mTORC1 (mammalian target of rapamycin complex 1) is controlled by diverse signals (e.g. hormones, growth factors, nutrients and cellular energy status) and regulates a range of processes including anabolic metabolism, cell growth and cell division. We have studied the impact of inhibiting mTOR on protein synthesis in human cells. Partial inhibition of mTORC1 by rapamycin has only a limited impact on protein synthesis, but inhibiting mTOR kinase activity causes much greater inhibition of protein synthesis. Using a pulsed stable-isotope-labelling technique, we show that the rapamycin and mTOR (mammalian target of rapamycin) kinase inhibitors have differential effects on the synthesis of specific proteins. In particular, the synthesis of proteins encoded by mRNAs that have a 5′-terminal pyrimidine tract is strongly inhibited by mTOR kinase inhibitors. Many of these mRNAs encode ribosomal proteins. mTORC1 also promotes the synthesis of rRNA, although the mechanisms involved remain to be clarified. We found that mTORC1 also regulates the processing of the precursors of rRNA. mTORC1 thus co-ordinates several steps in ribosome biogenesis.
mTOR (mammalian target of rapamycin) is a serine/threonine protein kinase. mTOR interacts with other proteins to form two distinct types of complex termed mTORC (mTOR complex) 1 and 2. In the short term, rapamycin, a macrolide antibiotic, inhibits some, but not all, of the functions of mTORC1. In some cell types, over longer treatment times, rapamycin can also interfere with mTORC2 .
mTORC1 and mTORC2 have distinct substrates, and this specificity is likely to be conferred by the partner proteins raptor (regulatory associated protein of mTOR) (mTORC1) and rictor (rapamycin-insensitive companion of mTOR) (mTORC2). For example, raptor interacts with substrates for mTORC1 through their TOS (target of rapamycin signalling) motifs. These substrates include S6K (ribosomal S6 kinase) 1 and 2, and the 4E-BPs [eIF (eukaryotic initiation factor) 4E-binding proteins]. Substrates for mTORC2 include PKB (protein kinase B) (also called Akt) and a related kinase, SGK (serum- and glucocorticoid-inducible kinase).
S6Ks are implicated in the control of cell and organismal growth [2–4]. Control of protein synthesis makes an important contribution to cell growth and division. This is for several reasons: (i) the bulk of the dry mass of cells is protein, so increased cell growth or cell division generally requires faster overall protein synthesis, and (ii) cell division, for example, requires specific proteins involved in checkpoints during the cell cycle, and their production is key to cell-cycle progression.
Many mitogenic, anabolic or hypertrophic stimuli activate protein synthesis. Initially (within minutes or a few hours), this involves the activation of existing components of the translational machinery, such as translation factors, and mainly the enhanced translation of existing mRNAs (i.e. improved ‘translational efficiency’). Over longer times (hours to days), additional mRNAs will be transcribed and the cell will generate more ribosomes, through activated ribosome biogenesis, as well as translation factors. Thus, over these longer times, there is an increase in the translational capacity of the cell.
The S6Ks phosphorylate several substrates which are linked to protein synthesis (mRNA translation) including ribosomal protein S6, a component of the 40S ribosomal subunit, eEF2K (eukaryotic elongation factor 2 kinase) , eIF4B (which promotes the activity of an RNA helicase, eIF4A)  and SKAR (S6K1 Aly/REF-like target)  (which is related to proteins involved in pre-mRNA splicing and mRNA transport). However, it remains unclear what contribution the phosphorylation of these S6K substrates makes to the control of mRNA translation.
The 4E-BPs interact with eIF4E, the protein that binds the 5′-cap structure of cytoplasmic mRNAs, and prevent it from engaging with the scaffold protein eIF4G to form initiation factor complexes. mTORC1 phosphorylates 4E-BP1 at several sites, resulting in its release from eIF4E, allowing it to bind eIF4G. eIF4G also binds eIF4A, and formation of eIF4E–eIF4G–eIF4A complexes is thought to be especially important for efficient translation of mRNAs that contain secondary structure within their 5′-UTRs (untranslated regions) (Figure 1). Recent findings show that 4E-BPs link mTORC1 to the control of cell proliferation, whereas S6Ks control cell growth .
