Evolutionary complementation for polymerase II CTD function

Evolutionary Complementation for Polymerase II CTDFunction

JOHN W. STILLER*, BETTY L. McCONAUGHY AND BENJAMIN D. HALL

Departments of Genetics and Botany, University of Washington, Box 357360, Seattle, WA 98195, USA

The C-terminal domain (CTD) of the largest subunit (RPB1) of eukaryotic RNA polymerase II is essential for polII function and has been shown to play a number of important roles in the mRNA transcription cycle. The CTD iscomposed of a tandemly repeated heptapeptide that is conserved in yeast, animals, plants and several protistanorganisms. Some eukaryotes, however, have what appear to be degenerate or deviant CTD regions, and others haveno CTD at all. The functional and evolutionary implications of this variation among RPB1 C-termini is largelyunexplored. We have transformed yeast cells with a construct consisting of the yeast RPB1 gene with 25 heptadsfrom the primitive protist Mastigamoeba invertens in place of the wild-type CTD. The Mastigamoeba heptads differfrom the canonical CTD by the invariable presence of alanines in place of threonines at position 4, and in place ofserines at position 7 of each heptad. Despite this double substitution, mutants are viable even under conditions oftemperature and nutrient stress. These results provide new insights into the relative functional importance of severalof the conserved CTD residues, and indicate that in vivo expression of evolutionary variants in yeast can provideimportant clues for understanding the origin, evolution and function of the pol II CTD. Copyright # 2000 JohnWiley & Sons, Ltd.

KEY WORDS Ð Saccharomyces; Mastigamoeba; RNA polymerase II; CTD; evolution

INTRODUCTION

The largest subunits of multi-subunit DNA-dependent RNA polymerase enzymes are homo-logous in all living cells. They share eight highlyconserved domains, designated regions A±H(Jokerst et al., 1989), most of which can bealigned among all eukaryotic and prokaryotichomologues (PuÈhler et al., 1989). The carboxyl-terminus of the largest subunit of eukaryoticpolymerase II (RPB1), however, contains anadditional conserved sequence that is absentfrom pol I, pol III and prokaryotic polymeraseenzymes (Corden, 1990). This C-terminal domain,or CTD, has become a focal point of research intoa variety of processes carried out by pol II andrelated transcription co-factors (for reviews, seeCarlson, 1997; Steinmetz, 1997; Shilatifard, 1998;Hampsey, 1998).

The CTD is composed of tandemly repeatedheptapeptides with the consensus sequence Tyr1±Ser2±Pro3±Thr4±Ser5±Pro6±Ser7. In animals andyeast, where it has been studied most extensively,the CTD is linked functionally to both pre- andpost-initiation steps in the mRNA transcriptioncycle. In its hypophosphorylated state, the CTD isassociated with the multi-protein mediator duringformation of the transcription initiation complex(Kim et al., 1994; Svejstrup et al., 1997), and isbelieved to enhance initiation through interactionswith chromatin remodelling SWI/SNF proteins(Wilson et al., 1996; Myer and Young, 1998) andspeci®c transcription activators (Scafe et al., 1990;Liao et al., 1991; Okamoto et al., 1996).

Once the initiation complex has been assembledand transcription begun, phosphorylation ofserine, tyrosine and threonine residues leads toan exchange of initiation-related proteins forelongation factors that bind to the hyperphos-phorylated CTD (Shilatifard, 1998; Otero et al.,1999). This change in the phosphorylation state ofthe CTD appears to trigger the release of pol IIfrom the promoter region, and a switch to

*Correspondence to: J. W. Stiller, Department of Genetics,University of Washington, Box 357360, Seattle, WA 98195,USA. E-mail: [email protected]

YEAST

Yeast 2000; 16: 57±64.

