Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Genomic approaches to explore virulence in the fungal pathogen Cryptococcus neoformans Tangen, Kristin Lynne 2006

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


831-ubc_2006-303160.PDF [ 7.27MB ]
JSON: 831-1.0105466.json
JSON-LD: 831-1.0105466-ld.json
RDF/XML (Pretty): 831-1.0105466-rdf.xml
RDF/JSON: 831-1.0105466-rdf.json
Turtle: 831-1.0105466-turtle.txt
N-Triples: 831-1.0105466-rdf-ntriples.txt
Original Record: 831-1.0105466-source.json
Full Text

Full Text

GENOMIC APPROACHES TO EXPLORE VIRULENCEIN THE FUNGAL PATHOGEN CRYPTOCOCCUSNEOFORMANSbyKRISTIN LYNNE TANGENB.Sc. (Microbiology), The University of Victoria, 1997A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS OF THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(MICROBIOLOGY AND IMMUNOLOGY)THE UNIVERSITY OF BRITISH COLUMBIAMarch 2006© Kristin Lynne Tangen, 2006ABSTRACTThe fungal pathogen Cryptococcus neoformans is the leading cause of encephalitis inpeople with the acquired immunodeficiency syndrome (AIDS). The range of drugs available totreat C. neoformans is limited in number and efficacy; therefore, there is a dire need for new,more powerful therapeutics. To achieve this we need to better understand how the fungus cancause disease, which will lead to the identification of factors relating to virulence and ultimatelymay provide new drug targets. The plethora of emerging genomic resources has allowed fortargeted biology to focus on key genes in important cellular processes that relate to virulence.The work described in this thesis contributes to the development of genomic resources forbiological investigations in Cryptococcus neoformans. This work has three specific components:1) physical mapping of the genomes for strains serotypes A and D; 2) analysis of the low irontranscriptome in a serotype A strain; and 3) characterization of a putative siderophore (iron)transporter gene, SIT], that was identified by transcriptome analysis. The first componentinvolved the hybridization of 125 markers to a set of 9,216 BAC clones from the strain JEC21(serotype D) and to 6,528 clones from the strain H99 (serotype A). These data provided the firstgenome-wide comparison of gene synteny between two strains of the fungus, and linked contigsto specific karyotype bands. The second component of the work involved the analysis of the lowiron transcriptome for the serotype A strain, H99 using serial analysis of gene expression(SAGE). A number of interesting genes were identified in the low iron transcriptome includingthose involved with the response to stress and mechanisms of iron uptake. A key finding wasthat the low iron transcriptome was remarkably similar to the in vivo library from cells grown inrabbit cerebral spinal fluid (CSF) and significantly distinct from the libraries grown in yeastnutrient broth (YNB). The third component of the work focused on the gene SIT], whichencodes a putative siderophore transporter. The gene was characterized in three strainbackgrounds of varying virulence. This work showed that SIT] was important for ironutilization in conditions of low iron or when a siderophore was provided as the sole iron source.Further, there were pleiotrophic phenotypes for a number of virulence-related attributesincluding melanization, cAMP signaling and cell wall integrity. Finally, throughout the entirebody of work, multiple differences were identified between strains of the same or differentserotypes on a genomic and biological level, and this variation may lend insight into differencesin virulence between strains.11TABLE OF CONTENTSABSTRACT iiTABLE OF CONTENTS iiiLIST OF TABLES viiiLIST OF FIGURES xLIST OF ABBREVIATIONS xiiACKNOWLEDGEMENTS xivCHAPTER ONE: INTRODUCTION 1A. Cryptococcus neoformans and Cryptococcosis 1a. C. neoformans varieties, serotypes and molecular types 1b. C. neoformans infections 2c. Environmental sources of Cryptococcus 2d. C. neoformans life cycle 3e. Transmission of C. neoformans and etiology of cryptococcosis inhumans 3f. Epidemiology of cryptococcosis 3g. The immune response to C. neoformans 4h. Treatment of cryptococcosis 5B. Genomic Analyses of C. neoformans 6a. Karyotype studies 6b. Physical mapping projects 7c. EST sequencing projects 8d. Genome sequencing efforts 8e. Serial analysis of gene expression (SAGE) 9C. C. neoformans Signaling and Virulence Factors 11a. The cAMP and PKC]/MAP kinase pathways 11b. Structure and biology of the polysaccharide capsule: contribution tovirulence 13c. Melanin and its effects on virulence 14d. Mating type specificity: prevalence in clinical strains 15e. Superoxide dismutase (SOD) and effects 15f. Resistance to oxidative stress 16g. Mannose production and virulence 17h. The ability to grow at 37°C 18i. Other virulence factors 18D. Iron and Siderophore Transport 20a. Importance of iron transport and iron responsive genes 20b. Iron acquisition 20111c. Reductive fungal iron transport systems in fungi 20d. Reductive fungal iron transport systems in C. neoformans 21e. Siderophores and siderophore transport 22E. Melanin Production 23a. Melanin synthesis and placement 23b. Redox interactions of iron and melanin 24F. Rationale and Aims of this Study 24CHAPTER TWO: Physical Mapping of the Genomes of Serotype A and D Strainsof C. neoformans 27INTRODUCTION 27MATERIALS AND METHODS 31A. Filter Preparation and Layout 31B. Probe Design 31C. Hybridization Protocol 31D. Data Collection and Analysis 32E. Data Integration 32F. Southern Analysis of HindIII Digested DNA and CHEF Gel SeparatedChromosomes 32RESULTS 33A. Summary of Hybridization Experiments 33B. Integration of the Markers into the BAC Fingerprint Maps 38C. Comparison of the H99 and JEC21 Fingerprint Maps 43D. Relationship of Specific Contigs to Chromosome-Sized Bands From the C.neoformans Electrophoretic Karyotype 48SUMMARY 54CHAPTER THREE: Serial Analysis of Gene Expression (SAGE) of C. neoformansunder Iron Limited Conditions 55INTRODUCTION 55MATERIALS AND METHODS 57A. Determination of Growth Conditions 57a. Fungal strains 57b. Media and growth conditions 57c. Capsule microscopy 57B. RNA Extraction 57C. Construction of a Serial Analysis of Gene Expression (SAGE) Library 58D. Sequencing and Data Processing 59E. Tag Identification 59ivRESULTS .61A. Determination of Growth Conditions for SAGE Library Construction 61B. Summary of SAGE Library Construction 62C. SAGE Tag Annotation 64D. Analysis of the Low Iron Transcriptome 65E. Pairwise and Multiple Library Comparisons 70a. Pairwise comparison of 37°C low iron vs. 25°C libraries 70b. Pairwise comparison of 37°C low iron vs. 37°C libraries 77c. Pairwise comparison of 37°C low iron vs. in vivo libraries 84d. Iron regulated tags: Three-way overall comparison of the libraries37°C low iron, 25°C and 37°C in YNB 89e. Comparison of multiple libraries including: 25°C, 37°C low iron,37°C and in vivo libraries 91SUMMARY AND DISCUSSION 93CHAPTER FOUR: Analysis of the Siderophore Transporter Gene, SIT1 97INTRODUCTION 97MATERIALS AND METHODS 98A. in silico Analysis of SIT1p and SIT] 98a. Determination of the coding sequence and exon-intron boundariesof CnSitlp 98b. Comparison of SIT] and Sitlp in serotypes A, B and D 98c. Identification of siderophore-related genes in C. neoformans 98d. Comparison of Sitip to fungal homologs 99B. Strains and Growth Conditions 99C. RNA Isolation and Northern Analysis 100D. Construction of sit]:: URA5 (serotype D) and sit]::NEO (serotype A)Alleles 101E. Plate Assays 102a. Siderophore utilization 102b. Melanin production 102c. Temperature sensitivity 103d. Cell wall integrity 103F. Microscopy 104a.DIC 104b. Transmission electron microscopy 104G. Antibiotic Susceptibility Assays 104H. Virulence Assays in the Murine Model 105RESULTS 106A. Characterization of SIT] 106a. Comparison of SIT] and Sitip in serotypes A, B and D 106vb. Identification of siderophore-related genes in C. neoformans 110c. Comparison of Sitip to fungal homologs 113d. Expression of SIT] in low iron and iron replete conditions 118B. Construction of Siderophore Transporter Mutants 119a. SIT] gene and Sitlp protein structure 119b. Construction and confirmation of null mutants 121C. The Role of SIT] in Iron Acquisition 125a. Growth in various iron conditions 125b. Siderophore utilization 133D. Involvement of the cAMP Pathway in Siderophore Utilization 135E. SIT] and Melanization 137a. Production of DOPA melanin 137b. Presence of extracellular melanin-like granules 137c. Effects of other factors on melanization 140F. Influences of SIT] on Cell Wall Integrity 143a. Temperature sensitivity and cell wall integrity 143b. Alteration of cell wall structure 145G. Summary of Additional Phenotypic Analyses of sit] Mutants 148H. Drug Susceptibility 149I. Virulence in the Murine Model 150a. SerotypeD 150b. SerotypeA 151SUMMARY AND DISCUSSION 153CHAPTER FIVE: GENERAL DISCUSSION 155A. Phenotypic and Genomic Differences Between Serotypes A and D Strains. 155B. Iron-regulated Genes in C. neoformans 157C. The Role of SIT] in Iron Acquisition 159D. The Role of SIT] in the Structure of the Cell Wall 161E. The Possible Role of SIT] in Endosomal Trafficking 162F. The Potential Influences of the cAMP and PKC] Signal TransductionPathways on SIT] 163G. Susceptibility of sit] Mutants to Antifungal Compounds 165H. SIT] is Not Required for Virulence in Serotype A 166I. Models of Possible Relationships Between SIT] and Cellular Processes in C.neoformans 166J. Future Directions 170REFERENCES 173APPENDIX I: Websites Used In These Studies 193APPENDIX II: Supplementary Physical Mapping Data (CHAPTER TWO) 194A. Probes sequences for hybridization experiments 194B. Hybridization strategy for pooled probes 199viC. Directions to interpret ResGen High density filters 200D. Raw hybridization experiment results 201i. JEC21: integrated markers 201ii. JEC21: pooled probes 203iii. H99: integrated markers 212iv. H99: pooled probes 214APPENDIX III: Supplementary SAGE Data (CHAPTER THREE) 225A. Construction of the SAGE library from H99 cells grown in low iron mediaat 37°C 225i. Figure lIla. agarose gel of RNA extraction 225ii. Figure IlIb. polyacrylamide gel of PCR Optimization 225iii. Figure IlIc. polyacrylamide gel of PCR for 102 bp band 226iv. Figure hId. polyacrylamide gel of digest of 26 bp band 226v. Figure hlIe. polyacrylamide gel of ligation to form concatemers 227vi. Figure IlIf. agarose gel of colony PCR results 227vii. Tag Identification for SAGE Analysis 228B. Raw SAGE data for H99 low iron media library 228APPENDIX IV: Supplementary Data Analysis of SIT] (CHAPTER FOUR) 229A. SIT] Sequence information 229a. SIT] nucleotide and amino acid sequence from serotype D (SGTC)and A (DUMRU) and B (MSGSC) 229b. Sequences of SIT] homologs in other fungi 232c. Sequence of knock out cassettes with primers 235d. Sequence of the reconstituted fragment with primers 237B. SIT] Mutant Construction 239a. Agarose gel showing PCR products of individual overlap knockout fragments 239b. Agarose gel showing amplification of knock out cassette 239C. Confirmation of homologous recombination of iXsit] allele 241a. Agarose gel showing positive colony PCR results 241D. Reconstitution of sit] with SIT] 243a. PCR: primers SIT1A and SIT1F to differentiate wild-type andmutant bands 243E. Raw Data and Calculations 244a. Raw growth curve data 244b. Sample calculation for cell count assays 247APPENDIX V: Construction of Yeast Two Hybrid Libraries 248APPENDIX VI: Author’s Contributions to Research and Development 249viiLIST OF TABLESTable 1: Summary of the hybridization of selected markers to BAC clones ofJEC21andH99 33Table 2: Summary of hybridization experiments for individual probes 35Table 3: Summary of the contigs in the fingerprint map of JEC2 1 39Table 4: Summary of the contigs in the fingerprint map of H99 40Table 5: Most abundant SAGE tags for H99 cells grown at 37°C in low iron 67Table 6: Comparison of SAGE data for strain H99 cells grown at (3 7°C low iron) vs.YNB (25°C) media 71Table 7: Differentially abundant SAGE tags from H99 cells grown at (3 7°C low iron)vs. YNB (25°C) media 73Table 8: Differentially abundant SAGE tags from H99 cells grown in YNB (25°C)vs. (3 7°C low iron) 75Table 9: Comparison of SAGE data for strain H99 cells grown at (3 7°C low iron) vs.YNB (37°C) media 78Table 10: Differentially abundant SAGE tags from H99 cells grown in at(37°C low iron) vs. YNB (37°C) media 80Table 11: Differentially abundant SAGE tags from H99 cells grown in YNB (37°C)vs. (3 7°C low iron) 82Table 12: Comparison of SAGE data for strain H99 cells grown at (3 7°C low iron)vs. cells isolated from rabbit cerebral spinal fluid (in vivo) 85Table 13: Differentially abundant SAGE tags from H99 cells grown at (3 7°C low iron)vs. cells isolated from rabbit cerebral spinal fluid (in vivo) 87Table 14: Strains used and constructed in these studies 100Table 15: Siderophore-related genes in C. neoformans 112Table 16: Summary of melanin production for sit] mutants in serotype D strains 142Table 17: Summary of cell wall defects for sit] mutants noted by TEM 146viiiTable 18: Minimum inhibitory concentrations of antifungal agents for sit] mutants.... 150Table 19: Summary of DBA1 mouse weight loss with wt and the sit] mutant of thestrain B3501A (serotype D) 150Table 20: Summary of AJJcr mouse survival with SIT] wild-type , sit] mutants andsit] + SIT] reconstituted strains of the strain H99 (serotype A) 5 x 1 0 cellinnoculum 151Table 21: Summary of AJJcr mouse survival with SIT] wild-type , sit] mutants andsit] + SIT] reconstituted strains of the strain 1199 (serotype A) 5 x 1 cell innoculum .151Table 22: Summary of pleiotrophic phenotypes for sit] mutants of JEC21, B3501Aand H99 backgrounds 154ixLIST OF FIGURESFigure 1: Example of an autoradiograph showing the results of a row pooi of 12overgo probes hybridized to a high-density BAC clone filter 34Figure 2: FPC display from the map of strain JEC21 41Figure 3: Conservation of gene synteny between the genomes of JEC21 and H99 45Figure 4: Determination of Hindlil fragment sizes for 26S rDNA 47Figure 5: Relationship between electrophoretically separated chromosomes andthe contigs of the JEC21 and H99 maps 52Figure 6: Genomic location of 26S rDNA and HIS3 53Figure 7: Elaboration of the polysaccharide capsule in low iron or iron repleteMedium 63Figure 8: Expression profiling comparing relative transcript levels of SAGE tags fromstrain H99 cells grown in YNB (25°C) and at (3 7°C low iron) media 72Figure 9: Expression profiling comparing relative transcript levels of SAGE tags fromstrain H99 cells grown in YNB (3 7°C) and at (3 7°C low iron) media 79Figure 10: Expression profiling comparing relative transcript levels of SAGE tags fromstrain H99 cells isolated from rabbit cerebral spinal fluid (in vivo) or grown at (37°C lowiron) media 86Figure 11: Comparison of SAGE tag occurrences for cells grown in low iron (37°C),YNB (25°C) or YNB (3 7°C) medium 90Figure 12: Comparison of SAGE tag occurrences for cells grown in YNB medium(25°C), isolated from rabbit cerebral spinal fluid (in vivo), low iron (3 7°C) medium orYNB (37°C) medium 92Figure 13: Amino acid alignment of the Sitip protein from serotypes A, B and Dstrains of C. neoformans 107Figure 14: Similarity tree of the Sitlp protein from serotypes A, B and D strains of C.neoformans 108Figure 15: Comparison of exon and intron nucleotide sequence similarity for theSIT] gene in serotypes A, B and D strains of C. neoformans 109xFigure 16: Amino acid alignment of the Sitlp protein from the serotype D strainB3501A of C. neoformans with fungal homologs 114Figure 17: Similarity tree of the Sitip protein from the serotype D strain B3501A ofC. neoformans with fungal homologs 117Figure 18: Transcript levels for the Sitlp gene in culture media with low or high ironlevels for strains JEC2 1, B3 501 A and H99 118Figure 19: Structure of the SIT] gene and Sitip protein 120Figure 20: Structure of the Asit] knock out constructs 122Figure 21: Confirmation of the iXsitl knock out in the serotype D strains B3501A andJEC21 123Figure 22: Confirmation of the isit] knock out in the serotype A strain H99 124Figure 23: Disruption of the SIT] gene significantly affects growth in LIM andLIM+BPDA+Deferoxamine in B3501A and H99 but not JEC21 126Figure 24: sit] mutants are unable to use a siderophore to acquire iron in all strainsas shown by siderophore utilization assays 134Figure 25: The cAMP pathway is involved in siderophore utilization in serotype Abut not serotype D strains 136Figure 26: Melanin production on DOPA medium 138Figure 27: DIC microscopy of melanized cells at four days growth on DOPAMedium 139Figure 28: Temperature sensitivity and cell wall integrity assays 144Figure 29: Transmission electron microscopy (TEM) of melanized cells forwild type and sit] mutant strains 147Figure 30: Virulence assays in the murine model for serotype A wild type and sit]mutant strains 152Figure 31: Model of possible relationships between SIT] and cellular processes inC. neoformans 169xiLIST OF ABBREVIATIONSABC ATP Binding CassetteAIDS Acquired ImmunoDeficiency SyndromeATP adenosine triphosphateBAC Bacterial Artificial ChromosomeBPDA Bathophenanthroline disulfonic acidBLAST Basic Local Alignment Search ToolcAMP cyclic adenosine mono-phosphateCFU colony forming unitsCHR chromosomecontig contiguous piece of DNACR congo redCSF cerebral spinal fluidCW calcofluor whiteDIC differential interference contrastDUMRU Duke University Mycology Research UnitEtBr ethidium bromideJR iron replete mediumGa1XM galactoxylomannanGFP green fluorescent proteinGTP guanine triphosphateGXM glucuronoxylomannanLGA low glucose asparagine mediumLiCl lithium chlorideLIM low iron mediumMAPK mitogen-activated protein kinaseMIC minimum inhibitory concentrationMP mannoproteinMSGSC Michael Smith Genome Sciences CenterNAT nourseothricinxiiNCCLS National Committee for Clinical Laboratory StandardsNEO neomycinNCBI National Center for Biotechnology InformationORF open reading framePBS phosphate buffered salinePCR polymerase chain reactionPKC protein kinase Cpoly-A poly-adenylatedSAGE Serial Analysis of Gene ExpressionSDS sodium dodecyl sulphateSSC saline sodium citrateSGTC Stanford Genome Technology CenterTEM transmission electron microscopyTIGR The Institute for Genomic ResearchYNB yeast nitrogen brothYPD yeast extract peptone dextrose mediumxliiACKNOWLEDGEMENTSI would first like to thank Dr. James W. Kronstad for his exceptional leadership,resources, advice, expertise and patience. His laboratory and mentorship provided aprogressive, supportive and positive environment that allowed me to develop skills bothpersonal and professional that will continue to provide me with strong guidance andknowledge in my future career.To my committee: Dr. B.B. Finlay, Dr. P. Hieter and Dr. S.J.M Jones, I thank you foryour input, encouragement and dedication to my studies. I also would like toacknowledge the Michael Smith Genome Sciences Center, a world-class institution thatwas an integral partner in much of my thesis work. Specifically, I would like to thank themapping and SAGE groups including J. Schein, S.J.M. Jones, M. Marra and S.Zuyderduyn for a successful collaboration and constructive experience. I thank Dr. J.Heitman from Duke University for many strains and vectors as well as expert advice onC. neoformans. Thank you to the UBC Bio-Imaging Facility: E. Humphrey, K. Rensing,G. Martens, D. Home for technical assistance and expertise and to L. Samuels for helpfuldiscussion on TEM. I would like to note institutions whose Cryptococcus databaseswithout which I could not have completed these studies: these include DUMRU, SGTC,the Broad Institute of Harvard and MIT, the University of Oklahoma and TIGR.To my family: Mom and Dad (Joy and Kenneth Tangen), Brother (Darren) and hisfamily: Nicole and Gabrielle Riley, Grandma Mary Tangen, Sage, all the other Tangensand Sinclairs. I appreciate every day how blessed I am to have such a wonderful familywith an incredible belief in me. Mom and Dad, your love and support have made my lifebeautiful and have taught me to follow my dreams and that with hard work, passion andcommitment anything is possible. To my friend in celebration of life Matt Galloway,thank you for being by my side and sharing in dreams and ideas for the present andfuture. To my friend and classmate Dr. M.D. Brazas your friendship was exemplary insupport and offering a sharing of experience both professional and in life.Thank you to members of the Kronstad laboratory past and present for an encouragingand collaborative environment. To Dr. B. R. Steen who provided friendship andmentorship, J. Klose (technical expertise and friendship), Dr. S, Kidd (gene genealogy,PCR fingerprinting, AFLP, RFLP), A.P. Sham (mouse work), Dr. T. Lian (Northerns andtechnical expertise), Dr. G. Hu (technical expertise), Dr. W. Jung (information on ironuptake and regulation genes) and Dr. C. D’Souza (technical expertise and comments).I would also like to thank many friends professional and personal who have enhanced mylife in numerous ways: Dr. P.R. Hardwidge, Dr. H. Ledebur, Dr. D.M. Martin, M and M.Shepard, Dr. C. Maxwell, M. Smith, LL.B., E. Thomsen, J. Hagel, S. Duck, A. Harbott, J.Keefer, D. and J. Madsen, C and B. Gannon, J. Prenter and The Vancouver Spurs.This work was funded through grants to J.W.K. from The Natural Sciences andEngineering Research Council of Canada (NSERC): mapping and the Canadian Institutesof Health Research (CIHR): SAGE and SIT].xivCHAPTER ONE: INTRODUCTIONA. Cryptococcus neoformans and Cryptococcosis.a. C. neoformans varieties, serotypes, molecular types and phylogeny.Cryptococcus neoformans is a saprophytic basidiomycete fungus that has adapted tosurvival in the mammalian host. Strains of the fungus are separated into five serotypes (A, B, C,D and AD) based on antigenic differences in their polysaecharide capsule and three varieties arealso recognized: neoformans (serotype D), grubii (serotype A) and gattii (serotypes B and C).Recently, the variety gattli has been designated as the separate species C. gattii. Cryptococcushas been further classified into molecular types by PCR-fingerprinting, URA5-RFLP (restrictionfragment length polymorphism) analysis and amplified fragment length polymorphism (AFLP)analysis (Meyer et a?., 1999; 2003; Kidd et a?., 2004). PCR-fingerprinting involves using asingle minisatellite specific primer specific to M13 phage core sequence, producing anelectrophoretic profile that differentiates strains within varieties and serotypes (Meyer et al.,1999; 2003). URA5-RFLP analysis separated molecular types based on polymorphisms in thegene for orotidine monophosphate pyrophosphorylase (URA5) resulting from a HhaI andSau961 double digest (Meyer et al., 2003). AFLP analysis is a multilocus genotyping methodand has been used to type Cryptococcus (Boekhout et al., 2001). These molecular typescorrespond to the same groupings found in PCR-fingerprinting and the URA5-RFLP method, andthe authors that performed these analyses were the first to recommend the two species divisionsof C. neoformans (serotype A and D) and C. basillisporus (serotypes B and C) rather than thenow accepted C. neoformans (serotype A and D) and C. gattii (serotypes B and C). Each ofthese methods group strains into eight succinct molecular types (VN or VG designations forPCR-fingerprinting and the URA5-RFLP method; and corresponding AFLP types). The accepteddesignations that have appeared are: C. neoformans serotype A var. grubii (VNI/AFLP1 andVNII/AFLP1A), serotype AD (VNIII/AFLP3), serotype D var. neoformans (VNIV/AFLP2),serotype B C. gattii (VGI/AFLP4), (VGII/AFLP6), (VGIII/AFLP5) and serotype C(VGIV!AFLP7). Molecular phylogenetic work revealed that the grubii and neoformans varietiesare separated by ‘—4 8.5 million years of evolution, and these diverged from the gattii variety —‘-37million years ago (Xu et a?., 2000). This divergence was based on the gene genealogies of fourgenes: the mitochondrial large ribosomal subunit RNA, the internal transcribed spacer region of1the nuclear rRNA, including ITS 1, 5.8S rRNA subunit and ITS2, orotidine monophosphatepyrophosphorylase (URA5) and diphenol oxidase (LA Cl) (Xu et al., 2000).b. C. neoformans infections.C. neoformans primarily affects people with a compromised immune system, althoughmultiple isolated infections have been documented in normal individuals (Christianson et al.,2003; D’Souza et al., 2004). C. neoformans is especially detrimental to those people with AIDS,or Diabetes mellitus, and in patients undergoing long-term corticosteroid therapy, organtransplantation or cancer treatment. Since the onset of the AIDS era, there has been a markedincrease in the prevalence of cryptococcal infections. Previously, the fungus was rarelyencountered as a human pathogen. Presently, approximately 10-3 0 % of AIDS patients contractC. neoformans and it is the leading mycological cause of death in these people (Forche et al.,2000). The average survival rate for an AIDS patient diagnosed with cryptococcosis is less than6 months (Casadevall and Perfect, 1998). C. gattii has been associated with a wider host rangeincluding immunocompetent humans, as well as wild and domestic animals (Krockenberger etal., 2002; Kidd et a!., 2004)c. Environmental sources of Cryptococcus.Cryptococcus neoformans is found in a wide range of environmental locationsincluding water, soil, air, trees and pigeon feces (Casadevall and Perfect, 1998). The prevalenceof the fungus in urban areas, specifically in pigeon guano, increases the risk of transmission tohumans. C. neoformans was initially hypothesized to exist in nature as a wood rotting fungus,however, extensive genomic and biological investigation has shown no evidence for wooddegradation enzymes such as cellulases or lignin degradation enzymes (Jacobson et a!., 2003).C. gattii has been more extensively studied with respect to environmental sources. To that effect,C. gattii was historically isolated from eucalypts (Sorrell et al., 1996; Krockenberger et a!.,2002), however now many tree species have tested positive for C. gattii even in temperateclimates (Davel et al., 2003; Kidd et a!., 2004; Granados et a!., 2005). Environmental isolates ofC. neoformans are largely found in soil and pigeon excreta (Baro et a!., 1999; Trilles et al., 2003:Granados et aL, 2005), but have also been associated with trees (Trilles et a!., 2003; Kidd et a!.,2004: Granados et a!., 2005).2d. C. neoformans life cycle.C. neoformans exists in the anamorphic yeast-like state as a or , mating type haploidstrains. The fungus can undergo sexual reproduction: conjugation occurs following signaling viaG-protein-coupled receptors and pheromones. A dikaryon with clamp connections forms as aresult of mating partner fusion. Fusion is followed by karyogamy and the development ofbasidia. Following meiosis, the fungus produces haploid basidiospores, which then germinate toform the yeast-like state. The fungus is most commonly found in the yeast form and this is theinfectious state of the organism. C. neoformans can also undergo sporulation in the anamorphicstate when nitrogen is limited through a process called haploid fruiting. The sexual stage has notbeen conclusively identified in nature and the bias for the a mating type in the environmentsuggests that mating is not a common event (Casadevall and Perfect, 1998). The vast majority ofclinical isolates are also of the a mating type, and it has been suggested that the a mating type isa virulence-associated factor.e. Transmission and etiology of cryptococcosis in humans.C. neoformans is commonly contracted by inhalation of the desiccated yeast cells,poorly encapsulated yeast forms, or basidiospores (Casadevall and Perfect, 1998). In individualswith a normal immune system, C. neoformans usually only produces an asymptomatic• respiratory infection. In immunocompromised individuals the infection may disseminate to causecryptococcosis. Once disseminated, the fungus may cross the blood brain barrier where it causesencephalitis in the form of gelatinous cysts with a high fungal load. Disseminated C. neoformansmay also cause lesions in the lung, heart, skin, and bone (Casadevall and Perfect, 1998).f. Epidemiology of Cryptococcus.C. neoformans serotypes A, D and AD are pandemic, the majority of clinical isolates inNorth America are serotype A strains; serotype D strains are more prevalent in specific Europeancountries, for example, Denmark, France and Italy (Bennett et al., 1984; Dromer et al., 1996;Franzot et al., 1998). Serotypes B and C were previously considered to be mainly restricted totropical and subtropical regions, although isolates have been obtained from temperate regions(Mitchell and Perfect, 1995; Sorrell, 2001). More recently, serotype B C. gattii has emerged as3the cause of cryptococcosis in humans and animals on the East Coast of Vancouver Island. Thisoutbreak (the fungus is now considered endemic to the region) is the largest of its kind in history.The emergence of C. gattii on Vancouver Island is unique for two reasons: 1) this variety hadpreviously mainly been identified in tropical regions, whereas Vancouver Island is a temperateregion; 2) the fungus was responsible for infections in a wide range of hosts including marine,wild and companion animals, in addition to immunocompetent humans. In the past C.neoformans was largely considered to be an opportunistic pathogen. Since the onset andidentification of this outbreak in 1999, significant strides have been made in the molecular,genetic, evolutionary and biological characterization of these strains including the noveldiscovery that these strains are a distinct molecular type to those found in tropical regions (Kiddet al., 2004).g. The immune response to C. neoformans.The immune response to C. neoformans involves both physical and nonspecificimmunity, as well as a specific acquired response. The initial physical factors encountered bythe fungus include: high temperature, slightly alkaline pH and nutrient depletion. The cells alsoundergo an attack by host effector cells of the innate immune system (Casadevall and Perfect,1998). If the fungus can overcome the physical barriers and initial immune response to surviveand persist in the human host, a specific immune response ensues. This response includes bothhumoral and cellular immunity. In humoral immunity, anticryptococcal antibodies are formed inresponse to fungal antigens. In the cellular response, T-lymphocytes are generated thatrecognize cryptococcal antigens (Casadevall and Perfect, 1998). Specific cellular immuneresponses are dependent on nonspecific immune mechanisms and opsonization by professionalphagocytes and complement. Previously, cell-mediated immune mechanisms were regarded asthe primary immune response to C. neoformans. Now it is accepted that a combination ofantibody- and cell-mediated responses are required for the control of C. neoformans (Perfect,2005; Casadevall and Pirofski, 2005). A strong cellular response results from cryptococcalinfection through Th- 1 and requires the cytokines tumor necrosis factor, interferon gamma andIL-2 in addition to the chemokines MCP- 1 and MIP- 1 a for recruitment of inflammatory cells(Aguirre et al., 1995; Kawakami et al., 1996; Huffnagle, 1997; Huffnagle, 2000). Effector cellsinvolved in the control of C. neoformans include CD4+ and CD8+ lymphocytes in addition to4activated macrophages (Hill, 1992; Levitz, 1994). The capsular polysaccharides andcryptococcal proteins can elicit an antibody response. The majority of research has focused onthe response to capsular polysaccharides (Casadevall and Perfect, 1998; Vecchiarelli, 2005). Themajor component of the polysaccharide capsule is glucuronoxylomannan (GXM). Thispolysaccharide is highly immunogenic when opsonized and elicits innate and adaptive immuneresponses (Vecchiarelli, 2005). Minimal macrophage binding occurs in the absence of opsonins,however, C. neoformans is a potent activator of complement, which results in the deposition ofthe final component C3, an opsonin, on the fungal cell. Macrophages recognize opsonized C.neoformans by three major complement receptors designated CD35 (CR1), CD11b/CD18 (CR3),and CD11c/CD18 (CR4) (Levitz, 2002). In the cellular response, GXM binding involves thefollowing receptors: toll-like receptor (TLR)4, CD14, TLR2, CD18 and Fc gamma receptor II(FcgammaRPi). This combination of receptor binding can actually lead to a suppressive effect.Interestingly, this can be reversed by the presence of GXM protective antibodies (Vecchiarelli,2005). Shed polysaccharide and mannoproteins represent a problem for the human immunesystem in that they use up many of the circulating opsonins in the serum and cerebral spinalfluid, thereby reducing the ability of the immune system to identify the whole pathogen vs. shedpolysaccharide.h. Treatment of cryptococcosis.The present therapy for cryptococcosis includes treatment with the polyene drugamphotericin B and azole drugs such as fluconazole and 5-flucytocine. Amphotericin B is oftenused in cases of rapidly progressing, life-threatening illness and fluconazole is used in chronicsuppression or prophylaxis in immunosuppressed individuals. Unfortunately, there is a highoccurrence of relapse even after apparent successful therapy (Casadevall and Perfect, 1998) aswell as a significant problem with increasing drug resistance worldwide (Pfaller et aL, 2005). Inaddition, antifungals such as amphotericin B can be highly toxic to the host. There is a greatneed for new anticryptoccocal compounds with higher efficacy and lower toxicity.Understanding pathogenesis in C. neoformans may allow researchers to identify new targets fordevelopment of more effective drug therapies for treating and controlling cryptococcosis.5B. Genomic Analyses of C. neoformans.There are notable differences between serotypes A (grubii) and D (neoformans) invirulence, epidemiology and clinical prevalence, with serotype A generally being the morevirulent serotype and the most common among clinical isolates. Due to these differences, manystudies have been initiated to compare the serotypes on a genomic level.a. Karyotype studies.Several studies have attempted to characterize the differences between serotype A and Dstrains. Thus, isolates have been characterized with respect to 26S rRNA sequences, PCRfingerprints, enzyme electrophoretic profiles and electrophoretic karyotypes (Perfect et al., 1989,1993; Guehó et a!., 1993; Meyer et a!., 1993; Brandt eta!., 1993; Wickes, eta!., 1994; Boekhoutand van Belkum, 1997; Bertout et a!., 1999). Strains of serotype A and D were distinguished bythe RFLP patterns obtained upon hybridization with a repeated element (CNRE-1) and by thenucleotide sequence analysis of specific genes (e.g. URA5) (Franzot et al., 1998; Xu et a!.,2000). Xu et al. (2000) performed a more detailed phylogenetic analysis of strains from thedifferent serotypes using sequence analysis of the mitochondrial large ribosomal subunit, theinternal transcribed spacer region of nuclear rRNA, and the genes encoding orotidinemonophosphate pyrophosphorylase (URA5) and diphenol oxidase (CnLACJ). This worksupports the current separation of strains into the three varieties and two species and provides aphylogenetic framework for understanding the evolution and geographic distribution of C.neoformans. The population structure of C. neoformans has also been examined in detail usingAFLP genotyping (Boekhout et at., 2001) and PCR fingerprinting (Meyer et at., 1999; Ellis etat., 2000). Finally, although strains of serotype A and D have been shown to be geneticallydistinct, it is possible to obtain mating between isolates from the two different serotypes (KwonChung, 1975).Several groups have performed electrophoretic karyotyping to determine the size andnumber of chromosomes in strains representing the different varieties of C. neoformans (Perfectet a!., 1989; Polachek and Lebens, 1989; Perfect et a!., 1993; Wickes et a?., 1994; Boekhout andvan Belkum, 1997; Boekhout eta!., 1997; Forche et at., 2000). In addition, Wickes eta?. (1994)and Spitzer and Spitzer (1997) assigned several markers to electrophoretically separated6chromosomes by hybridization with known genes or ESTs. At the time of these studies, the viewof the karyotype in C. neoformans indicated a genome size in the range of 15 to 27 Mb with anaverage chromosome number of 12 for variety neoformans and 13 for variety gattii. Forche etat. (2000) described the construction of a meiotic linkage map for C. neoformans serotype D. Amapping population of 100 progeny was employed with a total of 181 AFLP, RAPD and genemarkers to identify 14 major linkage groups. Six of the linkage groups were assigned to specificchromosomes. A more recent genetic linkage map has been reported for serotype D (Marra eta?., 2004). These studies further refined estimates of genome size to 20.2 Mb, with 14chromosomes ranging in size from 0.8-2.3 Mb.b. Physical mapping projects.The Kronstad laboratory in collaboration with the Michael Smith Genome SciencesCenter (MSGSC) initiated a physical characterization of the C. neoformans genome as part of aninternational effort (Heitman et a?., 1999) to obtain the complete genomic sequence of twostrains representing the A and D serotypes. Physical maps for one strain of each serotype werereported in 2002 (Schein et al., 2002), with 82 markers placed on the serotype D (strain JEC21)genome and 102 markers placed on the serotype A (strain H99) genome. A portion of the mapconstruction is detailed in Chapter Two of this thesis and has been published (Schein et a?.,2002). Conservation of gene synteny was generally good between the two genomes but therewere notable examples of chromosomal rearrangements such as inversions and translocations.The BAC fingerprinting technology first described by Marra et a?., (1997) was used to generatelarge contigs that formed the framework for assembly and finishing of the genomic sequences forthe serotype A and D strains (Loftus et a?., 2005). The ends of the fingerprinted BAC cloneswere also sequenced and the traces contributed to the shotgun sequence databases for bothstrains. The mapping approach was employed previously for whole-genome, random BAC clonefingerprinting projects that supported sequencing of the Arabidopsis thaliana (Marra et al., 1999;Mozo et a?., 1999) and human (McPherson et a?., 2001) genomes. Finally, the contributionsdescribed in Chapter Two placed markers on the BAC maps and these markers were employedboth to compare the conservation of synteny between the serotype A and D strains and to attemptto correlate BAC clone contigs with specific chromosomes. Five additional mapping projects fordifferent Cryptococcus genomes have been initiated and completed since the successful initial7maps. This resource allows the comparison of multiple strains and provides the framework foradditional sequencing efforts and genomic studies (detailed below).c. EST sequencing projects.Sequencing of expressed sequence tags (ESTs) assists in the annotation of genomes andin gene expression studies. ESTs are a single pass sequence read from cDNAs, thus representinggene transcripts from an organism. They are particularly useful because they represent the codingregions of the genes and provide key information to assist in the annotation of features such asintron and exon locations. They also provide transcript identification for expression-based dataderived from techniques such as serial analysis of gene expression (SAGE). EST sequencingprojects were completed for the serotype D strain B3501A and the serotype A strain 1199 at theUniversity of Oklahoma ( For B3501A, 4000 cloneswere sequenced from both ends (directionally cloned inserts) resulting in 8000 ESTs with anadditional 1700 clones sequenced directionally (‘-3300 EST5) from cells grown in low ironmedium. For the serotype A strain H99, 3750 ESTs were obtained from both ends of clonesresulting in 7500 reads. Additional ESTs were sequenced for H99 at the Duke UniversityMycology Research Unit (DUMRU) (!). These databases wereutilized in the total genome assembly projects, in addition to providing a crucial resource fortranscript identification in SAGE projects (Steen et at., 2002; 2003, Lian et at., 2005). Finally,the largest effort to collect EST sequences was performed more recently at The Institute forGenomic Research (TIGR) for the serotype D strain JEC2 1 (Loftus et at., 2005).d. Genome Sequencing Efforts.A number of genome sequencing efforts are now complete or near completion includingthe publication of two serotype D genomes JEC21 and B3501A (Loftus et at., 2005). JEC21 wassequenced, assembled and annotated at The Institute for Genomic Research (TIGR)( and B3501A was completed at the Stanford GenomeTechnology Center (SGTC) (http ://sequence-www. stanford.edulgroup/C. neoformansl). Inaddition to variation in virulence and genome structure between serotypes, major differences inthe genome structure of the closely related strains JEC2 1 and B3 501 A were reported, and this isinteresting because B3501A is a more virulent strain than JEC21. Specifically, chromosomal8rearrangements between these two strains were documented with a chromosomal translocationand segmental duplication identified in the strain JEC2 1 (Fraser et al., 2005). Other genomesequencing projects near completion are for the serotype A strain, 1199 at the Broad Institute ofHarvard and MIT (http ://, anenvironmental isolate, W1v1276 (serotype B strain), at the University of British Columbia(!about/news/crypto_public/view) and a Vancouver Island clinical isolate,R265 (serotype B strain), at the Broad Institute of Harvard and MIT(http :// Informationfrom mapping and sequencing projects has identified the number of chromosomes in C.neoformans for strains JEC21, B3501, H99 and WM276 to be 14. The availability andcomparative analysis of these genomes, in addition to many transcriptional studies in progress,will be powerful resources to provide further insight into key virulence factors in the fungus.e. Serial analysis of gene expression (SAGE).Serial analysis of gene expression (SAGE) is an innovative molecular-biology basedtechnique to analyze expressed genes in an organism. The SAGE technique, initially developedby Velculescu et a!. in 1995, provides a snapshot of the mRNA transcript abundance in a givencell state. The set of mRNA sequence “tags” in a SAGE library reflect the content and relativeabundance of mRNA in the cells isolated for that library. This profile of transcripts is knowncollectively as the transcriptome. A transcriptome is dynamic in contrast to the genome, whichis static. The technique involves isolation of mRNA, capture of the poly-A 3’ ends andconversion to cDNA. The DNA is cut with a restriction enzyme that has a 4 bp recognition site.This is the “anchoring” enzyme NlaIII (5-CATG-3’). A linker is added that contains therecognition site for a type ITS enzyme BsmJI (5’-GGGAC-3’), which cleaves the cDNA 15 bpdownstream of the recognition site. The fragments are blunt end ligated, linkers are removedand tags are concatenated, and then cloned into vectors for sequencing. Data output from aSAGE library is in the form of 10 base pair tags, each of which is 3’ adjacent to the 3’ mostNlaIII (CATG) site in a eDNA. The abundance of unique SAGE tags may be compared fromcells grown in different conditions. The corresponding transcript for each unique tag gives anindication of expression for that transcript in that cell state. The technique allows for tags frommultiple transcripts to be sequenced in a single read. This in turn increases the efficiency and9lowers the cost of transcriptome analysis. The tags are extracted and quantified throughautomated programs, they can then be annotated using EST and genomic sequence databases.Some of the advantages of SAGE include the real quantification of transcripts and the unbiasednature of the technique with respect to the identification of novel or unknown genes. Manyvariations of the technique have been developed since the original SAGE protocol such asMicroSAGE (Datson et al., 1999), a modified procedure to allow for smaller sample sizes, andLongSAGE (Saha et al., 2002) that produces longer tags (20 bp) using the TypellS enzymeMmeI (which has a recognition site further from the cut site than the original enzyme Bsmfl).This technique was used to confirm annotations from the original SAGE