Regulation of Nitrogen Metabolism in Fungi: the Impact of Nitrogen Catabolism on Virulence in the Fungal Pathogen Cryptococcus Neoformans

Regulation of Nitrogen Metabolism in Fungi: The Impact of Nitrogen Catabolism on Virulence in the Fungal Pathogen Cryptococcus neoformans

Ivor Russel Lee

Bachelor of Science (Honours Class 1) with specialisation in Microbiology

Bachelor of Science with double majors in Biomedical Science and Genetics

A thesis submitted for the degree of Doctor of Philosophy at The University of Queensland in 2013

School of Chemistry and Molecular Biosciences Abstract

Regulation of nitrogen catabolism is of immense interest in the kingdom Fungi, as nitrogen source utilisation influences key aspects of fungal biology including development, secondary metabolite production and pathogenesis. In the basidiomycete pathogen Cryptococcus neoformans that causes over 625,000 deaths annually, relatively little is known about nitrogen regulation. In view of this lack of knowledge, this thesis is articulated through four major aims to investigate the interplay between nitrogen catabolism and virulence in C. neoformans. The first two aims scrutinise the wiring of the global nitrogen regulatory circuit, while the last two aims examine specific nitrogen catabolic pathways of medical relevance.

In all species of fungi studied thus far, GATA transcription factors are the central regulators of nitrogen assimilation as they globally activate the expression of permease and catabolic enzyme-encoding genes required to degrade most complex/secondary/non-preferred nitrogenous compounds. The first aim is to characterise the function of the nitrogen-responsive GATA factor in C. neoformans. Here, the C. neoformans Gat1/Are1 factor was shown to regulate multiple key virulence traits, in addition to its evolutionary conserved role of controlling nitrogen utilisation. These virulence attributes include the formation of antiphagocytic capsule, deposition of antioxidant melanin in the cell wall, production of infectious basidiospores, and the ability to grow at high temperature.

Given the importance of the GATA factor Gat1/Are1 in pathogenicity, gaining insights into the molecular mechanism governing the regulation of Gat1/Are1 activity in response to nitrogen availability is pivotal. In certain fungal species, GATA factor activity is inhibited through interaction with co-repressor Nmr proteins when readily assimilated (preferred) nitrogen sources such as ammonium or glutamine are present. This regulatory phenomenon, nitrogen metabolite/catabolite repression, enables preferential utilisation of metabolically favoured compounds thereby conserving the organisms’ energy to maximise fitness. The second aim is to characterise the function of a potential Nmr homolog in C. neoformans. Here, unique evidence of divergence between different fungal species in the evolution of Nmr-related proteins was demonstrated. In addition to its functionally conserved role of inhibiting GATA factor

ii activity under repressing conditions, the C. neoformans Nmr homolog Tar1 also positively regulates GAT1/ARE1 expression under non-repressing conditions.

The transcriptional studies of catabolic genes from the proline degradation pathway were proven to be a fertile area of research for investigating the roles of Gat1/Are1 and Tar1 in modulating nitrogen metabolite repression. The third aim is to formally characterise the functions of these enzymes. Degradation of the multifunctional amino acid proline is associated with mitochondrial oxidative respiration. The two-step oxidation of proline is catalysed by proline oxidase and Δ1- pyrroline-5-carboxylate (P5C) dehydrogenase, which produce P5C and glutamate, respectively. In the C. neoformans genome, there are two paralogs (PUT1 and PUT5) that encode proline oxidases and a single homolog (PUT2) that encodes P5C dehydrogenase. Here, the expression of all three putative catabolic genes was shown to be inducible by the presence of proline, however, only Put5 and Put2 were required for proline utilisation. Additionally, Put2 was required to prevent excessive mitochondrial superoxide production probably by limiting the P5C-proline cycling that delivers electrons to the electron transport chain and to O2. Intracellular accumulation of reactive oxygen species is known to be a critical feature of cell death; consistent with this fact, the put2Δ mutant exhibited a slight, general growth defect. Further, the put2Δ mutant was avirulent during murine infection, making this the first report highlighting the importance of P5C dehydrogenase in enabling pathogenesis of a microorganism.

Multiple lines of evidence have also suggested that the enzymes of another catabolic pathway, uric acid, may be important for pathogenesis. The ecological niche of C. neoformans is uric acid-rich pigeon guano, and uric acid is also a well-known inducer of the virulence factor capsule formation. Assimilation of uric acid necessitates several enzymes including another bona fide virulence factor, urease, the only enzyme of the pathway that has been characterised. Since uric acid is the end product of purine metabolism in humans and it acts as an immune enhancer, uric acid utilisation by C. neoformans during infection may potentially serve as a method to modulate host immune response. The fourth aim is to characterise the functions of all the uric acid catabolic enzymes. Here, complete degradation of uric acid to ammonia was shown to require: URO1-encoded urate oxidase, URO2-encoded HIU hydrolase,

iii URO3-encoded OHCU decarboxylase, DAL1-encoded allantoinase, DAL2,3,3- encoded allantoicase-ureidoglycolate hydrolase, and URE1-encoded urease. Further studies however, are required to verify the roles of the C. neoformans uric acid catabolic enzymes during infection, ideally using a modified animal model system that lacks these biological enzymes for improved mimicking of human pathogenicity. Overall, these studies lay an excellent foundation for future research into dissecting the link between nitrogen sensing and elaboration of virulence composite in C. neoformans.

iv Declaration by author

This thesis is composed of my original work, and contains no material previously published or written by another person except where due reference has been made in the text. I have clearly stated the contribution by others to jointly-authored works that I have included in my thesis.

I have clearly stated the contribution of others to my thesis as a whole, including statistical assistance, survey design, data analysis, significant technical procedures, professional editorial advice, and any other original research work used or reported in my thesis. The content of my thesis is the result of work I have carried out since the commencement of my research higher degree candidature and does not include a substantial part of work that has been submitted to qualify for the award of any other degree or diploma in any university or other tertiary institution. I have clearly stated which parts of my thesis, if any, have been submitted to qualify for another award.

I acknowledge that an electronic copy of my thesis must be lodged with the University Library and, subject to the General Award Rules of The University of Queensland, immediately made available for research and study in accordance with the Copyright Act 1968.

I acknowledge that copyright of all material contained in my thesis resides with the copyright holder(s) of that material. Where appropriate I have obtained copyright permission from the copyright holder to reproduce material in this thesis.

v Publications during candidature

Peer-reviewed journal articles:

Lee IR, Chow EW, Morrow CA, Djordjevic JT, Fraser JA (2011). Nitrogen Metabolite Repression of Metabolism and Virulence in the Human Fungal Pathogen Cryptococcus neoformans. Genetics 188:309-23.

Morrow CA, Lee IR, Chow EW, Ormerod KL, Goldinger A, Byrnes EJ 3rd, Nielsen K, Heitman J, Schirra HJ, Fraser JA (2012). A Unique Chromosomal Rearrangement in the Cryptococcus neoformans var. grubii Type Strain Enhances Key Phenotypes Associated with Virulence. mBio 3:e00310-11.

Lee IR, Lim JW, Ormerod KL, Morrow CA, Fraser JA (2012). Characterization of an Nmr Homolog that Regulates GATA Factor-Mediated Nitrogen Metabolite Repression in Cryptococcus neoformans. PLoS One 7:e32585.

Morrow CA, Valkov E, Stamp A, Chow EW, Lee IR, Wronski A, Williams SJ, Hill JM, Djordjevic JT, Kappler U, Kobe B, Fraser JA (2012). De novo GTP Biosynthesis Is Critical for Virulence of the Fungal Pathogen Cryptococcus neoformans. PLoS Pathogens 8:e1002957.

Ormerod KL, Morrow CA, Chow EW, Lee IR, Arras SD, Schirra HJ, Cox GM, Fries BC, Fraser JA (2013). Comparative Genomics of Serial Isolates of Cryptococcus neoformans Reveals Gene Associated with Carbon Utilization and Virulence. G3- Genes/Genomics/Genetics DOI: 10.1534/g3.113.005660.

Lee IR, Lui EY, Chow EW, Arras SD, Morrow CA, Fraser JA (2013). Reactive Oxygen Species Homeostasis and Virulence of the Fungal Pathogen Cryptococcus neoformans Requires an Intact Proline Catabolism Pathway. Genetics 194:421-33.

Lee IR, Yang L, Sebetso G, Doan TH, Allen R, Blundell R, Lui EY, Morrow CA, Fraser JA (2013). Characterization of the Complete Uric Acid Degradation Pathway in the Fungal Pathogen Cryptococcus neoformans. PLoS One 8:e64292.

Lee IR, Morrow CA, Fraser JA (2013). Nitrogen Regulation of Virulence in Clinically Prevalent Fungal Pathogens. FEMS Microbiology Letters 345:77-84.

Publications included in this thesis

Lee IR, Morrow CA, Fraser JA (2013). Nitrogen Regulation of Virulence in Clinically Prevalent Fungal Pathogens. FEMS Microbiology Letters 345:77-84. – Incorporated as Chapter 1.

Contributor Statement of contribution

Lee IR (Candidate) Drafting and writing (90%)

Morrow CA Drafting and writing (5%)

Fraser JA Drafting and writing (5%)

Lee IR, Chow EW, Morrow CA, Djordjevic JT, Fraser JA (2011). Nitrogen Metabolite Repression of Metabolism and Virulence in the Human Fungal Pathogen Cryptococcus neoformans. Genetics 188:309-23. – Incorporated as Chapter 3.

Contributor Statement of contribution

Lee IR (Candidate) Conception and design (55%), experimental work (90%), analysis and interpretation of data (80%), drafting and writing (75%)

Chow EW Experimental work (5%), drafting and writing (5%)

vii Morrow CA Experimental work (5%), drafting and writing (5%)

Djordjevic JT Conception and design (5%)

Fraser JA Conception and design (40%), analysis and interpretation of data (20%), drafting and writing (15%)

Contributor Statement of contribution

Lee IR (Candidate) Conception and design (95%), experimental work (70%), analysis and interpretation of data (90%), drafting and writing (95%)

Lim JW Experimental work (20%), analysis and interpretation of data (5%)

Ormerod KL Experimental work (5%)

Morrow CA Experimental work (5%)

Fraser JA Conception and design (5%), analysis and interpretation of data (5%), drafting and writing (5%)

Contributor Statement of contribution

Lee IR (Candidate) Conception and design (45%), experimental work (40%),

viii analysis and interpretation of data (50%), drafting and writing (60%)

Lui EY Conception and design (30%), experimental work (40%), analysis and interpretation of data (40%), drafting and writing (35%)

Chow EW Experimental work (10%)

Arras SD Experimental work (5%)

Morrow CA Experimental work (5%)

Fraser JA Conception and design (25%), analysis and interpretation of data (10%), drafting and writing (5%)

Contributor Statement of contribution

Lee IR (Candidate) Conception and design (30%), experimental work (35%), analysis and interpretation of data (50%), drafting and writing (55%)

Yang L Conception and design (30%), experimental work (25%), analysis and interpretation of data (30%), drafting and writing (30%)

Sebetso G Experimental work (15%)

Doan TH Experimental work (10%)

Allen R Experimental work (5%)

Blundell R Experimental work (5%)

ix Lui EY Experimental work (5%)

Morrow CA Conception and design (5%), analysis and interpretation of data (5%)

Fraser JA Conception and design (35%), analysis and interpretation of data (15%), drafting and writing (15%)

Contributions by others to the thesis

Dr James Fraser contributed to the conception and experimental design of all projects, as well as interpretation of data and drafting and review of all manuscripts in Chapters 1, 3, 4, 5 and 6. Dr Carl Morrow assisted with the Caenorhabditis elegans killing assays and biolistic transformation in Chapter 3, molecular cloning of plasmids in Chapter 4, and overlap PCR for the generation of deletion constructs in Chapter 5. Mr

x Edmund Lui generated all of the proline catabolic deletion mutant and complemented strains in Chapter 5, and contributed to the generation of a single uric acid catabolic deletion mutant strain in Chapter 6. Ms Eve Chow assisted with the RNA extraction in Chapter 3, and performed the cross-species complementation in Chapter 5. Ms Liting Yang and Mr Gaseene Sebetso generated the majority of the uric acid catabolic deletion mutant strains in Chapter 6. Mr Jonathan Lim assisted with the yeast two- hydrid assays and performed the C. elegans killing assays in Chapter 4. Ms Thi Doan performed the C. elegans killing assays in Chapter 6. Ms Rebecca Allen performed the screening of the Agrobacterium library for uric acid catabolism defective mutants in Chapter 6. Ms Samantha Arras and Mr Ross Blundell assisted with the murine inhalation model of cryptococcosis experiments in Chapters 5 and 6, respectively. Ms Katherine Ormerod assisted with the molecular cloning of plasmids in Chapter 4. Dr Julianne Djordjevic demonstrated the protocol of the murine inhalation model of cryptococcosis in Chapter 3.

Statement of parts of the thesis submitted to qualify for the award of another degree

Approximately 30% of the experimental data generated in Chapter 3 was performed between 2008 – 2009, in partial fulfillment of the Bachelor of Science (Honours) degree, awarded by The University of Queensland. At least 80% of these (30%) experiments were re-performed for reproducibility sake or expanded upon during research higher degree candidature.

xi Acknowledgements

I’d like to thank my supervisor, James Fraser, for the opportunity to work in his laboratory as well as for his academic guidance. Thank you fellow lab rats, both past and present members, for all the good times: Edmund Lui, Carl Morrow, Sonia Rosenrauch, Andre Koh, Kelvin Ong, Kate Herd, John Dela Pena, Jonathan Lim, Kheng Wai Sitoh, Kate Ormerod, Samantha Arras, Linda Ly, Thomas Burger, Liting Yang, Yvette Emmanuel, Gaseene Sebetso, Thi Hoai Nam Doan and Jodeline Chua. You guys are great and have made working fun! In particular, I want to give a shout- out to my “Boss”, Edmund, cheers to the true and ever lasting friendship!

Thank you academics including my associate supervisor, Bernie Carroll, as well as my confirmation/mid-candidature/thesis review panel members, Mark Schembri, Peter O’Donoghue and Stuart Kellie, for their non-specialised but valuable general advice. I’d also like to thank my research collaborator, Julie Djordjevic, for demonstrating the protocol of the murine inhalation model of cryptococcosis. Definitely worth of mention, one of the facility/equipment specialists of the school, Bob Simpson, for teaching me anything and everything about quantitative real-time PCR – this technique became my “bread and butter” for many projects. Also, thank you Steve Mason, another facility/equipment specialist, I appreciate your patience in educating me about fluorescence microscopy. Thanks Dipti Vijayan from AIBN for the technical advice with flow cytometry.

Also deserving special praise are my best Brizzy housemates, Jason Flake and Leyla Smithers, thanks for always feeding my brain well with scrumptious, nutritious food. To my fun-loving mates in Singapore (you guys know who you are), thanks for the awesome parties, karaokes, movie dates, soccer matches and food-hunts, and of course Thai disco-ing, during my annual holidays – these entertainments have certainly helped me to recharge and press on with research away from home. Thank you to my self-proclaimed godmother Aunty Claire Lee and members of Kingdom Community Church for praying for me.

Importantly, a huge thank you to my mental and emotional pillars of strength: my wonderful family in Singapore. I’m sorry I’ve neglected you all for my research. Thanks Dad (Ronnie Lee) and Mom (Patricia Pereira) for the tremendous amount of

xii support and love, as well as constant prayers for more than two decades! My two brothers, Roy Lee and Wayne Lee, it’s amazing how we’re still close as ever.

Last but not least, thank you Jesus Christ for all the blessings; I’ll always keep the faith. “For I know the plans I have for you,” declares the Lord, “plans to prosper you and not to harm you, plans to give you hope and a future” – Jeremiah 29:11.

Ivor Russel Lee

August 2013

xiii Keywords

Cryptococcus neoformans, fungal pathogenesis, virulence, nitrogen metabolite repression, GATA factor, Nmr, proline, uric acid, ammonium

Australian and New Zealand Standard Research Classifications (ANZSRC)

ANZSRC code: 060104, Cell Metabolism, 30%

ANZSRC code: 060107, Enzymes, 30%

ANZSRC code: 060503, Microbial Genetics, 40%

Fields of Research (FoR) Classification

FoR code: 0601, Biochemistry and Cell Biology, 60%

FoR code: 0605, Microbiology, 40%

xiv Table of Contents

1.0 Introduction: Nitrogen Regulation of Virulence in Clinically Prevalent Fungal Pathogens 1

1.1 FEMS Microbiology Letters Mini-Review Article 2

2.0 Premise, Aim and Outline 13

3.0 Results: Nitrogen Metabolite Repression of Metabolism and Virulence in the Human Fungal Pathogen Cryptococcus neoformans 17

3.1 Preface 18

3.2 Genetics Research Article 19

4.0 Results: Characterization of an Nmr Homolog That Modulates GATA Factor-Mediated Nitrogen Metabolite Repression in Cryptococcus neoformans 56

4.1 Preface 57

4.2 PLoS One Research Article 58

5.0 Results: Reactive Oxygen Species Homeostasis and Virulence of the Fungal Pathogen Cryptococcus neoformans Requires an Intact Proline Catabolism Pathway 84

5.1 Preface 85

5.2 Genetics Research Article 86

xv 6.0 Results: Characterization of the Complete Uric Acid Degradation Pathway in the Fungal Pathogen Cryptococcus neoformans 113

6.1 Preface 114

6.2 PLoS One Research Article 115

7.0 Conclusions 142

7.1 Concluding Remarks 143

7.1.1 Characterisation of the GATA factor Gat1/Are1 143

7.1.2 Characterisation of the Nmr-related protein Tar1 144

7.1.3 Characterisation of the proline degradation pathway 145

7.1.4 Characterisation of the uric acid degradation pathway 146

7.1.5 Summary 146

7.2 Future Directions 147

7.2.1 Chromatin immunoprecipitation to determine the DNA-binding sites of GATA factors in the genome 147

7.2.2 Whole transcriptome sequencing to identify novel virulence- and stress adaptation-associated genes that are regulated by Tar1 and Gat1/Are1 148

7.2.3 Metabolic profiling of the proline catabolic deletion mutants using gas chromatography coupled to mass spectrometry 148

7.2.4 Virulence assays of the uric acid catabolic deletion mutants using urate oxidase-deficient mice as an animal model system 149

7.2.5 Investigating the arginine degradation pathway 149

7.3 Final Conclusions 151

xvi 7.4 References 151

Appendices 155

Appendix A mBio Research Article: A Unique Chromosomal Rearrangement in the Cryptococcus neoformans var. grubii Type Strain Enhances Key Phenotypes Associated with Virulence 155

Appendix B PLoS Pathogens Research Article: De novo GTP Biosynthesis Is Critical for Virulence of the Fungal Pathogen Cryptococcus neoformans 167

Appendix C G3-Genes/Genomics/Genetics Research Article: Comparative Genomics of Serial Isolates of Cryptococcus neoformans Reveals Gene Associated with Carbon Utilization and Virulence 184

xvii List of Figures & Tables

Figure 7-1 The predicted pathways of uric acid (black arrows), arginine (grey arrows) and proline (blue arrows), as well as central nitrogen metabolism (green arrows), in C. neoformans.

