The Bur1/Bur2 Complex Is Required for Histone H2B ...

Molecular Cell, Vol. 20, 589–599, November 23, 2005, Copyright ª2005 by Elsevier Inc.

DOI 10.1016/j.molcel.2005.09.010

The Bur1/Bur2 Complex Is Required for Histone H2B Monoubiquitination by Rad6/Bre1 and Histone Methylation by COMPASS Adam Wood,1 Jessica Schneider,1 Jim Dover,3 Mark Johnston,3 and Ali Shilatifard1,2,* 1 Department of Biochemistry 2 Saint Louis University Cancer Center Saint Louis University School of Medicine 1402 South Grand Boulevard St. Louis, Missouri 63104 3 Department of Genetics Washington University School of Medicine St. Louis, Missouri 63110

Summary To date, several classes of enzymes have been shown to affect transcription by catalyzing the modifications of nucleosomes via methylation. Employing our global proteomic screen, GPS, we have determined that the loss of Bur2, a component of the Bur1/Bur2 cyclindependent protein kinase, results in a decrease in histone H3K4 methylation catalyzed by COMPASS. Furthermore, Bur1/Bur2 is required for histone H2B monoubiquitination by Rad6/Bre1. The effect on histone monoubiquitination and methylation is the result of defective Bur1/Bur2-mediated phosphorylation of Rad6 on its serine residue 120 and proper recruitment of the Paf1 complex to chromatin. We have also demonstrated that serine 120 of Rad6 is required for histone H2B monoubiquitination and the regulation of gene expression in vivo. Our results identify in vivo substrates for Bur1/Bur2, thus linking its role to transcriptional elongation and demonstrating a potential activation mechanism for histone H2B monoubiquitination by the Rad6/Bre1 complex. Introduction The regulation of gene expression by way of nucleosomal modifications is a complex, multistage process requiring the concerted action of many proteins (Gerber and Shilatifard, 2003; Wood et al., 2005). In Saccharomyces cerevisiae, diverse groups of histone modifying enzymes exist either to ‘‘open’’ or ‘‘close’’ the chromatin and thus control access of the transcriptional machinery to the regulatory elements within chromatin. Likewise, transcription factors play pivotal roles by either altering the rate of RNA Polymerase II (Pol II) elongation or mediating interactions between Pol II and the surrounding chromatin. Mutations and/or translocations into the mammalian trithorax-related gene MLL result in the development of hematological malignancies (Tenney and Shilatifard 2005; Hess, 2004). To better define the role of MLL in the regulation of gene expression, we identified the Set1 protein from Saccharomyces cerevisiae as the MLL homolog and biochemically isolated its complex, which we call COMPASS (complex of proteins associated with Set1), as an MLL-related complex (Miller et al., *Correspondence: [email protected]

2001). We and others have demonstrated that the methyltransferase activity of COMPASS is specific for lysine 4 of histone H3 (H3K4), a chromatin mark that is important for both transcriptional activation and gene silencing (Briggs et al., 2001; Bryk et al., 2002; Krogan et al., 2002; Miller et al., 2001; Nislow et al., 1997; Roguev et al., 2001; Santos-Rosa et al., 2002; Gerber and Shilatifard, 2003; Wood et al., 2005). In addition to COMPASS, Dot1p also acts as a methyltransferase catalyzing the methylation of histone H3 at lysine 79 (H3K79), which is required for silencing of genes located near telomeres or within the rDNA repeats as well as the mating type loci (Feng et al., 2002; Lacoste et al., 2002; Ng et al., 2002a, 2003; Singer et al., 1998; van Leeuwen et al., 2002). To gain insight into the molecular machinery required for histone methylation by COMPASS and Dot1p, we exploited a global proteomic screen we call GPS to survey the set of yeast gene deletion mutants (Schneider et al., 2004). One of the first proteins identified in our screen was Rad6, an E2 ubiquitin-conjugating enzyme that had been shown to ubiquitinate histone H2B on lysine 123 (Dover et al., 2002; Gerber and Shilatifard 2003; Robzyk et al., 2000). We and others have since demonstrated that monoubiquitination of histone H2B is required for histone H3 methylation catalyzed by both COMPASS and Dot1p (Dover et al., 2002; Ng et al., 2002b; Sun and Allis, 2002). Almost every E2 ubiquitin-conjugating enzyme requires an E3 ligase for proper substrate selection. Employing GPS, we identified the RING fingercontaining protein Bre1 as the E3 ligase functioning with Rad6 in histone monoubiquitination and telomeric silencing (Wood et al., 2003a). A role in histone monoubiquitination by Rad6 and histone methylation by COMPASS and Dot1p has also been demonstrated for several components of the Paf1 complex (Paf1c), a Pol II elongation factor, thus linking chromatin modifications to transcription elongation (Krogan et al., 2003; Wood et al., 2003b; Gerber and Shilatifard, 2003). However, the molecular mechanism of activation for Rad6 in monoubiquitination of histone H2B has remained elusive. The BUR genes (bypass UAS requirement) were revealed in a genetic screen for genes involved in repressing transcription of SUC2 (Prelich and Winston, 1993). Removal of the SUC2 UAS abolishes SUC2 transcription, causing cells to be unable to grow on media with sucrose as the sole carbon source. Inactivation of any of the six BUR genes suppresses the requirement for the SUC2 UAS and allows growth on sucrose. Genes identified in this screen were shown to have diverse effects on transcriptional regulation. Several of the genes are members of the SPT (suppressor of Ty) class of genes, which encode transcription factors and other components of the transcriptional machinery (Fassler and Winston, 1988; Simchen et al., 1984; Winston et al., 1984, 1987). BUR1 is an essential gene that encodes a Cdc28-related cyclin-dependent protein kinase. BUR2 encodes a divergent cyclin of the cyclin T/cyclin C family that interacts with Bur1 and is required for its protein kinase activity both in vivo and in vitro (Yao et al., 2000). Deletion of BUR2 or mutations affecting the protein kinase activity

