Genetic Interactions With C-Terminal Domain (CTD) Kinases and the CTD of RNA Pol II Suggest a Role for ESS1 in Transcription Initiation and Elongation in Saccharomyces cerevisiae

Corresponding author: Steven D. Hanes, 120 New Scotland Ave., Albany, NY 12208., hanes{at}wadsworth.org (E-mail)

Communicating editor: F. WINSTON

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Ess1 is an essential prolyl isomerase that binds the C-terminal domain (CTD) of Rpb1, the large subunit of RNA polymerase II. Ess1 is proposed to control transcription by isomerizing phospho-Ser-Pro peptide bonds within the CTD repeat. To determine which step(s) in the transcription cycle might require Ess1, we examined genetic interactions between ESS1 and genes encoding the known CTD kinases (KIN28, CTK1, BUR1, and SRB10). Although genetic interactions were identified between ESS1 and all four kinases, the clearest interactions were with CTK1 and SRB10. Reduced dosage of CTK1 rescued the growth defect of ess1^ts mutants, while overexpression of CTK1 enhanced the growth defects of ess1^ts mutants. Deletion of SRB10 suppressed ess1^ts and ess1 mutants. The interactions suggest that Ess1 opposes the functions of these kinases, which are thought to function in preinitiation and elongation. Using a series of CTD substitution alleles, we also identified Ser5-Pro6 as a potential target for Ess1 isomerization within the first "half" of the CTD repeats. On the basis of the results, we suggest a model in which Ess1-directed conformational changes promote dephosphorylation of Ser5 to stimulate preinitiation complex formation and, later, to inhibit elongation.

THE expression of eukaryotic RNA polymerase II (pol II)-dependent genes is regulated at multiple levels: transcription initiation, capping, elongation, splicing, termination, and transcript cleavage (LEE and YOUNG 2000 ). Each step requires accessory proteins whose activities appear to be coordinated, in part, by binding to the C-terminal domain (CTD) of the pol II large subunit, Rpb1 (HIROSE and MANLEY 2000 ). How binding of these proteins to the CTD is regulated in such a way as to promote efficient transcription and mRNA processing is not understood.

The CTD is composed of multiple repeats of the consensus heptapeptide sequence YSPTSPS. In yeast, the CTD contains 26 or 27 repeats of this sequence, while in mammals the CTD contains 52 repeats (CORDON 1990 ). Changes in the phosphorylation state of this Ser-Pro-rich region accompany the transitions of RNA polymerase II as it moves from the promoter of a gene to its terminus (KOMARNITSKY et al. 2000 ). The CTD is hypophosphorylated during preinitiation complex (PIC) formation and is phosphorylated as pol II leaves the promoter and during elongation (LAYBOURN and DAHMUS 1989 ). The different phosphorylation states of the CTD might act as a signal to coordinate binding and release of the multiple proteins required for individual steps of transcription.

Four potential CTD kinases have been identified, each functioning as part of a kinase-cyclin complex. These kinases are thought to act at discrete steps in transcription. Kin28-Ccl1 (Cdk7-cyclin H in humans), a component of TFIIH, facilitates promoter clearance and mRNA capping (CISMOWSKI et al. 1995 ; AKHTAR et al. 1996 ; RODRIGUEZ et al. 2000 ). Ctk1-Ctk2, the kinase-cyclin subunits of CTDK-I, and Bur1-Bur2 (Cdk9-cyclin T in humans) are thought to promote efficient elongation (LEE and GREENLEAF 1997 ; MAJELLO et al. 1999 ; MURRAY et al. 2001 ). In addition, CTDK-I may help recruit pre-mRNA processing factors involved in 3'-end formation (SKAAR and GREENLEAF 2002 ). Srb10-Srb11 (Cdk8-cyclin C in humans), a component of the SRB/mediator, negatively regulates transcription of certain genes during exponential growth in rich media (HOLSTEGE et al. 1998 ) by inhibiting PIC formation (HENGARTNER et al. 1998 ) and by promoting the degradation of certain transcription activators (CHI et al. 2001 ). In addition to four CTD kinases, two CTD-specific phosphatases have been identified, Fcp1 (ARCHAMBAULT et al. 1997 ; KOBOR et al. 1999 ) and Ssu72 (K. SHANKARLING and M. HAMPSEY, personal communication).

Changing the phosphorylation state of the CTD is one mechanism by which binding of accessory proteins to the CTD of pol II may be regulated. Another mechanism might be conformational isomerization of the CTD by enzymes called peptidyl-prolyl cis/trans isomerases (PPIases; SCHIENE and FISCHER 2000 ). A PPIase could alter the structure of the CTD, changing the affinity of accessory proteins known to bind the CTD. One such enzyme, Ess1, which is essential for growth in yeast (HANES et al. 1989 ), binds phospho-Ser-Pro motifs (YAFFE et al. 1997 ), such as those found within the phosphorylated CTD. Ess1 has been shown to interact physically and genetically with Rpb1 (MORRIS et al. 1999 ; WU et al. 2000 , WU et al. 2001 ) and is thought to stimulate the binding of Fcp1, a CTD-specific phosphatase (LIN et al. 2002 ), to the CTD. Analysis of the Bye1 high-copy suppressor of ess1 mutants led to the finding that Ess1 acts negatively in transcription elongation (WU et al. 2003 ).

Here, we examined genetic interactions between ESS1 and genes encoding the four known or suspected CTD kinases. The goals were to help identify those step(s) during the transcription cycle in which Ess1 might act and to identify possible target sites within the CTD that are shared between the CTD kinases and Ess1. Genetic interactions with multiple CTD kinases indicate that Ess1 is likely to act at more than one step in transcription. In addition to a negative role in elongation, our results point to a positive role for Ess1 in preinitiation complex formation. We have also discovered that ess1^ts mutants are suppressed by serine-to-alanine mutations (YSPTAPS) in the amino-terminal half of the CTD repeats. Since Ser5-to-alanine substitution prevents phosphorylation at this position, suppression by this allele suggests that Ess1 may function by promoting dephosphorylation of Ser5 in the amino-terminal portion of the CTD.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Yeast strains, genetic methods, and media:
Yeast strains used in these experiments are listed in Table 1 and most are derived from the strain W303-1A (R. Rothstein), except YPR57, which is derived from W1021-7C and W961-5B (R. Rothstein), the rpb1 pRP112 strain (NONET and YOUNG 1989 ), and GY604 (MURRAY et al. 2001 ). The ess1 mutant alleles have been described previously (WU et al. 2000 ). Standard methods were used for transformation, mating, sporulation, and tetrad dissection (GUTHRIE and FINK 1991 ). Rich (YEPD), complete synthetic (CSM), and 5-fluoroorotic acid (5-FOA) media were prepared according to standard methods (ADAMS et al. 1997 ). For spot tests, cells were grown in liquid culture to an OD₆₀₀ of 0.5–0.8 (1–2 x 10⁷ cells/ml). An estimated 2–3 µl of cells (diluted to 0.5 OD₆₀₀ units) were spotted onto the equivalent solid media, using a multiprong device following serial 1:5 dilutions.

