Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The NADP-dependent glutamate dehydrogenase of Giardia lamblia: a study of function, gene structure, and… Yee, Janet 1993

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

Item Metadata


831-ubc_1993_fall_phd_yee_janet.pdf [ 3.63MB ]
JSON: 831-1.0086426.json
JSON-LD: 831-1.0086426-ld.json
RDF/XML (Pretty): 831-1.0086426-rdf.xml
RDF/JSON: 831-1.0086426-rdf.json
Turtle: 831-1.0086426-turtle.txt
N-Triples: 831-1.0086426-rdf-ntriples.txt
Original Record: 831-1.0086426-source.json
Full Text

Full Text

The NADP-dependent glutamate dehydrogenase of Giardia lamblia:a study of function, gene structure, and expressionbyJanet YeeB.Sc., The University of Toronto, 1983M.Sc., The Queen's University, 1987A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinFACULTY OF GRADUATE STUDIESDEPARTMENT OF BIOCHEMISTRYAND MOLECULAR BIOLOGYWe accept this thesis as conformingto the required standard THE UNIVERSITY OF BRITISH COLUMBIAJuly 1993© Janet Yee, 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 1316o-ft-I-I 151R`r^L7 HCL 1IFL-' 81C1-061The University of British ColumbiaVancouver, CanadaDate ^t.^`1"^(=I ) tri CC3DE-6 (2/88)iiAbstractGiardia lamblia is an interesting organism in several respects. Not only is it amedically important protozoan parasite, but its location in the phylogenetic tree is at acritical and informative position. Characterization of its subcellular structure andrRNA sequences suggest that Giardia is one the most primitive eukaryotes known.Giardia has an anaerobic metabolism that uses a NADP-dependent glutamatedehydrogenase (GDH) along with alanine aminotransferase to maintain anintracellular balance of NAD(P)+-NAD(P)H through the conversion of pyruvate toalanine.In the initial part of this study, the Giardia NADP-GDH gene was cloned andcharacterized. The Giardia NADP-GDH is encoded by a single copy gene in thehaploid genome. Transcript mapping and comparisons of the cDNA and genomicclones for the GDH gene did not detect the presence of introns or transsplicing. GDHtranscripts have short untranslated sequences and are initiated only three to sixnucleotides in front of the ATG translation initiation codon.In the latter part of this study, the 5' flanking sequence of the Giardia GDH gene wasanalyzed to identify possible promoter elements since core promoter elementsnormally associated with RNA polymerase II transcription in higher eukaryotes are notfound in the upstream sequence of the GDH gene or in other Giardia genes. Two novelsequence motifs were identified, an AT-rich element and a Giardia CAAT-box-likesequence called the g-CAB element. Their conservation and locations relative to thesites of transcription initiation suggest that these elements may be involved in theregulation of transcription. A 68 kD protein binds to the TTT trinucleotides found oneither one or both strands of the DNA encoding these elements from the upstreamregion of the GDH gene. This 68 kD protein is referred to as POT for its ply(T)binding ability. The function of POT may be to help denature the DNA at the promoterregion and to participate in the assembly of the RNA polymerase II pre-initiationcomplex in Giardia.Table of contentsAbsfruc-t-^ I iTable of contents^List of figuresList of tables viList of abbreviations^ viiAcknowledgments ixDedication^Introduction 1Life cycle 1Cell structure^ 1Metabolism and glutamate dehydrogenases^ 4Genome structure and organization^ 6rRNA and evolutionary implications 6Antigenic variation^ 7Transcription initiation 9TATA-binding protein (TBP)^ 11Initiator binding proteins and TBP-tethering factors^ 14Protein binding to single-stranded DNA^ 15Transcription initiation and promoters in Giardia^ 17Thesis Outline^ 18Materials and methods 19Culture conditions 19Isolation of nucleic acids^ 19Amplification and cloning of PCR products^ 20Filter hybridizations 20DNA sequencing^ 21Transcript mapping 21Enzyme assays 23Preparation of nuclear protein extracts^ 24Preparation of probes and competitors for band-shift assays^ 24Band-shift assays^ 25UV cross-linking 26Part 1: Isolation and characterization of a NADP-dependent^ 27glutamate dehydrogenase (GDH) gene^ 27Results^ 27Cloning of the GDH gene 27Characterization of the GDH gene transcripts^ 29Genomic Southern hybridization^ 31Enzyme activity^ 33Sequence alignments 34Discussion^ 37Function of the GDH protein^ 37Evolutionary relationships 39Structure and expression of the GDH gene^ 39ivPart 2: Identification and analysis of conserved sequence^ 42elements in the GDH gene promoter^ 42Results^ 42Identification of conserved sequence motifs^ 42Protein binding to double-stranded DNA 44Protein binding to single-stranded DNA 48DNA sequence requirements for protein binding^ 51Protein binding to RNA^ 55Discussion^ 56Distribution of the AT-rich and the g-CAB elements^ 56Sequence specificity of POT binding^ 57POT binding sites in the GDH PCR54 fragments^ 58Doublet band formations in some POT-DNA complexes^ 58Roles of POT in gene regulation^ 60References^ 63List of figuresFigure 1: Trophozoite and cyst forms of Giardia lamblia^ 2Figure 2: The reaction catalyzed by glutamate dehydrogenases^ 5Figure 3: A phylogenetic tree derived from theanalysis of the small subunit (16-18S) rRNA^ 8Figure 4: A comparison of a eubacterial promoterand a eukaryotic RNA polymerase II promoter^ 12Figure 5: Nucleotide sequence of the Giardia GDH gene 28Figure 6: Mapping of the 5' and 3' endsites of the GDH mRNA transcript^ 30Figure 7: Genomic Southern hybridization to identifyfragments encoding NADP-GDH or related proteins^ 32Figure 8: Glutamate dehydrogenase protein sequence alignments^ 35Figure 9: The role of NADP-GDH in Giardia metabolism^ 38Figure 10: Identification of two conserved sequenceelements in the upstream region of Giardia genes^ 43Figure 11: Alignment of the probes used in the band-shiftassay to the upstream region of the NADP-GDH gene^ 45Figure 12: Binding assays with probes containingeither the AT-rich or the g-CAB elements^ 46Figure 13: Band-shift assays with GDH PCR54containing both the AT-rich and the g-CAB elements^ 49Figure 14: Band-shift assays with duplex oligonucleotides 3a/bcontaining neither the AT-rich nor the g-CAB elements^ 50Figure 15: Band-shift assays with single-strandedoligonucleotides as either probes or competitors^ 52VList of tablesTable I: Comparison of homologue GDH proteins^ 36Table II: List of oligonucleotides used in band-shift assays^ 54viList of abbreviationsAMV^avian myeloblastosis virusATP adenosine triphosphatebp^base pair(s)cDNA DNA complementary to coding strandCTP^cytidine triphosphatedATP deoxyadenosine triphosphatedCTP^deoxycytidine triphosphatedGTP deoxyguanosine triphosphateDNA^deoxyribonucleic aciddN.TP deoxynucleoside triphosphateDTT^dithiothreitold'ITP deoxythymidine triphosphateEDTA^ethylenediamine tetraacetic acidGDH glutamate dehydrogenaseHEPES^N-2-hydroxyethylpiperazine-N'-2-ethane sulfonic acidkb kilobase(s)kD^kilodalton(s)Mb megabase(s)MF3^muscle factor 3MRE metal response elementmRNA^messenger RNANAD(H) nicotinamide adenine dinucleotideNADP(H)^nicotinamide adenine dinucleotide phosphateNTP nucleoside triphosphateviiviiiPAGE^polyacrylamide gel electrophoresisPBS 0.02 M potassium phosphate, pH 7.4; 0.155 M NaC1PCR^polymerase chain reactionPMSF phenylmethyl sulfonyl fluorideRACE^rapid amplification of cDNA endsRNA ribonucleic acidRNase^ribonucleaserRNA ribosomal RNAS^ Svedberg unit of sedimentation coefficientSDS sodium dodecyl sulphateSpl^soluble protein 1SSC (20X)^3.6 M NaCl; 0.3 M sodium citrate, pH 7TAF TBP associated factorTBE^90 mM Tris-borate, pH 8.3; 2 mM EDTATBP TATA-box binding proteinTE^10 mM Tris-HC1, pH 8.0; 1 mM EDTATEN 10 mM Tris-HC1, pH 8.0; 100 mM NaCl; 1 mM EDTATFII-I^RNA polymerase II transcription factor Itris tris (hydroxymethyl) aminomethanetRNA^transfer RNAUSF upstream stimulatory factorV^ volt(s)VSP variant-specific surface proteinixAcknowledgmentsI thank the members of my supervisory committee, Caroline Astell and PeterCandido for their help, advice and patience during the course of my studies. I thank mysupervisor, Patrick Dennis, for giving me the opportunity to work in his lab, forintroducing me to Giardia and Si nuclease mapping.I especially value the useful advice and protocols I have received from members ofthe Biochemistry Departments (Michel Roberge, Ivan Sadowski, Jan St. Armand, NinaSeto, Shu Hsu, Eve Stringham, Pat Tam, Don Jones, Steve, Rafferty, Louis Lefebvre) andother Departments (Rosie Redfield, Tom Cavalier-Smith, Rob McMaster, John Webb,Debbie Hays, Jennifer Couch).Last, but not least, I thank the members of the Dennis lab (Peter Durovic, PhalgunJoshi, Willa Downing, Bruce May, Craig Newton, Deidre de Jong-Wong, Daiqing Liao,Shanthini Mylvaganam, Janet Chow, Luc Bissonnette, Josephine Yau) for their supportand friendship. I would like to extend a special thank-you to Lawrence Shimmin whosenatural curiosity and love of science was an inspiration.xDedicationThis thesis is dedicated to my parents, Ken and Vicky Yee, and to my husband,Steve Rafferty for their encouragement and support throughout my Ph.D. studies.1IntroductionGiardia lamblia is a protozoan parasite that infects mammals, reptiles and birds.Giardia infection is one of the most common causes of severe and chronic diarrhea inhumans worldwide. In developing countries, infection of young children can lead tosevere dehydration and malabsorption resulting in death. Despite its medicalimportance, information about this organism is limited due to the relatively recentdevelopment of an in vitro culture method for Giardia lamblia (Meyer, 1976).Life cycleDuring its life cycle, Giardia can exist in two forms, as a cyst or as a trophozoite. Thecyst form is the infectious stage and the trophozoite form is the motile stage of theparasite. When a cyst is ingested by the host, by intake of either contaminated food orwater, the cyst travels through the host's digestive system where it excysts in the smallintestine. Two trophozoites emerge from one cyst after reproduction by binary fissionand attach to the epithelium of the small intestine. During the course of infection, someof the trophozoites encyst and are passed into the host's stool. Infection of another hostby these cysts allows the Giardia life cycle to be completed.Cell structureThe Giardia cyst is egg shaped and its dimensions are approximately five by eightpm (Fig. 1C). As a cyst, Giardia is protected inside a cell wall that is highly resistant toenvironmental factors. The cell wall is composed of glycoproteins with galactosamineand glucose as the major sugar moieties (Sheffield and Bjorvatn, 1977; Jarroll et al.,1989).Figure 1: Trophozoite and cyst forms of Giardia lamblia(A) The Giardia trophozoite. The ventral disk occupies the top third of the ventralsurface of the organism. The two nuclei are shown underneath the ventral disk and thepair of median bodies are shown as rod-like structures in mid-body. (B) The Giardiatrophozoite uses its ventral disk to attach to the surface of the epithelial cell. (C) TheGiardia cyst. A Giardia trophozoite is replicated while protected inside the cyst's cellwall.23The Giardia trophozoite is tear-drop shaped, and its width and length areapproximately five and ten pm, respectively (Fig. 1A). For its locomotion, Giardiamakes use of four pairs of symmetrically arranged flagella. Giardia, like othereukaryotic cells, also has a nucleus, cytoskeleton, endoplasmic reticulum, andlysosomes, In fact, each Giardia has two nuclei which are identical in size, and arebelieved to be functionally equivalent (Kabnick and Peattie, 1990). A structureresembling a golgi apparatus is observed in Giardia during encystation (Reiner et al.,1990).Giardia has an unique disk-like structure on its ventral surface; this ventral disk isused as a suction device by the organism to attach to the host's small intestine (Fig. 1B).The contractile proteins, actin, myosin and tropomyosin, have been identified in thedisk rim, and the structural proteins, tubulins and giardins, have been identified in thedisk proper (Feely et al., 1982; Peattie et al., 1989). Giardins are proteins foundexclusively in the Giardia ventral disks and range in size from 29 to 38 kD (Holberton,1981; Crossley and Holberton, 1983; Crossley and Holberton, 1985; Holberton et al.,1988; Peattie et al., 1989). A structure referred to as a median body is also found only inGiardia trophozoites. This structure, found as a pair in the mid-body inside theorganism, does not have a known function. However, the presence of giardins andtubulins in the median body suggest that it may be related to the ventral disk (Feely etal., 1982; Crossley et al., 1986).The most interesting aspect of Giardia is the organelles it does not have. Although itis a eukaryote, it does not have mitochondria or peroxisomes. The lack of theseorganelles suggests that the Giardia lineage either lost them or diverged from the maineukaryotic lineage before these organelles were acquired by endosymbiosis (Cavalier-Smith, 1987).4Metabolism and glutamate dehydrogenasesGiardia is an aerotolerant anaerobe. Associated with its lack of mitochondria, Giardialacks a Krebs' cycle and a cytochrome-mediated electron transport chain (Lindmark,1980; Weinbach et al., 1980). Most if not all of its energy requirements are met byanaerobic glycolysis (Muller, 1988). All the enzymes of carbohydrate metabolism inGiardia are present in the cytoplasm (Lindmark, 1988).Glutamate dehydrogenases (GDHs) are enzymes that play important roles incarbohydrate and amino acid metabolism, as well as ammonia assimilation. Theseenzymes catalyze the interconversion between a-ketoglutarate and L-glutamate andutilize either NADP or NAD as coenzyme (Fig. 2). Based on their metabolic specificityand their oligomeric structures, GDHs can be classified into two types (Smith et al.,1975).Class I GDHs are hexameric proteins which are involved in ammonia assimilation;they are generally NADP-dependent but may also be NAD-dependent or even havedual coenzyme specificity. Class I GDHs are found in all organisms and the genesencoding these proteins are highly conserved between eubacteria and eukaryotes(Mattaj et al., 1982; Hawkins et al., 1989). This high level of sequence conservation, andthe availability of homologous sequences from a number of different organismsrepresenting all three kingdoms, makes class I GDHs good candidates for phylogeneticanalysis (Benachenhou-Lahfa et al., 1993).Class II GDHs are tetrameric proteins which are NAD-dependent and are involvedin the conversion of glutamate to a-ketoglutarate for catabolic purposes. Thedistribution of the class II GDHs are limited to some unicellular eukaryotes such as theyeasts, Saccharomyces cerevisiae, and Neurospora crassa (Vierula and Kapoor, 1989; Millerand Magasanik, 1990). All organisms characterized to date which have a class II GDHalso have a class I GDH.NAD(P)+ NAD(P)HNH3^COOHICH, 2IDo CH 2I0=CICOOHCOOHICH,^2ICH, 2^-44IH2 N —C— HICOOHoxidative deaminationreductive amination5L-glutamate^ a-ketoglutarateFigure 2: The reaction catalyzed by glutamate dehydrogenasesGlutamate dehydrogenase catalyzed the interconversion of L-glutamate anda-ketoglutarate using either NAD(H), NADP(H), or both as coenzymes.6Genome structure and organizationThe number of different Giardia chromosomes has been estimated at four bymicroscopic examination of DAPI-stained nuclei and five by pulsed-field separation ofchromosomes (Adam et al., 1988; Kabnick and Peattie, 1990). Similarly, the genome sizeand the chromosome ploidy have not been resolved. Estimates of the haploid genomesize range from as large as 80 Mb by Cot analysis to as small as 12 Mb by size estimationof chromosomes separated by pulsed-field gel electrophoresis (Boothroyd et al., 1987;Fan et al., 1991). The estimation of chromosome ploidy has ranged from diploid topentaploid (Adam et al., 1988; Kabnick and Peattie, 1990).The Giardia genes characterized to date have tended to be GC-rich; the G+C contentis 67% for the I3-tubulin genes and 75% for the rRNA genes (Kirk-Mason et al., 1989;Healey et al., 1990) Surprisingly, the overall G+C content of the Giardia genome wasestimated to