|
|
Mol Cell Biol. 2006 July; 26(13): 4853–4862. doi: 10.1128/MCB.02367-05. | PMCID: PMC1489154 |
Copyright © 2006, American Society for Microbiology Fine-Structure Analysis of Ribosomal Protein Gene Transcription Yu Zhao,1 Kerri B. McIntosh,1 Dipayan Rudra,1 Stephan Schawalder,2† David Shore,2 and Jonathan R. Warner1* Department of Cell Biology, Albert Einstein College of Medicine, 1300 Morris Park Ave., Bronx, New York 10461,1 Department of Molecular Biology and NCCR Program “Frontiers in Genetics,” University of Geneva, Sciences III, 30, Quai Ernest-Ansermet, CH-1211 Geneva 4, Switzerland2 Received December 12, 2005; Revised January 30, 2006; Accepted April 17, 2006. The ribosomal protein genes of Saccharomyces cerevisiae, responsible for nearly 40% of the polymerase II transcription initiation events, are characterized by the constitutive tight binding of the transcription factor Rap1. Rap1 binds at many places in the yeast genome, including glycolytic enzyme genes, the silent MAT loci, and telomeres, its specificity arising from specific cofactors recruited at the appropriate genes. At the ribosomal protein genes two such cofactors have recently been identified as Fhl1 and Ifh1. We have now characterized the interaction of these factors at a bidirectional ribosomal protein promoter by replacing the Rap1 sites with LexA operator sites. LexA-Gal4(AD) drives active transcription at this modified promoter, although not always at the correct initiation site. Tethering Rap1 to the promoter neither drives transcription nor recruits Fhl1 or Ifh1, showing that Rap1 function requires direct DNA binding. Tethering Fhl1 also fails to activate transcription, even though it does recruit Ifh1, suggesting that Fhl1 does more than simply provide a platform for Ifh1. Tethering Ifh1 to the promoter leads to low-level transcription, at the correct initiation sites. Remarkably, activation by tethered LexA-Gal4(AD) is strongly reduced when TOR kinase is inhibited by rapamycin. Thus, TOR can act independently of Fhl1/Ifh1 at ribosomal protein promoters. We also show that, in our strain background, the response of ribosomal protein promoters to TOR inhibition is independent of the Ifh1-related protein Crf1, indicating that the role of this corepressor is strain specific. Fine-structure chromatin mapping of several ribosomal protein promoters revealed that histones are essentially absent from the Rap1 sites, while Fhl1 and Ifh1 are coincident with each other but distinct from Rap1. The 138 genes encoding the 79 ribosomal proteins (RP) of Saccharomyces cerevisiae are arguably the most coordinately regulated cluster of genes, spread throughout the yeast genome ( 7, 11). It was originally thought that the basis for much of this regulation lay in the presence of binding sites for the protein Rap1 upstream of a large majority of the RP genes ( 17, 21). However, Rap1 is a protein of many functions (reviewed by references 28 and 29). It is the primary transcription factor for the glycolytic genes and several translation factor genes. It acts as the major duplex DNA binding protein of telomeres. It nucleates the silencing of the HML and HMR mating-type loci. Genome-wide chromatin immunoprecipitation (ChIP) analysis revealed that Rap1 binds to about 5% of yeast genes and participates in the activation of 37% of RNA polymerase II transcripts in exponentially growing yeast cells ( 21). There is good evidence that the initial step of Rap1 is to clear nucleosomes from a patch of DNA ( 28, 40) and that the second step is to recruit specific factors to carry out the appropriate function. It is now clear that for the RP genes these factors are Fhl1 and Ifh1, which are found almost exclusively at RP genes. Gcr1 and Gcr2 are present at many glycolytic enzyme genes. Sir3, Sir4, and others are recruited for silencing at the silent MAT loci and telomeres ( 19, 21, 34). ChIP analysis of the RP genes showed that both Rap1 and Fhl1 are constitutively found at the promoters. Only occupancy by Ifh1 is correlated with active transcription, suggesting that Ifh1 plays a central role in the regulation of RP gene transcription ( 26, 33, 34, 39). Rap1 is one of the DNA binding proteins for which many consensus sequences have been suggested ( 29). Interestingly, the Rap1-binding sequences at RP gene promoters, termed RPG boxes, are quite different from those at the telomeres, while those at glycolytic gene promoters appear to be in between. Yet the basis for specificity remains obscure, although it has been suggested that Rap1 undergoes distinct conformational changes as a result of binding to somewhat different sequences ( 29). At the RP genes, it has been proposed that Rap1 recruits not only TAFs, which in turn recruit TATA binding protein to the RP genes that have characteristically poor TATA boxes ( 27), but also Esa1, which could acetylate either histone H4 or another participant in transcriptional activation ( 31). Yet, Esa1 probably provides little specificity since by ChIP analysis it is found upstream of many actively transcribed genes ( 30, 32). While genome-wide ChIP analysis revealed that Rap1, Fhl1, and Ifh1 are recruited to a majority of the RP gene promoters ( 19, 34, 39), neither the basis for the recruitment nor the role played by the factors in transcription of RP genes was clear. Furthermore, the binding sites for Fhl1 and for Ifh1 are elusive. When assayed in vitro, neither Fhl1 nor Ifh1 binds RP promoters, either