List of Abbreviations

L-DOPA 1-3,4-dihydroxyphenylalanine OHCU 2-oxo-4-hydroxy-4-carboxy-5-ureidoimidazoline HIU 5-hydroxyisourate

P5C Δ1-pyrroline-5-carboxylate AD activation domain AIDS acquired immunodeficiency syndrome N-terminus amino-terminus BD binding domain BHI brain heart infusion BLAST basic local alignment search tool BLASTp basic local alignment search tool protein-protein C-terminus carboxy-terminus cDNA complementary deoxyribonucleic acid CFU colony forming units CNS central nervous system CSF cerebrospinal fluid DMEM Dulbecco’s Modified Eagle Medium FACS fluorescence-activated cell sorting FAD flavin adenine dinucleotide (oxidised)

FADH2 flavin adenine dinucleotide (reduced) FBS fetal bovine serum FCS fetal calf serum FSC forward scatter

xviii GSA glutamate-γ-semialdehyde HIV human immunodeficiency virus LB lysogeny broth MAP mitogen-activated protein MAT mating-type MATa mating-type a

MATα mating-type α mRNA messenger ribonucleic acid MS Murashige-Shoog mtETC mitochondrial electron transport chain NAD+ nicotinamide adenine diphosphate (oxidised) NADH nicotinamide adenine diphosphate (reduced) NAT nourseothricin acetyl-transferase NEO neomycin phosphotransferase II NGM nematode growth medium OD optical density ORF open reading frame PBS phosphate buffered saline PCR polymerase chain reaction qRT-PCR quantitative real-time PCR RNS reactive nitrogen species ROS reactive oxygen species GSNO S-nitrosoglutathione SSC side scatter T-DNA transfer DNA TOR target of rapamycin VNI variety neoformans I YNB yeast nitrogen base YPD yeast peptone dextrose

xix

Try not to be a man of success but a man of value. – Albert Einstein

CHAPTER 1

Introduction Nitrogen Regulation of Virulence in Clinically Prevalent Fungal Pathogens

1 MINIREVIEW Nitrogen regulation of virulence in clinically prevalent fungal pathogens I. Russel Lee1,2, Carl A. Morrow1,2 & James A. Fraser1,2

1School of Chemistry and Molecular Biosciences, University of Queensland, Brisbane, Qld, Australia; and 2Australian Infectious Diseases Research Centre, University of Queensland, Brisbane, Qld, Australia

Correspondence: James A. Fraser, University Abstract of Queensland, 358 Molecular Biosciences Building, Cooper Road, Brisbane, Qld 4072, The habitats of fungal pathogens range from environmental to commensal, Australia. Tel.: +61 7 3365 4868; and the nutrient content of these different niches varies considerably. Upon fax: +61 7 3365 4273; infection of humans, nutrient availability changes significantly depending on e-mail: [email protected] the site and pathophysiology of infection. Nonetheless, a common feature enabling successful establishment in these niches is the ability to metabolise Received 24 April 2013; revised 17 May available nutrients including sources of nitrogen, carbon and essential metals 2013; accepted 18 May 2013. Final version published online 13 June 2013. such as iron. In particular, nitrogen source utilisation influences specific morphological transitions, sexual and asexual sporulation and virulence factor DOI: 10.1111/1574-6968.12181 production. All these physiological changes confer selective advantages to facilitate fungal survival, proliferation and colonisation. The three most well- Editor: Mark Schembri studied components of the nitrogen regulatory circuit that commonly impact fungal pathogenesis are the ammonium permeases (the nitrogen availability Keywords sensor candidate), ureases (a nitrogen-scavenging enzyme) and GATA tran- Saccharomyces cerevisiae; Candida albicans; scription factors (global regulators of nitrogen catabolism). In certain species, Aspergillus fumigatus; Coccidioides posadasii; the ammonium permease induces a morphological switch from yeast to inva- Cryptococcus neoformans. sive filamentous growth forms or infectious spores, while in others, urease is a bona fide virulence factor. In all species studied thus far, transcription of the ammonium permease and urease-encoding genes is modulated by GATA factors. Fungal pathogens therefore integrate the expression of different virulence-associated phenotypes into the regulatory network controlling nitrogen catabolism.

transcriptional regulatory mechanism that enables prefer- Background on fungal nitrogen ential utilisation of easily assimilated nitrogen sources such metabolism as ammonium and glutamine in order to conserve energy Microorganisms are constantly challenged with stressors (Wong et al., 2008). When metabolically favoured sources such as nitrogen deprivation in their ever-changing local are lacking or exhausted, nonpreferred/secondary/complex environment, and various regulatory mechanisms have nitrogenous compounds are then selectively consumed evolved to ensure that genes are appropriately expressed through the expression of permease and catabolic enzyme- according to their cellular state. In eukaryotes, studies encoding genes of their specific catabolic pathways. This into gene regulation in response to nitrogen availability metabolic response to nitrogen limitation typically necessi- were first carried out in the model ascomycetes, Saccharo- tates two distinct signals: (1) the absence of preferred myces cerevisiae, Aspergillus nidulans and Neurospora crassa. nitrogen sources leading to the activation of globally Unlike diazotrophic bacterial species, the free-living fungi acting GATA transcription factor(s), and (2) the presence are unable to fix gaseous nitrogen and therefore have to of a particular intermediate or substrate of the secondary

MICROBIOLOGY LETTERS MICROBIOLOGY obtain reduced forms of nitrogenous compounds from nitrogen pathway leading to activation of pathway-speciﬁc alternative sources. transcription factor(s). Thus, both global- and pathway- Pioneering research in the model fungi led to the under- speciﬁc factors are generally required to induce the standing of nitrogen catabolite/metabolite repression, a transcription of nonpreferred nitrogen catabolic genes.

Over the past decade, interest in nitrogen metabolic abrogates MEP2 expression and stability and results in gene regulation has expanded to encompass a number of pseudohyphal deficiency (Lorenz & Heitman, 1998). pathogens including clinical S. cerevisiae isolates, Candida Heterologous expression of MEP2 in gln3D and ure2D albicans, Aspergillus fumigatus, Coccidioides posadasii and mutants does not restore filamentation, indicating that Cryptococcus neoformans, a few of the several dozen fungal Gln3 and Ure2 have additional targets that regulate species that consistently cause disease in the human pop- dimorphism apart from Mep2 (Lorenz & Heitman, 1998). ulation. A common feature enabling these pathogenic Consistent with this notion, mep2D mutants of clinical fungi to survive in the environment and to cause disease S. cerevisiae isolates display equivalent virulence to wild in a host is inextricably linked to their ability to forage type in a murine model of disseminated disease, while nitrogen from the surroundings. This article focuses gln3D mutants are slightly attenuated in vivo (Kingsbury explicitly on our latest understanding of how nitrogen et al., 2006). This suggests that the ability to sense exter- regulation influences specific morphological changes, nal ammonium is not essential for the survival of S. cere- sporulation, virulence factor expression and catabolism- visiae in vivo and that the mep2D mutants may still have related pathogenesis activity of human pathogenic fungi. the ability to undergo pseudohyphal growth given the Such physiological responses are critical for successful availability of other serum nitrogen sources such as colonisation of a host during infection as they help evade amino acids and purines. Aside from ammonium, in the immune defence systems, facilitate attachment or invasion human blood glutamine, glutamate, asparagine, arginine, of tissues, promote dissemination or cause clinical symp- proline, uric acid, allantoin, urea and creatinine, among toms of the disease. others, can all potentially serve as a nitrogen source (Table S2). Whether the attenuated virulence of the gln3D mutants occurred through an inability to utilise a Fungal pathogens of humans wide range of nitrogen sources or through a defective pseudohyphal developmental pathway remains open to Saccharomyces cerevisiae speculation. The hemiascomycete S. cerevisiae is one of the most intensively studied eukaryotic model organisms in molec- Candida albicans ular and cell biology and has recently emerged as an opportunistic pathogen. This model yeast shares a rela- The phylogenetically related C. albicans diverged from tively close phylogenetic relationship with the more clini- S. cerevisiae c. 200–800 million years ago, and two-thirds cally prevalent Candida spp. Its natural habitat is on Oak of the C. albicans 6500 genes have clear orthologues in trees and fruits and can be widely found commercially in the S. cerevisiae genome (Jones et al., 2004). While this breweries, bakeries and scientific laboratories. Infection diploid commensal organism generally occupies the skin with S. cerevisiae presents with a variety of symptoms and and alimentary canal of healthy mammals, it can also is believed to be acquired nosocomially, through sexual cause opportunistic infections in most tissues and organs contact or by oral ingestion (McCusker, 2006). of an immunocompromised host. During infection, Under harsh conditions such as nitrogen starvation, C. albicans switches from the unicellular yeast form to S. cerevisiae a/a yeast diploid cells can undergo a dimor- multicellular, filamentous hyphal and pseudohyphal phic transition to an invasive filamentous pseudohyphal growth forms that serve as a means for tissue invasion growth form (Gimeno et al., 1992). The high-affinity (Lo et al., 1997). ammonium permease Mep2, but not the two other Neutral-to-alkaline extracellular pH and the presence ammonium permeases Mep1 and Mep3, is necessary for of CO2 are critical determinants that autoinduce hyphal differentiation into a pseudohyphal growth form during morphogenesis in C. albicans (Vylkova et al., 2011). In an ammonium-limiting conditions (Lorenz & Heitman, acidic, glucose-deprived environment (such as inside 1998). The mechanism of ammonium sensing and the phagocytic cells), C. albicans utilises amino acids as a car- signal transduction that activates filamentous growth in bon source and excretes the amino nitrogen as ammonia, S. cerevisiae, however, remains obscure. In a human host, causing the surrounding pH to rise (Vylkova et al., 2011). ammonium is present at varying concentrations in all the Mutations to STP2 (encodes a transcription factor that major tissue locations (Supporting Information, Table regulates the expression of amino acid permease genes), S1). DUR1,2 (encodes urea amidolyase that catabolises urea to

Activation of MEP2 expression requires at least one of ammonia and CO2) and ATO5 (encodes a putative the GATA factors Gln3 or Gat1, depending on the nitro- ammonia exporter), either abolish or delay pH neutralisa- gen source present (Marini et al., 1997). Mutations to tion (Vylkova et al., 2011). Consistent with this notion, Gln3 or its cytoplasmic interaction partner Ure2 genes for arginine metabolism are induced following

ª 2013 Federation of European Microbiological Societies FEMS Microbiol Lett 345 (2013) 77–84 Published by John Wiley & Sons Ltd. All rights reserved Nitrogen regulation of fungal pathogenesis 79 internalisation by macrophages; CAR1-encoded arginase isoenzyme SAP2 when proteins are the only nitrogen converts arginine to ornithine and urea for Dur1,2 to act source available (Dabas & Morschhauser, 2008). Secretion upon (Lorenz et al., 2004). dur1,2D mutants are defective of aspartic proteases is one of the earliest recognised viru- in germ tube formation in the presence of urea or argi- lence attributes of C. albicans as they directly modulate nine and are unable to pierce the macrophages cell mem- host immune response via their role in nutrient assimila- brane to facilitate escape following phagocytosis (Ghosh tion and degradation of extracellular matrix proteins and et al., 2009). Urea metabolism via the biotin-dependent immunoglobulins (Naglik et al., 2004). Consequently, Dur1,2 and dimorphic switching are therefore important sap2D mutants also exhibit reduced virulence in various for resistance against innate host immunity. In the animal models of disseminated candidiasis (Hube et al., human blood and cerebrospinal fluid, urea is present at 1997). very high concentrations (Table S2). Starvation for amino acids or ammonium is also a Aspergillus fumigatus potent morphogenetic signal that promotes switching from the budding yeast to filamentous hyphal form. In Aspergillus fumigatus is an opportunistic pathogen that support of this observation, the regulator of the general causes life-threatening invasive pulmonary infections in amino acid starvation response, Gcn4, has been impli- neutropenic and immunocompromised patients. This cated in dimorphism (Tripathi et al., 2002). In addition, saprotrophic fungus is found ubiquitously in decaying dimorphic switching is Mep2 ammonium permease- soils and on organic debris, where it plays an important dependent during nitrogen-limiting conditions (Biswas & role in the recycling of carbon and nitrogen, which Morschhauser, 2005). So, unlike S. cerevisiae mep2D demands metabolic diversity (Latge, 1999). Like its well- mutants that exhibit a filamentous growth defect during studied sister species A. nidulans, A. fumigatus has the ammonium-limiting conditions, the filamentation defect ability to utilise a wide array of nitrogen sources, and this of C. albicans mep2D mutants is under general nitrogen nutritional versatility has led to its widespread distribu- limitation independent of nitrogen source (Lorenz & tion. Transmission of A. fumigatus occurs through the Heitman, 1998; Biswas & Morschhauser, 2005). inhalation of germinated conidiospores that can lead to Similar to their counterparts in S. cerevisiae, both the hyphal invasion of lung parenchymal tissue through the GATA factors Gln3 and Gat1 activate the expression of bronchioles, often resulting in severe asthma, sinusitis MEP2 in C. albicans (Dabas & Morschhauser, 2007). While and even death. Gln3 is required for nitrogen limitation–induced filamenta- Nitrogen starvation has been demonstrated to be tion, Gat1 is surprisingly dispensable for filamentation under a prominent host-imposed stress during early-stage both repressing and nonrepressing conditions (Limjindap- A. fumigatus infection (McDonagh et al., 2008). Unlike orn et al., 2003; Dabas & Morschhauser, 2007). This C. albicans, whereby dimorphic transition from yeast to conundrum is explained by the observation that a func- filamentous form is required for pathogenesis, after spore tional Mep2 is not required for filamentation in gat1D germination A. fumigatus exists only in its monomorphic mutants, suggesting that the loss of GAT1 may lead to acti- mould state. Instead, proper regulation of pathways vation of other filamentous growth-inducing signalling involved in growth control and nutrient sensing are pathways (Dabas & Morschhauser, 2007). gat1D mutants the critical determinants that influence the ability of are avirulent in a murine model of disseminated candidia- A. fumigatus to cause disease. The cross-pathway control sis, and this attenuated phenotype is not associated with its system of amino acid biosynthesis is an example of a impaired serum nitrogen source utilisation ability, imply- growth control pathway that is important for pulmonary ing that Gat1 regulates additional unidentified genes aspergillosis. The bZIP transcription factor CpcA, an required for pathogenesis (Limjindaporn et al., 2003). orthologue of C. albicans Gcn4, activates the expression Indeed, more recent findings have shown that both Gat1 of target genes encoding components of the amino acid and Gln3 regulate the expression of DUR1,2-encoded urea biosynthetic pathways during amino acid starvation and amidolyase (Navarathna et al., 2011). Virulence of the other stresses (Rhodes & Brakhage, 2006). Although gln3D mutants is also reduced in comparison with wild cpcAD mutants display wild-type growth on ammonium type, albeit to a lesser extent than the gat1D mutants (Liao or nitrate as the sole nitrogen source, the mutants are et al., 2008). unable to derepress the expression of genes associated Additionally, while both Gln3 and Gat1 also regulate with amino acid biosynthesis when starved for amino the expression of the proteolytically activated transcrip- acids (Krappmann et al., 2004). Independent neutropenic tion factor-encoding gene STP1, Gat1 plays the predomi- murine models of pulmonary aspergillosis experiments nant role (Dabas & Morschhauser, 2008). Stp1 mediates have demonstrated that CpcA regulates pathogenesis the induction of the major secreted aspartic protease (Krappmann et al., 2004).

Nutrient sensing-related pathways also contribute to stresses. The production of toxic nitric oxide is mediated A. fumigatus virulence. RhbA, a member of the family of by the inducible nitric oxide synthase that requires argi- Ras-related proteins, is implicated in sensing nitrogen nine as a substrate. Like C. albicans, arginine metabolism availability (Panepinto et al., 2002). Transcription of rhbA by C. posadasii also serves as a means of innate immune occurs throughout the asexual developmental cycle, sug- evasion (Ghosh et al., 2009). One of the mechanisms gesting that RhbA functions in normal cellular processes employed by this respiratory pathogen to facilitate survival instead of specialised developmental programmes; rhbA in the hostile host environment is to induce the produc- expression markedly increases in response to nitrogen tion of macrophage cytosolic arginase I (Hung et al., starvation, however (Panepinto et al., 2002). rhbAD 2007). Arginase I hydrolyses arginine to ornithine, a mutants exhibit wild-type growth on rich media or on substrate for ornithine decarboxylase, a key enzyme in deﬁned medium supplemented with ammonium as the the polyamine biosynthetic pathway and a potential sole nitrogen source, but growth rates are impaired on regulator of C. posadasii parasitic cell differentiation secondary nitrogen sources such as proline, histidine and (Guevara-Olvera et al., 2000; Iniesta et al., 2002). Host- nitrate (Panepinto et al., 2003). A murine model of pul- derived ornithine at the sites of fungal infection may monary aspergillosis indicates that the nutrient acquisi- enhance C. posadasii growth and proliferation through tion role of RhbA is required for pathogenesis (Panepinto the provision of a pool of monoamines that the para- et al., 2003). sitic cells could utilise for the biosynthesis of glutamine, Additionally, rhbAD mutants display enhanced sensitivity proline and polyamines (Iniesta et al., 2002). In the to the antibiotic drug rapamycin, suggesting that RhbA macrophage, arginase I competes with nitric oxide syn- functions in the nutrient-sensing target of rapamycin thase for arginine, causing a reduction in the production (TOR) signalling pathway (Panepinto et al., 2003). The of nitric oxide but an increase in ornithine and urea highly conserved TOR pathway acts through the GATA production, thereby promoting intracellular survival of factors Gln3 and Gat1 in both S. cerevisiae and C. posadasii (Moali et al., 1998). Treatment of C. posad- C. albicans to control growth in response to nutrient asii-infected mice with the arginase physiological inhibitor, availability (Beck & Hall, 1999; Liao et al., 2008). Multi- N-hydroxy-nor-L-arginine, results in augmented host sur- ple lines of evidence suggest that mutants lacking the vival, indicating that arginase activity facilitates disease GATA factor known as AreA in A. fumigatus are also less development (Cole et al., 2006). virulent than wild type in murine models of invasive In C. albicans, arginine catabolism works in concert pulmonary aspergillosis (Hensel et al., 1998). According to with DUR1,2-encoded urea amidolyase to promote Hensel et al., these observations are likely caused by the hyphal switching that plays an important role in facili- reduced competitive ﬁtness of the areAD mutants given its tating escape from host defences (Lorenz et al., 2004; inability to utilise a wide array of nitrogen sources. Ghosh et al., 2009). Unlike hemiascomycetes that use Dur1,2 for urea utilisation, most other higher fungi including species of Coccidioides and Cryptococcus use Coccidioides posadasii the nickel-containing enzyme urease (Navarathna et al., Filamentous, saprobic Coccidioides spp. are members of 2010). Urease of C. posadasii catalyses the hydrolysis of the Onygenales that are found in deserts and semi-arid both fungus- and host-derived urea to produce ammo- soils (often several inches below the surface of the soil nia (Mirbod et al., 2002). Ammonia and urease released where nitrogen levels are higher) and have the ability to by C. posadasii during the endosporulation phase of the survive harsh environmental conditions such as alkaline parasitic cycle creates an alkaline microenvironment that soil with high salinity (Saubolle, 1996). Two species have likely interferes with host defence function and contrib- been characterised based on molecular and biogeographi- utes to exacerbated lung tissue damage (Cole, 1997; cal differences: Coccidioides immitis, which appears to Mirbod-Donovan et al., 2006). Consistent with this be restricted to San Joaquin Valley of California, and notion, the expression of URE-encoded urease is upreg- C. posadasii, which is widespread throughout regions of ulated during the parasitic cycle endosporulation stage endemicity in south-western USA and northern Mexico of C. posadasii (Mirbod et al., 2002). Mice intranasally (Fisher et al., 2002). The multicellular hyphal form of this infected with ureD mutants exhibit a better-organised ascomycete can septate into single-celled arthroconidia granulomatous response to infection and increase clear- and be inhaled, to cause potentially life-threatening pneu- ance of the pathogen, leading to better survival rates monia in both immunocompromised and immunocom- compared with mice inoculated with wild type (Mirbod- petent individuals (Cole et al., 2006). Donovan et al., 2006). Urease production is therefore The host macrophages in lungs kill invading microor- required for the pathogenesis of C. posadasii parasitic ganisms largely through nitrosative and oxidative cells.

2008). Although both mating in C. neoformans and Cryptococcus neoformans ammonium-dependent filamentous growth in S. cerevisiae Cryptococcosis is the leading invasive fungal infection in and C. albicans appear superficially similar, the morpho- the world today, responsible for c. 1 million cases of logical changes associated with mating are distinct as it cryptococcal meningitis per year. Over the past century, involves the formation of hyphae, basidia and spores this opportunistic basidiomycete has risen from the status (Morrow & Fraser, 2009). Like the S. cerevisiae ammo- of medical curiosity to a common but life-threatening nium permease Mep2, C. neoformans Amt2 is not central nervous system pathogen of the immunocompro- required for virulence (Kingsbury et al., 2006; Rutherford mised population, particularly patients with AIDS (Casa- et al., 2008). However, initiation of invasive growth in devall & Perfect, 1998). The primary ecological niche of response to nitrogen limitation is dependent on Amt2 the ubiquitous C. neoformans is pigeon guano, and the (Rutherford et al., 2008). majority of the nitrogen is present in the form of uric One particular nitrogen-scavenging enzyme in acid, with the remainder consisting primarily of xanthine, C. neoformans that has evoked as much interest as it has urea and creatinine (Staib et al., 1978). Normally isolated in C. posadasii is urease. Apart from converting the abun- as a budding yeast, C. neoformans can also undergo a dant urea to a usable nitrogen source, urease promotes dimorphic transition to filamentous growth form via yeast sequestration within the microcapillary beds of the mating or monokaryotic fruiting that leads to the pro- brain during haematogenous dissemination, facilitating duction of potentially infectious airborne basidiospores C. neoformans crossing of the blood-brain barrier (Morrow & Fraser, 2009). (Olszewski et al., 2004). Both murine intravenous and Progression through these sexual cycles is regulated by inhalation models of cryptococcosis revealed that mice a specialised region of the genome that confers cell-type infected with ure1D mutants lacking urease survived identity known as the mating-type locus, and environ- significantly longer than mice infected with wild type mental cues such as darkness, desiccation and nitrogen (Cox et al., 2000; Olszewski et al., 2004). However, direct starvation (Fraser et al., 2004). During ammonium limi- co-inoculation of wild-type and ure1D mutants into the tation, the high-affinity ammonium permease Amt2 is cerebrospinal fluid of corticosteroid-treated rabbits or into required for the induction of mating (Rutherford et al., the brain of mice revealed no significant differences in

Table 1. Roles of a selected panel of fungal nitrogen regulatory proteins on virulence-associated phenotypes

Mutant phenotype Gene Protein in mice Other phenotypes in vitro References

S.c. MEP2 Ammonium permease Wild type Pseudohyphal growth defect Lorenz & Heitman (1998) and Kingsbury et al. (2006) C.a. MEP2 Ammonium permease Unknown Pseudohyphal growth defect Biswas & Morschhauser (2005) C.n. AMT2 Ammonium permease Wild type Inability to initiate invasive growth form Rutherford et al. (2008) S.c. DUR1,2 Urea amidolyase Wild type Inability to utilise urea as a nitrogen source Kingsbury et al. (2006) C.a. DUR1,2 Urea amidolyase Unknown Inability to utilise urea as a nitrogen source. Ghosh et al. (2009) and Vylkova et al. Germ tube formation defect (2011) C.p. URE Urease Attenuated Inability to utilise urea as a nitrogen source Mirbod et al. (2002) and Mirbod-Donovan et al. (2006) C.n. URE1 Urease Attenuated Inability to utilise urea as a nitrogen source Cox et al. (2000) and Olszewski et al. (2004) S.c. GLN3 GATA factor Attenuated Inability to utilise a wide array of nitrogen Lorenz & Heitman (1998) and Kingsbury sources. Pseudohyphal growth defect et al. (2006) C.a. GLN3 GATA factor Attenuated Inability to utilise a wide array of nitrogen Dabas & Morschhauser (2007) and sources. Pseudohyphal growth defect Liao et al. (2008) C.a. GAT1 GATA factor Avirulent Inability to utilise a wide array of nitrogen Limjindaporn et al. (2003) sources A.f. areA GATA factor Attenuated Inability to utilise a wide array of nitrogen Hensel et al. (1998) sources C.n. GAT1 GATA factor Slight Inability to utilise a wide array of nitrogen Lee et al. (2011, 2012) hypervirulence sources. Basidiospore production defect. Overproduction of melanin. Enhanced growth at febrile temperature

S.c., Saccharomyces cerevisiae; C.a., Candida albicans; A.f., Aspergillus fumigatus; C.p., Coccidioides posadasii; C.n., Cryptococcus neoformans.