Gene deletions and disruptions:
KIN28, CTK1, and BUR1 gene deletions were generated by transformation of yeast strains W303-1A/B, CBW9, CBW15, and CBW32 with a polymerase chain reaction (PCR) product containing a G418^R cassette flanked by 45 bp homologous to the ends of the target open reading frame (WACH et al. 1994 ). Transformants were selected on YEPD containing 400 mg/liter geneticin (GIBCO BRL, Gaithersburg, MD). SRB10 was disrupted (and mostly deleted) by transformation of yeast with a SmaI-PstI restriction fragment of plasmid psrb10::TRP1 (gift of J. Dutko and R. Zitomer). This construct deletes base pairs from 300 to 1307 of the 1668-bp coding sequence of SRB10. The srb10 mutant strains generated using this construct are referred to as deletions (srb10) throughout this article. The presence of the knockouts was determined by PCR, using one primer complementary to sequence within the marker gene and another with complementarity to chromosomal DNA flanking the gene of interest. The identity of ess1^ts alleles in tetrad segregants was determined by PCR, using oligonucleotides specific to each allele, or by DNA sequencing. Oligonucleotide sequences are available on request.

Plasmids:
The high-copy plasmid pKIN28 was made by subcloning KIN28-HA from pGK13 (KIMMELMAN et al. 1999 ) into pRS425 (2µ, LEU2; CHRISTIANSON et al. 1992 ), using PstI and HindIII. Plasmids containing kin28^ts16 and kin28^T162A (in YCplac22; CEN TRP1) were from Mark Solomon (KIMMELMAN et al. 1999 ). CTK1 plasmids were from Jan Jones and Arno Greenleaf. The high-copy plasmid pJYC1501 contains a BglII-SalI fragment of the CTK1 locus inserted into the same sites of YEp24 (2µ, URA3). pJYC1501-HA3 and pJYC1501-K212A-HA3 are derivatives that encode HA-tagged wild-type and kinase-deficient Ctk1 (K212A), respectively. pJYC1511 is a centromeric CTK1 plasmid (CEN HIS3). Additional plasmids (2µ, URA3) encoding wild-type (YEp112CTK1WT-HA) and kinase-deficient Ctk1 (D324N; YEp112CTK1DN-HA) were from Denis Ostapenko and Mark Solomon. A high-copy SRB10 plasmid, YEp(195)SRB10 (2µ, URA3), was obtained from Richard Zitomer. Plasmids bearing CTD mutations are listed in Table 4 and were from Jeff Corden. BUR1 plasmids pGP232 (2µ, LEU2 BUR1), pGP162 (CEN TRP1 BUR1), and pGP163 (CEN TRP1 bur1-2) were from Gregory Prelich. YEpHESS1 has been described (HANES et al. 1989 ).

Plasmid-loss experiments:
We used a plasmid encoding Candida albicans Ess1, which complements Saccharomyces cerevisiae ess1 mutations (DEVASAHAYAM et al. 2002 ). The C. albicans version was used rather than the S. cerevisiae version to reduce possible recombination between the plasmid-borne copy of ESS1 and chromosomal ess1 mutations. The kin28 ess1 double-mutant strains were patched to solid 5-FOA – TRP medium to identify cells capable of growing after loss of pCaESS1 (2µ, URA3). For CTD mutant experiments, the haploid strains rpb1 pRP112 and ess1^H164R rpb1 pRP112 were transformed with LEU2 plasmids carrying rpb1-CTD mutations, and growth after loss of wild-type RPB1 carried on pRP112 (URA3) was monitored by replica plating to 5-FOA medium.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Allele-specific interactions between KIN28 and ESS1:
Kin28, an essential component of TFIIH, is thought to phosphorylate the CTD at the time of promoter escape (HENGARTNER et al. 1998 ). Because ESS1 is an essential gene, we used temperature-sensitive (ts) alleles that allow cell growth at 25° or 30° but not at 35° or 37°. The two we used, ess1^H164R and ess1^A144T, contain single-amino-acid substitutions in structurally distinct regions of the protein, one in the catalytic site (H164R) and one at the interface between the WW and prolyl isomerase domains (A144T; WU et al. 2000 ). To investigate a potential genetic interaction between ESS1 and KIN28, we first transformed ess1 mutant strains with a high-copy plasmid to overexpress KIN28. High-copy expression of KIN28 neither suppressed nor enhanced the ts lethality of either ess1^ts allele (data not shown), nor did lowering the dosage of KIN28 in diploid strains (kin28/KIN28 ess1^ts/ess1^ts; data not shown).

We next tested for synthetic lethality or suppression using kin28 haploid derivatives of strain W303-1A that carried the kin28 alleles on centromeric plasmids (e.g., kin28 ess1^H164R pkin28^ts16). The kin28 alleles used were kin28^ts16, which is temperature sensitive for growth and defective in kinase activity (CISMOWSKI et al. 1995 ), and kin28^T162A, which has a nonphosphorylatable T-loop mutation that severely reduces its kinase activity, but does not generally affect growth (KIMMELMAN et al. 1999 ). Genetic interaction was tested by monitoring the ability of different kin28 ess1 double-mutant strains to grow on 5-FOA medium after loss of a plasmid-borne copy of ESS1 (Fig 1). In the wild-type ESS1 background at 25°, the kin28^ts16 and kin28^T162A mutations appear to retain the pCaESS1 plasmid, as detected by moderate increases in 5-FOA sensitivity compared to KIN28 cells (Fig 1A, left, row 1). Alternatively, it is possible that the kin28 mutant cells lost the plasmid but are slow growing (see below), giving rise to the apparent 5-FOA sensitivity. In either case, the ess1^H164R mutation seemed to reverse this effect, facilitating loss of the pCaESS1 plasmid and subsequent growth on 5-FOA by both kin28^ts16 (at 25°) and kin28^T162A (most visible at 30°). Results with the ess1^A144T mutant cells were uninformative, as ess1^A144T appeared to be generally resistant to pCaESS1 plasmid loss (or slow growing on 5-FOA), even in KIN28 wild-type cells.

View larger version (79K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Complex allele-specific interaction between ess1 and kin28 mutations. (A) The chromosomal KIN28 gene was deleted in ESS1 and ess1 strains. KIN28 function was provided by a plasmid-borne copy of KIN28 or kin28 as indicated. Cells were patched onto 5-FOA – trp medium and tested for the ability to lose a 2µ, URA3-based ESS1 plasmid (pCaESS1) at 25° and 30°. Growth indicates ability to lose the ESS1 plasmid. The most prominent interaction observed is between ess1H164R and kin28T162A. In kin28T162A cells, 5-FOA resistance is maximal (i.e., highest frequency of loss of pCaESS1) in an ess1H164R background vs. ESS1 control cells. (B) ESS1 kin28 and ess1H164R kin28 cells bearing the indicated KIN28 or kin28 plasmids were spotted onto CSM – trp medium (1:5 serial dilutions) and grown for 3 days at the indicated temperatures. The ess1H164R mutation appears to suppress the ts-growth defects of both kin28 mutants.

