|Home | About | Journals | Submit | Contact Us | Français|
Messenger RNA processing is coupled to RNA Polymerase II (RNAPII) transcription through coordinated recruitment of accessory proteins to the Rpb1 C-terminal domain (CTD). Dynamic changes in CTD phosphorylation during transcription elongation are responsible for their recruitment, with serine 5 phosphorylation (S5-P) occurring towards the 5’ end of genes and serine 2 phosphorylation (S2-P) occurring towards the 3’ end. The proteins responsible for regulation of the transition state between S5-P and S2-P CTD remain elusive. We show that a conserved protein of unknown function, Rtr1, localizes within coding regions, with maximum levels of enrichment occurring between the peaks of S5-P and S2-P RNAPII. Upon deletion of Rtr1, the S5-P form of RNAPII accumulates in both whole cell extracts and throughout coding regions; additionally, RNAPII transcription is decreased and termination defects are observed. Functional characterization of Rtr1 reveals its role as a CTD phosphatase essential for the S5- to S2- P transition.
From yeast to mammals, there are three highly conserved RNA Polymerase complexes that are responsible for the transcription of all classes of cellular RNAs. RNA processing is closely tied to transcription in order to ensure the fate of nascent RNA. One unique mechanism for proper RNA processing involves the recruitment of a wide variety of accessory proteins to the C-terminal domain (CTD) of the largest subunit of RNAPII, Rpb1 (for review, see Phatnani and Greenleaf, 2006). The CTD consists of 27 repeats of the sequence Y1S2P3T4S5P6S7 in yeast, and is not conserved within the Rpb1 counterparts found in RNAP I and RNAPIII, thereby serving as a unique signaling platform for RNAPII. In order to form a competent initiation complex at the promoter of a target gene, the CTD must exist in a hypophosphorylated state. Following assembly of the initiation complex, the CTD exhibits increased phosphorylation on serine 5 (S5-P), carried out by the cyclin-dependent kinase Kin28, a subunit of the general transcription factor TFIIH (Komarnitsky et al., 2000; Schroeder et al., 2000). This phosphorylation event is responsible for the recruitment of the capping machinery, which begin processing of the nascent mRNA during early transcription (Cho et al., 1997; Fabrega et al., 2003; Komarnitsky et al., 2000; Schroeder et al., 2000). As transcription elongation progresses, there is a change in the modification state of the CTD as serine 2 phosphorylation (S2-P) increases through the action of the CTDK-I complex (Cho et al., 2001). Chromatin immunoprecipitation (ChIP) experiments have demonstrated that the increase in S2-P occurs as transcription progresses through the open reading frame (ORF) (Komarnitsky et al., 2000). As transcription approaches the 3’ end of the ORF, the termination and polyadenylation machinery are recruited, some of which interact with the S2-P CTD (Licatalosi et al., 2002; Meinhart and Cramer, 2004; Kim et al., 2004). Although this transition state from S5-P to S2-P during the transcription cycle is thought to distinguish different phases of RNAPII elongation, the proteins involved in the decrease of S5-P during elongation have yet to be identified.
In addition to the aforementioned CTD-kinases, the actions of CTD-phosphatases are also required to manage the different CTD-modification states. Two CTD phosphatases, Fcp1 and Ssu72, have been characterized in yeast (for review, see Meinhart et al., 2005).
Fcp1 has a preference for the S2-P modification, and has been shown by ChIP analysis to co-localize with RNAPII throughout coding regions (Cho et al., 2001). In addition, Fcp1 mutants show an increase in the level of S2-P in the coding region of genes, indicating that the phosphatase plays a role in dephosphorylation of S2-P during the transcription cycle (Cho et al., 2001). Fcp1 is also thought to play a major role in RNAPII recycling after the complex has dissociated from the coding region (Cho et al., 1999; Kong et al., 2005; Archambault et al., 1997; Chambers et al., 1995; Aygun et al., 2008). Ssu72, conversely, is a S5-P specific CTD phosphatase and a component of the yeast cleavage and polyadenylation factor (CPF), which is involved in mRNA processing at the 3’ ends of genes (Krishnamurthy et al., 2004; Reyes-Reyes and Hampsey, 2007). ChIP assays have revealed that Ssu72 is predominately enriched at the 3’ends of genes, with little to no enrichment found at the promoter (Nedea et al., 2003; Ansari and Hampsey, 2005). Although Fcp1 and Ssu72 have both been implicated in dephosphorylation of the RNAPII CTD, neither phosphatase has been shown to regulate the S5-P to S2-P transition during transcription elongation. It is therefore likely that an additional regulatory protein(s) is required to direct the S5-P to S2-P transition dephosphorylation event.