FEMS Microbiol Lett 345 (2013) 77–84 ª 2013 Federation of European Microbiological Societies Published by John Wiley & Sons Ltd. All rights reserved 82 I. Russel Lee et al. colony counts when strains were recovered (Cox et al., to utilise a wide array of nitrogen sources; not surprisingly, 2000; Olszewski et al., 2004). Urease is therefore impli- GATA factor mutants of S. cerevisiae, C. albicans and cated in the dissemination process of pathogenesis, but A. fumigatus exhibit attenuated virulence. The other major the importance of this virulence factor appears infection nitrogen-associated mediator of fungal adaptation and site specific as its activity is not required for C. neoformans pathogenesis, whose transcription is also dependent on proliferation in the host brain or cerebrospinal fluid. GATA factors, is the virulence factor and catabolic enzyme Instead, the enzyme laccase is crucial for C. neoformans urease that has been characterised in C. posadasii and survival in the brain (Nosanchuk et al., 2000). Both urease C. neoformans. Activities of the common proteins among and laccase act synergistically to promote invasion of the the fungal species described in this article are summarised host central nervous system. While urease facilitates inva- in Table 1. The complex interplay between the host envi- sion through brain microcapillaries, laccase enables yeast ronment and the fungal pathogen is also a critical determi- growth and proliferation within the brain parenchyma. nant; host immune cells generate highly toxic reactive Laccase catalyses the oxidation of catecholamines found in nitrogen species against microbial invaders, but the nitro- the central nervous system to produce the virulence factor sative stress response of the pathogen in catabolising and melanin that protects against toxic free radicals generated detoxifying these chemical compounds enable its survival by the host immune system (Nosanchuk et al., 2000). during infection. Clearly, it is becoming increasingly appar- Interestingly, the nitrogen responsive GATA factor Gat1/ ent that nitrogen regulation is a major factor modulating Are1 negatively regulates LAC1-encoded laccase expression pathogenesis in fungi. An important area for future and consequently melanin production (Lee et al., 2011, research is to dissect the nature and origin of intracellular 2012). Additionally, Gat1/Are1 impacts other virulence signals that govern the metabolic processes involved in attributes including infectious basidiospore production, nitrogen acquisition, in the context of host-pathogen antiphagocytic capsule biosynthesis and febrile body interactions. temperature growth (Lee et al., 2011). These findings are consistent with classic studies demonstrating that nitrogen Acknowledgements source utilisation affects cryptococcal virulence factor expression (Chaskes & Tyndall, 1975; Staib et al., 1978). We thank Michael Hynes for the critical reading of this MiniReview article. We also would like to apologise for being unable to discuss all the literature related to the Conclusions topic covered in this article. The authors report no Pathogenic fungi are an increasing public health problem conflict of interest. due to the growing immunocompromised population, particularly due to the advent of the AIDS pandemic. These fungi are found in diverse and often hostile conditions References where they are constantly challenged by nutrient availabil- Beck T & Hall MN (1999) The TOR signalling pathway ity such as nitrogen limitation. The ability to survive, pro- controls nuclear localization of nutrient-regulated liferate and colonise their respective niches is therefore transcription factors. Nature 402: 689–692. dependent on well-coordinated regulation of nitrogen Biswas K & Morschhauser J (2005) The Mep2p ammonium metabolism. While filamentous fungi have the ability to permease controls nitrogen starvation-induced filamentous utilise a wide variety of compounds as nitrogen sources, growth in Candida albicans. Mol Microbiol 56: 649–669. yeast-like growing fungi generally assimilate a narrower Casadevall A & Perfect JR (1998) Cryptococcus neoformans. range of nitrogen sources. Nonetheless, numerous fungal ASM Press, Washington, DC. species such as S. cerevisiae, C. albicans and C. neoformans Chaskes S & Tyndall RL (1975) Pigment production by have the ability to undergo a dimorphic transition that Cryptococcus neoformans from para- and ortho-Diphenols: 1 – involves the interconversion between multiple growth pat- effect of the nitrogen source. J Clin Microbiol : 509 514. terns in response to environmental cues. This morphologi- Cole GT (1997) Ammonia production by Coccidioides immitis cal developmental pathway, at least in part controlled by and its possible significance to the host-fungus interplay. ammonium permeases, is likely a scavenging mechanism Host-Fungus Interplay (Bossche H, Stevens DA & Odds FC, that enables nonmotile organisms to procure nutrients and eds), pp. 247–263. Plenum Press, New York, NY. has long been studied as a model of fungal pathogenesis. Cole GT, Xue J, Seshan K, Borra P, Borra R, Tarcha E, Expression of the ammonium permease genes is regulated Schaller R, Yu J & Hung C (2006) Virulence mechanisms of by GATA transcription factors that are master switches Coccidioides. Molecular Principles of Fungal Pathogenesis employed to globally control nitrogen metabolism. The (Heitman J, Filler SG, Edwards JE & Mitchell AP, eds), pp. lack of GATA factor most obviously results in an inability 363–391. ASM Press, Washington, DC.

Cox GM, Mukherjee J, Cole GT, Casadevall A & Perfect JR activator CpcA contributes significantly to the virulence of (2000) Urease as a virulence factor in experimental this fungal pathogen. Mol Microbiol 52: 785–799. cryptococcosis. Infect Immun 68: 443–448. Latge JP (1999) Aspergillus fumigatus and aspergillosis. Clin Dabas N & Morschhauser J (2007) Control of ammonium Microbiol Rev 12: 310–350. permease expression and filamentous growth by the GATA Lee IR, Chow EW, Morrow CA, Djordjevic JT & Fraser JA transcription factors GLN3 and GAT1 in Candida albicans. (2011) Nitrogen metabolite repression of metabolism and Eukaryot Cell 6: 875–888. virulence in the human fungal pathogen Cryptococcus Dabas N & Morschhauser J (2008) A transcription factor neoformans. Genetics 188: 309–323. regulatory cascade controls secreted aspartic protease Lee IR, Lim JW, Ormerod KL, Morrow CA & Fraser JA (2012) expression in Candida albicans. Mol Microbiol 69: 586–602. Characterization of an Nmr Homolog That Modulates Fisher MC, Koenig GL, White TJ & Taylor JW (2002) GATA Factor-Mediated Nitrogen Metabolite Repression in Molecular and phenotypic description of Coccidioides Cryptococcus neoformans. PLoS ONE 7: e32585. posadasii sp. nov., previously recognized as the Liao WL, Ramon AM & Fonzi WA (2008) GLN3 encodes a non-California population of Coccidioides immitis. Mycologia global regulator of nitrogen metabolism and virulence of 94:73–84. C. albicans. Fungal Genet Biol 45: 514–526. Fraser JA, Diezmann S, Subaran RL, Allen A, Lengeler KB, Limjindaporn T, Khalaf RA & Fonzi WA (2003) Nitrogen Dietrich FS & Heitman J (2004) Convergent evolution of metabolism and virulence of Candida albicans require the chromosomal sex-determining regions in the animal and GATA-type transcriptional activator encoded by GAT1. Mol fungal kingdoms. PLoS Biol 2: e384. Microbiol 50: 993–1004. Ghosh S, Navarathna DH, Roberts DD, Cooper JT, Atkin AL, Lo HJ, Kohler JR, DiDomenico B, Loebenberg D, Cacciapuoti Petro TM & Nickerson KW (2009) Arginine-induced germ A & Fink GR (1997) Nonfilamentous C. albicans mutants tube formation in Candida albicans is essential for escape are avirulent. Cell 90: 939–949. from murine macrophage line RAW 264.7. Infect Immun 77: Lorenz MC & Heitman J (1998) The MEP2 ammonium 1596–1605. permease regulates pseudohyphal differentiation in Gimeno CJ, Ljungdahl PO, Styles CA & Fink GR (1992) Saccharomyces cerevisiae. EMBO J 17: 1236–1247. Unipolar cell divisions in the yeast S. cerevisiae lead to Lorenz MC, Bender JA & Fink GR (2004) Transcriptional filamentous growth: regulation by starvation and RAS. Cell response of Candida albicans upon internalization by 68: 1077–1090. macrophages. Eukaryot Cell 3: 1076–1087. Guevara-Olvera L, Hung CY, Yu JJ & Cole GT (2000) Marini AM, Soussi-Boudekou S, Vissers S & Andre B (1997) A Sequence, expression and functional analysis of the family of ammonium transporters in Saccharomyces Coccidioides immitis ODC (ornithine decarboxylase) gene. cerevisiae. Mol Cell Biol 17: 4282–4293. Gene 242: 437–448. McCusker JH (2006) Saccharomyces cerevisiae: an emerging Hensel M, Arst HN Jr, Aufauvre-Brown A & Holden DW (1998) and model pathogenic fungus. Molecular Principles of Fungal The role of the Aspergillus fumigatus areA gene in invasive Pathogenesis (Heitman J, Filler SG, Edwards JE & Mitchell pulmonary aspergillosis. Mol Gen Genet 258: 553–557. AP, eds), pp. 245–259. ASM Press, Washington, DC. Hube B, Sanglard D, Odds FC, Hess D, Monod M, Schafer W, McDonagh A, Fedorova ND, Crabtree J et al. (2008) Brown AJ & Gow NA (1997) Disruption of each of the Sub-telomere directed gene expression during initiation of secreted aspartyl proteinase genes SAP1, SAP2, and SAP3 of invasive aspergillosis. PLoS Pathog 4: e1000154. Candida albicans attenuates virulence. Infect Immun 65: Mirbod F, Schaller RA & Cole GT (2002) Purification and 3529–3538. characterization of urease isolated from the pathogenic Hung CY, Xue J & Cole GT (2007) Virulence mechanisms of fungus Coccidioides immitis. Med Mycol 40:35–44. Coccidioides. Ann NY Acad Sci 1111: 225–235. Mirbod-Donovan F, Schaller R, Hung CY, Xue J, Reichard U Iniesta V, Gomez-Nieto LC, Molano I, Mohedano A, Carcelen & Cole GT (2006) Urease produced by Coccidioides J, Miron C, Alonso C & Corraliza I (2002) Arginase I posadasii contributes to the virulence of this respiratory induction in macrophages, triggered by Th2-type cytokines, pathogen. Infect Immun 74: 504–515. supports the growth of intracellular Leishmania parasites. Moali C, Boucher JL, Sari MA, Stuehr DJ & Mansuy D (1998) Parasite Immunol 24: 113–118. Substrate specificity of NO synthases: detailed comparison Jones T, Federspiel NA, Chibana H et al. (2004) The diploid of L-arginine, homo-L-arginine, their N omega-hydroxy genome sequence of Candida albicans. P Natl Acad Sci USA derivatives, and N omega-hydroxynor-L-arginine. 101: 7329–7334. Biochemistry 37: 10453–10460. Kingsbury JM, Goldstein AL & McCusker JH (2006) Role of Morrow CA & Fraser JA (2009) Sexual reproduction and nitrogen and carbon transport, regulation, and metabolism dimorphism in the pathogenic basidiomycetes. FEMS Yeast genes for Saccharomyces cerevisiae survival in vivo. Eukaryot Res 9: 161–177. Cell 5: 816–824. Naglik J, Albrecht A, Bader O & Hube B (2004) Candida Krappmann S, Bignell EM, Reichard U, Rogers T, Haynes K & albicans proteinases and host/pathogen interactions. Cell Braus GH (2004) The Aspergillus fumigatus transcriptional Microbiol 6: 915–926.

Navarathna DH, Harris SD, Roberts DD & Nickerson KW invasive growth and mating in Cryptococcus neoformans. (2010) Evolutionary aspects of urea utilization by fungi. Eukaryot Cell 7: 237–246. FEMS Yeast Res 10: 209–213. Saubolle MA (1996) Life cycle and epidemiology of Navarathna DH, Das A, Morschhauser J, Nickerson KW & Coccidioides immitis. Coccidiodomycosis (Einstein H & Roberts DD (2011) Dur3 is the major urea transporter in Catanzaro C, eds), pp. 1–8. National Foundation for Candida albicans and is co-regulated with the urea Infectious Diseases, Washington, DC. amidolyase Dur1,2. Microbiology 157: 270–279. Staib F, Grave B, Altmann L, Mishra SK, Abel T & Blisse A Nosanchuk JD, Rosas AL, Lee SC & Casadevall A (2000) (1978) Epidemiology of Cryptococcus neoformans. Melanisation of Cryptococcus neoformans in human brain Mycopathologia 65:73–76. tissue. Lancet 355: 2049–2050. Tripathi G, Wiltshire C, Macaskill S, Tournu H, Budge S & Olszewski MA, Noverr MC, Chen GH, Toews GB, Cox GM, Brown AJ (2002) Gcn4 co-ordinates morphogenetic and Perfect JR & Huffnagle GB (2004) Urease expression by metabolic responses to amino acid starvation in Candida Cryptococcus neoformans promotes microvascular albicans. EMBO J 21: 5448–5456. sequestration, thereby enhancing central nervous system Vylkova S, Carman AJ, Danhof HA, Collette JR, Zhou H & invasion. Am J Pathol 164: 1761–1771. Lorenz MC (2011) The fungal pathogen Candida albicans Panepinto JC, Oliver BG, Amlung TW, Askew DS & Rhodes autoinduces hyphal morphogenesis by raising extracellular JC (2002) Expression of the Aspergillus fumigatus rheb pH. MBio 2: e00055–e00011. homologue, rhbA, is induced by nitrogen starvation. Fungal Wong KH, Hynes MJ & Davis MA (2008) Recent advances in Genet Biol 36: 207–214. nitrogen regulation: a comparison between Saccharomyces Panepinto JC, Oliver BG, Fortwendel JR, Smith DL, Askew DS cerevisiae and ﬁlamentous fungi. Eukaryot Cell 7: 917–925. & Rhodes JC (2003) Deletion of the Aspergillus fumigatus gene encoding the Ras-related protein RhbA reduces virulence in a model of Invasive pulmonary aspergillosis. Supporting Information Infect Immun 71: 2819–2826. Rhodes JC & Brakhage AA (2006) Molecular determinants of Additional Supporting Information may be found in the virulence in Aspergillus fumigatus. Molecular Principles of online version of this article: Fungal Pathogenesis (Heitman J, Filler SG, Edwards JE & Mitchell AP, eds), pp. 333–345. ASM Press, Washington, Table S1. The presence or absence of a selected panel of DC. nitrogen sources in various tissues of a healthy individual. Rutherford JC, Lin X, Nielsen K & Heitman J (2008) Amt2 Table S2. The concentration (lM) of a selected panel of permease is required to induce ammonium-responsive nitrogen sources in various bioﬂuids of a healthy individual.

Supporting Information http://onlinelibrary.wiley.com/doi/10.1111/1574-6968.12181/suppinfo

Nitrogen Regulation of Virulence in Clinically Prevalent Fungal Pathogens

I. Russel Lee, Carl A. Morrow & James A. Fraser

DOI: 10.1111/1574-6968.12181

Table S1. The presence or absence of a selected panel of nitrogen sources in various tissues of a healthy individual. Ammonium, asparagine and proline are present in all major tissue locations.

N source

mine Urea Proline Arginine Uric acid Uric Glutamate Creatinine Tissue Gluta Asparagine Ammonium

Adipose tissue + + + + + + + - +

Adrenal cortex + - - + + + - - -

Adrenal medulla + - + + - + - - -

Bladder + - - + + + + - +

Brain + - - + - + + - -

Epidermis + - + + + + + - -

Erythrocyte + - - + - + + - -

Fibroblasts + + + + + + - - +

Gonads + - - + + + - - -

Gut + + - + - + - - -

Intestine + + + + + + + - -

Kidney + + + + + + + + +

Liver + - - + - + + + +

Muscle + + + + + + + - +

Myelin + + + + + + - - -

Nerve cells + - + + + + - - +

Neuron + + + + + + - - -

Pancreas + + + + + + - - +

Placenta + + + + + + + - +

Platelet + - + + + + + - +

Prostate + - - + + + - - +

Skeletal muscle + + + + + + - - +

Skin + - - + - + - + -

Spleen + + + + + + + - +

Stratum corneum + + + + + + - - -

Testes + + - + + + - - +

2 SI Thyroid gland + - - + - + - - +

+ Present; - Absent

Data obtained from the Human Metabolome Database website (http://www.hmdb.ca/).

3 SI Table S2. The concentration (µM) of a selected panel of nitrogen sources in various biofluids of a healthy individual. Urea is present at very high levels in both the human blood and cerebrospinal fluid (CSF).

N source

acid

Urea Proline Arginine Uric Allantoin Glutamate Creatinine Biofluid Glutamine Asparagine Ammonium Blood 29.0 586.0 24.0 41.0 80.0 168.0 256.8 28.5 6500.0 74.1

CSF 11.9 444.0 1.8 5.4 18.3 4.0 29.3 9.3 4160.0 65.2

Saliva - >10.0 >10.0 - - >10.0 <1.0 - - >10.0

Data obtained from the Human Metabolome Database website (http://www.hmdb.ca/).

4 SI

CHAPTER 2

Premise, Aim and Outline

13 Thesis premise and aim

From its humble beginning as an isolate obtained from peach juice in 1894 and its link to a single case of human disease a year later, the basidiomycetous pathogen Cryptococcus neoformans has evolved into a leading cause of life-threatening meningitis especially in the immunocompromised population today [1]. More than 30 million people are living with AIDS/HIV worldwide, a large portion of whom reside in sub-Saharan Africa. Recent estimates indicate that there are one million cases of cryptococcal infections globally each year in AIDS patients, leading to approximately 625,000 deaths [2]. Extensive research has been conducted towards understanding C. neoformans adaptation to key features within a mammalian host environment including carbon, iron and copper limitation, pH, as well as hypoxia [3,4,5,6,7,8].

However, there is generally little known about regulation of nitrogen metabolism by C. neoformans in the literature. As one would anticipate, nitrogen concentrations in the human lungs, brain and other anatomical sites are significantly lower compared to C. neoformans ecological niche of pigeon guano and decaying tree barks. Hence, in order to establish infection in a host, C. neoformans needs to sense and adapt to nitrogen limitation. The mechanisms involved in nitrogen sensing and adaptation to low nitrogen conditions have been extensively studied in the model ascomyctes Saccharomyces cerevisiae, Aspergillus nidulans and Neurospora crassa [9,10,11]. However, this important aspect towards understanding the pathobiology of C. neoformans remains elusive.

In view of this lack of knowledge, the broad aim of my research as a whole focuses on investigating the interplay between nitrogen catabolism and virulence in the hope of identifying novel genes and cellular processes important for C. neoformans pathogenesis. The main objective is to gain a deeper understanding of the nitrogen requirements of C. neoformans and its consequent metabolic gene regulation in response to nitrogen availability, that may contribute to new strategies for pathogen control via identification of potential therapeutic targets from key components of the nitrogen regulatory pathway. Unless otherwise stated, studies were almost exclusively carried out on C. neoformans H99, the type strain for variety grubii (most clinically prevalent form responsible for >90% of cryptococcal infections worldwide) that was disseminated to the research community [12].

14 Thesis outline

This thesis is presented in a slightly unconventional format, as a collection of minireview and research papers accepted in peer-reviewed journals. Accordingly, the introduction chapter consisted of an abstract, literature review and conclusion; each results chapter is a complete body of work comprising of an abstract, introduction, materials and methods, results, and discussion. Because the specific aim of most results chapters ensued from the findings of the previous, I have taken the liberty of writing a preface before the start of each results chapter to guide you through this comprehensive nitrogen story in a progressive fashion. Briefly, Chapters 3 and 4 explore C. neoformans global nitrogen regulation (GATA factor and Nmr proteins), while Chapters 5 and 6 dissect key nitrogen catabolic pathways of interest (proline and uric acid degradation). Finally, Chapter 7 summarises the research outcomes of the thesis and the new insights it yields to the field, and proffers some future directions.

References

1. Sanfelice F (1894) Contributo alla morfologia e biologia dei blastomiceti che si sviluppano nei succhidi alcuni frutti. Ann Igien 4: 463-495. 2. Park BJ, Wannemuehler KA, Marston BJ, Govender N, Pappas PG, et al. (2009) Estimation of the current global burden of cryptococcal meningitis among persons living with HIV/AIDS. AIDS 23: 525-530. 3. Chun CD, Liu OW, Madhani HD (2007) A link between virulence and homeostatic responses to hypoxia during infection by the human fungal pathogen Cryptococcus neoformans. PLoS Pathog 3: e22. 4. Jung WH, Sham A, White R, Kronstad JW (2006) Iron regulation of the major virulence factors in the AIDS-associated pathogen Cryptococcus neoformans. PLoS Biol 4: e410. 5. Kronstad J, Saikia S, Nielson ED, Kretschmer M, Jung W, et al. (2012) Adaptation of Cryptococcus neoformans to mammalian hosts: integrated regulation of metabolism and virulence. Eukaryot Cell 11: 109-118.

15 6. Lin X, Huang JC, Mitchell TG, Heitman J (2006) Virulence attributes and hyphal growth of C. neoformans are quantitative traits and the MATalpha allele enhances filamentation. PLoS Genet 2: e187. 7. O'Meara TR, Norton D, Price MS, Hay C, Clements MF, et al. (2010) Interaction of Cryptococcus neoformans Rim101 and protein kinase A regulates capsule. PLoS Pathog 6: e1000776. 8. Price MS, Betancourt-Quiroz M, Price JL, Toffaletti DL, Vora H, et al. (2011) Cryptococcus neoformans requires a functional glycolytic pathway for disease but not persistence in the host. MBio 2: e00103-00111. 9. Magasanik B, Kaiser CA (2002) Nitrogen regulation in Saccharomyces cerevisiae. Gene 290: 1-18. 10. Marzluf GA (1997) Genetic regulation of nitrogen metabolism in the fungi. Microbiol Mol Biol Rev 61: 17-32. 11. Wong KH, Hynes MJ, Davis MA (2008) Recent advances in nitrogen regulation: a comparison between Saccharomyces cerevisiae and filamentous fungi. Eukaryot Cell 7: 917-925. 12. Franzot SP, Salkin IF, Casadevall A (1999) Cryptococcus neoformans var. grubii: separate varietal status for Cryptococcus neoformans serotype A isolates. J Clin Microbiol 37: 838-840.

