Derepression of RNA Polymerase III Transcription by Phosphorylation and Nuclear Export of Its Negative Regulator, Maf1*

Maf1 is the global repressor of RNA polymerase III (Pol III) in yeast Saccharomyces cerevisiae. Transcription regulation by Maf1 is important under stress conditions and during the switch between fermentation and respiration. Under repressive conditions on nonfermentable carbon sources, Maf1 is dephosphorylated and located predominantly in the nucleus. When cells were shifted to glucose medium, Maf1 became phosphorylated and concomitantly relocated to the cytoplasm. This relocation was dependent on Msn5, a carrier responsible for export of several other phosphoproteins out of the nucleus. Using coimmunoprecipitation, Maf1 was found to interact with Msn5. When msn5-Δ cells were transferred to glucose, Maf1 remained in the nucleus. Remarkably, despite constitutive presence in the nucleus, Maf1 was dephosphorylated and phosphorylated normally in the msn5-Δ mutant, and Pol III was under proper regulation. That phosphorylation of Maf1 and Pol III derepression are tightly linked was shown by studying tRNA transcription in Maf1 mutants with an altered pattern of phosphorylation. In summary, we conclude that phosphorylation of Maf1 inside the nucleus acts both directly by decreasing of Maf1-mediated repression of Pol III and indirectly by stimulation of Msn5 binding and export of nuclear Maf1 to the cytoplasm.


Maf1 is the global repressor of RNA polymerase III (Pol III) in yeast Saccharomyces cerevisiae. Transcription regulation by
Maf1 is important under stress conditions and during the switch between fermentation and respiration. Under repressive conditions on nonfermentable carbon sources, Maf1 is dephosphorylated and located predominantly in the nucleus. When cells were shifted to glucose medium, Maf1 became phosphorylated and concomitantly relocated to the cytoplasm. This relocation was dependent on Msn5, a carrier responsible for export of several other phosphoproteins out of the nucleus. Using coimmunoprecipitation, Maf1 was found to interact with Msn5. When msn5-⌬ cells were transferred to glucose, Maf1 remained in the nucleus. Remarkably, despite constitutive presence in the nucleus, Maf1 was dephosphorylated and phosphorylated normally in the msn5-⌬ mutant, and Pol III was under proper regulation. That phosphorylation of Maf1 and Pol III derepression are tightly linked was shown by studying tRNA transcription in Maf1 mutants with an altered pattern of phosphorylation. In summary, we conclude that phosphorylation of Maf1 inside the nucleus acts both directly by decreasing of Maf1-mediated repression of Pol III and indirectly by stimulation of Msn5 binding and export of nuclear Maf1 to the cytoplasm.
Transcription in yeast is regulated by global nutrient-sensing signal transduction cascades. A central component of the eukaryotic growth regulatory network is the TOR (target of rapamycin) cascade, which controls all three polymerase systems involved in ribosome biogenesis (1). How the different environmental signals are integrated by the yeast transcription factor Maf1, a general and global repressor of RNA polymerase III (Pol III), 3 is of considerable interest (2)(3)(4). Maf1 was found to mediate several signaling pathways by repressing Pol III transcription in response to diverse stresses such as nutrient deprivation, rapamycin treatment, secretory defects, or DNA dam-age (5)(6)(7). Cells depleted of Maf1 remain alive under stress conditions but, surprisingly, Maf1-mediated regulation is essential during the transition from fermentative to glycerolbased respiratory growth. This observation underscored the critical function of Maf1 in the coupling of Pol III transcription to metabolic processes and/or energy production dependent on the carbon source (8).
The molecular mechanisms that trigger Maf1 activity in response to different signaling pathways are only partially understood. Maf1 is a phosphoprotein harboring two nuclear localization signals (2,9). Differentially phosphorylated forms of Maf1 can be resolved, and it is the least phosphorylated form that binds Pol III (3,4). In cells growing exponentially in rich glucose medium, Maf1 is highly phosphorylated and localized mostly in the cytoplasm. When cells are shifted to repressive conditions, Maf1 is dephosphorylated and relocated to the nucleus (3,4,8,9). Protein phosphatase 2A, a component of the TOR signaling pathway, is required for Maf1 dephosphorylation, nuclear accumulation, and Pol III repression in response to rapamycin (3). On the other hand, Pol III repression is blocked by artificially increased activity of cAMP-dependent protein kinase A (8,9), possibly via inhibition of Maf1 import to the nucleus (9, 10).