by itself or in the presence of Rap1 ( 33). Ifh1 appears to be recruited to RP promoters through its interaction with the “forkhead-associated” (FHA) domain of Fhl1 ( 9, 26, 33, 34). However, the story must be more complex. Although the FHA domain of Fhl1 can recruit Ifh1 to serve as a transcriptional activator of a GAL-based artificial reporter, a nearly full length Fhl1 recruits nearly as much Ifh1, but very little transcription ensues ( 34). Indeed, Fhl1 has been proposed as a repressor of RP gene transcription ( 5, 14). Furthermore, there is no direct evidence that Ifh1 functions as a transcriptional activator in the context of an RP gene promoter. Utilizing a minimally engineered promoter that drives the transcription of two RP genes oriented head-to-head, we have found that for Rap1 to recruit Fhl1 and Ifh1 and to activate transcription, it must bind DNA directly. Furthermore, Rap1 binding to sites from glycolytic genes, within the context of the RP genes, recruits neither Fhl1 nor Ifh1. Recruitment of Ifh1 by Fhl1 tethered to the promoter is also insufficient to drive transcription, although tethered Ifh1 alone does. By high-resolution ChIP analysis, we find that Fhl1 and Ifh1 are recruited to the RP promoters at a location distinct from the Rap1-binding sites. Strains and plasmid constructs. The strains used in this study are listed in Table 1. The epitope tagging of the proteins of interest was done by PCR-based gene targeting ( 23). Plasmid pRS306-ConI contains the intergenic region of RPL24A and RPL30, which controls the expression of G418 r and green fluorescent protein (GFP), respectively (Fig. ) ( 43). The RAP1 binding sites were replaced with two restriction sites (BglII and NheI) by PCR (pRS306-ConI-BN). To delete the Rap1-binding sites in the pRS306-ConI vector, primers (5′-CACTAAAATCTGAGATCAAAAATATGTGagatctgctagcAAGGTCTTTTTCCAAGAAACGTATC-3′) and (5′-GATACGTTTCTTGGAAAAAGACCTTgctagcagatctCACATATTTTTGATCTCAGATTTTAGTG-3′ (lowercase letters are the sites for BglII and NheI) were used to generate PCR products using pRS306-ConI as template. The PCR product was incubated at 37°C with DpnI for 1 h, followed by ligation (in the presence of 1 μl polynucleotide kinase) and transformation to the XL1-Gold competent strain of Escherichia coli. The RAP1 sites were reinserted into the restriction sites by ligating an annealed oligonucleotide containing the original RAP1 binding sites with pRS306-ConI-BN digested with BglII and NheI. Similarly, eight copies of LexA operators were amplified by PCR using pSH18-34 (Invitrogen) as the template and subcloned into the plasmid pRS306-ConI-BN. Full-length open reading frames (ORFs) (starting at residue 2) of RAP1, FHL1, and IFH1 were amplified by PCR using yeast genomic DNA as templates and subcloned in frame with the LexA DNA binding domain in the 2μm vector pBTM116 ( 1), generating pBTM116-RAP1, pBTM116-FHL1, and pBTM116-IFH1, respectively, which were verified by sequencing. The plasmid pLexA-pos contains LexA-Gal4(AD) under the control of the ADH1 promoter (Clontech). Northern and reverse transcription-PCR mRNA analysis. For Northern analysis, preparation of yeast total RNA and blotting was performed as described previously ( 43). The oligonucleotide JW61L was used to identify GFP and RPL30 mRNAs, and JW2258 was used to identify G418 r and RPL24A mRNAs (Table 2). For analysis by reverse transcription/real-time PCR, total RNA from 2 × 10 7 cells was isolated using the RNeasy minikit (QIAGEN). An 0.4-μg amount of DNase-treated total RNA was reverse transcribed. Quantitation of cDNAs was achieved using Taqman real-time PCR on an Applied Biosystems ABI Prism 7700 machine. Sequences of primer pairs and Taqman probes are available upon request. | TABLE 2.Oligonucleotide primers for Northern blotting and PCR |
ChIP. ChIP was carried out as described previously ( 16) with slight modifications. The antibodies used were anti-LexA rabbit polyclonal antibody (Santa Cruz), anti c-Myc mouse monoclonal antibody (9E10; Santa Cruz), antihemagglutinin (anti-HA) mouse monoclonal antibody (12CA5, hybridoma supernatant), anti-histone H3 C-terminal region antibody (ab1791; Abcam), and anti-Rap1 rabbit polyclonal antibody ( 34). Briefly, formaldehyde was added to 200 ml of culture (with an A600 of ~0.6 to 1.0) at a final concentration of 1% and incubated at room temperature for 20 min with occasional swirling and then 30 ml of 3 M glycine was added for 5 min. Chromatin was prepared as described previously ( 16). For the high-resolution ChIP (Fig. and ) the sonication time was extended, using 16 15-s bursts separated by 1 min of cooling on ice, so that the average size of sheared chromatin was ~200 bp. Extracts containing sheared chromatin were precleared with 20 μl UltraLink protein A/G beads (Pierce) for 3 hours at 4°C before being treated at 4°C overnight with specific antibody prebound to 20 μl protein A/G beads. Quantitative analyses were performed on an ABI PRISM 7900HT real-time PCR system (Applied Biosystems). To calculate the enrichment ( n-fold) of LexA (or LexA fusion protein), Rap1, Fhl1-HA 3, and Ifh1-Myc 13 occupancy at an individual promoter, the 2 −ΔΔCT method ( 22) was applied, using a control promoter of ACT1 that has been shown not to be occupied by Rap1, Fhl1p, or Ifh1p ( 21, 33). The oligonucleotide primers used in real-time PCR are listed in Table 2; those used for high-resolution analysis (Fig. and ) are available upon request. A