INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The yeast RNA polymerase III (Pol III) transcription system is well characterized. Small untranslated RNAs with essential housekeeping functions, such as tRNAs, 5S rRNA, or the U6 small nuclear RNA (snRNA) that is required for mRNA splicing, are synthesized by Pol III with the help of two general auxiliary factors, TFIIIC and TFIIIB. The large TFIIIC factor (six subunits) binds to the DNA promoter elements and assembles the initiation factor TFIIIB (three components) upstream of the start site. Once TFIIIB is in position, it recruits the Pol III enzyme (17 subunits) and directs accurate and multiple rounds of transcription. All of the polypeptide components of the Pol III apparatus (~1,500 kDa) have been characterized and found to be essential for cell viability (8, 23). The identification of the components of the Pol III system has facilitated the description of a cascade of protein-protein interactions that leads to the recruitment of the Pol III enzyme (reviewed in reference 55).

Detailed knowledge of the yeast Pol III transcription system contrasts with the limited information available on the control of class III gene expression in yeast. Cellular tRNA levels respond to cell growth rate (48, 49), to a nutritional upshift (27, 48) or to nitrogen starvation (36) but only modestly to amino acid starvation (41). Finally, Pol III transcription is repressed in secretion-defective cells (30). Although the mechanism of repression is not clear, it does involve activation of the cell integrity pathway (30). The effect of growth conditions on Pol III transcription is well mimicked in vitro with whole-cell extracts (11, 39). tRNA synthesis is downregulated in dense cell cultures approaching stationary phase, a result due essentially to reduced TFIIIB activity. The TFIIIB component Brf/TFIIIB70 was found to be the limiting factor in extracts from such cells (39). However, the occupancy of the TFIIIB binding site on the SUP53 gene encoding tRNA^Leu does not decrease in stationary-phase cells. Rather, in vivo footprinting data suggest reduced promoter occupancy by Pol III (25).

In higher eukaryotic cells, Pol III transcription responds to growth rate, developmental phase, cell cycle position, and a number of pathological conditions (reviewed in reference 55). The regulation operates principally at the level of TFIIIB and TFIIIC (17, 20, 42, 46, 52). The tumor suppressors Rb and p53 inhibit TFIIIB (9, 10, 28, 53). Therefore, it is likely that the control of Pol III transcription rate is important in restraining tumor cell proliferation (54). No equivalent negative regulator of Pol III transcription has been found in yeast.

Genes controlling tRNA synthesis in yeast can be identified by nonsense suppression approaches (22). One candidate for such a gene is Saccharomyces cerevisiae MAF1. It was originally identified by the isolation of maf1-1 as a temperature-sensitive mutation that decreases the efficiency of SUP11 (tRNA Tyr/UAA) suppression (34). A search for multicopy suppressors of maf1-1 revealed an intriguing genetic interaction between MAF1 and RPC160, the gene encoding the largest subunit of the RNA Pol III, C160. RPC160 genes with 3' deletions in their open reading frame suppress the maf1-1 phenotypes when overexpressed (6).

In the present work, we show that tRNA levels are elevated in maf1-1 cells and that spontaneous mutations in RPC160 which reduce tRNA levels also suppress the growth phenotype associated with maf1-1. maf1-1 cell extracts support increased levels of Pol III transcription in vitro compared to wild-type cells. Further, we show that Maf1p is a nuclear protein that physically interacts with RNA Pol III. Therefore, Maf1p appears to be a negative effector of Pol III activity, potentially regulating the level of cellular tRNA in response to external signals. A database search revealed that a variety of organisms have sequences similar to Maf1p, suggesting that this type of Pol III regulation may not be limited to yeast.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Media. The following media were used for growth of yeast: YPD (2% glucose, 2% peptone, 1% yeast extract), YPGly (2% glycerol, 2% peptone, 1% yeast extract), and W0 (2% glucose, 0.67% yeast nitrogen base without amino acids). W0ura, W0trp, and W0leu contained 20 µg of the amino acids per ml required for growth except for the single amino acid as indicated. 5-Fluoroorotic acid (5-FOA) medium was prepared as described previously (4). Sporulation medium (SP1) contains 0.25% yeast extract, 0.1% glucose, and 0.98% potassium acetate. Solid media contained 2% agar. All reagents used for media were Difco products.