The simplest hypothesis accommodating these observations would be that Maf1 activity is regulated via its phosphorylation state-dependent cellular localization (9). Alternatively, both states, phosphorylation and cellular localization, might contribute in synergy to the regulation of Maf1 activity. At this point, the factor(s) involved in Maf1 shuttling as well as the cellular site(s) of Maf1 phosphorylation-dephosphorylation reactions remain(s) to be identified. In addition, the link between the Maf1 phosphorylation state and its cellular location remains to be established to clarify the key determinant controlling Maf1 activity.
Here we identify Msn5 as the nuclear export receptor for Maf1. In the presence of Msn5, phosphorylated Maf1 is relocated out of the nucleus. However, in msn5-⌬ cells the cellular localization of Maf1 does not correlate with its phosphorylation state. We show that not Maf1 nuclear export but, rather, Maf1 phosphorylation is necessary for derepression of Pol III transcription.
Protein Extraction and Immunoblotting-To ensure that kinases or phosphatases were not functional during cell harvesting and protein extraction, yeast cells were rapidly harvested by centrifugation at 4°C, and 20% trichloroacetic acid was added to the cell pellet as described earlier (8). Cells were broken with acid-washed glass beads, the supernatant was retained, and trichloroacetic acid-precipitated proteins were pelleted by centrifugation. The pellet was resuspended in sample buffer (pH 8.8) and boiled for 5 min. Protein extracts were separated on SDS-PAGE using a modified acrylamide:bisacrylamide ratio of 33.5:0.3. One lane corresponded to the amount of proteins from 1 absorbance unit of cell culture (10 -20 g). The membrane was blocked for 30 min in 10 mM Tris, 150 mM NaCl, and 0.05% Tween 20 containing 5% fat-free dry milk and then incubated for 1 h with Maf1-specific antibody at a 1:10,000 dilution (8). The membrane was incubated with secondary antirabbit antibody coupled to horseradish peroxidase (DAKO), which was visualized by chemiluminescence using the ECL detection kit (Millipore).
Immunofluorescence-Cells were fixed in culture by adding formaldehyde to a final concentration of 3.7% for 1 h, spun down, washed and converted to spheroplasts, which were then applied on microscopic slides covered with polylysine, and subsequently treated as described previously (8). Maf1 antibody was added followed by secondary CY3-conjugated anti-rabbit antibody. Cells were viewed with a Microphot-SA microscope (Nikon). Images were collected using a Photometrix CH350 camera.
Coimmunoprecipitation-Yeast cells expressing wild-type or HA epitope-tagged versions of Msn5 were grown in YPD to exponential phase, transferred to prewarmed YPGly for 1 h, transferred into YPD, incubated for 30 min, and then pelleted and frozen in liquid nitrogen. Pellets, corresponding to 150 ml of culture, were thawed on ice and resuspended in 0.5 ml of buffer H (25 mM Tris-Cl, pH 7.4, 15 mM EGTA, 15 mM MgCl 2 , 1 mM dithiothreitol, 0.1% Triton X-100, 1 mM NaN 3 , 0.25 mM orthovanadate, 15 mM sodium fluoride, 50 mM disodium ␤-phosphoglycerate, and 5 g/ml Complete mixture (Roche Applied Science) supplemented with 10% glycerol). Cells were broken with glass beads using a Mini Beadbeater (Biospec Products) disruptor three times in 30-s pulses. Cellular debris was removed by centrifugation at 3,000 rpm for 10 min. The supernatant was clarified by centrifugation at 11,000 ϫ g for 10 min. Samples were immediately aliquoted in 1-mg portions and frozen at Ϫ80°C. Protein concentrations were determined with the Bio-Rad protein assay. Dynal PanMouse immunoglobulin G magnetic beads (50 l), washed twice with 0.5 ml of phosphatebuffered saline containing 0.5% bovine serum albumin, were incubated overnight with mouse monoclonal anti-HA 12CA5 antibody. After extensive washing in phosphate-buffered saline containing 0.5% bovine serum albumin and then in buffer H, the beads were incubated with 2 mg of protein extract for 4 h with gentle shaking at 4°C. Following supplementation with 5 M NaCl, the immune complexes were pelleted and washed with buffer H. Immunoprecipitated proteins were eluted by boiling for 3 min in the SDS gel electrophoresis loading buffer and analyzed by 8% SDS-PAGE followed by Western blotting.