Received 18 May 1999Accepted 17 August 1999

CCC 0749-503X/2000/010057±08$17.50Copyright # 2000 John Wiley & Sons, Ltd.

processive elongation (Greenleaf, 1993; Kang andDahmus, 1995). When hyperphosphorylated, theCTD also interacts with a wide array of co-factorsthat couple pol II transcription to mRNA proces-sing, including 5k-cap formation, intron-splicing,termination, polyadenylation, cleavage and DNArepair (Steinmetz, 1997; Shilatifard, 1998). Bycoordinating and facilitating pol II interactionswith processing-related factors, the CTD acts asa staging platform that helps to increase theef®ciency and quality of the transcript processing(Steinmetz, 1997); recent experiments show that italso can take the more active role as a co-factor insome processing reactions (Hirose and Manley,1998).

A canonical CTD, consisting of the repeatedY±S±P±T±S±P±S heptapeptide, is found in allanimal, plant and fungal pol II largest subunitsexamined to date, as well as in several protistanorganisms that are thought to be related closely tothese multicellular groups (Lam et al., 1992; Stillerand Hall, 1998); however, a growing number ofgenes isolated from diverse protists and algaedo not have C-terminal repeats. For example, theRPB1 C-termini of two different red algae containCTD-like motifs, but there is no evidence ofthe tandemly repeated heptad register (Stillerand Hall, 1998) that is a conserved feature ofthe canonical CTD (Corden, 1990). C-terminalsequences from putatively more ancient protists,such as Giardia, Trichomonas and trypanosomids,show almost no evidence of a CTD-like structure(Evers et al., 1989a, 1989b; Smith et al., 1989;Quon et al., 1996). At present, neither theevolutionary nor the functional signi®cance ofthese differences for Pol II transcription is under-stood.

In the course of our research into the evolu-tionary diversity of RPB1 genes among eukaryoticorganisms, we isolated a unique but repetitiveC-terminal sequence from the amitochondriateprotist, Mastigamoeba invertens. RPB1 fromMastigamoeba encodes 25 nearly identical hep-tads, but with the deviant sequence Y±S±P±A±S±P±A. Phylogenetic analyses and examination ofcodon usage are most consistent with an origin ofthese repeats independent from those of thecanonical CTD found in higher eukaryotes (Stilleret al., 1998). Very little is known about themolecular biology or evolutionary history of thisenigmatic amoeba; however, based on cytologicalfeatures and some molecular data, Mastigamoebamay belong to one of the earliest branches of

eukaryotic evolution (Stiller et al., 1998). Whetherthese tandem heptads can carry out the functionsof a canonical CTD, and what their role is in thetranscription cycle of such a simple and ancientunicellular organism, is unclear.

With its completely sequenced genome, rela-tively well-characterized mRNA transcriptioncycle, and established techniques for geneticmanipulation, Saccharomyces cerevisiae providesa model system for investigating the mechanisticimplications of broad-scale CTD variation amongeukaryotic organisms. Here we report results ofthe in vivo substitution, in yeast, of the Mastiga-moeba RPB1 C-terminus for the wild-type CTD.Our results provide new data regarding therelative importance of several conserved aminoacid positions in the yeast heptad; they alsosuggest that valuable insights into pol II functionand evolution can be gained by mining the worldof eukaryotic diversity for variant CTD sequences.

MATERIALS AND METHODS

Construction of the C-terminal substitution mutant

The RPB1 gene and 3k untranslated sequencewas isolated from Mastigamoeba using a PCR-based genomic-walking strategy described pre-viously (Siebert et al., 1995; Stiller et al., 1998).A clone encompassing the conserved G regionthrough the putative stop codon was used as PCRtemplate with primers designed to amplify thesequence encoding the 25 Mastigamoeba heptads(Figure 1). These primers (MiCTF: gataCCCGAG-

cgtccccggctcccctgctgcc; MiCTR: tcaCTCGGGatcg-taaggctccgctggatc) each include a different 5ktermimal AvaI restriction site recognition sequence(shown in capital letters), to allow directionalcloning of the Mastigamoeba repeats into a CTD-less yeast gene (see below). The resulting productswere digested with AvaI, and column puri®ed(Qiagen, Chatworth, CA) to remove primers andshort, terminal restriction fragments.

