UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Mutational and functional analysis of the SPT2 gene of Saccharomyces cerevisiae Lefebvre, Louis 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1994-894237.pdf [ 3.61MB ]
JSON: 831-1.0088077.json
JSON-LD: 831-1.0088077-ld.json
RDF/XML (Pretty): 831-1.0088077-rdf.xml
RDF/JSON: 831-1.0088077-rdf.json
Turtle: 831-1.0088077-turtle.txt
N-Triples: 831-1.0088077-rdf-ntriples.txt
Original Record: 831-1.0088077-source.json
Full Text

Full Text

Mutational and Functional Analysis of theSPT2 gene of Saccharomyces cerevisiae.byLouis LefebvreB.Sc., Université Lava!, 1987A Thesis Submitted in Partial Fulfillment ofthe Requirements for the Degree ofDoctor in PhilosophyinThe Faculty of Graduate StudiesDepartment ofBiochemistry andMolecular BiologyWe accept this thesis as confirmingto the required standardThe University ofBritish ColumbiaDecember 1993© Louis Lefebvre, 1993In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)___________________________Department of IDI’OC2t&94t(S77Y ifv.,iDThe University of British ColumbiaVancouver, CanadaDate_______________DE-6 (2/88)11AbstractThe Saccharomyces cerevisiae SPT2 gene was identified by mutants which are suppressors of Ty and 6insertional mutations at the HIS4 locus. The ability of spt2 mutations to suppress the transcriptional interferencecaused by the 6 promoter insertion his4-9125 correlates with an increase in wild-type HIS4 mRNA levels. TheSPT2 gene is identical to SIN], which codes for a factor genetically defined as a negative regulator ofHOtranscription. Mutations in SPT2/SINI suppress the effects of frans-acting mutations in SWI genes and of partialdeletions in the C-terminal domain of the largest subunit of RNA polymerase II. Nuclear localization and proteinsequence similarities suggested that the SPT2/S1N1 protein may be related to the nonhistone chromosomal proteinHMG1. In order to assess the significance of this structural similarity and identi1r domains of SPT2 functionallyimportant in the regulation of his4-912S I have studied recessive and dominant spt2 mutations created by in vitromutagenesis. Several alleles carrying C-terminal deletions as well as point mutations in the C-terminal domain ofthe SPT2 protein exhibit a dominant suppressor phenotype. C-terminal basic residues necessary for wild-typeSPT2 protein function which are absent from HMG1 have been identified. The competence of these mutant SPT2proteins to interfere with the maintenance of the His (Sptj phenotype of a his4-912ô strain is lost by deletion ofinternal BMG1-like sequences and is sensitive to the wild-type SPT2 gene dosage. Using cross-reactingantipeptide polyclonal antibodies, I demonstrate that the intracellular level of the wild-type SPT2 protein is notaffected in the presence of dominant mutations and furthermore that the reversion of the dominance by internaldeletion of HMG1-like sequences is not mediated by altered production or stability of the mutant polypeptides. Theresults suggest that the products of dominant alleles directly compete with the wild-type protein. On the basis ofprimary sequence similarities, it is proposed that a HMG-box-like motif is required for SPT2 function in vivo andthat this motif also is necessary for the dominant suppressor phenotype exhibited by some mutant SPT2 alleles.111Table of ContentsAbstract iiList of Tables viiList of Figures viiiAcknowledgements ixList of Abbreviations xDedication Xl1. INTRODUCTION 11.1 Spr phenotype: Suppression of 6 and Ty insertional mutations 31.1.1 Structure and transcription of yeast Tyl elements 41.1.2 Ty and 6 insertional mutations 121.1.3 Unlinked suppressors of inactivating Ty and 6 insertional mutations Class I SPT genes Class II SPT genes RNA polymerase II mutations 351.2 Sin phenotype: Suppression of swi mutations 371.2.1 Regulation ofHO gene expression 371.2.2 Positive regulators: SWI genes 391.2.2.1 SWI4 and SWI6, cell cycle regulators 391.2.2.2 SWI5, a mother-cell-specific activator 391.2.2.3 SWI1,2,3, SNF5 and SNF6, transcriptional coactivators 401.2.3 Negative regulators: SIN genes 431.2.3.1 S1N3 and SIN4 431.2.3.2 SIN1 and SIN2, chromatin components 451.3 Srb phenotype: Suppression of RNA polymerase 11 mutations 461.3.1 Yeast RNA polymerase II 471.3.2 Suppression of RNAPII mutations 481.4 Summary and thesis outline 49iv2. MATERIALS AND METhODS 522.1 Materials 522.2 Genetic nomenclature 522.3 Bacterial work 522.3.1 E. coil strains and growth 522.3.2 Transformation ofE coil 532.4 Yeast work 532.4.1 Yeast strains and growth 532.4.2 Genetic methods 552.4.3 Yeast transformation 562.5 Recombinant DNA techniques 572.5.1 Isolation of plasmid DNA 572.5.2 Isolation of single-stranded DNA 572.6 Plasmid constructions 582.6.1 YEp and YCp E.coii-yeast shuffle vectors 582.6.1.1 pLL1O (YEpSPT2j and pLL18 (YEpSPT2-1) 582.6.1.2 pLL32 (YCpSPT2j and pLL24 (YCpSPT2-1) 602.6.1.3 pLL77 602.6.2 spt2 deletion and disruption alleles (plasmids pLL8O, 81, 82) 602.6.3 spt2::lacZ fusion (plasmid pLL43) 622.6.4 pGAL4:SPT2 fusions (plasmids pLL94 to pLL97) 622.6.5 Internal deletion alleles 642.7 DNA sequencing 642.S Oligonucleotide-directed mutagenesis 652.9 Yeast strain constructions 672.10 PCR analysis of SFT2 alleles 692.10.1 Preparation of crude yeast DNA 692.10.2 Analytical PCR 70V2.10.3 Cloning of PCR products 712.10.4 Asymmetric PCR 712.11 Immunological techniques 722.11.1 Production of anti-SPT2-peptide antisera 722.11.2 ELISA2.11.3 Immunoaffimty purification 742.11.4 Immunoblotting 742.11.4.1 Yeast protein extracts 742.11.4.2 SDS-polyacrylaniide gel electrophoresis 752.11.4.3 Protein blotting and immunoprobing 752.12 Genomic DNA analysis 752.12.1 Preparation of yeast genomic DNA 752.12.2 Hybridization in dried agarose matrix 763. RESULTS AND DISCUSSION 773.1 In vivo assay system for SPT2 function and expression 773.1.1 Expression vectors 783.1.2 Isogenic yeast strains 803.1.3 Anti-SPT2-peptide antibodies 873.2 Mutational analysis of the SPT2 protein 913.2.1 Dominant carboxy-terininal deletions 913.2.2 Carboxy-terminal replacements 923.2.3 Definition of a dominance domain 963.2.4 Expression of mutant polypeptides in vivo 993.2.5 Subdomains of the SPT2 protein 1013.3 Characterization of spontaneous recessive spt2 mutations 1103.3.1 Analysis of genomic DNA 1103.3.2 Cloning and sequencing of mutant alleles 111v3.4 Studies on the mechanism of dominance of SPT2 alleles 1153.4.1 Gene dosage studies 1153.4.1.1 Construction of isogemc spt2 and SPT2 yeast strains 1173.4.1.2 Suppression in diploids 1213.4.1.3 Suppression in merodiploids 1223.4.2 Interference of SPT2 function by dominant SPT2 proteins 1244. CONCLUSIONS AND FUTURE DIRECTIONS 1285. APPENDICES 1315.1 NdeI fragment of SPT2 locus 1315.2 Synthetic peptides 1355.3 Glossary of gene symbols 1355.4 Genetic equivalencies 1366. BIBLIOGRAPhY 137List of TablesTable 1. Suppression pattern of spt mutations. 25Table 2. Properties and pleiotropic nature of class I and class II spt mutations. 26Table 3. S. cerevisiae strains used in this study. 54Table 4. Oligonucleotide-directed mutagenesis. 66Table 5. Carboxy-tenninal spt2 mutations. 95Table 6. Analysis of 4 recessive spt2 alleles. 113Table 7. Generation times of wt and mutant SPT2 strains. 120viiList of FiguresFigure 1. Structure of Ty element and solo ö sequence. 5Figure 2. The his4-912ö mutation. 20Figure 3, The SPT2 protein. 32Figure 4. Summary of the transcriptional effects associated with spt2 mutations. 50Figure 5. Construction of plasmids pLL1O (YEpSPT2j and pLL32 (YCpSPT2j. 59Figure 6. Construction of plasmid pLL77 and of the alleles spt2zi and spt2zl:: URA3. 61Figure 7. Construction of plasmid pLL43 (YEpspt2::lacZ). 63Figure 8. Yeast strain constructions. 68Figure 9. SPT2 locus and plasmid constructions. 79Figure 10. Analysis of SPT2 locus from strains SLL5 and SLL7. 82Figure 11. Analytical PCR amplification of the SPT2 and spt2zl:: URA3 alleles. 84Figure 12. PCR analysis of SLL4x7 p+ revertants. 86Figure 13. Immunological detection of the SPT2 protein. 88Figure 14. C-terminal truncations of SPT2 creating nonfunctional, dominant polypeptides. 93Figure 15. Residues 117 to 179 of SPT2 are essential for the dominance of truncated products. 98Figure 16. Production of mutant proteins with altered C-termini in vivo. 100Figure 17. Production of recessive and dominant mutant proteins in vivo. 102Figure 18. Primary structure of the SPT2 protein. 104Figure 19. SPT2 contains two regions of sequence similarity with the 14MG-box motif 107Figure 20. Restriction analysis of the SPT2 locus from mutant strains. 112Figure 21. Screening of gene replacement events at the spt2&: URA3 locus of strain SLL7 118by analytical PCRFigure 22. Western immunoblot analysis of mutant spt2 strains. 121Figure 23. Dosage dependence of the dominant Spr phenotype. 123Figure 24. Effects of dominant alleles onpGAL4::SPT2 function and SPT2 protein levels. 126ixAcknowledgementsMany friends and colleagues have, at one point or another in my training, contributed to the value of mylife as a student at UBC.Many thanks to my thesis supervisor, Michael Smith, for his continuous support and encouragement. Theintellectual stimulation and freedom in Michael’s lab provided me with a unique enviromnent to mature as aresearcher and individual. Michael has always demonstrated a lot of confidence in me. His help andunderstanding often reached beyond the call of duty. Thank you Michael. Special thanks to my thesis committeemembers, Patrick Dennis, Ivan Sadowski, and Caroline Astell, for their support and comments. I am particularlygrateful to Pat and Ivan for revising this manuscript.To my lab mates, Sarbjit Ner, Evan McIntosh, Ian Lorimer, Guy Guillemette, Mark Ring, Tom Atkinson,Jeanette Beatty-Johnson, Stephen Inglis, Chris Overall, Hailun Tang, Chi-Yip Ho and Marianne Huyer, manythanks for their help, friendship, and support. Thanks to Jeanette for her technical assistance and musicalinfluence. Thanks to Yip, with whom, over the years, I shared the frustration and excitement of research. He hasoffered me unconditional help with a philosophical touch.To my colleagues Peter Durovic, Steve Rafferty, and Eve Stringham, thanks for the laughs, dreams, anddiscussions that kept me going. Many thanks to Peter for his meticulous proofreading of the introduction and forsharing with me the possibilities of one’s imagination.My stay in Vancouver would not have been the same without the friendship of Hugh Miller, Jasbir SinghDhaliwal, John Fossum, Mary-Clare Zak, Stuart Robinson, Michel Lafleur, and Hélêne Côté.To my parents Robert and Céline, to whom I owe many years of love and support. With my recent“promotion” to fatherhood, I have come to appreciate more than ever their care, patience and intellectualstimulation.Finally, very special thanks to my partner, Jo-Anne Dillabough, for her love, support and encouragementthroughout these years. Thank you for sharing with me the challenges of a life as a student and a parent. ThanksDominique for your wonderful curiosity and innocence, both of which inspired me until the end.xList of AbbreviationsThe abbreviations used in this thesis are described below. A glossary of the genetic symbols is presentedin Appendix 3.Amp ampicillinbp base pairBSA bovine serum albuminCTD C-terminal domain of RNAPII largest subunit, RPO21 (heptapeptide repeat)DMF dimethylformamideds double-strandedDTT dithiotreitolEDTA ethylenediaininetetraacetic acid5-FOA 5-fluoroorotic acidIPTG isopropyl-f3-D-thiogalactopyranosidekb kilobasekDa kilodaltonLTR long terminal repeatOD optical densityPCR polymerase chain reactionpfu plaque forming unitR punneRNAP RNA polymeraseROAM regulated overproducing alleles responding to mating signalsrpm revolution per minuteSDS sodium dodecyl sulphatess single-strandedTBP TATA binding proteinUAS upstream activating sequence (or activation site)URS upstream regulatory sequenceUV ultraviolet lightVLP virus-like particlevol volumewt wild-typeX-Gal 5-bromo4-chloro-3-indoyl-3-D-galactosideY pyrimidineYCp yeast centromeric vector (low copy number)YEp yeast episomal vector (high copy number)xDedicationA mesparentRobert et Céline,et a ma nouvellefamille,Jo-Anne et Dominique.11. INTRODUCTIONIt has long been established that the regulation of gene expression constitutes a central mechanism for themodulation of cellular activities and fimctions. In multicellular organisms, genetically identical cells follow aprecise developmental program in response to complex spatial and temporal signals, generating the broad spectrumof specialized cellular structures and functions required for the establishment of tissues and organs. Many of thechanges occurring during such a process are likely to correlate with corresponding variations in the pattern of geneexpression such that each cell type may be characterized by a unique transcriptional state.Unicellular organisms, such as the budding yeast Saccharomyces cerevisiae, are equipped with animpressive network of true homeostatic sensors that often mediate their effects via the regulation of geneexpression. Their ability to respond to external stimuli, such as mating pheromones, temperature changes, oxygenlevels, nutrients, metals, etc., relies on complex regulatory pathways often converging towards changes in thelevels of expression of specific genes; again, for each condition a cell may be characterized by a uniquetranscriptional state.Recognition that the machinery responsible for the regulation of gene expression plays such a capital rolein cellular functions has fostered considerable research efforts in the last 30 years. The development of ourunderstanding of these processes since the demonstration of the role of DNA as the genetic material can bedescribed by three important phases.Phase I, mostly dominated by in vitro biochemical analyses and cytogenetics, led to the characterizationof important classes of nuclear enzymes and chromosomal proteins. The nuclear DNA molecule composing eacheukaryotic chromosome was found to be organized in a complex folded structure, known as chromatin. Thestructural role of histones in the formation of nucleosomes and higher levels of folding was established, but suchstatic representation of the packaging of the genetic material provided little insight into the molecular basis of generegulation.In phase U, the principal components of the transcriptional machinery as well as several transcriptionfactors involved in the regulation of gene expression were investigated. The fractionation of nuclear extracts andthe preparation of cell-free transcription systems offered and still offers an important avenue for probing theenzymology of transcriptional initiation, whereas a powerful combination of approaches from the fields of genetics2and molecular biology allowed the identification of several cellular factors affecting gene expression in a specificor general fashion. Transcription factors characterized by these approaches can be shown by genetic evidence toregulate positively and/or negatively the transcription of one or many genes. In organisms such as Saccharornycescerevisiae, it has been possible to develop genetic selection schemes for the identification of mutations affecting theexpression of specific promoter-reporter gene fusions for example, or suppressing specific cis- or trans-actingmutations. Together, these studies not only led to a detailed description of early events in the transcriptionalinitiation process but also revealed the complexity of interactions regulating this process. Although much has beenlearned about the modular structure of several transcription factors and although numerous functionallyindependent protein motifs or domains have been characterized, our current knowledge is far from providing aunifying molecular model of transcriptional control.One would like to exploit the reductionist approach to the point of reconstructing complex regulatoryswitches in vitro without the structural constraints imposed by the chromatin backbone. However, activators andrepressors of transcription, which exert their effects via their close association with other chromatin proteins, mustin fact be considered as true chromatin components. Several recent reviews have emphasized the importantconsequences of the packaging of DNA in nucleosomes on the function of the transcriptional machinery ineukaryotes (2, 71, 84, 126, 225, 229). In what will be tenned phase ifi, the chromatin is seen through its highlydynamic role, providing not only a scaffold or template to the regulatory processes, but also a functional frameworkdynamically involved in several steps leading to the control of transcriptional regulation.The work presented here summarizes mutational and functional studies of the SPT2 protein ofSaccharomyces cerevisiae. The SPT2 gene has been associated with negative transcriptional regulatory activitiesin three different contexts. In each of these systems, a primary cis- or trans-acting mutation interferes with thenormal transcription by RNA polymerase II of one or several yeast genes, by decreasing the level of transcriptionalinitiation from the normal start site(s). The resulting transcriptional defects can be suppressed by second-sitemutations at SPT2, which therefore acts as a modifier locus for these mutations. Several other modifier loci havebeen characterized for each of these primary transcriptional defects, and substantial evidence implicatescomponents of yeast chromatin in suppression. Within this group of suppressor genes, spt2 mutations are uniquein their broader specificity and ability to suppress all of these mutations. Knowledge of the nature of those primarymutations and of their effects on gene expression is necessary to discuss the role of SPT2 in transcriptional3regulation and the possible mechanisms of suppression by spt2 alleles. Emphasis will be put on the Spr phenotypeof spt2 mutations, which led to the characterization of this locus at the molecular level.1.1 Spt phenotype: Suppression of 3 and Ty insertional mutationsMutations at the SPT2 locus were originally identified by their ability to suppress the effects oftransposable element insertions interfering with gene expression. Transposable genetic elements consist ofrepetitive DNA sequences that have the ability to move (transpose) from one site to another in a genome. Initiallycharacterized in maize by genetic studies of unstable phenotypes, such elements have now been identified inprokaryotic and eukaiyotic cells and are thought to be present in the genomes of all higher eukaryotes. The studyof these genetic elements has shown that they belong to different structural and functional families. For thepurpose of the present study, the transposable elements of Saccharomyces cerevisiae belonging to the longterminal repeat (LTR)-containing retrotransposon family, the Ty elements (for transposon yeast), will be the focusof discussion. The term retrotransposition is used to designate the process whereby an element transposes via anRNA intermediate; this process therefore requires the synthesis of a complementary DNA molecule (cDNA) froman RNA template by the enzyme, reverse transcnptase (21). The Ty elements of yeast are closely related toelements of the copia family ofDrosophila, and although neither of these elements have been shown to participatein an infectious life cycle involving an extracellular phase, they do share several structural and functionalcharacteristics with animal retroviruses. One such similarity, the synthesis of a retroviral-like RNA molecule, iscritical to the retrotransposition process and plays a central role in the mutagemc effects of Ty element insertionmutations suppressed by spt2 alleles. Thus, the role of the SPT2 protein in the establishment of the Spt phenotypewill be discussed in relation to our current knowledge of the complex interactions between Ty elements and thehost genome.41.1.1 Structure and transcription ofyeast Tyl elementsSince their discovery, the transposable elements of Saccharomyces cerevisiae have been associated withseveral spontaneous mutations and much effort has been directed towards the understanding of their role asmutagemc agents. Ironically, the first characterization of these elements occurred independently of the phenotypicchanges associated with their insertion, and was based solely on the restriction fragment length polymorphismgenerated by a transpositional event into a new chromosomal site (33). In a study examining the chromosomalstructure of the repeated tyrosine tRNA genes, it was observed that the size of the EcoRI fragment of the sup4IRNATyr gene on chromosome X was highly variable from one strain to another (33). Subfragments of the clonedsup4 locus were then used as probes in Southern blot analyses ofyeast genomic DNA. Whereas some probes werespecific for sup4, others identified a new family of dispersed repetitive DNA sequences present at multiplelocations in the haploid genome. New polymorphisms associated with these repetitive sequences were alsoidentified during continuous growth of a single yeast clone, thus providing evidence for the transposition of theseelements.Following this initial, serendipitous, characterization of representative elements of the Tyl family, severalspontaneous mutations affecting gene expression were shown to be caused by the insertion of similar repeated Tyelements. Analyses of the nature of these mutations, of the role played by the Ty insertion in the resulting mutantphenotype, and of the structure of these elements themselves have contributed to our current understanding of Tyelements and their interactions with the yeast genome. The biology of retrotransposons in general, and of Tyelements in particular, has become an important field of research, and the reader should refer to the numerousreview articles available for a more exhaustive discussion of the subject (20, 22, 146, 184).DNA structure. The transposable elements of yeast all share the same general DNA structure, consisting of acentral domain approximately 5 kb long called epsilon (s), flanked by two direct repeats 330 to 370 bp long, calleddelta () sequences for Tyl and Ty2 elements (Fig. 1; ref. 33, 76). This genetic organization is very similar to thestructure of retroviral proviruses- the integrated form of retroviruses- and is also related to other repeated geneticelements of the retrotransposon family, such as the copia-like elements ofDrosophila (146). Anothercharacteristic shared by these elements is the presence of short duplicated sequences at the chromosomal site ofinsertion. These duplications, found as direct repeats of genomic sequences flanking the elements, are generated5by the transposition process itself and are 5 bp in length in the case of Ty elements (66, 76). By analogy to thenomenclature of retroviruses, the sequences are often called long terminal repeats or LTRs. The Ty elementscharacterized to date fall into 5 families, defined primarily by important sequence variations in the internal region(184, 20). The Ty3 elements (ito 4 copies per haploid genome) and their solo LTRs (called a) have only beenfound associated with tRNA genes. The single Ty5 element characterized so far was identified from the completeDNA sequence of chromosome Ill, where it was found in close association with the left-arm telomere (234). TheTyl and Ty2 families are the most abundant in the haploid yeast genome (25 to 35 and 5 to 15 copies respectively)and will be the focus of the following discussion.Ty 1-912o1 334 5585 5918recombination+soloFigure 1. Structure of Ty element and solo 6 sequence.Shown are the two genomic forms of the Tyi-9i2 element, the full-length LTR-containing element (proviral form),and the solo LTR (9126), obtained by homologous recombination between the terminal repeats. The terminallyredundant Ty RNA is shown initiating in the 5’ö and terminating in the 3’S. The numbering is from the sequenceof the Tyl-912 element.6Although they are subject to numerous rearrangements and were shown to exhibit extensive sequenceheterogeneity, Ty elements generally exist in one of two forms: complete elements with their LTRs, or isolated(solo) LTR sequences. The yeast genome contains more isolated LTR sequences than LTRs associated with fulllength elements. In fact, these solo LTRs may represent the most abundant family of dispersed repetitivesequences in yeast (22). The solo LTRs probably originate from a two-step process involving transposition of acomplete element into a new site, followed by excision of most of the element by homologous recombinationbetween the direct repeats (LTR-LTR recombination, Fig. 1). These solo LTRs represent genetic scars oftransposition events. Although these remnants of Ty insertions could be “repaired” by gene conversion in diploidcells heterozygous for the insertion, the conversion often seems to generate homozygotes for the LTR insertionitself (233). On the other hand, solo LTRs inserted in unique sequences are stable in laboratory haploid strains,where it is thought that the numbers of Ty and solo LTRs are continuously increasing (20). More than 4% of theyeast chromosome III is composed of Ty sequences (234).RNA structure. As mentioned previously, the studies on the Tyl elements inserted at the sup4 locus madeavailable Ty-specific probes which could then be used in Northern blot analysis of yeast RNA (33, 201). Both Tyland ö probes were shown to hybridize to an abundant RNA 5.7 kb long (61). Less abundant species 5.0 kb longand 2.2 kb long have also been identifIed (75). The Tyl RNA was shown to be polyadenylated and to constituteapproximately 0.1% of total yeast RNA (or 5 to 10% of poly[Aj RNA). That this RNA molecule is transcribed byRNA polymerase II (RNAPII) is suggested by studies demonstrating the sensitivity of its synthesis to o-amanitin aswell as to temperature shift in a temperature-sensitive rpo2l/rpbl background (mutation in the largest subunit ofRNAPII; ref. 118 and 165).The synthesis of the Tyl RNA was also found to be under mating-type control: haploid cells of a or cmating type synthesize 20 times more Tyl RNA than Wc diploids (201). Although the physiological andevolutionary significance of this regulation is unclear, Ty elements belong to the family of haploid-specific genes.The additional observation that mating pheromones posttranslationafly inhibit Ty transposition led to the proposalthat yeast cells have evolved mechanisms to limit transposition to the haploid phase, the shortest phase of theyeast’s life cycle in the wild (250). Other factors influencing Ty transcription and/or Ty P.NA levels include7irradiation with ultraviolet light (187), - and perhaps DNA-damaging agents in general (22) - carbon source (222),and metabolic state (201).Even though the function of the Ty RNA was initially unknown, the finding that at least some of the Tyelements of the yeast genome are transcribed constituted an important step in our understanding of the biology ofTy elements. It is now thought that the 5.7 kb long Ty RNA molecules fuffill a dual function, acting as bothtransposition intermediate and messenger RNA. That the transposition of Ty elements occurs through an RNAintermediate was suggested by the retroviral-like structures mentioned earlier and was convincingly demonstratedby a series of elegant experiments utilizing a recombinant Ty element expressed from the inducible GAL] promoter(pGTyH3, ref. 21). These experiments demonstrated that galactose induction stimulates transposition of thiselement and that an intron inserted in the element is efficiently spliced during transposition. The coding potentialof functional Ty RNAs is supported by the increase in virus-like particles (Ty-VLP5), reverse transcriptase activity,and TyA and TyB gene products in induced cells (77, 256).Ty transcription. Analysis of cDNA clones derived from the Tyl RNA established that the 5.7 kb long RNA isterminally redundant and has its 3’ and 5’ ends within the flanking 6 sequences (Fig. 1; ref. 61). The 6 sequences,by analogy to the retroviral LTRs, can be subdivided into three regions on the basis of their respective position inthe resulting RNA molecule: U3, unique to the 3’ end of the RNA; R, present at both tennini; and U5, unique tothe 5’ end. Each 6 sequence contains cis-acting DNA sequences required for transcriptional initiation andtermination. The orientation of these functional sites within the 6 sequences defines the transcriptional polarity ofthe Ty element. The orientation of a specific Ty element therefore refers to the direction of the Ty elementtranscription relative to a chromosomal reference point and is schematically represented by the boxed arrowheadsof 6 sequences (Fig. 1). Previous reports have represented the 6s pointing towards the adjacent gene, and awayfrom Ty transcription (61, 200); this representation has been replaced by the more logical one in which the boxedarrowhead points in the direction of Ty transcription (e.g., see 20 and 245). Since both 6s associated with a Tyelement are often identical in sequence and since the transcriptional initiation site is located upstream of thetermination site within a single LTR, determinants external to the 6 must inactivate the termination signal of the 5’S and the initiation signal of the 3’ 6. Indeed, there is no evidence supporting the production of small transcriptsinitiating and terminating within the 5’S, nor of transcripts emerging from the 3’S into the adjacent chromosomal8sequences. Sequences internal to the e region have been implicated in the activation of transcription from the 5’(see below), but the molecular basis for the silencing of the termination signal in the 5’ remains elusive. It shouldbe noted at this point that solo sequences in which the initiation signal (as in his4-912ô) or the termination signal(as in lys2-1288) is active have been characterized.In spite of the functional redundancy between the two LTRs, numerous studies have focused on theidentification of cis-acting elements required for the initiation and regulation of Ty transcription by RNApolymerase II. Since the mutagenic effects of Ty element insertions are often mediated by an interference withgene expression, the studies of Ty transcription and those of the effects of Ty insertions on adjacent geneexpression are closely linked. The elements directing Ty transcription will be described below - following a briefreview of yeast RNAPII promoters- whereas the role they play in Ty mutagenesis will be discussed in the nextsection (1.1.2).Yeast RItA polymerase II promoters. The transcription of yeast genes by RNA polymerase II typically requiresthree different cis-acting promoter elements: the initiator element (or I site) , the TATA element (or TATA box)and upstream regulatory sequences (URS; ref. 85, 218). In yeast, the TATA box directs the formation of theinitiation complex. Often related to the archetypal motif TATAAA, it is necessary for the transcription of most,but not all, yeast genes. The first step in the formation of the preimtiation complex is thought to be the binding ofthe TATA-box binding factor, component of TFIID, to the TATA box. The actual site of transcriptional initiation(nucleotide +1, corresponding to the 5’ end of the mRNA molecule) is located approximately 60 to 120 bpdownstream of the TATA sequence and is determined by the position of the initiator element. Initiator elementsare less conserved, but most are related to the consensus sequences TCRA or RRYRR. A single TATA box can infact direct initiation at more than one downstream I site (e.g. at CYCJ; ref. 143). The role of the TATA element inyeast is therefore different than in higher eukaryotes, where it is the primary determinant of the start site position,usually at a distance of 25 to 30 bp downstream of the TATA box.In vitro, the TATA box and the initiator are sufficient for accurate, albeit inefficient, transcription, knownas basal transcription. Regulated or activated transcription on the other hand requires additional upstreamregulatory or activating sequences (URS and UAS). These URS or UAS elements are present in most yeastpromoters, located between 100 to 1500 bp upstream of the TATA box. They are required for transcription in vivo9and constitute the major determinants or mediators of transcriptional regulation. Like mammalian enhancersequences, they can function at variable distances upstream of the TATA box, and in an orientation-independentfashion. However, yeast UASs usually do not activate transcription if located downstream of the initiation site. AUAS is composed typically of one or several repeated copies of short DNA sequences (10 to 30 bp long) acting asrecognition sites for sequence-specific DNA-binding proteins. The precise molecular mechanism responsible forthe activation of transcriptional initiation by RNA polymerase II is still unknown, but extensive structure-functionanalyses of the cis- and frans-acting determinants of transcriptional regulation in yeast - i.e. upstream regulatorysequences and the regulatory proteins binding to these elements - have led to the following conclusions:(i) in most yeast promoters, the constitutive or inducible (regulated) activation of transcription requiresshort cis-acting DNA sequences (URS, UAS) located upstream of the TATA box(es).(ii) UASs are modular regulatory determinants: deletion of a particular UAS in a promoter can eliminatea particular regulatory property of this promoter; addition of a foreign UAS to a promoter can add anew regulatory property to this promoter;(iii) the regulatory properties of a particular UAS are mediated by the binding of one or several sequence-specific, DNA-binding activator proteins;(iv) DNA-binding per se is not sufficient for transcriptional activation;(v) activator proteins often display a modular structure consisting of at least two domains, a DNA-binding domain, and an activation domain;(vi) activator proteins can act synergistically, with stronger UASs being composed of more binding sitesfor the same or different activator proteins;(vii) common regulatory pathways often utilize common UASs and therefore common regulatory proteins.The idea that the regulation ofyeast promoter activity is based solely on positive control or activationmechanisms would provide a simple, unifying model of transcriptional regulation. However, the actual situation iscomplicated by the existence of negative control or repression mechanisms. The cis-acting sequences responsiblefor this activity, called operator sequences (or 0 sites) by analogy to the prokaryotic systems, are generally thoughtto function according to the same basic principles as UASs: they can function away from the TATA box, mostoften being located between the TATA box and the UASs, and mediate their activity through the binding of10sequence-specific repressor proteins. The specific expression pattern of a yeast promoter is often the outcome ofcomplex interactions between positive and negative signals acting via its composite regulatory sequences. Thegeneral properties of RNAPII promoters presented here will provide a framework for the discussion of particularregulatory controls affecting the expression of Ty elements and that of other regulated genes such as HIS4 and HO.cis-acting determinants of Ty transcription. Because of the heterogeneity in Ty structure, it is important todefine a reference sequence from which relative positions of cis-acting elements can be assigned. All nucleotidepositions mentioned below refer to the nucleotide sequence of the element Tyl-912, which was the first completeTy element sequenced (40), and which for historical reasons became the archetypal Tyl element. Alignmentsbetween Tyl-912 and other Tyl and Ty2 elements have been presented elsewhere (22, 184). Numbering starts atthe 5’S (nucleotides 1 to 334) and extends through the c region (335 to 5584) to the 36 (5585 to 5918; see Fig. 1).As mentioned previously, the major Ty transcriptional product initiates and terminates in the 5’ and 3’ 3 sequences,respectively. Whereas the active basal promoter elements (I site and TATA box) appear to be located strictlywithin the 5’ 3, regulatory sequences have been identified both upstream and downstream of the initiation site, inthe S and c regions respectively.Primer extension experiments on yeast poly(A)+ RNA using a Tyl primer fragment showed that the major5’ end of Tyl RNA is located approximately at position 241, i.e. 94 bp upstream of the 5’ 3-c boundary (61). Thismajor initiation site corresponds to the first G in the sequence CTTGAG present at nucleotides 238 to 243 of Tyl912. It is located adjacent to an important polymorphism affecting a recognition site for the restrictionendonuclease XhoI (CTCGAG), present in several S sequences from Tyl and Ty2 elements (57, 184). Both ofthese sequences (TTGA and TCGA) are closely related to the consensus sequence TCRA found at many yeastinitiation sites.An AT-rich sequence located 83 bp upstream of the initiation site (position 157 to 174) is thought tocontain the main TATA element for Ty transcription (22). In the Tyl element Tyl-15, deletion of nucleotides 143to 185 of the S sequence abolishes Ty transcription, and the deleted region was proposed to contain or overlap witha TATA box (75). An internal deletion of residues 157 to 177 was also shown to greatly reduce the expression of aTy2-917: :lacZ fusion (137). Finer mutagemc data are also available for representatives of the Tyl and Ty2families. A transition at position 170 ofTy2-917 which changes the sequence ATATAAAA to ATATGAAA11reduces Ty transcription 5- to 59-fold in different hybrid elements (44). In the same region of the Tyl-912 Ssequence (position 166), another A-to-G transition, which changes the sequence ATATAAAA to GTATAAAA,also decreases the level of the S-initiated transcript (108).All the experimental evidence discussed thus far is based on in vivo data correlating specific S mutationswith their effects on Ty and/or Ty-adjacent gene expression. As mentioned above, the TATA element promotesthe formation of the initiation complex by providing a recognition sequence for the general transcription factor‘flJfl)’ A direct biochemical characterization of the TFIID binding sites within the S sequence has recently beenmade possible, following the cloning of the SPTIS gene, encoding yeast TFIID (see section; ref. 60, 90).DNase I protection analysis showed that nucleotides at position 153 to 203 of the Tyl-9125 sequence are protectedin the presence of purified TFIID (9). This region covers two consensus TATA box motifs (TATAAA) at positions159-164 and 167-172.The search for regulatory sequences involved in mating-type control and activation of Ty transcription orTy-adjacent gene expression (see next section) has revealed several unique features of Ty regulation. First, thesearch for typical upstream activating sequences in the S sequence (between positions ito 157) has yieldedconflicting results. Deletion studies on the Ty2 element, Ty2-9 17, has revealed the presence of an activatingsequence at positions 100 to 129 (137). However, similar studies on elements of the Tyl family (Tyi-15 and TyD15) suggested that the S sequence does not contain any UASs (75, 257). These results may point to truedifferences in the transcriptional regulation of Tyl and Ty2 elements and reflect the observed sequenceheterogeneity among elements of the same or different families. It is nevertheless clear from these studies, as wellas from the analysis of solo S insertional mutations, that the S sequence itself does not contain the regulatorydeterminants of mating-type control.Second, cis-acting sequences located downstream of the initiation site and overlapping the translatedsequences of the e region were shown to be necessary for the activation of Ty or Ty-adjacent gene transcriptionand its regulation by the mating-type locus (64, 67, 75, 185, 257). Since they are located downstream of the I site,these elements are not functionally related to typical yeast UASs, but rather appear to function similarly to‘Much confusion recently developed regarding the nomenclature of the RNApolymerase II transcription factor involved in TATA box recognition.As far as the biochemical fractionation oftranscription extracts is concerned, TFIID originally referred to a specific chromatographic fractioncontaining the TATA box recognition activity. Although a single protein was shown to exhibit this activity, TFIID was later found to consist oftheTATA box binding protein (now called TBP) and several associated factors (TAFs). In yeast, TFIID refers to a single protein, TBP (60,90).12maimnalian enhancers. That the functions of both of these regulatory elements are nevertheless connected issuggested by the observation that the Ty2 UAS element is dependent on the presence of the enhancer to allowmaximal levels of transcription. (137). However, the description of the structure and experimental characterizationof these elements is beyond the scope of this discussion. An important conclusion from this body of work is that,although the 5’ 6 of Ty elements provides the basal elements for transcriptional initiation, the 6 promoter requiresadditional activating sequences to direct the high levels of Ty transcription observed in haploid yeast cells.Although in Ty elements these activating and regulatory sequences seem to be located downstream of the initiationsite (perhaps as a constraint imposed by the very nature of the transpositional mechanism of these mobile elements,as proposed in ref. 22), solo 6 sequences may be rendered transcriptionally active by a nearby host (i.e. non-Ty)UAS.1.1.2 Ty and 6 insertional mutationsAs mentioned in the previous section, the first Ty element of yeast was cloned as a restriction fragment atthe sup4 locus (33). The availability of DNA probes specific for Ty elements soon led to the identification ofspontaneous mutations at the HIS4 and CYC7 loci caused by the insertion of Ty sequences (34, 63, 182). Sincethose early studies, much work has been directed towards the understanding of the mutagemc properties oftransposable elements and of their role in promoting spontaneous mutations. Two fundamentally differentmechanisms of mutagenesis can be distinguished. First, because they are repetitive in the yeast genome, Ty and 6sequences provide highly recombinogemc, homologous substrates for chromosomal rearrangements (deletions,inversions, translocations, duplications, gene conversions; ref. 34, 182, 200, 183). Second, through their ability totranspose and integrate at non-homologous sites in the genome, Ty elements and their solo 6 derivatives caninterfere with the expression of the transcriptional unit adjacent to the insertion site (referred to below as theadjacent gene). Both of these pathways can generate phenotypic variations, but only retrotransposition events areassociated with the formation of an insertion at a new chromosomal location.The integration of a Ty element clearly does not involve a sequence-specific step, as evidenced by the verydegenerate consensus integration site derived from more than one hundred independent insertional events (158,240). However, in addition to a preference for A/T rich sequences, it has been proposed that regions of openchromatin (often characterized by nuclease hypersensitivity) constitute preferred integration sites. This conclusion13is based on the observation of a statistically significant bias towards integration in the 5’ region of the mutatedlocus (58, 158, 240). A retrotransposition event consists of a simple insertional mutation, accompanied by theduplication of a 5 bp long sequence of the target gene (present as flanking direct repeats). As for other insertionalevents, it is clear that the transposition of a Ty element within a coding region (open reading frame) is likely todisrupt the function of the mutated gene and produce a null phenotype. The formation of such disruption alleles byTy transposition has been reported for several loci, notably for the counter-selectable markers URA3, LYS2, andCAN] (58, 158, 188, 211, 240).More interesting are the frequent transpositional events targeted to the 5’ noncoding regions of specificgenes. The functional consequences of such insertion are twofold: (I) the 6 or Ty insertion disrupts essentialpromoter elements, which are now displaced by 330 bp or 6 kb of intervening sequences (for 6 and Ty,respectively); (ii) new cis-acting sequences present within