technique. Recently, aneven more robust SAGE technique has been developed. SuperSAGE (Matsumura, 2005) utilizesthe Typelli enzyme EcoP 151, which produces 26 bp tags and allows for more definitivetranscript annotation.SAGE techniques have been successfully utilized in many transcriptional studies ofeukaryotic organisms thus far, including C. neoformans (Steen et al., 2002 and 2003; Lian et al.,2005). SAGE was initially developed to differentiate gene expression in cancer vs. normal cellsof humans (reviews by Porter and Polyak, 2003; Liu et al., 2004; Tuteja and Tuteja, 2004).SAGE has also been adopted for use in a wide variety of applications in other organisms. SAGEhas been extensively used in plant studies, for example, to investigate root nodule formation inthe legume Lotus japonicus (Asarnizu et al., 2005), virally infected Cassava plants (Manihotesceulenta) (Fregen et al., 2004) and tomato transcriptional regulation factors in Arabidopsisthaliana (Chakravarthy et al., 2003). SAGE has also been used to investigate gene expression inparasitic pathogens (Knox et al., 2005). After its development, SAGE was very quickly appliedto the model yeast S. cerevisiae and the SAGE transcriptome was reported in 1997 (Velculescuet al., 1997). This study was quickly followed by SAGE analysis that was adapted to askspecific questions for example, about the variation of transcriptome size (Holland, 2002),vinification (Varela et aL, 2005) or identification of transcriptional start sites (Zhang andDietrich, 2005). Many other fungi have been studied by SAGE for industrial processes e.g.Monascus aurantiacus to determine the efficient fermentation for citrinin production (Xiong etal., 2005) as well as other fungal pathogens e.g. Magnaporthe grisea for appressorium formation(Inc et a!., 2003) and barley mildew (Blumeria graminis) (Thomas et al., 2002). Multiple SAGEprojects and analyses are ongoing in the Kronstad laboratory and are detailed in Chapter Three.10There are advantages of SAGE over microarray technology. First, SAGE is unbiased inits ability to identify novel genes of unknown functions, this is not possible with microarrayswhere DNA sequence must be determined prior to array printing. Second, SAGE can also beused to effectively quantify fold differences in transcription by comparison of SAGE tagabundance between conditions. The abundance of SAGE “tags” corresponds directly with themRNA copy number in the cell. Until recently, a complete genomic sequence was not availablefor C. neoformans, therefore the construction of comprehensive microarrays was not possible.Since the onset of this project, total genome sequence has become available and has beenfollowed by the advent of microarray technology for C. neoformans (Fan et al., 2005; PukkilaWorley et al., 2005).C. C. neoformans Signaling and Virulence Factors.The remainder of this introductory chapter presents background information on thebiology and pathogenesis of C. neoformans. This information sets the stage to appreciate theapplication of genomic and genetic approaches to understand virulence in this importantpathogen.a. The cAMP and PKC1/MAP kinase pathway.Cell signaling plays an important role in the regulation of many cellular processes.Many virulence factors in C. neoformans are known to be regulated through the cAMP/PKA andFKC1/IVIAP kinase pathways (Alspaugh et al., 1998; Kronstad et al., 1998). Variousenvironmental signals such as nitrogen, glucose or iron deprivation activate the Ga proteincAMP-PKA pathway, which regulates differential transcription of genes involved with melaninformation, mating and capsule formation.Protein kinase A contains two regulatory and two catalytic subunits. The kinase isactivated when four molecules of cAMP bind to the two regulatory subunits (Pkrlp); this causesa conformational change in the regulatory subunits that consequently releases the catalyticsubunits (Pkalp). The catalytic subunits of protein kinase A are then active to phosphorylatedownstream targets that include factors related to expression of a number of virulence genes(D’Souza et al., 2001).11The Heitman laboratory at Duke University has performed extensive studiescharacterizing components of the cAMP pathway. Important components include products ofthe GPA1, CAC1, PKA1 and PKR] genes. GPA1 codes for the alpha subunit of a heterotrimericGTP-binding protein (Aispaugh et a!., 1997) and CAC] codes for adenylyl cyclase (Aispaugh eta!., 2002). The Heitman laboratory has characterized the PKA] gene in the serotype A strain H99that encodes the major cAMP-dependent protein kinase catalytic subunit in addition to the FKR]gene that encodes the protein kinase A (PKA) regulatory subunit (D’Souza et al., 2001). In theserotype D strain, the major catalytic subunit is PKA2 (Hicks et al., 2004). PKA2 is present inH99 and FKA 1 is present in JEC2 1, however it has been shown that the subunits have divergentroles between serotypes A and D (Hicks et a!., 2004). In the serotype A strain H99, pkal mutantswere sterile, failed to produce melanin or capsule and were avirulent. p/cr] mutants overproducedcapsule and were hypervirulent in animal models. The pkrlpkal double mutant had a similarphenotype to the pkal mutant indicating that the pkal mutation is epistatic to the pkrl mutation.Protein kinase A was also shown to function downstream of Gc protein and adenylyl cyclase andto regulate cAMP production by feedback inhibition (Alspaugh et a!., 2002). In serotype D,PKA2 not PKAJ is involved in the regulation of mating, haploid fruiting and virulence factorproduction, but is not a key regulator of virulence as in serotype A (Hicks et al., 2004).Recently a study of the transcriptional effects dependent on the G-alpha protein,encoded by GPA], was undertaken (Pukkila-Worley et a!., 2005). This study further identifiedand confirmed multiple genes that were controlled by the cAMP pathway, specifically focusingon those involved in capsule production and melanin formation. Two laccase genes (LA Cl andLAC2) encoding laccase enzymes were controlled by the cAMP pathway through GPA]. LAC]had been previously characterized in C. neoformans and was required for melanin formation(Williamson, 1997). This work showed that similar to LAC1, LAC2 was induced by low glucose,however deletion of the gene only resulted in a minor delay in melanization. Overexpression ofLAC2 was able to restore melanization to compensate for the loss of LAC] or GPAJ in additionto partially restoring virulence to lad and gpal mutants (Pukkila-Worley et al., 2005).The PKC] pathway is responsible for maintaining cell wall integrity in C. neoformans(Kraus et al., 2003; Gerik et a!., 2005). The terminal kinase in the pathway is Mpklp/Slt2p anddeletion mutants of mpk] are sensitive to temperature and cell wall damaging compounds and areattenuated for virulence in the murine model (Kraus et a!., 2003). More recent studies have12characterized other components of the PKC1 pathways including kinases and regulatory proteins(Gerik et al., 2005). These studies showed that the kinases Bcklp and Mkk2p are critical formaintaining cell wall integrity, in addition to the regulatory protein Lrglp and phosphatasePpglp.The PKC]/MAP kinase pathway has also been implicated in melanogenesis, where theloss of the Cl-domain (for diacyiglycerol) of PKC] led to loss of laccase activity and melaninproduction (Heung et aL, 2005). These changes were proposed to be mediated by a change incell wall integrity due to the displacement of laccase from the cell wall.There is also involvement of the C. neoformans PKC] pathway in cell wall integrity. It isknown that the cell wall integrity pathway is activated in C. albicans when Pkclp phosphorylatesthe MAP kinase Mkclp (Bates et a!., 2005). A homolog of Mkclp is present in the C.neoformans serotype D strain JEC21 (TIGR Identifier 162.m02645 ICNIOO41OI). The MAPkinase Mkc 1 p has already been implicated in regulation of cell wall integrity in response toantifungal drugs and loss of calcineurin function in C. neoformans (Kraus et a!., 2003). Asdescribed in Chapter Four, the work in this thesis documents a possible relationship between agene (SIT]) involved in iron uptake to melanogenesis and cell wall integrity, possibly throughPKCJ.b. Structure and biology of the polysaccharide capsule: contributions to virulence.The polysaccharide capsule of C. neoformans is generally regarded as the mostsignificant virulence factor in the fungus. The capsule is a distinguishing feature in comparisonto other medically important yeasts, and is often used diagnostically for the identification of C.neoformans using India ink stain. All acapsular strains produced thus far have reduced virulenceor are avirulent (Perfect, 2005). Extensive studies have been performed and summarized on thestructure, antigenicity, biology, synthesis and regulation of the capsule (Casadevall and Perfect,1998; McFadden and Casadevall, 2001; Bose, 2003; Janbon, 2004; Perfect, 2005). Three majorpolysaccharide components are described for C. neoformans exopolysaccharide:glucuronoxylomannan (GXM), galactoxylomannan (Ga1XM) and mannoprotein (MP). GXMaccounts for approximately 90% of the capsule, and it is differences in GXM structure leading toantigenic variation that led to the classical serotype designations A, B, C, D and AD (Casadevalland Perfect, 1998). The capsule contributes to virulence in a number of respects. Components13of the capsule do elicit a response from the immune system but can also be immunosuppressive.Mannoproteins are immunogenic and elicit a cell-mediated response in a delayed-typehypersensitivity reaction (Murphy et al., 1988). GXM and Ga1XM are poor antigens in theimmune response and immunization of mice with large amounts of these polysaccharidesresulted in a reduced ability to produce an antibody response (Kozel et at., 1977). In contrast,opsonized GXM is highly immunogenic and produces an antibody response characteristic of Tcell dependent antigens, with predominantly IgG isotypes (Casadevall et a!., 1992). Theimmunosuppressive effect of the capsule includes: antiphagocytosis (Kozel and Gotschlich,1982; Kozel et a!., 1988; Vecchiarelli et a?., 2003; Ellerbroek et al., 2004; Gates et al., 2004),complement depletion (Macher et al., 1978; Kozel, 1993), antibody unresponsiveness (Kozel etal., 1977), inhibition of leukocyte migration (Dong and Murphy, 1995; Lipovsky et a!., 1998;Murphy, 1999; Vecchiarelli, 2000; Ellerbroek et at., 2004), dysregulation of cytokine secretion(Retini et at., 1996; Vecchiarelli et at., 1995; 1996; 1998; Shoham et a?., 2001), brainenhancement of HIV infection (Pettoello-Mantovani et al., 1992; 1994), L-selectin and tumornecrosis factor receptor loss (Dong and Murphy, 1996; Murphy, 1999; Ellerbroek et a?., 2004),and negative charge (Nosanchuk and Casadevall, 1997).c. Melanin and its effects on virulence.Melanin production and the laccase enzyme represent another significant virulence factorin C. neoformans. Melanin and the production of laccase is necessary for survival within alveolarmacrophages (Liu et a?., 1999) and for extrapulmonary dissemination to the brain (Noverr et at.,2004). DOPA type melanin is synthesized from diphenolic catecholate compounds by the phenoloxidase enzyme, laccase (Lacip). Disruption of LAC1 leads to an “albino” phenotype onselective DOPA medium the mutant is and is attenuated for virulence in animal models (Salas etat., 1996). Melanin has multiple affects on cell stability and protection from the host includingits ability to act as an antioxidant, a contribution to cell wall integrity, reduction of drugsusceptibility by antifungal binding, interference with antibody-mediated phagocytosis andprotection from high temperatures (Casadevall and Perfect, 1998). The enzyme laccase has alsobeen implicated in protection from alveolar macrophages even in the absence of the DOPAsubstrate (Liu et at., 1999). This protection is proposed to be a result of the iron oxidase activity14of the enzyme, which allows for the oxidation of Fe2+ to Fe3+ thereby reducing the ability ofmacrophages to produce damaging free radicals.d. Mating type specificity: prevalence in clinical strains.The predominant mating type in environmental and clinical isolates is a, at an a:a ratioof 40:1 (Casadevall and Perfect, 1998). This led scientists to propose that asexual reproductionis the primary mode of reproduction supporting the idea that the a mating type is a virulence-associated factor (Kwon-Chung et a!., 1992). This hypothesis was reached because there wouldbe few mating type available to mate if the predominant mating type in the population was a.The mating type is even more infrequent among serotype A strains. The majority of clinicalisolates in North America are serotype A and a study of environmental isolates in New York Cityindicated a single pattern for CNRE- 1 repetitive DNA RFLP pattern which suggests a clonalorigin (Casadevall and Perfect, 1998; Currie et at., 1994). The haploid a mating type is alsoplausible as the infectious agent due to its ability to undergo haploid fruiting under conditions ofdryness and low ammonia concentration where many strains formed hyphae and basidiospores(Wickes et at., 1996). Haploid fruiting allows for the generation of particles that are smallenough to account for route of infection through the alveoli of the lung, without the need forsexual reproduction (Wickes et al., 1996). A very interesting recent finding was that mating canoccur between strains of the same mating type, a. (Fraser et al., 2005; Lin et a!., 2005). Thisunique biological phenomenon may further explain the prevalence of the alpha mating type inthe environment, in clinical prevalence and further would allow for genetic variation within thepopulation in the absence of the mating type.e. Superoxide dismutase (SOD) and effects.Superoxide dismutase (SOD) converts superoxide radicals into hydrogen peroxide and oxygen.This enzyme is widely believed to be key in the survival of bacterial and fungal pathogens in thehost. Cox et al. (2003) showed that a null mutant of the Cu,Zn SOD gene SOD] was moresusceptible to reactive oxygen species in vitro, had significantly reduced virulence in the murineinhalation model and had attenuated growth compared to wild-type in macrophages thatproduced reduced amounts of nitric oxide. These findings indicated that superoxide dismutasecontributes to virulence but is not absolutely required for pathogenesis in C. neoformans. The15authors hypothesize that the reduction in virulence may be due to an increased susceptibility tooxygen radicals within macrophages and may not be required if other antioxidant defensesystems in C. neoformans can compensate for the loss of the Cu,Zn SOD in vivo. Another gene,SOD2, encoding a mitochondrial superoxide dismutase is involved in resistance to oxidativestress, growth at high temperature and is required for virulence in the murine inhalation model(Giles et al., 2004). Similar work has been done in C. gattii, where the sod2 mutant was alsofound to be sensitive to growth at high temperature and avirulent in the intranasal model in mice(Narasipura et al., 2005). A SOD] homolog is also present in C. gattii and has been implicatedin virulence (Narasipura et a!., 2003). The sod] mutant was highly sensitive to the redox cyclingagent menadione, and showed fragmentation of the large vacuole in the cytoplasm, but no otherdefects were seen in growth, capsule synthesis, mating, sporulation, stationary phase survival orauxotrophies for sulphur-containing amino acids. The mutant was attenuated for virulence in themurine model, and it was significantly more susceptible to in vitro killing by human neutrophils(PIVINs) than wild type. The deletion of SOD] also resulted in defects in the expression of anumber of virulence factors, e.g. laccase, urease and phospholipase. Overall, these resultssuggested that the SOD] was required for virulence but not saprophytic survival. The virulencedefect of sod2 was not believed to be a result of increased killing by phagocytes. In addition, thesod2 mutant was also highly susceptible to redox-cycling agents, high salt and nutrientlimitations (Narisipura et a!., 2005). This group also constructed double mutants of sodlsod2.The sod]sod2 double mutants were also avirulent in mice (not surprisingly). The sodlsod2double mutants also showed a marked reduction in the activities of other known virulence factorsand they were more susceptible to killing by macrophages than was the sod2 single mutant. Thegroup had previously shown that sod] mutants of C. gattii were attenuated for virulence, likelydue to an increase in killing by phagocytes, and had a reduction of activities of other virulencefactors. Therefore this group hypothesizes that Sodlp and Sod2p play distinct roles in thebiology and virulence of C. gattii via independent modes of action.f. Resistance to oxidative stress.A number of genes have been identified that are implicated in the resistance to oxidativestress. Superoxide dismutases SOD] and SOD2 were discussed previously. Glutathioneperoxidases are responsible for the reduction of peroxides, which leads to resistance to oxidative16stress. Two glutathione peroxidases GPX1 and GFX2 have been characterized in C. neoformans(Missall et al., 2005a). Both genes were induced by t-butylhydroperoxide or cumenehydroperoxide stress and repressed during nitric oxide stress. Further, GPX2 was induced byhydrogen peroxide. The gpxl and gpx2 mutants were sensitive to peroxide stress but notsuperoxide or nitric oxide stressors. The mutants did show a slight sensitivity to killing bymacrophages but were dispensable for virulence in the murine model. The thioredoxin system isalso involved in resistance to oxidative stress in C. neoformans (Missall et al., 2005 b and c).This antioxidant system consists of thioredoxin, thioredoxin reductase and NADPH. Missall etal., showed that the thioredoxin proteins Trxlp and Trx2p were important for resistance to stressand for virulence. The trxl mutant was sensitive to multiple stresses and the trx2 mutant to nitricoxide stress. It was shown that TRXJ was necessary for virulence in mice and survival inmacrophages. They also showed that two putative transcription factors differentially regulate thethioredoxin system. Aftip is involved in oxidative stress induction and Yap4p for nitrosativestress induction of the thioredoxin genes. Two thiol peroxidases were identified andcharacterized in C. neoformans (Missall et al., 2004; 2005d). Tsalp and Tsa3p were identifiedin proteomic studies as more highly expressed at 37°C. Missall et al. showed that the TSA 1 andTSA3 were transcriptionally induced by hydrogen peroxide. The group also identified a thirdthiol peroxidase gene, TSA4 from genomic sequence. Through the construction of single, doubleand triple mutants they show that the tsal mutant is sensitive to hydrogen peroxide, tbutylhydroperoxide, nitric oxide, has reduced growth at 25 and 3 8.5°C and has reduced virulencein two mouse models. Finally, an alternative oxidase gene, AOX1 that showed increasedtranscription at 37°C was characterized with respective to virulence (Akhter et al., 2003).Alternative oxidases are involved with the cytochrome oxidative pathway to produce ATP andmay be involved with the response to oxidative stress in the mitochondria. The aoxl mutant wasmore sensitive to the oxidative stressor tert-butyl hydroperoxide, had reduced virulence in themurine inhalation model and had impaired growth in a macrophage-like cell line.g. Maunose production and virulence.Mannose production has been implicated in virulence in C. neoformans. The GDPmannose biosynthesis pathway is highly conserved in fungi and consists of three key enzymes:phosphomannose isomerase (PMI), phosphomannomutase (PMM) and GDP-mannose17pyrophosphorylase (GMP). Hills et al. (2001) produced a null mutant of the MAN] gene,encoding for the PMI enzyme, to block mannose synthesis. They showed that the gene wasrequired for virulence in both the rabbit and mouse model, resulting in the elimination of C.neoformans from the hosts. The mutants were also poorly encapsulated, had reducedpolysaccharide secretion and showed morphological abnormalities. These results indicated thatthe production of mannose is important for pathogenesis in C. neoformans.h. The ability to grow at 37°C.The ability to grow at 37°C is an inherent virulence factor of any human pathogen andthis trait can be attributed to many genes. A number of genes have been implicated in thesurvival of C. neoformans at 37°C thus far, including genes involved with stress resistancepathways, signaling networks and basic metabolism (Perfect, 2005). Some of these include:manganese superoxide dismutase, SOD2 (Giles et al., 2005), calcineurin, CNA] (Rhodes et al.,1986; Odom et al., 1997), a vacuolar ATPase, VPH] (Erickson et al., 2001) and a gene involvedin RNA/DNA splicing, CCNJ (Chung et al., 2003). The signaling genes RAS] (Alspaugh et al.,2000), CNA] (Odom et al., 1997), CNB] (Odom et al., 1997), MPK] (Kraus et al., 2003) andCTSJ (Fox et al., 2003) have been implicated in high temperature growth. Amino acidmetabolism genes ILV2 (Kingsbury et al., 2004a) and SPE3/LYS9 (Kingsbury et al., 2004b), andtrehalose synthesis genes TPS] (trehalose-6-phosphate synthase) and TPS2 (trehalose-6-phosphate phosphatase) (Wills et aL, 2003) are also required. Many other genes have beenidentified in transcriptional studies that are regulated by temperature but are not necessarilyrequired for high temperature survival (Steen et aL, 2002; Perfect et al., 2005). In S. cerevisiaeover 70 genes have been identified that are essential for growth at 3 7°C, indicating that there arelikely many more genes required for survival at 37°C in C. neoformans.Other virulence factors.A number of other virulence factors have been characterized in C. neoformans. Theseinclude the production of urease, phospholipase secretion, vacuolar acidification, and mannitolproduction, and an RNA helicase. Urease is a nickel metalloenzyme that catalyzes the hydrolysisof urea to ammonia and carbamate. C. neoformans is a prolific producer of urease and thischaracteristic is often used to diagnose the fungus (Casadevall and Perfect, 1998). Urease acts as18a scavenger enzyme and is important in conditions of low nitrogen. It is likely important duringinfection due to its ability to change the local pH in the host, for example when C. neoformans isin the phagolysosome of the macrophage (Casadevall and Perfect, 1998). Urease has beenimplicated in virulence during experimental cryptococcosis in the inhalation and intravenousmurine model but not the intercisternal rabbit model (Cox et al., 2000). Phospholipase secretionhas been linked to virulence because strains that produced higher levels of phospholipase causedmore significant infections than lower producers (Chen et a!., 1997). More recently, the genePLB1 encoding an enzyme that has phospholipase B (PLB), lysophospholipase hydrolase andlysophospholipase transacylase activities has been implicated in virulence (Cox et al., 2001). Thepib] mutants did not show defects for any other cryptococcal virulence factors in vitro (e.g.capsule formation, growth at 37°C, laccase activity or urease activity). However, they did showreduced virulence in the murine inhalation and rabbit meningitis model and exhibited a growthdefect in a macrophage—like cell line. Many virulence factors have been shown to be dependenton the vesicular proton pump (ATPase) VPH] (Erickson et a!., 2001). The vphl mutant wasdefective in capsule production, laccase and urease expression, growth at 37°C and was requiredfor virulence in the murine meningoencephalitis model. Mannitol has been implicated invirulence as a possible osmoticum or antioxidant during infection. Extracellular mannitol maycontribute to the elevation of intercranial pressure found in heavily infected patients because ahigh concentration of yeast has been implicated in this syndrome (Denning et al., 1991). Aserotype A, low-mannitol producing mutant, was isolated by UV mutagenesis from the serotypeA strain H99. The mutant was more susceptible to heat and NaCl and had reduced virulence inmice (Chaturvedi et a!., 1996). The DEAD-box RNA helicase-encoding gene (VAD]) has beenimplicated in virulence in C. neoformans (Panepinto et al., 2005), where the vad] mutant hadsignificantly reduced virulence in the murine model. The loss of VAD] resulted in theupregulation of NOT] which is a global repressor of transcription. Of particular interest, NOT]was found to be an intermediate repressor of laccase. The group further found that Vadip waslocated within macromolecular complexes that formed cytoplasmic granular bodies in maturecells and during infection of mouse brain. In addition, VAD] was shown by in situ hybridizationto be expressed in the brain of an AIDS patient coinfected with C. neoformans.19D. Iron and Siderophore Transport.a. Importance of iron transport and iron responsive genes.The low iron environment found in human serum, brain and cerebral spinal fluid mayprovide a nutritional cue for the induction of key virulence factors (e.g. capsule formation) in C.neoformans. Low iron conditions are not only important in relation to the elaboration of thepolysaccharide capsule (Casadevall and Perfect, 1998), but iron is also an essential cofactor forthe survival of virtually all organisms. Therefore, pathogens possess specific mechanisms thatallow survival in the mammalian host environment including the ability to acquire iron that istightly bound by host proteins.b. Iron Acquisition.The quest for iron by organisms is a complex relationship involving complicated ironchemistry, limited bioavailability, iron cytotoxicity and fierce competition for this valuablenutrient. Iron by its chemical nature is more predominant as the ferric (Fe3+) form. That is, inair, the ferrous (Fe2+) form will spontaneously oxidize to the Fe3+ form. In aqueous solutionFe3+ is essentially insoluble at neutral pH and thus not bioavailable. Fe2+ on the other hand isrelatively soluble at neutral pH. Fe3+ binds to its ligands more tightly than Fe2+, partly due tocharge difference, which makes it more difficult to exchange with an exogenous ligand becauseit has a larger kinetic barrier to overcome (Wilkins, 1991). This makes Fe3+ relativelyinaccessible to organisms in the absence of sophisticated iron acquisition mechanisms. The otherconsideration is that ferrous (Fe2+) iron is highly cytotoxic, and can actively participate in theformation of damaging superoxide and hydroxyl radicals. So, although iron is essential for thesurvival of virtually all organisms, iron acquisition and storage must be tightly regulated.Organisms have developed strategies to reduce Fe3+ to Fe2+ with plasma membrane reductases,but also must limit the level of Fe2+, sequester ferric iron from competitors through chelationsystems, and regulate iron levels by utilization of high and low affinity uptake systems (Kosman,2003).c. Reductive fungal iron transport systems.In fungi, the most developed understanding of fungal iron transport and accumulation isin the model organism S. cerevisiae. In this yeast, reductive iron acquisition occurs through the20cooperation of iron permeases, reductases and oxidases in both high and low affinity uptakesystems as well as non-reductive uptake through siderophore facilitators. Genes involved in ironuptake in S. cerevisiae are summarized in a review by Kosman in 2003. The ferric to ferrousuptake system is dependent on extracellular reduction of Fe3+ by ferric reductases and there areseven reductases in S. cerevisiae (FREJ- 7). Internalization of the iron can occur by directferrous uptake through divalent metal iron transporters Smfl -3p (Pinner et a!., 1997; Chen et al.,1999; Cohen et al., 2000). Another direct uptake system is through Fet4p, a low affinity process.Indirect ferrous uptake can occur though a collaboration of Ftrlp, a high affinity iron permeaseand Fet3p, a multicopper oxidase. Thus, this system is dependent of ferric to ferrous to ferricredox cycling prior to uptake, a process that is not well understood (Kosman, 2003). Theadvantage of the latter system is its high affinity, the KM value for Fet3p/Ftrlp uptake is 0.2 1iM(Dancis et a!., 1992) whereas the KM for Smflp and Fet4p transport of ferrous iron is 2 tM(Chen et a!., 1999) and 35 tM (Dix et a!., 1994:1997) respectively. In conditions wherebioavailable iron is abundant, the high affinity uptake system can be down regulated and ironwill be transported by the low affinity systems (Dancis, 1998; Leong and Winkelmann, 1998;Nelson, 1999; Radisky and Kaplan, 1999).d. Reductive iron transport mechanisms in C. neoformans.A number of iron transport systems have been identified in C. neoformans. Similar to S.cerevisiae, these include both high and low affinity uptake systems (Nyhus and Jacobson, 1999,Lian et a!., 2005). The high affinity uptake system is comprised of a high affinity iron permease,Ftrlp and the multicopper oxidase, Fet3p and is found on Chromosome 12 in the serotype Dstrain JEC2 1. Jacobson et a!., (1998), showed biochemically that there is a low-affinity ironuptake system in C. neoformans. Wonhee Jung in the Kronstad Laboratory has found a secondcluster of genes for a putative iron permease and an oxidase on chromosome three of the strainJEC2 1 and these are strong candidates for a low affinity uptake system. He is presentlycharacterizing this uptake system along with a putative iron responsive regulator protein, CIR](personal communication). He has found by Northern analysis that the high affinity uptakesystem on chromosome 12 is iron dependent and a target of Cirip, but the putative low affinitypermease is independent of iron and not a direct target of Cir ip. He also identified three putativehomologs of ferric reductases in C. neoformans, denoted by the TIGR identifiers (163.m03777,21163.m06361 and 177.m02865) Unlike S. cerevisiae, C. neoformans does not appear to have theability to undergo reductive iron uptake from siderophores, as it does not possess the FIT genesnecessary to reduce the tightly bound Fe3+ from siderophores (Protchenko et al., 2001).e. Siderophores and siderophore transport.Recently, the transcript for a homolog of a siderophore transporter (SIT]) was identifiedby SAGE (Lian et a!., 2005) as upregulated in low iron conditions. Siderophores are smallmolecules that bind ferric iron with high affinity and the resulting complexes can then betransported into cells. SIT] therefore may be necessary for the utilization of siderophore boundFe3+. The use of siderophores for the acquisition of iron is common for both fungi and bacteria(Barasch and Mori, 2004) and siderophores allow pathogens to steal tightly bound iron fromhuman proteins such as transferrin or ferritin or from nutritional competitors in theirenvironment. Enzymes necessary for siderophore synthesis but not uptake include an ornithineN5-monooxygenase and a non-ribosomal peptide synthase (Haas, 2003). Specific transportersare required for non-reductive uptake of ferric-siderophore complexes. Interestingly, someorganisms such as C. neoformans, Candida albicans and S. cerevisiae have homologs forsiderophore transporters even though they do not appear to possess the enzymes necessary forthe synthesis of these molecules (Hans, 2003). This may allow utilization of siderophoresproduced by competitors for efficient iron acquisition. In S. cerevisiae there are four siderophoretransporters ARN], ARN2/TAF], ARN3/SIT] and ARN4/ENB] (Lesuisse et al., 1998 and 2001;Haas, 2003). C. neoformans has many putative siderophore transporter homologs (straindependent) including SIT], which is a homolog of S. cerevisiae ARN3/SIT] and C. albicansARN1/SITA. Siderophore transporters are part of the major facilitator superfamily (MFS) ofsecondary transporters (Lesuisse et al., 1998; Kim et a!., 2005). The CnSitlp homologScArn3p/ Sitip has a high affinity for the hydroxamate siderophore ferrioxamine B (FOB;feroxamine) and lesser affmity for ferrichromes (FCH5) and ferricrocin (FC) (Lesuisse et a!.,2001).22E. Melanin Production.a. Melanin synthesis and placement.There are two major types of melanin in fungi: Dihydroxynapthalene (DHN) melanin issynthesized through the polyketide synthase pathway and DOPA melanin is synthesized throughphenol oxidase enzymes such as laccases and tyrosinases (Langfelder et a!., 2003). C.neoformans produces DOPA melanin through laccase enzymes. As mentioned above, two suchenzymes are present in the fungus encoded by LAC] and LAC2 with the former being thedominant enzyme responsible for the synthesis of melanin under glucose-starved conditions (Zhuand Williamson, 2003). Laccase can oxidize a wide range of substrates including ortho- and paradiphenols, aminophenols, diaminobenzenes and catecholamines, notably 3-4, dihydroxyphenylalanine (DOPA) and dopamine (Williamson, 1997). The reaction catalyzed by laccase involvesthe oxidation of the diphenolic compound e.g. DOPA to a highly reactive intermediate,dopaquinone (DQ). Spontaneous autooxidation reactions then take place to formleucodopachrome, then dopachrome (DC). This intermediate non—enzymatically decarboxylatesto form 5,6-dihydroxyindole (DHI) and 5,6-dihydroxyindole carboxylic acid (DHCI), which thenundergoes first a two electron oxidation to indole-5,6-quinones. The final step of synthesis isslow and involves the polymerization to melanochrome and then to melanin (Williamson, 1997).Melanin has been implicated in virulence in C. neoformans due to its antioxidant capabilities thatlikely protect the fungus against reactive nitrogen and oxygen species produced by host effectorcells. Since laccase can oxidize neurologic catecholamines such as DOPA, dopamine,norepinephrine and epinephrine, this has been suggested to partly explain its predilection forcolonization of the brain (Casadevall and Perfect, 1998). Melanization is in fact required forextrapulmonary dissemination to the brain (Noverr et al., 2004). The microstructure of thecryptococcal cell wall was recently reported (Eisenman et al., 2005) where it was shown thatmelanin is an integral part of the cell wall structure. The report suggests that melanin is placed inthe cell walls in concentric layers of spherical granular particles of about 40-130 nm in diameter.The incorporation of melanin into the cell walls lends a contribution to rigidity and may furthercontribute to the survival of the fungus in the human host.23b. Redox interactions of iron and melanin.Aside from laccase oxidizing a wide range of diphenolic compounds, it also carries ironoxidase activity (Williamson, 1997). Melanin has the ability to reduce or oxidize iron. Thesecharacteristics set the stage for possible redox interactions between melanin, laccase and iron andmay have implications in the already complex mechanisms involved in iron acquisition. Melaninand its redox buffering capability is responsible for neutralizing antioxidants released by hosteffector cells. Studies have suggested that electrons exported by the yeast to form extracellularFe(II) maintain the reducing capacity of the extracellular redox buffer (Jacobson and Hong,1997).F. Rationale and Aims of this Study.The onset of the genomic era and the advent of computer technology have exponentiallyincreased the efficiency and scope possible in genetic research. The work described in this thesisinvolved the identification and characterization of iron regulated genes in C. neoformans,through the development of genomic resources and functional biological studies.The first objective was to contribute to the construction of physical maps for a serotype Aand a serotype D strain. The first portion of this work involved the hybridization of 125 markersto set of 9,216 BACs from the JEC21 library and 6,528 BACs from the 1199 library on a high-density filter. First, a set of 96 overgo (Ross et al., 1999; Materials and Methods) probes(4Omers) was employed in a pooled format to rapidly match genes and ESTs with contigs.Second, overgo probes to BAC end sequences were employed to fill in missing data in the cross-reference analysis of the contigs. Finally, overgo and plasmid-derived probes for specificmarkers linked to the electrophoretic karyotype (Spitzer and Spitzer, 1997). This data wasintegrated into the fingerprint maps and was key in the determination of many joins in the maps,provided the first genome-wide comparison of gene synteny between two strains of the fungus,and linked contigs to specific karyotype bands. The physical maps for two serotypes, A and Dalso provided structural genomic information that was key in the assembly of the total genomicsequence for serotype D at the Stanford Genome Technology Center (SGTC) and The Institutefor Genomic Research (TIGR) and serotype A at the Broad Institute of Harvard andMassachusetts Institute of Technology (MIT). The robust assemblies provided the resourcesnecessary to annotate SAGE data.24The second objective was to analyze the low iron transcriptome of the serotype A strain,H99. This portion of the work involved the analysis of the low iron transcriptome for theserotype A strain, H99 using SAGE. Low iron conditions are important in relation to theelaboration of the polysaccharide capsule. Iron is also an essential cofactor for the survival ofvirtually all organisms and pathogens must possess mechanisms to acquire iron when low freeiron is available as is the case in the mammalian host environment. This work was specificallyfocused on iron regulated genes and mechanisms of iron regulation and uptake. A SAGE librarywas constructed from mRNA isolated from cells grown in low iron medium (LIM) using theMicroSAGE protocol (Datson et al., 1999). 19,278 high quality tags were obtained. The 100most abundant tags were annotated, in addition to the 100 more and less abundant tags incomparison to three other libraries (700 tags total): 1) from cells grown at 25°C on yeast nutrientbroth (YNB); 2) from cells grown at 37°C in YNB (Steen et al., 2002); 3) and from cells isolatedfrom rabbit cerebral spinal fluid (CSF) (Steen et a!., 2003), A number of interesting genes wereidentified in the transcriptome analysis including those involved with the response to stress andiron uptake. A key finding was that the low iron transcriptome was remarkably similar to the invivo library from cells grown in rabbit CSF and significantly distinct from the libraries grown inYNB. The functional characterization of the low iron transcriptome in the strain B3 501 A led tothe identification of a key gene necessary for iron uptake.The third objective was to characterize the gene SIT] with respect to iron uptake andcellular function in C. neoformans. Thus, this work focused on the gene SIT], which encodes aputative siderophore transporter. The gene was characterized through the construction of nullmutants in three strain backgrounds. Serotype D strains B3501A and JEC21 and the serotype Astrain, H99. The SIT] transcript was found to be upregulated in low iron vs. iron repleteconditions. These data were concordant with findings from previous SAGE analyses (Lian et a!.,2005). Deletion of SIT] led to pleiotrophic phenotypes involving melanization and cell wallstructure in the serotype D strains and defects in iron uptake for all strains. This work showedthat SIT] is required for the use of siderophore bound iron in all strains, for growth in the lowiron environment (for strain B35O1A and H99), for melanization in both serotype D strains(JEC2 1 and B3 501 A), and for proper cell wall organization in all strains. It was shown that thecAMP pathway affects siderophore utilization in serotype A but not D. The SIT] mutants haveincreased tolerance to the iron-dependent drug phleomycin and to amphotericin B in the JEC2125background. Finally, this work showed that SIT] is not required for virulence in the murinemodel in serotype A.In summary, the work described in this thesis contributed to the development of robustgenomic resources, provided information about gene transcription under an important nutritionalcue and further characterized a key gene involved in iron utilization. In addition, significantdifferences were noted in the phenotypes between strains of two different serotypes, A and D aswell as two strains within the same serotype, D. The accumulated information provides lastingresources for the Cryptococcus scientific community, reinforces the importance of the low ironnutritional cue, furthers our understanding of iron regulated genes and iron acquisition andprovides new insights about genomic and biological differences between strains of this importanthuman pathogen.26CHAPTER TWO: Physical Mapping of the Genomes of Serotype A and D Strains of C.neoformansINTRODUCTIONThe construction of physical maps was initiated for two C. neoformans genomes as partof an international effort to obtain the complete genomic sequences for strains representing the Aand D serotypes. In this chapter, hybridization experiments are presented that place markers onthe maps for both serotypes. These experiments contributed significant value to the maps andprovided the first genome-wide comparison of gene synteny for two strains of the fungus.Further, hybridization experiments allowed a correlation of specific contigs to electrophoretickaryotypes of C. neoformans. Molecular typing has documented numerous variations betweenthe serotype A and serotype D genomes (Meyer et al., 1993; 1999; 2003; Kidd et al., 2005). TheC. neoformans physical maps further investigated the variation by comparison of the twoserotypes to identify regions of polymorphism. Polymorphic regions are of interest becauseserotype A strains are more prevalent clinically and generally more virulent than the serotype Dstrains. Regions of difference in the genomes may contain virulence factors.The construction of the maps was performed at the MSGSC. The map construction andanalysis was comprised of multiple steps including: 1) BAC clone fingerprinting; 2) contig sizeestimation; 3) BAC end sequencing; 4) BAC end sequence alignment to TIGR shotgunsequence; 5) hybridization of probes to BAC clones and integration of markers to maps throughthe FPC program; 6) comparison of gene synteny; and 7) analysis of the relationship betweenelectrophoretically separated chromosomes and the contigs of the JEC2 1 and H99 maps. Thefirst four steps were not directly part of this thesis work but will be summarized as they relate tothe final three steps (5-7) that will be detailed in these results. The maps were published inGenome Research in 2002 (Schein et al., 2002). Note that J. Schein and K. Tangen contributedequally to this publication and were listed as co-first authors.C. neoformans BAC libraries were prepared at ResGen (Huntsville, AL) in the BACvector pBe1oBAC11 (Wang et al., 1997). For the BAC clones that were fingerprinted, theaverage insert size was reported to be 114.54 kb for the H99 library and 110.74 kb for the JEC21library (based on a sample of clones). High throughput, agarose gel-based BAC fingerprinting,fingerprint map assembly and manual editing was performed as described previously (Marra et27al., 1997; Marra et al., 1999; McPherson et al., 2001; Schein et al., 2004) with the exception thatrestriction fragment identification, fragment mobility and size determination was performedusing recently developed automated analysis software (Furhmann et al., 2003). For each library,a total of 3,072 bacterial clone glycerol stocks arrayed randomly into eight 384-well plates wereprocessed for fingerprint map construction. Each BAC clone was fingerprinted to determine thenumber and size of Hindill restriction fragments contained in the insert. Fingerprints weresuccessfully obtained for 2,642 JEC21 clones and 2,612 H99 clones. The average insert size forfingerprinted clones in the JEC21 library was 108,560 bp and 107,648 bp for the H99 library, asdetermined by the fingerprint analysis. A fingerprint database for each library was created andanalyzed using the program FPC (Soderlund et a?., 1997, 2000; Ness et a!. 2001;http://www.genome.clemson.edulfpc/). A high-stringency automated assembly was firstperformed in FPC to bin together clones with substantial overlap based on shared restrictionfragments. In order to maximize the likelihood that each bin represented a region of contiguousDNA, or “contig”, a minimum of 85 - 