The cytoplasmic and nuclear populations of the eukaryote tRNA-isopentenyl transferase have distinct functions with implications in human cancer.

Mod5 is the yeast tRNA isopentenyl transferase, an enzyme that is conserved from bacteria to humans. Mod5 is primarily cytoplasmic where it modifies the A37 position of a few tRNAs, and the yeast enzyme has been shown capable of forming heritable, amyloid-like aggregates that confer a selective advantage in the presence of specific antifungal agents. A subpopulation of Mod5 is also found associated with nuclear tRNA genes, where it contributes tRNA-gene mediated (tgm) silencing of local transcription by RNA polymerase II. The tgm-silencing function of Mod5 has been observed in yeast and a Mod5-deletion in yeast can be complemented by the plant and human tRNA isopentenyl transferases, but not the bacterial enzymes, possibly due to the lack of an extended C-terminal domain found in eukaryotes. In light of this additional nuclear role for Mod5 we discuss the proposed role of the human homologue of Mod5, TRIT1, as a tumor suppressor protein.

Division of labor among the yeast Sol proteins implicated in tRNA nuclear export and carbohydrate metabolism.

SOL1, the founding member of the S. cerevisiae SOL family, was previously identified as a multi-copy suppressor of the los1 defect in tRNA-mediated nonsense suppression. Here we report that the four-member SOL family is not essential and that individual family members appear to have distinct functions. SOL1-SOL4 are homologous to genes encoding 6-phosphogluconolactonase (6Pgl) involved in the pentose phosphate pathway. Both Sol3p and Sol4p affect this activity. However, Sol4p does not act as a los1 multi-copy suppressor. In contrast, neither Sol1p nor Sol2p, both of which correct the los1 defect in nonsense suppression, possess detectable 6Pgl activity. Rather, Sol1p and Sol2p appear to function in tRNA nuclear export as sol1 and sol2 mutants possess elevated levels of nuclear tRNA. Members of the Sol protein family appear to have different subcellular distributions. Thus, Sol3p and Sol4p likely function in carbohydrate metabolism, while Sol1p and Sol2p appear to have roles in tRNA function and nuclear export, thereby defining an unusual protein family whose individual members are biochemically distinct and spatially dispersed.

Maf1p, a negative effector of RNA polymerase III in Saccharomyces cerevisiae.

Although yeast RNA polymerase III (Pol III) and the auxiliary factors TFIIIC and TFIIIB are well characterized, the mechanisms of class III gene regulation are poorly understood. Previous studies identified MAF1, a gene that affects tRNA suppressor efficiency and interacts genetically with Pol III. We show here that tRNA levels are elevated in maf1 mutant cells. In keeping with the higher levels of tRNA observed in vivo, the in vitro rate of Pol III RNA synthesis is significantly increased in maf1 cell extracts. Mutations in the RPC160 gene encoding the largest subunit of Pol III which reduce tRNA levels were identified as suppressors of the maf1 growth defect. Interestingly, Maf1p is located in the nucleus and coimmunopurifies with epitope-tagged RNA Pol III. These results indicate that Maf1p acts as a negative effector of Pol III synthesis. This potential regulator of Pol III transcription is likely conserved since orthologs of Maf1p are present in other eukaryotes, including humans.

Role of nuclear pools of aminoacyl-tRNA synthetases in tRNA nuclear export.

