Upon cell-cycle arrest or nutrient deprivation, the cellular rate of ribosome production is reduced significantly. In mammalian cells, this effect is achieved in part through a co-ordinated inhibition of RP (ribosomal protein) synthesis. More specifically, translation initiation on RP mRNAs is inhibited. Translational regulation of RP synthesis is dependent on cis-elements within the 5′-UTRs (5′-untranslated regions) of the RP mRNAs. In particular, a highly conserved 5′-TOP (5′-terminal oligopyrimidine tract) appears to play a key role in the regulation of RP mRNA translation. This article explores recent developments in our understanding of the mechanism of TOP mRNA regulation, focusing on upstream signalling pathways and trans-acting factors, and highlighting some interesting observations which have come to light following the recent development of cDNA microarray technology coupled with polysome analysis.
What is a TOP (terminal oligopyrimidine tract) message? The vast majority of eukaryotic mRNAs have a cap structure at their 5′ ends, normally followed by an A residue (Figure 1) . However, transcription of TOP mRNAs is initiated with a C residue, which is followed by 4–14 uninterrupted pyrimidine residues (reviewed in ). It appears that the 5′-TOP motif is predominantly found in mRNAs that encode proteins that are involved in ribosome biogenesis, most notably the RPs (ribosomal proteins) . Crucially, the synthesis of these proteins is exquisitely sensitive to the growth rate of the cell . Growth arrest results in the inhibition of TOP mRNA translation. Experimentally, this effect can be observed as the shift of TOP mRNAs from the polysomes (the translating population) in growing cells into the sub-polysomes (the non-translating population) in growth-arrested cells. It is noteworthy that, when TOP mRNAs are associated with polysomes, they are loaded with a full complement of ribosomes. Thus TOP mRNAs appear to exist in two states: the repressed state and the active state. In growing cells, the proportion of TOP mRNAs present in the active state increases, and these mRNAs are translated at maximum efficiency. This ‘all-or-none’ binary control mechanism suggests that, in the repressed state, translation initiation on TOP mRNAs is blocked (reviewed in ). Importantly, this growth-dependent mechanism of translational regulation is absolutely dependent on the integrity of the 5′-TOP and on the position of the TOP immediately next to the cap structure . Furthermore, non-TOP mRNAs do not display this growth-regulated bimodal distribution between the polysomes and sub-polysomes. Therefore a 5′-TOP confers growth and nutrient-dependent expression on TOP mRNAs and thereby provides a mechanism for the co-ordinated expression of proteins that are required for ribosome biogenesis.
Structural characteristics of the cap structure
Signalling pathways upstream of TOP mRNA translation
Owing to the large energy costs that are involved in ribosome biogenesis, the synthesis of ribosomal RNAs and proteins is highly regulated [4,5]. Clearly, the cell must respond rapidly to extracellular and intracellular growth cues, and these signals must be relayed in order to regulate ribosome production accordingly. In mammalian cells, the rate of TOP mRNA translation is controlled by growth signals, and much attention has been focused on determining which signal transduction pathways are implicated in this process (reviewed in [6,7–8]) (Figure 2).
