Expression of 14-3-3 () is induced in response to DNA damage, and causes cells to arrest in G2. By SAGE (serial analysis of gene expression) analysis, we identified as a gene whose expression is 7-fold lower in breast carcinoma cells than in normal breast epithelium. We verified this finding by Northern blot analysis. Remarkably, mRNA was undetectable in 45 of 48 primary breast carcinomas. Genetic alterations at such as loss of heterozygosity were rare (1͞20 informative cases), and no mutations were detected (0͞34). On the other hand, hypermethylation of CpG islands in the gene was detected in 91% (75͞82) of breast tumors and was associated with lack of gene expression. Hypermethylation of is functionally important, because treatment of -non-expressing breast cancer cell lines with the drug 5-aza-2-deoxycytidine resulted in demethylation of the gene and synthesis of mRNA. Breast cancer cells lacking expression showed increased number of chromosomal breaks and gaps when exposed to ␥-irradiation. Therefore, it is possible that loss of expression contributes to malignant transformation by impairing the G 2 cell cycle checkpoint function, thus allowing an accumulation of genetic defects. Hypermethylation and loss of expression are the most consistent molecular alterations in breast cancer identified so far.A lthough many studies have identified critical genetic and epigenetic changes that mark the transformation of cells in tissues such as colon, pancreas, and lung, similar studies in breast cancer have met with limited success. In this paper we report the identification of a gene, 14-3-3 (), whose expression is undetectable in 94% (45͞48) of breast tumors.was originally identified as an epithelial-specific marker, HME1, which was down-regulated in a few breast cancer cell lines but not in cancer cell lines derived from other tissue types (1). Later studies showed that protein (also called stratifin) was abundant in differentiated squamous epithelial cells, but decreased by 95% in simian virus 40-transformed epithelial cells and in primary bladder tumors (2-4).We investigated the molecular mechanism underlying the low expression of in breast cancers. We find that genetic alterations such as loss of heterozygosity (LOH) and intragenic mutations are not major contributing mechanisms for lack of expression. Instead, we show that hypermethylation of the CpG-rich region in the gene is associated with its transcriptional silencing in the majority of breast tumors. Treatment of breast cancer cell lines that do not express with the DNA methyltransferase inhibitor, 5-aza-2Ј-deoxycytidine (5-aza-dC), leads to partial demethylation of this CpG island and synthesis of mRNA. Thus, hypermethylation appears to be responsible for silencing of gene expression.Recent studies have shed light on the function of . It was originally identified as a p53-inducible gene that is responsive to DNA damaging agents (5). apparently sequesters the mitotic initiation complex, cdc2-cyclin B1, in the cytoplasm after DNA damage (6). This prevents cdc...

Claudins are transmembrane proteins that seal tight junctions, and are critical for maintaining cell-to-cell adhesion in epithelial cell sheets. However, their role in cancer progression remains largely unexplored. Here, we report that Claudin-7 (CLDN-7) expression is lower in invasive ductal carcinomas (IDC) of the breast than in normal breast epithelium, as determined by both RT-PCR (9/10) and Western analysis (6/8). Immunohistochemical (IHC) analysis of ductal carcinoma in situ (DCIS) and IDC showed that the loss of CLDN-7 expression correlated with histological grade in both DCIS (Po0.001, n ¼ 38) and IDC (P ¼ 0.014, n ¼ 31), occurring predominantly in high-grade (Nuclear and Elston grade 3) lesions. Tissue array analysis of 355 IDC cases further confirmed the inverse correlation between CLDN-7 expression and histological grade (P ¼ 0.03). This pattern of expression is consistent with the biological function of CLDN-7, as greater discohesion is typically observed in high-grade lesions. In line with this observation, by IHC analysis, CLDN-7 expression was lost in the vast majority (13/17) of cases of lobular carcinoma in situ, which is defined by cellular discohesion. In fact, inducing disassociation of MCF-7 and T47D cells in culture by treating with HGF/scatter factor resulted in a loss of CLDN-7 expression within 24 h. Silencing of CLDN-7 expression correlated with promoter hypermethylation as determined by methylation-specific PCR (MSP) and nucleotide sequencing in breast cancer cell lines (3/3), but not in IDCs (0/5). In summary, these studies provide insight into the potential role of CLDN-7 in the progression and ability of breast cancer cells to disseminate.