In this study, we have characterized the interaction of a conserved protein of unknown function, Rtr1 (regulator of transcription (Gibney et al., 2008)), with RNAPII. Recent studies on Rtr1 revealed genetic interactions implicating the protein in the regulation of RNAPII transcription (Gibney et al., 2008). Our current study reveals that Rtr1 is a bona fide RNAPII-associated protein that copurifies with a transcriptionally competent form of the enzyme and can interact with CTD peptides in vitro. ChIP assays show that Rtr1 localizes to the coding regions of PMA1 and PYK1 (also known as CDC19), and that the highest level of association was seen between the peaks of S5-P and S2-P RNAPII. Deletion of Rtr1 results in the accumulation of S5-P RNAPII in whole cell extracts, as well as across the ORFs of PMA1 and PYK1 as shown by ChIP. In addition, we show that Rtr1 is able to dephosphorylate RNAPII that is present in a ternary complex supporting our hypothesis that Rtr1 is a CTD phosphatase that targets RNAPII during transcription elongation.
Our initial analysis focused on the identification of RNAPII interacting proteins through MudPIT analysis of RNAPII complexes. Using RNAPII subunits as bait, we were able to identify all twelve known subunits of the complex (Figure 1A). In addition, we identified a protein of unknown function, Rtr1. Purification of Rtr1 confirmed its interaction with RNAPII in agreement with data from large-scale proteomics studies in yeast (Gavin et al., 2002). Interestingly, the human homolog of Rtr1; RPAP-2, has also been shown to be a RNAPII-associated protein; indicating that the interaction is conserved in higher eukaryotes (Jeronimo et al., 2007). In addition to the identification of all twelve RNAPII subunits in the Rtr1-TAP purification, we were able to identify Rtr1 in four purifications using TAP-tagged RNAPII subunits as a bait, which has not been reported previously (Figure 1A). This information indicates that Rtr1 is associated with a significant fraction of RNAPII, and prompted us to further characterize the interaction. In order to determine if Rtr1 was a bona fide RNA Polymerase II-associated protein, RNAPII was purified via the TAP-tagged subunit Rpb3 in a strain that also contained Rtr1-His6-FLAG3-HA3 (Rtr1-HFH). The resulting RNAPII population was then fractionated by anion exchange chromatography performed as previously described (Hu et al., 2006). Rtr1 co-fractionates with Rpb3 and Rpb1 (using antibodies against either the S5-P or unmodified (UM) form of the CTD; Figure 1B). The western blots were confirmed by MudPIT analyses on the peak fractions as shown, which is visualized through a heat map displaying the spectral abundance factors (SAF) of the proteins (Figure 1C, Table S3). These data show that Rtr1 does indeed co-fractionate with all 12-subunits of RNAPII, indicating that they form a stable complex.
To determine if Rtr1 associates with functional RNAPII, electromobility shift assays (EMSA) and in vitro transcription experiments were performed. We used equal amounts of RNAPII purified through either Rpb3-TAP or an Rtr1-TAP strain as previously described and performed experiments using a C-tailed DNA template, which allowed for RNAPII binding in the absence of the general transcription factors (Carey et al., 2006). We found that the DNA binding activity of RNAPII to the C-tailed template was similar for both the Rpb3-TAP and Rtr1-TAP complexes (Figure S1). The amount of RNAPII used in the in vitro transcription experiments was normalized to this DNA binding activity. Both Rtr1-TAP and Rpb3-TAP purified RNAPII were able to produce full-length (FL) transcripts, indicating that Rtr1 interacts with a transcriptionally competent RNAPII (Figure 2A). However, we did observe a slight decrease in the total amount of FL transcript when comparing the highest concentrations of Rpb3-TAP to Rtr1-TAP (Figure 2A, bottom of panel). These data indicate that the association of Rtr1 with RNAPII may be inhibitory to transcription elongation.
