SIRT proteins play an important role in the survival and drug resistance of tumor cells, especially during chemotherapy. In this study, we investigated the potency, specificity, and cellular targets of three SIRT inhibitors, Sirtinol, Salermide, and EX527. Cell proliferative and cell cycle analyses showed that Sirtinol and Salermide, but not EX527, were effective in inducing cell death at concentrations of 50 μmol/L or over in MCF-7 cells. Instead, EX527 caused cell cycle arrest at G 1 at comparable concentrations. In vitro SIRT assays using a p53 peptide substrate showed that all three compounds are potent SIRT1/2 inhibitors, with EX527 having the highest inhibitory activity for SIRT1. Computational docking analysis showed that Sirtinol and Salermide have high degrees of selectivity for SIRT1/2, whereas EX527 has high specificity for SIRT1 but not SIRT2. Consistently, Sirtinol and Salermide, but not EX527, treatment resulted in the in vivo acetylation of the SIRT1/2 target p53 and SIRT2 target tubulin in MCF-7 cells, suggesting that EX527 is ineffective in inhibiting SIRT2 and that p53 mediates the cytotoxic function of Sirtinol and Salermide. Studies using breast carcinoma cell lines and p53-deficient mouse fibroblasts confirmed that p53 is essential for the Sirtinol and Salermideinduced apoptosis. Further, we showed using small interfering RNA that silencing both SIRTs, but not SIRT1 and SIRT2 individually, can induce cell death in MCF-7 cells. Together, our results identify the specificity and cellular targets of these novel inhibitors and suggest that SIRT inhibitors require combined targeting of both SIRT1 and SIRT2 to induce p53 acetylation and cell death. Mol Cancer Ther; 9(4); 844-55. ©2010 AACR.

Endometrial cancer is the most common malignancy of the lower female reproductive tract. The tumor suppressor FOXO1 is downregulated in endometrial cancer compared with normal endometrium but the underlying mechanisms are not well understood. Using microRNA (miR) target prediction algorithms, we identified several miRs that potentially bind the 3′-untranslated region of FOXO1 transcripts. Expression profiling of normal and malignant endometrial samples by quantitative real-time PCR and Northern blot analysis revealed an inverse correlation between the levels of FOXO1 protein and the abundance of several of the in silico-predicted miRs, suggesting that loss of FOXO1 expression in endometrial cancer may be mediated by miRs. To determine the role of candidate miRs, we used the endometrial cancer cell lines HEC-1B and Ishikawa, which express FOXO1 at high and low levels, respectively. Expression of miR-9, miR-27, miR-96, miR-153, miR-182, miR-183, or miR-186, but not miR-29a, miR-128, miR-152, or miR-486 mimetics in HEC-1B cells was sufficient to significantly reduce the abundance of FOXO1. Conversely, FOXO1 expression was efficiently restored in the Ishikawa cell line upon simultaneous inhibition of miR-9, miR-27, miR-96, miR-153, miR-183, and miR-186. Moreover, induction of FOXO1 in Ishikawa cells by miR inhibitors was accompanied by G 1 cell cycle arrest and cell death, and was attenuated by the small interfering RNA-mediated downregulation of FOXO1 expression. Our findings identify several miRs overexpressed in endometrial cancer that function in concert to repress FOXO1 expression. Further, aberrant miR expression results in deregulated cell cycle control and impaired apoptotic responses, and thus, may be central to endometrial tumorigenesis.

Vascular endothelial growth factor (VEGF) plays a central role in breast cancer development and progression, but the mechanisms that control its expression are poorly understood. Breast cancer tissue microarrays revealed an inverse correlation between the Forkhead transcription factor FOXO3a and VEGF expression. Using the lapatinib-sensitive breast cancer cell lines BT474 and SKBR3 as model systems, we tested the possibility that VEGF expression is negatively regulated by FOXO3a. Lapatinib treatment of BT474 or SKBR3 cells resulted in nuclear translocation and activation of FOXO3a, followed by a reduction in VEGF expression. Transient transfection and inducible expression experiments showed that FOXO3a represses the proximal VEGF promoter whereas another forkhead member, FOXM1, induces VEGF expression. Chromatin immunoprecipitation and oligonucleotide pull-down assays demonstrated that both FOXO3a and FOXM1 bind a consensus Forkhead response element (FHRE) in the VEGF promoter. Upon lapatinib stimulation, activated FOXO3a displaces FOXM1 bound to the FHRE before recruiting histone deacetylase 2 (HDAC2) to the promoter, leading to decreased histones H3 and H4 acetylation, and concomitant transcriptional inhibition of VEGF. These results show that FOXO3a-dependent repression of target genes in breast cancer cells, such as VEGF, involves competitive displacement of DNA-bound FOXM1 and active recruitment of transcriptional repressor complexes.