The Z26 yeast strain, as well as the pY1 shuttlevector (containing wild-type yeast RPB1) andpSBO (CTD-less subclone) plasmid that wereused to construct the yeast/Mastigamoeba hybridRPB1 gene, were all gifts from J. Corden and areexplained fully by West and Corden (1995). Thedigested Mastigamoeba PCR product describedabove and here designated `MI' was inserted intothe AvaI-digested pSBO subclone (Figure 2) andsequenced completely to ensure that there were no

58 J. W. STILLER, B. L. McCONAUGHY AND B. D. HALL

Copyright # 2000 John Wiley & Sons, Ltd. Yeast 2000; 16: 57±64.

shifts in reading frame or PCR-induced mutations.This pSBO±MI construct and the pY1 shuttlevector were both digested with a combination ofKpnI and SnaBI and the resulting fragments wereisolated and puri®ed on a 0.8% low-melt agarosegel. The KpnI±SnaBI fragment containing theMastigamoeba repeats was then ligated to thepY1 fragment containing the yeast RPB1 sequencewithout a CTD. The resulting clone, designatedpY1MI (Figure 2), was sequenced over the fulllength of the Kpn1±SnaB1 insert, including itsligation sites, to assure that the sub-cloning hadnot produced incidental mutations.

Transformation of yeast with CTD mutant

The yeast strain Z26 (Nonet et al., 1987) wastranformed with the pY1MI construct by lithium±acetate treatment (Ito et al., 1983) and selectedon synthetic complete (SC)-Leu-Ura medium.Leu+Ura+ transformants, which must containboth the URA3/RPB1 wild-type plasmid and theLEU2/pY1MI mutant plasmid, then were selectedon SC medium containing 5-¯uoro-orotic acid (5-FOA) (Boeke et al., 1987) and allowed to grow at30uC for several days. To survive, cells must losethe wild-type RPB1 gene linked to URA3; there-

fore, only cells incorporating the pY1MI mutantRPB1 into pol II should remain. As a positivecontrol, and for subsequent phenotypic compar-isons, cells were transformed in parallel with thepY1 vector containing intact wild-type yeastRPB1.

To be certain that the wild-type plasmid hadbeen lost, transformants were replica-plated alongwith a URA3/wild-type control, on YPD (2% yeastextract, 1% Bacto-Peptone, 2% glucose), SC±Leu,SC±Ura and SC+5-FOA. RPB1 sequencesfrom mutant colonies were PCR-ampli®ed andsequenced using primers that hybridize to authen-tic yeast sequences on either side of the insertedMastigamoeba heptads. These primers will amplifyboth the wild-type and pY1MI mutant sequencesif either is present, permitting a test for anypossible formation of hybrid yeast±MastigamoebaC-termini by recombination.

Characterization of mutant phenotype

Yeast cells containing the mutant RPB1 C-terminus were grown at 38uC and 15uC to deter-mine whether they were temperature-sensitive.Because proportional growth between wild-typeand mutant cells appeared to vary depending ontemperature, absolute growth rates were deter-mined for two to six replicate cultures of both pY1

Figure 1. The aligned coding sequence of the 25 tandemC-terminal repeats from Mastigamoeba. The consensus transla-tion is provided above each position and the single non-synonymous substitution (Ser2 to Gly) in heptad 22 is shown inbold.

Figure 2. Construction of pSBO±MI and pY1MI. TheMastigamoeba sequence is indicated by a black box in allthree panels. (A) Addition of ¯anking AvaI restriction sites tothe Mastigamoeba heptads (MI) by PCR, for subcloning intopSBO vector. (B) Subcloning of MI into the pSBO plasmidcontaining the yeast C-terminus with heptads previouslyremoved. (C) Insertion of the pSBO±MI construct into thepY1 shuttle vector by directional cloning of the KpnI±SnaB1fragment.

EVOLUTIONARY COMPLEMENTATION FOR pol II CTD 59


and pY1MI transformants at 15uC, 30uC and38uC. Flasks containing 100 mL of YPD wereinoculated with equal concentrations of wild-typeand mutant cells and allowed to acclimate for 1 h;1±2 ml of each culture then was sampled every2 h until cultures approached stationary phase.Because of slower growth at 15uC, samples weredrawn every 4 h at that temperature. Cell densitieswere measured by spectrophotometry (OD660) andplotted; doubling times, during log-phase, foreach culture were determined empirically fromthe resulting growth curve. Some CTD truncationmutants have been shown to be inositol auxo-trophs (Nonet and Young, 1989); therefore, wealso tested for conditional growth on SC-inositolplates at the same three temperatures.