We next sought to determine if Rtr1 interacts with a specific form of RNAPII that may direct its function during the transcription cycle. We performed western blot analysis on whole cell yeast extracts (WCE) and Rtr1-TAP purifications using antibodies directed against the different forms of the RNAPII CTD. Rtr1 interacts with both the unmodified (UM) and serine 5 phosphorylated (S5-P) forms of the RNAPII CTD in vivo, which was also recently shown by Gibney et al. (Figure 2B, Gibney et al., 2008). Since Rtr1 displays a binding preference for a particular form of the modified CTD, we next asked if Rtr1 was able to directly interact with the RNAPII CTD repeat sequence. CTD peptide binding assays were performed using recombinant Rtr1 (rRtr1) purified from bacteria and biotinylated CTD peptides containing four repeats of the Y1S2P3T4S5P6S7 sequence, which were unmodified (UM), or phosphorylated at S5, S2, or S2 and S5 (S2,5). These experiments show that rRtr1 is able to directly interact with the CTD peptides in vitro (Figure 2C). Although able to bind all four peptides, Rtr1 shows a preference for the S5-P form of the CTD and has the lowest affinity for the S2,5 modified peptide (Figure 2D).
To determine if Rtr1 co-localizes with elongating, initiating, or terminating RNAPII in vivo, chromatin immunoprecipitation (ChIP) assays were performed with an Rtr1-HFH strain using antibodies directed against HA. The genomic loci of PMA1 and PYK1 were analyzed for Rtr1, S5-P, and S2-P RNAPII occupancy using qPCR (Figure 3A & B).Both of these genes are highly expressed and have previously been used to study the localization of different phosphorylated forms of RNAPII (Komarnitsky et al., 2000). As shown in Figures 3A & B, a peak in S5 phosphorylated RNAPII is observed at the 5’ end of both PMA1 and PYK1 and decreases prior to an observed increase of S2-P, in agreement with previous results (Komarnitsky et al., 2000). Surprisingly, we found a strong and distinct peak of Rtr1 that localized to a region of both genes between the enriched regions of S5-P and S2-P. This peak was also observed after normalization of the levels of Rtr1 to the level of RNAPII occupancy (Figure S2). This localization, in combination with our other results could indicate a role for Rtr1 in the transition from S5-P CTD to S2-P CTD in vivo.
The Rtr1 deletion strain was analyzed for defects in CTD phosphorylation in vivo by western blot analysis of whole cell extracts. Loss of Rtr1 resulted in increased levels of S5-P CTD in vivo (Figure 4A, upper panel) and corresponded with a slight decrease in the cellular level of unmodified RNAPII (third panel, Figure 4B). The level of S2-P was not affected in rtr1Δ extracts nor was the level of Rpb3, which was used as a control for protein loading (bottom two panels). Quantitation was performed on triplicate experiments and is shown in Figure 4B. These data indicate that Rtr1 is involved in the regulation S5-P in vivo and supports the hypothesis that Rtr1 is involved in decreasing S5-P in wild type cells, since there is an accumulation of the S5-P form in rtr1Δ cells. In order to determine the effects of Rtr1 deletion on the levels of DNA-associated RNAPII, we performed ChIP analyses in the rtr1Δ strain and compared the levels of S5-P and S2-P RNAPII to those found in wild type. The levels of S5-P RNAPII dramatically increase in the rtr1Δ cells when compared to the levels observed in wild type cells at both the PMA1 and PYK1 ORFs (Figures 4C & D). This finding is especially true towards the 3’ ends of these coding regions, where S5-P on the CTD is normally very low. The level of S5-P RNAPII decreases near the polyadenylation sites of both genes to near wild type levels. This decline is likely due to decreased RNAPII occupancy in these regions, which was observed for Rpb3 in both the rtr1Δ and wild type cells (Figure 5A). The levels of S2-P RNAPII were also analyzed at the PMA1 and PYK1 coding regions (Figure S3). In the Rtr1 deletion, there was a slight increase in the amount of S2-P RNAPII at both PMA1 and PYK1 throughout their ORFs, suggesting that there may be local increases in S2-P although the total cellular level of S2-P remains unaffected in rtr1Δ cells.