At higher temperature (30°) the effects were similar but not identical (Fig 1A, right). Resistance to 5-FOA of the kin28^T162A ess1^H164R cells was more pronounced compared to kin28^T162A ESS1 cells, indicating that the double-mutant combination is favorable, allowing ESS1 plasmid loss. For kin28^ts16, 5-FOA resistance is no longer detected in either ESS1 background, probably because kin28^ts16 cells grow slowly at this temperature, rather than because of a failure to lose the plasmid. From these experiments, it appears that there may be genetic interactions between ESS1 and KIN28.

Genetic interaction was further examined by comparing the relative growth rates of cells that had lost the pCaESS1 plasmid on medium that lacked 5-FOA. The results indicated that ess1^H164R suppresses the ts-growth defects of both kin28^ts16 and kin28^T162A (Fig 1B). This is most easily seen at 30° for kin28^ts16, where there is partial suppression, and at 37° for kin28^T162A (compare to ESS1). Independent isolates behaved similarly (data not shown). The results also show that kin28^T162A suppressed the ts-growth defect of ess1^H164R cells at 37°. Given that Kin28 is involved in transcription initiation, the results may implicate a role for Ess1 in this step. Other genetic interaction experiments between ESS1 and KIN28 yielded results that suggest interactions between these genes are likely to be complex and allele specific (data not shown). The simplest interpretation of our results is that Ess1 might oppose (attenuate) Kin28-dependent initiation.

Reduced CTK1 gene dosage suppresses ess1^ts mutations:
Ctk1, another CTD kinase, has been shown to promote transcription elongation efficiency in vitro (LEE and GREENLEAF 1997 ) and is linked to 3'-end formation (CONRAD et al. 2000 ; SKAAR and GREENLEAF 2002 ). Ess1 has been proposed to inhibit elongation and promote termination (WU et al. 2003 ) and was recovered in a screen for mutations that cause defects in 3'-end formation (HANI et al. 1995 ). Therefore, Ess1 and Ctk1 are predicted to interact genetically, perhaps in an opposing manner.

Previous work has shown that CTK1 is not essential, although ctk1 mutants had a slow growth phenotype (LEE and GREENLEAF 1991 ; GIAEVER et al. 2002 ). In our strain backgound (W303-1A), however, we could not generate viable ctk1 haploid cells, even after the addition of SSD1-v, an allele of SSD1 that suppresses a number of other mutations in this background (LORENZ and HEITMAN 1998 ). In addition, high-copy expression of ESS1 did not suppress the ctk1 mutation (data not shown).

Instead, we worked with diploid strains in which one copy of CTK1 was deleted. The results show that ctk1/CTK1 suppressed the ts phenotype of ess1^H164R and ess1^A144T homozygous mutants at 34° (Fig 2A). The suppression by ctk1/CTK1 was much stronger for the ess1^H164R mutant, whose growth is generally more robust than that of ess1^A144T. Several independently derived ctk1 mutants were used for these experiments and the results were the same (data not shown). As a control we also showed that adding back CTK1 on a centromeric plasmid reversed this effect, rendering ctk1/CTK1 ess1^H164R/ess1^H164R cells more temperature sensitive than vector-only controls (Fig 2B). Western blot analysis indicated that the reduced dosage of CTK1 did not alter the expression levels of ess1^ts mutant proteins (data not shown). Decreased dosage of CTK1 did not, however, suppress a complete deletion of ESS1. This was tested using a ctk1/CTK1 ess1/ess1 strain carrying ESS1 on a plasmid (pCaESS1) and finding that cells were unable to lose the plasmid by plating to 5-FOA (data not shown).

View larger version (63K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Reduced dosage of CTK1 suppresses the growth defect of ess1ts cells. (A) Serial dilutions (1:5) of wild-type or isogenic ess1ts diploids containing one or two copies of CTK1 were spotted onto YEPD medium and grown for 2 days at 30° and 34°. ess1H164R/ess1H164R mutants with only one copy of CTK1 (CTK1/) grew nearly as well as ESS1/ESS1 wild-type cells, even at 34°. (B) As a control, cells were transformed with a centromeric plasmid (pJYC1511) encoding CTK1 or a vector control (pRS413), spotted to CSM – his medium (1:5 dilutions), and grown for 3 days. Results show that ess1H164R mutant cells with a reduced chromosomal dosage of CTK1 (CTK1/) but carrying a pCTK1 plasmid were restored to temperature sensitivity (at 34°), whereas the vector control cells were not.

The above results indicated that reduced levels of Ctk1 suppress the growth defects in ess1^ts mutants. We tested if the reciprocal was true, whether increased levels of Ctk1 might enhance the growth defects in ess1^ts mutants. Indeed, high-copy expression of CTK1 enhanced the temperature sensitivity of both the ess1^H164R and ess1^A144T cells (Fig 3A). This effect is detected at 30° and 32° but is most prominent at 34°. Note that the plasmid used in these experiments also contains a partial open reading frame (ORF; TGL1). However, we do not think this ORF could have caused the observed effect because three-fifths of the coding sequence is missing. In addition, independent experiments using plasmids without this ORF present gave similar results, and an equivalent plasmid carrying a CTK1 catalytic mutant did not produce this effect (both in Fig 3B, below).

View larger version (75K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Overexpression of kinase-active CTK1 enhances the growth defect of ess1ts cells. (A) Serial dilutions (1:5) of wild-type or isogenic ess1ts haploid cells containing vector alone or a high-copy CTK1 plasmid were spotted onto CSM – ura medium and grown for 3 days at the indicated temperatures. Both ess1H164R and ess1A144T mutants bearing the pCTK1 plasmid grew slower than cells bearing the control vector; this is particularly visible at 32° and 34°. (B) Cells were treated as in A except that plasmids encoding a kinase-deficient mutant were used, and the cells were grown and spotted on CSM – trp medium. Top, growth of ess1H164R cells expresssing the control wild-type CTK1, but not the kinase-deficient ctk1K212A, is inhibited at semirestrictive temperature (34°). Bottom, similar to the top except mutant allele ctk1D324N and a CTK1 control plasmid were used.

To test whether the kinase activity of high-copy-expressed Ctk1 was required for enhancing the growth defect of ess1^H164R cells, two kinase-deficient alleles of CTK1 were used, ctk1^K212A and ctk1^D324N. Plasmids carrying these alleles did not enhance the growth defect of ess1^H164R cells at semipermissive temperature (34°) as did wild-type control CTK1 plasmids (Fig 3B). These results indicate that the kinase activity of Ctk1 is required for enhancing the ts-growth defect of ess1^H164R cells.

In summary, reduced CTK1 dosage suppresses ess1^ts mutations, while increased CTK1 activity enhances ess1^ts mutations, indicating that Ess1 and Ctk1 may have opposing functions during transcription. Since CTK1 is known to stimulate elongation, these results suggest that Ess1 inhibits elongation, consistent with other recent studies (WU et al. 2003 ). In addition, since CTK1 has also been implicated in 3'-end processing (CONRAD et al. 2000 ; SKAAR and GREENLEAF 2002 ), Ess1 may also be important during this step.