Generation of Maf1 Mutants-Mutations in plasmid-borne MAF1 were created using a rapid method for localized mutagenesis (12). A fragment of the MAF1 gene was amplified with primers 5Ј-CGAGTTGCTTGTCAATCAGG-3Ј and 5Ј-CTGCTACTGCTCCTTCTTCT-3Ј under low-fidelity conditions using a Diversify PCR random mutagenesis kit (Clontech). The pRS315 (LEU2, CEN) plasmid carrying MAF1 was digested with BclI and BsgI. The gapped linear plasmid and product of the low-fidelity PCR were cotransformed into the YPH500 maf1-⌬ strain. Transformants, selected on minimal medium lacking leucine, were subsequently tested for Maf1 activity by replica plating on YPGly and incubating at 37°C for 3 days. Colonies unable to grow were expected to have compromised Maf1 function. Plasmids with mutated alleles encoding Maf1 were sequenced.

Maf1 Is Exported Out of the Nucleus in Response to Glucose-
Variations in glucose availability in its natural environment force yeast to undergo frequent physiological transitions between fermentation and respiratory growth by using nonfer-mentable carbon sources. The switch between fermentation and respiration causes widespread changes in gene transcription, including Maf1-mediated repression of Pol III transcription (8). We induced these metabolic changes by transferring yeast from a glucose-rich medium to a nonfermentable glycerol medium and then back to the glucose-rich medium to monitor the kinetics of Maf1 phosphorylation and cytoplasmic relocation. Polyclonal antibodies were used to localize Maf1 in yeast cells by immunofluorescence (Fig. 1, top panel). As shown previously (8), the Maf1-specific fluorescence signal was uniformly distributed throughout the cell in the glucose medium (YPD) and concentrated in the nuclear compartment in the glycerol medium (YPGly). Upon transferring the cells back to the glucose medium, Maf1 progressively relocated out of the nucleus (5 min to 1 h). At 30 min after their transfer to the glucose medium, the cells still had some Maf1 in the nucleus. After 1 h, the Maf1 signal became uniformly distributed throughout the cell. To investigate whether the cytoplasmic relocation of Maf1 was correlated with its altered phosphorylation, differentially phosphorylated forms of Maf1 were resolved by SDS-PAGE and identified by immunoblotting as a function of time after culture transfer from YPGly back to YPD (Fig. 1, bottom panel). As reported previously, Maf1 was dephosphorylated upon transition from glucose to glycerol medium (compare YPD and YPGly lanes), and the opposite transition resulted in gradual Maf1 phosphorylation (compare YPGly and YPD lanes; 5 min to 1 h). Remarkably, relocation of Maf1 to the cytoplasm in response to glucose appeared to be concomitant with Maf1 hyperphosphorylation, suggesting a link between the Maf1 phosphorylation state and its cellular location.
Msn5 Is Required for Export of Maf1 from the Nucleus-To identify the factors involved in Maf1 shuttling, we examined yeast strains carrying individual deletions of known importin or exportin genes (13). Nuclear transport receptors such as Mtr10, Kap114, Pdr6, Yrb4, Nmd5, Pse1, Srp1, and Sxm1 appeared to play no role in Maf1 transport (results not shown). In contrast, the Msn5 carrier was found to be required for cytoplasmic localization of Maf1. As shown in Fig. 2, msn5-⌬ mutant cells growing in a glucose-rich medium retained Maf1 in the nucleus, in contrast to the wild type. To test whether the expression of Mns5 was sufficient to relocate Maf1 to the cytoplasm, we transformed the msn5-⌬ mutant with a multicopy plasmid carrying the MSN5 gene. Indeed, under these conditions Maf1 was relocalized to the cytoplasm. We therefore conclude that Msn5 is an exportin for Maf1.
