The Role of the Carboxyterminal Domain of RNA Polymerase II in Regulating Origins of DNA Replication in Saccharomyces cerevisiae

Corresponding author: Krassimir Yankulov, University of Guelph, Guelph, ON N1G 2W1, Canada., yankulov{at}uoguelph.ca (E-mail)

Communicating editor: B. J. ANDREWS

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

MCM (minichromosome maintenance) proteins function as a replication licensing factor (RLF-M), which contributes to limiting initiation of DNA replication to once per cell cycle. In the present study we show that a truncation of the pol II CTD in a S. cerevisiae strain harboring a mutation in mcm5 partially reverses its ts phenotype and improves maintenance of CEN/ARS minichromosomes. We correlate this phenotype to effects on DNA replication rather than to effects on transcription or specific gene expression. We also demonstrate that a similar truncation of the CTD reduces minichromosome stability and impairs stimulation of DNA replication by trans-activators and that tethering of recombinant pol II CTD to an origin of replication has a significant stimulatory effect on minichromosome stability. Furthermore, we show that pol II is recruited to ARS1. We propose that in S. cerevisiae a mechanism of coordinating pol II transcription and DNA replication is mediated by the CTD of pol II.

DNA replication in eukaryotes initiates at DNA locations referred to as origins of replication. In Saccharomyces cerevisiae origins behave as autonomously replicating sequences (ARS) when placed on an extrachromosomal DNA. These contain one essential (A) and three auxiliary (B1, B2, and B3) elements (MARAHRENS and STILLMAN 1992 ). The A element is a core sequence that binds a protein complex called origin recognition complex (ORC; BELL and STILLMAN 1992 ), which in turn recruits CDC6 and six minichromosome maintenance (MCM) proteins to form the prereplicative complex (COCKER et al. 1996 ). Activation of these prereplicative complexes by protein kinases is absolutely required for initiation of replication. It is believed that disruption of the ORC-MCM interaction at the time of initiation is responsible for limiting origin firing to once per cell cycle (TYE 1999A ; LABIB and DIFFLEY 2001 ; LEI and TYE 2001 ). It has been shown that MCMs are associated with moving replication forks in yeast (APARICIO et al. 1997 ; LABIB et al. 2000 ; LABIB and DIFFLEY 2001 ), suggesting a postinitiation function in DNA replication. The precise biochemical role of the MCM proteins remains unclear. They do possess helicase activity (ISHIMI 1997 ), which could be involved in facilitating replication fork movement (LABIB and DIFFLEY 2001 ). Mammalian MCM2 binds to histones H3/H4 (ISHIMI et al. 1988 , ISHIMI et al. 1996 ) and to HBO1, a putative histone acetyl transferase (BURKE et al. 2001 ), indicating a potential role for MCMs in chromatin remodeling.

The function of the auxiliary (B1, B2, and B3) elements in yeast origins is not completely understood. The B1 element provides an additional binding site for ORC (RAO and STILLMAN 1995 ; ROWLEY et al. 1995 ). The role of the B2 element is not clear. The B3 element in ARS1 contains a binding site for the Abf1 protein, which also operates as a transcriptional activator or a transcriptional repressor in other contexts. The function of the Abf1-binding site in ARS1 can be replaced by binding sites for other transcriptional activators (MARAHRENS and STILLMAN 1992 ; LI et al. 1998 ; HU et al. 1999 ) or by tethering viral or mammalian transcriptional activation domains (BENNETT-COOK and HASSELL 1991 ; LI et al. 1998 ; HU et al. 1999 ; STAGLJAR et al. 1999 ) to the B3 element. Extensive evidence shows that transcriptional activators also stimulate viral origins of replication (DEPAMPHILIS 1988 ). It is obvious that transcriptional activators are positive regulators of origins of DNA replication; however, their mechanism of action in these contexts is largely unknown. Several reports correlate these positive effects to the binding of the trans-activator to the replication factor RPA (HE et al. 1993 ; LI and BOTCHAN 1993 ). Other reports show that the ability of trans-activators to stimulate replication is tightly linked to their ability to stimulate pol II transcription (BENNETT-COOK and HASSELL 1991 ; HE et al. 1993 ; LI et al. 1998 ) and chromatin remodeling (HU et al. 1999 ). Recently it was shown that recruitment of pol II or pol III complexes to ARS1 is sufficient to enhance replication from a minichromosome origin (STAGLJAR et al. 1999 ; BODMER-GLAVAS et al. 2001 ), presumably by remodeling chromatin.

Large pol II complexes, which contain some or all of the pol II general transcription factors, have been purified from a variety of sources and designated RNA polymerase II holoenzyme (KIM et al. 1994 ; KOLESKE and YOUNG 1994 ; HENGARTNER et al. 1995 ; OSSIPOW et al. 1995 ; CHAO et al. 1996 ; MALDONADO et al. 1996 ; PAN et al. 1997 ). Pol II holoenzyme complexes from S. cerevisiae contain the SRB and MED family of proteins, the SWI/SNF complex, RGR1, and GAL11. Temperature-sensitive (ts) mutations in genes for holoenzyme proteins show that some (SRB4 and SRB6) are essential for mRNA synthesis (THOMPSON and YOUNG 1995 ), whereas others contribute to the response to trans-activators (KIM et al. 1994 ; HENGARTNER et al. 1995 ; LI et al. 1995 ; WILSON et al. 1996 ; MYERS et al. 1998 ; HAN et al. 1999 ). Mammalian pol II holoenzyme complexes contain homologs of SRB, MED, and SWI/SNF proteins (OSSIPOW et al. 1995 ; CHAO et al. 1996 ; MALDONADO et al. 1996 ; PAN et al. 1997 ) but also DNA repair factors (MALDONADO et al. 1996 ; SCULLY et al. 1997 ; ANDERSON et al. 1998 ), replication factors (MALDONADO et al. 1996 ; YANKULOV et al. 1999 ), and cleavage/polyadenylation factors (MCCRACKEN et al. 1997 ). A direct involvement in mRNA synthesis for some of these factors is not evident. Similar factors are not detected in the yeast holoenzyme. These dissimilarities could result from different strategies of purification or might reflect some fundamental differences between yeast and higher eukaryotes.

Most of the yeast pol II holoenzyme components join pol II via interactions with the highly conserved carboxyterminal domain (CTD) of its largest subunit. The CTD is composed of heptapeptide repeats (26 in S. cerevisiae and 52 in higher eukaryotes) with a consensus YSPTSPT. Antibodies against the CTD dissociate the yeast holoenzyme into core pol II and another complex named the "mediator" (KIM et al. 1994 ; HENGARTNER et al. 1995 ; MYERS et al. 1997 ). The CTD, therefore, plays a key role in assembly of the pol II holoenzyme. As expected, the function of different mediator components is dependent on the CTD (NONET and YOUNG 1989 ; MYERS et al. 1997 ).

