mTOR and ERRα are key regulators of common metabolic processes, including lipid homeostasis. However, it is currently unknown whether these factors cooperate in the control of metabolism. ChIP-sequencing analyses of mouse liver reveal that mTOR occupies regulatory regions of genes on a genome-wide scale including enrichment at genes shared with ERRα that are involved in the TCA cycle and lipid biosynthesis. Genetic ablation of ERRα and rapamycin treatment, alone or in combination, alter the expression of these genes and induce the accumulation of TCA metabolites. As a consequence, both genetic and pharmacological inhibition of ERRα activity exacerbates hepatic hyperlipidemia observed in rapamycin-treated mice. We further show that mTOR regulates ERRα activity through ubiquitin-mediated degradation via transcriptional control of the ubiquitin-proteasome pathway. Our work expands the role of mTOR action in metabolism and highlights the existence of a potent mTOR/ERRα regulatory axis with significant clinical impact.

The utility of the folate receptor (FR) type A, in a broad range of targeted therapies and as a diagnostic serum marker in cancer, is confounded by its variable tumor expression levels. FR-A, its mRNA and its promoter activity were coordinately up-regulated by the glucocorticoid receptor (GR) agonist, dexamethasone. Optimal promoter activation which occurred at <50 nmol/L dexamethasone was inhibited by the GR antagonist, RU486, and was enhanced by coactivators, supporting GR mediation of the dexamethasone effect. The dexamethasone response of the FR-A promoter progressed even after dexamethasone was withdrawn, but this delayed effect required prior de novo protein synthesis indicating an indirect regulation. The dexamethasone effect was mediated by the G/C-rich (Sp1 binding) element in the core P4 promoter and was optimal in the proper initiator context without associated changes in the complement of major Sp family proteins. Histone deacetylase (HDAC) inhibitors potentiated dexamethasone induction of FR-A independent of changes in GR levels. Dexamethasone/HDAC inhibitor treatment did not cause de novo FR-A expression in a variety of receptornegative cells. In a murine HeLa cell tumor xenograft model, dexamethasone treatment increased both tumor-associated and serum FR-A. The results support the concept of increasing FR-A expression selectively in the receptor-positive tumors by brief treatment with a nontoxic dose of a GR agonist, alone or in combination with a well-tolerated HDAC inhibitor, to increase the efficacy of various FR-A-dependent therapeutic and diagnostic applications. They also offer a new paradigm for cancer diagnosis and combination therapy that includes altering a marker or a target protein expression using general transcription modulators. (Cancer Res 2005; 65(10): 4431-41)

Nuclear expression of CCAAT enhancer binding protein-α (C/EBPα), which supports tissue differentiation through several antiproliferative protein–protein interactions, augurs terminal differentiation of prostate epithelial cells. C/EBPα is also a tumor suppressor, but in many tumors its antiproliferative interactions may be attenuated by de-phosphorylation. C/EBPα acts as a corepressor of the classical androgen response element (ARE)-mediated gene activation by the androgen receptor (AR), but this is paradoxical as the genotropic actions of AR are crucial not only for the growth of the prostate but also for its maintenance and function. We show that DNA-bound C/EPBα recruits AR to activate transcription. C/EBPα-dependent trans-activation by AR also overrode suppression of AREs by C/EBPα elsewhere in a promoter. This mechanism was remarkable in that its androgen dependence was apparently for nuclear translocation of AR; it was otherwise androgen independent, flutamide insensitive and tolerant to disruption of AR dimerization. Gene response profiles and global chromatin associations in situ supported the direct bimodal regulation of AR transcriptional signaling by C/EBPα. This unique mechanism explains the functional coordination between AR and C/EPBα in the prostate and also shows that hormone-refractory AR signaling in prostate cancer could occur through receptor tethering.

The androgen receptor (AR) plays a critical role in the development and progression of prostate cancer (PC) and is a major therapeutic target in this disease. The transcriptional activity of AR is modulated by the co-regulators with which it interacts and consequently deregulation of cofactor expression and/or activity impacts the expression of genes whose products can play a role in PC pathogenesis. Here we report that ELF3, a member of the ETS family of transcription factors, is a repressor of AR transcriptional activity. Exogenous expression of ELF3 represses AR transcriptional activity when assessed using reporter-based transfection assays or when evaluated on endogenous AR target genes. Conversely, ELF3 knockdown increases the AR transcriptional activity. Biochemical dissection of this activity indicates that it results from the physical interaction between ELF3 and AR and that this interaction inhibits the recruitment of AR to specific androgen response elements within target gene promoters. Significantly, we observed that depletion of ELF3 expression in LNCaP cells promotes cell migration while increased ELF3 expression severely inhibits tumor growth in vitro and in a mouse xenograft model. Taken together, these results suggest that modulation of ELF3 expression and/or AR/ELF3 interaction may have utility in the treatment of PC.