Strains and plasmids. The strains used in this study are listed in Table 1. MT6-7, the maf1-1 mutant, was derived by ethyl methanesulfonate mutagenesis of T8-1D (34). The MAF1 gene was disrupted by replacing the internal NheI-HpaI fragment with the URA3 cassette, as described previously (6). KC3-4D, designated as maf1, was derived by sporulation of 4A × 11D diploid heterozygous disruption-deletion of the MAF1 gene. R16 and K2 were selected as spontaneous revertants of maf1-1 in MB123-2C and maf1 in KC3-4D, respectively. In MW671-HA the chromosomal copy of RPC160 is deleted, and C160 is provided by pC160-240 (TRP1 CEN4 HA-RPC160) plasmid carrying the RPC160 gene hemagglutinin HA-tagged at the 5' end (14). DNA coding for 13 repeats of the myc epitope was inserted just before the stop codon in the endogenous MAF1 in MW671-HA by a PCR-based gene deletion and modification system (31).

Identification of maf1-1 mutation. Two overlapping fragments of the maf1-1 allele and control wild-type MAF1 were amplified by PCR using specific primers, AmpliTaq DNA polymerase (Perkin-Elmer), and genomic DNA prepared from strains MT6-7 and T8-1D. The primer sequences were CGAGGATCCGATGAAATTTATTGATGAGCTAGAT, TGCTGTACCAGAACTG, GAGGATAAGAAGAAGGAG, and CGCGGATCCTACTGTAGGGATTCTTC. The PCR products were sequenced with the same primers in an automatic sequencer ABI PRISM (Perkin-Elmer).

Immunofluorescence. Immunofluorescence microscopy of Maf1p-HA protein was performed as described previously (37). Mouse 16B12 anti-HA (BabCo) was used in a 1:750 dilution as the primary antibody, and goat anti-mouse Cy3 (Jackson Immunoresearch Laboratories, West Grove, Pa.) at 1:250 dilution was used as the secondary antibody. Cells were stained with DAPI (4',6'-diamidino-2-phenylindole) at a final concentration of 4 µg/ml for 2 min. Samples were viewed at ×600 magnification on a Nikon Microphot-S.A. microscope equipped with filters for epifluorescence. Images were captured and saved electronically.

Preparation of total RNA and Northern analysis. Cells were grown in liquid YPD or YPGly at 30°C to log phase (A₆₀₀ = ca 0.8) and then shifted to a restrictive temperature for 2.5 h. Total RNA was isolated from cells disrupted with glass beads by phenol extraction using LETS buffer (0.1 M LiCl; 0.01 M Na₂EDTA; 0.01 M Tris-Cl, pH 7.4; 0.2% sodium dodecyl sulfate [SDS]) (26). A total of 10 µg of each RNA sample was resolved by electrophoresis in 7 M urea, 6% polyacrylamide gel electrophoresis (PAGE), or 1.2% Tris-borate-EDTA (TBE) agarose gels. The RNAs were transferred from agarose gel to a Zeta-probe membrane (Bio-Rad) with 10× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate) by capillary action. For quantitative studies, slot blots were prepared. The membrane was prehybridized in 4× SSC-20 mM EDTA-0.5% SDS-0.1 mg of denatured salmon sperm DNA per ml, and the RNA was hybridized at 37°C in the same solution with oligonucleotide probes labeled with [-³²P]ATP and T4 polynucleotide kinase. The probes were as follows: 5'-GCCATCTCCTAGAATCGAACCAGG-3', complementary to tRNA^His; 5'-GGTAAATTACAGTCTTGCG-3', complementary to tRNA^Tyr; 5'-CAGTTGATCGGACGGGAACA-3', complementary to 5S rRNA; and 5'-GGATTGCGGACCAAGCTAA-3', complementary to U3 snRNA. After hybridization, blots were washed 15 min in 6× SSC at room temperature, exposed to film, or exposed to a phosphorimager plate. RNA was quantified using the laser densitometer GelScan XL (Pharmacia). Arithmetic means and standard deviations of tRNA levels, corrected for snRNA U3 levels, obtained for at least three independent blots are presented.