Recently we reported that human pol II holoenzyme complexes interact with MCM proteins via the CTD of pol II (YANKULOV et al. 1999 ). We proposed that in higher eukaryotes MCM proteins are involved in pol II transcription. In this article we demonstrate that a deletion of pol II CTD partially reverses the ts phenotype in S. cerevisiae, which is caused by a mutation in MCM5. We also demonstrate that RNA polymerase II is recruited to ARS1.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Plasmids:
pARS1/wtA is an ARS1/CEN4/URA3-based vector described in MARAHRENS and STILLMAN 1992 . The derivative pARS1/-B23/G24 (LI et al. 1998 ) contains ARS1 in which the B2 and B3 elements are mutated and a GAL4-binding site is inserted next to the B3 element. pARS1wtA/2.2-kb/URA3 and pARS1/-B23/G24/2.2 kb/URA3 contain a 2.2-kb cDNA fragment derived from human MCM2, which is inserted in the SphI site between the ARS1 and the URA3 elements, respectively. This fragment contains no promoter elements. YCp50CDC46 is an ARS1/CEN4/URA3 plasmid containing a 15.6-kb MCM5 genomic fragment on the YCp50 vector (CHEN et al. 1992 ). YIp122CDC46 is a LEU2 integrating vector containing the 6.5-kb AflII MCM5 genomic fragment cloned into the SmaI site of YIp122. pFL35CDC46 is a TRP1 integrating vector containing the SalI-AflII genomic MCM5 fragment cloned in SalI/XhoI sites of pFL35. pFL26RPB1104 is a LEU2 integrating vector encoding rpb1104 (NONET et al. 1987 ). pFL38RPB1 (MCNEIL et al. 1998 ) is an ARS1/CEN4/URA3 plasmid containing the 6042-bp EcoRI-PstI fragment of RPB1. pGBKT7 (CLONTECH, Palo Alto, CA) is a 2µ/TRP1 vector encoding the DNA-binding domain (amino acids 1–147) of GAL4 under the control of the ADH1 promoter. pGBKT7-CTDwt encodes the mouse CTD (52 heptad repeats) fused to the DNA-binding domain of GAL4. pGBKT7-CTDmut encodes 15 mutant (S5 A) heptad repeats fused to the DNA-binding domain of GAL4. pGBKT7-dacB expresses the Escherichia coli DAC-B protein fused to the DNA-binding domain of GAL4.

Yeast strains and growth conditions:
The names and genotypes of the yeast strains used in this study are listed in Table 1. rpb1104mcm5 was produced by transforming the mcm5 strain with pFL26RPB1104 linearized by BsiWI and selecting on SC-Leu plates. MCM5 was produced by transforming the mcm5 strain with YIp122CDC46 linearized by BspHI and selecting on SC-Leu plates and then on YPD plates at 37°. rpb1104MCM5 was produced by transforming the rpb1104mcm5 strain with pFL35CDC46 linearized by BspHI and selecting on SC-trp plates.

Minichromosomes were introduced by electroporation. Yeast cultures were grown in SC (synthetic complete) medium plus 2% glucose or 2% galactose. Uracil, tryptophan, or leucine were omitted as indicated. Cells containing pARS1/-B23/G24 were grown on SC-Ura/Galactose medium.

Minichromosome stability assay:
The rate of plasmid loss per generation was estimated as described (DANI and ZAKIAN 1983 ) with modifications that improve precision of the measurement and accuracy of calculation (KRAMER et al. 2002 ). Briefly, cells grown in selective medium were diluted to <5 cells/ml in SC medium and 150 µl were dispensed in 96-well tissue culture plates. The plates were incubated at specified temperatures until single colonies were visible in the wells. Colonies were resuspended by pipetting and cells were counted in a hemacytometer chamber. Wells with the equal lowest number of total cells (A) corresponding to the same number of generations [N calculated as N = Log(A,2)] were selected as mini-cultures originating from a single cell and were further analyzed. Aliquots were briefly sonicated and cells were plated on selective and nonselective plates. Percentage of plasmid-containing cells (F) was calculated as F = number of colonies on selective plates/number of colonies on nonselective plates. We presumed that cultures that produced colonies on selective plates had originated from a single plasmid-containing cell. Plasmid loss was calculated as 1 - F/^1/N. A loss of 100% indicated a mini-culture that had originated from a single cell without a plasmid.

Measurement of total de novo RNA synthesis:
Cells were grown overnight under specified conditions to an early exponential phase and diluted with prewarmed medium to OD₆₀₀ = 0.2. [5,6-³H]Uridine was added to 15 µCi/ml final concentration. Aliquots of 0.2 ml were removed after 20, 40, and 60 min and immediately added to 1 ml ice-cold stop solution (15% trichloroacetic acid, 50 mM pyrophosphate) containing 0.2 ml unlabeled stationary-phase yeast culture. Cells were washed (five times for 10 min) in stop solution and once in EtOH. Radioactivity was determined by scintillation counting in a Beckman (Fullerton, CA) LS6500 counter.

Measurment of de novo poly(A)⁺ RNA synthesis:
Cells were grown and labeled for 1 hr as described for total RNA synthesis and then harvested in ice-cold water plus 0.5 ml of unlabeled stationary-phase yeast culture and washed four times with ice-cold water. RNA was isolated by a SV total RNA isolation system (Promega, Madison, WI) according to the instructions of the manufacturer. RNA yields were estimated by OD₂₆₀. Poly(A)⁺ containing RNA was isolated by a poly(A) tract mRNA isolation kit (Promega). Radioactivity in the RNA samples was determined by scintillation counting in a Beckman LS6500 counter. De novo synthesis of RNA and poly(A)⁺ RNA was expressed as counts per minute per microgram RNA.

Microarray analysis of gene expression:
Cells were grown under specified conditions to OD₆₀₀ = 0.2–0.5 and harvested on crushed ice. Total RNA was isolated by the lithium chloride method and cDNA was synthesized from 10 µg of total RNA by reverse transcribing with SuperScript II (GIBCO BRL, Gaithersburg, MD) in the presence of amino-allyl dUTP. N-hydroxysuccinimide Cy5 and Cy3 dyes (Amersham-Pharmacia) were coupled to the amine-modified cDNA according to the instructions of the manufacturer. Microarrays containing all 6200 open reading frames from the S. cerevisiae genome were purchased from the Microarray Centre at the Ontario Cancer Institute, Toronto. Hybridization was for 18 hr at 37°. The microarrays were scanned with the Axon GenePix 4000a microarray scanner and analyzed with the GeneSpring v4.0.1 software package (Silicon Genetics). Three different replica samples were analyzed. Differentially expressed genes were identified as twofold up- or downregulated.

