Identification of direct targets for the miR-17-92 cluster by proteomic analysis.

MicroRNAs (miRNAs) are small non-coding RNAs that post-transcriptionally repress the expression of target genes. Many miRNAs have been implicated in a number of diseases, including cancers. The miR-17-92 miRNA cluster is known as a body of oncogenic miRNAs, and has been shown to be overexpressed in several cancers, including lung cancer. Although the overexpression of miR-17-92 is clearly implicated in the development of lung cancer, only a few direct targets for the miR-17-92 cluster have been identified thus far. In this study, we examined miR-17-92 target profiles in SBC-3 small-cell lung cancer cells using a quantitative proteomic strategy to identify direct targets of the miR-17-92 cluster. By knocking down the expression of endogenous miR-19a, miR-20a and miR-92-1, which are contained in the cluster, 112 up-regulated proteins were detected and also identified as potential targets of these miRNAs. Among these candidate targets, we validated one direct target, RAB14. In conclusion, these findings suggest that proteomic approaches are valuable for identifying direct miRNA targets, and we were able to identify a novel direct target for the miR-92-1 using our proteomic strategy.

Single nucleotide polymorphism in the RAD18 gene and risk of colorectal cancer in the Japanese population.

The RAD18 gene, located on the human chromosome 3p24-p25, plays a crucial role in post-replication repair (PRR) in various organisms from yeast to humans. In the human RAD18 gene, one coding single nucleotide polymorphism (SNP) at codon 302, encoding either arginine (Arg, CGA) or glutamine (Gln, CAA), was reported. Although the molecular function of the RAD18 protein came to be elucidated, the association between the RAD18 Arg302Gln polymorphism and the risk of human cancer development was not examined. Therefore, we investigated the relationship between the polymorphism and the development of human primary colorectal cancer (CRC). The Arg302Gln polymorphism in 100 patients with CRC and 200 healthy controls were genotyped by the polymerase chain reaction with confronting two-pair primer (PCR-CTPP) assay. The Gln/Gln genotype was significantly more frequent in CRC (18.0%) than in the healthy controls (11.5%) (p=0.046). The increased risk was detected in CRC patients with the Gln/Gln genotype (Odds ratio [OR], 2.10; 95% confidence interval [CI], 1.00 to 4.40). When the relationship of the SNP with clinicopathological parameters of CRC was investigated, particularly in the well-differentiated grade and in the lymph node metastasis (N1) CRC patients, significantly higher risks were detected (OR, 7.00; 95% CI, 1.19-41.1 and OR, 3.71; 95% CI, 1.30-10.6, respectively). These results suggested that the RAD18 Arg302Gln polymorphism is associated with the risk of CRC. This report provides evidence for an association between the RAD18 Arg302Gln polymorphism and human CRC risk.

Single nucleotide polymorphism at codon 133 of the RASSF1 gene is preferentially associated with human lung adenocarcinoma risk.

The RASSF1 gene, a putative tumor suppressor gene located on human chromosome 3p21, garners much attention for the frequent allelic loss and gene silencing via promoter hypermethylation in a variety of human malignancies. An association between a single nucleotide polymorphism (SNP) at codon 133 of the RASSF1 gene, encoding either alanine (GCT) or serine (TCT), and human cancer risk remains undefined. We therefore, investigated the distribution of the Ala133Ser SNP in 101 patients with lung cancer, 63 with head and neck cancer, 72 with colorectal cancer, 56 with esophageal cancer and 110 healthy controls by polymerase chain reaction and restriction enzyme-digestion assay. The heterozygous Ala/Ser genotype was significantly more frequent in lung cancer patients than in healthy controls (P=0.028). The adjusted odds ratio (OR) for the patients with heterozygous Ala/Ser genotype as compared with the controls with the Ala/Ala genotype was 2.59 (95% confidence interval (CI); 1.11-6.04). The increased risk of the Ala/Ser genotype was found in lung cancer patients but not in other cancer patients we examined. The association was particularly strong in those lung cancer patients of male (adjusted OR; 3.33, 95% CI; 1.37-8.12), with adenocarcinoma (adjusted OR; 3.33, 95% CI; 1.36-8.15), early stages (adjusted OR; 3.42, 95% CI; 1.33-8.75) and with smoking habit (adjusted OR; 2.70, 95% CI; 1.06-6.83). These results suggest that the RASSF1 Ala133Ser SNP is associated with development of lung cancer, especially of lung adenocarcinoma. The increased risk of the heterozygous genotype is intriguing, implying a close relation with the dimerization feature of RASSF1 proteins.

Single nucleotide polymorphism of the AXIN2 gene is preferentially associated with human lung cancer risk in a Japanese population.