Expression of the p53 gene protects cells against malignant transformation. Whereas control of p53 degradation has been a subject of intense scrutiny, little is known about the factors that regulate p53 synthesis. Here we show that p53 messenger RNA levels are low in a large proportion of breast tumours. Seeking potential regulators of p53 transcription, we found consensus HOX binding sites in the p53 promoterS. Transient transfection of Hox/HOXA5 activated the p53 promoter. Expression of HOXA5 in epithelial cancer cells expressing wild-type p53, but not in isogenic variants lacking the p53 gene, led to apoptotic cell death. Moreover, breast cancer cell lines and patient tumours display a coordinate loss of p53 and HOXA5 mRNA and protein expression. The HOXA5 promoter region was methylated in 16 out of 20 p53-negative breast tumour specimens. We conclude that loss of expression of p53 in human breast cancer may be primarily due to lack of expression of HOXA5.

We have identi®ed 14-3-3 s (s) as a gene whose expression is lost in breast carcinomas, primarily by methylation-mediated silencing. In this report, we investigated the timing of loss of s gene expression during breast tumorigenesis in vivo. We analysed the methylation status of s in breast cancer precursor lesions using microdissection for selective tissue sampling. We found hypermethylation of s in 24 of 25 carcinomas (96%), 15 of 18 (83%) of ductal carcinoma in situ, and three of eight (38%) of atypical hyperplasias. None of the ®ve hyperplasias without atypia showed s-hypermethylation. Unexpectedly, patients with breast cancer showed s hypermethylation in adjacent histologically normal breast epithelium, while this was never observed in individuals without evidence of breast cancer. Also, samples of periductal stromal breast tissue were consistently hypermethylated, underscoring the importance of selective tissue sampling for accurate assessment of 14-3-3-s methylation in breast epithelium. These results suggest that hypermethylation of 14-3-3-s occurs at an early stage in the progression to invasive breast cancer, and may occur in apparently normal epithelium adjacent to breast cancer. These results provide evidence that loss of expression of s is an early event in neoplastic transformation. Oncogene (2001) 20, 3348 ± 3353.

Little is known about epigenetic silencing of genes by promoter hypermethylation in lobular breast cancers. The promoter methylation status of 5 cancer-related genes (RASSF1A, HIN-1, RAR-␤, Cyclin D2 and Twist) was evaluated in 2 types of lobular cancers, in situ (LCIS) and invasive lobular carcinomas (ILC) (n ‫؍‬ 32), and compared to ductal in situ (DCIS) and invasive (IDC) breast cancers (n ‫؍‬ 71). By using methylation-specific PCR (MSP), 100% of ILC and 69% of LCIS cases were found to have 1 or more hypermethylated genes among the panel of 5 genes (compared to 100% IDC and 95% of DCIS). Two or more hypermethylated genes were detected per tumor in 79% of invasive and 61% of in situ lobular carcinomas compared to 81% of IDC and 77% of DCIS. By contrast, DNA from nearly all normal reduction mammoplasty tissues (n ‫؍‬ 8) was unmethylated for the 5 genes. The methylation profiles of lobular vs. ductal carcinomas with respect to RASSF1A, Cyclin D2, RAR␤, and Hin-1 genes were similar, suggesting that gene silencing by promoter hypermethylation is likely to be important in both groups of diseases. Distinctly different, Twist was hypermethylated less often in ILC (16%, 3/19 cases) than in IDC (56%, 15/27 cases) (p ‫؍‬ 0.01). These results suggest that these 2 types of tumors share many common methylation patterns and some molecular differences. Additional studies might lend further understanding into the etiology and clinical behavior of this tumor type. © 2003 Wiley-Liss, Inc. Key words: methylation; breast cancer; lobular; ductal; DCIS; LCISComprehensive gene analyses such as serial analysis of gene expression (SAGE) and microarray analysis performed on breast cancer tissues revealed the expression profiles of thousands of genes and resulted in the identification of messenger RNAs that are over-and underexpressed in breast carcinomas compared to normal breast tissue. 1,2 We and others found that a major mode of tumor-specific downregulation of a number of these genes is by DNA hypermethylation. [3][4][5][6][7][8] The hypermethylation of cytosine residues in CpG-rich islands present in the promoter region of genes is an epigenetic alteration that leads to heritable changes in gene expression without changing the DNA sequence, most likely through the formation of repressive chromatin structure. Aberrant methylation of DNA is therefore believed to be an alternative pathway to cancer. 9 Studies of DNA hypermethylation in breast carcinoma have identified certain key genes as targets for epigenetic downregulation, 9 -13 including receptors such as the estrogen receptor 6,14,15 23 Identification of these genes has helped begin to elucidate the molecular pathogenesis of breast carcinoma, as it reveals intracellular pathways that are altered that likely contribute to oncogenesis. In addition, identification of methylated genes provides a potential target for molecular detection of breast carcinoma, as even small amounts of methylated sequences are readily detectable by methylation-specific PCR (MSP). 24 Because of the relat...