We next addressed the effects of Rtr1 deletion on the RNAPII transcription cycle by measuring the occupancy of RNAPII across the PMA1 and PYK1 open reading frames. For these experiments, we used antibodies directed against the RNAPII-specific subunit Rpb3 in both WT and rtr1Δ cells for ChIP analysis. As shown in Figure 5A and B, the total amount of Rpb3 occupancy at PMA1 and PYK1 decreases an average of 40% in the absence of Rtr1. A similar loss of RNAPII occupancy has been observed when analyzing Rpb3 occupancy in strains containing conditional mutants of the CTD phosphatase Fcp1 (Cho et al., 2001).
To address whether or not loss of Rtr1 may result in other transcriptional defects such as transcription read through, northern blot analysis was performed on total RNA isolated from either WT or rtr1Δ cells using a probe directed against the 3’ end of PMA1. Although no read through transcript was observed, we could detect a decrease in the total amount of PMA1 in the rtr1Δ cells when compared to the levels of a RNAPIII transcript, SCR1; or to the RNAPI transcript for the 18S rRNA (Figure 5C). Quantitation was performed on triplicate experiments and is shown in Figure 5D, confirming the decrease in the level of the PMA1 transcript. This observation was further verified using qRT-PCR analysis of PMA1, as well as PYK1 and ACT1 compared to the levels of SCR1 (Figure 5E). These data support that the loss of Rtr1 results in both loss of RNAPII occupancy and reduced transcription.
Since no read through transcripts were observed at the PMA1 gene, we performed additional northern blot experiments with WT or rtr1Δ polyA+ mRNA using probes directed against the NRD1 and MRPL17 loci. We chose these loci for analysis since they had previously been shown to display a read through phenotype in cells containing a thermosensitive allele of Ssu72; whose inactivation also causes accumulation of S5-P RNAPII in vivo (Ganem et al., 2003; Krishnamurthy et al., 2004). The loss of Rtr1 results in a read through transcript at NRD1 that can also be detected with probes directed against a downstream gene, MRPL17confirming that it was a read-through product (Figure 5F). The observation of this phenotype in rtr1Δ cells suggests that the activity of both Ssu72 and Rtr1 is required for proper RNAPII termination at some loci in vivo. Their requirement for proper termination at some loci may be related to the affinity of termination factors such as Pcf11 and Rtt103 for S2-P rather than S5-P CTD (Licatalosi et al., 2002; Meinhart and Cramer, 2004; Kim et al., 2004).
The effects of Rtr1 on RNAPII transcription mirror effects previously seen with thermosensitive alleles of Fcp1 and Ssu72, the two known CTD phosphatases in yeast. In light of these parallels, we sought to determine if Rtr1 could also function as a CTD phosphatase in vitro. For these experiments, Rtr1 was purified from bacteria using a two-step affinity purification to limit contaminating bacterial proteins (Figure S4). Recombinant Rtr1 was used for in vitro phosphatase reactions using TFIIK 32P-labeled CTD peptide. As shown in Figure 6A, Rtr1 is able to dephosphorylate the CTD peptide in vitro, indicating that it is capable acting as a CTD phosphatase. To test if Rtr1 has a preference for either S2-P or S5-P, we performed phosphatase reactions with GST-CTD that had been modified in vitro by either CTDK-I or MAPK2 and visualized the different forms of GST-CTD using western blotting with antibodies that detect either S2-P or S5-P. The kinases CTDK-I and MAPK2 were used for these experiments since they both exhibit robust activity in vitro and target both S2 and S5 for phosphorylation (Trigon et al., 1998; Jones et al., 2004). Rtr1 displays phosphatase activity on both CTDK-I and MAPK2 modified GST-CTD, as shown in Figures 6B and C; and appears to have a preference for the S5-P form of the protein (Figure 6D). Some dephosphorylation of a high-mobility form of S2-P GST-CTD was seen in the CTDK-I modified reactions, which are likely hyperphosphorylated (Figure S5). The dephosphorylation of S2-P may be due to the fact that the H5 monoclonal antibody has some cross-reactivity with the S5 modification, or it could indicate that Rtr1 also has some affinity towards S2-P (Jones et al., 2004).
