Over the past few years, there has been a growing interest in the interconnection between translation and metabolism. Important oncogenic pathways, like those elicited by c-Myc transcription factor and mTOR kinase, couple the activation of the translational machinery with glycolysis and fatty acid synthesis. Eukaryotic initiation factor 6 (eIF6) is a factor necessary for 60S ribosome maturation. eIF6 acts also as a cytoplasmic translation initiation factor, downstream of growth factor stimulation. eIF6 is up-regulated in several tumor types. Data on mice models have demonstrated that eIF6 cytoplasmic activity is rate-limiting for Myc-induced lymphomagenesis. In spite of this, eIF6 is neither transcriptionally regulated by Myc, nor post-transcriptionally regulated by mTOR. eIF6 stimulates a glycolytic and fatty acid synthesis program necessary for tumor growth. eIF6 increases the translation of transcription factors necessary for lipogenesis, such as CEBP/β, ATF4 and CEBP/δ. Insulin stimulation leads to an increase in translation and fat synthesis blunted by eIF6 deficiency. Paradoxycally, long-term inhibition of eIF6 activity increases insulin sensitivity, suggesting that the translational activation observed upon insulin and growth factors stimulation acts as a feed-forward mechanism regulating lipid synthesis. The data on the role that eIF6 plays in cancer and in insulin sensitivity make it a tempting pharmacological target for cancers and metabolic diseases. We speculate that eIF6 inhibition will be particularly effective especially when mTOR sensitivity to rapamycin is abrogated by RAS mutations.
It has long been known that many cancer cells undertake a metabolic glycolytic switch known as the Warburg effect . The Warburg effect causes the flow of glucose to be directed towards glycolysis thus generating all of the intermediates that fuel cell growth and cell cycle progression. In cancer cells, glycolysis is normally co-ordinated with nucleotide and fatty acid synthesis, the latter now appreciated as being essential for cancer cell malignancy . Glycolysis and fatty acid synthesis are almost invariably associated with cycling cells [3–6]. Given the strong association between glycolysis, fatty acid synthesis and cell cycle progression, it is not surprising that several oncogenic pathways regulate both metabolic fluxes. In addition, since cell cycle progression is preceded by an increase in cellular size, cycling cells require also a robust activation of the ribosomal and translational machineries.
c-Myc and mTOR orchestrate the translational machinery in both pathological and physiological conditions
Early observations unequivocally found increased ribosome synthesis in cancer cells, even in association with prognostic significance , but it was not noticed that changes in the translational machinery consistently associate with glycolysis. In the last years, thanks to high-throughput studies, it has become evident that major oncogenic pathways simultaneously regulate metabolism and translation. One of these pathways is represented by the action of the Myc family of cellular oncogenes. In tumor cells, the three members of the Myc family, L-Myc, N-Myc and in particular c-Myc, are frequently hyperactivated by amplification. C-Myc target genes include translational elongation factors, translational initiation factors, as well as nucleolar assembly components and ribosomal proteins belonging to the small and large subunits. Thus, Myc is an efficient positive regulator of protein synthesis capability . Indeed, Myc up-regulation is sufficient to cause nucleolar enlargement , which in itself was considered a sign of malignancy in cancer cells as early as the 1950s. Besides up-regulating the translational apparatus, Myc induces a glycolytic and fatty acid synthesis program through the pleiotropic induction of glucose transporters and transcriptional factors . Hence, the transcriptional activation driven by Myc co-ordinately regulates an increase in translation and in glycolysis. For years, the up-regulation of the translational machinery driven by Myc was interpreted as a by-product of transformation. However, restriction of Myc-induced up-regulation of protein synthesis, through haploinsufficiency of ribosomal protein L24, increases the overall survival rate of Myc-induced lymphomagenesis mice . Thus, the hijacking of the translational machinery by c-Myc is not only simply a by-product of transformation, but also a required and rate-limiting step for tumorigenesis.
Another strong set of observations that connects the translational machinery with the glycolytic and lipogenic pathways comes from studies on mTOR kinase activity. mTOR is the central sensor for nutrient and growth factor availability and assembles in two functionally distinct complexes, mTORC1 and mTORC2, fully described elsewhere . mTOR activation is able to elicit the Warburg effect by activating PKM2 and other glycolytic enzymes under normoxic conditions . Moreover, mTOR kinase activates a lipogenic program through a set of events that ultimately leads to expression of the sterol regulatory element-binding proteins . Importantly, perhaps the best characterized substrates of mTORC1 are S6K1 and the three protein isoforms of translational repressor 4E-BP, in mammals. Activated S6K1 phosphorylates ribosomal protein rpS6, while phosphorylation of 4E-BPs causes their release from the cap-binding protein eIF4E, an event which in turns activates cap-dependent translation . Biochemical evidence has convincingly shown that interruption of the inhibitory loop regulated by 4E-BPs can accelerate tumor progression and hamper sensitivity to mTOR inhibitors [16,17]. In conclusion, activation of a tyrosine kinase receptor signaling pathway converging on mTOR results in lipogenesis, glycolysis and an increase in translation.