Reports of nuclear tRNA aminoacylation and its role in tRNA nuclear export (Lund and Dahlberg, 1998; Sarkar et al., 1999; Grosshans et al., 20001) have led to the prediction that there should be nuclear pools of aminoacyl-tRNA synthetases. We report that in budding yeast there are nuclear pools of tyrosyl-tRNA synthetase, Tys1p. By sequence alignments we predicted a Tys1p nuclear localization sequence and showed it to be sufficient for nuclear location of a passenger protein. Mutations of this nuclear localization sequence in endogenous Tys1p reduce nuclear Tys1p pools, indicating that the motif is also important for nucleus location. The mutations do not significantly affect catalytic activity, but they do cause defects in export of tRNAs to the cytosol. Despite export defects, the cells are viable, indicating that nuclear tRNA aminoacylation is not required for all tRNA nuclear export paths. Because the tRNA nuclear exportin, Los1p, is also unessential, we tested whether tRNA aminoacylation and Los1p operate in alternative tRNA nuclear export paths. No genetic interactions between aminoacyl-tRNA synthetases and Los1p were detected, indicating that tRNA nuclear aminoacylation and Los1p operate in the same export pathway or there are more than two pathways for tRNA nuclear export.

WW domains of Rsp5p define different functions: determination of roles in fluid phase and uracil permease endocytosis in Saccharomyces cerevisiae.

Rsp5p, ubiquitin-protein ligase, an enzyme of the ubiquitination pathway, contains three WW domains that mediate protein-protein interactions. To determine if these domains adapt Rsp5p to a subset of substrates involved in numerous cellular processes, we generated mutations in individual or combinations of the WW domains. The rsp5-w1, rsp5-w2, and rsp5-w3 mutant alleles complement RSP5 deletions at 30 degrees. Thus, individual WW domains are not essential. Each rsp5-w mutation caused temperature-sensitive growth. Among variants with mutations in multiple WW domains, only rsp5-w1w2 complemented the deletion. Thus, the WW3 domain is sufficient for Rsp5p essential functions. To determine whether rsp5-w mutations affect endocytosis, fluid phase and uracil permease (Fur4p) endocytosis was examined. The WW3 domain is important for both processes. WW2 appears not to be important for fluid phase endocytosis whereas it is important for Fur4p endocytosis. In contrast, the WW1 domain affects fluid phase endocytosis, but it does not appear to function in Fur4p endocytosis. Thus, various WW domains play different roles in the endocytosis of these two substrates. Rsp5p is located in the cytoplasm in a punctate pattern that does not change during the cell cycle. Altering WW domains does not change the location of Rsp5p.

Splicing before import - an intein in a mitochondrially targeted preprotein folds and is catalytically active in the cytoplasm in vivo.

Nuclear-encoded mitochondrial proteins are cytoplasmically synthesized and imported into the organelle. The intein-containing RecA protein of Mycobacterium tuberculosis, with or without the CoxIVp mitochondrial targeting signal (MTS), was used to determine where a protein targeted to mitochondria folds and becomes catalytically active. Analysis of fractions from Saccharomyces cerevisiae cells expressing RecA without the MTS revealed that RecA and intein proteins remained cytoplasmic. With the MTS, most of RecA was directed to mitochondria, while most of the intein remained in the cytoplasm. The intein therefore folds into a catalytically active state in the cytoplasm prior to RecA import into mitochondria.

Defects in tRNA processing and nuclear export induce GCN4 translation independently of phosphorylation of the alpha subunit of eukaryotic translation initiation factor 2.

Induction of GCN4 translation in amino acid-starved cells involves the inhibition of initiator tRNA(Met) binding to eukaryotic translation initiation factor 2 (eIF2) in response to eIF2 phosphorylation by protein kinase GCN2. It was shown previously that GCN4 translation could be induced independently of GCN2 by overexpressing a mutant tRNA(AAC)(Val) (tRNA(Val*)) or the RNA component of RNase MRP encoded by NME1. Here we show that overexpression of the tRNA pseudouridine 55 synthase encoded by PUS4 also leads to translational derepression of GCN4 (Gcd(-) phenotype) independently of eIF2 phosphorylation. Surprisingly, the Gcd(-) phenotype of high-copy-number PUS4 (hcPUS4) did not require PUS4 enzymatic activity, and several lines of evidence indicate that PUS4 overexpression did not diminish functional initiator tRNA(Met) levels. The presence of hcPUS4 or hcNME1 led to the accumulation of certain tRNA precursors, and their Gcd(-) phenotypes were reversed by overexpressing the RNA component of RNase P (RPR1), responsible for 5'-end processing of all tRNAs. Consistently, overexpression of a mutant pre-tRNA(Tyr) that cannot be processed by RNase P had a Gcd(-) phenotype. Interestingly, the Gcd(-) phenotype of hcPUS4 also was reversed by overexpressing LOS1, required for efficient nuclear export of tRNA, and los1Delta cells have a Gcd(-) phenotype. Overproduced PUS4 appears to impede 5'-end processing or export of certain tRNAs in the nucleus in a manner remedied by increased expression of RNase P or LOS1, respectively. The mutant tRNA(Val*) showed nuclear accumulation in otherwise wild-type cells, suggesting a defect in export to the cytoplasm. We propose that yeast contains a nuclear surveillance system that perceives defects in processing or export of tRNA and evokes a reduction in translation initiation at the step of initiator tRNA(Met) binding to the ribosome.