reporter construct to study ribosomal protein gene transcription. Because the function of Rap1 depends significantly on its context, analysis of the way in which it influences Fhl1 and Ifh1 requires a test gene that differs minimally from the wild-type (wt) sequence. Yet ribosome synthesis is essential to cell viability; manipulation of endogenous RP gene promoters or their transcription factors leads to detrimental effects on growth. To avoid these problems, we generated a reporter construct that maintains as intact as possible the structure of the RP gene promoter elements including the transcription start site ( 43). RPL24A and RPL30 are divergently transcribed from an intergenic region of 609 bp that contains two Rap1-binding sites (Fig. ) ( 6). The ORFs of RPL24A and RPL30 were replaced with those of G418 r and GFP, respectively. Since the sequences of the transcription start sites and introns of RPL24A and RPL30 remained intact, the novel sequences are distant by 490 and 296 bp, respectively, from the sites of initiation of transcription. The construct was integrated as a single copy at the URA3 locus. The reporter is regulated in the same way as an RP gene ( 20, 43; data not shown) and yet is dispensable to the cell. The two Rap1-binding sites were replaced with two restriction sites by PCR mutagenesis (see Materials and Methods) (Fig. ). Either the original Rap1-binding sites or eight copies of the LexA operator ( 10) were inserted. Thus, the reporter has maintained almost all cis-acting elements of an RP promoter except the known Rap1-binding sites. LexA fusion proteins remain functional. Full-length ORFs of RAP1, FHL1, and IFH1 were subcloned in frame with the LexA DNA binding domain in the 2μm vector pBTM116 ( 1) and, after verification of sequence fidelity, were introduced into strains carrying the test gene with the LexA binding sites described above (Fig. ). Western blotting showed that the fusion proteins were expressed, although at substantially different levels (Fig. ). To determine whether the LexA fusion proteins were functional, we transformed the plasmids expressing LexA-Rap1 or LexA-Ifh1 into strains in which RAP1 or IFH1 was under GAL control. In both cases, the LexA fusion proteins supported growth in dextrose (Fig. ). Similarly, LexA-Fhl1 was able to rescue the slow-growth phenotype of an fhl1Δ strain (Fig. , bottom). Therefore, fusion of the LexA DNA binding domain at the N terminus of Rap1, Fhl1, and Ifh1 does not interfere with their function. LexA-Rap1 fails to activate. Total RNA was prepared from the strains with integrated constructs (Fig. ) and analyzed by Northern blotting ( 43). A single oligonucleotide detects transcripts of both GFP and RPL30 (Fig. ; see Materials and Methods), and another detects transcripts of both G418 r and RPL24A (Fig. ). Deletion of the Rap1-binding sites of the RPL30/RPL24A promoter leads to about 75% reduction in transcription (Fig. , lane 2). The residual transcription is likely due to the T-rich elements found between the Rap1 sites and the initiation site in most RP genes, which contribute to RP promoter function ( 12). Interestingly, the residual transcription in the RPL24A direction starts about 200 bp upstream of the normal initiation site. Reinsertion of Rap1-binding sites restores transcription almost to its original level, indicating that the restriction sites interfere neither with the Rap1-binding sites nor with Rap1 function (Fig. , lane 3). Strain YZ172, with the integrated construct containing the LexA hybrid promoter (Fig. ), was transformed with plasmids expressing the LexA DNA binding domain alone (Fig. , lane 4) or the LexA fusion proteins, LexA-Gal4(AD), LexA-Rap1, LexA-Fhl1, and LexA-Ifh1 (lanes 5 to 8). As expected, in cells carrying the LexA hybrid promoter, expression of the LexA DNA binding domain alone has no effect on transcription (Fig. , lanes 4). On the other hand, expression of LexA-Gal4(AD) restores transcription in both directions (Fig. , lanes 5), although the site of initiation in the RPL24A direction is still mostly anomalous (Fig. , lane 5). Nor does this transcript lead to G418 resistance (Fig. ), presumably because of aberrant splicing or aberrant translation initiation. These might lead to premature decay of the RNA, which would account for the apparent difference of LexA-Gal4(AD) activation in the two directions. On the other hand there is no reason to presume that Gal4(AD) will stimulate the same level of transcription in the two directions. LexA-Rap1 fails to activate transcription (Fig. , lane 6). Comparison of lanes 3 and 6 suggests that transcriptional activation of an RP gene requires that Rap1 bind DNA through its own DNA binding domain. Rap1 is known both to bend DNA ( 24) and to exclude nucleosomes ( 40). Presumably, one or both of these consequences of the binding of Rap1 are essential to recruit Fhl1 and Ifh1. The presence of the Rap1 polypeptide chain itself is insufficient. LexA-Fhl1 also fails to activate transcription (Fig. , lane 7). Thus, at the promoter of an RP gene, the presence of Fhl1, available to bind Ifh1, is not sufficient to drive transcription. This seems contradictory to previous work in which binding the FHA domain to DNA via the Gal4 binding domain led to recruitment of Ifh1 and activation of transcription ( 26, 33, 34). However, with such constructs no such activation occurs if the C-terminal region of Fhl1 is also present ( 34), as is the