Genetic interactions between between BUR1 and ESS1:
BUR1 encodes a kinase that may phosphorylate Ser5 residues within the CTD repeat and, like CTK1, acts positively to promote transcription elongation (MURRAY et al. 2001 ). Other studies confirm a role in elongation, but suggest that Bur1 kinase may have targets other than the CTD (KEOGH et al. 2003 ).

We tested for genetic interaction between BUR1 and ESS1, using a variety of experiments. High-copy BUR1 expression did not enhance or suppress the ts-growth defect of ess1^H164R (or ess1^A144T) mutant cells at various temperatures (25°, 30°, 34°, and 37°), nor did a reduction in BUR1 dosage, for example, in a bur1/BUR1 ess1^H164R/ess1^H164R diploid strain (data not shown). In addition, bur1/BUR1 ess1/ess1 mutant cells were inviable upon loss of an ESS1-containing (URA3) plasmid as indicated by the failure to grow on 5-FOA medium (data not shown). Thus, neither overexpression nor reduced dosage of BUR1 suppressed (or enhanced) ess1 mutations. We also generated a bur1 ess1 haploid mutant bearing a plasmid-borne copy of the bur1-2 allele (see below) and an ESS1 plasmid (pCaESS1). This strain could not lose the pCaESS1 plasmid, suggesting there is no suppression of ess1 by bur1-2 (data not shown).

Segregation analysis was then used to test whether ess1 bur1 double mutants are synthetic lethal. Because both BUR1 and ESS1 are essential, we used the bur1-2 slow-growth allele (MURRAY et al. 2001 ), combined with the ess1^H164R ts allele. The results are shown in Table 2. For controls, we analyzed the bur1-2 and ess1^H164R single mutants. As previously demonstrated (MURRAY et al. 2001 ), bur1-2/BUR1-2 diploids gave rise to four viable spores, and these showed 2:2 segregation for mild slow growth (Fig 4). As expected, ess1^H164R/ESS1 diploids gave rise to four viable spores at the permissive temperature, which grew equally well. In contrast, bur1-2/BUR1-2 ess1^H164R/ESS1 double mutants gave rise to a large number of tetrads bearing fewer than four viable spores. This result is consistent with two unlinked genes interacting to produce a synthetic enhancement of their individual growth defects (see also Fig 4 legend). However, if the double-mutant combination was always lethal, we would have expected far fewer than half the tetrads to yield four viable spores. Instead, slightly more than half the tetrads yielded four viable spores. This result suggests that the double mutants are most likely to be very slow growing (e.g., tiny colonies in tetrads 1, 4, and 7 in Fig 4), but that these mutants show some variability in their growth phenotype, ranging from inviability (e.g., dead spores in tetrads 2, 5, and 6; Fig 4) to slow/moderate growth (small/medium colony in tetrad 3; Fig 4).

View larger version (41K): [in this window] [in a new window] [Download PPT slide]   Figure 4. Segregation analysis of bur1 ess1 double mutants indicates a synthetic growth defect. Tetrads were dissected as described in Table 2. Tetrads were verified using independent markers. The darker color that is visible for some colonies is due to segregation of the ade2-1 allele. Very small and nongrowing colonies (bottom) are likely to be bur1-2 ess1H164R double mutants. For two unlinked mutations, we expected segregation patterns of 1:1:4 for PD:NPD:TT. If we assume that bur1-2 ess1H164R double mutants are extremely slow growing (or dead), then the following colony phenotypes would be expected for segregants of the double-mutant cross: parental ditype (PD), 2 large, 2 medium; nonparental ditype (NPD), 2 large, 2 small (or dead); tetratype (TT), 2 large, 1 medium, 1 small (or dead). Among the 26 tetrads that yielded four viable spores (data from Table 2) the ditypes were scored as follows: 5:3:18 for PD:NPD:TT. This ratio is close to the expected 1:1:4, supporting the assumption that bur1-2 ess1H164R double mutants are extremely slow growing or dead (in the case of 3:1 viable:inviable tetrads).

The slow-growing (and nonviable) colonies are inferred to be bur1-2 ess1^H164R double mutants. For ess1^H164R, ts growth at 37° was monitored and segregation patterns were consistent with the above inference. From these experiments, we conclude that bur1-2 and ess1^H164R mutations exhibit synthetic lethality or synthetic slow growth.

BUR1 has previously been shown to interact with other genes involved in elongation (SPT4/5), as well as with RNA pol II itself and a CTD phosphatase (FCP1; MURRAY et al. 2001 ). Because ESS1 also interacts with SPT4/5 (WU et al. 2003 ), as well as with RNA pol II and FCP1 (WU et al. 2000 ; see below), it was perhaps expected that we would detect interactions between BUR1 and ESS1. The genetic interaction observed between ESS1 and BUR1 is consistent with a role for ESS1 in elongation. However, in contrast to results with CTK1, the BUR1 results seem to indicate a positive role for Ess1 in elongation.

Disruption of SRB10 suppresses ess1^ts and ess1 mutations:
Phosphorylation of the CTD by Srb10 before PIC formation has been shown to inhibit transcription (HENGARTNER et al. 1998 ). We previously reported interactions between ESS1 and RPB1 that implicate Ess1 as a positive regulator of transcription (WU et al. 2000 ). This suggested that any interaction between Ess1 and Srb10 would be antagonistic. Since SRB10 is not essential (LIAO et al. 1995 ), we deleted SRB10 (srb10) in wild-type and ess1^ts haploid strains to determine whether there is a genetic interaction with ESS1. As expected, the srb10 ess1^H164R and srb10 ess1^A144T double-mutant strains grew at permissive temperatures (30° and 32°; Fig 5). However, the double mutants also grew at the semipermissive temperature (34°), and one of the mutants, srb10 ess1^H164R, grew at the restrictive temperature (37°). These results indicate that the srb10 mutation suppresses the growth defect of ess1^ts mutants.

View larger version (103K): [in this window] [in a new window] [Download PPT slide]   Figure 5. Deletion of SRB10 suppresses ess1ts and ess1 mutants. Serial dilutions (1:5) of wild-type or ess1ts cells containing SRB10 or srb10 or an ess1 srb10 double deletion strain were spotted onto YEPD medium and grown for 2 days at the indicated temperatures. The srb10 deletion alone resulted in only a slight ts-growth defect. Deletion of SRB10 allowed growth of the ess1H164R strains at restrictive temperature (37°) and of the ess1 strain at temperatures up to 34°.

The srb10 mutation also suppressed a complete deletion of ESS1. This was shown by deleting SRB10 in a haploid ess1 strain bearing a pCaESS1 plasmid and then curing cells of the plasmid (data not shown). These srb10 ess1 cells grew at 30°, 32°, and 34° (Fig 5). However, due to the possibility that suppressors might have arisen during the plasmid-loss procedure, we confirmed this result using standard segregation analysis with diploid cells of the following genotype: srb10/SRB10 ess1/ESS1. Suppression of the ess1 mutation by the unlinked srb10 mutation should alter the normal 2:2 viable:inviable segregation pattern observed for ESS1 disruption (HANES et al. 1989 ), generating 3:1 and 4:0 segregation patterns and allowing recovery of viable His⁺ segregants (i.e., that contain the ess1 deletion). Indeed, 3:1 and 4:0 segregation was observed (Table 3), indicating that srb10 suppresses the ess1 mutation. As expected, all the viable His⁺ segregants (ess1::HIS3) obtained were also Trp⁺, indicating that the srb10 mutation (srb10::TRP1) was also present. Viability was lower at 30° than at 25°, consistent with our observations that the requirement for ESS1 is stricter at higher temperatures (X. WU, C. B. WILCOX and S. D. HANES, unpublished results).