90% shared restriction fragments was required for clonesto be binned together. The automated fingerprint assembly resulted in the creation of 276contigs in the JEC21 database and 261 contigs in the H99 database. Additional contig integritywas achieved by manual interrogation and editing of each contig via tools within the FPCsoftware program, using fingerprint similarities to refine clone order and clone overlaps. Cloneswith fingerprints that appeared to be contaminated (comprised of DNA from more than oneclone) or partially digested clones were removed from the database during the editing process.Following the refined positioning of clones within all contigs, clones at the ends of each contigwere compared against all other clones within the FPC database at a reduced minimum requiredfingerprint overlap (approximately 50% shared restriction fragments) to identify potential joinsbetween contigs. Potential joins between contig ends were manually examined and permittedonly where the joins did not result in inconsistencies in the fingerprint data. Upon completion ofthese manual edits, the JEC21 map contained 2,322 clones, the H99 map contained 2,529 clones,and each map had been assembled into 20 sequence-ready contigs.The availability of fingerprinted BAC data from two closely related strains allowed a testof whether contigs could be assembled with orthologous clones from each genome. Specifically,FPC was used to analyze the combined set of fingerprints from both strains. No contigs could begenerated that were composed of clones from both strains. Thus, the genomes of the serotype A28strain H99 and the serotype D strain JEC2 1 are sufficiently divergent to preclude analysis ofsynteny based on HindlIl restriction digestion patterns.Finally, BAC clones comprising a minimally overlapping tiling set were manuallyselected for each contig in both databases. Great care was taken to ensure that shared restrictionfragments could be identified in the fingerprints of overlapping clone pairs. The selected tilingpath clones represent a collection of overlapping clones covering the genomes of JEC2 1 (165tiling path clones) and H99 (163 tiling path clones). These tiling sets will therefore be useful forassembling and finishing the genomic sequences of these strains.An automated algorithm was used to compare the restriction fragments of overlappingclone pairs in the tiling clone sets selected for each contig. The unique fragments for each tilingpath clone were identified and their sizes summed to estimate the overall size of the contigs.Specifically, the algorithm performed the following for each contig: 1) to the total contig size,added the sizes of all the fragments in the left-most tiling path clone in the contig; 2) identifiedfrom the next left-most tiling clone in the contig all unique fragments (not shared with theprevious clone) and added them to the total contig size; and 3) repeat (2) until all tiling clones inthe contig have been processed. Shared fragments were as defined by the FPC parameters suchthat two fragments are considered the same if their calculated mobilities are within 7 mobilityunits of each other.For the JEC21 map, the contigs range in size from 184,760 bp (6 clones) to 1,748,127 bp(321 clones). For H99, the smallest contig (84,272 bp) contains two clones and the largest of1,356,533 bp contains 246 clones.Correlation of shotgun sequence data with fingerprinted BAC clones allowed the contigsto provide a framework to assemble the existing shotgun sequence data for both strains. In orderto facilitate the alignment of the physical maps with the emerging genomic sequence data, theends of the fingerprinted BAC clones were sequenced and the traces were contributed to theshotgun sequence databases for both strains. The genomes for two serotype D strains, JEC2 1and B3501A of C. neoformans have now been completed and published (Loftus et a?., 2005) andthe serotype A strain, H99 is near completion at the Broad Institute of Harvard and MIT. Thephysical maps were a significant contribution to these efforts.BAC end sequence reactions were performed for both ends of all 3,072 clones from eachfingerprinted BAC library, for a total of 6,144 BAC attempted BAC end reads per strain. A total29of 4,772 (78%) successful BAC end sequences were obtained for JEC21 BACs with an averageread length of 540 bp. Of the successful reads, 4,186 were derived from clones that hadfingerprints in the map. Of the fingerprinted JEC21 BAC clones with successful BAC endsequences, 1,939 had both ends represented in the dataset (3,878 total end reads, or 93%) and308 clones had a single associated end read. For H99 clones, 4,908 (80%) successful BAC endsequences were obtained with an average read length of 560 bp. Of these successful reads, 4,390were derived from clones that had fingerprints in the H99 map. For the fingerprinted H99 BACswith end sequences, 1,957 had sequence represented from both ends (3,914 total reads, or 89%)and 476 had a single associated end read.Since the maps were produced prior to a finished genome, JEC2 1 BAC end sequencesderived from mapped BACs were correlated with the JEC2 I whole genome shotgun sequenceassembly generated at TIGR, representing nominally 3.5-fold coverage of the C. neoformansgenome and including the BAC end sequence data. Each of the BAC end sequences wascompared against the complete set of genomic sequence assembly contigs using the BLASTalgorithm (Altschul et al., 1990). Using this methodology a total of 7,643,886 bases of TIGRshotgun sequence were unambiguously correlated with the fingerprint contigs, or 48% coverageof the physical map.Finally, the hybridization experiments provided an important contribution to the use ofthe physical maps of JEC21 and H99 in the comparison of gene synteny between strains, theinitial placement of markers on genomic assemblies of total sequence and the preliminaryassignment of specific contigs to electrophoretically separated chromosomes. The results forthese experiments and their contribution to the maps are detailed below.30MATERIALS AND METHODSA. Filter Preparation and Layout.High Density filters containing C. neoformans BAC clones were purchased from ResGen(Huntsville, AL). The organization of the clones on the filters can be found in Appendix TIc.B. Probe Design.The Overgo method, as developed by Dr. J. D. McPherson (Ross et al., 1999; Volirath etal., 1999), and Overgo Maker ( were usedto design 123 probes for hybridization to the fingerprinted BAC clones. The sequences for thehybridization probes originated from known C. neoformans genes in GenBank(, putative genes identified in the JEC21 genomic database at the SGTC;http://baggage.stanford.edulgroup/C.neoformans/), expressed sequence tags (EST5;http://www.genome.ou.edulcneo.html), karyotype markers (Spitzer and Spitzer, 1997), and BACend sequences ( The 40mer overgo probes were checked forredundancy by searching against the JEC21 genomic database(http://baggage. stanford.edulgroup/C.neoformans/) and the H99 EST database( with the BLASTn algorithm. The pooled probes werepurchased in a 96 well format (2 x 96 well plates of 24mers) from Life Sciences, GIBCO-BRL(now Invitrogen, Inc.). The 25 additional overgo probes (2 x 25-24mers) for 26S rDNA, BACclone end sequences and karyotype specific markers were purchased from the Nucleic Acid andProtein Service Facility (NAPS) in The Michael Smith Laboratories (formerly the BiotechnologyLaboratory), University of British Columbia. Sequences for the PKA2 and HIS3 genes were on a4.5 kb fragment (PK42; pCD49) and a 604 bp cDNA in a TOPO TA vector (HIS3; pMJB54)supplied by Dr. Joseph Heitman from Duke University. Probe sequences can be found inAppendix ha.C. Hybridization Protocol.The Overgo protocol was used for hybridization (Ross et al., 1999) except that freenucleotides were removed with a nucleotide removal kit (Qiagen) and filter washes wereperformed in 50 mL of 4 X SSC/0.1 % SDS, 1.5 X SSC/0.1% SDS and 0.75%SSC/0.1% SDS at55 °C. Filters were exposed to film for three days at —80°C.31The plasmid-derived DNA fragments were labeled with an Oligonucleotide Labeling Kit(Amersham Pharmacia Biotech., Inc., Piscataway, NJ). The positive control was labeled as perthe Research Genetics High Density Filter manual by a Random Priming Reaction. A flowchartof the hybridization strategy for pooled probes can be found in Appendix lib.D. Data Collection and Analysis.All data was organized and processed in Microsoft Excel workbooks. Autoradiographswere read by eye with a 384 well grid. Data from the autoradiographs was interpreted accordingto the Research Genetics, Inc. manual. Each BAC clone was spotted twice on the filter. Thepositive clones were identified by a unique spotting pattern and field location. Filter layout andspotting patterns can be found in Appendix TIc. Results for probes were manually deconvoluted,and compared for each column and row. Clones found at the intersection of a column and rowwere assigned to the probe at that address. Ambiguous clones which had an anchor point; 1columnl>1 row or >1 columnll row were resolved manually. Results for all positive BAC clonesto individual probes can be found in Appendix lid.E. Data Integration.Integration of markers into physical maps was performed by the mapping group at theMichael Smith Genome Sciences Center (MSGSC) (Formerly the British Columbia GenomeSciences Center).F. Southern Analysis of HindilI Digested DNA and CHEF Gel Separated Chromosomes.Southern blot preparation and hybridization was performed as per Sambrook et al.,(1989). Genomic DNA was digested for 4 hours at 37°C with HindIII, separated on a 1% agarosegel and transferred to a nitrocellulose membrane overnight in 20X SSC. The CHEF blot wasprovided by Dr. Joseph Heitman. The 26S rDNA probe was labeled with 32P using the overgoprotocol (Ross et al., 1999) as in section C. Hybridization Protocol-above. The 26S overgo probe(Appendix ha) was used for hybridization to the HindlII and CHEF blots. The HJS3, plasmidderived DNA fragment (pMJB54) was labeled with an oligonucleotide labeling kit (AmershamPharmacia Biotech., Inc., Piscataway, NJ) for hybridization to the CHEF blot.32RESULTSA. Summary of Hybridization Experiments.Hybridization experiments were performed to place markers for known genes, ESTs andBAC ends onto the maps to identif’ corresponding contigs between the maps and to examine theconservation of synteny between the strains. Hybridization data from probes derived from BACend sequences were used as additional evidence for the identification and evaluation of potentialcontig merges. A summary of the marker data from the hybridization experiments is presentedin Table 1. Three sets of probes were used to identify hybridizing clones arrayed as a set of9,216 BACs from the JEC21 library and 6,528 BACs from the H99 library on a high-densityfilter. An example of an autoradiograph hybridized to a row pooi of 12 probes can be found inFigure 1. First, a set of 96 overgo (Ross et a!., 1999; Materials and Methods) probes (40mers)was employed in a pooled format to rapidly match genes and ESTs with contigs. Second, overgoprobes to BAC end sequences were employed to fill in missing data in the cross-referenceanalysis of the contigs. Finally, overgo and plasmid-derived probes for specific markers linkedto the electrophoretic karyotype (Spitzer and Spitzer, 1997) were also employed in an attempt tomatch contigs with specific chromosome-sized bands. Note that more BAC clones wereavailable on the high-density filter than were fingerprinted. Overall, 82 and 102 markers wereplaced on contigs for JEC21 and H99, respectively. Results for individual probes can be foundin Table 2. For all markers, a total of 1603 JEC21 and 1572 positive BAC clones wereidentified. Probes gave a more successful hybridization result to the BAC clones from the strainwhose sequence was used to design the probe.Table 1: Summary of the hybridization of selected markers to BAC clones of JEC2 1 and H99.Hybridization Probes Positive Average BACProbe Origin format Strain successful/attempted BAC clones/probe Total PositiveGenes/ESTs Pooled JEC21 79/96 1175 14.9 56H99 88/96 1307 14.9 76BAC Ends Individual JEC21 14/15 351 23.4 13H99 15/15 428 28.5 15IndividualKaryotype JEC21 14/14 505 36.1 13markers H99 11/14 280_________20____________11_____Probes were hybridized to a total of 9216 JEC2 1 and 6528 H99 BAC clones. Pooled probes were employed in the Overgoformat described by Ross et al. (1999). Unsuccessful probes failed to hybridize or gave ambiguous results.33Figure 1: Example of an autoradiograph showing the results of a rowpool of 12 overgo probes hybridized to a high-density BAC clone filter.Multiple 384 well plates are spotted in each field 1-6. Each BAC clone isspotted twice per well. A unique spotting pattern (Appendix Tic) indicates theoriginating plate number for the positive clone. The filter orientation isdetermined by the positive control probe that hybridizes to the filter edges.Identifying marks include the reference marks on two corners where one isabsent in the top left corner and an extra mark is present on the bottom rightcorner. Library identification marks are found on the right vertical edge of thefilter. A 384-well grid is placed over each field and the originating 384 wellmicrotitre plates for each positive clone are manually recorded in Excelworkbooks.+CDCDC)CDCDCDC)CDC.)CRows A-P34Table 2: Summary of hybridization experiments for individual probes.CPA2CPAIPLB1CBPIL1AI partialCAPS9ERG1ICNAIUREIMAT ALPHAPMAISTE12 ALPHAGPBIMP GENEGRASSTE2ORASICAP 10RHOI mRNALAdOPAlADE2GPDTORiPRRITOPOI‘PC’FICSIMDRIACTINTHY SYNCAP6OCELlTEFICHSIMAN OHPREICAP64GAL7IJBIQ‘322 330 09 3319 203— H993 H998 11993 H99CBS 132WSA-21JEC2IATCC63S2H99H99CBS 804H991199H9904500B3501010JEC2IB350IB350IB3501H991199H991-199MI-1061199B3501B4500B4500M 1-106H998 BAd Clones HybridioedName Gene Fonetion JEC2I 1-199 Strain Origin2663023 Iio5 200 025 Ii2 00 1929 260 10 537 170 024 6CYCLOPI-IILIN A REGULATOR OP TRANSLATION (CAMP rag.)CYCLOPHILIN A REGULATOR OF TRANSLATION (cAMP rcg.)SECRETED PHOSPHOLIPASE BCALCINEURIN-BINDING PROTEINP450 LANOSTEROL 1-4 ALPHA DEMETHYLASESEROTYPE AID CAPSULE GENELANOSTEROL -4 ALPHA DEMETHYLASECALCINEURIN A CATALYTIC SUBUNITUREASEPHEROMONE PRECURSORPLASMA MEMBRANE H (+) ATPaseSTERILE 12 KINASE0-PROTEIN BETA SUBUNITMATING RELATEDOROTATEPHOSPHORIBOSYLTRANSFERASEIOROTIDINE MONOPHOSPHATEPYROPI-IOSPHORYLASESTERILE 20 ALPHA KINASESMALL GTP BINDING PROTEINCAPSULE ASSOCIATED PROTEINRHO FACTORDIPHENOL DXI DASE (LACCASE)0-PROTEIN ALPHA SUBUNITPHOSHORIBOSYL AMINOIMIDAZOLE CARBOXYLASEGLYCERALDEHYDE-3-PHOSPHATE DEHYDROGENASEPHOSPHATIDYL INOSITOL-3-KINASEPKBPI2 MACROLIDE BINDING PROTEINTOPOISOMERASEINOSITOL PHOSPHORYLA CERAMIDE SYNTHASEGL[JCAN SYNTHASEMULTI DRUG RESISTANCE PROTEINACTIN GENETHYMIDYLATE SYNTHASECAPSULE ASSOCIATED PROTEINCELl PROTEINTRANSLATION ELONGATION FACTOR I-ALPHACHITIN SYNTHASEMANNITOL DEHYDROGENASEPROTEASOME SUBUNITCAPSULE ASSOCIATED PROTEINUDP-GLIJCOSE-D-GALACTOSE-I-PHOSPHATE URIDYLTRANSFERASEUBIQUITIN CARBOXY EXTENSION PROTEIN1992164113 ATCC 6352359 BAC Cones HybridiaedName Gene Fonclioo JEC2I HOE Stroio OriginCNIJB14 POLYUBIQUITIN 2 15 ATCC 6352TRPI PI4OSPHORIEOSYLANTHRANILATE ISOMERASE 0 0HIS3 IMIDAZOLE GLYCEROL PHOSPHATE OEHYRATASE 9CNRE-I REPETITIVE DNA ELEMENT E INMT N-MYRISTOYLTRANSFERASE EXONS I-I I 10 3 L2 10425ARP AOP-RIEOSYLATION FACTOR EXONS 1-7 4 5 L210425TRP2 TRANSLATION ELONGATION FACTOR 2 10 22IPI ANTIPHAOOCYTIC PROTEIN 26 35MPDI MANNFrOL- I-PHOSPHATE DEHYOROOENASE 24 19D11AI DELAYED TYPE SENSITIVITY ANTIGEN 31 36RHOI RHO FACTOR 0 0PIJKCI PUTATIVE URCI KINASE 12 10 JEC2I (U moydis homolog(PASAI PUTATIVE ASPARTATE SEMI-ALDEHYDE DEHYDROOENASE 26 17 JEC2I (U erojeh.s homalagiPHSPIO PUTATIVE HEAT SHOCK PROTEIN Spornhe 21 0 JEC2 I (S pornbs’ komoIog(PHSP78 PUTATIVE HEAT SHOCK PROTEIN Sposihe 22 II JEC2 I (S pen;ho homalag)PHSP9O PUTATIVE HEAT SHOCK PROTEIN 0. foo;penico 10 I JEC2I (0. /ao;po ice hoosolog)3EC21 (S corey/nec Hop70PHSPI04 PUTATIVE HEAT SHOCK PROTEIN Secrevisiac Hsp78 7 12 hossolog)PATPI PUTATIVE TRANSCRIPTION FACTORS pee;hc 7 I JEC2I (S porohe hornolog)PMTS1 PUTATIVE RIBOSOMAL PROTEIN LO 2 0 JEC2I (S penihc homalog)PPVPI PUTATIVE PROTEIN TYROSINE KINASE N. crane 4 2 IEC2I (N. eresse homolog)PPYP2 PUTATIVE PROTEIN TYROSINE KINASE N. crone 4 4 JEC2I (N. crease homolog)alaOOen.fl RET giil197059 (201351) ribosomal proleio L3 -I 360 3.Ie-33 3 I H99olo0Scn.fl ESTgoIIPIDIe2O6572 (Y00841) core prolcia II -3 166 4 9c-I I 7 9 H99olalOco.fl EST bbs1575 17 (557516) macil phosphorihosyl Icansfcrasc -2 311 4.Ia-27 II 6 H99olallee.rI EST goIIPIDIcII9O443 (A200I342( Polstioc S-phase-specific ptoleio +3 359 IKe-OS 12 32 H99albO4cnsl EST9iIII97O59 (201351) ribosossal proleio L3 +1 650 4.Ic-65 6 4 H99nlIsO7cn.rl EST0eIIPIDIdIOI4552 (D89194( similar to Ral ATP cilralo [(aso -3 224 6.4e-I0 2 13 H99alc0lcn.1l EST goIIPIDJoI23I300 (AJ003I97(adooiso oacloo6dc traoslocasa -t 207 I.3e-24 10 12 H99alcO2cn.ft EST9i(2029925 (AC002291) Sitoilarso Dns.I (heal shock protois) +2 206 I.Oe-15 32 17 H99slcO3cn.fl EST0IIPIDIe330964 (Z90762) [sIts acid ovolkelase alpha sabosil -I 206 2.3o-l4 12 9 H99sle04en.rl EST9i13420603 (AP075709( LsFA (Psoodomooss polida) (rokydrie) +1 343 I Se-30 4 11 H99alcO6en.fI EST gel(PIDI0I29I65I (AL023290) kypolkelical pmlcm ((ABC Irsosporter) 4-2 327 3.Sc-20 0 IS HOEalc09en.rI EST0oIIPIDIcII327O7 (Z99I63( kypolhclisal prose (L-aIlo-tkrceeieealdolasc(. +1 260 I (10-21 6 21 H99alctlcnSf EST geIIPIDI4IOI4I33 (0077351 ekosemal protelo LI4 -I 260’0.Oe-22 4 25 H99sldK0cn.fl EST oilS 120243 (AP004672( ribosamal pmseie L41 -I 523 l.3e-49 I 4 HOEofdOKcn.fl ESTgiI2I49ISO (U02536( faltyacidamide kydmlase+I 211 ISo-IS I 0 H99ald08rn.rI EST0t(2099797 (U9b695( deosyoridiocsripkasphatasc-3 349 3.6r-31 0 II H99nldO9cn.fl EST gol(PID(4I0I45I7 (009150) similarlo Saccharamycco cer... -3 113 l.4c-I6 2 2 II H99sIdlfcnSl ESTgoIIPIDIo2634O9 (Z79602( K09E9 2 (Cacoadsabdilis elogaos( -I 227 7 4e-I8 22 23 H99oleOlcn.rI EST9i1054566 (X07371) cmkamyl phosphale oyolhmc -2 330 I.9e-27 9 3 H99aleO4rn.ff ESTgoIIPIDIe334O27 (Z90529) Imoscriplico ieiñalios fhclor lIE (TPIIE( +2 171 9.4c-I2 0 12 H99ole06cn.ff EST90IIPIDI4IO32200 (AB00077O(callalasc +3 122 3.3o-06 12 22 H9936nlfOlcn.fIalfO2cn.fIulIO7cnSlalgO2ca.flalgO3cn.rIalgO4cn.fIalgO8cn.rlalgO8cnSIat gO9cn.f1ulgllcn.flulhOlcn.flnlbOlcn.fInllsl2cn.fl185 rDNA (intergenic)265 rDNAPKA2H153Spil6SpiZSSpi35Spi6Spi7Spi3ZSpISSSpil5Spi29H083109.F11004E17.R118153H04.F11003H22.Fl1003M21.F11003M21.RHOO2KII.F11003C03.F11807684.RH801024.F140152K09.F1100280SFH004008.FESTgil I 841864 (1)871011) nucleic acid binding protein (gtycmc-ricls) -3 ItO 2.Se-05E5Tgij2668565 (1)81804) translation elongation factor-I 663 I .7o-64EST gi11330376)U58758) ZKI 127.5 (Cuenodsabditis elecansl -3 59 I Sc-ICESTgnIIPIDIcI249769 (Z99163) hypothetical protein (turnorsoppressor) +1 290 6.7c-25EST9iI2088722 )AF003 139) strong siosiladey to 60S ribosomal protein -l 373 9.00-34EST bbs1t74324 pepninngeoA +3 56 4.4e-I0ESTciItO4O756 (X79206) ptd e20572t IS pombol +3 531 l.6e-50EST gi1832919 (X7t664) 0141-11 [Saccharomyccs core... -l 214 6.6e-17EST0nIIPIDIeI 173375 (Y13942) OTN Reductese A turoefaciens] -3 143 I .3c-08EST gn11P1D1e1285364 (AL022299) potatiyc cytoclsrome ct -3 330 3.7e-29ESTgnIIPtDIdtOI4629 )D8927t) racicosoore assembly protein +3 213 ICe-tOEST gnIJPtDjdtO222O2 )A0004530) prorcirr arginine N-ruothyl transferaso. -2 143 I.Ic-08EST8OIIPIDIdIO25092 (A00t0049)tmnsaldolase -2 395 4.5e-SO>gi13481 l6Igh)L2264t.IIPIORRI0S>g)61 pCt)49 rebate with dcl (I mglml)dsfosNjdukc.edo pMJB54 eDNA in TOPO TA mlcasc mills EcoRl>prl 153 I609b)AA05 1839.UAAO5 1039 Cn0016-5 Cryptneocens ncoformaoa>gil 153 t697)pbIAAOSlO47.IIAAOS 1847 Cn0025-5 Crvprococcas ocofomraos>gi)ISS t707)gb(AAO5 1057 IIAAOS 1857 Cn0035-5 Cryptococcan neofomsanscpu 153160 lIgbIAAO5 1031. flAAO5 1831 Crr1111116-5 Cwptococcus scofonrrairscpu 153 t602ipbIAAO5 1832. IIAAO3 1032 Cn110117-5 Cryplococcrrs uoofonssanscoil 153 I7O4lgbIAAOS 1854. IIAAO5 1854 Cn0532-5 Cryptecoccus noofnrmeas>gil 15317 I9gibAA05 1869. IjAAOS 1869 Cn0055-5 Crypcococcus neefnmmanscoil 153 I688I9NAAOS 1838. IIAAOS 1035 Cn0015-5 Cryptococcus oonfonrruns>gij 1531701 lgblAAOS 1851 .IIAAOS 1851 Cs0029-5 Cryptococcss srofonrraosBAC END Sequence www.begsc.bc.coBAC END Sequence END Sequence svsvw.bcgsc.6c.euBAC END Sequence www.bcgsc.bc.caEAC END Sequence www.bcgsc.bc.caBAd END Seqscsce snsnw,bcgsc.bc.caBAC END Sequence wsvw.bcgsc.bc.caEAC END Scquence wmnsv.bcgsc bc.caEAC END Sequence wss—sn.begsc.bc.caEAC END Sequence sswss.bcgse be.cuEAC END Sequence www.bcgsc.bc.caEAC END Seqoencc svww.bcgsc.bc.caEAC END Sequence www.bcgsc.bc.ce1415000 1724 133 628 383 1424 35tO 2312 198 BAC Clones HybridleedName Gene Fnnctiun JEC21 1499 Strain Origin270119911991-19911991199N991-19911991-1991199119911991199CES582CES88276 077 II47 2831 37Spitzcr aud Spitaor. 997 ESTs usasi42 37 as bars utypo maabemSpiseer and Spitecr. 1997 ESTs used26 28 as bervetype markemSpitaer and Spiteer, 1997 ESTs used19 16 as karyotype oserbemsSplIcer ced Spiseer, 1997 ESTs trsed26 26 em katyotypc merbersSpitzrr asd Spitaer. 1997 ESTs used38 24 as karynrype mmmarbemSpitzermtd Spiteor. 1997 ESTs esed21 0 as barvoevpc usaskersSpilaermsd Spitoor, 1997 ESTs ased22 24 en kars’orvpe madrersSpitaeread Sprsmr, 1997 ESTs used17 29 as bereotvpe moebemSpiteer aud Spiseer, 1997 ESTs speed37 20 as barvotvpe merkcm119936 28H9940 57119930 26119944 30119936 2911997 13H9928 32119926 3344 261199119911991199Total Clanea Nybridiaed 1603 157137B. Integration of the Markers into the BAC Fingerprint Maps.Summaries of the 20 assembled contigs for each strain can be found in Tables 3 and 4.These tables also list the number of markers that mapped to each contig by hybridization and theestimated total amount of DNA represented in each of the assembled contigs for both strains.With regard to genome size, the estimates of 15.79 Mb for JEC21 and 15.55 Mb for H99 are atthe lower end of the range (15 to 27 Mb) estimated from electrophoretic karyotyping of differentC. neoformans strains (Perfect et at., 1989; Wickes et at., 1994; Boekhout et al., 1997).The genome sizes estimated from the maps are likely to be underestimates because theremay be areas of the genomes that are not represented in the BAC libraries. These may includeareas that are difficult to clone or maintain in E. coil such as telomere or centromere sequencesor areas with an unusual distribution of Hindlil sites. Of course, estimates of genome size fromelectrophoretic karyotyping experiments can also be confounded by problems with chromosomesize determination and the co-migration of different chromosomes. These issues were largelyresolved with the completion of the serotype D genomes (Loftus et at., 2005)The fingerprint maps are available in FPC format at the web site of the MSGSC( In addition, the maps can be viewed with Internet Contig Explorer (iCE)and the BAC end sequences are available at this web site. An example of an FPC output forcontigil of in the JEC21 map containing the MATa locus, is shown in Figure 2.38Table 3: Summary of the contigs in the fingerprint map of JEC2 1.Contig Contig Number of Contig Number of Tiling Number ofNumber Size (bp) Clones Depth (a) Path Clones Markers1 1748127 321 20.16 20 77 1484186 196 14.87 14 811 1434639 250 19.23 14 115 1164507 185 17.04 13 48 1070654 167 17.73 12 610 1038718 163 16.13 11 616 975364 154 16.59 10 418 970748 129 14.57 10 44 853084 128 15.83 9 59 732482 108 16.01 9 319 708498 71 11.09 7 116 697149 98 15.09 7 1112 587336 68 11.95 6 220 503950 91 19.13 6 313 420308 65 17.14 4 415 407895 52 14.22 4 13 379639 49 14.05 4 217 222874 12 5.40 2 02 202053 9 4.67 1 014 184760 6 3.40 2 2Total 15786972 2322* 14.21 165 82***Number of singletons = 3* * 11 markers hybridize to >1 contigs(a) The contig depth is the sum of the actual fragment sizes for all of the fingerprinted clones inthe contig divided by the calculated size of the contig.39Table 4: Summary of the contigs in the fingerprint map of H99.Contig Contig Number of Contig Number of Tiling Number ofNumber Size (bp) Clones Depth (a) Path Clones Markers5 1356533 246 20.12 15 193 1221982 190 16.62 14 1412 1204660 228 20.37 12 1019 1199093 204 18.02 11 1117 1192998 246 22.16 12 141 1128612 212 20.30 11 1710 1080077 218 21.75 12 78 958050 161 18.41 10 720 957898 143 16.15 9 77 854305 155 19.81 8 1613 843814 106 13.56 9 84 746210 92 13.46 9 811 689677 92 13.99 7 99 631074 90 14.75 7 102 601000 87 14.92 7 316 236418 26 11.40 3 214 197615 13 7.21 2 118 190795 8 4.53 2 115 175905 20 11.77 2 16 84272 2 2.01 1 3Total 15551099 2539* 15.07 163 102****48 markers hybridize to >1 contigs(a) The contig depth is the sum of the actual fragment sizes for all of the fingerprinted clonesin the contig divided by the calculated size of the contig.40Figure 2: FPC display from the map of strain JEC21.Contig 11, containing the MAT locus, is shown on the left. Markers hybridizing to BAC clones inthis region are shown above the contig. Green boxes indicate tiling path clones. Gel images onthe right show some of the fingerprints of overlapping clones in the contig. Horizontal linesrepresent positive shared bands between adjacent clones.41TEF2H004E17,RH003H22.FSpiO7 DHA1GreyRamp!Whole!Zoom4ijj6!ITraillclip!8004000QHighGreenIClearTrail!IMove!Remove!liJIBandi!RedraIIRernoveAll!3008K083004C063008P053008F09ae,J008H12J007F12J008F23800I1100_a_;izize=::=I=___1200—1300___1400-1500———•i6001700—-—--—-1800——1900—2000——————2100=‘=—————--------2200————_j—————-—==,——=—2300——_2400::zZEE——2500=-..—=.==—=——2600——=—————2700:,:——2800”—2900——_.__==30003100_—=3200—Whole!Zoom:jj2,0Hidden:IBuriedi!Con1igureDisplay!Clone:!!EditContig!ITrail.,.!ClearFill!Merge!JAnalysisIIsnCtgCheckIsnCtgSteplMergeCtglB.MergeCtg4.MergeCtg43,MergeCtgl7,MergeCtg39.MergeCtg258.CtgilofJEC21_masterClones51of250LMarkersU.of11,Sequencedi5iSpiO6H003H04,FSpi55MF—GEHE STEI2—ALPHASTE20J008D03*3006110*J0.5±J2.4*J007N24+3006L14+3007K18*JOOSF2O*3004C06+3008P09*3006M133007F12*3001803*30081<08*3008009*1008H12*3007102*3004L07*J002H17+3006L08J003P03*1Q07J04+30031104*3006K013006806+3008306+,j006H23*3002M24*3008109*J002824*J003C22*3007E12*30050043007308+JOOILIS*3002319+3006P24*33003400—3500—36003700380039004000===—-——C. Comparison of the H99 and JEC21 Fingerprint Maps.Shared markers were found for 17 and 18 of the 20 contigs for JEC21 and H99,respectively (Figure 3). Note that no markers were found for contigs 2 and 17 of JEC21, butmarkers were found for all 20 contigs in strain H99. Hybridization with two different probes forC. neoformans rDNA sequences revealed that BAC clones carrying rDNA genes were found incontig 7 for JEC21, but were not present in the 6,528 clones from H99 BAC library. This resultmay indicate that the rDNA sequences from H99 have a different organization of HindlIIrestriction sites that precluded cloning of the region. DNA blot analysis of complete HindlIIdigests of H99 and JEC2 1 genomic DNA confirmed that the H99 rDNA contained fewer HindlIIrestriction sites (Figure 4), which may explain its inability to be cloned in pBe1oBAC1 1 using theprotocols employed for library construction. Lists of the sequences that were used as probes foreach marker are provided in Appendix ha.As shown in Figure 3, the hybridization data and the fingerprint information allowed usto match 18 of the 20 H99 contigs to 17 of the 20 JEC2 1 contigs. Markers were used for thecomparison only in situations where they could confidently be mapped to a single contig. Thecomparisons of marker positions between the maps revealed considerable conservation ofsynteny between the strains with several clusters of markers showing identical order. In additionto the overall similarities between the two maps, the conservation of marker order wasparticularly striking for contigs 13 (JEC21) and 3 (H99), for contigs 19 (JEC21) and 9 (H99),and contigs 11 (JEC21) and 5 (H99). As indicated earlier, the hybridization probes for genesknown to be in the MATh locus identified contig 11 in JEC21 (Figure 3; 5A) and contig 5 inH99 as arising from the mating-type chromosome (Figure 3; SB). The MATa region has beenthe focus of targeted sequencing (Lengeler et a!., 2002; Karos et a!., 2000) in part because of theassociation of the a mating-type with virulence (Fan et a!., 2005; Nielson et a!., 2005a and b;Kwon-Chung et a!., 1992). The combined FPC and hybridization data also revealed severalexamples of rearrangements between the two genomes. Specifically eight markers wereidentified whose positions did not agree between the two maps relative to flanking markers(Spi25, Spi35, alb07cn.rl, ale04cn.fl, H003C03.F, URA5 and H003M21.R; Figure 3). Thepositions of some of these markers suggest the presence of an inverted region between thegenomes (e.g. 5pi25 and Spi35) while the positions of others suggest translocations (e.g.43alb07cn.r 1 and al e04cn.fl). These results provided the first comparative view of genomeorganization between strains representing different varieties.The comparison of the locations of specific sets of markers between the two maps alsosuggested that certain contigs may be regions of the same chromosome. For example, thecomparisons of shared markers between the maps suggest that contigs 1, 3 and 13 could beregions on the same chromosome for JEC21 and contigs 3 and 17 could be joined in H99.Similar connections are suggested for contigs 10 and 15, and contigs 7 and 9 for JBC2 1. ForH99, contigs 4 and 18, 12 and 15, 1 and 19 and 7, 8 and 16 could be joined; the mapping dataalso suggest that contig 6 could be joined to contig 5 that contains the MATa locus. Overall,these joins would reduce the number of contigs to approximately 15 for the H99 map and 17 forthe JEC21 map. For comparison, the reported range for chromosome number for the majority ofstrains of C. neoformans was between 11 and 14 (Boekhout et al., 1997) and is now confirmed tobe 14 (Loftus et al., 2005).44Figure 3: Conservation of synteny between the genomes of JEC2 I and H99.The 20 contigs for each map are shown with the markers found on each contig. Note that onlymarkers that mapped to contigs in both strains are shown. Contigs are represented as verticallines with the contig numbers above. The relative positions of the markers within the contigs areapproximated based on hybridization results. Where the relative positions of adjacent markers ina contig could not be reliably interpreted they are grouped by a square bracket. Correspondingcontigs from JEC21 and H99, based on marker content and order, are placed adjacent to eachother for comparison. Differences in marker locations between the two strains are also indicated.No markers were found for the JEC21 contigs 2 and 17; these contigs, along with 14 and the 1199contigs 6 and 14, did not have shared markers between the genomes. The boxed and shadedmarkers also were used to relate the contigs to electrophoretically separated chromosomes (Fig.5).45JEC21 H99 JEC21 H99 JEC21 H9913 3 8 10 16alhl2cnfl alhl2cn.fl aI1 Icn,fl algi Icn.1l IH004001&F H(x)400aF IalgO4cn.f1 IalaO5cii.f1 alaO5cn.f1 algo4cn,11 ialbO4cri.rl albfl4cn.rl I 20alcO4cn.f 1 I I I-1003M21.RI1 GPD H003M21.FGPDTOPOIaklllcn.flIPI H003M21.RIP1a1i1knf1atgrncn ri/ialgO3cn.rI aleO6cii.i1 a1c(cn.f1K6FKS1FKSIMDRI 18MDRI11C}AL7GAl 7I CAP64_ CAP64 CAP1O] alcO4cn.fl{CA[N3O LJREICELl IalcO9cn,rlalcO9cn.rl11003C03.F I3 /9CAPIO‘I}/19CAP1 (11-1002K 1SFalbOlen.rl PLBI I PLB1H(X)2K15.FalbO7cn.rIPASA1 PASAIaliO2cn.i1 I alffl2cn.fI1131O9.FII Ho00Io9.r18 H003C0315/a1lcnf1 i aicoicn.nH007G04.R 11007G04.R9II 1RG11 lRG II12 13alcllcn.fII IA5 HOO2K11.F I I 1-10(J2K11.F12 16 PPYP2IHl)02K09.F}/alh02cn.fl alh02cn.f1 1T002K09.F8al cO3cn.[1 alc03cn.f1H001024.F JEC21 H99CPAI CPAI Contigs without shared markers11001024.F PATFI2* 6**I I7* 14PHSP7811PHS8 ii II} Jr110031-104.F H003H04.P I 14H004E17.R H004E17R MPDI M1’DITEF2 TEF2 P1JKCI PUKC1 I110031122,F H003H22.P20 2*No markers mapped to contigIMF-C,RN’ IMF.GENEI___________________________I STEZOI 1sTh20 PMAI PMAIDHAI_______NMT I I **Contig6may bcparlofcontig546__ __ _= 500kb1kb ladder A B1018•4— originFigure 4: Determination of Hindill fragment sizes for 26S rDNA.Southern Blot of Hindlil digested genomic DNA from C. neoformans. 26S rDNA probe.Size markers were 1 kb ladder from Invitrogen, Inc. Fragments carrying rDNA for JEC21were at the upper size limit of the ladder; the exact size can not be determined but thefragment can be approximated in the range of 6,000-12,000 bp. Fragments with H99rDNA were larger than the upper limit and noticeably larger than the JEC2 1 rDNAfragments, indicating fewer Hindill sites adjacent to rDNA in the H99 strain.A. .JEC21 B. H9912215090407230542036506,51747D. Relationship of Specific Contigs to Chromosome-sized Bands From the C. neoformansElectrophoretic KaryotypeThe positions of specific markers on the contigs were also compared to the publishedlocations of the same markers on electrophoretically-separated chromosomes and the meioticmap. For this analysis, the electrophoretic karyotypes of two progenitors of JEC21, B3501 andNIH12, were used to represent the chromosomes because these strains were used for previoushybridization experiments (Figure 5A; Wickes et al., 1994; Spitzer and Spitzer, 1997). Thepatterns of chromosome-sized bands for these strains appear to be similar or identical to thepattern of JEC21, as determined by Lengeler et al., (2000). The hybridization probes alsoincluded the genes URA5, CAP64, CnLACJ and STE2Oct that have been placed on a meiotic mapby Forche et al., (2000). The JEC21 contigs that hybridized with markers previously assigned tospecific chromosomes are shown in Figure 5A. Our results indicated that the three largestcontigs (1, 7 and 11) from JEC21 contained the same markers that map to the three largestchromosome-sized bands in the two other serotype D strains. The chromosome represented bythe third largest band in these strains contains the MAT locus and this chromosomal location hasalso been established in JEC2I (Lengeler et al., 2000).The rDNA markers are present on the second largest chromosome-sized band in strainsB3501 and NIH 12 (Wickes et al., 1994). Similarly, we found that the rDNA probes hybridizedto the second largest band on a blot of separated chromosomes (Figure 6 B and D) and to contig7 in JEC21 (Figure 5A). These results were in agreement with the hybridization data obtainedby Wang et al., (2001) for the CPA] and CPA2 genes; these genes hybridized to the secondlargest band in JEC21 and were found with the rDNA on contig 7 (Figure 5A). However, it wasfound that the HIS3 probe hybridized to a band equivalent to chromosome 7 (Figure 6 F and H)in contrast to the reported location of this gene on band 11 (Wickes et a!., 1994). This resultindicates that there are differences in the locations of some markers for JEC21 when comparedwith progenitor strains B3501A and N11112. Interestingly, it was later found that a chromosomalduplication event had taken place in JEC21 when compared to B3501A (Fraser et al., 2005). Theduplication included 62,872 identical nucleotides and 22 predicted genes. This duplicationoccurred in the contig 18 (which corresponds to chromosome 4 at TIGR) of JEC21 where thisstudy showed that HIS3 was located (Figure 3 and Figure 5A). This duplication in JEC21 mayexplain the differences in the marker placement on the karyotype when compared to the48progenitor strain B3501A. The sequence for HIS3 was retrieved from GenBank (U04329) andsearched against the TIGR JEC21 database using the BLASTn algorithm, surprisingly, there wasno significant hit. It is difficult to discern whether HIS3 resides in or adjacent to the duplicatedregion, however the lack of HIS3 in the TIGR database may explain its exclusion from thepublished list of duplicated genes in JEC21 if it does reside in that region.These observations suggest that caution should be exercised for some of the comparisonsbecause of possible differences in the karyotypes between JEC21 and the progenitor strains. Inthis regard, several reports have described the variability of the karyotype in C. neoformans(Perfect et al., 1989; Wickes et al., 1994; Boekhout and van Belkum, 1997). Loftus et al., 2005reported that the C. neoformans genome is rich in transposons, many near candidate centromericregions and these features may contribute to karyotype instability. However, in addition tocorrelating contigs with the largest chromosomes, the hybridization data may provide insight intothe contigs that represent the smallest chromosomes; this information may have utility for theanalysis of chromosome structure in C. neoformans. For example, both contig 4 and the 10thchromosome-size band of JEC21 hybridize with the Spi29 marker and have similar sizes (0.853Mb versus 1.02 Mb, respectively). Thus, contig 4 may represent most of chromosome 10 if thekaryotypes are equivalent for B-3501 and JEC21 for this band (Figure 5A).The estimated sizes of karyotype bands on Figure 6 were based on the NIH12 karyotypeby Wickes et al., (1994). They reported a chromosomal size range of 0.770 Mb-3.87 Mb for 12electrophoretically separated chromosomes. A more recent meiotic map was published by Marraet al., in 2004 that reports a size range of 0.80 Mb-2.3 Mb on 14 chromosomes for a total 20.2Mb genome. Finally, with the publication of the complete annotated genomes for strains B3501Aand JEC21 (Loftus et al., 2005) those estimates have been further refined to reach an estimate ofa 20 Mb genome.As mentioned, the complete sequence has now been published for the serotype D strainsB3501A and JEC21 (Loftus et al., 2005). In light of this, markers used in our correlation ofkaryotype bands were retrospectively matched to the final chromosomes of JEC21 using theTIGR database (Figure 5A). Specifically, the sequences used to generate the probes weresearched using the BLASTn algorithm against the TIGR database. The correlation was generallygood although there were again examples where the placement of a marker differed in JEC2149when compared to the B3501A karyotype. These results were in agreement with those notedfrom the initial correlation of contigs to karyotypes. CAP64 was indeed present on chromosome2 (contig 1 of the JEC21 physical map); the mating locus genes: MF gene (pheromone), STEJ2and STE2O in addition to the SpiO6 and Spi55 markers were found on chromosome 4 at TIGR(matching contig 11 in this map). The marker Spi32 was mapped to contig 4 of the JEC21 mapand found on chromosome 10 at TIGR. This marker is an example in which the JEC21placement did not correspond to B3501A. In the karyotype studies for l33501A, Spi32 wasmapped to the third largest band (and the same band as mating type genes), while it was locatedon contig 4 not 11 (where the mating type genes are located) in the JEC21 physical map. This isa good example in which the physical maps correctly located a difference between theorganization of the JEC21 and B3501A genomes. Further support for matching contig 4 tochromosome 10 is provided by the Spi29 marker that also mapped to contig 4 in the physicalmap and is found on chromosome 10 in the TIGR annotation. This marker was mapped to the10th band (Wickes et al., 1994) and 6th band (Spitzer and Spitzer, 1997). This result indicates thatthe Spi29 marker was possibly on the equivalent chromosome in JEC21 and B3501A. It was alsofound that the SpiOl probe hybridized to multiple contigs while Spitzer and Spitzer (1997) foundthat this marker is located on the smallest chromosome in B-3501. In the TIGR database, SpiOlis indeed located on one of the smallest chromosomes, number 13, suggesting that the results inthe initial mapping studies were due to cross hybridization of the SpiOl probe. Two additionalmarkers on the JEC21 map that did not correspond to the progenitor strains were URI45, foundon contig 15 and chromosome 7, where the marker was found on the 4” largest band of theSpitzer study and the 4t 5th or 6th largest bands of the Wickes study, and PKA2 mapped tocontig 16 of this map and chromosome 9 at TIGR where it was located on band 5 of the Spitzerstudy and band 7 of the Wickes study. It should be noted that the actual number ofchromosomes in these serotype D strains has been resolved at 14, so it is likely that these earlierkaryotype studies do not have the correct resolution of the chromosomes. Therefore, somedifferences in marker placement may be an affect of unresolved karyotype bands and notnecessarily different chromosomes when comparing the progenitor strains with the TIGRannotation. A limitation of karyotype analysis in C. neoformans is the possibility of doubletsand triplets that are difficult to resolve. In addition, there is variation of karyotypes even withinserotypes (Meyer et al., 1999; Kidd et al., 2004).50The markers for rDNA and the MAT locus are found on the largest chromosome-sizeband in the serotype A strain N1H371 (Wickes et al., 1994). The MAT locus is known to be onthe second largest karyotype band in H99 (Lengeler et al., 2000) and our hybridization data linkthis chromosome with contig 5 in the BAC map (Figure 5B). Results with the CAP64 and Spi35markers also suggest that contig 17 represents part of one of the largest chromosomes in 1199;this conclusion is supported by the comparison of the conservation of synteny (Figure 3) becausecontig 17 of H99 shares two other markers with contig 1 of JEC21. We noted earlier that therDNA markers did not hybridize to any of the clones in the 1199 BAC library; the rDNA clusteris therefore not represented on the contig map. However, hybridization of the rDNA probes toelectrophoretically separated chromosomes did locate the sequences on the largest band (Figure6 A and C). Furthermore, the location of the rDNA region between contigs 8 and 16 in H99 issuggested by the comparisons of the shared markers; i.e., these contigs carry the markersSpi 16!H002K09.F and CPA] that flank the rDNA on contig 7 in JEC21 (Figure 3). For H99, theURA5 and HIS3 probes each hybridized to two contigs although clones from one of the twocontigs (19 for URA5 and 11 for HIS3) were more frequently detected. Also hybridizationexperiments to electrophoretically separated chromosomes (Figure 6 E and G) showed that HIS3hybridized to a single chromosome in each strain. These results suggest that cross-hybridizingsequences may be present for these markers.51A4,5,6(1.61 Mb)7(135Mb)8 (1.29 Mb)9(1.18 Mb)10(1.02Mb)11(889 Kb)12 (770 Kb)Spitzer andSpitzer, 1997Bands B-35012Figure 5: Relationship between electrophoretically separated chromosomes and the contigs ofthe JEC21 and H99 maps.A. Diagrammatic representation of the electrophoretically separated chromosomes from theserotype D strains B-3501 and NIH12 (Spitzer and Spitzer, 1997; Wickes et al., 1994). Thepanel on the right shows the contigs of JEC21 that hybridized to the same markers used bySpitzer and Spitzer (1997) and Wickes et al. (1994) to identify specific chromosomes orgroups of chromosomes.B. Locations of chromosome specific markers on mapped and unmapped clones and specificcontigs of the serotype A strain H99. The number of BAC clones that hybridized with eachmarker is also presented for both the mapped BAC clones and the additional clones(unmapped) that were present on the high-density filter.Wickes et aL,1994Bands NIHI21(3.87 Mb)2 (3.34 Mb)3 (2.50 Mb) 3}Spiker Mapped UnmappedBand HindUl Map Contig JEC21 BAC JEC21 BAC(B-3501) Marker (Clones on Contig) Contig Size (bp) Clones Clones TIGR CHR1 Spil6 7 (12) 1,484186 13 29 No Hit2 CAP64 1 (2) 1,748,127 2 0 2rDNA18S Not2 intergenic 7 (9) 1,484186 16 60 RepresentedNot2 rDNA26S 7(9) 1,484186 18 59 Represented2 Spi25 1 (4) 1,748127 5 21 No Hit2 Spi35 1 (7) 1,748,127 9 10 No Hit3 Mato (MF gene) 11 (1) 1,434,639 2 3 43 Make (STEI2) 11 (1) 1,434,639 1 0 43 Matu (STE2O) 11 (1) 1,434,639 1 2 43 SpiO6 11 (6) 1,434,639 8 18 43 SpiO7 11 (10) 1,434,639 11 27 No Hit3 Spi32 4 (5) 853,084 5 16 103 Spi55 11 (2) 1,434,639 3 19 44 URA5 15 (4) 407,895 4 25 75 PKA2 16(12) 975,364 14 33 96 Spi15 3 (2) 379,639 2 15 No Hit6 Spi29 4 (5) 853,084 6 31 107 HIS3 18 (4) 970,748 6 25/9 No Hit8 SpiOl MULTIPLE N/A 3 23 13BUnmapped-Hind Ill MapConfig MappedH9g H99BACMarker (Clones on Conhig) Conhig Size (bp) BAC Clones ClonesSpil6 16(12) 236,418 16 21CAP64 17(7) 1,192,998 5 11rDNA 18Sintergenic NOTREPRESENTED N/A 0 0rDNA26S NOTREPRESENTED N/A 0 0Spi25 3(11) 1,221,982 12 16Spi35 17(8) 1,192,998 10 6Matu (MF gene) 5(1) 1,356,533 1 1Mata(STE12) NONE 1,356,533 0 2Mata(STE2O) 5(1) 1,356,533 1 2SpiO6 5(11) 1,356,533 14 12SpiO7 5(8) 1,356,533 14 10Spi32 NONE N/A 0 0Spi55 5(14) 1,356,533 15 9URA5 19(12)20(3) 1,199,093;957,898 16 10PKA2 20(10) 957,898 13 15Spil5 17(10) 1,192,998 12 17Spi29 4(10) 746,210 12 8HIS3 11 (17)5(4) 689,677;1,356,533 25 12SpiOl MULTIPLE N/A 6 552A BFigure 6: Genomic Location of 26S rDNA and HIS3.For JEC2 1 the corresponding bands to reported serotype D karyotypes are related to theJEC21 karyotype in addition to chromosome size estimates from Wickes et al., 1994.Probe from 26S rDNA hybridized to a CHEF gel of chromosomes CHEF gel A. H99 B.JEC21 and Blot probed with 26S rDNA C. H99 D. JEC21. 