the Ty element (transcriptional initiation andtermination signals, enhancers) are brought in close proximity to the adjacent gene. As a consequence, theinsertion of a transposable element within the 5’ noncoding region of a gene often interferes with the normalexpression of that gene. Most of the literature on the subject discusses insertional mutations characterized orselected on the basis of the phenotypic changes they mediate and is biased consequently towards interferinginsertions. To my knowledge, in all cases studied so far, phenotypic changes were found to correlate with specificchanges in the transcription of the adjacent gene. The actual effects of the insertion depend on several parameters,such as the position of the integration site with regard to the promoter elements of the adjacent gene, theorientation of the inserted element, the nature of the insertion (full Ty or solo 6) and its DNA sequence (whichdictates the transcriptional competence of its cis-acting sequences). Two opposite transcriptional effects, activationand inhibition, have been observed. Each of these can also be associated with other phenotypic or transcriptionalchanges such as cold sensitivity (cs), heat sensitivity (hs), deregulation and mating-type control. The followingsections summarize some important mechanistic aspects of each type of insertion, with emphasis on the inhibitoryinsertions, some of which are suppressed by spt2 mutations.Activation of transcription. Several examples of Ty-mediated gene activation or deregulation have beendescribed, extensively studied, and reviewed elsewhere (20, 22). These range from the reversion of a promoterdeletion mutant at HIS3 (202) to cis-dominant overexpressing alleles, notably at CYC7 (63, 62), ADH2 (37, 38,14241), HIS4 (183), DUR1 and DUR2 (135), CARl and CAR2 (56). Mutations belonging to this group represent avery interesting example of the relation between Ty sequences and adjacent gene expression. In most casesstudied, the insertional mutation consists of a full length Ty element oriented such that it is transcribed away fromthe adjacent gene (exceptions are discussed in ref. 20). The insertion often disrupts cis -acting regulatorysequences of the adjacent gene, but not its basal elements (TATA box and I site). Furthermore, in that particularorientation, the downstream regulatory sequences of the c region are proximal to those basal elements. Theconsequences of this new promoter configuration can be summarized as follows. First, the adjacent gene isexpressed either at higher levels, or in a deregulated fashion. Those effects have often constituted the basis of thephenotypic changes leading to the identification of these mutations in the first place. Two well characterizedexamples of those effects are the CYC7-H2 allele, which produces a 20-fold overexpression of iso-2-cytochrome cin haploid cells and allows cyci strains to grow on lactate (63), and the ADR3’ alleles, which constitutivelyexpress the glucose-repressible alcohol dehydrogenase II gene (ADR2 or ADH2) and allow strains deficient inAD}H and niADH to grow on glucose in the absence of respiration (37, 38, 241). Second, the expression of theadjacent gene comes under mating-type control and, like Ty transcription itsell is regulated as a haploid-specificgene. For example, the overexpression of iso-2-cytochrome c in CYC7-H2 strains is 10 times higher in a or chaploids than in a/cs diploids. Because of this unique feature, those insertional mutations have been described bythe acronym ROAM for regulated overproducing alleles responding to mating signals. Third, the overproducedtranscript initiates at the wt initiation site.The characteristic effects of ROAM mutations suggest a simple interference model based on a promoterreplacement mechanism: the basal elements of the adjacent gene promoter become isolated from their normalupstream regulatory sequences (deregulation), and simultaneously adopt new regulatory properties defined by thenearby downstream elements of the Ty s sequences (activation and mating control). The adjacent gene and the Tyelement appear to become co-regulated by the same cis-acting elements, which must therefore be able to act in bothorientations and on two divergent basal promoters. Following this view, a possible cellular role of Ty elements isto act as portable regulatory cassettes, capable of modifying the regulatory properties of any RNAPII promoter.Although this simple picture provides a working model for Ty-mediated gene activation, studies of cis- and transacting mutations modifying the effects of ROAM mutations have shown that the regulation of Ty transcription doesnot always parallel that of the adjacent gene (67, 44), and that the divergent basal promoters may themselves be in15direct competition for general imtiation factors (44). That internal (s) sequences are necessary for the observedphenotypes is supported by the fact that ROAM mutations are never caused by solo 6 insertions.Inhibition of transcription. The counter-selectable markers URA3, LYS2 and CAN], as well as enrichmentmethods for his4 auxotrophs by inositol starvation (34, 97), have all been exploited to select for inactivating Tyinsertional mutations into promoter regions (34, 158, 240, 58). Such Ty or 6 insertional mutations were shown tomediate their effects via inhibition of or interference with, normal transcription of the adjacent gene. In contrastto the stable insertional mutations into coding regions mentioned previously, inactivating Ty insertions intopromoter elements can revert at a high frequency by chromosomal rearrangements or mutations in unlinkedsuppressor genes (see below). Internal Ty sequences are apparently not required for inactivation, as mutations ofthis class caused by solo 6 sequences have also been characterized. This fact is also reinforced by the finding thatthere is no strong bias towards a particular orientation of the inserted Ty in interfering insertions.The mechanism of inhibition depends on the site of insertion and the sequence of the element. Overall,the study of several inactivating mutations and of their reversion by cis- or trans-acting mutations suggests that thetranscriptional interference can be caused by simple disruption of promoter elements, by promoter competition, orby premature termination. Three extensively studied mutations will be introduced to illustrate each of theseinactivating mechanisms.The first two mutations, his4-91 7 and his4-912, were obtained as spontaneous His auxotrophs derivedfrom a wt 1{is+ strain (34). Both of these mutations were shown to be highly unstable promoter mutations causedby the insertion of a transposable element in the 5 noncoding region of the HIS4 gene (34, 73, 182, 181). TheHIS4 gene encodes a multifunctional enzyme, catalyzing three steps in the biosynthesis of the amino acid histidine.Yeast strains carrying a mutation interfering with the production of this protein require exogenous histidine forgrowth (they are histidine auxotrophs) and are said to exhibit a His phenotype. Transcription of the HIS4 gene isregulated by two separate systems, the general amino acid control, and the basal control. The general amino acidcontrol co-regulates the production of several enzymes involved in amino acid biosynthetic pathways by activating(or derepressing) their synthesis in response to amino acid starvation (248). The activation ofHJS4 (as well as ofthe other genes under the general control) occurs at the level of mRNA synthesis (210) and requires thetranscriptional activator, GCN4 (reviewed in ref. 103). The basal control is responsible for the low levels of HIS416transcription under conditions of repression (or in the absence of GCN4) and for activation by phosphate oradenine limitations (10). This control is mediated by the trans-acting proteins, BAS1 and BAS2 (a.k.a. P1102).The 5 noncoding region of the HIS4 gene has been studied extensively and the sequence elementsresponsible for basal level expression (repressed state) and activation (derepression) have been identified. Thebasal promoter elements consist of a unique TATA box, located at position -60, that is, 60 nucleotides upstream ofa major I site (Ijjj, position +1), and 123 bp upstream of the ATG (55). The BAS1 and BAS2 sites are locatedbetween positions -180 and -158. The cis-acting determinants of the general amino acid control consist of fiverepeats of the GCN4 binding site, 5’-TGACTC-3’, located at positions -192 (site A), -180 (site B), -136 (site C), -113 (site D), and -85 (site B). Although these repeated sequences may act synergistically to mediate GCN4activation, site C represents the highest affinity site in vitro (8) and is primarily responsible for GCN4-dependenttranscription ofHJS4 in vivo (50). On the other hand, sites D and E are not sufficient to confer general control(54), and site B (at position -180) is in fact the BAS1 binding site (224).Both BAS1IBAS2- and GCN4-dependent transcription of HJS4 were shown to require the ubiquitousDNA-binding protein, RAP1 (50). RAP1 binds the HIS4 promoter in between the BAS2 site (at position -158) andthe GCN4 site C (with some overlap) and was proposed to assist the activators by preventing nucleosome fonnationover their respective binding sites. A strain lacking BAS 1LBAS2 and GNC4 activity (basi bas2 gcn4) is His,showing that HJS4 expression is dependent on these control systems. This information on the regulation of HIS4transcription (i.e. the nature of the trans-acting factors and cis-acting promoter elements) offers a uniqueopportunity to dissect the molecular events leading to transcriptional inhibition by transposable element insertions.his4-91 7. The his4-91 7 mutation is caused by the insertion of a Ty element of the Ty2 family (Ty2-9 17)in the HIS4 promoter, 9 bp upstream of‘JJIS4 (nucleotides -9 to -5 being terminally repeated; ref. 181). In contrastto most Ty elements, the Ty2-9 17 element does not have identical 6 repeats (181). In his4-91 7 the Ty2-9 17element is oriented such that it is transcribed divergently from the HIS4 transcriptional unit. Although thisconfiguration is identical to that of activating insertional mutations, his4-91 7 strains are His and express HJS4only at extremely low levels (181, 209).As discussed previously (44), an important feature of the his4-91 7 mutation is that the Ty2-917 element isinserted in between the HJS4 1 site and the TATA box. The disruption of promoter elements and the displacementof the TATA box by nearly 6 kb from1H184 are likely to be responsible for the observed transcriptional17interference mediated by Ty2-9 17. The analysis of His+ revertants of the his4-91 7 mutation, obtained asspontaneous or UV-induced Ura gene convertants of the Ty2-917 element marked with the URA3 gene (Ty2-917(URA3)), showed that the actual sequence of the element is the determining factor in this system (183, 185).Mutations affecting the Ty2-9 17 TATA sequence or activating a downstream enhancer sequence were shown torevert the inactivating phenotype to a ROAM-like phenotype (183, 185). Since the expression of the Ty2-917element is not affected by the base pair substitution in the region that activates HIS4 (Ty2467 element, ref. 44),it was concluded that the effects of a particular Ty element on the expression of the adjacent gene are not always areflection of the transcriptional competence of that element. Mechanistically, the c mutation in Ty2-467 probablyactivates a UAS that specifically acts on HIS4 transcription. In support of this model, a sequence-specific DNA-binding factor interacting with the region of Ty2-467, but not Ty2-917, has been detected in vitro (82) and invivo (70). This binding activity does not appear to be under mating-type control. It should also be noted that notall the His+ revertants of his4-91 7 are influenced by the mating-type locus as seen in typical ROAM mutations.Another implication of this model is that sequences in the 6 element of those Ty2-9 17(URA3) derivativesmay act as a fortuitous TATA box for the ff184 transcript initiating at‘ff184 and activated by the mutated UAS.Alternatively, Ty2-467-activated transcription may be TATA-independent as observed for low levels oftranscription of some ff184 promoter mutants (172). Since a point mutation in the normal Ty2-917 TATA boxdecreases transcription from 16 and snnultaneously increases transcription from1H1S4’ it was proposed that thesedivergent basal promoters are in competition for general initiation factors (44).Finally, derivatives of his4-91 7 in which the Ty element has been excised by 6-6 recombination have beenobtained (184). Since Ty2-917 contains non identical 6s that differ by 4 substitutions and a 1 bp deletion, the 6element of the resulting mutation, termed his4-91 7, can have any of several sequences: that of the 5’ or 3 ‘6, or ahybrid of the two. Interestingly, three of the his4-91 75 mutations characterized show different effects on HIS4expression (184). The resulting phenotype may reflect the ability of the solo 9176 to provide a fortuitous TATAbox for HIS4 initiation, or to allow activation by the HIS4 UAS, now displaced by only 333 or 334 bp. The ff184UAS responsible for general amino acid control was shown to act even when displaced by more than 100 bp inCYC1 fusions (104), or in the presence of a 6 insertion (9125 see below). The regulation of these His his4-91 75mutant strains by GCN4 has not yet been investigated.18his4-912. The his4-912 mutation is the first spontaneous insertional mutation shown to be caused by theinsertion of a repetitive Ty element (34, 182). In his4-912, a Ty element of the Tyl family (in fact, Tyl-9 12defines this family) is inserted 161 bp upslream of the HIS4 ATG. In relation to the HIS4 transcription initiationsite (position +1, 63 nucleotides upstream of ATG), this insertion occurs at position -97, nucleotides -97 to -93being terminally repeated (66). The Tyl-912 element is in the same transcriptional orientation as the adjacentHJS4 gene. As for most Ty insertions in that orientation, Tyl-912 interferes with the expression of the adjacentHIS4 gene and causes a recessive His phenotype (34). A his4-912 strain does not produce detectable amounts ofH184 mRNA, and does not present any evidence of a transcript emerging from the 3’ end of the Ty 1-9 12 element(209). With regard to the HJS4 promoter sequences, the Tyl-912 element is inserted upstream of the I site, theTATA box and the GCN4 site, E. Since this site is not sufficient to activate the HJS4 promoter (54), Tyl-912 maymediate transcriptional interference by promoter disruption and displacement of the general amino acid controland basal control UASs.In a study of Ura gene convertants of a Tyl-9 12 element marked with URA3 (Tyl-9 12(URA3)), a singleisolate in which part of Tyl-912 was replaced by another Ty element acquired a j5+ phenotype. The mechanismof this reversion remains unknown. The majority of Ura derivatives obtained in this selection system, as well asmost Ths revertants of his4-912, are generated by excision of the element by &ö recombination (34, 183, 186).Since both s of Tyl-912 are identical (66), the resulting mutation, called his4-912ö (previously known as his4-912R1), possesses a unique, defined sequence and generates a unique phenotype, i.e. cold-sensitive His+phenotype: yeast strains carrying the his4-9126 mutation are phenotypically j5+ at 37°C, but His at 30 and 25°C(34). This temperature dependence, notably the }31j5+ phenotype at high temperature, is somewhat affected by thegenetic background of strains carrying the his4-9125 mutation. The excision of Tyl-9 12 by LTR-LTRrecombination occurs at a frequency of approximately 1 in cells (245). Strains carrying the his4-912 mutationwere also shown to revert to Pjs by numerous chromosomal aberrations involving the Tyl-912 element (34, 182).The1j+ revertants of his4-91 2 and his4-912(URA3) caused by gene conversion, translocation or inversion wereall shown to express low levels of a HIS4 transcript of wt size. These transcripts are not regulated by the generalcontrol (209).19The his4-912ö mutation presents a particularly interesting configuration. The 9 l2, inserted in betweenthe HIS4 TATA box (TATAjjj4)and the UASs, provides a new basal promoter (TATA box and I site, ‘o) as wellas a new translational initiation signal (the TyA ATG, ATG). These cis-acting sequences are all oriented in thesame direction as the HJS4 transcriptional unit (Fig. 2). Northern blot analyses have shown that his4-9125 mutantstrains produce two transcripts homologous to HJS4. The smaller mRNA, present at very low levels, initiates at thewt1HIS4 These strains also produce a more abundant and longer transcript (209). Nuclease Si mapping andprimer extension experiments confirmed that this longer transcript initiates at the ö I site, located 94 nucleotidesinto the ö sequence (from the 3’ end). This nucleotide is identical to the initiation site of the major Ty mRNAproduced in haploid yeast cells (i08, 209).An obvious question arising from these observations is “What is the relationship between the observedtranscriptional pattern of the his4-912ö mutation and the phenotype of mutant strains carrying this mutation?”.The experiments described above, together with studies of cis- and trans-acting suppressor mutations of his4-9125have led to the conclusion that the His phenotype strictly correlates with the levels of the shorter mRNA initiatingat the wt‘HIS4• The 9i2 promotes the initiation of a nonfunctional mENA; the lack of function is probably dueto the presence of several translational start and stop codons in between the start codon of the TYA open readingframe (present within the ö sequence) and that of the HIS4 gene (40, 55). It is well documented that translationnormally initiates only at the 5’ proximal ATG in eukaryotes. For example, translation of the HJS4 mRNAinitiates at a downstream ATG only when the upstream ATG (5’ proximal) is mutated (36). Therefore, the longer&imtiated mRNA is thought to be translationally inactive.Another important aspect of his4-9]2Sis that both transcripts are regulated by the general amino acidcontrol (209). Mechanistically, this implies that the activator protein GCN4, acting via its recognition sites A to D(located 5’ of the 912ö insertion site), is able to activate the two downstream basal promoters simultaneously. Thisobservation and the proposed mechanism are also consistent with the inability of S sequences to activatetranscription on their own. The basal S promoter, normally silent when isolated from the downstream g enhancerin solo S insertions, therefore becomes regulated by the upstream HIS4 UASs in his4-912ö and adopts theregulatory properties of the H184 gene. Although this provides a basis for understanding the origin of the 5-initiated mRNA as well as the coregulation of both his4-912ömP.NAs, it does not explain the low levels of wtHJS4 niRNA (responsible for the His phenotype at 23°C), nor the temperature dependence of the His phenotype.20The molecuiar basis for the temperature dependence of his4-9128is still unknown, but at the transcriptional level,a temperature shift from 23 to 37°C (corresponding to a phenotypic change from His to Hisj is associated withan important increase in the levels of wt HIS4 mRNA and a twofold decrease in the levels of the longer S mRNA(108).HisHishis4 -9126>-I.I-Figure 2. The his4-912ö mutation.The structure of the insertional mutation his4-9128 has been expanded (above) to highlight the relative positions ofsignificant cis-acting sequences. Two transcripts are produced by the mutated HJS4 locus (below). Thepredominant niRNA (thick line) initiates at the 9125 promoter and is not functional for HIS4 activity probably dueto premature translational termination. cis- and trans-acting suppressor mutations of his4-9126 cause an increasein the shorter (thin line), functional wt HIS4 transcript.UASrTHIS421At least two different mechanisms, not necessarily exclusive, might account for the transcriptionalinterference imposed by the 9126 on the initiation of the wt HIS4 mRNA. The first possibility is that thetranscriptional activators acting through the UASs of HIS4 (GCN4, BAS1 and BAS2, along with RAP1) - whichare ultimately responsible for almost any transcriptional activity detected at the HIS4 locus (50)- preferentially actat the nearest downstream TATA box, the 6 TATA box in his4-912ö This model is supported by the observationthat the insertion of a synthetic wt HIS4 TATA box at position -98 of the HIS4 promoter results in the synthesis ofa new, longer mRNA, and in the inhibition of the wt HIS4 mRNA, normally promoted by the now downstreamHIS4 TATA box (at position -60; ref. 149). In other words, if the 111S4 TATA box is duplicated, only the upstreamTATA will be utilized and activated in response to amino acid starvation. The second possibility is that theobserved transcriptional pattern of his4-9125 reflects intrinsic differences in the 9128 and HJS4 TATA boxes.According to this model, both TATA elements are in competition for general transcription factors (includingTFIID) and the stronger basal promoter - the 9126 promoter - directs most of the RNAPII initiation. This model issupported by cis-acting (intragemc) suppressor mutations of the His phenotype of his4-9125at 23°C (108). Inthese experiments, a plasmid-borne his4-9125 allele was mutagenized and 6 independent mutations reverting theHis phenotype and increasing the levels of wt HIS4 mRNA were characterized. Three revertants carry the sameA-to-G transition at position 166 of the 9128, next to a TFIID consensus site. The other three mutations areclustered near the HIS4 TATA box at position -60 and were proposed to generate new functional TATA boxes.These mutations are thought to weaken the 6 promoter or strengthen the HIS4 promoter, respectively. A promotercompetition model was proposed to explain the effects of his4-9125 and its reversion by those cis-acting mutations.It is interesting to note that, as predicted by the first model presented above, the wt HIS4 mRNA produced in theserevertants does not appear to be regulated by the general control (108). Both mechanisms are thus likely toparticipate in the observed transcriptional interference.lys2-128. The lys2-128 mutation was isolated from a genetic screen selecting for spontaneous caminoadipate resistant colonies (211). The amino acid analog c-aminoadipate is a suicide substrate for wt yeastcells. Resistant mutants often carry mutations at the LYS2 locus, coding for the enzyme, 2-aminoadipate reductase,required for de novo lysine synthesis. Like HIS4, the LYS2 gene is regulated by the general amino acid control.The lys2-128 mutation was shown by Southern blot analysis to be caused by the insertion of a transposable elementof the Tyl family in the 5’ region of the LYS2 locus. The Tyl-128 element is in the same transcriptional22orientation as the adjacent LYS2 gene (211). Strains carrying the lys2-128 insertional mutation are auxotrophic forlysine, but in contrast to his4-912 strains, they revert to Lys+ only rarely. It was later found that the lys2-128mutation is not a promoter insertion, but is in fact an insertion into the 5’ portion of the LYS2 coding region (at+153 [43], or +158 [681). The solo derivative of Iys2-128, called lys2-128c5 also interferes with the expressionof the LYS2 gene: lys2-1285 strains are Lys, and do not produce the expected 4.2 kb long LYS2 mP.NA. Instead,a small transcript of ca 580 bp homologous only to the 5’ end of the normal LYS2 mRNA is detected in Northernblot analysis (42, 43, 221). This transcript initiates at the normal LYS2 I site and appears to terminate in the solo128 element. The conclusion from those experiments is that the lys2-128S mutation interferes with theexpression of the LYS2 gene by causing premature transcriptional termination. This mutation therefore providesevidence that ö sequences contain the cis-acting determinants sufficient to direct efficient termination of RNAPIItranscription and that solo sequences can act as portable termination signals. That 128S does not simply act bydisruption of the LYS2 coding region is shown by the transcriptional pattern of Iys2-128ö in strains carryingunlinked suppressors (see next section).So far, the discussion has focused on the structure and mutagenic potential of Ty elements. Through theirown transcriptional signals, some of them known to utilize components of the host transcriptional machinery, theseelements were shown to interact in complex ways with adjacent transcriptional units. In asking how a yeast cellcan modify, or even suppress the transcriptional effects caused by Ty insertions, molecular biologists learned thatTy insertional mutations in unique sequences never revert to the wt configuration in haploids. Since the Ty orcannot be fully excised, the cell must resort to other approaches toward this goal. One way to modulate theactivation or inhibition of Ty insertions is via cis-acting mutations in critical sequences. Several examples of suchmutations have been described above. Some of them act as true dr-acting suppressor mutations of the Ty orinsertion. Another way that the cell can revert the effects of Ty insertions is by mutations in unlinked suppressorloci. Many extragemc suppressor mutations of insertions such as his4-91 7, his4-9125 and lys2-128 have beencharacterized and define a group of yeast genes known as SPT genes (for SuPpressor of Ty). Since they cansuppress the transcriptional defects caused by Ty and/or insertions, these mutations are thought to define yeastgenes whose products are involved directly or indirectly in transcription initiation. The primary interest behind the23study of SPT genes therefore lies in the belief that SPT proteins belong to a group of host factors interacting withTy elements and/or the transcriptional machinery of yeast.1.1.3 Unlinked suppressors of inactivating Ty and 6 insertional mutationsIt has been noted previously that the hallmark of several Ty-mediated mutations is their extremely highfrequency of reversion. It was discussed in the previous section how this instability is mainly attributable to therecombinogemc potential of these repeated elements, associated with the heterogeneous transcriptional effects ofdifferent Ty sequences. The outcome of these recombination events, or of other mutagemc events occurring at theaffected locus, is a specific change in the DNA sequences of the insertional allele. Together, these molecularevents can be classified as cis-acting modifiers of Ty insertions. The study of His+ revertants of the his4-912 andhis4-91 7 mutations showed that in some instances the mutations responsible for the phenotypic reversion did notmap to the HIS4 locus. Those studies provided the first evidence that the Ty-mediated transcriptional inbibitioncould also be suppressed by uniiithed or trans-acting modifiers. Since they were thought to promote not only thesuppression of Ty mutations, but also the reversion of his4-912 by 6-6 recombination, the genes defined by thosemutations were originally named SPM by analogy to the Spm suppressor-mutator elements found in maize (181).It was later shown that the apparent “mutator” property of these mutations was the result of a bias in the geneticselection utilized: for example, spt3 mutations can suppress his4-9126 but not his4-912, such that the selection forHis+ revertants of spt3 his4-912 mutant strains (or SPT3 his4-912) will introduce an enrichment for 6-6recombination events (245). In order to represent more accurately the effects of these unlinked suppressormutations, the designation SPM was changed to SPT, for SuPpressor of Ty’s (245).The notion that spt mutations affect cellular factors involved in some general aspect of Ty transcription oreven of RNAPII transcription is supported by the fact that they rarely suppress non-Ty insertions, that they oftensuppress more than one rnsertional mutation at HIS4 and can also suppress insertions at the other genetic loci suchas LYS2 (211). The ability of a particular spt mutation to suppress particular Ty or solo 6 insertions, divergent ornot, defines the suppression pattern of that mutation (allele specificity), and is thought to reflect the cellularfunction of its wt gene product. It should be emphasized that although the wt SPT genes are known as suppressorsof Ty’s, it is the mutant fonns of those genes (or in some cases overexpression or deregulation) that actually lead tothe suppression of insertional mutations. This important yet confusing notion is perhaps best represented in terms24of the Spt phenotype of a strain: a Spt+ strain does not contain any spt mutation (but only wt SPT genes) andcannot suppress Ty insertions; conversely, a Spr strain carries a mutation in one or several SPT genes and cansuppress the effect of one or several Ty insertions. Of course this suppressor or Spt phenotype will vary inverselywith the phenotype associated with the insertion allele itself: for a His SPT his4-91 7 strain, a suppressing sptmutation will generate a Spt phenotype and consequently a Jfi+ phenotype (since it suppresses the inhibition ofHIS4 expression by Ty2-917).Three different categories of suppressor mutations will be discussed below: the spt mutations, originallycharacterized in genetic screens for suppressors of Ty insertions, mutations in histone genes and their regulators,and P.NA polymerase II mutations. Other unrelated mutations were also found to confer a suppressor Sprphenotype. An example of such mutations will be discussed in section 1.2.3.So far, 17 different SPT genes have been described. They have all been identified through genetic screensfor suppressors of one or several Ty insertions (42, 68, 157, 181, 245, 246). As Ty elements utilize differentmechanisms to inhibit the expression of the adjacent gene, spt mutations will generate different suppressionpatterns, dictated by the role of the wt SPT protein in transcription and Ty-mediated inhibition. Table 1summarizes the results from several studies and presents the suppression patterns of those 17 spt mutations. In allthe cases studied so far, the suppression was shown to occur at the level of transcription, although the actualtranscriptional changes underlying the phenotypic suppression can be quite different depending on the function ofthe SPT gene product mutated in the Spr suppressor strain. As expected if only a small number of mechanismscan suppress the Ty or 6 insertions, most spt mutations belong to one of three classes, on the basis of theirsuppression pattern (68, 96). Important genetic equivalencies and functional information on class I and class IIgenes are discussed below and summarized in Table 2. Since they do not share the 6 suppressor phenotype of spt2mutations and since little is known about their cellular function, the class III genes SPTJ3 and SPT14 will not bediscussed further.251 + +1-2 + -3 + -4 + -5 + -6 + -7 + -8 + -9 n.a. n.a.10 + n.a.11 + n.a.12 + n.a.13 — -14 - -15 + -16 + -21 + n.a.a Partially impenetrant (208)b Not availablea+1- ++1- ++ ++1- ++1- ++1- ++ ++ +1-+ ++ ++ ++ ++++ +- ++ +188, 208, 245188,211, 245188, 211,245,24668,24568,24568,24568,245,24668,2466868, 15742, 6842, 6868686820157Table 1. Suppression pattern of spt mutations.his4 lys2spt REF.912 912 91Th 917 128+n.a.+1-++n.a.n.a.n.a.+26Table 2. Properties and pleiotropic nature of class I and class II spt mutations.Class ISPT3,7,8 TBP-associated factors; coactivatorsSPTJS Codes for the TATA box binding factor of TFIID (TBP)Class IISPT2 SIN] : Sin and Srb phenotypes; HIVIG 1-likeSPT4,5,6 SPT6 =SSN2O =CRE2; Sin- phenotypeSPT1 1,12 HTAJ -HTB1 encoding histone H2A-H2B ; Sin phenotypeSPTJ ,]O,21 SPT1 =HIR2 ; regulators of histone gene expression1.1.3.1 Class I SPT genesMutations in SPT genes of the first class are suppressors of ö insertions and strong suppressors ofdivergent Ty insertions such as his4-91 7. Furthermore they are the only known suppressors of the solo insertionmutation, his4-9] 75(60, 246). The available evidence suggests that the SPT3, SPT7, SPT8, and SPTJ5 geneproducts either perform similar functions or act at the same step in transcriptional initiation.SPT3, SPT7, SPTS. The first developments in our understanding of the role played by the SPT3, SPT7,and SPT8 proteins came from studies on the spt3 mutations. Mutations in a gene previously called SPM3 werefirst isolated as suppressors of his4-912 and his4-917 (181), and three recessive spt3 mutations were later isolatedas suppressors of his4-917 (245). Both a deletion and a frameshift spt3 mutation (spt3-20] and spt3-]O],respectively) were shown to confer a recessive suppressor phenotype, suggesting that the SPT3 protein is notessential for mitotic growth and that suppression is caused by loss of SPT3 function. In addition to suppressingseveral Ty and ö insertional mutations at HIS4 and LYS2 (211), spt3 mutant strains also show a slow growth rate aswell as sporulation and mating defects: for example, spt3 strains can mate with SPT+ partners of the opposite27mating type, but fail to mate with other spt3 mutant strains (247). Several of these phenotypes were shown to becaused by changes at the level of transcription.Northern blot analysis of the expression of HIS4 in a his4-912ö spt3-1 strain revealed that the spt3mutation abolishes the production of the longer transcript initiating at I while simultaneously causing an increasein the wt HJS4 mRNA (209). This pattern of suppression of his4-9]2öis analogous to the one observed in Hisrevertants carrying a cis- acting mutation near the 912 TATA box (108). That the SPT3 protein may function asa Ty transcription initiation factor is also supported by the general effect of spt3 mutations on the expression of allthe Ty elements in a cell: spt3 haploids do not produce the major 5.7-kb Ty mRNA. Instead, a nonfunctional TyRNA, approximately 5.0 kb in length, initiating in the e region is produced (247). As a consequence, mutations inspt3 abolish the transposition of chromosomal Ty elements, thereby providing a very useful tool for the study ofartificial Ty elements expressed from the inducible GAL] promoter (which is SPT3-independent; ref. 23). Theseresults show that the SPT3 gene product is necessary for the synthesis of RNA molecules initiating in the Ssequence.Although spt3 mutations may have identified a host factor specific for Ty transcription, the additionalphenotypes of spt3 strains suggest that the SPT3 gene product is involved in a more general cellular function. Thishypothesis is supported by the observation that the mating defects of spt3 strains correlate with reductions in theexpression of the mating pheromone genes, MFaJ, MFa1 and MFa2 (107). The relationship between these genesand Ty elements, which could explain their common requirement for the SPT3 factor, is not clear. Interestingly,these 3 genes are, like Ty elements, haploid-specific and their promoters are all repressed by a]Icw2 in diploids(107).Following these studies, a new genetic screen was utilized in order to characterize additional yeast genesrequired for S-initiated transcription. This work showed that the products of the SPT7 and SPT8 genes arefunctionally similar to SPT3. Mutations in SPT7 and SPT8 produce phenotypic effects similar to those seen in spt3mutants and they do not affect SPT3 expression (246). Together, these results suggest that the SPT3, SPT7 andSPT8 proteins perform a common function in the transcriptional initiation process and may even interact in vivo.SPTJ5. Additional clues as to what that function may be came from the analysis of another member of this classof suppressors, the SPT]5 gene. Mutations in SFTI 5 first were obtained in the screen for spt3-like suppressor28mutations mentioned above (246). Haploid sptl5 mutant strains were also found to display several phenotypesassociated with spt3, spt7 and spt8 mutations. At the transcriptional level, the effects of sptl5 mutations on 6insertions and Ty elements are also similar to those caused by spt3, 78 mutations: they eliminate the fonnation ofthe Ty P.NA imtiating in the 6 sequence and reduce the cellular content of fi.ill-length Tyl RNA (9, 59, 60).Molecular characterization of the SPTI5 gene led to the important finding that SPTJ5 is essential for growth andencodes the yeast TATA binding protein, TF1ID (now called TBP; ref. 60, 90, 91, 111).Initially characterized as a specific nuclear fraction required for in vitro, promoter-dependent initiation byRNAPII, TFID is now known to consist of a TATA-binding protein (TBP) and several associated factors (TAFs;ref. 178, 204). Through its TBP component, TFIID can bind to the TATA box element of eukaryotic promotersand direct the formation of the preirntiation complex required for basal transcription. Activated transcription onthe other hand is thought to require specific coactivators (TAFs) that link TBP and the basal transcriptionalmachinery to gene-specific activator proteins (83, 178). The finding that particular point mutations in the yeastTBP (i.e. sptl5 mutations) can affect transcriptional imtiation in vivo (for example at the 6 promoter) provided thefirst genetic evidence supporting the profusion of in vitro work on TFIID.The actual mechanism whereby a sptls mutation can suppress his4-9128is unclear. By analogy to thecis-acting suppressor mutations of his4-9128 described previously, an obvious possibility was that the mutant TBPhas acquired a new DNA binding specificity, with increased affinity for the wt HJS4 TATA box. However, in vitrobinding studies do not support this simple model (9). In fact, these experiments have also shown that the wt TBPbinds to the 9126 and the HIS4 TATA boxes with similar affinities, even though in SPT15 his4-9126 strains, themost abundant transcript is the longer nonfunctional mRNA initiating from the 6 promoter.In order to reconcile the genetic and biochemical data, it could be argued that the strength of a promoter isnot determined by the affinity of its TATA box for TBP, but perhaps by another variable dependent on thecoactivators associated with the TBP bound at that promoter. This possibility was investigated by a genetic screenfor extragenic suppressors of a specific sptl5 point mutation. All suppressor mutations isolated mapped to theSPT3 locus (59). Coimmunoprecipitation experiments confirmed that the SPT3 protein interacts with TBP in vivo.These results suggest that the SPT3, SPT7 and SPT8 proteins act as transcriptional coactivators and represent theyeast counterpart of some of the proteins that copuri1’ with TBP to form TFIID (i.e. TAFs). With regard to thesuppression of the his4-9128 insertion, the SPT3,7,8 proteins may specifically recognize the TBP bound at the S29TATA box and participate in the GCN4-activated synthesis of the long non-functional mRNA. The production ofthis RNA, as well as that of all Ty elements and other genes utilizing the same coactivators, would be expected tobe eliminated in strains deficient in any of these coactivators (Spr phenotype).In conclusion, class I SPT genes code for an essential and well defined component of the basaltranscriptional machinery, the TATA-binding protein (TBP), and possibly a subset of coactivators (TAFs) thatinteract with TBP at certain regulated promoters. Further biochemical characterization of the role played by thesefactors in basal and activated transcription will be required to establish whether they perform a functionhomologous to the TFIID complexes of higher eukaryotes. Class II SPT genesThe class II spt mutations share several characteristics that distinguish them from the class I mutations.Mutations in class II SPT genes do not eliminate the -imtiated transcript. As a consequence, Ty transcription andtransposition are not impaired in these mutant strains. This important mechanistic difference is reflected by theinability of most class II mutations to suppress full Ty insertional mutations such as his4-91 7. In addition, thefunction of several class II genes is dosage-sensitive and is required for complete repression of SWI-dependentgenes in swi strains (see section 1.2). Since they include the hi stone H2A and H2B genes, members of this classare often collectively referred to as the “histone group” of SPT genes and are thought to encode nuclear proteinsinvolved in the maintenance of an inactive chromatin state (244). Whereas class I SPT genes code for componentsof the basal transcriptional machinery, class II genes seem to play a more general role in the transcriptionalprocess, perhaps through their direct or indirect role in the folding of the chromatin template.SPT2. The SPT2 gene, the focus of the present study, was the first class II SPT gene isolated. Formerly calledspm2 mutations, defects in this gene were isolated on the basis of their ability to suppress the His phenotype of ahis4-912 mutant strain (73, 181). It was observed that, in those His+ revertants, the Tyl-912 element hadundergone excision by S- recombination and that in fact, spt2 mutations did not suppress the full lengthinsertion, but rather the resulting solo insertion his4-9128 (73). Some spt2 alleles were also reported to suppressthe his4-91 7 mutation (73, 245), although in another study, all spt2 his4-91 7 mutants were categorized as being30His. It is not known at this time if this represents an allele- or a strain-specific phenomenon. That the role of theSPT2 protein is not merely related to the regulation of the HIS4 gene was suggested by the observation that spt2mutations can also suppress Ty or ö insertions at the LYS2 locus (211).These early genetic studies on SPT2 reveal some important features of its function. Interestingly, spt2mutations represent the major class of spt mutations isolated (85% in one screen; ref. 245). Although the reasonfor this bias is not known, it suggests that the SPT2 locus is particularly sensitive to mutations, or that spt2 mutantstrains had some growth advantage in the selection scheme utilized. It was also observed that mutations at SPT2can be recessive or dominant, whereas all other spt mutations are recessive2. Nearly 50% of the spt2 mutationsisolated show some degree of dominance for the Spr phenotype (245). A particular allele, spt2-]50, was found tobear a deletion of the entire SPT2 locus plus some sequences from the adjacent RAD4 gene on the right arm ofchromosome V (245). The spt2-150 deletion mutation is viable and confers a recessive suppressor Spt phenotype.This genetic information indicates that the SPT2 protein is not essential for mitotic growth and that, in this case,the suppressor phenotype corresponds to the loss of SPT2 function.As suggested by their suppressor phenotype, mutations in SPT2 cause transcriptional changes at the his4-9125 locus. Northern blot analysis showed that, in a his4-9125SPT2-1 strain, the levels of both the ö-initiatedtranscript and the wt HIS4 mRNA are increased (209). Both of these transcripts are still under general amino acidcontrol and GCN4 activation upon amino acid starvation. As mentioned above, this suppression pattern is in sharpcontrast with the effects of spt3 mutations, which abolish the ö transcript (209). From a mechanistic perspective,mutations in SPT2 appear to promote a general “derepression” of the his4-9125 locus. According to this view, thewt SPT2 protein acts as a repressor or negative regulator of transcription. Such a model is consistent with otherphenotypes associated with spt2 mutations, most notably their ability to bypass the requirement for generalactivator proteins (see section 1.2).The availability of dominant mutations greatly facilitated the molecular cloning of the SPT2 gene. Thedominant SPT2-1 allele was obtained from a clone that can revert the His phenotype of a his4-912SSPT2 strain2 The frequent misuse ofthe terms recessive and dominant warrants a briefnote to c1arit,’ the usage chosen here. The qualifiers recessive anddoninant should be used to modifj a specific phenotypic trait and not a particular allele or mutation. For example a pleiotropic mutation couldproduce the phenotype A, dominant over a, and simultaneously the phenotype b, recessive to B. However, since the discussion presented here focuseson a specific phenotype, i.e. suppressor of Ty or 6 insertions, this distinction will often be omitted for simplicity. Unless otherwise mentioned, adominant SPT mutation codes for a dominant SPT protein, conferring a dominant Spf phenotype.31and