Control of translational efficiency and capacity by mTORC1
mTORC1 also regulates ribosome biogenesis; it does so in at least two main ways. First, it promotes the translation of the mRNAs encoding all of the cytoplasmic ribosomal proteins (and some other components of the translational machinery, such as elongation factors ). The translation of these mRNAs is regulated through a 5′-TOP (terminal tract of pyrimidines), which suppresses their translation under basal (e.g. serum-starved) conditions; their translation is increased, e.g., by serum (Figure 1). This effect is inhibited by rapamycin and more strongly by inhibitors of PI3K (phosphoinositide 3-kinase) such as LY294002 . However, although the ability of rapamycin to impair 5′-TOP mRNA translation has been known for more than 15 years , it is not clear how mTORC1 controls their translation. At one time, this was thought to be mediated through the S6Ks and perhaps phosphorylation of S6 itself, but subsequent studies have disproved this [11,12].
Secondly, mTORC1 promotes the transcription of rRNAs, which is catalysed by Pol III (5S rRNA) or Pol I (the other three rRNAs) (reviewed in ). Although connections between mTORC1 and the transcriptional machinery that makes rRNA have been described (e.g. [14–16]), it is still not fully understood how mTORC1 controls rRNA synthesis (Figure 1). It is interesting to note that recent work has shown that mTORC2 associates with ribosomes [17,18], and appears to be controlled by this interaction, suggesting that ribosome number, for example, can control the activation of PKB signalling.
Recently, a number of compounds that inhibit the kinase activity of mTOR have been developed, including PP242  and AZD8055 . These compounds, called mTOR-KIs (mTOR kinase inhibitors) inhibit mTORC1 and mTORC2. Their effects are thus expected to be distinct from those of rapamycin, which only interferes with some functions of mTORC1 [19,21]. We have examined the effects of mTOR-KIs and rapamycin on protein synthesis and the production of rRNA.
mTOR-KIs inhibit overall protein synthesis more strongly than rapamycin
We studied the effects of rapamycin and mTOR-KIs in HeLa cells, using the standard approach of metabolic labelling of newly synthesized proteins with [35S]methionine. In serum-fed cells, treatment with rapamycin for 2 or 6 h had only a small effect on the rate of protein synthesis, inhibiting it by 10% or less. In contrast, mTOR-KIs exerted a much stronger inhibition, typically causing a 30–40% reduction in the rate of incorporation.
This difference could reflect either a role for mTORC2 in the short-term control of translation or effects of mTOR-KIs on components downstream of mTORC1 which are not affected by rapamycin. The effects of rapamycin and PP242 were similar in cells lacking Sin1, and thus devoid of mTORC2, and wild-type cells . This implies that the effects of mTOR-KIs on protein synthesis are mediated not through mTORC2, but through mTORC1.
We therefore studied the effects of rapamycin and mTOR-KIs on proteins known to be regulated through mTORC1. Rapamycin completely inhibited S6 phosphorylation, implying (i) that phosphorylation of S6 (and of other substrates for the S6Ks) plays only a minor role, at most, in maintaining protein synthesis rates, and (ii) that the phosphorylation of these proteins cannot explain the differing effects of rapamycin and mTOR-KIs on translation rates.
In contrast, rapamycin and mTOR-KIs had quite differing effects on the association of 4E-BP1 and eIF4G with eIF4E. Whereas the latter caused the dephosphorylation of 4E-BP1, resulting in its increased binding to eIF4E and loss of eIF4E–eIF4G complexes, rapamycin did not. Thus this branch of mTORC1 signalling is ‘resistant’ to rapamycin in HeLa cells. In principle, the ability of mTOR-KIs, but not rapamycin, to block eIF4E–eIF4G binding could account for their differing effects on overall protein synthesis. Further work is needed to understand how mTOR-KIs inhibit ongoing protein synthesis and also the mechanism by which rapamycin exerts its modest inhibitory effect.