26S rDNA is present on thelargest CHEF fragment in H99 and the second largest fragment in JEC2 1. CHEF gel E.H99 F. JEC21 and Blot probed with HIS3 G. H99 H. JEC21. HIS3 is present on the 7th(11th) CHEF fragment in H99 and the 5th (71h) fragment in JEC2 1.C D E F G H53SUMMARYThe BAC fingerprint maps described here for strains JEC21 and H99, along with thesequences of the ends of the mapped clones, provided a partial framework for the completion ofthe genomic sequences of these strains. The map with the minimum tiling set of BAC clones forthe genome and the end sequences of the BAC clones contributed to the genomic sequencingeffort for JEC21. The sequences for two serotype D strains JEC21 and B3501A are nowpublished in Science (Loftus et al., 2005). The genome for the serotype A strain, H99 has beencompleted at the Broad Institute of Harvard and MIT. The maps also provided the firstcomparison of the conservation of synteny between the genomes of C. neoformans strains fromthe A and D capsular serotypes that represent varieties grubii and neoformans, respectively.Furthermore, the maps provide the opportunity to use arrays of the minimum tiling sets of BACclones to make comparisons between genomes from different isolates from the same or differentvarieties. This approach has been used successfully to explore genome variability in theMycobacterium complex (Gordon et al., 1999). Five more mapping projects have been initiatedand successfully completed through collaborations of the Kronstad laboratory and the MSGSCusing similar techniques. The robust analyses of these first maps for serotypes A and D provideda strong platform to guide further efforts in the mapping of C. neoformans and C. gattii genomes.54CHAPTER THREE: Serial Analysis of Gene Expression (SAGE) of C. neoformans underIron Limited ConditionsINTRODUCTIONThe SAGE technique, originally developed by Velculescu et aL, 1995, provides asnapshot of the mRNA transcript abundance in a given cell state. The set of transcripts (knownas a transcriptome) in a SAGE library is unique to the mRNA in the cells isolated for that library.Data output from a SAGE library is in the form of 10 base pair tags, each of which is 3’ adjacentto the 3’ most NlaIII (CATG) site in a cDNA. The abundance of unique SAGE tags may becompared from cells grown in different conditions. The corresponding transcript for each uniquetag gives an indication of expression for that transcript in that cell state. It is this step ofidentifying a transcript and subsequent gene function from a SAGE tag that proves to be ratelimiting. The SAGE analysis has become more efficient and accurate with the completion ofsequencing, followed by assembly and annotation, for the JEC21 and B3501A genomes (Loftuset al., 2005). H99 genome sequence resources are available at Duke University and the BroadInstitute of Harvard and MIT, in addition to EST sequencing projects at the University ofOklahoma, Duke University and TIGR. At the onset of this project the first sequencing projectat Stanford University was in its infancy. SAGE data analysis at this time was slow, laboriousand often yielded limited results. The value of the genomic resources developed within the timeframe of this project have been clear in the efficiency of many analyses, both large scale (such asSAGE analysis) and on a gene by gene basis (Chapter Four). SAGE has been successfully usedfor a number of analyses in the Kronstad Laboratory and multiple libraries have been completed,sequenced, analyzed and published. For example, libraries were constructed to investigatethermotolerance in C. neoformans. JEC2 1 and H99 libraries were prepared from the mRNA ofcells grown in vitro at both 25°C and 37°C (Steen et al., 2002). Also the MicroSAGE protocolwas optimized for C. neoformans by Dr. Barbara Steen in order to construct SAGE libraries fromthe mRNA of C. neoformans cells that were isolated from the cerebral spinal fluid of rabbits. Amodified protocol was necessary due to the low number of cells isolated from this in vivo source.These libraries provided a view of in vivo C. neoformans transcription that revealed clues toimportant virulence genes (Steen et a!., 2003). Libraries were also constructed from the mRNAof cells grown in vitro on low or replete iron media for the serotype D strain, B3501A (Lian et55al., 2005). These libraries provided a view of differential gene transcription involved withcapsule synthesis and identified a set of iron regulated genes. It was in these libraries that theSIT] gene (characterized in Chapter Four) was initially identified. Additional libraries have beencompleted for cAMP mutants in the serotype A strain H99, for wild-type H99 under iron repleteconditions, for cells from in vivo sources such as lungs of mice and macrophages in culture.These data are currently being analyzed and will contribute, in conjunction with the SAGEanalysis detailed in this work, to large scale transcriptome analyses in C. neoformans. Multiplestudies have also been initiated to characterize genes identified in the SAGE analysis thus far andthis genetic follow-up work contributes biological relevance to these studies. Chapter Fourrepresents an example of this approach.At the time of map construction (described in Chapter Two), it was estimated that 8-10,000 genes existed in the 23 MB C. neoformans genome, by comparison to the fully sequencedS. cerevisiae genome which has 6,200 genes in a 12 MB genome (http://genomewww.stanford.edulSaccharomyces/; Velculescu et a!., 1997). Cryptococcus has a greater numberof introns per gene in comparison to Saccharomyces, and given an estimation for the averagelength of a fungal gene at 1000 base pairs, a smaller percentage of genes were predicted in C.neoformans, 35-43 % vs. 52% of the genome as coding sequence respectively. The completionof two serotype D genomes (Loftus et al., 2005) further refined these estimates to a genome sizeof approximately 20 MB and 6,500 intron-rich gene structures (revealing a smaller genome andfewer genes than initially predicted).The SAGE library constructed and analyzed in this work investigates the transcriptome ofthe serotype A strain H99 in low iron conditions. Low iron is an important nutritional signal forgene transcription in the fungus, resulting in the elaboration of the polysaccharide capsule, amajor virulence factor. In addition, the acquisition of iron, particularly in low iron conditions, isparamount for the survival of virtually all organisms. Pairwise comparisons were also included inthis analysis to compare the low iron transcriptome with those from cells grown in minimalmedia (YNB) at 25°C and 37°C (Steen et a!., 2002), in addition to a comparison with an in vivolibrary from rabbit CSF (Steen et a!., 2003). For simplicity in these comparisons, the librarieswill be referred to as 37°C low iron (cells grown at 37°C in low iron medium), 25°C (cells grownat 25°C in YNB), 37°C (cells grown at 37°C in YNB) and in vivo (cells isolated from rabbitCSF).56MATERIALS AND METHODSA. Determination of Growth Conditions.a. Fungal strains.The serotype A strains H99, H99Apkal and H99Apkrl were provided by Dr. J. Heitman(Duke University).b. Media and growth conditions.Low iron medium (LIM) was prepared as described with 20mM HEPES and 20 mMNaHCO3 (Vartivarian et al., 1993). The water used for LIM was treated with Chelex-100 resin(BioRad) to chelate iron. LIM+Fe was prepared by addition of ferric EDTA to 100 tM(FeEDTA;Sigma EDFS). Five mL of yeast extract peptone dextrose medium (YPD) wasinoculated with a single colony and grown overnight at 30°C in a gyratory shaker at 250 r.p.m.100 pL of culture was transferred to 5 mL of yeast nitrogen base (YNB) medium and grownovernight at 30°C. Cells were washed four times with sterile water and transferred to flaskscontaining 50 mL of LIM or LIM+Fe. The cultures were shaken at 250 r.p.m. at 30°C. Cells forSAGE library construction were harvested by centrifugation after 6 hours of growth in LIMflash frozen in an ethanol bath and lyophilized overnight at -20 °C.c. Capsule microscopy.For measuring the elaboration of capsule in low iron (LIM) and iron replete (IR) media, cellswere grown as detailed in the section on growth conditions except that cells were grown in LIMor JR and samples were obtained prior to flash freezing. Cells were stained on a microscope slidewith an equal volume of India ink (10 pL: 10 pL) covered with a glass slip and incubated atroom temperature for one hour. DIC microscopy was performed on a Zeiss Axioplan 2microscope. Cells were viewed at 1 000X in oil immersion.B. RNA Extraction.Lyophilized cell pellets were pulverized with glass beads for 10 minutes and the cellpowder was resuspended in 15 mL of Trizol extraction buffer (Invitrogen). RNA was isolatedaccording to the manufacturer’s recommendations with the additional step of LiCl precipitationat 4°C, following the standard ethanol precipitation step. Total RNA was dissolved in 1.0 mL of57lysis-binding buffer from a Dynabeads mRNA DIRECT kit (Dynal). Beads were prepared as perthe instructions of the manufacturer (Dynal), and mRNA was added to the beads. Poly(A)+RNA was isolated as described in the microSAGE method (Datson, et al., 1999 and version 1 .Oe; Construction of a Serial Analysis of Gene Expression (SAGE) Library.The SAGE library was constructed as per the microSAGE protocol (Datson, el al., 1999).SAGE primers were purchased from the Nucleic Acid and Protein Service Facility (NAPS) inThe Michael Smith Laboratories, University of British Columbia. A total of 300 PCR reactions(35 cycles) were performed to amp1if’ the ditags during the construction of the library. The firststep of SAGE library construction was the isolation of total RNA. This was achieved using theTrizol extraction method (described above). The RNA was checked for quality by running totalRNA on a 1% agarose gel and observing intact rRNA bands, the mRNA appears a smear(Appendix III;Figure lila). Following RNA isolation, mRNA is captured on oligo-dT beads andcDNA is synthesized. The cDNA is completely digested at the 3’ most NlaIII site (CATG) andonly the 3’ most portion of the cDNA remains attached to the beads. Linkers are then ligated atthe NlaIII site. The linkers contain the enzyme recognition site for the type II enzyme BsmJI andprimer sequences for the PCR amplification step. The cDNA is then released from the beadswith BsrnfI. This enzyme cuts downstream of its recognition site leaving a unique 10 bp fragmentfrom the transcripts adjacent to the NlaIII site. The fragments were blunt ended with Kienowenzyme and ligated to form ditags. These ditags represent unique tags from two unrelatedtranscripts. Note that sequence data was later checked for duplicate ditags that contain tags ofthe same two transcripts. A high number of duplicate ditags indicates an artifact of PCR inlibrary construction because the tags should independently sort at the ligation stage and shouldnot be present in the same ditag multiple times. Note that duplicate ditags were not abundant andtherefore not a concern in the data from the library constructed here. Following ditagconstruction, the tags were PCR amplified to produce a 102 bp band. The PCR conditions wereoptimized to identify the most robust number of PCR cycles and the optimal dilution of ditag.Reactions were run on a polyacrylamide gel for analysis (Appendix III;Figure Ilib). 35X cyclesof PCR and a 1/50 dilution of ditag was found to be optimal. 300 reactions were performed andpooled for the construction of the H99 low iron library. The bands were separated from residual58PCR products on a polyacrylamide gel (Appendix III;Figure Ilic). Bands were cut from the gel,pooled and purified. Linkers were then cut from the tags with NlaIlI. A 26 bp band containingonly the SAGE tags was then isolated by electrophoresis (Appendix III;Figure hId). Tags wereconcatemerized and run on a polyacrylamide gel (Appendix III;Figure ITIe). Various sizedsections of the concatemers were cut from the gel and ligated at the SphI site of the vectorpZERO for sequencing. One hundred colonies were screened by PCR to assess the insert sizeand frequency. The insert size range was 400-800 bp with an average of 600 bp and a 10%frequency of empty clones. Figures for all stages of SAGE library construction can be found inAppendix lilA i-vu.D. Sequencing and Data Processing.The SAGE library was sequenced, tags extracted and initial data processed by theSequencing and SAGE groups at The Michael Smith Genome Sciences Center (MSGSC). Thelibrary of clones was sequenced using BigDye primer sequencing and analysis on an ABIPRISM 3700 DNA analyzer. Sequencing chromatograms were processed using Phred software(Ewing and Green, 1998;Ewing et al., 1998) to achieve accurate base-calling. CROSS_MATCHsoftware was used to detect and remove vector sequence (Gordon et al., 1998). Tags 14 bp (10bp unique + 4-bp NlaIII site) were extracted from the vector-clipped sequence, and an overallquality score for each tag was derived based on the cumulative Phred score. Only tags with apredicted accuracy of 99% or greater were used in this study. Duplicate ditags and linkersequences were removed as described in the original SAGE protocol (Velculescu et al., 1995). Pvalues for comparison of tag abundance between libraries were determined using the methoddeveloped by Audic and Claverie, 1997. These statistically significant differences between tagabundance in different libraries were based on the probability that a difference in tag numberswas due to random fluctuation.E. Tag Identification.For preliminary assignments of tags to genes, EST databases and genomic databases were used.An EST database is available for the H99 strain at the University of Oklahoma’s AdvancedCenter for Genome Technology ( [funded under cooperativeagreement UOl Al 485 94-01]. When an EST was not available for a tag sequence, genomic59sequence was searched at the Duke University Center for Genome Technology; Duke UniversityMycology Research Unit (DUMRU) (!), or the Broad Institute ofHarvard and MIT (!annotationlfungi/cryptococcus_neoformans/). Tagsequence assignments were only reported if they could be unambiguously identified as matchingeither EST or genomic sequence. Gene assignments were recorded if they had significantsimilarity (<e5) to known sequences using the BLASTx (Basic Local Alignment Search Tool)algorithm against the non-redundant database at the National Center for BiotechnologyInformation ( Tentative gene assignments (including the specieswith the closest ortholog), expect values, percent similarity and positives were recorded in Excelspreadsheets and were assigned Gene Ontology terms (GO term) where possible(http://www.geneontology.orgl). Tags were only given a gene assignment in the analysis if theywere found at the 3’ most NlaIII site in the putative open reading frame or in a 3’ untranslatedregion. Detailed information about the tag annotation process can be found in Appendix IIIvii.Data from additional libraries used in pairwise analyses were provided by Dr. Barbara R. Steen(Steen et al., 2002 and 2003).60RESULTSA. Determination of Growth Conditions for SAGE Library ConstructionInitially, growth conditions were carefully optimized for three strains of C. neoformans, H99 wt,H99 iXpkal and H99 Apkrl. Different media were used to investigate phenotypes controlled bythe cAMP pathway as a prelude to SAGE analysis. These included medium that was low inglucose and contained the substrate for the production of melanin (Niger Seed and DOPA), orlow iron medium (LIM) to induce capsule synthesis and the expression of iron regulated genes(LIM and IR). Ultimately, LIM was chosen for a number of reasons. First, libraries wereunderway for the serotype D strain B3501A in low iron and iron replete conditions.Constructing the H99 libraries under the same conditions would provide for an opportunity in thefuture to compare annotated transcriptomes across the two serotypes. Secondly, iron is importantin capsule synthesis and growth in low iron a necessary adaptation for survival of the pathogen invivo. In this context, the Apka] mutant lacks capsule and the Apkr] mutant has enhanced capsuleformation. Finally, a sufficient number of cells could not be isolated from the melanin mediaeven in large volumes of medium and the melanized cells that were obtained were too rigid forefficient RNA extraction.A hallmark of C. neoformans under low iron conditions is the elaboration of the capsule.Therefore, cells were grown in low iron and iron replete media and assessed for the capsule byDIC microscopy at 0, 3 and 6 hours (Figure 7). Cells were grown in the same manner as used forthe libraries for the strain B3501A (Lian et al., 2005), and harvested at 6 hours when a markeddifference was noted in capsule size. The far right panel in Figure 7, shows the cells grown inLIM at 6 hours for H99 wild-type (A), H99 Apkal(B) and H99 iXpkrl (C); these photographsshow cells from the population used in final library construction. The capsule is apparent in theimages as a halo surrounding the cells as a result of exclusion of the India ink stain. For wild-type cells of H99, the capsule is induced in LIM and repressed under iron replete conditions(Figure 7A). As mentioned, the Apka] mutant lacks the catalytic subunit of the cAMPdependent protein kinase (PKA) and produces minimal capsule in low iron conditions (Figure7B); the Apkr] lacks the regulatory subunit of PKA, leading to a constitutively active signal fromthe catalytic subunits. This leads to a hyperencapsulated cell in both low iron and iron repleteconditions (Figure 7C).61There is still some uncertainty with regard to the most appropriate time for the isolationof cells to investigate the response to the low iron environment. The physical elaboration of thecapsule is likely the result of important transcriptional changes relating to the response to lowiron and the expression of capsule synthesis genes. For the experiments described here, thedecision was made to keep the conditions consistent with those used for the libraries fromB3501A to ensure continuity across the SAGE analysis and to allow for cross serotypecomparisons in the future.The wild-type library designated “3 7°C low iron” is presented in this work, the librariesfor the tpka1 and the Apkrl mutants of H99 were later completed by Dr. Guanggan Ru and willbe part of a larger future analysis that will include supporting functional studies of specificgenes.B. Summary of SAGE Library Construction (supporting data for library construction inAppendix III).The SAGE library was constructed according to the MicroSAGE protocol (Datson et al.,1999). The modified procedure was chosen because it was optimized for a small number of cells.LIM is a minimal medium and cells do not grow to high density. The MicroSAGE protocolincreases the efficiency of mRNA yield by immobilizing the poly-adenylated RNA on magneticbeads containing oligo-dT. The cDNA is synthesized directly on the beads and cDNA isrecaptured by placing the tubes on a magnet. In addition, MicroSAGE was already beingsuccessfully used in the laboratory prior to the construction of this library. After librarycompletion, colony PCR was performed on 100 clones to assess insert size and percent of emptyclones. An example of colony PCR results is found in Appendix III;Figure Ilif. The averageinsert size was 600 bp with a range of 400-800 bp and a 10% occurrence of empty clones.Approximately 3000 clones were sent to the MSGSC for sequencing (resulting in 19,278 highquality tags).620Hours3Hours6HoursC)C)A B CfronRepleteLowfronfronRepleteLowfronfronRepleteLowfronFigure7:Elaborationofthepolysaccharidecapsuleinlowironorironrepletemedium.CellsweregrowninElMorIRmediumandisolatedat0,3and6hours.CellswerestainedwithIndiainkandviewedbyDICmicroscopyunderoilemersionat1000X. A.H99, B. H99iXpka]andC.H99i.\pkrlC. SAGE Tag Annotation.The SAGE data were analyzed in collaboration with MSGSC. This analysis involved thefollowing: i) extraction of SAGE tags from raw sequence (MSGSC), ii) calculation of theabundance of each individual SAGE tag occurrence (MSGSC), iii) identification ofcorresponding transcript for 100 most abundant SAGE tags and 100 most differentiallytranscribed SAGE tags for each library (Kronstad Laboratory). A detailed explanation of tagprocessing can be found in Appendix IIIvii. The first step to identify the originating transcriptfrom a SAGE tag involved determining whether an EST was available at The University ofOklahoma ( EST database. An EST is ideal for SAGE tagidentification because it contains only the RNA coding sequence of the gene. However at thetime of these analyses many tags did not have corresponding ESTs. This may be in part a resultof different media conditions used to construct the EST libraries and differences in transcriptionthat would be present in a minimal media like LIM. Also, the number of ESTs available for theH99 strain is quite limited. More robust EST resources are now available at TIGR and willgreatly alleviate this difficulty in SAGE analysis for serotype D strains. If no EST was found,genomic databases for H99 were searched at Duke University (!)or the Broad Institute (http :// corresponding sequence plus about 2000 bp either side of the tag were retained for furtheranalysis. The positive contig and location was recorded in Excel workbooks. These sequenceswere then searched against the NCBI non redundant database using BLASTx analysis( to determine whether they contained putative genes.The gene sequences were then analyzed to identify the location of the SAGE tag and to ensurethat the SAGE tag was located in the 3’ region of the putative genes. Gene start sites, openreading frames, promoter regions and stop sites are identified within the putative gene in theVector NTI program. Analysis thus far has indicated that the SAGE tag often exists in the 3’untranslated region (UTR) of the genes. UTRs in C. neoformans have a length between 50-400bp in genes studied at present by the SAGE analysis in the Kronstad laboratory. In addition,genes appear to share a 3’ consensus sequence of 5’-TTTGTTTT-3’ which may represent apolyA addition sequence. A portion of the tag annotation found in the analysis presented herewere completed by Dr. Barbara R. Steen and included in a master Excel workbook database.Initially the tags from the low iron library were searched against this resource to identify64common tags that were already annotated (to avoid repetition of annotations). All additional tagswere annotated during this work and subsequently added to the database. This database hasproven to be a valuable resource for subsequent annotations of tags from new libraries.D. Analysis of the Low Iron Transcriptome.The top 100 most abundant tags in the low iron library were annotated and assigned GeneOntology (GO) terms where possible. Among the 100 most abundant, 49 were successfullyannotated with corresponding EST or genomic sequence, had a significant BLASTx hit in thenon-redundant database at NCBI (<e-5) and were successfully assigned associated GO terms. Sixtags (6%) had a corresponding genome sequence match, and had significant BLASTx hits, butwere most similar to a hypothetical protein. Another 29 tags (29%) had a corresponding genomesequence match but did not have significant BLASTx hits to known genes. The latter twocategories may still warrant further investigation as they may be novel genes important in ironregulation or acquisition. Nine (9%) of tags did not hit EST or genomic sequence. One (1%) taghit multiple sequences and therefore the correct transcript could not be unambiguously identified.Finally, six (6%) of the tags contained multiple adenosines suggesting that they were derivedfrom the polyA region of mRNA.Of the tags that were assigned GO terms, 20 genes were involved with protein biosynthesis.Transcripts for ribosomal proteins are generally found to be highly abundant in SAGE data andthe presence of these tags is an internal positive control in SAGE library analysis (Steen et al.,2002 and 2003; Lian et al., 2005). Two abundant tags matched genes that are involved in proteincatabolism including a carboxypeptidase and metalloprotease.Seven genes were identified in the most abundant tags that relate to the response to stress.This result is expected because low nutrient availability in LIM may provide a stressfulenvironment for the fungus. For example, four of the gene assignments were for heat shockproteins (two HSPJ2, HSP7O and HSP9O) and two were for genes for thioredoxin andglutathione peroxidase (HYR]), that have been studied in detail and implicated in virulence in C.neoformans (Missall et a!., 2005a-d; 2004). A glucose-1-dehydrogenase was also found to beabundant in low iron.The high affinity iron permease FTR] was highly abundant in the LIM library. The FTR]gene was identified in transcriptional studies of the B3 501 A strain (serotype D) and found to be65five-fold higher in LIM vs. IR media. These results were confirmed by Northern analysis (Lianet al., 2005).Four signal transduction genes were identified including a 14-3-3 protein and three GTPbinding proteins. Signal transduction is known to play an important role in the regulation ofmultiple virulence factors in C. neoformans (Casadevall and Perfect 1998; Kronstad et al., 1998;Wang and Heitman, 1999; Aispaugh et al., 2000; D’Souza and Heitman, 2001; Waugh et at.,2002; Kraus and Heitman, 2003; Pukkila-Worley and Aispaugh, 2004; Hicks et a!., 2004; Zhuand Williamson, 2004; Balm et at., 2004; Idnurm et a!., 2005; Gerik et at., 2005).Five genes were identified to be involved in carbohydrate metabolism and one in nucleic acidmetabolism. A chitin deacetylase, involved in cell wall synthesis, was also abundant. This genehas been studied in detail in our laboratory but no aberrant phenotypes were noted in a deletionmutant.Two genes involved with cell growth and maintenance were identified that encodedcyclophilin A and alpha-cop protein. There are two known cyclophilin A proteins in C.neoformans CPA] and CPA2 and this protein is important as the target of the antifungal drugcyclosporine A (CsA). Studies showed that Cpalp and Cpa2p play a role in cell growth, mating,virulence and CsA toxicity (Wang et at., 2001).Finally, three unclassified genes were identified: CIPC, an oxidoreductase and a gene for aphosphatidylglycerol/ phosphatidylinositol transfer protein. CIPC was the fourth most abundanttag in the low iron library and the second most abundant tag in the library constructed from cellsisolated from rabbit CSF (Steen et a!., 2004). CIPC has been named HOT] in C. neoformans andcharacterized in our laboratory; however, no significant phenotypes have been identified to date.66Table 5: Most abundant SAGE tags for H99 cells grown at 37°C in low iron.Sequence Low Iron BLAST HIT B-value % identitY I % similarity GO Termheat shock protein 12TATATGTGTA 280 Saccharornyces cereeisiae 1.OOE-06 45 61 Response to stress203 polyA tailheat shock proteIn 12TGACTGTTTA 123 Sacchsromyces paston3lus 9 e-04 55 73 Response to stressCipCATATGAAAGA 102 Emericella nidulans 1.OOE-08 52-70(3) 72-76 Unclassifiedno significant BLAST results in nrATTGAGATGG 72 database at NCBIno significant BLAST results in nrTAOTTGTGT 66 database at NCBISAGE tag does not hit C.neoformans cDNA or genomicTAAPATTGCT 66 sequenceGAAAAAAAAA 50 poiyA tailno significant BLAST results in nrTrGTAAkAA 48 database at NCB1ribosomal protein P2AGTACTCTTC 43 Podospora anserina 1.006-26 57 65 Protein biosynthesisGCAGATCTAT 38 ribosomal protein RPL39 ANNOTATED Protein biosynthesistranslation elongation factor ICGACAGACCG 38 (TEF1) ANNOTATED Protein biosynthesisno significant BLAST results In nrATGCACMTA 37 database at NCBIheat shock protein 90 homologTGTTATCGGT 37 Candida albicaris 2.006-32 60 70 Response to stressSAGE tag does not hit C.neoformans cDNA or genomicMGCCCCTTG 37 sequencehypothetical proteinAGGATGAGM 37 Agaricus bisporus 2.006-21 46 64SAGE tag hits more than oneACAATACCTA 36 sequence contigno significant BLAST results in nrGGAGACCAGG 36 database at NCBIhypothetical proteinTMiCGCATM 36 Schizosaccharomyces pombe 4.006-04 27 50405 ribosomal protein allGCATTCTTTA 35 Schizosaccharomyces pombe 8.OOE-54 72 80 Protein biosynthesisno significant BLAST results in nrCMCGATGAT 34 database at NCBISOs ribosomal protein L7a (L8)TATGATAGTG 34 Schizosaccharomyces pombe 3.OOE-82 66 79 Protein biosynthesisno significant BLAST results in nrTAGCGATCAC 34 database at NCBI40S nbosomal protein S6BAGMCTCAAA 34 Schizosaccharomyces pombe 4.006-75 65 74 Protein biosynthesisno significant BLAST results in nrATAAAAAAAA 33 database at NCBIADP,ATP carrier proteinTCTTTGATGT 31 Neixospora crassa 1.006-118 77-79 (2) 84-88 (2) Cellular respirationhypothetical proteinACWCGATA 31 Microbulbifer degradans 2-40 9,006-05 30 49heat shock protetn 70TATATATGCA 31 Cryptococcus curvatus 4.OOE-56 61 68 Response to stressSAGE tag does not hit C.neoformans cDNA or genomicA&AGAAGTT 29 sequenceno significant BLAST results in nrATATGACATA 29 database at NCBIno significant BLAST results in nrATATGTATCG 28 database at NCBITASAAAkkAA 28 polyA tailubtquinol-cytochrome-creductase SctiizosaccharomycesTGATGGAAGC 28 pombe 0.002-0.22 36-66 (3) 54-88 (3) Cellular respiration67Sequence Low Iron BLAST HIT E-value % Identity % similarity GO Term4Cc ribosomal protein s3ae (Si)TATACCTATG 28 Schizoseccharomyces pombe 2 006-85 68 81 Protein biosynthesis6-phosphogluconateTAGTGTCCCG 28 dehydrogenaee Aspergillus oryzse 0006.00 68 79 Csrbohydrste metsbolismno significant BLAST results In nrTAGCUAGGA 27 database at NCBITGAA.AAAAAA 26 polyA tail60s ribosomal protein L3OIL3OACTTTGSATCA 26 Schizossccheromyces pombe 4.006-23 79 92 Protein biosynthesisno significant BLAST results in nrMCTTGATTG 26 database at NCBIno significant BLAST results in nrCAGAGATGTG 26 database at NCBISAGE tag does not hit C.GAAAGCCMG 25 neoformans cDNA or genomicuracil phosphorlbosyltransferaeeCATATTGAGT 25 Nesrospora crsssa 2.006-37 65 84 Nucleotide metabolismchitin deacetylaseGACATTrTGA 25 Schizophylluw commune 5.00E-08 41 60 Cell wallthioredoxinTCAGAAGTTt3 24 Coprinus comatus 2.OOE-24 47 69 Response to stressribosomal protein LiGGCCGACCTG 24 Schizosecchsromyces pombe 5.006-56 52-58 (2) 61-64 (2) Protein biosynthesisoxidoreductaseATTGAATGTA 24 Clostridium perfringene 7.OOE-18 45 67 Unclassifiedribosomal protein L3ARCGGTTATG 24 Spodoptere frugiperda 1.006-44 65 75 Protein biosynthesis40S ribosomal protein S28TATGATTTTA 23 Neurospore cresse 1.006-59 80 85 Protein biosynthesiscytoplaemic ribosomal proteinMGGACTCTC 22 612 Podoepore eneerine 4006-43 60 67 Protein biosynthesisno significant BLAST results In nrCAGCAeJTTA 22 database at NCBI6ltS ribosomal protein L6ETCOTCTGAAG 21 Schizossccheromyces pombe 3006-98 61-66 (2) 78 Protein biosynthesisneoformans cDNA or genomicCATTCCTTCA 21 sequenceguanine nucleotide-bindingCAGATCTTCT 21 protein Neurospore crease 1006-140 72 85 Signal transductionno significant BLAST results In nrTCATAGGTAC 21 database at NCBI60S ribosomal protein L37TATTCATA4C 21 Seccheromyces cerevisiae 9.006-34 79 87 Protein biosynthesishypothetical proteinTATMGAGGT 21 Schizoseccheromyces pombe 4.006-52 41 61ADP,ATP carrier proteinTTCGGCAAGG 21 Neurospore crease 1.OOE-118 77-79)2) 84-88(2) Celluler respiration60s acidic ribosomal protein p1ACTACGTTCT 20 Schizoseccheromycee powbe 6.006-22 48 61 Protein biosynthesisno significant BLAST results in nrAAATGGTTTG 20 database at NCBIribosomal protein LISACACATTGATA 20 Xanthophyllomyces dendrorhous 2.006-80 72 81 Protein biosynthesisno significant BLAST results in nrTCTAAGTATA 20 database at NCBIno significant BLAST results in nrTGGATGGGCA 20 database at NCBIenolase 3, (bete, muscle)GTCGTAGAGT 20 Homo sapiens 4.OOE-48 69 81 Carbohydrate metabolismcarboxypeptidaseTCTCTTCCGT 19 Penicilliumjenthinellum 1.006-35 69 77 Protein cetebolism60S ribosomal protein L6TATTACAGCT 19 Arebidopsis theliene 5.006-08 41 60 Protein biosynthesisno significant BLAST results in nrTAGAATAGAG 19 database at NCBIalphe-amylaseGACATATGAA 18 Deberyomyces occidentelis 1.OOE-21 33 55 Cerbohydrete metebolism14-3-3 proteinMTTCGCTAT 18 Schizophyllum commune 1.OOE-123 96 98 Signet treneduction60S ribosomal protein L7TAACCCAeAT 18 Ceenorhebditie elegene 3.006-66 50 66 Protein biosynthesisglucose 1-dehydrogenaseCATTACTGCA 18 Bacillus subtilis 3.006-36 34-43)2) 51-62)2) Response to stress68Sequence Low Iron BLAST HIT E-value % Identity % similarity GO Termglutathione Peroxldase HYRICACTTGTTA 18 Saccharomyces cerevisiae 2.005-38 49 63 Response to stresscytochrome C oxldase subunit IVAGATGA5TGG IS Sct,izosaccharomyces pombe 8.OOE-16 48 58 Cellular respirationmetalloprotease (MEP)TGTATGGTCT 18 Asperglllus fumigatus 5.OOE-27 39 52 Protein catabolismSAGE tag does not hit C.neoformans cDNA or genomicTGA.AAATAAA 18 sequenceSAGE tag does not hit C.neolormans cONA or genomicCTT5AGTM 18 sequenceno significant BLAST results in nrCATATCATA 18 database at NCBICAkAAAAAAA 18 polyA tallribosomal protein L2DAMCCTTGCAT 17 Sctiizosaccharomyces pombe 2.OOE—47 56-68(3) 65-78(3) Protein biosynthesishypothetical protein (HomoGCATTGGCGT 17 sapIens) 25-29 28 48GTP-bindlng nuclear proteinCCMTGACGT 17 Spilp Sct,izosaccharomyces pombe 3.005-99 85 93 Signal transduction.hosphatidylglycerolIphosphatidylinositol transfer proteinCATCGTTACT 17 Aspergillusoryzae 1.006-19 39 65 UnclassifiedRhoZ GTP binding protein UstilagoMCCMTGTA 17 maydis 7.OOE-17 48 65 Signal transductionAAAAA°.A.AAG 17 polyA tailno significant BLAST results in nrCATTCGCATA 17 database at NCBIno significant BLAST results in nrGAATAGTGGG 16 database at NCBIaconitaseCCAATGATTA 16 Aspergillus terreus 0 48-66 66-74 Carbohydrate metabolismno significant BLAST results in nrAATTTATGAT 16 database at NCBIno significant BLAST results in nrAATTTTATCA 16 database at NCBISAGE tag does not hit C.neoformans cDNA or genomicGCCACAGCCA 16 sequencealpha-cop proteinAAATCTTATG 16 Sos primigenius 1.OOE-33 39 57 Vessicle transportATP citrate lyaseATTTATCAA.A 16 Schizosaccharomyces pombe 5.OOE-35 67 76 Carbohydrate metabolismSAGE tag does not hit C.neoformans cDNA or genomicGAGMAAAAA 16 sequenceno significant BLAST results In nrCATTTTATGT 16 database at NCBIno significant BLAST results In nrATGATTTGAA 16 database at NCBIplasma membrane iron permeaseCCCATCGTAT 16 Schizosaccharomyces pombe 7.OOE-14 43 57hypothetical proteinCTTGTAGAA 16 Neurospora crassa 5.OOE-06 56 70TAATGGTGC 16 nbosomal Protein RPL22 ANNOTATED Protein biosynthesisno significant BLAST results In nrCATTTCTTCA 16 database at NCBIno significant BLAST results in nrHATCATCCT 16 database at NCBIno signifIcant BLAST results in nrTAGCCGCGA’ 15 database at NCBI69E. Pairwise and Multiple Library Comparisons.a. Pairwise comparison of 37°C low iron vs. 25°C libraries.A pairwise comparison was completed for the low iron library (LIM), (designated 37°Clow iron) constructed here against the library from cells grown at 25°C in YNB (designated25°C) (Steen et al., 2002) (Table 6). Information for tag abundance classes can be found for eachlibrary included in the pairwise analysis show in Table 6. Total tags used in the analysis are19,278 for the 37°C low iron library and 30,468 for the 25°C library. At the 99.9% confidenceinterval, 0.82% (123 tags) were more abundant at 25°C vs. 37°C low iron and 1.28% (193 tags)were less abundant for a total of 2.09% (316 tags) differentially abundant. At the 99%confidence interval 2.15% (324 tags) are more abundant, 2.22% (335 tags) were less abundantfor a total of 4.37 % (659 tags). At the 95 % confidence interval 6.02% (908 tags) are moreabundant, 3.38% (510 tags) are less abundant for a total of 9.4 % (1,418 tags). Tags below the95% confidence interval are not considered to be significantly different. Therefore, 90.6 %(13,669 tags) are similarly expressed between these two libraries at alpha=0.05. A graphicalrepresentation of these data is found in Figure 8. The 100 most differentially abundant tags wereannotated for 37°C low iron vs. 25°C and assigned Gene Ontology (GO) terms where possible(Tables 7 and 8). All data was normalized to the library with the fewest tags. There were 32 tagsof the 100 more abundant (32%) and 31 tags of the 100 less abundant (31%) successfullyannotated with corresponding EST or genomic sequence, had a significant BLASTx hit in thenon-redundant database at NCBI (<e-5) and were successfully assigned associated GO terms.Note that differentially abundant tags in 37°C low iron vs. 25°C and 37°C low iron vs. 37°C werehighly similar, that is differentially abundant genes in 37°C low iron vs. YNB media regardlessof temperature had a similar profile. Therefore, genes of interest will be discussed only in thefollowing Results: section Eb. (the pairwise comparison of 37°C low iron vs. 37°C libraries) andthe Summary and Discussion.70Table 6: Comparison of SAGE data for strain H99 cells grown at (3 7°C low iron) vs. YNB(25°C) media.Abundance Classes for 25°C (Number of Tags)Interval 1. 2-4 5-9 10-99 100-999 1000+ TotalCount 4353 1976 644 492 26 0 7491Total 4353 5184 4078 12077 4776 0 30468Abundance Classes for 37°C low iron (Number of Tags)Interval 1 2-4 5-9 10-99 100-999 1000+ TotalCount 6602 2161 441 209 4 0 9417Total 6602 5424 2760 3784 708 0 19278Differentially Abundant (Number of Tags;% of Tags)999 % 99.0 % 95.0 %More Highly Abundant 25°C 123 0.82 % 324 2.15 % 908 6.02 °hLess Highly Abundant 2S°C 193 1.28 O/ 335 2.22 % 510 3.38 %Total 316 2.09% 659 4.37 °h 1418 9.40 °hSimilarly Abundant alpha=0.05 Tags PercentTotal 13669 90.60 %7120leOlTotal:15087Figure 8: Expression profiling comparing relative transcript levels of SAGE tagsfrom strain H99 cells grown in YNB (25°C) and at (37°C low iron) media.The trumpet shaped lines represent the separation of the data by confidence interval. Darkblue crosses inside the lines indicate tags that do not show a significant expressiondifference. The grey line represents the 95% confidence interval, the yellow line 99 % andthe black line 99.9%. The green dots indicate tags that show a difference that is significantwith >95% confidence. Tags in the top left quadrant of the graph represent tags moreabundant at 37°C in low iron and the bottom right quadrant represents tags more abundantat 25°C. The total includes the number of tags used in the analysis. The image wasproduced in Discovery SPACE Singleton tags were excluded.Crjptococcus (H99)5.01 e025.0I I I I99.9% .99%95% .Sij. UUIIi5.0U UU4. +4.+ 4.+ 4+ .++ ++ ++++ ++ + + ++ 4+ ++..+--+ +++-+-+++,-++ + + +++++++4++ + + ++++4++++4.++++ + + + +4.4.4.U— — —— •U+ + ++++..u_. —I IU.5.0 leOl 5.0 1e02 5.025oC72Table 7: Differentially abundant SAGE tags from H99 cells grown at (37°C low iron) vs. YNB(25°C) media. * data sorted by ascending p-valueTA1’ATGTGTA 31T4.&8ATTI3CT 0TTAIOTIUTGT 5ATATGWGA 23TTGTAMA&5 0ARGCCCCTTG 0TGACTOTTTA 57ACMTACCTA 0I3I3AGACCAI3G 0ACARCTCAOA 2ATTGAGATGG 27GMAGCCA6G 0TATATATGCA 3AGTACTCTTC itACMACGATA 4ATIG6ATGTA 1TCTPAGTATA 0TGTTATCGGT 9CTTTGGATCA 3AT6A5ASAnA 7GACATATGM 0GMAOAAW 21CCASTGATTA 5QCCACAGCCA 0ATGAT1TCA6 STCGI3GGCTGC 0TAOCAATGTA 0GACATTCTGA 0MCTIi3ATTG 5MASATCATC 0TATATGCGTA 0GACATTTTGA 8CA1TTTATi3T 1ATITATCW 1MTTTATGAT IAi3GATGAGM 15CTTASAGTM 2Gi3TCAGTCISA 0TATATGTGCA 0TGACTi3TCTA SGCATICT1TA ISAI3TTTCTTGT 0ATIATMCGA 0ATATTCATAA 0TATACGTGTA ICATATI3TGTA 1AOTTACTGQT 0CATTATATAT 0TAOITCGTGT S19.5 285 14300 56 unique to LIM3.2 66 25.914,5 102 7.00.0 48 uoiqun to LIM00 37 unique to LIM36.1 t23 3.40.0 36 uoique to LIM5.0 38 unique to LIM1.3 34 26.917.1 72 4.20.0 25 unique to LIM1.9 31 1637.0 43 6.22.5 31 12206 24 37900 20 uniquo In LIM5.7 37 6.51,0 26 13.744 33 7.50.0 18 uniqon In LIM13.3 50 3.00.0 18 unique to LIM00 16 unique to LIM0.0 16 unique In LIM0.0 15 unique to LIM0.0 15 unique to LIM0.0 15 unique to LIM3.2 26 8.20.0 14 uniqun to tIM0.0 14 unique to LIM3.8 25 6.80.9 16 25.30.9 16 25.30.6 16 20.395 37 391.3 19 1420.0 13 unique In LIM0.0 13 unique In LIM05 13 unique In tiM95 35 3.70.0 12 unique to LIM0.0 12 uniqun to LIM0.0 12 unlqun to LIM0.5 14 22.10.6 14 22.10.0 Ii unique In LIM0.0 Ii uniqun In LIM0.0 II unique In LIMtag hOn 2 nnnttgs at DUKEWrung orientation nt SAGEtagto BLAST 50 4.OOE-1361 5657 6530 4044 5300 7079 5233 55Snspunse lu utrounPruinir BiunynthnuisCarbotiydneln mulabolismRnnyunun to utrnnuProtein BiuuynlhosiuCarbohydrate mntubuliumCarbohydrate metabolism25nCSnquenue 25 u C Normalized Low bun Fntd Difference BLAST HIT , E-nalue to identity to similarity GO Term45 615270 131 72-7655 7355 74Renponne tu s0uusUnulunnifiedRnnpnnue In nonssProteir Bionynthosistitan Snook Protets 15Saruhtnumyuns unleviniunSAGE tag dnnn not hit C.snntnnmans nDNA or snnsminno signiflnant BLAST results inor database at NCBIC1pCEmnhnnlln nidnlnnnso signiflnant BLAST rnnuttn issr databasn at NCBtSAGE tag dons not hit C.seufsrznans nDNA or gennmisHeat Shook Protein 12Saunharumynnu pastodanunSAGE tag hits morn than onennqnnsoe nnntlgSAGE tag does nno hit C.nsmtnznsans nDNA or genemin405 rihosnmal protnis 0GBSubizusauubarornyuns pnrnbnno slgsifioant BLAST results innr datahann at NCBISAGE tag dunn not hit C.nsolnrmans nGNA on genominheat shook peooein 70Cryptunnonus uurvatundhosurrral protnis P2Pnduspum annnnirahyputlretinal pmteisMiurobulbifnn dnqraduns 2-40aloshol dehydrugenaseBanillun nnblilisno significant BLAST nasulls in nrdelebeun at 0091hnat shnuh protein 90 hnmnlngCandida albinansBits rihnssmnt protein L3OIL3OASuhizosauoharumyuns pombeon signiguant BLAST results inor database at NCBIalpha-amylaneDnhnryomyuns nuuidnnlslispolyA tailaunnitasa proteinBeninroidas fragileSAGE tag dons not hit C.sooturmans oDNA or ganom’mso siqnit’mant BLAST results isnr database at NCB1SAGE tag dons not hit C.seotnrsnans nDNA or sennm’mPutaoine S.AIdP-aotiuatnd,garrsna subunit familySubizunauubnrnmyuns pnmbeSAGE lag doss not hO C.nentnrmans nDNA or gennm’mns signiitsant BLAST resells insr database at NCBITag hits In wrung nrinetatinn onBLAST hitSAGE tag does nnt hit C.neotnrmans oDNA or gesominshitin dsanstylaseSuhizuphyllum nnmmnnnno signiBnast BLAST rnsnlts innr database at NCBIAlP nitrate lyasaOuhiznsannhannmynns pnmbnen signifloant BLAST rnsutts innr database nt NCBIhypothntinat proteinAqaninus bispomsSAGE tag doss not hit C.neotormans nDNA nr 5550mmhypothetical pmtsmnPlasmudium 50018 ynnhSAGE tag doss not hit C.seofurreans oDNA on gnnnmloSAGE tag dons not hit C.seotnrrsans nDNA on gesomlodOs rihosomal protein sli5. ynmbsSAGE tag does sat hit C.nantarmans nDNA on gennsdoprotnasnme nsmponnnt P052hnmnioqSuhinusauuharnmyues pumbeTag hits is mrnsg nrlentatins onBLAST hitBnta.tnbntin nntantor DHomo salnnnsnn signifloant BLAST resuBs isnr database at NCBIno slgniooant BLAST result isnr databasn at NCBI1.505-061.005-08S n.044 050-757.WE-264.565-501.OOS-269.005-OS4,005.222.WE-324555.331.OOE-213.OOE-061.560-375.000-000. OOE-3S2.OOE-211.060-170.WE-042. 000-275.000-4456 7534 58 Orqnnl Iransdununr41 80 Cell wall67 76 Carbnhydraln metebnlism40 6412-77121 50-0912172 85 Protein Blnsynsresis50-95121 74-95(2) Protein Catebulium32 47 Cell gnnoth andlor maintsnanuo33 49732SoC rSequence 25cc Normalland Low Iron Fold Difference [ BLAST HIT B-caine % density % similarity GO TermGTATGA°AGA 0CATCMCCTG 0MAuAAAeAG 3CAGTTATTM 2TrGTrAGATr3 2MCTGTTGTA 1CATATOATAA 4CTTGTAGA°A 3CTATGCACAA 0TTATGGGTAO 0TATGCA°A°A 0TGGPA°A°A0 0CT1’TGAGTCC 0TATATGTGTG 0A°ACTTTCM 0CTTCCTGTM STTTCGTCMT 0TATMOTGTA 0TATTAOAGCT 9TATGATTTTA 9TATATAOTTG ITATOTMOCA 1MCGMTGTA 4TATTCATMC 7TGP.OAOAkOA IiCAOTTACAGT 3ATOTATACCC 3CA&k°ACATC 3GTTTTATGGA 3T,OAIOAOAOA 13TOAGATTGCT 0TGCOA°A°A0 0GMTMTAGA 0TDTATGMTT SMTGTGCCAT 0T°ATCATAGC 0CATTTGCATA 0TTTTMGCTA 0TATTCTACAC 0TTGTATGATA SATATCCTOCA SGCTGTAGTAT SCAGAACGACG STAGGTTI3TCT 0TTCGAGAGCT STACATATATT 2ATATT7CTAT ICATATAGA°A IA°ATCTTATG 4CACATIGATA 7TGA°.