RPs (ribosomal proteins) are main components of the ribosome having essential functions in its biogenesis, function and structural integrity. Although most of the RP molecules are in the cytoplasm, being incorporated into translating ribosomes, some RPs have non-ribosomal functions when they are off ribosomal subunits. Notably, in eukaryotes, RPs are also present at transcription sites and some of these proteins have a function in transcription and pre-mRNA processing of specific genes. Although the consensus is that the proteins found at these sites are isolated RPs not assembled into ribosomal subunits, it has been proposed that ribosomal subunits might also be present. In the present paper, we review the available evidence for RPs at transcription sites and conclude that ribosomal subunits might be present, but additional studies will be required to solve this important issue.
- ribosomal protein
Ribosomal functions of RPs (ribosomal proteins)
Although rRNA contributes mostly to the structure and function of the ribosome, RPs are also important for viability in all cell types. RPs are mostly located on the surface of the ribosome, typically they contain an exposed N-terminal globular domain and a long C-terminal domain extending inwards to the rRNA [1–4]. RPs shape the rRNA into the correct tertiary structure and have essential functions in translation [5,6]. A second important function of RPs is in ribosome biogenesis, where they have specific roles in rRNA maturation, export and ribosomal subunit biogenesis [7–11]. RPs are some of the most abundant proteins in both prokaryote and eukaryotic cells and they are typically present stoichiometrically in the ribosome. These proteins are evolutionarily well conserved: approx. 30% of Escherichia coli RPs have orthologues in higher eukaryotes and archaea; archaeal ribosomes have an additional 30% RPs in common with eukaryotes [12,13]. Most RP genes in Saccharomyces cerevisiae and Schizosaccharomyces pombe are duplicated; the protein paralogues are either identical or very similar, yet deletions on the two paralogues have different phenotypes in S. cerevisiae . It has been suggested that there are different ribosome subtypes in the cell, distinguishable by the RP paralogues they carry . An alternative explanation for the different phenotypes is that one of the RP paralogues might have evolved a specific extra ribosomal function. In Drosophila and other multicellular organisms, heterozygous mutations in RP genes often lead to developmental irregularities [15,16]. Although these phenotypes can be explained by an insufficient level of ribosomes, in some instances they might also be due to extraribosomal functions [17–19].
In eukaryotes, newly synthesized RPs are rapidly imported into the nucleus and incorporated into pre-ribosomal subunits at the nucleolus, where they bind to the rRNA co-transcriptionally . The nucleolus is divided into three morphologically distinct subcompartments: transcription, early processing and rRNA modifications occur within the innermost FC (fibrillar component) and the surrounding DFC (dense fibrillar component), and RPs bind the rRNA either in the DFC or in the outer GC (granular component) layer [21,22]. A previous study shows that RPs accumulate in the nucleolus much more rapidly than other nucleolar proteins . However, the rate at which RPs are imported in the nucleus is higher than the rate at which they are exported back to the cytoplasm (presumably as ribosomal subunits) ; it appears that a large fraction of RPs is destroyed by the proteasome in the nucleus . Yet some RPs must escape degradation because many RPs are found to be associated with pre-mRNAs and proteins in the nucleus.
Non-ribosomal functions of RPs
Besides their main role in ribosome function and biogenesis, extraribosomal functions for RPs have been reported across organisms, from prokaryotes to eukaryotes [17–19]. There are several reports of RPs that bind their own mRNAs or pre-mRNAs and negatively autoregulate their own expression by affecting translation, splicing or transcription [17–19]. For example, RpS13 in mammalian cells represses its own gene expression by inhibiting splicing; RpS13 binds its own pre-mRNA close to the splice sites of the first intron and probably prevents spliceosome assembly (Figure 1, upper left panel) . Similar regulatory mechanisms have been reported for other RPs in both humans and yeast [25–28]. Notably, in mammalian cells, there is also evidence of RPs that regulate genes other than their own, by binding directly to specific transcription factors. RpL11, for instance, associates with a defined domain of the oncoprotein c-Myc and inhibits transcription activation of c-Myc target genes (Figure 1, lower left panel); the effect appears to be specific for RpL11 since other RPs did not show similar activities [29,30]. Similarly, it has also been reported that RpS3 specifically associates with the NF-κB (nuclear factor κB) DNA-binding protein complex and stabilizes the association of this transcription factor to its target sites . Interestingly, RpS3 functions also as a DNA glycosylase and interacts with OGG1 (8-oxoguanine DNA glycosylase 1), an enzyme involved in oxidative-stress-associated DNA repair [32–35]. In summary, there is abundant evidence for RPs not assembled into ribosomes having additional functions unrelated to the role of the ribosome in translation. In the present paper, we focus on examples of RPs with additional functions in the nucleus; however, there are also reports of RPs which bind mRNAs and affect translation. Classical demonstrations of extraribosomal functions in translation are RPs that bind specific RNA structures in the 5′-UTR (untranslated region) and suppress their own translation. This mechanism is widespread in E. coli, but there are also examples in eukaryotes. These repressor functions create negative-regulatory loops that synchronize synthesis of RPs with that of rRNA, preventing accumulation of free RPs in the cell . In prokaryotes, the RNA structure in the 5′-UTR typically resembles that of the rRNA domain that the RP binds on the ribosome [37,38]. Another well-documented case of an RP functioning as a translation repressor is RpL13a; in human cells during inflammation, RpL13a is phosphorylated at a specific residue and released from the ribosome, and as a free protein is incorporated into a complex that binds the mRNA 3′-UTR of inflammatory genes and represses translation [39,40].
