Koji Koinuma

Mutations of BRAF are associated with extensive hMLH1 promoter methylation in sporadic colorectal carcinomas

Activating mutations of BRAF have been frequently observed in microsatellite unstable (MSI+) colorectal carcinomas (CRCs), in which mutations of BRAF and KRAS are mutually exclusive. Previously, we reported that hypermethylation of hMLH1 might play an important role in the tumorigenesis of right‐sided sporadic CRCs with MSI showing less frequency of KRAS/TP53 alteration. Therefore, we have assumed that BRAF mutations might be highly associated with hMLH1 methylation status rather than MSI status. In this study, mutations of BRAF and KRAS and their relationship with MSI and hMLH1 methylation status were examined in 140 resected specimens of CRC. The methylation status was classified into 3 types: full methylation (FM), partial methylation (PM) and nonmethylation (NM). Only FM closely linked to reduced expression of hMLH1 protein. BRAF mutations were found in 16 cases (11%), all leading to the production of BRAFV599E. As for MSI status, BRAF mutations were found in 43% of MSI+ and 4% of MSI− cases (p < 0.0001). Among the MSI+ individuals, BRAF mutations were more frequent in cases with hMLH1 deficiency (58%) than those with hMSH2 deficiency (0%; p = 0.02). Moreover, they were found in 69% of FM, 4% of PM and 4% of NM, revealing a striking difference between FM and the other 2 groups (FM vs. PM or NM; p < 0.0001). These findings suggest that BRAF activation may participate in the carcinogenesis of sporadic CRCs with hMLH1 hypermethylation in the proximal colon, independently of KRAS activation.

Predictive factors for lymph node metastasis in T1 stage colorectal carcinomas.

PurposeSelective endoscopie resection may cure early colorectal cancer (Tl), but the management is controversial. There is concern about the small risk of lymph node metastasis, which will not be treated by endoscopie resection alone. The authors sought predictive markers of lymph node metastasis to assist patient management. METHODS: The authors retrospectively analyzed consecutive cases of Tl stage colorectal cancer resected using endoscopie resection or bowel surgery over the period 1979 to 2000. The risk of lymph node metastasis was analyzed using logistic regression model for the markers selected by univariate analysis: the type of initial treatment, depth of submucosal invasion, lymphatic channel invasion, differentiation of histology, and invasive front histology. RESULTS: Two hundred seventy-eight patients were available for study. Twenty-one had lymph node metastasis. Depth of submucosal invasion (2 2,000 yum) and lymphatic channel invasion significantly predicted risk of lymph node metastasis in multivariate analysis. When these two factors were adopted for the prediction of lymph node metastasis, sensitivity, specificity, positive predictive value, and negative predictive value were 100, 55.6, 15.6, and 100 percent, respectively. CONCLUSIONS: Depth of submucosal invasion and lymphatic channel invasion were accurate predictive factors for lymph node metastasis. These two factors could be used in selecting appropriate cases for surgery after endoscopie resection.

Evaluation of venous invasion by Elastica van Gieson stain and tumor budding predicts local and distant metastases in patients with T1 stage colorectal cancer.

Evaluation of pathologic predictors of metastases in T1 stage colorectal cancer may be difficult with hematoxylin and eosin (HE) staining alone. The aim of this study was to clarify the role of pathologic predictors by using immunohistochemical staining and Elastica van Gieson (EVG) staining. One hundred and twenty-four patients who underwent bowel resection for single T1 stage colorectal cancer from 1990 to 2004 in 1 institution were studied. D2-40, EVG staining, and CAM5.2 were used to detect lymphatic invasion, venous invasion, and tumor budding, respectively. These 3 factors were separately evaluated based on HE staining. Histology was reviewed by 1 pathologist. Lymph node metastases in the surgical specimen were the standard reference, and distant metastases were identified by periodic computed tomography for 2 years or more after surgery. A logistic regression model was applied to analyze risk factors for lymph node metastases and a Cox regression model for distant metastases. In predicting lymph node metastases, univariate analysis demonstrated significance for all predictors except venous invasion by HE staining. Multivariate analysis showed that venous invasion by EVG and tumor budding by HE showed significance as predictors. In predicting distant metastases, univariate analysis showed significance for lymphatic invasion shown by D2-40, tumor budding shown by CAM5.2 and HE, and lymph node metastases. Multivariate analysis showed only venous invasion by EVG stain as being significantly associated with distant metastases (P=0.001). In conclusion, venous invasion evaluated shown by EVG staining is a useful pathologic predictor for metastases in T1 stage colorectal cancer.