be only 48% by thermal denaturation analysis (Boothroyd et al., 1987).These observations imply that the non-coding regions of the Giardia genome must beAT-rich if the interpretation of the melting profile is correct (Adam, 1991).rRNA and evolutionary implicationsThe ribosome is a RNA-protein complex responsible for the translation of mRNAinto proteins. Ribosomes can be separated into two components which are referred toas the large (L) and the small (S) subunits. The ribosomal RNAs (rRNAs) are namedeither according to their sedimentation coefficient (i.e. 5S, 16S, 23S), or according to theirsubunit association (i.e. the 16S rRNA is also referred to as the small subunit rRNA).The ribosomal proteins are named according to both their subunit association and theirposition relative to the other ribomsomal proteins after separation on a 2-dimensionalpolyacrylamide gel. Therefore, ribosomal protein L12 is associated with the largesubunit and migrates to position twelve after 2-D gel electrophoresis.7Sequence analysis of its rRNA genes support the proposal that Giardia divergedearly from other eukaryotes. Giardia has the shortest rRNA known for any eukaryote;the large and small subunit rRNAs are approximately 2300 and 1300 nucleotides long,respectively (Boothroyd et al., 1987). These rRNAs are more similar in size to those fromeubacteria and are missing regions which were once thought to be absolutely conservedamong eukaryotes. Furthermore, sequence comparison of Giardia small subunit 16S-likerRNA to those from other organisms placed Giardia at the lowest branch of theeukaryotic phylogenetic tree (Fig. 3) (Sogin et al., 1989).Antigenic variationThe entire outer surface of the Giardia trophozoite, including the flagella, is covered bya yariant specific surface protein (VSP). These proteins vary in size from 35 to >200 kDand contain 11 to 12% cysteine which are arranged in cys-X-X-cys motifs throughout theprotein (Nash, 1992). Giardia possesses a family of genes encoding different VSPs butonly one of these genes is expressed at any one time in an individual trophozoite. VSPexpression changes spontaneously, and the frequency of change is a function both of theparticular Giardia isolate and of the VSP (Nash et al., 1990). This phenomenon isreferred to as surface antigenic variation.Although antigen variation is commonly thought to be a mechanism to evadeimmune response of the host, the biological significance of VSPs in Giardia has not beendetermined. Giardia isolates with different VSPs were observed to have variablesusceptibility to intestinal proteases (Nash et al., 1991). Recently, Giardia VSPs wereshown to be able to bind zinc through their multiple cys-X-X-cys motifs (Nash andMowatt, 1993). The biological importance of zinc binding is unknown but it wasproposed that this may be a method used by Giardia to prevent the detrimental effects ofoxidation (Nash and Mowatt, 1993). Zinc binding by VSPs has also been suggested to8[r Homo sapiensI— Xenopus laevis^ Saccharomyces cerevislaeZea maysChlamydomonas reinhardtiiOchromonas danicaProrocentrum micans^ Oxy1richa novaPlasmodium bergheiDictyostelium discoideum^ Eugfena grad/is^ Trypanosoma bruceiVairimorpha necatrixGiardia lamblia0.1^I' EukaryotesSulfolobus solfataricus^ Halobacterium volcaniiMethanococcus vannielii^ Bacillus subtilis Escherichia coil^ Anacystis nidulansArchaebacteriaEubacteriaFigure 3: A phylogenetic tree derived from the analysis of the small subunit (16-18S)rRNAThis unrooted phylogenetic tree contains organisms from the three domains:eubacteria, archaebacteria and eukaryota. The position of Giardia in the eukaryoticlineage is indicated by the arrow. The horizontal components (branches) of the treerepresents the evolutionary distance between the organisms. The short bar indicated onthe top left hand side of the figure corresponds to ten nucleotide changes per 100positions. This figure presents the tree of Sogin et al., (1989) derived by the distancematrix method.9be responsible for symptoms of zinc deficiencies in hosts infected with Giardia (Nashand Mowatt, 1993).The control of antigenic variation in Giardia trophozoites is thought to be at the levelof transcription or of RNA stability, since transcripts of expressed VSPs are observedwhereas transcripts of non-expressed VSPs are not observed (Adam et al., 1988; Adam etal., 1992; Mowatt et al., 1992; Nash and Mowatt, 1992). No other genetic alterationsappear to be associated with expressed versus non-expressed VSP genes (reviewed by(Nash, 1992). At present, the mechanism controlling antigen variation has not beendetermined.In the following sections, an overview of transcription initiation will be presented.Emphasis will be on specific aspects of this process relating to the analysis of putativepromoter elements in Giardia genes which will be discussed in the second half of thisthesis. These aspects include core promoter elements, transcription factors which bindthese elements, and proteins which can bind single-stranded DNA.Transcription initiationGene expression can be controlled at different levels by mechanisms that affectchromatin organization, DNA methylation, transcription, RNA decay, translation, post-translational modification, and protein degradation. The regulation of transcription isone of the early steps in the control of gene expression; the first step in the control oftranscription is the initiation event, a multi-step process catalyzed by RNA polymerase.This process in eubacteria has been characterized most completely through studies onEscherichia coli (reviewed by Reznikoff et al., 1985).The complete E. coli RNA polymerase, called the holoenzyme, consists of fivesubunits (a2PITa) which can be separated into the core enzyme (a21313') and the sigmafactor (a). The core enzyme has an intrinsic ability to bind non-specifically to DNA andto initiate transcription randomly on the DNA template. The sigma factor is required to10help stabilize the binding of the enzyme to a specific DNA sequence and to begintranscription at a particular site, usually an A or a G nucleotide, in front of a gene. Theregion of the DNA which contains the sequences needed for the specific binding andtranscription initiation by RNA polymerase is called the promoter.The E. coli standard promoter contains two sequence motifs which are namedaccording to their nucleotide position relative to the site of transcription initiation(Fig. 4A). The -10 sequence has the consensus, TATAAT, and the -35 sequence has theconsensus, TTGACA. These consensus sequences are found in the majority of E. coligenes and the sigma factor that allows the RNA polymerase to recognize them isdesignated 670 with the superscript indicating its molecular weight (in kD).Transcription of some E. co/i genes is induced under certain conditions and thesegenes have different consensus sequences in the -10 and -35 promoter region. Sincethere is only one type of core RNA polymerase in E. co/i, different sigma factors areused by the RNA polymerase to recognize these alternative promoter elements. Forexample, the promoters of the genes induced by heat shock and nitrogen starvation arerecognized by RNA polymerases with 632 and a60, respectively (reviewed by Helmannand Chamberlin, 1988).Eukaryotes, unlike eubacteria, have three different RNA polymerases to transcribethree different types of genes. Ribosomal RNA is transcribed by RNA polymerase I,mRNA by RNA polymerase II, and tRNAs and other small RNAs by RNA polymeraseIII. Similar to the eubacterial polymerase which requires a sigma factor, all threeeukaryotic polymerases require accessory proteins for the initiation of transcription atthe correct sites. These proteins are referred to as transcription factors. Formation of atranscription preinitiation complex requires the coordinate binding of thesetranscription factors to their respective recognition sites on the DNA template, and therecruitment of the correct RNA polymerase to this region (Buratowski et al., 1989).11In RNA polymerase II transcription, transcription factors and sequence motifs thatare found in the majority of genes and are sufficient to support a basal level oftranscription are called general initiation factors and core promoter elements,respectively. Core promoters usually include two sequence elements called the TATAbox and the initiator element, which are recognized by the transcription factors TBP andTF1I-I, respectively (Fig. 4B).In some genes, additional factors and their specific recognition sites are used tomodulate transcription regulation. In mammals, the fl-globin and the histone H2Bgenes have an additional promoter element, called the CAAT box, located between 40 to100 bp upstream of the transcription initiation site (reviewed by La Thangue and Rigby,1988). This element, with the consensus sequence GG(C/T)CAATCT, is bound by afamily of transcription factors which up-regulate transcription of these genes.TATA-binding protein (TBP) The TATA-box sequence, with the consensus, TATA(A/T)AAT, is located 25 to 30bp upstream of the transcription start site in the majority of RNA polymerase IIpromoters. The transcription factor which binds to this sequence is called TBP forTATA binding protein. TBPs are normally associated in a complex with a number ofcoactivators called TAFs for TBP associated factors (Nakatani et al., 1990; Peterson et al.,1990; Pugh and Tjian, 1990; Dynlacht et al., 1991). These TAFs-TBP associations arespecies specific and are indispensable for regulation by transcriptional activators (Pughand Tjian, 1990).The TBP gene has been cloned from a variety of eukaryotes including yeast, human,Drosophila, and the plant, Arabidopsis thaliana (reviewed by Dynlacht et al., 1991).Although the proteins encoded by these genes vary in size from 28 to 38 kD, they allshare a highly conserved (>80% sequence identity over 180 amino acids) carboxy-terminal domain.A: Eubacteria (E. coli) promoter12-101 ^I TATAATB: Eukaryotic RNA polymerase II promoter1 TTGACA-25TATA AAAT^  CACTcC C CT TT TTTTATA-box initiatorelementFigure 4: A comparison of a eubacterial promoter and an eukaryotic RNA polymeraseII promoter(A) An idealized eubacterial (E. coli) promoter with the -10 and -35 consensussequences indicated relative to the site of transcription initiation, at a purine nucleotide,which is indicated by the arrow at +1. (B) An idealized eukaryotic RNA polymerase IIpromoter containing both a TATA-box and an initiator element. Transcriptioninitiation, indicated by an arrow at +1, occurs within the initiator element.13This domain has been implicated in the DNA binding function of TBP (Horikoshi et al.,1989; Hoffman et al., 1990). The 180 amino acid carboxy-terminal domain of TBPcontains a number of distinct motifs (Hoffman et al., 1990). One of these is a set ofamino acid repeats which contributes to the symmetry of the folded TBP protein. Asecond motif is a region with limited sequence similarity (23% identity and 40%similarity over 30 amino acids) to all bacterial sigma factors. This similarity is greatestin the alignment of the TBP sequence to that portion of the sigma sequence which isimplicated in the binding of single-stranded DNA (Helmann and Chamberlin, 1988). Athird motif consists of a repeat of basic residues which has the potential to form ana-helical structure with the basic residues located on one face. The fourth motif is aregion that shows sequence similarity to the helix-loop-helix domains of a proteinfamily related to the myc oncogene family of proteins.In contrast to the high sequence conservation in the carboxy-terminus, the amino-terminal region of TBP from different eukaryotes varies markedly in both size andsequence and is responsible for the observed size variation of the protein. The amino-terminal of TBP is believed to be involved in interactions with transcriptional activatorsthat are species specific (Pugh and Tjian, 1990).It has long been known that TBP plays a pivotal role in the formation of thetranscription preinitiation complex in TATA-box containing RNA polymerase IIpromoters, but recently, TBP was found to be required for RNA polymerase IIpromoters lacking a TATA-box, and also for RNA polymerase I and III promoters (Pughand Tjian, 1991; Cormack and Struhl, 1992; Green, 1992; Sharp, 1992). How is TBPbrought to the promoter region in the absence of a DNA binding site for this protein?While TBP binds directly to DNA at the TATA-box sequence in TATA-containingpromoters, another transcription factor binds the DNA in TATA-less promoters andthen recruits TBP directly or indirectly via yet another transcription factor to the14promoter region (reviewed by Rigby, 1993). For the RNA polymerase II TATA-lesspromoters, TFII-I and YY1 are two of these TBP tethering factors.Initiator binding proteins and TBP-tethering factors The initiator element, a pyrimidine-rich DNA sequence with the consensus,(C/T)A(C/T)TC(C/T)(C/T)(C/T), is located at the transcription start site in someTATA-containing and TATA-less RNA polymerase II promoters (Smale and Baltimore,1989; Smale et al., 1990). A number of nuclear proteins from HeLa cells bindsspecifically to the initiator element. Two of these proteins are the transcription factors,TFII-I and YY1, which are approximately 120 and 68 kD in size, respectively (Roy et al.,1991; Shi et al., 1991).In the TATA-containing adenovirus major late promoter, TFII-I was found to bindboth the initiator and an upstream site called the E box, with the sequence CACGTG.Similarly, another transcription factor immunologically related to TFII-I, called USF(upsteam atimulatory factor), can also bind at both sites. TFII-I and USF weredemonstrated to interact cooperatively at these sites by band-shift assays and byDNaseI footprinting. These results suggest that TFII-I binding at multiple promoterelements may facilitate interactions between upstream regulatory factors and the corepromoter elements (Roy et al., 1991).In the TATA-less promoter of the terminal deoxynucleotidyl transferase gene inmammals, the initiator has been implicated as being essential for core promoter strengthand for determining the actual initiation site (Smale and Baltimore, 1989). Theseproperties of the initiator were suggested to be a consequence of the recruitment of TBPto the promoter region by the binding of to the initiator. (Roeder, 1991; Roy et al.,1991).In several respects, YY1 is similar to TFII-I. Firstly, YY1 binds both at the initiatorand at another upstream site in the adeno-associated virus P5 promoter (Shi et al., 1991).15Secondly, YY1 was demonstrated to be required for determining the transcriptioninitiation site and promoter strength in both the adeno-associated virus P5 andadenovirus major late promoters (Seto et al., 1991; Shi et al., 1991). Lastly, YY1 can alsobind at the initiator of the TATA-less terminal deoxynucleotidyl transferase genepromoter (Seto et al., 1991). It is probable that YY1, like TFII-I, can also function as aTBP tethering factor in TATA-less promoters.A subgroup of mammalian "housekeeping genes", which are expressedconstitutively at low levels in all tissue types, have promoters that are TATA-less(reviewed by Sehgal et al., 1988). These genes have another promoter element called theGC box, which has the consensus sequence CCGCCC. This element is usually found inmultiple copies within 100 bp upstream of the transcription start sites. A transcriptionfactor called Sp1 binds to the GC boxes and activates transcription (Dynan and Tjian,1983; Dynan et al., 1986). Another proposed function of Sp1 in these TATA-lesspromoters is to recruit TBP to this region indirectly by interacting with anothertranscription factor which has not yet been characterized (Pugh and Tjian, 1991). Theidentification of Sp1 has been restricted to mammalian cells although Sp1 can functionas a transcription activator when introduced into a Drosophila in vitro transcriptionassay. In contrast, Sp1 does not affect transcription in yeasts in similar experiments(Pugh and Tjian, 1990). These observations suggest that Sp1 is unlikely to function aseither a transcription activator or a TBP tethering factor in lower eukaryotes.Protein binding to single-stranded DNA Transcription initiation