In vivo labeling of RNA. Cells were grown in low-phosphate YPD at 30°C to an optical density at 600 nm of 0.8 and then pulse-labeled using 150 µCi of carrier-free [³²P]orthophosphate per ml of culture. After 5 min, unlabeled potassium phosphate was added to a final concentration of 0.8 M. The cell culture was further incubated for 15 min and then centrifuged and frozen prior to the preparation of total RNA. Labeled RNA (5 µg) was separated on a 7 M urea-6% PAGE gel. Gels were stained with ethidium bromide, dried, and exposed to a phosphor storage plate or autoradiographed. Quantitative analysis of RNA bands was done as described above.

In vitro transcription. Wild-type (T8-1D) and maf1-1 mutant (MT6-7) cell extracts were prepared in parallel essentially as follows. First, 6 to 8 g of cells grown in YPD at 30°C was collected in early stationary phase (A₆₀₀ = ca. 2.0), washed two times with solubilization buffer (200 mM Tris acetate, pH 8.0; 10 mM magnesium acetate; 10 mM -mercaptoethanol; 10% glycerol), resuspended at 1 ml/g (wet weight) of cells in buffer A800 (180 mM Tris acetate, pH 8.0; 9 mM magnesium acetate; 800 mM ammonium sulfate; 10 mM -mercaptoethanol; 10% glycerol, 2.5 mM phenylmethylsulfonyl fluoride; Complete protease inhibitor [Roche]), frozen at 70°C in dry ice and ethanol, and disrupted in an Eaton press. The homogenized paste was resuspended at 0.5 ml/g (wet weight) of cells in buffer A400 (same as buffer A800 but with 400 mM ammonium sulfate), and clarified by centrifugation at 20,000 rpm for 10 min ( Beckman roter JA20). The extract was then diluted with 1 ml of buffer A400/g (wet weight) of cells and centrifuged at 40,000 rpm for 150 min in a Beckman 45Ti ultracentrifuge rotor. The whole supernatant was collected, taking care to avoid the flaky white layer at the top of the tubes, and dialyzed for 3 h against 1 liter of transcription buffer (20 mM Tris acetate, pH 8.0; 8 mM magnesium acetate; 25 mM ammonium sulfate; 0.5 mM dithiothreitol [DTT]; 10% glycerol, 1 mM phenylmethylsulfonyl fluoride; 175 mM potassium glutamate). Insoluble material was removed by low-speed centrifugation.

Immunopurification experiments. Immunoprecipitation reactions were performed with cellular extracts (30 µg/µl of protein in a 50-µl reaction volume) in extraction buffer (HEPES [pH 7.5], 50 mM; EDTA, 0.5 mM; CH₃COOK, 500 mM; DTT, 1 mM; glycerol, 10%). The reactions were incubated while being agitated at 1,000 rpm in the presence of Dynal PanMouse immunoglobulin G beads (Dynal A.S.) preloaded with either -myc or -HA antibody. Incubation was for 3 h at 4°C. Routinely, 2 µg of antibody was incubated with 20 µl of the beads (4 × 10⁸/ml). Beads were recovered by use of a magnet and washed with extraction buffer. The proteins bound to the beads were separated on SDS-8% PAGE gels and analyzed by immunoblotting. The protein concentration was estimated by use of a Bio-Rad protein assay using bovine serum albumin as a standard.

Immunoblotting. Proteins separated by SDS-8% PAGE were transferred electrophoretically to nitrocellulose membranes. After incubation with a specific antibody as indicated, immune complexes were visualized with anti-mouse or anti-rabbit monoclonal antibody coupled to horseradish peroxidase (Amersham Life Sciences). The monoclonal antibodies 9E10 against myc and 12CA5 against HA were applied.

