Report

tRNA Actively Shuttles Between the Nucleus and Cytosol in Yeast

See allHide authors and affiliations

Science  01 Jul 2005:
Vol. 309, Issue 5731, pp. 140-142
DOI: 10.1126/science.1113346

Abstract

Previous evidence suggested that transfer RNAs (tRNAs) cross the nuclear envelope to the cytosol only once after maturing in the nucleus. We now present evidence for nuclear import of tRNAs in yeast. Several export mutants accumulate mature tRNAs in the nucleus even in the absence of transcription. Import requires energy but not the Ran cycle. These results indicate that tRNAs shuttle between the nucleus and cytosol.

Nuclear-encoded tRNAs are transcribed, processed in the nucleus, and exported to the cytosol to facilitate translation (1). Moreover, a nuclear pool of mature tRNAs also exists. In Saccharomyces cerevisiae, certain mutants defective in nuclear transport and tRNA processing accumulate mature tRNAs in the nucleus, which suggests that nuclear mature tRNAs are intermediates waiting for tRNA export (25). Before export, aminoacyl-tRNA synthetases may monitor their maturation (6). However, we recently found that tRNA-splicing endonuclease localizes to the mitochondrial surface in yeast and that an endonuclease mutant accumulates unspliced pre-tRNAs in the cytosol, which indicates that pre-tRNAs are exported to the cytosol and then spliced (7).

To understand the origin of nuclear mature tRNAs, we examined whether tRNAs are imported into the nucleus from the cytosol, using heterokaryon assays (8). In S. cerevisiae, a heterokaryon with two different nuclei sharing the same cytosol is formed from karyogamy-deficient MATa and MATa cells. To visualize tRNAs transcribed from one nucleus, a sup3+ gene (SptRNA-SerUGA::SptRNA-Meti) from Schizosaccharomyces pombe was introduced into the MATa strain. sup3+ is transcribed and processed to mature SptRNA-SerUGA and SptRNA-Meti in S. cerevisiae (9). Hetero-karyons between the MATa cells (target) and the MATa cells with sup3+ (donor) were subjected to fluorescence in situ hybridization (FISH). Pre-SptRNA-SerUGA was detected in only one nucleus in a heterokaryon (Fig. 1A, pre, and fig. S2). In a wild-type background for tRNA export, mature SptRNA-Meti was excluded from target nuclei like endogenous tRNAs (Fig. 1A and fig. S4B). Δlos1 cells, lacking yeast exportin-t (10, 11), exhibit a moderate defect in nuclear export of mature tRNAs (fig. S4B) (2). In Δlos1 heterokaryons, we observed more SptRNA-Meti in target nuclei (Fig. 1B). These data indicate that cytosolic mature tRNAs enter the nucleus.

Fig. 1.

Heterokaryon assay of nuclear import of tRNAs. (A) Heterokaryons formed from MATa cells with sup3+ and MATa cells were incubated in yeast extract, peptone, and dextrose (YPD) for the indicated times and subjected to FISH using probes against mature SptRNA-Meti (mature) and pre-SptRNA-SerUGA (pre). The nucleus and cell outlines are shown in the 4′,6′-diamidino-2-phenylindole (DAPI) and Nomarski images, respectively. Target nuclei are marked by arrows. (B) Heterokaryon assays similar to (A) were performed with Δlos1 haploids.

Next, we investigated nuclear import of endogenous tRNAs in cells growing vegetatively. If all nuclear mature tRNAs are newly transcribed, transcription arrest should reduce the nuclear pool of tRNAs through their export. If mature tRNAs are supplied from the cytosol, this should not necessarily be true. A complicating factor is that the signal intensity of mature tRNAs in the wild-type nucleus is lower than that in the cytosol. To circumvent this problem, we used a Δlos1 Δmsn5 double mutant. Msn5p is a homolog of mammalian exportin-5, which exports the tRNA-eEF1A complex and pre-miRNAs (microRNAs) (1214). Δmsn5 by itself did not alter tRNA localization, but the Δlos1 Δmsn5 mutant showed a strong synthetic defect in mature tRNA export (fig. S4), which suggests that Msn5p contributes to this export. Using this mutant, we analyzed the pool of nuclear tRNAs when transcription was blocked by thiolutin, an RNA polymerase inhibitor (15). In the presence of thiolutin, pre-tRNA-ProUGG disappeared within 1 hour (Northern hybridization, Fig. 2A; FISH, Fig. 2B, pre), whereas mature tRNA-ProUGG and tRNA-TyrGUA accumulated in the nucleus (Fig. 2B, mature). Because no detectable transcription occurred under these conditions, nuclear mature tRNAs must have been supplied from the preexisting cytosolic pool.

Fig. 2.

Treatment with thiolutin causes accumulation of mature tRNA-ProUGG in Δlos1 Δmsn5 cells. (A) Δlos1 Δmsn5 cells treated with the indicated concentrations of thiolutin were subjected to Northern hybridization with an anti-pre-tRNA-ProUGG probe. White triangle, primary transcripts; black triangle, end-matured pre-tRNAs. (B) The cells treated with 5 μg/ml thiolutin for the indicated times were subjected to FISH with probes against tRNA-ProUGG and tRNA-TyrGUA.