case in our experiment. Clearly, the large Fhl1 protein plays a more complex role in RP gene transcription than simply providing an FHA domain with which Ifh1 can interact. Finally, LexA-Ifh1 alone does activate transcription (Fig. , lane 8), although to a lesser extent than found in the wt gene. Furthermore, it restores transcriptional initiation in the RPL24A direction to the proper place, as confirmed by the resistance of the cells to G418 (Fig. ). This result suggests that Ifh1 is a transcriptional (co)activator for RP genes and that it is at least partly responsible for selection of the proper initiation site. The basis for the selection of the initiation site (often sites) in yeast is obscure. RP genes are said not to use a  canonical TATA box ( 16), and the region upstream is very AT rich. Transcription under the control of the LexA hybrid promoter is repressed by rapamycin. Loss of Ifh1 from an RP gene has been implicated as the proximal cause for the repression of transcription by rapamycin ( 26, 33, 34, 39). In addition, Martin et al. showed that Crf1, a protein resembling Ifh1, binds to Fhl1 at RP promoters following rapamycin treatment and is required for their repression ( 26). Remarkably, treatment of cells with rapamycin represses not only RPL30 but also the GFP transcript activated either by LexA-Ifh1 or by LexA-Gal4(AD) (Fig. ). Note that rapamycin has no effect on normal transcription driven by Gal4 (Fig. ). This result is consistent with our previous finding that rapamycin inhibits the very limited transcription of RP genes in cells lacking Fhl1 and Ifh1 ( 33). These observations suggest that the TOR pathway is required not only to promote the interaction of Ifh1 with Fhl1 but also to facilitate the residual transcription that occurs independently of the Rap1-Fhl1-Ifh1 transcription complex. To ask whether the Crf1 repressor might be involved in the rapamycin effect described above, we first examined the consequence of CRF1 deletion in an otherwise wild-type background. Surprisingly, we found that crf1Δ has no measurable effect on the repression at the three different RP genes tested (RPL9A, RPL30, and RPL37A) in our W303 strain background (Fig. ). We confirmed the role of Crf1 in repression of the same genes following rapamycin treatment in strain TB50 (Fig. ). Consistent with the absence of a role for Crf1 at RP gene promoters in W303, we have consistently failed to detect the protein at RP genes by ChIP (data not shown). We thus conclude that the effect of rapamycin on transcription driven by LexA-Gal4(AD) is unlikely to require Crf1. We have crossed W303 and TB50 in an attempt to characterize the genetic basis of the difference in Crf1 function in these two strains. However, the heterozygous diploid displayed an intermediate phenotype by transcript analysis, and the behavior of haploid segregants suggested that the difference is not due to a single gene. Rap1-RPG box interaction is necessary to recruit Fhl1 and Ifh1. To determine whether the binding of Rap1, Fhl1, and Ifh1 to the modified promoters of RPL24A/RPL30 was altered, we performed ChIP analysis on strains carrying Fhl1 and Ifh1 C-terminally tagged with HA 3 and Myc 13, respectively (Table 1; Fig. ). As controls, we also tested their binding to the ACT1, PGK1, and RPL11A genes. None of the three proteins are found at the ACT1 gene ( 19, 33), which was used as a negative control. As shown in Fig. , Rap1 is present at PGK1 but Fhl1 and Ifh1 are not. This is expected since the glycolytic genes are activated by Gcr1 in collaboration with Rap1 ( 8), rather than by Fhl1 and Ifh1 ( 19, 34). At the promoters of both RPL11A and RPL24A/RPL30 all three proteins are present, as expected from previous work ( 19, 33, 34). Deletion of the Rap1 sites abolishes binding not only of Rap1 but also of Fhl1 and Ifh1, indicating that Rap1 binding is essential for the recruitment of both. Reinsertion of the Rap1 sites restores binding of all three, but to a somewhat lesser extent. We cannot tell if this is due to the introduction of the restriction sites or is an artifact due to the specific oligonucleotide pair needed to distinguish the test gene from the wt gene. On the other hand, replacing the Rap1-binding sites of RPL24A/L30 with that of PYK1, a glycolytic enzyme gene, leads to strong recruitment of Rap1 but not of either Fhl1 or of Ifh1 (Fig. ). These results show once again that context is critical to the selection of the cofactors of Rap1 at individual promoters. Consequences of the binding of a LexA fusion protein. We next asked whether the failure of LexA-Rap1 to activate transcription from the LexA binding sites (Fig. , lanes 6) is due to its inability to recruit Fhl1 and/or Ifh1. In the presence of LexA alone none of the three proteins were found (Fig. ). In the presence of LexA-GAL4(AD), again Rap1, Fhl1, and Ifh1 were not present. LexA-Gal4(AD) was present, as predicted by the vigorous transcription that it activated (Fig. , lanes 5). In cells expressing both Rap1 and LexA-Rap1, ChIP using anti-LexA antibody, recognizing only the fusion protein, or anti-Rap1 antibody, recognizing both, shows the presence of LexA-Rap1 at the LexA-op site (Fig. ). (Note that our anti-Rap1 antibody poorly recognizes the LexA-Rap1 fusion protein [data not shown].) Yet, almost no Fhl1 or Ifh1 is associated with the LexA-Rap1 bound at the LexA-op sites. This result implies that it is not the presence of the Rap1 protein but the