Molecular Cell 590

of Bur1 result in several phenotypes, including sensitivity to heat, caffeine, formamide, and 6-azauracil (Keogh et al., 2003; Murray et al., 2001; Prelich and Winston, 1993; Yao et al., 2000; Yao and Prelich, 2002). In addition to the numerous chemical sensitivity phenotypes, cells containing Bur1 and Bur2 mutations are also unable to form spores and display a severe slow-growth phenotype (Yao et al., 2000). Overexpression of Bur1 compensates for several (but not all) of these phenotypes, suggesting that other cyclins may partially compensate for loss of Bur2. It has been proposed that the Bur1/Bur2 complex can impact transcriptional elongation. It was initially postulated that transcription elongation by RNA Pol II was affected through the phosphorylation of the C-terminal domain (CTD) of Pol II by Bur1/Bur2. However, it has recently been proposed that the CTD of Pol II may not be the sole target of Bur1/Bur2, raising the possibility that other substrates of Bur1/Bur2 can account for the role of this protein complex in transcriptional elongation (Keogh et al., 2003; Murray et al., 2001). Therefore, it is possible that additional in vivo substrates for the Bur1/ Bur2 complex could exist that link its activity to elongating Pol II. We report that the Bur1/Bur2 protein kinase affects several histone modifications involved in transcriptional elongation. We have identified serine 120 of Rad6 as a substrate for the kinase activity of this complex. Furthermore, we have also shown that the localization of the Paf1 complex, which is required for recruitment of COMPASS and activation of Rad6/Bre1 function (Krogan et al., 2003; Wood et al., 2003b), is partially regulated by the Bur1/Bur2 complex. Our study has demonstrated that phosphorylation of serine 120 of Rad6 is critical for fully catalyzing monoubiquitination of histone H2B in vivo and, therefore, regulation of histone methylation catalyzed by COMPASS and Dot1p. Results Identification of Bur2 as a Regulator of Histone Methylation By using our GPS, we surveyed the entire yeast gene deletion mutant collection to search for genes involved in COMPASS and Dot1p-mediated histone methylation (Figure 1A). By using antibodies directed against methylated histone H3K4 and methylated histone H3K79, we identified several mutants that appear to be deficient in these modifications. Deletion of BUR2, which encodes a cyclin possibly involved in transcriptional elongation, significantly reduces (but does not abolish) di- and trimethylation of lysine 4 (Figure 1B) and dimethylation H3K79 (Figure 1C). To demonstrate a direct role for Bur2 in this process, we were able to fully restore histone methylation on H3K4 in a bur2 mutant by reintroduction of full-length HA-tagged Bur2 (Figure 1B, lanes 9 and 10). To determine the effect of Bur2 loss on H3K79 dimethylation, we performed a comprehensive titration analysis of extracts from strains with BUR2 deletions and analyzed the extracts for the level of modified H3K79. We also used antibodies against acetylated histone H3 as our load control. On average, the deletion of BUR2 results in about 40% reduction in the level of dimethylated H3K79 (Figure 1C). Via chromatin immuno-

precipitation (ChIP) analysis, it has recently been reported that the loss of histone H2B monoubiquitination results in a severe reduction in the level of histone H3K79 trimethylation and a lowering in the level of H3K79 dimethylation by Dot1p. However, the level of H3K79 monomethylation is not affected by the loss of histone H2B monoubiquitination (Shahbazian et al., 2005). We have also demonstrated that the level of H3K4 monomethylation by COMPASS is not affected, but di- and trimethylation are reduced as the result of the loss of H2B monoubiquitination (Schneider et al., 2005). Therefore, it appears that the level of monomethylation of H3K4 and H3K79 are directly regulated by the enzymatic activities of COMPASS and Dot1p, respectively, and that monoubiquitination of histone H2B perhaps controls processive di- and trimethylation (Schneider et al., 2005; Shahbazian et al., 2005). We were also able to show that the histone H3 methylation phenotype genetically segregated with kanamycin resistance in a cross of the bur2D::kanMX mutant to a BUR2 (wild-type) strain, demonstrating that genetically bur2D is responsible for this observed methylation phenotype (Figure 2A, lanes 4–7). Histone H2B Monoubiquitination Is Affected by the Loss of Bur2 Because monoubiquitination of histone H2B is required for histone H3 methylation by COMPASS and Dot1p (Gerber and Shilatifard 2003; Dover et al., 2002; Ng et al., 2002b), and because deletion of BUR2 results in the reduction of both COMPASS and Dot1p-mediated histone H3 methylation, we entertained the possibility that Bur2 may be involved in the regulation of histone H2B monoubiquitination. To test this idea, we mated a strain whose sole copy of the gene encoding histone H2B is tagged at its N terminus with the FLAG epitope to a bur2::KanMX strain. The resulting diploid was sporulated, and histone H2B monoubiquitination was scored in the spores of the resulting tetrads. In the kanamycinresistant spore clones, histone H2B monoubiquitination was significantly reduced (Figure 2A, lanes 4 and 5). As expected, in spores in which bur2::KanMX segregated away from FLAG-H2B, the level of histone H2B monoubiqutination was not altered (Figure 2B, lane3). Overall, our data indicate that global histone monoubiquitination in the bur2::kanMX cells is reduced by as much as 70% compared to wild-type (wt). This reduction in histone H2B monoubiquitination, in turn, leads to a global reduction in histone H3 methylation on lysines 4 and 79 (Figures 1B, 1C, and 2B). To determine whether BUR1 (the kinase in the Bur1/Bur2 complex) is also involved in this process, we set out to determine the effect of BUR1 loss on histone H2B monoubiquitination and H3 methylation. Our data indicate that similar to the loss of BUR2, deletion of BUR1 also results in reduction of the level of histone H2B monoubiquitination and H3 methylation (data not shown). Thus, the Bur1/Bur2 complex is involved in Rad6/Bre1-mediated histone monoubiquitination and histone methylation by COMPASS. The Loss of Bur2 Does Not Affect the Expression of Known Factors Involved in the Histone H2B Monoubiquitination Pathway We previously demonstrated that several factors, such as Rad6, Bre1, and the components of the Paf1 complex,

The Role of Bur1/Bur2 in Histone Modifications 591

Figure 1. Deletion of BUR2 Results in Reduction of the Levels of Histone H3 Lysine 4 and 79 Methylation (A) Extracts prepared from plates containing single deletion mutants from the nonessential deletion consortium (lanes A–D) were subjected to SDS-PAGE and Western analysis by using antibodies directed against dimethylated lysine 4 and lysine 79 of histone H3. Blue arrows indicate the lane containing the BUR2 deletion mutant. Several other sites that are null for histone methylation represent slow-growing strains or plate markers. (B) Deletion of BUR2 results in a decrease in di- and trimethylation of H3K4 (lanes 3 and 4). After a CEN-URA3 plasmid containing HA-tagged Bur2 was introduced into the BUR2 deletion mutant, histone H3K4 methylation was restored to levels comparable to wild-type (wt) (lanes 9 and 10). (C) The loss of BUR2 results in reduction of the level of histone H3K79 methylation by Dot1p. A titration study was performed by the analysis of bur2 null strain extracts. Each extract was analyzed by its application to SDS-PAGE and tested for the presence of modified H3K79 by using appropriate antibodies. As load controls, antibodies against either unmodified or modified histone H3 and actin were also used.