RESULTS AND DISCUSSION

Our initial examination of an in vivo yeast trans-formation system, for the purpose of analysing thefunctional implications of RPB1 C-terminal varia-tion among eukaryotes, yielded promising results.Corden and collaborators have developed thissystem for expression of CTD mutants and haveused it to examine the effects of both CTDsubstitutions and deletions on yeast pol II func-tion (West and Corden, 1995; Yuryev and Corden,1996). Truncation of the CTD to less than eightrepeats, as well as substitution of the Tyr1 residueby phenylalanine, were shown to be lethal (Westand Corden, 1995); cells in which serine residues ateither position 2 or 5 are substituted with a non-phosphorylatable amino acid are also inviable(Yuryev and Corden, 1996). These and otherresults (Nonet et al., 1987) have demonstrated thecritical role of CTD length and preservation ofcertain key residues for yeast viability. ThepY1MI construct permitted us to test whethertwo additional conserved residues are also essen-tial for in vivo transcription.

Substitution of conserved CTD residues

Yeast cells tranformed with pY1MI are viable,growing at approximately 83% of the rate of thewild-type at 30uC (Figure 3). Replica-plating onYPD, SC-Ura and SC-Leu media demonstratedthat transformed cells carry only the mutantRPB1 gene. Furthermore, PCR ampli®cation andsequencing of the C-terminus of ®ve independentlytransformed clones showed only the pY1MIconstruct, but not the wild-type sequence or any

recombinant between the two (results not shown).Therefore, unlike the Tyr1, Ser2 and Ser5 residues,the conserved Thr4 and somewhat less conservedSer7 (Corden, 1990) are not essential for pol IItranscription in vivo. This is true even though theseresidues were substituted simultaneously. More-over, the mutants are viable at both 15uC and38uC (Figure 3 and see below), as well as in theabsence of inositol. Only when cells were bothdeprived of inositol and grown in the cold did themutant CTD appear to be lethal. No growth wasobserved on SC-inositol at 15uC after 3 weeks;however, wild-type cells grew extremely slowlyunder these conditions.

Phosphorylation of the CTD has been studiedextensively because it is correlated with mRNAtranscription initiation and the shift to elongation.Since alanines occur invariably at both substitutedpositions of the Mastigamoeba repeats, phosphor-ylation of neither Thr4 nor Ser7 is essentialfor mRNA transcription in yeast. Both in vivomeasurements of phosphoserine : phosphothreo-nine ratios (Zhang and Corden, 1991) and assaysof CTD kinase activities in vitro (Trigon et al.,1998) suggest that the Thr4 residue may not bephosphorylated during the normal transcriptioncycle. Although at least one CTD kinase canphosphorylate the serine at position 7 in vitro

Figure 3. Absolute growth rates (hatched bars) and growthrates relative to wild-type (solid bars) of pY1MI mutants atthree different temperatures. The left vertical axis scales growthrate as 1/doubling time in hours; the right axis indicates thepercentage of wild-type growth at the given temperature.Although mutant cells grow more slowly at high temperaturethan at 30uC in absolute terms, they do proportionally betterin relation to wild-type at 38uC.



(Trigon et al., 1998), our results indicate that nophosphoserine7 residues are necessary. It shouldbe noted, however, that the phosphorylation stateof the CTD in vivo appears to be labile in responseto growth-related changes and environmentalfactors (Dubois et al., 1997; Patturajan et al.,1998; Trigon et al., 1998). It is possible that eitheror both of the Thr4 and Ser7 residues are essentialphospho-acceptors under conditions other thanthose used in this study.

Responses to high and low temperatures

Although viable, cells transformed with pY1MIgrew at only 63% the rate of wild-type at 15uC(Figure 3). This was substantially less than theirproportional growth rate under more optimalconditions (83% of wild-type at 30uC). Surpris-ingly, transformants did not grow proportionallyworse at higher temperatures. Although theirabsolute growth rate at 38uC was lower than at30uC, pY1MI transformants grew at 89% of therate of wild-type at the higher temperature; that is,signi®cantly better than their comparative growthat the more optimal temperature (Figure 3).