Of the 57 double-mutant tetrads dissected, however, only 16 showed 3:1 segregation (28%), and 3 showed 4:0 segregation (5%) rather than the expected 67 and 17%, respectively. In addition, of the Trp⁺ segregants (srb10::TRP1), only 23% (rather than 50%) were His⁺ (ess1::HIS3). Random spore inviability may not be the cause, since neither ess1/ESS1 nor srb10/SRB10 single mutants showed spore-viability problems (Table 3). These results indicate that another gene might be required or that suppression by srb10 is not fully penetrant. It is also possible that ess1 srb10 double-mutant spores have germination defects. In any case, the results show that, in certain backgrounds, the deletion of SRB10 can relieve cells of their requirement for ESS1. Given that Srb10 inhibits PIC formation, it seems likely that Ess1 stimulates PIC formation.

While deletion of SRB10 suppressed the phenotype of ess1 mutants, deletion of ESS1 did not suppress the srb10 mutant phenotype, that of slow growth on galactose-containing medium. If anything, ess1 srb10 double mutants grew slower than srb10 single mutants on rich or synthetic media containing galactose as the carbon source (data not shown). We also tested whether high-copy expression of SRB10 enhanced or suppressed growth defects caused by ess1^ts mutations at restrictive temperature. No effects were detected (data not shown), as might be expected if Ess1 acts downstream of Srb10 (see DISCUSSION).

A CTD half-substitution allele that suppresses ess1 mutants:
The above results revealed genetic interactions between ESS1 and genes encoding CTD-modifying enzymes. This prompted us to investigate possible direct genetic interactions between ESS1 and the CTD. In otherwise wild-type yeast, the CTD can be truncated to 10 YSPTSPS repeats with no apparent effect on growth, and 8 repeats is sufficient for viability at 30°, but cells are cold sensitive, while mutants bearing more severe CTD truncation alleles are inviable at any temperature (NONET and YOUNG 1989 ; WEST and CORDEN 1995 ). These defects are suppressed, to a certain extent, by mutations in suppressor of RNA pol B (SRB) genes (NONET and YOUNG 1989 ), including SRB10 (THOMPSON et al. 1993 ; LIAO et al. 1995 ; HENGARTNER et al. 1998 ). Phenotypically, deletion of SRB10 "lengthens" the CTD by several repeats, augmenting Rpb1 function (HENGARTNER et al. 1998 ). Our previous work suggested that ess1 mutations do just the opposite, compromising Rpb1 function and sensitizing cells to the effects of truncated CTD alleles, i.e., effectively "shortening" the CTD (WU et al. 2000 ).

Here, we examined genetic interactions between ess1 mutations and various "half-substitution" CTD truncation/subsitution alleles. The CTDs are truncated so they contain only 10–14 heptad repeats rather than the usual 26–27, and they carry substitutions of Ser2 or Ser5 within the repeat to either Ala or Glu. However, these substitutions are restricted to either the "first half" or the "second half" (amino- or carboxy-terminal ends) of the CTD (see Table 4). In this way, we hoped to delineate the importance of first-half vs. second-half CTD repeats, as well as distinguishing the effects of Ser2 vs. Ser5 substitutions. Previous analysis of CTD half-substitution mutants indicated that serines at the same position (e.g., Ser5) within different heptad repeats may have distinct roles (WEST and CORDEN 1995 ; FONG and BENTLEY 2001 ). This is supported by the finding that suppressors of second-half mutations do not necessarily suppress first-half mutations (YURYEV and CORDEN 1996 ). Here, we tested whether the CTD half-substitution mutants could replace wild-type RPB1 in an ess1^ts background.

Plasmids expressing a CTD half mutant, wild-type RPB1, or an empty vector were transformed into rpb1 ESS1 and rpb1 ess1^H164R strains carrying RPB1 on a URA3-containing plasmid. Complementation of the rpb1 mutation was measured by patching cells to 5-FOA medium to detect RPB1 plasmid loss (Fig 6). This experiment allowed us to compare the ability of different CTD alleles to function in ESS1 vs. ess1^ts cells at the permissive temperature (30°) to identify possible synthetic-lethal interactions.

View larger version (61K): [in this window] [in a new window] [Download PPT slide]   Figure 6. Synthetic-lethal interactions between ess1H164R and specific CTD mutations. The indicated plasmids (rows 1–9) were transformed into ESS1 rpb1 pRPB1 or ess1H164R rpb1 pRPB1 strains. Three individual transformants for each plasmid were replica plated to leu– control medium and to 5-FOA medium to detect loss of the pRPB1 plasmid (2µ, URA3). Cells were incubated at 30° for 4 days. Three CTD mutant plasmids were unable to complement rpb1 in the ess1H164R mutant strain as detected by the inability to grow on 5-FOA vs. the control leu– medium (right side; rows 3, 7, and 9). Failure to grow on 5-FOA medium indicates synthetic lethality between these three mutations and ess1H164R.

Consistent with published observations (WEST and CORDEN 1995 ), all four of the serine-to-alanine mutants substituted for RPB1 in the control ESS1 strain (Fig 6, bottom left, rows 3, 4, 7, and 8). In the ess1^H164R mutant strain, however, only the Ser5-to-alanine substitutions complemented the rpb1 mutation, allowing cell growth (bottom right, rows 4 and 8), whereas Ser2-to-alanine substitutions did not (rows 3 and 7). These results identify a synthetic-lethal interaction between ess1^H164R and Ser2, but not Ser5 mutations. Similar results were obtained with plasmid-loss experiments using liquid cultures (data not shown). The lack of sensitivity between ess1^H164R and the Ser5-to-alanine mutation (i.e., mutating both is no worse than mutating either one alone) points to Ser5 as a possible direct target of Ess1 (see also below).

Mutation of Ser2 or Ser5 to glutamic acid in the first half of the CTD (Fig 6, bottom, rows 5 and 6) did not support growth in either ESS1 or ess1^H164R cells. While the inability of these mutants to complement rpb1 may indicate that dephosphorylation of Ser2 and Ser5 in the first half of the CTD is essential for viability, this result is relatively uninformative with respect to the role of Ess1. However, mutation of Ser2 or Ser5 to glutamic acid in the second half of the CTD (rows 9 and 10) did support growth in the ESS1 cells, but not in ess1^H164R cells (i.e., they are synthetic lethal). This result is consistent with the idea that ess1 mutations sensitize cells to second-half mutations, as if Ess1 and these residues of the CTD function in the same pathway but at different steps.