The above-mentioned connections are not unique to cancer cells. It is not surprising that specialized cell populations, which respond rapidly and vigorously to extracellular signals, may have also the capability to employ the translational machinery for rapid metabolic adaptation. The regulation of the immune system depends on establishing efficient immunoresponse while preventing from excessive inflammation and/or autoimmunity. Such modulations range from increased production of cytokines and cytolytic molecules to the ability to undergo cell division and migration. For instance, upon activation, a naive T cell undergoes a radical metabolic shift, down-regulating the expression of fatty acid oxidation genes and shifting towards a glycolytic program. The important role of mTOR in the activation of CD4+ T cells is well known  However, the commitment of an immune cell to a specific metabolic pathway depends on its acquired function. For example, energy production relies on aerobic glycolysis in CD4+ effector T cells and Th17 cells, while it requires fatty acids oxidation in memory and regulatory T cells (Treg) . Interestingly, several recent articles highlighted a role for mTOR in coupling the immune response with metabolic demands (Figure 1). mTOR-dependent regulation of metabolic programs is mediated through the regulation of key transcriptional factors, such as Myc, which promotes the expression of enzymes involved in aerobic glycolysis, and HIF-1α, which regulates the expression of components of the glycolytic pathway . Given the extent of metabolic changes after T-cell activation, it is possible that specific mechanisms of translational regulation may be involved in the metabolic reprogramming of T cells.
Metabolic reprogramming upon activation and differentiation of different T-cell subsets.
eIF6 is a central regulator of metabolism independent from mTOR and Myc
Translation occurs in four phases: initiation, elongation, termination and recycling. Initiation of translation is controlled by eukaryotic initiation factors (eIFs) and is a rate-limiting event because mRNAs must successfully compete for ribosomes and for eIFs. eIFs activity can be regulated by signaling pathways . It is not the purpose of this review to connect the known connections of translation with metabolism that involve mTOR kinase , but we will focus on (eukaryotic initiation factor 6) eIF6. The existence of eIF6 is known from the late 1970s, when it was shown that eIF6 acts as an antiassociation factor that binds free 60S ribosomal subunits and impairs improductive ribosomal joining . Later, genetic studies in yeast cells unveiled a role for eIF6 in the maturation of 60S ribosomal subunits, but failed to demonstrate a role for eIF6 in translational control . To date the specific mechanism of eIF6 release from the 60S subunit, an event necessarily required for subsequent 80S formation and translation initiation remains open, and is discussed elsewhere [22–24]. Strikingly, in mammalian cells, it was evident that eIF6 is necessary for efficient translation downstream of the activation of insulin and growth factors . Early work suggested that the activation of eIF6 is independent of mTOR activity, but dependent on the RAS-PKC cascade . These studies, among others, suggested the general principle that translational activity can be insensitive to mTOR inhibition, if RAS is active. Clinical studies have confirmed this prediction as cancer cells with mutated RAS are totally insensitive to the inhibition of mTORC1 . These data suggest the importance of the RAS-PKC/eIF6 axis in tumorigenesis and that eIF6 could be a relevant target for the treatment of specific types of cancers.