PI3K and mTOR signalling pathway and the known interactions with translation machinery
Initially, it was noted that there is a correlation between the phosphorylation of RPS6 and translational activation of TOP mRNAs . In response to mitogens, RPS6 is phosphorylated on multiple sites by the closely related kinases S6K (S6 kinase) 1 and 2 (reviewed in ). Blocking S6K activity using the mTOR (mammalian target of rapamycin) inhibitor rapamycin repressed the translation of TOP mRNAs in a number of studies [10,11,12–13]. In addition, a dominant-negative S6K1 mutant partially inhibited the recruitment of TOP mRNAs to the polysomes . This evidence led to the proposal of a model in which phosphorylation of RPS6 results in the selective translation of TOP mRNAs . However, various findings have challenged this model. First, during MEL (mouse erythroleukaemia) cell differentiation, TOP mRNAs become translationally repressed, but RPS6 appears to be constitutively dephosphorylated in these cells . Furthermore, in several cell lines, rapamycin causes complete inhibition of S6K activity, but, in the same cell lines, this inhibitor has only moderate effects on TOP mRNA translation [15,16]. More convincingly, the targeted disruption of S6K alleles does not affect TOP mRNA translational regulation . In one study, an S6K1 nullizygous cell line was shown to have constitutively dephosphorylated RPS6, and yet these cells display normal regulation of TOP mRNA translation . More recently, it has been reported that TOP mRNA translation remains responsive to mitogens in embryonic stem cells with a disruption in both alleles of S6K1 and S6K2 . The combined knockout of S6K1 and S6K2 revealed the presence of another S6K . However, while the mitogen-dependent translational regulation of TOP mRNAs is inhibited by rapamycin, this residual S6K is rapamycin-insensitive. Therefore S6K activity does not correlate with the regulation of TOP mRNA translation in the double-knockout cell line . Together, these data indicate that TOP translational regulation does not depend on S6K or S6 phosphorylation.
As outlined above, a number of studies have implicated the mTOR signalling pathway in the translational activation of TOP mRNAs (reviewed in ). Signalling through the mTOR pathway is controlled by growth factors and nutrients, and impinges on protein synthesis through the phosphorylation of a number of proteins that are involved in translation initiation, including the eIF4E (eukaryotic initiation factor 4E) inhibitor protein, 4E-BP1 (eIF4E-binding protein 1), eIF4B and eIF4G (reviewed in [19,20]) (Figure 2). Furthermore, genetic evidence indicates that mTOR is important in the regulation of cell size [21,22–23]. Therefore mTOR is ideally positioned to transduce growth signals to ribosome biogenesis. However, it is possible that the contribution of this pathway towards the control of TOP mRNA translation varies depending on the cellular context, since some studies suggest a major role for mTOR in TOP mRNA translational regulation, whereas, in other studies, the inhibition of mTOR has little or no effect on TOP mRNA translation (reviewed in ).
Evidence has also emerged that another growth-factor-regulated pathway, the PI3K (phosphoinositide 3-kinase) pathway, is involved in controlling TOP mRNA translation. PI3K is activated in response to numerous growth stimuli and results in elevated levels of the lipids PtdIns(3,4,5)P3 and PtdIns(3,4)P2. As a consequence, the downstream effector kinase PKB (protein kinase B) is activated via PDK1 (phosphoinositide-dependent kinase 1) (reviewed in ). Several lines of evidence indicate that TOP mRNA translation is regulated by the PI3K pathway. First, specific inhibitors of PI3K completely block the growth-factor-dependent translational activation of TOP mRNAs. In addition, overexpression of proteins that interfere with signalling through the PI3K pathway inhibit the recruitment of TOP mRNAs to the polysomes on serum stimulation. Dominant-negative inhibitors of PI3K and PDK1, PTEN (phosphatase and tensin homologue deleted on chromosome 10), a cellular inhibitor of the PI3K pathway, and a kinase-inactive mutant of PKBα all selectively block growth-factor-dependent TOP mRNA translation. Furthermore, the translational suppression of TOP mRNAs in growth-inhibited cells can be relieved by constitutively active forms of PI3K or PKB. Together, these data strongly suggest that the effect of growth factors on TOP mRNA translation is mediated through the PI3K pathway. It must be noted that there is now a significant body of evidence which indicates that the PI3K pathway can directly activate mTOR (reviewed in ). However, at least one study has shown that the translational activation of TOP mRNAs is fully dependent on PI3K, but is only moderately affected by the inhibition of mTOR . Thus it seems that growth factors can signal to TOP mRNA translation through the PI3K/mTOR pathways, but the precise role of each pathway in this process has not been defined at present.