These results are of interest in light of ®ndingsthat heat-shock inactivates the TFIIH-associatedprotein kinase, leads to increased activity of stress-induced CTD kinases and results in a change inthe pattern of CTD phosphorylation (Duboiset al., 1997; EgyhaÂzi et al., 1998; Patturajan et al.,1998). Heat-shock has also been shown todecrease the activity of the major CTD phospha-tase (Dubois et al., 1999). Together, these resultssuggest that changes in the phosphorylation stateof the CTD may be an important controlmechanism of gene expression in response tohigh temperature stress (Dubois and Bensaude,1999). Therefore, the substitution of alanines forThr4 and Ser7 residues may have relatively fewerconsequences for pol II function at high tempera-tures, either because one or both of these positionsare hypophosphorylated in response to heat stress,or because they are less important for substraterecognition by heat-shock-activated kinases thanby kinases present at the cold or permissivetemperatures.

It is also possible that either or both of theseheptad residues do not need to be phosphorylatedunder any conditions, or at any stage, of cellgrowth; rather, they may help to conserve a localphysical environment that is indirectly importantfor CTD function. For example, the subtitution of

Ser5 by threonine decreases phosphorylation byDNA-PK in vitro, despite the fact that this CTD-kinase does not phosphorylate the Ser5 residue(Trigon et al., 1998). These authors suggestedthat changes of charge and/or in the stericenvironment that result from this substitutionmay be responsible for the observed decrease inkinase activity. Therefore, the presence of Thr andSer as potential phospho-acceptors at heptadpositions 4 and 7 may be coincidental. Acumulative inhibition, without complete disrup-tion, of kinase and/or other CTD-related func-tions by such indirect effects, could explain thegenerally reduced growth rates we found in cellscontaining the Mastigamoeba C-terminal repeats.

Evolutionary conservation of non-essentialresidues?

Although the Ser7 is the most commonlysubstituted residue in CTD sequences (Corden,1990) the threonine at position 4 is conservedstrongly in most organisms. For example, of 24well-conserved heptads in yeast, only two do nothave a Thr4 residue. In mammals, only eight of 52repeats have a substitution of this threonine; incomparison, the serine at position 2, which isessential (West and Corden, 1995), is substitutedin 9 of the 52 repeats. The reason for the strongconservation of the Thr4 in most heptads is notimmediately apparent based on our results; how-ever, in natural populations where modi®cation ofthe CTD has been shaped by evolutionary forces,conditions may occur frequently in which theseresidues are essential. Alternatively, although notessential in a physiological sense, the Thr4 and/orSer7 residues may be conserved in evolution due tothe relatively low ®tness associated with their loss.

CTD in eukaryotic evolution

Most RPB1 sequences from protists that arebelieved to represent ancient lineages do notcontain heptad repeats, although all largest sub-units isolated to date have an extended C-terminalsequence beyond that present in Pol I, Pol III orprokaryotic homologues. These primitive pol IIC-termini typically are enriched in the tyrosine,serine, proline and threonine residues that com-prise the CTD (see Stiller and Hall, 1998). It isunclear, however, whether these sequences areCTD regions that have accumulated enoughsubstitutions to make them unrecognizable, orwhether they represent the ancestral condition



from which the CTD arose. This lack of clarity isdue, in part, to the fact that many of theseputatively ancient organisms are parasites andmay represent highly derived forms of otherwisewell-characterized protist groups.

The presence of C-terminal repeats, albeit ofa different sequence, in the non-parasitic butpossibly ancient protist Mastigamoeba raises thepossibility that the CTD dates from the earlieststages of eukaryotic evolution. Based on phylo-genetic analyses of RPB1 sequences (Stiller et al.,1998), however, a number of eukaryotic groupsmust have lost a canonical CTD if the Mastiga-moeba heptads share a common origin with theCTD repeats. Because of its centrality to thetranscription cycle, we consider wholesale loss ofCTD heptads in multiple eukaryotic taxa to beunlikely. Indeed, all plant, animal or fungal RPB1genes yet examined encode a CTD.