Chromatin immunoprecipitation:
This was performed according to the procedure described in STRAHL-BOLSINGER et al. 1997 with some modifications. Cells containing pARS1/2.2 kb/URA3 or pARS1/-B23/G24/2.2 kb/URA3 were grown in SC-Ura medium to O_D600 = 1.5, crosslinked with 1% formaldehyde, and sonicated to an average DNA size of 100–1000 bp. Sheared chromatin was immunoprecipitated with monoclonal antibodies against pol II CTD (8WG16; THOMPSON et al. 1989 ) or with the corresponding amount of fetal bovine serum IgG as a control. The immunoprecipitated chromatin was eluted with 1% SDS, DNA was uncrosslinked and precipitated, and aliquots were subjected to 20–23 cycles of PCR in the presence of 2 µCi [-³²P]dCTP with the primers described below. Amplification of CEN4, 2.2 kb, and ARS1 DNA was performed in multiplex PCR reactions with three pairs of primers. The URA3 fragment was amplified separately. PCR products were resolved on native polyacrylamide gels and exposed to X-ray films.

PCR primers:
These were designed to specifically amplify the plasmid-borne, but not the endogenous ARS1, CEN4, and URA3 elements. One of the primers annealed to ARS1, CEN4, and URA3, respectively, while the corresponding reverse primers annealed to the pUC119 backbone (see Fig 8A). Another pair of primers was designed to amplify a 400-bp fragment from the 2.2-kb insert positioned between the ARS1 and URA3 elements. The amplified fragment is 1 kb away from both ARS1 and URA3. The CEN4 amplified fragment is 1.2 kb away from ARS1. The sequences of the used primers and the PCR conditions are available upon request.

View larger version (86K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Truncation of pol II CTD partially suppresses the ts phenotype of mcm5. rpb1104mcm5, mcm5, MCM5, and rpb1104MCM5 and an unrelated wild-type strain were streaked on SC or SC-Leu and grown at the temperatures indicated below each plate. The positions of the strains are diagrammed above each column of plates. A and B represent two separate experiments.

View larger version (24K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Analysis of cell growth of mcm5, rpb1104mcm5, and MCM5. Cells were grown at 30° in SC or SC-Ura liquid cultures as indicated. Time course measurements of OD600 are plotted. (A) Growth in SC medium. (B) Complementation of rpb1104mcm5 by expressing RPB1 from an ARS1/CEN4 minichromosome retards its growth in selective medium. rpb1104mcm5 cells were transformed with YCP50 (ARS1/CEN4/URA3), YCP50CDC46, and pFL38RPB1 (ARS1/CEN4/URA3) and grown in SC-Ura medium. (C) Growth rate of mcm5, rpb1104mcm5, and MCM5 cells transformed with pARS1/wtA in SC-Ura medium.

View larger version (33K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Minichromosome stability in rpb1104mcm5, mcm5, MCM5, and rpb1104MCM5. Cells were transformed with pARS1/wtA. Single colonies were grown in SC-Ura medium as stock cultures, diluted in SC medium at 5 cells/ml, dispensed in 96-well tissue culture plates, and grown in SC at 30° or 37°. Minichromosome loss per generation (%) was calculated as (F, percentage of minichromosome-containing cells; N, number of generations) as described in MATERIALS AND METHODS. Each bar represents the calculation of minichromosome stability from an individual mini-culture in the strain indicated below the graph.

View larger version (44K): [in this window] [in a new window] [Download PPT slide]   Figure 4. Rate of de novo total RNA and mRNA synthesis at 30° and 37°. (A) Rate of de novo RNA synthesis. Cultures were grown in SC medium at 30° (top) or 37° (bottom) to OD600 = 0.2. [5,6-3H]Uridine was added at 15 µCi/ml. Aliquots were collected at the 20th, 40th, and 60th minute. Incorporation of the label was measured as described in MATERIALS AND METHODS. (B) Rate of de novo mRNA synthesis at 30°. (C) Rate of de novo mRNA synthesis at 37°. Cultures were grown in SC medium at 30° (B) or 37° (C) to OD600 = 0.2. [5,6-3H]Uridine was added at 15 µCi/ml and cells were grown for 1 hr. Isolation of total RNA and mRNA is described in MATERIALS AND METHODS. Incorporation of the label is plotted as counts per minute per micrograms RNA.

View larger version (46K): [in this window] [in a new window] [Download PPT slide]   Figure 5. Stability of pARS1/-B23/G24 in rpb1104mcm5, mcm5, MCM5, and rpb1104MCM5. Cells were transformed with pARS1/-B23/G24 and single colonies were grown in SC/Galactose. Stock cultures were diluted in SC/Glucose or SC/Galactose medium at 5 cells/ml and dispensed in 96-well tissue culture plates. Minichromosome loss per generation (%) was calculated as (F, percentage of minichromosome containing cells; N, number of generations) as described in MATERIALS AND METHODS. Each bar represents the calculation of minichromosome stability from an individual mini-culture in the strain indicated below the graph.

View larger version (26K): [in this window] [in a new window] [Download PPT slide]   Figure 6. Stability of pARS/wtA in strains containing RPB1 with 10 and 12 CTD repeats. pYwt(10), pYwt(12), and Z551(wt) strains were transformed with pARS1/wtA and selected on SC-Ura-Leu. Single colonies were grown in SC-Ura-Leu and diluted in SC-Leu. Minichromosome loss per generation (%) was calculated as (F, percentage of minichromosome containing cells; N, number of generations) as described in MATERIALS AND METHODS. Each bar represents the calculation of minichromosome stability from an individual mini-culture in the strain indicated below the graph.

View larger version (28K): [in this window] [in a new window] [Download PPT slide]   Figure 7. Recombinant CTD stimulates stability of pARS1/-B23/G24. DF5 cells were transformed with pGBKT7 plasmids encoding different GAL4-fusion proteins and pARS1/-B23/G24 and selected on SC-Ura-Trp. Single colonies were grown in SC-Ura-Trp and diluted in SC-Trp. Minichromosome loss per generation (%) was calculated as (F, percentage of minichromosome containing cells; N, number of generations) as described in MATERIALS AND METHODS. Each bar represents the calculation of minichromosome stability from an individual mini-culture in the strain indicated below the graph. GAL4-DBD, GAL4 DNA-binding domain; GAL4-dacB, GAL4 DNA-binding domain fused to the E. coli DAC-B protein; GAL4-CTDwt, GAL4 DNA-binding domain fused to wt pol II CTD; GAL4-CTDmut, GAL4 DNA-binding domain fused to 15 mutant (S5 A) heptad repeats.