are required for proper monoubiquitination of histone H2B (Dover et al., 2002; Wood et al., 2003a, 2003b; Krogan et al., 2003; Gerber and Shilatifard, 2003). The Bur1/ Bur2 complex is known to be involved in transcriptional regulation. In order to demonstrate that the loss of Bur2 does not affect the expression of these genes, we performed RT-PCR in wt and bur2 null strains and tested for the presence of transcripts for these genes. As shown in Figure 3A (top), the loss of Bur2 does not affect the

expression of these genes. Furthermore, our biochemical studies have demonstrated that the stability of Rad6 protein is also not altered in a bur2 null strain (Figure 3A, bottom). Rad6 Is a Phosphoprotein and Can Be Phosphorylated by the Bur1/Bur2 Kinase Rad6 obtained from wt strains migrates in twodimensional (2D) gel electrophoresis in multiple forms

Molecular Cell 592

Figure 2. The Loss of Bur1/Bur2 Results in the Reduction in the Levels of Histone H2B Monoubiquitination (A) A strain carrying a plasmid bearing a single copy of histone H2B FLAG tagged at its N terminus was mated to a bur2 null strain, and the resulting viable tetrads were dissected. Because the KanMX cassette was used to disrupt the BUR2 open reading frame, strains carrying the bur2 deletion will be resistant to G418. The top panel displays the growth analysis of the mating cross between a bur2 null and wt strain. Growth phenotypes associated with Bur2 mutations migrate with kanamycin resistance (lanes A1 and A2, top). Strains produced from the cross that are not resistant to G418 are wt progeny (lanes A3 and A4). Protein extracts were also prepared from the parental strains and each of the four strains resulting from the tetrad dissection. After SDS-PAGE, the blots were probed with antisera against the FLAG epitope, dimethylated H3K4, trimethylated H3K4, and acetylated histone H3 (anti-FLAG antibody was used to visualize the amount of monoubiquitinated histone H2B). (B) Comparative analysis between the bur2 mutant and wt progeny of a separate mating cross. Two spores, one bearing the KanMX::bur2 deletion (B1) and one containing the wt BUR2 gene (B2), were also analyzed for defects in histone H2B monoubiquitination and histone H3K4 methylation.

(Figure 3B, top). To demonstrate whether such a pattern of migration for Rad6 is due to its phosphorylation, we treated Rad6 with phosphatase and subjected the resulting mixture to 2D gel electrophoresis. As shown in Figure 3B (bottom), treatment of Rad6 with phosphatase results in an alteration of its mobility on 2D gel electrophoresis toward an unmodified Rad6. To determine whether Rad6 can be phosphorylated in vitro by Bur1/ Bur2 kinase, we affinity purified Bur1, Bur2, and Rad6TAP (Figure 3C), using them to perform an in vitro kinase assay (Figure 3D). Indeed, affinity-purified Bur1/Bur2 complex catalyzes phosphorylation of Rad6-TAP (Figure 3D, lanes 3–5 and 8–10), resulting in a band that migrates at the same molecular weight as Rad6-TAP in SDS-PAGE that is sensitive to phosphatase treatment (Figure 4B, lane 2). Because Rad6 is a phosphoprotein in vivo and is phosphorylated by Bur1/Bur2 in vitro, it is likely that Rad6 may be an in vivo target of the Bur1/ Bur2 protein kinase activity. Rad6 Serine 120 Is Required for Rad6 Function We set out to determine the amino acid residue within Rad6 that is modified by Bur1/Bur2. It has been demonstrated that the Bur1/Bur2 complex can catalyze phosphorylation of serine 5 of the heptapeptide repeats (YSPTSPS) of Pol II CTD in vitro (Keogh et al., 2003; Murray et al., 2001); therefore, the consensus sequence PxSP is a plausible site of phosphorylation by Bur1/ Bur2. One of the several evolutionarily conserved serine

residues of Rad6 is within such a sequence (Figure 4A, top). This residue is in close proximity to the active site cysteine (C88) of Rad6 and could play a role in substrate recognition and/or ubiquitin transfer by Rad6 (Worthylake et al., 1998). Our studies have indicated that serine 120 of Rad6 is a substrate for Bur1/Bur2 kinase activity in vitro. In our in vitro assay, the mutation of serine 120 of Rad6 to alanine results in a substantial reduction in phosphorylation of Rad6 by Bur1/Bur2 kinase (Figure 4A, bottom, and Figure 4B). However, to determine whether serine 120 of Rad6 is a substrate for Bur1/Bur2 in vivo, we tested strains harboring a Rad6 (S120A) mutation for their ability to monoubiquitinate histone H2B. Our study demonstrates that this point mutation results in a significant reduction in the levels of histone H2B monoubiquitination in vivo, similar to that of a bur2D mutant (compare Figures 2 and 4C). Consistent with the notion that Rad6 is a phosphoprotein in vivo, mutation of serine 120 of Rad6 to alanine or the loss of BUR2 both result in a change in the abundance of the phosphorylated form of Rad6 on 2D gel electrophoresis, similar to that of Rad6 phosphatase treated sample (please compare Figures 3B and 4D). We wanted to further investigate whether serine 120 of Rad6 is required for its localization to chromatin. In doing so, we preformed ChIP analysis with wt Rad6TAP and Rad6-TAP(S120A) mutant. The Rad6 (S120A) mutant is still localized to chromatin and demonstrates