Competition between a sterol biosynthetic enzyme and tRNA modification in addition to changes in the protein synthesis machinery causes altered nonsense suppression.

The Saccharomyces cerevisiae Mod5 protein catalyzes isopentenylation of A to i(6)A on tRNAs in the nucleus, cytosol, and mitochondria. The substrate for Mod5p, dimethylallyl pyrophosphate, is also a substrate for Erg20p that catalyzes an essential step in sterol biosynthesis. Changing the distribution of Mod5p so that less Mod5p is present in the cytosol decreases i(6)A on cytosolic tRNAs and alters tRNA-mediated nonsense suppression. We devised a colony color/growth assay to assess tRNA-mediated nonsense suppression and used it to search for genes, which, when overexpressed, affect nonsense suppression. We identified SAL6, TEF4, and YDL219w, all of which likely affect nonsense suppression via alteration of the protein synthesis machinery. We also identified ARC1, whose product interacts with aminoacyl synthetases. Interestingly, we identified ERG20. Midwestern analysis showed that yeast cells overproducing Erg20p have reduced levels of i(6)A on tRNAs. Thus, Erg20p appears to affect nonsense suppression by competing with Mod5p for substrate. Identification of ERG20 reveals that yeast have a limited pool of dimethylallyl pyrophosphate. It also demonstrates that disrupting the balance between enzymes that use dimethylallyl pyrophosphate as substrate affects translation.

ADEPTs: information necessary for subcellular distribution of eukaryotic sorting isozymes resides in domains missing from eubacterial and archaeal counterparts.

Sorting isozymes are encoded by single genes, but the encoded proteins are distributed to multiple subcellular compartments. We surveyed the predicted protein sequences of several nucleic acid interacting sorting isozymes from the eukaryotic taxonomic domain and compared them with their homologs in the archaeal and eubacterial domains. Here, we summarize the data showing that the eukaryotic sorting isozymes often possess sequences not present in the archaeal and eubacterial counterparts and that the additional sequences can act to target the eukaryotic proteins to their appropriate subcellular locations. Therefore, we have named these protein domains ADEPTs (Additional Domains for Eukaryotic Protein Targeting). Identification of additional domains by phylogenetic comparisons should be generally useful for locating candidate sequences important for subcellular distribution of eukaryotic proteins.

Nuclear tRNA aminoacylation and its role in nuclear export of endogenous tRNAs in Saccharomyces cerevisiae.

Nuclear tRNA aminoacylation was proposed to provide a proofreading step in Xenopus oocytes, ensuring nuclear export of functional tRNAs [Lund, E. & Dahlberg, J. E. (1998) Science 282, 2082-2085]. Herein, it is documented that tRNA aminoacylation also occurs in yeast nuclei and is important for tRNA export. We propose that tRNA aminoacylation functions in one of at least two parallel paths of tRNA export in yeast. Alteration of one aminoacyl-tRNA synthetase affects export of only cognate tRNA, whereas alterations of two other aminoacyl-tRNA synthetases affect export of both cognate and noncognate tRNAs. Saturation of tRNA export pathway is a possible explanation of this phenomenon.