used as a probe to clone the wt SPT2 gene from a genomic library (180). The determination of the DNAsequence of the fragment cloned revealed that the wt SPT2± locus contains a 999 bp long open reading frameinterrupted at codon 213 by an ochre nonsense codon in the SPT2-1 allele. The interpretation of these results wasthat SPT2 encodes a 333-amino-acid protein and that production of a truncated polypeptide by prematuretermination (the 212-amino-acid SPT2-1 mutant protein) generates a non-functional protein, unable to maintainthe His phenotype of a his4-9128 mutant strain. Analysis of the amino acid sequence of SPT2 showed that itencodes a novel, highly charged protein (Fig. 3). Although basic (with a theoretical p1 of 108), SPT2 contains twocarboxy-terminal acidic regions with a net charge of -15 each, the second of which is part of a longer poiar regionof predicted helical structure. These features are all absent from the predicted SPT2-l truncated protein. Twosequences of 22 amino acids were proposed to show similarity with the helix-turn-helix DNA binding domain(180). It was later suggested that SPT2 may be distantly related to the maimnalian HIVIG1 non-histonechromosomal protein (see section 1.2.3; ref. 127). More properties of SPT2 will be discussed in the section on theSin phenotype of spt2 mutations (1.2). Since mutations in these genes promote transcriptional changes similar tothose caused by spt2 mutations, the other class II SPT genes provide valuable information concerning the possiblecellular function of the SPT2 protein.SPT4, SPT5, SPT6. Mutations in the SPT4, SPT5, and SPT6 genes share most of the phenotypes associated withspt2 mutations. However, they appear to belong to a separate group since several lines of genetic and biochemicalevidence suggest that the products of these three genes are functionally related and may act as a complex in vivo.Mutations in SPT4, 5,6 were obtained as suppressors of the his4-912ö and Iyc2-128ö mutations (68, 245). Likesome spt2 mutations, they appear to be only partial suppressors of the full Ty insertion his4-91 7. The phenotypicchanges associated with spt5 and spt6 mutations correlate with specific effects at the transcriptional level (43, 220).For example, in a his4-9125 strain, they lead to an increase in the wt HIS4 transcript, correlating with the Hisphenotype, and have variable effects on the &initiated transcript (but do not eliminate it).32AMSFLSKLSQI RKSTTASKAQ VQDPLPKKND EEYSLLPKNY IRDEDPAVKR1 50LKELRRQELL KNCALAKKSG VKRKRGTSSC SEKKKIERND DDEGGLGIRF100KRSIGASHAP LKPVVRKKPE PIKKMSF’EEL MKQAENNEKQ PPKVKSSEPV150TKERPHFNKP GFKSSKRPQK KASPGATLRG VSSCONSIKS SDSPKPVKLN200LPTNGFAQPN RRLKEKLESR KQKSRYQDDY DEEDNDMDDF IEDDEDEGYH* 250SKSKHSNGPG YDRDEIWANF NRCKKRSEYD YDELEDDDME ANEMEILEEE300EMARKNARLE DKREEAWLKK HEEEKRRRKK LUR333BPSi PS2r’til iII I I I ‘ I0 100 200 300 (aa)Helix-Turn-Helix motif (60 to 81 and 250 to 271)Acidic (226 to 249 and 277 to 303)[Fi]] Polar, helical (304 to 333)Figure 3. The SPT2 protein.Primary sequence (A) and schematic representation (B) of the 333-amino-acid (aa) product of the SPT2 gene. Theasterisk indicates the position of the predicted carboxy-terminal residue (Arg-212) in the product of the dominantnonsense mutation SPT2-1. In (B), regions of the protein with significant primary sequence or predicted secondarystructures are highlighted. Sequences from which the synthetic peptides, PSi and PS2, were designed areindicated as solid bars.33Several spt6 alleles were also shown to cause temperature-sensitive lethality; these observations have beenreinforced by the finding that the integration of a null allele of SPT6 constructed in vitro is lethal (43, 162, 220).That SPT5 is also essential for mitotic growth was similarly demonstrated by the inviability of spores canying asptY disruption allele (221). Some implications of the lethality of sptY and spt6 null alleles are that the sptS andspt6 suppressor alleles isolated are probably only partial loss of function mutations and that alleles such as his4-9128 are more dependent on or sensitive to SPT5 and SPT6 functions than are the essential gene(s) also regulatedby these proteins. A similar argument could be made with regard to the sptl5 mutations (affecting TBP; seesection and the RNA polymerase II mutations with Spr phenotype (section genetic results that link the fimctions of these three proteins include the lethality of certain spt4spt5, spt4 spt6, and spt5 spt6 double mutants, and the inability of some spt4, spt5 and spt6 alleles to complementeach other (221a, 245). Since the expression of each of these genes is not affected by mutations in the other two,the possibility that their products interact to perform a common function was assessed by immunoprecipitationexperiments. These studies showed that the SPT5 and SPT6 proteins coimmunoprecipitate and may therefore forma complex in vivo (22 la). Gene dosage studies revealed that an increase or a decrease in the copy number of the wtSPTS or SPT6 gene causes an Spt phenotype (43, 221). These observations suggest that variations in the properstoichiometry of the SPT5 and SPT6 proteins perturb their function. As will be discussed below, the histonesgenes display an analogous dosage-sensitive behavior.What is the function of the putative SPT4,5,6 complex in transcriptional initiation and what are theimplications of this function with regard to the possible role of class II SPT genes like SPT2? Although theavailable information does not provide a definitive answer to this question, independent work, which also led to theisolation of spt6, mutations revealed an important link between these SPT proteins and chromatin components.Indeed, the molecular characterization of SPT6 demonstrated that it is allelic to the previously identified SSN2Oand CRE2 loci (48, 161). Mutations at SSN2O were obtained as suppressors of the transcriptional defects at SUC2(the structural gene for invertase) caused by primary mutations in the activator genes, SNF2 and SNF5 (160, 162).The CRE2 gene was identified by mutations that allow the expression of the ADH2 gene under conditions ofrepression by glucose (48). Together with other results discussed in section 1.2, these genetic equivalencies uniteseveral genetically defined regulatory systems in a model identifying three functional groups of proteins: (1) gene-34specific activator proteins, such as ADR1, GAL4 and SWI5; (ii) general coactivators, including SWI1/ADR6,SWI2ISNF2, SWI3, SNF5 and SNF6; and (iii) general repressors of transcription. According to this model, onepathway of activation by gene-specific activators requires a coactivator complex (SWI-SNF complex) and mediatesits transcriptional stimulation by antagonizing general repressors of transcription (244). The group of repressorproteins include SPT2/SIN1, SPT6/SSN2O/CRE2, SIN4, as well as the core histones H2A, H2B, H3 and H4. Thisfunctional relationship between SPT2, SPT6 and the core histones is also supported by a distinct line of geneticevidence showing that mutations affecting histone genes can generate a class II Spr phenotype.Histone genes and regulators of their expression. In an attempt to characterize other SPT genes that, like SPT5and SPT6, exhibit a dosage-sensitive phenotype, a strain carrying the mutations his4-9125 and lys2-1285 wastransformed with a high-copy-number yeast library and }s Lys+ transformants were selected (42). This selectionsystem led to the finding that the previously identified SPTII and SPT12 genes can also suppress insertions whenoverexpressed (68). The molecular characterization of these loci revealed an important genetic equivalence: SFTi]and SPTJ2 were found to be allelic to the HTA 1-HTB1 locus, which encodes the nucleosome core histones, H2Aand H2B (42, 98). The genome of Saccharomyces cerevisiae contains four pairs of divergently transcribed histonegenes: HTA 1-HTBJ and HTA2-fITB2 each code for histones H2A and H2B, whereas HHTI-HHF1 and HHT2-HHF2 each encode histones H3 and H4 (reviewed in ref. 170). With regard to the Spt phenotype, it was foundthat overexpression on a high-copy-number vector of any of these four loci can suppress insertions. Similarly, areduction of histone gene dosage by deletion of the HTA 1-HTB1 locus confers a Spt phenotype. At thetranscriptional level, those changes in histone gene dosage cause an increase in the wt HJS4 transcript and a smalldecrease in the ö transcript at the his4-9125 locus.Talen together, these results suggest (but do not demonstrate) that an imbalance in histone proteinscauses changes in chromatin and that these changes affect transcription. Since the nucleosomal histones areknown to be complexed as a histone H3-H4 tetramer plus two histone H2A-H2B dimers (230), the experimentsdescribed above suggest that it is the imbalance of the structural nucleosomal units (H3-H4 tetramer, or H2A-H2Bdimer) that leads to the observed transcriptional changes. This prediction is supported by two additional results:(I) whereas overexpression of the whole HTA 1-HTB1 locus - i.e. overproduction of H2A-H2B dimers - causessuppression of his4-9126 overexpression of individual histone genes (HTA 1 or HTBI) does not; (ii) simultaneous35overexpression of all four histone genes (HTA 1-IITBJ plus HHT1-HHF1) does not confer a Spt suppressorphenotype.Another prediction that arises from these results is that mutations affecting the expression of the histonegenes could generate a similar histone imbalance and suppress insertional mutations. Several recent results haveconfirmed this hypothesis. Mutations in the HIR], HIR2, and HIR3 genes, identified by histone regulatoiymutations (171), were shown to exhibit a suppressor Spt phenotype. In fact, HIR2 was found to be allelic to SPT1,one of the first SPT genes identified (73, 208). Similarly, mutations in SPTIO (68) and SPT2J (157) were shown tocause a Hir phenotype, characterized by a decrease in the expression of the HTA2-HTB2 locus (208). That thesespt mutations may mediate the suppression of insertions indirectly, via their effects on histone gene expression,was shown by the ability of a low-copy-number HTAJ-HTB1 construct to compensate for their transcriptionaleffects at the other H2A-H2B locus and to eliminate their Spt phenotype. All the SPT genes were similarly testedfor their effects on histone gene expression, and mutations in other class II genes, such as SPT2, SPT4, SPT5, andSPT6, did not exhibit a Hir phenotype. In a different study, mutations in HCP2, encoding a regulatory proteininvolved in the cell-cycle regulation of HTA 1-HTBJ, were also shown to suppress ö insertional mutations (251).Here again, spt2/sinl mutations were specffically tested and failed to exhibit a Hcp phenotype, corresponding toconstitutive derepression of J-ITA1-HTBJ.In conclusion, class II SPT genes comprise known chromatin components (the core histones) as well as agroup of new nuclear proteins involved in transcriptional repression. As with the class I genes, none of these classII genes defines a Ty-specific factor. On the other hand, in contrast to class I genes, they do not encodecomponents of the transcriptional machinery, but seem to code for general factors regulating the accessibility of theDNA template to the transcriptional apparatus. RNA polymerase II mutationsThe studies presented in the previous sections have shown that Ty insertional mutations have provided avery powerful genetic tool to characterize cellular factors involved in or affecting transcriptional initiation. One ofthe conclusions derived from this work is that it is possible to obtain viable mutations in components of the basictranscriptional machinery with altered function. That mutations in subunits of RNA polymerase II could also36affect transcriptional initiation or start-site selection, and suppress Ty or 6 insertional mutations, has beendemonstrated by a specific genetic screen (96). In this study, the cloned RPO21/RPB1 and RP022/RPB2 genes,coding for the two largest subunits of RNAPII (254), were chemically mutagenized in vitro and the mutant librarywas used to transform a yeast strain carrying two 6 insertional mutations, his4-9125 and lys2-]28c5 The sptmutations in RPO2J and RP022 were obtained by screening for His+ Lys+ transfonnants. Several pointmutations, clustered in conserved regions of the subunits, were shown to generate a suppressor Spr phenotype. Athis4-912ö they cause an increase in the wt HJS4 mRNA, but no significant changes in the 6-initiated transcript.Even though the actual mechanism of suppression by the mutant RNAPII is unknown, these results areinteresting in several respects. First, they clearly demonstrate the role of RNAPII in the selection of transcriptionalstart sites. Perhaps the mutant subunits affect the pattern of transcription of his4-912ôbecause of changes in thepreference for initiation sites or for preinitiation complexes. Second, these results validate the use of mutationssuch as his4-912Sin genetic screens for components of the transcriptional machinery. Third, they provideadditional examples of specific mutations in RNAPII subunits which, since they are viable, affect the expression ofonly a subset of genes. In this regard, it is interesting to note that another class of RNAPII mutations with alteredfunction was obtained in a genetic screen selecting for His+ revertants of a strains deficient in the activatorproteins GCN4, BAS1 and BAS2, required for HIS4 expression (10). Specific mutations in the genes for the twolargest subunits of RNAPII were shown to cause TATA-dependent transcription of HIS4 in the absence ofactivators. Other examples of RNAPII mutants with altered function include mutations in the C-terminal domainof the largest subunit. These mutations, some of which can be suppressed by spt2 mutations, will be discussedfurther in section 1.3. The RNAPII mutations obtained in these three systems affect different regions of thesubunits, do not display common phenotypes, and may define distinct functional domains within the enzyme.371.2 Sin phenotype: Suppression of swi mutationsWhen this project was initiated, the only known phenotype associated with spt2 mutations was thesuppression of Ty element insertional mutations (Spr phenotype). In early 1989, the molecular characterization ofthe SIN] gene, isolated as a negative regulator of HO transcription, led to the finding that SIN] is identical toSPT2. How the sin] mutations were isolated and what is known about the transcriptional role of the SIN1 proteinthus became valuable information for the study of SPT2. This section as well as the next one (1.3) briefly presentsome important aspects of two new phenotypes associated with spt2 mutations, the Sin and Srb phenotypes.1.2.1 Regulation of HO gene expressionThe life cycle of budding yeast defines three specialized cell types: a and a haploids, specialized formating, and a/ce diploids, specialized for meiosis and sporulation. The differences between these cell types aredetermined by a master regulatory mating-type locus, the M4 T locus, for which two functional alleles exist, M4 Taand MA Ta. In heterothallic (ho) strains, these alleles are stable, such that diploidization requires mating betweenhaploids of different clonal origin (and opposite mating type). Homothallic (HO) yeasts can progress through theirlife cycle independently and alternate from the stable diploid state to the temporary haploid state. Thischaracteristic of homothallic yeasts rests on their ability to generate haploid partners of both mating types by amitotic process called mating type switching or interconversion. Since this system provides an opportunity todissect the molecular mechanisms underlying cell specialization in eukaryotes, the generation and maintenance ofthese cell types have been extensively studied (reviewed in ref. 101, 100 and 155).For the purpose of this discussion, a specific question will be considered: “How does mating-typeswitching occur and how is it regulated during the homothallic life cycle?”. DNA sequences homologous to theMAT locus are present at three locations on chromosome III of yeast: at the expressed MAT locus and at two silentloci, HAIL and HMR, located near the left and right telomeres, respectively. In most yeast strains, HML contains acopy ofMA Ta (HMLa) and HAIR, a copy ofM4Ta (HMRa). These two alleles are repressed or silenced by aheterochromatin-like complex requiring the activity of several nuclear proteins (SIR1 to SIR4, RAP 1, histone H4).Even though they are not expressed in wt cells, the silent mating-type loci, HMLa and HMRa, play crucial roles in38the life cycle of homothallic yeasts: they serve as storage locations for the two MAT alleles, thereby assuring thepropagation of both alleles in a population, and act as donors during mating type switching. This process isthought to involve the transposition of a silent “cassette” from one of the silent loci to the M4 T locus by a geneconversion event. The nonreciprocal exchange is initiated by a double-strand break at the AL4T locus, which iscatalyzed by a site-specific endonuclease encoded by the HO gene. The outcome of such a rearrangement is thereplacement of the MAT allele at the MAT locus by the opposite allele.The emergence of HO activity must be tightly regulated to generate the mitotic cell lineage observed inhomothallic strains. This regulation was shown to occur at the level of the transcription of the HO gene: the HOgene transcription is regulated by three independent signals conferring cell-type control, cell-cycle control andmother-daughter or asymmetric control (150). The outcome of these control pathways, ultimately translated interms of the transcriptional properties of the HO promoter, is the regulated production of the HO endonuclease inlate Gi phase of haploid mother cells. Since the transcriptional regulation ofHO is responsible for the generationof cell diversity by asymmetric cell division (113), considerable efforts have been expended on defining cis- andtrans-acting elements involved in this complex regulatory pathway.The cis-acting sequences responsible for HO regulation are present at positions -ito -1400 and comprise aTATA-like element (positions -90 to -130) and two functional sub-regions or upstream regulatory sequences, URS1(-1000 to -1400) and URS2 (-150 to -900), defined by deletion analysis (151, 194, 216). These two regulatoryelements play different and overlapping roles in regulating HO transcription: (I) only URS 1 is essential for HOexpression; (ii) the URS1 is responsible for mother-daughter control; (iii) the URS2 is responsible for cell-cyclecontrol; and (iv) both URS1 and URS2 contain sequences required for haploid-specific expression (al-ct2operators).The products of at least thirteen genes are known to affect HO expression in trans. In diploids, theheterozygous MAT alleles produce the al and (12 proteins, which together repress the expression of haploidspecific genes such as HO. The expression ofHO also requires the products of six SWI genes (SWitch), which actas positive regulators, and five SIN genes (Swi-INdependent), which act as negative regulators. Through generalor specific interactions with the HO 5’-flanking region, these regulatory proteins participate in one or severalaspects of HO expression.391.2.2 Positive regulators: SWI genesThe .swi mutations were isolated in searches for non-switching homothallic strains (i.e. heterothallicmutants) specifically deficient in HO expression, or for strains preventing the expression of a chromosomalho::lacZ fusion (29, 88, 214). Haploid strains canying a mutation in any one of these SWI genes fail to switchmating type because they do not express the HO gene. The participation of SWI proteins in more general cellularfunctions was suggested by the pleiotropic nature of some swi mutations (214). With regard to HO expression, theSWI proteins appear to fall into one of three functional groups. SWI4 and SWI6, cell cycle regulatorsThe SWI4 and SWI6 proteins are involved in cell-cycle regulation of HO. They act through the URS2element, which contains 10 repeats of the sequence, CACGA4. Each of these octamers functions as anindependent, cell-cycle regulatory, cis-acting sequence (cell-cycle box, CCB; ref. 29, 152). The SWI4 and SWT6proteins are components of the DNA-binding, cell-cycle box factor (CCBF; ref. 6, 7, 28, 177) implicated in thecell-cycle expression of HO and of a subset of Gi cyclins (168). The expression of the SWJ4 gene is itself undercell-cycle regulation, and constitutive SWI4 expression abolishes the cell-cycle expression of HO (28). The SWI6protein is also a component of another DNA-binding activity, which specifically recognizes the MluI cell cycle box(30, 142), and is required for the cell-cycle expression of SWI4 and several DNA synthesis genes (53). The abilityof SWI4 and SWI6 to participate in DNA-binding activities and to activate transcription from isolated CACGA4sequences (i.e. from a CA CGA4::lacZ fusion) does not require SWI1,2,3 or SWI5 activity (6, 29). That otherfactors negatively regulate the activity of SWI4 and SWI6 on the HO promoter was suggested by the observationthat in the URS2 context, the cell-cycle boxes lose their UAS activity (see section; ref. 151). SWI5, a mother-cell-specific activatorThe SWI5 gene encodes a TFIIIA-like zinc-finger transcription factor required for mother-cell-specifictranscription of HO (156, 216). The SWT5 protein recognizes sequences in the URS 1 element and acts as a transacting transcriptional activator. Promoter sequences within IJRS1 (from position -1349 to -1290) were shown tocontain a SWI5 binding site in vitro and to confer SWI5-dependent UAS activity to a CYC1::lacZ fusion in vivo(215, 216). Deletion analysis on the whole HO promoter shows that these sequences are required to direct40detectable levels of HO transcription (151). The molecular basis for the mother-cell-specific inheritance of SWI5is still unknown, but two additional regulatoiy steps- the cell-cycle regulation of SWJ5 transcription (154) and ofSWI5 nuclear entry (153) - are thought to participate in this asymmetric distribution of the SWI5-dependenttranscriptional activity at mitosis. Consistent with this view, it was found that constitutive expression of SWJ5leads to HO expression and mating-type switching in both mother and daughter cells (154). A deletion of theURS2 element renders HO transcription SWJ4- and SWI6-independent. Such a deleted promoter (hozlURS2) is stillunder asymmetric and haploid-specific controls, demonstrating the independence of the cIs-acting signals presentin URS 1 and URS2 (151). In the absence of URS2, the HO promoter therefore remains SWJ5-dependent (as wellas SWI1,2,3-dependent). SWI1,2,3, SNF5 and SNF6, transcriptional coactivatorsThe SWI1, SWI2 and SWI3 proteins appear to perform similar functions in the transcriptional activationprocess and are thought to form a complex in vivo (referred to below as SWI1,2,3). In addition to their inability toexpress the HO gene, swil, swi2, or swi3 mutant cells exhibit additional, identical phenotypes such as slow growthon complete medium and nonfermentable carbon sources, aberrant cell morphology and lethality in a swi4background (174, 214). These observations suggested that SWI1,2,3 are also required for the expression ofadditional yeast genes. Several lines of evidence have now confirmed this hypothesis. The SWI] gene was foundto be identical toADR6, previously characterized as a positive effector ofADH1 and ADH2 transcription (174,223). This genetic equivalence provides an explanation for the growth defects of swil mutant strains onnonfermentable carbon sources. The SWII/ADR6 gene encodes a 13 14-amino-acid nuclear protein withpolyglutamine and polyasparagine segments (169). The SWI2 gene is identical to SNF2, a positive transcriptionalregulator of SUC2, the stmctural gene for invertase (134, 174). Haploid snf2 mutant strains were isolated on thebasis of their inability to derepress SUC2 in response to glucose deprivation (Suc phenotype; ref. 160). TheSWI2/SNF2 protein is a 1703-amino-acid nuclear protein with a glutamine-rich region. It is highly conserved ineukaryotes and is homologous to helicases (134). Recent experiments have shown that 5W12/SNF2 possesses aDNA-stimulated ATPase activity required for transcriptional activation by SWI2/SNF2 (132). The SWJ3 gene,which has only been characterized in the HO system, encodes a nuclear protein of 825 amino acids. Two striking41features of the SWI3 protein are a highly acidic N-terminus and a heptapeptide repeat in the C-terminal region(174).What is the function of the SWI1,2,3 complex in the regulation of HO? What is the molecular basis of itsapparent lack of specificity? Several studies, conducted in the last five years or so, have addressed these questions.Although they only provide partial answers, an interesting picture is emerging.As mentioned previously, it appears that the products of the SWJJ/ADR6, SWI2/SNF2, and SWI3 genesform a complex in vivo. The experimental evidence supporting this hypothesis is as follows: (1) none of these SWIproteins is essential for mitotic growth, but their null alleles all share several characteristic phenotypes (174, 175,214, 253); (ii) single and triple swil,2,3 mutants display identical phenotypes, although none of these SWI genesshow SWI-dependent expression (174); (iii) the stability of the SWI3 protein in vivo is decreased in swil or swi2mutant strains (174); (iv) the SWI3 protein was shown to interact with activator proteins in vivo, and thisinteraction is disrupted in swiP or swi? strains (253); (v) direct biochemical evidence for the existence of aSWI1,2,3 complex has been presented recently (173). The SWI1, SWI2, and SWI3 proteins were shown tocopurily in a large complex (Ca 2 MDa) extracted from yeast homogenates by 0.35 M NaC1. Other members of thiscomplex include the SNF5 and SNF6 gene products. As in the case of SNF2/SWJ2, the SNF5 and SNF6 genes werecharacterized by sucrose non-fermenting mutations that prevent activated expression of SUC2 (160). TheSNF2/SWI2, SNF5 and SNF6 proteins were also shown to be required for the activated expression of several yeastgenes and to be functionally interdependent (93, 131, 134). The SNF5 and SNF6 proteins are both localized in thenucleus (65, 133). The SNF5 gene was shown to encode a glutamine- and proline-rich protein (133).The postulated SWII,2,3-SNF5,6 complex (SWI-SNF) has been implicated in the expression of severalregulated yeast genes, including HO, INOJ (175), SUC2, GAL genes and many other glucose-repressible genes (1,131, 174), as well as Ty elements (93). In fact, mutations in SNF2/SWI2 and SNF5 were independently isolated assuppressors of Ty-mediated ADH2 activation (Iye3 and tye4, respectively j39]). Since swil, 2,3 mutants areviable, not all yeast genes are SWI-dependent for their expression. The transcription of several genes, such asURA3, LYS2, CLNJ,2 and 3, SWJ1,2,3,4 and 5, as well as SINJ/SPT2, does not appear to be affected by mutationsin SWJ-SNF genes. Nevertheless, in contrast to mutations affecting gene-specific regulators, the phenotypes of theswi-.snfmutations described here suggest a more general function in the process of transcriptional activation.Consistent with such a function, none of the known components of this complex possesses sequence-specific DNA42binding activity. It is therefore thought that the SWI-SNF complex acts as a coactivator or mediator and functionscoordinately with gene-specific transcriptional activators. Indeed, SWI-SNF facilitates or enhances activatedtranscription, but does not affect basal transcription. This implies that each gene affected by swi mutations isregulated by a SWI-SNF-dependent activator: SWI5 for HO, GAL4 for the GAL genes, ADR1 for ADH2, and1N02 and 1N04 for INO]. The SWI1,2,3 proteins are also required for transcriptional activation mediated by therat glucocorticoid receptor (OR) in yeast. In this system, the SWI3 protein was shown to coimmunoprecipitatewith GR, but only if SWI1 and SWI2 were also present (253). These experiments suggest that the SWI-SNFcomplex contact its target transcriptional activators directly by protein-protein interactions (at least transiently) toparticipate in the transcriptional activation process. A simple model in which the SWI-SNF complex acts as anessential adaptor mediating the activation signal from dedicated transcriptional activators to the generaltranscriptional machinery can be rejected on the basis of two observations: (I) strong UASs composed of multipleactivator binding sites can confer SWI-SNF-independent activation to the downstream basal promoter (131, 173),and (ii) overexpression of a gene-specific activator can alleviate its requirement for SWI-SNF (173).A model for SWI-SNF function accounting for all the available information could propose that gene-specific activators function through two independent yet complementary pathways, one of which would require theaction of the accessory SWI-SNF complex (244). Saturation of the SWI-SNF-independent pathway byoverexpression of the activator or multiplication of its target site in the UAS could mitigate the effects of swi-snfmutations on the activator’s function. Similarly, direct targeting of the SWI-SNF complex to DNA could bypassthe requirement for the gene-specific activator. In support of this view, it was found that fusion proteins consistingof SNF2ISWI2 or SNF5 fused to the DNA binding domain of LexA could activate a lexA operator-IacZ fusion in aSWI-SNF-dependent fashion (134). It is currently unknown what the mechanistic differences between the twopostulated pathways of transcriptional activation might be. An attractive proposal implicates the SWI-SNFcomplex in the elimination of repressive effects maintained by chromatin components (244). This anti-repressionmodel of activation is supported by genetic studies that have led to the description of the SiN genes.431.2.3 Negative regulators: SIN genesThe previous section described how specific transcriptional properties of the HO promoter depend on theaction of regulated activator proteins. Two independent pathways of activation were identified, the SWI5 pathway,responsible for asymmetric control, and the SWI4, SWI6 pathway, responsible for cell cycle control. Thetranscriptional activator SWI5, like several other yeast regulators, was also shown to require the accessory SWISNF complex for maximal activity. How are the two pathways of activation coordinately regulated to generate theobserved pattern of HO expression? A partial answer to this question came from the characterization ofsuppressors of swi mutations. The SWI proteins act as positive regulators of HO. In a specific swr strain, one ofthe SWI proteins is deficient, and the HO gene is not expressed. Extragemc suppressor mutations of a particularswr mutation therefore bypass the requirement for that particular SWI protein. In these double mutant strains, theHO gene is expressed even though an essential activator is missing. The suppressor loci are termed SIN, for SWIindependent. The SIN proteins act as negative regulators ofHO, normally antagonized by the SWI activators. Thesin mutations do not interfere with the diploid-specific repression ofHO by al-cQ (215). Experimentally, the sinmutations were obtained as blue revertants of a white swi ho::lacZ strain (156, 215). On the basis of theirsuppression pattern with regard to swi mutations, the SIN genes define two functional groups. S1N3 and SIN4Mutations in the SIN3 and SIN4 genes were obtained as suppressors of swiY mutations (156, 215). TheSIN3 gene has also been referred to as SDI1 (156). Genetic characterization of their pattern of suppression initiallyshowed that sin3 and sin4 do not bypass the requirements for the other SWI proteins in HO transcription (156,215). It was later found that a null sin4 allele can alleviate partially the requirement for SWI4 and SWI2 in HOexpression (119). Additional phenotypes of sin3 and sin4 mutants, such as larger cell size, slow growth in swi4and swi6 backgrounds and very low sporulation efficiency of homozygous diploids, suggested that the S1N3 andS1N4 proteins are involved in the regulation of other yeast genes in addition to HO.Since the SWI5 activator has been implicated in mother-daughter control of HO expression and matingtype interconversion, an important question to address was whether or not these SiN proteins are involved inasymmetric control. Wild-type daughter cells (of SWI5 SJN phenotype) switch mating types in less than 0.01%of cell divisions, whereas mother cells switch at high frequencies (>75%). Although they do express HO, swiY44s1n3 mutant cells show an aberrant pedigree of mating type switching: both mother and daughter cells expressHO and switch mating type at comparable, but low, efficiencies (10-15% per division; ref. 156). It was alsoobserved that SWI5 s1n3 daughter cells switch mating types at significant levels (16-18% per cell division; ref.156 and 215), presumably due to SWI5-independent HO transcription.A current model for SIN3 function suggests that it regulates the activity of a daughter-specific DNA-binding activity, involved in the repression of HO (216, 236). Molecular characterization of the SIN3 gene showedthat it encodes a 175-kDa polypeptide, localized to the nucleus, and containing four copies of a paired amphipathichelix motif involved in protein-protein interactions, necessary for S1N3 activity (235, 237). These studies alsorevealed that SIN3 is identical to UME4 and RPD1, identified as negative regulators of enes involved insporulation and potassium transport (217, 232). This demonstrates, as suggested by the pleiotropic effects of as1n3 mutation, that STN3 may play a general role in the repression of transcription in yeast. Other studies alsoimply that S1N3 participates in positive regulatory activities (discussed in ref. 237).Originally, the SIN4 protein was thought to act in conjunction with SIN3, but recent experiments havedemonstrated that sin4 mutants possess several unique phenotypes (119). The SIN4 gene has been cloned andshown to be identical to TSF3/GAL22, identified as a negative regulator of the GAL] and GAL1O promoters (35,119). A sin4/tsJ3 mutant strain is temperature sensitive and displays an abnormal cell morphology and defects inprocesses, such as mating and sporulation (35). Interestingly, a null sin4 allele also appears to promotetranscriptional effects characteristic of ssn and spr mutations. These include the partial suppression of thetranscriptional defects of a swi2/snJ2 mutant strain and the suppression of the ö insertional mutations, his4-9128and lys2-]286 (119). These parallels between SSN, SPT and SIN genes as well as histone gene mutations will bediscussed in greater detail below. Like in the case of SIN3, the SIN4 protein has also been implicated in positiveregulation of transcription (119). Finally, some indirect evidence may point to alink between SIN4 function andchromatin structure: (1) sin4 mutations cause a decrease in the linking number of plasmid DNA, which is thoughtto reflect the nucleosome density of the plasmid (119); similar effects are seen in cells depleted for histones H2B orhistones H4 (92, 124); (ii) the bulk chromatin was shown to be severely disturbed in sin4 mutant strains, asdemonstrated by the loss of the characteristic DNA ladder obtained upon micrococcal nuclease digestion ofchromatin (189); (iii) mutations in SIN4, like mutations in the N-terminal domain of histone H4, disrupt c2repression and, as a consequence, derepress ot2-regulated promoters (35, 190).451.2.3.2 SIN1 and SIN2, chromatin componentsMutations in the SIN] and SIN2 genes were obtained as suppressors of swil, or swil and swiYmutations, respectively. Mutations in these two SIN genes were later shown to exhibit the same suppressionspectrum: they can alleviate the requirement for SWI5 and SWI1,2,3 on HO expression (215). In view of theprevious discussion on the transcriptional properties of the HO promoter, the SIN) and SIN2 gene products wouldappear to couple events occurring at URS 1 and URS2. This statement is based largely on the observation that, incontrast to its consequences on the entire HO promoter, a sin] mutation does not suppress the effects of a swilmutation on the UAS activity of the isolated URS 1, which is the site of action of the SWI1,2,3-dependent SWI5activator (215). In view of these results, it was proposed that the role of SIN1 might be to maintain the cell-cycleboxes of URS2 in an inactive state, and that at least one of the roles of the SWI1,2,3-dependent pathway ofactivation of SWI5 is to antagonize this repression. In sin] cells, the CCBs of URS2 are accessible, and the SWI4and SWI6 activators can regulate HO in the absence of the other SWI proteins (6, 127). As discussed elsewhere(101), this model implies that S1N1 functionally links the two activation pathways acting on the HO promoter: itrepresses the function of URS2 and is controlled by regulators acting in URS1.The molecular characterization of SIN] and SIN2 led to the conclusion that both are identical to genespreviously identified: SIN] is identical to SPT2 (127), and SIN2 is identical to HHT], one of the two histone H3genes in Saccharomyces cerevisiae (128, 213, 253). A sin] null allele was shown to restore HO transcription of aswil strain without altering the position of the transcriptional start site (127). The SIN1 protein was also found tobe concentrated in the nucleus and to show significant sequence similarity with the mammalian HIvIG1 proteinsover more than 50% of its residues (33% similarity; ref. 127). These HMG1 proteins constitute an important classof non-histone chromosomal proteins of eukaryotes (reviewed in ref. 120 and 230). Although their cellularfunction remains unknown, they have been implicated in nuclear processes such as transcription (212, 227), DNAreplication (115), and chromatin assembly (25). At the functional level, the parallel between SPT2ISIN1 andHMG1-like proteins was reinforced by demonstrating that a TrpE-SIN1 fusion protein produced in bacteriabehaves as a non-specific DNA binding protein in vitro (127).Studies on the regulation of the HO gene have uncovered a new function of the SPT2/SIN1 protein.Mutations in spt2 can suppress not only cis-acting insertional mutations caused by Ty elements, but also transacting swi mutations. What are the implications of this finding with regard to the role of SPT2 in the cell? It46appears that SPT2/S1N1 belongs to a group of yeast proteins that, together with the histones themselves, areinvolved in the maintenance of the repressive transcriptional effects of chromatin structure. A series ofobservations, summarized in two recent review articles (225, 244), suggests that certain SPT, SIN, and SSN geneproducts, including SFT2/SINJ , SPT6/SSN2O, SPT4, SPT5, and perhaps SIN4, may participate in the establishmentof a chromatin-mediated inactive transcriptional state at several loci. As described above, mutations in any ofthese genes can suppress Ty insertions (Spr phenotype) as well as swil, 2,3 and snJS, 6 mutations (Sin phenotype;although not all combinations have been experimentally tested, the interdependence of SWI1,2,3 and SNF5,6supports this view). The genetic evidence that links histone mutations to the Spt phenotype has already beendiscussed (section Similarly, several observations suggest that the SWI-SNF complex participates in theactivation of transcription by antagonizing the repressive effects of chromatin components, and that by inference,suppressor mutations that alleviate or bypass the requirement for a functional SWI-SNF complex affect proteinsinvolved in this repression. Some of these observations are as follows: (I) the SIN2 gene encodes histone 113; (ii) asin4 mutation affects yeast chromatin (see previous section); (iii) histones H2A and H2B mutations can suppresssnJ2/swi2 and snJ5 mutations (106); (iv) the chromatin structure of the SUC2 promoter is affected in snJ2/swi2 andsnJ5 mutants and this defect is suppressed by deficiencies in SPT6 or H2A-H2B (106, 243); (v) specific histone H4mutations with a Sin phenotype have been isolated (173).1.3 Srb phenotype: Suppression of RNA polymerase II mutationsThe transcriptional defects caused by the absence of functional SWIJ,2,3 gene products can be suppressedby loss of SPT2/SIN1 function. These antagonistic functions, although initially characterized as modulators ofHOexpression, were also shown to regulate the expression of the INO] gene (175). This observation led to thediscovery of an important functional interaction between SPT2/SIN1 and a key player in the transcriptionalmachinery, the RNA polymerase II itself Indeed, several studies had shown already that inositol auxotrophy anddefective INOJ transcription were often associated with mutations in genes coding for subunits of RNAPII. TheIno phenotype of swi strains and of some RNAPII mutant strains turned out to be SPT2/S1N1-dependent. A briefdiscussion of the current understanding of RNAPII structure and function-with emphasis on the generation of47inositol auxotrophy and on the nature of the alleles suppressible by spt2 - will be necessary to appreciate hilly theimplications of this third genetic system linked to SPT2 activity.1.3.1 Yeast RNA polymerase IIIn prokaryotic cells, the synthesis of all cellular RNAs is performed by a single RNA polymerase.Eukaryotic cells have developed a more complex system for RNA synthesis. Three different forms of nuclear RNApolymerase, RNA polymerase I, II, and III, are responsible for the synthesis of rRNA, pre-mRNA, and small stableRNAs, respectively. These polymerases are highly conserved in structure and sequence among eukaryotes, andconsist of complex multisubunit enzymes. Each enzyme contains two nonidentical large subunits (130 to 220 kDa)and 8 to 11 smaller proteins (12 to 80 kDa). The two large subunits share extensive similarity amongst themselvesas well as with the two large subunits of the prokaryotic enzyme, 13 and 13’, but are unique to each class ofpolymerase. Some of the smaller subunits are shared by two or all three polymerases.Yeast R1’TAPII has been studied extensively, first at the biochemical level, and more recently at the geneticlevel (196, 254). The subunit composition of RNAPII from yeast has been defined empirically as the smallestnumber of polypeptides copuril’ing with the DNA-dependent RNA polymerase activity in classical biochemicalfractionation experiments. From these studies, 11 polypeptide subunits with molecular masses ranging from 10 to220 kilodaltons have been assigned to the enzyme (254). The genes encoding these 11 polypeptides have beencloned and sequenced (254). The two largest subunits are part of the “core enzyme” and are thought to containmany of the structural detenninants involved in the definition of the catalytic site required for RNA synthesis. Inyeast, they are encoded by the RPO2J/RPBJ and RP022/RPB2 genes (114, 255).The largest subunit of RNAPII was found to contain an unusual structural feature: the C-tenninal domain(CTD) ofRPO2l consists of several tandem repeats of the consensus heptapeptide sequence, Tyr-Ser-Pro-Thr-SerPro-Ser (4, 46). This C-tenninal extension is absent in the corresponding subunits of RNAPI, RNAPIII and in thebacterial polymerases. Although this structure is conserved in several eukaiyotic RNAPII, the number ofheptapeptide repeats was found to vary among species such that the enzyme has 26 or 27 repeats in yeast (4, 166),42 to 44 in Drosophila (5), and 52 in mouse and hamster (5, 46).The conservation and uniqueness of this structure suggested that it may participate in an essential andunique property of RNAPII. That the CTD is essential for RNAPII function in vivo was demonstrated by genetic48studies in yeast, Drosophila and mice (5, 166, 258). In yeast, extensive deletion studies have shown that the CTDmust contain at least 50% of the repeats (13/26-27) to confer a pseudo-wild-type phenotype. Whereas a largesubunit with less than 10 repeats is not functional and cannot support growth in a rpo2l background, yeast cellscontaining mutant RPO2 1 subunits with 11 to 13 heptapeptide repeats are viable, but display diverse phenotypessuch as general growth defects, cold- and heat-sensitivity (5, 166). These deletions do not appear to affect thestability of the enzyme. It was also found that these partially