Our current view of the nucleolus has been shaped by the concept that the organization of this prominent compartment within the nucleus is primarily dictated by its function, the making of ribosome subunits. Whether ribosome biogenesis is framed by a dedicated nucleolar scaffold has remained unclear. In this issue of The EMBO Journal, Caudron‐Herger and colleagues present evidence for a nucleolar skeleton composed of non‐coding RNA enriched in Alu repeat elements.
See also: M Caudron‐Herger et al (November 2015)
Production of ribosomes is vital for life, and eukaryotic cells evolved a specialized compartment, the nucleolus, where ribosomal RNA genes are transcribed, pre‐ribosomal RNA is processed, and pre‐ribosomes are assembled (Pederson, 2010). Mammalian cells typically contain a few nucleoli that disappear during mitosis and form de novo in early G1 around ribosomal genes (rDNA). In humans, long stretches of rDNA arrays located on five distinct chromosomes coalesce to form between one and three nucleoli (Savino et al, 2001). The rDNA transcription units coupled with binding sites for the HMG‐box protein UBF are sufficient to direct formation of a nucleolus (Grob et al, 2013). However, what drives the clustering of rDNA transcription units into only one or a few nucleoli per nucleus remains elusive. The new study by Caudron‐Herger and colleagues identifies non‐coding RNAs enriched in Alu repeat elements as potentially involved in this process.
There are over 1 million copies of Alu elements in the human genome, many of which localize within introns of protein‐coding genes (Deininger, 2011). According to Caudron‐Herger et al (2015), stable short intron‐derived RNAs containing Alu repeat elements accumulate in the nucleolus and are essential for its integrity. In the absence of so‐called aluRNAs, nucleoli dispersed into smaller domains and rDNA transcription was severely reduced. Nascent intronic Alu RNAs were previously shown to escape degradation after pre‐mRNA splicing, being processed and packaged into metabolically stable small ribonucleoprotein particles (Jády et al, 2012). However, these previously described RNA species were exclusively detected in the nucleoplasm (Jády et al, 2012).
Caudron‐Herger et al (2015) also carried out RNA tethering experiments using a lacO array consisting of multiple copies of the binding sequence for the lac repressor protein, LacI (Shevtsov & Dundr, 2011). AluRNA was tagged with MS2 stem loops and immobilized at the lacO array through binding to MS2 protein fused to LacI. This resulted in recruitment of the lacO genomic locus to the nucleolus. Nucleolin and nucleo‐phosmin, two abundant structural proteins of the nucleolus, were further shown to interact in vivo with aluRNAs. Based on these observations, the following model was proposed: (i) pre‐mRNA splicing of protein‐coding genes containing intronic Alu repeat elements generates small RNA species termed aluRNAs; (ii) these RNAs act as a scaffold that recruits nucleolin and nucleophosmin; and (iii) interactions between nucleolin, nucleophosmin, and UBF result in coalescence of rDNA arrays that would otherwise disperse throughout the nucleus (Fig 1).
A remarkable feature of this model is that it explains the long‐debated conundrum that specific inhibition of RNA polymerase II (Pol II) causes disintegration of nucleoli, although very little Pol II activity is detected in this nuclear compartment. Inevitably, however, this new model raises a number of open questions. For example, Alu elements are primate specific, and although targeted disruption of Alu‐related B1‐containing RNAs causes nucleolar dispersion in mouse cells (Caudron‐Herger et al, 2015), it is intriguing how distinct RNA species co‐evolved a common nucleolar scaffolding role. It is also puzzling that a subset of intron‐encoded Alu RNAs generate nucleoplasmic box H/ACA RNPs (Jády et al, 2012), while others are essential for nucleolar organization. Additional mechanistic issues concern specificity of the interaction between nucleolar aluRNAs and the very abundant nuclear proteins nucleolin and nucleophosmin. A final remaining enigma is how nucleolar aluRNAs are functionally coupled to Pol I activity. Clearly, the aluRNA‐scaffold model paves the way for further research that will likely expand our understanding of how ribosome biogenesis is controlled by nucleolar biology.
- © 2015 The Author