View larger version (99K): [in this window] [in a new window] [Download PPT slide]   Figure 8. RNA polymerase II is crosslinked to ARS1 in MCM5, rpb1104mcm5, and mcm5. (A) A diagram of pARS1wtA/2.2 kb/URA3. The locations of the CEN4, ARS1, the 2.2-kb insert, and the URA3 gene are shown in scale. The positions of the pairs of PCR primers and the distance between the amplified fragments are shown below the diagram. In pARS1/-B23/G24/2.2 kb/URA3 the ARS1 element is replaced by ARS1/-B23/G24. (B) PCR amplification of minichromosome-borne CEN4, ARS1, 2.2-kb, and URA3 fragments immunoprecipitated with anti-pol II CTD. Cells were transformed with pARS1wtA/2.2 kb/URA3 and grown in SC-Ura. The strain used for preparation of crosslinked chromatin is shown above each panel of lanes. In lanes 4, 10, and 16 the samples were immunoprecipitated with anti-pol II CTD antibody without prior crosslinking with formaldehyde. In lanes 5, 11, and 17 the samples were crosslinked and immunoprecipitated with control antibody. In lanes 6, 12, and 18 samples were crosslinked and immunoprecipitated with anti-pol II CTD. The following amounts of DNA were amplified and resolved on polyacrylamide gels: 1% of the input in lanes 1, 7, and 13; 0.02% of the input in lanes 2, 8, and 14; 0.02% of the input in lanes 3, 9, and 15; 10% of the immunoprecipitate in lanes 4, 5, 6, 10, 11, 12, 16, 17, and 18. Amplification of CEN4, 2.2-kb, and ARS1 DNA was performed in multiplex PCR reactions with three pairs of primers. The URA3 fragment was amplified separately. The position of each amplified fragment is indicated on the left. (C) PCR amplification of minichromosome-borne CEN4, ARS1/-B23/GAL4, 2.2-kb, and URA3 fragments immunoprecipitated with anti-pol II CTD. MCM5 cells transformed with pARS1/-B23/G24/2.2 kb/URA3 and grown in SC-Ura/glucose (lanes 1–6) or SC-Ura/galactose (lanes 7–12). In lanes 4 and 10 the samples were immunoprecipitated with anti-pol II CTD antibody without prior crosslinking. In lanes 5 and 11 the samples were crosslinked and immunoprecipitated with control antibody. In lanes 6 and 12 the samples were crosslinked and immunoprecipitated with anti-pol II CTD. The following amounts of DNA were amplified and resolved on polyacrylamide gels: 1% of the input in lanes 1 and 7; 0.02% of the input in lanes 2 and 8; 0.02% of the input in lanes 3 and 9; 10% of the immunoprecipitate in lanes 4, 5, 6, 10, 11, and 12. Amplification of fragments and description are as in Fig 8B.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Truncation of RNApol II CTD partially reverses mcm5 phenotype:
The biochemical interaction between the human pol II holoenzyme and MCM proteins raises the issue of a similar interaction in S. cerevisiae. We explored this possibility by disrupting the RPB1 gene with an rpb1104 encoding 11 out of 26 CTD repeats (NONET et al. 1987 ) in strains that harbored mcm mutations. The parental mcm strains, mcm2-1, mcm3-3, mcm5-1, and mcm5-3, were produced by ethyl methanesulfonate mutagenesis and have conditional ts phenotypes (MAINE et al. 1984 ; HENNESSY et al. 1991 ; TYE 1999B ). Strains that encode rpb1 with 10–12 CTD repeats can also display conditional cs and ts phenotypes (NONET et al. 1987 ; WEST and CORDEN 1995 ). Initially we compared the cs and ts phenotypes of single mcm mutants vs. double mcmrpb1104 mutants. We did not see any obvious alteration of the phenotypes of rpb1104mcm2-1, rpb1104mcm3-3, and rpb1104mcm5-1 relative to the corresponding parental mcm strains. The rpb1104mcm5-3 mutant (from now on referred to as rpb1104mcm5), grew significantly better at 37° than did the parental mcm5 strain (Fig 1). We focused our studies on this strain.

The phenotype of rpb1104mcm5 could be specific to mcm5 or, alternatively, could be a consequence of mutations introduced during the mutagenesis of the parental strain. We addressed this issue by complementation with MCM5. To avoid any effects from poor maintenance of plasmids, we inserted the MCM5 gene in the genomes of mcm5 and rpb1104mcm5 strains, respectively. The resulting isogenic strains were designated MCM5 and rpb1104MCM5. The growth of all strains was comparable at room temperature and at 30° (Fig 1). As expected, insertion of MCM5 in the mcm5 strain completely reversed its ts phenotype (Fig 1). Introduction of MCM5 in the rpb1104mcm5 strain resulted in some growth advantage (Fig 1), yet rpb1104MCM5 did not grow as fast as MCM5 at 37° (Fig 1B and data not shown). The temperature sensitivity of rpb1104MCM5 could be attributed to the rpb1104 mutation, which had shown a similar phenotype in an unrelated strain (NONET et al. 1987 ). Neither of the mutant strains grew at 38.5° (data not shown). These initial results indicated that the observed phenotypes of mcm5 and rpb1104mcm5 strains were specific to the mcm5 and rpb1104 mutations.

Minichromosome stability is enhanced in rpb1104mcm5:
In a separate set of experiments we attempted to complement each of the mutations in the double rpb1104/mcm5 strain by expressing wild-type (wt) RPB1 or MCM5 from CEN4/ARS1/URA3 minichromosomes. As expected, expression of MCM5 significantly increased the growth rate of both mcm5 (not shown) and rpb1104mcm5 (Fig 2B) strains in SC-Ura medium at 30°. Surprisingly, rpb1104mcm5 cells expressing RPB1 from a plasmid-borne gene (pFL38RPB1) grew slightly slower relative to cells with a control plasmid (Fig 2B). One possibility for the observed kinetics could be that expression of RPB1 may interfere with the maintenance of pFL38RPB1 in rpb1104mcm5, resulting in slower growth. We tested this possibility by analyzing cell growth of the rpb1104mcm5 and mcm5 strains containing the same pARS1/wtA (CEN4/ARS1/URA3) plasmid. Fig 2C shows that the mcm5 strain grew significantly slower than rpb1104mcm5 in selective SC-Ura medium. These results are consistent with the idea that truncation of the CTD in RPB1 partially reverses the effect of mcm5-3 on minichromosome stability (Fig 2C). Hence, we specifically analyzed the loss rate of pARS1/wtA in mcm5, rpb1104mcm5, MCM5, and rpb1104MCM5 strains grown in nonselective medium at both normal and restrictive temperatures.

Minichromosome stability is estimated by measuring the percentage of minichromosome-containing cells after a period of growth in nonselective medium (TYE 1999B ). The major advantage of this assay is that it is direct and no assumption is made about other phenotypes that may or may not be associated with some deficiency in DNA replication (TYE 1999B ). At 30° the mcm5 strain showed a loss rate of 22.3 ± 2.2%/generation, whereas the MCM5 strain had a loss rate of 5.2 ± 0.8% (Fig 3). These data are in good agreement with previously reported studies on pARS1/wtA maintenance in other strains (MARAHRENS and STILLMAN 1992 ; LI et al. 1998 ; HU et al. 1999 ; STAGLJAR et al. 1999 ) or maintenance of similar minichromosomes in the mcm5 strain (MAINE et al. 1984 ; CHEN et al. 1992 ). The loss rate in rpb1104mcm5 was 16.1 ± 2.6% (Fig 3), which was higher than that in MCM5, but lower than that in the single mcm5 mutant. The loss rate in rpb1104MCM5 was 9.67 ± 1.5% (Fig 3). Similar relative levels of minichromosome loss per generation were observed at 37°. The MCM5 strain continued to lose plasmids at 5%/generation (Fig 3). The loss rate in the rpb1104mcm5, mcm5, and rpb1104MCM5 mutants increased to 25.3 ± 3.4%, 36.7 ± 0.4%, and 16.2 ± 1.38%, respectively (Fig 3).