Although Rtr1 does not contain a previously described phosphatase motif, it does have a number of conserved residues that could contribute to its catalytic activity. In fact, studies by Gibney et al. revealed that C73 is required to compensate for loss of Rtr1 in yeast (Gibney et al., 2008). In order to address if the defects seen with C73 mutation correlate with changes in the total levels of RNAPII phosphorylation, we performed western blot analysis on yeast strains containing either WT or C73A TAP-tagged Rtr1 integrated into its chromosomal locus. Using increasing concentrations of extract, we observed increased levels of S5-P in the Rtr1-C73A-TAP strain as compared to Rtr1-TAP (Figure 6E & F). The level of S5-P in the C73A-TAP strain was similar to that observed in rtr1Δ cells, indicating that the mutant does not efficiently regulate S5-P levels in vivo In addition, copper induction experiments were performed in rtr1Δ cells that had been transformed with a CUP1 promoter driven plasmid containing either wild type or C73S RTR1. Upon copper induction, there was a decrease in the total level of S5-P RNAPII in rtr1Δ cells containing the RTR1 wild-type plasmid. However, induction of the C73S-Rtr1 mutant did not result in decreases in the total level of S5-P in the cells; which again suggests that C73 is required for Rtr1 function in vivo (Figure S6A & B).
Our experiments suggest that Rtr1 is a CTD phosphatase which functions during the transcription cycle to target S5-P for dephosphorylation. For this to occur, Rtr1 would have to display phosphatase activity on an intact RNAPII complex that was engaged on DNA. In order to test if Rtr1 was able to dephosphorylate RNAPII present in an elongation complex, we prepared ternary complexes in vitro as previously described on a biotinylated DNA template with a 3’ overhang (Figure 7A, Kong et al., 2005). For these experiments, we used Rpb3-TAP purified RNAPII, which contains all twelve subunits (Figure 1A) and phosphorylated the complex in the presence of [γ-32P]-ATP with either TFIIK or MAPK2. As shown in Figures 7B and C, Rtr1 is able to dephosphorylate RNAPII that is present in an elongation complex with similar efficiency as seen with the free GST-CTD protein (Figures 6D and and7D).7D). These data support our hypothesis that Rtr1 is a CTD phosphatase that targets RNAPII during the transition in the CTD modification state.
We have shown that Rtr1 is a phosphatase that preferentially dephosphorylates the S5 modification on the RNAPII CTD in vivo. In addition to Rtr1, the phosphatases Ssu72 and Fcp1 also play an important role in the regulation of CTD phosphorylation. The depletion of Fcp1 in vivo has been shown to result in the specific accumulation of S2-P RNAPII, which is enriched at the 3’ end of genes (Cho et al., 2001). The depletion of Ssu72, conversely, has been shown to lead to an accumulation in the S5-P form of RNAPII by western blot analysis (Krishnamurthy et al., 2004). It has been unclear, however, if Ssu72 is the transition phosphatase since, as a member of the CPF complex, its major site of localization is towards the 3’ end of the gene (Nedea et al., 2003); and since inactivation of Ssu72 has not been shown to increase S5-P across open reading frames. The localization of Rtr1 by ChIP indicates that the protein is not enriched at the promoter, but is highly enriched at a region preceding the phosphorylation of serine 2, which is consistent with Rtr1 being the transition phosphatase. This hypothesis is also supported by the observation that Rtr1-TAP purifications are enriched in RNAPII complexes that contain some combination of non-phosphorylated and S5 phosphorylated repeats. These data provide the first evidence of a phosphatase specifically involved in the regulation of the S5-P to S2-P transition; and indicate that Rtr1 is involved in S5-P dephosphorylation during elongation.
We have shown that mutation of C73 in Rtr1 lead to increased S5-P of the RNAPII CTD in vivo to levels that rival the rtr1Δ full deletion strain. This cysteine residue is completely conserved in Rtr1 homologs in higher eukaryotes and our data indicate that it is required for proper Rtr1 function. Rtr1 does not contain a characterized phosphatase motif, but there is precedence for a low level of sequence homology between known phosphatases and the other identified CTD phosphatases, Fcp1 and Ssu72. Fcp1 was originally isolated through its ability to dephosphorylate the CTD, not by sequence homology (Chambers and Dahmus, 1994; Chambers et al., 1995; Archambault et al., 1997). When Fcp1 was characterized, it was described to contain an “unusual” phosphatase domain that was unique for eukaryotic protein phosphatases (Kobor et al., 1999). By contrast, Ssu72 was found to contain a known phosphatase motif, CX5R, but had no other significant homology to known phosphatases (Meinhart et al., 2003).