CHAPTER 3

Results Nitrogen Metabolite Repression of Metabolism and Virulence in the Human Fungal Pathogen Cryptococcus neoformans

17 3.1 Preface

The fungal pathogen Cryptococcus neoformans infects humans via inhalation of basidiospores or desiccated yeast cells from the environment. When effective immune containment is compromised, the initial pulmonary infection has a unique predisposition to disseminate to the central nervous system resulting in meningitis. This insidious fungus must therefore make the transition from its ecological niche of pigeon feces and eucalyptus trees to the different mammalian niche of intracellular locale of phagocytic cells and extracellular sites in the lungs, bloodstream and central nervous system. Well-characterised mechanisms of adaptation during this transition include the expression of virulence factors such as capsule, melanin and urease. The remodeling of central nitrogen metabolism and the expression of specific nitrogen acquisition systems however, remains poorly defined; these concepts are particularly relevant as specific macro- and micronutrient availability is scarce in a human host relative to bird droppings. This chapter investigates the global regulation of nitrogen catabolism and its impact on C. neoformans virulence.

Nitrogen Metabolite Repression of Metabolism and Virulence in the Human Fungal Pathogen Cryptococcus neoformans

I. Russel Lee,* Eve W. L. Chow,* Carl A. Morrow,* Julianne T. Djordjevic† and James A. Fraser*,1

*Australian Infectious Disease Research Centre, School of Chemistry and Molecular Biosciences, University of Queensland, Brisbane, QLD 4072 Australia and †Centre for Infectious Diseases and Microbiology, Westmead Millennium Institute, University of Sydney at Westmead Hospital, NSW 2145 Australia* Manuscript received March 8, 2011 Accepted for publication March 22, 2011

ABSTRACT Proper regulation of metabolism is essential to maximizing ﬁtness of organisms in their chosen environmental niche. Nitrogen metabolite repression is an example of a regulatory mechanism in fungi that enables preferential utilization of easily assimilated nitrogen sources, such as ammonium, to conserve resources. Here we provide genetic, transcriptional, and phenotypic evidence of nitrogen metabolite repression in the human pathogen Cryptococcus neoformans. In addition to loss of transcriptional activation of catabolic enzyme-encoding genes of the uric acid and proline assimilation pathways in the presence of ammonium, nitrogen metabolite repression also regulates the production of the virulence determinants capsule and melanin. Since GATA transcription factors are known to play a key role in nitrogen metabolite repression, bioinformatic analyses of the C. neoformans genome were undertaken and seven predicted GATA-type genes were identiﬁed. A screen of these deletion mutants revealed GAT1, encoding the only global transcription factor essential for utilization of a wide range of nitrogen sources, including uric acid, urea, and creatinine—three predominant nitrogen constituents found in the C. neoformans ecological niche. In addition to its evolutionarily conserved role in mediating nitrogen metabolite repression and controlling the expression of catabolic enzyme and permease-encoding genes, Gat1 also negatively regulates virulence traits, including infectious basidiospore production, melanin formation, and growth at high body temperature (39°–40°). Conversely, Gat1 positively regulates capsule production. A murine inhalation model of cryptococcosis revealed that the gat1D mutant is slightly more virulent than wild type, indicating that Gat1 plays a complex regulatory role during infection.

ITROGEN is a major component of macromole- ammonium and glutamine are usually the preferred N cules ranging from proteins to nucleic acids that nitrogen sources of fungi, although in the absence of are essential to all living organisms (Carroll and Salt these sources, less easily assimilated nitrogen sources 2004). Yet while almost 80% of the earth’s atmosphere such as amines, amides, purines, and pyrimidines may is made up of diatomic nitrogen gas, very few species also be utilized (Marzluf 1997). In members of the are able to break the strong triple bond that exists be- phylum Ascomycota, the selective utilization of these tween the two atoms to reduce gaseous nitrogen into secondary nitrogen sources is typically tightly controlled a state that is biologically available (Postgate 1998). In by transcriptionally regulating the synthesis of the en- organisms unable to utilize nitrogen directly from the zymes and permeases required for their scavenging and atmosphere, the ability to scavenge reduced forms of degradation (Mitchell and Magasanik 1984; Marzluf nitrogen from alternative sources is crucial and elabo- 1997; Fraser et al. 2001; Magasanik and Kaiser 2002). rate control mechanisms are needed to ensure its con- Achieved through a global regulatory circuit known stant supply. as either nitrogen metabolite repression or nitrogen As major decomposers of organic material in the catabolite repression, this process ensures secondary biosphere, fungi are well known for their ability to nitrogen source degrading pathways are not expressed utilize numerous compounds as nitrogen sources by when more easily assimilated nitrogen sources are avail- expressing catabolic enzymes and permeases in a range able, thereby maximizing the ﬁtness of the organism in of pathways. Readily assimilated compounds such as its ever-changing local environment (Wiame et al. 1985; Marzluf 1997; Fraser et al. 2001). By combining nitro- Supporting information is available online at http://www.genetics.org/ gen metabolite repression with pathway-speciﬁc induc- cgi/content/full/genetics.111.128538/DC1. tion mediated by an array of dedicated transcription 1Corresponding author: 358 Molecular Biosciences Building, Cooper Road, The University of Queensland, Brisbane, QLD 4072 Australia. factors, resources are even more tightly conserved until E-mail: [email protected] preferred nitrogen sources are depleted (Davis et al.

Genetics 188: 309–323 ( June 2011) 310 I. Russel Lee et al.

1993; Feng and Marzluf 1998; Berger et al. 2006, ences in nitrogen availability that C. neoformans often 2008). In this way, the fungus activates the transcription encounters as part of its infection cycle, there are more of catabolic genes only when their substrates are imme- subtle indications that nitrogen metabolism may also diately available. play a role in the regulation of virulence. The antipha- In ascomycete species where it has been observed, gocytic polysaccharide capsule is a well-known virulence nitrogen metabolite repression is mediated by tran- factor that is highly induced in the presence of uric acid scription factors belonging to the GATA family. These (Staib et al. 1976, 1978). Assimilation of uric acid as proteins are distinguished by their highly conserved a nitrogen source requires a number of catabolic en- DNA-binding domain consisting of a Cys2/Cys2-type zymes including the virulence factor urease, a nitrogen- zinc finger motif followed by an adjacent basic region scavenging enzyme shown to play a role in central and are named after the DNA sequence 59-GATA-39 that nervous system invasion (Cox et al. 2000; Olszewski they recognize (Ko and Engel 1993; Merika and Orkin et al. 2004; Shi et al. 2010). While evidence that nitrogen 1993; Ravagnani et al. 1997). GATA transcription factors availability may play a role in pathogenicity of C. neofor- are widely found in eukaryotes, including metazoans and mans is mounting, little is known about this process. plants, and have diverse biological functions. For exam- In this study we provide the first proof of canonical ple, in the nematode worm Caenorhabditis elegans, the ascomycete-like nitrogen metabolite repression in the GATA factor ELT-2 is required for immunity to bacterial basidiomycete C. neoformans. We also reveal that beyond and fungal pathogens such as Salmonella enterica and C. the traditional role in nitrogen scavenging, this regula- neoformans (Kerry et al. 2006). In the plant Arabidopsis tory mechanism controls the coordinated production of thaliana, the GATA factor HAN regulates the develop- the virulence factors capsule and melanin. Study of de- ment of flower and shoot apical meristems (Zhao et al. letion mutants of every predicted GATA-type gene in 2004). the C. neoformans genome revealed the existence of only The first GATA factor mutant, however, was identified one positively acting nitrogen regulatory GATA factor: in Neurospora crassa on the basis of its central role in ni- Gat1. Gat1 mediates nitrogen metabolite repression trate metabolism (Fincham 1950). Subsequent analyses and is required for utilization of most nitrogen sources showed that these positively acting nitrogen regulatory including the preferred ammonium; this represents an GATA factors often act synergistically with pathway- anomaly from archetypical nitrogen metabolite repres- specific transcription factors to activate the expression sion. Gat1 also functions to negatively regulate mating, of catabolic enzyme and permease-encoding genes re- melanin production at human body temperature (37°), quired to assimilate secondary nitrogen sources when and growth at high body temperature (39°–40°) and is preferred sources are lacking (Davis et al. 1993; Feng required for capsule synthesis. Importantly, this com- and Marzluf 1998; Berger et al. 2006, 2008). As a conse- plex series of changes in the regulation of C. neoformans quence, loss of such global trans-acting GATA regulatory virulence composite combines to create an unexpected genes including areA in Aspergillus nidulans and nit-2 in phenotype during murine infection: the gat1D mutant is N. crassa render these species incapable of utilizing ni- slightly more virulent than wild type. Together, these trogen sources other than ammonium or glutamine studies indicate that in addition to controlling the regu- (Arst and Cove 1973; Stewart and Vollmer 1986; Fu lon of genes for nitrogen acquisition, Gat1 is a key co- and Marzluf 1987; Kudla et al. 1990; Caddick 1992; ordinator of multiple virulence-associated phenotypes. Marzluf 1997). These GATA factors have also been shown to affect virulence in a number of pathogenic fungi. Aspergillus fumigatus areA∆,clinicalSaccharomyces cer- MATERIALS AND METHODS evisiae gln3D, and Candida albicans gat1D or gln3D mutants Strains and media: Cryptococcus strains (supporting infor- all exhibit reduced virulence compared to wild type in mu- mation, Table S1) were grown in YPD (1% yeast extract, 2% rine virulence assays (Hensel et al. 1998; Limjindaporn Bacto-peptone, 2% glucose) or YNB (0.45% yeast nitrogen base et al. 2003; Kingsbury et al. 2006; Liao et al. 2008). w/o amino acids and ammonium sulfate, 2% glucose, 10 mm The importance of nitrogen metabolism in the study nitrogen source) unless specified otherwise. Biolistic transform- of C. neoformans, an opportunistic pathogen of the phy- ants were selected on YPD medium supplemented with 100 µl/ml G418 (Sigma) and/or nourseothricin (Werner BioAgents). lum Basidiomycota that causes life-threatening menin- Pigeon guano medium was prepared as described previously goencephalitis predominantly in immunocompromised while the derivative pigeon guano (unfiltered) medium was individuals, is also evident (Casadevall and Perfect prepared with the exclusion of the filtration step to ensure 1998). The primary ecological niche of C. neoformans is that insoluble nitrogen sources such as uric acid would not be nutrient-rich pigeon guano; 70% of the nitrogen pres- removed (Nielsen et al. 2007). l-3,4-Dihydroxyphenylalanine ent is in the form of uric acid with the rest consisting (l-DOPA) medium was prepared as described previously with the original nitrogen source (asparagine) replaced with 10 mm primarily of xanthine, urea, and creatinine (Staib et al. of the specified nitrogen source (D’souza et al. 2001). V8 (5% 1976; Staib et al. 1978; Casadevall and Perfect V8 juice, 3 mm KH2PO4, 0.1% myo-inositol, 4% Bacto-agar) and 1998). In contrast, nitrogen availability becomes scarce Murashige & Skoog (MS) mating media (PhytoTechnology Lab- upon infection of humans. Beyond these obvious differ- oratories) at pH 5.0 were prepared as described previously Nitrogen Metabolite Repression 311

(Kwon-Chung et al. 1982; Xue et al. 2007). Christensen’s urea an Applied Biosystems 7900HT fast real time PCR System with agar was prepared as described previously (Cox et al. 2000). the following cycling conditions: denaturation at 95° for Nematode growth medium (NGM) was prepared as described 10 min, followed by amplification and quantification in 45 previously (Brenner 1974). Escherichia coli Mach-1 cells served cycles at 95° for 15 sec and 60° for 1 min, with melting curve as the host strain for transformation and propagation of all profiling at 95° for 2 min, 60° for 15 sec, and 95° for 15 sec. plasmids, which were carried out according to methods of Dissociation analysis confirmed the amplification of a single Sambrook et al. (1989). C. elegans strain N2 was maintained at PCR product for each primer pair and an absence of primer 15° and propagated on its normal laboratory food source E. coli dimer formation. Relative gene expression was quantified using OP50 cells as described previously (Brenner 1974; Honda et al. SDS software 1.3.1 (Applied Biosystems) on the basis of the 1993; Garsin et al. 2001). 2-∆∆CT method (Livak and Schmittgen 2001). Several house- Bioinformatic analyses: The genome sequence of C. neoformans keeping genes including GPD1 (glyceraldehyde-3-phosphate var. grubii reference strain, H99 (http://www.broadinstitute.org/ dehydrogenase), HHT1 (histone H3), and TUB2 (b-tubulin) annotation/genome/cryptococcus_neoformans/MultiHome.html), were tested, but the ACT1 (actin) gene was eventually used as was used in our study to gain insights into the regulation of a control for normalization since its gene expression demon- nitrogen metabolism and virulence in C. neoformans. Sequence strated the highest consistency across all growth conditions analyses were performed using BLAST and MacVector 9.5 tested. Statistical analysis of variance (one-way analysis of vari- (MacVector Inc, Cary, NC) (Altschul et al. 1990). Sequence ance) was performed using the unpaired, two-tailed t-test in alignments were created using ClustalW v. 1.4 within MacVec- GraphPad Prism Version 5.0c. P values ,0.05 were considered tor (Thompson et al. 1994). Sequence traces generated at the statistically significant. Australian Genome Research Facility (Brisbane, Queensland, Capsule assays: Strains were inoculated onto YNB plates Australia) were analyzed using Sequencher 4.7 (Gene Codes supplemented with 10 mm specified nitrogen source or into Corporation, Ann Arbor, MI). RPMI/DMEM supplemented with 10% fetal calf serum Construction and complementation of mutant strains: (Gibco), and incubated at 37° for 2 days. To visualize capsule, GATA-type gene deletion mutants were created using overlap cells were stained with India ink (Becton Dickinson) and ob- PCR and biolistic transformation as described previously served using a ZEISS Axioplan 2 epifluorescent/light micro- (Davidson et al. 2000, 2002). All primers used for the gen- scope. Pictures were taken with an Axiocam grayscale digital eration of deletion mutants are listed in Table S2. For example, to construct the gat1∆ mutant, the 4239-bp GAT1 camera using the AxioVision AC imaging software. Quantita- (CNAG00193.2) coding sequence was replaced with the neo- tive analysis of capsule diameter was performed as described mycin phosphotransferase II-encoding selectable marker NEO previously (Zaragoza et al. 2003). using a construct created by overlap PCR combining an 1-kb Growth, melanization, and urease assays: Starter cultures fragment upstream from the GAT1 start codon, theNEO were prepared by growth in YPD at 30° overnight with shaking, diluted to OD 0.05 in water, and then further marker, and an 1-kb fragment downstream from the GAT1 595 nm ¼ stop codon. Strain H99 genomic DNA and the plasmid pJAF1 diluted 10-fold in series. For nitrogen utilization, melaniza- were used as PCR templates (Fraser et al. 2003). The con- tion, and urease assays as well as growth test on pigeon guano struct was transformed into the wild-type strain H99 via par- and at human body temperature, each diluted cell suspension ticle bombardment. Transformants were selected on YPD was spotted onto YNB or l-DOPA supplemented with 10 mm fi plates supplemented with 100 µl/ml G418. Deletion of speci ed nitrogen source, Christensen’sureaagar,25%(wt/vol) GAT1 was confirmed by Southern blot, creating strain RL1 pigeon guano medium, or YPD medium, respectively. Results (Southern 2006). To complement the gat1∆ mutant, the were imaged after 2 days’ incubation at 30° (nitrogen utiliza- wild-type GAT1 gene was amplified from genomic DNA via tion, growth test on pigeon guano, and urease assays), 30° and PCR, the product cloned into pCR2.1-TOPO (Invitrogen) to 37° (melanisation assays), or 37°–40° (growth test at human give pIRL1, and sequenced. The 5.95-kb AscI/SacI GAT1 frag- body temperature). ment of pIRL1 was subcloned into the AscI/SacI sites of the Mating assays: Mating assays were conducted as described Cryptococcus nourseothricin resistance vector pPZP-NATcc, previously (Nielsen et al. 2003). Briefly, strains were pregrown creating the complementation construct pIRL3 (Walton on YPD plates for 2 days, after which a small amount of cells et al. 2005). pIRL3 was subsequently linearized with AscI and was removed using a flat-end toothpick and patched onto V8 XmaI, then biolistically transformed into the gat1∆ mutant. and MS mating media (pH 5.0) either alone or mixed in equal Stable transformants were selected on YPD plates supple- proportions with a strain of the opposite mating-type (MAT) mented with 100 µl/ml nourseothricin and transformants (Kwon-Chung et al. 1982; Xue et al. 2007). Plates were then containing a single copy of the wild-type GAT1 gene were incubated at room temperature in the dark for 1 week and identified by Southern blot. assessed by light microscopy for formation of filaments and Quantitative real-time PCR: Strains were grown in YNB basidia. supplemented with 10 mm of the specified nitrogen source or C. elegans killing assays: Starter cultures of C. neoformans YPD and shaken at 30° for 16 hr. Overnight cultures were strains were prepared by growth in YPD at 30° overnight with harvested, cell pellets frozen and lyophilised, total RNA iso- shaking. Overnight cultures, 10 µl, were spread onto both lated using TRIzol reagent (Invitrogen), and cDNA generated brain–heart infusion (BHI) (Becton Dickinson) and 2.5% pi- using the SuperscriptIII first-strand synthesis system (Invi- geon guano agar plates (35 mm), and incubated at 25° over- trogen). Primers for genes URO1 (CNAG04305.2), DAL1 night. Approximately 50 young adult C. elegans worms were (CNAG00934.2), URE1 (CNAG05540.2), GDH1 (CNAG01577.2), then transferred from a lawn of E. coli OP50 on NGM to the GLN1 (CNAG00457.2), GLT1 (CNAG04862.2), GDH2 lawn of BHI and pigeon guano media-grown C. neoformans (CNAG00879.2), AMT1 (CNAG00235.2), AMT2 (CNAG04758.2), (Mylonakis et al. 2002). Plates were incubated at 25° and PUT1 (CNAG02049.2), PUT5 (CNAG02048.2), and PUT2 worms examined for viability at 24-hr intervals using a dissect- (CNAG05602.2)(Table S2)weredesignedtospanexon–exon ing microscope, with worms that did not respond to a touch boundaries and tested to verify that they bind specifically to with a platinum wire pick considered dead. Each experimen- their respective cDNA genes but not H99 genomic DNA (data tal condition was performed in triplicate. Survival was plotted not shown). Quantitative real-time PCR (qRT-PCR) was per- against time, and P values were calculated by plotting a Kaplan– formed using SYBR green supermix (Applied Biosystems) and Meier survival curve and performing a log-rank (Mantel–Cox) 312 I. Russel Lee et al. test using Graphpad Prism v. 5.0c. P values ,0.05 were nitrogen sources, the nitrogen assimilation profile of considered statistically significant. Cryptococcus more closely resembled that of Saccharo- Murine virulence assays: For murine killing assays, C. neofor- myces; both species are largely restricted to utilization mans strains were used to infect 7-week-old female BALB/c mice by nasal inhalation (Cox et al. 2000). Ten mice were of ammonium and more complex macromolecules inoculated each with a 50-µl drop containing 1 · 105 cells of such as amino acids and purines (Kinsky 1961; Hynes each strain. Mice were weighed before infection and daily 1970). Only a subset of strains was able to utilize xan- thereafter; animals were sacrificed by cervical dislocation once thine, lysine, and nicotinamide as the sole nitrogen their body weight had decreased by 20% of the preinfection source. None of the strains were able to utilize cadaver- body weight. Survival was plotted against time, and P values were calculated by plotting a Kaplan–Meier survival curve and ine, nitrite, or nitrate as sole nitrogen sources. Impor- performing a log-rank (Mantel–Cox) test using Graphpad tantly, while interstrain growth variability on different Prism v. 5.0c. P values ,0.05 were considered statistically nitrogen sources was observed, this did not correlate fi signi cant. with molecular type. Ethics statement: This study was carried out in strict The explanation for some of these significant differ- accordance with the recommendations in the Australian Code of Practice for the Care and Use of Animals for Scientific ences between Cryptococcus and Aspergillus or Neu- Purposes by the National Health and Medical Research Council. rospora is evident at the genomic level. As previously The protocol was approved by the Molecular Biosciences reported, analysis of the C. neoformans genome con- Animal Ethics Committee of The University of Queensland firmed an absence of genes required for molybdenum (AEC approval no. SCMB/473/09/UQ/NHMRC). Infection cofactor biosynthesis, an essential component in all pre- and sacrifice were performed under methoxyflurane anesthesia, and all efforts were made to minimize suffering through viously characterized fungal nitrate metabolism path- adherence to the Guidelines to Promote the Wellbeing of ways (Heck et al. 2002; Cultrone et al. 2005). Genes Animals Used for Scientific Purposes as put forward by the encoding homologs of key catabolic enzymes for cadav- National Health and Medical Research Council. erine, nitrate, and nitrite utilization were also not found in the genome of VNI strain H99 via BLASTp analysis, explaining the lack of growth on these nitrogen sour- RESULTS ces. We are therefore unable to explain the Cryptococ- Cryptococcus assimilates a limited subset of cus growth phenotype associated with nitrate utilization nitrogen sources: Work on Cryptococcus has demon- that was observed by Jiang et al.; consistent with our strated the ability of this organism to assimilate a number bioinformatic analyses, none of the 16 Cryptococcus of different nitrogen sources (Chaskes and Tyndall strains tested were able to utilize nitrate or nitrite as 1975; Staib et al. 1976; Polacheck and Kwon-Chung the sole nitrogen source ( Jiang et al. 2009). 1980; Fiskin et al. 1990; Kmetzsch et al. 2010). However, Nitrogen metabolite repression of transcription ex- these studies have not taken into consideration the sig- ists in C. neoformans: In many members of the phylum nificant advances that have been made in understanding Ascomycota, nitrogen metabolite repression ensures the population structure of the species: the former vari- that of the options available, favored nitrogen sources ety Cryptococcus gattii has now been classified as a sepa- such as ammonium are assimilated before all others. To rate species, and eight distinct Cryptococcus haploid address if the same regulatory mechanism exists in the molecular types have been identified, each of which dis- most clinically prevalent form of C. neoformans (VNI), plays distinct physiological properties and virulence fea- we bioinformatically identified several predicted cata- tures (Kwon-Chung and Varma 2006; Litvintseva et al. bolic genes belonging to the uric acid degradation path- 2006). To begin to better understand nitrogen metab- way in the unpublished H99 genome and employed olism in this pathogen, we sought to extensively iden- qRT-PCR to analyze their transcriptional regulation: tify the panel of usable nitrogen sources in the context URO1 (encoding urate oxidase), DAL1 (encoding allan- of this complex population structure. toinase), and URE1 (encoding the virulence factor We assessed the ability of an array of 16 strains that urease) (Figure 1A). When compared to an ammonium- represent all eight molecular types of the Crypto- grown control, the level of expression of URO1 and coccus species complex to assimilate a diverse range DAL1 was significantly upregulated (.10-fold) during of 42 different compounds as sole nitrogen sources on growth in uric acid as the sole nitrogen source, while YNB defined medium, including the common l-amino that of URE1 was only slightly upregulated ( 3-fold) acids as well as a range of purines, pyrimidines, amides, (Figure 1B). When both uric acid and ammonium were and amines (Table S3). As commercial formulations of present, this upregulation was abolished, the hallmark agar have been shown to contain substantial concentra- of traditional nitrogen metabolite repression (URO1, tions of nitrogen as impurities (up to 4 mm when 20 g/ P 0.0224; DAL1, P 0.0058; URE1, P 0.1448) (Fig- ¼ ¼ ¼ liter is present), we performed all nitrogen assimilation ure 1B). It is worth noting that nitrogen metabolite reg- assays using agarose (SeaKem LE) as the gelling agent ulation of URE1 was not statistically significant, indicating (Scholten and Pierik 1998). Unlike other well-known that nitrogen metabolite repression may not affect every fungal genera such as Aspergillus and Neurospora, nitrogen catabolism-associated gene in C. neoformans. which can assimilate a remarkably large collection of However, it is possible that differential expression of Nitrogen Metabolite Repression 313