A CTD is present even in microsporidianparasites, protists that were once thought to beamong the most ancient eukaryotes due to theirsimple cell structure, but now are widely held to behighly modi®ed fungi (Hirt et al., 1999). Micro-sporidia are greatly reduced as an adaptation totheir intracellular parasitic habit, and have thesmallest documented eukaryotic genomes (Biderreet al., 1995). Their ribosomal RNAs, moleculesthat are strongly conserved in evolution, havebeen modi®ed to an extreme; they are now wellbelow the typical size found even in prokaroticorganisms, having lost all but the most essentialdomains for basal function of the translationmachinery (Peyretaillade et al., 1998). The reten-tion of tandem heptads in the Microsporidia (Hirtet al., 1999) provides further evidence of theimportance of conservation of the CTD in thecourse of eukaryotic evolution.

Despite the indirect evidence against CTD loss,it is not possible as yet to determine whether theMastigamoeba repeats are evolutionarily homo-logous with the canonical CTD. The fact that theMastigamoeba RPB1 C-terminus functions so wellin the yeast system reinforces the possibility ofhomology; however, nothing is known about therole of these heptads in the Mastigamoebatranscription cycle. Likewise, it is unclear whetherheptad-like sequences in red algae (Stiller andHall, 1998) and certain other protists (Li et al.,1989; Gieseke et al., 1991) represent independentevolutionary variants or highly modi®ed CTDs,and whether they are capable of substitutingfunctionally for the canonical heptads.

Far more RPB1 sequences are needed fromdiverse organisms to provide a clearer picture ofthe distribution of heptad repeats throughout theeukaryotic world, and to determine what otherC-terminal variants exist. Furthermore, if thesedeviant sequences are examined in a functionalcontext, much more can be learned than ispossible by simple comparative evolutionarymethods. Our initial results using the Mastiga-moeba C-terminus suggest that in vivo expressionin yeast can be a valuable tool for deciphering theevolutionary history of the C-terminal domain ofRPB1, as well as for providing further insightsinto the function of the CTD in yeast transcrip-tion.

ACKNOWLEDGEMENTS

We thank J. Corden for providing the yeast strainand plasmids, and for advice, and E. C. S.Duf®eld for Mastigamoeba culturing.

REFERENCES

Biderre C, Pages M, Metenier G, Canning EU, VivaresCP. 1995. Evidence for the smallest nuclear genome(2.9 Mb) in the microsporidian Encephalitozooncuniculi. Mol Biochem Parasitol 74: 229±231.

Boeke JD, Trueheart J, Natsoulis G, Fink GR. 1987.5-Fluoro-orotic acid as a selective agent in yeastmolecular genetics. Methods Enzymol 154: 164±175.

Carlson M. 1997. Genetics of transcriptional regulationin yeast: connections to the RNA polymerase II CTD.Ann Rev Cell Dev Biol 13: 1±23.

Corden JL. 1990. Tails of polymerase II. TIBS 15:383±387.

Dubois MF, Bensaude O. 1998. Phosphorylation ofRNA polymerase II C-terminal domain (CTD): a newcontrol for heat-shock gene expression? Cell StressChaparones 3: 147±151.

Dubois MF, Marshall NF, Nguyen VT, Dahmus GK,Bonnet F, Dahmus ME, Bensaude O. 1999. Heatshock of HeLa cells inactivates a nuclear proteinphosphotase speci®c for dephosphorylation of theC-terminal domain of RNA polymerase II. NucleicAcids Res 27: 1338±1344.

Dubois MF, Vincent M, Vigneron M, Adamczewski J,Egly JM, Bensaude O. 1997. Heat-shock inactivationof the TFIIH-associated kinase and change in thephosphorylation sites on the C-terminal domain ofRNA polymerase II. Nucleic Acids Res 25: 694±700.

EgyhaÂzi E, Ossoinak A, Lee JM, Greenleaf AL, MaÈkelaÈTP, Pigon A. 1998. Heat-shock-speci®c phosphoryla-tion and transcriptional activity of RNA polymeraseII. Exper Cell Res 242: 211±221.