functional RNAPII mutations cause inositolauxotrophy, due to a decrease in JNOJ transcription (167). A similar Ino phenotype, also thought to reflectdefects in INO] transcription, has been observed in strains carrying several other mutations in different subunits ofRNAPII (11, 198, 199, 249). Similarly, progressive deletions in the CTD of RPO2 1 have been shown to affect theexpression of other yeast genes. These transcriptional defects were found to correlate with a decrease in the abilityof the enzyme to respond to activator proteins (3, 136, 197).How the CTD affects RNA polymerase II function is still unclear, but recent biochemical studies haveprovided important evidence suggesting a role in transcription initiation itself Details of these studies have beenreviewed and are beyond the scope of this discussion (45, 176, 254). Central to the development of these ideas wasthe observation that RNAPII exists in two separable forms in vivo, HA and 110, corresponding to thenonphosphorylated and hyperphosphorylated forms of the largest RPO21 subunit, respectively. This extensivephosphorylation was shown to occur at serine and threonine residues of the CTh. The two forms appear to reflecttwo crucial functional states of RNAPII: (1) only the nonphosphorylated form can participate in the formation ofpreimtiation complexes; (ii) the elongating, polymerizing enzyme is phosphorylated. These results suggest amodel in which the phosphorylation of the CTD ofRPO2l constitutes an important step in the initiation oftranscription, perhaps promoting the release of RNAPII from the preinitiation complex. Consistent with such amodel, the TATA-binding protein (TBP) of TFIID was found to interact specifically with the nonphosphorylatedenzyme (228). Furthermore, the general transcription factor, TFIIH, as well as a yeast initiation factor, factor b,were shown to possess a CTD- kinase activity (69, 140).1.3.2 Suppression of RNAP II mutationsThe link between the function of the CTD of RNAPII’s largest subunit and the SPT2 protein was providedby the observation that swil, 2,3 mutant strains, in addition to their defects in HO transcription, are also Ino49auxotrophs due to impaired INO] transcription (175). This parallel between HO and INO] transcriptional controlswas expanded by demonstrating that, like in the case ofHO, a deletion of SPT2/SIN1 can revert partially thedefects in JNOJ transcription and bypass the requirement for the SWI1,2,3 complex. Similarly, spt2/sinlmutations were shown to suppress partially the JNO1 transcriptional defects caused by truncations in the C-terminal domain of RPO21. This new phenotype caused by spt2 mutations is called Srb, for suppressor of RNApolymerase II (a.k.a. RNA polymerase B). That the suppression is specific to CTD deletions and does not reflect ageneral effect on the INO] promoter is supported by two observations: spt2 mutations do not suppress a deletion ofthe transactivator protein 1N04 or of the RNAPII subunit RPB4, both of which also cause a Ino phenotype andINO1 transcriptional defects. Finally, loss of function mutations in SPT2 partially suppress other phenotypescaused by partial deletions in the CTD, such as the growth defect and cold-sensitive phenotype of strainscontaining a CTD with only 10 heptapeptide repeats, and the lethality of a CTD deletion with only 9 repeats.This functional interaction between SPT2 and the CTD of RNA polymerase II can be interpreted indifferent ways in light of what is known about SPT2 and related proteins. First, there is a clear link betweenRNAPII function and other SPT genes, especially SPTJ5. As described previously, SPT15 codes for the TATAbinding protein, TBP, which seems to provide a docking site for the nonphosphorylated RNAPII at promoterelements. Assuming that a mutant RNAPII with deleted CTD recognizes TBP with lower affinity, SPT2 couldparticipate in this process by limiting access to TBP: loss of SPT2 function could allow a functional RNAPII-TBPinteraction. Second, SPT2 could act at the level of the elongation step and provide a barrier for nonphosphorylatedor partially phosphorylated RNAPII. Both of these models are consistent with the involvement of other chromatinproteins, such as histone H3 (S1N2) and histones H2A-H2B (SPT11,12).1.4 Summary and thesis outlineMutations at the SPT2 gene of yeast have been identified by two independent genetic screens. The Sptphenotype of spt2 mutations represents their ability to suppress the transcriptional effects of inhibitory insertionalmutations caused by sequences, such as his4-912ö and lys2-1285 The Sin phenotype can bypass therequirement for positive effectors of transcription encoded by the SWIJ, 2,3 genes, notably for the expression of the50HO and INOJ genes. The link between SPT2 and the function of the SWI activators was revealed by theequivalency between SPT2 and SJNJ. It was also shown that spt2/sinl mutations can partially suppress thetranscriptional defects caused by deletions in the RNAPII largest subunit (Srb phenotype). These three phenotypesare summarized in figure 4.his4 -9123 OFFhis4-9123spt2 ONhis4-9123rpo21 ONFigure 4. Summary of the transcriptional effects associated with spt2 mutations.The three known phenotypes caused by spt2 mutations and discussed in the last three sections are schematicallyrepresented: 1.1, Spr phenotype; 1.2, Sin phenotype; and 1.3, Srb phenotype. Each phenotype presents a gene(open arrow, for HIS4, HO, or JNO1 coding region) whose expression is affected by a primary cis- or trans-actingmutation (1), which in turn can be suppressed by spt2 mutations (U). Implication of the transcriptional machineryis emphasized by also presenting the Spt phenotype of some RNA polymerase II mutations (rpo2l). The symbolswi denotes mutations in SWI], SWI2, and/or SWI3, and rpo2l CTDzI is a large RNAPII subunit with only 11 to 13heptapeptide repeats.GenotypeHIS4TmnscriptionON1.11.21.3-HO ONHO swi OFFHOswispi2 ONINOI ONINOI swi OFFINO] swi spil ON+ r ----—--HEX)-_ INOI rpo2] CTDA OFFL ------1NO1 rpo2l CTD spt2 ON51What do these phenotypes tell us about the possible cellular function of the SPT2 protein? An importantconclusion arising from the studies described in this introduction is that SPT2 may belong to a new family of yeastnuclear factors involved in the repression of transcription through changes in chromatin structure. Theobservations that histone mutations themselves can lead to Spt or Sin phenotypes, and that other related genessuch as SPT5, SPT6/SNN2O and SIN4 appear to act via changes in chromatin structure, are consistent with such anhypothesis.While these genetic studies suggest that the SPT2 protein is involved in the maintenance of negativetranscriptional effects, it is not yet known how this protein mediates this activity nor what is the functionalsignificance of the sequence similarities with BMG1-like proteins. Based on the available information regardingthe structure of the wt SPT2 protein and the nature of the dominant SPT2-1 mutation, mutational and functionalstudies on the SPT2 gene product were initiated.The mutational analyses were primarily guided by the following question: What are the structuralrequirements for SPT2-mediated repression and for the dominant suppressor phenotype of some SPT2 mutations?Three specific tools were developed to create an in vivo assay system for SPT2 function and expression. Theseinclude: (I) appropriate low-copy number (YCp) and high-copy number (YEp) vectors suitable for in vitromutagenesis and for the expression of spt2 alleles in yeast; (ii) isogenic yeast strains carrying the suppressibleallele his4-9125as well as the wild type SPT2 allele or a recessive null allele (spt2zi:: URA3) for dominance andcomplementation tests respectively; and (iii) anti-SPT2 polyclonal antibodies to monitor the expression of mutantgene products in vivo. Several in vitro mutagenesis approaches were utilized to alter the cloned SPT2+ and SPT2-1 genes in an attempt to identify functionally important regions of the polypeptide. Mutant alleles were assayed bycomplementation and dominance tests for the Spt phenotype, and the expression of several mutants proteins wasalso assayed. The results of this directed mutagenesis approach were also complemented by the analysis of fourspontaneous recessive spt2 alleles.The functional studies were directed towards an analysis of the dominance exhibited by some SPT2alleles. Different recessive and dominant alleles were integrated at the SPT2 locus by allele replacement, and theeffects of wt gene dosage on the function of dominant suppressor alleles was studied. The final phase of the projectinvolved constructing promoter and translational fusions to assess the effects of dominant alleles on thetranscription and translation of the wt SPT2 gene.522. MATERIALS AND METHODS2.1 MaterialsAll DNA modifying enzymes were purchased from Pharmacia, BRL, Promega, New England Biolabs orUSB. Deoxynucleotides and dideoxynucleotides were obtained from Pharmacia. The pyrimidine analog, 5-fluoroorotic acid (5-FOA), was obtained from PCR Inc., through a consortium bulk purchase sponsored by theGenetics Society of America. Media components were from Difco, all other chemicals were from Sigma, Aldrich,BDH and Boehringer-Mannheim. Radiolabeled nucleotides were purchased from New England Nuclear (Dupont).Zymolyase-20T was purchased from ICN Immunobiologicals.2.2 Genetic nomenclatureThe standard rules of genetic nomenclature have been respected for the designation of the in vitrogenerated alleles (206). The dominant alleles are denoted by capital italicized letters (e.g. SPT2-324) and aredefined as the alleles whose products confer a Spr suppressor phenotype (Hisj when expressed in a his4-912bSPT2 background. With the exception of the wt allele, which is designated SPT2, all silent mutations andrecessive alleles are denoted with lowercase italics. The product of the SPT2 gene will be referred to as the SPT2protein, or simply SPT2. A glossary of gene symbols utilized in this thesis as well as important geneticequivalencies are presented in appendices 3 and 4, respectively.2.3 Bacterial work2.3.1 E. coli strains and growthEscherichia coli JM1O 1: supE thizl(lac-proAB) F’[fraD36proAB lacJq lacZAIvIl5] (252) was used forplasmid constructions, yeast vector propagation and preparation of single-stranded DNA for sequencing. E. coliRZ1032: HFrKL16 PO/45[lysA(61-62)] duti ungi thu reM] Zbd-279::Tn]O supE44 (129) was the host for53isolation of uracil-contaimng single-stranded DNA. JM1O1 stocks were maintained on M9+B1 plates (50 mMNa2HPO4,25 mM KH2PO4,8.5 mM NaC1, 20 mlvi NH4C1, 10 mM glucose, 0.1 mM CaCl2, 1 mM MgSO4,0.0001% thiamine [vitamin Bi], 1.5% Bacto-agar). The medium utilized for all bacterial growth in liquid culturesor on plates was YT (0.8% Bacto-tryptone, 0.5% Bacto-yeast extract, 0.5% NaCI, 1.5% Bacto-agar for plates),supplemented with ampiciflin (Amp) at 100 J.Lg/ml for selection of Amp’S transformanta. For f3-galactosidase (Xpeptide complementation, JM1O1 transformants were plated on YT-Amp plates on which was spread 50 .d ofbothX-GAL (2% in DMF) and IPTG (2.2% ‘2°)2.3.2 Transfonnation of E. coilBacterial strains were transformed by the calcium chloride procedure or by electroporation. CaC12-competent cells are prepared and transformed as described (195). Electroporation was performed essentially asrecommended by the manufacturer (Bio-Rad). Briefly, cells are grown to early- to mid-log phase (O])600 0.5-0.8)in YT and successively washed with 1 and 0.5 vol ice-cold dH2O, followed by 0.02 vol ice-cold 10% glycerol. Thefinal pellet is resuspended to a final volume of 0.002 to 0.003 vol in 10% glycerol and aliquots are stored at -70°C.Cells (40 j.d) and DNA (0.1-10 ng in 1-5 j.d) are mixed, incubated 1 mm on ice, and pulsed at 2.5 kV, 25 p.F and200Q using 0.2 cm cuvettes (time constant of 4-5 msec). One ml of YT is added to the mixture and aliquots areplated on YT+Amp plates and incubated overnight at 37°C.2.4 Yeast work2.4.1 Yeast strains and growthSaccharomyces cerevisiae strains used in this study are listed in Table 3. The original his4-912ö strainscarrying the wt SPT2 allele (strain SR26-12C) or the deletion allele spt2-150 (strain S704) were provided byG.S.Roeder. Strain S49 was obtained from the laboratory stock (A.M. Spence). Four strains carrying recessivespt2 mutant alleles (FW157, FW186 to 188) and a strain with suppressible insertional mutations at HIS4 andLYS2 FW1237) were provided by Fred Winston.54Table 3. S. cerevisiae strains used in this study.aName GenotypeSR26-12C MATh his4-912ö ura3-52 leu2 -3 leu2 -112S704 MATa his4-9]25 ura3-52 leu2-3 lysl-1 spt2 -150S49 MATa leu2-1 trpl-]FW1237 MATa his4-9]2Sura3-52 lys2-12&5FW157 MATa his4-912ö1eu2-3 can] -100 ade2-] lys2-1 SUP4-o spt2 -122FW186 MATahis4-9l2öleul can] -100 ade2-1 trpS SUP4-o spt2-22FW187 MA Ta his4-91231eu1 can] -1 00 ade2-1 trps SUP4-o spt2-23FW188 MATa his4-9121eu1 can] -100 ade2-1 trp5 SUP4-o spt2 -204SLL4 MATa his4-912ö ura3-52 leu2 lys]-]SLL5 MATa his4-9125ura3-52 leu2 tip]-]SLL7 MATa his4-9128 ura3-52 leu2 trpl-1 spt2&:URA3SLL9 MATa his4-9l2öura3-52 leu2 trpl-l spt2ziSLL1O MATa his4-9125 ura3-52 leu2 lysi-] spt2zl::URA3SLL1O1 MATa his4-9]23ura3-52 leu2 trpl-1 SPT2 -324SLL1O2 MATa his4-9]2Sura3-52 leu2 trpl-1 spt2-324z3SLL2O8 MATc ura3-52 leu2 lys]-] spt2 -150SLL2O9 MATh ura3-52 leu2 lysi-]a Strains SLL4, SLL5, SLL2O8, and SLL2O9 were obtained as meiotic segregantsfrom the diploid SLL2 (S49xS7O4).Strains SLL5, SLL7, SLL9, SLL1OI, and SLL1O2 are isogenic.Strain SLL1O was obtained as a tneiotic segregant form diploid SLL4x7 (SLL4xSLL7).55Yeast strains were grown and maintained in complete YPD medium (1% yeast extract, 2% peptone, 2%dextrose). Prototrophic transformants and diploids were selected for and maintained in minimal media. Themedium SD is a synthetic minimal medium (0.67% Bacto-yeast nitrogen base [without amino acids], 2% dextroseand 2% Bacto-Agar) and SC is a synthetic complete medium (SD supplemented with the appropriate amino acidrequirements, described in ref. 206). For example, SC-Ura (minimal uracil dropout medium) includes allnutritional requirements except uracil and selects for Ura+ cells. The Ura auxotrophs were selected on mediumcontaining the pyrimidine analog, 5-fluoroorotic acid (5-FOA medium: SC with 0.1% 5-FOA), as describedpreviously (24).The His phenotype of the different mutants was monitored as follows: three independent transformantswere grown to saturation in dropout medium selecting for maintenance of the vector(s) (SC-Ura, SC-Leu or SCLeu-Ura for cotransformants); fixed aliquots of the cultures were used to inoculate fresh medium and the cultureswere grown for 12 to 24 h until a cell density of 2.0 to 4.0 0D600 was reached; equal amounts of cells were spottedon selective dropout plates lacking histidine and incubated at 30°C for two to three days.The generation (doubling) time (g) of SLL strains was determined by monitoring the growth of haploidcells in 100 ml of niinimal medium (SC or SC-His) at 30°C, by measurement of the optical density of the culture at600 nm. When required, aliquots were diluted in culture medium to obtain absorbances in the 0.1 to 1 range. Thegrowth rate was estimated from the slope (m =4[log0D600]/zlt) of the linear portion of the growth curve (plot oflog 0D600 f[time]), which corresponds to the exponential growth phase. For each yeast strain, the generationtime was calculated from the formula, g = log2/m.2.4.2 Genetic methodsThe genetic methods used for yeast strain constructions and analysis have been described previously (87,205, 206). AU diploids were obtained by mating stable (ho) haploid cells of opposite mating types in solution andselecting for complementing auxotrophic markers on appropriate dropout plates. A typical mating reaction isdescribed below for the SLL strain series. The MA Ta and MA Ta haploids were grown to saturation in YPD,diluted in fresh medium, and grown to mid-log phase (1-5x107cells/mi; OD600=0.3-1.7). Aliquots of eachculture (50 .tl) were mixed in 1 ml of fresh YPD in a sterile culture tube and incubated 4 to 6 h at 30°C withoutshaking to allow mating. The success of the reaction was monitored by microscopic observation of zygotes. For56the SLL strain series, all M4Ta cells are also LYS1 trp]-] and all MA Ta cells are lysl-l TRP] (see Table 3,page 54). Therefore, all diploids derived from this series were selected for complementation of these two markerson SC-Lys-Trp dropout plates. The plates were incubated at 30°C for 2 to 3 days and individual diploids wereisolated by restreaking Lys Trp cells on SC-Lys-Trp plates. All diploids were tested for their ability to spomlateas described below.For diploid and random spore analyses, sporulation was induced by nitrogen starvation in liquid medium.Diploids were grown to an 0D600 of 2.5 to 3.0 in YPD, and cells from 1 ml of culture were collected bycentrifugation in a clinical centrifuge (5 mm at 3000 rpm). The cells were washed in 5 ml of sterile water. Thefinal cell pellet was resuspended in 1 ml of sporulation medium (SPO: 1% potassium acetate, 0.1% yeast extract,0.05% dextrose, supplemented with the required nutrients as described above for SC medium). The diploids wereincubated 2 to 3 days at 30°C with shaking (350 rpm) and sporulation was monitored by microscopic observationfor the appearance of asci. In random spore analysis, the sporulation mixture was treated to release isolated sporesand kill unsporulated diploids. Cells and asci were harvested by centrifugation (5 mm at 3000 rpm) and washedthree times with 5 ml of sterile water. The final pellet was resuspended in 5 ml of water and the ascus walls weredigested with Zymolyase (0.5 ml of Zymolyase-20T at 1 mg/nil in d1120) in the presence of approximately 25 mMf3-mercaptoethanol (10 j.tl of 14.4 M stock) for 12 h at 30°C with gentle shaking. Spores from tetrads wereseparated by adding 5 ml of 1.5% nomdet P-40, incubating 15 mm on ice and sonicating 3 times for 30 sec at 50 to75% of full power. If required, the spores were collected by centrifugation (10 mm at 3000 rpm), resuspended in 5ml of 1.5% nomdet P-40, vortexed vigorously, and somcated again. The spores were centrifuged as above andwashed twice with 5 ml of di120. The spore density was estimated using a hemocytometer, and 100 j.d aliquots ofserial dilutions were plated on YPD plates to obtain 20 to 200 spores per plate. Isolated haploids werecharacterized by mating with reporter strains and replica plating on dropout plates.2.4.3 Yeast transformationTwo different methods were used for the introduction of DNA into yeast cells. For high efficiencies oftransformation, yeast spheroplasts were transformed according to published procedures (13, 87, 105, 206). Routinetransformations with replicating vectors were performed by electroporation of whole cells as described by themanufacturer (Bio-Rad), following a protocol adapted from published procedure (16).572.5 Recombinant DNA techniques2.5.1 Isolation of plasmid DNAPlasmid DNA was isolated from bacteria by the alkaline lysis procedure (13). For small scalepreparations (ininipreps), a modified version of this protocol, in which the phenol extraction step has been replacedby an ainmonium acetate precipitation, was utilized (195). Larger amounts of plasmid DNA were prepared from25 mi cultures, by a scaled-up version of the same protocol. Cleaner DNA preparations were obtained by treatingthe final DNA solution with RNase A (20 g/ml, 15 mm at 37°C) and with proteinase K (50 j.tg/ml, 15 mm at 37°C), followed by phenol extraction and ethanol precipitation.2.5.2 Isolation of single-stranded DNASingle-stranded DNA templates for DNA sequencing and oligonucleotide-directed mutagenesis wereprepared following the same procedure, the only difference being the host strain: JM1O1 for ssDNA, RZ1032 forssUDNA (uracil-containing template). All preparations were done on a small scale (1.5 ml cultures) and largeramounts of DNA were obtained from multiple minipreps.The production of single-stranded DNA molecules from vectors containing the origin of replication ofphage fi requires trans-acting factors produced by a helper phage. In this study, all superinfections wereperfonned with R408, a mutant strain of phage fi carrying cis-acting mutations leading to inefficient packaging ofits own genomic DNA (193). The methods for the propagation of this helper phage and the determination of thetiter of the stock (plaque-forming unit [pfu] per ml) were carried out as described (49, 193).The preparation of ssDNA involves growth of a single transformant, superinfection with R408,precipitation of the viral particles released in the culture medium with polyethyleneglycol, extraction with phenoland ethanol precipitation of the ssDNA. The detailed protocol has been published (49).582.6 Plasmid constructions2.6.1 YEp and YCp E. coil-yeast shuffle vectors2.6.1.1 pLL1O (YEpSPT2j and pLL18 (YEpSPT2-1)Original clones of the wt SPT2 locus (plasmid R559: 4.3 kb long PstI-PvuI fragment of SPT2 inserted intothe PstI-PvuI site of pBR322) and of the dominant allele SPT2-1 (plasmid R159: 5.2 kb longX7zol-BamHlfragment of SPT2-1 inserted into the SaiI-BamHI site of pBR325) were provided by G. S. Roeder (180). The SPT2coding region, along with 553 bp of 5’-noncoding sequences and 905 bp of 3’-flanking regions, was subcloned as a2460 bp long NdeI fragment (in which the 3 ‘-recessed ends were filled in using theE. coil DNA polymerase IKienow fragment) in the unique Smal site of the multiple cloning site of pEMBLYe3O (15), to obtain plasmidpLL1O (YEpSPT2, Fig. 5). This yeast shuffle vector is a 2tm-based, high-copy-number vector (YEp) canyingthe LEU2 marker for selection in yeast and the fi replication origin, which allows propagation and packaging ofsingle-stranded DNA carrying the coding (mRNA-like or sense) strand of SPT2. The SPT2-1 allele was similarlysubcloned in pEMBLYe3O to form pLL18 (YEpSPT2-1).Most nonsense mutations, point substitutions and internal deletions were obtained as pLL1O or pLL18derivatives by oligonucleotide-directed mutagenesis (see section 2.8). Seven sequencing primers (15-mers)complementary to the sense strand of SPT2, packaged by pLL1O and pLL18, were designed to span the entire SPT2coding region (see Appendix 1). In the original sequencing of the SPT2 gene, the nucleotide sequence of a 2025 bplong Hindu-AvaT fragment encompassing the SPT2 gene was published (180). This fragment does not contain theNdeI sites defining the fragment subcloned in the present study. Together with unpublished sequencing dataobtained in the course of this study, information available from other sequencing projects carried out in the vicinityof the SPT2 locus on chromosome V, allowed us to define the DNA sequence of the entire NdeI fragment. Thissequence is included in Appendix 1 for reference.59P N MV N P B N HpPR559 L.01_j_1_I.J—j. ....1j°fl— T ITcr SPT21)NdeI2) Fill in3)HpaT+ 4) Purify 2.46 kb long SPT2 fragment+ pEMBLYe3O/SnzalSPT2+4, Ligate (Smh) (ISm)E CE CP H,P,B P H EpLL1Ofi--* 2i onor, LEIJ2 Apr SPT21) Ba,nHI + CIaI2) Purify SPT2’ fragmentB PH E C+ YCpSO IBarnHI + ClaI—b. .SPT2+ fflLigateEP P B PH E C,EpLL15on ARS1 CEN4 on1)EcoRl+ 2) LigateEP P B P HEpLL32on ARS1 CEN4 •—Ap URA3 SPT21 kbFigureS. Construction of plasmids pLL1O (YEpSPT2j and pLL32 (YCpSPT2j.The 2460 bp long NdeI fragment of the SPT2 locus was subcloned in the multiple cloning site of pEMBLYe3O toobtain pLL1O (YEpSPT2j. A 4.2 kb long BamHI-C/al fragment containing the SPT2 gene and the fi origin ofreplication was subcloned in the centromeric plasmid YCp5O to form pLL15. In pLL32 (YCpSPT2j, the EcoRIfragment of the fi origin has been deleted. Restriction enzymes are abbreviated as follows: B, BamHI; C, C/al; E,EcoPJ; H, HindlII; Hp, HpaI; N, NdeI; P, PstI; Sm, Smal; V, PvuI. Parentheses are used to denote sites lost duringsubclomng steps, whereas solid lines cover blunt-ended sites.602.6.1.2 pLL32 (YCpSPT2jand pLL24 (YCpSPT2-1)Low-copy-number (YCp) derivatives canying the SPT2 and SPT2-1 alleles were constructed bysubclomng the 4.2 kb long BamHl-ClaI fragment of pLL1O and pLL18 in the BamHI-CiaI site of YCp5O (122) toform pLL15 (YCpSPT2j and pLL24 (YCpSPT2-]) respectively. Plasmid pLL32 was obtained by deleting theEcoPJ fragment of pLL15 carrying the fi origin (Fig. 5). pLL77The EcoPJ-Sail fragment of pLL1O canying the SPT2 gene was isolated, treated with E. coil DNApolymerase I (Klenow fragment) to fill in the 5’ extensions, and ligated to the 6.1 kb long SphI fragment ofpVT100-U2 (a derivative of the YEpURA3 vector pVT100-U obtained from T. Vernet [231], in which the uniqueEcoPJ site was destroyed) whose 3’-protruding ends had been removed by the exonuclease activity of E. coil DNApolymerase I. This subclomng strategy regenerates the flanking EcoRI and Sail sites and produces a plasmid(pLL77) in which the HindIII and PstI sites, internal to the SPT2 coding region, are both unique (Fig. 6). LikepLL1O, pLL77 is a high-copy-number, E. coil-yeast shuttle vector allowing the packaging of the coding (sense)strand of SPT2. However, pLL77 is maintained in yeast by complementation of uracil (ura3) auxotrophy, andcarries unique restriction sites used for the construction of internal spt2 deletion and disruption alleles.2.6.2 spt2 deletion and disruption alleles (plasmids pLL8O, 81, 82)Plasmid pLL77 was used as a parental vector from which were derived two spt2 alleles, the deletion spt2ziand the disruption (deletion-insertion) spt24:: URA3. The subcloning strategy for the construction of these allelesis presented in figure 6.The plasmid pLL77 was digested to completion with ThndIII and PstI, treated with E. coil DNApolymerase I and ligated in presence of excess Hindlil linkers to create a 8 10-bp deletion in the SPT2 codingregion. As confirmed by DNA sequencing, the resulting plasmid, pLL8O, carries the first 46 codons of SPT2 fusedto codons for the amino acids Lys, Leu, Val, Gly, and Stop (ochre), referred to below as the spt2-46 allele (or spt2ziin strain constructions). This allele was then subcloned as a Bam}{I-EcoPJ fragment in pLL1O to form pLL91(YEpspt2-46).61E E S P S,B PH EpLL1Oft_ 2ji onon LEU2 Apr SPT21) SaIl + EcoRl2)FilIin3) Purify 2.4 long kb SPT2 fragmentB P H+ pVT100-U2/ 7iSPT2Ligate(Sp) (Sp) (Sp)S,B PH E G G S,BpLL77onf1. 2iSPT2 Apr on UIA31) Psl + Hindffl2) Generate blunt ends+ 3) Ligate with Hindffl linkersS,B H E G G S,BpLLSOon -—* 2)1sj,ti.i Apr on IJRA31)Bglfl4 2) LigateS,B H E G S,BpLLS1 _Lcn.. + 1.1-kb Hindffl URA3sp124 on Apr on fragment from YEp24LigateS,B H H E G S,BpLL82______I* on 4 2)1spt2.:URA3 Apr on 1 kbFigure 6. Construction of plasmid pLL77 and of the alleles spt2zI and spt2zl:: URA3.The 2460 bp long fragment ofSPT2 was subcloned as a blunt-ended Sail-EcoPJ fragment from pLL1O into theYEpURA3 vector pVT100-U2 (231) to form pLL77. The deletion allele spt24 (or spt2-46) of pLL8O was obtainedby deletion of the internal PstI-HindIII fragment of pLL77 and ligation of the blunt-ended vector in presence ofHindu linkers. The disruption allele spt2A:: URA3 was constructed by deletion of the URA3 gene of pLLSO (1.1kb long BglII fragment) followed by insertion of the URA3 gene into the unique Hindlil site of spt2zl. Restrictionenzymes are abbreviated as described in the legend of Fig. 5, and as follows: G, Bglll; S, Sail; Sp, SphI.62Plasmid pLL8 1 was obtained by deleting the 1.1 kb long BglII fragment of pLL8O containing the URA3gene. The disruption allele spt2/i:: URA3 was created by inserting the URA3 gene, as a 1.1 kb long HindIIIfragment from YEp24 (27), in the unique Hindill site ofpLL8l, present within spt2il. In pLL82, the orientationof the URA3 gene with regard to the deleted SPT2 locus is such that the two transcriptional units are in oppositedirections (Fig. 6). The resulting allele, spt24:: URA3, was used as a disruption allele in strain constructions(section 2.9).2.6.3 spt2::lacZ fusion (plasmid pLL43)The spt2::lacZ fusion was constructed by subclomng the 1.7 kb long SalI-MluI fragment of pLL4O (seeTable 4, page 66) in the Sail-BamHI site of pLG-fl (a derivative of pLG669-Z [86], constructed by S. S. Ner andcanying the II replication origin) in the presence ofMIuI-BamHI adapters (oLL48a 5’-CGCGGCGGCG-3’ andoLL48b 5’-GATCCGCCGC-3’) that fuse the two coding regions (Fig. 7). In the resulting construct, the fusioncoding region, containing the SPT2 gene at the 5’-end and the lacZ gene at the 3’-end, is expressed from the SPT2promoter on a high-copy-number vector.2.6.4 pGAL4:SPT2 fusions (plasmids pLL94 to pLL97)For the cotransformation experiments, the wt SPT2+ gene was expressed from the weak GAL4 promoteron vectors carrying the URA3 gene as a selectable marker. The plasmid pPS42B-2 containing a 3.4 kb longBamHI-EcoRI fragment of GAL4 (with an additional EcoRI site inserted upstream of the ATG, in the AccI site) inpBR322 (obtained from I. Sadowski) was digested with BamHI and EcoRI to completion, and the 0.4 kb longpGAL4 fragment was subcloned in the BamHl-EcoPJ site of YCplac33 and YEplac195 (81) to create pLL94 (YCppGAL4) and pLL95 (YEp-pGAL4). The 1.9 kb long EcoPJ fragment from pLL16 (see Table 4), carrying the SPT2coding region, was subcloned in the unique EcoRI site of pLL94 and pLL95 to form pLL96 (YCp-pGAL4:SFT2+)and pLL97 (YEp-pGAL4:SPT2+) in which the insertions are oriented to create a transcriptional fusion.63pLG669Z7 T_______________cycl::lacZon Apr LIRAJ1)HindllJ2) Partial EcoRI3) Purify 7.9 Ui long fragmentuspvE+ pVT1O3-U lEcoRI + Hindfflcycl:dacz Ap4, LigatepLG4I ‘‘cycl:thwlon Apr on URA31) Sail + BamHl2) Purify 11.4 kh long fragment3) Ligate with 1.7 Ui long Sail + Mlul SPT2 fragment from pLL4O+ in presence of oLL48A+B MluI-Bamffl adapterspLL43 711° Anr on URA3—I—11kb7.7.__GGC ATA [cGC GGC GGC GA TCC GGACCG TAT GCG ØCG CCG CCT 4fG CCTG I R G G G S GSPT2 oLL4S IacZFigure 7. Construction of plasmid pLL43 (YEpspt2::lacZ).Plasmid pLO-fl was constructed by S. S. Ner by replacing the 2p.-URA3 fragment of pLO669-Z (86) with the fiorigin-2p.-URA3 fragment of pVT1O3-U (231). The spt2::lacZ fusion of plasmid pLL43 was constructed byreplacing the Sail-BamHl fragment of CYC] with the SalI-MluI fragment of pLL4O in presence ofMluI-BamHIadapters. The structure of the resulting junction point has been expanded to show the in-frame fusion between theterminal R333 codon of SPT2 and lacZ, through the adapter oLL48 (boxed). Restriction enzymes are abbreviatedas described in the legend of Fig. 5, and as follows: 0, BgllI; M, MluI; Pv, PvuII; 5, Sail.642.6.5 Internal deletion allelesThree different sets of internal deletion alleles of SPT2 were constructed. In each case, the deletionmutation was introduced in the wit gene and at least one dominant allele. The alleles are symbolized as follows: zi1, for the deletion of residues 60 to 81; 42, residues 48 to 179; 43, residues 117 to 179. Alleles spt2zil and SPT2-]41 were obtained by oligonucleotide-directed mutagenesis on pLL1O (YEpSPT2j and pLL18 (YEpSPT2-1)respectively. Similarly, alleles spt243, spt2-324z13 and spt2-35043 were obtained by looping-out the desirednucleotides with a single mutagenic primer extended on pLL1O, pLL49 (YEpSPT2-324) and pLL64 (YEpSPT2-350), respectively (see section 2.8).The deletion of residues 48 to 179 (42) was constructed by PCR mutagenesis. The 3’ portion of the SPT2coding region (from codon 180) was amplified with the PCR primer oLL34 (see section 2.10) and the mutagenicprimer oLL5 1 (5’-CCCCTGGCGCAACATCTGCAGGAGTATCTrCTGGAGGCAATAGC-3’). The latterintroduces a PstI site (bold) immediately upstream of codon 180 (underlined) and in the same position relative tothe SPT2 reading frame as the PstI site present in the SPT2 coding region after codon 47. The protocols followedfor the PCR reaction as well as for the treatment of the PCR product for cloning are described in section 2.10. The484-bp long PCR product obtained with oLL34 and oLL5 1 was digested with PstI and HincLIII and cloned into thePstI-HindIII site of pLL77, in which these restriction sites, internal to the SPT2 coding region, are both unique. Inthe resulting plasmid (pLL85), codons 47 and 180 are fused at the PstI site to form allele spt242. The allele, spt2-324z12, of plasmid pLL86 was obtained by oligonucleotide-directed mutagenesis on pLL85 (see section 2.8). Bothalleles were subcloned in pLL1O as BamHI-EcoPJ fragments to create pLL89 (YEpspt242) and pLL93 (YEpspt2-32442).2.7 DNA sequencingFor all the mutations introduced in the SPT2+gene, the nucleotide sequence of the entire coding regionwas determined using the Sanger dideoxy chain termination method on ssDNA templates. Two different protocols,utilizing distinct polymerases were routinely used. The DNA polymerase large fragment cKlenow) was used in arapid droplet method previously published (164), whereas modified T7 polymerase (Sequenase) was used asrecommended by the manufacturer (LISB).652.8 Oligonucleotide-directed mutagenesisOligonucleotide-directed mutagenesis (260) was performed essentially as described (130, 13), exploitingthe uracil-containing template selection (129). Oligonucleotides were synthesized on an Applied Biosystems 380Bor 380A DNA synthesizer and purified on C18 Sep Pak cartridges (Millipore) as described (12). Mutagemcprimers were phosphorylated chemically during synthesis (112) or enzymatically with T4 polynucleotide kinase(195). Table 4 presents all the mutations created by oligonucleotide-directed mutagenesis in the present study,along with the template and oligonucleotide utilized in each case. The SPT2-335 allele was obtained by replacingthe C-terminal HindllH-MluI fragment of pLL4O by a synthetic cassette of two 52-mers introducing the mutationsK325M, P.3261, R327L, R328L, K329M, and K330M to the V334 silent mutation spt2-334 of pLL4O. For eachmutant allele presented here, the DNA sequence of the entire coding region was confirmed by dideoxy sequencing,as described above.Several different protocols have been used to perform oligonucleotide-directed mutagenesis. In all cases,the same fundamental approach was exploited: a single 5’-phosphorylated mutagemc primer is annealed to auracil-containing ssDNA template (ssUDNA, prepared from the dur ung strain RZ1032) and extended by a DNApolymerase in the presence of deoxynucleotides (dNTP). Fully extended products are ligated by T4 DNA ligaseand transformed in JM1O1. This dut+ ung+ strain removes the uracil from the template strand and renders itbiologically inactive, thus creating an enrichment for the in vitro synthesized mutant strand. A procedure thatconsistently gave good results is described below.The phosphorylated mutagenic primer (1 pmol) is mixed with the ssUDNA template (0.1 to 0.5 pmol) inthe presence of 1xSSC (150 mM NaC1, 15 mMNa3citrateH2O,pH 7.0) in a final volume of 25 tl. Annealing ispromoted by heating 5 ruin at 65°C followed by progressive cooling to room temperature over 30 mm and 5 mmincubation on ice. The extension/ligation step is catalyzed by 2 units of DNA polymerase large fragment (Klenow)and 2 units of T4 DNA ligase in the presence of polymerase mix (20 J.Ll of a 5x stock: 100 mM Tris•HC1, pH 8.8,10 mlvI DTT, 50 mM MgC12,2.5 mlvi dNTP, 5 mM ATP) in a final volume of 100 1.d. The reaction is incubated 5mm on ice, 5 mm at room temperature and 2 h at 37°C. Aliquots of 10 and 50 .tl are used to transform competentJM1O 1 cells. A control reaction performed in the absence of oligonucleotide primer is always included to assessthe purity of the ssUDNA preparation.66Table 4. Oligonucleotide-directed mutagenesis.aTemplate Mutagenic primer Plasmid Mutation (s) AlleleoLL37oLL38oLL39oLL4OoLL41oLL42oLL43TAGCGTATGCCTGACTI’ACGGCGTCGTATGCCCITfGAACGGCGTCTCATGCCCITCITACTGCGTCTCYFCCTCfTACGGCTfCTCTCFCCTTfCTfACGGCGACCflCTCCECTTTACGGCGTCTFGACTCCTCfCATCCAGAAGATrATI’AACGCAATGTpLLl6 EcoRl site in promoterpLL38 R272OCHREpLL4O V334MluI Site at 3 endpLL44 A60to81pLL47 R3O4OCHREK3O5OCHREpLL48 K319OCHREK32OAMBERpLL49 K325OCIIRER326OPALpLL63 K117OCIIREK118OCHREpLL57 K330SpLL58 K329SpLL59 R328SpLL6O R327SpLL61 R326SpLL62 K325SpLL68 G1800CI-IREV18 1OCHREpLL73 K325RpLL75 A117to179spt2-S01SPT2 -271spt2 -334spt2ii]SPT2 -303SPT2 -318SPT2 -324.cpt2 -116spt2-330spt2-329spt2-328spt2-327spt2-326spt2 -325SPT2 -1 79spt2 -336spt2ii3L2I3OCHREA6Oto 81K325OCHRER326OPALA1 17 to179pLL64 K325S K330SpLL65 K325S K329SpLL76 K325S K330S8117 to179pLL86 K325OCHRER326OPAL M8 to 179SPI’2 -141.spt2-32443SPI’2 -350SPT2 -359spt2 -350113pLL1O oLL2 CITfGAATECGACTCAAGoLLI 8 TCTTGCCITAA1TGAACAoLL23 TCAATAITCCUAAACGCGTATGCCCflCoLL26 CTATlTCTlTFTCfCCAGTCCTGCCGCCTCoLL3O TrGCCA3TfATFATGCCA1TlCoLL3 1 CTCATGCTA1TATAACCAAGCoLL3Z TACGGCGTCATfACTCCTC11toLL36 GTfCAGGTrATrACCTTACAACoLL47 TfACGGCGTCTTCTCFCCCtTCAoLL49 TGClAHGCCTCCAGAAGATACfCCCCTFACAACFGGCTrGAGTGGCGCApLL18 oLL26 See above pLL45pLL49 oLL49 Idem pLL74p11.62 oLL37 IdemoLL38 IdempLL64 oLL49 IdempLL85 oLL32 Idema For each mutation created by oligonucleotide-directed mutagenesis are presented the template and primer(name of oligonucleotide and sequence from 5’ to 3 end) used for the mutagenesis reaction,as well as the designation ofthe resulting plasmid, mutation and allele. Mutant bases are in bold.spt2 -32442672.9 Yeast strain constructionsIn the course of this work, several haploid and diploid yeast strains of specific genotypes have beengenerated. These yeast strain constructions, summarized in figure 8, were oriented towards the generation of threeoverlapping sets of strain. The first set consists ofM4 Ta and M4 Ta haploid strains canying the suppressiblemsertional allele his4-912ö and the wt SPT2 gene and allows phenotypic selection for diploid formation. Thesestrains are SLL3, SLL4, SLL5, and SLL6 (see Table 3 and Fig. 8). The second set consists of isogenic his4-9125SPT2 and his4-9]28spt2 strains for dominance and complementation tests (SLL5 and SLL7). The third setconsists of isogemc his4-9125 strains carrying different spt2 alleles integrated at the SPT2 locus (SLL5, SLL7,SLL9, SLL1O1, SLL1O2).The diploid SLL2 was obtained by mating S49 (MA Ta leu2-1 lip]-’) with S704 (MA Ta his4-9]2ö ura3-52 leu2-3 lysi-] spt2-150) and selecting for Trp Lys Ura }Jj+ mating products on SD+Leu plates. StrainsSLL3 to SLL6 were obtained as 5rnAr (Ura) meiotic segregants from SLL2. In this first step, the segregation ofthe ura3-52 marker was followed for two reasons: (I) the selection kills all unsporulated diploids (which are Ura+,therefore 5-FOA); (ii) the resulting set of haploid cells should all carry the ura3-52 marker, frequently used forvector selection in transfonnations. In the second step, independent 5-FOA’ spores were streaked on SC-His platesto identily }{is cells, which should have the genotype his4-9125SPT2. The other allele combinations obtainedfrom the diploid SLL2 all lead to a Ths phenotype (HIS4 SPT2, HIS4 spt2-]50, and his4-9]2öspt2-]50).Finally, 5FoAr His cells were streaked on SC-Lys and SC-Trp plates to determine their Lys and Tip phenotypes,and mated to S49 and S704 to identiiy their mating type.For the growth studies, two HJS4 strains carrying the SPT2 or the spt2-]50 allele were also derivedfrom the diploid strain SLL2. These strains, SLL2O8 and SLL2O9 (see Table 3), were obtained as 5-FOA’ HisTrp meiotic segregants of SLL2. The presence of the wt HIS4 allele was scored by the His phenotype of diploidsobtained by mating with SLL5, and the segregation of the spt2-]5O allele was followed by radiation sensitivity (74,245) and analytical PCR.68S49 x S7041) mate2) select for diploids on SC-Lys-Trp platesSLL21) sporulate2) select for Ura haploids on 5-FOA plates3) screen for His haploids4) determine mating type and phenotypesMATa MAToLys Trp SLL3 SLL4Lys Trp SLL5 SLL61) transform SLL5 with spt2iA::URA32) select for Ura+ transformantsSLL71) transform with spl2 alleles / 1) mate with SLIA2) select for Ura /‘ \ 2) select for Lys Trp diploidstransforinants /on 5-FOA plates /SLL4X71) sporulateSLL9 2) select for His haploidsSLL1O1 3) screen for MATaSLL1O2 Lys Trp cellsSLL1OFigure 8. Yeast strain constructions.The diagram shows the origin of the yeast strains SLL3 to SLL7, SLL9, SLL1O, SLL1O1 and SLL1O2, constructedin the present study. Strains SLL3 to SLL6 are his44128SPT2, the only allele combination leading to a Hisphenotype in the spores from the diploid SLL2 (HJS4/his4-9125SPT2/spt2-15O). The spt2 alleles used toreplace the spt2/J:: URA3 allele of SLL7 are as follows: spt2zi for SLL9, SPT2-324 for SLL1O1, and spt2-324/J3 forSLL1O2. Strain SLL1O is phenotypically identical to SLL4 (except for the Ura and His phenotypes), but carries thedisruption allele spt2&: URA3, the only allele forming His spores from the diploid SLL4X7 (his4-9125/his4-9125SPT2/spt2A:: URA3).69The haploid strain SLL5 (Table 3) was used as a SPT2 parental strain from which was derived a set offour isogenic strains carrying different alleles at the SPT2 locus. SLL5 was transformed with 3 g of the 2.6 kblong EcoRI-BamHl fragment of pLL82 containing the spt2zl:: URA3 disruption allele. Stable Ura transformantswere selected as described by the one-step gene disruption procedure (191). The substitution of the wt SPT2 geneby the mutant allele spt2zi:: URA3 in SLL7 was confirmed by genomic DNA analyses and genetic methods(cosegregation of the Ura+ and recessive His+ phenotypes in random spore analysis from the diploid obtained bymating SLL4 and SLL7, SLL4x7).A cotransformation procedure allowing direct allele replacements of the spt24:: URA3 allele in SLL7 by 5-FOA counter-selection and fast screening by analytical PCR analysis was developed (in collaboration with Chi-YipHo). SLL7 spheroplasts were cotransformed with ito 10 j.tg of digested plasmid DNA carrying the allele ofinterest and 0.1 to 1 ig of pEMBLYe3O, and plated overnight on SC-Len plates. The regeneration agar was thentransferred to a 5-FOA-Leu plate, which selects for loss of URA3 function (24). The desired transformants werescreened by analytical PCR amplification of the SPT2 locus (section 2.10) and positives clones were cured of thepEMBLYe3O vector by growth in YPD. Following this procedure, the isogenic strains SLL9 (spt2A, from pLL8O),SLL1O1 (SPT2-324, from pLL49) and SLL1O2 (spt2-324z13, from pLL74) were constructed.2.10 PCR analysis of SPT2 allelesDuring the course of this work, the development of the polymerase chain reaction (PCR) revolutionizedmolecular biology techniques. The advantages of this technique were exploited in three different ways in thecourse of this work: (I) for analytical or diagnostic purposes; (ii) for the cloning and mutagenesis of DNAfragments; (iii) for direct sequencing of genomic sequences.2.10.1 Preparation of crude yeast DNACrude yeast genomic DNA samples suitable for PCR amplification were prepared by a fast procedurebased on a method originally developed for yeast plasmid rescue (110). In this protocol, yeast cells are simplybroken open by vortexing in the presence of glass beads, detergents, and phenol. Yeast were grown overnight in 570ml of medium, and the cells from 1.5 ml of culture were harvested by 5 sec centrifugation in a microfuge. The cellpellet was resuspended in 200 il of yeast DNA buffer (YDB: 2% Triton X-100, 1% SDS, 100 mM NaCl, 10 mMTris•HC1 [pH 8.0], 1 mM EDTA) and 200 i.tl of phenol/chloroform and 200 jil of glass beads were added. Thecells were lysed by vortexing for 2 mm. The cell lysate was recovered by pelleting the glass beads and cell debrisby centrifugation for 5 mm in a microfuge.2.10.2 Analytical PCRAnalytical or diagnostic PCR was used to analyze qualitatively the structure of the SPT2 locus. Theprimers used for this purpose were oLL34 (5’-CGCGGATGCCCITCITAIGGIGTITCITCTCCTCTTCATGCTTTTTTAACCA-3’, from the ThndIII-MluI cassette used to form allele SPT2-335 [where I represents inosineresidues introduced for random cassette experiments not described herel ) and oLL5O (5’-AAGTTGAATTCAGGGACAGGGACTrGAGTC-3), which hybridizes to the 5’ noncoding region of SPT2. Figures hA and 21A presentthe location of the oLL34 and oLL5O priming sites with regard to different spt2 alleles and the size of theamplification product expected in each case.The polymerase chain reactions were performed in a Perkin Elmer Cetus DNA Thermal Cycler using 1 ngof plasmid DNA or 2 1il of total yeast lysate as template, 50 pmol of each primer, 10 nmol of each dNTP(Pharmacia), 1 unit of Taq DNA polymerase (obtained from Cetus, Pharmacia or Promega) and 10 tl of 10 X PCRbuffer [100 mM TrisHCl (pH 8.4), 6 mM MgC12,0.5% Tween 20, 0.5% nomdet P-40] in a total volume of 25 or50 j.tl. The reaction mixtures were overlaid with two drops of oil (Stanley) and subjected to a denaturation step of 2mm at 94°C followed by 20 to 30 amplification cycles (10 sec at 94°C, 30 sec at 53°C and 1 mm at 72°C) and afinal extension of 5 mm at 72° C. The aqueous phase was recovered and 2 il was analyzed on an agarose gel.The amplification of the CYC] gene was achieved similarly using the primers, CYH7 (5’-GAGGGCGTGAATGTAAGCGTGACATAACTA-3’) and CYH8 (GCATAAATrACTATACTrCTATAGACACGC).712.10.3 Cloning of PCR productsThe use of PCR as a preparative tool to generate restriction fragments suitable for cloning experiments isbased essentially on the same principle and methodology as the analytical PCR. Since the purpose of theamplification reaction is not simply to produce a DNA fragment of a specific size, but to generate copies of atemplate molecule for cloning, several parameters must be considered. The mutagemc potential of Taq DNApolymerase must be minimized to avoid the cloning of so-called “PCR mutations’. This can be achieved by usinghigh-quality reagents and determining the optimal conditions for a specific pair of PCR primers. For example,imbalances in the deoxynucleotide pools are known to promote misincorporations. Amplification cycles are alsokept to a minimum (no more than 20) and the reaction is scaled up to produce enough material for cloning (100 .ilreactions).The actual cloning of PCR products was carried out essentially as for restriction fragments, although theseproducts were found to be particularly resistant to ligation. This problem was overcome by treating the PCRproducts with proteinase K as described (47), digesting them with the appropriate enzymes, and purifying themusing the Geneclean Kit (BlO 101) after electrophoresis on a 1.5% agarose gel. The protease treatment is thoughtto help degrade and remove the tightly bound polymerase, which interferes with subsequent enzymaticmodification of the PCR product.2.10.4 Asymmetric PCRThe generation of PCR products as ssDNA template suitable for direct sequencing by the dideoxy chaintermination procedure was based on a published protocol (141). This asymmetric PCR was performed essentiallyas described above, but for 35 cycles and in the presence of a 100-fold excess of the PCR primer of the desiredtemplate strand. A typical reaction, which produces an enrichment of the sense strand of SPT2 (identical to thepackaged strand ofpLLlO), contained 50 pmol of oLL5O and 0.5 pmol of oLL34. The successful utilization of thistemplate molecule in sequencing reactions was also highly dependent on the complete removal of unreacteddeoxynucleotides and oligonucleotide primers. This was done by extracting the reaction mixture with phenol andsubmitting the aqueous phase to three successive ultrafiltration steps by centrifugation 2 to 4 mm at 2000g on72Ultrafree-MC filters (Millipore) with a molecular weight cut-off of 30,000 (which corresponds approximately to assDNA molecule of 100 nucleotides).2.11 Immunological techniques2.11.1 Production of anti-SPT2-peptide antiseraPolyclonal antibodies were raised against two synthetic peptides (provided by Ian Clarke-Lewis) designedfrom the predicted amino acid sequence of the SPT2 protein: PSi and PS2, spanning amino acids 209 to 238, and297 to 325 respectively (see Fig. 3, page 32, and Appendix 2). In addition to the SPT2 segments, each peptidecontains, at the amino terminus, two linker glycines and a reactive cysteine, used for couplings to carrier proteinsor chromatographic supports. The peptides were synthesized using Na-tert-butyloxycarbonyl-protected aminoacids with appropriate side chain protections, on an Applied Biosystems 430A synthesizer, using methodsdescribed elsewhere (41). The peptides were purified by reverse phase HPLC and coupled to the carrier proteinkeyhole limpet hemocyanin (KLH) as described (259).For each peptide (1 and 2), two New Zealand White rabbits (1A, lB and 2A, 2B) were immunizedfollowing the same immunization schedule: primary injections with 500 ig of peptides, followed by secondaryinjections with 200 .tg and subsequent boosts with 100 g of peptides, all at 1 month intervals. The primaryinjections were performed subcutaneously, whereas all subsequent injections were performed intramuscularly.Each injection (200 to 500 i.tl) consisted of an emulsion prepared from equal volumes of antigen solution andFreund’s adjuvant. Complete Freund’s adjuvant (containing killedM tuberculosis) was used only for the primaryinjections.Test bleeds of 5 to 10 ml (from the marginal ear vein) were collected before the first injection (preimmune sera, samples #0) as well as 10 days after the secondary injections (samples #1) and subsequent boosts(samples #2 and 3). Serum was prepared by allowing the blood to clot 60 mm at 37°C, followed by overnightincubation at 4°C to allow the clot to contract. The serum was collected after centrifugation at 10 000g for 10 mmat 4°C. The activity of each antisera against the peptide antigens were determined by immunoassays.732.11.2 ELISAThe presence of specific anti-peptide antibodies in the different bleeds as well as during purification stepswas monitored and quantified by enzyme-linked immunosorbent assay (ELISA). This immunoassay, also calledantibody capture assay, can be described as follows: the antigen is attached to a solid support, and specificantibodies are allowed to bind to their epitope, present in excess. The amounts of bound antibodies are quantifiedindirectly through an anti-rabbit-IgG enzyme conjugate.The protocol used to perform these assays is essentially as described elsewhere (94). The peptides aredissolved in 10 mlvi acetic acid to 1 mg/mI and diluted to 10 .tg/mi with phosphate-buffered saline (PBS: 137 mMNaC1, 2.7 mlvi KC1, 4.3 mlviNa2HPO4,1.4 mM KH2PO4)containing 0.02% sodium azide. The wells of amicrotiter plate (polystyrene ELISA/EIA plate from Corning) are first coated with 100 .tl of peptide solution (1 jig,as recommended for maximal binding [94] ) and incubated overnight at 4°C to promote antigen binding. Theexcess antigen is removed by three successive washes with wash buffer (0.05% Tween 20 in PBS). The remainingnon-specific binding sites are blocked or quenched with 200 j.tl of quenching buffer (5% skim milk in PBS) byincubation 1 h at room temperature, followed by three washes. The antibody solutions to test (primary antibody)are prepared from sera or purified fractions diluted in 1:10 quenching buffer (0.5% skim milk in PBS). 