Figure 3. The Bur1/Bur2 Complex Can Phosphorylate Rad6 (A) The mRNA levels of the several known genes involved in histone H2B monoubiquitination were measured by using RT-PCR in wt and bur2 null strains with primers generated against the indicated genes (top). When BUR2 is deleted, Rad6 protein expression is not affected (bottom). These data indicate that the expression or stability of Rad6 is not altered in a bur2 null strain. (B) Rad6 exists as a phosphoprotein in vivo. Lysates from a wt strain containing TAPtagged Rad6 were subjected to 2D gel analysis and probed with antibody directed against Rad6-TAP. Addition of phosphate groups causes Rad6 to migrate toward the anode (positive electrode). Treatment of the extracts with potato acid phosphatase (Sigma) results in a loss of the phosphorylated forms of Rad6 (bottom). A red dot denotes the free form of Rad6, and the asterisks indicate crossreacting proteins from the phosphatase mixture. (C) In order to ensure the purity of the purified Rad6 substrate used in the in vitro kinase assays and that no other contaminating proteins migrate with Rad6 in the Bur1 or Bur2 purification, a fraction of the TAP-purified proteins was resolved by SDS-PAGE and was silver stained. (D) TAP-tag-purified Bur1, Bur2, and Rad6 from (C) were used to reconstitute an in vitro kinase assay. Increasing amounts of either Rad6 (lanes 3–5) or Bur1/Bur2 (lanes 8–10) were titrated into the reaction as described in the Experimental Procedures. Individual reactions were resolved by SDS-PAGE, and kinase activity was determined by using a phosphoimager plate (Molecular Dynamics).

only a marginal reduction in its localization (Figure 4E) that is within range of the wt. Loss of factors such as Bre1, which is required for recruitment of Rad6 to chromatin, results in the total abolishment of localization of Rad6 to chromatin (Wood et al., 2003a). To determine whether the loss of BUR2 has any effect on localization of Rad6 to chromatin, similar ChIP analyses were performed with Rad6-TAP in wt and bur2 null backgrounds. Our data indicate that the loss of Bur2 does not significantly alter the recruitment of Rad6 to chromatin on PMA1 gene (Figure 4E). In a bur2 null background, it is possible that the basal Bur1 kinase activity is still functional and can, perhaps, partially phosphorylate Rad6. Therefore, the Rad6 (S120A) mutants demonstrated a stronger effect on their localization to chromatin. Nevertheless, both Rad6 (S120A) and Rad6 in bur2 null strains are still localized to chromatin. Collectively, our data indicate that phosphorylation of serine 120 of Rad6 is required for the activation of its catalytic activity in monoubiquitination of histone H2B. Bur2 Is Required for Localization of the Paf1 Complex to Chromatin Although the mutation of serine 120 of Rad6 to alanine results in the loss of histone H2B monoubiquitination in vivo (Figure 4C), our genetic and biochemical studies demonstrate that the histone methylation properties of bur2 null strains and Rad6 (S120A) are a bit different

(data not shown). Because we previously demonstrated that the Paf1 complex is required for the activation of Rad6 and recruitment of COMPASS to the transcribing Pol II (Wood et al., 2003b; Krogan et al., 2003), we have tested a possible role for Bur2 in the recruitment of the Paf1 complex to chromatin. Our ChIP studies have demonstrated that, in addition to regulation of phosphorylation of Rad6, Bur2 is also partially required for proper localization of the Paf1 complex to chromatin (Figure 5B). However, it does not appear that the Paf1 complex is required for the localization of Bur2 to chromatin (Figure 4A). The effect on the localization of the Paf1 complex observed on the PYK1 gene in the absence of Bur2 is much more exacerbated than that observed on the ADH1 gene. This may indicate that Bur1/Bur2 could function in the localization of the Paf1 complex on selected regions of chromatin or that the localization of the complex may depend on the level of transcriptional activity of a given gene, the chromosomal context, or environmental factors. A complete ChIP-Chip analysis on the localization of the Paf1 complex throughout the genome in the presence and absence of Bur2 under different environmental conditions should address this question. Furthermore, it is not clear at this time whether the localization of Paf1 complex in the absence of Bur2 is a direct effect of the Bur2 loss. Given the fact that bur2 null cells have slower growth phenotypes due to the reduction in density of

Molecular Cell 594

Figure 4. Mutation of Rad6 on Serine 120 Reduces Histone H2B Ubiquitination In Vivo (A) Serine 120 of Rad6 is an in vitro substrate for Bur1/Bur2 kinase. The top panel demonstrates the sequence alignment of Rad6 and its homologs ranging from yeast to humans. Residues colored with red are completely conserved in all species. The asterisk denotes the position of serine 120. TAP-purified Bur1/Bur2 was used to reconstitute an in vitro kinase assay. Rad6-TAP or Rad6-TAP carrying the S120A mutation (3C) was purified and used as substrates for the assay. Kinase and substrates were either incubated alone (lanes 1–3) or in combination (lanes 4 and 5). The individual assays were resolved by SDS-PAGE, and kinase activity was determined by using a phosphoimager screen after 6 hr of exposure. (B) An additional kinase assay was run by using the identical conditions, and one reaction containing wt Rad6 was then incubated with phosphatase (middle lane). (C) Strains carrying plasmids (URA3 CEN) with either the TAP-tagged Rad6 or Rad6 (S120A) (pRAD6-TAP) or (pS120A-TAP) were assayed for their ability to compliment histone H2B monoubiquitination (lanes 3 and 4). The parental strain contains a sole copy of FLAG-tagged histone H2B (lane 1). When RAD6 gene is deleted, the level of the slow-migrating band (ubH2B) is significantly reduced. Wt Rad6 can compliment


Figure 5. The Localization of the Paf1 Complex Is Partially Reduced When BUR2 Is Deleted (A) ChIPs were performed to determine if the localization of Bur2 to chromatin on constitutive genes such as ADH1 and PYK1 is altered when the Paf1 complex is dissociated (Paf1 null). (B) The localization of Paf1 to the same regions of chromatin was determined in a bur2 deletion background following the ChIP method as described in Figure 4. In (A) and (B), error bars represent the standard deviation from the mean. (C) The transcription of BUR2, ADH1, PYK1, and actin genes were measured by using RT-PCR with primers generated against the indicated genes in wt and BUR2 strains.