Nucleocytoplasmic transport: Inside out regulation.

Cells can respond to extracellular signals by redistributing transcription regulators between the nucleus and cytosol. Recent findings in budding yeast indicate that the nuclear transport receptor Msn5p mediates the nuclear export of several transcription regulators after their phosphorylation in the nucleus.

The human WASP-interacting protein, WIP, activates the cell polarity pathway in yeast.

WIP, the Wiskott-Aldrich syndrome protein-interacting protein, is a human protein involved in actin polymerization and redistribution in lymphoid cells. The mechanism by which WIP reorganizes actin cytoskeleton is unknown. WIP is similar to yeast verprolin, an actin- and myosin-interacting protein required for polarized morphogenesis. To determine whether WIP and verprolin are functional homologues, we analyzed the function of WIP in yeast. WIP suppresses the growth defects of VRP1 missense and null mutations as well as the defects in cytoskeletal organization and endocytosis observed in vrp1-1 cells. The ability of WIP to replace verprolin is dependent on its WH2 actin binding domain and a putative profilin binding domain. Immunofluorescence localization of WIP in yeast cells reveals a pattern consistent with its function at the cortical sites of growth. Thus, like verprolin, WIP functions in yeast to link the polarity development pathway and the actin cytoskeleton to generate cytoskeletal asymmetry. A role for WIP in cell polarity provides a framework for unifying, under a common paradigm, distinct molecular defects associated with immunodeficiencies like Wiskott-Aldrich syndrome.

Saccharomyces cerevisiae Mod5p-II contains sequences antagonistic for nuclear and cytosolic locations.

MOD5 encodes a tRNA modification activity located in three subcellular compartments. Alternative translation initiation generates Mod5p-I, located in the mitochondria and the cytosol, and Mod5p-II, located in the cytosol and nucleus. Here we study the nucleus/cytosol distribution of overexpressed Mod5p-II. Nuclear Mod5p-II appears concentrated in the nucleolus, perhaps indicating that the nuclear pool may have a different biological role than the cytoplasmic and mitochondrial pools. Mod5p contains three motifs resembling bipartite-like nuclear localization sequences (NLSs), but only one is sufficient to locate a passenger protein to the nucleus. Mutations of basic residues of this motif cumulatively contribute to a cytosolic location for the fusion proteins. These alterations also cause decreased nuclear pools of endogenous Mod5p-II. Depletion of nuclear Mod5p-II does not affect tRNATyr function. Despite the NLS, most Mod5p is cytosolic. We assessed whether Mod5p sequences cause a karyophilic reporter to be located in the cytosol. By this assay, Mod5p may contain more than one region that functions as cytoplasmic retention and/or nuclear export sequences. Thus, distribution of Mod5p results from the presence/absence of mitochondrial targeting information and sequences antagonistic for nuclear and cytosolic locations. Mod5p is highly conserved; sequences responsible for subcellular distribution appear to reside in "accessory" motifs missing from prokaryotic counterparts.

Antagonistic effects of NES and NLS motifs determine S. cerevisiae Rna1p subcellular distribution.

Nucleus/cytosol exchange requires a GTPase, Ran. In yeast Rna1p is the GTPase activating protein for Ran (RanGAP) and Prp20p is the Ran GDP/GTP exchange factor (GEF). RanGAP is primarily cytosolic and GEF is nuclear. Their subcellular distributions led to the prediction that Ran-GTP hydrolysis takes place solely in the cytosol and GDP/GTP exchange solely in the nucleus. Current models propose that the Ran-GTP/Ran-GDP gradient across the nuclear membrane determines the direction of exchange. We provide three lines of evidence that Rna1p enters and leaves the nuclear interior. (1) Rna1p possesses leucine-rich nuclear export sequences (NES) that are able to relocate a passenger karyophilic protein to the cytosol; alterations of consensus residues re-establish nuclear location. (2) Rna1p possesses other sequences that function as a novel nuclear localization sequence able to deliver a passenger cytosolic protein to the nucleus. (3) Endogenous Rna1p location is dependent upon Xpo1p/Crm1p, the yeast exportin for leucine-rich NES-containing proteins. The data support the hypothesis that Rna1p exists on both sides of the nuclear membrane, perhaps regulating the Ran-GTP/Ran-GDP gradient, participating in a complete RanGTPase nuclear cycle or serving a novel function.

