Date Published: May 12, 2013
Publisher: Impact Journals LLC
Author(s): Ruchi Singhal, Jonathan E. Bard, Norma J. Nowak, Michael J. Buck, Eugene S. Kandel.
Phosphoinositol-3-kinase (PI3K) pathway is a crucial modulator of many physiological and pathophysiological phenomena, including aging, diabetes and cancer. Protein kinase Akt, a downstream effector of PI3K, controls a plethora of cellular functions, including gene transcription. A key mechanism connecting Akt activity to changes in gene expression is inhibitory phosphorylation of FOXO family of transcription factors. Accordingly, altered expression of FOXO targets may account for many biological consequences of PI3K/Akt signaling. While the previous efforts focused on FOXO-dependent regulation of protein-coding genes, non-coding RNA genes have emerged as equally important targets of many transcription factors. Therefore, we utilized a regulated form of FOXO1 to profile FOXO1-dependent changes in miRNA expression in human cells. Both microarray hybridization and next-generation sequencing revealed changes in the products of a miRNA cluster on X chromosome. Rapid induction of these miRNAs occurred independently of de novo protein synthesis. Furthermore, inhibition of PI3K in cancer cell lines caused derepression of these miRNAs, as would be expected for FOXO-regulated genes. Members of the major oncogenic cascades are significantly overrepresented among the predicted targets of the miRNAs, consistent with tumor-suppressive role of FOXO1. The discovered miRNAs represent new candidate mediators of FOXO1 functions and possible biomarkers of its activity.
Activation of the signal transduction cascade that proceeds through phosphoinisitol-3-phosphate kinase and protein kinase B (also known as AKT) is a frequent theme in signaling by various growth factors and a common subject of oncogenic alteration in cancer cells [1, 2]. A detailed knowledge of the consequences of activation of this pathway is essential for a better understanding of the function of a normal cell and may provide insights into the nature and possible mitigation of various disease states, including cancer, diabetes, and many others. Activation of AKT effects cellular changes through two broad groups of mechanisms. It modulates a large cohort of biochemical processes in a manner independent of de novo RNA synthesis. It also affects the levels of various cellular components through controlling the functions of transcription factors. Arguably, the best characterized AKT targets among the transcription factors are the proteins from the FOXO family . AKT phosphorylates FOXO proteins on a series of highly conserved sites, leading to nuclear exclusion and inactivation of these molecules . It is commonly accepted that many of the hallmark effects of AKT activation are directly attributable to FOXO inhibition, and that reactivation of FOXO influences cell response to therapies that target PI3K pathway . This highlights the need to identify and investigate the targets of these transcription factors, especially in the view of the role of PI3K-AKT-FOXO connection in cancer, ageing and diabetes.
We have engineered a derivative of human embryonic kidney cell line HEK-293T for the expression of FOXO1-AAA-ER, a regulated form of human FOXO1 (also known as FKHR). The expressed protein contains T24A S256A S319A mutations, which make it insensitive to inhibition by AKT . Furthermore, the protein is produced as a fusion with the ligand-binding domain of mouse estrogen receptor (ER). Therefore, the entire protein remains sequestered in the cytoplasm, until a suitable ligand is added. An additional mutation ensures that the ER fragment retains affinity to some artificial ligands (e.g. 4-hydroxitamoxifen or 4HT), but not to the natural estrogens . Overall, the system provides tight regulation of FOXO1 activity under regular cell culture conditions in the presence of fetal bovine serum, which might contain Akt-activating factors or estrogen.
Our findings demonstrate that expression of a cluster of miRNAs on chromosome X is susceptible to changes in the activity of FOXO1 transcription factor. The phenomenon depends on transcription by RNA polymerase II, but not de novo protein synthesis, suggesting that FOXO1 acts as a classical transcription factor and an immediate regulator of the miRNA expression. At present, the full structure of the corresponding transcripts is unknown. The miRNA may be generated from a single primary transcript or from multiple RNAs, each with its separate promoter. In our experiments we did not observe induction of miR-514b (data not shown), which is located approximately 10 kbps upstream of miR-508. Thus, it is tempting to speculate that at least one of the relevant promoters is located within that region. However, this does not necessarily mean that the relevant FOXO-binding sites have to be located within that area: there are ample examples of genes being affected by the DNA binding events at considerable distances [23, 24]. In fact, a recent study has indicated that FOXO3, which is functionally similar to FOXO1, may transactivate its target genes from distant enhancers via DNA looping .