Sequence homology searches. S. cerevisiae Maf1p (GI:642810) was searched against several databases using the BLAST (1) server at The National Center for Biotechnology Information (NCBI). Similar proteins were exhaustively searched against the same databases until all similar proteins were identified. The identified protein sequences were aligned using CLUSTAL X (45) based on the BLOSUM 62 scoring matrix (19). Complete MAF1-like genes were identified in Caenorhabditis elegans (GI:3786409), Candida albicans (Con 5 2883), and Arabidopsis thaliana (GI:7529279). Partial genomic sequences resembling MAF1 were identified in the following organisms: Plasmodium falciparum (AL035476), Cryptosporidium parvum (AQ522040, B83774, and G35235), and Schizosaccharomyces pombe (GI:2388951, GI:2388963, and GI:1507665). Translations of expressed sequence tags (ESTs) from the following organisms appear to be Maf1p homologs: human, AL040071, AA352022, AA179500, AA486502, AA143586, and others; mouse, AA671675, AA066820, AA467315, AA154074, AA042010, and others; rat, AA899627, H34701, H32342, AI172177, and AA817806; zebrafish, AI794541; Drosophila melanogaster, AA142300, AA140704, AI403188, AA539678, and AI109742; Bombyx mori, AU006267; Brugia malayi, AA841076 and AA471615; Arabidopsis thaliana, AL096524, R83973, AI992545, and AA042353; rice, C26657, D24548, C26708, AU030720, and AQ050286; and Saccharum spp., AA525694.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Maf1p is a nuclear protein conserved from yeast to humans. Yeast MAF1 gene (6) encodes a hydrophilic protein of 395 amino acids rich in serine and asparagine residues with a predicted molecular mass of 44.7 kDa. Comparison of the Maf1p sequence with multiple databases using the BLAST server at the NCBI revealed potential orthologs in human, animals, plants, and lower eukaryotes (Fig. 1). No potential prokaryotic orthologs were identified. The eukaryotic proteins identified share three regions of high similarity that are shown in Fig. 1 (labeled regions A, B, and C). Within these regions signature sequences for this protein family can be identified (PDYDFS and WSfnYFFYNkklKR). These motifs have not been previously reported in the PROSITE database (21). Interestingly, in the majority of Maf1p homologs, the second motif includes a putative nuclear targeting signal. The PSORT program found two possible nuclear targeting signals in S. cerevisiae Maf1p: KRRK (position 204) and a double signal RKRK-KRKR (positions 327 and 328).

View larger version (75K): [in this window] [in a new window]   FIG. 1.   Schematic representation of Maf1p and potential orthologs. Abbreviations: Homo sapiens, Hs; Drosophila melanogaster, Dm; Caenorhabditis elegans, Ce; Arabidopsis thaliana, At; Schizosaccharomyces pombe, Sp; Saccharomyces cerevisiae, Sc; Plasmodium falciparum, Pf. Conserved domains are boxed and labeled by letters A (dashed box), B (black box), and C (gray box) in the upper panel. Sequence alignments of each conserved domain (A, B, and C) are shown in the lower panel. Amino acids conserved in at least two sequences are boxed. A star indicates the maf1-1 mutation substituting a stop codon for W319. The two Maf1p-specific signature sequences are underlined.

View larger version (40K): [in this window] [in a new window]   FIG. 2.   Nuclear localization of Maf1p. maf1 (KC3-4D) cells were transformed with the control YCp111 plasmid (A and B) or YCpMAF-4HA, a single-copy plasmid encoding HA-tagged Maf1p (C and D). The cells were stained with DAPI (A and C) or tagged with the 16B12 antibody that recognizes the HA epitope in Maf1p-HA (B and D).