requires the unwinding of the duplex DNA in the promoterregion to allow the RNA polymerase access to the single-stranded DNA template.Therefore, the ability to bind single-stranded DNA may be one of the properties of theRNA polymerase and any associated transcription factors (Horikoshi et al., 1989).16The bacterial sigma factor has been shown by genetic criteria to bind single-strandedDNA (Helmann and Chamberlin, 1988). As the sigma factor allows the RNApolymerase to bind to the -10 and -35 sequences and to initiate transcription at theproper sites, the eukaryotic TBP recruits the RNA polymerase to the promoter region bybinding to the TATA-box, allowing the formation of a preinitiation complex, and thestart of transcription within the initiator element. Due to this functional analogy and itssequence similarity to the sigma factor, TBP may also be expected to bind single-stranded DNA (Horikoshi et al., 1989).A protein called muscle factor three (MF3), which was isolated from an embryonicchicken muscle nuclear protein extract, can bind to single-stranded DNA containingpromoter elements from muscle genes (Santoro et al., 1991). These promoter elementsare the E-box motif from the mouse creatine kinase M gene, and the MCAT and MRE(metal response element) motifs from the chicken a-actin gene. The binding by MF3 tothese elements was usually detected as a doublet in band-shift assays. In the separationof MF3 from other proteins in the nuclear extract by heparin-Sepharosechromatography, the activity that forms the lower band of the doublet elutes slightlybefore the activity that forms the upper band in band-shift analysis. Mutations in theMCAT motif which diminished binding by MF3 on band-shift assays also diminishedthe ability of the MCAT motif to activate transcription from a DNA construct containingthe chloramphenicol acetyltransferase reporter gene in a transient transfection assay inmuscle cells (Santoro et al., 1991).Attempts to identify a common sequence required for MF3 binding to the threedifferent promoter elements by interference footprinting and band-shift assays onmutant probes were unsuccessful. Despite the lack of sequence similarity among thedifferent DNA probes recognized by MF3, the protein does discriminate in its bindingto different DNA. MF3 binds both strands of the DNA containing the MCAT sequence,but binds only to the (-) strand of the DNA containing either the E-box or the MRE17sequence. Furthermore, MF3 does not bind to DNA containing other promoterelements such as the c-Fos and the cytoskeleton actin serum response element. Santoroet al. (1991) speculate that MF3 may recognize and stabilize altered conformations ofDNA that may arise in the promoter region during muscle cell development.MyoD is a transcription factor which can induce muscle differentiation in a widevariety of primary cells and transformed cell lines (Weintraub et al., 1989). Similar toMF-3, MyoD can also bind single-stranded DNA such as the (-) strand of the DNAcontaining the creatine kinase M E-box. However, in contrast to MF-3, MyoD does notbind either DNA strand containing the MCAT or the MRE motifs (Santoro et al., 1991).Transcription initiation and promoters in Giardia A limited number of genes have been characterized from Giardia; they include genesencoding giardins, tubulins, triosephosphate isomerase (TIM), ADP-ribosylation factor(ARF), and a number of variant cell surface proteins (Adam et al., 1988; Kirk-Mason etal., 1989; Gillin et al., 1990; Mowatt et al., 1992; Murtagh et al., 1992). Thecharacterization of the transcripts from these genes by a combination of Si nucleaseprotection, primer extension and RNA sequencing indicate that the transcripts have ashort 5' leader sequence that is initiated one to six nucleotides in front of the ATGtranslation initiation codon (reviewed by Adam, 1991). This is in contrast to highereukaryotes where transcription initiates 20 to 100 or more nucleotides upstream of thetranslation initiation codon (Kozak, 1989). Furthermore, the examination of the regionsupstream of the 5' transcript end sites has failed to identify core promoter elementsassociated with RNA polymerase II transcription in other eukaryotes.Analyses of promoter sequences in the other protozoan parasites such asTrypanosomes and Leishmania have been confounded by the phenomenon oftranssplicing, in which a leader sequence is added to the 5' end of every RNA transcript.Because this transsplicing process is thought to occur at the same time as transcription,18the identification of transcriptional start sites and associated promoter sequences hasbeen unsuccessful in these parasites (Huang and Van der Ploeg, 1991). Sincetranssplicing has not been detected in Giardia lamblia, and sites of transcription initiationhave been mapped for a number of genes, sequences associated with transcriptionalregulation may be more easily identified and characterized in this organism.Thesis OutlineThis thesis has been organized into two parts, each with its own Results andDiscussion sections. In the first part, I present the characterization of both genomicDNA and cDNA clones of the Giardia class I NADP-dependent GDH gene. Southernblot analysis and enzyme assays were performed to determine the copy number of thegene encoding the NADP-GDH and whether Giardia has a class II NAD-dependentGDH. Transcripts of the NADP-GDH gene were analyzed by primer extension and Sinuclease protection experiments to determine the sites of transcription initiation andpoly(A) addition. The protein sequence derived from the NADP-GDH gene wasaligned to homologous proteins from Neurospora crassa and E. coli. to obtain an estimateof the degree of similarity among these proteins.The second part describes my investigation of two conserved sequence motifs,referred to as the A/T-rich and the g-CAB elements, that I identified in the 5' flankingregion of the GDH gene and other Giardia genes. The ability of proteins from a Giardianuclear extract to bind specifically to DNA containing these elements in band-shiftassays was investigated. The sequence specificity for protein binding was determinedby cross-competition experiments and binding assays on DNA probes containing pointmutations. The size of the protein bound to these DNA probes was determined by UVcross-linking experiments. These results and their significance will be discussed.19Materials and methodsCulture conditions Giardia lamblia (Portland-1 strain, American Type Culture Collection) was grown at37°C in modified TYI-S-33 medium (Keister, 1983). The composition of one litre of TYI-S-33 medium is as follows: 20 g casein digest, 10 g yeast extract, 10 g D-glucose, 2 gNaC1, 2 g L-cysteine HC1, 1 g K2HPO4, 0.6 g KH2PO4, 0.75 g bovine bile, 0.2 g L-ascorbicacid, 22.8 mg ferric ammonium citrate, and 100 ml heat-inactivated calf serum. All theabove components, except for the serum, were dissolved in deionized H20 and the pHof the solution adjusted to 7.0-7.2 with 4 M NaOH. The medium was then sterilized bypassage through a 0.2 p.m filter, and calf serum was added before use. To limit theamount of oxygen present in the culture flasks after inoculation, the flasks were filled tocapacity with medium before capping.Cells were harvested by chilling the culture flasks on ice for 15 to 30 mm to allowdetachment of the Giardia from the flask surfaces, followed by centrifugation of theculture at room temperature (1,200 rpm for 15 mm in a Western H-103N centrifuge).The cells were washed twice in cold PBS buffer (0.02 M potassium phosphate, pH 7.4,0.155 M NaC1).Isolation of nucleic acidsFor the isolation of DNA, washed G. lamblia cells were resuspended in TEN buffer(10 mM Tris-HC1, pH 8.0, 100 mM NaC1, 1 mM EDTA) and lysed by addition of SDS(0.5%) and proteinase K (500 lig/m1). The mixture was incubated at 50°C for 1.5 h. Thelysate was extracted once with phenol, once with phenol:chloroform (1:1) and once withchloroform. After dialysis overnight against TEN buffer at 4°C, the DNA wasconcentrated by ethanol precipitation and resuspended in TE buffer (10 mM Tris/HC1,pH 7.5, 1 mM EDTA). Contaminating RNA was removed by treatment with RNaseA.20For preparation of RNA, washed cells were lysed in 6 M guanidine HC1 in0.1 M sodium acetate at pH 5. The RNA was precipitated by the addition of 0.5 volumeof ethanol and was centrifuged at 4°C for 1 h (14,000 rpm in a Beckman TL-100ultracentrifuge).Amplification and cloning of PCR products PCR amplification was performed on 5 ng of poly(G) tailed cDNA using the RACEprotocol as described by Frohman et al. (1988). The primers used were5'GTCGAA(G/C )AG(G / C)CCGAA(G/ C)C CC AT3' (oLW 72) and 5'GCGCTCTAGAC143'(adapter primer XbaI-dC14). The PCR reaction was annealed at 57°C and extended at72°C, repeated for 35 cycles. The amplification products were treated with DNApolymerase (Klenow fragment) and dNTPs, and ligated into the Hindi site of the pUC13 vector. One of the clones, designated pJYPCR10.5, was characterized further.Filter hybridizationsAll fragments that were used as probes in Southern filter hybridizations werepurified from agarose or acrylamide gels and radiolabeled by the random primermethod (Feinberg and Vogelstein, 1982). Unincorporated radioactivity was removed byeither ethanol precipitation of the DNA, or by passage of the reaction mixture througha Sephacryl S-200 spin column (Pharmacia).For Southern blot analysis, 8 [tg of Giardia DNA was digested with variousrestriction enzymes and electrophoresed in a 0.8% agarose gel. The DNA wastransferred onto Hybond-N membranes (Amersham) according to the manufacturer'sinstructions. In the analysis to determine the copy number of the NADP-GDH gene, the2.4 kb Pvull fragment containing the full length Giardia NADP-GDH gene was used toprobe a genomic blot prepared as described above. After hybridization at 68°C, the blotwas washed under conditions of high stringency (0.2X SSC, 0.1% SDS at 68°C). In theanalysis to determine the presence of a G. lamblia gene homologous to the Saccharomyces21cerevisiae NAD-GDH, a 2.6 kb Sall fragment containing about three quarters of the entireyeast genomic NAD-GDH gene was used to probe a G. lamblia genomic blot. Thegenomic clone containing the S. cerevisiae NAD-GDH gene was a gift from Steve Millerand Boris Magasanik (Miller and Magasanik, 1990). After hybridization at 42°C, theblot was washed initially at 2X SSC, 0.1% SDS at 42°C and subjected toautoradiography. The blot was rewashed with 2X SSC, 0.1% SDS at 52°C andautoradiography was repeated.A G. lamblia genomic library in XEMBL3 and a G. lamblia cDNA library in Xgt10were gifts from D. Peattie (Peattie et al., 1989). Duplicate filters from each library wereprobed at 60°C with 32P-labeled cDNA insert from pJYPCR10.5. The filters from bothlibraries were washed at intermediate stringency (2X SSC, 0.1% SDS at 60°C); thegenomic library filters were rewashed at higher stringency (0.2X SSC, 0.1% SDS at65°C).DNA sequencingDNA fragments were cloned into pGEM-Zf vectors (Promega) and single-strandedtemplates were prepared for dideoxy chain termination sequencing (Sanger et al., 1977)with T7 DNA polymerase. Regions of secondary structure were resolved by theincorporation of either deoxy-7-deazaguanosine triphosphates or deoxyinosinetriphosphates in the sequencing reactions. Both strands of the genomic DNA and thecDNA containing the GDH gene were sequenced from overlapping clones obtainedafter progressive exonuclease III deletion of the full length DNA or cDNA clone(Henikoff, 1984).Transcript mappingTo determine the location of the 5' transcript ends by primer extension, theoligonucleotide oJY1 (51CCACGACCTCCTCGACGGC3') was 5' end-labeled with T4polynucleotide kinase and y32P-ATP. The radiolabeled oligonucleotide was hybridized22to 10 lig of G. lamblia total RNA and extended with AMV reverse transcriptase anddNTPs (Newman, 1987). The resulting extension products were electrophoresed onpolyacrylamide gels alongside a sequence ladder generated using the same end-labeledoligonucleotide as a primer in a dideoxy chain termination sequencing reaction using anappropriate template.Probes for Si nuclease protection experiments were (i) a 441 bp HindIII-HincIIfragment which was end-labeled on the (-) strand at the HincII site using T4polynucleotide kinase and 732P-ATP and (ii) a 296 bp EcoRI-Nan fragment which wasend-labeled on the (-) strand at the EcoRI site by a32P-dATP, a32P-dTTP and DNApolymerase (Klenow fragment). The radiolabeled fragments were hybridized with 10lig of G. lamblia total RNA and digested with Si nuclease (Favaloro et al., 1980). Theproducts were electrophoresed on denaturing polyacrylamide gels alongside amolecular length standard (pGEM-7Z4+), digested with Mspl and 3' end-labeled withdCTP and a32P-dGTP), or a sequence ladder generated from the original radiolabeledfragment by the chemical degradation method performed on Amersham's Hybond M &G membranes (Maxam and Gilbert, 1980).GDH sequence alignments The protein alignment was performed using a DNASTAR alignment computerprogram utilizing a Lipman-Pearson algorithm. The alignment was checked by visualinspection.To calculate the degree of amino acid identity in pairwise comparison, the sum ofthe number of positions with identical amino acids were calculated as a percentage ofthe total number of positions shared between the two sequences.23Enzyme assays Washed cells were resuspended in 100 mM potassium phosphate buffer (pH 7.3)and lysed by two passages through a Yeda Press, once at 700 psi and once at 400 psi.The cell lysate was centrifuged at 4°C to remove particulate matter and cellular debris(12,000 rpm for 10 min in an Eppendorf 5412 centrifuge). For the forward reaction (thereductive amination of a-ketoglutarate) the supernatant was added to a glutamatedehydrogenase assay mixture containing 20 mM a-ketoglutarate, 20 mM NH4C1, and0.1 mM of either NADPH or NADH in 100 mM potassium phosphate buffer, pH 7.3.The assay conditions for determining endogenous background NADPH and NADHoxidation were the same as for the glutamate dehydrogenase assays except that 40 mMNaCl replaced the substrates, a-ketoglutarate and NH4C1.For the reverse reaction (the oxidative deamination of L-glutamate), the supernatantof the Giardia lysate was added to an assay mixture containing 20 mM L-glutamate and0.1 mM NAD+ in 100 mM potassium phosphate buffer, pH 7.3. To determineendogenous background NAD÷ reduction, the same assay conditions were used exceptthat 20 mM NaCl replaced the substrate L-glutamate.The reactions were monitored at room temperature (on a Perkin-Elmer Lambda 3Bspectrophotometer) by the change in optical density at 340 nm following the oxidationof NAD(P)H or the reduction of NAD+. All assays were performed in duplicate. A unitof activity (U) is defined as the amount of enzyme which catalyzed the oxidation of1 ;mole of NAD(P)H or the reduction of 1 umole of NAD÷ per minute under the assayconditions. Specific activity is defined as units per milligram of protein. The proteinconcentration of the crude extract was estimated by the Bradford assay using bovineserum albumin as the standard (Bradford, 1976).24Preparation of nuclear protein extractsNuclear extracts were prepared from freshly harvested Giardia cultures using amodification of the procedure of Andrews and Faller (1991). Giardia was lysed inhypotonic buffer (10 mM HEPES-KOH, pH 7.9, 1.5 mM MgCl2, 10 mM KC1, 0.5 mMDT'T, 0.2 mM PMSF) and by the addition of 0.2% NP40. After incubation on ice for 10min, a nuclear pellet was recovered by centrifugation of the cell lysate at 4°C (12,000rpm for 15 sec. in an Eppendorf centrifuge). The nuclear pellet was resuspended inextraction buffer (20 mM HEPES-KOH, pH 7.9, 25% glycerol, 420 mM NaCl, 1.5 mMMgCl2, 0.2 mM EDTA, 0.5 mM DTT, 0.2 mM PMSF), and incubated on ice for 20 minwith occasional inversion of the sample tube. The extraction mixture was centrifugedfor 4 min at 4°C (12,000 rpm in an Eppendorf centrifuge) and the extracted protein wasrecovered in the supernatant. The protein concentration of the extract was estimated bythe Bradford assay using bovine serum albumin as the standard (Bradford, 1976). Theextract was divided into small aliquots, frozen in