Because exportins have been shown to bind directly to their target proteins, we used coimmunoprecipitation experiments to test whether Msn5 and Maf1 were able to interact with each other. Crude extracts were prepared from Msn5-HA-tagged cells and control wild-type cells with untagged Msn5. Cells were grown in glycerol medium, transferred to prewarmed glucose medium, and harvested after 30 min. HA-tagged Msn5 was immunoprecipitated from cell extracts with anti-HA 12CA5 antibodies, and the immunoprecipitates were examined for the presence of Maf1 by immunoblotting. As shown in Fig. 3 Maf1 Shuttling and Phosphorylation State-Subsequent studies of msn5-⌬ cells showed that in these cells Maf1 was retained in the nucleus upon transition from respiratory conditions to glucose, and incubation was prolonged for up to 3 h (Fig. 4A), thus confirming the requirement of Msn5 for Maf1 export. Most remarkably, in the msn5-⌬ mutant, the Maf1 localization and phosphorylation state were totally uncoupled as Maf1 was dephosphorylated and phosphorylated normally. During the transitions from glucose medium to a nonfermentable carbon source and back to glucose (compare Figs. 1 and  4A), both phosphorylation and dephosphorylation reactions FIGURE 1. Maf1 is phosphorylated and exported out of the nucleus after transferring cells from respiratory conditions to glucose. The BY4741 wild-type strain was grown to exponential (Exp) phase in rich glucose medium (YPD), transferred to prewarmed glycerol medium (YPGly), grown for 2 h, transferred again to glucose medium (YPD), and further incubated at 30°C. Cells were harvested as indicated. Maf1 localization was analyzed by immunofluorescence microscopy using polyclonal anti-Maf1 antibodies. Nuclei were stained with 4Ј,6-diamidino-2-phenylindole (DAPI) (top panel). Crude extracts prepared from the same cells by trichloroacetic acid precipitation were analyzed using SDS-PAGE with a modified acrylamide:bisacrylamide ratio followed by immunoblotting with polyclonal anti-Maf1 antibodies (bottom panel). The slower-migrating diffuse band corresponds to phosphorylated forms of Maf1.
Msn5 overproduction did not perturb the nuclear accumulation of Maf1 under respiratory conditions (Fig. 4B, YPGly lane), although the Maf1 nuclear import rate was slightly decreased (data not shown). Conversely, Msn5 overproduction increased to some extent the Maf1 nuclear export rate. Thirty minutes after the shift from glycerol to glucose medium, the Maf1 signal became uniformly distributed throughout the cell overproducing Msn5, in contrast to the wild-type cell, which needed 1 h to achieve the similar state (compare Figs. 1 and 4B, top panels). The phosphorylated form of Maf1 appeared concomitantly with its relocation out of the nucleus, 30 min after the shift of cells overproducing Msn5 from glycerol to glucose medium (Fig. 4B, bottom panel), suggesting that export and phosphorylation are linked.
Maf1 Nuclear Export Is Not Required for Derepression of Transcription-Knowing that nuclear export of Maf1 repressor was mediated by Msn5, we wondered whether the constitutive nuclear location of Maf1 in msn5-⌬ cells might result in constitutive repression of Pol III transcription. We therefore compared the derepression of Pol III transcription in wild-type and msn5-⌬ cells upon transition from respiratory conditions to glucose. The msn5-⌬ mutant and isogenic wild-type control cells were grown at 30°C to exponential phase in glucose medium (YPD), transferred to prewarmed glycerol medium (YPGly) for 1 h, and then transferred back to YPD. RNA was isolated from cells harvested from the glycerol medium or after 2 or 3 h of growth in glucose-rich medium and analyzed by Northern blotting. To estimate the amounts of newly synthesized tRNAs, we used intron probes complementary to pre-tRNA Tyr , pre-tRNA Leu , and pre-tRNA Phe (Fig. 5A). Quantification of the intron-containing tRNA precursors revealed a similar repression of Pol III activity in both wild-type and msn5-⌬ strains upon transition of glucosegrown cells to glycerol and a similar derepression level upon transition back to glucose medium (Fig. 5B)   . Msn5 mediates nuclear export of Maf1 in response to glucose but does not affect kinetics of Maf1 phosphorylation. Cells were grown to exponential phase in rich glucose medium (YPD), transferred to prewarmed glycerol medium (YPGly), grown for 2 h, transferred again to glucose medium (YPD), further incubated at 30°C, and harvested as indicated. By using polyclonal anti-Maf1 antibodies, Maf1 localization was analyzed by immunofluorescence, and Maf1 phosphorylation was tested by Western blotting as described in the legend to Fig. 1. A, msn5-⌬  pAG49 (Asn 17 3 His and Leu 43 3 Ser), which had constant phosphorylation of Maf1, even when cultivated in glycerol medium (Fig. 6, top panel). An altered pattern of Maf1 phosphorylation observed in mutants may be related to the new serines generated by the amino acid substitutions. To inspect the effect of pAG23 and pAG49 on Pol III activity, RNA isolated from the respective cells grown in YPD and transferred to YPGly was analyzed by Northern blotting using a probe complementary to pre-tRNA Leu . Clearly, both mutants were defective in their ability to repress tRNA Leu transcription upon transfer from glucose to glycerol medium (Fig. 6, bottom panel). Thus, the constant phosphorylation of Maf1 precluded Pol III repression.

