison of the expression ratios obtained with those measured using quantitative real-time polymerase chain reaction (PCR) as a reference method revealed that cDNA probes designed from a sequence within the coding region provided a highly accurate expression profile, whereas results obtained from oligonucleotide probes showed no correlation with real-time PCR data, which might be caused by inadequate immobilization of oligonucletotide probes on the nylon membrane. In addition, we observed that cDNA probes targeting different regions of a gene yielded different signal intensities. Most cDNA probes designed from a sequence within the coding region showed detectable signals, whereas few cDNA probes designed from 3'-UTR did.

mRNA expression, functional profiling and multivariate classification of colon biopsy specimen by cDNA overall glass microarray

AIM

To understand the local pathophysiological alterations and gene ontology-based functional classification of colonic biopsies into inflammatory and neoplastic diseases.

METHODS

Total RNA was extracted from frozen biopsies and amplified by T7-method. Expression profile was evaluated by Atlas Glass 1K microarrays. After microarray quality control, applicable data were available from 10 adenomas, 6 colorectal adenocarcinomas (CRCs), and 6 inflammatory bowel diseases (IBDs). Multivariate statistical and cell functional analyses were performed. Real-time RT-PCR and immunohistochemistry were used for validation.

RESULTS

Discriminant analysis of selected genes, could correctly reclassify all 22 samples using 4 parameters (heat shock transcription factor-1, bystin-like, calgranulin-A, TRAIL receptor 3). IBD samples were characterized by overregulated chemokine (C-X-C motif) ligand 13, replication protein A1, E74-like factor 2 and downregulated TNF receptor-associated factor 6, BCL2-interacting killer genes. In adenomas upregulation of TNF receptor-associated factor 6, replication protein A1, E74-like factor 2 and underexpression of BCL2-associated X protein, calgranulin-A genes were found. CRC cases had significantly increased epidermal growth factor receptor, topoisomerase-1, v-jun, TNF receptor-associated factor 6 and TRAIL receptor 3, and decreased RAD51 and RAD52 DNA repair gene, protein phosphatase-2A and BCL2-interacting killer mRNA levels. Epidermal growth factor receptor RT-PCR and immunohistochemistry, topoisomerase-1 RT-PCR confirmed the chip results.

CONCLUSION

Different histological alterations can be reclassified by functional, multivariate analysis using cDNA microarrays. Further studies with expanded sample number are needed for subclassification of pathological alterations.

The results of the expression array studies correlate and enhance the known genetic basis of gastric and colorectal cancer

Gastric and colorectal cancers belong to the most frequent cancer types in the world today. This fact emphasizes the importance of identification of useful diagnostic and prognostic markers, in the earliest stage of the disease. The examination of gene expression profile in gastric and colorectal cancer may develop the bases of early diagnosis and of individual therapeutic strategies. In the microarray examinations done so far for these types of cancers, the expression of hundreds and thousands of genes were studied, however, both the sample collection and the results showed wide variations. The diversity of expression array methods and data analysis makes the comparison of microarray results difficult. Beside the exposition of the practical aspects of the chip technology, our aims are the systematization of data that are currently available in the international scientific literature and the description of the results in a comprehensive way. Microarray results show that the gene expression pattern, detected in gastric and colon cancers, highly depends on the histological type and heterogeneity of the sample, array type, and softwares, used for data analysis. Recent experiments point out not just the changes of the alterations of tumor suppression, apoptosis, cell-cycle regulation, and signal transduction, but tumor cell metabolism and cell-microenvironment interactions also. Results show connection to and make more complete the already known molecular background of gastric and colorectal cancers. Based on the accumulation of recent and further data, such kind of multifunctional diagnostic microarrays that can be suited for completing the conventional histological diagnostics and subtypization will certainly become available in the near future.

Synergistic effect of hypoxia and TNF-alpha on production of PAI-1 in human proximal renal tubular cells

BACKGROUND

Chronic hypoxia has been newly proposed as a common mechanism of tubulointerstitial fibrosis in the progression of various chronic inflammatory renal diseases, where plasminogen activator inhibitor-1 (PAI-1) plays an important role in the accumulation of extracellular matrix (ECM) through inhibition of plasmin-dependent ECM degradation. In the present study, we investigated the presence of PAI-1 in renal tubular cells by immunostaining renal biopsy samples. We also closely examined the effects of hypoxia and tumor necrosis factor-alpha (TNF-alpha) on PAI-1 expression in cultured human proximal renal tubular cells (HPTECs).