What are the consequences of misregulated CTD modification during the transcription cycle? In addition to the hyperphosphorylation of RNAPII, loss of Rtr1 leads to diminished transcription of a number of RNAPII target genes and causes transcription read-through at some loci, indicating a termination defect. A decrease in RNAPII occupancy has previously been reported in studies which made use of different thermosensitive alleles of Fcp1 (Cho et al., 2001). The loss of RNAPII occupancy may be a result of RNAPII hyperphosphorylation, either at S5-P as seen with deletion of Rtr1; or S2-P as seen with mutant alleles of Fcp1. It has been suggested that CTD hyperphosphorylation may inhibit RNAPII reinitiation, leading to decreased levels of RNAPII in the ORF (Lux et al., 2005; Max et al., 2007; Payne et al., 1989). In addition to the reduced transcription observed at the RNAPII loci tested in our study, Gibney et al. observed that induction of GAL1 and GAL7 was not detectable in rtr1Δ cells grown with galactose as the sole carbon source. Similar reductions in transcription have also been observed using a thermosensitive allele of Ssu72, ssu72-2. Although RNAPI transcripts were not affected at the restrictive temperature, the levels of several RNAPII transcripts were found to decrease in ssu72-2 cells (Dichtl et al., 2002). The loci tested included ACT1 which decreased 52%, which was very similar to what we observed for the same loci in rtr1Δ cells (Figure 5E). The transcription defects of Fcp1 thermosensitive alleles have been more extensively studied and have been shown to cause a global defect in transcription. Using expression arrays, approximately 77% of all yeast genes were found to have more than a 2-fold reduction in their expression levels when mRNA was extracted from fcp1-1 cells grown at the restrictive temperature (Kobor et al., 1999). These data show that the regulation of the RNAPII phosphorylation level is essential for the maintenance of proper transcription in vivo.
In conclusion, our experiments have uncovered a previously uncharacterized player in the RNAPII transcription cycle, Rtr1, a phosphatase required for the regulation of the CTD modification state during early RNAPII elongation events. In addition to the recruitment of mRNA processing factors, the different combinations of S5-P and S2-P CTD have been shown to be involved in the recruitment of a number of accessory factors that are involved in histone modification and chromatin remodeling processes during transcription (Li et al., 2007a). It remains to be determined if the correct arrangement of S5-P RNAPII across the open reading frame is involved in the management of the chromatin state during the transcription cycle, and how alterations in the regulation of this process may affect mRNA processing events other than termination.
The yeast strains used in this study are given in Table S1. All strains were derived from BY4741.
Antibodies used in these studies are listed in Table S2.
All TAP-tagged strains were obtained from Open Biosystems (Ghaemmaghami et al., 2003) and were purified as previously described with slight modification (Puig et al., 2001). In order to isolate complexes from chromatin solubilized extracts; extracts were treated with 120ug of heparin (sodium salt, Sigma) and 100 units of DNase I (amplification grade, Sigma) for 10 minutes at room temperature after cell lysis. Detailed methods are provided in the supplementary experimental procedures.
Rpb3-TAP purified RNAPII was isolated from the Rpb3-TAP Rtr1-His6-FLAG3-HA3 strain via TAP purification and was then subjected to fractionation on a Uno-Q1 anion exchange column (Biorad). Proteins were eluted with a gradient of column buffer (50 mM Tris·HCl, pH 7.8/10% glycerol/1 mM EDTA/10 µM ZnCl2) starting at 150 mM (NH4)2SO4 and ending at 500 mM (NH4)2SO4 as previously described (Banks et al., 2007).
EMSAs and in vitro transcription experiments were performed as previously described (Carey et al., 2006) using equal concentrations of Rpb3-TAP or Rtr1-TAP as illustrated. EMSAs were performed using the 32P-labeled C-tailed nucleosomal template as described. The products were separated on 4% native polyacrylamide gels (37.5:1 acrylamide-bis acrylamide) electrophoresed in 0.5 X TBE at 4°C. For the in vitro transcription reactions, the 32P-RNAs were separated on an 8% polyacrylamide/urea gel. The resulting gels were dried, exposed to a Phosphorimager or film.