Mutations in RPC160 suppress the maf1 defects. To gain a better understanding of Maf1p function, we searched for second-site suppressors of maf1-1 and spontaneous bypass suppressors of a maf1 disruption. maf1 mutants are temperature sensitive on nonfermentable carbon sources (6). We selected for suppressor mutations that allowed colonies to grow at the restrictive temperature of 37°C on glycerol-containing medium (YPGly). Three spontaneous revertants of maf1-1, i.e., R2, R9, and R16, were isolated. R2 and R9 were obtained from MT6-7, while R16 was obtained from MB123-2C (Fig. 3A). All partially complemented the antisuppressor phenotype of maf1-1 (not shown). In a parallel screen, one revertant, K2, was isolated in maf1 KC3-4D, the MAF1 deletion strain (Table 1). All four suppressors were haploid maf1 mutants with additional chromosomal suppressor mutations that gave a cold-sensitive phenotype. The phenotype of R16 was the most severe since cell growth was reduced at the permissive temperature. Mating of yeast containing the suppressors to each other yielded cold-sensitive diploids, indicating that all mutant alleles were members of a single complementation group.

View larger version (43K): [in this window] [in a new window]   FIG. 3.   Mutations in the RPC160 gene are second site suppressors of maf1 mutants. (A) The R16 suppressor (rpc160-16 maf1-1) and its isogenic maf1-1 strain (MB123-2C), The K2 suppressor (rpc160-2 maf), and the isogenic maf strain (KC3-4D) were grown on YPD and replicated on YPD and YPGly prior to incubation at the indicated temperatures. (B) In vivo content of small RNA species in the rpc160-2 mutant carrying the wild-type MAF1 (strain MB156-1B; Table 1) transformed with empty vector ([ ]) or pC160-6 plasmid ([C160]) that complements its cold-sensitive phenotype. Cells were grown in glucose selective medium and shifted to a nonpermissive temperature (2.5 h at 16°C). RNA was extracted and separated by electrophoresis on a 7 M urea-6% PAGE gel using equal amounts of RNA per lane (10 µg). The gel was stained with ethidium bromide.

Increased tRNA levels in maf1 mutants correlate with increased tRNA synthesis rate in vitro. The recovery of RPC160 mutant alleles as suppressors of maf1-1 prompted us to investigate tRNA levels in maf1 mutants. Total RNA was isolated from maf1-1 (MT6-7), maf1 (KC3-4D), and the isogenic parental strains (T8-1D and 11D, respectively) grown with a shift to the restrictive growth conditions. Separation of total cellular RNAs by electrophoresis and examination of ethidium-stained gels revealed a substantial increase of tRNA levels in maf1-1 cells, whereas the levels of 5S and 5.85 rRNA remained stoichiometric and seemed to be unaffected by the maf1-1 mutation (Fig. 4A).

View larger version (37K): [in this window] [in a new window]   FIG. 4.   tRNA levels are elevated in maf1-1 strains and subsequently reduced in a maf1 rpc160 cold-sensitive suppressor strain. Total RNA was isolated from yeast cells after a shift to a nonpermissive temperature in YPGly (2.5 h at 37°C). (A) Small RNA species from maf1-1 (MT6-7) and the parental strain (T8-1D) were separated on a 7 M urea-6% PAGE gel using equal amounts of RNA per lane (10 µg) and stained with ethidium bromide. (B) Northern analysis of RNA from the wild-type (11D WT), maf1-1 (MT6-7), and parental strain isogenic to maf1-1 (T8-1D) using labeled oligonucleotide probes complementary to U3 RNA, 5S rRNA, tRNAHis, or tRNATyr, as indicated. (C) Northern analysis of RNA from the wild-type (11D WT), maf1-1 (MB123-2C), and maf1-1 strains with a suppressor mutation in C160 (R16) using the probes indicated as in panel B.