100 il ofeach antibody solution is added to the coated wells and incubated 1 h at room temperature to promote theformation of specific antibody-antigen complexes. Non-reacting antibodies are removed by three washes. Thebound antibodies are detected and quantified by repeating the antigen binding step using an affinity purified goatanti-rabbit IgG alkaline phosphatase conjugate (Bio-Rad) as secondary antibody. 100 jil of a 1:1000 dilution in1:10 quenching buffer is used per well, incubated 1 h at room temperature and washed three times as above. Theenzymatic reaction is initiated by the addition of 100 .tl of the substrate p-nitrophenyl phosphate (PNPP, Sigma104-0) at 1 mg/mI in 10 mM diethanolamine (pH 9.5), 0.5 mMMgCl2 The reaction is stopped with 50 jil of 0.1M EDTA after approximately 30 mm at room temperature (or after appearance of yellow coloration). Theconcentration of the product is measured by optical densitometiy at 405 nm.742.11.3 Immunoaffinity purificationPolyclonal antibodies specific to both SPT2 peptides were purified from high titer sera by immunoaffinitychromatography. The peptides PSi and PS2, which contain terminal cysteine residues, were covalently coupled tothe chromatographic matrix, activated thiol-sepharose 4B (Pharmacia), as recommended by the manufacturer. Thepurification of anti-peptide antibodies was performed essentially as described (94). The peptide-bead conjugateswere washed thoroughly with PBS and 2 ml of beads were mixed with 2 ml of serum and 8 ml of PBS in a 15-mitube. Antibody binding was achieved by rotation at 4°C for 2 h. The mixture was poured in a small column,washed with 20 ml of PBS and 20 ml of 0.2 M NaC1 in PBS. The antibodies were eluted with 8 ml of 100 mMglycine (pH 2.5). 900 tl fractions were collected in 1.5 mi-tubes containing 100 jil of 1.0 M Tris•HC1 (pH 8.0) toneutralize the pH of the eluate. The columns were washed with 10 ml of 100 mM Tris•HC1 (pH 8.0) and 10 ml of0.02% sodium azide in PBS for storage. All the fractions were assayed for total protein concentration by UVabsorption at 280 nm and for reactivity against the specific antigen by ELISA. Fractions containing the antipeptide antibodies were pooled, concentrated by ultrafiltration (Amicon Centricon concentrators) and stored inPBS containing 1% BSA and 0.02% sodium azide.2.11.4 Immunoblotting2.11.4.1 Yeast protein extractsTotal yeast protein extracts were prepared by mechanical lysis with glass beads in the presence of SDS.Yeast cells were grown to late log phase at 30°C in 5 ml of minimal medium, selecting for the vector if required.Aliquots of the cultures were used to determine the optical density at 600 urn, and equal amounts of cells (inca 1.0ml) were harvested by centriliigation in a microfuge (12, 000 g) for 5 sec at 4°C. The cell pellet was resuspendedin 500.tl of ice-cold 50 mM Tris.HC1 (pH 8.0), transferred to a twist-cap 1.5 ml tube and harvested as above. Cellswere resuspended in 200 l of yeast lysis buffer [YLB: 20 mM Tris•HCI (pH 8.0), 0.1% Triton X-i00, 0.5% SDSIand 200 l of acid-washed glass beads were added to the cell suspension. Yeast cells were lysed by 5 cycles ofvigorous vortexing (20 sec), each followed by 1 mm incubation on ice. The protein extracts were recovered bypuncturing the bottom of the tube with a hot needle (23 -gauge) and centrifugation in a recipient 1.5 ml tube in a75clinical centrifuge at 4°C. The extracts were cleared by centrifugation in a niicrofhge for 5 mm at 4°C and storedat -20°C. SDS-polyacrylamide gel electrophoresisThe electrophoretic separations of proteins were performed using the SDS-polyacrylamide gelelectrophoresis (PAGE) system of Laenmili, following published procedures (94, 195). All separations wereexecuted with the mini-protean II apparatus of Bio-Rad, following the manufacturer’s instructions. Protein blotting and immunoprobingFor Western immunoblot analysis, the proteins resolved by SD S-PAGE were electroblotted onto amtrocellulose membrane (Hybond-ECL, Amersham) using the Mini trans-blot apparatus of Bio-Rad, following themanufacturer’s instructions. The electrophoretic transfers were performed in a buffer containing 50 mlvi Tris, 380mM glycine, 20% methanol and 0.1% (wtlvol) SDS (94) at 4°C for 1 h, at a constant voltage of 100 V. Themembranes were treated successively with blocking agent (5% skim milk in PBS), primary antibodies, secondaryantibodies and detection reagents as described for the enhanced cheniiluminescence (ECL) protocol of Amersham.The prnnaiy antibodies consisted of 1:800 to 1:1000 dilutions of both affinity-purified anti-peptide antisera. Agoat anti-rabbit IgG horseradish peroxidase conjugate (Bio-Rad) was used at a 1:6000 dilution as secondaryantibody. Unless otherwise specified, the membranes were exposed for 15 to 30 sec to a Hyperfllm-ECL(Amershain).2.12 Cenomic DNA analysis2.12.1 Preparation of yeast genotnic DNAYeast genomic DNA for restriction analyses was prepared from 5-mi cultures as described elsewhere (13).In this procedure, yeast spheroplasts were prepared by Zymolyase treatment and lysed in the presence of SDS.Genomic DNA was recovered after precipitation of proteins and SDS with potassium acetate. The final DNApellet was treated with ribonuclease, proteinase, and extracted with phenol. This protocol yields approximately 10.tg of yeast DNA.762.12.2 Hybridization in dried agarose matrixAll the restriction analyses of yeast genomic DNA presented here were performed by hybridization indried agarose gels using oligonucleotide probes. This fast procedure, developed from a protocol provided by EvanMcIntosh, takes advantage of the ability of small oligonucleotide probes to diffuse freely and rapidly inside apartially dehydrated and collapsed agarose gel matrix.The oligonucleotide probes (1 pruol) were prepared by 5’-end-labeling with T4 polynucleotide kinase (1unit) in the presence ofy-32PdATP (10 .tCi, 3.3 pmol) and 1 x PNK buffer (0.5 M Tris•HC1 [pH7.5], 0.1 MMgC12,50 mM DTT) in a 20 .tl reaction for 30 mm at 37°C.2 to 4 j.tg of yeast genomic DNA were digested with 25 units of enzymes for 5 to 8 h at 37°C asrecommended by the manufacturer, and separated on a 0.8% agarose gel. The DNA was denatured by successivelysoaking the gel in 0.5 M NaOH and 0.5 M Tns•HC1 (pH 8.0), both for 30 mm at room temperature with shaking,followed by water rinses. The agarose gel was placed on 2 sheets of Whatman 3MM paper, covered with SaranWrap and dried on a gel drier for 30 mm with no heat, followed by 30 to 60 mm at 60 °C. The gel was separatedfrom the filter paper by briefly soaking in 5 x SSC. Prehybndization was performed in a sealable bag containing0.035% ATP, 6 x SSC, 0.5% SDS, 0.05% heparin and 0.05% sodium pyrophosphate in 20 ml for 30 mm at 40 to50°C. The crude labeled oligonucleotide was added and incubated for 2 to 16 h. The gel was washed twice in 6 xSSC at room temperature for 15 mm followed by stringent washes in 1 to 6 x SSC at 58°C for 2 mm. Thestringency of the washes was increased by decreasing the salt concentration and the washes were repeated until nocounts above background were detected over most of the gel. This procedure was adopted since the dried agarosematrix starts to collapse at temperatures above 60° C. The gel was dried briefly without heat and exposed to a Xray film (X-OMAT, Kodak) at -70°C.Four different oligonucleotide probes were used in this study. The sequences of these oligonucleotides aregiven below.oLL8 5’- GCTTGACGGGCTTGG -3’oLL27 5’- GGGTAACAAGGAATACTCTT’CGTCArTC-3’oLL5 1 5’-CCCCTGGCGCAACATCTGCAGGAGTATCI7CTGGAGGCAATAGC-3’CHY-26 5’- GTGCATGATATTAAATAGCTTGGCA-3’773. RESULTS AND DISCUSSION3.1 In vivo assay system for SPT2 function and expressionA yeast strain carrying the his4-9l2Sinsertional mutation shows a His phenotype at 30°C. A completedeletion of the SPT2+ gene is a recessive suppressor of the promoter insertion, resulting in a His+ phenotype.Therefore, the suppressor phenotype corresponds to a loss of function of the 333-amino-acid SPT2 protein. Severaldominant and partially dominant suppressor spt2 mutations have been characterized genetically and one suchallele, SPT2-1 , was shown to be caused by a nonsense ochre mutation (TAA) at the Leu 213 codon ‘ETA.To gain a better understanding of the molecular basis of SPT2 function in yeast, a mutational analysis ofthis protein was undertaken. The basic experimental approach involved three steps: (1) manipulate the clonedSPT2 gene in vitro using different mutagenesis protocols; (ii) reintroduce the mutant alleles in his4-9128 mutantstrains; and (iii) study the effects of the mutations on SPT2 function. Since several SPT2 alleles were shown tobehave as dominant alleles, two different questions were asked for each mutant allele created: (i) is the mutantprotein still functional or does it suppress the insertional mutation his4-9126? (ii) if it is not functional forrepression (and therefore a suppressor), does it confer a dominant or recessive suppressor phenotype? From agenetic perspective, these two questions can be answered by complementation and dominance tests, respectively.First, an in vivo system appropriate for these studies was established. Such a system would consist ofefficient E. coil-yeast shuttle vectors for in vitro mutagenesis and expression in yeast, and isogenic SPT2+ and spt2strains carrying a Ty or insertional mutation suppressible by spt2 mutations. Another important molecular toolin structure-function studies similar to the one presented here is the availability of a specific assay or probe toassess the expression and stability of the mutant proteins in vivo. Consequently, the preparation of polyclonal antiSPT2 antibodies also was undertaken. The next three sections present the development of each of the componentsof this in vivo assay system.783.1.1 Expression vectorsPrevious studies on the SPT2 gene have led to the cloning and sequencing of two different alleles of theSPT2 locus: the wt SPT2 gene and the dominant mutant allele, SPT2-1 (180). In the course of this work, it wasshown that a 43 kb long PstI-PvuI fragmentof the wt SPT2 locus was sufficient to complement a null spt2mutation (spt2-150) when introduced on the replicating yeast vector YRp1O. The DNA clones of the SPT2(plasmid R559) and SPT2-1 (plasmid R159) alleles were obtained and subclomng steps in E. coil-yeast shuttlevectors were undertaken. The primary goals were to identify a small fragment of the SPT2 locus sufficient forappropriate expression in yeast and to subclone it in low-copy-number (YCp) and high-copy-number (YEp) vectorsallowing the production of ssDNA for oligonucleotide-directed mutagenesis and sequencing. Another objectivewas to avoid part of the Tyl element discovered 926 bp downstream of the SPT2 coding region (see map of SPT2locus, Fig 9A[180]).It was found that a 2460-bp NdeI fragment of the SPT2 locus can complement the spt2-150 mutation ofstrain S704 when introduced on the high-copy-number YEp vector, pEMBLYe3O (pLL1O, Fig. 9B; ref. 15). Thisfragment contains part of the RAD4 coding region (see below), 283 bp of intergemc sequences, the 999 bp longSPT2 open reading frame, and 905 bp of 3’-flanking region. It also avoids all Ty sequences (see Appendix 1).It should be noted that, since the SPT2 gene has been cloned, the nearby RAD4 gene, located within 0.25cM of SPT2, has also been characterized (74, 80). The RAD4 coding region terminates 283 bp upstream of theSPT2 translation initiation codon, suggesting that the promoter elements required for SPT2 expression are locatedin a small region of the 5’ flanking sequences. The cloned RAD4 locus is apparently lethal in E. coil andundergoes frequent rearrangements and mutagenesis when introduced into bacteria (74). In the original SPT2clone, the nearby RAD4 open reading frame was disrupted by two frameshift mutations (74, 180). Similarly, theyeast genomic fragment containing the SIN1/SPT2 gene isolated in E. coil harbored a transposable elementinsertion (Tn) within the RAD4 coding sequences (127). The NdeI fragment of SPT2 contains the last 272 bp ofthe 2265-bp RAD4 coding region (80) and is not functional for RAD4, since it does not complement the Radphenotype of strain S704, in which regions of both SPT2 and RAD4 are at least partially deleted (not shown).79A2.46kbXN H P N G BI I I I I I[4’ I I ——4 4Tyl SPT2+ p4JJ4 I I1 kbB (N) (N)E E P H,P,B P H EpLL1O___L______________i-i-I-.’’2j.t onon LEU2 bla SPT21 kbCEP P 11 PH EpLL32-•-•* on ARS1 CEN4II kbD(P)spt2A::URA3_URA3Figure 9. SPT2 locus and plasmid constructions.(A) Physical map of the SPT2 and RAD4 loci on the right arm of chromosome V. Also shown are the location of anearby Tyl element and the position of the 2460-bp NdeI fragment subcloned for these studies. (B) Map ofplasmid pLL1O (YEpSPT2j, a derivative of pEMBLYe3O, which carries the NdeI fragment of the SPT2 locusinserted in the multiple cloning site (at Smal site). (C) Map of plasmid pLL32 (YCpSPT2j, a derivative ofYCp5O, canying the Bam}{I-EcoRI SPT2 fragment of pLL1O. (B) Structure of the disruption allele spt2iJ:: URA3used in strain construction. A 810 bp long fragment of the SPT2 coding region (PstI-ThndIll fragment) wasdeleted and replaced by a 1.1 kb long URA3 fragment. Restriction enzymes are abbreviated as follows: B, BamH[;E, EcoPJ; G, BglII; H, HindJII; N, NdeI; P, PstI; X, XhoI. Parentheses are used to highlight restrictionendonuclease recognition sites lost during subcloning steps.80The same SPT2 fragment, along with a fragment of the fi origin required for ssDNA propagation, wasthen subcloned in the low-copy-number centromeric vector, YCp5O (pLL15). For some unknown reason, thisvector does not direct enough SPT2+ expression to complement the Spr phenotype of strain S704. It was howeverfound that deletion of thefi origin fragment did produce an efficient YCpSPT2 vector for the present studies(pLL32, Fig. 9C). Similarly, the YEpSPT2-1 (pLL18) and YCpSPT2-1 (pLL24) plasmids were constructed byrepeating those subcloning steps with the cloned dominant allele SPT2-i (from plasmid R159). Both of theseconstructs could suppress the }{is phenotype of strain SR26-12C (SPT2 his4-912b) as expected from a dominantsuppressor allele. That this phenotype is caused by the acquisition of a Spr phenotype was confirmed by the abilityof these plasmids to suppress both the his4-912ö and /ys2-l28ömutations of strain FW1237.As demonstrated by the experiments described above, this set of four vectors is suitable for the necessarycomplementation and dominance tests. Most mutations described below were obtained as derivatives of pLL1O(YEpSPT2j or pLL18 (YEpSPT2-1) and some were also subcloned into YCp5O to obtain their low-copy-numberderivatives.3.1.2 Isogemc yeast strainsEarly studies on the construction of the expression vectors described above emphasized an importantaspect of SPT2 function, namely the extent to which the phenotypic changes monitored at 30°C can vary from onegenetic background to another. Although the reasons for these variations are unknown, the elimination of thisadditional parameter was attempted by constructing SPT2 and spt2 strains with the same genetic background(isogenic strains). To achieve this goal, a set of SPT2 his4-912ö strains ofboth mating types (MA Ta or M4 Ta),carrying ura3 and leu2 mutations for vector selection, and a lysi or a frpl mutation for diploid selection wereconstructed (Table 3, page 54, and Fig. 8, page 68). These strains are His at 30°C and become His whentransformed with pLL18 or grown at 37°C, confirming the presence of the suppressible solo insertion at HIS4.Strain SLL5 (MA Ta his4-9125ura3-52 leu2 trpl-l SPT2) was used as the wt reporter strain. The potential toperform one-step gene replacement by homologous recombination in yeast was utilized to obtain an isogenic spt2derivative of SLL5. The feasibility of this approach in a haploid strain like SLL5 was supported by the fact that acomplete deletion of SPT2 has been shown to be viable in haploid cells (245).81A disruption allele was constructed by replacing 810 bp of the SPT2 coding region (81%) with the URA3gene (allele spt2iI:: URA3 on plasmid pLL82; see Fig. 9D, page 79). A linear DNA fragment of spt2ii:: URA3,containing the URA3 marker flanked by approximately 700 bp of SPT2 sequences on either side (2.6 kb longEcoPJ-BamHI fragment of pLL82) was used to transform SLL5 to uracil prototrophy. Several independent Ura+transformants were recovered, streaked on SC-Ura plates to obtain isolated colonies, and tested on dropout platesfor the presence of the expected markers. Since a deletion of SPT2+ constitutes a recessive suppressor allele ofhis4-9126 it was expected that the desired integrants- those in which the SPT2+ gene had been disruptedsuccessfully - would exhibit a Pjs phenotype at 30°C. More than 50% of the Ura+ transformants were also His+.A single Ura+ His+ clone, named SLL7, was subjected to further genetic and molecular characterization.SLL7 cells were crossed to SLL4 (MATcc his4-9128ura3-52 leu2 lysl-1 SPT2j and mating products wereobtained by selection for Lys Trp cells. The resulting diploids (SLL4x7) were Ura and His, consistent withthe recessiveness of the disruption allele. Similarly, the introduction of the wt SPT2 allele on pLL1O in strainSLL7 yielded Leu+ transformants that displayed Ura+ and His phenotypes, confirming that the Spr suppressorphenotype of SLL7 (i.e. the His phenotype) was caused solely by the loss of SPT2 function. The diploid SLL4x7also was sporulated and meiotic products were analyzed by random spore analysis. This experiment confirmed thecosegregation of the Ura and His phenotypes. The haploid strain, SLL1O, was obtained as aMAT(x Ura HisTrp Lys+ segregant of SLL4x7 and was used as a reporter M4 Tc spt2zl:: URA3 strain in crosses (see section3.4.1.2).The substitution of the spt2zi:: URA3 construct for the wt SPT2 gene was confirmed by analysis ofgenomic DNA digests. Genomic DNA was prepared from the isogemc strains SLL5 and SLL7, digested with therestriction enzymes NdeI or X7ioI and BglII, and run on a 1% agarose gel. The agarose gel was dried down and thegenomic digests were analyzed by three successive hybridizations in the dried agarose matrix with differentoligonucleotide probes (Fig. 10). The results were interpreted as follows:(I) probe oLL27 is specific for the N-terminus of the SPT2 coding region, present in both the SPT2 andthe spt24:: URA3 alleles. The oLL27 hybridization patterns showed the presence of a 300-bp insertion at the SPT2locus of SLL7 and confirmed the presence of an extra NdeI site in the inserted fragment;87respectively: SLL5 cells transformed with pLL18 (YEpSPT2-1) are His, demonstrating that high-copy-numberSPT2-1 dominates over the chromosomally expressed Spt phenotype, and SLL7 cells transformed with pLL1O(YEpSPT2j are His, confirming the complementation of the spt2zl:: URA3 allele by the cloned SPT2 gene andthe recessive phenotype of the chromosomal disruption. The same results were obtained in all SPT2+ and spt2strains tested, as well as with low-copy-number (YCp) derivatives of pLL1O and pLL18 (see Fig. 23, page 123).No additional phenotype was associated with the increased dosage of the wt SPT2 allele.3.1.3 Anti-SPT2-peptide antibodiesPolyclonal antibodies were raised against two synthetic peptides designed from non-overlapping regions ofthe predicted C-terminal sequence of the SPT2 gene product: PSi and PS2, spanning amino acids 209 to 238 and297 to 325, respectively (Fig. 3, page 32, and Appendix 2). The presence of specific antibodies in the sera fromfour immunized rabbits was monitored in immunoassays (ELISA) against the purified peptides. Both peptide-KLHconjugates were found to elicit an inunune response and exhibit comparable antigenicity. Anti-peptide antibodieswere immunoaffimty purified from high-titer bleeds on peptide-columns and used in Western immunoblot analysesof total yeast extracts. The polyclonal antibodies raised against each peptide were shown to recognize the SPT2+gene product (see below). The immunoaffinity step was found to be critical in reducing the non-specificbackground masking the low-level expression seen from the chromosomally expressed SPT2 protein. Otherparameters that greatly increased the sensitivity of the immunoblots include the combination of both purified antipeptide antibodies (therefore increasing the number of detected epitopes) and the use of a sensitive enhancedchemiluimnescence detection method.The combined immunoaffinity-purified sera recognize a protein migrating at an apparent molecularweight of 44,000 in Western blot analysis of total yeast protein extracts (Fig. 13). This protein band is identifiedconclusively as being SPT2 on the basis of four observations: (I) it is present in total protein extracts from SLL5(SPT2j but not from SLL7 (spt2&: URA3) (Fig. 13, lanes 3 and 4); (ii) extracts from strain SLL7 transformedwith the SPT2 high-copy-number plasmid pLL1O (YEpSPT2j reveal an overproduction of the 44-kDa band(Fig. 13, lane 2); (iii) the migration of the cross-reacting protein is affected by partial deletions of the SPT2 geneintroduced in vitro (Fig. 13, lane 1, and Fig. 17); (iv) the same protein is detected by the antisera raised against thetwo different peptides when used individually, but not by the preimmune sera (data not shown).86indistinguishable from the homozygous diploids SLL4x5 (SPT2’7SPT2j with regard to the PCR product obtainedand their non-reverting His phenotype.M 1 2 3 4 5 6 7 8 910111213M1.64— 4— spt24::URA31.02— 4—SP120.51 —Figure 12. PCR analysis of SLL4x7 1’js revertants.The PCR primers oLL34 and oLL5O were used as shown in Fig. ii to ampli1y the SPT2 locus from the followingdiploid yeast strains: lanes ito 10, 10 independent His revertants of SLL4x7; lane 11, SLL4x5 (SPT27SPT2j;lane 12, SLL4x7 (SPT2/spt2zi:: URA3); lane 13, SLL1Ox7 (spt2zi:: URA3/spt2/J:: URA3). The products of thePCR reactions were separated on a 1.2% agarose gel and stained with ethidium bromide. The molecular weightmarkers are from the 1 kb ladder (BRL). Fragment sizes are given in kilobases.The studies described above confirmed that the haploid strain SLL7 carries a disruption allele of SPT2behaving as a recessive suppressor of the insertional mutation his4-9126 In diploids homozygous for his4-912ô asingle copy of the wt SPT2+ gene is sufficient and essential to maintain the His phenotype in presence of spt2zJ:: URA3. With respect to the mutational analysis, it was finally important to establish whether the isogenic strains,SLL5 and SLL7, could be utilized to monitor the activity of plasmid-borne alleles. Each strain was transformedwith three plasmids: pEMBLYe3O, as a control, pLL1O, carrying the wt SPT2 allele, and pLL18, carrying thedominant allele SPT2-1. As shown in figure 14 on page 93, the ability of SLL5 and SLL7 transformants to grow inthe absence of histidine provides easily scored dominance and complementation tests for SPT2 activity,85This PCR protocol was used to analyze the heterozygous diploids SLL4x7. As noted previously, thesediploids were created in order to assess the ability of the SPT2 allele of SLL4 to complement the disruption alleleof SLL7 and confirm the recessive suppressor phenotype associated with spt2zl:: URA3. The His phenotype ofSLL4x7 yeast cells supported these predictions. After non-selective growth in complete medium and spreading onSC-His dropout plates, it was also observed that Jfi+ revertants emerge in the cell population (approximately 1revertant in i04 to cells). Since these diploids are homozygous for the insertional allele his4-9125 it wasassumed that the reversion is mediated by the acquisition of a suppressor Spr phenotype. Several mechanismscould be responsible for the emergence of these revertants, such as the loss of the wt SPT2+ function by mutation,mitotic recombination or chromosomal loss, or the creation of a dominant SPT allele by spontaneous mutation inany other SPT gene.In an attempt to identify the possible cause(s) of this reversion and establish whether the reversionphenomenon is related to the complementation of spt24:: URA3 by SPT2, the fate of both spt2 alleles wasfollowed by diagnostic PCR in 10 independent His revertants of SLL4x7, As shown in figure 12, all the Hisrevertants show an identical pattern of PCR products: they have all lost the smaller SPT2+ fragment and retainonly the larger spt2zl:: URA3 PCR product. This pattern is identical to the one seen for SLL1Ox7 homozygousdiploids (spt2.&: URA3/spt2il:: URA3; Fig. 12, lane 13) and could originate from the loss of the wt SPT2 allele bymitotic recombination.If mitotic recombination (reciprocal recombination or gene conversion) is indeed responsible for theobserved reversions, it was reasoned that the reciprocal loss of the spt24::URA3 disruption allele to create diploidshomozygous for the wt allele should also occur. This possibility was tested by taking advantage of a counter-selection for URA3 activity: 5-fluoro-orotic acid (5-FOA) is a pyrimidine analog metabolized to a lethal form byUra+ strains. Since only Ura cells (ura3) can survive on 5-FOA plates, this approach selects for loss of URA 3function. With regard to the experiment on SLL4x7 diploids, the 5-FOA plates impose a selection converse to theone effective on SC-His plates: selection for histidine prototrophy demands loss of SPT2 function in a his4-9]2östrain (i.e. a suppressor Spr phenotype), whereas selection for 5-FOA resistance enriches for loss of URA3 function(i.e. a Ura phenotype). It was observed that, following growth in complete medium, 5-FOA resistant revertants ofSLL4x7 arise at a frequency comparable to that observed for the His+ revertants. PCR analysis confirmed that theresistant cells had lost the disruption allele spt2zl:: URA3 (not shown). Those revertants were found to be844— spI2&:URA34— SPT2Figure 11. Analytical PCR amplification of the SPT2 and spt24:: URA3 alleles.(A) Schematic representations of the spt2zl:: URA 3, SPT2, and CYCJ genes, showing the priming sites for thePCR oligonucleotides and the size of the expected PCR products. (B) and (C) Ethidium bromide-stained agarosegels of analytical PCR products separated by electrophoresis. (B) SPT2 (oLL5O + 0LL34) and CYC] (CYH8 +CYH7) amplifications of plasmid and genomic DNA. The template DNAs used for the PCR reactions were asfollows: lane 1, pLL1O (YEpSPT2j; 2, pING4 (YEpCYCJ); lanes 3 to 6, genomic DNA from strains S49 (CYC]SPT2j, S704 (CYC] spt2-150), SLL4, and SLL5, respectively. (C) spt2zl:: URA3 and SPT2 amplifications(oLL5O + oLL34) on the following DNA samples: lane 1, pLL1O; 2, pLL82 (YEpspt2zi::URA3); 3, pLL1O +pLL82; lanes 4 to 7, crude genomic DNA samples from strains SLL5 (SPT2j, SLL7 (spt2zl:: URA3), SLL4(SPT2j, and SLL4x7 (SFT2/spt2,i:: URA3). The molecular weight markers are from a X174IHaeIII digest(Pharmacia) and the fragment sizes are given in number of base pairs.oLL5Ospt2&:URA3SPT2oL5OL341420 bp1056 bp452 bpABCoLL34CYH8cyci --7Ml 2345 6M4— SP724— crc)83(ii) probe oLL5 1 is specific for a C-tenrnnal region of the SPT2 open reading frame deleted in spt2zi:: URA3. Only SLL5 DNA was found to hybridize to oLL5 1, consistent with the loss of SPT2 sequences inchromosome V of strain SLL7;(iii) The CYH-26 probe (provided by Chi-Yip Ho) is specific for the URA3 gene. In addition to thehybridizing fragments from the mutant ura3-52 allele present on the left arm of chromosome V of SLL5 and SLL7,strain SLL7 contains another URA 3-specific fragment, that co-migrates with the SPT2 fragments detected byoLL27 in SLL7. Taken together, these data confirmed the physical replacement of the wt SPT2 allele of SLL5 bythe URA3 disruption construct spt2&: URA3 in SLL7.Having established the presence of a unique URA3 insertion site and the linkage between the UR.43marker and the SPT2 locus in strain SLL7, a rapid diagnostic PCR protocol was developed for the analysis of theSPT2 locus in wt and mutant strains. This tool was developed primarily for the analysis of SLL4x7 diploids andtheir His+ revertants (see below), and for the characterization of mutant strains obtained by gene replacement inSLL7 (see section 3.4.1). As shown in figure 11, the oligonucleotides, oLL34 and oLL5O, constitute an effectivepair of primers for the amplification of both the SPT2 and the spt2ii:: URA3 alleles from crude, whole-cellextracts. The usefulness of this rapid approach lies in its ability to distinguish the alleles on the basis of the size ofthe amplified products: the SPT2 gene gives a 1056-bp PCR product whereas the deletion-insertion allele, spt2zl:: URA3, gives a longer product of 1420 bp.Using CYCJ -specific primers as internal controls, it was established first that oLL34 and oLL5O can beused to amplify the wt allele from plasmid (pLL1O) or genomic DNA (strains S49, SLL4 and SLL5), and that thisamplification is specific to the SPT2 gene: DNA from strain S704, carrying a complete deletion of SPT2, allows forthe amplification of the CYC] gene, but not of SPT2 (Fig. 1 1B, lane 4). Then, it was demonstrated that the sameprimers can be used to amplify and distinguish the wt and disruption allele, spt2iI:: URA3, from plasmid orgenomic DNA (Fig. 1 1C, lanes 1, 2, 4, 5, and 6). It was found also that both alleles can be amplifiedsimultaneously in plasmid mixtures or in crude extracts from heterozygous diploids, such as SLL4x7 (Fig. 1 1C,lanes 3 and7).BXN H PI I ISLL5 ::: ‘4 I I’Tyl 4SPT2XNSLL7 :::____Tyl7.13 —5.09 —4.07 —3.05 —1.7kb4.15kbI kbFigure 10. Analysis of SPT2 locus from strains SLL5 and SLL7.(A) Structure and restriction map of the SPT2 locus from the isogenic strains SLL5 (SPT2j andSLL7 (spt2zl:: URA3). Also shown are the locations of the sequences from which the oligonucleotide probesoLL27, oLL5 1 and CYH-26 were designed, as well as the size of the expected NdeI and BglJI+XhoI fragments.(B) Autoradiographs of three successive hybridizations in the dried agarose gel using the indicated [32P] end-labeled oligonucleotides. Restriction enzymes are abbreviated as described in the legend of Fig. 9. The molecularweight markers are from the 1 kb ladder (BRL), and the fragment sizes are given in kilobases,A oLLS1U U82N GI I2.46kb3.85kbCYH-26 oLL27.HNI I_UH NURA3GN X+G5757N X÷GM5 757— —2.04— —1.M—N X÷G5757oLL27 oLL51 CYH-26+ +2t X1234-43.0— -— —+2 X1234Figure 13. Immunological detection of the SPT2 protein.Coomassie blue-stained gel (A) and Western blot analysis (B) of total protein extracts resolved by electrophoresison a SDS-10% polyacrylamide gel. Yeast extracts: lane 1, SLL7 (spt2ii:: URA3)/pLL75 (YEpspt2zI3); lane 2,SLL7/pLL1O (YEpSPT2j; lane 3, SLL5 (SPT2j; lane 4, SLL7. For the immunoblot analysis, the electroblottedproteins were reacted with immunoaffinity-purified anti-SPT2 peptide antisera and cross-reacting antibodies werevisualized by the enhanced chemiluminescence dectection system. The arrow indicates the position of the SPT2protein. Designations: 2.t, high-copy-number (YEp) transformants; X, chromosomal expression.An additional cross-reacting band with an apparent molecular weight of 31,000 was also observed inYEpSPT2 transformants (Fig. 13B, lane 2). This polypeptide is probably a derivative of SPT2 protein since itspresence seems dependent on the overproduction of SPT2 and its concentration varies considerably from oneprotein extract to another (e.g., see Fig. 17B, lane 1). It contains the C-terminal epitopes recognized by thepolyclonal antibodies and thus could originate from translational initiation at an internal methionine (Met 125) orfrom N-tenninal proteolysis of the SPT2 protein. This hypothesis is also supported by the observation that themigration of this cross-reacting polypeptide is affected by the deletion of 9 carboxy-terminal residues in alleleSPT2-324 (see Fig. 16, page 100, lane 2).A88kDa— 200BkDa200 —97.4—68.0 —43.0—29.0 ——97.4—68.089It should be noted at this point that, although the anti-peptide sera detect a protein of ca 44,000, thepredicted molecular weight of the 333-amino-acid SPT2 protein is 36,000 Da. A possible explanation for thediscrepancy in molecular weight is that there is read-through of the 56 in-frame codons that follow the stop codonat position 334 of the SPT2 gene product (180). However, as shown below (section 3.2, allele SPT2-324), theintroduction of a nonsense mutation at position 325 of the SPT2+ coding region in vitro produces a mobility shiftin the resulting mutant polypeptide consistent with the termination of translation at codon 334 (Fig. 16B, 17B and24C). Hence, the discrepancy between the apparent molecular weight of the SPT2 protein on SDS-polyacrylamidegels and the molecular weight calculated from the primary sequence of the SPT2+ gene is probably a consequenceof posttranslational modifications of the nuclear protein or of its high content of charged residues.The results presented in figure 13 also point to an important aspect of SPT2 function. It was observed thatthe expression of the SPT2+ gene from a high-copy-number vector causes a significant overproduction of the SPT2protein. Densitometiy analysis estimated a 5- to 8-fold increase in SPT2 protein levels compared to those achievedby expression from the single-copy chromosomal gene, without considering the contribution of the 3 1-kDa species.Such levels of overproduction are comparable to those reported for most proteins encoded by genes maintained inyeast on 2-.tm-based vectors, which are overproduced approximately 10-fold (179).With regard to the suppression of the ö insertion his4-912d it was found that strains transformed with theoverexpressing YEpSPT2+ construct are functional for SPT2 activity and exhibit a non-suppressing Spt+phenotype (e.g., see Fig. 14). It was discussed previously how the suppression of the transcriptional defects causedby Ty and ö insertional mutations can be mediated by mutations in one of several unlinked SPT genes (245).Furthermore, particular class II SPT genes exhibit a dosage-dependent suppression phenotype: alteration of the wtSPT5 (221), SPT6 (43, 161) or SPTJ 1/HTA 1 plus SPTJ2/HTBI (42) gene dosage causes suppression ofinsertional mutations at the HJS4 and LYS2 loci. These observations are consistent with the involvement of therespective gene products in complexes where a particular stoichiometry of the components is critical for function.In the case of the HTA 1-HTB1 locus, encoding histones H2A and H2B, it was concluded that an imbalance in therelative levels of histone dimers causes transcriptional changes leading to the suppression of the insertionalmutation (42). The results presented here show that the SPT2 protein is not subject to the same restrictions andthat S insertion alleles are unaffected by a nearly ten-fold range of SPT2 concentrations. While these observationsargue against SPT2 being part of a complex sensitive to SPT2 protein levels (such as proposed for SPT5 and90SPT6), they do not eliminate the possibility that, as in the case of histones H2A and H2B, a Spt phenotypeproduced by increased SPT2+ gene dosage may require simultaneous overexpression of all the members of acomplex. The overexpression of another HMG1-like nonhistone nuclear protein of yeast, encoded by the NHP6Bgene (125), was shown to be lethal (139).The results presented above convincingly demonstrated that the polyclonal antibodies raised against twodifferent peptides designed from the predicted amino acid sequence of the SPT2 gene cross-react with the SPT2gene product on Western immunoblots. Unfortunately, the immunoaffinity-purified sera did not bind to the nativeSPT2 protein with detectable affinity in immunoprecipitation or indirect immunofluorescence experiments. Thesenegative results suggest that in the native (folded) protein, the antigenic determinants are buried or inaccessible tothe antibodies. Another possibility is that the epitope, although present on the surface of SPT2, is structurallyrestrained in a non-reactive conformation or masked by a ligand.In any case, the sera obtained constitute a highly efficient probe to monitor the expression of mutant andwt SPT2 proteins in yeast. Of course, as is the case for monoclonal antibodies, their usage is restricted to thedetection of polypeptides bearing the small antigenic determinants. Notwithstanding these limitations, it should beemphasized that an important advantage of these anti-peptide antibodies over polyclonal antibodies raised againstthe whole purified protein, for example, is that the epitopes recognized are present in small, defined regions of thetarget protein. This characteristic implies that large portions of the protein under study can be mutated, or evendeleted, without affecting the stoichiometry of the protein-antibody complexes (and therefore the intensity of thesignals on Western blots) for drastically different mutant derivatives of SPT2 carrying the epitope(s) recognized.Consequently, the anti-peptide antisera (anti-PS 1 and anti-PS2 Ab) can be utilized to estimate the relative steadystate levels of various SPT2 variants, as will be appreciated from the mutational analysis presented below.913.2 Mutational analysis of the SPT2 proteinThe predicted translational product of the wt SPT2 gene is a highly charged protein of 333 amino acids.It was found that the dominant suppressor allele, SPT2-1, is caused by a T to A transversion at nucleotide position638 of the coding region (180). This substitution changes the leucine codon (TTA) at amino acid position 213 ofSPT2 to a nonsense ochre codon (TAA). The SPT2-1 allele is therefore thought to encode a nonfunctionaltruncated polypeptide consisting of the first 212 amino acids of SPT2 (see Fig. 3, page 32). Since the suppressionof his4-912ö is caused by the loss of SPT2 function (such as in the strain SLL7, carrying the disruption allele spt2zi:: URA3) the mutational analysis of SPT2 was guided primarily by the following questions: (I) what are thestructural elements, missing from the truncated product of the SPT2-] allele, that are necessary for wt SPT2function and maintenance of the Spt+ phenotype? (ii) what is the structural basis for the dominance of SPT2suppressor mutants? The next three sections present the mutational analysis carried out on the SPT2 gene toaddress these questions. Section 3.2.4 consolidates the conclusions of those studies with in vivo data on theexpression of some mutant polypeptides.3.2.1 Dominant carboxy-terminal deletionsThe truncated polypeptide presumably encoded by the SPT2-1 mutant allele lacks several distinctivestructural motifs present in the C-terminal third of SPT2 protein (Fig. 3, page 32). The first goal of this analysiswas to identify elements of this region of the protein, absent in SPT2-], that are necessary for wt function andmaintenance of the Spt+ phenotype. Using oligonucleotide-directed mutagenesis, nonsense mutations wereintroduced in the cloned SPT2 gene carried on pLL1O (YEpSPT2j in order to create mutant alleles encodingtruncated polypeptides of 271, 303, 318, and 324 amino acids. These mutations were designed to incorporateprogressively the features of the primary sequence of the C-terminus of SPT2 protein missing in the nonfunctionalSPT2-1 polypeptide. The mutagenesis reactions were based on the extension of mutagenic primers (18 to 22nucleotides long and introducing 2 or 3 point mutations) on uracil-containing single-stranded pLL1O templates.Positive clones were identified by DNA sequencing. The expected mutant alleles were obtained at efficiencies92ranging from 22 to 66%. For each allele, the entire coding region was sequenced to confirm the absence ofsecondary mutations.The mutant alleles were introduced into the isogenic his4-912öSPT2 and spt2zi:: UR.43 strains, SLL5and SLL7, and Leu+ transformants were tested for their growth in the absence of histidine at 30°C (Fig. 14). Noneof the nonsense mutations were able to complement spt2zi:: URA3 and each mutation retained the strong dominantcharacter exhibited by the SPT2-1 allele when transformed into SLL5. In order to confirm that the suppressorphenotype of SLL5 transformants is dependent on the plasmid-borne alleles, plasmid segregation studies wereperformed. The transformants were grown for several generations in complete medium, which does not select forthe maintenance of the pEMBLYe3O derivatives. In these conditions, such YEp plasmids - dependent on the 2imorigin for replication- are lost at a rate of approximately 1% per generation (26). Isolated colonies were obtainedon complete medium and tested for their ability to grow in absence of leucine or histidine (on SC-Leu or SC-Hisplates). In all cases, Leu SLL5 cells that had been cured of the plasmid vector regained the expected Hisphenotype (not shown). These results confirm that the suppressor phenotype observed in SLL5 transformants isdependent on the presence of the plasmid-borne alleles, which behave as dominant SPT alleles.3.2.2 Carboxy-tenrnnal replacementsAs seen from the smallest C-terminal deletion (allele SPT2-324, Fig. 14), residues at the very end of theSPT2 protein are critical in determining its effects on the phenotype of the 6 insertion allele. Nine amino acidresidues of SPT2 are absent from the truncated polypeptide product of the dominant allele SPT2-324: Lys 3 25-Arg-Arg-Arg-Lys-Lys-Gly-Ile-Arg 333. The importance of the six consecutive basic residues was demonstrated byreplacement of Lys 325 to Lys 330 using a cassette-mutagenesis approach in a pseudo-wt allele with aphenotypically silent terminal valine insertion (see alleles spt2-334 and SPT2-335, Table 5). These multiplereplacements generated a non-complementing and strongly dominant suppressor allele.93SPT2 +SPT2-1SPT2-2 71SPT2-303SPT2-318SPT2 -324Suppression of his4-912 6++++++++++Figure 14. C-terminal truncations of SPT2 creating nonfunctional, dominant polypeptides.Nonsense mutations were introduced in the cloned SPT2+ gene to generate alleles encoding truncated proteins of271, 303, 318, and 324 amino acid (aa) residues. Derivatives of the high-copy-number vector pEMBLYe3O,carrying the LEU2 gene as a selection marker and a spt2 allele (left), were used to transform the isogenic strainsSLL5 (his4-9J2öSPT2 Ieu2) and SLL7 (his4-9128spt2&: URA3 leu2) to leucine prototrophy. For each allele,three independent Leu+ transformants were monitored for their ability to grow on SC-Leu-His plates (right).Symbols: +, His+; -, His at 30°C. The central panel provides a schematic representation of the expectedpolypeptide products.SPT2 derivative 100 200I I i INONESLL5300 (aa) (SPT2 +)SLL7(spt2 zi ::URA3)+F II II II I94A finer mutational analysis was performed through systematic serine replacements at each of the six basicresidues. Experimentally, the introduction of each amino acid replacement in the SPT2 protein involved themutation of one or three nucleotides in the SPT2+ gene by oligonucleotide-directed mutagenesis, followed by thesequencing of the whole coding region. Using primers of 23 to 25 nucleotides, efficiencies of mutagenesis rangingfrom 12.5 to 50% were obtained. Four of these point substitutions can complement the spt2/i:: URA3 null allele ofSLL7 and show a wt Spt phenotype (Table 5, alleles spt2-326 to spt2-328, spt2-330). The K329S mutant (allelespt2-329) is only partially functional in SLL7 (weak His phenotype) and correspondingly is semi-dominant whenassayed in SLL5. On the other hand, the spt2-325 allele carrying the K325S replacement shows a stronger mutantphenotype associated with a more pronounced dominant behavior. The conservative change K325R (allele spt2-336) has no apparent effect on SPT2 protein function. The double amino acid replacements, K325S K330S andK325S K329S, lead to dominant Spr phenotypes indistinguishable from the C-terminal deletant SPT2-324.95Table 5. Carboxy-terminal spt2 mutationsaProtein Suppression of his4 -9128 CSPT2 derivative‘ b SLL5 SLL7Designation equence (SPT2 +) (spt2 zl ::URA3)NONE — +SPT2 vt EKRRRKKGIR - -SPT2-324 1-324 E + +spt2-334 V334 EKRRRKKGIRV - -SPT2-335 cassette EMILLMMGIRV + +spt2-330 K330S EKRRRKSGIR — —spt2-329 K329S EKRRRSKGIR +1.