Evers R, Hammer A, KoÈck J, Waldemar J, Borst P,MeÂmet S, Cornelissen AWCA. 1989a. Trypanosomabrucei contains two RNA polymerase II largestsubunit genes with an altered C-terminal domain.Cell 56: 585±597.

Evers R, Hammer A, Cornelissen AWCA. 1989b.Unusual C-terminal domain of the largest subunitof RNA polymerase II of Crithidia fasciculata.Nucleic Acids Res 17: 3403±3413.

Giesecke H, Barale JC, Langsley G, Cornelissen AW.1991. The C-terminal domain of RNA polymerase IIof the malaria parasite Plasmodium berghei. BiochemBiophys Res Commun 180: 1350±1355.

Greenleaf A. 1993. Positive patches and negativenoodles: linking RNA processing to trancription.Trends Biochem Sci 18: 117±119.

Hampsey MH. 1998. Molecular genetics of the RNApolymerase II general transcriptional machinery.Micro Molec Biol Rev 62: 465±503.

Hirose Y, Manley J. 1998. RNA polymerase II is anessential mRNA polyadenylation factor. Nature 395:93±96.

Hirt RP, Logsdon JM, Healy B, Dorey MW, DoolittleWF, Embley TM. 1999. Microsporidia are related tofungi: evidence from the largest subunit of RNApolymerase II and other proteins. Proc Natl Acad SciU S A 96: 580±585.

Ito H, Fukuda Y, Murata K, Kimura A. 1983.Transformation of intact yeast cells treated withalkali cations. J Bacteriol 153: 163±168.

Jokerst RS, Weeks JR, Zehring WA, Greenleaf AL.1989. Analysis of the gene ecoding the largest subunitof RNA polymerase II in Drosophila. Mol Gen Genet215: 266±275.

Kang ME, Dahmus ME. 1995. The unique C-terminaldomain of RNA polymerase II and its role intranscription. In Advances in Enzymology and RelatedAreas of Molecular Biology, Vol. II. Meister A (ed).Wiley: New York; 41±77.

Kim Y-J, BjoÈrklund S, Li Y, Sayre MH, Kornberg RD.1994. A multiprotein mediator of transcriptionalactivation and its interaction with the C-terminalrepeat domain of RNA polymerase II. Cell 77:599±608.

Lam TY, Chan L, Yip P, Siu CH. 1992. The largestsubunit of RNA polymerase II in Dictyostelium:conservation of the unique tail domain and geneexpression. Biochem Cell Biol 70: 792±799.

Li W-B, Bzik DJ, Gu H, Tanaka M, Fox BA, InselburgJ. 1989. An enlarged largest subunit of Plasmodiumfalciparum RNA polymerase II de®nes conserved andvariable RNA polymerase domains. Nucleic AcidsRes 17: 9621±9636.

Laio S-M, Taylor ICA, Kingston RE, Young RA. 1991.RNA polymerase II carboxy-terminal domain con-tributes to the response to multiple acidic activators invitro. Genes Dev 5: 2431±2440.

Myer VE, Young RA. 1998. RNA polymerase II

holoenzymes and subcomplexes. J Biol Chem 273:27757±27760.

Nonet M, Sweetser D, Young RA. 1987. Functionalredundancy and structural polymorphism in the largesubunit of RNA polymerase II. Cell 50: 909±915.

Nonet ML, Young RA. 1989. Intragenic and extragenicsuppressors of mutations in the heptapeptide repeatdomain of Saccharomyces cerevisiae RNA polymeraseII. Genetics 123: 715±724.

Okamoto H, Sheline CT, Corden JL, Jones KA, PeterlinBM. 1996. trans-Activation by human immuno-de®ciency virus Tat protein requires the C-terminaldomain of RNA polymerase II. Proc Natl Acad SciU S A 93: 11575±11579.

Otero G, Fellows J, Li Y, de Bizemont T, Dirac AMG,Gustafsson CM, Erdjument-Bromage J, Tempst P,Svejstrup JQ. 1999. Elongator, a multisubunit com-ponent of a novel RNA polymerase II holoenzymefor transcriptional elongation. Mol Cell 3: 109±118.