binding of Rap1 to its sites that somehow modifies the DNA, or the chromatin, to induce the association of Fhl1 and Ifh1 As noted above (Fig. ), even binding of Rap1 through a site from the PYK1 promoter is not sufficient for Fhl1 and Ifh1 recruitment. LexA-Fhl1 binds to the LexA-op sites and appears to recruit Ifh1, but not Rap1 (Fig. ). This result is consistent with the observation that Fhl1 is able to coimmunoprecipitate Ifh1 (reference 33 and data not shown). Nevertheless, the recruitment of Ifh1 by LexA-Fhl1 does not activate transcription (Fig. , lane 7), suggesting that a particular geometry of the DNA-Fhl1-Ifh1 complex is required. Although this could involve the binding of Fhl1 to the DNA, we have found that the putative DNA binding “forkhead” domain of Fhl1 is of little importance for cell growth ( 32). An intriguing observation is that the endogenous Fhl1-HA 3 is also recruited to the LexA-op sites. This result suggests that Fhl1 forms a dimer in vivo, although we have been unable to confirm this with coimmunoprecipitation experiments on extracts (data not shown). The interaction may be transient, observable only in extracts of formaldehyde-fixed cells. LexA-Ifh1 binds to the LexA-op site (Fig. ) and in this case brings about substantial transcription (Fig. , lanes 8). This result suggests that Ifh1 can act as a transcriptional activator, and conversely, that neither Rap1 nor Fhl1 itself is a transcriptional activator of RP genes. However, comparison of the levels of LexA-Gal4(AD) and LexA-Ifh1 at the promoter with the levels of transcription by each suggests that Ifh1 by itself is a rather weak activator. Again, the endogenous Ifh1-Myc13 is also present at the LexA-op promoter, suggesting possible dimerization. Rap1 and Fhl1/Ifh1 are at distinct locations in RP gene promoters. Previously, mapping of Fhl1 and Rap1-binding sites across three RP promoters, RPL12A, RPS11B, and RPL40A, suggested that there are promoter-specific differences in the relative locations of these proteins ( 39). Fhl1 was found either at the same location as Rap1 or between Rap1 and the transcription start site. To determine more precisely the locations of Rap1, Fhl1, and Ifh1 at RP promoters, we performed high-resolution ChIP on three intergenic regions at which RP genes were head to head either with another RP gene ( RPL24A/L30 and RPS22A/L39) or with a non-RP gene ( PRE2/RPL11A), using primer pairs covering every 100 bp of the intergenic region (Fig. ). In this case, the chromatin was sheared to a mean size of ~200 bp. The peak of Rap1 binding is sharply defined, directly over the predicted Rap1-binding sites ( 17). However, Fh1l and Ifh1 are distinct from Rap1 at these RP gene promoters. The location of Ifh1 coincides with that of Fhl1, consistent with the notion that Ifh1 is recruited to the RP promoter through its interaction with Fhl1 ( 33, 34, 39). In both cases of head-to-head RP genes, while the Rap1 sites are asymmetrical with respect to transcription initiation, Fhl1 and Ifh1 are located in a single peak approximately equidistant from the initiation sites. This result suggests that there is no directionality to their function. Yet, in the case of the RPL11A/PRE2 pair, the transcription of PRE2 is regulated entirely differently from that of RPL11A ( 11). The key conclusion from Fig. is that Ifh1 and Fhl1 are found at the same site on RP promoters, rather distant from the site of transcription initiation, and clearly 100 to 200 bp distinct from the site of Rap1 binding. In particular, Fhl1 and Ifh1 need not necessarily lie between Rap1 and the initiation site. Employing the same chromatin preparations, we asked about the presence of nucleosomes at the RP promoters, using antibody directed against the unmodified C terminus of histone H3 (Fig. ). It is evident that the Rap1 sites are almost completely clear of nucleosomes, as suggested by the genome-wide data recently published ( 3, 30, 41) and the observation that RP gene promoters are particularly deficient in Htz1, an H2A variant implicated in regulation ( 42). Based on recent work, a simple model for the transcription of RP genes is that Rap1 recruits Fhl1, which in turn recruits the transcriptional activator Ifh1. Regulation of transcription involves the control of the Fhl1-Ifh1 interaction ( 26, 33, 34, 39). Using a minimally modified promoter driving the transcription of two RP genes in head-to-head configuration, we set out to examine in more detail the role of Rap1 in active transcription of the RP genes and to evaluate the degree to which the simplest form of the model can account for the observations. Indeed, deletion of Rap1 sites from the promoter of a pair of RP genes (Fig. ) leads to a major loss of transcription (Fig. ) as well as to the loss of Fhl1 and Ifh1 binding (Fig. ). Reintroduction of the Rap1 sites restores transcription and the binding of all three factors. However, replacing the native Rap1 sites with the Rap1 site from the glycolytic gene PYK1 restores the binding of neither Fhl1 nor Ifh1 (Fig. ), although some transcription ensues, in only one direction, possibly through the recruitment of Gcr1 (data not shown). This result mirrors the published genome-wide analyses that found neither Fhl1 ( 19) nor Ifh1 ( 34) at glycolytic genes, where Gcr1 and Gcr2, rather than Fhl1 and Ifh1, act as coactivators (also Fig. , data for PGK1) ( 8, 37). Since Rap1 operates at so many genes, in both positive and negative