No abstract

A new tumor suppressor gene PTEN/MMAC1 was recently isolated at chromosome 10q23 and found to be inactivated by point mutation or homozygous deletion in glioma, prostate and breast cancer. PTEN/MMAC1 was also identi®ed as the gene predisposing to Cowden disease, an autosomal dominant cancer predisposition syndrome associated with an increased risk of breast, skin and thyroid tumors and occasional cases of other cancers including bladder and renal cell carcinoma. We screened 345 urinary tract cancers by microsatellite analysis and found chromosome 10q to be deleted in 65 of 285 (23%) bladder and 15 of 60 (25%) renal cell cancers. We then screened the entire PTEN/MMAC1 coding region for mutation in 25 bladder and 15 renal cell primary tumors with deletion of chromosome 10q. Two somatic point mutations, a frameshift and a splicing variant, were found in the panel of bladder tumors while no mutation was observed in the renal cell carcinomas. To screen for homozygous deletion, we isolated two polymorphic microsatellite repeats from genomic BAC clones containing the PTEN/MMAC1 gene. Using these new informative markers, we identi®ed apparent retention at the gene locus indicative of homozygous deletion of PTEN/MMAC1 in four of 65 bladder and 0 of 15 renal cell tumors with LOH through chromosome 10q. Identi®cation of the second inactivation event in six bladder tumors with LOH of 10q implies that the PTEN/MMAC1 gene is occasionally involved in bladder tumorigenesis. However, the low frequency of biallelic inactivation suggests that either PTEN/MMAC1 is inactivated by other mechanisms or it is not the only target of chromosome 10q deletion in primary bladder and renal cell cancer.

Klotho is a single pass transmembrane protein, associated with premature aging. We identified tumor suppressor activities for klotho, associated with reduced expression in breast cancer. We now aimed to analyze klotho expression in early stages of breast tumorigenesis and elucidate mechanisms leading to klotho silencing in breast tumors. We studied klotho expression, using immunohistochemistry, and found high klotho expression in all normal and mild hyperplasia samples, whereas reduced expression was associated with moderate and atypical ductal hyperplasia. Promoter methylation and histone deacetylation were studied as possible mechanisms for klotho silencing. Using bisulfite sequencing, and methylation-specific PCR, we identified KLOTHO promoter methylation in five breast cancer cell lines and in hyperplastic MCF-12A cells, but not in the non-tumorous mammary cell line HB2. Importantly, methylation status inversely correlated with klotho mRNA levels, and treatment of breast caner cells with 5-aza-2-deoxycytidine elevated klotho expression by up to 150-fold. KLOTHO promoter methylation was detected in 8/23 of breast cancer samples but not in normal breast samples. Chromatin immunoprecipitation revealed that in HB2 KLOTHO promoter was enriched with AcH3K9; however, in breast cancer cells, H3K9 was deacetylated, and treatment with the histone deacetylase inhibitor suberoylanilide bishydroxamide (SAHA) restored H3K9 acetylation. Taken together, these data indicate loss of klotho expression as an early event in breast cancer development, and suggest a role for DNA methylation and histone deacetylation in klotho silencing. Klotho expression and methylation may, therefore, serve as early markers for breast tumorigenesis.