A comparison of conventional endoscopy, chromoendoscopy, and the optimal-band imaging system for the differentiation of neoplastic and non-neoplastic colonic polyps

BACKGROUND The diagnostic accuracy of conventional endoscopy for small colonic polyps is not satisfactory. Optimal band imaging (OBI) enhances the contrast of the mucosal surface without the use of dye. OBJECTIVE To evaluate the diagnostic accuracy for the differentiation of neoplastic and non-neoplastic colorectal polyps by using magnified OBI colonoscopy. DESIGN An open prospective study. SETTING Jichi Medical University, Japan. PATIENTS A total of 133 colonoscopy cases. MAIN OUTCOME MEASUREMENT A comparative study of the overall accuracy, sensitivity, and specificity for the differentiation of neoplastic and non-neoplastic colorectal polyps < or =5 mm in size by capillary-pattern diagnosis by using conventional colonoscopy, capillary-pattern diagnosis in OBI, and pit-pattern diagnosis in chromoendoscopy with low magnification. RESULTS A total of 107 polyps, composed of 80 neoplastic and 27 non-neoplastic polyps, were evaluated. OBI clearly showed the capillary network of the surface mucosa of neoplastic polyps at low magnification, whereas the surface mucosa of non-neoplastic polyps showed up as a pale lesion. The capillary pattern in conventional colonoscopy had 74% accuracy, 71% sensitivity, and 81% specificity for neoplastic polyps. The accuracy and sensitivity were significantly lower than those that used the capillary pattern in OBI (accuracy 87% and sensitivity 93%) and the pit pattern in chromoendoscopy (accuracy 86% and sensitivity 90%). There were no significant differences in specificity (OBI 70% and chromoendoscopy 74%). The kappa analysis indicated good agreement in both OBI and chromoendoscopy. CONCLUSIONS Capillary-pattern diagnosis in OBI is superior to that in conventional endoscopy and is not significantly different from pit-pattern diagnosis for predicting the histology of small colorectal polyps.

Epigenetic silencing of AXIN2 in colorectal carcinoma with microsatellite instability.

Mutation or epigenetic silencing of mismatch repair genes, such as MLH1 and MSH2, results in microsatellite instability (MSI) in the genome of a subset of colorectal carcinomas (CRCs). However, little is yet known of genes that directly contribute to tumor formation in such cancers. To characterize MSI-dependent changes in gene expression, we have now compared transcriptomes between fresh CRC specimens positive or negative for MSI (n=10 for each) with the use of high-density oligonucleotide microarrays harboring >44 000 probe sets. Correspondence analysis of the expression patterns of isolated MSI-associated genes revealed that the transcriptome of MSI+ CRCs is clearly distinct from that of MSI− CRCs. Such MSI-associated genes included that for AXIN2, an important component of the WNT signaling pathway. AXIN2 was silenced, apparently as a result of extensive methylation of its promoter region, specifically in MSI+ CRC specimens. Forced expression of AXIN2, either by treatment with 5′-azacytidine or by transfection with AXIN2 cDNA, resulted in rapid cell death in an MSI+ CRC cell line. These data indicate that epigenetic silencing of AXIN2 is specifically associated with carcinogenesis in MSI+ CRCs.

DNA microarray analysis of stage progression mechanism in myelodysplastic syndrome

Summary. Myelodysplastic syndrome (MDS) is a clonal disorder of haematopoietic stem cells. Despite the high incidence of MDS in the elderly, effective treatment of individuals in its advanced stages is problematic. DNA microarray analysis is a potentially informative approach to the development of new treatments for MDS. However, a simple comparison of ‘transcriptomes’ of bone marrow mononuclear cells among individuals at distinct stages of MDS would result in the identification of genes whose expression differences only reflect differences in the proportion of MDS blasts within bone marrow. Such a ‘population shift’ effect has now been avoided by purification of haematopoietic stem‐like cells that are positive for the cell surface marker AC133 from the bone marrow of healthy volunteers and 30 patients at various stages of MDS. Microarray analysis with the AC133+ cells from these individuals resulted in the identification of sets of genes with expression that was specific to either indolent or advanced stages of MDS. The former group of genes included that for PIASy, which catalyses protein modification with the ubiquitin‐like molecule SUMO. Induction of PIASy expression in a mouse myeloid cell line induced apoptosis. A loss of PIASy expression may therefore contribute directly to the growth of MDS blasts and stage progression.