We considered the possibility of recombination between the direct repeats, which were produced from the integration of rpb1104 or MCM5 in the genome of the recipient strains. If this was the case, the LEU2 and TRP1 marker genes would be lost from the rpb1104mcm5 and MCM5 or rpb1104MCM5 strains, respectively. We controlled against such recombination events by selecting for the Leu+ and Trp+ phenotypes before each experiment and confirming it after growth in nonselective SC medium. In five independent experiments we consistently observed lower levels of minichromosome loss in the rpb1104mcm5 strain relative to the single mcm5 mutant (data not shown). We also consistently observed increased minichromosome loss in the rpb1104MCM5 relative to MCM5 (data not shown).

Analysis of transcription:
The suppression of the ts phenotype and of minichromosome loss in rpb1104mcm5 could be a consequence of aberrant transcription resulting from the truncation of pol II CTD. Initially we tested this possibility by analyzing the rate of de novo total RNA and mRNA synthesis and steady-state mRNA levels at different temperatures. The rate of total RNA synthesis was assessed by incorporation of [5,6-³H]uridine in exponentially growing cells at 30° and 37°. Cells were harvested at the 20th, 40th, and 60th minute after addition of the label and incorporation was measured by scintillation counting. At 30° the rpb1104mcm5 and mcm5 incorporated [5,6-³H]uridine at comparable rates (Fig 4A). rpb1104MCM5 incorporated the label at higher levels relative to rpb1104mcm5 and mcm5, but did so more slowly than MCM5 (Fig 4A). A similar rate of total RNA synthesis was observed at 37° with the exception of the significant difference between MCM5 and the mutant strains (Fig 4A, bottom). Again, rpb1104mcm5 and mcm5 incorporated the label at similar levels, while the rpb1104MCM5 strain incorporated at a higher rate. Rates of mRNA synthesis were assessed by exposing cells to [5,6-³H]uridine for 1 hr and isolating mRNA on oligo(dT) magnetic beads. De novo total RNA and mRNA synthesis were expressed as counts per minute per microgram RNA. As shown in Fig 4B and Fig C, no substantial difference in the levels of total RNA and mRNA synthesis between rpb1104mcm5 and mcm5 was observed at both temperatures. rpb1104MCM5 synthesized RNA at slightly higher levels than did rpb1104mcm5 and mcm5 probably because of its slightly higher growth rate (not shown). In summary, our results did not point out any significant variation in the ratio of mRNA/total RNA in the mutant strains. Importantly, they did not reveal any considerable differences in total or mRNA transcription between rpb1104mcm5 and mcm5 that might explain the difference in growth rate and plasmid maintenance.

Analysis of specific gene expression in rpb1104mcm5 and mcm5:
The differences in cell growth and minichromosome maintenance between rpb1104mcm5 and mcm5 could result from changes in the expression of specific genes, which cannot be detected by global analysis of mRNA. We therefore performed analysis of gene expression using microarrays. Expression profiles of rpb1104mcm5 and mcm5 were compared at both 30° and 37°. Control experiments with mcm5 at 30° and 37° and rpb1104mcm5 at 30° and 37° were also performed. We analyzed three independent replicas for each couple of samples. The number of differentially expressed genes in mcm5 vs. rpb1104mcm5 was 89 at 30° and 173 at 37° (see supplementary data at http://www.uoguelph.ca/mbgwww/faculty/yankulov/appendix_ky082001/appendix_ky082001.html). Most of these genes encode ribosomal proteins and proteins involved in the regulation of metabolic processes, RNA metabolism, and translation and are referred to as environmental stress response (ESR) genes (GASCH et al. 2000 ). Three of the genes that were not found in the ESR cluster (GASCH et al. 2000 ) and were upregulated in rpb1104mcm5 only at 37° (see supplementary data at http://www.uoguelph.ca/mbgwww/faculty/yankulov/appendix_ky082001/appendix_ky082001.html) had been previously implicated in regulation of DNA replication and cell growth. POL32 is a subunit of DNA polymerase . PSP1 and YAC1 are high-copy-number suppressors of cell growth (GARRETT et al. 1991 ; AKADA et al. 1997 ; FORMOSA and NITTIS 1998 ). Another group of genes (ZDS1, CYC8, TUP1, POP2, SPT5, SPT8, SNF5, and GAL11), which positively or negatively regulate pol II transcription, were upregulated in rpb1104mcm5 at both temperatures. Analysis of gene expression in rpb1104mcm5 at 30° vs. 37° and in mcm5 at 30° vs. 37° showed a significantly broader range of differentially expressed genes in both strains (not shown), which probably reflects the combination of the effects of temperature change, slower growth, and mutations in mcm5 and rpb1.

In conclusion, the comparison of gene expression profiles of rpb1104mcm5 relative to mcm5 did not show a gene or a group of genes whose expression pattern could explain the increased stability of minichromosomes in rpb1104mcm5.

Activation of DNA replication by GAL4 is abolished in mcm5, rpb1104mcm5, and rpb1104MCM5:
We performed three additional experiments, which addressed the effects of pol II CTD on DNA replication. Earlier reports demonstrated a direct role of an array of transcriptional activators in stimulating origins of DNA replication (MARAHRENS and STILLMAN 1992 ; LI et al. 1998 ; HU et al. 1999 ; LI 1999 ). In most cases the effects on replication were measured by the stability of pARS1/-B23/G24 in which the ABF1-binding site in ARS1 is transformed to a GAL4-binding site (LI et al. 1998 ). pARS1/-B23/G24 is very poorly maintained; however, its loss is dramatically decreased if cells are grown in galactose, presumably because GAL4 replaces the function of ABF1 (LI et al. 1998 ; STAGLJAR et al. 1999 ). First we tested whether this effect of GAL4 is influenced by the truncation of pol II CTD by measuring the loss rate of pARS1/-B23/G24 in rpb1104mcm5, mcm5, rpb1104MCM5, and MCM5 strains in the presence or absence of galactose. Because the mutant strains did not grow in galactose at 37° the experiment was performed at 30° only. In SC/GLU the mutant strains were losing the minichromosome at a very high rate of 40–42% (Fig 5). MCM5 was losing the pARS1/-B23/G24 at 29.2 ± 4.4% (Fig 5). When the strains were grown in SC/GAL, pARS1/-B23/G24 gained significant stability only in MCM5 (Fig 5). We observed similar results in three independent experiments. We concluded that truncation of pol II CTD or a mutation in MCM5 completely abolished the positive effect of GAL4 on the activity of a GAL4-responsive synthetic origin of DNA replication.