Figure 1.—The predicted catabolic enzyme- encoding genes of uric acid, URO1 and DAL1, are sensitive to nitrogen metabolite repression. (A) Scheme representing the predicted (partial) uric acid degradation pathway of C. neoformans. (B) cDNA from wild-type H99 grown in YNB supplemented with ammonium, uric acid, or uric acid plus ammonium (10 mm each nitrogen source) were amplified via qRT-PCR using primers against URO1 (urate oxidase), DAL1 (allantoinase), URE1 (urease), and the control gene ACT1 (actin). In the presence of uric acid as the sole nitrogen source, the expression of URO1 and DAL1 was significantly increased while that of URE1 was slightly increased, but this upregulation was abolished when ammonium was also present. This nitrogen metabolite repression sensitivity of URO1 (* denotes P , 0.05) and DAL1 (** denotes P , 0.01) was statistically significant. Error bars represent standard errors across three biological replicates.

URE1 may be observed under other untested nitrogen supplemented with the same panel of nitrogen sources. conditions such as nitrogen starvation. Again, varying virulence factor production was observed Nitrogen metabolite repression regulates well- in response to the available nitrogen source (Figure established virulence factors: Expanding our analysis, 2B). Creatinine, uric acid, and alanine-grown cells syn- we sought to determine whether nitrogen metabolite thesized the least melanin, proline synthesized an in- repression plays a role in regulating the expression of termediate amount, and glutamine, asparagine, urea, other key virulence factors. Early studies have shown and ammonium were the most prolific producers of that the virulence factors capsule and melanin are melanin. It was interesting to note that the effect of expressed differently during growth on different nitro- creatinine and uric acid on capsule production was in- gen sources (Chaskes and Tyndall 1975; Staib et al. versely proportional to melanin production: while uric 1976, 1978). To expand on these classic analyses, we acid and creatinine induce capsule formation to the grew C. neoformans on a wider selection of easily assim- largest extent, melanization on these nitrogen sources ilated (e.g., ammonium) and less easily assimilated (e.g., was the poorest. Similarly, while ammonium induces uric acid) nitrogen sources and found that varying cap- melanization to the largest extent, capsule formation sule size was indeed observed in the presence of differ- was extremely poor on this nitrogen source. Consistent ent nitrogen-yielding substrates (Figure 2A and Figure with conservation of the nitrogen metabolite repression S1). The size of capsule was uniformly small when C. regulatory paradigm, simultaneous addition of ammo- neoformans was grown on ammonium, glutamine, pro- nium restored melanin production to the same level as line, or alanine as the sole nitrogen source. In contrast, that of the ammonium control (Figure 2B). The only capsule production was much more prolific when cells unusual exception occurred with proline-grown cells, were grown on asparagine, compounds from the uric which melanized to the same extent with or without acid degradation pathway (urea, uric acid), or creati- ammonium, suggesting that proline catabolism in C. nine. Significantly, simultaneous addition of ammo- neoformans may be insensitive to the repressing effects nium to each of these nitrogen sources repressed of ammonium. This observation was not seen with the capsule production to a level equivalent to that seen capsule assays previously, as both proline and ammo- on ammonium alone, providing the first evidence that nium were equally poor inducers of capsule synthesis. nitrogen metabolite repression directly influences the Overall, we observed that the general effect of nitrogen formation of the virulence determinant capsule (Figure metabolite repression on melanin was opposite to that 2A and Figure S1). observed for capsule—rather than blocking virulence We next investigated C. neoformans formation of the factor production, the presence of ammonium induces dark antioxidant pigment melanin on l-DOPA medium the highest level of melanin formation. 314 I. Russel Lee et al.

GATA transcription factors are distinguished by their highly conserved DNA-binding domain consisting of fi a Cys2/Cys2-type zinc nger motif followed by an adjacent basic region (Fu and Marzluf 1990; Marzluf 1997). Within the consensus “Cys–X2–Cys–Xvariable–Cys– X2–Cys,” there are always two residues separating cysteines 1 and 2, and always two residues separating cysteines 3 and 4 of the zinc finger motif. In contrast, the number of residues comprising the central loop separating cysteines 2 and 3 can vary. For example, the zinc finger motif of the A. nidulans nitrogen regulatory protein AreA contains a 17-amino-acid central loop while that of N. crassa blue light response factor WC1 contains 18 residues (Kudla et al. 1990; Ballario et al. 1996). We sought to establish a more precise description of this consensus domain in fungi to assist in accurately predicting all of the bona fide GATA factor-encoding genes in the genome of C. neoformans strain H99. To more accurately define the sequence constraints associated with fungal GATA factors, we aligned the DNA-binding motif sequences of all 38 known characterized GATA factors, originating from 17 fungal species (Figure S2). By comparing the number of amino Figure 2.—Nitrogen metabolite repression influences cap- acid residues at the central loop of the zinc finger motif sule and melanin formation. (A) India ink cell staining under and the distribution of basic residues within 25 amino light microscopy showed that wild-type H99 produces capsules acids adjacent to the zinc finger motif, we were able to that vary in size when grown on YNB supplemented with dif- fi ferent nitrogen sources (10 mm each). Capsule size increased de ne a more stringent GATA motif consensus for in the following order: (ammonium, glutamine, proline, ala- fungi: all reported examples matched the motif “Cys– nine), asparagine, urea, uric acid, and creatinine. Upon cocul- X2–Cys–X17–20–Cys–X2–Cys 1 5–12 basic residues within the ture of each of these nitrogen sources with ammonium, capsule 25 beyond the most C-terminal cysteine.” size was restored to that of the ammonium control. Scale bar, Parallel searches were subsequently taken to identify 10 µm. (B) Wild-type H99 produces varying amounts of mela- potential GATA factors in the C. neoformans genome. In nin when grown on l-DOPA supplemented with different ni- fi trogen sources (10 mm each). Melanization increased in the the rst approach, the protein sequences of all 38 known following order: (creatinine, uric acid), alanine, proline and characterized fungal GATA factors were searched against (glutamine, asparagine, urea, ammonium). Upon coculture the Broad Institute genome sequence database of strain of each of these nitrogen sources with ammonium, melanin H99 for potential homologs using BLASTp, leading to production was restored to that of the ammonium control, with fi fi the exception of the proline-grown cells, which melanized to the identi cation of ve potential GATA factor-encoding the same extent with or without ammonium. genes (Table S4). However, this approach was heavily biased in requiring the GATA factors detectable in the C. neoformans genome to have first been identified in The C. neoformans genome encodes up to seven other fungi. In the second approach, we searched the GATA factors: Our phenotypic analyses showed that C. neoformans genome for the consensus “Cys–X2–Cys– beyond controlling nitrogen catabolism, nitrogen me- X17–20–Cys–X2–Cys” using a Perl script-based approach, tabolite repression plays a role in regulating at least two and hits analyzed for a match with the “5–12 basic residues important virulence factors in C. neoformans: capsule within the 25 beyond the most C-terminal cysteine” consensus. and melanin. The production of these factors varied This broader analysis identified the same five GATA on different nitrogen sources, with the addition of am- factor-encoding genes as the BLASTp analysis, plus monium dominating over other nitrogen signals (apart two additional candidates. In the third approach, we from proline) to either abolish (capsule) or highly ac- BLASTed all seven hits back against the H99 genome, tivate (melanin) virulence factor production. In the finding no additional homologs. Through these com- phylum Ascomycota, this dominance of ammonium in bined bioinformatic analyses, we therefore identified controlling a physiological response is mediated by tran- a total of seven possible GATA factor-encoding genes. scription factors belonging to the GATA zinc finger Of the seven predicted GATA ORFs identified, all four family. We therefore chose to pursue potential GATA previously characterized C. neoformans members of this factors of C. neoformans as likely candidates for the un- family were present: Bwc2, Cir1, Gat201, and Gat1—thus derlying mechanism controlling these responses to ni- validating the combination approach taken (Idnurm trogen availability. and Heitman 2005; Jung et al. 2006; Liu et al. 2008; Nitrogen Metabolite Repression 315

Figure 3.—The gat1∆ mutant is unable to utilize a wide variety of nitrogen sources. Tenfold spot dilution assays of wild-type H99 and GATA-type deletion mutants for nitrogen utilization showed that the gat1∆ mutant is unable to grow on YNB supplemented with 10 mm ammonium, uric acid, urea, or creatinine but exhibits only a slight growth defect compared to wild type on 10 mm proline. Complementation of the gat1∆ mutant with the GAT1 gene restored wild-type nitrogen utilization phenotype (Figure S3).

Kmetzsch et al. 2010). Furthermore, sequence searches variety of nitrogen sources does not ipso facto prove that of these candidates in Pfam confirmed the presence of Gat1 mediates nitrogen metabolite repression. To in- a predicted GATA zinc finger, supporting the hypothesis vestigate the regulatory mechanism of Gat1, we per- that all seven are bona fide GATA factors. Consistent formed qRT-PCR to analyze how the deletion of GAT1 with the observation of Wong et al., all C. neoformans affects transcriptional regulation of genes associated GATA factors lack putative leucine zippers that are pres- with nitrogen metabolism. The unusual observation ent in negatively acting nitrogen regulatory GATA fac- that loss of GAT1 merely reduces growth on the less tors such as Dal80 and AreB of S. cerevisiae and A. easily assimilated nitrogen source proline rather than nidulans, respectively (Wong et al. 2008). completely abolish it provided an ideal condition under Only one positively acting GATA factor controls global which to study the transcriptional effects of the loss of nitrogen source utilization in C. neoformans: Despite the this regulatory gene. Growth in proline enabled the fact that four GATA factors have been previously isolation of RNA from defined medium without con- characterized in C. neoformans, to our knowledge any cern for starvation-related artifacts of the gat1∆ mutant potential role in nitrogen metabolite repression has associated with growth in nonutilizable nitrogen sour- not been explored in these transcription factors. It is ces such as uric acid. Hence, by choosing proline as the possible that a GATA factor may play a regulatory role sole source of nitrogen for culturing of both the wild- in more than one global response or that more than one type and mutant strains, we were able to isolate RNA GATA transcriptional activator may play a role in nitro- from a steady-state, actively growing culture with a de- gen regulation as in systems of S. cerevisiae and C. albicans fined nitrogen source. (Stanbrough et al. 1995; Limjindaporn et al. 2003; Liao Bioinformatic analyses of the H99 genome identified et al. 2008). All seven ORFs containing our strict consen- several genes predicted to encode enzymes required for sus for a fungal GATA DNA-binding domain were degradation of proline: PUT1 and PUT5, two paralogs therefore deleted via homologous recombination. that each encode proline oxidase, and PUT2, which The seven GATA mutants were viable and had similar encodes pyrroline-5-carboxylate dehydrogenase (Figure growth rate as the wild-type strain on YPD medium at 4A). qRT-PCR using RNA isolated from wild-type cells 30° (Figure 3). However, all but one of the C. neoformans revealed that only one of these three proline catabolic GATA mutant strains exhibited growth equivalent to genes was regulated by nitrogen metabolite repression, wild-type H99 on YNB defined medium supplemented with higher levels of transcription (.15-fold) on pro- with a wide panel of 38 usable nitrogen sources (Figure line than when both proline and ammonium were pres- 3 and Table S5). Consistent with previous studies con- ent (PUT1, P , 0.0001) (Figure 4B). In the gat1∆ ducted by Kmetzsch et al. (2010), deletion of ORF mutant, transcription of PUT1 was not highly activated CNAG00193.2 (GAT1) rendered C. neoformans incapa- but instead showed the same level of transcription irre- ble of utilizing a wide repertoire of nitrogen sources. spective of the presence of ammonium, confirming the The gat1D mutant, however, could still utilize a limited role of Gat1 as the mediator of nitrogen metabolite re- selection of amino acids, notably proline, strengthening pression (Figure 4B). Unlike PUT1, both PUT5 and our hypothesis that proline utilization is largely indepen- PUT2 were unaffected by nitrogen metabolite repres- dent of nitrogen metabolite repression. Importantly, we sion, explaining why the gat1∆ mutant can utilize profound that the gat1D mutant was also unable to utilize line as the sole nitrogen source, albeit less effectively uric acid and creatinine, two predominant nitrogen con- than wild type. Although the expression of PUT1 was stituents found in the ecological niche (pigeon guano) drastically reduced in the gat1∆ mutant, PUT5 was still of C. neoformans (Staib et al. 1976, 1978; Casadevall expressed at wild-type levels sufficient to produce proline and Perfect 1998). oxidase, as was PUT2 (encoding pyrroline-5-carboxylate Gat1 mediates nitrogen metabolite repression: The dehydrogenase), enabling Gat1-independent growth on fact that the gat1∆ mutant does not grow on a wide proline as the sole nitrogen source. 316 I. Russel Lee et al.

Figure 4.—Gat1 regulates nitrogen metabolite repression. (A) Scheme representing the predicted proline degradation pathway of C. neoformans. (B) cDNA from wild-type H99 and gat1∆ mutant grown in YNB supplemented with proline or proline plus ammonium (10 mm each nitrogen source) were ampliﬁed via qRT-PCR using primers against PUT1 (proline oxidase), PUT5 (proline oxidase), PUT2 (pyrroline- 5-carboxylate dehydrogenase), and the control gene ACT1 (actin). One of the predicted catabolic enzyme-encoding genes of proline, PUT1, was sensitive to nitrogen metabolite repression in wild type (*** denotes P , 0.0001) but not in the gat1∆ mutant. The remaining two catabolic genes, PUT5 and PUT2, were nitrogen metabolite repression insensitive in both strains. Error bars represent standard errors across three biological replicates.

Gat1 positively regulates the expression of key permease-encoding genes AMT1 and AMT2 in both ammonium assimilation and permease genes: While the wild-type and gat1D mutant strains (Figure 5B) certain phenotypic similarities to ascomycete nitrogen (Rutherford et al. 2008). When grown in rich unde- regulatory GATA factor mutants were clear, this basid- fined YPD medium that contains a wide array of nitrogen iomycete ortholog mutant exhibited an ammonium sources including the generally preferred ammonium, phenotype much more extreme than that seen in glutamine, and glutamate, the level of expression of previously described GATA factor mutants. The gat1D the major ammonium assimilation enzyme-encoding mutant is unable to utilize the most easily assimilated gene GDH1 as well as the permease-encoding genes nitrogen source in both ammonium replete and suf- AMT1 and AMT2 was significantly lower ( 10–20-fold) ficient conditions (1–100 mm) (data not shown), para- in the gat1D mutant compared to wild type (GDH1, P , doxically suggesting that the nitrogen metabolite 0.0001; AMT1, P 0.0086; AMT2, P 0.0308). The level ¼ ¼ repression mechanism that occurs in the presence of of expression of the other central nitrogen metabolism ammonium blocks the ability to utilize ammonium genes was also consistently lower ( 1.5-fold) in the itself. While a slight reduction in growth on ammonium gat1D mutant than in wild type, although these differ- has been observed for other fungal nitrogen regulatory ences for the remaining two ammonium assimilation GATA mutants, the complete loss of growth is unprec- enzyme-encoding genes GLN1 and GLT1 were not edented and counterintuitive (Christensen et al. 1998; quite statistically significant (GLN1, P 0.0666; ¼ Hensel et al. 1998; Liao et al. 2008). To investigate this GLT1, P 0.0523; GDH2, P 0.1308). This result ¼ ¼ difference in phenotype, we bioinformatically identified indicates that Gat1 controlstheexpressionoftheam- the predicted central nitrogen metabolism genes in the monium assimilation enzyme-encoding gene GDH1 H99 genome: GDH2 [encoding NAD1-dependent glu- and permease-encoding genes AMT1 and AMT2 in tamate dehydrogenase (NAD-GDH)] as well as the am- C. neoformans. monium assimilation enzyme-encoding genes GDH1 Gat1 is not required for the growth of C. neoformans [encoding NADP1-dependent glutamate dehydroge- on pigeon guano: The gat1∆ mutant is unable to utilize nase (NADPH-GDH)], GLN1 [encoding glutamine syn- uric acid, urea, or creatinine, the most abundant nitro- thetase (GS)], and GLT1 [encoding NAD1-dependent gen sources found in the natural habitat of C. neofor- glutamate synthase (GOGAT)] (Figure 5A). mans. This raises the question of whether the GAT1 qRT-PCR was again employed, this time to analyze gene is essential for optimum growth on pigeon guano, the transcriptional regulation of these central nitrogen a growth medium previously validated for both C. neo- metabolism genes as well as the two known ammonium formans and C. gattii (Nielsen et al. 2007). We therefore Nitrogen Metabolite Repression 317

Figure 5.—Gat1 regulates the expression of the ammonium assimilation enzyme and permease-encoding genes. (A) Scheme representing the predicted pathway of central nitrogen metabolism in C. neoformans. (B) cDNA from wild-type H99 and gat1∆ mutant grown in YPD were amplified via qRT-PCR using primers against GDH1 (NADPH-GDH), GLN1 (GS), GLT1 (GOGAT), GDH2 (NAD-GDH), AMT1 (ammonium transporter 1), AMT2 (ammonium transporter 2), and the control gene ACT1 (actin). The expression of GDH1, AMT1, and AMT2 was significantly lower in the gat1∆ mutant compared to wild type, while that of GLN1, GLT1, and GDH2 was slightly lower. This difference in level of expression for GDH1 (*** denotes P , 0.0001), AMT1 (** denotes P , 0.01), and AMT2 (* denotes P , 0.05) was statistically significant. Error bars represent standard errors across three biological replicates. investigated whether the loss of GAT1 affected growth Gat1 positively regulates capsule synthesis, but neg- on this substrate that mimics the environmental niche. atively regulates melanin formation and growth at high Our growth assays indicate that the wild-type and gat1∆ body temperature: Given that nitrogen metabolite mutant strains displayed equivalent growth on both fil- repression plays a role in regulation of capsule synthe- tered and unfiltered 25% pigeon guano agar media, sis, we next chose to examine capsule production by the revealing that pigeon guano contains at least one nitro- gat1D mutant. Strains were cultured in amino-acid-rich gen source that supports robust growth and whose uti- RPMI and DMEM media supplemented with 10% fetal lization is Gat1 independent (Figure S4). calf serum; both media supported growth of all strains Gat1 negatively regulates mating in both mating-type including the gat1∆ mutant (Zaragoza et al. 2003). In- a and a strains: During nitrogen limiting or starvation dia ink staining revealed that while the wild-type strain conditions, C. neoformans can undergo sexual reproduc- produced characteristic halos around its cells represent- tion that leads to the production of potentially infec- ing enlarged capsule, the gat1D mutant produced re- tious air-borne basidiospores (Kwon-Chung 1976; duced amounts of capsule under these serum-induced Lengeler et al. 2000; Giles et al. 2009). To address conditions (Figure 7A and Figure S5). This result indi- whether Gat1 plays a role in this sexual development, cates that Gat1 plays a positive role in the regulation of we created a gat1D derivative of MATa C. neoformans capsule synthesis. strain KN99a (VNI) by replacing the GAT1 ORF with Like capsule, melanin synthesis is also under the the nourseothricin acetyl-transferase (NAT ) selectable influence of nitrogen metabolite repression control. marker, to serve as a mating partner for our existing The role of Gat1 was therefore evaluated for melanin gat1::NEO of MATa strain H99. Wild type (wild type · production at both 30° and 37° on l-DOPA medium wild type), unilateral (gat1 · wild type), and bilateral supplemented with proline as the sole nitrogen source, (gat1 · gat1) crosses were then performed by mixing since proline supports the most robust growth in the strains of opposite mating type on V8 and MS media. gat1∆ mutant on defined medium. Interestingly, the The ability to mate was examined by microscopically gat1∆ mutant melanized to the same extent as wild type inspecting the production of filaments and basidia at at 30° but produced more melanin than wild type at 37° the edges of the mating patches. Unexpectedly, fila- (Figure 7B). This result indicates that Gat1 plays a neg- mentation and production of basidia were enhanced ative role in regulating melanin production at human in both unilateral and bilateral gat1D crosses in compar- body temperature. ison to wild-type crosses (Figure 6). Furthermore, de- We also evaluated the role of Gat1 in regulating the letion of GAT1 in both the MATa and MATa strains led activity of the nitrogen-scavenging enzyme urease on to the most robust filament formation and production Christensen’sureaagar.Althoughboththegat1∆ and of basidia, indicating that the effect of this gene dele- negative control ure1∆ mutants are unable to utilize urea tion is additive. This result indicates that Gat1 plays as the sole nitrogen source, the mutants could still grow a critical role in regulating morphological differentia- on Christensen’smediumasimpuritiesfoundinagar tion in both mating types of C. neoformans. likely provide alternative sources of consumable nitrogen. 318 I. Russel Lee et al.