ways (see the introduction), the sequence elements that designate which coactivators or corepressors associate with Rap1 at a particular locus are of key importance but have yet to be clearly delineated. Although some attempts have been made in silico, by examining the relationship between sequence and coregulation ( 2, 39), the proposed sequences are not found at every RP gene. While there are close matches to the IFHL sequence element suggested by Wade et al. ( 39) within both the RPL24A-RPL30 and the RPS22A-RPL39 promoters, right at the peak of Fhl1/Ifh1 binding, there is none upstream of the RPL11A gene. Thus, while the IFHL motif is promising, it seems unlikely to be key to the recruitment of Fhl1 and Ifh1. A useful insight comes from our observation that for the recruitment of Fhl1 and Ifh1 to a RP promoter, Rap1 must bind directly to its sites on the DNA (Fig. ). The presence of the Rap1 polypeptide chain, held at the promoter by LexA, is insufficient. A primary result of the direct binding of Rap1 to DNA is undoubtedly to clear nucleosomes from the region ( 28, 40). Indeed, nucleosomes are almost entirely absent from the Rap1 sites in front of RP genes (Fig. ) ( 4, 18, 30). Nevertheless, we remain ignorant about the specificity with which Rap1 selects the appropriate cofactors at any given class of genes. The importance of Rap1 in bringing about an arrangement of chromatin hospitable to Fhl1 and Ifh1 is demonstrated by the observation that deletion of the Rap1-binding sites leads to loss of the two proteins (Fig. ), even though the region of the promoter with which they are associated, some 100 to 200 bp distant, remains intact (Fig. ). Although Fhl1 has a domain that is related to the DNA binding domain of the Drosophila melanogaster Forkhead protein, Fhl1 does not bind DNA in vitro, nor is the Forkhead domain important for RP gene transcription or cell growth ( 33). Thus, the binding of Fhl1, and by implication Ifh1, to a normal RP promoter is likely to be indirect, either secondary to or facilitated by another protein. Is this simply Rap1 itself? If so, what determines the specificity? Two observations raise questions about the arrangement of factors necessary to drive transcription at the RP genes. One is that there seems to be a single site on the DNA for the association of Ifh1 and Fhl1, even when they are driving two RP genes (Fig. ). Previous work identified Fhl1 as coincident with Rap1 or slightly proximal to the transcription initiation site ( 39), as we see for RPL11A (Fig. ). However, in the cases where two RP genes are driven from the same promoter, Fhl1 and Ifh1 are located in a single, albeit relatively broad, site asymmetrical with respect to the Rap1 sites, yet approximately equidistant from the two initiation sites. Transcriptome data suggest that both genes are transcribed to roughly the same extent ( 15). Thus, Ifh1 appears to be able to drive transcription from a position distal to Rap1. How can it overcome the barrier that the two Rap1 molecules seem to impose on chromatin structure? The other observation is that although LexA-Fhl1 recruits Ifh1 to the promoter, no transcription ensues. There are two possible explanations. One is that the geometry of the site is incorrect in this configuration, i.e., that Fhl1 must hold Ifh1 in just the right orientation and does not do so as a LexA fusion. However, this seems unlikely, as the presumably flexible LexA-Ifh1 is able to drive transcription in both directions. Another possibility is that the eight LexA sites led to congestion on the promoter, such that the activation region of Ifh1 was unable to address the transcription site. However, when the experiments of Fig. were repeated using only two LexA sites, essentially identical results were obtained, except that the levels of transcripts of both L24A-G418 r and L30-GFP were substantially reduced. Both the aberrant transcription initiation by LexA-Gal4(AD) and the correct one by LexA-Ifh1 were reproduced. An alternative possibility is that, although the FHA domain of Fhl1 can recruit Ifh1 to promote transcription of a test gene ( 26, 34), Fhl1 in its intact form acts as a repressor, as was originally suggested based on the observation that deletion of FHL1 restored the viability (although barely) to a strain without IFH1 ( 5). Thus, we must consider Fhl1 as more than a platform to present an FHA domain with which Ifh1 can interact. Other parts of Fhl1 may affect the rate of transcription of the RP gene or may transduce other regulatory signals. A final point that casts some doubt on the simple model described above is that rapamycin inhibits the transcription driven by LexA-Ifh1 as well as by LexA-Gal4(AD). Thus, the scenario in which the mechanism of action of rapamycin is either to modify Ifh1 ( 33, 34) or to cause Crf1 to move from the cytoplasm to the nucleus to replace Ifh1 ( 26) seems inadequate. Rather, our data suggest that although rapamycin does act to reduce the interaction of Ifh1 with Fhl1, it must also work in some way independently of Fhl1 and Ifh1, as suggested previously for cells devoid of both ( 33). Further underscoring the complexity of the rapamycin response, as well as its variability, we show that the Crf1 repressor has no apparent role in our strain background. Thus, Ifh1 leaves RP gene promoters in the absence of competition by Crf1 in W303 (strain YSS120, Table 1), indicating that the rapamycin response at the level of Fhl1 and