These results suggest that Ess1 targets a subdomain within the CTD (Ser5, first half). For example, Ess1-dependent isomerization might promote dephosphorylation of Ser5 in the first half of the CTD. If true, then a Ser5-to-alanine substitution (which mimics the dephosphorylated form of Ser5) in the first half of the CTD might relieve the requirement for Ess1, whereas a Ser5-to-alanine substitution in the second half of the CTD would not, nor would a substitution of Ser2 to alanine. This is exactly what we observed (Fig 7); a Ser5 first-half mutation (S5A₅ctd₇) suppressed the temperature-sensitive growth defect of ess1^H164R cells (and ess1^A144T, data not shown) at 34°, whereas a Ser5 second-half mutation (ctd₇S5A₇) and a Ser2 first-half mutation (S2A₄ctd₇) did not. We could not test Ser5-to-glutamic acid substitutions (which mimic the phophorylated forms) because they do not support cell growth in an rpb1 background (Fig 6). The suppression of ess1^ts mutations by S5A₅ctd₇ suggests that Ess1 binding to the CTD promotes dephosphorylation of first-half Ser5 residues, perhaps by isomerization of Ser5-Pro6 dipeptide bonds. This could block the action of CTD kinases on Ser5 or expose phospho-Ser5 to the action of a CTD phosphatase.

View larger version (58K): [in this window] [in a new window] [Download PPT slide]   Figure 7. Ser5 in the first half of the CTD may be the target of Ess1 within the CTD. ESS1 rpb1 or ess1H164R rpb1 strains carrying plasmids (CEN LEU2) encoding wild-type RPB1 or CTD mutants with the indicated serine-to-alanine substitutions were spotted (1:5 serial dilutions) onto YEPD medium and grown for 2 days at 25° or 34°. Ser5-to-alanine substitution in the first half of the CTD (pS5A5ctd7) suppressed the temperature sensitivity of ess1H164R, allowing growth at 34°.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

In this article, we present genetic evidence that Ess1 interacts with all four known CTD kinases, indicating that it is likely to act at multiple stages of the transcription cycle. The clearest genetic interactions indicate that Ess1 opposes the effects of Ctk1 and Srb10. Ess1 may also oppose Kin28 and work positively with Bur1. The types of genetic interactions observed (summarized in Table 5), combined with the known substrate preference of Ess1/Pin1 prolylisomerases for phospho-Ser-Pro motifs, suggest a model for Ess1 function. In this model, Ess1 binds the CTD after the CTD is phosphorylated by Srb10. Ess1 then catalyzes a conformational change in the CTD that promotes dephosphorylation by CTD-specific phosphatases, such as Fcp1 and Ssu72. This dephosphorylation would reverse the negative effects of Srb10 and stimulate PIC formation. Ess1 would also be required later in the transcription process, possibly during promoter clearance (Kin28 step), but more likely for elongation and termination/3'-end formation (see below). Here, Ess1 may act by helping Bur1-dependent elongation and later by antagonizing the effects of Ctk1, which promotes elongation. Thus, our model suggests that Ess1 and CTD kinases work together to coordinate multiple steps in transcription and that both covalent (phosphorylation) and noncovalent (isomerization) modifications of the CTD are crucial to this process.

Ess1 acts positively in transcription initiation:
In this study, we observed genetic interactions between ESS1 and SRB10, which regulates the formation of the preinitiation complex (HENGARTNER et al. 1998 ). Deletion of SRB10 suppressed both ts mutants and a complete deletion of ESS1. Thus, when Srb10's negative effect on PIC formation is removed, Ess1 is no longer needed, suggesting that the normal function of Ess1, which occurs genetically downstream of Srb10, is to overcome the effects of Srb10 and stimulate PIC formation. Because Srb10 phosphorylates the CTD, we presume that Ess1 isomerization promotes CTD dephosphorylation by making the phospho-CTD a better conformational substrate for a CTD-phosphatase. In addition, genetic interactions observed with KIN28 suggest that Ess1 may also play a role later during initiation and promoter clearance, although these interactions appear to indicate a negative role for Ess1 during these stages.

Ess1 may inhibit elongation and promote 3'-end formation:
Previous work has implicated Ess1 in 3'-end pre-mRNA processing by an unknown mechanism (HANI et al. 1995 ). Here, we provide genetic data that might help explain this finding. We found that reducing the dosage of CTK1 suppresses the lethality of ess1^ts mutations at restrictive temperature, indicating that Ess1 and Ctk1 have opposing functions. Ctk1 has been shown to increase elongation efficiency and may play a role in 3'-end processing (LEE and GREENLEAF 1997 ; SKAAR and GREENLEAF 2002 ), presumably by phosphorylating the CTD. Since Ctk1 acts positively in elongation, the results point to a negative role for Ess1 in elongation, consistent with the findings of another recent study (WU et al. 2003 ). We propose the following scenario: Ess1 binds the phospho-CTD on the elongating polymerase and catalyzes an isomerization that in turn promotes dephosphorylation by Fcp1 or Ssu72. By doing so, Ess1 would oppose the pro-elongation action of Ctk1 and perhaps enhance the binding of pre-mRNA processing factors to the CTD, thus stimulating 3'-end processing.

Isomerization by Ess1 likely promotes the dephosphorylation of the first half of the CTD at serine 5:
Results of complementation tests with Ser2 and Ser5 substitutions (Fig 6) and the finding that the CTD mutation, S5A₅ctd₇, suppresses ess1^ts mutants (Fig 7) imply that the mutation of Ser5 to alanine (S5A) in the first half of the CTD compensates for the loss of Ess1. Together with genetic results showing that Ess1 opposes the actions of at least two CTD kinases (Srb10 and Ctk1), these data suggest that one role of Ess1 isomerization is to prevent phosphorylation of Ser5 or to promote its dephosphorylation. Therefore, when Ess1 function is compromised, Ser5 would be inappropriately phosphorylated, preventing PIC formation and, in later steps, interfering with proper elongation and 3'-end formation.

While this model nicely fits our data, Srb10 and Ctk1 have been shown to phosphorylate the CTD on Ser2 to a greater degree than that on Ser5 (HENGARTNER et al. 1998 ; PATTURAJAN et al. 1999 ; CHO et al. 2001 ; RAMANATHAN et al. 2001 ; S. BURATOWSKI and M. KEOGH, personal communication). At this point, it is not clear how to explain this difference. There are many possibilities. One is that these kinases also phosphorylate Ser5 to a minor degree, but that these modifications have large effects on Rpb1 function. A second possibility is that Ess1 acts indirectly to promote dephosphorylation of Ser5 (in the first half of the CTD) and that this mechanism might involve another CTD kinase, such as Kin28 or Bur1, which are known to be important for phosphorylation of Ser5. For example, Ser2 phosphorylation by Srb10 (or Ctk1) might stimulate Ess1 binding to Ser2-Pro3 and subsequent isomerization of the CTD. The resulting conformational change might reduce Ser5 phosphorylation (by Kin28 or Bur1) or increase dephosphorylation by Ssu72 phosphatase, which seems to have a preference for phospho-Ser5 (K. SHANKARLING and M. HAMPSEY, personal communication). This might explain the observation that S2A mutations are synthetic lethal (same pathway, different step) with the ess1^H164R mutation (Fig 6, rows 3 and 7), whereas an S5A mutation suppresses ess1^H164R (Fig 7), suggesting Ser5 might be the functional target.

