The fatty acid synthase (FAS) gene is significantly up-regulated in various types of cancers, and blocking the FAS expression results in apoptosis of tumor cells. Therefore, FAS is considered to be an attractive target for anticancer therapy. However, the molecular mechanism by which the FAS gene is up-regulated in tumor cells is poorly understood. We found that FAS was significantly up-regulated by hypoxia, which was also accompanied by reactive oxygen species (ROS) generation in human breast cancer cell lines. The FAS expression was also activated by H 2 O 2 , whereas N-acetyl-L-cystein, a ROS inhibitor, suppressed the expression. We also found that the hypoxia significantly up-regulated sterol regulatory-element binding protein (SREBP)-1, the major transcriptional regulator of the FAS gene, via phosphorylation of Akt followed by activation of hypoxia-inducible factor 1 (HIF1). Moreover, our results of reporter assay and chromatin immunoprecipitation analysis indicate that SREBP-1 strongly bound to the SREBP binding site/E-box sequence on the FAS promoter under hypoxia. In our xenograft mouse model, FAS was strongly expressed in the hypoxic regions of the tumor. In addition, our results of immunohistochemical analysis for human breast tumor specimens indicate that the expressions of both FAS and SREBP-1 were colocalized with hypoxic regions in the tumors. Furthermore, we found that hypoxia-induced chemoresistance to cyclophosphamide was partially blocked by a combination of FAS inhibitor and cyclophosphamide. Taken together, our results indicate that FAS gene is up-regulated by hypoxia via activation of the Akt and HIF1 followed by the induction of the SREBP-1 gene, and that hypoxia-induced chemoresistance is partly due to the up-regulation of FAS.

To characterize the molecular feature in prostate carcinogenesis and the putative transition from prostatic intraepithelial neoplasia (PIN) to invasive prostate cancer (PC), we analyzed gene-expression profiles of 20 PCs and 10 high-grade PINs with a cDNA microarray representing 23,040 genes. Considering the histological heterogeneity of PCs and the minimal nature of PIN lesions, we applied laser microbeam microdissection to purify populations of PC and PIN cells, and then compared their expression profiles with those of corresponding normal prostatic epithelium also purified by laser microbeam microdissection. A hierarchical clustering analysis separated the PC group from the PIN group, except for three tumors that were morphologically defined as one very-high-grade PIN and two low-grade PCs, suggesting that PINs and PCs share some molecular features and supporting the hypothesis of PIN-to-PC transition. On the basis of this hypothesis, we identified 21 up-regulated genes and 63 down-regulated genes commonly in PINs and PCs compared with normal epithelium, which were considered to be involved in the presumably early stage of prostatic carcinogenesis. They included AMACR, OR51E2, RODH, and SMS. Furthermore, we identified 41 up-regulated genes and 98 down-regulated genes in the transition from PINs to PCs; those altered genes, such as POV1, CDKN2C, EPHA4, APOD, FASN, ITGB2, LAMB2, PLAU, and TIMP1, included elements that are likely to be involved in cell adhesion or the motility of invasive PC cells. The down-regulation of EPHA4 by small interfering RNA in PC cells lead to attenuation of PC cell viability. These data provide clues to the molecular mechanisms underlying prostatic carcinogenesis, and suggest candidate genes the products of which might serve as molecular targets for the prevention and treatment of PC.

CD82, also known as KAI1, was recently identified as a prostate cancer metastasis suppressor gene on human chromosome 11p1.2 (ref. 1). The product of CD82 is KAI1, a 40- to 75-kDa tetraspanin cell-surface protein also known as the leukocyte cell-surface marker CD82 (refs. 1,2). Downregulation of KAI1 has been found to be clinically associated with metastatic progression in a variety of cancers, whereas overexpression of CD82 specifically suppresses tumor metastasis in various animal models. To define the mechanism of action of KAI1, we used a yeast two-hybrid screen and identified an endothelial cell-surface protein, DARC (also known as gp-Fy), as an interacting partner of KAI1. Our results indicate that the cancer cells expressing KAI1 attach to vascular endothelial cells through direct interaction between KAI1 and DARC, and that this interaction leads to inhibition of tumor cell proliferation and induction of senescence by modulating the expression of TBX2 and p21. Furthermore, the metastasis-suppression activity of KAI1 was significantly compromised in DARC knockout mice, whereas KAI1 completely abrogated pulmonary metastasis in wild-type and heterozygous littermates. These results provide direct evidence that DARC is essential for the function of CD82 as a suppressor of metastasis.

Fatty acid synthase (FAS) has been found to be overexpressed in a wide range of epithelial tumors, including breast cancer. Pharmacologic inhibitors of FAS cause apoptosis of breast cancer cells and result in decreased tumor size in vivo. However, how the inhibition of FAS induces apoptosis in tumor cells remains largely unknown. To understand the apoptotic pathway resulting from direct inhibition of FAS, we treated breast tumor cells with or without FAS small interfering RNA (siRNA) followed by a microarray analysis. Our results indicated that the proapoptotic genes BNIP3, tumor necrosis factor-related apoptosis-inducing ligand (TRAIL), and death-associated protein kinase 2 (DAPK2) were significantly up-regulated on direct inhibition of the FAS gene. We also found that the knockdown of FAS expression significantly increased ceramide level in the tumor cells, and this increase was abrogated by acetyl-CoA carboxylase inhibitor. In addition, carnitine palmitoyltransferase-1 (CPT-1) inhibitor up-regulated the ceramide and BNIP3 levels in these cells, whereas treatment of tumor cells with FAS siRNA in the presence of a ceramide synthase inhibitor abrogated the upregulation of BNIP3 and inhibited apoptosis. Furthermore, we found that treatment of cells with BNIP3 siRNA significantly counteracted the effect of FAS siRNA-mediated apoptosis. Consistent with these results, a significant inverse correlation was observed in the expression of FAS and BNIP3 in clinical samples of human breast cancer. Collectively, our results indicate that inhibition of FAS in breast cancer cells causes accumulation of malonyl-CoA, which leads to inhibition of CPT-1 and up-regulation of ceramide and induction of the proapoptotic genes BNIP3, TRAIL, and DAPK2, resulting in apoptosis. (Cancer Res 2006; 66(11): 5934-40)

Interferon regulatory factors (IRFs) regulate transcription of interferon genes through DNA sequence-speci®c binding to these targets. Using a dierential display method for examining gene expression in p53-defective cells infected with adenovirus containing wild-type p53, we found that expression of interferon regulatory factor 5 (IRF-5) mRNA was increased in the presence of exogenous p53. An electrophoretic mobility-shift assay showed that a potential p53 binding site (p53BS) detected in exon 2 of the IRF-5 gene could in fact bind to p53 protein. Moreover, a heterologous reporter assay revealed that the p53BS possessed p53-dependent transcriptional activity. Expression of IRF-5 was induced in p53+/+ cells (MCF7 and NHDF), but not in p537/7 cells (H1299) when DNA was damaged by girradiation, UV-radiation, or adriamycin treatment in a wild-type p53-dependent manner. These results suggest that IRF-5 is a novel p53-target, and that it might mediate the p53-dependent immune response.