Pol II and lower overall transcription, we set out to determine whether deletion of BUR2 has any effect on ADH1 or PYK1 gene expression when the level of mRNA is normalized in both backgrounds. Our RT-PCR studies (Figure 5C) indicate that the loss of Bur2 does not significantly affect the transcription of ADH1 or PYK1 genes. Serine 120 of Rad6 Is Required for Cellular Growth To determine the effect of mutation at serine 120 of Rad6 on cellular growth under different conditions, we ana-

lyzed the growth phenotypes of RAD6 and BUR2 deletions and/or mutants as well as other mutants previously shown to affect histone H2B monoubiquitination. The Rad6 (S120A) mutation and bur2 null cells demonstrate sensitivity to hydroxyurea (HU), with the bur2 null strain exhibiting the stronger phenotype on HU (Figure 6A). The rad6D, bre1D, and rtf1D (one of the components of the Paf1 complex) mutants also demonstrate growth phenotypes similar to Rad6 (S120A) mutants on HU (Figure 6A). This suggests that serine 120 of Rad6 plays a role in the

the monoubiquitination phenotype, but the S120A substitution results in a substantial decrease in histone H2B monoubiquitination when compared to wt (compare lanes 3 and 4). (D) Two-dimensional gel analysis of Rad6-TAP, Rad6-TAP in a bur2 null strain, and Rad6-TAP mutated at serine residue 120 (S120A) in a wt background. When either the BUR2 gene is deleted or serine 120 of Rad6 is mutated to alanine, the abundance of phosphorylated Rad6 shifts drastically toward the unmodified form. (E and F) ChIP demonstrates that the S120A substitution or the deletion of BUR2 does not significantly alter the localization of Rad6 to chromatin on the PMA1 gene. To monitor the presence of Rad6 on chromatin at the PMA1 gene, chromatin was immunoprecipitated with rabbit IgG-agarose from a strain containing either wt Rad6-TAP or Rad6-TAP(S120A). (F) ChIP analyses for Rad6 were also performed in wt or bur2 null backgrounds. PCR amplifications were carried out by using primer pairs recognizing the early transcribed region of the PMA1 gene. Each PCR reaction contained a second primer pair that amplified a region of chromosome V devoid of ORFs, thus providing an internal control for background. The ratio of the experimental to the control signal for the precipitated DNA was divided by the ratio of the experimental to the control signal for the input DNA. Error bars represent the standard deviation from the mean.

Molecular Cell 596

Figure 6. Serine 120 of Rad6 Is Required for Several Functions In Vivo (A) To determine if Rad6 serine 120 is required for several known functions of Rad6, we tested the Rad6 (S120A) mutant alongside other strains deleted for known components of the histone H2B monoubiquitination machinery to assay phenotypic similarities. Cells were tested for sensitivity to HU, formamide, and heat shock at 37ºC. (B) Serine residue 120 of Rad6 is also required for telomeric silencing. Components of the machinery required for proper histone H2B monoubiquitination were knocked out in a strain with the URA3 reporter inserted into the telomere of chromosome 7. Either wt or Rad6 (S120A) was tested for its ability to grow on 5FOA.

same pathway as Bre1 and the Rtf1 component of the Paf1 complex: regulation of histone H2B monoubiquitination. However, the Rad6 (S120A) mutant or strains deleted for other known components of the histone H2B monoubiquitination pathway (such as Bre1 and Rtf1) are not sensitive to formamide-like bur2 null cells (Figure 6A). Our results, taken together, suggest a common in vivo role for serine 120 of Rad6 and factors involved in the regulation of histone H2B monoubiquitination. Point mutation of Rad6 on serine 120 does not result in sensitivity to heat shock observed in rad6 null strains (Figure 6A). Furthermore, a rad6D mutant is much more sensitive to formamide than the Rad6 (S120A) mutant. These observations, in addition to the observation on the stability of Rad6 in a bur2 null background (Figure 3A) and the stability of Rad6 (S120A) mutants (Figure 4C), collectively indicate that the mutation of serine 120 of Rad6 does not result in the overall loss of stability or the enzymatic activity of Rad6. Rad6 Serine 120 Is Required for Silencing of Telomeric-Associated Genes Histone H3 methylation by COMPASS and Dot1p and monoubiquitination of histone H2B are linked to silencing of genes located near telomeres (Huang et al., 1997; Krogan et al., 2003; Nislow et al., 1997; Singer et al., 1998; Wood et al., 2003a). To determine if mutation of serine 120 of Rad6 plays a role in Rad6-dependent regulation of telomeric silencing, we scored expression of a URA3 gene located near the telomere of chromosome VII (telVII::URA3) (Figure 6B). In wt cells, URA3 expression is repressed. Mutations that disrupt silencing allow expression of URA3, which leads to sensitivity to 5FOA (Wood et al., 2003a). Mutation of Rad6 serine 120 to alanine leads to a loss of URA3 silencing comparable to that seen in mutants deleted for RAD6, BRE1, or RTF1, all of which are required for histone H2B monoubiquitination and histone H3 methylation by COMPASS and Dot1p. Discussion Activation of transcription is dependent on the proper recruitment and activation of chromatin remodeling fac-

tors. In yeast, the Rad6/Bre1 complex is required for proper regulation of histone H2B monoubiquitination and for histone H3 methylation by COMPASS and Dot1p (Dover et al., 2002; Ng et al., 2002b; Robzyk et al., 2000). Furthermore, the involvement of the Paf1 complex in the regulation of histone H2B monoubiquitination via Rad6 and association of COMPASS with the elongating form of Pol II both link posttranslational modification of histones to transcriptional elongation (Gerber and Shilatifard, 2003). The Bur1/Bur2 complex is a cyclin-dependent protein kinase involved in response of yeast cells to pheromone, cell cycle regulation, and regulation of transcription. It has been proposed that Bur1/Bur2 is involved in the regulation of transcriptional elongation. Although the BUR genes were identified many years ago (Prelich and Winston, 1993), surprisingly little is known about the roles of Bur1 and Bur2. It has been suggested that, like several related cyclin-dependent protein kinases in yeast, Bur1/ Bur2 activity is directed toward the heptapeptide repeats of the RNA Pol II CTD (Murray et al., 2001). However, it appears that the Bur1/Bur2 complex does not contribute a significant amount to CTD phosphorylation in vivo, raising the possibility that there are other substrates for Bur1/Bur2 linking its activity to transcription elongation (Keogh et al., 2003). Because Bur1/Bur2 has an effect on overall transcription, it seems reasonable that the target(s) of this complex would also have defined roles in transcriptional regulation. Our results suggest that one role for Bur1/Bur2 is the activation of Rad6 by phosphorylation. Not only might this stimulate Rad6 activity, but it could also serve as a regulatory switch for both Dot1p and COMPASS, which have been linked to transcriptional elongation control (Krogan et al., 2003; Gerber and Shilatifard 2003). Based on the severity of bur1 mutant phenotypes and the extent of histone methylation in bur2 null cells when compared to Rad6 (S120A), it is likely that Rad6 is not the only in vivo substrate for Bur1/Bur2. It is possible that other transcription-associated factors such as the Paf1 complex, components of COMPASS, Set2, and other histone modifying enzymes (or elongation factors) could serve as substrates of Bur1/Bur2 as well.