METHODS

Confluent cells growth-arrested in Dulbecco's modified Eagle's medium (DMEM) for 24 hours were exposed to hypoxia (1% O(2)) and/or TNF-alpha at 10 ng/mL for up to 48 hours. Amounts of PAI-1 protein and mRNA after stimulation were measured by enzyme-linked immunosorbent assay (ELISA) and TaqMan quantitative polymerase chain reaction (PCR) or cDNA array analysis, respectively, and compared to those in cells incubated under control conditions (18% O(2) without TNF-alpha). Hypoxia-inducible factor-1alpha (HIF-1alpha) was demonstrated by immunoblot and immunofluorescence analyses. Human PAI-1 promoter activity was estimated by luciferase reporter gene assay.

RESULTS

In crescentic glomerulonephritis, clusters of proximal tubules were specifically stained for PAI-1. cDNA array analysis identified PAI-1 as a major gene highly induced by hypoxia in HPTECs. Treatment of 24 hours with hypoxia, TNF-alpha, and their combination induced a 2.8-fold, a 1.8-fold, and a 4.6-fold increase in PAI-1 protein secretion, and produced a 3.6-fold, a 3.3-fold, and a 12.1-fold increase at the PAI-1 mRNA level, respectively. Immunoblot analysis and immunocytochemistry revealed that hypoxia-inducible factor-1alpha (HIF-1alpha) was markedly accumulated in the cell lysates and exclusively translocated to nuclei after 16 hours' exposure of HPTECs to hypoxia but not to TNF-alpha. Luciferase reporter gene assay showed that hypoxia, TNF-alpha, and their combination increased PAI-1 transcription activity by 1.8-fold, 1.4-fold, and 2.2-fold, respectively. A dominant-negative form of HIF-1alpha significantly suppressed PAI-1 transcription activity induced by hypoxia. Inhibition of nuclear factor-kappaB (NF-kappaB) caused a moderate decrease in PAI-1 production under hypoxia.

CONCLUSION

Hypoxia induces PAI-1 expression via remarkable nuclear accumulation of HIF-1alpha and partially via NF-kappaB activation in HPTECs. TNF-alpha can synergistically enhance this hypoxia-induced PAI-1 expression.

Development of a functional cDNA array for evaluation of the Th1/Th2 balance

The immune balance controlled by CD4(+) helper T cell subsets (T helper 1 (Th1) and T helper 2 (Th2)) is crucial for immunoregulation and its imbalance causes various immune diseases including infections, allergic disorders and autoimmune diseases. Therefore, it is of great importance to develop a system of diagnosing Th1/Th2 imbalances for curing immune diseases. Here we developed a functional cDNA array filter useful for assessing the Th1/Th2 balance in mice. To overcome the disadvantages of conventional microarrays carrying thousands of genes, we prepared an array filter containing 40 Th1-specific and 32 Th2-specific genes, which were selected from over 8700 genes based on (i) the specificity of expression in Th1 or Th2 cells and (ii) an expression level which is high enough for detection using a DNA array. This array filter provided a prompt and precise evaluation for the skewing of the Th1/Th2 balance combined with our calculation algorithm. The bias toward Th1 or Th2 was evaluated visually at a glance by aligning the genes on the filter. Moreover, we succeeded in evaluating the skewing of the Th1/Th2 balance in vivo during acute graft versus host disease (GVHD). Thus, this array filter will provide a novel tool for evaluation of the Th1/Th2 balance in a variety of immune diseases.

Expression of Siva-1 protein or its putative amphipathic helical region enhances cisplatin-induced apoptosis in breast cancer cells: effect of elevated levels of BCL-2

cis-Diaminedichloroplatinum (II) (cisplatin) is routinely used to treat various types of cancers; however, a significant number develop resistance. One of the underlying factors that contribute to cisplatin resistance is the elevated level of BCL-2 and/or BCL-XL, which promotes cell survival. A potential method of overcoming such resistance is to use a potentiator that is capable of neutralizing the antiapoptotic effects of BCL-2/BCL-XL, such as Siva-1. We previously cloned the proapoptotic protein Siva-1 and showed a possible role for it in both extrinsic and intrinsic apoptosis. Using an adenovirus-based expression system, we now show that Siva-1 can synergize with cisplatin in inducing apoptosis in MCF7 and MDA-MB-231 breast cancer cells. In an anchorage-independent clonogenicity assay, MCF7/caspase-3 cells stably expressing Siva-1, but not the control cells, showed a dramatic decrease in the number of colonies formed on one-time cisplatin treatment. Further, we show that the unique putative amphipathic helical region (SAH) in Siva-1 (amino acid residues 36-55) is necessary and sufficient for the observed enhancement in cisplatin-induced apoptosis by Siva-1. Although cisplatin treatment results in significant elevation in the expression of Fas ligand and intracellular p21 levels, expression of Siva-1 has no additional benefit. Instead, the enhancement in apoptosis seems to be due to activation of intrinsic pathway that involves caspase-9 activation. Moreover, Siva-1 augments cisplatin-mediated cell death in MCF7 cells stably expressing BCL-2. We therefore propose that Siva-1 or its SAH region can be used as a potentiator of cisplatin-based chemotherapy.