Trans-acting factors and TOP mRNA translational regulation
It has been suggested that the 5′-TOP motif is recognized by specific trans-acting factors, and it is these factors that regulate TOP mRNA translation initiation (reviewed in ). Indeed, the selective repression of a TOP mRNA in translation extracts can be relieved by a synthetic TOP RNA oligonucleotide, arguing for the existence of a titratable repressor protein . Although proteins have been shown to interact with 5′-TOP sequences, the functional significance of these interactions remains unclear . However, several lines of evidence suggest that the La autoantigen may play a role in TOP mRNA translational regulation. In vitro binding studies demonstrated that La can interact with 5′-TOPs from humans and Xenopus laevis mRNAs [26,27]. Subsequently, Xenopus La was shown to co-sediment with polysomes in an RNA-dependent manner . Moreover, La is associated specifically with TOP mRNAs in the polysome fraction . Evidence for the functional significance of this interaction was provided by overexpressing variants of La. In growth-arrested cells, overexpression of wild-type La shifted RPL4 mRNA from the sub-polysomes to the polysomes, whereas, in growing cells, expression of truncated forms of La resulted in a greater proportion of RPL4 mRNA in the sub-polysomal fraction . Taken together, these data suggest a model in which the interaction between La and TOP mRNAs specifically stimulates the translation of these mRNAs. Nevertheless, conflicting evidence exists which challenges this model. Although La is predominantly a nuclear protein, recently, the subcellular localization of this protein has been examined in more detail. La phosphorylated on Ser366 is distributed throughout the nucleoplasm, whereas non-phosphorylated La (npLa) is found in both the nucleolus and the cytoplasm . It is perhaps no surprise that TOP mRNAs associate preferentially with npLa . More interestingly, however, the overexpression of a S366A mutant La protein increases the amount of RPL37 that is associated with La and also increases the abundance of sub-polyribosomal RPL37 . Hence, in contrast with the data obtained in Xenopus cells, these data suggest that interaction of La with a TOP mRNA can inhibit TOP mRNA translation. In support of this hypothesis, it has been shown that La can repress the translation of a TOP mRNA in vitro . Evidently, La can interact with TOP mRNAs and is somehow implicated in the regulation of TOP mRNA translation, but its precise role in this process remains to be determined.
TOP mRNA and polysome profiling
Currently, the number of mRNAs that are regulated in a TOP-like manner is not known. In fact, transcription initiation occurs at a C residue in 17% of mRNAs within a eukaryotic cell . With the development of cDNA microarray, the search for mRNAs that are subject to translational regulation has accelerated greatly. Interestingly, the application of this technology has led to the identification of numerous mRNAs whose translation appears to be regulated in a TOP-like fashion. However, further investigation is required to clarify whether these mRNAs possess a TOP motif at their 5′-terminus.
Polysome profiling is a technique in which the sub-polysome/polysome distribution of a large number of mRNAs is analysed using cDNA microarray to identify mRNAs that are subject to differential translational regulation under certain cellular conditions ([33,34–35], and M. Bushell, M. Stoneley, Y.-W. Kong, T.L. Hamilton, P. Sarnow and A.E. Willis, unpublished work). In one such study, polysome profiling was performed on T-cells treated with the mTOR inhibitor rapamycin . Not surprisingly, among the mRNAs that are translationally repressed by rapamycin are the RP mRNAs and the eEF (eukaryotic translation elongation factor) mRNAs, which are known to contain 5′-TOP sequences. In addition, numerous other mRNAs are translationally suppressed by the inhibition of mTOR, most notably a group of mRNAs that encode subunits of the proteasome. It remains to be determined how many of the non-RP mRNAs identified in this screen are bona fide TOP mRNAs . Using a similar approach, we have analysed mRNAs that are subject to differential translational regulation during apoptosis, when cap-dependent protein synthesis is compromised (M. Bushell, M. Stoneley, Y.-W. Kong, T.L. Hamilton, P. Sarnow and A.E. Willis, unpublished work). Many of the mRNAs that are severely translationally repressed during cell death were also identified in the rapamycin study, implying that TOP or TOP-like translational control is important under different physiological conditions and lending support to the notion that TOP-like regulation may not be restricted to mRNAs that are involved in ribosome biogenesis.