°ATASA 60.0 II unique to LISA0.0 11 unique to LISA1.9 17 So1.3 15 11.91.3 15 11.90.9 13 2002.5 18 7,11.9 16 9.40.0 10 unique to LIM0.0 10 unique to LIM0.0 10 unique to LIM0.0 10 unique to LIM0.0 10 unique to LIM0.0 10 unique to L1M0.0 IS unique to L1M0.0 10 unique to L1M0.0 10 unique to LIM0.0 10 unique to LISA32 19 9.80.1 23 4.50.9 12 19.009 12 1902.5 17 6.74.4 21 4.77.0 26 3.71.9 19 7.91.9 15 - 7.91.9 15 7.91.9 15 7.98.2 29 3.40.0 9 unique to LIM5.0 9 unique to L1M0.0 9 unique to LIM0.0 9 unique to LIM0.0 9 unique io LIM0.0 9 unique to LIM0.0 9 unique to LIM0.0 9 unique to LIM0.0 9 oniqoo to LIM0.0 9 unique to LIM0.0 9 unique to LIM0.0 9 unique to LIM0.0 9 unique to LIM0.0 9 oniqoe to LIM0.0 9 unique to LiM1.3 13 10.30.6 ii 17.45.8 ii 1742.9 16 834.4 20 4.03.8 18 4.7SAGE lag dccc not hit C.seoformaos oDNA or ganomioSAGE tag dccc not hit C.oeofermans oDNA or gnnomioeoioA tailno significant BLAST results inor database at NCBiNo eignitioant BLASTu hits neartagno significant BLAST rasoito inor datahasa at NCBIno signifinani BLAST resaits iner database at NCBIno significant BLAST resaito innr databana at NCBItag bite 2 oontigs at DUKEBod2OpOaooharonnyoeo oeneolnlanOS ribosomai protein S23Nooroupora onaooaSAGE log dens not hit C.nnofnrmans oSNA or gnnomioaiginate iyasnHatotin dinous htnnaiGenemio Hit too oiona to S Endof ContigSAGE tag does not hil C.naofnrnnans oSNA or gnnomioSAGE tag dens not hit C.nnafamqans oSNA or gonomloSAGE tag hits mam than onesaqoenne oentigSAGE log dens not hit C.naoformans oSNA or genemio905 nibonomai protein LBAnabidopsin thatanadOS elbonomai protein S25Naunonpota or0000no nigoifroant BLAST results inor database at NCBIooosneond hypothntinai protein.yeastCandida oibioaonBho2 GTP binding pmlnlnUntilaqo may40fibS nibonomai proteIn L37aSaooharornyoan oereoisiaeSAGE lag dens not hit C.neofarmans oSNA or gnnomiohypothetical proteinNeur0050na onaooaTranslation Eiongatias Faotor3nytonhroma ci pr000rnorNeuronyona oraooaHeal shoob prolsis 60Oaooharowyosn oareoloiaapeivA tallLong-ohain-taBy-aoid-CoArgaso-rho proteinArobidofuin thalianeSAGE tag does not hit C.nnolnrmass oDNA or gasomioso oignifinant BLAST reeoits Inor database al NCBIno nigni9naot BLAST results inno database at NCB1Wrong odsntatien of SAGEtagto BLAST 55hypothotloal iorntuisNeuronpona ora000no significant BLAST reunil inor databasa at NCBIWroog onlonlatlon of SAGEtagto BLAST hfipradiotad proteinNeuronpora onannaWrong annotation at SAGEtaglo BLAST h9SAGE tag does not hit C.naotormann oDNA ergonomicno signifroant BLAST result loor database at NCBiSAGE fag does nol hit C.naoformann oDNA or ganomloSAGE tag does net hit C.oaafoomaoo nSNA or genomlodihydrotiaoonoi 4-eaduotasaS0000anomyoen onrnoioiaeno nignifloasl BLAST nosuito inor database at NCBInobaryolin translation Initiationfaotor 4A2Mon monouiuoso significant BLAST mnoits Inor database al NCBIaipha.oop proloinBos yrimiqaniosribonomal prolnin L13AXanthophytiomyoas dnndroAoooSAGE lag does sot hit C.aeetormaos oDNA or ganemlo7.OOE-tt3.009-452.0GB-IS5.000-OSlOSE-OSS.OOE-097.OOE-179,00E-341.000-toANNOTATED3.000-554.000-569.000-290 000-W0.OOE-129. 000-05I.OOE-087.000-231.000-332.000-8044 57 Call growih and/on maiotonanoe59 61 Protein Biosyothesiu30-40 45-93 Carbohydrate mntabolism41 60 Protein Biosynthesis80 85 Froteio Biooyothaoio27-29 41-4648 60 Siqoal inaosdooSon79 87 Pnoteio Siooynihaols26 43Pnoieio Siooynthaols04 67 cellular neopiraiioo99 54 Oe000noetouSoos34 49 Lioid and Sterol Meiaboi’om31-56 43-7545 077 731 53 ResE0000 lo o0ooc77 90 Protoin Biooonihnsin39 57 V000iola Oansyod72 51 Pnoiain Biosyoihaoio74Table 8: Differentially abundant SAGE tags from H99 cells grown in ‘[NB (25°C) vs. (37°Clow iron). * data sorted by ascending p-value250 CSequence 250 C NormalIzed Low Iron Fold Difference BLAST HIT E-nalue % identity % similarito DO Term1]CAGCAGGC 490 310.0 II 28,2CTCAGCGATG 384 243.0 9 27.0CATTCGCATA 342 210.4 17 12.7GCCANCGCCG 213 134.8 I 134.0GCTCTCCAGG 189 110.8 I 119.6CATCTGTTCC 187 118.3 3 39.4ATAAGC1TTC 162 102.9 6 17.1GTTICCGCTG ISO 94.6 5 19.0TCTDGTCGAG 131 82.8 2 41.4CGCGDMAGG 167 109.7 9 11.7GTGGACACGA 132 83.0 3 27.0AGCGAGCACT 124 78.9 2 39.2ATATGACATA 232 146.8 29 5.1GTATTGACCC 119 79.3 2 37.6GCTGCCTACA 96 62.0 0 ooiqce to 25ATGATCGGGC 119 72.0 3 24.3ACGDTGGCM 95 60.1 0 unique to 29GGTTACGCCQ 93 58.8 0 unique to 28TAGCCGCGAN 162 102.5 15 6.8CGACAGACCG 239 148.7 36 3.9GCGTTCTCGG 87 85.0 0 uoiqueto2sCCGCGACCGT 102 64.9 4 16.1MTGMTCTT 132 83.9 11 7.6AaAzACGCGT 109 69.0 6 11.5OTCGGTGGTA 149 86.6 14 6.3ACTCAGGTTG 91 07.6 3 19.2ATGCATTTCD 83 92.5 2 28.3ACCGTCGITG 75 47.5 I 47,5GCTCGCOACG 77 48.7 2 24.4GGTATCCTCG 62 39.2 0 unique 1020ACGOCCGTTA 61 38.6 0 ueiquo to 20AGCOCTOCTO 61 30.6 0 uoiqoo to 20CATCACTCTT 107 07.7 11 6.2CACGTTCACG 64 40.5 1 40.5TTCGGCPAGG 137 86.7 21 4.1MCOTCTGCC 80 50.8 5 10.1CCTCANCGGC 54 34.2 0 unique 1025CGTGTCANGC 54 34.2 0 unique to 25TCTTTCCOAG 65 41.1 2 20.6AT0000TCCC 53 33.0 0 unique to 25GCTGGTTTGA 93 33.5 0 unique to 25TGGTGGGMA 53 33.5 0 unique to 28ANGCCCGfl’O ‘ 59 37.3 1 37.3OCTTTTGCCC 59 37.3 1 37.3CATCACGCTT 62 39.2 2 19.6GTTGGCMCG 50 31.6 0 unique to 25CAGACGTANC 49 31.0 0 unique to 25CCGAGACMC 60 389 2 190ACACGTCTOO 92 50.2 10 5.8CACCTCANOC 58 36.7 2 18.3potatine motel traeaporterSchizosaooharomyoes pombe 6.000-52 36 50oytokiee ieduolog-glyooprsteieFilobaoidiella reotormano ANNOTATBDno significant BLAST meulte leerdatabase at NCBICyolophille A ANNOTATEDSAGE lag doee not hit C.eeotnrmans oDNA or genowlono slgnllloaet BLAST reeslle leerdatabase at NCBImaeeltot-1-phusphatedehydrogesase ANNOTATEDso slgsifloaet BLAST resells leerdatabase at NCBIHistoee 54 proteleMoo wuoculos 9,008-30 75 80so sigeifioaet BLAST resells leerdatabase at NCBIeuolaeslde dlphosphate-eugarhydrolase Snhizosaooharomyoes 7.008-26 35 Stso elgnifloaet BLAST beetle ieeedatabase at NCBIso slgeifloant BLAST results in erdalabaee at NCBIphosphebetolaseThermosyooohoyooyos elortatos 2.000-32 46 615TP eyethase ohgomyole seeeltloityooeterral proteinN. onassa 6.00849 43 63SAGE tag does not hit C.eeeformaes eDNA or genomleen sigeifioaet BLAST results In erdatabase at NCBImitnohondrial malatedehydrosesase 0. 3 OOE-37 65 79no sigelfloent BLAST results leerdatabase at NCBITranslation eleogatlee factor I(TEPI) ANNOTATEDtranealdolaseSoh’oosaooharomyces rombe t.009-lt4 68 78hypothesoat proteinNeurospona crassa 4.008-09 48 74no sigeifioaet BLAST results in ordalabese at NCBImyo-ieoe9ol i-phosphate synthaseNeorospona onassa 2.008-67 54 70Ft ATFase beta subunslyoyyeromnoes taotis 3,00B-03 80 91Nuotose biphesphale aldolasePorecoooidioides brasitiensis 1.008-40 70 82no significant BLAST results in ordutahase at NCBItñosephosphate isomeraseAsporgillos oryzee 9.008-50 52 72Ribesomal Protein LtTenopus laenis 5.OOB-102 72 8203 smell nuoleolarriboeeolaoprutein prolate IMP3S oeran’s’ao 9 008-50 59 74SAGE tag hits more thae oeasequeeoeooetigHlstsse HI; HhotpSaooharomynes conenistee 9.008-05 45 61pyrenata deoarbooylaseCandida glabrata 5.008-26 52 66thioredoole perooldaseS. pombe 2.008-63 BI 76ADPATP oaerlar proteinNosrospora onassa l.OOB-llo 77-79(21 84-88(21probable membrane proteinYLRIS2o 3 OOE-04 35 51RIRBNcDNA -Moe masoolos 5.008-42 62 75no slgoifioaet BLAST results In ccdatabase at NCBIglyoeraldehyde3-phesphatedehydrogeoase Fitobasldiolta 4.008-63 96 100ATP synthase gamma ohaleSnhiz0500ohanomyoeo pombe t.OOE-66 47 70glutamate eoaloaoetatetraesamieaee Dante rndo 8.008-98 56 72eterol-C5-daeaturaseMon mosoolso 7.000-17 62 73no slgoifloaet BLAST resells leerdatabase at NCBIhypothetloal protele -Sohtaosaoohanomycos pombe 8.OOE-06 36 60seed maturation protele PM27Gtycioe mao 2.000-07 29 46related to Ypt-ieteraotleg proteleTIP2 Nesrospora orassa 9.OOE-30 44 58related to aide-hole radestase TPRINeurosBsra orasea 9.OOE-13 38 59UnhoeweStreot000500s a9alaoriao 2.OOE-07 42 57ATP syethase G chaIn,mitoohoedrial Sacyharonnycos 7.OOE-05 31 50SAGB tag does not hit C.oeotormaes oDNA or genomloTranspodUnolassifledSignal TransductionCarbohydrate metabolismChromosome organ’oohon andbiosynthesisNucleobase MetabolismCarbohydnato motaboliomCanbohydrato metabolismFnotein biosynthesisCarbohydrate metabol’smCarbohydrate metabolismCellolar RespirationCarbohydrate metabolismCarbohydrate metabolismProtein biosynthesisCell qromlh and/or mainleoaoooChromosome organization andbiosynthesisCarbohydrate metabolismRespoesa to stressCellolar RespirationTransportUnyl055ibedCarbohydrate metabolismCarbohydrate metabolismAmino acid metabolismLipid and storol metabolismUnclassifiedCell growth and/or maintenancoCarhohydnata metabolismCellulan Respiration7525 a CSequence 25cc Normalized Law Iron Fcld Difference BLAST NIT 0-cube % identity % nimitarity GO TermTAGCGATCAC 159 100.6 34 3.0CCTGATCGCG 44 27.6 0 unique Ic 20OCOOACMCT 50 31.6 1 31.6TCCTGGCANG 43 27.2 0 unique to 26CACCAGGCAT 42 26.6 0 oniqun to 25TCGOTCGTGT 42 26.6 0 unique to 25AOGCTTOGAQ 42 26.6 0 oclque to 25TGC4ANCGCG 73 46.2 7 6.6GGCCTCGGTT 57 36.1 3 12.0ACCTTGA000 47 29.7 1 29.7TCTGCATCAT 47 29.7 1 29.7TCTGTCGAGG 52 32.9 2 16.5TCAGANGTTG 123 77.6 24 3.2ANGCCTGACG 67 42.4 6 7.1OCC4ACTCTC 55 34.8 3 11,6TA’rfcCGGTC 59 37.3 4 9.3TCCACCAYI’A 39 24.7 5 unique to 25ANGCCCOACT 39 24.7 0 uniqun to 29ATCGGTACCC 58 36.7 4 9.2CGAGOTATCA 38 24.0 0 unique to 25OAGMGCGTG 37 23.4 0 unique to 25AGCANOGAGG 64 40.0 6 6.7UCTCCTCTTA 73 46.2 9 5.1CAGMCCCCG 58 36.7 S 7.3ASGGGTOOTG 69 437 9 4.9GAGGAGGAGG 47 29,7 3 9.9OOCCcAGACA 31 19.5 0 unique tn 25CCCcGTcAcT 31 19.6 0 unique tn 29CCTCGTATCG 30 19.0 0 unique to 25ANGCGAflTt’ 71 44.9 11 4.1TCGTTATCTI’ 28 18.3 0 uniqun to 29ACGGCCAuAC 29 18.3 0 uniqun to 29GACGACTCTA 47 29.7 4 7,4CCACCARTGC 35 22.1 1 22.1O4ATAGT000 83 52.5 18 3.3ATGCACCCAT 28 17.7 0 unique to 25TCCGACCACT 28 17.7 0 unique to 25CCOc1TITGC 28 17.7 0 unique to 25CAATTCG 28 17.7 5 unique to 29GMGTAGAuA 42 266 3 99MGMGACCG 27 17.1 0 unique tn 25GTCCOANGcA 32 20.2 1 20.2GCCATcTTCA 44 27.8 4 7.0GACTCOACGA 26 16.5 0 unique to 25ANGGAGATTC 26 16.5 0 uniqun to 25TGC1’TCTGTG 26 165 0 unique ro 25CGANGACTCA 26 16.5 5 unique to 25GARTGGMTG 31 19.6 1 19.6OATGGCAGUG 31 19.6 I 19.6OTCAAGANGc 25 15.8 0 0010cc to 25nc siqsifioant BLAST results in nrdatabase at NcBIunnamed ponteis productPodospora rnnndnn 5.000-32 30 54sccteostde diphcsphate binasnEmndcnlla nidulacs 2.000-46 57 77profilinMalon 0 domeslioa 0.OOE-22 42 60PSGoscypium hiroutom 1.000-20 44 69predicted proteinNaurosporu oranno 9.008-00 29 38nn significant BLAST rnsottq in nrdatabase at NCBtpnrooisnmat membrane proteinPmp2Op 1.000-21 40 52no significant BLAST results in srdatabase at NCBtSAGE tag dnns not hit C.ncntormans cGNA or gnncmiohypotbntioal proteinBorkholdnria tongorum 3.000-06 31 46dbcsomal prcteis S12Branohiostoma bobbed 2.050-43 56 81ThiorndooisCopdnun ucmatus 2.OOE-24 47 69nc signifinest BLAST results In nrdatabase at NcBIRNA-bindtng protein AoRNBPAmbystoma meoioanum 3.000-14 60 78arqlnincosocinatn nynthasnStuohatomyonc unrevisirn 3.SOE-04 g 83mangannsn sspnrooide dismotasePhanoruchanto chrynonyorium 4.SOB-73 08 76Lg-Hspl2pSacoharomycen pastorianon 0.001 55 73phosphate transport proteinS. cereolsian 4.OOE-97 00 71NADPH-dopendnnt atdnhydnredoutase Sporidiobotos oolmonioolor 8.000-60 40 01Ribucomal Protein RPL2IAS. onrnuisiao 4.000-52 54 76Acyl-CnA-binding proteinChactophraotos villosus 2,508-10 47 65ATP synthasn - alpha chainN. orassa 1.000-66 77 84hypothntlcat pratninNeorospora orossa 2.000-58 76 86ribosumat protein L24Etuo’unromuons lactic 2.OOB-34 54 07SAGE tag hits more than onesngonnon contigno significant BLAST resolts in nrdatabase at NCBIWrong onientntlun SAGEtag tnBLAST hithypothetical prctnlsNnurospora upacca 1.OOE-13 47 03MIS rlbcsomal protein 926Soh’oophylluw commune 5.000-20 56-79131 66-97131nc significant BLAST rnnolts io nrdatabase at NCBI6-phcsphcqlocooatndehydrcgecasoAspnrqilloo oryzan 0 g 79hypothnhcal protninSch’oosacoharomyoos pombe g.OOE-21 47 81brain prntnin (3F505)Caerortrabdiss etogacs 0008-14 38 47nc significant BLAST rnsclts is nrdatabase at NcBISOhAAcpnrqillos nidolans 6.000-23 45 63nc signifloast BLAST results is nrdatabose at NCBInc signlgoost BLAST results in firdatabase at NcBIPotyohiqoiin ANNOTATEDmitnchcnddat ATPase alpha-ssbcnitN. cracsa 1.000-56 77 84SAGE tag dons not hit C.nnntcrmanc cGNA cr genomlocpmbtcsis-rntetnd proteinLaccoria bioolor 3.005-56 35 40pro-mRNA splicing tautorS. pombn 7.008-52 45 60hppcthntical proteinNnorospora crassa 9.000-30 44 50utathris tight chainSch’oosacoharomucec pombn 0.001 32 49SAGS tag hits morn than cnnseqonncnocntigDNA Unwinding Factor DUPO7Xnnopos landis 3.000-04 20 45gtyccgen branching enzymeAcpnrqilbon orqzae 0.OOE-42 44 56SAGS tag dcns not hit C.nnotcrmans cSNA cr gnnnmtcpotatioe protein*--,----. 4.000-28 34 49Nuolnobace Untnbutsmcoil growth and/or maintenanceUnotassiledResponse to stressProtein biocunthesisResponse to chessCell growth andtor waietccaccoAmino auid metabolismResponse to ctressResponse to strocsTraespodCarbuhydratn metabolismProtcin biosonthecisLipid and sterol metabolismCnlbolan RespirationProtein hiosyethoslsProtein bioopnthesisCarhohcdmte motabulicmUnclassifiedUculassiradprotein uatabollnmCcttutrr RespirstinoUeolassiledCr11 gnowth and/or maintnnnnueTmacspooDNA reptuationCarbohydrate metabolism76b. Pairwise comparison of 37°C low iron vs. 37°C libraries.A pairwise comparison was completed for the 37°C low iron library against the libraryfrom cells grown in YNB (3 7°C) (Table 9). Information for tag abundance classes can be foundfor each library included in the pairwise analysis (Table 9). Total tags used in the analysis are19,278 for 37°C low iron and 38,988 for 37°C. At the 99.9% confidence interval 2.3 1% (370tags) were more abundant in the 37°C vs. 37°C low iron and 1.81% (290 tags) were lessabundant for a total of 4.13% (660 tags) differentially abundant. At the 99% confidence interval4.78% (765 tags) are more abundant, 2.72% (435 tags) were less abundant for a total of 7.50 %(1200 tags). At the 95 % confidence interval 7.93% (1269 tags) are more abundant, 4.33% (693tags) are less abundant for a total of 12.27 % (1962 tags). Interestingly, this library from cellsgrown at 37°C is less similar to the low iron library also grown at 37°C than the cells from 25°C.Tags below the 95% confidence interval are not considered to be significantly different.Therefore, 87.73 % (14,031 tags) are similarly expressed between these two libraries atalpha=0.05. A graphical representation of these data is found in Figure 9. The 100 mostdifferentially abundant (more and less) tags were annotated for 37°C low iron vs. 37°C andassigned Gene Ontology (GO) terms where possible (Tables 10 and 11). All data werenormalized to the library with the fewest tags. There were 40 tags of the 100 more abundant(40%) and 55 tags of the 100 less abundant (55%) successfully annotated with correspondingEST or genomic sequence, had a significant BLASTx hit in the non-redundant database at NCBI(<e-5) and were successfully assigned GO terms. For the purpose of this study only these tagswill be discussed. As mentioned in Results section Ba, (the pairwise comparison of 37°C lowiron vs. 25°C), differentially abundant tags in 37°C low iron vs. 25°C and 37°C low iron vs. 37°Cwere highly similar. Therefore the genes will be discussed in this section only and the 25°C and37°C libraries will be referred to collectively as “YNB.” Tags of interest that were moreabundant in 37°C low iron vs. YNB were (Fold change): iron permease (16), HSP12, 60, 70 and90 (5-62), glutathione peroxidase (12), CipC (10), chitin deacetylase (25), 5’AMP activated ysubunit (15), Rho2 GTP binding proteins (17)Interesting tags that were less abundant in 37°C low iron vs. YNB were (Fold change):polyubiquitin (28), Ubiquinol-cytochrome-c reductases (3 8-42), high affinity copper transporter(30), CuJZn superoxide dismutase (17), carbohydrate metabolic genes (6-93) and a phosphatetransport protein (11).77Table 9: Comparison of SAGE data for strain H99 cells grown at (3 7°C low iron) vs. YNB(3 7°C) media.Abundance Classes for 370C (Number of Tags)Interval 1 2-4 5-9 10-99 100-999 1000+ TotalCount 4183 1899 841 642 48 1 7614Total 4183 4996 5467 14799 8383 1160 38988Abundance Classes for 370C low iron (Number of Tags)Interval 1 2-4 5-9 10-99 100-999 1000+ TotalCount 6602 2161 441 209 4 0 9417Total 6602 5424 2760 3784 708 0 19278Differentially Abundant (Number of Tags;% of Tags)999 0/0 99.0 % 95.0 %More Highly Abundant 37°C 370 2.31 % 765 4.78 % 1269 7.93 %Less Highly Abundant 37°C 290 1.81 % 435 2.72 % 693 4.33 %Total 660 4.13 % 1200 7.50 % 1962 12.27 %Similarly Abundant alpha=0.05 Tags PercentTotal 14031 87.73 %785.01 02: 5.02leOl5.0® Total:1599399.9% .99% I95% ISiq. •Figure 9: Expression profiling comparing relative transcript levels of SAGE tagsfrom strain H99 cells grown in YNB (37°C) and at (37°C low iron) media.The trumpet shaped lines represent the separation of the data by confidence interval.Dark blue crosses inside the lines indicate tags that do not show a significant expressiondifference. The grey line represents the 95% confidence interval, the yellow line 99 % andthe black line 99.9%. The green dots indicate tags that show a difference that issignificant with >95% confidence. Tags in the top left quadrant of the graph representtags upregulated at 37°C in low iron and the bottom right quadrant represents tagsupregulated at 37°C. The total includes the number of tags used in the analysis. Theimage was produced in Discovery SPACE Singleton tags were excluded.5.0 leOl 5.0 1e02 5.0 1e03370 C79Table 10: Differentially abundant SAGE tags from H99 cells grown in at (37°C low iron) vs.YN’B (37°C) media. * data sorted by ascending p-valueHeat Shack Preteht 12Se000aromooes oereuieiee I ODE-OSHeat Shack Prulele 12Sac000rurnoons eastoniorna S e.04SAGE tag does not hA C.neotetmoes eDNA He geeeedesequaneeCIpCEmeooete o.drdaesnu slgsltloanl BLAST results Iner databaso at NC BIcubA tellSAGE lag dues net hIt C.neotettnaes eDNA cc genumleeeguuneeSAGE leg hAs mere thee netsegaenoE oentlgNeoN soreeeIoA tailBe slgsllleanl BLAST results InCr dulabase at NCBIno elgnltloeet BLAST results iner database at NOEl605 dbusumal prutein L7a Lu)Ootiausaooherumuoes combe 3.00E-A2Norrothetloal preleleMorabslbtendenradaus 2-40 S.OOE-00eolmA tallne nlgnltloanl BLAST results Innr databaseatNOEl4n5 dbosamal peetein StESohensacoherumoces oombe 4.OOE-75nn nlgnltloanl BLAST tasulls Inen database aINCBISAGE lag dues nut hIt C.rtentotmans eDNA oe enumIosetsuaeoodes Hbcsaeral peeleln BlIS 00,4-recubA laBboat shook ptulele TOCttmloceucus crrualuuealmb taBSAGE tag dues not hA C.nrnutetnuane eDNA me geriatriceclduredsutasoCleslnidkan cenlrieooes 7.660-15nlbusattnal pruleltt P2Pudoseana ansan’na 1030-26nu slgrdtloanl BLAST results InttrdatabasealNCBldOs dbusaesat prolate s2oa (SI)SulAoosacoln.rumyces curIAe 203k-nohyputhelloal prencltrSch’eosacch.mrnooos eomhe 4,000-04hypeshetloal proteleMarcus biecorus 2.000-21heat shuck ptetcln BE NomolugCsndid. elbicone 2.000-32no elgnltlosot BLAST nnnolts Innrdatabeseal 5081Ire slgsltlosnt BLAST results Innrdatabase at NCBIBOs Ahosunrol Erotoln L3BIL3AASeldeuuucoharomcoee cumbe 4,000-23uhItIn deacetylaseSohionehollum commune 0,000-OStie nlgslBceet BLAST results Innt dalaheneat NCBIcc nlgnltloant BLAST nesslts Innrdatabaso at NCBIsypulhetloal proteInSuhiaosaooharumocos cumbe 4.000-s2BAGE tag dues nut hIt C.eeetonmann aDNA or gonumluId-OS pnoleln Schieaphyllamcommune t OOE-123SAGE tag dues cut hIt C.neetorenaes eDNA at Baomndosagueeeecsouhteme c eeldasc rebuntIV Ssdiauuauukammcoos cumbe 0000-t6alpba-areytasuAmyAEme.tota nidriaco 2000-45SAGE lag does nut hA C.rnnmtoeesaars eDNA en Benetnlosuguenceno slgnlscanl BLAST tnnuttslnerdatabaso at NCBInedoutaue Schizuoaoulnunmnyonuemnbe 0002.0228hu2 GTP bArdleg pretclntlsraaoumaydis 7000-17no slgnlsuanl BLAST mucus Innn database at NC RINAS dbenumal pruleln LOT000charumoces oooeoisbae 9030-34sEC Abusumal peotele L5ftuabidoouislbslisn. 5.000-onATP eltease lyaseSohioeeacohuromvcns remAn 5.000-35nn slgsltluarnl BLAST results Innn database a500BITATATGTGTA 0 0.0 290 unAnmo lu Lea IrumTOACTOITTA 4 20 123 02.2Suqunsoo 37cC 2? cC nunrnagaed Lea Ian Fold Dtttorcuce BLAST HIT B-nellIe % Idanrtlny to sbtdlaAny GO TmmAl7345501.020-SB 52-70(31 72-TA66 7930 4055 74n-s4 72 5042*-OS 01 60TAeASTTGCTATATOPuIuSGAITGTA&4°A5ou00000aoo4500000110ACMTACCTAQQASACCAI3GATOCACASTACAS005TSATTATSATASTSACA&SCGATATPAI,aAkaASTA0CTTOGGAA0ASCTCPASATAT0ACATA414-5045010SCOTT CTIATGAt.IA5ASATATATATGCAATAeAnAa,aAGpAa000450ATTGAeJGTAO0TACTCT’TCATATGTA000TATAC COATSTAS0000TAOAGOATSASMTOTTA0000TASCTTSATTSCASAGATGTSCIFTOSATCAGACAT]TTSATCTASSTATATASASTASA0TATMOASOTT0ASASTAIAASTOCSCTATCTTMAGTMASATOASTOG000ATATGAACAITACTGCACPGCM’rrTATOA000MOCASCSASTSTATCA0000TASTATOCATASCTAIOACASCTATTTATCAASCCCATCOTAT0 65 65 aniraluLewIrun21 t04 102 0.0 46 ainunlnLosnlrunn3 1.5 55 30.7O 00 37 uninue lu Lou ImsO 00 36 uniuun lu Lou Iruno o.o 30 unineo to Lea IrusISA 03t 203 2.4I 0.B 37 74.Ao 0.0 34 uninun to Lou Iruo0 0.0 34 uniquu Ia Lou Iano o.o 31 unioLto Ia Lom Iano o.o 28 uoiquo to Lee Iano 0.0 27 uninun le Lou Iron3 t.5 34 22.9I 0.5 29 5071 0.5 29 55.74 20 35 17.7o oo 20 unbnseloLonnlrorn2 to at 31.33 15 33 223o o.o 25 uninun to Lea Iano o.o 24 Le0500loLewlrun11 5.4 43 7.52 to 2A 0032 10 28 253T 35 30 104O 40 37 9.4o to 37 S.42 t.O 25 26.32 1.0 20 2632 1.0 26 2632 1.0 25 203o o.o 20 uniuue lo Lou lieno 0.0 19 unique In Lea Iran1 0.0 21 42.5O 00 18 unique Ic Luo Irono 0.0 IS unique In Lom Ian0 0.5 15 rainue to Lea leao 0.5 - iS raetLetuLealanO 0.0 10 roetuctuLeatron0 0.0 10 uesuuotoLealruo2 1.0 22 22.26 3.0 28 5.40 05 17 undooetuLealtorn2 IS 21 2122 1.0 21 2121 55 lB 3045 0.5 18 unison to Lea Ian0 0.0 16 ensue to Lou roeRescosue to streusRoueaouo to stressUratoqodtedUnoleseiSedProtein biouomhes’ePatein biosenthenisPatois to stress45 57 Urclassioed57 65 Protein bkrsqnitrnoieGA 81 Protein bbenetdtnenb27 0040 0465 70 Respunsr to stress79 92 Protein biosordhosis41 65 Cell aell41 6196 9k S’mnal treosduGino44 50 Cmeokodmi31-59 44-00 Cnnbolrodnele motabuBsrn36-66(31 54-50131 Cdltdar ascbnocedo 69 SarAlcorurixflon70 87 Protnin biosordheett41 66 Protein blosorthesis5? 76 Carb000drabe mntebolrsm80GCCACAGCCAATGATTTGMGAGM067A0 0 0.0CTTGTAGW 0 0.0CATITTATOT 0 00AMTCTIATG 0 0.0CCASTGATTA 0 0.0TWTGGTGC 0 0.0MTTTATGAT 0 0.0TSGAT000CA 2 1.0cMAaAoMA 1 0.5TAGCGATCAC 13 6.4TCGGGGCTGC 0 0.0TTGTTAGATG 0 0.0SAGE tag does not hit C.neotonnans cONS on gnoomlcno slgnlfinant BLAST nosolto tonn datahaso at BCBISAGE tag does not hit C.naotorntaos cOttA or gonomlohypothntinal protainNouroscora donna 5.006-COno significant BLAST tosolts Inon datahase at NC Elalpha-cop protoinBoo n6mine6us 1.066-33Aopergiiius lorrous 0fithosotnal Pnotnin BPLO2 ANNOTATEDno significant BLAST macfin inor datahaseat hoston signIficant BLAST rooctts inordatahaso at ttoBtcomA tallso significant BLAST resells innrdatahaso at hostSAGE tog dons not hit C.nontormans oGNA or gonomlogo significant BLASTo NIlePotatlvo 5-AMP-actinatodgamma nahunit familySnhiaosaccharnmcoos nomho 1.000-37on signltinant BLAST resells isor datahaso at NCBIdlonnlaotnna hydmiaso familySnhioosaoclraromyonn pomho 1.060-15no nigniticant BLAST msatts inor database at NOElnytonhromn ci pmoutnorSAGE tag hits mom than nonsnqannoe000tlgno signinnant BLAST msults Isor datahasnat NOElBeta-tsbslln selector 0CIIITIN SYNTHASE IONhOTATEGSAGE tag dons oct hit C.nooformans oGNA or gonoinlosogoancono significant BLAST monolts inno databaso at NOElno slgnlfioant BLAST moolts innr datahaso at NOElrihnsomal pmtaio L3Scodonlnra Nuoinnrda 1.000-44no significant BLAST msolts Inor database at NOElno signlfioant BLAST macits innrdatabasn at lostno significant BLAST macits innodatahaseat hostno significant BLAST macIts Innrdatahaso at 11051SAGE tag doan oct hit C.naotntmats oGNA or gonomlonaltoonnaoropnrphyrlnogondonarh000lana Dade mnrio 2.000-51SAGB tag dons not nib.noolormats sOfiA or gonominsan00000SAGE tag dons oct hit C.nnolntmats nGNA or fi050micGlatathlonn Pnnooldaso hTfiISamharomconscnmacisiae 2.OOE-36dhosomal proteIn LISAXanlhonhsllomccns dondrorhooc 2.OOE-80SAGE tag does not hit C.saotormana cGNA cr genominSaocharomycnscnmcislan 4.OOE.56Translation Elcngatlnn Factor 3 ANNOTATEDdbg rlhnsomal paotnln ggboumnsooracrasns 1,000-55protoasoma componnot POPShomologSctdnosaochammooon nombe 6.006-22pnrnibln S.c3000dhancoloasn,dsp1Aansrgillcs Siminalm 6.OOE.S0no sIgnificant BLAST msolts Innodatahaso at NOEldbonaclooprntnln, F4lknSohiacoacohammycos nomhn 5.006-07no stgnifinant BLAST mnolts inor datahasnat NOElSAGE tag dons not hoC.nonlormans oENA or gnnnmlcsngnancano ntgnlfioast BLAST msolts isfiaquancn 32eO 07 nO nonnalleod Lom Ims Pold Gittomnon BLAST SIT fl-salon NO Idoomp NO similarIty GG Tarm I56 7031 5748.66 66-7434 5630Vnsniole lrannnortCarbohcdraln molohoismProtein biosonthnsisSional lrancdcwion0.0 16 odoco to Los Iron0.0 16 onioco In Low ImcIS onhon to Low Ironto oohon to Loo Ironon’moo In Loo Iron16 oni000loLoolmn16 oninon to Low Imnto oninon In Loo Imn16 on’nlon In Low Iwn20 20218 36.434 5.315 enrIco lo Lao Iron15 onlnooloLoolwn0 0.0 15 onmnooloLaolron0 0.0 15 onlnon to Low Iwn0 0.0 15 06100 to Low Iwn0 0.0 15 oninoeloLoolron0 0.0 15 onl000 to Low Iwn0 0.0 14 enrIco bLow Iron0 0.0 14 udoco to Low Iwno os 14 codon to Low lroo0 0.0 14 06000 Ic Low Irono co odaoa to Low Iron0 0.0 oniqoo to Low Iwo0 0.0 on’moatnLcwlron6 3.0 8.1I 0.0 3241 0.0 3242 1.0 17.2o o.o udoon to Low Iwno o.o undoo to Low Irono o.o 06000 to Low Iron0.0 13 oninoo to Leo Iwn54 Awwalio oowcocnd dorrradatioo3.006-55 54 67 Oolhdarrnsciranon5.000-44 32 4714141424161617131313TANGAATGTAOAOTTATTANOWITCATTGANTGTACPAOA&8NOATCCATTTATGMTMCATMTGTATACGTGTAAAAOATCATCTATAT005TATAOCATACTTCATATGTOTAM000TTATOMmTATCATTATCATCOTCAT100CATAMCTGTTGTATATATSTOOATATOTTITACTGACTSTCTATA00000ATACARCTTGTTACACAITOATAGAOATTCTGAOTTTTATGGAATOTATA000TATGATTTTAATTATPACGAATATTCATMAGPAT000AOTATGTSTAATTATATAC’TTG0 0.03 1.54 2.0I 0.5I 0.5057 3.50 0.0o o.oO 0013to20t5151523t21212Coil nrowlh andlorwainlonanonCoil oaIIProloir bmnsordhnsisEnmoy and motahotawRnscomn IoslrnssProloin biosynlhasisOmnomo to OrcnsProlnin bionynlhnsinProloin hiosoolhos’nProlnmn Ioldira and dooradotiooNmlnobaso wolaholiswNmloobasn wolaboliamanmnun In Los iroo12.110.130.330.330.36.6odnun In Low Ironudoun to Low Ironuoiqun to Low Iron65 7542-70 60-7441 6372 6165 8480 8560-55 74-5555-70 06-5251-60 71-710.0 12 uninun to Low Iron0.0 12 06000 to Low IronAGTTTCTTGT 0 0.0 12 uni000 In Low IronMTOTOOTAT 0 0.0 12 udquo Ic Low Iron81Table 11: Differentially abundant SAGE tags from H99 cells grown in YNB (37°C) vs. (37°Clow iron). * data sorted by ascending p-valueSequence 37cc 37 oC normalized Low rca Fold Difference BLAST HIT E-vaiue % identity % similarity GO TermCOACADACCD 1160 573.0 38 15.1GOCCTCDGTT 430 212.6 S 70.9GCCAACDCCD 359 177.5 I 177.5CACGTTCACD 299 147.3 1 147.3OCTCGCDACG 290 143.4 2 71.7AACGTCTDCC 303 149.0 S 30.0COCDD4AADD 294 146.4 9 16.2DTTTCCDCTD 240 110.7 5 237TCTDTCDAOG 210 103.0 2 51.9AADCCCOTTG 199 98.4 1 98.4DADANOCOTO 180 92.0 0 unique to 37OTCOOTOOTA 267 127.1 14 9.1AN000TOOTO 229 1132 9 12.0OCTCTCCAOG 171 84.6 I 84.0ACOOCCOTTA 148 73.2 0 unique to 37TAOGCCGTCT 101 89.9 4 22.4OOCCGACCTO 286 141.4 24 9.9TCTOOTCOAO 164 01.1 2 40.5OCTOCCTACA 149 692 0 unique to 37CTCADCOATO 199 97.9 9 10.9OTATTOACCC 149 73.7 2 36.6AT000CTCCC 127 62.8 S uniquc to 37TCTTTCCDAO 144 71.2 2 35.6CAOANCCCCG 196 77.1 S 15.4CACOOCOCAT 185 91.5 11 8.3TCOOTCOTOT 012 55.4 0 unique to 37GCTTTOCTOC 129 63.8 2 31.0ATOATC000C 934 66.3 3 22.9TCCATCCOAT 140 69.2 4 17.3CACOTCCACO 140 69.2 4 17.3GOTTACOCCO 108 53.4 0 unique to 37OCTTTTOCCC 917 57.9 1 57.9CCTCTTCCTO 106 92.4 0 unique 0037OACOACTCTA 136 67.2 4 16.8OCOTTCTCOO 104 59.4 0 unique 1037COTOTCAAOC 104 51.4 0 unique to 37OOTATCCTCG 903 50.9 0 unique to 37OTTOOCANCO 902 50.4 0 uniquetu37OCCOTCCGAA 141 69.7 6 11.6OTCAAOMOC 89 49.0 0 unique to 37ACTCAOOTTO 122 60.3 3 20.1CTCTTCCCCT 145 71.7 7 10.2COADOTATCA 96 47.5 0 unique to 37TCCTOOCMO 95 47.0 0 unique to 37OCOGACOACT lOt 49.9 I 49.9TCTOCCTCCO 90 44.5 9 uniquetu37CCOCOACCOT 117 57.8 4 14.9OCCOCTTCTO 85 42.0 0 unique to 37Translation eiongaooo tactor I1TEPII ANNOTATEDno significant BLAST results in nrdatabaoe at NCB1Cycicphflin A ANNOTATEDThioredoolo perooidaseSchizosacuhunonnyces pombr 2.000-03500 RIBOSOMAL PROTEIN L6Xenopus laeyis 0-902SAGE tag does net Nit C.neolormans cDNA or gesomicno significant BLAST results in ordatabase at NCSIno signtficant BLAST results in ordatabase at NC6iRibosonsal protein S12Bnanclrioctoma boloheri 2.000-43tag too close to end at canda600 RIB050MAL PROTEIN Lot-ASaocharomycesoereuisiae 4.OOE-34Fl ATPasa beta subunitKtuyvoromyoos iaoys 3.000-03en stgolficant BLAST results Is srdatabase at NCBtSAGE tag does eat hit C.oeotormans cDNA or genemicno significant BLAST resaits in srdatabase at NC6tSAGE tag does not All C.noolormans cBNA or genomicribnsomat proteIn LISohioosacchanumyces oombe 9.009-56Histose h4 proteInMus musculus 9.000-39ATP syotNaso ollgomycinsasoitinity conlerral pretelsN. urama 0.000-39cytobine isdscing-giyceprotelnFilobaddirta oeotormans ANNOTATEDphosphoketoiasoThanmocynoohocoocus alongalus 2,00E-32Nypethetlcai proteinMionobolbilen degradons 3.000-17aiteroative onldaseCry010coccus neolonmans oar. grubi AoootatodADPIATP carriar protoinNeurosponaorassa -ItORibasomal protein Ldt ANNOTATEDprediotod prstsioNeun0500ra crassa 9.000-08putatine hydrolase; dienelactenebydralase tamlly; possibly inoeioedIn oblcrocatocol degradatIonS. combo 9.000-09SAGE tag does not Nit C.oaatarmans cDNA or gonomlchypotNeticai proteInNeurosoora orasna 6.000-95CelOn saporoolde dlsmutaseCrypl000ccus nrolormans oar. gmbb 6.OOE-g9mltocboodrlai malatedeNydrogonase 0. 3.000-37hypotboticai proteinSchiaosaocharomyoes pombe 0.SOE-06predIcted protalsNeurosoora omssa 6.SOE-05conserved hypotbatical proteinSohiaosacoharomycns combo 9.000-21transaidniaseSohizosacoharomycos oombo 1.OOE-114so signiitcant 6LAST results in nrdatabase at NCBI03 small nucieoiarrlbosuoloopratein prelate 1MP3S. oerovidae 5.0DB-SOmtatad ta Ypt-Istoracting proteinYIP2 Nourosouro crassa 9.000-30dbossrrnal pralels S6; 405ribosomal protein 05 Homo sapiens 2.OOE-75putatlue proteinAnobidoosis thahaoa 4D-28Sucrose biphesphate aldatasePanauoocidioideo bradtends lOSE-dOdbosomai protoie L37Sacuharomyoesoereuisiae 4.OOE-32atdohydn reduolase INASPHISponidobolus salmonicolun 0.005-60prottilsMalus a domossoa 0.OOE-22seciaoslde dlpNosphale kleaseEmerluelia oidulaos 2.OSE-46dbosomai proteIn 513 4ibo; protnioArabidoosistrataoa 2.005-38hypolNetloat proteioNeunospona uraosa 4.005-09ablqulnel-oytochromo-u roductaseNeunosooracnassa 2.000-75Protein Bi0500AosisSignal iraosducoonN0500nse Ic oSessProtein BiosyoArdsPnotoin BiosynthesisProtein OrosynthocisCellular respirationProtein EiosyoAnnisChromosome organioa9on andbiosynAesicCollular recyinasooOnclacsifirdCarbohydrale metabotsmCellular respiradonProtein SiosynthedsUnolasolnadRonponse to nErnsCellular reooir0900Carbohydrate moiabotsmCr0 gruoth aodlom mdateoaocoCall grooth andloro mdnieoaoceProtein BiosynthedsCarbohydrate motabo9smProtein BionynorosisCarbohydrate metabolismCell gruuu9r anryoro mdntenanoaNucleohaso metabolismProtein BiosynthesisCellular resoiraSon01 7672 6266 8964 7680 9152-50(21 694412175 gO43 6346 0930 9377-79 0029 3820 d776 8969-96 96-90065 7936 6060 6447 6968 7650 7444 5871 8434 4970 g259 0140 6142 6057 7773-65 66-9448 7468 8082Seooasoa 37 nO 37o0 normalized Low iron Fold Oirrennnoo BLAST HIT E-oaloe % Identttn % sindlaeito GO TerOr140 69.2 9 7799 49.0 2 24570 38.6 0 unique to 3778 38.6 0 uniqon to 3775 37.1 S uoiqoo to 37161 86.5 22 4173 36.1 S onique to 3772 35.6 S wiqsn to 3769 34.1 S oi4qoo to 378.4 41.5 2 20876 37.6 1 37.665 38.6 2 15865 32.1 5 uorqoe to 3773 36.1 I 36.164 31.6 0 oniqon to 3764 31.6 0 ooiqoe to 3784 41.5 3 13862 35.7 0 oniqor to 3761 30.2 0 redqsn to 3787 43.6 4 108123 60.6 12 5.175 37.1 2 10.560 39.6 3 13256 27.7 S oniqon to 3756 27.7 6 ooiqon to 3767 33.1 2 16.660 29,7 1 29.7S CT C CTCTTACACCTCMOCSC CO 000TCGCÁO CACACC STO CÁO CATTAAAGGA000TCA000TTOOAGAGO OCT OCTOTCCM000TAACCMGCTTOACC ST C 5170MCTTOITNGAATOTACC CCCTACAOC000TC CT C GTATC SOCTANCOCCOGAOOAGOAOOCOAATTATOOAGCACCA&AOATCOO3TACCGOAOITOTTGACAOATGGAGA0CCAhCTCTCCWTTTFCGAAOATOCGAGCAGTACCAGGGATOOCAI300GCTO OTTTOAACC SC C OAT 0CGMCC0000OTOACOT1TCC CT GATC SC 0GGTATGAACOCGOTGT7GATAGO GAOCACTPAOMOACCOTCTOAOCOCOASOCTTTCTCCO3ACCCCTCGCCACCAATGCMTGCCGGMAAGCCTGACOOGAAGATCGCTTCACCACCTAOCT000CMGACTCGACGACCGCTTTTISCGTGTTMGGCATCGACTTGGTTCAGCAGGCCC OTAAC GOTATP s09thann - alplra StrainHonassa 109646 71 64SAGE tag does oot hit C.nnnlnrmaos oDNA or gnnotnishypnthe6oal protoioMagentospiriiioooognototeobourn 3.006-tI 37 54obtqutsot-oytoohromn 0 rndootaseoomptno ooboottSohiooseoohorooyoes poobo 2.006-22 41 64moog anose soperootde dismotasePhaooooohooto oinysosoooom 4.006-73 66 76oytoptaomio ribosornat protoio S12Podosporo ansedno 4.006-43 65 67no sigoificoot BLAST resoits in ordatabase at NCBtHtstoroe HI; HbnlpSaodcsomyons onreoteon 9.006-05 45 61SAGE tag dons ont hOC.nnotoomaos oSNA or gonondsdOS dbosornat protein SOSobizopbylom oonerooen 1.006-70 67 76trtnseploosptrate inonroraseAopnogdoo ooyaan 9.506-50 52 72hypotnrottoat proceinNeorosporo 000550 0.005 32 43toypotho6sat proteinMiooobotbiloo degrodaos 9.05E-18 45 65no sigsifi000t BLAST resoits in ordatabase at 6661hypothetloat proteinNoorosporo wosso 1.006-13 47 63poptidyt-protyt ots-traos isoworasoNeorospore 000550 3.006-77 45 OSSAGE tag hIts more mao onesegoenoe condopstaboe synaptobreoinAsoergtioshimiqosis I 006-11 57 616166 artadty sepper transpoetorSotrioooaooirewnryons poobe 4.006-25 40 92ptroopbate teansport proteinS. ononoteon 4,006-97 60 71hppotttetioat proteinNoorosoooa ooasso 5.00640 43 97no signinoast BLAST rnsoits in ordatabaso at NCBtBNA-hteoeng protein OoBNBPAodoysronoa nmoroarum 3,006-Id 60 76Poiysbtgsiun ANNOTATEDspnaptobreotn hnmotoglSoh’oosaoohooooyoes p0060 4.006-26 HI 72beta-tebottn 2Hy0000aaoiroos 4.006-38 71 81SAGE tag dons not hit C000lormans sONA or genomioetotamato oaaloaoetateOanio redo 6.006-96 56 72spiioing taotor; rnsI7a ribosomalprotein Sohizonaooharooryons pombe S 006-42 82 91PIroHOp0000irmonryons onrnoisioo 4.096-37 47 65hypothodoat proteinhourosoora orassa 5.096-22 48 62onnamnd protoin peodootPodospora ansnAoa 5.096-32 30 94Protein wIth 3 RNA bindingdomains Sobioosoostrwooyoos 2.506-19 52 63dhosomal protein S21Ceoddo albroors 1.006-26 62 02no signifioont BLSST rosoita in nedatabose et NCBino significant BLAST rnsoits in nrdatabase at NCBiSAGE tag bits mote then nonsegoense sontigprotoasome sobonit iotaMis mosoulos 1.006-36 31-72 91-94brain protein (3PtOt)Ceeooohobdtrs nleqans 9.506-14 30 47no signifioant BLAST resoita is nrdatabase at NCB1SAGE tag hits more than consngsoenoe snntlgatiergonHatesonoio symoodails 2.006-11 26 43dehydrogeoasa Ssnetomyoesooolioolor 2.006-Sr 34 53hypotholtoat protewNoro0500ra wassa 9096-36 44 59no significant BLAST rnsotts In srdatabase at huntteanslormer sednndargintse-dohrihonooieoprnteln, pota000Arabidopsis theliere 2.006-09 40-SI 57-70SAGE tag hits more Ihon onesegoenoe oontigpotatioe metal ttanspottnrSohioosaooharooyons pombo 6.006-52 35-36 90on signinoent BLAST renoits io ordatabase at NCBiAoyi-CoA-binding proteinChaetoohraotus orllosus 1006-16 47 69Co*rdorrnspirahonCnlioiaroospirotorOnsooese ta slressProtein BiosynthesisCnraomosooe orgorloaton nodbiosynthesisProtelir BiosynthesisCwhobydoole enoloboisonCnll growth and/nm mainteneooeCell groooth ondloro maintenanceTrwrsowlTmnrneoelCoo m000th ordloro oowetnraeooProtein CotebolismCell growth aedloro mortnnonooCell growth aedloro noinlenenonAmino cold mntabohsoProtein BiosynthesisTransonoNrwieobasn rentthobsoProtein BicsysthnsisProtein CarabolismSeolossrtedUndoes/fledCarhohs&ate metabsismUnniassAndTranspod52 25.758 26.775 37.150 24.750 242do 24.256 27762 30.7dO 23.7do 23.747 23.247 23254 26.759 29.279 39.145 22345 22344 21.944 21,844 21.843 21343 21.396 47.560 29.774 36,6S ooiqun 10372874 930 ra000n to 37S seiwieto3lS uerrpoe to 371 2772 153S uniqon to 37S onigunta370 unrgueta3l0 unigue Is 37I 26.72 1466 65S oriqun to 37o uoiqonto370 raiqse to 37S tedqoets37o uniqon to 370 seiqun to 37S seiqun to 3711 4.33 9,9H 6.183c. Pairwise comparison of 37°C low iron vs. in vivo librariesA pairwise comparison was completed for the low iron library against the library fromcells isolated from rabbit CSF, designated as the “in vivo” library (Table 12). Information for tagabundance classes can be found for each library included in the pairwise analysis (Table 12).Total tags used in the analysis were 19,278 for 37°C low iron and 49,048 for in vivo. At the99.9% confidence interval, 0.20% (44 tags) were more abundant in cells from in vivo vs. 37°Clow iron and 0.1% (23 tags) were less abundant for a total of 0.31% (67 tags) differentiallyabundant. At the 99% confidence interval 0.80% (176 tags) are more abundant, 0.32% (71 tags)were less abundant for a total of 1.13 % (247 tags). At the 95 % confidence interval 5.42%(1188 tags) are more abundant, 0.68% (149 tags) are less abundant for a total of 6.1 % (1337tags). Tags below the 95% confidence interval are not considered to be significantly different.Therefore, 93.9 % (20592 tags) are similarly expressed between these two libraries atalpha=0.05. A graphical representation of this data is found in Figure 10. Notably, there werefew tags that were significantly differentially expressed indicating that the libraries were highlysimilar. Pairwise comparisons were annotated for genes that were more abundant in 37°C lowiron vs. in vivo and assigned GO terms where possible (Table 13). All data was normalized to thelibrary with the fewest tags. Only 16 tags with higher abundance in 37°C low iron vs. in vivo(16%) were successfully annotated with corresponding EST or genomic sequence, had asignificant BLASTx hit in the non-redundant database at NCBI (<e-5) and were successfullyassigned associated GO terms. The low number of annotations may be partially due to a lowlevel of stress-related ESTs in the H99 database, and lack of a completely assembled H99genome at the time of SAGE data analysis. For the purpose of this study only the fully annotatedtags will be discussed. Interesting genes encode a zinc finger protein, a ubiquitin degradationprotein, a putative 5’-AMP activated gamma subunit signaling molecule, a putative ABCtransporter and a white collar protein homolog. White collar protein, encoded by BWC] isinvolved in light sensing and has been characterized in C. neoformans (Idnurm and Heitman2005). Mating and haploid fruiting is normally repressed when light is present but null mutantsof BWC] have been shown to mate regardless of the presence of light. It is possible that Bwclhas a repressive affect on mating and therefore may be more highly expressed in the presence oflight. In this regard, it is plausible that this tag may be more abundant in 37°C low iron vs. invivo due to a higher exposure to light during growth in vitro vs. in vivo. All other genes identified84are involved in protein synthesis, cellular metabolism and maintenance. Genes that were moreabundant in vivo vs. 37°C low iron were annotated by Dr. Barbara Steen and will become part ofa larger scale SAGE analysis in the future. Genes found in vivo but not in vitro will likelyprovide interesting targets for further study.Table 12: Comparison of SAGE data for strain H99 cells grown at (37°C low iron) vs. cellsisolated from rabbit cerebral spinal fluid (in vivo).More Highly Abundant in vivoIess Highly Abundant in vivoTotalDifferentially Abundant (Number of Tags;% of Tags)9g % 99.0 °/o44 0.20 % 176 0.80 %23 0.10 % 71 0.32 %67 0.31 % 247 1.13%qçfl %5.42 %0.68 %6.10 %Similarly Abundant alpha=O.05 Tags PercentTotal 20592 93.90 %Abundance Classes for in vivo (Number of Tags)Interval 1 2-4 5-9 10-99 100-999 1000+ TotalCount 10305 4183 1007 685 27 0 16207Total 10305 10426 6436 16651 5230 0 49048Abundance Classes for 37°C low iron (Number of Tags)Interval 1 2-4 5-9 10-99 100-999 1000+ TotalCount 6602 2161 441 209 4 0 9417Total 6602 5424 2760 3784 708 0 19278118814913378599_9% .99%95% B1e02 Si!j •5.0leOlI.-.5.0) Total:21929Figure 10: Expression profiling comparing relative transcript levels of SAGE tagsfrom strain H99 cells isolated from rabbit cerebral spinal fluid (in vivo) or grown at(37°C low iron) media.The trumpet shaped lines represent the separation of the data by confidence interval.Dark blue crosses inside the lines indicate tags that do not show a significant expressiondifference. The grey line represents the 95% confidence interval, the yellow line 99 % andthe black line 99.9%. The green dots indicate tags that show a difference that issignificant with >95% confidence. Tags in the top left quadrant of the graph representtags upregulated at 37°C in low iron and the bottom right quadrant represents tagsupregulated in vivo.The total includes the number of tags used in the analysis. The imagewas produced in Discovery SPACE Singleton tags were excluded.86Crptococcus (H995.0 leOl 5.0 1e02 5.0 1e03in vivoOTTTGADTOOCTTOCTGTMTATAT000TACATCMCCTGCATTATATATATATCCTGOACTATGCACASTATDCMw6PV6ATCGAGACGGTASGAVIGCTCAmGGTATTAATACGCGACCACTGCOGCASITCUSATG0-°ATATGSACATATDSAGTA9.OAWA6GTAOAGCCPAGTGAGAGGWTPAGPATGTATATACDTGTATAaAOCAOTCTGACTGTCTACTI’WCACATWGAO.OAItmGGP.0ACC1TCTAGTGTA000’TTI’GTCATATGTPAGCATTGTTAGATGCAGCDTADTATTAODGC1TTDTGTACGTAGCGATCDMDGACCTCGDATCACCDTTGAOAOTTGTAAacOAATGTATGGTTTATATDAGTAATATGTATGACATADTGTAGCATIGACACTTGCAOAAOOOAAr1A,rrw’rnATA.OATtAIASACGTGATTGCATAGCAOCCTATGATASTAGGTATATAGTMTTGDATACATMCC1TTCAS1TCAGPAGTTTWTACT1TIGTCATTASOACTGCTA1TITGGT500TTC0.0 10 wcquntotow’won0.0 10 teiouototowwoo1.2 14 11.654 II 29.0OA 11 29002 9 aqouotolow’won0.4 10 25.454 10 2540.0 9 unjoco to low woo0.4 5 2290.0 7 uoiqeotolowiroo0.0 7 uoiqoo to low iroo0.0 7 uoi000 to low iroo0.0 7 uoiqootolowiroo00 7 uoiooo to low iroo00 7 unison to lowiroo59 17 4.304 9 20.40.4 9 20.4at to 4.92.9 14 5.106 9 hA24 13 5.50.0 6 wdnwetobw’woo0.0 6 uoinoo to low iwo0.0 6 urwtobwiow0.0 6 tr*twtolowfrow0.0 6 wnwtolownow1.2 12 10.255 15 4.231 14 4.504 7 17.60,4 7 17.6OA 7 17.00.4 7 17.00.4 7 17.00.4 7 t7.604 7 17.600 6 10200 8 10.220 11 56t2 0 761.2 9 7612 0 760.0 5 aique to tow ‘eon0,5 5 wiqunwtow’eoo0.5 5 Lecque to tow ‘eon0.5 5 toqw to bow ‘wooas 5 utiour to tow iron0.0 0 tatiocw to ton. ‘eon0.0 0 tat our to bow ‘woo0.0 0 wiqtwtwbo’eoo0.0 5 teiqurwbweoo5.0 5 0010.0 totow woo0,0 0 ooiour to tow too0.0 0 ooiouo to tow too38 5447-65 02-75UteleooiOod60000000 to tbhtSigoat TrsosdtwtiooCat month eodloe merarrenceProtoit sto4hesisTable 13: Differentially abundant SAGE tags from H99 cells grown at (37°C low iron) vs. cellsisolated from rabbit cerebral spinal fluid (in vivo). * data sorted by ascending p-valueSoqeenee 11 otoo St .1gw soreseozed Late row Fold Sfftereooe BLAST lIT E-oabe % tdrrdtty 1’. stntttarty GO Term0to200003S230000Wrong orioslanon SAGE tag to BLAST hBSAGE tag dons not htt C, newtortrtanooSIdA or 0000mb nonoenweSAGE ta dons 60666 C. nnwlorrntarscOttA or g000rtdo noquesceSAGE rag doss not hOC. oeotoemaescGldA on gewsedc seqoeoosSAGE rag hits noon. than one swqoeooetontigSAGE tag does nut hit C. neotorrenetsoEINA or 0060ndo nmwnsonSAGE tag hAs mom than one sequenceoonttgWrons edentatton SAGE son to BLAST 66SAGE lag dons not htt C. iteolermenscOttA or neonede snnoenoedoom. otinowtion SAGE tao to ELAST hASAGE tag dons not hit C. seolermaneoGNA or nesendosnnaesssSAGE tag dons not hit C. neolormansoGNA ornenondcsenoesseWrong odentution SAGE tag 1w BLAST hisSAGE tag hAs mom than onoseqasnonsonttgExpressed pmtetn-stmttar to RIKEN oGNANnhidoosis lhsliaoa aSSE-37 36 StWHtTE COLLAR I PROTEIN (WCI)Souwsituta utsssa S.SSE-S4 4t 09poltA witWrong ndantaAoo SAGE tag to BLAST hASAGE tag does got hit C, nsclorntaesoGNA or ganondo sequencesuhuris lamttySchwosaucharonycen eowbs 1006-37 34 56Bwea4tahegn cwtaxtor SHoowsaeieou 5526-44 32 47565 rittohontal pecteis LITSuhtoooauultatuntycan owobe A.