DISCUSSION
RNA Pol III produces essential components of the protein biosynthesis machinery, and therefore its activity is tightly coupled to cell growth and metabolism. The mechanism for this coordination is provided by Maf1, a general negative regulator conserved from yeast to man (2,14,15). Maf1 of Saccharomyces cerevisiae contains 15.7% of serines, which provide nearly 30 potential phosphorylation sites specific for several kinases (data according to Swiss-Prot PROSITE). Thus, phosphorylation of different sites may be used to regulate Maf1 function under various conditions. However, so far there has been no proof that Maf1 activity is controlled by phosphorylation, although indirect indications were obvious. First, protein phosphatase 2A was shown to be required for Maf1-generated repression of Pol III and for nuclear accumulation of Maf1 in response to rapamycin (3,4). Second, it was shown that under conditions of favorable growth, Maf1 was largely phosphorylated and located mostly in the cytoplasm, although a low level of phosphorylated and nuclearly localized Maf1 was also observed (8,9). In this report we present evidence that both phosphorylation and dephosphorylation of Maf1 may occur inside the nucleus. Concomitantly with phosphorylation, Maf1 is exported to the cytoplasm by the Msn5 carrier. However, it is not the export but the phosphorylation of Maf1 that is essential for the derepression of Pol III transcription in response to favorable growth conditions. As shown previously, Maf1 is relocated out of the nucleus in response to a shift of respiratory growing culture to a medium with glucose (8). We used this observation to examine yeast mutants defective in various exportins and found constitutive nuclear localization of Maf1 in msn5-⌬ cells. Two lines of evidence show that Msn5 functions as an exportin for Maf1. When the msn5-⌬ cells are transformed with an Msn5-encoding plasmid, relocation of Maf1 to the cytoplasm is observed again. Furthermore, using coimmunoprecipitation we were able to show that Maf1 interacts with Msn5.
It is worth mentioning that the Msn5 carrier, nuclearly localized under standard conditions in glucose, becomes redistributed to the cytoplasm upon transfer to a medium with a nonfermentable carbon source (16). It was shown that whenever Msn5 was detected in the cytoplasm, Saa4 protein, another Msn5 cargo, was concentrated in the nucleus (16). It was suggested that when Msn5 was sequestered in the cytoplasm, it was unable to ferry its cargo from the nucleus. The location of Msn5 may affect distribution of Maf1 in a similar manner.
It has been suggested previously that Msn5 specifically exports phosphoproteins because several Msn5 cargoes, such as Pho4, Mig1, Crz1, Aft1, Cdh1, and HO, have to be phosphorylated to be relocated out of the nucleus (17)(18)(19)(20)(21)(22)(23). However, binding of Msn5 to other cargoes, such as Far1, does not depend on the cargo phosphorylation state (24).  msn5-⌬ cells. A, Northern blot analysis. msn5-⌬ and wild-type (wt) control cells were grown to exponential (Exp) phase in rich glucose medium (YPD), transferred to prewarmed rich glycerol medium (YPGly) for 2 h, transferred again to YPD, and incubated at 30°C. RNA (10 g) isolated from cells at the indicated growth conditions was separated on 8 M urea-containing 10% polyacrylamide gel followed by hybridization using labeled oligonucleotide probes complementary to pre-tRNA Tyr , pre-tRNA Leu , pre-tRNA Phe , and U3 small nucleolar RNA (snoRNA) (loading control). B, quantification of pre-tRNA Leu synthesis. pre-tRNA Leu bands were quantified as described under "Experimental Procedures." Bars represent the levels for each pre-tRNA Leu transcript normalized to U3 small nucleolar RNA transcript. -Fold change was calculated relative to expression in exponential glucose cells. The experiment was performed in triplicate to estimate S.D. FIGURE 6. Maf1 phosphorylation works against Pol III repression. Wildtype (wt) strain YPH500 and isogenic maf1 mutants pAG23 and pAG49 were grown to exponential phase in glucose medium (YPD) and then transferred to glycerol medium (YPGly) for 2 h. Protein extracts of lysed cells were analyzed by Western blotting polyclonal anti-Maf1 antibodies. Total RNA isolated from cells was tested by Northern blotting with a pre-tRNA Leu probe. Values are the ratios of pre-tRNA Leu transcription in YPD over the transcription in YPGly.