View larger version (45K): [in this window] [in a new window]   FIG. 5.   Increased transcription of different Pol III genes in maf1-1 deficient cell extracts. Transcription assays were performed as described in Materials and Methods using a mixture of two genes: a gene transcribed by Pol III and a mini-35S rRNA gene transcribed by Pol I. Transcription was carried out using cell extract (65 µg) from maf1-1 (MT6-7) or parental (T8-1D) strains. RNA was analyzed by urea-PAGE and quantified with a PhosphorImager (Molecular Dynamics). RNA levels transcribed by Pol III were normalized to the level of mini-35S rRNA transcript. (A) Specific transcription of various genes transcribed by Pol III in crude extracts from wild-type (odd lanes) or maf1-1 cells (even lanes). The different DNA templates are indicated. Lane 9, control experiment with no Pol III gene; lane 10, negative control experiment with no added template DNA. (B) Arithmetic means and standard deviations of Pol III transcript levels, corrected for mini-35S rRNA levels, using five independent crude extracts per strain. The different DNA templates are indicated. All of the RNA bands shown in panel A were used for quantification. Gray bars, wild-type crude extract; black bars, maf1-1 crude extracts; AU, arbitrary units.

View larger version (17K): [in this window] [in a new window]   FIG. 6.   In vivo 32P labeling of RNA. The wild-type (T8-1D) and mutant (MT6-7) cells grown in low-phosphate medium were pulse-labeled for 5 min and supplemented with an excess of nonradioactive phosphate for 15 min to allow for the maturation of 5.8S rRNA. RNAs were extracted and analyzed as described in Materials and Methods. Loading of equal amounts of RNA in each lane was confirmed by staining with ethidium bromide prior to drying of the gel and autoradiography. Synthesis rate of tRNA and 5S rRNA is expressed relative to the 5.8S rRNA species.

Maf1p interacts with RNA Pol III. These genetic studies showed that the effect of Maf1p deficiency causing elevated tRNA levels is suppressed by alterations of C160. This genetic interaction suggested that Maf1p and Pol III might physically interact. Coimmunopurification experiments were therefore performed to investigate the possibility of a complex formation between Maf1p and RNA Pol III.

View larger version (45K): [in this window] [in a new window]   FIG. 7.   Maf1p interacts with RNA Pol III. Crude extracts from cells expressing HA-tagged C160 (RPC160-HA), myc-tagged Maf1p (MAF1-myc) or both (RPC160-HA MAF1-myc) were incubated with magnetic beads coated with anti-myc (left panels) or anti-HA (right panels) antibodies. The beads were washed, and the bound polypeptides were eluted and analyzed by SDS-PAGE, followed by immunoblotting using specific antibodies, as indicated on the left. For clarity, the polypeptides revealed by the specific antibodies are identified on the right. (A) The beads coated with anti-myc antibodies retained specifically Maf1p-myc; background retention of C160-HA was undetectable. However, C160-HA copurified with Maf1p-myc (left panel). Reciprocally, Maf1p-myc was specifically retained on the anti-HA beads only in the presence of C160-HA (right panel). (B) Pol III subunits C82, AC40, and C34 specifically copurified with Maf1p-myc (left panel), as well as with C160-HA (right panel).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Regulation of Pol III-directed tRNA biosynthesis is a potentially important mechanism that may contribute to the cell metabolic economy and coordination of translation with growth. The genetic and biochemical data presented here support the model that in S. cerevisiae Maf1p is a negative effector of Pol III: (i) mutations in the MAF1 gene increased the level of tRNAs in vivo and the rate of Pol III RNA synthesis in vitro; (ii) Maf1p, a nuclear protein, coimmunopurified with the components of the Pol III complex; and (iii) mutations in the largest Pol III subunit, C160, suppressed mutations in MAF1 and restored normal levels of tRNA. The presence of proteins similar to yeast Maf1p in a wide variety of organisms suggests that Maf1p homologs may provide a mechanism for the coordination of Pol III transcription with cell growth rate in eukaryotes.