Truncation of pol II CTD impairs minichromosome stability:
In Fig 3 we show that the loss rate of pARS1/wtA was higher in rpb1104MCM5 than in the corresponding MCM5 strain (Fig 3). We furthered these observations by testing whether truncation of the CTD would have similar effects in the unrelated Z26 strain (NONET et al. 1987 ), which carries a disrupted genomic RPB1. The Z26 strain is complemented by RPB1, which contains 10 [strain pY1WT(10)], 12 [strain pY1WT(12)], or 26 (wt, Z551) CTD repeats, respectively (NONET et al. 1987 ; WEST and CORDEN 1995 ). The strains containing truncated CTD did not exhibit any ts or cs phenotype (WEST and CORDEN 1995 ). pY1WT(10), pY1WT(12), and Z551 were transformed with pARS1/wtA and minichromosome stability was assessed as described earlier. Z551 lost pARS1/wtA at 2.75 ± 0.59%/generation (Fig 6). Truncation of pol II CTD to 12 or 10 repeats [pY1WT(12) and pY1WT(10)] increased the loss rate to 5.67 ± 0.90% and 8.82 ± 1.18%, respectively (Fig 6). Thus, truncation of pol II CTD resulted in a small but consistent decrease in minichromosome stability in two unrelated strains (Fig 3 and Fig 6). Furthermore, stability of the test minichromosme was lower in the strain with the shorter CTD (Fig 6).

The observed deficiency in minichromosome stability in pY1WT(12) and pY1WT(10) could be a consequence of the aberrant transcription of genes, which are involved in the regulation of DNA replication. We tested this possibility by comparing the gene expression profiles of pY1WT(10) and Z551. There were significant differences in the expression of numerous genes from the ESR cluster (GASCH et al. 2000 ), but no alteration in the expression of genes involved in DNA replication was observed (see supplementary data at http://www.uoguelph.ca/mbgwww/faculty/yankulov/appendix_ky082001/appendix_ky082001.html). The possibility of minor changes in the expression of such genes still exists; however, we obtained no evidence in support of this idea.

Artificial recruitment of CTD stimulates origins of replication:
If the CTD can influence DNA replication independently of its role in pol II transcription, then recruitment of CTD to origins of replication may have an effect on plasmid stability. We tested this possibility by measuring the loss rate of the pARS1/-B23/G24 in a gal4 strain (DF5), in which we expressed the DNA-binding domain of GAL4 (GAL1-147) fused to the wild-type mouse CTD (GAL4-CTDwt), to 15 synthetic mutant CTD repeats (GAL4-CTDmut), or to the E. coli dacB gene product (GAL4-dacB). These recombinant proteins were expressed from pGBKT7 (TRP/2µ). DF5 cells were cotransformed with the test minichromosome and the expression plasmid and stability of pARS1/-B23/G24 were measured. Loss rate of the test minichromosome in the presence of GAL4(1-147) and GAL4-dacB was 32–35% (Fig 7). Upon expression of GAL4-CTDwt and GAL4-CTDmut the loss rate decreased by 12 and 16% relative to the controls. It is noteworthy that both the mutant and wild-type CTD exerted a positive effect on the activity of the synthetic origin of pARS1/-B23/G24, whereas only GAL4-CTDwt was reported to stimulate transcription from the GAL4-responsive promoter in vivo (YURYEV et al. 1996 ).

RNA polymerase II is recruited to ARS1:
A key question in our study was whether pol II itself is recruited to origins of DNA replication. Previous studies indicated that artificial tethering of pol II or pol III complexes can substitute for transcriptional activators (STAGLJAR et al. 1999 ; BODMER-GLAVAS et al. 2001 ), but direct association of pol II to unmodified origins has not been shown. We addressed this question by performing chromatin immunoprecipitation experiments on minichromosomes with antibodies against the CTD of pol II. Cells containing minichromosomes were grown in selective medium, crosslinked with formaldehyde, and sonicated to average size of DNA of 100–1000 bp as in STRAHL-BOLSINGER et al. 1997 . Sheared chromatin was immunoprecipitated with anti-pol II CTD antibodies and DNA was un-crosslinked and analyzed by 20–23 cycles of PCR. Under these conditions the PCR of the immunoprecipitated DNA was within the linear range as judged by the at least 200 times higher signal obtained in parallel PCR with 10 ng of pARS1wtA/2.2 kb/URA3 (not shown). An important part of our experimental design was the elimination of noise signals coming from the pol II transcribed URA3 gene. We cloned a 2.2-kb DNA fragment in the SphI site between ARS1 and URA3 on the pARS1/wt and pARS1/-B23/GAL4, respectively. The subsequent PCR analysis was performed with primers that specifically amplified minichromosome-borne DNA elements, which were 1000 bp away from each other (Fig 8A). In these experiments signals from the amplification of the URA3 and ARS1 elements and the absence (or significant decrease) of signals from the amplification of the CEN4 and the 2.2-kb insert elements means that pol II is independently crosslinked to URA3 and ARS1.

Initially we performed experiments with pARS1wtA/2.2 kb/URA3 in MCM5, rpb1104mcm5, and mcm5 cells (Fig 8B). In all immunoprecipitates we observed very low or no signals from the amplification of the CEN4 and the 2.2-kb fragment as compared to the strong input signals (Fig 8B, lanes 6, 12, and 18). Similar low signals from the amplification of the URA3 and ARS1 elements were detected in the control immunoprecipitates without crosslinking (Fig 8B, lanes 4, 10, and 16) and with control antibody (Fig 8B, lanes 5, 11, and 17). In the anti-pol II CTD precipitates there was a clear increase in the signals resulting from amplification of the URA3 and ARS1 fragments (Fig 8B, lanes 6, 12, and 18). These results indicate that pol II was independently crosslinked to ARS1 and URA3. We did not attempt to measure the amounts of immunoprecipitated DNA relative to the input signal between the three strains because the truncation of the CTD in rpb1104mcm5 could contribute to altered efficiency of immunoprecipitation with the anti-CTD antibody. In addition, the proportion of minichromosomes relative to genomic DNA between the three strains is different (see Fig 3), which could further complicate the interpretation of data.