The AXIN2 gene, a negative regulator gene of Wnt/beta-catenin signaling, is a putative tumor suppressor gene on human chromosome 17q24. In the genomic locus on which the AXIN2 gene is located, allelic loss and rearrangement were frequently detected in many cancers. An association between human cancer risk and a single nucleotide polymorphism (SNP) at codon 50 of the AXIN2 gene, encoding either proline (CCT) or serine (TCT), remains undefined. We, therefore, investigated the distribution of the SNP at codon 50 in 110 healthy controls and 160 patients with non-small-cell lung cancer, 113 patients with colorectal cancer, and 63 patients with head and neck cancer. We found that the frequency of the homozygous T/T (Ser/Ser) genotype was significantly less in lung cancer patients (5.0%) than in healthy controls (13.6%) (p=0.005). As compared with the C/C (Pro/Pro) genotype of the controls, lung cancer patients with the T/T genotype showed reduced risk of cancer; the adjusted odds ratio (OR) for patients with the homozygous T/T (Ser/Ser) genotype was 0.31 (95% confidence interval (CI), 0.12-0.79). The association was particularly strong in lung cancer patients with lung adenocarcinoma (LAD) (adjusted OR, 0.24; 95% CI, 0.07-0.81), with well-differentiated grade cancer (adjusted OR, 0.12; 95% CI, 0.01-0.99) and with moderately-differentiated grade cancer (adjusted OR, 0.18; 95% CI, 0.04-0.85). These results suggest that the AXIN2 Pro50Ser SNP is associated with development of lung cancer as a protective SNP, while an association between the AXIN2 SNP and risk of colorectal cancer and of head and neck cancer was not observed. This is the first report to show an association between the AXIN2 SNP and lung cancer risk.

Uncovering Direct Targets of MiR-19a Involved in Lung Cancer Progression.

Micro RNAs (miRNAs) regulate the expression of target genes posttranscriptionally by pairing incompletely with mRNA in a sequence-specific manner. About 30% of human genes are regulated by miRNAs, and a single miRNA is capable of reducing the production of hundreds of proteins by means of incomplete pairing upon miRNA-mRNA binding. Lately, evidence implicating miRNAs in the development of lung cancers has been emerging. In particular, miR-19a, which is highly expressed in malignant lung cancer cells, is considered the key miRNA for tumorigenesis. However, its direct targets remain underreported. In the present study, we focused on six potential miR-19a target genes selected by miRNA target prediction software. To evaluate these genes as direct miR-19a target genes, we performed luciferase, pull-down, and western blot assays. The luciferase activity of plasmids with each miR-19a-binding site was observed to decrease, while increased luciferase activity was observed in the presence of anti-miR-19a locked nucleic acid (LNA). The pull-down assay showed biotinylated miR-19a to bind to AGO2 protein and to four of six potential target mRNAs. Western blot analysis showed that the expression levels of the four genes changed depending on treatment with miR-19a mimic or anti-miR-19a-LNA. Finally, FOXP1, TP53INP1, TNFAIP3, and TUSC2 were identified as miR-19a targets. To examine the function of these four target genes in lung cancer cells, LK79 (which has high miR-19a expression) and A549 (which has low miR-19a expression) were used. The expression of the four target proteins was higher in A549 than in LK79 cells. The four miR-19a target cDNA expression vectors suppressed cell viability, colony formation, migration, and invasion of A549 and LK79 cells, but LK79 cells transfected with FOXP1 and TP53INP1 cDNAs showed no difference compared to the control cells in the invasion assay.

Reduced expression of the REIC/Dkk-3 gene by promoter-hypermethylation in human tumor cells.

The human REIC gene is a recently found mortalization-related gene and a candidate tumor suppressor gene expression of which is largely attenuated in many immortalized and tumor-derived cell lines (Biochem. Biophys. Res. Commun. 268 (2000) 20-24). To gain insight into the mechanisms of the down-regulation, we investigated the genomic structure and promoter activity of the human REIC gene. The gene, identical with the DKK-3 gene, resides on chromosome 11p15.1, consists of nine exons, and has two promoters. Methylation in the main promoter region was detected in 11 out of 21 cell lines tested (52%) derived from a variety of human tumors, in which the expression of the REIC gene was decreased. In ten of these 11 cell lines the minor promoter was also methylated. Similarly, the REIC gene expression was decreased in 14 of 24 fresh non-small cell lung cancer specimens (58%) compared to that in corresponding non-cancerous tissue, though allelic loss and tumor-specific mutation were rare. Of these 14 tumors, at least five tumors exhibited heavy methylation of the REIC promoter region. These results indicate that the down-regulation of the REIC gene expression is ascribed to the aberrant promoter hyper-methylation at least in a subset of human tumors. The expression was restored upon treatment of SQ5 cells with 5-aza-deoxycytidine, confirming DNA methylation as the mode of downregulation. A notable single nucleotide polymorphism in the coding region (cSNP) with an amino acid substitution of glycine (GGG) to arginine (AGG) was found at codon 335 of the REIC gene. However, the distribution of the cSNP showed no significant difference between lung cancer patients and healthy population.

Novel direct targets of miR-19a identified in breast cancer cells by a quantitative proteomic approach.