-3d 55 66SAGE tag dose cot hit C. nooloenontscOttA ow sonetrdw seqneewemoons odmttation SAGE log to ELAST he 4.Wh-56 26-50 36-57topoisortwrona 5ANNOTATEDSAGE tag dons not hAG. sowlomontsoGNA or ennoedo emoeooeSAGE tag doss not hit C. nooleenraesoGNA or gotsondo sequ000eSAGE tag dons cot hit C. nooroeoaesoGNA or osnoardo smaetweconsomad hypuehotioct protais- yeastCandito atoloam 5.026-SO 27-29 41-46ho stenihoact BLASTo hAs nero tanhypothetIcal AlT-family proteinhieDsacchatum9ces oomhn 2666-07 37-05 48-49 htwleobena metebothtoSAGE tag dons not hit C. seolormacsoGNA or genendo sequenceNo slgntfeaet BLASTs hits near fagNe sigetficant BLASTo hAn near tag6SS dhosomrt nmtois L34-B 4.SSE-S6 ST-7t 76O2 Protoio synthesisNo etgotft006t BLASTt hAs sear tagSAGE tag does net hitC. nonlormansoGNA or owonmic sequenoeWrong odentatton SAGE tan to BLAST hit 1.596-to 26-30 42-49SAGE tag does not hit C. neoformannsGNA or qenonde sequenceSAGE tag hits more than one seqoewseocnttgSAGE rag hits mum than woe sequeooe000iigSAGE rag hire emra than one seqotece000tigSAGE rag dons not hit C. seuluroranssENA or gerwmio sasoonoeSAGE rag hits mum that one seqsonoeeontigSAGE tag deco nwt he C, eenlonnonsnENA or geewedc sntrnrrsondoors eeiwosaAon SAGE leg to BLAST hit 4.660-tiWrors edestation SAGE ton to BLAST he 1000-167SAGE sag dwos sot he C. ntotorrnansoGNA or gwewodo setssmtoeSAGE sag dons not hA C. noolormanseSNA or genomtosegurneepltmphattdytinoseoi hhiana-mmalndhEster DtserAa towbtw 2266-66 L6-d cod Sotot metabolismSAGE tag dons nut hA C. neolornransnSesA or geeondo seqomweSAGE tag duos ont hit C. nsnloennanseShlA or oaeondo neqoenonSAGE lag duas tot hit C. neoeernonssoGNA or gorsoedo saguorwagiso Nnqor proteinOnh’aosaoohawmyces cembe i.66E-55 Pwwin cetsbnhonSAGE tag dccc not hA C. nAelermaosoSNA or genomic sequenceSAGE tag dues not hA C, neofermansoSNA or aenomic smusnee8744 6327-34 48-52Sequctrnw hi utoo Is slow noenretteed Low trwr Fold Dilterenco BLAST HIT E.oehw Irs Idnntiy % niedleiky [ G TermGTATCSACGG 0 0.0 5 unIson to low ironGCGGTGTICG 9 0.0 5 Lrmtw In low ‘nonTCGGCACGGT 0 00 5 Lei000tolnw’nonWTGGTCTG 0 00 5 wniqwn to Ice ironCATTATAGCT 0 00 5 unique to low IronGGAAACDTGG 0 0.0 5 unique to low ironCATTITCTIA 0 0.0 5 unique In mw ironGACTGTTGTA 0 0.0 5 unioun Intro ironCTTGCTTTAG 0 0.0 5 unique Intro ironCGTTACCGM 0 0.0 5 unique Intro ironCTC000CTAO 0 0.0 5 uruoue to low irnnACCAACCT7G 5 2.4 11 4.7TCC0.6CGACT I 0.4 B 15.3GCATICDTCC 1 0.4 B 15.3TCTCTGTACC I 5.4 6 15.34AITTCTGTT 1 0.4 6 15.3AGTATGTAGC I 5.4 6 15.3GAT1GATCCC I TA 6 15.3GCG1TICCTA I 0.4 6 15.3MTAGCTCAG 1 0.4 6 10.3GMAO7ACTG 1 0.4 6 15.3ACOACTCAOA 1 0.4 6 10.2TTGTMG0.SA I 0.4 6 10.3A40r0TCATC 15 3.9 14 3.6TMTCATAGC 4 1.6 9 5.7CAGMCGACG 4 1.6 9 5.7CAO.AGGTACC 2 0.6 7 6.9GACITTCMT 3 12 6 6.0AGT0AMVI 3 12 6 6.0GACATATGM 17 6.7 16 2.7TGACAOSATG 0 0.0 4 urqotw In low eon11TTGTCAGC 0 0.0 4 rerique to lore ‘eonATAGDTACGG 0 0.0 4 Lewrw to low hornCTMCTTTCA 9 00 4 rerhiuetolow’eon1TIGTTAku.A 9 09 4 reiuuetulow’eonDTGMTCGCS 0 05 4 uoioue to low ron5ACTTDGGCA 0 0.0 4 unique to low iron0GT1ACTGCA 0 0.0 4 unique In low iwnATACCTTCPA 0 0.0 4 unique to low ironITCATCATCA 0 0.0 4 oniqun to low ironGA000TGATT 0 0.0 4 unlqun to low loreATATTGDATG 0 0.0 4 unique to low IronGGAAGACCCG 5 0.0 4 onique to low IronWCGGTCAG 9 50 4 unique to low ironSAGE lug dote not SOC. nuolnnnreeerOSA nr qeewedo eeeueeonSAGE tug dote rot 9SC. ewolonereesoG1IA or eenusedc sequ000nNo elgelloerd BLASTu she new teeSAGE tug duos set hOC. eeeloreraneoGNA or gnreoedc sneowecoSAGE lag does net 911G. nuetereranocGBA or grncndc seeueeoeSAGE tag does net hit C. enolormuesrOSA or gencetic soquenonWrowe oduetuttee SAGE tag to BUST hit l.OOE-19 25 37No stontncunt Bt.AST0 hon eeoc 109SAGE tog dnnn not 99G. neetonmeesoGNA wr eunomtn uequeeoeGSA tepnteomoraee IIIGrosoplritumuleonouster I.OOE.62 40.100 64.150 DNA rudioosnnSAGE tug does nut httC. nootuewernoGNA or qenuetmo sequ000eSAGE tag does nut hltC. neotnewonseGNA or qeeuwto soquunonebtquflte testes dugrodatlun prwtetnSohwoescchar00005e nowhe 400.22 313-IA 49.62 Prurojo ottuholieroSAGE tag dews not 9IIC. eeodomrecrscElIA or gmrue,4o suqAenceSAGE tug dnns sot hi C. eootooeresrsoGl4A on geswndo sequenceh5pwthnocal semeheane proteInNeronepuro o’uoou 3000.16 41 04Wrong eeleetanoe SAGE tag to BUST hi 2005.47 35.49 55.68SAGE tag dote sot hIt C. eootwrnraesoGleS or emeusde euesotwetanoy peoteteGolnl000anctruewer005 conrhn 1.000.110 26.54 42.77 Cdl orowftr uodlnr esuirtoterweSAGE tog dens not lid C. neuloninaneeDNA or genoetto sequenceSo togettleent Bt.ASTO hits Beer toeSAGE tog dote not 911G. newtwrwunnoDNA ne eeromto seguenceSAGE toe dots not 9IIC. neotorenenecENA we ennomlo sequenceWrong odentetico SAGE 1001w BUST 69hypothetIcal proteInseuroAtore crouce 6,000.00 31.56 43.75SAGE tue dots rot hItC. seotoenunscDNA or enromlo sequenceputetioe ATP.htrdtsg cassette trenspoflerorococ4dioidu hrusitisnn’n 2,006.32 40 04 Ttensnmtacelyltranetcreeu500chorornooesoereuieiae 1 050.13 56 71 Cstt wattlcaAohudroto b’wuonthesisSAGE tag dote sot 611G. eeutoesaessItNA or qenuetic soeuecoselplru.acerylaswAuryAEmeticefle redriens 2.35545 31.59 44.60 Cwholrodrelr nreueEoternSAGE lee dote nolh9C. neoloenransoDNA en genundc sueueeoeSAGE tag dots nut hOC. nnolwnenunsoGNA ce osnorntc seeueeceSAGE tog does sue 911G. neotoenanerENA wn ernondo sequenceNo sknrlllcald BLASTo hfle ecer tagSAGE leg dote cr199 C. seulusoetruoGNA orueeordo ewouwroeWrneq cdrnlattcn SAGS lee to BUST 99 1.OSC.50 27 41No stenlloent BLASTu hits neer legSAGE leg dues oct 6IIC. esotosneescGNA or gencmlc sequenceSAGE leg hits now then one suquenserocheSAGE leg hfls more then unu sequenceccsltgSAGE leg duns sot htl C. nenlnsnenecENA crossumtc sequenceSAGE leg dote sot hit C. oeotosneesrOSA oneenomloeequencwSAGE leg dote set 611G. neolceneneoGNA or ennomle 510051ccSAGE leg dose eel 611G. seotormunscGNA or genomlo sequence88d. Iron-regulated tags: Three-way overall comparison of the libraries 37°C low iron,25°C and 37°C in YNB.SAGE tag occurrence was compared between the libraries from 37°C low iron, 25°C and3 7°C. Data is represented in a Venn diagram (Figure 11). The low iron library has a far highernumber of unique tags at 7,329 vs. 3304 for 25°C in YNB and vs. 4,210 at 37°C in ‘(NB. Theseresults further indicate that the low iron nutritional signal has a greater effect on transcriptionthan a temperature change in ‘(NB. A total of 771 tags were shared by all libraries, some ofwhich likely represent essential metabolic genes. There were a higher number of tags shared bythe 25°C and 37°C libraries (2,366) than 37°C low iron and 25°C (1,050) or 37°C (267) libraries.The same interesting observation that the 37°C low iron library is more similar to 25°C than 37°Cwas observed in this multiple library comparison.89low ironA cZ5oC 37oCOccurrence of a tag in a SAGE library for A. 2 5°C, B. 37°C low iron and C. 3 7°C.Intersections of the Venn diagram represent shared tag sequences between libraries.Numbers represent occurrence of a tag and do not reflect relative abundance.Figure 11: Comparison of SAGEtag occurrences for cells grown in low iron(37°C), YNB (25°C) or YNB (37°C) medium.90e. Comparison of multiple libraries including: 25°C, 37°C low iron, 37°C and invivo.A multiple comparison of SAGE tag occurrence was performed between the librariesdesignated : 25°C, 37°C low iron, 37°C and in vivo . The data are represented in a Venn diagram(Figure 12). In this comparison, the in vivo library from rabbit CSF has by far the most uniquetags at 10,290, followed by37°C low iron at 5,388, 37°C at 3,737 and 25°C at 2,331. Theseresults suggest that the in vivo environment induce multiple genes necessary for survival in thislow nutrient environment; some of these may help the fungus thrive in the hostile hostenvironment. It should be noted that these comparisons do not take into account library size andthe in vivo library had far more tags (49,048 vs. 19,278) than the low iron library. This sizedifference may account for some of the difference in the number of unique tags. In total, 683tags are shared by all libraries, again likely reflecting the many tags for potentially essentialmetabolic genes. Genes that are essential would presumably have some level of transcript presentunder any condition. This is similar to the number of shared tags in the comparison of low ironwith 25°C in YNB and 37°C in YNB (771). This is not surprising because the shared tags likelyrepresent tags that are essential for cell growth in any condition. The greatest number of tagsshared by only two libraries is that of the in vivo and low iron libraries at 1,941 tags compared to973 for in vivo and 25°C, 473 for in vivo and 37°C, 138 for low iron vs. 25°C, 108 for low ironvs. 37°C and 1,590 for 25°C and 37°C in YNB. This global analysis of tag occurrence isconcordant with findings from the pairwise comparisons where the in vivo and low iron librariesare more similar than the libraries from cells grown in YNB.91Occurrence of a tag in a SAGE library for A. 25°C, B. in vivo C. 37°C low ironand D. 37°C. Intersections of the Venn diagram represent shared tag sequencesbetween libraries. Numbers represent occurrence of a tag and do not reflectrelative abundance.AFigure 12: Comparison of SAGE tag occurrences for cells grown in YNBmedium (25°C), isolated from rabbit cerebral spinal fluid (in vivo), lowiron (37°C) medium or YNB (37°C) medium.92SUMMARY AND DISCUSSIONA striking result in this SAGE analysis is the similarity of the low iron library to that of cellsisolated from rabbit cerebral spinal fluid. Tag abundance compared between these two librariesshow only 0.31 % difference in abundance at the 99.9% confidence level. This indicates that thelibraries are nearly identical (<0.10 % at 99.9% would be considered identical). This resultsuggests that the low iron nutritional influence on gene transcription in C. neoformans may berelevant to conditions during infection. Further, the data suggest that that low iron medium(LIM) may provide a useful in vitro model for studying virulence-related gene transcription.Another interesting result was that the 37°C library for cells grown in low iron was more similarto the library for cells grown in ‘(NB at 25°C than grown at the same temperature, (37°C) inYNB. Cells are stressed under low iron conditions as partially indicated by the number of stressrelated genes that are highly or more abundant under these conditions. It is possible that strainH99, originally a human clinical isolate, is less stressed at 37°C than 25°C. Therefore, morestress-related genes are induced at the lower temperature and this response may more closelymirror the stress of low nutrient availability that is found in low iron conditions. Notsurprisingly the same phenomenon was noted in pairwise comparisons of the in vivo library fromrabbit CSF to the ‘(NB libraries (Steen, B.R. personal communication).Genes that were notable in relation to the response to stress in the low iron transcriptomeencoded the heat shock proteins (HSPs). HSPs are involved in the modulation of apoptotic death(Garrido et al., 2001) acting as cellular chaperones (Parcellier et al., 2003). HSP]2 is ofparticular interest because it was highly abundant in 37°C low iron. There are two copies ofHSP]2 in the strain H99. The first and third most abundant tags showed similarity to HSP]2 ofS. cerevisiae (designated HSF]2-1) and Saccharomyces pastorianus (designated HSP]2-2) andhad 280 and 123 tags, respectively. HSP]2-l and HSPJ2-2 were highly upregulated in the 37°Clow iron library in comparison to the 25°C library (HSP]2-1 at 14.3 fold ; HSPJ2-2 at 3.4 fold).Both genes were upregulated compared to the 37°C library (HSP]2-1 was unique to 37°C lowiron; HSPJ2-2 at 62.2 fold). HSP]2 was also found to be highly abundant in the in vivo library,with 637 tags corresponding to HSP]2-1 and 468 tags for HSPJ2-2 (Steen et al., 2003). The tagscorresponding to HSPJ2 in the in vivo library were more abundant in comparison to the 25°Clibrary (HSPJ2-1 at 31 copies ; HSP]2-2 at 57 copies) and the 37°C library (HSP]2-1 at 093copies; HSPJ2-2 at 4 copies). These results suggest that HSPJ2 may play an important role inthe response to stressful situations such as the low iron or the in vivo environment, butinterestingly not to temperature for C. neoformans. In fact, SAGE tag numbers for HSPJ2 arelower at 37°C (HSP12-1 at 0 copies; HSPJ2-2 at 4 copies) than 25°C (HSP]2-1 at 31 copies;HSP]2-2 at 57 copies) in YNB, which is counterintuitive when considering the gene designationas a “heat shock protein”. The same trend was noted for one of the HSP]2 genes from SAGEtemperature libraries for the strain B3 501 A (serotype D) (Steen et al., 2002). A second copy ofHSPJ2 was identified in genomic DNA of B3501A, however no corresponding SAGE tag wasidentified in this SAGE temperature data (Steen et a!., 2002).HSPJ2 has been implicated in the response to many stressors in C. albicans and S.cerevisiae. The well studied process of vinification represents a number of stresses to yeast cellsincluding heat, ethanol, oxidative, osmotic and glucose starvation stress; in all of these casesHSP]2 has been shown to be upregulated (Carrasco et al., 2001; Sales et a!., 2000).Transcriptional and biological studies in S. cerevisiae have identified HSP12 as required for thefollowing processes: 1) biofilm formation (Zara et a!., 2002); 2) upregulated in response tocadmium stress (Momose and Iwahashi, 2001); 3) involved in barotolerance and cell wallintegrity (Motshwene et a!., 2004; Chauhan et a!., 2003); 4) survival during near freezingtemperatures (Kandror et al., 2004); 5) transcriptional response to the herbicide 2,4dichlorophenoxyacetic acid (Simoes et a!., 2005); and 6) resistance to ethanol and acetaldehyde(Aranda et a!., 2002). C. albicans mutants that were resistant to the drug fluphenazine showedan increase in transcription for HSPJ2 in conjunction with the genes for the multidrug effluxpumps CDRJ and CDR2 (Karababa et a!., 2004). Finally, Hspl2p has been suggested as apotential medicinal component in yeast extract for the treatment of burns and wounds of the skin(Schlemm et a!., 1999). Alcoholic extracts of yeast have been used as a naturopathic treatment ofthese ailments for a number of years, biochemical analysis by this group identified a number ofcandidate peptides related to the stress response including Hsp l2p, copper and zinc superoxidedismutase and ubiquitin. Further analysis on the separate compounds would be necessary toidentify which proteins carry definative medicinal attributes.HSPJ2 is regulated by a number of well-studied signal transduction systems. In S.cerevisiae, HSPJ2 is regulated by cAMP through the zinc finger proteins Msn2p/Msn4p(Ferguson et a!., 2005; Praekelt and Meacock, 1990) and by the two component signal94transduction system that includes the putative response regulator Ssklp and mitogen-activatedprotein kinase (MAPK) Hogip (Chauhan et al., 2003; Alepuz et a?., 2001). In the absence of afunctional Msn2p/Msn4p transcription factor and PKA catalytic subunits HSPJ2 has been shownto be controlled through the heat shock factor Hsflp (Ferguson et al., 2005). There are manyuncharacterized putative zinc finger proteins in the genome of C. neoformans that could behomologs of MSN2/MSN4. The HOG pathway has been characterized in C. neoformans whereserotype A null mutants are attenuated for virulence (Bahn et a?., 2005) but the pathway has notbeen studied with respect to the stress response. The global regulator in the C. neoformans HOGpathway is thought to be Pbs2p rather than Sskl p as is the case in S. cerevisiae and C. albicans.A strong candidate for an HSFJ homolog is present in C. neoformans, on chromosome 2 ofJEC2 1 (1 .4e-27 and H99 (6e-2 1 Although a great deal is known about what inducesand regulates HSPJ2, little is known about its specific function except that it is located in the cellwall and cytoplasm (Motshwene et a?., 2004; Sales et a?., 2000) and is proposed to be a cellularchaperone (Parcellier et a?., 2003).HSP]2 clearly plays a role in a wide range of stress responses in fungi as is apparentfrom the multitude of publications that identify the gene as upregulated in response to stressorsand in the sensitivities of hsp]2 null mutants of S. cerevisiae to stressors. It is also clear from thisSAGE analysis that HSP]2 likely plays a role in the response to low iron in C. neoformans. In C.neoformans, homologs are present or systems are characterized for the HSP]2-related signalingpathways found in S. cerevisiae and C. albicans, however, many of these pathways have notbeen well studied in C. neoformans with respect to low iron stress. It would therefore be veryinteresting to elucidate the function, localization and regulation of HSP]2 in C. neoformans withrespect to low iron and other stressors such as glucose starvation, oxidants, high osmolarity, cellwall damaging compounds, ethanol, high pressure or antifungal agents.A number of genes that were more highly abundant in the low iron transcriptome havebeen further investigated in the laboratory. These include the iron permease FTR], a gene ofunknown function HOT] (homolog of CIFC from Emericella nidulans) and a chitin deacetylasegene. Due to the similarity of the low iron library to the in vivo library, many of these genes hadalready been identified as potentially related to virulence by studies of the transcriptome fromcells isolated from the CSF of rabbits (Steen et a?., 2003) or from iron regulation in the serotype95D strain, B3 501 A (Lian et al., 2005). The following chapter presents the results of a detailedfunctional analysis of the SIT] gene identified in iron regulation studies of the serotype D strainB3501A (Lian et al., 2005).96CHAPTER FOUR: Analysis of the Siderophore Transporter Gene SITJINTRODUCTIONPathogens possess specific mechanisms that allow survival in the mammalian hostenvironment including the ability to sequester iron that is tightly bound by host proteins. Thelow iron environment found in human serum, brain and cerebral spinal fluid also provides anutritional cue for induction of key virulence factors (e.g. capsule formation) in C. neoformans.The work described in this chapter is specifically focused on the mechanisms of iron sensing andtransport in C. neoformans. Fungi have both reductive and non-reductive mechanisms to acquireiron. In C. neoformans a high affinity, reductive iron system has been identified andcharacterized that includes the high affinity iron permease, FTR] and the multicopper oxidase,FET3 (Lian et al., 2005). In conditions where iron is scarce or tightly bound, more specializedmechanisms are often employed by organisms to acquire this essential nutrient. Recently, asiderophore transporter homolog was identified in a serotype D strain by SAGE; this gene wasdesignated SIT] and it was found to have a higher transcript level under low iron vs. iron repleteconditions in SAGE analysis (Lian et al., 2005). This relationship was confirmed in this chapterusing Northern analysis. Siderophores are small molecules that bind ferric iron with highaffinity. Due to this high binding affinity, iron bound to siderophores is largely unavailable tomore generalized reductive uptake systems (such as that encoded by FTR]/FET3) and istherefore reserved for organisms with the ability to uptake such molecules. The siderophorebound iron can then be non-reductively transported into the cells by specialized ABCtransporters. In C. neoformans, SIT] is a homolog of the S. cerevisiae ARN3/SIT] gene thatencodes a ferrioxamine B permease from the major facilitator superfamily (Lesuisse et al.,1998). The high transcript level of SIT] in low iron vs. iron replete conditions and its possiblerole in iron acquisition in vivo led to the initiation of gene characterization studies of SIT] in C.neoformans as detailed in this chapter.97MATERIALS AND METHODSA. In silico Analysis of Sitip and SIT].a. Determination of the coding sequence and exon-intron boundaries of CnSitlp.The coding sequence and exon and intron boundaries were manually determined in thenucleotide sequence by comparison of six frame translations of the C. neoformans SIT] gene tothe S. cerevisiae homolog Sitlp/Arn3p (gi6320770refNP_010849.) protein sequence. Exonsand intron locations were further resolved by identification of C. neoformans exonlintronconsensus sequences: donor 5’-GT(XX)GY-3’ and acceptor 5’-YAG-3’ (See sequences inAppendix IV-Aa.). Intron and exon location and coding regions were later supported for theserotype D strain in the annotated genome at TIGR where the SIT] gene was given the identifier181.m08534 and designated CNA07920. Note that the TIGR database is a JEC21 database andthat JEC21 Sitip has 605 predicted amino acids, however closer inspection revealed that the stopcodon had been included as an amino acid in the prediction and the published sequence (whendownloaded) had 604 amino acids. No EST was available for SIT] in the University ofOklahoma databases for B3501A or H99.b. Comparison of SIT] and Sitlp in serotypes A, B and D.SIT] sequences were identified in genomic sequence databases for the serotype D strainB3 501 A (http://www-sequence.stanford.edulgroup/C.neoformans/index.html), the serotype Astrain H99 ( and the serotype B strain WM276( Amino acid sequence alignments were performed with Clustal W.http://www.ddbj p/search!clustalw-e.html. The similarity tree was produced from theClustal W alignment file using the Clustal W program:http://www.ddbj . nig. ac.j p/search/clustalw-e.html. Nucleotide similarity comparison from exonand intron sequence of SIT] was achieved by comparing sequences using the BLAST 2sequences algorithm at NCBI (http :// .cgi).c. Identification of siderophore-related genes in C. neoformans.Siderophore-related genes were identified by searching known fungal genes involved insiderophore synthesis and usage (Haas, 2003) against genomic databases for C. neoformansusing the tBLASTx algorithm. Homologs were searched against the following databases:98Serotype D-at SGTC (http:Hwww-sequence.stanford.edulgroup/C.neoformans/index.htrnl) orTIGR(, Serotype A at DUMRU(!) and serotype B at MSGSC ( Only fungalhomologs that had been fully characterized were included in the analysis. The genomic locationof hits and expect values were recorded in an Excel spreadsheet.d. Comparison of Sitip to fungal homologs.Sequences for fungal homologs were supplied by NCBI ( andGenBank Accession numbers can be found in Appendix IV-Ab. C. neoformans SIT] sequencewas from the Serotype D strain B3501A (!group/C.neoformans/index.html) (Appendix IV-Aa). Amino acidalignments were performed with Clustal W http://www.ddbj similarity tree was produced from the alignment file using the Clustal W programhttp://www.ddbj Abbreviations of fungi are: An-Aspergillusnidulans, Ca-Candida albicans, Cn-Cryptococcus neoformans, Sc-Saccharomyces cerevisiae andUrn- Ustilago maydisB. Strains and Growth Conditions.Cryptococcus neoformans strains used in this study are listed in Table 14. The serotypeD strain B3501A was provided by Dr. J. Kwon-Chung (National Institutes of Health). Theserotype D strain JEC21, JEC43 (ura5), the serotype A strain H99 and serotype D and A cAMPmutants were provided by Dr. J. Heitman (Duke University). The ura5 mutant of B3501A wasisolated by plating the wt strain on 5-fluororotic acid (5-FOA) plates; the spontaneous mutantsobtained were restreaked twice for single colonies on 5-FOA. LIM was prepared as described inChapter Three. LIM+Fe was prepared by addition of ferric EDTA to 100 M (FeEDTA;SigmaEDFS). LIM+BPDA+Deferoxarnine was prepared by addition of bathophenoanthrolinedisulfonic acid to 550 tM (BPDA; Sigma B1375) and deferoxamine to 100 tM (Deferoxamine;Sigma D9533). For measuring growth in liquid media, 5 mL of yeast extract peptone dextrosemedium (YPD) was inoculated with a single colony and grown overnight at 30°C in a gyratoryshaker at 250 r.p.m. 100 iL of culture was transferred to 5 mL of yeast nitrogen base (YNB)medium and grown overnight at 30°C. Cells were washed four times with sterile water and99transferred to flasks containing 50 mL of LIM, LIM+Fe or LIM+BPDA+Defereoxamine at aconcentration of 1 x 106 cells per milliliter. The cultures were shaken at 250 r.p.m. at 3 0°C. Thegrowth of wild-type and mutant strains were monitored by measuring absorbance at 600 nm.Results were an average of three separate trials.Table 14: Strains used in this study.Name Genotype Serotype ReferenceJEC21 MATawt D J.HeitmanJEC43 MATa JEC21 ura5 (FOAR) D J. 1-leitmanje-si MATa,Isitl::URA5 ura5 D This studyje-siRl MATotAsitl::URA5 ura5 + SIT1::NAT D This studyB3501A MATcz wt D J. Kwon-Chungb-ui MATa B3501A ura5 (FOAR) D This studyb-s10 MATaAsitl::URA5 ura5 D This studyb-s42 MAToAsit1::URA5 ura5 D This studyb-s42R1 MATaAs1t1::UR,45 ura5 + SIT1::NAT D This studyb-s42R2 MATczAsitl::LIRA5 ura5 ÷ SIT1 D This studyH99 MATcz wt A J. Heitmanh-si MATczAsitl::NEO A This studyh-s2 MATaASItI::NEO A This studyh-siRl MATaAs1t1::NEO + SIT1::NAT A This studycdc40 MATaJEC21pka1::ADE2ade2 D J. Heitmancdc 68 MATa JEC21 pkrl::URA5 ura5 D J. Heitmancdc 99 MATa JEC21 pka2::URA5 ura5 D J. Heitmancdc 103 MATa 3EC21 pkal::ADE2 pka2::URA5 ade2 ura5 D J. Heitmancdc 2 MATa H99 pkal::ADE2 ade2 A J. HeitmanJF-13 MATcL H99 pka2::UP.A5 ura5 A 3. Heitmancdc2+ PKA1 MATa H99 pkal::ADE2 ade2 + PKA1 A J. HeitmanGPA1 MATccH99gpal::ADE2ade2 A J. Heitmancdc 7 MATcz H99 pkrl::URA5 ura5 A J. HeitmanC. RNA Isolation and Northern Analysis.Cells were grown as described for the measurement of growth in liquid media. Cells wereharvested by centrifugation after 6 hours growth in LIM or IR, flash frozen in an ethanol bathand lyophilized overnight at -20 °C. Cell pellets were pulverized with glass beads for 10 minutesand the cell powder was resuspended in 15 mL of Trizol extraction buffer (Invitrogen). RNAwas isolated according to the manufacturer’s recommendations with the additional step of LiC1precipitation at 4°C , following the standard ethanol precipitation step. Northern blot preparationand hybridization was performed as described (Sambrook et al. 1989) and all hybridizationexperiments were performed with two independent preparations of RNA from cells grown inLIM or LIM+Fe. The hybridization probe was prepared with polymerase chain reaction (PCR)amplified 409 bp DNA fragment from exon 6 of the SIT] gene using primers: SITEXON6F100(GTTATGATCCAGTCTGCCGT) and SITEXON6R (TGCCGAGAAGCTCGAGAAGG). Theprobe was labeled with 32P using an Oligolabeling kit (Amersham Pharmacia Biotech).D. Construction of SIT]:: URA5 (serotype D) and SIT]::NEO (serotype A) Disruption Alleles.For serotype D strains JEC21 and B3501A, a PCR overlap strategy was used to constructdisruption alleles (Davidson et al. 2002). First, DNA fragments containing the 5’ (742 bp) and3’ (853 bp) regions flanking the SIT] gene were amplified from genomic DNA from strainJEC21. The primers adjacent to the gene contained sequence for SIT] and the URA5 marker5’: SIT1A (ACTCACTTCCTCCGATTCAG) andSIT 1 C (AAGGTCGAGCAACTTCGCTCAGGACTAAGACGTTGGCAAG);3’: SIT 1 D (CCCACCTCCTGGAGGCAAGTCGGTGCGCTGTTATATGAG) andSIT1 F (CTATGTCATCAGGTGAGTGG).The URA5 marker was amplified using hybrid primers for SIT] and URA5:SIT1 B (CTTGCCAACGTCTTAGTCCTGAGCGAAGTTGCTCGACCTT) andSIT1 E (CTCATATAACAGCGCACCGACTTGCCTCCAGGAGGTGGG) amplified fromthe plasmid pJHM973 containing the URA5 marker encoding orotidine monophosphatepyrophosphorylase. The three fragments were joined by overlapping PCR as per Davidson et al.,2002 using the primers SIT1A and SIT1F. For the serotype A strain H99, a modified overlapstrategy was employed (Yu et al,. 2004) that increased the yield of the final construct. A DNAfragment containing the 5’ (1151 bp) and 3’ (1162 bp) portions flanking the SIT] gene wereamplified from genomic DNA from strain H99. The primers adjacent to the gene were hybrid forthe neomycin (NEO) marker:5’: SITA1 (AATCCGCACTCTCTCCATCA) andSITA3 (AGCTCACATCCTCGCAGCCCAAGATGTTGGCAAGTGGA);3’: SITA4 (TAGTTTCTACATCTCTTCATGTACCATAGCTGCGGCTG) andSITA6 (CTGTGTGCTGATAACTGTCG).The NEO cassette was amplified using hybrid primers for SIT] and NEO:SITA2 (TCCACTTGCCAACATCTTGGGCTGCGAGGATGTGAGCT) andSITA5 (CAGCCGCAGCTATGGTACATGAAGAGATGTAGAAACTA) from the plasmidpJAF1 containing the resistance gene for neomycin. Fragments were used in 2nd and 3 roundPCR reactions as per Yu et al. 2004. Nested primers were used in 3’ round PCR:101SITANF (TGCGCATGCTAAGAACTTCC) andSITANR (AAGGCCGAAGCGGAGAGGAT).Constructs were introduced into the appropriate parent strain by biolistic transformation(Toffaletti et a!. 1993). Transformants were screened by colony PCR using primers:SIT1UP (TCACTTCCTCCGATTCAGAC) and 6328 (GGTCGAGCAACTTCGCTC) forserotype D; andSITAUP (GGACAGAGAATTGCCTTCGT) and NEOPS (AGCTCACATCCTCGCAGC) forserotype A.Transformants in which the wild-type allele was replaced were confirmed by Southern analysis.One disruption strain for JEC21, two for B3501A and two for H99 were retained for analysis. Tocomplement the disruption mutations, the wild type SIT] gene was reintroduced by biolistictransformation into one mutant of each strain background on the vector pCH233 that confersresistance to nourseothricin (NAT). Serotype D reconstituted mutants were confirmed by PCRusing primers SIT1A and SIT1F (Appendix IVD) and Southern analysis with the enzyme Scal(Figure 21 D) to ensure the return of the wt band. Reconstituted strains were resistant to NATand regained the ability to use a siderophore as a sole iron source (Figure 24A and B). Theserotype D strains were also 5FOAR and unable to grow on uracil deplete medium (return ofura5- phenotype). For serotype A, reconstituted strains were resistant to NAT and regained theability to use siderophores as a sole iron source (Figure 24C).E. Plate Assays.a. Siderophore utilization assays were performed as described by Heymann et a!. 2002.Iron free medium plates (IFM) were prepared in the same manner as LIM with the addition of20g!L of Bacto-agar and 550 tM BPDA chelator. 200 iL of 1 0 CFU/mL cells were spread onIFM. A disk of Whatman paper was saturated with 10 iL of 100 M deferoxamine then placedin the center of the plate. Plates were incubated for two days at 30 °C. Zones of growth wererecorded by digital photography.b. Melanin production assays: DOPA medium was prepared with 20g/L bacto-agar in 900mL dH2O (autoclaved for 40 minutes). Separately, 1 g of L-asparagine (Sigma A0884), 1 g ofglucose, 3 g of KH2PO4, 0.25 g of MgSO4-7H20, 200 mg of L-DOPA (3,4102dihydroxyphenylalanine Sigma D9628) were dissolved in 100 mL dH2O. The pH was adjusted to5.6. One mg of Thiamine-HC1 (Sigma T4625) and 5 tg of Biotin (Sigma B4501) was thenadded. The mixture was filter sterilized and then added to the autoclaved agar after cooling to50°C. The following modifications were made to DOPA medium prior to filter sterilization totest various parameters:1. 4X [DOPA]: 4mM DOPA (Sigma D9628) was added instead of 1mM2. 4X [DOPA] + glucose: 4mM DOPA (Sigma D9628) instead of 1mM + 1.0% glucoseinstead of 0.1 % was added3. DOPA pH9: media was raised to pH9 with NaOH4. DOPA + Fell Ascorbic acid: 100 iM of Fell Ascorbic acid (Sigma A0207) was added5. DOPA + BPDA: 100 tM of BPDA (Sigma B1375) was added6. DOPA + BPDA + Cu(S04)2:100 1iM BPDA (Sigma B1375) was added stirred for 1hour; then 200 pM of Cu(S04)2,was added7. DOPA + FeEDTA: 100 iM of FeEDTA (Sigma EDFS) was added8. DOPA + Deferoxamine: 100 pM of deferoxamine (Sigma D9533) was added9. 2,5 Dihydroxybenzene diacetic acid: 100 tM (Aldrich D10,920-7) was added in lieuof DOPA10. 3,4 dihydroxybenzoic acid: 100 !IM (Sigma P5630) was added in lieu of DOPA11. hydroxyquinone: 100 iM (Sigma H9003) was added in lieu of DOPATo test for melanin production, 10 pL of 1 CFU/mL cell suspension was spotted on plates.Plates were incubated for two days at 30°C, then assessed for pigment.c. Temperature sensitivity assays: Low glucose asparagine medium (LGA) was preparedin the same maniier as DOPA without the addition of DOPA. DOPA medium was prepared asdescribed in the DOPA assays. 5 tL of 10-fold serial dilutions of cells were spotted on eitherLGA or DOPA plates, initial inoculum was 1 0 CFU/mL. Plates were incubated for two days at30°C or 37°C.d. Cell wall integrity assays: Assays were performed as above except 0.015 % SDS, 300tg/mL Congo Red or 30 pg/mL of calcofluor white was added to the media.103F. Microscopy.a. Differential interference contrast (DIC) microscopy was performed on a ZeissAxioplan 2 microscope. Melanized cells were prepared by placing 10 L of sdH2O on amicroscope slide. A loop of cells was scraped from a DOPA plate (4 days growth), resuspendedin sdH2O and covered with a glass slip. Cells were viewed at 1 000X in oil immersionb. Transmission electron microscopy (TEM) was performed on a Hitachi H7600 TEM.The cells were grown as described for the DOPA plate assays on DOPA or low glucoseasparagines (LGA) grown for 4 days at 30°C. The full 10 pL spot of cell growth from DOPA orLGA medium was resuspended in lmL sdH2O, vortexed for 10 minutes then centrifuged for 5minutes at 13,200 rpm. Cells were fixed in 2.5% glutaraldehyde in 0.05M cacodylate buffer at28°C under vacuum. Microwave (Ted Pella Microwave) processing was employed at 100W with2 minutes on and 2 minutes off (repeated twice). Samples were washed with cacodylate buffer at28°C (twice) for 40 seconds using power level 2. Cells were fixed in osmium tetroxide at 28°Cunder vacuum by microwave processing at 100W for 2 minutes on and 2 minutes off (repeatedtwice). Samples were dehydrated in increasing concentrations of ethanol as follows: 50%, 70%,90% and 100% (the final step was repeated three times). Samples were infiltrated with Spurrsresin in increasing ratios of Spurrs;Acetone under vacuum power level 3 for 3 minutes: 1:3, 1:1and 3:1. Resin was polymerized by baking the sample 0/N at 60 °C. Samples were cut into 70nm sections on a Leica Ultracut T Ultramicrotome and stained with 2% uranyl acetate for 14minutes and lead citrate for seven minutes.G. Antibiotic Susceptibility Assays.Antifungal compounds were purchased from Sigma-Aldrich Corporation as follows:streptonigrin (Sigma S1014), amphotericin B (Sigma A4888), ciclopirox olamine (SigmaC04 15), glyphosate (Sigma PS 1051) and phleomycin (Sigma P9564). Antibiotic susceptibilityassays for determining the minimum inhibitory concentration of antifungal agents (MIC assays)were performed as per the National Committee for Clinical Laboratory Standards (NCCLS)procedure for micro-broth dilution of yeasts (approved standard M27-A). For these assays, 96well plates (Falcon 353072) were employed with a starting concentration of 16 ig/mL for all104compounds except phleomycin (which was 4 ig/mL). Assays were performed in SabouraudDextrose Broth (SDB) with a starting inoculum of 5 X CFU/mL. Microtitre plates wereincubated for two days at 30 °C and growth was evaluated visually. The MIC assays wereperformed in triplicate.F. Virulence Assays in the Murine Model.Cells were prepared by inoculating 5 mL of YPD with a single isolated colony and growingstrains overnight on a gyratory shaker (250 r.p.m) at 30 °C. Cells were washed three times withsterile phosphate buffered saline (PBS), then resuspended in 5mL of PBS. Cells were countedwith a hemocytometer and cell counts were confirmed by plating serial dilutions of strains onYPD agar with subsequent growth for two days at 30 °C and colony forming units (CFU) werecounted. Virulence assays were performed by Anita Sham. For serotype D, the B3 501 A wt strainand the sit] mutants were tested initially using DBA1 mice (Jackson Laboratories, 16-20 g). Fivemice were used for wild-type B3 501 and four mice were inoculated with the sit] mutant. Themice were inoculated by intranasal inhalation with 5 x 106 cells per mouse in 50 p1 of PBS. Forserotype A, A!Jcr mice (NIH Program, 16-20 g). were infected intranasally with an inoculum of5 x serotype A cells per mouse in 50 p1 of PBS. Ten mice each were used for wt, sit] andsit] + SIT]. An additional experiment was performed using a ten fold smaller iiinoculum 5 xcells in 50 p1 of PBS and 5 mice per strain. Animals that appeared moribund or in pain wereeuthanized. Protocols followed approved standards of the UBC Animal Care Facility.105RESULTSA. Characterization of SIT].a. Comparison of SIT] and Sitlp in serotypes A, B and D.The Clustal W program was used to compare the nucleotide and amino acid sequencesimilarity and produce similarity trees of SIT] and Sitip from serotype D (JEC21), serotype A(H99) and serotype B (WM276) strains of C. neoformans (Figure 13 and 14). As found inprevious cross serotype comparisons, the genes from serotype D and from A are more closelyrelated to each other than either is to serotype B (Tanaka et al., 2005). There is also a notableinsertion in the serotype B Sitip protein (605 amino acids) in comparison to serotypes A and D(604 amino acids). This insertion appears to take place in the carboxy-terminus at amino acid600 where serotype B Sitip appears to have an additional glutamic acid residue (E) (Figure 13-inbold). The SIT] gene was further compared across serotypes for nucleotide divergence inseparate exons and introns (Figure 15). As expected, the majority of nucleotide sequencedivergence occurs in intronic sequence. Nucleotide similarity in exons is> 80 % in all pairwisecomparison of strains. Intron sequence similarity varied widely, with little to no similarity insome introns for serotype B vs. A or D strains (e.g. introns 1, 4 and 5).106SerotypeD MPDQLYPEYELQRAISNTKNADDSTAFGEKSPGVRRIEI IAASFTTWHRWVLFISVFFMSSerotypeA MPDQLYPEHELERAISSTKNADDSTAFGEKSPGVRRIEIIAACFTTWHRWVLFISVFLMASerotypeB MPDQIJYPEAEIJERAI SNTKNADDSTAFGEKSPGVRRIELIAASFTTWHRWVLFISVFLNA** .Serotypeo CNYGTJDGSVRYTYQAEALSELGTSAQVSTVTVVRSIVAAAAQPAFAICVSDYFGRVSILI ISerotypeA CNYGLDGSVRFTYQSEALSELGTSAQVSTVTVVRS IVAAAAQPGFAKI SDYFGRVS ILVISerotypeB CNYGLDGSVRYTYQAEALSELGTSAQVSTVTVVRSIVAAAAQPCFAKVSDYFGRISILVISerotypeD SVILYVVGTIVTATSTNLAAFCGGSVLYQFGYTGVQLLVEVLIADVTSLRSRLLFSYI PASerotypeA SVILYVVGTIVTATSTNLAAFCGGSVLYQFGYTGVQLLVEVLIADVTSLRSRLI FSYI PASerotypeB SVILYVVGTVVTATSTNLAAFCCCSVLYQFGYTGSQLLVEVLIADVTSLRSRLLFSYI PA*********:************************ ******************:******SerotypeD TPFLINAWI SGNVASAVLTHSTWGWGIGMWAI IFPVTVIPLLFSLIQAEWRAHRKGLLRBSerotypeA TPFLINAWI SGNVASAVLTHSTWGWGIGMWAIIFPVTVIPLLFSLIQAEWRAHRKGLLRESerotypeB TPFLINAWI SCNVASAVLTHSTWGWGIGMWAIIFPVTVIPLVVSLVQAEWRAHRRGLLRESerotypeD I PSPLRTFGDRBMWADIFWQIDLVGLLLLAAVLALILLPFTLAGGVAS IWRTARVIAPLVSerotypeA I PSPLRTLGNRHMWADIFWQI DLLGLLLLAAVLALILLPFTLAGGVGSIWRTARVIAPLVSerotypeB I PSPLRTLGDRHMWADIFWQVDLMGLLLLAAVLSLILLPFTLAGGVASIWRTARVIAPLV*******:*:**********:**:*********:*************************SerotypeD VGFVVALPLFVFWELKVARHPMLPFRILKDRQVLASLFIANLLNTAWYTQGDYLYYTLLVSerotypeA VGFVVALPLFVIWELKFARHPMLPFRILKDRQVLASLFIANLLNTAwYTQGDYLYYTLLVSerotypeB VGFVVALPLFVIWELKVARHPMLPFKILKDRQVLASLFIAMLLNTAwYTQGDYLYYTLLVSerotypeo AFDRDIISATRVQNIYSFTSVVIGVCLGLIIRKVRRLKwFIVAGTLLFVLAFGLLIRYRGSerotypeA AFDRDI ISATRVQNIYSFTSVVIGVCLGLI IRKVRRLKWFIVAGTLLFVLAFGLLIRYRGSerotypeB AFDRDI ISATRVQNIYSFTSVVIGVCLGFIIRKVRRLKWFIVAGTLLFVLAFGLLIRYRGSerotypeD GYS I SDFAGLVAAEVVLGIAGGLFPYPTQVMIQSAVQHERTAVVTSLYLASYSVGSALGNSerotypeA GYSVSDFAGLVAAEVVLGIAGGLFPYPTQVMIQSAVQHERTAVVTSLYLASYSVGSALGNSerotypeB GYSVSDFAGLVGAEVVLGIAGGLFPYPTQVMIQSAVQBERTAVVTSLYLASYSVGSALGN***:*******************************************************SerotypeD TIAAAIWTNTMPSHLYNDFIRAGLSTTDASTLQALAYASPLQFI IDYPPGTPEREAVGSASerotypeA TIAAAIWTNTMPSBLYNDLLRAGLSTVDASTLQALAYASPLQFIAEYAPGTPEREAvGSASerotypeB TIAAAIWTNTMPSHLYNDFLRAGLSAADASTLQALAYASPLEF IVEYPPGTPEREAVGSA******************: :*****: .**************:** :*.************SerotypeD YREVQRYLTITGICI STVIVFMALSLRNPRLGDEQSLPDAEKLEKVPSEMSGATEETTETSerotypeA YREVQRYLTITGICLSTVIVFVSLTLRNPRLGDEQSLPDAEKLEKVASEVSGTTEETTKTSerotypes YREVQRYLTITGICI SSVIVLVALTLRNPRLGDEQSLPEAEKLEQVPSKMSGANNEGTEE**************:*:***:..*:*************:*****:**..**::* *:SerotypeD KQKNSerotypeA KHENSerotypeB TKQKNFigure 13: Amino acid alignment of the Sitip protein from serotypes A, B and D strains ofC. neoformansAlignments performed with Clustal W http://www.ddbj Serotype D strain B3501A Sitip (604 amino acids)(http://www-seciuence.stanford.edulgroup/C.neoformans/index.html)B. Serotype A strain 1199 Sitlp (604 amino acids) ( Serotype B strain W1v1276 Sitip (605 amino acids) (www.bcgsc.bc.caThe additional glutamic acid residue (E) in the C-terminus of serotype B Sitip has been bolded.107SerotypeDSerotypeASe’rotyp eBFigure 14: Similarity tree of the Sitip protein from serotypes A, B and D strains of C.neoformansSimilarity tree produced from alignment (Figure 13) in Clustal W.http://www.ddbj, C..)60SerotypeAvsD50•SerotypeAvsB40LlSerotypeBvsD30a, 20Z0--ExonIntronExonIntronExonIntronExonIntronExonIntronExon11223344556GeneFeatureFigure15:Comparisonof exonandintronnucleotidesequencesimilaritiesfortheSIT1geneinserotypesA,BandDstrainsof C.neoformansExonsandintronsweremanuallydeterminedinthenucleotidesequencebycomparisontotheS.cerevisiaehomologSIT1p1/Arn3pproteinsequenceandidentificationof C. neoformansexonlintronconsensussequencesdonor5’-GT(XX)GY-3’andacceptor5’-YAG-3’(SeesequencesinAppendixIVAa).NucleotidesimilaritywasdeterminedbycomparingsequencesusingtheBLAST2sequencesalgorithmatNCBI (,4and5).b. Identification of siderophore related genes in C. neoformans.The completion of total genomic sequencing projects for strains representing threeserotypes of C. neoformans (serotypes A, B and D) in conjunction with the growing databaseresources in other fungi, allowed an in depth initial in silico analysis to search for siderophorerelated genes in C. neoformans. A review by Haas (2003) summarized siderophore synthesis andtransport in fungi. The review included a table of fungi, detailing the presence or absence ofsiderophore related genes. These included structural genes for siderophore transport andsynthesis (OrnithineN5-monooxygenase and non-ribosomal peptide synthase) as well asregulatory genes of siderophore systems in fungi (transcriptional activators and repressors).Using this table as a guide, the C. neoformans databases were searched using all fullycharacterized fungal homologs involved in siderophore synthesis, transport and regulation. C.neoformans databases were searched for serotype A at Duke: (,D Stanford-(http://www-seguence. stanford.edulgroup/C.neoformans/index.html) or TIGR:( and serotype B at MSGSC ( Results ofthese searches are summarized in Table 15.Through this analysis, multiple siderophore transporter homologs were identified in C.neoformans: SIT]/ARN3 of S. cerevisiae (characterized in this study and found on chromosome 1of B3501A, JEC21 and H99) and ARN4/ENB] of S. cerevisiae and MIRB of A. nidulans.ARN4/ENB] and MJRB may be equivalent homologs resulting in one gene because they hit thesame genomic location in C. neoformans, where MIRB is the closest homolog (chromosome 6 ofB3501A, 7 of JEC21 and 8 of H99). There is also a putative homolog of MIRA of A. nidulans inthe serotype D strains (chromosome 4 of B3501A and 5 of JEC21) and MJRC in JEC21(chromosome 3). Other transporters do register significant hits, however the homologassignment has been given to the gene that has the lowest e-value for that genomic locationwhere more than one transporter hits a given location.The serotype A and D strains do not have putative homologs for the ornithinemonooxygenase enzyme involved in siderophore synthesis. They do have putative homologs fora non-ribosomal peptide synthase, however the e-values are fairly high and the top annotated hitis for an aminoadipate-semialdehyde dehydrogenase. The putative homologs are found on thesame chromosome as the MIRB homologs (chromosome 6 in B3501A, 7 in JEC21 and 8 in H99).110In contrast the serotype B strain WM276 (C. gattii) does have putative homologs for bothenzymes necessary for siderophore utilization. The assembly and annotation of the serotype Bstrain WM276 has not been assigned chromosome numbers at this time, therefore gene locationsand the number of homologs can not be accurately determined.Although serotype A and D strains of C. neoformans do possess putative siderophoretransporters, they do not appear to have homologs for both of the enzymes necessary tosynthesize siderophores, with the exception of the serotype B of C. gattii. The samephenomenon is noted in the fungi S. cerevisiae and C. albicans (Table 15) where transporters butnot synthesis enzymes are present. It is likely that these fungi capitalize on siderophoreproduction by nutritional competitors in their environment by possessing these transporters.These results are interesting because the more evolutionarily similar basidiomycete Ustilagomaydis can synthesize siderophores.In contrast to their similarities in siderophore related structural genes, C. neoformansdoes not have a homolog of the well characterized S. cerevisiae transcriptional activator AFT]. Itdoes however have a putative homolog of the transcriptional repressor GAF2/FEP] ofSchizosaccaromyces pombe (chromosome 1 of B3501A, 10 of JEC21 and 10 of H99). Thehomolog to this gene has been named CIR] (Cryptococcus Iron Regulation) and is beingcharacterized by post-doctoral fellow, Wonhee Jung, in the Kronstad laboratory. A secondputative repressor that is a homolog of Penicillium chysogenum SREP is present in B3501Achromosome 9 and JEC2 1 chromosome 1. An additional repressor homolog is also present onchromosome 1 of H99, however it is still most closely related to GAF2/FEP1 of S. pombe andtherefore may represent a second homolog of that gene. Like C. neoformans, U maydis has ahomolog for the transcriptional repressor called URBSJ, but not a homolog of the yeasttranscriptional activator AFT].111Table 15: Siderophore-related genes in C. neoformans.CHR= chromosome followed by e-value for the related database: tBLASTx for serotype A andD; BLASTn for serotype BSerotypn DC. nnofonoans Scrotype A- Serolyite B @BCGSCStr,tetnral Gnneo Organism Gene Seqoenee homolog (a) Stanford or TIGR DUKE DUMRC BLASTn only>YHL04OC CIte 8 CHR Staoford ln-60Sidnro1,ltorn Transporter ,1.acsbara,,,vco.vwac,Waiae ARNI CHR I TIGR 2,6r-106 CUR I 3n-77 No HitCHR 6 Stanford lc-30ENBt/ARN4 ‘YOLI58C Ctn tS CHR 7 TIOR 7.Oe-51? CHR8 30-3 t No HitCUR I Stanford 3n-89 No Hit/cootig_284 withSITI/ARN.1 >YEL065W Chr 5 CUR t TIGR 4.2n-t 15 CUR I 4n-89 onrotype D StTt IBLASTodR t Stanford e-75TAFt/ARN2 ‘YHLO47C Chr 8 9.So-93 CUR t TIGR 5n-37 CHR t 2n.35 No Hit.S’chiaoaaccarootycc.opo,,,ho ALO3 1534AL033t27Zt/13 12C0ARN I/CaStCaodeda a/b/stow TACHR4 Stanford 30-tOAaposgi/bcroidu/aaa airA ‘gi117062tl84lgblAY027565.tI CHR S TICS 2.5r—I9 CURS te-tI No HitCHR6 Stanford 4n-121ojrB sgil3O/158789] CHR 7 TIOR 5.30-133 CHRS o.IOb eontig_218 t.le—113Sgil3(6/5879lIgbIAYI35 152. t( CHR I Stooford Io-26ntird CHR 3 TICS 1,4n-40 CHR S 3e-22 No He?Aopcrg/llo.eJ’ooo’gatoo *Nouraspora eras at *Ornithine N’-monoooygenase Uall/ago,oavdio sidl >gilt70588lgblM98520.Ij No Hit No Hit eonlig_Itt tOe-SbSocohanooycea cerevislac-.Vchizasaccaroorvceapo,,shs- AL 138854Caw/ido o/hicaa.r.