We performed similar experiments with pARS1/-B23/G24/2.2 kb/URA3 in MCM5 cells grown on glucose and galactose, respectively (Fig 8C). It was previously shown that the GAL4-binding site, which replaces the B3 element in wild-type ARS1, can be activated when cells are grown on galactose (MARAHRENS and STILLMAN 1992 ; LI et al. 1998 ). In Fig 8C we show virtually no crosslinking of pol II to pARS1/-B23/G24 when cells are grown in glucose (Fig 8C, lane 6). In cells grown in galactose we observed signals from crosslinking of pol II to pARS1/-B23/G24, which were significantly higher than the signals from the 2.2-kb linker fragment (Fig 8C, lane 12). This result indicates that the crosslinking of pol II to ARS1 is mediated by association of a trans-activator to the B1 element.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

RNA polymerase II is involved in regulation of origins of DNA replication:
Previous studies have indicated that in S. cerevisiae transcriptional activators regulate origins of DNA replication (MARAHRENS and STILLMAN 1992 ; LI et al. 1998 ; HU et al. 1999 ). Furthermore, artificial recruitment of pol II holoenzyme/pol II transcription factors and also pol III transcription factors can substitute for the function of transcriptional activators (STAGLJAR et al. 1999 ; BODMER-GLAVAS et al. 2001 ). Whereas the recruitment of pol II complexes can be linked to their ability to bind trans-activators, the way by which pol III complexes stimulate replication is not clear. In all these cases the likely cause of stimulation is chromatin remodeling (HU et al. 1999 ; BODMER-GLAVAS et al. 2001 ); however, the mechanisms by which remodeling factors are recruited under normal conditions are not known.

The present study provides a significant advancement toward understanding these mechanisms. First, we show that RNA polymerase II is recruited to ARS1 (Fig 8). Second, we show a genetic interaction between a component of the prereplicative complex, MCM5, and the CTD of pol II (Fig 1 Fig 2 Fig 3). Third, we correlate the phenotypes of rpb104mcm5 and mcm5 to the stability of an ARS1/CEN4 minichromosome and to the response of an artificial origin of replication (ARS1/-B23/G24) to trans-activators (Fig 5). Taken together, these experiments indicate that RNA polymerase II could be directly involved in regulating origins of DNA replication. Truncation of the pol II CTD improves chromosome stability in the mcm5 strain (Fig 2 and Fig 3), but decreases chromosome stability if the CTD is truncated in strains with no mutations in MCM5 [rpb1104MCM5, pYwt(10), pYwt(12); Fig 3 and Fig 6]. Thus, truncation of the pol II CTD can have a positive or a negative effect on minichromosome stability depending on the genetic context of the strain. While the actual mechanism by which pol II exerts these effects on replication origins is still enigmatic, one possibility is that the correct recruitment and arrangement of the chromatin remodeling factors is mediated at least in part by pol II.

Previous reports have indicated that some CTD deletions, which leave 8–14 heptapeptad repeats (NONET et al. 1987 ; WEST and CORDEN 1995 ) in rpb1, produce conditional ts, cs, or other growth phenotypes presumably because pol II is not transcribing correctly in vivo. Our results suggest that some of these phenotypes could be caused in part by concomitant effects on DNA replication. The subtle decrease in minichromosome stability in strains with 10–12 CTD heptad repeats [rpb1104MCM5, pYwt(10) and pYwt(12); Fig 3 and Fig 6] could explain why these effects had not been noticed in screens for mcm mutants (TYE 1999B ) and in analyses of strains with CTD truncation where minichromosome stability has not been exclusively tested.

Does the CTD truncation directly affect DNA replication?
A central issue in this study is whether the partial deletions of CTD directly affected DNA replication or if the observed effects were a consequence of aberrant pol II transcription. It is important that the minichromosome assay directly measures efficiency of DNA replication and is independent of other phenotypes that may be associated with deficiencies in DNA replication or in other processes (TYE 1999B ). Therefore, if truncation of the CTD affects minichromosome stability indirectly, it should be through altered expression of genes, which are directly involved in DNA replication. We therefore conducted microarray analyses to compare the expression profiles of mcm5 and rpbD104mcm5 and of Z551 and pY1WT(10), respectively. We did not see any altered gene expression that could explain the differences in minichromosome stability between Z551 and pY1WT(10) (see supplementary data at http://www.uoguelph.ca/mbgwww/faculty/yankulov/appendix_ky082001/appendix_ky082001.html). A more complex situation was observed between mcm5 and rpb104mcm5 (see supplementary data at http://www.uoguelph.ca/mbgwww/faculty/yankulov/appendix_ky082001/appendix_ky082001.html). Several genes (ZDS1, CYC8, TUP1, POP2, SPT5, SPT8, SNF5, and GAL11), which function in repression/activation of transcription and in chromatin remodeling (http://www.proteome.com/databases/YPD/YPDsearch-quick.html and the references therein), are upregulated in rpb1104mcm5 vs. mcm5 at both 30° and 37°. To our knowledge these genes have never been implicated in direct regulation of DNA replication. SNF5 encodes a component of the SWI/SNF global transcription activator complex. Inactivation of SWI/SNF specifically cripples the maintenance of minichromosomes containing ARS121, but not the maintenance of ARS1, ARS309, or ARS307 minichromosomes (FLANAGAN and PETERSON 1999 ). The overexpression of another gene, ZDS1, can enhance the stability of linear, but not circular minichromosomes (ROY and RUNGE 1999 ). Because we use circular ARS1 containing plasmids (pARS1/wtA and pARS1/-B23/G24) upregulation of SNF5 and ZDS1 is unlikely to directly influence their maintenance. Three other genes that function in cell growth and DNA replication were expressed above the twofold-increase threshold in rpb1104mcm5 vs. mcm5 at 37° (see supplementary data at http://www.uoguelph.ca/mbgwww/faculty/yankulov/appendix_ky082001/appendix_ky082001.html). POL32 encodes a subunit of DNA polymerase . It is present at higher concentrations than the catalytic subunit POL3 and its overproduction in vivo does not result in an increase of DNA polymerase activity (BURGERS and GERIK 1998 ). PSP1 is a suppressor of a cdc17 mutation (FORMOSA and NITTIS 1998 ) and its overproduction leads to growth inhibition (AKADA et al. 1997 ). YAK1 is a protein kinase that might work in controlling exit from G₁ to G₀ (GARRETT et al. 1991 ). Its overexpression also inhibits cell growth (GARRETT et al. 1991 ). Our current understanding of the function of these genes argues that their increased production cannot enhance cell growth and DNA replication in rpb1104mcm5. While the possibility that changes in gene expression contribute to the observed characteristics of rpb1104mcm5 is still applicable, we have not obtained any evidence pointing in this direction. In addition, subtle limitations in the production of some replication factor(s) that might be caused by truncation of the CTD cannot explain why we see positive or negative effects in different genetic contexts (Fig 3 and Fig 6). This reasoning leads us to the hypothesis that the truncation of pol II CTD could affect DNA replication independently of transcription.