Figure 7. The Bur1/Bur2 Kinase Complex Is Required for Partial Recruitment of the Paf1 Complex and the Activation of Rad6 Activity via Its Phosphorylation These activities of the Bur1/Bur2 complex are required for proper monoubiquitination of histone H2B and methylation of histone H3 by COMPASS and Dot1p linking Bur1/Bur2 complex to transcriptional elongation.

In addition to its effects on transcriptional initiation, monoubiquitination of histone H2B by Rad6/Bre1 also plays a role in silencing gene expression. We propose that during the initiation stage of transcription, Rad6 is recruited to active genes via its E3 ligase, Bre1 (Kao et al., 2004; Wood et al., 2003a). The RNA Pol II-associated Paf1 complex could then allow for an interaction between RNA Pol II, Rad6, and the histone methyltransferase COMPASS (Krogan et al., 2003; Wood et al., 2003b) (Figure 7). The activation of histone H2B monoubiquitination by Bur1/Bur2 may provide for the beginning of transcriptional elongation, because mutations affecting histone H2B monoubiquitination drastically alter the kinetics of initiation (Kao et al., 2004). The removal of the ubiquitin moiety from histone H2B has been implicated as yet another crucial step in the process of transcriptional regulation. This reversal of monoubiquitination is catalyzed by the SAGA subunit Ubp8, and the overall increase in histone H2B ubiquitination caused by Ubp8 deletion results in decreased transcription of SAGA-regulated genes (Daniel et al., 2004; Henry et al., 2003). In this regard, the Bur1/Bur2 kinase could be responsible for moderating several chromatin-associated factors capable of activating and/or repressing transcription based on the audience of transcription factors gathered during the early stages of initiation. We have presented evidence that the Bur1/Bur2 cyclin-dependent protein kinase is required for activation of Rad6 in monoubiquitination of histone H2B. This, in turn, negatively affects the accumulation of both Dot1p and COMPASS-mediated histone methylation. We have demonstrated that (1) deletion of the genes encoding Bur1 and its cyclin Bur2 results in reduction in the amount of di- and trimethylated H3K4 and dimethylated H3K79 as well as significant reduction in the level of monoubiquitinated histone H2B; (2) the loss of Bur2 results in a partial reduction in recruitment of the Paf1 complex to chromatin; (3) Rad6 can be phosphorylated by Bur1/Bur2 in vitro, and the Bur1/Bur2 complex localizes with Rad6 on chromatin; (4) the phosphorylation of serine 120 of Rad6 catalyzed by Bur1/Bur2 is required for histone H2B monoubiquitination in vivo; (5) mutation of serine 120 of Rad6 results in phenotypes similar to those caused by the loss of histone H2B monoubiquitination; and (6) similar to Bre1, Rad6, and the components of

the Paf1 complex, serine 120 of Rad6 is required for proper regulation of telemeric silencing. Collectively, our data suggest that monoubiquitination of histone H2B via Rad6 is regulated by the Bur1/Bur2 complex and identifies a long sought after in vivo substrate of the Bur1/Bur2 complex that links its activity to transcriptional elongation. Experimental Procedures GPS Analysis of the deletion mutant consortium was carried out as previously described (Schneider et al., 2004). Yeast Strains, Plasmids, and Media Used Plasmid pGP466 (CEN URA3 3XHA-BUR2) was obtained from G. Prelich. Plasmids pRAD6 and pS120A (CEN URA3 RAD6) were made by inserting the wt or S120A mutant allele of Rad6 into pRS316 and PRS315. All media and plates, including YPD (rich media) and YPD + G418, were made as previously described (Wood et al., 2003a). Synthetic dropout plates were made according to the specifications of the manufacturer (Qbiogene). All plates used in chemical sensitivity assays were made as previously described (Wood et al., 2003a). Isolation of Yeast Extracts at Larger Quantities To obtain yeast cell extracts in larger quantities, the cells were grown to midlog phase in YPD medium, pelleted, washed with distilled water, and resuspended in lysis buffer (20 mM Tris [pH 7.5], 50 mM KCl, 1 mM EDTA, 0.1% NP40, and 1 mM DTT). Cells were then disrupted by vortexing with glass beads (0.5 mm, Biospec Products) for 15 min at 4ºC. The bottoms of the microcentrifuge tubes were punctured, and cell extracts were recovered in a larger tube by brief centrifugation. The pellet was resuspended in water and 4X Laemmli loading buffer before being briefly vortexed and heated at 95ºC for 5 min. The protein extracts were resolved by SDS-PAGE and subjected to Western analysis by using antisera specific for mono-, di-, or trimethylated H3K4, monoubiquitinated histone H2B (using anti-FLAG), and unmodified histone H3 and/or actin as load controls. Protein Purification and Kinase Aassay Yeast cells containing either Rad6-TAP, Bur1-TAP, or Bur2-TAP were inoculated into 1.5 L of YPD and grown to an OD600 of about 1.2. The cells were harvested and extracts were prepared by freezing the cell pellets, pulverizing them with a coffee grinder, and resuspending the powder in an equal volume of calmodulin binding buffer (50 mM HEPES [pH 7.5], 150 mM NaCl, 10 mM 2-mercaptoethanol, 1 mM magnesium acetate, 3 mM CaCl2, 1 mM imidazole, and 0.1% NP-40). This crude extract was clarified by centrifugation at 15,000 3 g for 20 min, and the supernatant was added to 300 mL