Effects of mTOR-KIs and rapamycin on the synthesis of specific proteins
To explore the effects of rapamycin and mTOR-KIs on the synthesis of specific proteins, we made use of a newly developed heavy-isotope-labelling procedure which we call pSILAC (pulsed stable-isotope labelling with amino acids in cell culture). At the start of the study period, cells are transferred to medium containing heavy-isotope (13C or 15N)-tagged versions of arginine and lysine. In our experiments, where appropriate, mTOR inhibitors were added 30 min before this.
Newly synthesized proteins will contain ‘tagged’ versions of lysine and arginine. After cell lysis and tryptic digestion of proteins, peptides are analysed by MS, where the ‘light’ and ‘heavy’ versions of each peptide can be resolved, identified and quantified. The proportion of any given peptide present as the heavier form indicates the rate of accumulation of the parent protein from which it is derived. Analysis of the data reveals the impact of signalling inhibitors on the synthesis of specific polypeptides.
Our results show that rapamycin and mTOR-KIs have different effects on the synthesis of many proteins. In general, mTOR-KIs inhibit their synthesis more strongly than does rapamycin, in line with the effects of these compounds on overall protein synthesis rates.
For proteins known to be encoded by 5′-TOP mRNAs, such as ribosomal proteins, rapamycin inhibited synthesis quite strongly (typically by 20–50%), which is roughly in line with results from earlier studies (which mostly focused on the association of such mRNAs with polyribosomes rather than actual rates of synthesis of the corresponding proteins). In contrast, mTOR-KIs exerted much stronger inhibition, with inhibition of more than 50% being seen for most of these proteins. Importantly, this suggests that mTOR, probably mTORC1, but perhaps via another type of mTORC , controls the synthesis of these proteins by rapamycin-sensitive and -insensitive mechanisms. A key issue is to identify how mTORC1 modulates the translation of 5′-TOP mRNAs. The rapamycin-insensitive input to mTOR regulation of 5′-TOP mRNAs may contribute to the inhibitory effects of the PI3K inhibitors mentioned above, since LY294002 also inhibits mTOR activity .
Rapamycin only rather weakly inhibited the synthesis of most other proteins, with the mTOR-KIs slowing their production by between 20 and 50%. Thus mTOR inhibition has less impact on the synthesis of most proteins that are not encoded by non-5′-TOP mRNAs. Interestingly, within the set of proteins initially categorized as encoded by non-5′-TOP mRNAs are a number whose synthesis is strongly affected by mTOR-KIs or rapamycin and which appear (as judged from the database of transcriptional start sites at http://dbtss.hgc.jp/) to contain 5′-terminal sequences resembling a TOP. These may be previously unrecognized 5′-TOP mRNAs. An earlier study pointed to the existence of many human mRNAs that contain a 5′-TOP, although, for most of them, this feature was not ‘conserved’ in mice . Nevertheless, taken together, these data point to the possibility that signalling through mTORC1 may, via 5′-TOP-mediated translational control, promote the synthesis of a larger number and diversity of proteins than has hitherto been recognized. This mechanism may therefore mediate the regulation by mTORC1 of other processes in addition to ribosome biogenesis and translational capacity.
Control of ribosome biogenesis by mTOR
As noted above, mTORC1 is known to control the transcription of rRNA as well as the synthesis of ribosomal proteins . Rapamycin impairs the formation of transcription-initiation complexes involved in transcribing the 45S pre-rRNA genes . The major rRNAs (5.8S, 18S and 28S) are derived from a single primary transcript, the 45S pre-rRNA, which is made by Pol (RNA polymerase) I. Rapamycin causes the inactivation of TIF-1A, a component of the Pol I transcriptional machinery, via changes in its phosphorylation. Rapamycin causes the dephosphorylation of one site in TIF-1A and increased phosphorylation of another one. It also leads to accumulation of TIF-1A in the cytoplasm. It is important to note that mTORC1 has been shown to associate with the promoters for genes transcribed by Pol I and Pol III [25,26]. mTORC1 (and its yeast counterpart) are thought to phosphorylate components involved in Pol III function directly, but it is not clear how mTORC1 controls Pol I.