Mechanism of CTD effects on DNA replication:
A simple and straightforward explanation of our results is that pol II is recruited to origins of replication where its CTD participates in the formation of prereplicative complexes (Fig 9). This idea is in tune with observations in several previous studies. For example, artificial tethering of pol II holoenzyme (in which the CTD plays a central role; KIM et al. 1994 ; HENGARTNER et al. 1995 ; MYERS et al. 1997 ) to ARS1 significantly stimulates origin function (STAGLJAR et al. 1999 ; BODMER-GLAVAS et al. 2001 ). In addition, a tight correlation between the potency of trans-activators to enhance pol II transcription and DNA replication has been well documented (MARAHRENS and STILLMAN 1992 ; LI et al. 1998 ; HU et al. 1999 ; LI 1999 ). Most of the trans-activators used in these studies directly interact with pol II holoenzyme in the absence of promoter DNA (GOLD et al. 1996 ; ANDERSON et al. 1998 ; NEISH et al. 1998 ; YANKULOV et al. 1999 ) and stimulate transcription presumably by recruiting pol II holoenzyme to promoters. It is conceivable that the trans-activators can establish the same contacts at origins of DNA replication. If this is the case, truncation of CTD, which destabilizes pol II holoenzyme, will prevent the effect of trans-activators on DNA replication, as is the case in the experiment shown in Fig 5, and will impair plasmid maintenance mediated by Abf1, as shown in Fig 3 and Fig 6. In addition, recruitment of pol II holoenzyme components that bind the CTD would have a stimulatory effect on origin function as shown in Fig 7.

View larger version (17K): [in this window] [in a new window] [Download PPT slide]   Figure 9. A model for the possible interaction between MCM5 and pol II CTD. TA, trans-activator; ORC, origin recognition complex; CTD, carboxyterminal domain of pol II.

Our hypothesis presumes a contact between pol II holoenzyme and the MCM protein complex, which could explain the genetic interaction between mcm5 and the pol II CTD (Fig 1 Fig 2 Fig 3). Such a contact has been described in metazoan cells (YANKULOV et al. 1999 ); however, it was not revealed in yeast extracts by the methods we used in higher eukaryotes (X. SONG and K. YANKULOV, unpublished data). Alternatively, pol II holoenzyme and MCM proteins may independently bind a factor(s), which is important for origin function. In this case improper contacts with either MCM5 or CTD would have a negative effect on minichromosome stability, as is the case in the mcm5, rpb1104MCM5, pYwt(10), and pYwt(12) strains, while mutations in both MCM5 or CTD could reverse this effect, as is the case in rpb1104mcm5. Both possibilities suggest the intriguing idea that in S. cerevisiae there is a mechanism of coordinating pol II transcription and DNA replication, which is mediated by the CTD of pol II. More in vivo studies are needed to address this question in detail.

	FOOTNOTES

¹ These authors contributed equally to this study.
² Present address: Department of Biochemistry and Molecular Biology, University of Calgary, 3330 Hospital Dr. N.W., Calgary AB T2N-4N1, Canada.
³ Present address: Pharmacia Corporation, AA215/AA2C, 700 Chesterfield Parkway, Chesterfield, MO 63198.

	ACKNOWLEDGMENTS

We thank B. Tye, R. Scalfani, J. Corden, R. Young, D. Mangroo, and R. Lu for yeast strains; R. Li for the pARS1/wtA and pARS1/B23/G24 plasmids; and J. Bag, A. Wildeman, D. Evans, L. Holland, and J. Philips for valuable suggestions and discussion. This study was supported by a grant from the Canadian Institutes of Health Research (MOP-36371) to K. Yankulov.

Manuscript received May 15, 2002; Accepted for publication August 29, 2002.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AKADA, R., J. YAMAMOTO, and I. YAMASHITA, 1997  Screening and identification of yeast sequences that cause growth inhibition when overexpressed. Mol. Gen. Genet. 254:267-274.[Medline]

ANDERSON, S. F., B. P. SCHLEGEL, T. NAKAJIMA, E. S. WOLPIN, and J. D. PARVIN, 1998  BRCA1 protein is linked to the RNA polymerase II holoenzyme complex via RNA helicase A. Nat. Genet. 19:254-256.[Medline]

APARICIO, O. M., D. M. WEINSTEIN, and S. P. BELL, 1997  Components and dynamics of DNA replication complexes in S. cerevisiae: redistribution of MCM proteins and Cdc45p during S phase. Cell 91:59-69.[Medline]

BELL, S. P. and B. STILLMAN, 1992  ATP-dependent recognition of eukaryotic origins of DNA replication by a multiprotein complex. Nature 357:128-134.[Medline]

BENNETT-COOK, E. R. and J. A. HASSELL, 1991  Activation of polyomavirus DNA replication by yeast GAL4 is dependent on its transcriptional activation domains. EMBO J. 10:959-969.[Medline]

BODMER-GLAVAS, M., K. EDLER, and A. BARBERIS, 2001  RNA polymerase II and III transcription factors can stimulate DNA replication by modifying origin chromatin structures. Nucleic Acids Res. 29:4570-4580.[Abstract/Free Full Text]

BURGERS, P. M. and K. J. GERIK, 1998  Structure and processivity of two forms of Saccharomyces cerevisiae DNA polymerase delta. J. Biol. Chem. 273:19756-19762.[Abstract/Free Full Text]

BURKE, T. W., J. G. COOK, M. ASANO, and J. R. NEVINS, 2001  Replication factors MCM2 and ORC1 interact with the histone acetyltransferase HBO1. J. Biol. Chem. 276:15397-15408.[Abstract/Free Full Text]

CHAO, D. M., E. L. GADBOIS, P. J. MURRAY, S. F. ANDERSON, and M. S. SONU et al., 1996  A mammalian SRB protein associated with an RNA polymerase II holoenzyme. Nature 380:82-85.[Medline]

CHEN, Y., K. M. HENNESSY, D. BOTSTEIN, and B. K. TYE, 1992  CDC46/MCM5, a yeast protein whose subcellular localization is cell cycle-regulated, is involved in DNA replication at autonomously replicating sequences. Proc. Natl. Acad. Sci. USA 89:10459-10463.[Abstract/Free Full Text]

COCKER, J. H., S. PIATTI, C. SANTOCANALE, K. NASMYTH, and J. F. DIFFLEY, 1996  An essential role for the Cdc6 protein in forming the pre-replicative complexes of budding yeast. Nature 379:180-182.[Medline]

DANI, G. M. and V. A. ZAKIAN, 1983  Mitotic and meiotic stability of linear plasmids in yeast. Proc. Natl. Acad. Sci. USA 80:3406-3410.[Abstract/Free Full Text]

DEPAMPHILIS, M. L., 1988  Transcriptional elements as components of eukaryotic origins of DNA replication. Cell 52:635-638.[Medline]

FLANAGAN, J. F. and C. L. PETERSON, 1999  A role for the yeast SWI/SNF complex in DNA replication. Nucleic Acids Res. 27:2022-2028.[Abstract/Free Full Text]

FORMOSA, T. and T. NITTIS, 1998  Suppressors of the temperature sensitivity of DNA polymerase alpha mutations in Saccharomyces cerevisiae. Mol. Gen. Genet. 257:461-468.[Medline]

GARRETT, S., M. M. MENOLD, and J. R. BROACH, 1991  The Saccharomyces cerevisiae YAK1 gene encodes a protein kinase that is induced by arrest early in the cell cycle. Mol. Cell. Biol. 11:4045-4052.[Abstract/Free Full Text]<