Are ribosomal subunits present at chromosomal sites?
As described above, the current understanding is that free proteins not assembled into ribosomal subunits mediate extraribosomal functions of RPs in the nucleus. RPs are often found together in biochemical preparations of complexes involved in transcription and pre-mRNA processing [41–43]; however, the presence of RPs in these complexes is typically dismissed as a contamination with proteins that associate after cell lysis . The observation that the full complement of 40S or 60S RPs is never found in these complexes also suggests that the RPs are contaminants. The caveat of this argument is that RPs or whole complexes might have been lost during the purification. For instance, in the study referred to above , RpL11, but not other RPs, associates with c-Myc. While finding only RpL11 is a strong argument for the interaction with c-Myc being specific, it is still possible that RpL11 was recruited to c-Myc as a complex rather than as an individual protein (Figure 1, lower right panel). Most of the RpL11 in the cell is probably bound to 5S rRNA together with RpL5; most of the 5S rRNA is probably associated with the 60 S subunits, but a substantial fraction is not and might have a non-ribosomal function . Several factors may have released RpL11 from the 5S rRNA during the biochemical purification procedure. For example, the solutions used by Dai et al.  contained EDTA, which is known to disassociate the ribosome and release RPs from the subunits; in particular, addition of EDTA readily releases the 5S rRNA and associated proteins from the 60S subunits [44,45]. Using a similar argument, we could speculate that RpS13 is recruited to the splice sites as part of the 40 S subunit, but the rest of the subunit and associated RPs are lost during the purification (Figure 1, upper right panel) .
It has also been reported that at least 20 RPs and rRNA are present at many transcription sites in Drosophila melanogaster polytene chromosomes; the kinetics of the recruitment to the transcription site and RNA sensitivity suggested that the RPs are associated with nascent Pol II (RNA polymerase II) transcripts ; Figure 2 shows polytene chromosomes immunostained for RpS30. This study demonstrates that RPs are present at most transcription sites; in addition, the concurrent presence of many RPs and rRNA seems to be a good indication that these proteins are probably recruited as components of ribosomal subunits. Yet it has been speculated that some of the antibodies used in the polytene study are not sufficiently specific and might cross-react with common components of transcription sites . Although some of the antibodies recognized multiple bands on Western blot of whole-protein extracts, our view is that it is unlikely that the 20 antibodies could produce a similar polytene banding pattern by non-specific binding. In a later study, it was reported that RpS7b, RpL7b, RpL26a and RpL34b associate with genes also in S. cerevisiae . Although this finding confirms that RPs are present at transcription sites also in budding yeast, Schroder and Moore  found that the RPs associate both to coding and non-coding genes and it was concluded that the association is probably with free proteins rather than ribosomal subunits.
Many studies clearly demonstrate that RPs are present at transcription sites and have important functions in transcription and RNA processing. The association of RPs with transcription sites is probably a general cellular feature; an important issue is whether these proteins are recruited to transcription sites as individual RPs or ribosomal subunits, as indicated by the polytene study . The consensus is that the extraribosomal functions are always mediated by free RPs; however, future studies might change this view. For example, in E. coli, the RP S10 is a classic example of a protein moonlighting. During lytic growth of bacteriophage λ on E. coli, the λ N gene product modifies host RNA polymerase molecules that are transcribing certain λ genes so they cannot be prematurely terminated by the host. To do this, N protein forms a complex with several host proteins, known as Nus (N-utilization substance) factors, and one of these is the free S10 RP (also known as NusE) [17,49]. Another of these Nus factors is NusG, a well-characterized bacterial transcription elongation factor, which forms a bridge between NusE and transcribing RNA polymerase . Interestingly, in non-infected E. coli, NusG associates with most elongating RNA polymerase molecules  and appears to contact the NusE (S10) protein in the small subunit of the first ribosome that loads on the nascent transcript . This interaction thus couples bacterial transcription to translation, and this is critical for minimizing transcriptional polarity and counteracting RNA polymerase pausing . In summary, similarly to S10 in E. coli, it might be that in eukaryotes, individual RPs can affect transcription on and off the ribosome, depending on the specific gene. Ribosomal subunits are clearly present in the nucleus; they are assembled in the nucleolus during ribosome biogenesis . Although these ribosomal subunits might not be able to join together to form translation competent 80S ribosomes before export to the cytoplasm, they can still have a nuclear function; this perhaps is independent of translation and it could only be that of tethering RPs which have non-ribosomal roles in transcription or pre-mRNA processing of specific genes. Future studies should investigate these important open issues.
S.D. is supported by a Darwin Trust scholarship and S.B. is supported by a Royal Society University Research Fellowship.
We thank Steve Busby for critically reading the paper.
Post-Transcriptional Control: mRNA Translation, Localization and Turnover: A Biochemical Society Focused Meeting held at University of Edinburgh, U.K., 8–10 June 2010. Organized and Edited by Matthew Brook (Edinburgh, U.K.), Mark Coldwell (Southampton, U.K.), Simon Morley (Sussex, U.K.) and Nicola Gray (Edinburgh, U.K.).
Abbreviations: DFC, dense fibrillar component; Nus, N-utilization substance; RP, ribosomal protein; UTR, untranslated region
- © The Authors Journal compilation © 2010 Biochemical Society