-lapergellus ole/it/oats stdA ‘gil327093921gb1AY2235t1 I No Hit No Hit No Hit>gil227792371dbjlAB07 1287..tsporgil/eaa srsoac dITA I No Hit No Hit COotig_t It 5.2r.12.4spcrge//oofo,olgalu.r *Nesrospora crania *..1.ipsrgi//o.s pot/a/sos 1.1859119CHR 6 Staoford 5n-06Nonribonomtd topside CUR? TIGR 8.20-18synthrino. Ust,lagscoasat/s oid2 >gi127316321gb(U62738.tI atoinaodipate-ootoiatdnlo’de CHt?53e-/t6 cootig III 7.7r-t0Saccharo,oyccx cerews,00--.Schtaaaaccaesoec-.spaohc. ALt38ttS4Cain?lila a/bleats,-CHR 7 TICS 5.3E-21antinoadipatn-tennatdnttyde.4apergi/ttceoldu/aoa sidC >gip327093941gb1AY2238l2.II detlydmgonasr CI-tR 83e-05 No Hit.itapo-rgillusJiso/gatos +Neoroapora crasaa *.‘lspor/T//oopa//o/aoa 1.185909Regulator-encoding genesIron-reaponsir’r agil t290691gbJL5-t05O.Itronooriplionat repreonor Use lags s,avd/s ob,ot promotnr No Hit No Ho No HitSaccharo,ovcoa cerceioiae->gi17034671gb1L29t?5t.tI CHR t Stanford Is-IS CHR tO7n-t8&hlaoaeesresro’ceapaoche GAF2/FEPI CHR tO TIGR I.4e-22 attd CHR I tn-t6 No Hit‘gi12t74532/NSbIAF52t/9?3,tl CHR I Stanford 2e-1l CHR tO3e-tJCaodola a/b/oars SFUI CUR tO TIOR I 2r-21 and CUR t 2n-t2 ooobg 45 2.2e.6>gi45852l2gbAFO95898.I dO Stanford 7n-10 CHR lIt Sn-Itt.‘tspeglhis ,bdo/sos amA CUR I TIGR md CUR I Sn-9 enolig 214 90-8Aapergilhsofa,oigatua *CHR 9 Stanford c 90-13 CHR tO to-t3Neut-oapca-aeraara arc cgii3552tl27gbAFO87I30,IJ CHR tO TIGR 2n-08 andCHR I 2e-t18 sootig_214 tin-ISi-gi1l5179t51gb1U484t4.II CHR 9 Stanford 4e-I5 CUR It) 6r-l3Pc-old//la,,, chysagc-aost amP CHR Itt TIOR l.8e-07 and CHR I 4r-8 No Hit>giIt3620I72IembA3309O5l. CHR 9 Stanford 4n-t5 CHR to 3e-t4&cryliaciarc-c.a birt II partiat CUR I/I TICR 1.8e-tS and CHR I 4e-9 No Hittrait reoponoivntranscriptional activator Saccharo,,,cces co-,oria’iao AFTI >YCLO7IW Cltr 7 No Hit No Hit No HitSaeeharo,ovco.,cc.,-cclslae AFr2 ‘YPL2O2C Chr 16 No Hit No Hit No HitCandle/a a/b/coos *.1sjorg,l/a.a sic/a/any-.4ope,iltccofcaoigotoa-Neoraapara crassa-112c. Comparison of Sitip to fungal homologs.C. neoformans Sitlp was compared to orthologs in the other fungi by amino acidsequence alignment and similarity trees (Figure 16 and 17). Interestingly, we found that the C.neoformans homolog clustered with the ascomycetes S. cerevisiae and C. albicans and not themore evolutionarily similar basidiomycete U maydis. These results are concordant with thefindings from homolog searches of siderophore related genes (Table 15), where C. neoformans ismore similar to S. cerevisiae and C. albicans than U maydis with respect to the presence orabsence of siderophore synthesis genes. The fungi clustered in a pattern that relates to whetherthey appear to produce and take up siderophores, as is the case with U maydis and Aspergillusnidulans (Haas, 2003), or whether they only have homologs for siderophore transporters but notsynthesis such as C. neoformans, S. cerevisiae and C. albicans (Haas 2003). The closesthomolog, Arn3p/ Sitip of S. cerevisiae has 39 % identity and 55 % positive amino acid matchesto C. neoformans Sitip (resulting in an e value of-.l 12).113Figure 16: Amino acid alignment of the Sitip protein from the serotypes D strain B3501Aof C neoformans with fungal homologs.Alignments performed with Clustal W. http ://www.ddbj!clustalw-e.html Fungalhomolog sequences and GenBank Accession numbers can be found in Appendix IVAb. C.neoformans SIT] sequence was from the Serotype D strain B3501A (http:I/ Coding sequence was manuallydetermined. (Appendix IV-Aa).An-Aspergillus nidulansCa-Candida albicansCn-Cryptococcus neoformansSc-Saccharomyces cerevisiaeUrn- Ustilago maydisCLUSTAI W (1.83) Multiple Sequence AliqnmentsSequence type explicitly set to ProteinSequence format is PearsonSequence 1: CnSIT1p 604 aaSequence 2: ScSITlp/ARN3 628 aeSequence 3: ScARN1 627 aaSequence 4: ScARN2/TAF1 637 aaSequence 5: ScARN4/ENB1 606 aaSequence 6: CaSITA/CaARN1 604 aaSequence 7: AnmirA 609 aaSequence 8: AnmirB 604 aaSequence 9: AnmirC 607 aaSequence 10: UmSitl 583 aaCLUSTAL—Alignment file created [clustalw.aln]ScARN1 MESVHSRDPVKEEKKHVFMGMEHELNPETHNDSNSDSYGLPQL.nA V1flJJflScARN2 /TAF 1 MIEVPEDNRSSQTKRKNTEKNCNELMVDEKMDDDSSPR DEMKDKLKGCaSITA/CaARN1 MTSYQSSNNHSSEEDKHLSGDEKTFSPSDIVEKAIVEScSIT lp/ARN3 MDPGIANHTLPEEFEEVVVPEMLEKEVGAKVDVKPTLTTSSPAPSYIECnSIT1p MPDQLYPEYEIJQRAISNTKNADDSTAFGScARN4/ENB1 MLETDHSRNDNLDDKSTVCYSEKTDSNVEKSTTSGLRRIDAnmirC MPLLEPSATAYGTFGDMRPDTEDEGERLLTDGYVSDDDCSAVTSVDUmSitl MSRPFDAENVEHRDTSHDSMDMTDQLAAnmirA MALDDISAVPKGALDTDPAVERPPPLLDAORSDSEAnmirs MTIGSKFSLLAGTRKTDGPTEI SASSPPDVETPSAEKTATASAGNKEVGINDNSSDEALPScARN1 NRSLI IQQTEI IG-SAYNKWYLQAILLLSAFICGYGYGLDGNIRYIYTGYATSSYSEHSLScARN2 /TAF1 TKSLIIRKSELMA-KKYDTWQLKAIFLFSAFICTFAYGLDSSIRGTYMTYAMNSYSAHSLCaSITA/CaARN1 EKSICVEKA3ILANQWKHTFWFKJLLGFSAFLCGYAYGLDSQTRYVYTAYATASWSEM5LScS IT lp/ARN3 LIDPGVHNIEIYA-EMYNRPIYRVALFFSLFLIAYAYGLDGNIRYTFQAYATSSYSQHSLCnSIT lp EKSPGVRRIEI IA--ASFTTWHRWVLFI SVFFMSCNYGLDGSVRYTYQAEALSELGTSAQScARN4 /ENB1 AVNKVLSDYSSFTAFGVTFSSLKTALLVALFLQGYCTGLGGQISQSIQTYAANSFCKHSQAnmirC SVQEGVRKIEAIN--ITWTTRSLVIAYI SIFLMAFCTSLEGQTIMSLSAYATSAFSKHSLUmSit 1 DKQICVVAAEAGR---TVADWTLWMGILGIALVAYLYGLDNNTMWAWQTYATTSFNDYPAAnmirA RLQPCVKRAEMLR--KGWTRQGLI IAFTGLFLATLSINFGDYSTQVYVPYATSAFKQMSAAnmirB SQHVQTGVQKIQAVTLVWSKWSLVAVFCLLWLVTLANGFRQS ILYSLTPYATS SFQSMSL114SCARN1 LSTINVINAVVSAASQI IYARLSDVFGRLYLFISAVILYVVGTI IQSQAYDVQRYAAGAIScARN2 /TAF1 ISTVSVIVLMISAVSQVIFGGLSDIFGRLTLFLVSIVIJYIVGTIIQSQAYDVQRYAAGAVCaSITA/CaARN1 LTTVNAITGVVAAASQPVYARLSDVLGRLELFIVAVLFYVVGTI IECQSPTINAYVAGAVScSIT1p/ARN3 LSTVNCIKTVIAAVGQIFFARLSDIFCRFSIMIVSIIFYSNGTIIESQAVNITRFAVGGCCnSIT1p VSTVTVVRSIVAAAAQPAFAKVSDYFGRVSILI ISVIIYVVGTIVTATSTNLAAFCGGSVScARN4 /ENB1 VGSINTVKSIVASVVAVPYARISDRFGRIECWIFALVIJYTIGEII SAATPTFSGLFAGIVAnmirC I STVLVVQNVVNAVIKPPMAKIADVFGRFEAFCVSILIYVLGYIQMAASTWVQTYASAQIUmSit 1 YTAVSVVQAVI IAVGKFPIAKLADVFGRAQAYALSVFLWVIGFVIIAIAQNTRYVAGGTVAnmirA MSAARVVGNITRIAAYPI IAKLGDVFGRAENFILS IVFQAVGYAIYAGCKNVGQYIAGGIAnmirB LTVINIVSSAFIVSAIYIPVAKVVDVWGRAEGWLVMVGLSTLGLIMMAASKNLETYCAADV* ** :*SCARN1 FYNAGYVGVILILLI ILSDFSSLKWRLLYQFVPTWPFIINTWIAGNITSRANP---VVNWSCARN2 /TAF1 FYYVGLVGVk4LQVVLMLSDNSSLKWRLFYTLIPSWPSIITTWVSGSVVEAANP---LENWCaSITA/CaARN1 LFQIGYSGI I IMLLFITJSDFSSLRWRLFFTLCPSFPFIINTWISGNVTAAVG TRWSCSIT1p/ARN3 FYQLGLTGI ILILEVIASDFSNLNWRLLALFIPALPFIINTWISCNVTSAIflCnSIT1p tYQFGYTGVQLLVEVLIADVTSLRSRLLFSYIPATPFLINAWISGNVASAVLT---HSTWScARN4/ENB1 IQQFGYSGFRIJLATALTGDLSGLRDRTFAMNIFLIPVIINTWVSGNIVSSVAGNVAPYKWAnmirC FYAAGSTGLQILQQVFIADSSSLLNRALLAILPELPFLVTVWIGPT VV E---NSSWUmSit 1 IYAFGNTGVQIMQQIVLADYI STKWRGAAIGLVSLPYVINFWASAQIYPKIVA----VNWAnmirA FEAIGSTGFGLTQQVFVADVTNIJINRAVWSTLPDSLTVIPALYLGTEIAEAVLE--KNEWAnmirB FYSVGFAGNNYILCVLAADITNLRNRGIAFAFTSSPYMITAFAGSKAAEJ<FLVN---VNW* * * * *ScABI41 SWDVGMWAFIFPLSCVPIVLCMLHMQWRARKTPEWHALKGQKSY--YQEIJGFIKILRQLFScARN2 /TAF1 SWNIANWAFIFPLCCIPLILCMLHMRWKVRNDVEWKELQDEKSY--YQTHGLVQMLVQLFCaSITA/CaARN1 KWGIGMWAFILPLSCIPLVCCMIHMRWLAGKTEEWRVFRQRKTK--FQELGVAGFSKYLFSCSIT1p/ARN3 KWGIGMWAFILPLACIPLGICMLHMRYLARKHAKDRLKPEFEAL--NKLIcWKSFCIDIAFCnSIT1p GWGIGMWAI IFPVTVIPLLFSLIQAEWRABRKGLLREIP---SP--LRTFGDRHMWADIFSCARN4 /ENB1 RWGYGIFCIIVPISTLILVIJPYVYAQYISWRSGKLPPLKLK EKGQTLRQTLWKFAAnmirc RWGYGMWSI IIJPASFLPIJAIJSLLLNQRKAKRLNLIKERP HHRRGFVAAVRRTWUmSitl RWGPGFFSICAPVAAFSI IFALAINQRRARAMGIjVPSRP YKHMSFLAAVYNFLAnmirA RWGFGNWAI IEPVC SVIJLVCTMLYYQKRARKDPSPAEFASEPTERNVDDGWWKRIYNLVWAnmirs RWGFGAFAI IFPFVASPVYFVLKVGLNRAEKQGIIQPRIJR SGRTLSQNFKYYF•*ScARN1 WMLDVVGVLLMGCSLCCILVPLTLAG-GVKTTWNDSRLIGPFVLGFVLIP--ILWIWEYRScARN2 /TAF1 WKLDVVGVLLFTAGVGCILVPLTLAG-GVSTNWRNSKIIGPFVLGFVLVP--GFIYWESRCaSITA/CaARN1 WRLDVIGLLLLTVSLCCLLVPLTLAG-GIRETWKKAHI IVPIVIGGVLIP--VFLLWEGYScSIT1p/ARN3 WKLDI IGMLLITVFFGCVLVPFTLAG-GLKEEWKTAHIIVPEVIGWVVVLP-LYMLWEII(CnSITlp WQIDLVGLLLLAAVLALILLPFTLAG-GVASIWRTARVIAPLvvGFvvALP-LFvFWELKSCARN4 /ENB1 DDINLIGVILFTAFLVLVLLPLTIAG-GATSKWREGHIIAMIvvGGCLGF--IFLIWELKAnmirc YDLDIFGLALLSAAVTLILVPLTLAA-NTIQJGWK5NSIVAMIVIQVVCLI LLPFWETSIcI(UmSitl IDIDALGLFLICVGFLLILLPvNLAK-LQPNGWSTGWIIAMLVICGVMLIS---FCVWECAnmirA VQLDAFGAILLLLGL5LFLVPL5LTGSGN5DDWBRGSFIAJ4LVLCVVIFVA--FLAWD’pWAnmirs FAFDIPGVILLAGCLTVFLLPflLAT-RAPNGWK5DYIIANIVTGFVVMVL--FVLYQAY* *: *:*:: *•: * *ScARN 1 FARDPILPYRLVKDRAVWSSMGI SFLIDFIflNAAD-flflVMIVAVNE5VK5ATRIAflScARN2 /TAF 1 LALVPFAPFKLLKDRGVWAPLGIMFFICflYQNAAG-YLYTILWAVDE5ASSATRI 11ThCaSITA/CaARN1 GARDPI IPLHLMEDRGIWSVV5I5LFFDFVFAVE5N-FLYTVId4VAVNESQ55ATRIASLScSITlp/ARN3 YSRHPLTPWDLIQDRGIFFAI,LIAFFINFNWYNQGD-YMYTVIJVVAVHESIKSATRITSLCnSIT1p VARHPMLPFRILKDRQVLASLFIANLLNTAWYTQGD-YLYYTLLVAFDRDIISATRvQNIScARN4/ENB1 FAZNPFIPRVYLGDPTIYVALLMEFVWRLCLQIELE-YLVTVIJMVAFGESTLSAQRIAQLAnmirC LAPKPLLSLHLLKQRTALAGCCLAFFYFMAFYFSVQPYLY5yLQVVQGyDVATAGRVTQTUmSit 1 VAPKPILNRRWRLI4HDVHFATAIGFFDFFSFYASWVPAYYWSLIVI4G-yDT4TGATyFSNCAnmirA CAKKPF IPYRMIKNRTVAAACLJflILDFflIYSVFSV-FFTSYLQVAAHHGAQPATRIDN5AnairB WAPQPFLKYEFLTNRTVLGACLIDATYQNSYYCWN5-YFNSFLQVVCNLPVAEAGYVG5T*: * *115SCARN1 SSFVSTVASPFFAI.LVTRCTRLKPFIMFGCALWMVANGLLYHFRG----GSQSHSGIIGASCARN2 /TAF1 YSFVTAVVAPFLGLIVTRSSRLKSYIIFGGSLYFITNGLFYRYRS----GQDADGGIIAGCaSITA/CaARN 1 SSFVSVVTGFIFGLFVVYFRRLKGFVVFGCANWNVAFGIMYHFRS----QLHANAGIICCScSIT1p/ARN3 YSFVSVIVGTILGFILIKVRRTKPFIIFGISCWIVSFGLLVHYRG-----DSGAHSGIIGSCnSIT1p YSFTSVVIGVCLGLI IRKVRRLKWFIVAGTLLFVLAFGLLIRYRGGY--SI SDFAGLVAAScARN4 /ENB1 YNFLQSCTNIVVGIMLHFYPHPKVFVVAGSLLGVIGMGLLYKYRV----VYDGISCLIGAAnmirc FAFTSTIAAFGVS ILIKYTRRYRVYVTLGCVIYMTGLLLMLLYRK----EGSSPLQVLGTUmSiti QSLALTVFGIAAGFLSLGTKNYKWIMISGACIRLLGIGLMIKYRS----SGSSNVQAVFPAnmirA LRVAFQVAGI FAAYFNRFTKRSQVWVFTGVPLCVIGMGVISLYLVDMGEGRVGNEAAFVTAAnmirB FQVVSGVLIJFMVGFAIRKTGYFRWLLFIGVPLYI FAQGLMIHFRQ----PNQYIGYIVMC*ScARN 1 LCVWGVGTTLFTYPVTVSVQS-AVSHENMATVTALNYTLYRIGSAVGSAVSGAIWTQTLYScARN2 /TAF1 MVIWGLSSCLFDYPTIVSIQS-VTSHENMATVTALNYTVFRIGGAVAAAISCAIWTQSLYCaSITA/CaARN1 MCLMGFGTGFFSYPINVSAQS-CVSHEHMAVISSALYTTYRIGYAVGSSVAGAIWSQMLYSCSIT lp/ARN3 LCLLGFGAGSFTYVTQASIQASAKTHARMAVVTSLYLATYNIGSAFGSSVSGAVWTNILPCnSIT ip EVVLGIAGGLFPYPTQVMIQS-AVQHERTAVVTSLYLASYSVGSALGNTIAAAIWTNTMPScARN4 /ENB1 EIVVCIAGGMIRFPMWTLVHA-STTHNEMATVTGLLMSVYQIGDAVGASIAGAIWTQRLAAnmirC QVIVGNGGGLLNVPVQLGVQA-SASHQEVAAATAMFLTSMEMGGAVGAAI SGAVWTHNIPUmSit 1 QVLQGMGGGFLGITLQVAAQV-SVRHQOVATVTAYFLLLTEMGCACGNALVGAVQTNVLPAnmirA KSLIGIGRGFYQTASQVSVQA-KVSRGEVSVVTAVFFAAMS IGGAIGTSVAGAIWRSTIJPAnmirB EIFISIGGSIFVLLQQLAVLV-AVDHQYVAAALAVLFISGGIGGAVGNAISGAIWTNTFL.* *SCARN1 KQILKRNG DVAIJATTAYESPYTFIETYTWGTPQRNALNNAYKYVQRLETIVASCARN2/TAF1 PKLLHYMG DADLATAAYGSPLTFILSNPWGTPVRSANVEAYRHVQKYEVIVACaSITA/CaARN1 SRLVKYLG DSTLATSVYTDPYTFIAQYVWGTPEREAAVKAYGEVQRVLMSVCScSIT1p/ARN3 KEI SKRI S DPTLAAQAYGSPFTFITTYTWGTPERIALVNSYRYVQKILCIICCnSIT1p SHLYNDFIRAGLSTTDASTLQALAYASPLQFIIDYPPGTPEREAVGSAYREVQRYLTITGScARN4 /ENB1 KELIQRIJG SSLGMAIYKSPLNYLKKYPIGSEVRVQMIESYSKIQRLLIIVSAnmirC RKJJNLYLP DEYKSEAGAI FGKLTKALSYEMGTPVRSAINRSYQETMNKLLVLAUmSiti GYYEKYLP MLNATQRAAIYASPFTAVSQYPIGTPERTGMINAYNDYMRILLIVAAnmirA PKLAQHLP AELKOQAQAX FGS IVVAQKYEVGTPARDAIDMCYRQSQRMLAIAAAnmirS PALMRNLP ESAKARAVAIYGDLRVQLSYPVNSPERIAIQESYGYAQARMI1* *SCARN 1 LVFCVPLIAFSLCLRDPKLTDTVAVEYIEDGEYVDTKDNDP ILDWFEKLPSKFTFKRE --ScARN2 /TAF 1 LVFSAPMFLLTFCVRDPRLTEDFAQKLP-DREYVQTKEDDPINDwIAKRFAKALGGHKKDCaSITA/CaARN1 IAFVAPMIVSALFMRDHKLTNEQSLEDVEKQEEKDSLANFFGNFKKKRVAVScSITlp/ARN3 LVFCFPLLGCAFNLRNHKLTDSIALEGNDULESKNTFEIEEKEESFLKNKFFTHFTSSKDCnSITlp ICISTVIVFMALSLRNPRLGDEQSLPDAEKLEKVPSEMSCATEETTETKQKNSCARN4/ENB1 ISFAAFNAVLCFFLRGFTVNKKQSLSAEEREKEKLKIKQQSWLRRVIGYAnmirC LLATLPLIPLSLLMSNYKLOKMSESSDHDDASPRNGLGPCERAKRTUmSit 1 IVLAVPPI ILGLVVNDRKLNDNQNCVSNELACMRRMSSENESTDNKAnmirA LAALAPMLIIMFFLENVPLTDETflIELHGNREAVKKNS5GQEGKEAR5AnmirS TGLMAIMFIWMFMVKNYNVKNM5QTKGMVFSCARN1SCARN2 /TAF1 LQNPNRICVRKNDLCaSITA/CaARN1ScSIT1p/ARN3 RIOCnSIT1pSCARN4/ENB1AnmircUmSitiAnmirAAnmirB116Figure 17: Similarity tree of the Sitlp protein from the serotypes D strain B3SO1A of C.neoformans with fungal homologsSimilarity tree produced from alignment (Fig. 16.) in Clustal W.http://www.ddbj nidulansCa-Candida albicansCn-Cryptococcus neoformansSc-Saccharomyces cerevisiaeUrn- Ustilago maydis117ScARN1ScARN2/TAF ICa,SITA/CaARN ISoSITIp/ARN3CnSITIpScARN4/ENB IAmm’irAAnmirBAnmirCUmSitiFigure 17: Similarity tree of the Sitip protein from the serotypes D strainB3501A of C. neoformans with fungal homologsSimilarity tree produced from alignment (Figure 16) in Clustal W.http:/Iwww.ddbj!clustalw-e.htmlAn-Aspergillus nidulansCa-Candida albicansCn-Cryptococcus neoformansSc-Saccharomyces cerevisiaeUrn- Ustilago maydis117B3501A JEC2I H99rRNAFigure 18: Transcript levels for the SIT1 gene in culture media with low or high iron levelsfor strains JEC21, B3501A and 1199.Cells were grown in low iron media, ElM (designated-) or iron replete media LIM+Fe(designated +). The Northern blot was prepared from total RNA isolated from cells grown inElM or LIM+Fe medium at 37°C for 6 hours. Blots were hybridized with a DNA fragment fromExon 6 of the SIT] gene of C. neoformans.A. SIT] transcript levelsB. rRNA (18S and 28S) bands as a loading controld. Expression of SIT] in low iron and iron replete conditions.The SIT] gene was identified by SAGE as having an elevated transcript in cells grown inlow iron vs. iron replete conditions (Lian et a!., 2005). The SIT] transcript was detected byNorthern blot during growth of cells in low iron conditions but absent during growth conditionsof high iron in all strains (Fig. 18). These results suggest that the Sitip may be important in theacquisition of iron when C. neoformans encounters a low iron environment. These results wereconsistent with SAGE analysis in B3501A (Lian et a!. 2005) where a 3.33 fold increase of theSIT] transcript was noted from cells grown in LIM vs. LIM+Fe.ASITIB+ + +118B. Construction of Siderophore Transporter Mutants.a. SIT] gene and Sitip protein structure.SIT] is 2144 nucleotide bp in length in serotype D and serotype A and the predicted Sitippolypeptide 604 amino acids in length. The gene contains six exons and five introns (Figure1 9A), and the encoded polypeptide has 13 predicted transmembrane regions by TmHMM( The predicted structure includes an extracytosolicN-terminus, 6 regions in the cytosolic face, 6 regions in the extracytosolic face and a cytosolic C-terminus (Figure 19B). There is an unusually large extracytosolic loop at the C-terminus (58amino acids). The Am family of protein in S. cerevisiae (of which Sitlp is a homolog) possess alarge loop at the C-terminus (44 amino acids) that has been shown to be involved with thelocalization of the protein, more specifically the cycling of the protein from the plasmamembrane to endosomes (Kim et al., 2005). The loop in C. neoformans Sitlp thereforerepresents a strong candidate for site-directed mutagenesis in the future to investigate itspotential role in the localization of the protein.119AFigure19: StructureoftheSIT1geneandSitipproteinA.SIT1genestructureB.PredictedtopologyofSitip48bp62bp6lbp113174bp185bpB84bp I7jTA(AIG)pJ C197bp240bp469bp230bp494bpEExonIlntron2144nucleotidebplinus195-203351 -369134-136488-546160-17166-110Extracytosolicface314-332393-398445-464604aminoacids(aa)227-262PredictedTransmembraneRegionsC-terminusIc___faceb. Construction and confirmation of null mutants.Null mutants were constructed in three strain backgrounds to investigate the function ofSIT]: serotype D strains JEC21 and B3501A and serotype A strain H99. The entire codingregion (6 exons; 5 introns) was deleted by using a knock out cassette constructed by PCRoverlap (Davidson et al., 2002). A URA5 marker was used for the serotype D strains (Varma eta?., 1992). A modified overlap strategy was used (Yu et a?., 2004) with a neomycin marker forthe serotype A strain. The structure of the knock-out constructs for serotypes A and D includingprimer locations is detailed in Figure 20. One homologous integrant was characterized forJEC21 and two independent homologous integrants were obtained for both B3501A and H99.Mutants were confirmed by colony PCR (Appendix IVC) and Southern analysis (Figure 21 and22). PCR confirmation showed homologous integration by using one primer upstream of theintegration site and one primer internal to the marker (Appendix IVC). Southern analysisunequivocally determined that a single homologous integration event had occurred in the strains,ensuring that no additional ectopic integrations resulted from the transformation (Figure 21 and22). For serotype D, three enzyme combinations were used to differentiate wild-type and mutantalleles: EcoRl/NcoI, Seal and StuI. A schematic diagram of restriction enzyme sites in theknock-out construct and the wild type gene, as well as images of Southern blots, can be found inFigure 21. For serotype A, FspI was used to distinguish the wild-type and mutant alleles (Figure22). sit] mutants were reconstituted with the wild-type SIT] gene cloned into a vector containinga marker for resistance to nourseothricin (NAT).121A SITIUPSITIA-SITIBSITIC853bpColonyPCRScreen1411bpSITNFSITAI-SITA2SITA31151bpURA51760bp1—63283355bpNeomycin1876bp1-___4189bpMarkerSITIESITID-SITIF1-B742bpSITA5SITA4-SITNR1- SITA61162bpSITAUPColonyPCRScreen1310bp5’Flank3’FlankFigure20:StructureoftheEssitlknockoutconstructsPrimersforknockoutconstructionandcolonyPCRscreeningaredenotedwithblackarrows.A.serotypeDandB.serotypeAStul Seal‘r4rNcoI4r5’742bp URA5176Obp 3’ 853bpB.EcoRIProbeSIT1 Genomic Sequence3’853bpA. EcoRlStul ScaTAsiti ConstructStul Seal Seal Ncol Stul4r ‘4r ‘4r5’ 742 bpProbeSIT] 2290 bpD.—L)Lfl-,- —iD rJ)• IIi) ).- .-C N Nc/) Z1I IC.4571 bpE.->6000 bpm—40010-—2723 bpNc-) C c-ci r1_e2218 bpFigure 21: Confirmation of the Asitl knock outin the serotype D strains B3501A and JEC21.Southern analysis to confirm homologousintegration of the tXsit] knock out construct. TheSouthern blot was hybridized with the 5’ arm of theAsit] knock out construct. A. Enzyme recognitionsites in Asit] construct. B. Enzyme recognition sitesin the SIT] gene. C. Southern blot of genomic DNAdigested with: C. EcoRl/Ncol D. Seal and E. StuIand probed with SIT].‘i bp123iXsitl Constructr’B.FspIProbeSIT1 Genomic SequenceFspI5’ 1151 bp SIT] 2258 bp 3’ 1162bpProbeC.7000 bp___.•* UndigestedDNA2000 bpFigure 22: Confirmation of the 1!isitl knock out in the serotype A strain H99.Southern analysis to confirm homologous integration of the iXsit] knock out construct.The Southern blot was hybridized with the 5’ arm of the Asit] knock out construct. A.FspI enzyme recognition sites in Asit] construct. B. FspI enzyme recognition sites in theSIT] gene. C. Southern blot of genomic DNA digested with FspI and probed with SIT].A.FspI¶4, FspINeomycinl876bp 3’ 1162bp124C. The role of SIT] in Iron Acquisition.a. Growth in various iron conditions.The disruption of SIT] resulted in greatly reduced growth of C. neoformans under lowiron conditions (LIM), or when a siderophore was provided as the sole iron source (LIMcontaining the chelator bathophenanthroline disulfonic acid [BPDA] supplemented with thesiderophore deferoxamine) for strains B3501A (serotype D) and H99 (serotype A) (Figure 23A,B,C,D). No strains are able to grow in LIM and the chelator BPDA alone. Interestingly, theJEC2 1 sit] mutant did not show reduced growth in comparison to the wt strain in these growthassays (Figure 23 A,C). All strains showed similar growth patterns in LIM+Fe, although growthwas slightly delayed for the sit] mutants. These strains eventually reached wt levels by 36 hours(Figure 23. E,F). These results indicated that Sitip plays an important role in iron acquisition inthe low iron environment and in the use of a siderophore as a sole source for iron acquisition forstrains B3501A and H99, but not JEC21. It is possible that JEC21 has additional or redundantmechanisms for acquisition of iron. Another interesting observation is that the serotype D strainsattained more robust growth than serotype A in LIM and deferoxamine (Figure 23 A,B,C,D).The strains were tested to determine whether loss of Sitip resulted in a more generalized uptakedefect by growing the cells in minimal media with urea as the sole nitrogen source. sit] mutantsin the serotype D strains JEC21 and B3501A but not the serotype A H99 strain had retardedgrowth in this media at 24 hours (0D600 of 8 vs. 12). However the OD of the culture wasequivalent for all strains by 48 hours (0D600 of 12) (data not shown). This result may suggestthat a defect in Sitip has a general influence on membrane transporter function in the serotype Dbackground.125Figure 23. Disruption of the SIT1 gene significantly affects growth in LIM andLIM+BPDA+Deferoxamine in B3501A and H99, but not JEC21.All strains have comparable growth in LIM+Fe. A comparison is shown for serotype D strains:JEC21 wt, the SIT]:: URA5 disruption mutant (je-si), reconstituted SIT]:: URA5 + SIT] (jesiRl), B3501 wt, two SITJ:. URA5 disruption mutants (b-slO and b-s42) and reconstitutedstrains SIT]:: URA5 + SIT] (b-s42R1 and b-s42R2) in LIM (A), LIM+BPDA+Deferoxamine (C)and LIM+Fe (E). A comparison is also shown of serotype A strains H99 wt, two SIT]::NEOdisruption mutants (h-si and h-s2) and the reconstituted strain SIT]::NEO + SITJ (h-slRl) inLIM (B), LIM+BPDA+Deferoxamine (D) and LIM+Fe (F). These growth experiments representthe average of three trials. Raw growth curve data can be found in Appendix IVE.126A13—+—JEC21(wt)12—R—je-sl(sitl::URA5)11—-—je-s1R1(sitl::UPA5+SIT1::NAT)—f—B35O1A(wt)10*—b-slC(sitl::URAS)9——-—b-s42(sitl::URAS)E8——+-—b-s42R1(sitl::UR.A5+SIT1::NAT)—b-s42R2(sitl::URJk5+SIT1::NAT)- N)0 D6-053 2 1-0061218243036424854HoursccCzI—U’)UIL.0+(N000UJ Ui;-; ;‘; ‘to4-’ 4-’ %—Cfl (it ,-4.— —0” - r.4 -rcc0f (N - o cc N ‘.0 If) (N C-I -l v-IWU 009 aoC12813—+—JEC21(we)C12———je-s1(sitl::URAS)11-—*—--je-slP.1(sitl::URAS+SIT1::NAT)10—E—B35O1A(wt)—*—-b-slO(siti::URA5)9—s——b-s42(siti::URA5)E8—+—b-s42R1(sitl::URA5+SIT1::NAT)o7—b-s42R2(sitl::URA5+SIT1::NAT)-0 1.0CD051061218243036424854Hours13D12 11H99Ih-si(sfti::NEO)1Ah-s2(sitl::NEO)Xh-slRi(siti;:NEO+SIT1::NAT)9E87.0O6.054 3 2 1-01—F061218243036424854Hours13•JEC21(wt)E12•je-si(sitl::URA5)Aje-siRl(sitl::URA5+STT1::NAT)11XB3501A(wt)101Xb-sb(sitl::URA5)•b-s42(sitl::URA5)Iib-s42R1(sitl::URA5+SIT1::NAT)8b-s42R2(sitl::URA5+SIT1::NAT)E a-0C)-04 3 2 1*0T061218243036424854Hours13 12—+—H99F———h-sl(sitl::NEO)11—*-—h-s2(sitl::NEO)10———h-s1R1(sftl::NEO+SIT1::NAT)9 805 43-/2////1/0061218243036424854Hoursb. Siderophore utilization.Siderophore utilization assays were performed on iron free plates with deferoxamine tofurther determine whether a siderophore could be used by C. neoformans to acquire iron when noother source was available. Zones of growth were apparent for all wt and reconstituted strainsbut no zones of growth were present for the sit] mutants (Figure 24. A,B,C). These siderophoreutilization assays showed that sit] mutants in all strains were unable to use deferoxamine as asole source of iron. There is a notable difference in the JEC2 1 results in the siderophoreutilization assay vs. the liquid growth curves where no growth was noted on the plate assay andwild-type growth was present in the liquid medium. Growth of the JEC2 1 sit] mutant was notedon the plates after 7 days incubation indicating again that this strain may have additionalmechanisms to acquire iron from a siderophore that are not expressed on agar plates. The loss ofthe high affinity permease, FTR], in serotypes A (H99) or D (ATCC 24067) was alsoinvestigated to determine if it led to a change in siderophore utilization. Wild type, null andreconstituted strains were tested in the assay and it was found that all strains could usesiderophore-bound iron as a sole source (data not shown). All experiments were repeated threetimes to ensure consistency in the results.133Asiti Asitl + SITIFigure 24: siti mutants are unable to use a siderophore to acquire ironin all strains as shown by siderophore utilization assays.To test for siderophore utilization, a culture of 200 iL of 1 o CFU/mL cells werespread on LIM plates containing 550 tM BPDA. A disk containing 10 iL of 100 tMdeferoxamine was placed in the center of each plate and the cultures were incubatedtwo days at 30°C. A. JEC21 B. B3501A and C. 1199.wtABC134D. Involvement of the cAMP Pathway in Siderophore Utilization.In S. cerevisiae, the cAMP pathway has been implicated in the regulation of iron uptakegenes (Robertson et al., 2000). It was shown that one of three PKA catalytic subunits, Tpk2p,negatively regulated a number of genes involved in iron uptake including the ARN3/SIT] gene.We were therefore interested in the affect of the cAMP pathway on siderophore utilization in C.neoformans. Serotype A and serotype D cAMP mutants (defective in the catalytic subunit ofPKA (PKAJ) or the regulatory subunit (PKRJ) were tested in the siderophore utilization assay todetermine if specific components were necessary for the utilization of deferoxamine as a solesource of iron. The results suggest that FKA 1 was necessary for siderophore utilization in theserotype A H99 background (Fig. 25). That is, no growth with deferoxamine was present for thepica] mutant (catalytic subunit knockout of PKA) and growth was enhanced in the pkr] mutant(regulatory subunit knockout of PKA-constitutively active) indicating that siderophore use ispositively regulated by the cAMP pathway in this background. This relationship is opposite tothat observed in S. cerevisiae (Robertson et al., 2000). No difference in growth was noted forthe pka2, pica] + PKA1 (data not shown) mutants in comparison to wild type. The gpal mutanthad slightly reduced growth in comparison to the wt type strain (data not shown). GPA] encodesthe alpha subunit of a heterotrimeric GTP binding protein that is involved in the induction of thecAMP pathway and is epistatic to PKAJ (Aispaugh et al., 1997). A reduction in growth would beexpected in the gpal mutant where PKA 1 is necessary for siderophore utilization. In strikingcontrast to the serotype A results, no significant difference in growth compared to wild-type wasnoted for mutants of the cAMP pathway in the serotype D, strain JEC2 1 background for each ofmutants (pica], pka2, pkalpka2 [data not showni or pkr]) (Fig. 25). These results suggest thatcAMP pathway is involved in siderophore utilization in serotype A but not serotype D, at leastfor the representative strains tested here. The cAMP mutants were checked for presence of theappropriate marker and a subset of published phenotypes to confirm that they retained theirexpected defects in cAMP signaling (Alspaugh et a?., 1997; D’Souza et a?., 2001; Hicks et a?.,2004). Specifically, the strains were tested for melanin and capsule production where in serotypeA, melanization and capsule was reduced for pica], unchanged for pka2 and increased for pkrl incomparison to the wt strain. In serotype D, melanization was unchanged for pka], reduced forpka2 and slightly increased for pkr] in comparison to the wt strain. Capsule size was similar in135all the serotype D strains tested. These results are concordant with published phenotypes forthese strains. All experiments were repeated three times to ensure consistency in the results.Figure 25: The cAMP pathway is involved in siderophore utilization in serotype A but notserotype B strains.To test for siderophore utilization, a culture of 200 iL of 1 CFU/mL cells were spread on LIMplates containing 550 iM BPDA. A disk containing 10 L of 100 p.M deferoxamine was placedin the center of each plate and the cultures were incubated two days at 30°C. A. JEC21 B. H99.wt Apkal Apka2 ApkrlAB136E. SIT] and Melanization.a. Production of DOPA melanin.Production of melanin by the enzyme laccase is important for virulence in C. neoformansand a variation in production or deposition may alter the pathogenicity of the fungus. Laccaseexpression is activated through the cAMP pathway, the same pathway that is thought to beactivated in low iron conditions. Therefore, the sit] mutants were tested for melanin productionand a striking phenotype was noted in the serotype D strains JEC21 and B3501A. Specifically,evaluation of melanin production on DOPA medium plates revealed a variation in melanizationbetween wt and sit] for strains JEC21 and B3501A (Fig. 26A). In JEC21, the sit] mutant wassignificantly more melanized than wt. In B3501A, the sit] mutant produced melanin morerapidly than wt and the surface of the colony appeared coarser with a higher sheen in comparisonto wt (Fig. 26A). B3501A wt cells were able to invade the agar whereas the sit] mutant couldnot (Table 16). No defect in melanization was noted for the H99 sit] mutants (Fig 26A). Allexperiments were repeated three times to ensure consistency within the results.b. Presence of extracellular melanin-like granules.A closer inspection of cells grown in DOPA by DIC microscopy (1000X) revealed anaccumulation of extracellular granules in the sit] mutants that may be related to melaninpolymers (Fig. 27A and B). Cells grown in low glucose asparagine medium (LGA) lacking theDOPA substrate did not produce granules. If these granules are related to melanin, this resultindicated that SIT] may play a role in melanogenesis or melanin placement in serotype D strains.Further, this may represent a mislocalization of melanin from the cell walls and may indicate thatSitip plays a role in membrane transport. Interestingly, no difference was noted in the serotypeA strain H99 for either melanin production or the accumulation of extracellular granules (Fig.26A;27C). This illustrates another difference between serotypes. All experiments were repeatedthree times to ensure consistency within the results.137CNNIcr12riA B C00Fig26: MelaninProductiononDOPAMedium.Totestformelaninproduction,acultureof10tLofi0CFU/mLcellswerespottedonto0.1%glucoseplatescontainingA.1mMDOPAB.4mMDOPAandC.4mMDOPA+1.0%glucose.Plateswereincubatedtwodaysat30°C.CellsinFocusAsitiGranulesinFocus- (‘3 CDA B CToinvestigatethemelanindefectofsitimutants,acultureof10tLofi0CFU/mLcellswerespottedonto0.1%glucoseplatescontaining1mMDOPA.Plateswereincubatedfourdaysat30°C.CellswerescrapedfromplatesandresuspendedinsdH2O.10pLofthesuspensionwasviewedat1000Xinoilimmersion.Twoimagesarerepresentedforeachstrainwithcellsinthefieldoffocusandthentheextracellulargranulesinfocus.Insetframesshowamagnffiedimageofthecellsorgranules.A.JEC21B.B3501AC.H99.Fig27:DICmicroscopyofmelanizedcellsatfourdaysgrowthonDOPAmedia.c. Effects of other factors on melanization.A surprising result in these studies was the alteration of melanin deposition that wasnoted in SIT] mutants for the serotype D strains JEC21 and B3501A. There have been previousreports of melanin existing in a granular form (both intracellular and extracellular) in associationwith C. neoformans (Eisenman et al., 2005, Chaskes and Tyndall, 1975). Numerous parameterswere tested to try to further determine what role SIT] may be playing with respect tomelanogenesis because melanin has been implicated in virulence. Specifically, additionalsubstrate compounds and conditions were tested for effects on the melanin phenotype of sit] andwt C. neoformans by plate assay and investigation by DIC microscopy to monitor theaccumulation of the extracellular granules. A number of physiological effects were investigated.These included the effects on melanin levels and deposition by: excess substrate, high glucose,high pH, exogenous ferric vs. ferrous iron and ferric and ferrous iron chelators and a variety ofdiphenolic substrates. The results are summarized in Table 16.The strains were tested at a higher substrate concentration (Fig. 26B) to determine if wtmelanin levels would be restored to mirror the sit] mutant phenotype. If the increased level ofmelanin noted in the sit] mutants was a result of additional substrate available to laccase at thecell surface due to lack of Sitlp, then excess substrate should allow wild-type melanin levels toequal those of the mutants. A restoration of melanin levels by high substrate could indicate thatthe aberrant melanization noted between wt and mutant was the result of higher substrateavailability to laccase at the cell surface caused by lack of DOPA/melanin influx by Sitip. Thehigher DOPA concentration restored the phenotype for B3501A but not JEC21. These resultssuggest that substrate concentration may be a factor of the aberrant melanization in B3 501 A butnot JEC21.Melanization in C. neoformans is partially controlled by the cAMP pathway (Alspaugh et al.,2002). In conditions of low glucose, the cAMP pathway is believed to be activated leading toexpression of laccase and production of melanin. Melanization is repressed by high glucose.Strains were tested for glucose repression using 4X [DOPA] and 1.0 % glucose (Fig.26C). Wefound that glucose could repress melanization in all strains indicating that the SIT] mutation didnot affect the regulatory influence of glucose on melanization.Laccase requires an acidic environment for activity, so not surprisingly, high pHrepressed melanization in all serotype D strains.140Melanin is able to both reduce and oxidize iron (Jacobson, E.S. 2000). It was thereforetested whether different redox conditions would influence the melanin phenotype when an irontransporter (Sit 1 p) was absent. Iron sources with different oxidation states Fell or Feill wereadded to DOPA medium. DOPA with chelators of both Fell (BPDA) or FelIl (deferoxamine)were also tested. The addition of ferric iron (FeEDTA) did not alter the melanin phenotypes ofwt or sit] mutants. The Fell chelator BPDA led to albino phenotypes, however addition ofexcess copper restored the melanization in all strains indicating that loss of melanization byBPDA addition was likely a result of copper not iron chelation. Four molecules of copper arenecessary for a functional laccase enzyme (Williamson, 1997). In contrast, the addition offerrous iron (Fell + the antioxidant ascorbic acid) or a ferric chelator (Felil + deferoxamine)caused the loss of the extracellular granules in sit] mutants. This may indicate that the oxidationstate of iron can contribute to the placement or solubility of melanin. This result was notsurprising as many redox interactions are possible between iron and melanin (Jacobson andHong, 1997).Past research has shown that structurally varied diphenolic substrates can lead to differentsolubilities and placement of melanin resulting in either extracellular and/or intracellularlocalization (Chaskes and Tyndall, 1975). These authors noted that para substrates often lead toextracellular melanin, while ortho compounds (like DOPA-3,4, dihydroxyphenylalanine)normally lead to intracellular melanin. Since these phenotypes appeared to relate to observationsfor the melanin phenotype of serotype D sit] mutants the effects of a variety of diphenolicsubstrates were tested in place of DOPA. Specifically the para compounds 2,5 dihydroxybenzenediacetic acid and hydroxyquinone (1,4) and the ortho compound 3,4 dihydroxybenzoic acid weretested. A general reduction in pigmentation with these diphenolic compounds was noted and areduction in the presence of extracellular granules was found for all substrates (Table 16). Theexpected increase in extracellular granules with para substrates was not observed. These resultsindicated that the melanin phenotype that is observed in the sit] mutants is not affected in thepresence of other diphenolic substrates and the results contradict the findings of Chaskes andTyndall (1975) at least in the strain backgrounds tested. All experiments were repeated threetimes to ensure consistency within the results.141Table 16: Summary of melanin production for sit] mutants in serotype D strains.minimal pigmentformationmedium pigmentformationexcessive pigmentformation5Symbols for ExtracellularGranulesCharacteristic JEC2I je-s1 je-siRl B3501A b-slO b-s42 b-s42R1 b-s42R2DOPA Melanization ++ +++ ++ +++ +++ -1-++ +++ +++ExtracellularGranulesb + +++ + + +++ -i--t-+ + +Agar Invasion - - - + - - + +4X (DOPAI Melanization ++ -H-+ ++ +++ +-f-+ +++ ++4- +++ExtracelIularGranuIes ++ ++ ++ ++ ++ ++ ++ ++Agar Invasion - - - + - - + +4X IDOPAI + Glucose MeIanization - - - - - - - -Extracellular Granules” - - - - - - - -Agar Invasion - - - - - - - -DOPA pH9 MeIanization - - - - - - - -Extracellular Granules” - - - - - - - -Agar Invasion -- - - - - - -DOPA + Fell Ascorbic Acid MeIanization ++ +++ ++ +++ +++ +++ +++ +++Extracellular Granulesb - - - - - - -Agar Invasion - - - + - - + +DOPA + BPDA MeIanization- -- + -- + +Extraceflulartjranules” + - + + - - + +Agar Invasion- - - - - - --DOPA + BPDA + Cu(SQ), MeIanization +++ +++ +++ +++ ++-f- +++ +++ +++Extracellular Granulesb - - - - - - -Agar Invasion -- -- + + - -DOPA + FCEDTA Melanizalion ++ -t-++ ++ +++ +++ +++ +++ +4—I-Extracellular Granulesb - + -- ++ ++ - -Agar Invasion- -- + -- + +DOPA + Deferoxamine Melanization’ ++ -H-f- ++ +++ +++ +++ +++ +++Extracellular Granules”- --- + +--Agar Invasion- -- +- - + +2,5 Dihydroxybenzenediacetic acid Melanization - + - + + + + +Extracellular Granulesb - - - - - - - -Agar Invasion - -- - - -- -3,4 dihydroxybenzoic acid Melanization - ++ - -t-+ ++ ++ ++ ++Extracellular Granulesb + + + + + + + +Agar Invasion - - - + + + + +hydroxyquinone(1,4 pars) Melanization’ - +-f - ++ ++ ++ ++ ++Extracellular Granulesh — — — — - — — —Agar Invasion- -- + + + + +Symbols For Melanization - albino - no extracellular granules*sgg Fig. 26 for visualassessment ofpigmentation++++++* see Fig. 27 for visual assessment of + minimal extracellulargranulesextracellular granules ++ medium extracellulargranules+++ excessive extracellulargranules142F. Influences of SIT] on Cell Wall Integrity.a. Temperature sensitivity and cell wall integrity.Because an accumulation of extracellular granules may be a result of blockedincorporation or displacement of melanin from within the cell wall, alteration to the structure ofthe cell wall was investigated. Cells were tested for temperature sensitivity (a phenotypeassociated with changes in cell wall intrgrity) by spotting serial dilutions of cells on LGA (Figure28A) or DOPA (Figure 28B) media and incubating plates at 30°C or 37°C. The sit] mutantsshowed slightly less growth than wt at 30°C and significantly less growth at 37°C for strainsJEC21 and B3501A (Fig. 28. A,B). The temperature sensitivity is particularly severe in theB3501A background. No difference was noted for 1199 (Fig. 28.A,B). The addition of DOPA tothe medium to allow melanin production did not change the results (Fig. 28B). The compoundscalcofluor white, sodium dodecyl sulphate (SDS) and congo red were also tested in low glucoseor DOPA plates to examine cell wall integrity. Cells with cell wall integrity problems aregenerally more sensitive to these compounds. Only a slight increase in sensitivity was observedfor wt cells vs. sit] mutants to these compounds in JEC21 and B3501A and no difference wasnoted for H99 (Fig. 28 A,B). All experiments were repeated three times to ensure consistencywithin the results.143AC4 w 0 C) C4ID0 w -, OiU,0Totestfortemperaturesensitivityandcellwallintegrity,acultureof5LofserialdilutionsofcellswerespottedontoA.0.1%glucoseorB. 0.1%glucose+1mMDOPAplates. Serialdilutionswere107-102CFU/mL.Plateswereincubatedfortwodaysat30°Cor37°C.SDS-sodiuni dodecylsulfate;CR-congored;CW-calcofluorwhiteI30°C37°C0.015%300igImL30ig/mL37°C37°C37°CSDSCRCWa, I- -1b. Alteration of cell wall structure.Because the cell wall may be compromised by the loss of SIT], a closer examination ofthe cell wall of sit] mutants vs. wt for JEC21 and B3501 was initiated to determine if thestructure had been changed in the sit] mutants. Transmission electron microscopy (TEM) ofmelanized cells (grown for four days on DOPA) and non-melanized cells (grown four days onLGA) were used to further elucidate any changes to the cell wall. There was a striking reductionin cell wall organization between the sit] mutants and wt cells for JEC21 (Figure 29A) andB3501 (Figure 29B) and H99 (Figure 29C). For wild-type and mutant cells, the figure shows twomagnifications of the same cell for each strain background: 30,000X magnification of the fullcell and 150,000X magnification of the cell wall. A 150,000X magnification of the cell wall ofan independent mutant cell is shown to illustrate the consistency of the microscopy findings. Oneimage of 1 50,000X magnification is shown for a reconstituted strain in each background. Inwild-type and reconstituted cells, there are distinct dense concentric rings in the cell walls andordered polysaccharide fibrils are apparent, this is consistent to what was reported by Eisenmanet al., 2005. In contrast, the mutant cell walls were much less dense, varied in thickness and hadunordered truncated fibrils. The cell wall appeared to be porous in comparison to wild-typesuggesting a loss of integrity to the cell wall structure. The mutant cells were also far moredistorted by the TEM processing. For 40 cells documented in each strain, 100 % of JEC21 sit]mutants were significantly distorted (not spherical) compared to 12.5 % of wt. In B35O1A 89 %of sit] mutants were distorted vs. 7.1 % of wt (data not shown). An independent preparation forTEM with a prolonged dehydration steps reduced the distortion in the mutant cells, however thedefect in the cell wall was still apparent (data not shown). These results suggest that the loss ofsit] caused an altered organization of the cell wall for strains JEC21 and B3501A. H99 was notinitially tested in TEM because there was no indication of aberrant melanization, temperature orcell wall sensitivity in the H99 sit] mutant. However, the H99 sit] mutant was later tested and adefect in the cell wall was apparent in this background as well (Fig 29C). The cell walls of thewild type strains increase in thickness in the order of JEC21, B3501A then H99. It is possiblethat the greater thickness in the H99 cell wall may have made it more resistant to temperaturestress and the cell wall integrity compounds even though a defect in the cell wall was present asnoted by TEM. Resin infiltration in the wt H99 cells was also less effective than JEC2 1 orB3501A, further suggesting that more rigidity may be present in the cell wall of this strain.145Finally, the defect in the cell wall noted in the sit] mutants was apparent even in non-melanizedcells that were grown on LGA. The non-melanized cells were more distorted by the TEMprocessing, and the wild type had less density in their cell walls compared to the melanized cells,which may indicate some contribution to rigidity of the cell wall by the incorporation of melanin.However, taken as a whole, these results may indicate that the cell wall defect is not a direct orcomplete result of loss of melanin from the cell wall. A summary of cell wall defects noted byTEM can be found in Table 17.Table 17: Summary of cell wall defects for sit] mutants noted by TEM.