Finally, it is possible that the phosphorylation/dephosphorylation states of Ser2 and Ser5 are coupled. For example, phosphorylation at Ser2 might stimulate phosphorylation at Ser5, perhaps by a processive mechanism involving one or more kinases. Thus, in the absence of Srb10 or Ctk1 function, Ser2 would not be phosphorylated, causing loss of Ser5 phosphorylation, thereby reducing or eliminating the requirement for Ess1. In this scenario Ess1 would normally act on Ser5-Pro6.

Conclusions:
Ess1 was originally proposed to regulate cell division at mitosis. This was based primarily on the mitotic defects observed in yeast ess1 mutants (HANES et al. 1989 ; LU et al. 1996 ). Its human counterpart, Pin1, was proposed to control mitosis in human cells by isomerization of mitotic phosphoproteins such as Cdc25 (CRENSHAW et al. 1998 ; SHEN et al. 1998 ). However, our previous work (AREVALO-RODRIGUEZ et al. 2000 ; WU et al. 2000 , WU et al. 2001 ) suggested an alternative model, one in which Ess1 (and Pin1) controls the transcription of genes necessary for mitosis, and the mitotic-arrest phenotype in ess1 mutant cells is an indirect consequence of the loss of Ess1 function. Work presented here supports a transcription-based model for mitotic regulation in which Ess1 interacts with the CTD of RNA polymerase II and along with CTD kinases controls the binding of transcription and mRNA processing cofactors. Loss of proper transcriptional control in ess1 mutants may trigger mitotic arrest pathways. We suggest this model might hold true in other organisms, including humans.

	FOOTNOTES

¹ Present address: Magee-Womens Research Institute, Ovarian Cancer Center of Excellence, 204 Craft Ave., Pittsburgh, PA 15213-3054.

	ACKNOWLEDGMENTS

We thank Jeffry Corden, James Dutko, Arno Greenleaf, Joe Heitman, Gregory Prelich, Rod Rothstein, Mark Solomon, Michael Stark, and Richard Zitomer for plasmids and yeast strains. We also thank Jessica Matthias for helping to generate the srb10::TRP1 strains, Marisa Foehr and Danielle Lebrecht for technical assistance, and the Wadsworth Center's Media Facility and Molecular Genetics Core Facility (for oligonucleotides/DNA sequencing). We are grateful to Xiaoyun Wu, Gina Devasahayam, Taryn Phippen, and Randy Morse for helpful discussions and/or reading of the manuscript. This work was supported by a grant from the National Institutes of Health (R01-GM55108) to S.D.H.

Manuscript received August 28, 2003; Accepted for publication January 22, 2004.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ADAMS, A., D. E. GOTTSCHLING, C. A. KAISER and T. STEARNS, 1997 Methods in Yeast Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

AKHTAR, A., G. FAYE, and D. L. BENTLEY, 1996  Distinct activated and non-activated RNA polymerase II complexes in yeast. EMBO J. 15:4654-4664.[Medline]

ARCHAMBAULT, J., R. S. CHAMBERS, M. S. KOBOR, Y. HO, and M. CARTIER et al., 1997  An essential component of a C-terminal domain phosphatase that interacts with transcription factor IIF in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 94:14300-14305.[Abstract/Free Full Text]

ARÉVALO-RODRÍGUEZ, M., M. E. CARDENAS, X. WU, S. D. HANES, and J. HEITMAN, 2000  Cyclophilin A and Ess1 interact with and regulate silencing by the Sin3-Rpd3 histone deacetylase. EMBO J. 19:3739-3749.[CrossRef][Medline]

CHI, Y., M. J. HUDDLESTON, X. ZHANG, R. A. YOUNG, and R. S. ANNAN et al., 2001  Negative regulation of Gcn4 and Msn2 transcription factors by Srb10 cyclin-dependent kinase. Genes Dev. 15:1078-1092.[Abstract/Free Full Text]

CHO, E. J., M. S. KOBOR, M. KIM, J. GREENBLATT, and S. BURATOWSKI, 2001  Opposing effects of Ctk1 kinase and Fcp1 phosphatase at Ser 2 of the RNA polymerase II C-terminal domain. Genes Dev. 15:3319-3329.[Abstract/Free Full Text]

CHRISTIANSON, T. W., R. S. SIKORSKI, M. DANTE, J. H. SHERO, and P. HIETER, 1992  Multifunctional yeast high-copy-number shuttle vectors. Gene 110:119-122.[CrossRef][Medline]

CISMOWSKI, M. J., G. M. LAFF, M. J. SOLOMON, and S. I. REED, 1995  KIN28 encodes a C-terminal domain kinase that controls mRNA transcription in Saccharomyces cerevisiae but lacks cyclin-dependent kinase-activating kinase (CAK) activity. Mol. Cell. Biol. 15:2983-2992.[Abstract]

CONRAD, N. K., S. M. WILSON, E. J. STEINMETZ, M. PATTURAJAN, and D. A. BROW et al., 2000  A yeast heterogeneous nuclear ribonucleoprotein complex associated with RNA polymerase II. Genetics 154:557-571.[Abstract/Free Full Text]

CORDON, J. L., 1990  Tails of RNA polymerase II. Trends Biochem. Sci. 15:383-387.[CrossRef][Medline]

CRENSHAW, D. G., J. YANG, A. R. MEANS, and S. KORNBLUTH, 1998  The mitotic peptidyl-prolyl isomerase, Pin1, interacts with Cdc25 and Plx1. EMBO J. 17:1315-1327.[CrossRef][Medline]

DEVASAHAYAM, G., V. CHATURVEDI, and S. D. HANES, 2002  The Ess1 prolyl isomerase is required for growth and morphogenetic switching in Candida albicans.. Genetics 160:37-48.[Abstract/Free Full Text]

FONG, N. and D. L. BENTLEY, 2001  Capping, splicing, and 3' processing are independently stimulated by RNA polymerase II: different functions for different segments of the CTD. Genes Dev. 15:1783-1795.[Abstract/Free Full Text]

GIAEVER, G., A. M. CHU, L. NI, C. CONNELLY, and L. RILES et al., 2002  Functional profiling of the Saccharomyces cerevisiae genome. Nature 418:387-391.[CrossRef][Medline]

GUTHRIE, C., and G. R. FINK (Editors), 1991 Methods in Enzymology. Academic Press, New York.