a dry ice/ethanol bath, and stored at-700C . The nuclear protein preparation procedure was monitored by fluorescentstaining (Hoechst 33258 dye diluted 1:1000 from a 1 mg/ml stock solution) of theGiardia nuclei at various stages during its isolation and extraction.Preparation of probes and competitors for band-shift assaysTable II lists the oligonucleotides used. Oligonucleotides were used either alone, assingle-stranded DNA, or annealed together with the complementary strand, as double-stranded DNA. For annealing, the two complementary oligonucleotides were added inequimolar amounts to a 200 ill volume of an annealing buffer (50 mM Tris-HC1, pH 8.0,10 mM MgCl2, 50 mM NaCl), heated in a 90°C water bath for 5 min and cooled slowlyto room temperature over 1 h. Both single-stranded and double-strandedoligonucleotides were radiolabeled with T4 polynucleotide kinase and y32P-ATP.25A 245 bp HindIII-Sacl fragment containing the GDH sequence (from -223 to +21relative to the zero position assigned to A of the ATG translation initiation codon; ornucleotide position 1 to 245 in the sequence presented in figure 5) was cloned into apGEM3Zf(-) vector (Promega). A 54 bp fragment, called GDH PCR54, was preparedfrom the pGEM clone by using oligonucleotides 2a and lb as primers in a PCRamplification reaction. The primers were annealed at 500C and extended at 720C for 30cycles in a 100 i.t1 reaction. The 54 bp PCR product was gel purified and its sequencewas checked by direct sequencing using the CircumVent thermal cycle dideoxy DNAsequencing kit (New England Biolabs). The fragment was radiolabeled by thesubstitution of half the total concentration of either dATP or dCTP with oc32P-dATP oroc32P-dCTP, respectively, in the PCR amplification reaction. For radiolabeling purposes,amplification was performed for 15 cycles in a 25 ill reaction. The radiolabeled GDHPCR54 fragment was gel purified prior to use.Transcription of RNA in vitro was from the 245 bp Hindu/ Sad fragment cloned intothe pGEM vector described above. The pGEM clone was linearized at the EcoRI site inthe vector's polylinker and transcription was initiated at the SP6 promoter by theaddition of NTPs and SP6 RNA polymerase. The RNA was radiolabeled by theincorporation of oc32P-CTP. The DNA template was removed after transcription bytreatment with RQ1-RNase free DNase (Promega) and the RNA was gel purified priorto use.Poly(dI-dC) and oligo(dT) 12-18 were obtained from Pharmacia.Band-shift assaysBinding reactions contained 2.5 lig of poly(dI-dC), 3-5 ug of nuclear extract protein,and 0.02-0.05 pmoles of radiolabeled probe in a 20 1J1 volume of binding buffer (20 mMTris-HC1, pH 7.5, 100 mM NaCl, 10% glycerol, 1 mM DTT, 0.2 mM PMSF). Incompetition experiments, all components, including probe and competitor DNA, were26pre-mixed before the final addition of the protein extract. In the studies on single-stranded DNA binding, the poly(dI-dC) was first heated at 100°C and cooled quickly onice prior to its addition to the binding reaction. The reactions were incubated on ice for20 min. A 5% polyacrylamide gel was pre-electrophoresed for 1-2 h at 200 V in 0.5XTBE/11)/0 glycerol with a replacement of fresh running buffer prior to sample loading.Samples were loaded onto the gel and subjected to electrophoresis at 150 V for 1.5 - 2 hat room temperature. After electrophoresis, the gel was fixed (10% acetic acid/10°/0methanol for 30 min), dried onto filter paper, and subjected to autoradiography.UV cross-linkingThe binding reactions using oligonucleotides 1a/b and 2a/b as probes were scaledup five-fold and performed in duplicate. After incubation on ice, samples weresubjected to UV irradiation by a pulsed Nd:YAG laser (5 nsec pulse at 266 nm, 60mJ/pulse; Quanta-Ray, GCR14). One set of each duplicate reaction was subjected toelectrophoresis through a standard band-shift gel. After electrophoresis, the gel wassubjected to autoradiography and gel slices corresponding to the shifted bands on theautoradiograph were removed. The gel slices were placed in the sample wells of adiscontinuous 10% SDS-polyacrylamide gel (Laemmli, 1970) along with the remainingset of UV treated binding reactions and a pre-stained protein size marker (BRL). Afterelectrophoresis, the gel was fixed, dried and subjected to autoradiography as describedabove.27Part 1: Isolation and characterization of a NADP-dependentglutamate dehydrogenase (GDH) geneResultsCloning of the GDH gene A partial cDNA clone encoding a polypeptide homologous to eukaryotic glutamatedehydrogenases was fortuitously isolated during an attempt to clone the Giardia lambliagene encoding the protein equivalent to the E. colt L12 ribosomal protein. Mostcharacterized eukaryotic L12 equivalent genes contain the sequence MGFGLFD at theircarboxy termini (Newton et al., 1990). A 21-mer oligonucleotide complementary tomRNA encoding this heptapeptide (oLW72) was used along with an oligonucleotideXbaI-d(C)14 as primers in a PCR reaction using poly(G) tailed cDNA as template (seeMaterials and Methods). One amplification product, cloned into the Hindi site ofpUC13, was 385 nucleotides in length. This clone, designated pJYPCR10.5, was foundto encode a polypeptide that exhibits 77% identity over a region of 69 amino acidresidues to the NADP-specific glutamate dehydrogenase of Neurospora crassa. I chose tocharacterize this GDH-like gene more fully because very few Giardia genes have beenstudied in detail.The cDNA clone pJYPCR10.5 was used to screen a XEMBL3 G. lamblia genomiclibrary. Eight positive clones were identified; all were found to contain a 2.4 kb Pvullfragment that hybridized to the pJYPCR10.5 cDNA fragment. This 2.4 kb fragment wassubcloned in both orientations into the SmaI site of pGEM3Zf(-) to give pJY1 and pJY11.The sequence of a 1691 nucleotide long portion of the 2.4 kb fragment, from an internalHin dill site to the downstream Pvull site, was determined (Figure 5). The sequencedregion was found to contain an open reading frame specifying a 449 amino acid long28^Hindi!!^20^40^60^80^100^120ANCTICACGTCGTCGTTCTCGTCTCCGCTGTCTGCATCCTTTACTTATCGCTTATAGARTGTCCACAAACTAAAAGTATACCGGTCARACTTATTCTGAGGGITACGGRACAAGTGTGC^140^160^180^200^....^240GCGTAGAGAGCCGGGGGCTCCCGAGGACCTTCGGGGGAGATGGCTGTTCTTGFIAAAGCTGACCACARATAACGCCTTTAATTACAGGCGCCCCAGATTTTAAAATGCCTGCCCAGACGATAPACITISad l^260^280^300^oJY 1^320^340^360CGAGGAGCTCATCGCGGTGATCAAGCAGAGGGATGGCCACATGACGGAGTTCCGCCAGGCCGTCGAGGAGGTCGTGGACTCUTCAAGGTTATCTTCGAGCGCGAGCCGRAGTATATCCCEELIAUIKURDOHATEFAQAUEEUUDSLKUIFEREPKYIP380^400^420^ Hincll^460^480AATCTTCGAGAGGATGCTTGAGCCGGAGCGCGTCATCATCTTCCGCGTGCCCTGGATGGATGACGCTGGACGCATCAACGTCAACCGCGGCTTCCGTGTCCAGTACAACTCTGCTCTCGGIFERALEPERUIIFRUPWADDAGRIHUHRGFRUQYHSALG500^520^540^560^580^0LW 72^600CCCCTACAAGGGTGGCCTCCGCTTCCACCCCTCTGTCAATUTTCGATTCTCAAGTTCCTCGGTTTCGAGCAGATCCTGRAGRACTCCCTCACCACGCTCCCGATGGGCGGCGGCAAGGGPYKGGLRFHPSUNLSILKFLGFE(1ILKHSLTTLPAGGGKG620^640^660^680^700^720CGGETCCGACTTTGACCCARAGGGCRAGTCCGACAACGAGGTCATOCGCTTCTGCCAGTCCTTCATGACCGAGCTCCAGAGGCACGTCGGCGCCGACACTGACGTTCCTGCCGGCGACATGSDFDPKOKSONEUMAPCOSFATELORHUGADTDUPAGDI740^760^780^800^820^840CGGCGTCGGCGCCCGCGAGATCGGGTACCTGTACGGACAGTACAAGCGCCTGAGGRACGAGTTCACAGGCGTCCTCACAGGCAAGAACGTCAAGTGGGGCGGGICTTTCATCAGGCCGGAGUGAREIGYLYGOYKALRNEFTGULTGKHUKUGGSFIRPE860^880^900^920^940^960GGCCACGGGCTATGGCGCTGTCTACTTCCTGGAGGAGATGTGCAAGGACAACAACACTGTGATCAGGGGTAAGRACGTCCTTCTTTCTGGCTCCGOCAACGTTGCCCAGTTTGCTTGCGAATGYGAUYFLEEMCKONNTUIRGKHULLSGSGHUACIFACE980^1000^1020^1040^1060^1080GARGCTCATTCAGCTCGGCGCARAGGICCTCACCTTCTCAGACTCCAACOGGACCATTGTCGACAAGGACGGGITCAACGAGGAGRAGCTGGCCCACCTCATGTACCTCAAGAACGAGAAKLIQLGAKULTFSDSNGTIUDKDGFHEEKLAHLAYLKHEK1100^1120^1140^1160^1180^1200GCGTGGGCGTGTTTCCGAGTICAAGGACAAGTATCCCAGCGTCGCGTACTACGAGGGCAAGAAGCCTTGGGAGTOCTTCGAGGGCCAGATGGATTGCATCATGCCTTGCGCCACTCAGAAAGAUSEFKOKYPSUAYYEGKKPLIECFEGUADCIAPCATOH1220^1240^1260^1280^1300^1320CGAGGTUCCGGGGACGATGCGACGCGCCTGGTCGGCCTCGGCCTCAAGTTCGTGGCCGAGGGTGCGARCATOCCCTCCACGGCAGAGGCTOTTCACGTCTATCACGCCAAGGGCGTGATEUSGDDATALUGLGLKFUREGAHAPSTAEAUHUYHAKGUM1340^1360^1380^EcoRI^1400^1420^1440GTACGGACCCGCCAAGGCCAGCAACGCGGGCGGTOTTCTGTCTCCGGCCTCGAGATGTCCCAGRATTCCGTGAGGCTCCAGTGGACGGCTGAGGAGGTCGACCAGAAGCTCCGCGGCATYGPAKASHAGGUSUSGLEASONSURLQUTREEUDOKLAGI1960^1480^1500^1520^1540^1560CATGAGGGGCATCTTCGTCGCCTGCCGCGACACTGCCAAGRAGTATGGGCACCCCAAGAACTACCAGATGGGCGCGRACATCGCCGGGTTCCTGAAGGICGCAGACTCTATGATCGAGCAARGIFUACROTAKKYGHPKNYQMGANIAGFLKUADSMIEQ1580^0000^ 1620^1640^1660^MarlGGGCTGCGTGTGAGGGCTGAAGTGAATATTTACCTTUCCGGGITCGACGGGCGTGCTTGACGATGCGGRAGTCGCTGTCCGTTCCAACGGCTATCTGCCCCAGGACCCCCATCCCGGCGCCUPvulICCGATCAGCTGFigure 5: Nucleotide sequence of the Giardia GDH geneThe nucleotide sequence of the 1691 nucleotide long HindlII-Pvull fragment isillustrated. The location of the indicated restriction enzyme sites are: HindlIl (1), Sad(245), Hindi (441), EcoRI (1384), Nan (1677) and PvuI1(1686). These are the sites used togenerate fragments for the Si nuclease protection experiments and other experimentsdiscussed in the main text. The sequences complementary to oligonucleotides oWL72and oJY1 are indicated by overlines; the interruptions in the oWL72 overline representnucleotides that are not complementary to the GDH sequence. The position of 5'(•) and3'(o) transcript end sites are indicated. A putative polyadenylation signal is underlinedat positions 1581-1586.29GDH-like protein. The protein-coding portion of this gene has a G+C content of 61%.The oligonucleotide oLW72, used in the initial PCR amplification, matches the GDHsequence at 14 out of 21 nucleotide positions located between 584-604. The other end ofthe pJYPCR10.5 cDNA clone corresponds to genomic nucleotide position 220 and thecDNA sequence agrees precisely with the genomic sequence between positions 220 and583.The insert from the clone pJYPCR10.5 was also used to probe a G. lamblia cDNAlibrary in Ast10. The insert of one of the hybridizing cDNA clones was subcloned intothe EcoRl site of pGEM3Zf(-), designated pJYc4-7, for sequencing. The cDNA sequenceextended from nucleotide position 234 to the EcoRI site at 1389 and was colinear andidentical in sequence to the genomic clone, implying that there are no introns withinthis region of the Giardia GDH-like gene.Characterization of the GDH gene transcripts The 5' ends of the transcripts derived from this open reading frame werecharacterized by primer extension and Si nuclease protection. A 441 bp HindIII-Hincllfragment overlapping the 5' portion of the open reading frame was 5' end-labeled in the(-) DNA strand at nucleotide position 442. Following its hybridization to total GiardiaRNA and Si nuclease digestion of the heteroduplex, a 222 nucleotide long doublet,representing the protected product, was observed (Fig. 6B). This places the 5' mRNAend in the vicinity of nucleotide position 221, three nucleotides in front of the ATGtranslation initiation codon.This result was confirmed using primer extension analysis using a 20-meroligonucleotide (oJY1), complementary to the (+) DNA strand of the genomic sequencebetween nucleotide position 299-319, as presented in figure 5. Using total Giardia RNAas template, a major extension product 98 nucleotides in length was observed; minorproducts one to three nucleotides longer were also apparent (Fig. 6A). These 5' end sites30Figure 6: Mapping of the 5' and 3' end sites of the GDH mRNA transcriptPrimer extension or Si nuclease protection assays were used to locate the 5' and 3' endsof the GDH mRNA. Radioactive products of these assays were separated on denaturingpolyacrylamide gels and subjected to autoradiography. (A) Primer extension analysisusing 5' end-labeled oJY1 oligonucleotide and total Giardia RNA was carried out asdescribed in the "Materials and Methods". The length of the product was estimatedusing a DNA sequence ladder generated using the same oJY1 oligonucleotide as asequencing primer. The major end site (large arrow) occurs three nucleotides in front ofthe ATG translation initiation codon. Minor end sites (small arrows) occur four to sixnucleotides in front of the ATG codon. The A of the ATG codon, at position 224, isindicated within the sequence shown. (B) Nuclease Si protection by mRNA of an end-labeled DNA probe. A 441 nucleotide long HindIII-HincII fragment was 5 end- labeledin the (-) DNA strand at the Hindi site at position 443. Following hybridization to totalGiardia RNA and Si digestion, a protected product of about 222 nucleotides in lengthwas observed. The molecular lengths of size standards in nucleotides are indicated onthe left. (C) Nuclease Si protection by mRNA of an end-labeled DNA probe. A 296nucleotide long EcoRI-NarI fragment was labeled at the 3' end of the (-) DNA strand atthe EcoRI site at positions 1385-1388. The molecular length standard was generated byemploying the G and G+A Maxam-Gilbert sequencing reaction on the same end-labeledfragment. The protected product 207-210 nucleotides in length positions the 3'transcript end sites near nucleotide positions 1594-1597. Nucleotide position 1593 isindicated within the sequence shown.31occur at nucleotide positions 218-221 in the genomic DNA sequence, and correspond tothe 5' end sites determined by Si nuclease protection. Furthermore, they alsocorrespond to the 5' end of the original pJYPCR10.5 cDNA insert at nucleotide position221.The 3' transcript end sites were also located by Si nuclease protection (Fig. 6C). The296 bp EcoRI-NarI fragment overlapping the 3' end of the open reading frame was 3'end-labeled in the (-) DNA strand at the E coRI site (nucleotide position 1285-8).Following hybridization to total Giardia RNA and digestion with Si nuclease, protectedfragments of 207-210 nucleotides in length were observed. This places the 3' transcriptend site at nucleotide positions 1594-1597, preceded by a putative Giardiapolyadenylation signal sequence AGTGAA at nucleotide position 1478-1483.These results demonstrate that the major transcript of the GDH gene excluding thepoly(A) tail is about 1380 nucleotides in length. The major 5' end site is located onlythree nucleotides in front of the putative ATG translation initiation codon and thetranscript is probably polyadenylated at a site that is 21-24 nucleotides beyond thetranslation termination site. The 5' untranslated leader is not extended by transsplicingas indicated by primer extension analysis; presumably, the observed 5' end siterepresents the position of transcription initiation.Genomic Southern hybridizationTo determine if the Giardia haploid genome contains other copies of the cloned GDHgene or related genes, restriction enzyme digests of genomic DNA were probed withthe 2.4 kb Pvull fragment under high stringency conditions in Southern hybridizations(Fig. 7). Only single hybridizing bands, all greater than 6 kb in length, were evident inBamHI, BglII, EcoRI, and PstI digests. These fragments span a region greater than 9 kbin length and encompass the entire GDH gene, which implies that the haploid genomeprobably contains a single GDH-like sequence. The partial restriction enzyme map of32IlfTIJr uunFigure 7: Genomic Southern hybridization to identify fragments encoding NADP-GDH or related proteins(A) Genomic G. lamblia DNA was digested with different restriction enzymes andseparated on a 0.8% agarose gel. Following transfer to nylon membrane, the DNA wasprobed with the 2.4 kb PvuII fragment from the pJY11 clone. (B) A partial restrictionmap of the genomic DNA containing the GDH gene was constructed from the Southernhybridization results as illustrated and from other hybridization results using the insertfrom pJYPCR10.5 as the probe (data not shown). The nucleotide scale is presentedabove, the restriction map in the middle, and the position and polarity of the GDH geneon the 2.4 kb Pvull fragment on the bottom. Restrictions sites are as follows: E, EcoRl;B, Bam H, Hind III; P, Pvull; C,33the region, along with the position and orientation of the GDH gene is included infigure 7. It is interesting to note that only a single EcoRT fragment was detected in theseexperiments in spite of the fact that this enzyme cuts once asymmetrically within the2.4 kb PvuII probe. I suspect that the second fragment is very large because, byempirical observations, Giardia DNA appears to be