Molecular Cell 598

of Calmodulin-Sepharose 4B (Amersham Biosciences) that had been washed two times with 10 mL of binding buffer and rocked at 4ºC for 2 hr. The extract-bead slurries were passed over Miniprep purification columns (BioRad) to isolate the resin, and the beads were washed three times with 2 mL of calmodulin binding buffer. Seven elutions were performed with 300 ml of calmodulin elution buffer (50 mM HEPES [pH 7.5], 150 mM NaCl, 10 mM 2-mercaptoethanol, 1 mM MgAc, 4 mM EDTA, and 0.1% NP-40). Eluent fractions were analyzed by SDS-PAGE and Western analysis using purified rabbit immunoglobulin that binds the protein A component of the TAP tag. Kinase assays were performed by adding the indicated amounts of immunoprecipitated material to 28 mL of kinase assay buffer (50 mM Tris [pH 7.5], 10 mM MgCl2, and 0.5 mM DTT). After the addition of 5 mCi of 32P ATP, the reactions were incubated for 40 min at 30ºC. Each reaction was quenched by the addition of SDS-PAGE loading buffer and a brief heating of the samples at 90ºC for 5 min. Reactions were then immediately resolved by SDS-PAGE and subjected to autoradiography. Growth Sensitivity Assays Cells were grown to an OD600 of 1.0 in 5 mL of rich medium or synthetic dropout medium (Qbiogene). 5-fold serial dilutions of each culture were made and spotted onto YPD plates or plates containing either 2% formamide, 0.05% MMS, or 100 mM HU. Cells were then allowed to grow at 30ºC for 48 hr. To measure UV sensitivity, cells were exposed to UV irradiation for 10 s and allowed to recover at 30ºC for 48 hr. Assay for Telomeric Silencing For telomeric silencing assays, the same cultures as above were also spotted onto synthetic complete plates containing 5FOA and allowed to grow for 48 hr at 30ºC. Heat sensitivity was scored by spotting cultures onto YPD plates and incubating at 37ºC for 48 hr.

COMPASS requires ubiquitination of histone H2B by Rad6. J. Biol. Chem. 277, 28368–28371. Fassler, J.S., and Winston, F. (1988). Isolation and analysis of a novel class of suppressor of Ty insertion mutations in Saccharomyces cerevisiae. Genetics 118, 203–212. Feng, Q., Wang, H., Ng, H.H., Erdjument-Bromage, H., Tempst, P., Struhl, K., and Zhang, Y. (2002). Methylation of H3-lysine 79 is mediated by a new family of HMTases without a SET domain. Curr. Biol. 12, 1052–1058. Gerber, M., and Shilatifard, A. (2003). Transcriptional elongation by RNA polymerase II and histone methylation. J. Biol. Chem. 278, 26303–26306. Henry, K.W., Wyce, A., Lo, W.S., Duggan, L.J., Emre, N.C., Kao, C.F., Pillus, L., Shilatifard, A., Osley, M.A., and Berger, S.L. (2003). Transcriptional activation via sequential histone H2B ubiquitylation and deubiquitylation, mediated by SAGA-associated Ubp8. Genes Dev. 17, 2648–2663. Hess, J.L. (2004). MLL: a histone methyltransferase disrupted in leukemia. Trends Mol. Med. 10, 500–507. Huang, H., Kahana, A., Gottschling, D.E., Prakash, L., and Liebman, S.W. (1997). The ubiquitin-conjugating enzyme Rad6 (Ubc2) is required for silencing in Saccharomyces cerevisiae. Mol. Cell. Biol. 17, 6693–6699. Kao, C.F., Hillyer, C., Tsukuda, T., Henry, K., Berger, S., and Osley, M.A. (2004). Rad6 plays a role in transcriptional activation through ubiquitylation of histone H2B. Genes Dev. 18, 184–195. Keogh, M.C., Podolny, V., and Buratowski, S. (2003). Bur1 kinase is required for efficient transcription elongation by RNA polymerase II. Mol. Cell. Biol. 23, 7005–7018. Krogan, N.J., Dover, J., Khorrami, S., Greenblatt, J.F., Schneider, J., Johnston, M., and Shilatifard, A. (2002). COMPASS, a histone H3 (Lysine 4) methyltransferase required for telomeric silencing of gene expression. J. Biol. Chem. 277, 10753–10755.

ChIP ChIP studies were performed by using strains containing either Rad6-TAP, Bur2-TAP, Paf1-TAP, or Rad6-TAP, Rad6-TAP(S120A), and other tagged strains as indicated in the figures. ChIPs were then carried out as previously described (Wood et al., 2003a).

Krogan, N.J., Dover, J., Wood, A., Schneider, J., Heidt, J., Boateng, M.A., Dean, K., Ryan, O.W., Golshani, A., Johnston, M., et al. (2003). The Paf1 complex is required for histone H3 methylation by COMPASS and Dot1p: linking transcriptional elongation to histone methylation. Mol. Cell 11, 721–729.

Acknowledgments

Lacoste, N., Utley, R.T., Hunter, J.M., Poirier, G.G., and Cote, J. (2002). Disruptor of telomeric silencing-1 is a chromatin-specific histone H3 methyltransferase. J. Biol. Chem. 277, 30421–30424.

We are grateful to Steve Burtowski for conversation regarding this work and for insight in making a bur1 null strain. We also thank Kristen Tenney for critically reading the manuscript. This work was supported in part by grants from a National Institutes of Health (1R01GM069905) award to A.S. A.W. is supported by a predoctoral fellowship from American Heart Association. A.S. is a scholar of the Leukemia and Lymphoma Society.

Miller, T., Krogan, N.J., Dover, J., Erdjument-Bromage, H., Tempst, P., Johnston, M., Greenblatt, J.F., and Shilatifard, A. (2001). COMPASS: a complex of proteins associated with a trithorax-related SET domain protein. Proc. Natl. Acad. Sci. USA 98, 12902–12907.

Received: May 24, 2005 Revised: July 19, 2005 Accepted: September 13, 2005 Published: November 22, 2005

Ng, H.H., Feng, Q., Wang, H., Erdjument-Bromage, H., Tempst, P., Zhang, Y., and Struhl, K. (2002a). Lysine methylation within the globular domain of histone H3 by Dot1p is important for telomeric silencing and Sir protein association. Genes Dev. 16, 1518–1527.

References

Ng, H.H., Xu, R.M., Zhang, Y., and Struhl, K. (2002b). Ubiquitination of histone H2B by Rad6 is required for efficient Dot1p-mediated methylation of histone H3 lysine 79. J. Biol. Chem. 277, 34655– 34657.

Briggs, S.D., Bryk, M., Strahl, B.D., Cheung, W.L., Davie, J.K., Dent, S.Y., Winston, F., and Allis, C.D. (2001). Histone H3 lysine 4 methylation is mediated by Set1 and required for cell growth and rDNA silencing in Saccharomyces cerevisiae. Genes Dev. 15, 3286–3295.