tRNA nuclear export in saccharomyces cerevisiae: in situ hybridization analysis.

To understand the factors specifically affecting tRNA nuclear export, we adapted in situ hybridization procedures to locate endogenous levels of individual tRNA families in wild-type and mutant yeast cells. Our studies of tRNAs encoded by genes lacking introns show that nucleoporin Nup116p affects both poly(A) RNA and tRNA export, whereas Nup159p affects only poly(A) RNA export. Los1p is similar to exportin-t, which facilitates vertebrate tRNA export. A los1 deletion mutation affects tRNA but not poly(A) RNA export. The data support the notion that Los1p and exportin-t are functional homologues. Because LOS1 is nonessential, tRNA export in vertebrate and yeast cells likely involves factors in addition to exportin-t. Mutation of RNA1, which encodes RanGAP, causes nuclear accumulation of tRNAs and poly(A) RNA. Many yeast mutants, including those with the rna1-1 mutation, affect both pre-tRNA splicing and RNA export. Our studies of the location of intron-containing pre-tRNAs in the rna1-1 mutant rule out the possibility that this results from tRNA export occurring before splicing. Our results also argue against inappropriate subnuclear compartmentalization causing defects in pre-tRNA splicing. Rather, the data support "feedback" of nucleus/cytosol exchange to the pre-tRNA splicing machinery.

MDP1, a Saccharomyces cerevisiae gene involved in mitochondrial/cytoplasmic protein distribution, is identical to the ubiquitin-protein ligase gene RSP5.

Alteration of the subcellular distribution of Mod5p-I, a tRNA modification enzyme, member of the sorting isozyme family, affects tRNA-mediated nonsense suppression. Altered suppression efficiency was used to identify MDP genes, which, when mutant, change the mitochondrial/cytosolic distribution of Mod5p-I,KR6. MDP2 is the previously identified VRP1, which encodes verprolin, required for proper organization of the actin cytoskeleton. MDP3 is identical to PAN1, which encodes a protein involved in initiation of translation and actin cytoskeleton organization. We report here the cloning and characterization of wild-type and mutant MDP1 alleles and the isolation and characterization of a multicopy suppressor of mdp1 mutations. MDP1 is identical to RSP5, which encodes ubiquitin-protein ligase, and mdp1 mutations are suppressed by high copy expression of ubiquitin. All four characterized mdp1 mutations cause missense changes located in the hect domain of Rsp5p that is highly conserved among ubiquitin-protein ligases. In addition to its well-known function in protein turnover, ubiquitination has been proposed to play roles in subcellular sorting of proteins via endocytosis and in delivery of proteins to peroxisomes, the endoplasmic reticulum and mitochondria. mdp1, as well as mdp2/vrp1 and mdp3/pan1 mutations, affect endocytosis. Further, mdp1 mutations show synthetic interactions with mdp2/vrp1 and mdp3/pan1. Identification of MDP1 as RSP5, along with our previous identification of MDP2/VRP1 and MDP3/PAN1, implicate interactions of the ubiquitin system, the actin cytoskeleton and protein synthesis in the subcellular distribution of proteins.

Actin-binding verprolin is a polarity development protein required for the morphogenesis and function of the yeast actin cytoskeleton.