extremely deficient in EcoRIrecognition sites. This large fragment may transfer inefficiently during the blottingprocess or may hybridize poorly because it overlaps the probe by only 300 nucleotides,and therefore has escaped detection. The other end of the detectable EcoRI fragment hasbeen mapped to a position about 9 kb upstream of the GDH coding region.Enzyme activity The presence of mRNA transcripts from the putative NADP-dependent GDH genepredicts that Giardia cell extracts should exhibit this enzymatic activity. In the presenceof the substrates a-ketoglutarate and NH4C1, a crude cell lysate exhibited a specificactivity of 210 ± 20 mU per mg of protein. No significant NAD-dependent GDH activitywas detected in either the direction of reductive amination or oxidative deamination.The fungi Neurospora crassa and Saccharomyces cerevisiae and the protozoanTrypanosoma cruzi have, in addition to a class I NADP-dependent GDH, a class II NAD-dependent GDH (Kinnaird and Fincham, 1983; Cazzulo, 1984; Moye et al., 1985; Nagasuand Hall, 1985; Cazzulo et al., 1988; Vierula and Kapoor, 1989; Miller and Magasanik,1990). A clone of the gene encoding the S. cerevisiae NAD-GDH was used to probe theGiardia genome by Southern blot analysis to identify the homologous gene. No specifichybridization was detected. Together with the enzyme assays, these results indicatethat Giardia has the single NADP-dependent GDH.34Sequence alignments The amino acid sequence specified by the Giardia GDH gene was aligned over itsentire length to the amino acid sequence of the NADP-dependent GDHs fromNeurospora crassa (a fungus) and Escherichia coli (a eubacterium) (Figure 8). Thealignments are disrupted by 11 separate gaps presumed to reflect deletion or insertionevents during the course of evolution. Of the 11 gap positions, seven are shared jointlybetween E. coli and Giardia and only three are shared by Neurospora and Giardia(Table I). Neurospora and E. coli have no gaps in common. The remaining gap at 310-312, because it is different in each sequence, is phylogenetically uninformative.It is also surprising to see that in pairwise comparisons at common positions, theGiardia sequence is 55% and 58% identical to the N. crassa and E. co/i sequence,respectively (Table I). The E. co/i sequence exhibits 56% identity with the Neurosporasequence, and nearly 47% of the commonly shared amino acid positions are identical inall three sequences. These amino acid identity data argue that the Giardia GDH proteinsequence is at least as closely related to the eubacterial sequence as it is to the fungalsequence. This conclusion is further supported by the distribution of the gap positionsin the alignment of the proteins.3510^ 20^ 30Gla^MP AQT 1 EEL 1 AU 1 KQADGHMTEFRQAUEEUUDSL••KU 1Hcr SHLPSEP--E--S,KOLAYT-EHSSLEco^11 D QTYSL-SFLHHUQK--PHQ---A---R--MTT-••LIPF40^ 50^ 60^ 70Gla^FEREPKY•• IP I FERMLEPERU I I FRUPL1 MDDAGR I HUHRHcr^-QKH-E-••R T AL TURS I ^ Q^U^E^11^HUQ-__Eco^L -0^AQMSLL--L U  Q^u^U^RHQ-Q -- -^80 90^ 100 110Gla^6 FRUQYHSFILGPYKGGL RFHPSUHLS I LKFLGFEQ I LKHSHcr^Y F^ L ^  F^AEco^ALI- --FS^1 ^ t1  IF--A120 130^ 140^ 150Gla^L TTLPI1GG 6 KGGSDFDPKGKSDHEUMRFCQSFMTELQRHUticr^--GS ^ R  A^1 R- --CAF- A --HK- 1Eco   EG L^Y^L160^ 170^ 180^ 190Gla^GAD TDUP AGD 1 GUGARE I GYLYGQYKRLRHEFTGULTGKHliar G^ ri F^A^AKAR^R^E ^ 6Eco 6^ U^Ft1 A-ri 11 -K -S-HTAC-F - -- G200^ 210 220^ 230Gla^UKL1GGSF IRPEATGYGAUYFLEEMCK•DHHTUIRGKHULLHcr^LS ----L  L^VUGH-LEYSGFIGSYA--A-R-Eco^L SF^L ^  L T^A^L^•RHGMGFE-1113 -SU240 250^ 260 270Gla^SGSGHURQFACEKL I QL 6 AKUL TFSDSHGT I U••DKDGFHHcr 1?^ AL-- I E--- T-USL---K -AL -A TOES- I TEco   Y^1^AMEF--R- I -A- --S- -U-••-ES-- T280 290^ 300^ 310610^EEKL AHLMYLKNEKRGRUSEFKDKYPSUAYYEGKKPIJECFHcr^U-D I MAU-A I -•-A-QSLTS-QHA•GHLKIJI --Ft R--L•HEco^K----R- I E I -F1 SAD- --ADYRKEFG•LU-L--QQ--•••320^ 330^ 340^ 350Gla^EGQMDC I MPCATQHEUSGDDATRLUGLGLKFUAEGAHMPSHer^U-KU- 1 AL  KEE^EG^L AA^C ^ S--CCEco^SLPU- I AL ^  U^A^HQ^1 AN^U^R  T360 370 380Gla^TAEAUHUYH ^ AKGUMYGPAKASHROGUSUSGLE 11Hcr^-L-- I E-FEHHFIKEKKGEA^A^G^A^C^A ^Eco^- 1 --TELFQ ^ QAG^LFA^G^A  AT 390^ 400^ 410 420Gla^SQHSURLQUTAEEUDQKLRG I MR 6 I FUACRDTAKKYGHPKHer^A -- -Q-- 11 --QA---E--KD--KHA-FHGLH-- - T-UEF1 AEco^A-- A A - -G-K --K --AR-HH--LD-HH--UEHGGE•-EQT430^ 440Gla^HYQM ^ GAN I AGFLKUADSM I EQGCUHcr^EGELPSLUR^S ^ U^Q A -HD --DU ll S K HEco^--UQ     U A^L A --U IFigure 8: Glutamate dehydrogenase protein sequence alignmentsThe NADP-glutamate dehydrogenase amino acid sequences from Neurospora crassa(Ncr) and Escherichia coli (Eco) are aligned to the predicted Giardia lamblia (Gla) proteinsequence. Positions in N. crassa and E. coil sequences that are identical to the G. lambliaamino acid are indicated by dashes (-); positions of deletions or insertions necessary tomaintain optimal alignment are indicated by filled circles (.). The amino acidnumbering is based on the G. lamblia protein sequence. The amino acids that arespecified by codons which are interrupted by introns in the N. crassa gene are boxed.36Table I: Comparison of homologue GDH proteinsComparisonsl Common AminoAcid Positions2Amino AcidIdentities3Gaps inCommon4Gla Ncr 432 239 (55.3 `)/0) 3Gla Eco 442 255 (57.7 `)/0) 7Ncr Eco 428 240 (56.1 %) 0Gla Ncr Eco5 428 200 (46.7 %)lAbbreviations are: Gla, Giardia lamblia; Eco, Escherichia coli; Ncr, Neurospora crassa.2Common amino acid positions represent the number of shared amino acid positions ina pairwise alignment.3Amino acid identities are the number of positions where identical amino acids occur inthe pairwise alignments. The percentages of amino acid identities are in brackets.4Gaps in common are those that occur at the same position in two of the sequences thatare required to maintain overall alignment with the third sequence.5The bottom row illustrates the common amino acid positions and the amino acididentities shared between the three sequences.37DiscussionFunction of the GDH proteinIn both eukaryotic and prokaryotic organisms, the primary role of NADP-dependentglutamate dehydrogenase is normally to produce glutamate for the support of proteinsynthesis. In contrast, anaerobic Giardia utilize this enzyme almost exclusively tomaintain a redox balance within the intracellular NAD(P) pool (Paget et al., 1993). In theanaerobic metabolism of Giardia, glucose or other carbohydrates are converted topyruvate via the Embden-Meyerhoff pathway with the concomitant reduction ofNAD(P) (Lindmark, 1980; Jarroll et al., 1989). Pyruvate is then converted to a variety ofend products including acetate, ethanol, CO2, and alanine (Fig. 9) (Paget et al., 1993).The conversion to alanine utilizes two coupled reactions:glutamatedehydrogenase^a—ketoglutarate + NH4+ + NADPH^ H20 + glutamate + NADP+andalanirteaminotransferase^pyruvate + glutamate ^ alanine + a-ketoglutarateA number of unicellular eukaryotes (fungi and protozoa) possess a second GDH thatutilizes NAD rather than NADP as coenzyme (Cazzulo et al., 1988; Stevens et al., 1989;Vierula and Kapoor, 1989; Miller and Magasanik, 1990). This activity is associated withthe conversion of glutamate to a-ketoglutarate for catabolic purposes. Southern blotanalysis of Giardia genomic DNA using the S. cerevisiae NAD-GDH as a probe did notdetect the presence of a related sequence. Furthermore, no NAD-dependentinterconversion between a-ketoglutarate and glutamate was evident in Giardia cellextracts; only the NADP-dependent activity was detectable.38AcetateEthanolAlanineNAD+ NADHCarbohydrate PyruvatealanineaminotransferaseGlutamate^a-KetoglutarateNADP-glutamatedehydrogenaser NADP+^ NADPH + NH3Figure 9: The role of NADP-GDH in Giardia metabolism.Giardia uses anaerobic glycolysis to generate pyruvate which is then converted to threeseparate end products, acetate, ethanol, and alanine. The role of NADP-glutamatedehydrogenase and alanine aminotransferase in the conversion of pyruvate to alanine isillustrated.39Evolutionary relationshipsThe deduced amino acid sequence of the Giardia GDH protein exhibits substantialsequence identity to numerous fungal and eubacterial NADP-GDH sequences. Figure 8illustrates the amino acid sequence alignment of the Giardia protein to the N. crassa andthe E. coli proteins. Comparison of amino acid gap positions and amino acid identitiesindicate that the Giardia GDH is no more similar to the Neurospora enzyme than it is tothe E. co/i enzyme. The same conclusion is obtained when the Giardia sequence iscompared to other fungal or other eubacterial NADP-GDH sequences (data not shown).If inheritance of the NADP-GDH gene has been entirely vertical from a commonancestor, these protein comparisons provide strong support for the hypothesis thatGiardia diverged very early from the main eukaryotic lineage. However, these sequencecomparisons were performed on GDHs from just three organisms by rudimentaryanalysis. Later studies by Benachenhou-Lahfa et al., (1993) using the Giardia GDHalong with class I GDHs from 20 other organisms in a more extensive analysis,suggested that the class I GDHs can be further sub-divided into two families with theGDHs from Giardia,Neurospora, and E. coll. belonging to the same family. Dependingon the portion of the GDH sequence and the method of tree reconstruction used, theGiardia sequence either groups with the eubacteria or the fungi sequences. This andother discrepancies in the trees derived from the GDH sequence comparisons byBenachenhou-Lahfa et al., suggest that the evolution of GDHs may be unusual and morecomplex than expected.Structure and expression of the GDH geneThe N. crassa NADP-GDH gene contains two spliceosomal introns that disrupt thegene within the codons at alignment positions 29 and 119 (Figure 8). In contrast, theGiardia genomic and cDNA clones (covering 87% of the coding region) are colinear andidentical. Together with Si nuclease protection experiments, these results demonstrate40that there are no introns in the Giardia GDH gene and that at least 87%, and probablythe entire RNA transcript, is unaffected by post-transcriptional RNA editing. Featuressuch as introns, RNA editing and transsplicing have yet to be detected in Giardia.However, since the number of fully characterized genes and transcripts is small, it isstill not possible to make general statements regarding the evolutionary origins of thesefeatures and their frequency and distribution within Giardia.There are only three to six nucleotides in front of the AUG translationinitiation codon on the Giardia GDH mRNA. Primer extension, Si nucleaseprotection and RNA-PCR experiments all indicate that this short 5' transcriptleader is not extended by a transsplicing event. The DNA sequence overlappingthe 5' transcript end site and the translation initiation codon is AT-rich(5'...AGAT1MAAATG...'3) and provides a striking contrast to the generally GC-richGiardia coding sequences. Other Giardia mRNA transcripts have also been shown tohave 5' leaders of one to six nucleotides in length (reviewed by Adam, 1991). For anumber of these, the 5' transcript end sites fall within AT-rich sequences which exhibit ahigh degree of sequence similarity to this GDH AT-rich sequence. In part 2 of thisreport, I present evidence that this motif is common to other Giardia genes and containsa specific binding site for a Giardia nuclear protein.It seems likely that Giardia uses a unique mechanism for translation initiation. Theribosome scanning model for translation initiation in other eukaryotes requires thesmall ribosomal subunit to scan along the mRNA and identify the AUG initiation codonwithin an appropriate context sequence (Kozak, 1989). Because the AUG initiationcodon is so close to the 5' end of the mRNA, ribosome scanning may not be required fortranslation initiation of Giardia mRNAs.The sequence AGTPuAA has been found in the 3' untranslated region of all Giardiagenes that have been examined. This sequence has been proposed to function as apolyadenylation signal because of (i) its location relative to the translation termination41codons and the site of polyadenylation, and (ii) its similarity to the eukaryoticconsensus polyadenylation signal (5'...AATAAA...3') (Holberton et al., 1988; Peattie et al.,1989). A sequence (5'...AGTGAA...3') matching the putative Giardia polyadenylationsignal occurs in the GDH transcript eight nucleotides downstream from thetranslational termination codon. Nuclease Si protection experiments suggest that thepolyadenylation site is located eight to 11 nucleotides beyond this putativepolyadenylation signal.The presence of a poly(A) tail on the GDH mRNA has not been verified directly inthis study. However, the Xgt10 library from which the GDH cDNA clone was isolatedwas constructed using oligomer d(T)n as first strand primer. Therefore, the ability toisolate a GDH clone from this library implies that the GDH mRNA is polyadenylated.During the construction of the cDNA library, the EcoRI sites within the cDNAproducts were protected by EcoRI methylase before EcoRI linkers were added to theends. The ligation products were then cut with EcoRI and ligated into the X vector. Thefinding that one end of the GDH cDNA clone from this library contains a naturallyoccurring EcoRI site within the GDH gene suggests that this site was not protected bythe methylase during the construction of the library.In summary, the results indicate that Giardia contains a single class I GDH gene.Similar to Giardia genes characterized previously, the GDH gene is GC-rich (61%), andits transcripts have short (three to six nucleotides) untranslated 5' sequences. Asequence matching a putative Giardia-specific poly(A) signal is found at in the 3'flanking region of the gene. The sites of poly(A) addition on the GDH mRNA, locatedby Si nuclease protection, is consistent with the usage of the putative poly(A) signal.Comparisons of amino acid identities suggests that the Giardia GDH is no more closelyrelated to homologous fungal enzymes than it is to eubacterial enzymes. The failure toidentify introns, transsplicing or RNA editing in the transcript of the GDH gene mayindicate that these molecular features are rare or absent in the Giardia lineage.42Part 2: Identification and analysis of conserved sequenceelements in the GDH gene promoterResultsIdentification of conserved sequence motifsThe upstream sequences from eleven Giardia genes, including the NADP-dependentglutamate dehydrogenase gene (GDH), are aligned in Fig. 10. It is evident from thealignment that core promoter elements such as the TATA-box (5TATA(A/T)AAT3') andthe pyrimidine-rich initiator element 5(C/T)A(C/T)TC(C/T)(C/T)(C/1)3' that arenormally associated with RNA polymerase II promoters of higher eukaryotes aremissing from these regions. However, two other conserved sequence motifs are found(Fig. 10). One is an AT-rich sequence element adjacent to the ATG translation initiationcodon and encompasses the transcription start sites. This element is present in ten ofthe eleven sequences examined. The AT -richness of this element resembles the TATA-box but its location resembles the initiator sequence of higher eukaryotes.The second conserved element is a CAAAT sequence which is referred to as theg-CAB element for the Giardia CAAT-box like sequence. This g-CAB motif shareslimited sequence identity with the CAAT-box found in mammalian promoters whichhave the consensus sequence, GG(C/T)CAATCT. This g-CAB element, in its completeform in either orientations, CAAAT or GTTTA, is found at least once betweennucleotides -50 and -25 in the promoter region of seven Giardia genes. In itstetranucleotide abbreviated forms, CAAA, AAAT, GTTT and TTTA, the g-CAB elementoccurs at least three times in nine of the eleven promoter sequences including once ortwice within most of the longer AT-rich motifs.