'Junk' DNA meets the p53 network.

A major part of the genome of higher eukaryotes consists of non-coding sequences. In former times, these sequences were called ‘junk-DNA' as no specific function could be attributed to them. More recent research has shown that small non-coding RNAs are contained in these parts of the genome. These non-coding RNAs have a fundamental role in gene regulation.

Regulation of gene expression in eukaryotes is multifaceted and additional layers of complexity are constantly discovered. In a recent article published in Molecular Systems Biology, Varda Rotter et al describe a network that combines several of such regulatory layers by integrating the concerted actions of the transcription factors p53, E2F1, their respective targets and three clusters of microRNAs (Brosh ).

MicroRNAs (miRNAs) are a relatively recently identified means for gene regulation. They are small, endogenous non-coding RNAs, between 19 and 25 nt in length. In some cases, miRNAs are organised in clusters and transcribed as polycistrons, which are cleaved in the nucleus into shorter precursor miRNAs. These precursor miRNAs are exported into the cytoplasm where they are further processed. Mature miRNAs are loaded onto the RISC complex, which guides them to their mRNA target. Usually, miRNAs base-pair imperfectly in the untranslated region of their target mRNA, and control gene expression at the post-transcriptional level (reviewed in Le Sage and Agami, 2006). Unlike siRNA, miRNAs are of endogenous origin and alterations in their expression are associated with a number of diseases, including cancer.

One example of how miRNAs contribute to the control of malignant transformation is demonstrated in the work by Ran Brosh and co-workers who investigated three clusters of miRNA, including the miRs-106b/93/25, miR-17-92 and miR-106a-92 polycistrons. These three clusters are coordinately regulated by the activity of the transcription factor p53, one of the key tumour suppressor proteins. Together with the retinoblastoma (pRb) protein, p53 controls G1/S transition during the cell cycle. Transition of cells into S phase requires the activity of the E2F1 transcription factor. Normally, this protein is sequestered and kept inactive by the pRb protein. At the end of G1, pRb is phosphorylated by cyclin-dependent kinases (CDKs), which release E2F1 from its constraint (Sun ; Calzone ). The transcription factor p53 controls pRb phosphorylation by inducing expression of p21, a CDK inhibitor. In fact, p21 is one of the many p53 targets, which collectively mediate the antiproliferative and pro-apoptotic response induced by oncogenic stress signals or by cellular senescence (Riley ).

Ran Brosh and co-workers show that the three miRNA clusters above (miR-17-92, miR-106b/93/25 and miR-106a-92) are downregulated by p53 and that this set of miRNAs also targets a group of antiproliferative factors, including pRb, p21 and p57. Their regulation by p53 turns out, however, to be indirect. Thus, none of the three miRNA clusters harbours a p53-binding site. Instead, they are controlled by the transcription factor E2F1. Strikingly, genes that are silenced by these E2F1-regulated miRNAs are themselves directly controlled by E2F1. The transcription factor E2F1, the miRNA cluster and their common target genes appear, therefore, to form a regulatory feed-forward loop (Figure 1). In unstressed proliferating cells, p53 is inactive and pRb is phosphorylated. E2F1 can therefore drive transcription of antiproliferative genes, while at the same time it ensures that only low levels of protein are made through the activation of transcription of corresponding miRNAs (Figure 1A). Conversely, in senescent cells, p53 is active, which leads to the interruption of this loop: on activation, p53 reduces E2F1 activity by stimulating transcription of the p21 gene, which prevents phosphorylation of pRb thus keeping E2F-1 in check (Figure 1B). In addition, p53 enhances the transcription of miRNAs from the miR-34 cluster, which controls E2F1 abundance (Kumamoto ). As a consequence, transcription of the miRNA clusters is no longer activated by E2F1 and expression of their targets is therefore upregulated.

Figure 1

In unstressed proliferating cells, p53 is inactive, whereas E2F1 induces expression of antiproliferative genes. At the same time, E2F1 stimulates transcription of a set of miRNAs that silence those antiproliferative factors (A). During replicative senescence, p53 is activated and stimulates expression of antiproliferative genes. Simultaneously, it inhibits E2F1 activity, thus preventing transcription of miRNAs that target antiproliferative regulators (B).

With a detailed analysis of this regulatory network, Ran Brosh and co-workers have not only characterised the interplay of several components for gene regulation, they have also identified a new mechanism by which p53 inhibits proliferation and controls cell fate. Nevertheless, several questions remain open. Although the E2F-centered feed-forward loop motif may appear at first sight as potentially wasteful, this motif has been proposed to have potential consequences on regulatory kinetics and sensitivity to noise (Shalgi ). What are the physiological implications of this specific arrangement within the context of p53-regulated proliferation? Furthermore, why does p53 interrupt only this feed-forward loop during senescence and not when it is activated by other conditions such as treatment with doxorubicin? Certainly, future investigations will provide us with answers to these questions. Moreover, with the current rapid advances in genomics and proteomics and further developments in our understanding of miRNA biology, we will hopefully gain deeper insight into the physiological significance of increasingly larger networks that combine multiple layers of regulation into integrated cellular decision mechanisms.