Yeast verprolin, encoded by VRP1, is implicated in cell growth, cytoskeletal organization, endocytosis and mitochondrial protein distribution and function. We show that verprolin is also required for bipolar bud-site selection. Previously we reported that additional actin suppresses the temperature-dependent growth defect caused by a mutation in VRP1. Here we show that additional actin suppresses all known defects caused by vrp1-1 and conclude that the defects relate to an abnormal cytoskeleton. Using the two-hybrid system, we show that verprolin binds actin. An actin-binding domain maps to the LKKAET hexapeptide located in the first 70 amino acids. A similar hexapeptide in other acting-binding proteins was previously shown to be necessary for actin-binding activity. The entire 70- amino acid motif is conserved in novel higher eukaryotic proteins that we predict to be actin-binding, and also in the actin-binding proteins, WASP and N-WASP. Verprolin-GFP in live cells has a cell cycle-dependent distribution similar to the actin cortical cytoskeleton. In fixed cells hemagglutinin-tagged Vrp1p often co-localizes with actin in cortical patches. However, disassembly of the actin cytoskeleton using Latrunculin-A does not alter verprolin's location, indicating that verprolin establishes and maintains its location independent of the actin cytoskeleton. Verprolin is a new member of the actin-binding protein family that serves as a polarity development protein, perhaps by anchoring actin. We speculate that the effects of verprolin upon the actin cytoskeleton might influence mitochondrial protein sorting/function via mRNA distribution.

Regulation of mRNA export in response to stress in Saccharomyces cerevisiae.

The response of eukaryotic cells to heat shock and other forms of stress occurs at both transcriptional and post-transcriptional levels. We used in situ hybridization to determine whether stress affected the subcellular distribution of poly(A)+ RNA in Saccharomyces cerevisiae. Following induction of stress by either heat shock (42 degrees C) or addition of a high concentration of ethanol (10%), the nucleocytoplasmic export of most poly(A)+ RNA was blocked. In situ hybridization indicated that heat-inducible SSA4 and SSA1 mRNAs were exported from nuclei under these same conditions. On the other hand, both GAL1 and URA3 transcripts expressed from the SSA4 promoter accumulated in nuclei following heat shock. Sequences within either the 5' 1600 or the 3' 500 nucleotides of SSA4 mRNA were sufficient to direct GAL1 mRNA to the cytoplasm during stress. The export of SSA4 mRNA following stress required functional nuclear pore complexes, as SSA4 mRNA accumulated in nuclei following heat shock of cells containing temperature-sensitive nucleoporins. However, the selective export of SSA4 mRNA was maintained in heat-shocked cells carrying temperature-sensitive alleles of RNA1, PRP20, or an inducible dominant-negative allele of GSP1, the S. cerevisiae homolog of RAN/TC4. The results reported here suggest that there is selective export of mRNA in yeast.

Nucleus-associated pools of Rna1p, the Saccharomyces cerevisiae Ran/TC4 GTPse activating protein involved in nucleus/cytosol transit.

Rna1p is the GTPase activating enzyme for Ran/TC4, a Ras-like GTPase necessary for nuclear/cytosolic exchange. Although most wild-type Rna1p is located in the cytosol, we found that the vast majority of the mutant Rna1-1p and, under appropriate physiological conditions, a small portion of the wild-type Rna1p cofractionate with yeast nuclei. Subnuclear fractionation studies show that most of the Rna1p is tightly associated with nuclear components, and that a portion of the active protein can be solubilized by treatments that fail to solubilize inactive Rna1-1p. To learn the precise nuclear locations of the Rna1 proteins, we studied their subcellular distributions in HeLa cells. By indirect immuno-fluorescence we show that wild-type Rna1p has three subcellular locations. The majority of the protein is distributed throughout the cytosol, but a portion of the protein is nucleus-associated, located at both the cytosolic surface and within the nucleoplasm. Mutant Rna1-1p is found at the outer nuclear surface and in the cytosol. We propose that a small pool of the wild-type Rna1p is located in the nuclear interior, supporting the model that the same components of the Ran/TC4 GTPase cycle exist on both sides of the nuclear membrane.