^-70^-60^-50^-40^-30^-20^-10^I I I I I I ICITTGGGGATTACAATTCAAAGTTGGTCAATTTGAGCTICAAAAAATACCGCCAAAAAATGGCAATTCGG•••• 00 0TTTTAAAAATGGATGCATAGGATAAGCGAACTCATGATGGAAATTCAAATTACCTTAAAATATATTTTCTGAGCGCTTACA TTTAGAAAATGTCATTATTGCATATTTATTGTTACGCAATCTGAACTIGGCCGCCGCCAAATCCAGATTCCGCCGGCGCCGGATTATATGATTGCCATTTAAATTGCAATCTGAACTTGGCCGCCGCCAAATCCAGATTCCGCCGGCGCCGGTCGAGCTICTTCTGGAGCAGAAAACAATTTAGAATTCAAATCAGCAAATTCCAGAGICTGGACGGGCGGGACCITCGGGGGAGATGGCTGTTCTTGAAAAGCTGACCACAAATAACGCCUTAATTACAGGCGCCCCAGTAAAGGGTTATGCCCTGGACACCATTGGCTGCCTCGGATCAAGACTTCAAATTAGAAATTCTAGCCAGAAATTTAAAAATGATTTAAAAATGAAATAAAAATGATTTTAAAATGAATAAATCATGATGCTAAATAAAAGGCTAGTCGTTGTACAGCTTTGGTTTAAAAACCCTGATTCGAAAATTCTGATCAGAATAGAAGTCCTGCAGATCCTTIGGCATCCTGTACAAACTATTATGATTTCAAAAAATAGAAGGCTGGCCAT•ATAACCAGATGAATCTITTTAGCTITAAGTACCUTACTAACAGACTCTTACCCTATAATAAGCTGAGGG AAACTTCAATGAATAAATTATGTAAAAAGAATG43al-glardIna2-glardlnP-tubulln I0-tubulln IIa-tubullnNADP-GDHTIM (WB)TIM (GS)ARFVSP 1267ISA 417 •GGACACGCAAGAAGCTGTCTGTGGTAGCCTGGCCCCGGGCTTTGCGTTGGAAGCGCCACCCAGCAGGTCGGCGGCCTAATGTTCGGCAGA.10CCGAAGGTCCCGAAGCTACGTGAGATTCGTGAGATTCGTGAGTGCCCTGCCCAGCCTGCTCGTCCTGCTCGTGGCCAAGGCTTGTTGATAAT-rich element consensus: AATTAAAAATGFigure 10: Identification of two conserved sequence elements in the upstream regionof Giardia genes.The 5' sequences of the Giardia genes presented in the alignment are as follows from topto bottom: a1- and a2 giardins (Peattie et al., 1989; Alonso and Peattie, 1992), 13I-, r3II-and a-tubulins (Kirk-Mason et al., 1989), NADP-dependent glutamate dehydrogenase(GDH) (Yee and Dennis, 1992), triosephosphate isomerase (TIM) from the WB and GSGiardia isolates (M. Mowatt, pers. comm.), ADP-ribosylation factor (ARF) (Murtagh etal., 1992), cell surface antigens VSP 1267 (Mowatt et al., 1992) and TSA 417 (Gillin et al.,1990). The ATG translation initiation codon (0) and the sites of transcription initiation(.) are indicated. The transcripts of the first nine genes are initiated three to sixnucleotides in front of the ATG codon and the transcripts from the last two genesinitiate one nucleotide in front of the ATG codon. A conserved AT-rich DNA sequenceelement overlapping the 5' transcript end sites and the translation initiation codon isenclosed by the large box on the right. A second conserved sequence element g-CAB(with the sequence CAAAT or its inverted complement ATTTG) located between -25and -50, is indicated by heavy overlines and tetranucleotide sequences related to g-CABare underlined. The numbering of the nucleotides in the sequence is relative to the zeroposition assigned to the A of the ATG translation initiation codon.44The 5' sequences that appear most aberrant with respect to these motifs are from thevariant-specific antigen genes ISA 417 and VSP 1267. The former completely lacks theAT-rich element and contains only a single copy of a tetranucleotide version of theg-CAB element whereas the latter contains an AT-rich element with limited identity tothe consensus and no g-CAB-like elements. These observations suggest that theputative promoters of the variant-specific antigen genes differ from the putativepromoters of genes encoding structural proteins and enzymes involved in centralmetabolism. In this study I have examined the role of the AT-rich and the g-CABsequences in protein binding to the putative promoter region of the GDH gene.Protein binding to double-stranded DNATo investigate whether the AT-rich and the g-CAB motifs are recognition sites forprotein binding, two probes were constructed to represent each of these motifs for usein band-shift assays (Fig. 11). The oligonucleotide duplex 1a/b represents the region ofthe GDH sequence between -13 and +9, and contains the AT-rich element. Theoligonucleotide duplex 2a/b represents the region of the GDH sequence between-44 and -24, and contains the g-CAB element. Each of these oligonucleotide duplexeswas radiolabeled and incubated with nuclear protein extracts prepared from Giardia(Fig. 12A, B). On a non-denaturing polyacrylamide gel, both probes in the presence ofnuclear extract exhibited two bands with retarded mobilities relative to the mobility ofthe free probes. The two complexes, referred to as the upper and lower bands,represent specific binding of the nuclear proteins to the probes. Formation of thesecomplexes is inhibited by the addition of an excess of unlabeled probe and not by theaddition of the oligonucleotide duplex G4-a/b that contained the binding site for theyeast GAL4 protein (Johnston and Davis, 1984). (See Table II for oligonucleotidesequences). Surprisingly, oligonucleotides 1a/b and 2a/b, containing the AT-rich andthe g-CAB motifs respectively, can also cross-compete with each other with45-50^-40^-30^-20^-10^0^+10I I I I I I IAARGCTGACCACARATFIRCGCCITTRFITTFICRGGCGCCCCRGRTITTARAFITGCCTGCCCR--0.- .4--duplex oligo la/bduplex oligo 2a/bduplex oligo 3a/Bduplex oligo 4a/bGDH PCR54 fragmentFigure 11: Alignment of the probes used in the band-shift assays to the upstreamregion of the NADP-GDH gene.The top line represents the upstream region from the GDH gene with the AT-rich andthe g-CAB elements indicated by heavy overlines. Tetranucleotide sequences related tog-CAB are indicated by heavy underlines. The opposing arrows represent invertedrepeat sequences. The numbering of the nucleotides in the sequence is relative to thezero position assigned to the A of the ATG translation initiation codon. The duplexprobes are represented by double lines. See table II for the sequences of theoligonucleotide probes.46Figure 12: Binding assays with probes containing either the AT-rich or the g-CABsequence elements.Band-shift assays with either the duplex oligonucleotide 1a /b containing the AT-richelement (A) or with the duplex oligonucleotide 2a /b containing the g-CAB element (B)as probes. The doublets in the shifted complexes, referred to as the upper (U) and lower(L) bands, observed with each of the two probes, are indicated by the arrows.Increasing amounts of the unlabeled duplex oligonucleotides (20, 40 and 80 fold molarexcess relative to the probe) used as competitors are indicated at the top. (C) Duplexoligonucleotides la/b and 2a/b were used either alone or together as probes, with orwithout nuclear proteins (pro.) in binding assays as indicated in the table at the top.The amount of probe used in each binding assay was either 0.05 pmoles (1X) or 0.10pmoles (2X). All other conditions in the binding assays are as described in "Materialsand Methods". 03) UV cross-linking of proteins bound to duplex oligonucleotides 1a/band 2a/b was done as described in "Materials and Methods". The autoradiograph fromthe SDS-polyacrylamide gel is presented. The two outer lanes are the total UV cross-linked binding reactions and the middle lanes are the upper and lower bands in eachcomplex isolated after UV cross-linking and electrophoresis through a standard band-shift gel. The sizes from a protein molecular weight standard are indicated on the left.The arrow indicated the position of the 68 kD protein.47approximately equal affinities. This observation suggested that these two probes arebinding the same protein or proteins in the nuclear extract. To examine thisphenomenon further, the two duplex probes were incubated together in a band-shiftassay (Fig. 12C). No additional shifted complexes were observed in this experiment;this suggests that there is no interaction between the protein binding to these twoprobes.The results from the experiments illustrated in figure 12 indicate (i) that the twoduplex probes, containing the AT-rich and the g-CAB element respectively, bind to thesame protein or proteins in the nuclear extract and (ii) that each probe forms twodifferent complexes with slightly different electrophoretic mobilities. The relationshipbetween these different complexes and the nature of the protein or proteins bound toeach probe was analyzed by UV cross-linking. Duplicates of the binding reactions withprobes 1a /b and 2a/b were subjected to UV irradiation and one set of each reaction wassubjected to electrophoresis through a standard band-shift gel. The upper and lowershifted bands that were observed with each probe were cut out separately from the geland were loaded on to a SDS-polyacrylamide gel along with the remaining set of UVtreated binding reactions. The autoradiograph of this SDS-polyacrylamide gel showedthat the upper and lower bands obtained from both oligonucleotide probes migrated atthe same molecular weight of approximately 68 kD (Fig. 12D). A band of this molecularweight was also apparent in the unfractionated binding reactions. In a controlexperiment, gel slices of the shifted complexes were cut from band-shift gels of identicalbinding reactions that were not cross-linked. DNA recovered from these gel slices wassubjected to electrophoresis on a SDS-polyacrylamide gel. No radiolabeled bands otherthan the bands representing the free probes were observed in the autoradiograph of thisgel (data not shown).The next set of experiments used a 54 bp fragment (GDH PCR54) containing boththe AT-rich and the g-CAB motifs (Fig. 11). When this fragment was radiolabeled and48used in band-shift assays, two shifted complexes were observed which I refer to as Cland C2 (Fig. 13). These complexes represented specific binding of proteins by GDHPCR54 as the formation of these complexes was not inhibited by the addition of anexcess of duplex oligonucleotides containing either the yeast GAL4 binding site (G4-a /b) or the adenovirus EIB TATA-box (E1B-a/b). Formation of the complexes wasinhibited by the addition of oligonucleotides containing either the AT-rich element(1a/b) or the g-CAB element (2a/b). Furthermore, the two complexes appear to exhibitdifferential competition, with the upper C2 band being more sensitive to competitionthan the lower Cl band (Fig. 13A). This differential competition of the two complexesmay be related to the ordered appearance of the complexes in the presence of increasingprotein concentration in binding reactions with the GDH PCR 54 probe (Fig. 13B). Theappearance of the upper complex (C2) requires a higher concentration of protein thanthat needed for the appearance of the lower complex (Cl).A duplex oligonucleotide (3a /b), representing the region between the g-CAB andthe AT-rich motifs (Fig. 11) was radiolabeled and used in band-shift assays to determineits ability to bind proteins (Fig. 14A). Protein binding by this probe cannot be inhibitedby duplex oligonucleotides containing the yeast GAL4 binding site (G4-a/b) but isinhibited by oligonucleotides containing the AT-rich (1a/b) and the g-CAB (2a/b)elements. However, the competition for protein binding to the 3a/b probe by an excessof unlabeled 3a /b is weak as compared to the competition by 1a /b and 2a/b.Oligonucleotide 3a/b also acted as a weak competitor against protein binding tooligonucleotides 1a/b, 2a/b and GDH PCR54 (Fig. 14B).Protein binding to single-stranded DNASince I had originally interpreted the above results to indicate that the 68 kD proteinbinds specifically to double-stranded DNA, I next wanted to determine whether theprotein can also bind single-stranded DNA. Single-stranded oligonucleotides were49Figure 13: Band-shift assays with GDH PCR54 containing both the AT-rich and theg-CAB elements.(A) Competitors used against protein binding to the GDH PCR54 probe are indicated atthe top. Competition was at 20, 40 and 80 fold molar excess of the unlabeled DNAsrelative to the probe. The two shifted complexes, Cl and C2, are indicated by thearrows. (B) Increasing amounts of protein extract was added to binding reactions withthe GDH PCR54 probe. The amount of NaC1 added to each of the binding reactionswas adjusted to maintain a constant concentration of 100 mM to compensate for theincreasing amounts of NaC1 introduced with the increasing amounts of the proteinextracts added.50Figure 14: Band-shift assays with duplex oligonucleotide 3a/b containing neither theAT-rich nor the g-CAB elements.(A) Competitors used against protein binding to the duplex 3a/b probe are indicated atthe top. Competition was at 20, 40, and 80 fold molar excess of the unlabeledoligonucleotides relative to the probe. The shifted complex representing specificbinding is indicated by the arrow. (B) Duplex oligonucleotide 3a/b was used as acompetitor in protein binding assays against the duplex probes listed at the top. Lanes:a, probe alone; b, probe plus proteins; c and d, reactions with 40 and 80 fold molarexcess of unlabeled duplex 3A/B as competitor, respectively.51used as either probes or competitors in band-shift assays (Fig. 15). Single-strandedoligonucleotides (both (+) and (-) strands) containing either the AT-rich (la, lb) or theg-CAB (2a, 2b) element can compete for protein binding to GDH PCR54 fragment (Fig.15A). In contrast, only the (+) strand of the region between the AT-rich and the g-CABelement (oligonucleotide 3a) can compete for binding. Neither strand of the DNAscontaining the yeast GAL4 binding site (G4-a, G4-b) or the adenovirus ElB TATA box(El-a, El-b) can act as competitor for protein binding to GDH PCR54.When the single-stranded oligonucleotides were used as probes in band-shift assays,both (+) and (-) strands containing the AT-rich (la, lb) and the g-CAB (2a, 2b) elementsbound proteins whereas only the (+) strand of the region between the two elements (3a)bound protein. The protein binding to these single-stranded probes can be inhibited bythe addition of an excess of unlabeled GDH PCR54 (Fig. 15B).DNA sequence requirements for protein bindingExamination of the oligonucleotides used in the binding assays revealed that there isa TTT sequence in all the single-stranded DNA probes that bind protein (Table II). ThisTTT sequence is absent in the probes that did not bind. To investigate the possibilitythat a TIT sequence is required for the binding of the 68 kD protein, oligonucleotidescontaining point mutations were tested in binding assays (Fig. 15B). Probe 3a*, whichdiffers from oligonucleotide 3a by a single point mutation that converts a TTTtrinucleotide to a TGT sequence, was unable to bind the 68 kD protein. Probe 3b*,which differ from oligonucleotide 3b by a TGT to ITT conversion, did bind the 68 kDprotein.Similarly, complementary oligonucleotides 4b and 4a, representing a sequence of theupstream region of the GDH gene containing one and no TTT sequence, respectively(Fig. 11), can bind the 68 kD protein when annealed together as duplex DNA but onlyoligonucleotide 4a can bind as a single-stranded DNA probe (Fig. 15C). Furthermore,52Figure 15: Band-shift assays with single-stranded oligonucleotides as either probesor competitors.Oligonucleotides representing the (+) DNA strand are given the 'a' designation andthose representing the (-) strand are given the 'b' designation in the name. (A) Proteinbinding to the duplex GDH PCR54 probe was inhibited by increasing amounts of thesingle-stranded oligonucleotides listed at the top. For competitors la, lb, 2a, and 2b, theamounts used were 20,40 and 80 fold molar excess over the probe. For 3a, 3h, G4-a, G4-b, El-a, and El-b, 40 and 80 fold molar excess over the probe were used. (B) Thedifferent single-stranded probes used in the binding assays are listed at the top. Theprobes 3a* and 3h* are oligonucleotides containing point mutations (see main text andtable I). Lanes: a, probe alone; b, probe plus proteins; c and d, reactions with 40 and 80fold molar excess of unlabeled duplex GDH PCR54 as competitor, respectively. (C) Thedifferent probes used in the binding assays are listed at the top. Lanes: (-) probe alone;(+) probe plus proteins. (D) .4a /4b represents the binding reaction with the duplexprobe consisting of radiolabeled oligonucleotide 4a annealed to unlabeledoligonucleotide 4b. Similarly, 4a/.4b represents the binding reaction with unlabeled 4aannealed to radiolabeled 4b.53binding of the 68 kD protein to 4a can be inhibited by the addition of oligo(dT)12_18(data not shown). Although these results suggest that a minimum sequence of a TTTtrinucleotide is required for the binding of the 68 kD protein, I wanted to determine ifthere are other sequence requirements. I tested the binding ability of the 68 kD proteinto a pair of complementary oligonucleotides (5A and 5B) derived from within thecoding region of the GDH gene (Fig. 15C). These oligonucleotides represents a portionof the GDH gene encoding amino acids 103 to 109 (LKFLGFE) at nucleotide position531-550 in the sequence presented in figure 5. Other than a single trinucleotide TTTsequence in 5A, there is no sequence similarity between these oligonucleotides andthose tested previously. Since the single-stranded oligonucleotide 5A