— +1—spt2-328 R328S EKRRSKKGIR — —spt2-327 R327S EKRSRKKGIR — —spt2-326 R326S EKSRRKKGIR — —spt2-325 K325S ESRRRKKGIR +1— +1—spt2 -336 K325R ERRRRKKGIR — —SPT2-350 K325S K330S ESRRRKSGIR + +SPT2-359 K325S K329S ESRRRSKGIR + +a Single and multiple missense mutations were introduced in the wild-type SPT2+ gene and assayed forcomplementation of spt2z::URA3 and dominance over SPT2 as described in the legend of Fig.14.b Structure of the cm-boxy terminus from residue E324. Amino acid insertions and replacements are underlined.The replacements performed in SPT2 -335 by cassette mutagenesis are K325M, R3261. R327L, R328L, K329M,and K330M, introduced in the pseudo-wild-type allele spt2 -334.C Symbols: +, [+; -, His; +/-, partial growth on SC-Leu-His at 30°C. spt2 -325 exhibits a stronger suppressorphenotype than does spt2 -329 (see text).96Within the C-terminus of SPT2 protein, six consecutive basic residues were mutated to serine. MutationsK325S and K329S exhibited partial loss of his4-912ó repression by SPT2. These results established that pointsubstitutions in the carboxy-tenninal region of SPT2 can mimic the effects of carboxy-terminal deletions removing9 (SPT2-324 allele) to 121 amino acids (SPT2-1 allele) and suggested that it is not the net positive charge of thisdomain, but rather the presence of positively charged residues at specific positions that is critical for SPT2function. Thus, the mutation K325R conserves the positive charge at position 325 of SPT2 and is phenotypicallysilent. This strict requirement for positively charged residues at specific terminal positions in repression isreminiscent of the role of N-terminal amino acids of hi stone H4 in the repression of the silent mating loci, HMLcLand HMRa (121, 171a). Mutational analysis of histone H4 has shown that, as in the case of SPT2 protein, a smallterminal deletion - in this case of amino acids 4 to 19 - results in loss of repression and expression of the normallysilent mating loci (17 la). Within this region, four out of eight positively charged residues - including K16 andH18- that are known to undergo reversible modification (acetylation and phosphorylation respectively) arerequired for repression. As shown for position 325 of SPT2, elimination of the positive charge at position 16 ofhistone H4 causes derepression: the substitution K16G, but not K16R, leads to expression of HMLc and H./vlRa.Acetylation of K16 could neutralize its positive charge and this residue is thought to be deacetylated at HMLcL andHMRa (121). Similarly, the positranslational modification of K325 and/or K329 of SPT2 could regulate therepressor activity of SPT2 by neutralization of those critical positive charges. Therefore, such a mechanism ofderepression could play a role in the activation of transcription brought about by SPT2/SIN1 antagonists, such asSWI1, SWI2/SNF2 and SWI3 (174).The results presented above demonstrate the importance of specific C-terminal basic residues in thenegative regulation of his4-9]2öby the SPT2 protein and furthermore establish a correlation between the severityof the effects of C-terminal mutations and their potential to dominate over the Spt+ phenotype3.2.3 Definition of a dominance domainThe dominant phenotype of mutants at the C-terminus of the SPT2 protein suggests that there are at leasttwo regions of the protein required for normal function: the C-terminal basic sequence between residues 325 and333, and a positively acting “dominance domain” responsible for the ability of C-terminal mutants to interfere withthe maintenance of the Spt+ phenotype. This domain must lie within the 212-amino acid sequence of the SPT2-]97protein. As shown in Figure 3 on page 32, a region of similarity with the DNA-binding, helix-turn-helix motif ofprokaryotic repressors and homeo-domain proteins is present in the N-terminus of SPT2-1 protein (residues 60 to81, ref. 180). Since such a functional motif constitutes a potential candidate for the dominance domain, a completein-frame deletion of these residues (referred to as LI]) was generated in the wt SPT2+ gene and in the dominantsuppressor allele, SPT2-]. This deletion was accomplished by looping out the SPT2 sequences coding for residues60 to 81 using a single oligonucleotide, hybridizing to 16 nucleotides on either side of the deletion, in anoligonucleotide-directed mutagenesis reaction. In contrast to what could have been expected for the introduction ofsuch a drastic mutation, the efficiency of this reaction was rather high; more than 50% positive deletants wereobtained in each case.An examination of the phenotypes of SLL5 and SLL7 transfonnants disproved the hypothesis that aminoacids 60 to 81 play a role in dominance. These residues are not required for the dominance of SPT2-1, nor are theynecessary for the wt SPT2+ function, at least as far as the Spt phenotype is concerned (Fig. 15, alleles spt2ziI andSPT2-lzll).In order to localize the dominance domain, further truncations of the SPT2-1 protein were produced tocreate C-terminal deletion mutants containing 179, 116, and 46 N-terminal amino acids. Alleles SPT2-] 79 andspt2-116 were created by oligonucleotide-directed mutagenesis as described above for the other nonsensemutations. Allele spt2-46 was obtained by deletion of the internal PstI-HindIII fragment of SPT2. The 179-amino-acid product of allele SPT2-] 79 still confers a strong dominant suppressor phenotype, but deletionsremoving residues 47 to 179 or only 117 to 179 create recessive spt2 alleles (Fig. 15, alleles SPT2-] 79, spt2-1 16and spt2-46). It should be noted at this point that, although the spt2-1 16 and spt2-46 mutant polypeptides appearto have lost the ability to interfere with the maintenance of the Spt+ phenotype, it is not known if they are producedin a stable form in vivo. Since they do not include the C-terminal half of the SPT2 protein, these mutant productsdo not contain the epitopes recognized by the anti-SPT2 antisera and could not be studied further with theseantibodies.These problems were circumvented by asking whether a similar loss of dominance could be produced inC-terminal mutants by internal deletions. The amino acids removed from the truncated SPT2-179 protein in spt2-46 (residues 47 to 179) or in spt2-1 16 (residues 117 to 179) were deleted in the wt gene and in dominant alleleswith altered C-termini. The former deletion was achieved by PCR mutagenesis, the latter by oligonucleotide98directed mutagenesis. For this reaction, which involves the deletion or looping out of 189 nucleotides (63 codons)in the cloned alleles, a mutagenic primer hybridizing to 25 nucleotides on either side of the deletion was utilized.The expected deletion was identified in 18 to 25% of the clones analyzed.Suppression of his4-9123SLL5300 (aa) (SPT2 +)I_____________________________spt2zl 1SPT2-JA 1I II II I + +SPT2-1 79spt2-116 I Ispt2-46 I I+ +++spt2A2 I Ispt2-324 A 2 Ispt2A3 I Ispt2-324A3 I Ispt2-350A3 I I+++++Figure 15. Residues 117 to 179 of SPT2 are essential for the dominance of truncated products.C-terminal and internal deletion alleles of SPT2+ were constructed by in vitro mutagenesis and assayed forcomplementation and dominance as described for Fig. 14. Symbols: ill, in-frame deletion of residues 60 to 81; 42,deletion of residues 48 to 179; 43, deletion of residues 117 to 179. The double missense mutation of allele SPT2-350 (K325S K330S) is represented by the dotted box. an, amino acids.SPT2 derivativeNONE100 200I ISLL7(spt2 A ::URA3)+I II II II II ILI99The results demonstrate that deletions of amino acids 48 to 179 (z12) or of amino acids 117 to 179 (43)have the same effects: they revert the dominance caused by the C-terminal alterations of the SPT2 protein and inthemselves create recessive suppressor alleles (see Fig. 15, 42 and 43 deletions). Thus, residues 117 to 179 arenecessary for the dominant phenotype of SPT2 alleles with altered C-termini and define a region of the SPT2protein essential for wt function, referred to as the dominance domain (in a functional, not a structural sense).3.2.4 Expression of mutant polypeptides in vivoAs a general rule, it is expected that mutant polypeptides that constitute functional products (pseudo-wtSpt+ phenotype in a spt2 background) or nonfunctional dominant products (suppressor Spr phenotype in a SPT2background) must be produced in a stable form in vivo on the basis of their phenotypic effects. These predictionswere confirmed for all the point substitution mutations that code for pseudo-wt or dominant products (presented inTable 5, on page 95).Total protein extracts were prepared from SLL7 transformants, and the proteins were separated on a SDSpolyacrylamide gel and immunoblotted using the anti-SPT2 antisera (Fig. 16). Controls included extracts fromSLL7 transformed with the parental vector pEMBLYe3O as a negative control (lane 4, labeled “none”) and fromSLL5 to show the low levels of chromosomal expression of SPT2 (lane 13). The results show that in general, thesteady state levels of the mutant polypeptides are comparable to that of the wt allele (produced by pLL1O, lane 1).The alleles containing the C-terminal valine codon (spt2-334 and SPT2-335, lanes 3 and 14) lead to the productionof lower levels of proteins. Nevertheless, all the mutant proteins are produced in quantities exceeding by far thelevels of SPT2 obtained from the chromosomal locus, which are sufficient for function. Similarly, the partial ortotal loss of function of certain single and double serine substitutions respectively does not correlate with specificalterations in the steady state levels.100Figure 16. Production of mutant proteins with altered C-termini in vivo.(A) Coomassie blue-stained gel and (B) Western iinmunoblot analysis of total yeast protein extracts, as describedfor Fig. 13. Yeast extracts are all from strain SLL7 (spt2i±: URA3) transformed with the indicated alleles onpEMBLYe3O derivatives, except lane 13, which contains extracts from strain SLL5 (SPT2j as a controls forchromosomal expression of SPT2 (X). Refer to Table 5 for the structure and properties of the mutant allelesanalyzed here.A+I4.kDa Ml 2 3 4 5 6 7 8 9101112131497.4—68.0—43.0—29.0—B+I I0kDa 1413121110 9 8 7 6 5 4 3 2 197.4—43.0—’29.0—-101Since the complete deletion of SPT2+ (therefore absence of its product) is recessive, any mutation thatdecreases the production or stability of the mutant protein has the potential to generate a recessive Spr phenotype.That the deletion of the internal dominance domain does not revert the dominance of C-terminal mutationsthrough this mechanism was demonstrated by monitoring the intracellular levels of wt and mutant proteins byWestern blot analysis of total yeast protein extracts (Fig. 17). The internal deletions of 63 (A3) and 132 residues (2) in the wt protein do not significantly affect the production of the mutant polypeptides (Fig. 17, lanes 1, 3 and6). Similarly, the reversion of the dominance of the SPT2-324 protein by those deletions is not accompanied byaltered protein levels (lanes 2, 5, and 7). The deletion of residues 117 to 179 (t3) in the double point substitutionmutant polypeptide SPT2-350 slightly reduces the levels of the product (lane 4). However, as for all the otherproducts analyzed here, it is produced in much larger quantities than the wt chromosomal gene product present inSLL5, the strain used for the dominance tests (Fig. 17B and C, lanes 8).These results clearly demonstrate that the reversion of the dominance of SPT2 alleles by internal deletionscannot be attributed to impaired production or stability of the resulting polypeptides. It can therefore be concludedthat the deletion of residues 117 to 179 produces polypeptides that have lost the potential to interfere with themaintenance of the Spt+ phenotype, even when over-produced in the wt background.3.2.5 Subdomains of the SPT2 proteinGenetic analyses of spt2 suppressor mutations showed that the suppressor phenotype could be recessive ordominant to the wt function (211, 245). The observation that a spt2 null allele confers a recessive suppressorphenotype implies that suppression is caused by the loss of SPT2 activity and that the wt SPT2 protein plays anessential role in the transcriptional interference mediated by the ö insertional mutation. A dominant SPT2suppressor allele is capable of interfering with the maintenance of the Spt+ phenotype102+x123456789Figure 17. Production of recessive and dominant mutant proteins in vivo.(A) Coomassie blue-stained gel, (B) and (C) Western immunoblot analyses (15-s [B] and 150-s [C] exposure) oftotal yeast protein extracts, as described for Fig. 13. Yeast extracts: lanes ito 7, strain SLL7 (spt2zi:: URA3)transformed with the indicated alleles on pEMBLYe3O derivatives; lanes 8 and 9, strain SLL5 (SPT2j and SLL7respectively. (X, chromosomal expression). The arrow indicates the position of the SPT2 protein.BA ::‘ci+k —xkDa M1234 5678997.4—68.0—43.0—29.0—C+x89- ——103The existence of two different Spt2 suppressor phenotypes, recessive and dominant, provided the basis foraddressing two related questions regarding the transcriptional effects mediated by SPT2: (I) what are the structuralelements, missing from the truncated product of the dominant SPT2-1 allele, that are necessary for themaintenance of the Spt+ phenotype? (ii) what is the structural basis for the dominance of SPT2 suppressormutants? The mutational analysis of the SPT2 gene has identified the two regions of its product containing thesestructural elements- a carboxy-terminal polar domain and an internal dominance domain- and defined threefunctionally distinct regions of the protein (see Fig. 18).Region I: dispensable N-terminus. The spt2-1 16 allele, which codes for the first 116 amino acids ofSPT2 protein, behaves as a recessive suppressor. This result shows that region I of SPT2 protein, comprised ofresidues 1 to 116, does not contain structural determinants sufficient for repression nor for interference with the wtphenotype. That a substantial part of this region (5 8%) is dispensable for SPT2 activity was shown by theobservation that the deletion of residues 60 to 81 is silent in both wt and dominant products, as well as by previousstudies demonstrating that residues 2 to 51 are not required for SJN1ISPT2 activity (127).Region II: dominance domain. The finding that the dominant SPT2-1 allele is caused by theintroduction of a nonsense mutation at amino acid position 213 and that this allele is transcribed at nearly wt levels(180) suggests that the production of a truncated SPT2 protein interfering with the wt Spt+ phenotype constitutesthe basis for the dominance. Three observations confinning that a C-terminal truncation of SPT2 is sufficient forthe dominant mutant phenotype are presented: (i) the SPT2-1 polypeptide is not detected by the antisera andtherefore does not contain the C-terminal epitopes covered by the peptide antigens (see section 3.4.2, and Fig. 24C,page 126); (ii) the introduction of nonsense mutations at codons 180, 273, 304, 319 and 325 of the wt SPT2 genein vitro produces non-complementing dominant alleles (Fig. 14 and 15); (iii) the dominant allele, SPT2-324,produced by the introduction of a stop codon at position 325, encodes a mutant protein detected by the anti-peptideantisera and migrating with a slightly higher mobility than the wt protein in SDS-PAGE analysis (Fig. 16 and 17).10450I DDEGGLGIRF1002KRSIGASHAP LKPVVRjKKPE PIKKMSFEEL MKQAENNEKQ PPKVKSSEPVA 150spt2 -116 JJ—U——.——.TKERPHFNKP GFKSSKRPQK KASPGATLI4G VSSGGNSIKS SDSPKPVKLNA 200SPT2 -1 79LPTNGFAQPN RRLKEKLESR KQKSRYQDDY DEEDNDMDDF IEDDEDEGYH250IIISKSKHSNGPG YDRDEIWAMF NRGKKRSEYD YDELEDDDE ANEMEILEEE300EMARKMARLE DKREEAWLKK HEEKRRRKK GIA 333SPT2 -324Figure 18. Primary structure of the SPT2 protein.The predicted amino acid sequence of the SPT2+ gene product is divided here in three functional regions: region I,residues ito 116; region II, residues 117 to 179; and region III, residues 180 to 333. Shaded boxes highlight tworegions dispensable for SPT2 activity, and open boxes identify two functional determinants mapped by themutational analysis: the dominance domain (amino acids 117 to 179) and C-terminal residues (amino acids 324 to333). Thick lines cover the two HMG-box-like motifs as presented in figure 19. The C-terminal extension ofBMG box 2 is shown by the broken line (see text). Thin lines underline the two acidic sequences of region III.Arrows point to the carboxy-terminal residues of the products of three nonsense mutations studied here: therecessive allele, spt2-116, and the dominant alleles, SPT2-1 79 and SPT2-324.105It was shown here that the ability of the products of dominant SPT2 suppressor alleles to interfere with thenegative regulation of the insertional mutation his4-912ö requires a functional domain lost by the deletion ofresidues 117 to 179. This functional region of SPT2 protein is referred to as the dominance domain (region II, Fig.18). The data do not however define the N-terminal boundary of the smallest dominant polypeptide and cannotreject the possibility that part of region I is also involved in the structure of a competing polypeptide. The productsof recessive alleles carrying a deletion of the dominance domain, although present at high levels in vivo, have lostthe ability to interfere with the wt Spt phenotype.It was noticed that the position of the dominance domain in the SPT2 protein (amino acid residues 117 to179) overlaps with the sequences showing similarity with the HMG1-like proteins (127) and more specifically withthe DNA-binding domain termed the HMG-box (116). This motif defines a DNA-binding domain found on itsown or as a repeated unit in several regulatory proteins (HMG-box-containing factors) and abundant nuclearfactors (HMG1/2-like nonhistone chromosomal proteins; reviewed in reference 163).The SPT2 protein contains two regions showing sequence similarity to this HMG-box motif: amino acidresidues from positions 26 to 88 (boxi) and 98 to 159 (box2)(Fig. 19), The highest degree of primary sequencesimilarity was found with members of the HMG1/2-like protein family (163), although the weak similarity betweenSPT2 protein and hUBF has been reported previously (116). The Monte Carlo feature of the COMPARE program(159) was used to assess the statistical significance of the sequence similarities between the putative 11MG-boxes ofSPT2 protein and a group of 15 HMG-boxes previously identified (163). Several pairwise alignments were shownto be statistically significant, with actual alignment scores ranging from 3.8 to 4.3 standard deviations away fromthe mean score of 150 randomizations. The most significant similarities were seen between SPT2 boxi. and the C-terminal box of I{MG1, box2 (aligmnent score of 3.8 SD, with 32% of sequence similarity), as well as betweenSPT2 box2 and the N-terminal box of HMG1, boxi (3.9 SD, 34% similarity). At least 50% of the highlyconserved residues found in HMG 1/2-like proteins are present in each of the two putative 11MG-boxes of SPT2(163). Although the sequence alignment presented in figure 19 shows that the BMG-box2 of SPT2 coverspositions 98 to 159, the similarity between this region of SPT2 and 11MG-box-containing transcription factorsinvolved in sex determination extends to amino acid 170 (see Fig. 18 and ref. 163). The motif PXYK (where Xrepresents any amino acid) is conserved between SPT2 (positions 160 to 164) and most of these proteins (SRY, al,106IRE-ABP, Mata-1, Mc, stel 1; ref. 163), and the 11MG-box of the mating type protein Mc of S. pombe is 73%similar to SPT2 in this region (8/11 conserved residues; ref. 123).It is clear that the BMG-boxes of SPT2 are more divergent than other members of this family, but thestructural basis for the DNA-binding activity of this protein domain is still unknown. The fact that the bindingcharacteristics of previously studied HMG-box-containing proteins range from sequence-specific (219, 226) to nonspecific, and even structure-specific (18, 72) suggests that subtle structural differences can dictate the recognitionproperties of this domain.In view of the mutational analysis of SPT2, the sequence similarity raises the possibility that the role ofamino acids 117 to 179 (dominance domain, Fig. 18) in the transcriptional repression exerted by the SPT2 proteinas well as in the dominance of truncated polypeptides could be mediated by protein-DNA interactions through anintact BMG-box2. Recent studies on nuclear proteins containing one or several HMG-boxes suggest that a generalproperty of this DNA-binding domain is the interaction with structurally distorted DNA (72). 11MG 1 and otherHMG-box proteins have been shown to bind to cruciform DNA, bent DNA, supercoiled DNA, or DNA modified bythe antitumor drug cisplatin (18, 32, 52, 72). Several HMG-box-containing proteins have been implicated incellular functions that may be relevant to the role of SPT2 in chromatin structure and/or transcription initiation.For example: (I) the human CCG] gene encodes a HMG1-like factor involved in cell-cycle regulation recentlyshown to be a TBP-associated factor (TAF) and to participate in the formation of the TFIID complex (109, 192,203); (ii) the HMG1-like ACP2 protein of yeast was found to be a subunit of RNA polymerase III (89, 147); (iii)the hUBF protein is an RNA polymerase I activator that interacts with a TFIID complex at the rRNA promoterelements (117); (iv) the IIMG1 protein is thought to affect RNA polymerase II transcription (212, 227) and tointeract with the core histones (17).107ASPT2 boxl (26-88)HMGlbox2 (95157) PR PPSR?FZFCSEYRPKX GHPGXS-JGDNHP6 (2183) PR ALSA(MFFANENRDX S2NDXjT_-[GQHMGlboxl (973) PRGKMS FFVQTCREEH KHPDASVNFjSESPT2 box2 (98-159) I_____ ____IRVARK GMWNi9ADDKQ pYEKKAK:LKER1EKIJGKK GKWK TPEEKQPYEAKQA KKESF SKKCS ERWKTMSAKEKGEiFEDMAKA KARYERBDOMINANCEDOMAINS P12 I I 33311 III G 1 I 21 4N H P6 I NL1Figure 19. SPT2 contains two regions of sequence similarity with the HMG-box motif(A) Alignment between the amino acid sequences of the HMG-box motifs of human I{MG1 nonhistonechromosomal protein (239) and of the product of the S. cerevisiae NHP6A gene (125) and two regions of the SPT2protein. Numbers in parentheses give the amino acid positions of the sequences in the respective proteins.Identical or conserved amino acids between SPT2 and the HMG boxes are shaded. Conservative amino acidsubstitutions are as follows: H=K=R; F=Y; D=E; AI=LV; S=T. In SPT2 boxi, v69’SGV. This alignment isbased on the best match obtained for those six sequences by using the CLUSTAL program (102) from thePCGENE software packaged (Intelligenetics Inc.). (B) Schematic representation of the three proteins aligned inA. The HMG boxes of HMG1, NNP6, and the proposed homologs in SPT2 are shaded and numbered as definedfor panel A. Hyperacidic sequences are highlighted by black boxes. Also shown are the total number of aminoacids in each protein as well as the location of the dominance domain of SPT2 mapped by the mutational analysispresented here.108Even if the putative }IMG-box2 of SPT2 is involved in DNA binding, it should be emphasized that theactual mechanism of dominance could still involve protein-protein interactions. For example, the HMG-box2 maybe required to target SPT2 to specific chromosomal locations, where it may contact other chromatin components orthe transcriptional machinery. The products of dominant alleles could prevent those contacts once recruited to theDNA target, a property dependent on their functional DNA-binding domain. Alternatively, protein sequencesadjacent to the HMG-box2 of SPT2 may promote dimer formation. It has been observed previously that fragmentsof HMG1 consisting of amino acids 91 to 176 and 88 to 164 can both bind to DNA through their intact HMG-box(amino acids 95 to 157; see Fig. 19A), but only the largest fragment appears to be dimeric in solution (19, 238).According to the sequence alignment presented in figure 19, the residues of HMG1 required for dimerization(amino acids 165 to 176) would correspond to residues 167 to 178 of SPT2, which are present in the product of thedominant allele, SPT2-1 79.Saccharornyces cerevisiae contains several other genes encoding proteins showing sequence similaritieswith HMG1 and/or the HMG-box motif. In addition to the ACP2 gene already mentioned, these include theduplicated genes NHP6A and NHP6B (125), ABF2 (51, 52), ROX] (14), and JXRJ (31). The NHP6 genes encodesimilar proteins of 11.4 kDa, each of which essentially consist of a single HMG-box (Fig. 19). The role of theseproteins is still unknown, although it was observed recently that nhp6a nhp6b double mutant strains show growthdefects at high temperature and reduced induction of the P1-105 gene (124a). The ABF2 protein was originallycharacterized on the basis of its specific binding to the autonomous replicating sequence ARS] (51). This 20-kDaprotein, composed of two HMG-boxes, is concentrated in the yeast mitochondria and appears to play a role in theexpression and maintenance of the mitochondrial genome. ABF2 is not required for mitotic growth and exhibitsHMG1-like DNA-binding properties such as DNA bending of an ARS] fragment and high-affinity binding ofsupercoiled DNA (52). The ROX] gene is of particular interest here since, like SPT2, it was geneticallycharacterized as a negative regulator of transcription. The 40-kDa Roxi protein acts as a sequence-specificrepressor of genes required under partially anaerobic conditions (145). The basic protein was found to contain anamino-terminal 11MG-box, but the role of this domain in the DNA-binding activity has not yet been studied (14).The JXR] gene of yeast was isolated by screening a yeast expression library with platinated DNA (31). Consistentwith the observation that HMG1-like proteins can specifically recognize such cisplatin-modified DNA (32), the 80-kDa Ixrl protein was shown to be a member of the 11MG-box protein family and to contain two HMG-box motifs109(31). Haploid cells disrupted for JXRJ show no growth defect and the cellular function of this protein is stillunknown.Region UI: polar C-terminus. The SPT2-1 79 allele, which codes for the first 179 amino acids of SPT2protein, behaves as a dominant suppressor. Since the suppression of ö insertional mutations is a consequence ofthe loss of SPT2 protein function, amino acids 180 to 333 of SPT2 protein (region III, Fig. 18) must contain atleast one functional domain which, in conjunction with the dominance domain, is essential for SPT2 activity. Theobservation that the SPT2-324 allele, which codes for the first 324 amino acids of SPT2 protein, also exhibits adominant suppressor phenotype provided for the identification of such a functional domain in the C-terminalamino acids 325 to 333. Among these amino acids, specific basic residues (Lys 325 and Lys 329) were shown tobe required for SPT2 protein function.At present, the N-terminal boundary of this domain has not been mapped, but several features of the C-terminal third of region III suggest that residues 325 to 333 are part of a larger functional structure, mutations inwhich could also yield dominant suppressor alleles. The last 56 amino acids of SPT2 protein consist of a longpolar tail of predicted u-helical structure when analyzed by the secondary structure prediction method of Gamier(78). Composed of 66.5% charged residues, this region also contains an acidic subdomain (residues 277 to 303;net negative charge of 15) similar to the one found in several HMG-like proteins (human HMG1 [239], xUBF[144] and hUBF [116]) and previously shown to promote contacts between HMG1 and histone H2A-H2B dimers invitro (17). A second acidic region (residues 226 to 249; net negative charge of 15) is also present in region III ofSPT2 (Fig. 18). It is possible that the mutations studied here act indirectly by affecting the structure of theseadjacent acidic domains. However, the fact that point mutations at specific residues can mimic the effect of adeletion of residues 325 to 333 suggests that these amino acids may also participate in specific intra- orintermolecular interactions critical for SPT2 protein function. That this positively charged C-terminus does notsimply act as a nuclear localization signal for SPT2 is supported by the observation that the product of a SJN]:lacZfusion lacking the last 17 amino acids of SIN1/SPT2 is, like the wt protein, concentrated in the yeast nucleus (127).1103.3 Characterization of spontaneous recessive sptr mutationsThe experiments discussed in the previous section led to the identification of a structural determinantdifferentiating the products of suppressor spt2 alleles with recessive and dominant phenotypes. One of theconclusions of these studies is that amino acids 117 to 179 define or overlap with a region of SPT2 essential for thedominant suppressor phenotypes of some mutant alleles. It has been established previously that a completedeletion of SPT2+ causes a recessive suppressor phenotype. A similar result was obtained here with a disruptionallele (spt2Ii:: URA3), with two truncated forms of the protein (spt2-46 and spt2-116), and with internal deletantsmissing codons 117 to 179 (z13 alleles). It was therefore demonstrated that recessive spt2 suppressor alleles canencode a nonfunctional product and do not need to be a deletion of the entire locus.In the course of previous genetic studies directed towards the identification of SPT suppressor genes,several spontaneous mutations belonging to the spt2 complementation group have been obtained (245). Mutationsconfering recessive or dominant suppressor phenotypes have been characterized. In an attempt to support theconclusions of the in vitro mutational analyses presented in section 3.2, four recessive spt2 alleles obtained duringthese genetic studies have been investigated. The primary goal of this analysis was to identify spontaneousmutations that generate recessive spt2 suppressor alleles. The results from the mutational analysis would predictthat such mutations should alter the functional domain required for dominance, disrupted by deletion of residues117 to 179. Since deletions or rearrangements involving the SPT2 locus could also create recessive suppressoralleles, the strains analyzed here (provided by Fred Winston) were selected on the basis not only of their genotype(i.e., recessive spt2 allele), but also of the absense of noticeable chromosomal rearrangements involving the SPT2locus.3.3.1 Analysis of genomic DNAThe integrity of the SPT2 locus was first analyzed for four spt2 strains: FW157 (spt2-1 22), FW186 (spt2-22), FW187 (spt2-23), and FW188 (spt2-204). Genomic DNA was isolated and digested with PstI, which cutswithin the 5’-end of the SPT2 coding region and in flanking genomic sequences, generating SPT2 fragments ofapproximately 2.9 and 12 kb in length (Fig. 20A). Controls included strains S704 and SR26-12C, carrying a111complete deletion of SPT2 and the wt gene, respectively. The genomic digests were separated on an agarose gel,and two successive hybridizations were performed in the dried agarose gel matrix using[32P] end-labeledoligonucleotides. These hybridization probes were designed from SPT2 sequences located on either side of the PstIsite (0LL8 and oLL27, Fig. 20A).The results showed that none of the strains analyzed carries a gross deletion or rearrangement of the SPT2locus. The four FW strains with recessive spt2 alleles show two hybridizing bands indistinguishable from thoseobtained with DNA isolated from the wt SPT2 strain, SR26-12C. Since the 2.9-kb band recognized by oLL8 isdefined by a PstI site present in the adjacent Tyl element, it appears that all those strains harbor a similar elementin the 3 Lend of the SPT2 locus on chromosome V. However, these results do not eliminate the possibility that theactual DNA sequence of the nearby Tyl element differs in the strains analyzed.3.3.2 Cloning and sequencing of mutant allelesThe molecular characterization of the mutations in the SPT2 coding region of the four spt2 strains wasachieved in two steps. First, the SPT2 coding region of each mutant strain was amplified by preparative PCR,cloned in the vector pEMBL 18+ (49), and sequenced. Since amplification with Taq polymerase is known to bemutagenic, and because this approach provides the sequence of a single isolated clone obtained from the PCRreaction, the presence in the genomic DNA of the mutation(s) sequenced had to be confinned. This was performedin a second step in which sequencing templates were prepared directly from genomic DNA using asymmetric PCR.In this technique, genomic DNA is amplified by a PCR reaction in which one of the PCR primers is present inlimiting amounts and the other is added with 100-fold excess. The limiting primer is depleted after a fewamplification cycles such that additional extensions can only occur using the excess primer. Although theseadditional cycles no longer select for full-length products, they do allow for the preparation of short ssDNAtemplates suitable for direct sequencing. In contrast to the sequencing information derived from a single, clonedPCR product, the data obtained by direct sequencing of a population of amplified products is thought to reflect the“consensus” primary structure of the genomic sequences. This procedure was exploited to confirm the presence ofeach mutation in the genomic DNA.112p —2.9kb p >12kbI I‘4_____Tyl 4SPT2M3.05 —2.04 —1.64— —Figure 20. Restriction analysis of the SPT2 locus from mutant strains.