Pleiotropic phenotype of maf1 mutants. maf1-1 was originally isolated as a mutation that decreases the efficiency of the tRNA suppressor SUP11. Although one might have anticipated that increased cellular tRNA levels would improve the efficiency of tRNA-mediated nonsense suppression, our data show the opposite. There are a number of mechanisms that could account for this counterintuitive result. First, since SUP11 tRNA is a minor tRNA Tyr/UAA, even if SUP11 tRNA is proportionally elevated in the maf1 tRNA population, it may compete less efficiently for tyrosyl-tRNA synthetase, EF1, or ribosome binding, resulting in the observed antisuppressor phenotype. This explanation is supported by the observation that increasing the copy number of a SUP11 tRNA encoding gene overcomes the antisuppression of maf1-1 cells (K. Pluta and M. Boguta, unpublished results). Alternatively, reduced suppression could result from altered SUP11 tRNA modification. We recently reported that cells have limiting quantities of dimethylallyl pyrophosphate (3), the precursor for the i⁶A modification found at position 37 of tRNA^Tyr. Since suppression is dependent upon this modification and tRNAs are overproduced, it is possible that not all i⁶A-containing tRNAs are completely modified in maf1 cells. Our preliminary results indicate that tRNAs in maf1-1 strains contain i⁶A (M. Boguta and N. C. Martin, unpublished results); nevertheless, partial modification would result in reduced nonsense suppression.

Effect of Maf1p on Pol III transcription. Steady-state tRNA levels in maf1 mutants are significantly increased relative to 5S rRNA levels that reflect the cellular content in the 60S ribosomal subunit. Northern analysis confirmed that the level of two different tRNAs, tRNA^His and tRNA^Tyr, increased about fourfold but the 5S rRNA level remained the same relative to U3 RNA that is a Pol II transcript. That result is supported by pulse-labeling experiments showing a threefold increase of tRNA synthesis in maf1-1 relative to 5.8S rRNA (Fig. 6). It is likely that the population of tRNAs was increased as a whole, as also suggested by the trailing of the tRNA species in ethidium bromide-stained stained gels (Fig. 4A).

Interaction of Maf1p with RNA Pol III. The increased in vitro Pol III transcription of maf1-1 extracts suggests a direct interaction of Maf1p with the Pol III complex. The observation that mutations in the largest subunit of Pol III, C160, counteract MAF1 inactivation led us to investigate a possible physical interaction between Maf1p and Pol III. Immunopurification experiments demonstrated that Maf1p is directly or indirectly tightly associated with C160 in cell extracts. The precise binding target of Maf1p on Pol III, however, is undefined since C160 is part of a 17-subunit enzyme. That Maf1p is bound to Pol III and not to a hypothetical free pool of C160 polypeptides was confirmed by showing that three additional subunits of Pol III (C34, AC40, and C82) copurify, together with C160, with Maf1p. This finding strongly suggests that Maf1p represses tRNA synthesis by a direct action on the Pol III enzyme.

Maf1p, the founding member of a new class of Pol III negative effectors. There is a remarkable conservation of the yeast and human Pol III machineries. The most conserved components are those involved in transcription complex assembly: subunit of TFIIIC, 131, all components of TFIIIB (TBP, Brf/TFIIIB70, and B"/TFIIIB90), and the tetrad of Pol III-specific subunits (C82, C34, C31, and C17) all have structural and functional homologs in human cells (2, 15, 24, 44, 50, 51). Since Maf1p is conserved across a wide range of organisms, it is probable that the orthologs of Maf1p are involved in Pol III regulation in higher cells. This would be very interesting in light of the connection between Pol III regulation and the malignancy process (29). The abundance of Pol III transcripts is abnormally elevated in many types of transformed and tumor cells (55). In mammals, two tumor suppressors, Rb (53) and p53 (7, 9), act as global repressors of Pol III transcription. Rb and p53 interact with and inactivate TFIIIB (9, 10, 28). Remarkably, Rb mutations occurring in tumors also release TFIIIB from repression (10). Deregulation of the control of Pol III and Pol I transcription is clearly an important step in tumor development. Therefore, it will be important to determine whether Maf1p orthologs really represent a new class of Pol III regulators in mammals.