and duplexoligonucleotide 5a/b can also bind the 68 kD protein, I must conclude that a single TITsequence is sufficient for binding of this protein to both single and double-strandedDNA. Table II presents a summary of the band-shift assay results.Although I assumed that the complexes observed with the duplex oligonucleotideprobes represented binding of the 68 kD protein to the double-stranded DNA, it ispossible that the protein is binding to single-stranded DNA melted from the duplexes.To investigate this possibility, oligonucleotide 4a, which does not contain any TTTsequences, was radiolabeled and annealed to its unlabeled complementaryoligonucleotide 4b, which contains one TTT sequence. After annealing, this mixturewas subjected to electrophoresis through a 20% non-denaturing polyacrylamide gelbeside a sample of single-stranded radiolabeled 4a. Examination of the autoradiographobtained from this gel revealed that >90% of the radiolabeled oligonucleotide 4a was inthe duplex form (data not shown). When this mixture was used in the band-shift assay,no shifted complex was observed. In the complementary experiment with the duplexprobe containing unlabeled oligonucleotide 4a and radiolabeled oligonucleotide 4b, ashifted complex was observed (Fig. 15D). Thus, only single-stranded DNA binding bythe 68 kD protein was detected in this experiment.54Table II: List of oligonucleotides used in band-shift assayssingle-strand sequence2 double-strand mobilitynamel (5'—>3') (duplex) name shiftla CCCAGATTTTAAAATGCCTGla/byes1b GGGCAGGCATTTTAAAATCTGGG yes2a GACCACAAATAACGCCTTT yes2a /b2b TTAAAGGCGTTATTTGTGGTC yes3a CGCCTTTAATTACAGGCGC yes3a /b3b GGGGCGCCTGTAATTAAAGGCG no•3a* CGCCTGTAATTACAGGCGC no•3b* GGGGCGCCTTTAATTAAAGGCG yes4a AAGCTGACCACAAATAACG no4a /b4b GGCGTTATTTGTGGTCAGCTT yes5a TCAAGTTCCTCGGTTTCGAG yes5a /b5b TGCTCGAAACCGAGGAACTTGA noG4-a TCGACGGAGTACTGTCCTCCG noG4-a/bG4-b TCGACGGAGGACAGTACTCCG noE1B-a CTAGAGGGTATATAATGCGCCAGCT noE1B-a/bE1B-b GGCGCATTATATACCCT no101igonucleotides representing the (+) DNA strand are given the 'a' designation andthose representing the (-) strand are given the 'b' designation in the name.2Sequences with three or more T's in a row are underlined and (0) indicates thenucleotide which was altered in the mutant oligonucleotides.55Protein binding to RNA The ability of the 68 kD protein to bind to RNA was examined. Transcripts derivedfrom the 5' flanking region of the GDH gene (position -223 to +21 relative to the zeroposition assigned to the A of the ATG translation initiation codon; nucleotide position 1to 245 in the sequence presented in figure 5) were synthesized in vitro using an SP6promoter system. The transcribed RNA contained three separate UUU sequencescorresponding to three TTT sequences on the (+) DNA strand. When the radiolabeledRNA was incubated with proteins from the nuclear extract, complexes were obtainedbut these could not be inhibited by an excess of unlabeled oligo(dT)12 _18 or byoligonucleotide 4b (data not shown). Similarly, an excess of unlabeled RNA could notinhibit protein binding to radiolabeled oligonucleotide 4b (data not shown). Theseresults suggest that the protein which is binding to TTT sequences on single-strandedDNA does not bind RNA containing UUU sequences.56DiscussionDistribution of the AT-rich and the g-CAB elements Two conserved sequences, referred to as the AT-rich and g-CAB elements, have beenidentified in the region upstream of the GDH gene and in other Giardia genes (Fig. 8).The AT-rich element has the consensus A(A/T)TTAAAAATG and contains both thestart sites for transcription initiation and the ATG translation initiation codon. Theg-CAB element with the consensus CAAAT, is present in one or more copies between25 to 50 nucleotides upstream of the sites of transcription initiation. Both of theseelements have a high AT content, in contrast to the normally GC-rich content of theGiardia genome. Melting of DNA at sites within these elements may facilitate theassembly of a RNA polymerase II preinitiation complex. In addition, the upstreamregion of the GDH gene contains two inverted repeats: one 5 bp long andencompassing the AT-rich element and the other 9 bp long with one mismatch (Fig. 9).Inverted repeat sequences are common recognition motifs for a number of regulatoryproteins including transcription factors. Furthermore, these AT-rich and g-CABelements are found in approximately the same position in the upstream region of allGiardia genes characterized, with the exception of the genes encoding variant surfaceantigens (Fig. 8).The absence of one or both of these elements from the variant-specific surfaceantigen genes (VSP 1267 and TSA 417) may reflect a different mechanism for theregulation of these genes. This is not entirely unexpected since the expression of theantigen genes is involved in the phenomenon of antigenic variation, a process thatmight require more complex regulation. Another difference between these genes is thatthe transcripts from the antigen genes contain an untranslated RNA leader sequenceconsisting of just a single nucleotide whereas the transcripts from the other genescontain leaders that are three to six nucleotides in length.57In the parasite Trypanosoma brucei, in which antigenic variation also occurs, theantigen genes are suspected to be transcribed by RNA polymerase I, as theirtranscription is a-amanitin resistant (Rudenko et al., 1989; Zomerdijk et al., 1991). This isin contrast to the a-amanitin sensitive transcription of other protein-encoding genes.Thus, the A/T rich and the g-CAB elements may only be prevalent in a class of Giardiagenes that excludes the variant antigen genes because of differences in either regulationrequirements or transcription by different RNA polymerases, or a combination of both.Sequence specificity of POT bindingA 68 kD protein that can bind in a sequence specific manner to single-stranded DNAfrom the upstream region of the GDH gene was identified from a Giardia nuclearextract. Sequence comparison of probes that are recognized by this 68 kD protein andbinding studies on mutant probes indicated that this protein requires a minimumrecognition sequence of TTT. This protein is named POT for its poly(T) binding ability.In the upstream region of the GDH gene there are four potential binding sites forPOT; two TTT sequences are on the (+) strand and two are on the (-) strand. Of these,two are within the AT-rich element, one on each strand, and one is within the g-CABelement on the (-) strand. The fourth site is located on the (+) strand between the twoelements. The locations of these binding sites are consistent with the observation thatoligonucleotide probes representing sub-sections of this region can cross-compete witheach other for the binding of POT. Duplex oligonucleotides la/b and 2a/b contain twopotential binding sites, one on each DNA strand, and 3a/b contains only a singlebinding site on its (+) strand. These differences in the numbers of binding sites areconsistent with the observation that oligonucleotides la/b and 2a/b haveapproximately equal affinities for POT whereas 3a /b has a lower affinity in band-shiftcompetition assays. When the different single-stranded oligonucleotides (1a, lb, 2a, 2b,3a, 3b*, 4a, 5a), each containing a single potential POT binding site, were used in cross-58competition binding assays, they all appear to have approximately equal affinities forPOT (data not shown).POT binding sites in the GDH PCR54 fragments Although there are four potential binding sites for POT in the GDH PCR54fragment, only two complexes (Cl and C2) are observed in gel mobility shift assays(Fig. 11). The differential competition and dependency on protein concentration for theformation of Cl and C2 suggest that these complexes represent two levels of POTbinding. If POT binds as a monomer, then Cl and C2 may represent binding at one andtwo sites, respectively. Steric hindrance could prevent the binding of POT to more thantwo of the four potential sites at any one time. Alternatively, if POT is binding as adimer, then Cl may represent binding of one dimer to two of the four sites and C2 mayrepresent binding of two dimers, one subunit binding at each of the four sites. In theabove explanations, I assumed that the GDH PCR54 probe remained as a duplex withlocalized melting at the TTT containing sequences where POT binds. If the GDH PCR54is completely melted in the complexes, then Cl may represent binding of one of the twosites available on each DNA strand and C2 may represent binding at both sites.Doublet band formations in some POT-DNA complexes The binding of POT to some DNA probes resulted in shifted complexes consisting ofdoublet bands. These doublets are distinct from the Cl and C2 complexes observedwith the GDH PCR54 probe by three criteria. First, the doublet bands are not subjectedto differential competition. Second, their appearance is simultaneous with increasingprotein concentrations and third, they cannot be correlated with the number of TTTsites on the probe. Finally, the doublets appear with some but not all single and double-stranded probes. For example, doublets were observed with probes la /b, 2a /b, lb, 2b,593b* and 4b but not with probes 3a /b, 5a/b, la, 3a and 5a. Thus, these bands areunlikely to represent different levels of protein binding.The doublet bands formed with duplex oligonucleotides la/b and 2a /b wereanalyzed separately by UV cross-linking (Fig. 10D). Both components of the doubletsexhibit cross-links to a protein of 68 kD. Since it was possible to separate the two bandsof the doublet on a non-denaturing polyacrylamide gel (used in the band-shift analysis)but not on a denaturing gel, I can speculate about some possible explanations for theseobservations.One possibility is that the two bands represent DNA binding by two differentproteins with the same molecular weights but with different conformations undernative gel conditions. The probe specific appearance of the doublet formation mayindicate that at least one of the proteins has a more restricted recognition site.Alternatively, the two bands may represent DNA binding by the same protein butwith modifications affecting the net charge of the protein. One such modification couldbe either a phosphorylation or a dephosphorylation so that the two protein formswould be separated under native gel conditions due to the differences in net charges butwould not be separated under denaturing gel conditions due to the slight change in thesizes of the two forms of the protein.A third possibility is that protein binding to some DNA sequences may induce DNAbending which results in two protein-DNA isomers with slightly different gelmobilities. If this conformational change is dependent on the nucleotide composition ofboth the protein binding site and the flanking sequences, then doublet formation wouldonly be observed upon protein binding to a subset of probes. The binding of theproducts of the oncogenes Jun and Fos to the AP-1 site is an example of proteinsinducing a conformational change to the DNA. Binding of Fos-Jun heterodimers bendDNA toward the major groove whereas Jun homodimers bend DNA toward the minor60groove and these species can be distinguished by gel mobility assays (Kerppola andCurran, 1991). I am unable to distinguish between these alternative from our results.Roles of POT in gene regulationRegulatory proteins that can interact with both single and double-stranded DNA ina sequence specific manner to control either replication or transcription have beenidentified in mammalian muscle cells, HeLa cells, and yeast (Hofmann and Gasser,1991; Santoro et al., 1991; Bergemann and Johnson, 1992). It is clear from the results thatPOT can bind to TTT sequences on single-stranded DNA but it is uncertain whetherPOT can also bind to double-stranded DNA. Although POT binding is observed inband-shift assays with double-stranded oligonucleotide probes, it is possible that POT isbinding to DNA that remains single-stranded after annealing of the complementaryoligonucleotides. This is less likely to be the explanation in the experiments with thedouble-stranded GDH PCR54 fragment which was generated by PCR and was gelpurified away from any possible contaminating single-stranded DNA prior to its use.Since GDH PCR54 is both a substrate and a competitor for POT binding, a helicaseactivity in the nuclear extract may be inferred. However, I do not know if this helicaseactivity is associated with POT or with another protein in the nuclear extract.Results from band-shift assays with the 4a /b duplexes labeled on only one strandsuggest that short oligonucleotide probes are completely denatured when bound toPOT. However, I do not know if GDH PCR54 is completely denatured by this helicaseactivity or whether there is only localized melting. Nevertheless, the ability of POT tobind to single-stranded DNA may play an important role in Giardia gene regulationsince the melting of duplex DNA and its maintenance in a single-stranded form is anessential feature in both transcription and DNA replication in other eukaryotes. Thebinding of POT to the TTT sequences found in the AT-rich elements, the g-CABelements, and to other TTT sequences in the upstream region of the GDH gene may61facilitate the melting of the DNA and allow the formation of a transcription initiationcomplex to occur within this region.Is POT a homologue of a transcription factor previously characterized in highereukaryotes? One possibility is that POT is related to the TATA-binding protein (TBP)and the AT-rich sequence is functionally equivalent to the TATA-box. Anotherpossibility is that the AT-rich motif functions as an initiator element and Giardia geneshave TATA-less promoters. In this scenario, POT binds to the initiator and recruits aTBP-like protein to this putative promoter region in an analogous manner as the TBP-tethering factors in others eukaryotes. A final alternative is that POT is unrelated to anyknown transcription factor and Giardia utilizes an unique system for the control oftranscription initiation.To my knowledge, this is the first example of a DNA binding protein with asequence specificity of only three nucleotides. Our results suggest that POT has thepotential to bind to every site in the Giardia genome where there are at least three T's ina row. Although the protein encoding regions in the Giardia genome have a strong GCbias, TTT sequences do occasionally occur in these and other non-promoter regions.What would be the function of POT if it binds to all these sites? One alternativepossibility is that POT is not directly involved with transcription but may be associatedwith other functions such as recombination or chromatin organization. The problemwith this alternative is that these types of proteins, for example topoisomerases andhistones, normally do not have a specific sequence requirement for their binding toDNA. Another possibility is that I am overlooking other physical requirements in ourin vitro binding assays that limit the binding of POT to a subset of TTT sequences in theGiardia genome in vivo. Physical requirements could include chromatin structure andDNA methylation. Yet another possibility is that POT binds cooperatively as multimersto a number of TTT sites which are located close together as in the upstream putativepromoter region of the GDH gene and not to relatively isolated TTT sites as found62within non-promoter regions. A final alternative is that the binding or activity of POT ismediated by other protein factors that I was not able to detect in these binding assays.In summary, two conserved sequence motifs, the AT-rich and the g-CAB elements,are found in front of a variety of Giardia genes. Their conservation and location relativeto the sites of transcription initiation suggest that these elements may be involved in theregulation of transcription. A 68 kD protein, called POT, binds to the TIT trinucleotidesfound on either one or both strands of the DNA encoding these elements in theupstream region of the GDH gene. Because of a current lack of a transformation systemor an in vitro transcription system for Giardia, it is not yet possible to define the precisefunction of these elements