(A) Structure and restriction map of the wt SPT2 locus on chromosome V. Also shown are the positions of thesequences from which the oligonucleotide probes oLL8 and oLL27 were designed, as well as the size of theexpected hybridizing PstI (P) fragments. (B) A composite picture of the autoradiographs from two successivehybridizations in the dried agarose gel using the indicated[32P] end-labeled oligonucleotides (arrows). The PstIgenomic digests were from 4 spt2 strains (FW157, 186, 187 and 188), four SPT2 strains (FW158, 160, 201 and202, not studied here), and the control strains S704 (spt2-150, a deletion allele) and 5R26-12C (SPT2j.oLLS oLL27ABp112.2 —4.07 —I I1kb4...... oLL274— oLL8113The results of this analysis are summarized in table 6. Each PCR clone of the spt2 alleles (calledpFW157, pFW186 to 188) was found to contain at least one mutational alteration. The clones, pFW157 andpFW187, each contain a unique mutation. The sequencing of asymmetric PCR templates obtained from genomicDNA confirmed the presence of these single mutations in the SPT2 coding region of strains FW157 and FW187.The clones, pFW 188 and pFW 186, show two and three mutations, respectively. However, asymmetric PCRsequencing demonstrated that both mutations identified in pFW 188 and two of the three mutations cloned inpFW186 are not present in the spt2 allele of the corresponding strains. These mutations are likely to representPCR artefacts introduced during the amplification reaction in the specific molecule cloned. With regard to thefidelity of Taq DNA polymerase, the cumulative error frequency is about 0.1% (4/4000) after 20 cycles of PCR.Assuming 20 doublings, this corresponds to an average mutation rate of 1/10,000 misincorporations per nucleotidepolymerized per cycle (1 x 10k), which is comparable to the published values of 1.7 x i04 to 1.2 x i0 (79).Table 6. Analysis of 4 recessive spt2 alleles.Strain Allele Mutation Expected productFW157 spt2-122 A1toC noneFW186 spt2-22 AC425 1-141°FW187 spt2-23 2x (62 to 81) 1271FW188 spt2-204 nonea Followed by the 8 amino acids QKLSHRNP.b Followed by the 15 amino acids CKIHYPSAKSISQEE.114The sequencing analysis presented here has failed to identil’ any mutation in the coding region of thespt2-204 allele of strain FW188. That the spt2-204 allele may be a regulatory mutation, which affects theexpression of the gene, is suggested by the absence of SPT2 protein in FW188 extracts (not shown).The only mutation found in the spt2-23 allele of strain FW187 was a duplication of nucleotides 62 to 81 ofthe coding region. This duplication alters the coding region of SPT2 such that, if translated, it would code for thefirst 27 amino acids of SPT2, followed by the residues CKLHYPSAKSISQEE. Allele spt2-23 can therefore beconsidered a null allele of SPT2, which is consistent with its recessive suppressor phenotype.Allele spt2-122 of strain FW157 was shown to carry an A to C transversion at position 1 of the codingregion. This mutation changes the AUG initiator codon to CUG, which codes for leucine, and is thus likely tointerfere with the production of the SPT2 protein. Indeed, the SPT2 open-reading frame does not contain anotherAUG codon in the vicinity of codon 1 (the second AUG being present at position 86 and in a different frame) and itis a well documented fact that translation initiates exclusively at AUG codons in eukaryotes. For example, studieson cyci mutations have shown that 17 out of 490 cyci mutants lacking iso-1-cytochrome c carried single-sitemutations in the AUG initiator codon (207), Revertants of those mutations carried an additional substitutionintroducing a new AUG codon in the same reading frame, 1 to 4 codons away from the original AUG. Theefficient translation of cycl in these intragenic revertants was promoted therefore by the new AUG. From thesestudies, it can be concluded that, in the absence of nearby and in-frame AUG codons, mutations altering the AUGinitiator codon interfere with the production of the gene product. Similar to the case of the allele spt2-23, the spt2-122 allele provides another example of a spontaneous null spt2 mutation.The only mutation identified in the spt2-22 allele of strain FW186 was a single base-pair deletion atposition 425 of the SPT2 coding region. At the translational level, this mutation introduces a frameshift aftercodon 141. The expected translational product of spt2-22 consists of the first 141 amino acids of SPT2, followedby the amino acid sequence QKLSHRNP. The in vitro mutational analysis presented in section 3.2 showed thatwhereas a truncated polypeptide of 116 amino acids confers a recessive suppressor phenotype, a longer product of179 amino acid residues can compete with the maintenance of the Spt+ phenotype As a consequence, it confers adominant suppressor phenotype. The nature of the spontaneous, recessive, spt2-22 mutation is consistent with thein vitro data and further extends conclusions derived from them. Both spt2-1 16 (amino acid ito 116) and spt2-22115(amino acids ito 141) are nonfunctional polypeptides with regard to the repression of his4-9126 Since neither ofthem can interfere with the Spf’ phenotype (since they are recessive), the spt2-22 allele product defines a new C-terminal boundary - amino acid 141 - for nonfunctional recessive SPT2 derivatives. Conversely, it appears thatamino acids 142 to 179 of SPT2 comprise structural determinants that are required for the synthesis of anonfunctional and competing polypeptide. The structure and phenotype of spt2-22 are also consistent with thehypothesis that an intact HMG-box2 (amino acids 98 to 159 or 170) may be required for a nonfunctional SPT2derivative to confer a dominant suppressor phenotype.3.4 Studies on the mechanism of dominance of SPT27 allelesThe previous sections dealt with the definition of functionally important regions of the SPT2 proteinthrough the analysis of in vivo- and in vifro-generated mutant spt2 alleles. Two independent genetic assays wereexploited to characterize the spt2 alleles under study: a complementation assay and a dominance test. Thecomplementation assay answered the following question: Does the mutant protein constitute a functional repressorof his4-9126’? The outcome of this phenotypic test was the definition of two distinct categories of SPT2derivatives: the functional repressors (including the wt protein and pseudo-wt mutant repressors) and thenonfunctional repressors (which confer a suppressor phenotype in a spt2 background). The dominance testallowed a further partition of the members of the latter class by asking the question: Can this nonfunctional SPT2derivative interfere with the repression of his4-912ö in a SPT+ background? On the basis of this second test, thesuppressor phenotype associated with a nonfunctional repressor (and by extension, the allele or repressor itself) isclassified as recessive or dominant. This section focuses on the functional basis of the dominant phenotype.3.4.1 Gene dosage studiesFrom a genetic perspective, the dominance of a phenotypic trait over another one often is associated witha gain of function. For example, the repressor phenotype caused by the wt SPT2+ gene (His phenotype of a his4-912ö strain) is dominant over the suppressor phenotype of a spt2 null allele (His+ phenotype of a his4-9]2östrain). In SPT2/sptr heterozygous diploids, the dominance of SPT2 correlates with the gain of repressor116function. Conversely, the recessiveness of a phenotypic trait over another is perceived in terms of a loss offunction: the suppressor phenotype of a spt2 null allele is recessive to the repressor phenotype of the wt SPT2+gene because it correlates with the loss of repressor function.This logical argument, and the underlying nomenclature, can lead to apparent contradictions, as in thecase of the dominant suppressor phenotype of non-complementing SPT2 alleles. On one hand, the dominance oftheir suppressor phenotype over the wt repressor phenotype suggests that dominant alleles constitute a gain offunction. On the other hand, their suppressor phenotype would be consistent with a loss of repressor function, asseen for null spt2 alleles. This inconsistency can be reconciled by the introduction of two categories of gain-of-function mutations that have acquired the ability to interfere with the maintenance of the phenotype associatedwith the wt allele, thus conferring a new phenotype identical to the one associated with a null loss-of-functionallele. These non-complementing, interfering (or dominant) alleles are known as neomorphic and antimorphicmutations (148). Although they can induce similar phenotypic changes in heterozygous diploids with the wt allele,they operate by different mechanisms. A neomorphic mutation has acquired the ability to perform a new function,not shared by the wt gene product, that allows its mutant product to interfere with another activity, also requiredfor the wt phenotype. An antimorphic mutation has acquired the ability to act as an antagonist of the wt geneproduct and to interfere directly with the maintenance of the wt phenotype by the wt protein. A third class of gain-of -function dominant mutations, called hypennorphs, will not be discussed further here since it was shown thatoverproduction of SPT2 does not confer a mutant suppressor phenotype.Since the functions of several unlinked SPT genes are required to maintain the Spt+ phenotype of a yeastcell, the product of dominant SPT2 alleles could act via one of two distinct mechanisms. As neomorphicmutations, SPT2 alleles could have gained the ability to compete with the function of another SPT protein. Asantimorphic mutations, SPT2 alleles could have acquired an activity opposite to that of the wt SPT2 protein, ananti-SPT2 activity. Genetically, these two possibilities can be distinguished by gene dosage studies in which thedominant alleles are challenged by different levels of wt product. Whereas the activity of antimorphic alleles isexpected to be sensitive to the presence of wt protein, that of true neomorphic alleles should be unaffected byvariations in the wt protein levels.Two different experimental approaches were taken to effect the gene dosage variations necessary toidentify the nature of dominant SPT2 alleles: mating experiments, which examine suppression in diploids, and117yeast transformations, in which suppression in merodiploids2is studied. These experiments were made possibleby the construction of isogenic yeast strains carrying different in vitro-generated spt2 alleles substituted for theSPT2+ gene on chromosome V. Construction of isogenic spt2 and SPT2 yeast strainsThe haploid yeast strain SLL7 carries the disruption allele spt2A:: URA3 substituted for the SPT2 locus onchromosome V. This null allele is a recessive suppressor of the insertional mutation his4-9128 such that SLL7cells are phenotypically FTiis+ and Ura+. The possibility to impose a selection for the loss of URA3 function usingthe analog 5-FOA (discussed on page 85) was utilized in the development of a protocol for direct allelereplacement in SLL7. The procedure, based on the one-step gene replacement protocol used to create SLL7 fromSLL5 (the SPT2 parent), consisted of transforming SLL7 spheroplasts with DNA fragments containing differentspt2 alleles and imposing 5-FOA selection after a period of non-selective growth. This outgrowth was found to benecessary to allow the expression of the Ura phenotype.Two different events were expected to generate the desired 5-FOA’ (Ura) phenotype: spontaneousmutations in the URA3 marker of the spt2&: URA3 allele, and replacement of spt2zl:: URA3 with the transformedallele by homologous recombination. The emergence of false positives due to spontaneous ura3 mutations wasreduced significantly by using a cotransformation procedure and primary selection for transformed cells. Positiveintegrants were identified by analytical PCR amplification on genomic DNA, in which the structure of the allelepresent at the SFT2 locus can be determined readily (Fig. 21A).Using this approach, three non-complementing spt2 alleles were transformed in SLL7 and successfullysubstituted for the SPT2 locus: the recessive alleles, spt2zi (strain SLL9) and spt2-324z13 (strain SLL1O1), and thedominant allele, SPT2-324 (strain SLL 102). The first attempts to perform allele replacement in SLL7 by 5FOArselection involved transformations with the spt2iJ allele. Unexpectedly, three as opposed to two different kinds ofalleles were detected at the SPT2 locus of Ura cells. Some transformants carried an allele structurally similar tothe original disruption allele spt24:: URA3 of SLL7; these are thought to be spontaneous ura3 mutants (Fig. 21B,2 The term merodiploid, which means partly diploid’, refers to haploid cells in which only part ofthe genome (usually present on a centromericepisome) is duplicated.Aspt2A.:URA3oLL5OoLL5OoLL5OoLL34oLL34SPT2-324_______oLLS0oLL34hioLL34oLL5OoLL341420 bp1056 bp1056 bp118BC1.64—1.02 —0.51 —0.30 —Ml 2 3 4 5 6 789 101112MM 1 2 3 4 5 6 7 8 9 10 M4— spt2A::URA3+—Sp72+(?).—.spt2i4— spt2zl::URA3— SPT2-324— spt2-324z13Figure 21. Screening of gene replacement events at the spt2/J:: URA3 locus of strain SLL7 by analytical PCR.(A) Schematic representations of the alleles spt2zl:: URA3, SPT2, SPT2-324, spt2-324z13, and spt2zl, showing thepriming sites for the PCR primers oLL34 and oLL5O, and the size of the expected PCR product for each allele. (B)and (C) Ethidium bromide-stained agarose gel of analytical PCR products. (B) Analysis of spt2zl integration. ThePCR templates were as follows: lanes ito 9, genomic DNA from 5FOAr transformants; lanes 10 to 12, DNA fromplasmids pLL82 (YEpspt2zl:: URA3); pLL8O (YEpspt2zi), and pLL1O (YEpSPT2j respectively. (C) Analysis ofSPT2-324 (lanes 1 to 6) and spt2-32443 integrations (lane 7). The PCR templates were as follows: lanes ito 7,genomic DNA from 5FoAr transformants; lanes 8 to 10, DNA from plasmids pLL82; pLL1O, and pLL74(YEpspt2-32433) respectively.SPT2 _-1 >—spt2-324z13spt2A867 bp254 bp119lane 1). Others have lost the disruption allele and substituted the desired spt24 construct (Fig. 21B, lanes 4 and 9);one of these was named SLL9. However, most 5-FOA resistant cells were found to harbor a different allele, givinga PCR product similar to that of the wt allele (Fig. 21B, lanes 2,3,5-8). That those undesired transformantsoriginated from SLL5 contaminations was suggested by further experiments in which a SLL7 clone isolated from aSC-Ura plate was utilized. Transformation of this Ura+ SLL7 isolate with DNA fragments from the alleles spf2-324zi3 and SPT2-324 gave fewer 5FOAr colonies, all of which had substituted the transformed alleles (Fig. 21C).Although the structure of the SPT2-324 allele cannot be distinguished from that of the wt allele by the PCRanalysis of genomic DNA, further analyses confirmed the production of a truncated nonfunctional SPT2 derivativein these 5_FOAr clones (see below).The three isogenic strains obtained by allele replacement experiments - SLL9, SLL1O1, and SLL1O2 -were subjected to phenotypic characterization to confirm the presence of the expected genetic markers. As shownin figure 23 (page 123), they all exhibit a His phenotype at 30°C, which is consistent with the presence of asuppressor allele at the SPT2 locus.The levels of suppression of his4-9l2öby each allele was also estimated by growth studies. The isogenicstrains SLL5, SLL9, SLL1O1, and SLL1O2, as well as the control strains SLL2O8 (HIS4 SPT2j and SLL2O9(HIS4+ spt2-150), were grown in minimal medium, and the growth of the haploid strains at 30°C was monitoredby measurement of the optical density at 6OO. Table 7 presents the generation time (doubling time) obtained foreach strain grown in complete minimal medium (SC) or in histidine dropout medium (SC-His). Whereas theisogenic strains all show similar generation times in complete minimal medium (140±20 mm), only SLL5 isdeficient for growth in absence of histidine, with a 11.5-fold decrease in growth rate. Furthermore, the growth ofSLL9, SLL 101, and SLL 102 in SC-His is comparable to that of the HJS4 strains SLL2O8 and SLL2O9. Theseresults show that the isogenic strains constructed from SLL5 by allele replacement experiments have acquired asuppressor phenotype allowing normal growth in absence of histidine. Otherwise, the loss of SPT2 functionappears to have very little effect on the mitotic growth of haploid cells in complete synthetic medium.120Table 7. Generation times of wt and mutant SPT2 strains.’1Generation time (mm) in minimal mediaStrain Relevant genotypeSC SC-HisSLL5 his4-912SSPT2 120 1374 (11.45)SLL9 his4-9l2öspt2zi 156 204 (1.3)SLL1O1 his4-9126SPT2-324 150 198 (1.3)SLL1O2 his4-.912öspt2-324z13 138 204 (1.5)5LL208 HIS4SPT2 150 204 (1.4)SLL2O9 HIS4spt2-15O 192 246 (1.3)a The ratio ofthe generation times (SCHis/SC) is given in parentheses.The expression of the mutant SPT2 protein produced by these strains was also analyzed by Westernimmunoblot (Fig. 22). None of these strains was found to produce the wt SPT2 protein or a cross-reacting specieswith similar electrophoretic mobility. Extracts from strains SLL1O1 and SLL1O2 were also shown to contain a newprotein recognized by the anti-peptide antisera (Fig. 22, lanes 1 and 2). The electrophoretic mobility of eachprotein band is consistent with the production of the expected mutant protein: allele SPT2-324 of strain SLL1O 1codes for a SPT2 derivative truncated of 9 C-terminal amino acids (see Fig. 14, page 93), and allele spt2-324A3 ofstrain SLL1O2 codes for a deleted version of SPT2-324, missing amino acids 117 to 179 (see Fig. 15, page 98).Strain SLL9 does not produce any cross-reacting polypeptide from the SPT2 locus as expected from the deletion ofmost of the SPT2 coding region in the allele spt2zl (Fig. 22, lane 4).The results presented here continued that the isogenic haploid strains SLL5, SLL9, SLL 101 and SLL 102carry different spt2 alleles at the SPT2 locus, and that each of these mutant alleles directs the production of theexpected polypeptide product. The suppressor phenotype of strains SLL9, SLL1O1, and SLL1O2 also demonstratedthat the alleles spt2zi, SPT2-324, and spt2-324iJ3 encode nonfunctional SPT2 derivatives leading to thesuppression of his4-9125 even when expressed from a single chromosomal allele.12168.0—43.0—29.0—18.4—12345Figure 22. Western inununoblot analysis of mutant spt2 strains.Extracts from the isogenic strains SLL1O1 (SPT2-324, lane 1), SLL1O2 (spt2-324z13, lane 2), SLL5 (SPT2, lane3) and SLL9 (spt2zl, lane 4) were analyzed by immunublotting as described in the legend of figure 13, page 88.An extract from SPT2 overexpressing cells (SLL7 [spt2zl:: URA3]/pLL1O [YEpSPT2j, lane 5) was included as acontrol. The positions of protein markers are shown to the left (kDa). Suppression in diploidsThe possibility that dominant suppressors mediate their effects by inhibition of an unrelated SPT geneproduct, and not of the SPT2 protein, was tested first in heterozygous diploid strains. The four isogenic MATahis4-9125 strains carrying different spt2 alleles were mated with MA Ta his4-9128 strains with (SLL4) or without(SLL1O spt2zl:: URA3) the wt SPT2 gene. All the haploid cells and their pairwise mating products wereanalyzed for their ability to grow in absence of histidine (Fig. 23 A).The three mutant alleles tested (spt24 spt2-324z13 and SPT2-324) showed a strong suppressor phenotypeboth in haploid cells and in combination with the null allele spt2zl:: URA3 in heterozygous diploids. This resultsuggests that the action of these suppressor alleles is not affected by doubling the dosage of other SPT loci. It wasobserved also that, whereas wt homozygous diploids show a stable His phenotype, heterozygous diploids carryinga non-complementing recessive allele and a wt allele (e.g. spt2zl/SPT2j exhibit a background His phenotype withhigh frequency of f{j5+ reversion. This result is consistent with the full complementation of the recessive alleles by122SPT2 and the emergence of homozygous suppressor diploids by mitolic recombination (e.g. spt24/SPT2becomes spt24/spt2ii). A similar phenomenon has been described already for the heterozygote SLL4x7 (section3.1.2). The SPT2-324/spt2zi:: URA3 diploids clearly show a stronger F{is phenotype than the SPT2-324/SPT2diploids, indicating the semi-dominant phenotype for the SPT2-324 allele in diploids. This dominance wasabolished when residues 117 to 179 were deleted (allele spt2-324z13). Although the background His phenotype ofSPT27spt2 diploids can be masked by the high density ofHis revertants (notably in the case of SPT2/spt2zi:: URA3 diploids, Fig. 23A), spreading of those cells or spotting at lower cell density confinned the phenotypicdistinction between recessive and dominant alleles in heterozygous diploids with the wt allele (not shown). Thepartial dominance of several spt2 alleles in diploids has been observed previously and in fact was utilized to assignthese alleles to the spt2 complementation group (245). Suppression in merodiploidsThe phenotypic differences between the heterozygous diploids SPT2-324/spt2zi:: URA3 and SPT2-324/SPT2 suggest that the extent of suppression is sensitive to the relative levels of mutant and wt products, andthat the two polypeptides are in direct competition in heterozygotes. This conclusion is supported also byexperiments in which recessive and dominant alleles are challenged directly by the wt gene in merodiploids,obtained by transformation of isogenic strains with a YCpSPT2 construct (Fig. 23B). The presence of the wtSPT2 protein has no effect on the repression of the his4-9126 mutation in a SPT2 strain, but is sufficient tocomplement the recessive suppressor mutation, spt2-324z13. When the dominance domain is present, as in thetruncated product of SPT2-324, the complementation of the Spr phenotype by YCpSPT2 is only partial. On theother hand, expression of the dominant allele SPT2-1 in a wt background (SFT2 [YCpSPT2-11) leads to a strongHis phenotype, indistinguishable from that of a spt2ii:: URA3 strain (Fig. 23B). Therefore, heterozygousmerodiploids show different levels of suppression when chromosomal or episomal expression of dominant and wtalleles are compared (SPT2 [YCpSPT2-11 vs SPT2-324 [YCpSPT2j). This observation suggests that the relativelevels of wt and mutant products are different in these transfonnants and that the dominant suppressor phenotypeis sensitive to the levels of wt protein.123SPT2spt2-324A3SPT2-324SPT2(YCpSPT2-1)spt2Z::URA3YCp5O YCpSPT2Figure 23. Dosage dependence of the dominant Spt phenotype.(A) Suppression in diploids. The isogenic MA Ta his4-9125 strains SLL5 (SPT2j, SLL9 (spt2A), SLL1O 1(SPT2-324), and SLL 102 (spt2-324z13) were mated to the MA Ta his4-9125 strains SLL4 (SPT2j and SLL 10(spt2iI:: URA3) by selection on SC-Lys-Trp dropout plates. Haploid cells and their mating products were grown inYPD and equal amounts of cells were spotted on SC-His plates and grown at 30°C for 3 days.(B) Suppression in merodiploids. The isogenic strains SLL5 (SPT2j, SLL1O2 (spt2-324z13), and SLL1O1 (SPT2-324) were transformed to uracil prototrophy with the centromeric plasmid YCp5O or pLL15 (YCpSPT2j. Twoindependent transformants were grown in SC-Ura, and equal amounts of cells were spotted on SC-Ura-His andgrown at 30°C for 3 days. Controls for the dominance (SLL5 transformed with plasmid pLL24 [YCpSPT2-1] ) andfor the suppression by a null allele (SLL7 [spt2iJ:: URA3] ) are also presented.A BSPT2spt2ASPT2-324spt2-324A3SC-His SC-Ura-His124From a genetic perspective, these results suggest that dominant SPT2 alleles do not behave as neomorphicmutations, which have acquired a new activity, but rather as negatively acting antimorphs (also called dominantnegative mutations; ref. 99), which compete against the wt protein and act as anti-SPT2 polypeptides (148). Thesequalitative data do not eliminate the possibility of an increase in affinity in dominant polypeptides, but argueagainst a change in specificity. It is therefore predicted that high levels of wt protein could completely reverse thedominant phenotype. Knowledge of the mechanism and stoichiometry of the interaction(s) responsible for thedominance will be required to predict the levels necessary for this reversion. The dominance domain defined byresidues 117 to 179 thus allows nonfunctional polypeptides with an altered C-terminus to act as competitors forSPT2-mediated regulation of his4-91263.4.2 Interference of SPT2 function by dominant SPT2- proteinsThe genetic experiments described so far have shown that dominant SPT2 alleles can interfere with themaintenance of the Spt+ phenotype when expressed in a SPT+ background, and that this activity is sensitive to thelevel of wt SPT2 protein. Since the absence of SPT2 activity causes a suppressor phenotype, a model for the actionof SPT2- polypeptides consistent with these results proposes that dominant alleles interfere with the production ofthe wt protein. The following section presents three experimental approaches designed to address this possibility.A yeast strain carrying a spt2zl null allele and cotransformed with a low-copy-number vector carrying theSPT2 gene and a dominant mutant SPT2 allele on a high-copy-number plasmid exhibits a His phenotype. TheSPT2+ gene was expressed from a heterologous promoter, the weak glucose-repressible GAL4 promoter, and thesame dominance effect was observed in cotransformation experiments (Fig. 24A). The pGAL4:SPT2+ fusion wassubcloned in low- and high-copy-number vectors (YCp and YEp), and three qualitatively different levels ofexpressions were achieved under repressed (glucose) or derepressed (galactose) conditions. When cotransformedinto a his4-9]2öspt2zl strain along with a control parental vector, only the high-copy-number fusion underderepression produces enough SPT2 protein to complement fully the null allele and generate a His phenotypesimilar to the one observed in cotransformants with a YEpSPT2+ vector. At the other extreme, the repressed lowcopy-number fusion does not produce enough SPT2 protein to complement spt2zl. The repressed YEppGAL4:SPT2 and derepressed YCp-pGAL4:SPT2 lead to an intermediate His phenotype at 30°C, suggestingthat under these conditions partial suppression of the ö insertion is achieved. However, the presence of the125dominant allele SPT2-] in the cotransfonued vector always yields a strong His+ phenotype. It has therefore beenconcluded that the dominance occurs even when SPT2+ is expressed from a heterologous promoter and that inthese experiments the competence of the high-copy SPT2-1 allele to dominate is not affected by the level ofexpression of SPT2+. In conjunction with the experiments showing that dominant polypeptides interfere with wtSPT2 function as antimorphic mutations, these results suggest that the dominance mechanism does not involve aspecific transcriptional effect on the SPT2 promoter.It was also found that the intracellular levels of the wt protein are not affected by the presence of adominant allele. The strain SLL9 (spt2zl Ura Leu) was cotransfonned with vectors expressing the wt SPT2protein and/or one of two dominant products, SPT2-l and SPT2-324. Total protein extracts of cotransformedyeasts were analyzed by Western immunoblotting using the combined anti-SPT2-peptide antisera as a probe (Fig.24). The results show that the dominance of the SPT2-1 or SPT2-324 allele is not caused by transcriptional ortranslational inhibition of the wt SPT2gene. It should be noted that the dominant negative product of the SPT2-]allele is a truncated polypeptide lacking the C-tenrnnal third of the SPT2 protein, since it does not contain the C-terminal epitopes of the cross-reacting antipeptide sera.The conclusion of the cotransformation experiment described above was also supported by examining theexpression of an in-frame spt2::lacZ fusion in SPT2 and SPT2-] backgrounds. When expressed from a high-copy-number vector, the fusion produced 22 to 25 U of f3-galactosidase, irrespective of the cotransformed allele.The experiments described in this section show that, although the products of dominant SPT2 allelesinterfere with the function of the wt SPT2 protein, they do not appear to inbibit the production of SPT2. Thesimultaneous production of wild type and dominant proteins in Spt cells implies that the mutant polypeptidescompete with SPT2 function post-translationally, a mechanism consistent with the mutagenesis studies. Indeed,the implication of a putative 11MG-box DNA-binding domain in the dominance of polypeptides encoded bydominant SPT2 alleles raises the possibility that the competition between wt and dominant products could occur atthe level of DNA-binding.126AURA3 pGAL4:SPT2YEp YCp — YEp YCp — YEp YCp —— 29.0 —Figure 24. Effects of dominant alleles on pGAL4:SPT2 function and SPT2 protein levels.(A) Strain SLL9 (his4-9l2öspt2zl) was cotransformed to leucine prototrophy with pLL1O (YEpSPT2j, pLL18(YEpSPT2-1), or the parental vector pEMBLYe3O and uracil prototrophy with pLL97 (YEp-pGAL4:SPT2j,pLL96 (YCp-pGAL4:SFT2j or pLL95 (YEp-pGAL4). Two independent transformants were grown in SC-UraLeu dropout medium, and equal amounts of cells were spotted on the indicated media, selecting for both vectors,and grown for 3 days at 30°C. Coomassie blue-stained gel (B) and Western blot analysis (C) of total proteinextracts, as described for Fig. 13, but using an SDS-15% polyacrylamide gel. Yeast extracts are from sLL9cotransformed with pEMBLYe3O, pLL18, or pLL49 (YEpSFT2-324) and pVT100-U or pLL77 (YEpSPT2j. Thedouble arrow points to the wt and SPT2-324 gene products.SPT2SPT2-1GLUCOSE GALACTOSE GALACTOSE—His —HisBSPT2- + - + - +SPT2-1 - - ++ - -SPT2-324 - - -- + +C- + - + - + SPT2- -++- -SPT2-1kDa - - - - + + SPT2-32497.4— 68.01— 43.0 —123456 M— ——Kspi’24123456127A potential explanation consistent with this analysis implies that wt and dominant negative products areboth able to bind DNA through the HMG-box motifs (alone or in combination with other specific factors) and thatan intact C-terminal domain is necessary to promote the additional interactions mediating the transcriptionalnegative regulatory effects of SPT2 protein. Since it has been shown previously that a bacterially expressedTrpE:S1N1 fusion binds DNA non-specifically in vitro (127), it is conceivable that additional nuclear proteinsinteract with SPT2 protein and restrict its activity by directing the repressor protein to its target genes. Theproduct of a dominant negative allele may also retain the ability to interact with other component(s) of a repressorcomplex once bound to DNA and disnipt their activity due to a nonfunctional polar tail. This could provide apossible explanation for the stronger Sin phenotype of particular dominant negative SJN1/SPT2 alleles (127) whencompared with the phenotype of a null deletion allele. According to this model, the HMG-boxes of SPT2 proteinwould not be functionally redundant and the HMG-box 2, which overlaps with the dominance domain, would berequired for high-affinity binding. Studies on the transcriptional activator hUBF, which contain at least 4 11MG-boxes, have shown that these domains may play different roles not only in DNA binding, but also in transcriptionalactivation (117).The proposed molecular mechanism of dominance bears some resemblance to the mode of action ofnomnducible LexA and repressor proteins (138): both wt and mutant repressors can bind to the operator sitesand the dominant mutant phenotype requires an intact DNA-binding domain. An important distinction betweenthe action of these prokaryotic repressors and the model of SPT2 protein negative regulation is that the products ofdominant SPT2 alleles are not functional for repression. Association with other proteins of the transcriptionalmachinery or chromatin components may be essential for this activity, relieved in the presence of activators and anintact RNA polymerase II C-terminal domain (175).1284. CONCLUSIONS AND FUTURE DIRECTIONSThe primary goal of this work was to initiate mutational and functional studies on the SPT2 gene of yeast.The development of anti-peptide polyclonal antibodies provided a confirmation that the 333-codon SPT2 codingregion is expressed in yeast and produces a protein that is required for transcriptional inhibition of the insertionalleles such as his4-9126 Site-directed mutagenesis techniques were utilized to introduce defined mutationalalterations into the cloned SPT2+ gene and construct several suppressor alleles conferring recessive and dominantphenotypes. This mutational analysis identified two functionally important regions of the SPT2 protein, a basic C-terminal region (residues 325 to 333) and a central region (residues 117 to 179). The deletion of each region ofSPT2 creates a nonfunctional polypeptide that confers a suppressor phenotype (Spr) in a spt2 strain, even whenthe mutant gene is overexpressed. That these two regions are functionally distinct is suggested by the observationthat the dominance of some C-terminal deletion alleles requires residues 117 to 179, which overlap a region ofhomology with the }{MG-box, DNA-binding domain. The characterization of a spontaneous recessive spt2 alleleexpected to produce a polypeptide containing amino acids ito 141 of SPT2 suggested that residues 142 to 179overlap with a functional domain responsible for the dominance phenomenon. Results also demonstrate that pointsubstitutions in specific C-terminal basic residues can mimic the effects of C-terminal truncations and createdominant suppressor alleles.With regard to the mechanism of dominance, it was found that an increase in the levels of wt protein caninterfere with the dominant phenotype, suggesting that the dominant alleles act as antimorphic or dominantnegative mutations. Even though the Spr suppressor phenotype can be caused by the absence of SPT2 protein,cotransformation experiments showed that dominant alleles do not inhibit SPT2 production. These results suggesta mechanism of dominance whereby the products of dominant alleles interfere with the function of the wt proteinpost-translationally, perhaps at the level of DNA binding through the 11MG-box-like region.The definition of two functional regions of the SPT2 protein has raised several questions concerning thefunction of SPT2, its regulation by genetically defined antagonists, and the mechanism of dominance of particularalleles. Although it is clear that much remains to be learned about the biochemical properties of SPT2 and that129numerous approaches could shed light on the possible role of this nuclear protein, some of the implications of thismutational study suggest new experimental avenues.Studies examining both the structure and the function of a small dominant polypeptide could provideimportant insight into the transcriptional role of SPT2. It has been shown here that amino acids 1 to 179 of SPT2contain the primary information necessary and sufficient for the folding of an interfering polypeptide that cancompete with the function of SPT2. Another conclusion of the mutational analysis is that the N-terminus of SPT2(residues 1 to 81) may be dispensable for its repressor function. It would therefore be interesting to determine theN-terminal boundary of a dominant SPT2 derivative by progressive deletions toward amino acid 179. Thisapproach could lead to the definition of the smallest dominant polypeptide and provide a test for the hypothesisthat the HtvIG-box 2 is involved in the mechanism of dominance. Alternatively, the functional significance of theproposed similarity could be tested directly by expression and characterization of a HMG-box 2-containingpolypeptide.The implication of specific C-terminal residues in the repressor activity of SPT2 suggested that this regionof the SPT2 protein may be the target of its proposed antagonist, the SWI-SNF complex. According to this model,the dominant alleles studied here generate a non-repressing, “activated” form of SPT2, similar to the onetransiently produced by the action of the SWI-SNF complex during activation of transcription. Otherinterpretations consistent with the results presented here could implicate these C-terminal residues in intra- orintermolecular interactions required for repression. Although these models are not necessarily exclusive, therole(s) of these residues in SPT2 function could be addressed by characterizing suppressors of their effects.Following a gene replacement procedure similar to the one described here to introduce mutant alleles at the SPT2locus, dominant point substitution mutations could be integrated in a his4-91251ys2-l28Sbackground. Using aclassical genetic approach, these 1j+ Lys+ mutant strains could be mutagenized to isolate His Lys revertants.The genetic and molecular analysis of these revertants could determine whether intragenic, second-site spt2mutations can revert the dominant Spt phenotype and create pseudo-wt alleles. Of course, this possibility couldalso be addressed directly through non-specific mutagenesis of the cloned dominant alleles in vitro. Alternatively,the Spt+ phenotype of the revertants could be caused by extragenic mutations affecting the function of factorsinteracting with SPT2 or modil’ing its function. This strategy, which is similar to the one used to reveal the130interaction between SPT3 and SPT15 (the TATA-box binding factor), for example, has great potential and couldlead to a better understanding of the cellular role of SPT2.As mentioned in the introduction, mutations in the SPT2 gene were also found to suppress swi and rpomutations (Sin and Srb phenotypes respectively). Although the work presented in this thesis focused on the Sptphenotype of spt2 mutations, the dominant alleles constructed could also be tested for their ability to conferdominant Sin and Srb phenotypes. That the same structural and functional domains of SPT2 may be involved inat least two of these phenotypes is suggested by the observation that the sinl-2 allele exhibits dominant Spr andSin phenotypes (127). The Sin and Srb phenotypes associated with particular spt2 alleles could be studied bylooking at the levels ofINOI mRNA in the appropriate swil and rpozJCTD genetic backgrounds. Thisexperimental design is based on the observation that a deletion allele of SPT2/SINJ can partially suppress thetranscriptional defects noted at the JNO] locus in swi and rpoziCTD strains (175). Alternatively, the dominanceof spt2 alleles for the Sin phenotype could be assayed with a ho::lacZ reporter fusion, or even by monitoring theexpression of the SUC2 gene, which is a well characterized SWI-dependent transcriptional unit (244). Thesestudies could establish whether or not the three phenotypes associated with spt2 mutations respond similarly to themutational alterations leading to recessive and dominant Spr phenotypes and lead to the elaboration of a unifyingmodel of SPT2 function.1315. APPENDICES5.1 NdeI fragment of SPT2 locusAll the mutational and subclomng experiments presented in this thesis were performed using a 2460 bplong DNA fragment from chromosome V of S. cerevisiae. The sequence of this NdeI restriction fragment, whichencompasses the 3’ end of the RAD4 gene, the entire SPT2 coding region, and 905 bp of its 3’-flanking sequences,is presented below. The sequencing data compiled to form this entire sequence come from the published sequencesof the SPT2 gene (180), of the RAD4 gene (80), and of an adjacent Tyl element (95). It was extracted from thesequence of a larger fragment of chromosome V (15 978 bp; locus YSCSYGP1) obtained through the GenBankdatabase (NIH) under accession number L10718. The Tyl element is not present on the NdeI fragment; accordingto the numbering presented below, it is inserted at position 2492.The double-stranded DNA sequence is presented here numbered from the upstream NdeI site with regardto the SPT2 gene. The translational products of part ofRAD4 (position ito 272) and of SPT2 (positions 554 to1555) are shown above the sequence. Underlined nucleotide bases highlight the positions of important restrictionsites (sense strand) and sequencing primers (anti-sense strand).NdeIY G K IA E E E P N VT K EQ N IA DCATATGGTAAAATTGCCGAGGAAGAACCTAACGTTACGAAGGAACAGAATATTGCGGACAGTATACCATTTTAACGGCTCCTTCTTGGATTGCAATGCTTCCTTGTCTTATAACGCCTGT10 20 30 40 50 60N H D NT E T F MG G G FL PG IAN HATCACGATAATACGGAGACTTTTATGGGAGGTGGGTTCCTACCAGGTATAGCAAACCACGTAGTGCTATTATGCCTCTGAAAATACCCTCCACCCAAGGATGGTCCATATCGTTTGGTGC70 80 90 100 110 120EAR P Y SEP SEP ED S L DY VS VAAGCAAGGCCGTATAGTGAACCTTCAGAGCCAGAAGATAGTTTAGATTATGTTTCTGTTGTTCGTTCCGGCATATCACTTGGAAGTCTCGGTCTTCTATCAAATCTAATACAAAGACAAC130 140 150 160 170 180D K A E ES AT D D DV GE DY SD FMACAAAGCGGAGGAAAGTGCTACAGACGACGATGTCGGGGAGGATTATTCGGATTTTATGATGTTTCGCCTCCTTTCACGATGTCTGCTGCTACAGCCCCTCCTAATAAGCCTAAAATACT190 200 210 220 230 240K EL EMS E ES D #AAGAACTAGAGATGTCAGAGGAATCAGACTGAAATGAGGCTGAAACGGTTTGAATAATTATTCTTGATCTCTACAGTCTCCTTAGTCTGACTTTACTCCGACTTTGCCAAACTTATTAAT250 260 270 280 290 300GGAAAGTATGTTTTTAATAAAGAAATTCTATGTTCAGGAATTTTGTATATACTTTGTAATCCTTTCATACAAAAATTATTTCTTTAAGATACAAGTCCTTAAAACATATATGAAACATTA310 320 330 340 350 360GAATGAGAACTTAGTTGGCTTCAAACTTTTTCGTTTAACATGATTATTTTTCTTGTTCGACTTACTCTTGAATCAACCGAAGTTTGAAAAAGCAAATTGTACTAATAAAAAGAACAAGCT370 380 390 400 410 420132CTAAGATATTCCCACATGGACAAGTGCCACAGATTAATATATGAATACAATAAAATAACTGATTCTATAAGGGTGTACCTGTTCACGGTGTCTAATTATATACTTATGTTATTTTATTGA430 440 450 460 470 480AGTGTAATTTGAMATAAAAGTTGATGAGAGGGACAGGGACTTGAGTCCTATTCAAAGTGTCACATTAAACTTTTATTTTCAACTACTCTCCCTGTCCCTGAACTCAGGATAAGTTTCAC490 500 510 520 530 540MS FL S K L SQ IRKS T TAAAATATTTTAGTTATGAGTTTTCTTTCCAAACTTTCCCAAATACGAAAATCAACGACTGCTTTATAAAATCAATACTCAAAAGAAAGGTTTGAAAGGGTTTATGCTTTTAGTTGCTGACG550 560 570 580 590 600S K A Q V Q D P L P K K N DEE Y S L LATCAAAAGCCCAAGTGCAAGATCCATTACCCAAGAAGAATGACGAAGAGTATTCCTTGTTTAGTTTTCGGGTTCACGTTCTAGGTAATGGGTTCTTCTTACTGCTTCTCATAAGGAACAA610 620 630 640 650 660PstIP K NY I RD ED PA V KR L K EL R RACCCAAAAATTACATAAGAGACGAAGATCCTGCAGTAAAAAGATTGAAGGAGCTGAGGCGTGGGTTTTTAATGTATTCTCTGCTTCTAGGACGTCATTTTTCTAACTTCCTCGACTCCGC670 680 690 700 710 720oLL5Q EL L K N GALA K K S G V KR KR GGCAGGAACTGTTAAAGAATGGTGCTTTGGCTAAAAAAAGTGGTGTAAAACGGAAACGTGGCGTCCTTGACAATTTCTTACCACGAAACCGATTTTTTTCACCACATTTTGCCTTTGCACC730 740 750 760 770 780T S S G SE K K K I ER ND DDE G G LCACCTCATCTGGATCTGAGAAAAAGAAAATAGAAAGGAATGACGATGATGAAGGTGGCCTGTGGAGTAGACCTAGACTCTTTTTCTTTTATCTTTCCTTACTGCTACTACTTCCACCGGA790 800 810 820 830 840oLL6G I R F KR SI GASH A P L K P V V RTGGAATTAGGTTTAAGAGGTCTATTGGAGCAAGTCATGCGCCACTCAAGCCAGTTGTAAGACCTTAATCCAAATTCTCCAGATAACCTCGTTCAGTACGCGGTGAGTTCGGTCAACATTC850 860 870 880 890 900K K PEP 1K KM S FEEL M K Q A ENGAAGAAACCTGAACCTATCAAAAAGATGTCATTTGAAGAGCTAATGAAACAAGCGGAAAACTTCTTTGGACTTGGATAGTTTTTCTACAGTAAACTTCTCGATTACTTTGTTCGCCTTTT910 920 930 940 950 960NE K Q PP K V K S SEP VT K ER PHTAATGAGAAACAGCCCCCAAAAGTTAAGTCATCGGAACCCGTAACTAAGGAACGCCCACAATTACTCTTTGTCGGGGGTTTTCAATTCAGTAGCCTTGGGCATTGATTCCTTGCGGGTGT970 980 990 1000 1010 10200LL7F N K PG F K S S KR P Q K K A S PG ATTTTAACAAGCCAGGTTTCAAAAGTTCAAAGACCACAAAAGAAAGCATCCCCTGGCGCAAAATTGTTCGGTCCAAAGTTTTCAAGTTTTTCTGGTGTTTTCTTTCGTAGGGGACCGCG1030 1040 1050 1060 1070 1080133T L R G VS S G G N S 1K S SD S P K PAACATTGCGTGGAGTATCTTCTGGAGGCAATAGCATAAAATCATCAGACTCACCCAAGCCTTGTAACGCACCTCATAGAAGACCTCCGTTATCGTATTTTAGTAGTCTGAGTGGGTTCGG1090 1100 1110 1120 1130 1140oLL8V K L N L PT N G FAQ P N R R L K E KCGTCAAGCTCAACTTGCCCACAAATGGATTTGCTCAACCTAATAGGAGATTAAAAGAAAAGCAGTTCGAGTTGAACGGGTGTTTACCTAAACGAGTTGGATTATCCTCTAATTTTCTTTT1150 1160 1170 1180 1190 1200XbaIL ES R K Q K SR Y Q D DY DEED NDGTTAGAATCTAGAAAACAGAAATCAAGATACCAGGATGACTATGATGAAGAAGATAACGACAATCTTAGATCTTTTGTCTTTAGTTCTATGGTCCTACTGATACTACTTCTTCTATTGCT1210 1220 1230 1240 1250 1260MD D F I ED D ED E G Y H SK S K H STATGGATGATTTTATAGAAGACGATGAAGATGAAGGTTACCACAGCAAATCGAAACACAGATACCTACTAAAATATCTTCTGCTACTTCTACTTCCAATGGTGTCGTTTAGCTTTGTGTC1270 1280 1290 1300 1310 1320oLL9N G PG Y DR D El WA M F N R G K KRCAATGGTCCCGGATATGATCGTGACGAAATTTGGGCTATGTTCAATAGAGGCAAGAAGCGGTTACCAGGGCCTATACTAGCACTGCTTTAAACCCGATACAAGTTATCTCCGTTCTTCGC1330 1340 1350 1360 1370 1380SE Y DY DEL ED D D MEAN EM ElGTCAGAATACGATTACGATGAGCTTGAGGATGATGATATGGAAGCAAATGAGATGGAAATCAGTCTTATGCTAATGCTACTCGAACTCCTACTACTATACCTTCGTTTACTCTACCTTTA1390 1400 1410 1420 1430 1440Hindl IILEE E EM AR K MARL ED K RE E ACTTGGAAGAGGAGGAAATGGCAAGAAAAATGGCAAGGTTAGAGGATAAACGTGAGGAAGCGAACCTTCTCCTCCTTTACCGTTCTTTTTACCGTTCCAATCTCCTATTTGCACTCCTTCG1450 1460 1470 1480 1490 1500oLL1OW L K K HE E E KR R R K KG I R #TTGGTTAAAAAAGCATGAAGAGGAGAAGAGACGCCGTAAGAAGGGCATACGCTAAGGAATAACCAATTTTTTCGTACTTCTCCTCTTCTCTGCGGCATTCTTCCCGTATGCGATTCCTTA1510 1520 1530 1540 1550 1560ATTGATATATGTTTTGATATATGGACGTGAAATGACTAATGAAGTCGTAGAGAGTTTGGGTAACTATATACAAAACTATATACCTGCACTTTACTGATTACTTCAGCATCTCTCAAACCC1570 1580 1590 1600 1610 1620oLL11AACTGTTTCGAGGCACTGTTTCACTTCTTACATTCATTTTCATACCCTTTGTAATTGCGTTTGACA1AGCTCCGTGACAAAGTGAAGAATGTAAGTAAAAGTATGGGAAACATTAACGCA1630 1640 1650 1660 1670 1680134TTTCCATTTATCCAGTTTGCCTGTCCGATTTTCAAACGTACAGTGATGATATGCATCAGTAAAGGTAAATAGGTCAAACGGACAGGCTAAAAGTTTGCATGTCACTACTATACGTAGTCA1690 1700 1710 1720 1730 1740TGGGTTAGAACATTTATATTGTGTATCGCCCATAATTCTATAAACTTTACTATGTAAAAAACCCAATCTTGTAAATATAACACATAGCGGGTATTAAGATATTTGAAATGATACATTTTT1750 1760 1770 1780 1790 1800TAAAAATGAACCTTCACTATTCTTTCAAGACGGACTGAAAATTTAAAGACTTGGTTGTTGATTTTTACTTGGAAGTGATAAGAAAGTTCTGCCTGACTTTTAAATTTCTGAACCAACAAC1810 1820 1830 1840 1850 1860CAGTTGATCGATTATACAAGACTAACAATTCCAGTATCATTTTTGCCTTAATTTGAGACCGTCAACTAGCTAATATGTTCTGATTGTTAAGGTCATAGTAAAAACGGAATTAAACTCTGG1870 1880 1890 1900 1910 1920TTTTTCAACAAGATTCGCGGCAACGTAGTTGTATTTTTTTTTTTCCACAACCCGTTTCCTAAAAAGTTGTTCTAAGCGCCGTTGCATCAACATAAAAAAAAAAAGGTGTTGGGCAAAGGA1930 1940 1950 1960 1970 1980TACAAAAGCATTCGGAAACTAAACATAAATATGGACCAGCTTTACAAGAGCTATGGTATGATGTTTTCGTAAGCCTTTGATTTGTATTTATACCTGGTCGAAATGTTCTCGATACCATAC1990 2000 2010 2020 2030 2040TTCATATTATTAGGATATATTAGGTGAGATATTAAAAAATGAAACAAATTGTGTCACCAGAAGTATAATAATCCTATATAATCCACTCTATAATTTTTTACTTTGTTTAACACAGTGGTC2050 2060 2070 2080 2090 2100TTAGATAGGATTCAAGTAGTCATTAAAATAGAAACAAGCGTTTAGGGTATGCGTTAAAAGAATCTATCCTAAGTTCATCAGTAATTTTATCTTTGTTCGCAAATCCCATACGCAATTTTC2110 2120 2130 2140 2150 2160AAACTCTAGCAACCTCCAATTGCCAGTGAAAAACTTCCCGAGAAATACTACAACGACAGTTTTGAGATCGTTGGAGGTTAACGGTCACTTTTTGAAGGGCTCTTTATGATGTTGCTGTCA2170 2180 2190 2200 2210 2220GATACATCATACACTTAATAACACTGTAAGGTCCTCAGTTTTTCCAGGTGGAAGGATCAACTATGTAGTATGTGAATTATTGTGACATTCCAGGAGTCAAAAAGGTCCACCTTCCTAGTT2230 2240 2250 2260 2270 2280ATACATACCCTTAATCAATATAAGTAAGTCGAAGGAAGAGATTCGAGCAATGCATTAAAATATGTATGGGAATTAGTTATATTCATTCAGCTTCCTTCTCTAAGCTCGTTACGTAATTTT2290 2300 2310 2320 2330 2340TATGATATTTCAACACATTGACTAAAACGGTTGTATATTCTTAGCCCACTGTGTTGTATCATACTATAAAGTTGTGTAACTGATTTTGCCAACATATAAGAATCGGGTGACACAACATAG2350 2360 2370 2380 2390 2400NdeITCAAAATGAGATACGTCAGTATGACAATACGTCATCCTAAACGTTCATAAAACACATATGAGTTTTACTCTATGCAGTCATACTGTTATGCAGTAGGATTTGCAAGTATTTTGTGTATAC2410 2420 2430 2440 2450 24601355.2 Synthetic peptidesThe primary structures of the synthetic peptides used to raise polyclonal antibodies are given below.Amino acid residues and chemical groups not present in SPT2 are shown in bold.PSi (amino acids 209 to 238)H-Cys-Gly-GIy-Pro-Asn-Arg-Arg-Leu-Lys-Glu-Lys-Leu-Glu-Ser-Arg-Lys-Gln-LysSer-Arg-Tyr-Gln-Asp-Asp-Tyr-Asp-Glu-Glu-Asp-Asn-Asp-Met-Asp-NHPS2 (amino acids 297 to 325)H-Cys-Gly-Gly-Leu-Glu-Glu-Glu-Glu-Met-Ala-Arg-Lys-Met-Ala-Arg-Leu-Glu-AspLys-Arg-Glu-Glu-Ala-Trp-Leu-Lys-Lys-His-Glu-Glu-Glu-Lys-NH5.3 Glossary of gene symbolsacp Acidic proteinadr Alcohol dehydrogenase regulation defectivears Autonomously replicating sequencecan Canavamne resistancecen Centromerecre Carbon catabolite repression effectorcyc Cytochrome c deficiencygal Galactose non-utilizergcn General control of amino acid synthesis non-depressiblehht Histonehir Histone cell cycle regulation defectivehis Histidine requiringhml Mating type cassette - lefthmr Mating type cassette - rightho Homothallic switchinghpc Histone promoter controlhta Histone 2A geneshtb Histone 2B genesleu Leucine requiringlys Lysine requiringrad Radiation sensitiverpb RNA polymerase IIrpo RNA polymerase II, III, IVsin Switch independentsdi SWI dependence inducersir Silent mating type information regulationsit Suppression of initiation of transcriptionsnf Sucrose nonfermentingspt Suppressor of Tysrb Suppressor of rpb]ssn Suppressor of snflsuc Sucrose fermentationsup Suppression of nonsense mutationssuE Homothallic switching deficientTy Transposable element (Transposon yeast)lye Ty-mediated expressionura Uracil requiring1365.4 Genetic equivalenciesSJN2 = HHT]SIN3 SDI] = RPDJ = UME4SJN4 = TSF3SIN6 = nonsense suppressor tRNA geneSWI] =ADR6SWI2 = SNF2 = TYE3.SNF5 = TYE4 = SWI] 0 (?)SPT] = HIR2SPT2 = SIN]SPT6 = SSN2O = CRE2SPT]O = CRE]SPT]] =HTA]SPT]2 = HTB]SPT]3 = GAL]]SPT]5 codes for TFIIDSPT]6 = CDC68HIR3 = HPC]1376. BIBLIOGRAPHY1. Abrams,E., Neigeborn,L., and Carlson,M. (1986). Molecular analysis of SNF2 and SNF5, genes requiredfor expression of glucose-repressible genes in S cerevisiae. Mol. Cell. Biol. 6:3643-3651.2. Adams,C.C. and Workman,J.L. (1993). Nucleosome displacement in transcription. Cell 72:305-308.3. Allison,L.A. and Ingles,C.J. (1989). Mutations in RNA polymerase II enhance or suppress mutations inGAL4. Proc. NatI. A cad. Sci. USA 86:2794-2798.4. Allison,L.A., Moyle,M., Shales,M., and Ingles,C.J. (1985). Extensive homology among the largestsubunits of eukaryotic and prokaryotic RNA polymerases. Cell 42:599-610.5. Allison,L.A., Wong,J.K.