Chromosome copy number analysis in screening for prognosis-related genomic regions in colorectal carcinoma

Colorectal carcinoma (CRC) remains the major cause of cancer death in humans. Although chromosomal structural anomaly is presumed to play an important role in the carcinogenesis of CRC, chromosomal copy number alterations (CNA) and loss of heterozygosity (LOH) have not yet been analyzed extensively at high resolution in CRC. Here we aim to identify recurrent CNA and LOH in human CRC with the use of single nucleotide polymorphism‐typing microarrays, and to reveal their relevance to clinical outcome. Surgically resected CRC specimens and paired normal mucosa were obtained from a consecutive series of 94 patients with CRC, and both of them were subjected to genotyping with Affymetrix Mapping 50K arrays. CNA and LOH were inferred computationally on every single nucleotide polymorphism site by integrating the array data for paired specimens. Our large dataset reveals recurrent CNA in CRC at chromosomes 7, 8, 13, 18, and 20, and recurrent LOH at chromosomes 1p, 4q, 5q, 8p, 11q, 14q, 15q, 17p, 18, and 22. Frequent uniparental disomy was also identified in chromosomes 8p, 17p, and 18q. Very common CNA and LOH were present at narrow loci of <1 Mbp containing only a few genes. In addition, we revealed a number of novel CNA and LOH that were linked statistically to the prognosis of the patients. The precise and large‐scale measurement of CNA and LOH in the CRC genome is efficient for pinpointing prognosis‐related genome regions as well as providing a list of unknown genes that are likely to be involved in CRC development. (Cancer Sci 2008; 99: 1835–1840)

Experimental trial for diagnosis of pancreatic ductal carcinoma based on gene expression profiles of pancreatic ductal cells.

Pancreatic ductal carcinoma (PDC) remains one of the most intractable human malignancies, mainly because of the lack of sensitive detection methods. Although gene expression profiling by DNA microarray analysis is a promising tool for the development of such detection systems, a simple comparison of pancreatic tissues may yield misleading data that reflect only differences in cellular composition. To directly compare PDC cells with normal pancreatic ductal cells, we purified MUC1‐positive epithelial cells from the pancreatic juices of 25 individuals with a normal pancreas and 24 patients with PDC. The gene expression profiles of these 49 specimens were determined with DNA microarrays containing >44 000 probe sets. Application of both Welchs analysis of variance and effect size‐based selection to the expression data resulted in the identification of 21 probe sets corresponding to 20 genes whose expression was highly associated with clinical diagnosis. Furthermore, correspondence analysis and 3‐D projection with these probe sets resulted in separation of the transcriptomes of pancreatic ductal cells into distinct but overlapping spaces corresponding to the two clinical classes. To establish an accurate transcriptome‐based diagnosis system for PDC, we applied supervised class prediction algorithms to our large data set. With the expression profiles of only five predictor genes, the weighted vote method diagnosed the class of samples with an accuracy of 81.6%. Microarray analysis with purified pancreatic ductal cells has thus provided a basis for the development of a sensitive method for the detection of PDC. (Cancer Sci 2005; 96: 387–393)

Proteomic analysis of hematopoietic stem cell-like fractions in leukemic disorders

DNA microarray analysis has been applied to identify molecular markers of human hematological malignancies. However, the relatively low correlation between the abundance of a given mRNA and that of the encoded protein makes it important to characterize the protein profile directly, or ‘proteome,’ of malignant cells in addition to the ‘transcriptome.’ To identify proteins specifically expressed in leukemias, here we isolated AC133+ hematopoietic stem cell-like fractions from the bone marrow of 13 individuals with various leukemic disorders, and compared their protein profiles by two-dimensional electrophoresis. A total of 11 differentially expressed protein spots corresponding to 10 independent proteins were detected, and peptide fingerprinting combined with mass spectrometry of these proteins revealed them to include NuMA (nuclear protein that associates with the mitotic apparatus), heat shock proteins, and redox regulators. The abundance of NuMA in the leukemic blasts was significantly related to the presence of complex karyotype anomalies. Conditional expression of NuMA in a mouse myeloid cell line resulted in the induction of aneuploidy, cell cycle arrest in G2–M phases, and apoptosis. These results demonstrate the potential of proteome analysis with background-matched cell fractions obtained from fresh clinical specimens to provide insight into the mechanism of human leukemogenesis.

DNA microarray analysis of in vivo progression mechanism of heart failure

Dahl salt-sensitive rats are genetically hypersensitive to sodium intake. When fed a high sodium diet, they develop systemic hypertension, followed by cardiac hypertrophy and finally heart failure within a few months. Therefore, Dahl rats represent a good model with which to study how heart failure is developed in vivo. By using DNA microarray, we here monitored the transcriptome of >8000 genes in the left ventricular muscles of Dahl rats during the course of cardiovascular damage. Expression of the atrial natriuretic peptide gene was, for instance, induced in myocytes by sodium overload and further enhanced even at the heart failure stage. Interestingly, expression of the gene for the D-binding protein, an apoptotic-related transcriptional factor, became decreased upon the transition to heart failure. To our best knowledge, this is the first report to describe the transcriptome of cardiac myocytes during the disease progression of heart failure.