It has been suggested that ligand-dependent gene activation by the progesterone receptor (PR) can result from recruitment of PR by the promoter bound Sp1. A detailed investigation of the Sp1-dependent agonistic activity of RU486 and R5020 on the folate receptor (FR) type alpha, p27, thymidine kinase 1 and p21 genes reveals a different mechanism. The FR-alpha P4 promoter and the endogenous FR-alpha gene were up-regulated by the PR agonist R5020 through either PR-A or PR-B. The classical antagonist RU486 also activated the promoter but only through PR-B. The most proximal (essential) G/C-rich (Sp1 binding) element and the initiator region constituted the minimal promoter responsive to PR regulation; substitution with a stronger cluster of G/C-rich elements enhanced the magnitude of the PR response. In contrast, substitution of the G/C-rich element with a TATA box resulted in the loss of regulation by PR. Overexpression of Sp1 and Sp4 but not Sp3 enhanced activation of the FR-alpha promoter by PR, knocking down Sp1 decreased the activation in a manner that was reversed by ectopic Sp1 or Sp4. The ligand-dependent action of PR on the promoter was delayed compared with its activation of a classical glucocorticoid response element-driven promoter and activation of both the promoter and the endogenous FR-alpha gene by PR required new protein synthesis. Activation by PR paralleled RNA polymerase II recruitment but was not accompanied by either association of PR or a change in the association of Sp1 with the endogenous FR-alpha P4 promoter. Similar observations were made for PR regulation of the genes encoding p27, thymidine kinase 1, and p21. The results contradict the current view of Sp1-dependent gene regulation by PR and point to the existence of one or more PR target genes whose promoter and cell context(s) must thus be key determinants of the agonistic activity of RU486 on a large group of important Sp1-dependent downstream target genes.

A need for androgen response elements (AREs) for androgen receptor (AR)-dependent growth of hormone depletion-insensitive prostate cancer is generally presumed. In such cells, androgen-independent activation by AR of certain genes has been attributed to selective increases in basal associations of AR with putative enhancers. We examined the importance of AR binding to DNA in prostate cancer cells in which proliferation in the absence of hormone was profoundly (∼ 90%) dependent on endogenous AR and where the receptor was not up-regulated or mutated but was predominantly nuclear. Here, ARE-mediated promoter activation and the binding of AR to a known ARE in the chromatin remained entirely androgen dependent, and the cells showed an androgen-responsive gene expression profile with an unaltered sensitivity to androgen dose. In the same cells, a different set of genes primarily enriched for cell division functions was activated by AR independently of hormone and significantly overlapped the signature gene overexpression profile of hormone ablation-insensitive clinical tumors. After knockdown of endogenous AR, hormone depletion-insensitive cell proliferation and AR apoprotein-dependent gene expression were rescued by an AR mutant that was unable to bind to ARE but that could transactivate through a well-established AR tethering protein. Hormone depletion-insensitive AR binding sites in the chromatin were functional, binding, and responding to both the wild-type and the mutant AR and lacked enrichment for canonical or noncanonical ARE half-sites. Therefore, a potentially diverse set of ARE-independent mechanisms of AR interactions with target genes must underlie truly hormone depletion-insensitive gene regulation and proliferation in prostate cancer.

The FDA-approved DNA hypomethylating agents (DHAs) like 5-azacytidine (5AC) and decitabine (DAC) demonstrate efficacy in the treatment of hematologic malignancies. Despite previous reports that showed histone acetylation changes upon using these agents, the exact mechanism underpinning these changes is unknown. In this study, we investigated the relative potency of the nucleoside analogs and non-nucleoside analogs DHAs on DNA methylation reversal using DNA pyrosequencing. Additionally, we screened their effect on the enzymatic activity of the histone deacetylase sirtuin family (SIRT1, SIRT2, SIRT3, SIRT5 and SIRT6) using both recombinant enzymes and nuclear lysates from leukemia cells. The nucleoside analogs (DAC, 5AC and zebularine) were the most potent DHAs and increased the enzymatic activity of SIRT6 without showing any significant increase in other sirtuin isoforms. ChIP-Seq analysis of bone marrow cells derived from six acute myeloid leukemia (AML) patients and treated with the nucleoside analog DAC induced genome-wide acetylation changes in H3K9, the physiological substrate for SIRT6. Data pooling from the six patients showed significant acetylation changes in 187 gene loci at different chromosomal regions including promoters, coding exons, introns and distal intergenic regions. Signaling pathway analysis showed that H3K9 acetylation changes are linked to AML-relevant signaling pathways like EGF/EGFR and Wnt/Hedgehog/Notch. To our knowledge, this is the first report to identify the nucleoside analogs DHAs as activators of SIRT6. Our findings provide a rationale against the combination of the nucleoside analogs DHAs with SIRT6 inhibitors or chemotherapeutic agents in AML due to the role of SIRT6 in maintaining genome integrity and DNA repair.

Estradiol (E 2 ) acts through the estrogen receptor (ER) to downregulate many genes, and tamoxifen (Tam) largely reverses this repression but the underlying mechanisms are unclear. Repression of the folate receptor (FR)-a P4 core promoter by ER is enhanced by E 2 and reversed by Tam. This effect was unaffected by inhibition of new protein synthesis and required the E/F and the DNA-binding domains of ER without direct binding of ER to DNA. The repression by E 2 /ER was not specific for either Sp1 or TATA elements but was loosely selective for the initiator and flanking sequence. Insertion of a response element or a relatively strong Sp1 cluster to recruit ER upstream of the core promoters caused a switch to activation by E 2 /ER that was inhibited by Tam. In nuclear extracts, association of ER with a biotinylated core promoter fragment was promoted by E 2 but Tam blocked this effect. Repression/ de-repression of the P4 promoter and endogenous FR-a expression by E 2 /Tam required SMRT and/or NCoR. ER associated with the chromosomal P4 promoter and SMRT and NCoR associated with it in an ER-dependent manner; these associations were favored by E 2 but disrupted by Tam, in the short term, without changes in ER expression. TAFII30 was required for optimal P4 promoter activity and for the repressive association of ER. E2 may thus maintain a low transcriptional status of genes by favoring direct TAFII30-dependent association of ER with the core promoter in a co-repressor complex containing SMRT and/or NCoR; this repression is overridden in target genes containing an upstream element that strongly recruits ER. In addition to suppressing the activation of classical E2 target genes, Tam may upregulate genes by passively dissociating the ER co-repressor complex.