To determine whether nuclear import of mature tRNAs requires energy, Δlos1 Δmsn5 cells were incubated with NaN3 and 2-deoxyglucose (2-dG). Under these conditions, both the primary transcript of tRNA-ProUGG and the gradient of mature tRNA-ProUGG across the nuclear envelope (NE) were no longer detected (Fig. 3A and Fig. 3B, e). Reappearance of the primary transcript 10 min after NaN3 and 2-dG removal indicates rapid replenishment of intracellular adenosine triphosphate (Fig. 3A, –thiolutin). Nuclear accumulation of the mature tRNA was also reestablished after removal of NaN3 and 2-dG, even in the presence of thiolutin (Fig. 3B, h and k). We observed similar results with tRNA-IleAAU encoded by intronless genes (Fig. 3B, i and l). These results indicate that the import of both intron-containing and intronless tRNAs requires energy.

Fig. 3.

Nuclear import of tRNAs is energy dependent. (A) Δlos1 Δmsn5 cells grown in YPD (lane 1) were incubated with NaN3 and 2-dG (YPAdG) for 1 hour (lane 2). The cells were chased without (lanes 3 to 8) or with (lanes 9 to 14) thiolutin for the indicated times. Pre-tRNA-ProUGG was detected by Northern hybridization. White triangle, primary transcripts; black triangle, end-matured pre-tRNAs. (B) Δlos1 Δmsn5 cells were treated as in (A) and were harvested before the energy poison treatment (YPD), after the treatment (YPAdG), and after a 2-hour chase without (–thio.) or with (+thio.) thiolutin. Mature tRNA-ProUGG and tRNA-IleAAU were visualized with FISH. (C) Δlos1 Δmsn5 cells grown in YPD (lane 1) were incubated in YPAdG (lane 2) and chased with thiolutin for 1 hour (lane 3) or 2 hours (lane 4). Aminoacylation states of tRNA-ProUGG and tRNA-TyrGUA were analyzed with acidic urea-PAGE. Lane 5 contains deacylated tRNA. White and black triangles represent aminoacylated and deacylated tRNAs, respectively.

We analyzed aminoacylation states of mature tRNAs imported into the nucleus using acidic urea–polyacrylamide gel electrophoresis (urea-PAGE). Mature tRNA-ProUGG and tRNA-TyrGUA in Δlos1 Δmsn5 cells were present primarily as slower migrating forms even after the chase with thiolutin (Fig. 3C, lanes 1, 3, and 4). These forms were converted to faster migrating forms by base treatment (lane 5), which indicates that the slower migrating tRNAs are aminoacylated. Therefore, imported tRNAs exist mainly in aminoacylated forms in the nucleus.

We asked whether 3′-truncated tRNAs are imported. cca1-1 cells are deficient in both de novo formation and repair of 3′-terminal CCA ends of tRNAs (16), and also in tRNA export (4). We therefore treated cca1-1 cells with NaN3 and 2-dG at a restrictive temperature to abolish the mature tRNA gradient across the NE and chased them with thiolutin. tRNA-TyrGUA and tRNA-IleAAU reaccumulated in the nucleus to similar levels (Fig. 4A, q to t), although the proportions of the 3′-truncated forms of these tRNAs were different (Fig. 4A, bottom). Taken together with the fact that tRNAs in Δlos1 Δmsn5 cells were full-length tRNAs, these results indicate that various forms of tRNAs are imported.

Fig. 4.

Various tRNA species are imported into the nucleus, and import is Ran independent. (A) cca1-1 cells grown at 23°C were incubated at 37°C for 3 hours, treated with NaN3 and 2-dG for 3 hours at 37°C (YPAdG), and then chased without (–thio.) or with (+thio.) thiolutin for 1.5 hours. Samples were analyzed using FISH (top) and Northern hybridization (bottom) with indicated probes. White, black, and gray triangles represent aminoacylated, deacylated, and 3′-truncated tRNAs, respectively. (B) rna1-1 cells grown at 23°C were incubated at 37°C for 1.5 hours to induce defects and then processed as in (A).

To determine the contribution of Ran guanosine triphosphatase (GTPase) (17), we examined tRNA import in rna1-1 cells defective in RanGAP (Ran GTPase activating protein). The rna1-1 cells accumulated mature tRNA-TyrGUA in the nucleus at 37°C (2), and this tRNA gradient disappeared upon treatment with NaN3 and 2-dG (Fig. 4B, 37°C, YPAdG). When the cells were transferred to medium with thiolutin, mature tRNA-TyrGUA was reaccumulated (Fig. 4B, +thio.). Because protein import ceased at 37°C in rna1-1 cells (fig. S5), these results suggest that tRNA import is Ran independent.

Our results indicate that nuclear mature tRNAs are supplied from the cytosol and that tRNAs shuttle between these two compartments. Shuttling may contribute to tRNA quality control, because tRNAs have long lifetimes and may run the risk of inappropriate modifications (18). A quality-control system in the nucleus may repair or filter out inactive tRNAs from those shuttling through the nucleus and provide only active tRNAs back to the cytosol. Another controversial possibility would be to supply tRNAs for nuclear translation (19).

Supporting Online Material

www.sciencemag.org/cgi/content/full/1113346/DC1

Materials and Methods

SOM Text

Figs. S1 to S6

Tables S1 to S3

References

References and Notes

View Abstract

Navigate This Article