lian host and in its ecological niche. While growth up to 38° was equivalent in the wild-type and gat1∆ mutant strains, the mutant exhibited better growth in comparison to wild type at 39°, the febrile temperature commonly experienced by most patients upon the onset of cryptococcal meningoencephalitis (Figure 7D). Fur- thermore, at 40°, growth of the wild-type strain was completely abolished while the gat1∆ mutant was still able to proliferate, albeit at a slow rate. This unantici- pated finding indicates that Gat1 plays a negative role in regulating growth at high temperature. Gat1 is not required for the killing of C. elegans but modestly represses virulence in a murine inhalation model of cryptococcosis: Our ultimate interest in C. neoformans lies in better understanding its pathogenicity. In this context our data thus far were confusing, as Gat1 arguably contributes to production of some viru- Figure 6.—Gat1 inhibits filament formation during mating. lence attributes but repression of others. For example, Filamentation assays on V8 and MS mating media (pH 5.0) the gat1D mutant’s reduction in capsule synthesis and showed that filament formation was enhanced in unilateral abated ability to utilize a variety of nitrogen sources gat1D crosses in comparison to the wild-type crosses. Filamen- suggest that this strain should exhibit a decrease in tation was enhanced even further in a bilateral gat1D cross. virulence, just like that seen in equivalent ascomycete Complementation of the H99a gat1D and KN99a gat1D mutants with the GAT1 gene restored wild-type mating in all GATA factor mutants. In contrast, the increase in mel- tested crosses (data not shown). anin production and enhanced growth at high temperature would support an increase in virulence. To determine the effect of Gat1 on virulence, we per- In support of our previous qRT-PCR data indicating that formed C. elegans killing assays and a murine inhalation urease is not significantly regulated by nitrogen metabo- model of cryptococcosis. lite repression, the gat1D mutant still produces urease C. neoformans presumably interacts with C. elegans in even though the mutant is unable to utilize the ammonia the environment and killing of C. elegans by C. neofor- resulting from urea catabolism (Figure 7C). mans in vitro has previously been validated as a model To complete a thorough phenotypic analysis of the for studying pathogenesis (Mylonakis et al. 2002). We gat1D mutant, we evaluated the role of Gat1 for its abil- performed C. elegans virulence assays using two different ity to grow at high temperature (37°–40°), a common media: the standard BHI medium for nematode killing stress that C. neoformans encounters both in a mamma- experiments, as well as 2.5% pigeon guano medium to

Figure 7.—Gat1 is required for capsule synthesis, but negatively regulates melanization and growth at 39° or 40°. (A) India ink cell staining under light microscopy revealed that the wild-type H99 and gat1D 1 GAT1 strains produce enlarged capsules while the gat1∆ mutant produces resid- ual amount of capsule when strains were cultured under serum-induced growth conditions. Scale bar, 10 µm. (B) The gat1∆ mutant produces more melanin compared to both the wild-type and gat1∆ 1 GAT1 strains when grown on l-DOPA medium supplemented with 10 mm proline at 37°. In contrast, all three strains melanized to the same extent when grown at 30°. (C) Unlike the negative control ure1∆ mutant, the wild-type, gat1∆ mutant and gat1∆ 1 GAT1 strains all had the ability to produce urease when grown on Christensen’s urea agar as reﬂected by the bright pink clearing surrounding the aliquot of spotted cells. (D) Tenfold spot dilution assays on YPD medium at human body temperature showed that the gat1∆ mutant exhibits enhanced growth compared to both the wild-type and gat1∆ 1 GAT1 strains at 39° and 40°. Nitrogen Metabolite Repression 319

mimic C. neoformans ecological niche. Under both growth conditions, killing of C. elegans by the gat1D mutant [LT (time for half of the worms to die) 6 days 50 ¼ for both BHI and pigeon guano media] was not significantly different to that observed for wild type (LT 6 50 ¼ days and 7 days, for BHI and pigeon guano media, respectively) (Figure 8A). Gat1 is therefore not required for C. neoformans-mediated killing of the invertebrate C. elegans. On the other hand, the murine inhalation model of cryptococcosis more closely mimics human infection by C. neoformans—inhaled cryptococcal cells ﬁrst infect the lungs before disseminating to the brain to cause meningoencephalitis. Interestingly, mice infected with the gat1D mutant succumbed to infection slightly faster (between 14 and 21 days postinfection, median survival of 19 days) than mice infected with the wild-type strain (between 17 and 28 days postinfection, median survival of 23 days) (wild type vs. gat1D,P 0.0151) (Figure 8B). ¼ This result suggests that the gat1D mutant is modestly more virulent than wild type during murine infection. Thus, in addition to regulating nitrogen metabolism and virulence factor expression in vitro, Gat1 also represses virulence in a vertebrate host.

DISCUSSION The ability to acquire and catabolize nutrients from the environment is imperative for survival of an organism. While fungi are well known for their ability to utilize a broad range of nitrogen sources, the study of molecular mechanisms controlling nitrogen acquisition and catabolism in model fungi have until now been primarily limited to members of the phylum Ascomy- cota (Wong et al. 2008). Our studies of transcriptional regulation and phenotypic responses to the presence of ammonium have shown that utilization of nitrogen Figure 8.—The gat1D mutant kills C. elegans as efficiently as sources by C. neoformans is controlled in a similar fash- wild-type H99, but exhibits modestly enhanced virulence in ion. The results we report here represent the first de- a murine host. (A) C. elegans infection: 50 nematode worms finitive characterization of the regulatory phenomenon D D 1 were transferred to a lawn of wild-type, gat1 , or gat1 of nitrogen metabolite repression in a second phylum GAT1 cells as the sole food source on both BHI and 2.5% pigeon guano media, and survival was monitored at 24-hr of the kingdom fungi, the Basidiomycota. In addition to intervals. There was no observable difference in C. elegans functional conservation between these two phyla, the killing by all three strains on both BHI (wild type vs. gat1D, repertoire of physiological responses controlled by ni- P 0.1066; gat1D vs. gat1D 1 GAT1, P 0.9230) and pigeon trogen metabolite repression in C. neoformans has ex- ¼ D ¼ D guano media (wild type vs. gat1 ,P 0.0250; gat1 vs. panded to encompass regulation of virulence factor gat1D 1 GAT1, P 0.3614). (B) Mus musculus¼ infection: 10 mice were infected¼ intranasally with either 1 · 105 cells of wild expression. We have also shown that the GATA factor type, gat1D, or gat1D 1 GAT1 strains, and survival was moni- Gat1 is responsible for mediating nitrogen metabolite tored daily. Mice infected with the wild-type and gat1D 1 repression in C. neoformans. GAT1 strains progress to morbidity at the same rate (wild type Bioinformatically, the predicted GATA factor of C. D 1 vs. gat1 GAT1, P 0.6691), whereas mice infected with the neoformans Gat1 shows little similarity to nitrogen regu- gat1D strain progress¼ to morbidity more rapidly (wild type vs. gat1D,P 0.0151). latory GATA factors of the Ascomycota for the majority ¼ of the protein, except within the GATA motif itself (Table S6). With 19% overall protein identity, it showed poor but still highest similarity to the various nitrogen regulatory GATA factors of the Ascomycota. This 320 I. Russel Lee et al. similarity, however, was largely restricted to the pre- and Wickner 1995; Xu et al. 1995; Hall and Dietrich dicted GATA zinc finger where identity to A. nidulans 2007; Wong et al. 2008). Our analyses did not identify AreA and N. crassa Nit2 was 88 and 90%, respectively. a Ure2-like candidate encoded in the C. neoformans ge- The dependence of C. neoformans nitrogen metabo- nome. Together, these data support the model that lism on a single GATA factor is in stark contrast to S. C. neoformans and the filamentous ascomycetes share cerevisiae or C. albicans, where multiple nitrogen regula- a nitrogen regulatory mechanism that more closely tory GATA factors are required (Stanbrough et al. resembles the ancestral nitrogen metabolism regulatory 1995; Limjindaporn et al. 2003; Liao et al. 2008). It pathway than that seen in S. cerevisiae. We therefore is important to note that the existence of the two propose that the last common ancestor of the Ascomy- positively acting GATA factors, Gln3 and Gat1, in the cota and Basidiomycota likely had one positively acting hemiascomycete S. cerevisiae is not the result of the GATA factor (the AreA/Nit2/Gat1/Are1 ortholog) and whole-genome duplication event that occurred around corepressor protein Nmr; after the phyla separated, 100 million years ago, since these two orthologs are a second GATA factor (Gln3) and Ure2 coevolved in also present in C. albicans, a species that did not un- S. cerevisiae, and the nmr gene was lost. dergo this event (Wong et al. 2008; Gordon et al. While the effect of the loss of GAT1/ARE1 on nitro- 2009). C. neoformans global nitrogen regulatory circuit gen metabolism largely met our predicted phenotype, instead more closely resembles the single positively act- a key aspect of it did not. One of the most confusing, ing GATA factor systems in A. nidulans and N. crassa and seemingly contradictory, phenotypes of the gat1/ (Stewart and Vollmer 1986; Kudla et al. 1990). are1∆ mutant is its inability to grow on ammonium as Given that N. crassa nit-2 was named after the mutant’s the sole nitrogen source. We subsequently gained inability to utilize nitrate, which C. neoformans is unable insights into this paradox by proving that along with to do, and the apparent existence of only one GATA its established role as an activator of secondary catabolic factor responsible for nitrogen utilization, we therefore gene expression in the absence of ammonium, Gat1/ recommend an alternative and more informative name Are1 is also functional in the presence of ammonium. to this gene: ARE1. Like A. nidulans AreA, C. neoformans Gat1/Are1 also The similarities in the global nitrogen regulatory regulates the expression of the major ammonium as- circuit between C. neoformans and A. nidulans or N. similation enzyme-encoding gene GDH1 as well as the crassa continue when NmrA and Nmr1 are considered. permease-encoding genes AMT1 and AMT2 The Nmr proteins are inhibitors of the functions of (Christensen et al. 1998; Monahan et al. 2002, 2006; A. nidulans AreA and N. crassa Nit2, and a potential Rutherford et al. 2008). Since the alternative pathway homolog of this protein named Tar1 has recently been of ammonium assimilation (via GLN1 and GLT1) is not characterized in C. neoformans, suggesting that the reg- significantly impaired in the gat1/are1D mutant, we ulation of GATA factor activity may operate in a similar speculate that the basal levels of transcription of fashion to these filamentous ascomycetes (Pan et al. AMT1 and AMT2 may be insufficient for ammonium 1997; Andrianopoulos et al. 1998; Jiang et al. 2009). uptake. It is worth noting that our nitrogen metabolite However, this model is confounded by conflicting repression study followed the traditional use of ammo- observations between the studies of Jiang et al. and nium as the “repressing” nitrogen source. In reality, it our own. Most notably, we have bioinformatics and phe- is likely that the true signal affecting Gat1/Are1 activity notypic evidence indicating that C. neoformans is unable may be intracellular concentrations of glutamine and/ to utilize nitrate as a nitrogen source. Confusingly, or glutamate since these metabolites reflect the nitro- Jiang et al. observed growth changes on this same nitrogen status of the cell in other species. Further work gen source as evidence that Tar1 plays a role in nitro- using glutamine, glutamate, and perhaps even nitrogen metabolism, an observation that we are unable to gen starvation will be required to dissect this aspect of explain. Furthermore, C. neoformans Gat1/Are1 lacks the nitrogen-sensing mechanism. the highly conserved C-terminus sequence of AreA The opposing regulatory effect of Gat1/Are1 on and Nit2, which is involved in Nmr recognition (Platt melanin and capsule production has been observed in et al. 1996; Pan et al. 1997; Andrianopoulos et al. previously characterized GATA factors of C. neoformans: 1998). Nevertheless, Nmr proteins also interact with Cir1 and Gat201 ( Jung et al. 2006; Liu et al. 2008). the DNA-binding domain of AreA and Nit2, and further Notably, although LAC1 is the major contributor to study will be required to determine if Tar1 is a true melanin biosynthesis, we were unable to find putative functional ortholog of NmrA/Nmr1. Gat1/Are1 HGATAR binding sites within a 1-kb region Unlike regulation of AreA and Nit2 by Nmr, the S. upstream from the start codon of LAC1 (Pukkila- cerevisiae GATA transcriptional activators Gln3 and Gat1 Worley et al. 2005). In contrast, numerous HGATAR are negatively regulated by the structurally unrelated sites were found in the promoter regions of CAP60 and prion-forming glutathione-S-transferase Ure2, pre- CAP64, two key genes that are regulated to control cap- dicted to have been horizontally transferred from the sule biosynthesis (Chang et al. 1996; Chang and Kwon- bacterial species Burkholderia vietnamiensis (Masison Chung 1998). Together, these data suggest that Gat1/ Nitrogen Metabolite Repression 321

Are1 may activate the transcription of capsule biosyn- enzyme and permease-encoding genes required for thesis genes directly, but indirectly regulate melanin nitrogen assimilation, it is also a key regulator of es- production, perhaps by regulating the expression of sential virulence traits in this important human patho- a repressor of LAC1 transcription. gen. While the significance of Gat1/Are1 is now clearly To gain insights into the role of Gat1/Are1 in the demonstrated, full appreciation of its role awaits fur- environment we employed a novel approach, combin- ther analysis of the gene targets and processes regu- ing an established virulence model based on a known lated by this global transcription factor as well as its predator (C. elegans) with a medium designed to emu- potential interactions with other GATA factors. Cer- late the environmental niche (pigeon guano). This tainly, a more complete understanding of the complex study enabled us to make an important observation: regulatory circuit governing nitrogen metabolism and Gat1/Are1 does not appear to have a significant effect virulence mechanisms in C. neoformans will require fur- on virulence in the environmental niche; instead it may ther study. play a role in nitrogen scavenging. However, in a murine We thank Bob Simpson for assistance with qRT-PCR, Jason Stajich host, this gene has likely been co-opted into regulating for assistance with Perl script-based bioinformatic analyses, Jim various aspects of the virulence composite. Kronstad for the provision of the cir1D strains, and Gary Newell of It is nevertheless worth noting that the C. elegans kill- the Queensland Racing Pigeon Federation, Inc., for the provision of ing assays were conducted at 25° instead of mammalian pigeon guano. body temperature. Whereas the enhanced growth and melanization abilities of the gat1/are1D mutant may be exhibited during the infection process in mice, these LITERATURE CITED virulence attributes could not be displayed in C. elegans. Altschul, S. F., W. Gish, W. Miller, E. W. Myers and D. J. Lipman, While the slightly quicker progression to morbidity in 1990 Basic local alignment search tool. J. Mol. Biol. 215: 403– mice infected with the gat1/are1D mutant would not be 410. considered great enough to classify this strain as “hyper- Andrianopoulos, A., S. Kourambas, J. A. Sharp, M. A. Davis and M. J. Hynes, 1998 Characterization of the Aspergillus nidulans virulent,” it certainly highlights the complexity of the nmrA gene involved in nitrogen metabolite repression. J. Bacter- role of Gat1/Are1 in gene regulation of C. neoformans iol. 180: 1973–1977. during infection. We note that Kmetzsch et al. (2010) Arst, H. N. Jr., and D. J. Cove, 1973 Nitrogen metabolite repression in Aspergillus nidulans. Mol. Gen. Genet. 126: 111–141. recently reported that the gat1/are1D mutant exhibited Ballario, P., P. Vittorioso, A. Magrelli, C. Talora, A. Cabibbo equivalent virulence to wild type when high doses (1 · et al., 1996 White collar-1, a central regulator of blue light re- 107 cells per strain) were used, which resulted in mice sponses in Neurospora, is a zinc finger protein. EMBO J. 15: 1650– 1657. succumbing to infection as early as the third day post- Berger, H., R. Pachlinger, I. Morozov, S. Goller, F. Narendja infection. In contrast, we employed the more tradi- et al., 2006 The GATA factor AreA regulates localization and tional inoculum of 1 · 105 cells to ensure the gradual in vivo binding site occupancy of the nitrate activator NirA. Mol. Microbiol. 59: 433–446. progression of disease. We believe this increased sensi- Berger, H., A. Basheer, S. Bock, Y. Reyes-Dominguez, T. Dalik tivity helped identify the subtle difference in virulence et al., 2008 Dissecting individual steps of nitrogen transcription between the wild-type and gat1/are1D mutant strains factor cooperation in the Aspergillus nidulans nitrate cluster. Mol. Microbiol. 69: 1385–1398. that had previously been missed. Brenner, S., 1974 The genetics of Caenorhabditis elegans. Genetics 77: Notably, all other known GATA factors of C. neofor- 71–94. mans (Bwc2, Cir1, and Gat201) play a role in virulence Caddick, M. X., 1992 Characterization of a major Aspergillus regula- factor expression in vitro and affect virulence during tory gene, areA, pp. 141–152 in Molecular Biology of Filamentous Fungi, edited by U. Stal and P. Tudzynski. VCH Press, Wein- murine infection (Idnurm and Heitman 2005; Jung heim, Germany. et al. 2006; Liu et al. 2008). Bwc2, Cir1, Gat201, and Carroll, S. B., and S. D. Salt, 2004 Ecology for Gardeners. Timber Gat1/Are1 together with the three other uncharac- Press, Portland, OR. Casadevall, A., and J. R. Perfect, 1998 Cryptococcus neoformans. terized GATA factors (CNAG03401.2, CNAG04263.2, ASM Press, Washington, DC. CNAG06762.2) we have identified are predicted to bind Chang, Y. C., and K. J. Kwon-Chung, 1998 Isolation of the third to 59-GATA-39 sites in the genome (Ko and Engel 1993; capsule-associated gene, CAP60, required for virulence in Crypto- coccus neoformans. Infect. Immun. 66: 2230–2236. Merika and Orkin 1993; Ravagnani et al. 1997). Chang, Y. C., L. A. Penoyer and K. J. Kwon-Chung, 1996 The Whether these GATA factors antagonize and compete second capsule gene of Cryptococcus neoformans, CAP64, is essential for the same 59-GATA-39 binding sites in the promoter for virulence. Infect. Immun. 64: 1977–1983. Chaskes, S., and R. L. Tyndall, 1975 Pigment production by Cryp- regions of genes, act in synergy to activate transcription tococcus neoformans from para- and ortho-Diphenols: effect of the of genes, or act interchangably remains to be deter- nitrogen source. J. Clin. Microbiol. 1: 509–514. mined. Notwithstanding, it is becoming increasingly Christensen, T., M. J. Hynes and M. A. Davis, 1998 Role of the regulatory gene areA of Aspergillus oryzae in nitrogen metabolism. apparent that Bwc2, Cir1, Gat201, and Gat1/Are1 can Appl. Environ. Microbiol. 64: 3232–3237. act in parallel to regulate multiple virulence pathways in Cox, G. M., J. Mukherjee, G. T. Cole, A. Casadevall and J. R. C. neoformans. Perfect, 2000 Urease as a virulence factor in experimental cryptococcosis. Infect. Immun. 68: 443–448. In summary, Gat1/Are1 not only regulates nitrogen Cultrone, A., C. Scazzocchio, M. Rochet, G. Montero-Moran, C. metabolite repression and the expression of catabolic Drevet et al., 2005 Convergent evolution of hydroxylation 322 I. Russel Lee et al.

