tRNA trafficking along the TOR pathway

Adaptation of metabolic activity to nutrient availability is among the most ancestral of cellular behaviors. Resource-intensive processes such as growth and protein synthesis are particularly tightly coupled to nutrient levels, ensuring that cells devote resources to these functions only under appropriate conditions. The target of rapamycin (TOR) protein kinase is an essential and conserved component of such regulation, and orchestrates a comprehensive set of cellular responses to nutrient levels. Under favorable conditions, TOR signaling promotes protein synthesis through upregulation of ribosome biogenesis and direct activation of the translation machinery. In response to nutrient withdrawal, reduction of TOR activity inhibits biosynthesis and causes activation of autophagy, in which cytoplasmic components are degraded in the lysosomal compartment. These responses promote cell survival during periods of starvation by reducing demand and generating an intracellular source of recycled nutrients. How the TOR pathway senses nutrient status is only partially understood. One upstream signal involves the Rag family of small GTPases, which regulate the association of TOR with the GTPase Rheb, a direct activator of TOR. The more proximal upstream steps in this pathway by which amino acid levels are assessed have remained elusive, with nutrient transporters, uncharged tRNAs, and intracellular levels of ATP or calcium each being proposed as potential mediators of the signal.1 tRNAs are essential intermediates of protein synthesis, translating the mRNA ribonucleotide code into polypeptide sequence. tRNAs synthesized in the nucleus are actively exported to the cytoplasm by specific transport factors. Recent studies in yeast and mammalian cells have also identified a retrograde pathway of tRNA import into the nucleus.2,3 Interestingly, the balance between nuclear and cytoplasmic pools of tRNA is regulated by nutrient availability, with starvation causing a rapid and reversible accumulation of tRNA in the nucleus. This response has been proposed to provide an alternative means of translational control under nutrient-poor conditions, by reducing cytoplasmic pools of charged tRNAs available for polypeptide chain elongation. A report by Huynh et al. in a previous issue of Cell Cycle4 provides further insight into this process, and identifies a new role for tRNA trafficking in transducing nutrient signals and controlling TOR-dependent responses. These authors manipulated tRNA localization by targeting the karyopherin exportin-t (Xpo-t), a tRNA-specific nuclear export receptor. Depletion of Xpo-t in human fibroblasts led to accumulation of tRNA in the nucleus, consistent with previous studies in yeast. In Xpo-t depleted cells, phosphorylation of several TOR-dependent targets (as well as TOR itself) was significantly reduced, suggesting that altered tRNA localization leads to downregulation of TOR activity. Xpo-t depletion also caused activation of autophagy in these cells, consistent with the observed reduction in TOR signaling. How might nuclear accumulation or cytoplasmic depletion of tRNA lead to a decrease in TOR activation? This response would not be expected to result indirectly from reduced translational capacity, as inhibitors of protein synthesis generally have a positive effect on TOR activation, presumably by increasing the intracellular concentration of free amino acids. Interestingly, recent reports have described a number of non-canonical functions of tRNAs including transcriptional regulation, mRNA degradation, translation inhibition, and suppression of apoptosis.5-7 In addition, stimuli such as viral infection, DNA damage and oxidative stress can also lead to nuclear accumulation or specific cleavage of tRNA.5,7,8 These studies implicate the processing and trafficking of tRNA as potential intermediate steps in a number of responses to different cellular stresses, and present a wide range of possible mechanisms by which tRNA exerts regulatory effects on cellular nutrient and energy balance. Taken together, the data suggest that, in addition to its passive role as an adaptor molecule for protein synthesis, tRNA could be yet another in the family of non-coding regulatory RNA molecules that have emerged as potent regulators of cell and developmental biology. The study by Huynh et al. also raises the question of how nutrient conditions affect tRNA localization. In yeast, retrograde import of tRNA has been shown to be constitutive, whereas re-export of imported tRNA is responsive to nutrient levels.9 Although TOR activity is reduced by amino acid starvation, inhibition of TOR with rapamycin does not induce nuclear accumulation of tRNA. Rather, rapamycin was unexpectedly found to block nuclear tRNA accumulation in response to deprivation of amino acids, and had no effect on nuclear tRNA accumulation in response to glucose deprivation.10 These results indicate that different stresses can signal to the tRNA export machinery using distinct pathways. This response may be regulated in part at the level of tRNA aminoacylation by tRNA synthetases, since defects in this process can also block tRNA export.11 Recent genetic screens in Drosophila identified mutations in aminoacyl-tRNA synthetases and nuclear transporters as causing reduction of cell size and activation of autophagy, consistent with a decrease in TOR activity.12 Regardless of mechanism, the influence of tRNA trafficking on TOR signaling presents an interesting parallel with current models of nutrient-dependent TOR regulation, in which amino acids promote the Rag-dependent trafficking of TOR to its site of activation on the surface of the endosomal-lysosomal compartment. An important goal for future studies will be to understand the mechanisms by which nutrients affect the itineraries of these molecular journeys.