eIF6 role in physiological conditions
In mice, the loss of one allele of eIF6 causes a somewhat spectacular phenotype. In brief, mice are leaner, presenting less adipose tissue , and are resistant to Myc-induced lymphomagenesis without overt signs of pathology . At the cellular level cells derived from mice haploinsufficient for eIF6 have normal translational rates in unstimulated conditions, but fail to elicit translation when challenged with insulin . For this reason a characterization of the mRNAs whose translation was affected by eIF6 levels, upon insulin stimulation, was carried out. The global translatome of cells with down-regulated eIF6 was revealed by microarray studies of the mRNAs that were associated with polysomes. The underlying assumption is that mRNAs that are strongly translated are enriched on translating ribosomes, i.e. polysomes. Obviously, this assumption cannot be generalized because also mRNAs stalled at elongation enrich on polysomes, in spite of being less translated. Coupling polysomal microarray studies with independent validation protocols like western blotting or reporter studies, helped therefore to unveil the genes translationally regulated by eIF6, upon insulin stimulation . eIF6 expression was found associated with the translation of mRNAs that contain regulatory sequences with high G/C-rich regions or upstream open reading frames (uORF). High G/C-rich regions in the 5′-UTR of mRNAs impair translation by generating high-energy secondary tracts that require the melting activity of helicases. uORFs diminish translation of the underlying ORF unless they are overcome by either leaky scanning of ribosomal subunits or by reinitiation . Interestingly, expression of eIF6 causes an increase in translation of transcription factors that contain uORFs in their 5′-UTRs and which positively regulate lipogenesis, like ATF4 , C/EBPδ and C/EPBβ . We wish to note here that in spite of being ATF4 generally described as a transcription factor associated with the stress response , ATF4 null organisms have a phenotype very similar to eIF6 heterozygotes and involving impaired lipid synthesis [31,34]. The reduction in the translational efficiency of transcription factors involved in lipid synthesis upon eIF6 depletion, caused a change in steady-state mRNA levels of (almost) all the enzymes involved in fatty acid synthesis and glycolysis. In parallel, also the relative protein levels were reduced. Accordingly, metabolic studies demonstrated that the changes in gene expression driven by eIF6 resulted in metabolic changes. In conclusion, eIF6 activation induces glycolysis and fatty acid synthesis in a cell-autonomous fashion (Figure 2). A tentative model is that upon insulin administration, eIF6, a master regulator of translation initiation, triggers a metabolic reprogramming by coupling translation to transcription and by overall regulating the steady-state levels of mRNAs involved in glycolysis and lipogenesis. Of note, eIF6 depletion was shown to increase the steady-state levels of TGF-β1 mRNA. In the present study, a direct role of eIF6 in transcription was proposed . Rigorous, future work is required in order to define direct and indirect effects driven by eIF6, due to the strong effects driven by its gene dosage.
eIF6 regulation of translation downstream of insulin administration.
eIF6 in disease
In target cells, insulin elicits both short- and long-term responses, including rapid glucose uptake and a transcriptional cascade that induces biosynthesis of glycogen, lipids and proteins. In the past, the increase in protein synthesis elicited by insulin was thought to be part of the wider cell growth program, i.e. a necessary but rather unselective translational burst required for hypertrophic growth. However, the fact that eIF6 activity downstream of insulin induces the preferential translation of transcription factors necessary for lipid synthesis and for a glycolytic switch demonstrates that insulin-regulated translation acts as a feed-forward stimulus to amplify insulin lipogenic action (Figure 2). Given the physiological role of eIF6 in the regulation of fatty acid synthesis, the next step was to test its involvement in the metabolic syndrome. The metabolic syndrome is the status by which increased glycemia is observed in spite of high insulin levels and is directly associated with high risk of diabetes and obesity.
Interestingly, eIF6 inhibition was found to have an anti-diabetic effect, as it reduced weight gain upon a high fat diet, and increased insulin sensitivity . Moreover, a full analysis of eIF6 effects on triglycerides synthesis unveiled that eIF6 depletion protects from hepatic steatosis and reduces circulating lipids; accordingly, in humans, methylation levels of eIF6 gene in fat tissues highly correlate with obesity .
Besides its involvement in the metabolic syndrome, the metabolic reprogramming induced by eIF6 has also strong implications in tumorigenesis. It is well known that cancer cells preferentially increase aerobic glycolysis to have a growth and proliferation advantage (Warburg effect). In addition, the fact that eIF6 directly regulates the mRNA translation of Fasn, a fundamental enzyme of de novo lipogenesis that is required for membrane building during oncogenesis, suggests that eIF6 could have a crucial role in neoplastic lipogenesis. Accordingly, eIF6 gene is amplified in some breast human cancers  and its overexpression is a marker of aggressive tumors, such as malignant pleural mesothelioma (MPM)  and colorectal cancer . eIF6 depletion impairs lactate and ATP production in MPM cells, leading to a disadvantage for tumoral cell growth . For all of these reasons, we can envisage that drugs which target eIF6 in the translational machinery could block the feed-forward mechanism triggered by insulin, control metabolic reprogramming and limit the glycolytic switch of tumor cells. A new assay may help to screen for eIF6 antagonists by exploiting the quantitative and functional binding of eIF6 to the 60S .
eIF6 capability to modulate tumor growth and metabolism is impressive. One exciting perspective is that different initiation factors act at different levels providing specific selectivity of translational control. For instance, we do not know if factors that increase start codon accuracy like eIF1 and eIF1A have an impact on uORF translation and could perhaps modulate metabolism in opposite ways compared with eIF6. Future and exciting studies describing the multiple ways by which translational control shapes gene expression in human cells will be certainly paved by the impressive array of new technologies that are available nowadays.
This work has been possible through grants from the Italian Association for Cancer Research (AIRC, IG 2014) and the European Research Council (ERC TRANSLATE).
The Authors declare that there are no competing interests associated with the manuscript.