Patturajan M, Schulte RJ, Sefton BM, Berezney R,Vincent M, Bensaude O, Warren SL, Corden JL.1998. Growth-related changes in phosphorylation ofyeast RNA polymerase II. J Biol Chem 273;4689±4694.

Peyretaillade E, Biderre C, Peyret P, Duf®eux F,Metenier G, Gouy M, Michot B, Vivares CP. 1998.Microsporidian Encephalitozoon cuniculi, a unicellulareukaryote with an unusual chromosomal dispersionof ribosomal genes and a LSU rRNA reduced to theuniversal core. Nucleic Acids Res 26: 3513±3520.

PuÈhler G, Leffers H, Gropp F, Palm P, Klenk H-P,Lottspeich F, Garret RA, Zillig W. 1989. Archae-bacterial DNA-dependent RNA polymerases testifyto the evolution of the eukaryotic nuclear genome.Proc Natl Acad Sci U S A 86: 4569±4573.

Quon DVK, Delgadillo MG, Johnson PJ. 1996.Transcription in the early diverging eukaryoticTrichomonas vaginalis: an unusual RNA polymeraseII and a-amanitin-resistant transcription of protein-coding genes. J Mol Evol 43: 253±262.

Scafe C, Chao D, Lopes J, Hirsch JP, Henry S, YoungRA. 1990. RNA polymerase II C-terminal repeatin¯uences response to transcriptional enhancer sig-nals. Nature 347: 491±494.

Seibert PD, Chenchik A, Kellogg DE, Lukyanov KA,Lukyanov SA. 1995. An improved PCR method forwalking in uncloned genomic DNA. Nucleic AcidsRes 23: 1087±1088.

Shilatifard A. 1998. The RNA polymerase II generalelongation complex. J Biol Chem 379: 27±31.

Smith JL, Levin JR, Ingles CJ, Agabian N. 1989. Intrypanosomes the homologue of the largest subunit ofRNA polymerase II is encoded by two genes and hasa highly unusual C-terminal domain structure. Cell56: 815±827.

Steinmetz EJ. 1997. Pre-mRNA processing and theCTD of RNA polymerase II: the tail that wags thedog? Cell 89: 491±494.



Stiller JW, Duf®eld ECS, Hall BD. 1998. Amitochondri-ate amoebae and the evolution of DNA-dependentRNA polymerase II. Proc Natl Acad Sci USA 95:11769±11774.

Stiller JW, Hall BD. 1998. Sequences of the largestsubunit of RNA polymerase II from two red algaeand their implications for rhodophyte evolution.J Phycol 34: 857±864.

Svejstrup JQ, Li Y, Fellows J, Gnatt A, Bjorklund S,Kornberg RD. 1997. Evidence for a mediator cycle atthe initiation of transcription. Proc Natl Acad SciUSA 94: 6075±6078.

Trigon S, Serizawa H, Weliky Conaway J, ConawayRC, Jackson SP, Morange M. 1998. Characterizationof the residues phosphorylated in vitro by differentC-terminal domain kinases. J Biol Chem 273:6769±6775.

West ML, Corden JL. 1995. Construction and analysisof yeast RNA polymerase II CTD deletion andsubstitution mutants. Genetics 140: 1223±1233.

Wilson JC, Chao DM, Imbalzano AN, Schnitzler GR,Kingston RE, Young RA. 1996. RNA polymerase IIholoenzyme contains SWI/SNF regulators involved inchromatin remodeling. Cell 84: 235±244.

Yuryev A, Corden JL. 1996. Suppression analysisreveals a functional difference between the serines inpositions two and ®ve in the consensus sequence ofthe C-terminal domain of yeast RNA polymerase II.Genetics 14: 661±671.

Zhang J, Corden JL. 1991. Identi®cation of phos-phorylation sites in the repetitive carboxyl-terminaldomain of the mouse RNA polymerase II largestsubunit. J Biol Chem 266: 2290±2296.



Documents

Evolutionary complementation for polymerase II CTD function