mechanisms in the fungal kingdom: molybdenum cofactor- Jiang, N., D. Xiao, D. Zhang, N. Sun, B. Yan et al., 2009 Negative independent hydroxylation of xanthine via alpha-ketoglutarate- roles of a novel nitrogen metabolite repression-related gene, dependent dioxygenases. Mol. Microbiol. 57: 276–290. TAR1, in laccase production and nitrate utilization by the basid- Davidson, R. C., M. C. Cruz, R. A. Sia, B. Allen, J. A. Alspaugh iomycete Cryptococcus neoformans. Appl. Environ. Microbiol. 75: et al., 2000 Gene disruption by biolistic transformation in sero- 6777–6782. type D strains of Cryptococcus neoformans. Fungal Genet. Biol. 29: Jung, W. H., A. Sham, R. White and J. W. Kronstad, 2006 Iron 38–48. regulation of the major virulence factors in the AIDS-associated Davidson, R. C., J. R. Blankenship, P. R. Kraus,M.de Jesus Berrios, pathogen Cryptococcus neoformans. PLoS Biol. 4: e410. C. M. Hull et al., 2002 A PCR-based strategy to generate integra- Kerry, S., M. TeKippe, N. C. Gaddis and A. Aballay, 2006 GATA tive targeting alleles with large regions of homology. Microbiology transcription factor required for immunity to bacterial and fungal 148: 2607–2615. pathogens. PLoS One 1: e77. Davis, M. A., J. M. Kelly and M. J. Hynes, 1993 Fungal catabolic Kingsbury, J. M., A. L. Goldstein and J. H. McCusker, 2006 Role gene regulation: molecular genetic analysis of the amdS gene of of nitrogen and carbon transport, regulation, and metabolism Aspergillus nidulans. Genetica 90: 133–145. genes for Saccharomyces cerevisiae survival in vivo. Eukaryot. Cell D'Souza, C. A., J. A. Alspaugh, C. Yue, T. Harashima, G. M. Cox 5: 816–824. et al., 2001 Cyclic AMP-dependent protein kinase controls viru- Kinsky, S. C., 1961 Induction and repression of nitrate reductase in lence of the fungal pathogen Cryptococcus neoformans. Mol. Cell Neurospora crassa. J. Bacteriol. 82: 898–904. Biol. 21: 3179–3191. Kmetzsch, L., C. C. Staats, E. Simon, F. L. Fonseca, D. L. Oliveira Feng, B., and G. A. Marzluf, 1998 Interaction between major ni- et al., 2010 The GATA-type transcriptional activator Gat1 regu- trogen regulatory protein NIT2 and pathway-specific regulatory lates nitrogen uptake and metabolism in the human pathogen factor NIT4 is required for their synergistic activation of gene Cryptococcus neoformans. Fungal Genet. Biol. fi expression in Neurospora crassa. Mol. Cell Biol. 18: 3983–3990. Ko, L. J., and J. D. Engel, 1993 DNA-binding speci cities of Fincham, J. R. S., 1950 Mutant strains of Neurospora deficient in the GATA transcription factor family. Mol. Cell Biol. 13: 4011– aminating ability. J. Biol. Chem. 182: 61–74. 4022. Fiskin, A. M., M. C. Zalles and R. G. Garrison, 1990 Electron Kudla, B., M. X. Caddick, T. Langdon, N. M. Martinez-Rossi, C. F. cytochemical studies of Cryptococcus neoformans grown on uric Bennett et al., 1990 The regulatory gene areA mediating nitrogen metabolite repression in Aspergillus nidulans. Mutations af- acid and related sources of nitrogen. J. Med. Vet. Mycol. 28: fi – fecting speci city of gene activation alter a loop residue of a 197 207. fi Fraser, J. A., M. A. Davis and M. J. Hynes, 2001 The formamidase putative zinc nger. EMBO J. 9: 1355–1364. gene of Aspergillus nidulans: regulation by nitrogen metabolite Kwon-Chung, K. J., 1976 Morphogenesis of Filobasidiella neoformans, – repression and transcriptional interference by an overlapping the sexual state of Cryptococcus neoformans. Mycologia 68: 821 833. upstream gene. Genetics 157: 119–131. Kwon-Chung, K. J., and A. Varma, 2006 Do major species concepts Fraser, J. A., R. L. Subaran, C. B. Nichols and J. Heitman, support one, two or more species within Cryptococcus neoformans? FEMS Yeast Res. 6: 574–587. 2003 Recapitulation of the sexual cycle of the primary fungal Kwon-Chung, K. J., J. E. Bennett and J. C. Rhodes, 1982 Tax- pathogen Cryptococcus neoformans var. gattii: implications for an onomic studies on Filobasidiella species and their anamorphs. outbreak on Vancouver Island, Canada. Eukaryot. Cell 2: 1036– Antonie van Leeuwenhoek 48: 25–38. 1045. Lengeler, K. B., P. Wang, G. M. Cox, J. R. Perfect and J. Heitman, Fu, Y. H., and G. A. Marzluf, 1987 Characterization of nit-2, the 2000 Identification of the MATa mating-type locus of Cryptococ- major nitrogen regulatory gene of Neurospora crassa. Mol. Cell cus neoformans reveals a serotype A MATa strain thought to have Biol. 7: 1691–1696. been extinct. Antonie Van Leeuwenhoek 97: 14455–14460. Fu, Y. H., and G. A. Marzluf, 1990 nit-2, the major positive-acting Liao, W. L., A. M. Ramon and W. A. Fonzi, 2008 GLN3 encodes nitrogen regulatory gene of Neurospora crassa, encodes a sequence- fi a global regulator of nitrogen metabolism and virulence of C. speci c DNA-binding protein. Proc. Natl. Acad. Sci. USA 87: albicans. Fungal Genet. Biol. 45: 514–526. 5331–5335. Limjindaporn, T., R. A. Khalaf and W. A. Fonzi, 2003 Nitrogen Garsin, D. A., C. D. Sifri, E. Mylonakis, X. Qin, K. V. Singh et al., metabolism and virulence of Candida albicans require the GATA- 2001 A simple model host for identifying Gram-positive viru- type transcriptional activator encoded by GAT1. Mol. Microbiol. lence factors. Proc. Natl. Acad. Sci. USA 98: 10892–10897. 50: 993–1004. Giles, S. S., T. R. Dagenais, M. R. Botts, N. P. Keller and C. M. Litvintseva, A. P., R. Thakur, R. Vilgalys and T. G. Mitchell, Hull, 2009 Elucidating the pathogenesis of spores from the 2006 Multilocus sequence typing reveals three genetic sub- human fungal pathogen Cryptococcus neoformans. Infect. Immun. populations of Cryptococcus neoformans var. grubii (serotype A), 77: 3491–3500. including a unique population in Botswana. Genetics 172: Gordon, J. L., K. P. Byrne and K. H. Wolfe, 2009 Additions, losses, 2223–2238. and rearrangements on the evolutionary route from a recon- Liu, O. W., C. D. Chun, E. D. Chow, C. Chen, H. D. Madhani et al., structed ancestor to the modern Saccharomyces cerevisiae genome. 2008 Systematic genetic analysis of virulence in the human fun- PLoS Genet. 5: e1000485. gal pathogen Cryptococcus neoformans. Cell 135: 174–188. Hall, C., and F. S. Dietrich, 2007 The reacquisition of biotin Livak, K. J., and T. D. Schmittgen, 2001 Analysis of relative gene prototrophy in Saccharomyces cerevisiae involved horizontal gene expression data using real-time quantitative PCR and the 2(-Delta transfer, gene duplication and gene clustering. Genetics 177: Delta C(T)). Method. Methods 25: 402–408. 2293–2307. Magasanik, B., and C. A. Kaiser, 2002 Nitrogen regulation in Sac- Heck, I. S., J. D. Schrag, J. Sloan, L. J. Millar, G. Kanan et al., charomyces cerevisiae. Gene 290: 1–18. 2002 Mutational analysis of the gephyrin-related molybdenum Marzluf, G. A., 1997 Genetic regulation of nitrogen metabolism in cofactor biosynthetic gene cnxE from the lower eukaryote Asper- the fungi. Microbiol. Mol. Biol. Rev. 61: 17–32. gillus nidulans. Genetics 161: 623–632. Masison, D. C., and R. B. Wickner, 1995 Prion-inducing domain of Hensel, M., H. N. Arst,Jr., A. Aufauvre-Brown and D. W. Holden, yeast Ure2p and protease resistance of Ure2p in prion-containing 1998 The role of the Aspergillus fumigatus areA gene in invasive cells. Science 270: 93–95. pulmonary aspergillosis. Mol. Gen. Genet. 258: 553–557. Merika, M., and S. H. Orkin, 1993 DNA-binding specificity of Honda, S., N. Ishii, K. Suzuki and M. Matsuo, 1993 Oxygen- GATA family transcription factors. Mol. Cell Biol. 13: 3999–4010. dependent perturbation of life span and aging rate in the nem- Mitchell, A. P., and B. Magasanik, 1984 Regulation of glutamine- atode. J. Gerontol. 48: B57–B61. repressible gene products by the GLN3 function in Saccharomyces Hynes, M. J., 1970 Induction and repression of amidase enzymes in cerevisiae. Mol. Cell Biol. 4: 2758–2766. Aspergillus nidulans. J. Bacteriol. 103: 482–487. Monahan, B. J., J. A. Fraser, M. J. Hynes and M. A. Davis, Idnurm, A., and J. Heitman, 2005 Light controls growth and de- 2002 Isolation and characterization of two ammonium perme- velopment via a conserved pathway in the fungal kingdom. PLoS ase genes, meaA and mepA, from Aspergillus nidulans. Eukaryot. Biol. 3: e95. Cell 1: 85–94. Nitrogen Metabolite Repression 323

Monahan, B. J., M. C. Askin, M. J. Hynes and M. A. Davis, Scholten, H. J., and R. L. M. Pierik, 1998 Agar as a gelling agent: 2006 Differential expression of Aspergillus nidulans ammonium chemical and physical analysis. Plant Cell Rep. 17: 230–235. permease genes is regulated by GATA transcription factor AreA. Shi, M., S. S. Li, C. Zheng, G. J. Jones, K. S. Kim et al., 2010 Real- Eukaryot. Cell 5: 226–237. time imaging of trapping and urease-dependent transmigration Mylonakis, E., F. M. Ausubel, J. R. Perfect, J. Heitman and S. B. of Cryptococcus neoformans in mouse brain. J. Clin. Invest. 120: Calderwood, 2002 Killing of Caenorhabditis elegans by Cryptococ- 1683–1693. cus neoformans as a model of yeast pathogenesis. Proc. Natl. Acad. Southern, E., 2006 Southern blotting. Nat. Protoc. 1: 518–525. Sci. USA 99: 15675–15680. Staib, F., S. K. Mishra, T. Able and A. Blisse, 1976 Growth of Nielsen, K., G. M. Cox, P. Wang, D. L. Toffaletti, J. R. Perfect Cryptococcus neoformans on uric acid agar. Zentralbl Bakteriol. et al., 2003 Sexual cycle of Cryptococcus neoformans var. grubii 236: 374–385 [Orig A]. and virulence of congenic a and alpha isolates. Infect. Immun. Staib, F., B. Grave, L. Altmann, S. K. Mishra, T. Abel et al., 71: 4831–4841. 1978 Epidemiology of Cryptococcus neoformans. Mycopathologia Nielsen, K., A. L. De Obaldia and J. Heitman, 2007 Cryptococcus 65: 73–76. neoformans mates on pigeon guano: implications for the realized Stanbrough, M., D. W. Rowen and B. Magasanik, 1995 Role of ecological niche and globalization. Eukaryot. Cell 6: 949–959. the GATA factors Gln3p and Nil1p of Saccharomyces cerevisiae in Olszewski, M. A., M. C. Noverr, G. H. Chen, G. B. Toews, G. M. the expression of nitrogen-regulated genes. Proc. Natl. Acad. Sci. Cox et al., 2004 Urease expression by Cryptococcus neoformans USA 92: 9450–9454. promotes microvascular sequestration, thereby enhancing central Stewart, V., and S. J. Vollmer, 1986 Molecular cloning of nit-2, nervous system invasion. Am. J. Pathol. 164: 1761–1771. a regulatory gene required for nitrogen metabolite repression in Pan, H., B. Feng and G. A. Marzluf, 1997 Two distinct protein- Neurospora crassa. Gene 46: 291–295. protein interactions between the NIT2 and NMR regulatory pro- Thompson, J. D., D. G. Higgins and T. J. Gibson, 1994 CLUSTAL teins are required to establish nitrogen metabolite repression in W: improving the sensitivity of progressive multiple sequence Neurospora crassa. Mol. Microbiol. 26: 721–729. alignment through sequence weighting, position-specific gap Platt, A., T. Langdon, H. N. Arst,Jr., D. Kirk, D. Tollervey et al., penalties and weight matrix choice. Nucleic Acids Res. 22: 1996 Nitrogen metabolite signalling involves the C-terminus 4673–4680. and the GATA domain of the Aspergillus transcription factor Walton, F. J., A. Idnurm and J. Heitman, 2005 Novel gene func- AREA and the 39 untranslated region of its mRNA. EMBO J. tions required for melanization of the human pathogen Crypto- 15: 2791–2801. coccus neoformans. Mol. Microbiol. 57: 1381–1396. Polacheck, I., and K. J. Kwon-Chung, 1980 Creatinine metabo- Wiame, J. M., M. Grenson and H. N. Arst,Jr., 1985 Nitrogen ca- lism in Cryptococcus neoformans and Cryptococcus bacillisporus. tabolite repression in yeasts and filamentous fungi. Adv. Microb. J. Bacteriol. 142: 15–20. Physiol. 26: 1–88. Postgate, J., 1998 Nitrogen Fixation. Cambridge University Press, Wong, K. H., M. J. Hynes and M. A. Davis, 2008 Recent advances in Cambridge, UK. nitrogen regulation: a comparison between Saccharomyces cerevisiae Pukkila-Worley, R., Q. D. Gerrald, P. R. Kraus, M. J. Boily, M. J. and filamentous fungi. Eukaryot. Cell 7: 917–925. Davis et al., 2005 Transcriptional network of multiple capsule Xu, S., D. A. Falvey and M. C. Brandriss, 1995 Roles of URE2 and melanin genes governed by the Cryptococcus neoformans cyclic and GLN3 in the proline utilization pathway in Saccharomyces AMP cascade. Eukaryot. Cell 4: 190–201. – fi cerevisiae. Mol. Cell Biol. 15: 2321 2330. Ravagnani, A., L. Gor nkiel, T. Langdon, G. Diallinas, E. Adjadj Xue, C., Y. Tada, X. Dong and J. Heitman, 2007 The human fungal et al., 1997 Subtle hydrophobic interactions between the sev- pathogen Cryptococcus can complete its sexual cycle during a path- enth residue of the zinc finger loop and the first base of an fi ogenic association with plants. Cell. Host Microbe 1: 263–273. HGATAR sequence determine promoter-speci c recognition by Zaragoza, O., B. C. Fries and A. Casadevall, 2003 Induction of the Aspergillus nidulans GATA factor AreA. EMBO J. 16: 3974– capsule growth in Cryptococcus neoformans by mammalian serum 3986. and CO(2). Infect. Immun. 71: 6155–6164. Rutherford, J. C., X. Lin, K. Nielsen and J. Heitman, 2008 Amt2 Zhao, Y., L. Medrano, K. Ohashi, J. C. Fletcher, H. Yu et al., permease is required to induce ammonium-responsive invasive 2004 HANABA TARANU is a GATA transcription factor that growth and mating in Cryptococcus neoformans. Eukaryot. Cell 7: regulates shoot apical meristem and flower development in Ara- – 237 246. bidopsis. Plant Cell 16: 2586–2600. Sambrook, J., E. F. Fritsch and T. Maniatis, 1989 Molecular Clon- ing: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Communicating editor: A. P. Mitchell GENETICS

Supporting Information http://www.genetics.org/cgi/content/full/genetics.111.128538/DC1

Nitrogen Metabolite Repression of Metabolism and Virulence in the Human Fungal Pathogen Cryptococcus neoformans

I. Russel Lee, Eve W. L. Chow, Carl A. Morrow, Julianne T. Djordjevic and James A. Fraser

FIGURE S1.Quantitative measurements of relative capsule diameters of wild-type H99 cells grown on YNB supplemented with various nitrogen sources (10 mM). A total of 60 cells (20 cells from three independent experiments) were measured for each nitrogen condition. Error bars represent standard errors. I. R. Lee et al. 3 SI

FIGURE S2.ClustalW multiple sequence alignment of GATA DNA-binding domains from various fungi retrieved either from GenBank or Broad Institute H99 database. Residues highlighted in yellow represent the universally conserved cysteines involved in zinc ion chelation. Amino acids at the central loop separating the second and third cysteines are in the range of 17-20 residues in length. Residues highlighted in blue indicate basic amino acids within 25 residues beyond the most C- terminal cysteine. Accession numbers or Broad annotations of protein sequences: Saccharomyces cerevisiae (Sc) Gln3 P18494.2, Candida albicans (Ca) Gln3p XP_721981.1, Cryptococcus neoformans (Cn) Gat1 CNAG00193.2, Neurospora crassa (Nc) Nit2 P19212.2, Gibberella fujikuroi (Gf) AreA P78688.1, Aspergillus parasiticus (Ap) AreA AAD37409.1, Aspergillus nidulans (An) AreA XP_681936.1, Aspergillus niger (As) AreA O13412.1, Aspergillus oryzae (Ao) AreA XP_001816951.1, Magnaporthe grisea (Mg) Nut1 XP_366679.1, Penicillium chrysogenum (Pc) Nre Q01582.1, Ca Gat1p AAP50501.1, Sc Nil/Gat1 NP_116632.1, Sc Dal80 CAA42757.1, Sc Gfz3 NP_012425.1, Debaryomyces hansenii (Dh) Gfz3 CAJ78410.1, Nc Asd4 Q9HEV5.1, An AreB XP_663825.1, Pc NreB XP_002557977.1, An SreA XP_657780.1, Pc SreP XP_002567610.1, Ustilago maydis (Um) Urbs1 XP_757197.1, Histoplasma capsulatum (Hc) Sre1 ABY66603.1, Cn Cir1 CNAG04864.2, Nc Sre XP_961978.1, Aspergillus fumigatus (Af) SreA XP_753523.1, Bipolaris oryzae (Bo) Blr2 BAF47401.1, Nc WC2 P78714.1, Cn Bwc2 CNAG02435.2, Schizosaccharomyces pombe (Sp) Gaf2 AAB38022.1, Sp Ams2 NP_588400.2, An NsdD XP_660756.1, Cn Gat201 CNAG01551.2, Sc Ash1 NP_012736.1, Ca Ash1p CAC48044.1, Nc WC1 Q01371.2, Sc Srd2 NP_015304.1 and Sc Srd1 NP_009944.2. 4 SI I. R. Lee et al.

FIGURE S3.Complementation of nitrogen utilisation phenotype to wild-type levels upon the re-introduction of GAT1 into the gat1 mutant. 10-fold spot dilution assays for nitrogen utilisation showed that the gat1 + GAT1 strain exhibited wild-type growth on 10 mM ammonium, uric acid, urea, creatinine and proline. I. R. Lee et al. 5 SI

FIGURE S4.Gat1 is dispensable for growth on pigeon guano. 10-fold spot dilution assays on 25% (wt/vol) pigeon guano media showed that both the gat1 and gat1 + GAT1 strains exhibited similar growth rates to wild-type H99. 6 SI I. R. Lee et al.

FIGURE S5.Quantitative measurements of relative capsule diameters of wild-type H99, gat1 and gat1 + GAT1 cells grown in RPMI and DMEM (10% fetal calf serum). A total of 60 cells (20 cells from three independent experiments) of each strain were measured. Error bars represent standard errors. I. R. Lee et al. 7 SI

TABLE S1

Cryptococcus strains used in this study

Molecular Country of Strain Strain details Source type origin

H99 Laboratory strain VNI USA J. Perfect

RL1 H99 gat1 VNI N/A This study

CM127 H99 gat1+GAT1 VNI N/A This study

RL2 H99 gat201 VNI N/A This study

RL3 H99 bwc2 VNI N/A This study

RL4 H99 CNAG04263.2 VNI N/A This study

RL5 H99 CNAG06762.2 VNI N/A This study

RL6 H99 CNAG03401.2 VNI N/A This study

H9CIR4 H99 cir1 VNI N/A J. Kronstad

RL8 KN99a gat1 VNI N/A This study

YL5 H99 ure1 VNI N/A This study

125.91 Clinical isolate VNI Tanzania J. Heitman

8-1 Clinical isolate VNII USA J. Heitman

I57 Clinical isolate VNII India B. Fries

JEC21 Laboratory strain VNIV USA J. Kwon-Chung

NIH430 Clinical isolate VNIV Denmark Kavanaugh

Bt33 Clinical isolate VNB Botswana A. Litvintseva

Bt63 Clinical isolate VNB Botswana A. Litvintseva

WM276 Environmental isolate VGI Australia W. Meyer

E566 Environmental isolate VGI Australia J. Fraser

R265 Environmental isolate VGII Canada J. Fraser

CBS 1930 Veterinary isolate VGII Aruba W. Meyer

NIH312 Clinical isolate VGIII USA J. Kwon-Chung 8 SI I. R. Lee et al.

B4546 Environmental isolate VGIII Australia J. Fraser

MMRL2651 Clinical isolate VGIV India W. Schell

Bt201 Clinical isolate VGIV Botswana A. Litvintseva

C. neoformans consists of variety grubii and variety neoformans, comprising four molecular types, VNI, VNII, VNB and VNIV (variety neoformans). C. gattii comprises molecular types VGI, VGII, VGIII and VGIV (variety gattii). I. R. Lee et al. 9 SI

TABLE S2

Primers used in this study

Primer name Purpose Sequence (5-3)

UQ613 URO1 qRT-PCR TGAGTACACCCTCCGAGTCCTT

UQ733 URO1 qRT-PCR TTTACAGTGTCGGTAGCAACGAC

UQ609 DAL1 qRT-PCR CTTATCGACACCCACGTCCA

UQ736 DAL1 qRT-PCR GCATGTCGATCAGGGTGGT

UQ739 URE1 qRT-PCR TAAGATCAAACTCGCTGACATGG

UQ740 URE1 qRT-PCR CTCGAATGACCTTACCTCCACC

UQ1070 PUT1 qRT-PCR AAAAGCTAAAGAGAACGACGTTGC

UQ1071 PUT1 qRT-PCR TCTCACTCTTTGACTTTGGAGGTTT

UQ1072 PUT5 qRT-PCR AGAAGGCTAAGGAGAACAACATCATT

UQ1073 PUT5 qRT-PCR GTCCAGATCTCCTCCTTGGAAG

UQ1074 PUT2 qRT-PCR TCATTAACGGTGAGGAGGTCAAG

UQ1075 PUT2 qRT-PCR GCAAGAGCACCGTCAATGG

UQ1586 GDH1 qRT-PCR GCTTCCCAGAATGAGCTTAACG

UQ1587 GDH1 qRT-PCR CGTCAAGTGTACAGCCCATGTT

UQ1588 GDH2 qRT-PCR CTTATGCTTATGTCGGATGGATCTT

UQ1589 GDH2 qRT-PCR TGGACTCATCCAGAGCATCCT

UQ1590 GLN1 qRT-PCR TCGAATCCCCAGGCATGT

UQ1591 GLN1 qRT-PCR CGACGAGGATGGCTGTGACT

UQ1592 GLT1 qRT-PCR CCGTTGGATCAGAGAAATTCTACC

UQ1593 GLT1 qRT-PCR CTGCTGAGGACCGAGGAAAC

UQ1630 AMT1 qRT-PCR TTATGTTTTGCTGGACTACCTTGGT

UQ1631 AMT1 qRT-PCR CGCCACCTGCATAGTCCAAT

UQ1632 AMT2 qRT-PCR TGAGGTTAAGGGCTAGCGAAGA 10 SI I. R. Lee et al.