Earlier work indicated that S6K1 played a role in promoting the activity of Pol I . The mechanism by which rapamycin impairs rRNA transcription appears to involve dephosphorylation of the C-terminal tail of UBF (upstream binding factor), which binds directly to rDNA and interacts with other components of the RNA Pol I machinery, in particular SL-1 .
In the case of Pol III, which makes 5S rRNA, regulation by TORC1 in yeast involves the phosphorylation of Maf1, a repressor of Pol III . Phosphorylated Maf1 is excluded from the nucleus, thus alleviating the inhibition of Pol I transcription. TORC1 appears to promote the phosphorylation of Maf1 both directly (i.e. it can phosphorylate Maf1)  and through the protein kinase Sch9, a homologue of the mammalian S6Ks [29–31]. The latter apparently mediates the nuclear exclusion of Maf1, whereas TORC1 regulates the association of Maf1 with chromatin .
Mammalian Maf1 undergoes dephosphorylation in response to inhibition of mTORC1 [26,32] and has been reported to be a direct substrate for mTORC1 which can phosphorylate it at several sites . Knockdown or knockout of Maf1 rendered Pol III-mediated tRNA transcription insensitive to inhibition of mTORC1 [26,32], confirming its importance for the control of Pol II by mTOR. In the case of mammalian Maf1, which does not contain an obvious nuclear localization signal, mTORC1 mainly acts by ensuring the exclusion of Maf1 from chromatin rather than from the nucleus, which contrasts with the situation in yeast. Another difference between yeast and mammalian Maf1 is that the Sch9 phosphorylation sites in the yeast protein are absent from the shorter mammalian orthologue. However, a rapamycin-resistant mutant of S6K1 can partially reverse the dephosphorylation of Maf1 caused by mTOR inhibition .
Kantidakis et al.  have confirmed that mTOR associates with Pol III genes and have shown that it binds to TFIIIC (transcription factor IIIC), which binds to Pol III promoters. This mechanism for localization is likely to be mediated through a TOS motif in the TFIIIC component TFIIIC63, thereby allowing mTORC1 to phosphorylate Maf1 on chromatin.
The 45S pre-rRNA must be processed, through multiple steps, to yield the mature 5.8S, 18S and 28S rRNAs. Our own recent results suggest that mTORC1 also regulates this process. Treatment of cells with rapamycin causes the decay of newly made rRNA and the accumulation of intermediates in the rRNA-processing pathway, indicating that it blocks specific steps in that process.
Although important advances continue to be made in understanding how mTOR controls cell function, there remain important gaps in our understanding. The present article has highlighted the fact that we do not understand how it regulates steps involved in ribosome biogenesis, such as the translation of 5′-TOP mRNAs, the activity of Pol I or the processing of rRNA. It is also still unclear how mTOR-KIs impair general protein synthesis in HeLa cells, given that this seems not to be mediated through inactivation of eIF4E by hypophosphorylated 4E-BPs.
Signalling 2011: a Biochemical Society Centenary Celebration: A Biochemical Society Focused Meeting held at the University of Edinburgh, U.K., 8–10 June 2011. Organized and Edited by Nicholas Brindle (Leicester, U.K.), Simon Cook (The Babraham Institute, U.K.), Jeff McIlhinney (Oxford, U.K.), Simon Morley (University of Sussex, U.K.), Sandip Patel (University College London, U.K.), Susan Pyne (University of Strathclyde, U.K.), Colin Taylor (Cambridge, U.K.), Alan Wallace (AstraZeneca, U.K.) and Stephen Yarwood (Glasgow, U.K.).
eukaryotic initiation factor
mammalian target of rapamycin
mTOR kinase inhibitor
protein kinase B
regulatory associated protein of mTOR
rapamycin-insensitive companion of mTOR
ribosomal S6 kinase
transcription factor IIIC
terminal tract of pyrimidines
target of rapamycin signalling
This work was supported by grants to C.G.P. from AstraZeneca, the European Union and the British Heart Foundation.