+DOPA -DOPAJEC21 NO NOje-si YES YESje-siRi NO NOB35O1A NO NOb-s42 YES YESb-s42Ri NO NOH99 NO NOh-si YES YESh-siRl NO NO146Ce112Topreparecellsfor TEM,acultureofiOCFU/mLcellswerespottedonto0.1%glucoseplatescontaining1mMDOPA.Plateswereincubatedfourdaysat30°C.Cellswereglutaraidehydeandosmiumfixedandstainedwithuranylacetateandleadcitrate.A.JEC21,B.B3501AandC.H99.Imagesshowacrosssectionofthecellwall,cytoplasmandpolysaccharidecapsule.A B C30,000X150,000X150,000XFigure29:Transmissionelectronmicroscopy(TEM)ofmelanizedcellsforwildtypeandsitimutantstrains.150,000XG. Summary of Additional Phenotypic Analyses of sit] Mutants.Overall, this work has suggested possible connections between SIT] and the cAMP pathwayand has shown that the SIT] gene plays a role in iron acquisition during low iron conditions.Therefore phenotypes were tested that are affected by cAMP signaling or the low iron nutritionalcue. Specifically, the mutants were assessed for the elaboration of the polysaccharide capsule inlow iron and iron replete conditions. There are two obvious possible connections here. First, ironsensing may be affected by the loss of SIT], thereby altering the induction of the capsule.Second, a melanin phenotype is observed in the serotype D sit] mutants. Melanin productionthrough the regulation of laccase is also controlled by the cAMP pathway and capsule formationmay also be affected if SIT] has a generalized effect on the cAMP pathway. The wild-type,mutant and reconstituted cells were tested in all three strain backgrounds (JEC21, B3501A andH99). Cells were grown in the same manner as for the Northern analysis in LIM or JR media andharvested at 6 hours. The cells were stained in India ink and viewed at 1000X under oilimmersion. No difference in capsule size or induction was noted between wt or mutant cells inany background. All strains produced capsule in low iron and repressed capsule in iron repleteconditions. These results suggest that SIT] is not required for iron sensing that leads toelaboration of the polysaccharide capsule.To further investigate a generalized cAMP relationship to SIT], mating was tested for wt,mutant and reconstituted cells with the congenic MATa mating strain JEC2O (serotype D) orBt63 (serotype A). Mating is also partly controlled by the cAMP pathway in C. neoformans(Casadevall and Perfect, 1998; Kronstad et al., 1998; D’Souza et al., 2001; Aispaugh et al.,2002). The assays were performed by spotting 10 iL of1O8CFU/mL of the a and a strain on V8agar, then incubating the plates in the dark for 14 days. The plates were then observed forfilamentation, which is indicative of mating. All strains in all backgrounds could mate with thecongenic MATa mating strain indicating that no unilateral mating defect was caused by the lossof SIT]. These results, in conjunction with the lack of any effect on capsule induction, reducethe possibility of a generalized effect of the loss of SIT] on the cAMP pathway.In this study an alteration of the structure of the cell wall has been noted by TEM, by anincrease in temperature sensitivity on minimal media and a slight increase to the cell walldamaging compounds SDS, calcofluor white and congo red. It was therefore plausible that themutants may be more sensitive to osmotic or oxidative stress. The strain set was tested for148oxidative stress against: 30% H20 , menadione and hypochiorite in addition to the osmoticstressors sorbitol and sodium chloride. Interestingly, the mutants did not have increasedsensitivity to any of the oxidative or osmotic stressors tested.H. Drug Susceptibility.C. neoformans is a human pathogen and the availability of effective drugs is limited. It wastherefore of interest to determine whether the loss of a putative siderophore transporter wouldaffect the minimum inhibitory concentration (MIC) of C. neoformans to antifungal agents.Further, SIT] is part of the major facilitator superfamily that also includes multidrug effluxpumps and therefore was a possible candidate for drug efflux activity. The sit] mutants weretested for susceptibility to a number of compounds. Compounds were tested that are used totreat cryptococcosis, that require iron as a cofactor or that were implicated in an increase of SITJtranscription in C. albicans (Lee et al. 2005). Antifungal susceptibility assays were performedon the strain set using the standard microbroth dilution method (NCCLS) for minimuminhibitory concentrations (MIC) in Sabouraud Dextrose Broth (SDB). RPMI was not usedbecause the sit] mutants were unable to grow in this media. (possibly due to the low nutrientcontent). The mutants were able to grown in V2 RPMI; V2 YPD medium suggesting that the lownutrient content and not inhibitory effects were responsible for the lack of growth. Fivecompounds: streptonigrin, ciclopirox olamine, glyphosate, amphotericin B and phleomycin weretested (Table 18) and the experiment was repeated three times. No difference in MIC was notedfor streptonigrin, ciclopirox olamine or glyphosate. JEC21 sit] showed a four-fold increase intolerance to amphotericin B in comparison to wt, whereas the B3501A sit] mutant did not showa difference in MIC. No difference was noted for H99. For phleomycin, there was an eight-foldincrease in tolerance for JEC21 sit], four-fold for B3501A sit] and two-fold for H99 sit]. Nodifference in the phleomycin MIC was noted with either the addition of exogenous iron(FeEDTA) or the chelator (BPDA) (data not shown). This result is important becausephleomycin requires intracellular iron as a cofactor for efficacy and an increase in tolerance ofthe sit mutants to phleomycin is therefore consistent with reduced intracellular iron levels.149Table 18: Minimum inhibitory concentrations of antifungal agents for sit] mutants (tg/mL).CiclopiroxStrain Streptonigrin Amphotericin B Olamine Glyphosate PhleomycinJEC2I 8 0.125 0.5 16 0.125je-sI 8 1 0.5 16 1je-siRi 8 0.125 0.5 16 0.125B3501A 2 0.25 0.5 16 0.03b-sb 2 0.25 0.5 16 0.125b-s42 2 0.25 0.5 16 0.125b-s42Ri 2 0.25 0.5 16 0.03b-s42R2 2 0.25 0.5 16 0.03H99 1 1 0.5 >16 0.25h-si 1 1 0.5 >16 0.5h-s2 1 1 0.5 >16 0.5h-slRi 1 1 0.5 >16 0.25I. Virulence in the Murine Model.Because the ability to acquire iron is paramount to survival and the mammalian hostenvironment is low in free iron, it was appropriate to test whether the sit] mutant had analteration in virulence in comparison to wt cells.a. Serotype D.Initially, experiments were performed for serotype D B3501A wt and the sit] mutant inthe murine model of C. neoformans. DBA1 mice were used because these mice have a defect inC3 complement. In this trial, five mice were inoculated intranasally with wt B3 501 and fourmice with the sit] mutant using 5 x 106 cells per mouse. Unfortunately, due to low virulence ofstrains of the serotype D background, this murine model was not useful. Specifically the micelost weight initially but gained it back by 30 days and appeared healthy for both the wt and thesit] mutant strains (Table 19). JEC21 was not tested because it is known to be even less virulentthan B3501A.Table 19: Summary of DBA1 mouse weight loss with wt and the sit] mutant of the strainB3501A (serotype D).Weight in grams. Mice were euthanized at day 28.Day 0 5 6 7 8 9 12 13 14 15 19 26 28B35OiA 18.08 16.22 15.80 15.50 15.18 14.57 15.19 14.88 15.04 15.18 16.66 17.58 17.74b-s42Asitl 18.53 16.68 16.10 15.72 15.20 15.08 14.68 14.68 14.60 15.25 17.15 19.01 19.50150b. SerotypeA.Virulence was tested for serotype A strains using the murine model. In the first trial,A/Jcr mice were infected intranasally with an inoculum of 5 x 1 serotype A cells per mouse.Ten mice each were used for wt, the sit] mutant and sit] + SIT] reconstituted strain but nodifference in survival was noted in virulence between these strains. All mice were dead or weresacrificed between day 18 and 22 (Table 20; Fig. 30). Fungal cell counts from brain and lungtissue were performed for two mice each of the wt and the sit] mutant strain and one mouse ofthe sit] + SIT] reconstituted strain. Cell counts were equivalent for the brains (1 0 cells/gtissue) and lungs (106 cells/g tissue) of wt, sit] mutants and the reconstituted strain, furtherindicating that SIT] is not necessary for extrapulmonary dissemination in this murine model.The sit] mutant cells isolated from the brain were retested in the siderophore utilization assayand this confirmed that the strain was unable to use deferoxamine (data not shown). In thesecond trial, an experiment was performed using a ten fold smaller innoculum 5 x 1 cells and 5mice per strain. Again there was no difference in virulence between sit] mutants and wt strains.The mice survived slightly longer with the ten fold lower inoculums but were dead or sacrificedat day 26-27 (Table 21) giving further evidence that SIT] is not required for virulence of C.neoformans in this murine model.Table 20: Summary of A/Jcr mouse survival with SIT] wild-type , sit] mutants and sit] + SIT]reconstituted strains of the strain H99 (serotype A) 5 x 1 cell inoculum.days of survival 0 6 8 9 10 15 17 21 22H99 10 10 10 10 10 10 10 2 0h-sI 4sitl 10 10 10 10 10 10 9 2 0h-slRl h-sI Asitl+SITI 10 10 10 10 10 10 9 2 0Table 21: Summary of A!Jcr mouse survival with SIT] wild-type , sit] mutants and sit] + SIT]reconstituted strains of the strain H99 (serotype A) 5 x 1 cell inoculum.Days of Survival 0 15 21 23 26 27H99 5 5 5 5 2 0h-sI 4sitl 5 5 5 4 0 0h-si RI zlsitl + SITI 5 5 5 4 2 015110 9 8 7C.) 0 L-5.D E4z3 2 I 0N)—+-H99wt——h-slsiti—h-slRIsiti+SITI0510152025Daysof SurvivalFigure30:Virulenceassaysinthemurinemodel forserotypeAwildtypeandsitimutantstrains.FemaleA/Jcrmicewereinoculatedintranasally.A.H99wt5XiOcells.B.H99sit]mutant(h-si)5XiOcells. C.H99reconstitutedSIT]+SIT](h-siRi)5XiOcells.Tenmiceperstrainweretested.SUMMARY AND DISCUSSIONThrough the construction and analysis of null mutants in serotype D (JEC2 1 andB3501A) and serotype A (H99) strains of C. neoformans, this work showed that SIT] is requiredfor the use of siderophore bound iron in all strains, for growth in the low iron environment (forstrain B3501A and H99), and for melanin and proper cell wall organization in both serotype Dstrains (JEC21 and B3501A) and cell wall organization in serotype A (H99). Furthermore,siderophore utilization assays show that the cAMP pathway affects siderophore utilization inserotype A but not D. The sit] mutants have increased tolerance to the iron-dependent drugphleomycin and to amphotericin B in the JEC21 background. Finally, it is shown that SIT] isnot required for virulence in the murine model for serotype A. Notable phenotypes for all strainbackgrounds are summarized in Table 22.These results indicate that SIT] does indeed play a role in iron acquisition, particularly inthe utilization of a siderophore as a sole iron source. The increased tolerance of all strains to thedrug phleomycin further suggested that the level of intracellular iron is reduced in sit] mutants.SIT] also appears to have a more generalized affect on iron uptake for the strains B3 501 A andH99 but not JEC21. In B3501A and H99, the cells were unable to achieve robust growth invitro, when iron was low, (400 nM in LIM). Perhaps SIT] has some effect on other uptakemechanisms such as the high affinity system involving FTR] and FET3 or may play a moregeneralized role in nutrient uptake. For example, slower growth for serotype D strains was notedwhen urea was the sole nitrogen source. This observation, in addition to a possiblemisincorporation of melanin into the cell wall, may indicate a more complex role for SIT].Even though SIT] does not appear to have an effect on virulence in the murine model forthe strain H99 (serotype A), it is clear by the in vitro results that it does play a role in effectiveiron utilization when iron is strictly limited. It is possible that the tightly bound or basal level ofiron found in vivo is sufficient for cell survival by other iron uptake mechanisms, such as thehigh affinity uptake system. In addition, other siderophore transporter homologs have beenidentified in C. neoformans as shown in table 15: four in JEC21, three in B350l and two in H99.These additional transporters may be able to compensate for the loss of SIT] during infection.The fact that JEC2 1 does not have impeded growth under these conditions indicates that it maypossess redundant or alternate methods of iron utilization, for example, this strain possesses themost putative siderophore transporter homologs. This is also a good example of how the153function of genes among strains may have diverged through evolution. Another example ofdivergence of gene function is the effect of SIT] on melanization noted in the serotype D but notserotype A strains. SIT] appears to play a role in proper melanin placement in B3 501 A andJEC2 1 and, furthermore, the loss of SIT] resulted in temperature sensitivity and sensitivity to cellwall damaging compounds.Interestingly, a clear alteration of the cell wall structure is noted in all strain backgroundsby TEM even in the absence of melanin. These results indicate that SIT] does play some role incell wall integrity in all backgrounds and that this role may be independent of reinforcement ofthe cell wall by melanin. A final example of divergent roles of the gene is the effect of thecAMP pathway on siderophore utilization noted in serotype A but not D strains. In serotype A,the catalytic subunit knock-out pka] is unable to use siderophores as a sole iron source,suggesting positive regulation of SIT] by cAMP in the serotype A strain. Possibilities for rolesthat SIT] may have with respect to iron utilization, effects involving the signaling pathwayscAMP and PKC]/MAP kinases and cellular processes involved in endosomal trafficking and cellwall integrity will be investigated in more detail in the General Discussion in Chapter Five.Another interesting set of observations from these analyses reveal the diverse array offunction for the SIT] gene in serotype A and D and within closely related strains of the sameserotype, D. Serotype and strain differences are commonly observed in C. neoformans and areparticularly prevalent in studies of iron-regulated genes in our laboratory. Given the variation ofpathogenicity amongst these strains, this phenomenon warrants detailed investigation in thefuture.Table 22: Summary of pleiotrophic phenotypes for sit] mutants of JEC21, B3501A and H99backgrounds (notable phenotypes in bold)._____________________________________________________________1EC21 je-si je-siRi B3501A b-s42 b-s42Ri H99 h-si h-siRlSiderophore Uptake YES DELAYED YES YES NO YES YES NO YEScAMP Regulated SiderophoreUptake NO N/A N/A N/A N/A N/A YES N/A N/AGrowth In UM YES YES YES YES NO YES YES NO YESGrowth in Deferoxamine YES YES YES YES NO YES YES NO YESGrowth in IR YES DELAYED YES YES DELAYED YES YES DELAYED YESMelanization YES INCREASED YES YES ALTERED YES YES YES YESExtracellular Granules FEW MANY FEW FEW MANY FEW FEW FEW FEWIncreased TemperatureSensitivity N/A YES NO N/A YES NO N/A NO NOHigher Sensitivity to Cell WallDammaging Compounds N/A SLIGHT NO N/A YES NO N/A NO NODefect in Cell Wall by TEM NO YES NO NO YES NO NO YES NOIncreased Tolerance toAmphotericin B N/A YES NO NO NO NO NO NO NOIncreased Tolerance toPhleomycin N/A YES NO N/A YES NO N/A YES NO154CHAPTER FIVE: GENERAL DISCUSSIONA. Phenotypic and Genomic Differences Between Serotypes A and D Strains.Since the onset of the studies presented in this thesis, genomic and biological resourcesfor C. neoformans have increased dramatically. Contributions from the work presented here,particularly the physical maps for serotypes A and D (Chapter Two) were of great value to thewhole genome sequence assemblies. As well they identified translocations and inversionsbetween the two genomes (Schein et al., 2002) and identified a region of difference between theprogenitor strain B3501A and JEC21 (Chapter Two) which was later identified to be a segmentalduplication in JEC21 (Fraser et al., 2005). The comparison of strains on a genomic level isbecoming an important area of investigation and a number of sequencing projects are nowunderway or completed for C. neoformans. These include the recent publication of two genomesfor serotype D strains JEC21 and B3501A (Loftus et al., 2005). Three other projects nearcompletion are for the serotype A strain H99 and the serotype B strain R265 at the BroadInstitute of Harvard and MIT, and the serotype B strain WM276 at the University of BritishColumbia and MSGSC. These genome sequences will allow for much broader comparisonsacross strains.A major push of these comparisons is due to the notable differences in virulence thathave been documented between the C. neoformans strains of serotype A, D and AD in themurine model (Barchiesi et al., 2005) and as reflected in clinical prevalence (Casadevall andPerfect, 1998). Biological comparison of the different serotypes and strains will hopefullycontribute to the elucidation of the mechanisms needed to cause disease. A number ofphenotypic differences have already been noted between strains serotypes A and D including therole of components in the cAMP pathway (Hicks et al., 2004) and the HOG pathway (Balm eta!., 2005). The cAMP pathway partially controls virulence factors such as the elaboration of thepolysaccharide capsule and the expression of laccase, the enzyme necessary for the production ofmelanin. Low iron conditions also lead to induction of the capsule and low glucose activates theexpression of laccase via the cAMP pathway (Aispaugh et a!., 1997). The HOG pathway, whichis important in the stress response (including regulation of HSPJ2 in S. cerevisiae) that leads tocell wall integrity has been characterized in C. neoformans where serotype A null mutants areattenuated for virulence but serotype D mutants are not (Balm et al., 2005).155The acquisition of iron is a key process for the survival of pathogens such as C.neoformans in the host environment. The uptake of siderophores may be an importantcomponent of iron acquisition. Therefore, this work characterized a siderophore transporter SIT]in three strains of C. neoformans to identify its role in iron acquisition and the biology of thefungus. Interestingly, marked differences were noted for SIT] phenotypes of serotypes D and Ain addition to two strains within the D serotype. For example, there were differences inmelanization in serotype D but not A; restoration of the alteration in melanin by excess substratein B3 501 but not JEC2 1; differences in sensitivity to temperature and cell wall damagingcompounds; lack of growth in low iron for B3501A and H99 but not JEC21; evidence of cAMPinvolvement in siderophore utilization for serotype A but not D and evidence of reduced uptakeof urea in serotype D but not A. The only common phenotypes to all three strains was the loss ofintegrity to the cell wall structure by TEM and loss of the ability to use siderophores as a soleiron source, although in JEC2I the growth response was simply delayed. These variations inphenotypes between B3501A and JEC21, two serotype D strains, is becoming a commonobservation in the Cryptococcus literature.Notable differences between genome content and structure have also been recentlyreported (Loftus et at., 2005; Fraser et at., 2005). Specifically a segmental duplication has beenidentified in the JEC2I strain in comparison to the B3501A strain (Fraser et at., 2005). One ofthe duplicated genes; 8MG] plays a role in melanin formation, where it was identified as asuppressor of the gpa] mutation (Pukkila-Worley et at., 2005). These results suggest that thisduplicated region, and particularly the additional copy of the 5MG] gene may partly account forthe differences in melanization noted between the JEC2 1 and B3 501 A strains. These results areof great interest given the difference in levels of virulence between the strains, despite their closerelatedness. B3501A is a progenitor of JEC21, where JEC21 resulted from nine backcrosseswith the MATa strain B3502 (Heitman et al., 1999). The lack of aberrant melanization in theserotype A SIT] and the presence of a cAMP influence on siderophore utilization in serotype Abut not D are further examples of phenotypic differences observed between knock out strains inserotype D vs. A (Hicks et at., 2004; Balm et al., 2005). Structural differences in the genome ofthe serotypes have also been noted in Chapter Two (Schein et at., 2002). In this case thedifference in virulence is dramatic, the serotype A strain H99 is far more pathogenic thanserotype D strains B3501A or JEC21 (Barchiesi et al., 2005).156An observation in the Kronstad laboratory is that this variation of phenotypes amongststrains appears to be particularly prevalent in iron regulated genes and factors involved with cellwall integrity: for example in the genes SIT], CIG] and CIR] (presented here and personalcommunications). The comparison of the serotype D and A genomes (Chapter Two), theanalysis of the serotype A low iron transcriptome (Chapter Three) in addition to the variationdetailed in the sit] mutant phenotypes (Chapter Four) of this study, in combination with ongoingstudies in the Kronstad Laboratory and Cryptococcus scientific community, will contribute to amuch broader picture of strain variation in the future for C. neoformans. The value of theseresources is already apparent here, leading from physical mapping, EST projects and karyotypestudies to genome sequence assemblies, that allow transcriptional studies and targeted biologyfor investigation of virulence in C. neoformans and further cross strain comparisons on agenomic and biological level. It should be noted that the variation in phenotypes and genomestructure noted within strains of this study should invoke caution that these strains may notrepresent all strains within a given serotype and gene structure and function should beinvestigated on a strain by strain basis.B. Iron-regulated Genes in C. neoformans.SAGE analysis of iron induced genes in this work showed a remarkable similarity to theSAGE findings for a library of cells isolated from rabbit CSF. These results suggest animportance of the low iron signal on gene transcription in vivo. Of particular interest are highlyabundant transcripts for genes involved in iron transport e.g. FTR]; the stress response e.g.HSPs, thioredoxin, glutathione peroxidase; signaling components e.g. guanine nucleotidebinding protein, 14-3-3 protein, GTP-binding nuclear protein SPIJ, and Rho2 GTP-bindingprotein; cell wall synthesis e.g. chitin deacetylase; and the potential antioxidant oxidoreductase.Further characterization of these genes may lead to important insights into the ability of thefungus to survive in vivo specifically in relation to survival in a low iron environment. Several ofthe genes have been characterized and shown to be important in the stress response and/orvirulence, including the genes for glutathione peroxidase (Missall et al., 2005a) and thioredoxin(Misall et al., 2005b; c). HSP]2 would be an excellent candidate to investigate because itstranscript is highly abundant in low iron conditions and in vivo in C. neoformans and is inducedin the presence of many stressors in S. cerevisiae and C. albicans (as discussed in Chapter157Three). Many of the same candidate genes have been highlighted in the work done by Steen etal., 2003 that characterized the transcriptome of C. neoformans from rabbit CSF.Comparison of the serotype A low iron transcriptome (characterized in this work) to thatof the serotype D strain B3501A (Lian et al., 2005) showed that similar genes were inducedunder low iron conditions in this strain, for example: the genes for iron permease, oxidoreductaseand thioredoxin peroxidase. Interestingly, the SIT] tag was not identified in the serotype A lowiron transcriptome as being abundant, although the putative tag was identified in the data at twocopies in low iron. In serotype D, there were 17 tags corresponding to the SIT] gene in low ironand 3 under iron replete conditions. Comparison of data from this study with forthcoming datafrom iron replete medium in this serotype A strain will provide a basis for further comparisonswith the data from the serotype D strain. It is certain that SIT] is induced in low iron vs. ironreplete conditions in serotype A (H99) as shown by Northern analysis in this study. It is possiblethat the transcript is not highly abundant (low expression) or it may not be possible to obtain atag for this gene in serotype A by the SAGE procedure; for example if an NlaIII recognition siteis not present in the transcript or the site adjacent to a poly-A tail.In C. albicans, the response to low iron was investigated by a genome-wide expressionchanges by microarray experiments (Lan et al., 2004). This group found that 526 open readingframes (ORFs) were more highly expressed in low iron and 626 in iron replete conditions. Therewas a wide range of genes induced by low iron including those involved iron transport and ironregulation. The group further focused on a the SFU] gene, a homolog of U maydis URBS],which is a transcriptional repressor of siderophore synthesis. Comparison of wt and sfu] mutantsidentified 139 putative targets of the gene, many of which were iron responsive. A homolog ofthis gene has been identified in C. neoformans designated CIR] (Cryptococcus iron regulation)and as mentioned is being characterized by deletion analysis and microarray experiments by apost-doctoral fellow, Wonhee Jung in the Kronstad laboratory. Multiple phenotypes have beenidentified and the SIT] gene characterized in Chapter Four of this work does have a putativerecognition site for the CIR] gene product. CIR] is a putative transcriptional repressor of ironresponsive genes, the cir] deletion mutant is more sensitive to phleomycin (possible increase inintracellular iron), and has defects in melanin in serotype D and cell wall integrity in serotypes Dand A (W. Jung personal communication). This work will further elucidate an important ironregulatory system in C. neoformans that includes the regulation of the SIT] gene.158Another group investigated the transcriptional changes in C. albicans in response to theexposure to human blood (Fradin et al., 2003). Interestingly, a similar suite of genes to thosefound in the low iron studies were induced by exposure to blood, including those involved withthe stress response and antioxidative response. These results are reminiscent of the finding inthis study that the low iron environment induces many genes similar to those found in cellsisolated from rabbit CSF.Related work underway in the Kronstad laboratory will use SAGE and microarraytechnology to further investigate the relationship of the low iron transcriptome (characterized inthis study) with respect to iron regulation (by comparison to iron replete SAGE transcriptome);cAMP signaling (by comparison to pka] and pkr] SAGE transcriptomes); pathogenesis (bycomparison to SAGE transcriptomes of cells from mouse lung or macrophage cell line) andmicroarray experiments using cir] mutants in comparison to wt cells. These analyses takentogether should further develop the foundation of resources for genomic and genetic approachesto understand virulence. The value of using genomic resources leading to targeted biology hasbeen demonstrated by the range of interesting phenotypes identified in the study of the SIT] genein Chapter Four.C. The Role of SIT] in Iron Acquisition.The SIT] gene proved to be necessary for the use of the siderophore ferrioxamine B(FOB) as the sole source of iron in all strains. Although it is interesting that C. neoformans doesnot appear to synthesize siderophores, it is possible that siderophore uptake plays an importantrole during saprophytic growth in the environment. The nature of virulence genes inopportunistic pathogens potentially includes a primary role during growth in the environment(Casadevall and Perfect, 1998). C. neoformans may be capitalizing on the production ofsiderophores by nutritional competitors for iron acquisition. In the host environment (e.g. blood,cerebral spinal fluid and the brain) where free iron is low, this transport system could be playingan important role for the acquisition of tightly bound iron. However, if Cryptococcus does notproduce its own siderophores, where would the fungal cells acquire these molecules? For otheropportunistic fungal pathogens (such as Aspergillusfumigatus) that do synthesize siderophores, agene involved in siderophore synthesis SITA that codes for an ornithine monooxygenase isessential for virulence in the rat model (Schrettl et al., 2004). These authors also found that159reductive iron assimilation by the high affinity uptake system through FTR] is not required forvirulence.Although microbes possess ingenious ways to sequester iron, mammals also havemechanisms to regulate iron levels to limit the availability of this essential metal to microbes.Iron is tightly regulated in mammalian fluids, e.g. free iron in blood serum is only present at 1 026M (Otto et aL, 1991). Mammals can also reduce the amount of transferrin (an iron bindingprotein), shut off iron recycling in macrophages (Barasch and Mori, 2004) and recently it hasbeen discovered that mammals can steal bacterial siderophores through the lipocalin 2 protein(Flo et a!., 2004). This protein has a high affinity for binding enterochelin (a bacterialsiderophore) (10’°M) (Goetz et al., 2002), which led the investigators to believe that thesiderophore may be the true ligand of lipocalin 2. C. neoformans does have a putative homologfor Arn4p/Enbl of S. cerevisiae, a transporter that has been shown to have affinity forenterochelin (also known as enterobactin) (Heymann et al., 2000). Alternatively, lipocalin 2could shuttle an unidentified ligand, perhaps even a putative mammalian siderophore (Yang eta?., 2002). If this were true, it is plausible that C. neoformans could make use of theseendogenous compounds in the mammalian host. It is also possible that C. neoformans cansynthesize its own siderophores through an undiscovered biochemical pathway. Bioinformaticanalysis in this study showed that serotype A and D C. neoformans do have putative homologsfor at least one of the two siderophore synthesis genes. Although, the presence of siderophores inthe host environment may be in question, it was shown in this work that in two strains of C.neoformans, B3501A and H99, SIT] is required for growth in low iron medium even in theabsence of a siderophore. Perhaps SIT] has evolved a further role for the acquisition of ironwhen low iron is available for these strains through an effect on the high or low affinity uptakesystems.There is a notable difference in the JEC2 1 results in the siderophore utilization assay onagar medium vs. growth in liquid medium. Lack of growth is noted for the JEC2 1 sit] mutant inthe utilization assays but not in the liquid assays. Growth of the JEC21 sit] mutant was found onthe plates after seven days incubation further indicating that JEC2 1 may have additionalmechanisms to acquire iron. For example, as mentioned in Chapter Four, JEC2 1 does have fourputative homolog for siderophore transporters, where B350 1 A has three and H99 two; it is160possible that a transporter is present in JEC21 that has redundant affinity with Sitlp forferoxamine.D. The Role of SIT] in the Structure of the Cell Wall.The recent publication of the microstructure of the cell wall of C. neoformans (Eisenmanet a!., 2005) suggests that melanin is placed in the wall in concentric layers of spherical granularparticles of about 40-130 nm in diameter. In light of these results, I wondered whether Sit 1 p wasplaying a role in the deposition of melanin in the wall and whether the accumulation of granulesoutside sit] cells was a result of blocked deposition. If this were the case then there wouldpresumably be a decrease in the rigidity or integrity of the cell wall in the sit] mutants.Although the mutants did not exhibit significant increased sensitivity to the compoundscalcofluor white, congo red or sodium dodecyl sulphate (SDS), they did show a marked increasein sensitivity to temperature on minimal medium. Temperature sensitivity can be caused byfactors other than reduced integrity of the cell wall, however the H99 sit] mutants that did notexhibit aberrant melanization, also did not show any temperature sensitivity in comparison to wt.Given the possibility of an alteration of the cell wall structure, transmission electronmicroscopy (TEM) was used to more closely examine the wall. A significant difference wasnoted in the cell wall appearance of mutant vs. wt cells. In contrast to the previous results, allstrains did show a decrease in cell wall organization by TEM. This was interesting because itsuggested that SIT] does play a role in the organization of the cell wall in all strain backgrounds.Whether this is partly due to a lack of melanin in serotype D strains remains to be investigated;however this defect in the cell wall was noted even in the absence of melanin and was alsopresent in the serotype A (H99) background that had no apparent alteration to melanization.Therefore, it seems that melanin is not the sole cause for the cell wall defect. The greaterdistortion of non-melanized cells during processing for TEM and reduction of cell wall density inwild type cells suggested that melanin does have some contribution to cell wall rigidity. Theabsence of melanin specific stains and the insolubility of the polymer makes it difficult tothoroughly detect and characterize. This is illustrated by the need for harsh treatments such asproteases, 4 M guanidinium isothiocyanate followed by 6 M HC1 at 100 degrees C (in addition tostudies with monoclonal antibodies) to definitively identify the presence of melanin in vivo(Rosas et a!., 2000). It is possible that SIT] plays a role in delivery of components to the cell wall161that may be required for both integrity and laccase localization. In this context, the potential roleand interactions of SIT] with the PKC]/MAP kinase pathway is discussed below.E. The Possible Role of SIT] in Endosomal Trafficking.In S. cerevisiae, the Am family of proteins (that includes the CnSITJ ortholog) are uniquein their structure among the major facilitator superfamily (Kim et al., 2005). They possess anextracytosolic domain at their carboxy terminus believed to be responsible for the localization ofthe protein in the cell (Kim et a!., 2005). In S. cerevisiae, the Am proteins cycle from plasmamembrane to the endosomes depending on the concentration of siderophore (Kim et a!., 2005).When no siderophore is present the protein resides in the endosomal compartment. In thepresence of low levels of siderophore, the transporter relocalizes to the plasma membrane butlittle siderophore uptake is noted. When siderophore levels are elevated, the protein cycles fromthe plasma membrane to the endosome to transport iron bound siderophores. It is likely thatthese proteins are actively internalizing the siderophore bound iron for storage or reductionwithin the cell. Thus, it is possible that CnSitlp may be playing a role in endosome traffickingof components to the cell wall in C. neoformans. For example the lack of growth in low iron mayindicate that SIT] has a connection to the high or low affinity uptake systems. Perhaps SIT] isinvolved in the localization of key components from these systems such as the iron permeaseFTR] or the oxidase FET3. Although SIT] may affect the high and low affinity uptake systems,it has been shown here that the loss of the high affinity permease FTR] does not affectsiderophore utilization. It is possible that in serotype D strains, C. neoformans SIT] has evolvedto play a role in the internalization or transport of additional substrates such as melanin ormelanin precursors for deposition into the cell wall. It was also found that the serotype D strainshave reduced growth in minimal medium when urea is the sole nitrogen source indicating thatSIT] may be playing a more generalized role in uptake systems of the fungus. All strains exhibita defect in the integrity of the cell wall when viewed by TEM. A defect in endosome traffickingcould be responsible, if SIT] were responsible for shuttling key components between endosomesand the cell walls. A prediction of the transmembrane regions of the protein showed that Sitipdoes indeed have a strong candidate for a carboxy-terminus extracytosolic loop of 58 aminoacids (discussed in Chapter Four), this provides suggestive evidence that SIT] may play a role inendosomal trafficking like its homologs, the Am family of proteins from S. cerevisiae.162F. The Potential Influences of the cAMP and PKC1 Signal Transduction Pathways on SIT].The PKC]/MAP kinase pathway has been recently implicated in melanogenesis, wherethe loss of the Cl-domain (for diacyiglycerol binding) of PKC] led to loss of laccase activity andmelanin production (Heung et a!., 2005). The Cl-mutants displayed a similar phenotype to theserotype D SIT] mutants in that they had disorganized melanization and a reduced integrity tothe cell wall. Heung et al. noted that the biological effects were mediated by a change in cellwall integrity due to the displacement of laccase from the cell wall. SIT] has six putative Pkc 1 pphosphorylation sites as determined by the motif predicting program Prosite( SIT] may therefore be involved in the translocation of laccaseto the plasma membrane, so in the absence of the phosphorylation of SIT] by Pkclp or in theabsence of the Sitl p protein, the laccase does not reach the plasma membrane or is misplacedfrom the membrane, possibly into the supernatant. The cell wall integrity pathway is activated inC. albicans when Pkclp phosphorylates the terminal MAP kinase in the pathway, Mkclp (Bateset a!., 2005). A homolog of Mkclp is present in the C. neoformans serotype D strain JEC21(TIGR Identifier 162.m02645 ICNIOO41OI) and has been characterized in C. neoformans (Krauset a!., 2003) where the study showed that the Mkclp homolog, (Mpklp/Slt2p) plays a role inmaintenance of cell wall integrity during elevated temperatures and in the presence of cell wallinhibitors. Further components of the PKC] signal transduction pathway have now beenextensively analyzed in C. neoformans with respect to its affect on cell wall integrity (Gerik eta!., 2005). This study included the deletion and analysis of genes involved in the PKC] pathway.The group found that the kinases Bcklp and Mkk2p were essential for maintaining integrity ofthe cell wall suggesting that they act downstream of Pkc 1 p in the pathway. A regulator proteinLrglp and phosphatase Ppglp were also implicated in cell wall integrity. It is interesting to notethat a decrease in melanin production was found for the mkk2 mutant and for the ppg] mutantsuggesting further involvement of the PKC]/MAP kinase pathway in melanogenesis. Theabsence of the melanin phenotype in serotype A sit] mutants may indicate that Sitlp plays a rolein melanogenesis via the PKC]/MAP kinase pathway in serotype D but not A. In conclusion,given the loss of cell wall integrity in all strain backgrounds caused by the loss of SIT], it ispossible, that Sitlp may interact at some level with the PKC] pathway leading to the retention ofintegrity in the cell walls.163The potential role of cAMP signaling on or by SIT] was preliminarily investigated in thisstudy. It was found that siderophore utilization is influenced by the cAMP pathway in theserotype A H99 background but not the serotype D JEC2 1 strain. Siderophore utilization ispositively affected by the induction of the cAMP pathway in serotype A through the catalyticsubunit Pkalp. Divergence in the role of cAMP components have been previously documentedbetween serotype A and D strains of C. neoformans (Hicks et al., 2004). The PKA pathway hasalso been implicated in the regulation of iron uptake in S. cerevisiae (Robertson et al., 2000). Inthis case the loss of the catalytic subunit TPK2, results in an increase of transcription ofcomponents of the high affinity uptake system (FRE2, FET3, FTR], CCC2) and the siderophoretransporters (SIT] and ARNJ). So in contrast to preliminary results in C. neoformans where ironuptake appears to be induced or positively regulated by the catalytic subunit (PKA]), iron uptakeis negatively regulated by the cAMP pathway through TPK2 in S. cerevisiae.Experiments were performed to determine whether the melanin phenotype observed inserotype D strains may be a result of a cAMP signaling connection. Low iron is a nutritionalsignal for induction of the cAMP pathway in C. neoformans leading to capsule formation(Alspaugh et al., 1997). Melanization is also partly regulated by this pathway. In the low glucoseenvironment the cAMP pathway is activated leading to the expression of laccase, the phenoloxidase enzyme that oxidizes DOPA, in the rate limiting step in the production of melanin.Therefore, it was plausible that the loss of an iron transporter may lead to a reduction inintracellular iron and activation of the cAMP pathway. This activation could then lead to theupregulation of the LAC] gene encoding laccase. High glucose represses the cAMP pathwayand melanization (Nurudeen and Ahearn, 1979; Alspaugh et a?., 1997)Strains were tested to see if the repression of melanization by glucose was interrupted inthe SIT] mutants. No difference was noted in the suppression of melanization by glucose. Thisresult suggests that the influence of SIT] on melanin formation is not related to the suppressionmechanisms of glucose. We tested other factors controlled by the cAMP pathway such as theability to mate with the congenic MATa strain and the elaboration of capsule in the low ironenvironment (LIM medium) or the repression of the capsule in iron rich conditions (IR medium)(data not shown). No difference was noted in mating or capsule formation, all strains couldmate, all strains produced a large capsule in LIM and a small capsule in IR medium. These dataindicated that a general relationship (general induction of the cAMP pathway) is not likely for164SIT] and melanogenesis through the cAMP pathway. There is however an affect of the cAMPpathway on siderophore utilization in the serotype A strain H99. However, this cellular processmay likely be independent as the melanin phenotype is noted in the serotype D strains.G. Susceptibility of sit] Mutants to Antifungal Compounds.Streptonigrin is an anticancer drug that possesses antifungal activity and that also requiresiron as a cofactor. Since we have characterized SIT] as a putative iron transporter, we wereinterested in whether the susceptibility to this iron-requiring drug would be affected by deletionof the gene. Ciclopirox olamine was tested because transcriptional microarray experiments withC. albicans cells treated with this drug show an upregulation of the siderophore transporterhomolog ARN]/SITA (Lee et al., 2005). Glyphosate was tested because it has been shown toinhibit melanization in C. neoformans and to prolong the survival of mice infected with thefungus (Nosanchuk 2001). No difference in MIC was noted between wt, the sit] mutants orreconstituted cells for streptonigrin, ciclopirox olamine or glyphosate. Amphotericin B is thefront line drug used in the treatment of cryptococcosis. It is an amphipathic compound that bindsergosterol forming a pore in the cell membrane leading to lysis. Melanin provides protection tocryptococcal cells from amphotericin B, presumably through binding of the drug (Ikeda et a!.2003). JEC21 sit] showed a four-fold increase in tolerance to Amphotericin B in comparison towt where B3501A sit] did not show a difference in MIC. This result was not surprisingconsidering there was a far more significant difference in the level of melanization between wtand sit] in JEC2 1. No difference was noted for H99. The drug phleomycin was tested, whichrequires intracellular iron for efficacy. There was an eight-fold increase in tolerance for theJEC21 sit] mutant, four-fold for the B3501A sit] mutant and two-fold for the H99 sit] mutant.These results were expected because the sit] mutants likely had a reduction in the level ofintracellular iron caused by the loss of the Siti p siderophore transporter. This in turn reduced theefficacy of phleomycin. It is also possible that SIT], an ABC tran