Ifh1 displays remarkable strain variability. An understanding of the genetic basis of this difference might reveal interesting features of the TOR pathway as it impinges upon RP genes. Another factor implicated in the regulation of RP gene transcription is Sfp1 ( 25). Although ChIP analysis showed little binding of Sfp1 to the L24A-L30 intergenic region ( 19, 25), bioinformatics approaches suggested that both RPL24A and RPL30 could be down-regulated by the transfer of Sfp1 from the nucleus to the cytoplasm caused by the presence of rapamycin and other repressive conditions ( 25). Therefore, it seems at least possible, if not likely, that the transcription caused by LexA-Gal4 and LexA-Ifh1 was repressed by rapamycin through its effect on Sfp1. Based on the results in Fig. and on our results previously reported that rapamycin represses the residual transcription of RP genes in cells without Fhl1 ( 33), we suggest that the action of Sfp1 on RP gene transcription is independent of the Rap1-Fhl1-Ifh1 transcriptional complex and may indeed involve other factors or other sequences than the Rap1-binding sites. In summary, we have shown that direct binding of Rap1 to the promoter of RP genes excludes nucleosomes and recruits Fhl1 and Ifh1 to drive transcription, consistent with the simple model presented above. However, a number of observations suggest that the simple model is insufficient. Recruitment of Ifh1 by LexA-Fhl1 does not drive transcription, suggesting either that the geometry of the arrangement is important or that Fhl1 can act as a repressor. The binding sites of Fhl1 and Ifh1 are coincident but at some distance from the Rap1 sites, which themselves are asymmetrically arranged in the two head-to-head pairs of RP genes that we have examined. Rapamycin inhibits transcription of RP genes that are driven by the Gal4(AD), suggesting that something beyond the interaction of Fhl1 and Ifh1 is responsive to the TOR pathway. We dedicate this paper to the memory of Stephan Schawalder, a most promising young scientist. We are grateful to Rodolfo Negri for discussions and access to unpublished data and to Dietmar Martin and Mike Hall for providing strains and advice. This research was supported in part by grants from the Human Frontiers Program (to both D.S. and J.R.W.), by the NIH (GM-25532 to J.R.W. and CAI-3330 to the Albert Einstein Cancer Center), by the Swiss National Fund (through the ESF euroDYNA program to D.S.), and by funds provided by the Canton of Geneva (to D.S.). 1. Bartel, P. L., C. Chien, R. Sternglanz, and S. Fields. 1993. Using the two-hybrid system to detect protein-protein interactions, p. 153-179. In D. A. Hartley (ed.), Cellular interactions in development: a practical approach. IRL, Oxford, United Kingdom. 2. Beer, M. A., and S. Tavazoie. 2004. Predicting gene expression from sequence. Cell 117:185-198. [PubMed] 3. Bernstein, B. E., C. L. Liu, E. L. Humphrey, E. O. Perlstein, and S. L. Schreiber. 2004. Global nucleosome occupancy in yeast. Genome Biol. 5:R62. [PubMed] 4. Boeger, H., J. Griesenbeck, J. S. Strattan, and R. D. Kornberg. 2003. Nucleosomes unfold completely at a transcriptionally active promoter. Mol. Cell 11:1587-1598. [PubMed] 5. Cherel, I., and P. Thuriaux. 1995. The IFH1 gene product interacts with a fork head protein in Saccharomyces cerevisiae. Yeast 11:261-270. [PubMed] 6. Dabeva, M. D., and J. R. Warner. 1987. The yeast ribosomal protein L32 and its gene. J. Biol. Chem. 262:16055-16059. [PubMed] 7. DeRisi, J. L., V. R. Iyer, and P. O. Brown. 1997. Exploring the metabolic and genetic control of gene expression on a genomic scale. Science 278:680-686. [PubMed] 8. Drazinic, C. M., J. B. Smerage, M. C. Lopez, and H. V. Baker. 1996. Activation mechanism of the multifunctional transcription factor repressor-activator protein 1 (Rap1p). Mol. Cell. Biol. 16:3187-3196. [PubMed] 9. Durocher, D., and S. P. Jackson. 2002. The FHA domain. FEBS Lett. 513:58-66. [PubMed] 10. Estojak, J., R. Brent, and E. A. Golemis. 1995. Correlation of two-hybrid affinity data with in vitro measurements. Mol. Cell. Biol. 15:5820-5829. [PubMed] 11. Gasch, A. P., P. T. Spellman, C. M. Kao, O. Carmel-Harel, M. B. Eisen, G. Storz, D. Botstein, and P. O. Brown. 2000. Genomic expression programs in the response of yeast cells to environmental changes. Mol. Biol. Cell 11:4241-4257. [PubMed] 12. Goncalves, P. M., G. Griffioen, R. Minnee, M. Bosma, L. S. Kraakman, W. H. Mager, and R. J. Planta. 1995. Transcription activation of yeast ribosomal protein genes requires additional elements apart from binding sites for Abf1p or Rap1p. Nucleic Acids Res. 23:1475-1480. [PubMed] 13. Goncalves, P. M., K. Maurer, G. van Nieuw Amerongen, K. Bergkamp-Steffens, W. H. Mager, and R. J. Planta. 1996. C-terminal domains of general regulatory factors Abf1p and Rap1p in Saccharomyces cerevisiae display functional similarity. Mol. Microbiol. 19:535-543. [PubMed] 14. Hermann-Le Denmat, S., M. Werner, A. Sentenac, and P. Thuriaux. 1994. Suppression of yeast RNA polymerase III mutations by FHL1, a gene coding for a fork head protein involved in rRNA processing. Mol. Cell. Biol. 14:2905-2913. [PubMed] 15. Holstege, F. C., E. G. Jennings, J. J. Wyrick, T. I. Lee, C. J. Hengartner, M. R. Green, T. R. Golub, E. S. Lander, and R. A. Young. 1998. Dissecting the regulatory circuitry of a eukaryotic genome. Cell 95:717-728. [PubMed] 16. Kuras, L., and K. Struhl. 