and the role of POT in the regulation of gene expression inthis organism. However, it may be possible to gain a better idea about the function ofPOT when a more purified preparation of POT is available and when the gene encodingPOT is obtained.Cavalier-Smith (1987) first suggested that Giardia has retained characteristics of aprimitive eukaryote from the observation that mitochondria and other organelles areabsent in this organism. This idea was supported later by phylogenetic analysis ofGiardia's 16S-like rRNA gene ( Sogin et al., 1988). Giardia is, therefore, a valuableresource for studying aspects of cellular function in an early eukaryote. Theidentification of a nuclear protein which can bind specifically to conserved sequenceelements in Giardia gene promoters may be the first step in understanding theregulation of transcription, and ultimately the control of gene expression in thisorganism.63ReferencesAdam, R.D. (1991). The biology of Giardia spp. Microbiol. Rev., 55: 706-732.Adam, R.D., A. Aggarwal, A.A. Lal, V.F. de la Cruz, T. McCutchan, and T.E. Nash(1988). Antigenic variation of a cysteine-rich protein in Giardia lamblia. J. Exp. Med.,167: 109-118.Adam, R.D., T.E. Nash, and T.E. Wellems (1988). The Giardia lamblia trophozoitecontains sets of closely related chromosomes. Nucleic Acids Res., 16: 4555-4567.Adam, R.D., Y.M. Yang, and T.E. Nash (1992). The cysteine-rich protein gene family ofGiardia lamblia: loss of the CRP170 gene in an antigenic variant. Mol. Cell. Biol.,12: 1194-1201.Alonso, R.A. and D.A. Peattie (1992). Nucleotide sequence of a second alpha giardingene and molecular analysis of the alpha giardin genes and transcripts in Giardialamblia. Mol. Biochem. Parasitol., 50: 95-104.Andrews, N.C. and D.V. Faller (1991). A rapid micropreparation technique forextraction of DNA-binding proteins from limiting numbers of mammalian cells.Nucleic Acids Res., 19: 2499.Benachenhou-Lahfa, N., P. Forterre, and B. Labedan (1993). Evolution of glutamatedehydrogenase genes: evidence for two paralogous protein families and unusualbranching patterns of the archaebacteria in the universal tree of life. J. Mol. Evol.,36: 335-346.Boothroyd, J.C., A. Wang, D.A. Campbell, and C.C. Wang (1987). An unusually compactribosomal DNA repeat in the protozoan Giardia lamblia. Nucleic Acids Res., 15: 4065-4084.Bradford, M.M. (1976). A rapid and sensitive method for the quantitation of microgramquantities of protein utilizing the principle of protein-dye binding. Anal. Biochem.,72: 248-254.Buratowski, S., S. Hahn, L. Guarente, and P.A. Sharp (1989). Five intermediatecomplexes in transcription initiation by RNA polymerase II. Cell, 56: 549-561.Cavalier-Smith, T. (1987). Eukaryotes with no mitochondria. Nature, 326: 332-333.Cazzulo, J.J., C. Nowicki, J.A. Santome, C. Wernstedt, and U. Hellman (1988). Aminoacid composition and N-terminal sequence of the NADP-linked glutamatedehydrogenase from Trypanosoma cruzi. FEMS Microbiology Letters, 56: 215-220.64Cormack, B.P. and K. Struhl (1992). The TATA-binding protein is required fortranscription by all three nuclear RNA polymerases in yeast cells. Cell, 69: 685-696.Crossley, R. and D.V. Holberton (1983). Characterization of proteins from thecytoskeleton of Giardia lamblia. J. Cell Sci., 59: 81-103.Crossley, R. and D.V. Holberton (1985). Assembly of 2.5 nm filaments from giardin, aprotein associated with cytoskeletal microtubules in Giardia. J. Cell Sci., 78: 205-231.Crossley, R., J. Marshall, J.T. Clark, and D.V. Holberton (1986). Immunocytochemicaldifferentiation of microtubules in the cytoskeleton of Giardia lamblia usingmonoclonal antibodies to alpha-tubulin and polyclonal antibodies to associated lowmolecular weight proteins. J. Cell Sci., 80: 233-252.Dynan, W.S., S. Sazer, R. Tjian, and R.T. Schimke (1986). Transcription factor Sp1recognizes a DNA sequence in the mouse dihydrofolate reductase promoter. Nature,319: 246-248.Dynan, W.S. and R. Tjian (1983). The promoter-specific transcription factor Sp1 binds toupstream sequences in the SV40 early promoter. Cell, 35: 79-87.Dynlacht, B.D., T. Hoey, and R. Tjian (1991). Isolation of coactivators associated with theTATA-binding protein that mediate transcriptional activation. Cell, 66: 563-576.Fan, J.B., S.H. Korman, C.R. Cantor, and C.L. Smith (1991). Giardia lamblia: haploidgenome size determined by pulsed field gel electrophoresis is less than 12 Mb.Nucleic Acids Res., 19: 1905-1908.Favaloro, J., R. Treisman, and R. Kamen (1980). Transcription maps of polyoma virus-specific RNA: analysis by two-dimensional nuclease 51 gel mapping. MethodsEnzymol., 65: 718-749.Feely, D.E., J.V. Schollmeyer, and S.L. Erlandsen (1982). Giardia: distribution ofcontractile proteins in the attachment organelle. Exp. Parasitol, 53: 145-154.Feinberg, A. and B. Vogelstein (1982). A technique for radiolabeling DNA restrictionendonuclease fragments to high specific activity. Anal. Biochem., 132: 6-13.Frohman, M.A., M.K. Dush, and G.R. Martin (1988). Rapid production of full-lengthcDNAs from rare transcripts: amplification using a single gene-specificoligonucleotide primer. Proc. Natl. Acad. Sci. U.S.A., 85: 8998-9002.Gillin, F.D., P. Hagblom, J. Harwood, S.B. Aley, D.S. Reiner, M. McCaffery, M. So, andD.G. Guiney (1990). Isolation and expression of the gene for a major surface proteinof Giardia lamblia. Proc. Natl. Acad. Sci. U.S.A., 87: 4463-4467.65Green, M.R. (1992). Transcriptional transgressions. Nature, 357: 364-365.Hawkins, A.R., S.J. Gurr, P. Montague, and J.R. Kinghorn (1989). Nucleotide sequenceand regulation of expression of the Aspergillus nidulans gdhA gene encoding NADPdependent glutamate dehydrogenase. Mol. Gen. Genet., 218: 105-111.Healey, A., R. Mitchell, J.A. Uperoft, P.F.L. Boreham, and P. Uperoft (1990). Completenucleotide sequence of the ribosomal RNA tandem repeat unit from Giardiaintestinalis. Nucleic Acids Res., 18: 4006.Heimann, J.D. and M.J. Chamberlin (1988). Structure and function of bacterial sigmafactors. Ann. Rev. Biochem., 57: 839-872.Henikoff, S. (1984). Unidirectional digestion with exonuclease III creates targetedbreakpoints for DNA sequencing. Gene 28: 351-359.Hoffman, A., E. Sinn, T. Yamamoto, J. Wang, A. Roy, M. Horikoshi, and R.G. Roeder(1990). Highly conserved core domain and unique N terminus with presumptiveregulatory motifs in a human TATA factor (TFIID). Nature, 346: 387-390.Holberton, D., D.A. Baker, and J. Marshall (1988). Segmented a-helical coiled-coilstructure of the protein giardin from the Giardia cytoskeleton. J. Mol. Biol., 204: 789-795.Holberton, D.V. (1981). Arrangement of subunits in microribbons from Giardia. J. CellSci., 47: 167-185.Horikoshi, M., C.K. Wang, H. Fujii, J.A. Cromlish, P.A. Weil, and R.G. Roeder (1989).Cloning and structure of a yeast gene encoding a general transcription initiationfactor TFIID that binds to the TATA box. Nature, 341: 299-303.Huang, J. and L.H.T. Van der Ploeg (1991). Maturation of polycistronic pre-mRNA inTrypanosoma brucei: analysis of trans splicing and poly(A) addition at nascent RNAtranscripts from the hsp70 locus. Mol. Cell. Biol., 11: 3180-3190.Jarroll, E.L., P. Manning, A. Berrada, D. Hare, and D.G. Lindmark (1989). Biochemistryand metabolism of Giardia lamblia. J. Protozool., 36: 190-197.Jarroll, E.L., P. Manning, D.G. Lindmark, J.R. Coggins, and S.L. Erlandsen (1989). Giardiacyst wall-specific carbohydrate: evidence for the presence of galactosamine. Mol.Biochem. Parasitol., 32: 121-132.Johnston, M. and R.W. Davis (1984). Sequences that regulate the divergent GAL1-GAL10 promoter in Saccharomyces cerevisiae. Mol. Cell. Biol., 4: 1440-1448.66Kabnick, K.S. and D.A. Peattie (1990). In situ analyses reveal that the two nuclei ofGiardia lamblia are equivalent. J. Cell Sci., 95: 353-360.Keister, D.B. (1983). Axenic culture of Giardia lamblia in TYI-S-33 medium supplementedwith bile. Trans. R. Soc. Trop. Med. Hyg., 77: 487-488.Kerppola, T.K. and T. Curran (1991). Fos-Jun heterodimers and Jun homodimers bendDNA in opposite orientations: implications for transcription factor cooperativity.Cell, 66: 317-326.Kirk-Mason, K.E., M.J. Turner, and P.R. Chakraborty (1989). Evidence for unusuallyshort tubulin mRNA leaders and characterization of tubulin genes in Giardia lamblia.Mol. Biochem. Parasitol., 36: 87-100.Kozak, M. (1989). The scanning model for translation: an update. J. Cell Biol., 108: 229-241.La Thangue, N.B. and P.W.J. Rigby (1988). Trans-acting protein factors and theregulation of eukaryotic transcription, pp. 1-41. In Transcription and Splicing,Frontiers in Molecular Biology, B.D. Hames and D.M. Glover (Eds.). IRL Press,Oxford, England, and Washington, D.C.Laemmli, U.K. (1970). Cleavage of structural proteins during the assembly of the headof the bacteriophage T4. Nature, 227: 680-685.Lindmark, D.G. (1980). Energy metabolism of the anaerobic protozoon, Giardia lamblia.Mol. Biochem. Parasitol., 1: 1-12.Lindmark, D.G. (1988). Giardia lamblia: localization of hydrolase activities in lysosome-like organelles of trophozoites. Exp. Parasitol., 65: 141-147.Mattaj, I.W., M.J. McPherson, and J.C. Wootton (1982). Localisation of a stronglyconserved section of coding sequence in glutamate dehydrogenase genes. FEBSLetters, 147: 21-25.Maxam, A.M. and W. Gilbert (1980). Sequencing end-labeled DNA with base-specificchemical cleavages. Methods Enzymol., 65: 499-560.Meyer, E.A. (1976). Giardia lamblia: isolation and axenic cultivation. Exp. Parasitol.,39: 101-105.Miller, S.M. and B. Magasanik (1990). Role of NAD-linked glutamate dehydrogenase innitrogen metabolism in Saccharomyces cerevisiae. J. Bacteriol., 172: 4927-4935.67Mowatt, M., A. Aggarwal, and T. Nash (1992). Carboxy-terminal sequence conversionamong variant-specific surface proteins of Giardia lamblia. Mol. Biochem. Parasitol.,49: 215-227.Muller, M. (1988). Energy metabolism of protozoa without mitochondria. Ann. Rev.Microbiol., 42: 465-488.Murtagh, J.J., Jr., M.R. Mowatt, C.-M. Lee, F.-J.S. Lee, K. Mishima, T.E. Nash, J. Moss,and M. Vaughan (1992). Guanine nucleotide-binding proteins in the intestinalparasite Giardia lamblia. J. Biol. Chem., 267: 9654-9662.Nakatani, Y., M. Horikoshi, M. Brenner, T. Yamamoto, F. Besnard, R.G. Roeder, and E.Freese (1990). A downstream initiation element required for efficient TATA boxbinding and in vitro function of TFIID. Nature, 348: 86-88.Nash, T. (1992). Surface antigen variability and variation in Giardia lamblia. Parasitol.Today, 8: 229-234.Nash, T.E., S.M. Banks, D.W. Alling, J.W. Merrit Jr., and J.T. Conrad (1990). Frequencyof variant antigens in Giardia lamblia. Exp. Parasitol., 71: 415-421.Nash, T.E., J.W.M. Jr., and J.T. Conrad (1991). Isolate and epitope variability insusceptibility of Giardia lamblia to intestinal proteases. Infect. Immun., 59: 1334-1340.Nash, T.E. and M.R. Mowatt (1992). Characterization of a Giardia lamblia variant-specificprotein (VSP) gene from isolate GS/M and estimation of the VSP gene repertoiresize. Mol. Biochem. Parasitol., 51: 219-228.Nash, T.E. and M.R. Mowatt (1993). Variant-specific surface proteins of Giardia lambliaare zinc-binding proteins. Proc. Natl. Acad. Sci. U.S.A., In press.Newman, A. (1987). Specific accessory sequences in Saccharomyces cerevisiae intronscontrol assembly of pre-mRNA into spliceosomes. EMBO J., 6: 3833-3839.Newton, C.H., J. Yee, L.C. Shimmin, and P.P. Dennis (1990). A family of genes encodethe multiple forms of the yeast L12 equivalent proteins and a single form of the L10equivalent proteins. J. Bacteriol., 172: 579-588.Paget, T.A., M.L. Kelly, E.L. Jarroll, D.G. Lindmark, and D. Lloyd (1993). The effects ofoxygen on fermentation in Giardia lamblia. Mol. Biochem. Parasitol., 57: 65-72.Peattie, D.A., R.A. Alonso, A. Hein, and J.P. Caulfield (1989). Ultrastructural localizationof giardins to the edges of disk microribbons of Giardia lamblia and the nucleotideand deduced protein sequence of alpha giardin. J. Cell Biol., 109: 2323-2335.68Peterson, M.G., N. Tanese, B.F. Pugh, and R. Tjian (1990). Functional domains andupstream activation properties of cloned human TATA binding protein. Science,248: 1625-1630.Pugh, B.F. and R. Tjian (1990). Mechanism of transcriptional activation by Sp1: evidencefor coactivators. Cell, 61: 1187-1197.Pugh, B.F. and R. Tjian (1991). Transcription from a TATA-less promoter requires amultisubunit TFIID complex. Genes & Dev., 5: 1935-1945.Reiner, D.S., M. McCaffery, and F.D. Gillin (1990). Sorting of cyst wall proteins to aregulated secretory pathway during differentiation of the primitive eukaryote,Giardia lamblia. Eur. J. Cell Biol., 53: 142-153.Reznikoff, W.S., D.A. Siegele, D.W. Cowing, and C.A. Gross (1985). The regulation oftranscription initiation in bacteria. Ann. Rev. Genet., 19: 355-387.Rigby, P.W.J. (1993). Three in one and one in three: it all depends on TBP. Cell, 72: 7-10.Roeder, R.G. (1991). The complexities of eukaryotic transcription initiation: regulationof preinitiation complex assembly. Trends Bio. Sci.,16: 402-408.Roy, A.L., M. Meisterernst, P. Pognonec, and R.G. Roeder (1991). Cooperativeinteraction of an initiator-binding transcription initiation factor and the helix-loop-helix activator USF. Nature, 354: 245-248.Rudenko, G., D. Bishop, K. Gottesdiener, and L.H. Van der Ploeg (1989). alpha-amanitinresistant transcription of protein coding genes in insect and bloodstream formTrypanosoma brucei. EMBO J., 8: 4259-4263.Sanger, F., S. Nicklen, and A.R. Coulson (1977). DNA sequencing with chainterminating inhibitors. Proc. Natl. Acad. Sci. U.S.A., 74: 5463-5467.Santoro, I.M., T.-M. Yi, and K. Walsh (1991). Identification of single-stranded-DNA-binding proteins that interact with muscle gene elements. Mol. Cell. Biol., 11: 1944-1953.Sehgal, A., N. Patil, and M. Chao (1988). A constitutive promoter directs expression ofthe nerve growth factor receptor gene. Mol. Cell. Biol., 8: 3160-3167.Seto, E., Y. Shi, and T. Shenk (1991). YY1 is an initiator sequence-binding protein thatdirects and activates transcription in vitro. Nature, 354: 241-245.Sharp, P.A. (1992). TATA-binding protein is a classless factor. Cell, 68: 819-821.69Sheffield, H.G. and B. Bjorvatn (1977). Ultrastructure of the cyst of Giardia lamblia. Am. J.Trop. Med. Hyg., 26: 23-30.Shi, Y., E. Seto, L.-S. Chang, and T. Shenk (1991). Transcription repression by YY1, ahuman GLI-Kruppel-related protein, and relief of repression by adenovirus E1Aprotein. Cell, 67: 377-388.Smale, S.T. and D. Baltimore (1989). The "initiator" as a transcription control element.Cell, 57: 103-113.Smale, S.T., M.C. Schmidt, A.J. Berk, and D. Baltimore (1990). Transcriptional activationby Sp1 as directed through TATA or initiator: Specific requirement for mammaliantranscription factor IID. Proc. Natl. Acad. Sci. U.S.A., 87: 4509-4513.Smith, E.L., B.M. Austen, K.M. Blumenthal, and J.F. Nyc (1975). Glutamatedehydrogenases, pp. 293-367. In The enzymes, Vol. XI, P. D. Boyer (Eds.), NewYork, Academic Press.Sogin, M.L., J.H. Gunderson, H.J. Elwood, R.A. Alonso, and D.A. Peattie (1989).Phylogenetic meaning of the kingdom concept: an unusual ribosomal RNA fromGiardia lamblia. Science, 243: 75-77.Stevens, L., D. Duncan, and P. Robertson (1989). Purification and characterisation ofNAD-glutamate dehydrogenase from Aspergillus nidulans. FEMS MicrobiologyLetters, 57: 173-178.Vierula, P.J. and M. Kapoor (1989). NAD-specific glutamate dehydrogenase ofNeurospora crassa. J. Biol. Chem., 264: 1108-1114.Weinbach, E.C., C.E. Claggett, D.B. Keister, and L.S. Diamond (1980). Respiratorymetabolism of Giardia lamblia. J. Parasitol., 66: 347-350.Weintraub, H., S.J. Tapscott, R.L. Davis, M.J. Thayer, M.A. Adam, A.B. Lassar, and A.D.Miller (1989). Activation of muscle-specific genes in pigment, nerve, fat, liver, andfibroblast cell lines by forced expression of MyoD. Proc. Natl. Acad. Sci. U.S.A.,86: 5434-5438.Yee, J. and P.P. Dennis (1992). Isolation and characterization of a NADP-dependentglutamate dehydrogenase gene from the primitive eucaryote Giardia lamblia. J. Biol.Chem., 267: 7539-7544.Zomerdijk, J.C., R. Kieft, and P. Borst (1991). Efficient production of functional mRNAmediated by RNA polymerase I in Trypanosoma brucei. Nature, 353: 772-775.


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