-C., Fitzpatrick,V.D., Moyle,M., and Ingles,J. (1988). The C-terminal domain ofthe largest subunit of RNA polymerase II of Saccharomyces cerevisiae, Drosophila melanogaster, andmanunals: a conserved structure with an essential function. Mol. Cell. Biol. 8:321-329.6. Andrews,BJ. and Herskowitz,I. (1989). Identification of a DNA binding factor involved in cell-cyclecontrol of the yeast HO gene. Cell 57:21-29.7. Andrews,B.J. and Herskowitz,I. (1989). The yeast SWI4 protein contains a motif present indevelopmental regulators and is part of a complex involved in cell-cycle-dependent transcription. Nature342:830-833.8. Arndt,K.T. and Fink,G.R. (1986). GCN4 protein, a positive transcription factor in yeast, binds generalcontrol promoters at all 5’ TGACTC 3’ sequences. Proc. Nath Acad. Sci. USA 83:8516-8520.9. Arndt,K.M., Ricupero,S.L., Eisenmann,D.M., and Winston,F. (1992). Biochemical and geneticcharacterization of a yeast TFIID mutant that alters transcription in vivo and DNA binding in vitro. Mol.CelL Biol. 12:2372-2382.10. Arndt,K.T., Styles,C.A., and Fink,G.R. (1987). Multiple global regulators control HJS4 transcription inyeast. Science 237:874-880.11. Arndt,K.T., Styles,C.A., and Fink,G.R. (1989). A suppressor of a HIS4 transcriptional defect encodes aprotein with homology to the catalytic subunit of protein phosphatases. Cell 56:527-537.12. Atkinson,T. and Smith,M. (1984). Solid-phase synthesis of oligodeoxyribonucleotides by the phosphitetriester method. In Oligonucleotide synthesis. A practical approach. M.J.Gait (Ed.), IRL Press, Oxford.pp. 35-81.13813. Ausubel,F.M. eta! (Eds.) (1991). Current protocols in molecular biology. John Wiley & Sons, Inc. N.Y.14. Balasubrarnanian,B., Lowry,C.V., and Zitomer,R.S. (1993). The Roxi repressor of the Saccharornycescerevisiae hypoxic genes is a specific DNA-binding protein with a high-mobility-group motif Mol. Cell.Biol. 13:6071-6078.15. Baldari,C. and Cesarem,G. (1985). Plasmids pEMBLY: new single-stranded shuttle vectors for therecovery and analysis of yeast DNA sequences. Gene 35:27-32.16. Becker,D.M., and Guarente,L. (1991). High-efficiency transformation of yeast by electroporation.Methods Enzymol. 194:182-187.17. Bernues,J., Espel,E., and Querol,E. (1986). Identification of the core-histone-binding domains ofHMG1and HMG2. Biochim. Biophys. Acta 866:242-25 1.18. Bianchi,M.E., Beltrame,M., and Paonessa,G. (1989). Specific recognition of cruciform DNA by nuclearprotein HMG1. Science 243: 1056-1059.19. Bianchi,M.E., Falciola,L., Ferrari,S., and Lilley,D.M.J. (1992). The DNA binding site of HMG1 protein iscomposed of two similar segments (HMG boxes), both of which have counterparts in other eukaryoticregulatory proteins. EMBO .L 11:1055-1063.20. Boeke,J.D. (1989). Transposable elements in Saccharomyces cerevisiae. In Mobile DNA. D.E.Berg andM.M.Howe Eds., American Society for Microbiology, Washington, D. C. pp. 335-374.21. Boeke,J.D., Garfinkel,D.J., Styles,C.A., and Fink,G.R. (1985). Ty elements transpose through an RNAintermediate. Cell 40:491-500.22. Boeke,J.D. and Sandmeyer,S.B. (1991). Yeast transposable elements. In The molecular biology oftheyeast Saccharomyces. Genome dynamics, protein synthesis, and energetics. J.B.Broach, J.R.Pringle, andE.W.Jones Eds., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. pp. 193-261.23. Boeke,J.D., Styles,C.A., and Fink,G.R. (1986). Saccharomyces cerevisiae SPT3 gene is required fortransposition and transpositional recombination of chromosomal Ty elements. Mo!. Cell. Biol. 6:3 575-3581.24. Boeke,J.D., Trueheart,J., Natsoulis,G., and Fink,G.R. (1987). 5-fluoroorotic acid as a selective agent inyeast molecular genetics. Methods Enzymo!. 154:164-175.25. Bonne-Andrea,C., Harper,F., Sobczak,J., and De Recondo,A.-M. (1984). Rat liver HMG1: a physiologicalnucleosome assembly factor. EMBO J. 3:1193-1199.13926. Botstein,D. and Davis,R.W. (1982). Principles and practice of recombinant DNA research with yeast. InThe molecular biology ofthe yeast Saccharomyces. Metabolism and gene expression. J.N. Strathern,E.W.Jones, and J.B.Broach Eds., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. pp.607-636.27. Botstein,D., Falco,S.C., Stewart,S.E., Brennan,M., Scherer,S., Stinchcomb,D.T., Struhl,K., andDavis,R.W. (1979). Sterile host yeast (SHY): a eukaryotic system of biological containment forrecombinant DNA experiments. Gene 8:17-24.28. Breeden,L. and Mikesell,G.E. (1991). Cell cycle-specific expression of the SWI4 transcription factor isrequired for the cell cycle regulation of HO transcription. Genes & Dev. 5:1183-1190.29. Breeden,L. and Nasmyth,K. (1987). Cell cycle control of the yeast HO gene: Cis- and trans-actingregulators. Cell 48:389-3 97.30. Breeden,L. and Nasmyth,K. (1987). Similarity between cell-cycle genes of budding yeast and fission yeastand the Notch gene of Drosophila. Nature 329:651-654.31. Brown,S.J., Kellett,P.J., and Lippard,S.J. (1993). Ixrl, a yeast protein that binds to platinated DNA andconfers sensitivity to cisplatin. Science 261:603-605.32. Bruhn,S.L., Pil,P.M., Essigmann,J.M., Housman,D.E,, and Lippard,S.J. (1992). Isolation andcharacterization of human cDNA clones encoding a high mobility group box protein that recognizesstructural distorsions to DNA caused by binding of the anticancer agent cisplatin. Proc. Nati. Acad. Sci.USA 89:2307-2311.33. Cameron,J.R., Loh,E.Y., and Davis,R.W. (1979). Evidence for transposition of dispersed repetitive DNAfamilies in yeast. Cell 16:739-751.34. ChaleflD.T. and Fink,G.R. (1980), Genetic events associated with an insertion mutation in yeast. Cell21:227-237.35. Chen,S., West,R.W.Jr., Johnson,S.L., Gans,H., Kruger,B., and Ma,J. (1993). TSF3, a global regulatoryprotein that silences transcription of yeast GAL genes, also mediates repression by c2 repressor and isidentical to SIN4. Mol. Cell. Biol. 13:831-840.36. Cigan,A.M., Pabich,E.K., and Donahue,T.F. (1988). Mutational analysis of the HJS4 translationalinitiator region in Saccharomyces cerevisiae. Mol. Cell. Biol. 8:2964-2975.37. Ciriacy,M. (1976). Cis-dominant regulatory mutations affecting the fonnation of glucose-repressiblealcohol dehydrogenase (ADHII) in Saccharomyces cerevisiae. Mol. Gen. Genet. 145:327-3 33.14038. Ciriacy,M. (1979). Isolation and characterization of further cis- and trans-acting regulatory elementsinvolved in the synthesis of glucose-repressible alcohol dehydrogenase (ADHII) in Saccharoinycescerevisiae. Mol. Gen. Genet. 176:427-431.39. Ciriacy,M., Freidel,K., and Lohning,C. (1991). Characterization of trans-acting mutations affecting Tyand Ty-mediated transcription in Saccharomyces cerevisiae. Curr. Genet. 20:44 1-448.40. Clare,J. and Farabaugh,P. (1985). Nucleotide sequence of a yeast Ty element: Evidence for an unusualmechanism of gene expression. Proc. Nati. A cad. Sd. USA 82:2829-2833.41. Clark-Lewis,I., Moser,B., Walz,A., Baggiolini,M., Scott,G.J., and Aebersold,R (1991). Chemicalsynthesis, purification, and characterization of two inflammatory proteins, neutrophil activating peptide 1(interleukin-8) and neutrophil activating peptide 2. Biochemistry 30:3128-3135.42. Clark-Adams,C.D., Norris,D., Osley,M.A., Fassler,J.S., and Winston,F. (1988). Changes in histone genedosage alter transcription in yeast. Genes & Dcv. 2:150-159.43. Clark-Adams,C.D. and Winston,F. (1987). The SPT6 gene is essential for growth and is required for -mediated transcription in Saccharomyces cerevisiae. Mol. Cell. Biol. 7:679-686.44. Coney,L.R. and Roeder,G.S. (1988). Control of yeast gene expression by transposable elements:Maximum expression requires a functional Ty activator sequence and a defective Ty promoter. Mol. Cell.Biol. 8:4009-4017.45. Corden,J.L. (1990). Tails of RNA polymerase II. TJBS 15:383-387.46. Corden,J.L., Cadena,D.L., Ahearn,J.M.Jr., and Dahmus, M.E. (1985). A unique structure at the carboxyltenninus of the largest subunit of eukaryotic RNA polymerase II. Proc. Nati. A cad. Sd. USA 82:7934-7938.47. Crowe,J.S., Cooper,H.J., Smith,M.A., Sims,M.J., Parker,D., and Gewert,D. (1991). Improved cloningefficiency of polymerase chain reaction (PCR) products after proteinase K digestion. Nucleic Acids Res.19:184.48. Denis,C.L. and Malvar,T. (1990). The CCR4 gene from Saccharomyces cerevisiae is required for bothnonfermentative and spt-mediated gene expression. Genetics 124:283-291.49. Dente,L., Cesareni,G., and Cortese,R. (1983). pEMBL: A new family of single stranded plasmids. NucleicAcids Res. 11:1645-1655.50. Devlin,C., Tice-Baldwin,K., Shore,D., and Arndt,K.T. (1991). RAPI is required for BAS1/BAS2- andGCN4-dependent transcription of the yeast HIS4 gene. Mol. Cell. Biol. 11:3642-3651.14151. Diifiey,J.F.X. and Stillman,B. (1991). A close relative of the nuclear, chromosomal high-mobility groupprotein IIMG1 in yeast mitochondria. Proc. Nat!. A cad. Sd. USA 88:7864-7868.52. Diffiey,J.F.X. and Stillman,B. (1992). DNA binding properties of an HMG1-related protein from yeastmitochondria. J Blot. Chein. 267:3368-3374.53. Dirick,L., Moll,T., Auer,H., and Nasmyth,K. (1992). A central role for SWI6 in modulating cell cyclestart-specific transcription in yeast. Nature 357:508-5 13.54. Donahue,T.F., Daves,R.S., Lucchini,G., and Fink,G.R. (1983). A short nucleotide sequence required forthe regulation of HJS4 by the general control system of yeast. Ce!! 32:89-98.55. Donahue,T.F., Farabaugh,P.J., and Fink,G. (1982). The nucleotide sequence of the HJS4 region of yeast.Gene 18:47-59.56. Dubois,E., Jacobs,E., and Jauniaux,J.-C. (1982). Expression of the ROAM mutations in Saccharornycescerevisiae: involvement of frans-acting regulatory elements and relation with the Tyl transcription.E!vIBOJ 9:1133-1139.57. Eibel,H., Gather,J., Stotz,A., and Philippsen,P. (1980). Characterization of the yeast mobile element Ty].Co!d Spring Harbor Symp. Quant. Bio!. 45:609-617.58. Eibel,H. and Philippsen,P. (1984). Preferential integration of yeast transposable element Ty into apromoter region. Nature 307:386-388.59. Eisenmann,D.M., Arndt,K.M., Ricupero,S.L., Rooney,J.W., and Winston,F. (1992). SPT3 interacts withTFIID to allow normal transcription in Saccharomyces cerevisiae. Genes & Dev. 6:1319-1331.60. Eisenmann,D.M., Dollard,C., and Winston,F. (1989). SPTJ5, the gene encoding the yeast TATA bindingfactor TFIID, is required for normal transcription initiation in vivo. Ce!! 58:1183-1191.61. Elder,R.T., Loh,E.Y., and Davis,R.W. (1983). RNA from the yeast transposable element Ty] has bothends in the direct repeats, a structure similar to retrovirus RNA. Proc. Nati. Acad. Sd. USA 80:2432-2436.62. Errede,B., Cardillo,T.S,, Sherman,F., Dubois,E., Deschamps,J., and Wiame,J.-M. (1980). Mating signalscontrol expression of mutations resulting from insertion of a transposable repetitive element adjacent todiverse yeast genes. Cell 25:427-436.63. Errede,B., Cardillo,T.S., Wever,G., and Shennan,F. (1980). Studies on transposable elements in yeast.ROAM mutations causing increased expression of yeast genes: their activation by signals directed towardconjugation functions and their formation by insertion of Tyl repetitive elements. Cold Spring HarborSymp. Quant. BioL 45:593-602.14264. Errede,B., Company,M., Ferchak,J.D., Hutchison III,C.A., and Yarnell,W.S. (1985). Activation regions ina yeast transposon have homology to mating type control sequences and to mammalian enhancers. Proc.Nat!. A cad. Sd. USA 82:5423-5427.65. Estruch,F. and Carlson,M. (1990). SNF6 encodes a nuclear protein that is required for expression of manygenes in Saccharomyces cerevisiae. Mo!. Ce!!. Bio!. 10:2544-2553.66. Farabaugh,P.J. and Fink,G.R. (1980). Insertion of the eukaryotic transposable element Tyl creates a 5-base pair duplication. Nature 286:352-356.67. Farabaugh,P., Liao,X.-B., Belcourt,M., Zhao,H., Kapakos,J., and Clare,J. (1989). Enhancer andsilencerlike sites within the transcribed portion of a Ty2 transposable element of Saccharomycescerevisiae. Mo!. Ce!!. Biol. 9:4824-4834.68. Fassler,J.S. and Winston,F. (1988). Isolation and analysis of a novel class of suppressor of Ty insertionmutations in Saccharomyces cerevisiae. Genetics 118:203-212.69. Feaver,W.J., Gileadi,O., Li,Y., and Kornberg,R.D. (1991). CTD kinase associated with yeast RNApolymerase II initiation factor b. Cell 67:1223-1230.70. Feaver,W.J. and Pearlman,R.E. (1990). Blockage to exonuclease III digestion in the chromatin ofSaccharomyces cerevisiae maps to the in vitro-determined binding site of a trans-acting regulatory factor.Current Genetics 18:17-22.71. Felsenfeld,G. (1992). Chromatin as an essential part of the transcriptional mechanism. Nature 355:219-224.72. Ferrari,S., Harley,V.R., Pontiggia,A., Goodfellow,P.N., Lovell-Badge,R., and Bianchi,M.E. (1992). SRY,like HN4G1, recognizes sharp angles in DNA. ElvIBO J 11:4497-4506.73. Fink,G.R., Farabaugh,P.J., Roeder,G.S., and ChaleffD.T. (1980). Transposable elements (Ty) in yeast.Cold Spring Harbor Symp. Quant. Bio!. 45:575-580.74. Fleer,R., Nicolet,C.M., Pure,G.A., and Friedberg,E.C. (1987). RAD4 gene of Saccharomyces cerevisiae:Molecular cloning and partial characterization of a gene that is inactivated in Escherichia co!i. Mol. Cell.Bio!. 7:1180-1192.75. Fulton,A.M., Rathjen,P.D., Kingsman,S.M., and Kingsman,A.J. (1988). Upstream and downstreamtranscriptional control signals in the yeast retrotransposon, TY. Nucleic Acids Res. 16:5439-5458.76. Gafher,J. and Philippsen,P. (1980). The yeast transposon Ty] generates duplications of target DNA oninsertion. Nature 286:414-418.14377. Garfinkel,D.J., Boeke,J.D., and Fink,G.R. (1985). Ty element transposition: Reverse transcriptase andvirus-like particles. Cell 42:507-517.78. Garnier,J., Osguthorpe,D.J., and Robson,B. (197$). Analysis of the accuracy and implications of simplemethods for predicting the secondary structure of globular proteins. I Mo!. Biol. 120:97-120.79. Gelfand,D.H. and White,T.J. (1990). Thermostable DNA polymerases. In PCR protocols. A guide tomethods and applications. M.A.Innis, D.H.Gelfand,J.J.Sninsky and T.J.White Eds., Academic Press, SanDiego, CA. pp.129-141.80. Gietz,R.D. and Satya Prakash. (1988). Cloning and nucleotide sequence analysis of the Saccharonzycescerevisiae RAD4 gene required for excision repair of UV-damaged DNA. Gene 74:535-541.81. Gietz,R.D. and Sugino,A. (1988). New yeast-Escherichia coli shuttle vectors constructed with in vitromutagenized yeast genes lacking six-base pair restriction sites. Gene 74:527-534.82. Goel,A. and Pearlman,R.E. (1988). Transposable element-mediated enhancement of gene expression inSaccharomyces cerevisiae involves sequence-specific binding of a trans-acting factor. Mo!. Cell. Biol.8:2572-2580.83. Greenblatt,J. (1991). Roles of TFIID in transcriptional initiation by RNA polymerase II. Cell 66:1067-1070.84. Grunstein,M. (1990). Nucleosomes: regulators of transcription. Trends Genet. 6:395-400.85. Guarente,L. (1987). Regulatory proteins in yeast. Ann. Rev. Genet. 21:425-452.86. Guarente,L. and Ptashne,M. (1981). Fusion of Escherichia coli !acZ to the cytochrome c gene ofSaccharomyces cerevisiae. Proc. Nati. Acad. Sci. USA 78:2199-2203.87. Guthrie,C. and Fink,G.R. (Eds.) (1991). Guide to yeast genetics and molecular biology. Methods.Enzymol. 194.88. Haber,J.E. and Garvik,B. (1977). A new gene affecting the efficiency of mating type interconversions inhomothallic strains of S. cerevisiae. Genetics 87:33-50.89. Haggren,W. and Kolodrubetz,D. (1988). The Saccharomyces cerevisiae ACP2 gene encodes an essentialHMG1-like protein. Mo!. Cell. BioL 8:1282-1289.90. Hahn,S., Buratowski,S., Sharp,P.A., and Guarente,L. (1989). Isolation of the gene encoding the yeastTATA binding protein TFIID: A gene identical to the SPT]5 suppressor of Ty element insertions. Cell58:1173-1181.14491. Hahn,S., Buratowski,S., Sharp,P.A., and Guarente,L. (1989). Yeast TATA-binding protein TFIID bindsto TATA elements with both consensus and nonconsensus DNA sequences. Proc. Nat!. A cad. Sd. USA86:5718-5722.92. Han,M., Chang,M., Kim,U.-J., and Grunstein,M. (1987). Histone H2B repression causes cell-cycle-specific arrest in yeast: effects on chromosomal segregation, replication, and transcription. Cell 48:5 89-597.93. Happel,A.M., Swanson,M.S., and Winston,F. (1991). The SNF2, SNF5 and SNF6 genes are required forTy transcription in Saccharomyces cerevisiae. Genetics 128:69-77.94. Harlow,E. and Lane,D. (1988). Antibodies. A laboratory manual. Cold Spring Harbor Laboratory, ColdSpring Harbor, N.Y.95. Hauber,J., Nelboeck-Hochstetter,P., and Feldmann,H. (1985). Nucleotide sequence and characteristics of aTy element from yeast. Nucleic Acids Res. 13:2745-2758.96. Hekmatpanah,D.S. and Young,R.A. (1991). Mutations in a conserved region of RNA polymerase IIinfluence the accuracy of mRNA start site selection. Mol. CelL Biol. 11:5781-5791.97. Henry,S., Donahue,T., and Culbertson,M. (1975). Selection of spontaneous mutants by inositol starvation.Mo!. Gen. Genet. 143:5-11.98. Hereford,L., Fahrner,K., Woolford Jr,J., Rosbash,M., and Kaback,D.B. (1979). Isolation of yeast histonegenes H2A and H2B. Cell 18:1261-1271.99. Herskowitz,I. (1987). Functional inactivation of genes by dominant negative mutations. Nature 329:219-222.100. Herskowitz,I. (1988). Life cycle of the budding yeast Saccharomyces cerevisiae. Microbiol.Rev. 52:536-553.101. Herskowitz,I. (1989). A regulatory hierarchy for cell specialization in yeast. Nature 342:749-757.102. Higgins,D.G. and Sharp,P. M.. (1988). CLUSTAL: a package for performing multiple sequencealignment on a microcomputer. Gene 73:237-244.103. Hinnebusch,A.G. (1988). Mechanisms of gene regulation in the general control of amino acid biosynthesisin Saccharomyces cerevisiae. Microbiol. Rev. 52:248-276.104. Hinnebusch,A.G., Lucchini,G., and Fink,G.R. (1985). A synthetic HIS4 regulatory element confersgeneral amino acid control on the cytochrome c gene (CYC]) of yeast. Proc. Natl. Acad. Sci. USA 82:498-502.145105. Hinnen,A., }{icks,J.B., and Fink,G.R. (1978). Transformation of yeast. Proc. NatL A cad. Sd. USA75:1929-1933.106. Hirschhorn,J.N., Brown,S.A., Clark,C.D., and Winston,F. (1992). Evidence that SNF2/SWI2 and SNF5activate transcription in yeast by altering chromatin structure. Genes & Dev. 6:2288-2298.107. Hirschhorn,J.N. and Winston,F. (1988). SPT3 is required for normal levels of a-factor and x-factorexpression in Saccharomyces cerevisiae. Mo!. Cell. BioL 8:822-827.108. Hirschman,J.E., Durbin,K.J., and Winston,F. (1988). Genetic evidence for promoter competition inSaccharornyces cerevisiae. Mo!. Cell. Biol. 8:4608-4615.109. Hisatake,K., Hasegawa,S., Takada,R., Nakatani,Y., Horikoshi,M., and Roeder,R.G. (1993). The p250subunit of native TATA box-binding factor TFIID is the cell-cycle regulatory protein CCG1. Nature362:179-181.110. Hoffman,C.S. and Winston,F. (1987). A ten-minute DNA preparation from yeast efficiently releasesautonomous plasmids for transformation of Escherichia coli.Gene 57:267-272.111. Horikoshi,M., Wang,C.K., Fujii,H., Cromlish,J.A., Weil,P.A., and Roeder,R.G. (1989). Cloning andstructure of a yeast gene encoding a general transcription initiation factor TFIII) that binds to the TATAbox. Nature 341:299-303.112. Horn,T. and Urdea,M. S. (1986). A chemical 5-phosphorylation of oligodeoxyribonucleotides that can bemonitored by trityl cation release. Tetrahedron Lett. 27:4705-4708.113. Horvitz,H.R. and Herskowitz,I. (1992). Mechanisms of asymmetric cell division: Two Bs or not two Bs,that is the question. Ce!! 68:237-255.114. Ingles,C.J., Himmelfarb,H.J., Shales,M., Greenleaf,A.L., and Friesen, J.D. (1984). Identification,molecular cloning, and mutagenesis of Saccharomyces cerevisiae RNA polymerase genes. Proc. Nat!.Acad. Sd. USA 81:2157-2161.115. Isackson,P.J., Fishback,J.L., Bidney,D.L., and Reeck,G.R. (1979). Preferential affinity of high molecularweight high mobility group non-histone proteins for single-stranded DNA, I Biol. Chern. 254:5569-5572.116. Jantzen,H.-M., Admon,A., Bell,S.P., and Tjian,R. (1990). Nucleolar transcription factor hUBF contains aDNA-binding motif with homology to HMG proteins. Nature 344:830-836.117. Jantzen,H.-M., Chow,A.M., King,D.S., and R. Tjian. (1992). Multiple domains of the RNA polymerase Iactivator hIJBF interact with the TATA-binding protein complex hSL1 to mediate transcription. Genes &Dev. 6:1950-1963.146118. Jerome,J.F. and Jaehning,J.A. (1986). mRNA transcription in nuclei isolated from Saccharornycescerevisiae.Mol. Cell. Blot. 6:1633-1639.119. Jiang,Y.W. and Stillman,D.J. (1992). Involvement of the SIN4 global transcriptional regulator in thechromatin structure of Saccharomyces cerevisiae. Mo!. Cell. Blot. 12:4503-45 14.120. Johns,E.W. (ed.). (1982). The HMG chromosomal proteins. Academic Press, New York.121. Johnson,L.M., Kayne,P.S., Kahn,E.S., and Grunstein,M. (1990). Genetic evidence for an interactionbetween SIR3 and histone H4 in the repression of the silent mating loci in Saccharomyces cerevisiae.Proc. Nath Acad. Sd. USA 87:6286-6290.122. Johnston,M. and Davis,R.W. (1984). Sequences that regulate the divergent GAL1-GALJO promoter inSaccharornyces cerevisiae. Mo!. Cell. Blot. 4:1440-1448.123. Kelly,M., Burke,J., Smith,M., Klar,A., and Beach,D. (1988). Four mating-type genes control sexualdifferentiation in the fision yeast. EMBO 1 7:1537-1547.124. Kim,U.-J., Han,M., Kayne,P., and Grunstein,M. (1988). Effects of histone H4 depletion on the cell cycleand transcription of Saccharomyces cerevisiae. EMBO 1 7:22 11-2219.124a. Kolodrubetz,D. (1993). Personnal conunumcation.125. Kolodrubetz,D. and Burgum,A. (1990). Duplicated NHP6 genes of Saccharomyces cerevisiae encodeproteins homologous to bovine high mobility group protein 1. 1 Blot. Chem. 265:3234-3239.126. Kornberg,R.D. and Lorch,Y. (1991). Irresistible force meets immovable object: transcription and thenucleosome. Cell 67:833-836.127. Kruger,W. and Herskowitz,I. (1991). A negative regulator of HO transcription, S1N1 (SPT2), is anonspecific DNA-binding protein related to HMG1. Mo!. Cell. Blot. 11:4135-4146.128. Kruger,W. and Herskowitz,I. (1993). Personnal communication.129. Kunkel,T.A. (1985). Rapid and efficient site-specific mutagenesis without phenotypic selection. Proc.Nat!. A cad. Sd. USA 82:488-492.130. Kunkel,T.A., Roberts,J.D., and Zakour,R.A. (1987). Rapid and efficient site-specific mutagenesis withoutphenotypic selection. Methods Enzymol. 154:367-382.131. Laurent,B.C. and Carlson,M. (1992). Yeast SNF2/SWI2, SNF5, and SNF6 proteins function coordinatelywith the gene-specific transcriptional activators GAL4 and Bicoid. Genes & Dev. 6:1707-1715.147132. Laurent,B.C., Treich,I., and Carlson,M. (1993). The yeast SNF2/SWT2 protein has DNA-stimulatedATPase activity required for transcriptional activation. Genes & Dev. 7:583-591.133. Laurent,B.C., Treitel,M.A., and Carlson,M. (1990). The SNF5 protein of Saccharomyces cerevisiae is aglutamine- and proline-rich transcriptional activator that affects expression of a broad spectrum of genes.Mol. Cell. Biol. 10:5616-5625.134. Laurent,B.C., Treitel,M.A., and Carlson,M. (1991). Functional interdependence of the yeast SNF2, SNF5,and SNF6 proteins in transcriptional activation. Proc. NatI. Acad. Sd. USA 88:2687-2691.135. Lemoine,Y., Dubois,E., and Wiame,J.-M. (1978). The regulation of urea amidolyase of Saccharomycescerevisiae. Mating type influence on a constitutivity mutation acting in cis. Mol. Gen. Genet. 166:251-258.136. Liao,S.-M., Taylor,I.C.A., Kingston,R.E., and Young,R.A. (1991). RNA polymerase II carboxy-terminaldomain contributes to response to multiple acidic activators in vitro. Genes & Dev. 5:2431-2440.137. Liao,X.-B., Clare,J.J., and Farabaugh,P.J. (1987). The upstream activation site of a Ty2 element of yeastis necessary but not sufficient to promote maximal transcription of the element. Proc. Nati. A cacL Sd.USA 84:8520-8524.138. Little,J.W. and Mount,D.W. (1982). The SOS regulatory system of Escherichia coli. Cell 29:11-22.139. Liu,H., Krizek,J., and Bretscher,A. (1992). Construction of a GALl-regulated yeast cDNA expressionlibrary and its application to the identification of genes whose overexpression causes lethality in yeast.Genetics 132:665-673.140. Lu,H., Zawel,L., Fisher,L., Egly,J.-M., and Reinberg,D. (1992). Human general transcription factor ITHphosphorylates the C-terminal domain of RNA polymerase II. Nature 358:641-645.141. McCabe,P.C. (1990). Production of single-stranded DNA by asymmetric PCR. In PCR protocols. A guideto methods and applications. M.A.Innis, D.H.Gelfand, J.J.Sninsky, and T.J.White Eds., Academic Press,San Diego, CA. pp.76-83.142. Mclntosh,E.M., Atkinson,T., Storms, R.K., and Smith,M. (1991). Characterization of a short, cis-actingDNA sequence which conveys cell cycle stage-dependent transcription in Saccharomyces cerevisiae. Mol.Cell. Biol. 11:329-337.143. McNeil,J.B. and Smith,M. (1986). Transcription initiation of the Saccharomyces cerevisiae iso-icytochrome c gene: multiple, independent T-A-T-A sequences. I Mo!. BioL 187:363-378.144. McStay,B., Frazier,M.W., and Reeder,R.H. (1991). xUBF contains a novel dimerization domain essentialfor RNA polymerase I transcription. Genes & Dev. 5:1957-1968.148145. Mehta,K. D., Leung,D., Lefebvre,L., and Smith,M. (1990). TheANBJ locus of Saccharoinyces cerevisiaeencodes the protein synthesis initiation factor elF-4D. J. Blot. Chem. 265:8802-8807.146. Mellor,J., Kingsman,A.J., and Kingsman,S.M. (1986). Ty, an endogenous retrovirus of yeast? Yeast2:145-152.147. Mosnn,C., Riva,M., Beltrame,M., Cassar,E., Sentenac,A., and Thuriaux,P. (1990). The RPC31 gene ofSaccharomyces cerevisiae encodes a subunit of RNA polymerase C (III) with an acidic tail. Mo!. Cell.Blot. 10:4737-4743.148. Muller,H.J. (1932). Further studies on the nature and causes of gene mutations. Proc. 6th mt Congr.Genet. 1:213-255.149. Nagawa,F. and Fink,G.R. (1985). The relationship between the “TATA” sequence and transcriptioninitiation sites at the HIS4 gene of Saccharornyces cerevisiae. Proc. Nat!. A cad. Sd. USA 82:8557-8561.150. Nasmyth,K. (1983). Molecular analysis of a cell lineage. Nature 302:670-676.151. Nasmyth,K. (1985). At least 1400 base pairs of 5’-flanking DNA is required for the correct expression ofthe HO gene in yeast. Cell 42:213-223.152. Nasmyth,K. (1985). A repetitive DNA sequence that confers cell-cycle START (CDC28)-dependenttranscription of the HO gene in yeast. Cell 42:225-23 5.153. Nasmyth,K., Adolf,G., Lydall,D., and Seddon,A. (1990). The identification of a second cell cycle controlon the HO promoter in yeast: Cell cycle regulation of SWI5 nuclear entry. Cell 62:631-647.154. Nasmyth,K., Seddon,A., and Ammerer,G. (1987). Cell cycle regulation of SWI5 is required for mother-cell-specific HO transcription in yeast. Cell 49:549-558.155. Nasmyth,K. and Shore,D. (1987). Transcriptional regulation in the yeast life cycle. Science 237:1162-1170.156. Nasmyth,K., Stillman,D., and Kipling,D. (1987). Both positive and negative regulators of HOtranscription are required for mother-cell-specific mating-type switching in yeast. Cell 48:579-587.157. Natsoulis,G., Dollard,C., Winston,F., and Boeke,J.D. (1991). The products of the SPTJO and SPT2] genesof Saccharomyces cerevisiae increase the amplitude of transcriptional regulation at a large number ofunlinked loci. New Biol. 3:1249-1259.158. Natsoulis,G., Thomas,W., Roghmann,M.-C., Winston,F., and Boeke,J.D. (1989). Ty] transposition inSaccharornyces cerevisiae is nonrandom. Genetics 123:269-279.149159. Needleman,S.B. and Wunsch,C.D. (1970). A general method applicable to the search for similarities inthe amino acid sequence of two proteins. I Mo!. Blot. 48:443-453.160. Neigeborn,L. and Carlson,M. (1984). Genes affecting the regulation of SUC2 gene expression by glucoserepression in Saccharoinyces cerevisiae. Genetics 108:845-856.161. Neigeborn,L., Celenza,J.L., and Carlson,M. (1987), SSN2O is an essential gene with mutant alleles thatsuppress defects in SUC2 transcription in Saccharomyces cerevisiae. Mo!. Cell. Blot. 7:672-678.162. Neigeborn,L., Rubin,K., and Carlson,M. (1986). Suppressors of snJ2 mutations restore invertasederepression and cause temperature-sensitive lethality in yeast. Genetics 112:741-753.163. Ner,S.S. (1992). HMGs everywhere. Current Blot. 2:208-210.164. Ner,S.S., Goodin,D.B., Pielak,G.J., and Smith,M. (1988). A rapid droplet method for Sanger dideoxysequencing. BioTechniques 6:408-412.165. Nonet,M., Scafe,C., Sexton,J., and Young,R. (1987). Eucaryotic RNA polymerase conditional mutant thatrapidly ceases mRNA synthesis. MoL Cell. Biol. 7: 1602-1611.166. Nonet,M., Sweetser,D., and Young,RA. (1987). Functional redundancy and structural polymorphism inthe large subunit of RNA polymerase II. Cell 50:909-915.167. Nonet,M.L. and Young,R.A. (1989). Intragenic and extragenic suppressors of mutations in theheptapeptide repeat domain of Saccharornyces cerevisiae RNA polymerase II. Genetics 123:715-724.168. Ogas,J., Andrews,B.J., and Herskowitz,I. (1991). Transcriptional activation of CLNJ, CLN2, and aputative new Gi cyclin (HCS26) by SWI4, a positive regulator of G 1-specific transcription. Cell 66:1015-1026.169. O’hara,P.J., Horowitz,H., Eichinger,G., and Young,E.T. (1988). The yeastADR6 gene encodeshomopolymeric amino acid sequences and a potential metal-binding domain. Nucleic Acids Res.21:10153-10169.170. Osley,M.A. (1991). The regulation of histone synthesis in the cell cycle. Ann. Rev. Biochem. 60:827-86 1.171. Osley,M.A. and Lycan,D. (1987). trans-acting regulatory mutations that alter transcription ofSaccharomyces cerevisiae histone genes. Mol. Cell. Blot. 7:4204-4210.171a. Park,E.-C. and Szostak,J.W. (1990). Point mutations in the yeast histone H4 gene prevent silencing of thesilent mating type locus HML. Mol. Cell. Biol. 10:4932-4934.150172. Pellman,D., McLaughlin,M.E., and Fink,G.R. (1990). TATA-dependent and TATA-independenttranscription at HJS4 gene of yeast. Nature 348:82-85.173. Peterson,C.L. (1993). Personnal communication.174. Peterson,C.L. and Herskowitz,I. (1992). Characterization of the yeast SWI], SWI2 and SWJ3 genes, whichencode a global activator of transcription. Cell 68:573-583.175. Peterson,C.L., Kruger,W., and Herskowitz,1. (1991). A functional interaction between the C-terminaldomain of RNA polymerase II and the negative regulator SIN1. Cell 64:1135-1143.176. Peterson,M.G. and Tjian,R. (1992). The tell-tail trigger. Nature 358:620-621.177. Primig,M., Sockanathan,S., Auer,H., and Nasmyth,K. (1992). Anatomy of a transcription factor importantfor the start of the cell cycle in Saccharomyces cerevisiae. Nature 358:593-5 97.178. Pugh,B.F. and Tjian,R. (1992). Diverse transcriptional functions of the multisubunit eukaiyotic TFIIDcomplex. I Blot. Qiem. 267:679-682.179. Rine,J. (1991). Gene overexpression in studies of Saccharo,nyces cerevisiae. Methods Enzymol. 194:239-251.180. Roeder,G.S., Beard,C., Smith,M., and Keranen,S. (1985). Isolation and characterization of the SPT2gene, a negative regulator of Ty-controlled yeast gene expression. Mol. Cell. Blot. 5:1543-1553.181. Roeder,G. S., Farabaugh,P.J., Cha1eflD.T., and Fink,G.R. (1980). The origins of gene instability in yeast.Science 209: 1375-1380.182. Roeder,G.S. and Fink,G.R. (1980). DNA rearrangements associated with a transposable element in yeast.Cell 21:23 9-249.183. Roeder,G.S. and Fink,G.R. (1982). Movement of yeast transposable elements by gene conversion. Proc.NatI. Acad. Sd. USA 79:5621-5625.184. Roeder,G.S. and Fink,G.R. (1983). Transposable elements in yeast. In Mobile genetic elements.,J.A.Shapiro Ed. Academic Press, New York. pp. 299-328.185. Roeder,G.S., Rose,A.B., and Pearlman,R.E. (1985). Transposable element sequences involved in theenhancement of yeast gene expression. Proc. Nati. A cad. Sci. USA 82:5428-5432.186. Roeder,G.S., Smith,M., and Lambie,E.J. (1984). Intrachromosomal movement of geneticaly markedSaccharomyces cerevisiae transposons by gene conversion. Mol. Cell. Blot. 4:703-711.151187. Rolfe,M., Spanos,A., and Banks,G. (1986). Induction of yeast Ty element transcription by ultravioletlight. Nature 319:339-340.188. Rose,M. and Winston,F. (1984). Identification of a Ty insertion within the coding sequence of the S.cerevisiae URA3 gene. Mo!. Gen. Genet. 193:557-560.189. Roth,S.Y. (1993). Personnal conummication.190. Roth,S.Y., Shimizu,M., Johnson,L., Grunstein,M., and Simpson,R.T. (1992). Stable nucleosomepositioning and complete repression by the yeast c2 repressor are disrupted by amino-terminal mutationsin histone H4. Genes & Dev. 6:411-425.191. Rothstein,R.J. (1983). One-step gene disruption in yeast. Methods Enzymol. 101:202-211.192. Ruppert,S., Wang,E.H., and Tjian,R. (1993). Cloning and expression of human TAF11250: a TBPassociated factor implicated in cell-cycle regulation. Nature 362:175-179.193. Russel,M., Kidd,S., and Kelley,M.R. (1986). An improved filamentous helper phage for generatingsingle-stranded plasmid DNA. Gene 45:333-338.194. Russell,D.W,, Jensen,R., Zoller,M.J., Burke,J., Errede,B., Smith,M., and Herskowitz,I. (1986). Structureof the Saccharomyces cerevisiae HO gene and analysis of its upstream regulatory region. Mo!. Cell. Biol.6:4281-4294.195. Sambrook,J., Fritsch,E.F., and Maniatis,T. (1989). Molecular cloning: A laboratory manual, 2nd Ed.Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.196. Sawadogo,M. and Sentenac,A. (1990). RNA polymerase B (II) and general transcription factors. Ann.Rev. Biochem. 59:711-754.197. Scafe,C., Chao,D., Lopes,J., Hirsch,J.P., Henry,S., and Young,R.A. (1990). RNA polymerase II C-terminal repeat influences response to transcriptional enhancer signals. Nature 347:491-494.198. Scafe,C., Martin,C., Nonet,M., Podos,S., Okamura,S., and Young,R. (1990). Conditional mutations occurpredominantly in highly conserved residues of RNA polymerase II subunits. Mol. Cell. Biol. 10:1270-1275.199. Scafe,C., Nonet,M., and Young,R. (1990). RNA polymerase II mutants defective in transcription of asubset of genes. Mo!. Cell. Biol. 10:1010-1016.200. Scherer,S. and Davis,R.W. (1980). Recombination of dispersed repeated DNA sequences in yeast. Science209:1380-1384.152201. Scherer,S. and Davis,R.W. (1980). Studies on the transposable element Ty] of yeast. II. Recombinationand expression of Tyl and adjacent sequences. Cold Spring Harbor Symp. Quant. Blot. 45:584-59 1.202. Scherer,S., Mann,C., and Davis,R.W. (1982). Reversion of a promoter deletion in yeast. Nature 298:8 15-819.203. Sekiguchi,T., Nohiro,Y., Nakamura,Y., Hisamoto,N., and Nishimoto,T. (1991). The human CCGJ gene,essential for progression of the G1 phase, encodes a 2 10-kilodalton nuclear DNA-binding protein. Mo!.Cell. Blot. 11:33 17-3325.204. Sharp,P.A. (1992). TATA-binding protein is a classless factor. Cell 68:819-821.205. Sherman,F. (1991). Getting started with yeast. Methods Enzyinol. 194:3-21.206. Sherman,F., Fink,G., and Hicks,J.B. (1986). Methods in yeast genetics. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.207. Sherman,F. and Stewart,J.W. (1982). Mutations altering initiation of translation of yeast iso-1-cytochromec: contrasts between the eukaryotic and prokaryotic initiation process. In The molecular biology of theyeast Saccahromyces. Metabolism and gene expression. J.N.Stratern, E.W.Jones, and J.B.Broach Eds.,Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. pp. 301-333.208. Sherwood,P.W. and Osley,M.A. (1991). Histone regulatory (hir) mutations suppress insertion alleles inSaccharo,nyces cerevisiae. Genetics 128:729-738.209. Silverman,S.J. and Fink,G.R. (1984). Effects of Ty insertions on 11154 transcription in Saccharomycescerevisiae. Mo!. Cell. Blot. 4: 1246-1251.210. Silverman,S.J., Rose,M., Botstein,D., and Fink,G.R. (1982). Regulation of HJS4-lacZ fusions inSaccharomyces cerevisiae. Mo!. Ce!!. BioL 2:1212-1219.211. Simchen,G., Winston,F., Styles,C.A., and Fink,G.R. (1984). Ty-mediated gene expression of the LYS2and HJS4 genes of Saccharomyces cerevisiae is controlled by the same SPT genes. Proc. Nat!. Acad. Sd.USA 81:2431-2434.212. Singh,J. and Dixon,G.H. (1990). High mobility group proteins 1 and 2 function as class II transcriptionfactors. Biochemistry 29:6295-6302.213. Smith,M.M. and Andresson,O.S. (1983). DNA sequences of yeast H3 and H4 histone genes from two nonallelic gene sets encode identical H3 and H4 proteins. .1 Mo!. Bio!. 169:663-690.214. Stern,M., Jensen,R., and Herskowitz,I. (1984). Five SWI genes are required for expression of the HO genein yeast. I Mo!. Biol. 178:853-868.153215. Sternberg,P.W., Stern,M.J., Clark,I., and Herskowitz,I. (1987). Activation of the yeast HO gene by releasefrom multiple negative controls. Cell 48:567-577.216. Stillman,D.J., Bankier,A.T., Seddon,A., Groenhout,E.G., and Nasmyth,K.A. (1988). Characterization of atranscription factor involved in mother cell specific transcription of the yeast HO gene. EMBO 1 7:485-494.217. Stritch,R., Salter,M.R., and Esposito,R.E. (1989). Identification of negative regulatory genes that governthe expression of early meiotic genes in yeast. Proc. Natl. Acad. Sd. USA 86:10018-10022.218. Struhl,K. (1989). Molecular mechanisms of transcriptional regulation in yeast. Ann. Rev. Biochem.58:1051-1077.219. Sugimoto,A., Iino,Y., Maeda,T., Watanabe,Y., and Yamainoto,M. (1991). Schizosaccharoinyces pombestel 1 + encodes a transcription factor with an HMG motif that is a critical regulator of sexualdevelopment. Genes & Dev. 5:1990-1999.220. Swanson,M.S., Carlson,M., and Winston,F. (1990). SPT6, an essential gene that affects transcription inSaccharomyces cerevisiae, encodes a nuclear protein with an extremely acidic amino terminus. Mol. Cell.Biol. 10:4935-4941.221. Swanson,M.S., Malone,E.A., and Winston,F. (1991). SPT5, an essential gene important for normaltranscription in Saccharomyces cerevisiae, encodes an acidic nuclear protein with a carboxy-terminalrepeat. Mol. Cell. Biol. 11:3009-3019.221a. Swanson,M.S. and Winston,F. (1992). SPT4, SPT5 and SPT6 interactions: effects on transcription andviability in Saccharomyces cerevisiae. Genetics 132:325-336.222. Taguchi,A.K.W., Ciriacy,M., and Young,E.T. (1984). Carbon source dependence of transposable element-associated gene activation in Saccharomyces cerevisiae. Mol. Cell. BioL 4:61-68.223. Taguchi,A.K.W. and Young,E.T. (1987). The identification and characterization ofADR6, a generequired for sporulation and for expression of the alcohol dehydrogenase II isozyme from Saccharo,nycescerevisiae. Genetics 116:523-530.224. Tice-Baldwin,K., Fink,G.R., and Arndt,K.T. (1989). BAS1 has a Myb motif and activates HIS4transcription only in combination with BAS2. Science 246:931-935.225. Travers,A.A. (1992). The reprogramming of transcriptional competence. Cell 69:573-575.226. Travis,A., Amsterdam,A., Belanger,C., and Grosschedl,R (1991). LEF-1, a gene encoding a lymphoidspecific protein, with an HMG domain, regulates T-cell receptor x enhancer function. Genes & Dev.5:880-894.154227. Tremethick,D.J. and Molloy,P.L. (1986). High mobility group proteins 1 and 2 stimulate transcription invitro by RNA polymerase II and III. I Biol. Chem. 261:6986-6992.228. Usheva,A., Maldonado,E., Goldring,A., Lu,H., Houbavi,C., Reinberg,D., and Aloni,Y. (1992). Specificinteraction between the nonphosphorylated form of RNA polymerase II and the TATA-bining protein.Cell 69:871-881.229. van Holde,K. (1993). The omnipotent nucleosome. Nature 362:111-1 12.230. van Holde,K.E. (1989). Chrornatin. A. Rich Ed., Springer-Verlag Press, New York.231. Vernet,T., Dignard,D., and Thomas,D.Y. (1987). A family of yeast expression vectors containing thephage fi intergenic region. Gene 52:225-233.232. Vidal,M., Buckley,A.M., Hilger,F., and Gaber,R.F. (1990). Direct selection for mutants with increased K+transport in Saccharornyces cerevisiae. Genetics 125:313-320.233. Vincent,A. and Petes,T. (1986). The effect of mitotic and meiotic gene conversion on the number of Tyelements in yeast. Yeast 2:s408234. Voytas,D.F. and Boeke,J.D. (1992). Yeast retrotransposon revealed. Nature 358:717.235. Wang,H., Clark,I., Nicholson,P.R., Herskowitz,I., and Stillman,D.J. (1990). The Saccharomycescerevisiae SIN3 gene, a negative regulator of HO, contains four paired amphipathic helix motifs. Mol.Cell. Biol. 10:5927-5936.236. Wang,H. and Stillman,D.J. (1990). In vitro regulation of a SIN3-dependent DNA-binding activity bystimulatory and inhibitory factors. Proc. Nat!. Acad. Sci. USA 87:9761-9765.237. Wang,H. and Stillman,D.J. (1993). Transcriptional repression in Saccharomyces cerevisiae by a SIN3-LexA fusion protein. Mo!. CelL Biol. 13:1805-18 14.238. Weir,H.M., Kraulis,P.J., Hill,C.S., Raine,A.R.C., Laue,E.D., and Thomas,J.O. (1993). Structure of the11MG box motif in the B-domain of HMG1. EMBO I 12:1311-1319.239. Wen,L., Huang,J.-K., Johnson,B.H., and Reeck,G.R. (1989). A human placental cDNA clone that encodesnonhzistone chromosomal protein 11MG-i. Nucleic Acids Res. 17:1197-1214.240. Wilke,C.M., Heidler,S.H., Brown,N., and Liebman,S.W. (1989). Analysis of yeast retrotransposon Tyinsertions at the CAN] locus. Genetics 123:655-665.155241. Williamson,V.M., Young,E.T., and Ciriacy,M. (1981). Transposable elements associated with constitutiveexpression of yeast alcohol dehydrogenase II. Cell 23:605-614.242. Winston,F. (1988). Genes that affect Ty-mediated gene expression in yeast. In Eukaryotic transposableelements as mutagenic agents. M.E.Lambert, J.F.McDonald ,and I.B.Weinstein Eds., Cold Spring HarborLaboratory Press, Banbuiy report #30, Cold Spring Harbor, N.Y. pp. 145-153.243. Winston,F. (1993). Personnal communication.244. Winston,F. and Carlson,M. (1992). Yeast SNF/SWI transcriptional activators and the SPT/SIN chromatinconnection. Trends Genet. 8:387-391.245. Winston,F., Chaleff,D.T., Valent,B., and Fink,G.R. (1984). Mutations affecting Ty-mediated expressionof the HIS4 gene of Saccharomyces cerevisiae. Genetics 107:179-197.246. 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 115:649-656.247. Winston,F., Durbin,K.J., and Fink, G.R. (1984). The SPT3 gene is required for normal transcription of Tyelements in S. cerevisiae. Cell 39:675-682.248. Wolfner,M., Yep,D., Messenguy,F., and Fink,G.R. (1975). Integration of amino acid biosynthesis into thecell cycle of Saccharomyces cerevisiae. I Mol. Biol. 96:273-290.249. Woychik,N.A. and Young,R.A. (1989). RNA polymerase II subunit RPB4 is essential for high- and low-temperature yeast cell growth. Mol. Cell. Biol. 9:2854-2859.250. Xu,H. and Boeke,J.D. (1991). Inhibition of Tyl transposition by mating pheromones in Saccharomycescerevisiae. Mol. Cell. Blot. 11:2736-2743.251. Xu,H., Kim,U.-J., Schuster,T., and Grunstein,M. (1992). Identification of a new set of cell cycle-regulatory genes that regulate S-phase transcription of histone genes in Saccharomyces cerevisiae. Mo!.Cell. Biol. 12:5249-5259.252. Yanisch-Perron,C., Vieira,J., and Messing,J. (1985). Improved M13 phage cloning vectors and hoststrains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33:103-119.253. Yoshinaga,S.K., Peterson,C.L., Herskowitz,I., and Yamamoto,K.R. (1992). Roles of SWI1, SWI2, andSWI3 proteins for transcriptional enhancement by sterol receptors. Science 258:1598-1604.254. Young,R.A. (1991). RNA polymerase II. Ann. Rev. Biochem. 60:689-715.156255. Young,R.A. and Davis,R.W. (1983). Yeast RNA polymerase II genes: Isolation with antibody probes.Science 222:778-782.256. Youngren,S.D., Boeke,J.D., Sanders,N.J., and Garfinkel,D.J. (1988). Functional organization of theretrotransposon Ty from Saccharomyces cerevisiae: Ty protease is required for transposition. Mol. Cell.Biol. 8:1421-1431.257. Yu,K. and Elder,R.T. (1989). A region internal to the coding sequences is essential for transcription of theyeast Ty-D15 element. Mo!. Cell. Biol. 9:3667-3678.258. Zehring,W.A., Lee,J.M., Weeks,J.R., Jokerst,R.S., and GreenleafA.L. (1988). The C-tenninal repeatdomain of RNA polymerase II largest subunit is essential in vivo but is not required for accuratetranscription initiation in vifro. Proc. Nath Acad. Sd. USA 85:3698-3702.259. Ziltener,H.J., Clark-Lewis,I., Hood,L.E., Kent,S.B.H., and Schrader,J.W. (1987). Antipeptide antibodiesof predetermined specificity recognize and neutralize the bioactivity of the pan-specific hemopoietininterleukin 3. f Immunol. 138:1099-1104.260. Zoller,M.J. and Smith,M. (1982). Oligonucleotide-directed mutagenesis using M13-derived vectors: Anefficient and general procedure for the production of point mutations in any fragment of DNA. NucleicAcids Res. 10:6487-6500.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items