The Proapoptotic Gene SIVA Is a Direct Transcriptional Target for the Tumor Suppressors p53 and E2F1

The p53 tumor suppressor gene is believed to play an important role in neuronal cell death in acute neurological disease and in neurodegeneration. The p53 signaling cascade is complex, and the mechanism by which p53 induces apoptosis is cell type-dependent. Using DNA microarray analysis, we have found a striking induction of the proapoptotic gene, SIVA. SIVA is a proapoptotic protein containing a death domain and interacts with members of the tumor necrosis factor receptor family as well as anti-apoptotic Bcl-2 family proteins. SIVA is induced following direct p53 gene delivery, treatment with a DNA-damaging agent camptothecin, and stroke injury in vivo. SIVA up-regulation is sufficient to initiate the apoptotic cascade in neurons. Through isolation and analysis of the SIVA promoter, we have identified response elements for both p53 and E2F1. Like p53, E2F1 is another tumor suppressor gene involved in the regulation of apoptosis, including neuronal injury models. We have identified E2F consensus sites in the promoter region, whereas p53 recognition sequences were found in intron1. Sequence analysis has shown that these consensus sites are also conserved between mouse and human SIVA genes. Electrophoretic mobility shift assays reveal that both transcription factors are capable of binding to putative consensus sites, and luciferase reporter assays reveal that E2F1 and p53 can activate transcription from the SIVA promoter. Here, we report that the proapoptotic gene, SIVA, which functions in a broad spectrum of cell types, is a direct transcriptional target for both tumor suppressors, p53 and E2F1.

Phenotypic differences of proliferating fibroblasts in the stroma of lung adenocarcinoma and normal bronchus tissue

Fibroblasts in tumor tissue are thought to interact with tumor cells directly and/or indirectly and to have important roles in tumor invasion and metastasis. To characterize the phenotype of proliferating fibroblasts in pulmonary adenocarcinoma, we established short-term fibroblast cell lines from both normal bronchus and adenocarcinoma tissues obtained from the same patients and compared the gene expression profiles. Four sets of fibroblast cell lines (eight cell lines in total) were used in the analysis. Total RNA was extracted from each cell line and hybridized with 550 cancer-related RNAs blotted on a cDNA filter array. Five up-regulated genes and 12 down-regulated genes (total of 17 genes) were detected in the fibroblast cell lines from the tumor tissues compared with those from normal bronchus. Using real-time quantitative RT-PCR methods, the expression profile of each gene was examined; five genes, one up-regulated (MLH1) and four down-regulated (Cox1, FGFR4, p120, and Smad3), were confirmed. Furthermore, the protein expression levels of the five genes in the cancerous and normal tissues were examined immunohistochemically, and the up-regulation of MLH1 and the down-regulation of Cox1 in cancerous tissue were confirmed in vivo. These results indicate that the proliferating fibroblasts in pulmonary adenocarcinomas are phenotypically different from fibroblasts in normal bronchus tissues.

Mice lacking a transcriptional corepressor Tob are predisposed to cancer

tob is a member of antiproliferative family genes. Mice lacking tob are prone to spontaneous formation of tumors. The occurrence rate of diethylnitrosamine-induced liver tumors is higher in tob(-/-) mice than in wild-type mice. tob(-/-)p53(-/-) mice show accelerated tumor formation in comparison with single null mice. Expression of cyclin D1 mRNA is increased in the absence of Tob and is reduced by Tob. Tob acts as a transcriptional corepressor and suppresses the cyclin D1 promoter activity through an interaction with histone deacetylase. Levels of tob mRNA are often decreased in human cancers, implicating tob in cancer development.

Classification of Dukes' B and C colorectal cancers using expression arrays

PURPOSE

Colorectal cancer is one of the most common malignancies. Substaging of the cancer is of importance not only to prognosis but also to treatment. Classification of substages based on DNA microarray technology is currently the most promising approach. We therefore investigated if gene expression microarrays could be used to classify colorectal tumors.