The miR-17-92 cluster encodes 7 miRNAs inside a single polycistronic transcript, and is known as a group of oncogenic miRNAs that contribute to tumorigenesis in several cancers. However, their direct targets remain unclear, and it has been suggested that a single miRNA is capable of reducing the production of hundreds of proteins. The majority of reports on the identification of miRNA targets are based on computational approaches or the detection of altered mRNA levels, despite the fact that most miRNAs are thought to regulate their targets primarily by translational inhibition in higher organisms. In this study, we examined the target profiles of miR-19a, miR-20a and miR-92-1 in MCF-7 breast cancer cells by a quantitative proteomic strategy to identify their direct targets. A total of 123 proteins were significantly increased after the endogenous miR-19a, miR-20a and miR-92-1 were knocked down, and were identified as potential targets by two-dimensional electrophoresis and a mass spectrometric analysis. Among the upregulated proteins, four (PPP2R2A, ARHGAP1, IMPDH1 and NPEPL1) were shown to have miR-19a or miR-20a binding sites on their mRNAs. The luciferase activity of the plasmids with each binding site was observed to decrease, and an increased luciferase activity was observed in the presence of the specific anti-miRNA-LNA. A Western blot analysis showed the expression levels of IMPDH1 and NPEPL1 to increase after treatment with anti-miR-19a, while the expression levels of PPP2R2A and ARHGAP1 did not change. The expression levels of IMPDH1 and NPEPL1 did not significantly change by anti-miR-19a-LNA at the mRNA level. These results suggest that the IMPDH1 and NPEPL1 genes are direct targets of miR-19a in breast cancer, while the exogenous expression of these genes is not associated with the growth suppression of MCF-7 cells. Furthermore, our proteomic approaches were shown to be valuable for identifying direct miRNA targets.

The association between RAD18 Arg302Gln polymorphism and the risk of human non-small-cell lung cancer.

PURPOSE: The repair enzyme RAD18 plays a key role in the post-replication repair process in various organisms from yeast to human, and the molecular function of the RAD18 protein has been elucidated. Single nucleotide polymorphism (SNP) of arginine (Arg, CGA) or glutamine (Gln, CAA) at codon 302 is known on RAD18; however, the association between the SNP and the risk of any human cancers including non-small-cell lung cancer (NSCLC) has not been reported. We therefore investigated the relationship between the polymorphism and the development and progression of human NSCLC. METHODS: The study population included 159 patients with NSCLC and 200 healthy controls. The SNP was genotyped by polymerase chain reaction with the confronting two-pair primer (PCR-CTPP) assay. Genotype frequencies were compared between patients and controls, and the association of genotypes with clinicopathological parameters was also studied. RESULTS: The Gln/Gln genotype was significantly more frequent in NSCLC patients (20.7%) than in healthy controls (11.5%)(P = 0.003). The increased risk was detected in NSCLC patients with the Gln/Gln genotype [Odds ratio (OR) = 2.63, 95% confidence interval (CI)=1.38-4.98]. As to the relationship of the SNP with clinicopathological parameters of NSCLC, significantly higher risks were detected in lung squamous cell carcinoma (LSC) (OR = 4.40, 95% CI = 1.60-12.1). CONCLUSIONS: Our results suggested that Gln/Gln genotype of the RAD18 SNP has the increased risk of NSCLC, especially of LSC. This is the first report to provide evidence for an association between the RAD18 Arg302Gln polymorphism and human NSCLC risk.

Single nucleotide polymorphisms in the EXO1 gene and risk of colorectal cancer in a Japanese population.

EXO1 is a member of the RAD2 nuclease family and functions in DNA replication, repair and recombination. We investigated the relationship of single nucleotide polymorphisms (SNPs) at exon 10 (T439M) and exon 13 (P757L) of the EXO1 gene with development, progression and metastasis of colorectal cancer. For T439M, the Thr/Met genotype [odds ratio (OR) = 2.03, 95% confidence interval (CI) 1.04-3.98] and Thr/Met and Met/Met genotypes combined (OR = 2.37, 95% CI 1.23-4.56) demonstrated significant association with the development of colorectal cancer after adjusting for age, gender and smoking status. For P757L, patients with the Leu/Leu genotype showed a reduced risk of colorectal cancer (adjusted OR = 0.398, 95% CI 0.183-0.866) when the Pro/Leu and Pro/Pro genotypes were combined and used as the reference. The Leu/Leu genotype also had a reduced risk (adjusted OR = 0.373, 95% CI 0.164-0.850) when the Pro/Leu genotype was used as the reference. Individuals who carried both putative risk genotypes (Thr/Met and Met/Met for T439M and Pro/Leu for P757L) showed an adjusted OR of 4.95 (95% CI 1.56-15.7) compared with those who carried both low risk genotypes. Analysis of microsatellite instability (MSI) revealed that tumors from individuals who carried both putative risk genotypes tended to have a higher frequency of MSI positives than those from patients who carried both low risk genotypes, although a significant correlation was not found between EXO1 genotype and MSI status. This is the first report to provide evidence for an association of EXO1 gene polymorphisms with colorectal cancer risk. The EXO1 genotypes were not associated with any clinicopathological characteristics in colorectal cancer patients.