Comparison of the data from the different polysome profiling studies suggests that some TOP mRNAs can be subject to more than one level of translational regulation; perhaps the most striking example is that of the nucleophosmin mRNA. First, the nucleophosmin mRNA has been shown previously to possess the hallmarks of a TOP mRNA, in that there is there is a C residue and a continuous stretch of eight pyrimidine residues adjacent to the cap structure . Moreover, the mRNA is translationally repressed in serum-starved cells . These observations are supported by data from the polysome profiling of rapamycin-treated T-cells . As one would expect of a TOP mRNA, the inhibition of mTOR results in the translational suppression of nucleophosmin mRNA. However, in another study, nucleophosmin mRNA was found to be among a group of mRNAs that are preferentially associated with heavy polysomes in mitotic cells . Global protein synthesis is inhibited during mitosis, and this event is accompanied by a shift of the majority of cellular mRNAs to polysomes containing fewer ribosomes . It appears that the nucleophosmin mRNA can resist the general inhibition of protein synthesis because the 5′-UTR (5′-untranslated region) of this mRNA contains a cis-element that can recruit ribosomes to a site downstream of the cap structure. This complex structural element is known as an IRES (internal ribosome entry segment). Hence, although cap-dependent protein synthesis is compromised during mitosis, the synthesis of nucleophosmin is maintained by internal initiation . In contrast, during apoptosis, when it has been shown that cellular IRESs function to maintain the translation of certain mRNAs, we have noted that the nucleophosmin mRNA is severely translationally suppressed (M. Bushell, M. Stoneley, Y.-W. Kong, T.L. Hamilton, P. Sarnow and A.E. Willis, unpublished work). Hence our data suggest that the nucleophosmin mRNA is regulated in a TOP-like manner, rather than via internal initiation during cell death. Together, these studies suggest that not only is nucleophosmin mRNA regulated by a 5′-TOP, but also, depending on the cellular conditions, its translation can be controlled by internal initiation. It is worth noting that translation initiation via internal ribosome entry is not a general property of TOP mRNAs, since the 5′-UTR of RPS5 mRNA was found to be unable to promote internal initiation . Nevertheless, translation initiation on a number of RP mRNAs was found to be resistant to the inhibition of capdependent protein synthesis during poliovirus infection . Thus RP mRNAs are translated in a manner that is less dependent on the cap structure. It will be important to address whether this cap-independence is due to the 5′-TOP or whether it is due to some other common feature of RP mRNAs.
The regulation of TOP mRNAs has been studied for a number of years, and, consequently, an extensive body of literature has accumulated, focusing in particular on the signalling pathways that lie upstream of TOP mRNA translational control. However, little progress has been made towards understanding precisely how the 5′-TOP regulates translation initiation. Recently, it has become clear that, in addition to controlling ribosome biogenesis, TOP-like translational regulation may be employed by a plethora of cellular mRNAs under different cellular conditions. Given that this mechanism of regulating gene expression may be considerably more widespread than was previously suspected, it is all the more important that we gain a deeper understanding of TOP translational regulation.
Translation UK: Focused Meeting and Satellite to BioScience2005, held at Western Infirmary, Glasgow, U.K., 21–23 July 2005. Organized and Edited by M. Bushell (Nottingham, U.K.), S. Newbury (Newcastle upon Tyne, U.K.), G. Pavitt (Manchester, U.K.) and A. Willis (Nottingham, U.K.).
eukaryotic initiation factor
internal ribosome entry segment
mammalian target of rapamycin
phosphoinositide-dependent kinase 1
protein kinase B
terminal oligopyrimidine tract
We thank Ian Powley and Helen Dobbyn for critical reading of the manuscript. This work was supported by the BBSRC (Biotechnology and Biological Sciences Research Council) (to M.B., a David Phillips Fellowship, and T.H.) and the Wellcome Trust (M.S. and K.S).
1These authors contributed equally to this work.