1999. Binding of TBP to promoters in vivo is stimulated by activators and requires Pol II holoenzyme. Nature 399:609-613. [PubMed] 17. Lascaris, R. F., W. H. Mager, and R. J. Planta. 1999. DNA-binding requirements of the yeast protein Rap1p as selected in silico from ribosomal protein gene promoter sequences. Bioinformatics 15:267-277. [PubMed] 18. Lee, C. K., Y. Shibata, B. Rao, B. D. Strahl, and J. D. Lieb. 2004. Evidence for nucleosome depletion at active regulatory regions genome-wide. Nat. Genet. 36:900-905. [PubMed] 19. Lee, T. I., N. J. Rinaldi, F. Robert, D. T. Odom, Z. Bar-Joseph, G. K. Gerber, N. M. Hannett, C. T. Harbison, C. M. Thompson, I. Simon, J. Zeitlinger, E. G. Jennings, H. L. Murray, D. B. Gordon, B. Ren, J. J. Wyrick, J. B. Tagne, T. L. Volkert, E. Fraenkel, D. K. Gifford, and R. A. Young. 2002. Transcriptional regulatory networks in Saccharomyces cerevisiae. Science 298:799-804. [PubMed] 20. Li, B., C. R. Nierras, and J. R. Warner. 1999. Transcriptional elements involved in the repression of ribosomal protein synthesis. Mol. Cell. Biol. 19:5393-5404. [PubMed] 21. Lieb, J. D., X. Liu, D. Botstein, and P. O. Brown. 2001. Promoter-specific binding of Rap1 revealed by genome-wide maps of protein-DNA association. Nat. Genet. 28:327-334. [PubMed] 22. Livak, K. J., and T. D. Schmittgen. 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(−ΔΔC(T)) method. Methods 25:402-408. [PubMed] 23. Longtine, M. S., A. McKenzie III, D. J. Demarini, N. G. Shah, A. Wach, A. Brachat, P. Philippsen, and J. R. Pringle. 1998. Additional modules for versatile and economical PCR-based gene deletion and modification in Saccharomyces cerevisiae. Yeast 14:953-961. [PubMed] 24. Lopez, M. C., J. B. Smerage, and H. V. Baker. 1998. Multiple domains of repressor activator protein 1 contribute to facilitated binding of glycolysis regulatory protein 1. Proc. Natl. Acad. Sci. USA 95:14112-14117. [PubMed] 25. Marion, R. M., A. Regev, E. Segal, Y. Barash, D. Koller, N. Friedman, and E. K. O'Shea. 2004. Sfp1 is a stress- and nutrient-sensitive regulator of ribosomal protein gene expression. Proc. Natl. Acad. Sci. USA 101:14315-14322. [PubMed] 26. Martin, D. E., A. Soulard, and M. N. Hall. 2004. TOR regulates ribosomal protein gene expression via PKA and the Forkhead transcription factor FHL1. Cell 119:969-979. [PubMed] 27. Mencia, M., Z. Moqtaderi, J. V. Geisberg, L. Kuras, and K. Struhl. 2002. Activator-specific recruitment of TFIID and regulation of ribosomal protein genes in yeast. Mol. Cell 9:823-833. [PubMed] 28. Morse, R. H. 2000. RAP, RAP, open up! New wrinkles for RAP1 in yeast. Trends Genet. 16:51-53. [PubMed] 29. Pina, B., J. Fernandez-Larrea, N. Garcia-Reyero, and F. Z. Idrissi. 2003. The different (sur)faces of Rap1p. Mol. Genet. Genomics 268:791-798. [PubMed] 30. Pokholok, D. K., C. T. Harbison, S. Levine, M. Cole, N. M. Hannett, T. I. Lee, G. W. Bell, K. Walker, P. A. Rolfe, E. Herbolsheimer, J. Zeitlinger, F. Lewitter, D. K. Gifford, and R. A. Young. 2005. Genome-wide map of nucleosome acetylation and methylation in yeast. Cell 122:517-527. [PubMed] 31. Reid, J. L., V. R. Iyer, P. O. Brown, and K. Struhl. 2000. Coordinate regulation of yeast ribosomal protein genes is associated with targeted recruitment of Esa1 histone acetylase. Mol. Cell 6:1297-1307. [PubMed] 32. Robert, F., D. K. Pokholok, N. M. Hannett, N. J. Rinaldi, M. Chandy, A. Rolfe, J. L. Workman, D. K. Gifford, and R. A. Young. 2004. Global position and recruitment of HATs and HDACs in the yeast genome. Mol. Cell 16:199-209. [PubMed] 33. Rudra, D., Y. Zhao, and J. R. Warner. 2005. Central role of Ifh1p-Fhl1p interaction in the synthesis of yeast ribosomal proteins. EMBO J. 24:533-542. [PubMed] 34. Schawalder, S. B., M. Kabani, I. Howald, U. Choudhury, M. Werner, and D. Shore. 2004. Growth-regulated recruitment of the essential yeast ribosomal protein gene activator Ifh1. Nature 432:1058-1061. [PubMed] 35. Sikorski, R. S., and P. Hieter. 1989. A system of shuttle vectors and yeast host strains designed for efficient manipulation of DNA in Saccharomyces cerevisiae. Genetics 122:19-27. [PubMed] 36. Thomas, B. J., and R. Rothstein. 1989. Elevated recombination rates in transcriptionally active DNA. Cell 56:619-630. [PubMed] 37. Uemura, H., M. Koshio, Y. Inoue, M. C. Lopez, and H. V. Baker. 1997. The role of Gcr1p in the transcriptional activation of glycolytic genes in yeast Saccharomyces cerevisiae. Genetics 147:521-532. [PubMed] 38. Vilardell, J., and J. R. Warner. 1994. Regulation of splicing at an intermediate step in the formation of the spliceosome. Genes Dev. 8:211-220. [PubMed] 39. Wade, J. T., D. B. Hall, and K. Struhl. 2004. The transcription factor Ifh1 is a key regulator of yeast ribosomal protein genes. Nature 432:1054-1058. [PubMed] 40. Yu, L., and R. H. Morse. 1999. Chromatin opening and transactivator potentiation by RAP1 in Saccharomyces cerevisiae. Mol. Cell. Biol. 19:5279-5288. [PubMed] 41. Yuan, G. C., Y. J. Liu, M. F. Dion, M. D. Slack, L. F. Wu, S. J. Altschuler, and O. J. Rando. 2005. Genome-scale identification of nucleosome positions in S. cerevisiae. Science 309:626-630. [PubMed] 42. Zhang, H., D. N. Roberts, and B. R. Cairns. 2005. Genome-wide dynamics of Htz1, a histone H2A variant that poises repressed/basal promoters for activation through histone loss. Cell 123:219-231. [PubMed] 43. Zhao, Y., J. H. Sohn, and J. R. Warner. 2003. Autoregulation in the biosynthesis of ribosomes. Mol. Cell. Biol. 23:699-707. [PubMed]
|
This article has been 