Murray, S., Udupa, R., Yao, S., Hartzog, G., and Prelich, G. (2001). Phosphorylation of the RNA polymerase II carboxy-terminal domain by the Bur1 cyclin-dependent kinase. Mol. Cell. Biol. 21, 4089–4096.

Ng, H.H., Ciccone, D.N., Morshead, K.B., Oettinger, M.A., and Struhl, K. (2003). Lysine-79 of histone H3 is hypomethylated at silenced loci in yeast and mammalian cells: a potential mechanism for positioneffect variegation. Proc. Natl. Acad. Sci. USA 100, 1820–1825.

Bryk, M., Briggs, S.D., Strahl, B.D., Curcio, M.J., Allis, C.D., and Winston, F. (2002). Evidence that Set1, a factor required for methylation of histone H3, regulates rDNA silencing in S. cerevisiae by a Sir2independent mechanism. Curr. Biol. 12, 165–170.

Nislow, C., Ray, E., and Pillus, L. (1997). SET1, a yeast member of the trithorax family, functions in transcriptional silencing and diverse cellular processes. Mol. Biol. Cell 8, 2421–2436.

Daniel, J.A., Torok, M.S., Sun, Z.W., Schieltz, D., Allis, C.D., Yates, J.R., 3rd, and Grant, P.A. (2004). Deubiquitination of histone H2B by a yeast acetyltransferase complex regulates transcription. J. Biol. Chem. 279, 1867–1871.

Prelich, G., and Winston, F. (1993). Mutations that suppress the deletion of an upstream activating sequence in yeast: involvement of a protein kinase and histone H3 in repressing transcription in vivo. Genetics 135, 665–676.

Dover, J., Schneider, J., Tawiah-Boateng, M.A., Wood, A., Dean, K., Johnston, M., and Shilatifard, A. (2002). Methylation of histone H3 by

Robzyk, K., Recht, J., and Osley, M.A. (2000). Rad6-dependent ubiquitination of histone H2B in yeast. Science 287, 501–504.


Roguev, A., Schaft, D., Shevchenko, A., Pijnappel, W.W., Wilm, M., Aasland, R., and Stewart, A.F. (2001). The Saccharomyces cerevisiae Set1 complex includes an Ash2 homologue and methylates histone 3 lysine 4. EMBO J. 20, 7137–7148. Santos-Rosa, H., Schneider, R., Bannister, A.J., Sherriff, J., Bernstein, B.E., Emre, N.C., Schreiber, S.L., Mellor, J., and Kouzarides, T. (2002). Active genes are tri-methylated at K4 of histone H3. Nature 419, 407–411. Schneider, J., Dover, J., Johnston, M., and Shilatifard, A. (2004). Global proteomic analysis of S. cerevisiae (GPS) to identify proteins required for histone modifications. Methods Enzymol. 377, 227–234. Schneider, J., Wood, A., Lee, J.S., Schuster, R., Dueker, J., Maguire, C., Swanson, S.K., Florens, L., Washburn, M.P., and Shilatifard, A. (2005). Molecular regulation of histone H3 trimethylation by COMPASS and the regulation of gene expression. Mol. Cell 19, 849–856. Shahbazian, M.D., Zhang, K., and Grunstein, M. (2005). Histone H2B ubiquitylation controls processive methylation but not monomethylation by Dot1p and Set1. Mol. Cell 19, 271–277. Simchen, G., Winston, F., Styles, C.A., and Fink, G.R. (1984). Tymediated gene expression of the LYS2 and HIS4 genes of Saccharomyces cerevisiae is controlled by the same SPT genes. Proc. Natl. Acad. Sci. USA 81, 2431–2434. Singer, M.S., Kahana, A., Wolf, A.J., Meisinger, L.L., Peterson, S.E., Goggin, C., Mahowald, M., and Gottschling, D.E. (1998). Identification of high-copy disruptors of telomeric silencing in Saccharomyces cerevisiae. Genetics 150, 613–632. Sun, Z.W., and Allis, C.D. (2002). Ubiquitination of histone H2B regulates H3 methylation and gene silencing in yeast. Nature 418, 104–108. Tenney, K., and Shilatifard, A. (2005). A COMPASS in the voyage of defining the role of trithorax/MLL-containing complexes: linking leukemogensis to covalent modifications of chromatin. J. Cell. Biochem. 95, 429–436. van Leeuwen, F., Gafken, P.R., and Gottschling, D.E. (2002). Dot1p modulates silencing in yeast by methylation of the nucleosome core. Cell 109, 745–756. Winston, F., Chaleff, D.T., Valent, B., and Fink, G.R. (1984). Mutations affecting Ty-mediated expression of the HIS4 gene of Saccharomyces cerevisiae. Genetics 107, 179–197. Winston, F., Dollard, C., Malone, E.A., Clare, J., Kapakos, J.G., Farabaugh, P., and Minehart, P.L. (1987). Three genes are required for trans-activation of Ty transcription in yeast. Genetics 115, 649– 656. Wood, A., Krogan, N.J., Dover, J., Schneider, J., Heidt, J., Boateng, M.A., Dean, K., Golshani, A., Zhang, Y., Greenblatt, J.F., et al. (2003a). Bre1, an E3 ubiquitin ligase required for recruitment and substrate selection of Rad6 at a promoter. Mol. Cell 11, 267–274. Wood, A., Schneider, J., Dover, J., Johnston, M., and Shilatifard, A. (2003b). The Paf1 complex is essential for histone monoubiquitination by the Rad6-Bre1 complex, which signals for histone methylation by COMPASS and Dot1p. J. Biol. Chem. 278, 34739–34742. Wood, A., Schneider, J., and Shilatifard, A. (2005). Cross-talking histones: implications for the regulation of gene expression and DNA repair. Biochem. Cell Biol. 83, 460–467. Worthylake, D.K., Prakash, S., Prakash, L., and Hill, C.P. (1998). Crystal structure of the Saccharomyces cerevisiae ubiquitinconjugating enzyme Rad6 at 2.6 A resolution. J. Biol. Chem. 273, 6271–6276. Yao, S., and Prelich, G. (2002). Activation of the Bur1-Bur2 cyclindependent kinase complex by Cak1. Mol. Cell. Biol. 22, 6750–6758. Yao, S., Neiman, A., and Prelich, G. (2000). BUR1 and BUR2 encode a divergent cyclin-dependent kinase-cyclin complex important for transcription in vivo. Mol. Cell. Biol. 20, 7080–7087.