The Rtr1-His6-HA1-Protein A bacterial expression vector was constructed and protein purified as described in the supplementary experimental procedures. The UM, S2-P, and S5-P peptides were synthesized, and the S2,5P peptide was prepared as previously described (Phatnani and Greenleaf, 2004),(Phatnani et al., 2004),(Kizer et al., 2005). For CTD binding assays, biotinylated CTD peptides (2.5 µg) were incubated with 0.5 mg of streptavidin-coated Dynabeads M280 (Invitrogen) in binding buffer (50 mM Tris-HCl pH 6.5, 300mM NaCl, 1 mM dithiothreitol, 0.5% Nonidet P-40, 1mM PMSF) at 4 °C for 2 h. Approximately 1µg recombinant Rtr1-His6-HA1 and 100 µg of BSA were added in binding buffer and allowed to interact with the peptides for 3 hours at 4°C. The beads were washed four times in binding buffer containing 10 µg/mL BSA using a Dynal MPC (Invitrogen) prior to addition of 4x SDS-loading buffer for elution of bound proteins.
Northern blots were performed as previously described using either total or polyA+ mRNA as indicated, isolated from WT or rtr1Δ as given in the text (Carrozza et al., 2005; Li et al., 2007b). Probes were generated by PCR with primers listed in Table S4.
Quantitative RT-PCR was performed on total RNA isolated from WT or rtr1Δ using the RNeasy® Mini Kit (Qiagen) according to the manufacturer’s protocol. First strand synthesis was performed using Superscript™ II Reverse Transcriptase (RT, Invitrogen) and random hexamers with 2.5 µg total RNA. Real-time PCR was performed with samples generated in the presence and absence of RT using a Biorad iCycler and SYBR Green master mix (Stratagene) and primers directed against the gene of interest (Table S4). Total nanograms of transcript was determined by comparison to a standard curve generated with the same primers and 5 10-fold serial dilutions of yeast genomic DNA.
TFIIK was generated by purification of the complex from yeast through Kin28-TAP. CTDK-I was purified through the Ctk1-TAP subunit. Active MAPK2/Erk2 was obtained from Millipore. For kinase reactions, approximately 400 ng of RNAPII, 200 ng of GST-CTD, or 5 µg of biotylated CTD peptide was incubated with the indicated kinase for 1 hour at 3°C in kinase buffer (40 mM Hepes, pH 7.5, 10mM MgCl2, 5 mM dithiothreitol, and either 10 µ5Ci of [γ-32P]-ATP (6000 Ci/mmol; Perkin Elmer) or 500 µM ATP (Roche). Kinase reactions were stopped by removal of the unincorporated ATP through an Illustra MicroSpin Column (GE Healthcare).
Reactions were performed with approximately 5 pmol of phosphorylated RNAPII or GST-CTD (Thompson et. al, 1993) as indicated or 1ug of CTD-peptide in phosphatase buffer (50mM Tris-HCl, pH 6.5, 10 mM MgCl2, 20mM KCl, and 5 mM dithiothreitol). Reactions were quenched by the addition of SDS loading buffer and incubation at 98°C for 5 minutes prior to loading on a gel. Reactions performed with RNAPII ternary complex or GST-CTD were separated on a 10 or 12% SDS-PAGE gel. Reactions performed with the CTD-peptides were separated on a 16% Tricine gel. After separation, 32P-labeled protein gels were dried and exposed to either a PhosphoImager screen or BioMax film with an intensifying screen. Reactions performed in the absence of 32P were subjected to western blot analysis as described.
Ternary complexes were prepared as described in (Kong et al., 2005). Approximately 5 pmol of RNAPII present in a ternary complex was used for phosphatase assays.
We are grateful to all the members of the Washburn and Workman laboratories for useful discussions and technical suggestions and would also like to thank Drs. Joan and Ron Conaway for critical reading of our manuscript and useful suggestions. We would like to thank Dr. Richard Young for the gift of the GST-CTD plasmid and Dr. Kevin Morano for the CUP1-Rtr1 plasmids. This work was supported by a postdoctoral fellowship from the NIGMS to A.L.M. (F32 GM075541) and funding from the Stowers Institute.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.