METHODS

We used the Affymetrix oligonucleotide arrays to analyze the expression of more than 5,000 genes in samples from the sigmoid and upper rectum of the left colon. Five samples were from normal mucosa and five samples from each of the Dukes' stages A, B, C, and D. Expression data were filtered based on either covariance or a selection of the most significantly varying genes between tumor stages.

RESULTS

A nearest neighbor classifier was used to classify normal, and Dukes' B and C samples with less than 20% error, whereas Dukes' A and D could not be classified correctly. A number of interesting gene clusters showed a discriminating difference between Dukes' B and C samples. These included mitochondrial genes, stromal remodeling genes, and genes related to cell adhesion.

CONCLUSION

Molecular classification based on gene expression of one of the most common malignancies, colorectal cancer, now seems to be within reach. The data indicates that it is possible at least to classify Dukes' B and C colorectal tumors with microarrays.

Male specific expression of a cytochrome P450 (Cyp312a1) in Drosophila melanogaster

Using cDNA array techniques, the transcriptional levels of Drosophila cytochrome P450 (P450) genes were compared between male and female flies. Of the 86 P450s, 15 isoforms were picked up and the levels of transcription were confirmed by the real time quantitative RT-PCR. Cyp315a1 and Cyp302a1, which had been reported as P450s involved in the ecdysteroid biosynthesis, were included in the 15 isoforms and expression of these P450s was 8.6- and 7.9-fold higher in females than in males, respectively. In addition, we confirmed that expression of Cyp312a1 was 82-fold higher in adult males than females. This gene expression was observed mostly in the abdomen and its transcription level gradually increased from pupal stage and peaked in the 5-day-old adult. Furthermore, the male specific expression of Cyp312a1 was universally observed in three Drosophila strains originated from the USA (Oregon R), China (Canton S), and Japan (HKJ), suggesting possible involvement of this P450 in significant endogenous catalytic reaction(s). This is the first report of a P450 enzyme being predominantly expressed in male Drosophila.

Analysis of gene expression in human colorectal cancer tissues by cDNA array

BACKGROUND

The development and progression of cancer are accompanied by complex changes in patterns of gene expression. The purpose of this study was to clarify the relevance of macroarray analysis of human colorectal cancer tissues.

METHODS

Hybridization of cDNA macroarray filters on which 550 genes had been spotted was performed with biotin-labeled cDNA targets that were prepared from mRNA extracted from 20 pairs of colorectal cancer and corresponding noncancerous tissues. Expression of differentially expressed genes was further studied by semiquantitative RT-PCR.

RESULTS

Fourteen (2.5%) of the 550 genes were differentially expressed and up- or downregulated in cancer tissues by at least threefold compared with matched noncancerous tissues in 10 or more of the 20 patients. The genes that were upregulated in cancer tissues were associated with transcription, cell cycle, growth factor receptor, cell adhesion, extracellular matrix-degrading enzymes, and angiogenesis, and the downregulated genes were those involved in apoptosis and immune recognition. Semiquantitative RT-PCR analysis of these differentially expressed genes gave results consistent with those by cDNA array analysis.

CONCLUSIONS

Although the macroarray used in this study contained only a small number of genes, our results support the feasibility and usefulness of this approach to study variation in gene expression patterns in human colorectal cancer tissues. The results also suggest the possibility of a diagnostic application of cDNA macroarrays in daily clinical settings.

Siva-1 binds to and inhibits BCL-X(L)-mediated protection against UV radiation-induced apoptosis

We previously cloned Siva-1 by using the cytoplasmic tail of CD27, a member of the tumor necrosis factor receptor family, as the bait in the yeast two-hybrid system. The Siva gene is organized into four exons that code for the predominant full-length Siva-1 transcript, whereas its alternate splice form, Siva-2, lacks exon 2 coding sequence. Various groups have demonstrated a role for Siva-1 in several apoptotic pathways. Interestingly, the proapoptotic properties of Siva-1 are lacking in Siva-2. The fact that Siva-1 is partly localized to mitochondria despite the absence of any mitochondrial targeting signal, it harbors a 20-aa-long putative amphipathic helical structure that is absent in Siva-2, and that its expression is restricted to double-positive (CD3(+), CD4(+), CD8(+)) thymocytes like BCL-X(L), prompted us to test for a potential interaction between Siva-1 and BCL-X(L). Here, we show that Siva-1 binds to and inhibits BCL-X(L)-mediated protection against UV radiation-induced apoptosis. Indeed, the unique amphipathic helical region (SAH) present in Siva-1 is required for its binding to BCL-X(L) and sensitizing cells to UV radiation. Natural complexes of Siva-1/BCL-X(L) are detected in HUT78 and murine thymocyte, suggesting a potential role for Siva-1 in regulating T cell homeostasis.