HANES, S. D., P. R. SHANK, and K. A. BOSTIAN, 1989  Sequence and mutational analysis of ESS1, a gene essential for growth in Saccharomyces cerevisiae.. Yeast 5:55-72.[CrossRef][Medline]

HANI, J., G. STUMPF, and H. DOMDEY, 1995  PTF1 encodes an essential protein in Saccharomyces cerevisiae, which shows strong homology with a new putative family of PPIases. FEBS Lett. 365:198-202.[CrossRef][Medline]

HENGARTNER, C. J., V. E. MYER, S. M. LIAO, C. J. WILSON, and S. S. KOH et al., 1998  Temporal regulation of RNA polymerase II by Srb10 and Kin28 cyclin-dependent kinases. Mol. Cell 2:43-53.[CrossRef][Medline]

HIROSE, Y. and J. L. MANLEY, 2000  RNA polymerase II and the integration of nuclear events. Genes Dev. 14:1415-1429.[Free Full Text]

HOLSTEGE, F. C., E. G. JENNINGS, J. J. WYRICK, T. I. LEE, and C. J. HENGARTNER et al., 1998  Dissecting the regulatory circuitry of a eukaryotic genome. Cell 95:717-728.[CrossRef][Medline]

KEOGH, M.-C., V. PODOLNY, and S. BURATOWSKI, 2003  Bur1 kinase is required for efficient transcription elongation by RNA polymerase II. Mol. Cell. Biol. 23:7005-7018.[Abstract/Free Full Text]

KIMMELMAN, J., P. KALDIS, C. J. HENGARTNER, G. M. LAFF, and S. S. KOH et al., 1999  Activating phosphorylation of the Kin28p subunit of yeast TFIIH by Cak1p. Mol. Cell. Biol. 19:4774-4787.[Abstract/Free Full Text]

KOBOR, M. S., J. ARCHAMBAULT, W. LESTER, F. C. P. HOLSTEGE, and O. GILEADI et al., 1999  An unusual eukaryotic protein phosphatase required for transcription by RNA polymerase II and CTD dephosphorylation in S. cerevisiae. Mol. Cell 4:55-62.[CrossRef][Medline]

KOMARNITSKY, P., E. J. CHO, and S. BURATOWSKI, 2000  Different phosphorylated forms of RNA polymerase II and associated mRNA processing factors during transcription. Genes Dev. 14:2452-2460.[Abstract/Free Full Text]

LAYBOURN, P. J. and M. E. DAHMUS, 1989  Transcription-dependent structural changes in the C-terminal domain of mammalian RNA polymerase subunit IIa/o. J. Biol. Chem. 264:6693-6698.[Abstract/Free Full Text]

LEE, J. M. and A. L. GREENLEAF, 1991  CTD kinase large subunit is encoded by CTK1, a gene required for normal growth of Saccharomyces cerevisiae.. Gene Exp. 1:149-167.

LEE, J. M. and A. L. GREENLEAF, 1997  Modulation of RNA polymerase II elongation efficiency by C-terminal heptapeptide repeat domain kinase I. J. Biol. Chem. 272:10990-10993.[Abstract/Free Full Text]

LEE, T. I. and R. A. YOUNG, 2000  Transcription of eukaryotic protein-coding genes. Annu. Rev. Genet. 34:77-137.[CrossRef][Medline]

LIAO, S. M., J. ZHANG, D. A. JEFFERY, A. J. KOLESKE, and C. M. THOMPSON et al., 1995  A kinase-cyclin pair in the RNA polymerase II holoenzyme. Nature 374:193-196.[CrossRef][Medline]

LIN, P. S., N. F. MARSHAL, and M. E. DAMUS, 2002  CTD phosphatase: role in RNA polymerase II cycling and the regulation of transcript elongation. Prog. Nucleic Acid Res. Mol. Biol. 72:333-365.[Medline]

LORENZ, M. C. and J. HEITMAN, 1998  Regulators of pseudohyphal differentiation in Saccharomyces cerevisiae identified through multicopy suppressor analysis in ammonium permease mutant strains. Genetics 150:1443-1457.[Abstract/Free Full Text]

LU, K. P., S. D. HANES, and T. HUNTER, 1996  A human peptidyl-prolyl isomerase essential for regulation of mitosis. Nature 380:544-547.[CrossRef][Medline]

MAJELLO, B., G. NAPOLITANO, A. GIORDANO, and L. LANIA, 1999  Transcriptional regulation by targeted recruitment of cyclin-dependent CDK9 kinase in vivo.. Oncogene 18:4598-4605.[CrossRef][Medline]

MORRIS, D. P., H. P. PHATNANI, and A. L. GREENLEAF, 1999  Phospho-carboxyl-terminal domain binding and the role of a prolyl isomerase in pre-mRNA 3'-end formation. J. Biol. Chem. 274:31583-31587.[Abstract/Free Full Text]

MURRAY, S., R. UDUPA, S. YAO, G. HARTZOG, and G. PRELICH, 2001  Phosphorylation of the RNA polymerase II carboxy-terminal domain by the Bur1 cyclin-dependent kinase. Mol. Cell. Biol. 21:4089-4096.[Abstract/Free Full Text]

NONET, M. L. and R. A. YOUNG, 1989  Intragenic and extragenic suppressors of mutations in the heptapeptide repeat domain of Saccharomyces cerevisiae RNA polymerase II. Genetics 123:715-724.[Abstract/Free Full Text]

PATTURAJAN, M., N. K. CONRAD, D. B. BREGMAN, and J. L. CORDEN, 1999  Yeast carboxyl-terminal domain kinase I positively and negatively regulates RNA polymerase II carboxyl-terminal domain phosphorylation. J. Biol. Chem. 274:27823-27828.[Abstract/Free Full Text]

RAMANATHAN, Y., S. M. RAJPARA, S. M. REZA, E. LEES, and S. SHUMAN et al., 2001  Three RNA polymerase II carboxyl-terminal domain kinases display distinct substrate preferences. J. Biol. Chem. 276:10913-10920.[Abstract/Free Full Text]

RODRIGUEZ, C. R., E. J. CHO, M. C. KEOGH, C. L. MOORE, and A. L. GREENLEAF et al., 2000  Kin28, the TFIIH-associated carboxy-terminal domain kinase, facilitates the recruitment of mRNA processing machinery to RNA polymerase II. Mol. Cell. Biol. 20:104-112.[Abstract/Free Full Text]

SCHIENE, C. and G. FISCHER, 2000  Enzymes that catalyse the restructuring of proteins. Curr. Opin. Struct. Biol. 10:40-45.[CrossRef][Medline]

SHEN, M., P. T. STUKENBERG, M. W. KIRSCHNER, and K. P. LU, 1998  The essential mitotic peptidyl-prolyl isomerase Pin1 binds and regulates mitosis-specific phosphoproteins. Genes Dev. 12:706-720.[Abstract/Free Full Text]

SKAAR, D. A. and A. L. GREENLEAF, 2002  The RNA polymerase II CTD kinase CTDK-1 affects pre-mRNA 3' cleavage/polyadenylation through the processing component Pti1p. Mol. Cell 10:1429-1439.[CrossRef][Medline]

THOMPSON, C. M., A. J. KOLESKE, D. M. CHAO, and R. A. YOUNG, 1993  A multisubunit complex associated with the RNA polymerase II CTD and TATA-binding protein in yeast. Cell 73:1361-1375.[CrossRef][Medline]

WACH, A., A. BRACHAT, R. POHLMANN, and P. PHILIPPSEN, 1994  New heterologous modules for classical or PCR-based gene disruptions in Saccharomyces cerevisiae.. Yeast 10:1793-1808