The good synchronization of phosphorylation and export to the cytoplasm observed for Maf1 in wild-type cells suggests that both processes might be interdependent. This relation becomes more understandable when studied in strains with altered Msn5 levels. Examination of msn5-⌬ cells shows that export to the cytoplasm is not necessary for Maf1 phosphorylation. This experiment clearly documents that Maf1 can be dephosphorylated and phosphorylated in the nucleus, suggesting that phosphorylation of Maf1 precedes its export. This is not questioned by examination of a strain overproducing Msn5. Although the rate of Maf1 export was enhanced with no change in the kinetics of its phosphorylation, a slight amount of the phosphorylated form appeared at the same time when relocation to the cytoplasm became visible. Most likely, the excess of Msn5 titrates phosphorylated Maf1 more quickly. In summary, we concluded that during the transfer of cells from a nonfermentable carbon source to glucose, nuclear Maf1 becomes phosphorylated and subsequently exported to the cytoplasm by Msn5. Thus, Maf1 belongs to the major category of Msn5 substrates, which are bound by Msn5 as phosphoproteins. Similar to the other Msn5 cargoes, Maf1 does not contain an obvious leucinerich nuclear export signal. A comparative bioinformatic analysis of all Msn5 cargoes showed no common motifs or domains, thus suggesting Msn5 binding to a specific secondary structure containing phosphate residues.
The kinase that phosphorylates Maf1 upon transferring cells from a nonfermentable carbon source to glucose has not been identified yet. The nuclear localization of this kinase has been suggested by examination of Maf1 phosphorylation in msn5-⌬ cells. However, it cannot be excluded that the unknown kinase is also Msn5 cargo. In this case, the kinase would likewise be retained in the nucleus of msn5-⌬ cells but transported from the nucleus to the cytoplasm in cells that do express Msn5. Thus, it remains to be shown whether Maf1 can be phosphorylated within the nucleus of wild-type yeast.
Maf1 was previously found as a substrate of protein kinase A kinase (9,25). However, our previous data indicated that the protein kinase A-directed phosphorylation of Maf1 was not involved in the coupling of Pol III regulation to the carbon source (8). Moreover, protein kinase A is considered, rather, a cytoplasmic kinase, although its nuclear localization is also possible (26). Previously, Moir et al. (9) reported that protein kinase A-directed phosphorylation of Maf1 prohibited its nuclear import under favorable growth conditions. It may be that phosphorylation of protein kinase A sites in Maf1 prevents its entry to the nucleus, whereas phosphorylation of other sites by an unknown nuclear kinase works against interaction of Maf1 with Pol III.
Interestingly, despite the unusual, constitutively nuclear localization of Maf1 in msn5-⌬ cells, Pol III-transcribed genes are under proper regulation. This indicates that phosphorylation of nuclear Maf1 causes Pol III derepression in glucosegrown cells. This hypothesis is further supported by our finding of a direct link between constant Maf1 phosphorylation and the defect in Pol III regulation observed in pAG23 and pAG49 mutants. Thus, we conclude that phosphorylation of Maf1 by an unknown nuclear kinase is a key part of the mechanism controlling Pol III (3). The detailed mechanism of this regulation remains to be elucidated.
A study of the Msn5-dependent cytoplasmic accumulation of Maf1 allowed us to learn important aspects of Maf1 function; a proposed scheme is presented in Fig. 7. Following a transfer of cells to favorable growth conditions, nuclear Maf1 is phosphorylated by an unknown kinase, which prevents its interaction with Pol III and leads to derepression of transcription. The phosphorylated Maf1 is relocated out of the nucleus by the Msn5 exportin. In parallel, protein kinase A-directed phosphorylation at different sites prevents nuclear import of the newly synthesized cytoplasmic Maf1 by interference with its nuclear localization signal sequence (9). This means that Maf1 phosphorylation operates at several levels. Apart from decreasing direct Maf1 binding to Pol III (3), phosphorylations act both to relocate the nuclear pool of Maf1 and to prevent import of cytoplasmic Maf1 to the nucleus.