UQ1633 AMT2 qRT-PCR CGGGATCTGTGCCAACATAGT

UQ482 ACT1 qRT-PCR CCTACAACTCTATCATGAAGTGTGATCTC

UQ728 ACT1 qRT-PCR TCTGCATACGGTCGGCAATAC

UQ856 GAT1 deletion GTTGCGGGATGCTGGTGAAAT

UQ857 GAT1 deletion AGCTCACATCCTCGCAGCCCTAGCCGCCGCTAGCTG

UQ858 GAT1 deletion GTGTTAATACAGATAAACCGCGAATGAATGGAAGCA

UQ859 GAT1 deletion GCCCAGCAACTACCATCGTGT

UQ946 GAT1 deletion CAGCTAGCGGCGGCTAGGGCTGCGAGGATGTGAGCT

UQ947 GAT1 deletion TTGCTTCCATTCATTCGCGGTTTATCTGTATTAACA

UQ864 GAT201 deletion CCGCTAAGGAAGGAAGTGGTG

UQ865 GAT201 deletion AGCTCACATCCTCGCAGCTGTGCTGGCAAGTAGGGA

UQ866 GAT201 deletion GTGTTAATACAGATAAACCTGAGAATAGGAAAGAAC

UQ867 GAT201 deletion CAGGGTAAAGGCCGTTAGATA

UQ963 GAT201 deletion TCCCTACTTGCCAGCACAGCTGCGAGGATGTGAGCTGGAGAGCG

UQ964 GAT201 deletion GCTGTTCTTTCCTATTCTCAGGTTTATCTGTATTAACACGGAAGAGATGTAG

UQ872 BWC2 deletion TACTCCAGTATCGCTGTCTCC

UQ873 BWC2 deletion AGCTCACATCCTCGCAGCTGGCCGTGCTATAGGTGA

UQ874 BWC2 deletion GTGTTAATACAGATAAACCGAGTAGTACTTTGGACT

UQ875 BWC2 deletion TCAGCCTCGCACATAATCTAA

UQ966 BWC2 deletion TCACCTATAGCACGGCCAGCTGCGAGGATGTGAGCTGGAGAGCG

UQ967 BWC2 deletion CAGTCCAAAGTACTACTCGGTTTATCTGTATTAACACGGAAGAGATGTAG

UQ876 CNAG04263.2 deletion CTTGAAGCTGCGTTAGAACAT

UQ877 CNAG04263.2 deletion CTCACATCCTCGCAGCTGTGGTTTTTGTAGGC

UQ878 CNAG04263.2 deletion GTTAATACAGATAAACCGGAGTGATTAAGAGG

UQ879 CNAG04263.2 deletion TTAGCTTCTGGCGTCTGATGA

UQ968 CNAG04263.2 deletion CCGCCTACAAAAACCACAGCTGCGAGGATGTGAGCTGGAGAGCG

UQ969 CNAG04263.2 deletion CCACCTCTTAATCACTCCGGTTTATCTGTATTAACACGGAAGAGATGTAG I. R. Lee et al. 11 SI

UQ888 CNAG06762.2 deletion GGGAAGATTAGCGCACCAAAC

UQ889 CNAG06762.2 deletion AGCTCACATCCTCGCAGCTGTTGGGGAAAAAGGAAA

UQ890 CNAG06762.2 deletion GTGTTAATACAGATAAACCGTGTTGTAACGAGTCGA

UQ891 CNAG06762.2 deletion CGCTGCTCTTATCATGTCCAT

UQ974 CNAG06762.2 deletion TTTCCTTTTTCCCCAACAGCTGCGAGGATGTGAGCTGGAGAGCG

UQ975 CNAG06762.2 deletion ATCGACTCGTTACAACACGGTTTATCTGTATTAACACGGAAGAGATGTAG

UQ852 CNAG03401.2 deletion CTTGAAGGTTAATCGTGACGG

UQ853 CNAG03401.2 deletion AGCTCACATCCTCGCAGCCCTATAATTCTCTTTTAT

UQ854 CNAG03401.2 deletion TGTTAATACAGATAAACCAATGCCAGGGAGAGAATT

UQ855 CNAG03401.2 deletion ATCCGCCAACAACAGTACCAC

UQ960 CNAG03401.2 deletion ATAAAAGAGAATTATAGGGCTGCGAGGATGTGAGCTGGAGAGCG

UQ961 CNAG03401.2 deletion AATTCTCTCCCTGGCATTGGTTTATCTGTATTAACACGGAAGAGATGTAG

UQ1234 GAT1 complementation GTGGCGCGCCGCTGGTGAAAT

UQ1235 GAT1 complementation GCCCAGGAGCTCCCATCGTGT

M13F General sequencing GTAAAACGACGGCCAG

M13R General sequencing CAGGAAACAGCTATGAC

UQ1027 GAT1 sequencing GAAAGGCGAATAAAGGCTGTT

UQ1028 GAT1 sequencing TCTGCCGCGTCTTCCCGGCCC

UQ1029 GAT1 sequencing CCAAAAGCTGCTAGGGGCACC

UQ1030 GAT1 sequencing TTGCCAAACGGTTTGTCTCTT

UQ1031 GAT1 sequencing GCAGCCGCAGCAACGGCTGCG

UQ1032 GAT1 sequencing AAAAAAAAGCATAAATAAAAA

UQ1033 GAT1 sequencing TGACCGCACAGACGTATTTTT

UQ1034 GAT1 sequencing AGCTTTCGGTAATGGCCGGAT

UQ1035 GAT1 sequencing GTGGTCGAAGGATGTCTTCCG

qRT-PCR primers were designed using Primer Express software 2.0 (Applied Biosystems). All remaining primers were designed using Oligo 6.8 (Molecular Biology Insights, West Cascade CO). 12 SI I. R. Lee et al.

TABLE S3

Nitrogen utilisation profile of various Cryptococcus strains

Molecular VNI VNII VNIV VNB VGI VGII VGIII VGIV type

Strains

8-1 I57 H99

Bt33 Bt63 E566 R265 Bt201 B4546 JEC21 125.91 WM276 WM276 NIH430 NIH312 CBS 1930 CBS Nitrogen MMRL2651 sources

Prolinea +++ +++ ++ +++ ++ +++ + +++ +++ +++ +++ +++ ++ ++ +++ +++

Glutaminea +++ +++ ++ ++ +++ +++ + +++ +++ +++ +++ +++ ++ + +++ +++

Asparaginea +++ +++ ++ +++ +++ +++ + ++ ++ +++ +++ +++ +++ + +++ +++

Glutamatea +++ +++ ++ +++ +++ +++ + +++ +++ +++ +++ +++ ++ + ++ ++

GABAa +++ +++ + +++ +++ +++ +/- +++ +++ +++ +++ +++ ++ + +++ +++

Creatininea +++ + ++ ++ +++ +++ + + +++ +++ +++ +++ +++ + +++ +++

Glycinea +++ +++ + ++ ++ +++ + ++ +++ +++ +++ +++ ++ + +++ +++ I. R. Lee et al. 13 SI

Ammoniuma +++ +++ + ++ +++ +++ + +++ +++ ++ +++ +++ + +/- +++ ++

Argininea ++ ++ + ++ + ++ ++ ++ +++ +++ +++ +++ ++ + +++ ++

Aspartic acidb ++ ++ + + + +/- + +++ +++ +++ +++ +++ + /- +++ +++

Ureab ++ ++ ++ ++ + + ++ ++ ++ ++ ++ ++ + +/- ++ ++

Alanineb ++ + + ++ ++ ++ +/- + ++ ++ ++ ++ + +/- ++ ++

Tyrosineb ++ ++ + ++ + ++ + ++ ++ ++ + ++ +/- /- ++ ++

Guanineb ++ ++ + + ++ + +/- + + ++ ++ ++ + +/- ++ ++

Serineb + + +/- + ++ ++ +/- + + ++ ++ ++ ++ + ++ +

Uric acidb ++ ++ + + /- /- + ++ ++ ++ ++ ++ + ++ ++ ++

Kynurenineb + + + + + + +/- + ++ ++ ++ ++ + +/- ++ +

Citrullineb ++ ++ +/- + + + ++ ++ + + + + +/- +/- + +

Histidineb + + + + + + + + ++ + + + + +/- + +

Tryptophanc ++ +/- + + +/- + +/- ++ + + + + +/- +/- + +

Threoninec + + + + + + +/- + + + + + +/- +/- + +

Leucinec + +/- + + + + +/- + ++ + + + +/- /- + + 14 SI I. R. Lee et al.

Ornithinec + + +/- +/- +/- +/- ++ + + + + + +/- +/- ++ +

Allantoinc + + + + +/- +/- + +/- + + + + +/- +/- + +

Lysinec ++ ++ + + +/- +/- +/- ++ + + + + - - + +

Isoleucinec + + + +/- + + +/- +/- + + +/- + +/- +/- +/- +

Methioninec + +/- +/- +/- +/- + +/- +/- + + + + +/- +/- + +

Valinec +/- +/- +/- +/- +/- +/- +/- +/- ++ + + + + +/- + +

Guanosinec /- /- + + + + + + + + +/- +/- +/- +/- + +

Phenylalaninec + + +/- +/- +/- +/- +/- + + +/- +/- + /- /- + +/-

Adeninec +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/-

Adenosinec +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/-

Glucosamined +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- /- /- +/- +/-

Acetamided +/- +/- +/- +/- +/- +/- /- +/- +/- +/- +/- +/- +/- /- +/- +/-

Formamided +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- +/- /- /- /- +/-

Nicotinamided +/- +/- /- +/- +/- +/- /- +/- - +/- +/- +/- /- /- +/- +/-

Uracild +/- +/- +/- +/- /- /- /- +/- +/- +/- +/- +/- /- /- +/- +/- I. R. Lee et al. 15 SI

Niacinamided +/- +/- /- +/- +/- /- /- +/- /- /- /- +/- /- /- +/- +/-

Xanthined - - /- /- - - /- /- - /- /- /- /- /- - -

Cadaverinee ------

Nitritee ------

Nitratee ------aGood nitrogen source; bAverage nitrogen source; cPoor nitrogen source; dVery poor nitrogen source; eNot a nitrogen source

+++ denotes excellent growth; ++ denotes good growth; + denotes moderate growth; +/- denotes poor growth; /- denotes very poor growth; - denotes no growth

10-fold spot dilution assays revealed that the strains exhibited varying growth rates on different nitrogen sources (10 mM) that were mainly limited to ammonium, purines and amino acids. 16 SI I. R. Lee et al.

TABLE S4

Reciprocal BLASTp analyses of characterised fungal GATA factors against the C. neoformans

genome

Species Protein Function Reference Best hit in C. E value f N. crassa Nit2 Nitrogen metabolism Stewart 1986 CNAG00193.2 1.12689E-21 (Positively acting) (Gat1) G. fujikuroi AreA Mihlan 2003 1.28176E-19

A. parasiticus AreA Chang 2000 1.16799E-24

A. nidulans AreA Kudla 1990 9.05971E-25

A. niger AreA MacCabe 1998 2.24874E-23

A. oryzae AreA Christensen 1998 1.16799E-24

A. fumigatus AreA Hensel 1998 1.23235E-21

M. grisea Nut1 Froeliger 1996 1.81053E-34

P. chrysogenum Nre Haas 1995 3.09042E-23

C. albicans Gat1p Limjindaporn 2003 1.14722E-20

S. cerevisiae Gln3 Minehart 1991 5.14654E-18

C. albicans Gln3p Liao 2008 1.29509E-15

S. cerevisiae Gat1 Stanbrough 1995 2.77123E-20

C. neoformans Gat1 Kmetzsch 2010 0.0

D. hansenii DhGzf3 Nitrogen metabolism Garcia-Salcedo 2006 9.62535E-19 (Negatively acting) S. cerevisiae Gzf3 Soussi-Boudekou 1997 5.37613E-17

A. nidulans AreB Conlon 2001 9.70406E-20

P. chrysogenum NreB Haas 1997 1.3125E-18

S. cerevisiae Dal80 Cunningham 1992 6.18739E-16

H. capsulatum Sre1 Iron uptake Chao 2008 8.20534E-17

A. fumigatus SreA Schrettl 2008 2.02048E-16

N. crassa Asd4 Sexual development Feng 2000 3.01803E-17

U. maydis Urbs1 Iron uptake Voisard 1993 CNAG04864.2 4.96038E-20 (Cir1) A. nidulans SreA Haas 1999 1.23395E-13

P. chrysogenum SreP Haas 1997 1.35569E-17

N. crassa Sre Zhou 1998 3.6177E-19

S. pombe Gaf2 Hoe 1996 1.00542E-26

C. neoformans Cir1 Jung 2006 0.0 I. R. Lee et al. 17 SI

N. crassa WC1 Blue light signal Ballario 1996 CNAG05181.2 0.0 (Bwc1) *

N. crassa WC2 Linden 1997 CNAG02435.2 2.91186E-12 (Bwc2) B. oryzae Blr2 Moriwaki 2008 2.86005E-13

C. neoformans Bwc2 Idnurm 2005 0.0

S. cerevisiae Ash1 Pseudohyphal growth Chandarlapaty 1998 0.00163479

A. nidulans NsdD Sexual development Han 2001 CNAG01551.2 6.06295E-11 (Gat201) S. cerevisiae Srd1 Pre-rRNA processing Hess 1994 4.0921E-5

S. cerevisiae Srd2 Pseudohyphal growth Canizares 2002 2.55412E-7

C. neoformans Gat201 Anti-phagocytosis Liu 2008 0.0

S. pombe Ams2 Cell cycle Chen 2003 CNAG03401.2 4.1389E-9

C. albicans Ash1p Pseudohyphal growth Munchow 2002 0.00266078

* C. neoformans Bwc1 is not strictly a GATA factor due an absence of the zinc finger DNA-binding domain in this ortholog

Protein sequences of fungal GATA factors retrieved from GenBank were queried against the C. neoformans strain H99 genome sequence available at the Broad Institute. 18 SI I. R. Lee et al.

TABLE S5

Growth responses of wild-type H99 and GATA-type deletion mutants on various utilisable

nitrogen sources

Strains

Nitrogen source H99 cir1 gat1 bwc2 gat201 CNAG04263.2 CNAG06762.2 CNAG03401.2

Proline +++ ++ +++ +++ +++ +++ +++ +++ Glutamine +++ +/- +++ +++ +++ +++ +++ +++ Asparagine +++ +/- +++ +++ +++ +++ +++ +++ Creatinine +++ - +++ +++ +++ +++ +++ +++ Glutamate +++ +/- +++ +++ +++ +++ +++ +++ GABA +++ +/- +++ +++ +++ +++ +++ +++ Glycine +++ - +++ +++ +++ +++ +++ +++ Ammonium +++ - +++ +++ +++ +++ +++ +++ Arginine ++ + ++ ++ ++ ++ ++ ++ Uric acid ++ - ++ ++ ++ ++ ++ ++ Urea ++ - ++ ++ ++ ++ ++ ++ Aspartic acid ++ +/- ++ ++ ++ ++ ++ ++ Alanine ++ +/- ++ ++ ++ ++ ++ ++ Tyrosine ++ - ++ ++ ++ ++ ++ ++ Guanine ++ - ++ ++ ++ ++ ++ ++ Citrulline ++ - ++ ++ ++ ++ ++ ++ Tryptophan ++ - ++ ++ ++ ++ ++ ++ Lysine ++ +/- ++ ++ ++ ++ ++ ++ Serine + - + + + + + + Kynurenine + - + + + + + + Histidine + - + + + + + + Threonine + - + + + + + + Leucine + - + + + + + + Ornithine + +/- + + + + + + Isoleucine + - + + + + + + Allantoin + - + + + + + + Methionine + - + + + + + + Phenylalanine + - + + + + + + Valine +/- - +/- +/- +/- +/- +/- +/- Adenine +/- - +/- +/- +/- +/- +/- +/- Adenosine +/- - +/- +/- +/- +/- +/- +/- Guanosine +/- - +/- +/- +/- +/- +/- +/- Glucosamine +/- - +/- +/- +/- +/- +/- +/- Formamide +/- - +/- +/- +/- +/- +/- +/- I. R. Lee et al. 19 SI

Nicotinamide +/- - +/- +/- +/- +/- +/- +/- Uracil +/- - +/- +/- +/- +/- +/- +/- Acetamide +/- - +/- +/- +/- +/- +/- +/- Niacinamide +/- - +/- +/- +/- +/- +/- +/-

- denotes no growth; +/- denotes poor growth; + denotes moderate growth; ++ denotes good growth; +++ denotes excellent growth

Media was supplemented with 10 mM of each nitrogen source. Summarised results of the nitrogen utilisation assays. 20 SI I. R. Lee et al.

TABLE S6

Sequence identity of characterised fungal GATA factors in comparison to predicted C. neoformans members

% DNA-binding domain (Protein) identity to C. neoformans GATA factor candidates

Species Protein Function (Cir1) (Gat1) (Bwc2) (Gat201) CNAG00193.2 CNAG01551.2 CNAG02435.2 CNAG03401.2 CNAG04263.2 CNAG04864.2 CNAG06762.2

N. crassa Nit2 Nitrogen metabolism (Postively-acting) 90 (19) 31 (14) 32 (13) 40 (18) 36 (17) 64 (20) 28 (15) G. fujikuroi AreA 88 (19) 31 (17) 32 (15) 40 (18) 36 (18) 66 (19) 26 (18) A. parasiticus AreA 88 (19) 31 (18) 32 (15) 34 (18) 36 (19) 66 (21) 26 (17) A. nidulans AreA 88 (18) 31 (17) 32 (16) 40 (17) 36 (20) 66 (21) 26 (17) A. niger AreA 88 (18) 31 (19) 32 (15) 40 (18) 36 (19) 66 (20) 26 (19) A. oryzae AreA 88 (18) 31 (18) 32 (15) 34 (17) 36 (17) 66 (20) 26 (17) M. grisea Nut1 88 (18) 31 (16) 32 (15) 34 (18) 36 (18) 66 (21) 26 (16) P. chrysogenum Nre 88 (17) 31 (17) 32 (18) 34 (18) 36 (19) 66 (19) 26 (17) C. albicans Gat1p 86 (15) 39 (15) 34 (14) 44 (18) 36 (17) 66 (18) 38 (16) S. cerevisiae Gat1 78 (13) 29 (16) 38 (17) 40 (15) 38 (16) 64 (13) 28 (19) C. albicans Gln3p 76 (17) 39 (17) 34 (16) 32 (17) 36 (16) 60 (19) 26 (19) S. cerevisiae Gln3 72 (18) 39 (17) 34 (16) 32 (17) 36 (17) 58 (18) 16 (17) I. R. Lee et al. 21 SI

D. hansenii DhGzf3 Nitrogen metabolism (Negatively-acting) 74 (17) 31 (21) 34 (20) 36 (20) 38 (18) 66 (16) 22 (17) S. cerevisiae Gzf3 74 (16) 29 (18) 32 (19) 32 (17) 38 (18) 68 (18) 26 (17) A. nidulans AreB 74 (10) 21 (20) 38 (19) 34 (15) 40 (7) 68 (12) 28 (17) P. chrysogenum NreB 74 (8) 37 (19) 38 (20) 34 (16) 40 (16) 70 (12) 28 (18) S. cerevisiae Dal80 72 (4) 29 (11) 32 (3) 36 (8) 38 (5) 62 (4) 24 (10) U. maydis Urbs1 Iron uptake 68 (22) 35 (10) 34 (15) 36 (19) 42 (19) 76 (19) 22 (16) A. nidulans SreA 64 (16) 35 (18) 32 (19) 38 (20) 36 (20) 76 (17) 24 (22) H. capsulatum Sre1 64 (17) 35 (20) 30 (18) 38 (22) 40 (17) 74 (17) 22 (19) P. chrysogenum SreP 74 (17) 35 (21) 30 (17) 36 (18) 36 (18) 74 (17) 26 (20) A. fumigatus SreA 44 (16) 39 (18) 36 (18) 42 (19) 32 (20) 46 (16) 28 (19) N. crassa Sre 44 (16) 37 (18) 38 (19) 34 (18) 34 (17) 44 (16) 28 (18) S. pombe Gaf2 46 (17) 33 (20) 38 (18) 28 (19) 38 (18) 40 (17) 30 (22) C. neoformans Cir1 64 (16) 39 (11) 32 (8) 36 (14) 36 (13) 100 (100) 24 (10) B. oryzae Blr2 Blue light signal transduction 32 (12) 47 (23) 63 (21) 50 (19) 32 (18) 28 (17) 36 (18) N. crassa WC2 36 (16) 47 (21) 58 (22) 46 (20) 28 (18) 28 (16) 34 (19) C. neoformans Bwc2 34 (9) 45 (19) 100 (100) 50 (15) 34 (12) 32 (8) 32 (15) N. crassa WC1 44 (17) 39 (14) 40 (11) 38 (17) 34 (17) 38 (17) 30 (14) N. crassa Asd4 Sexual development 74 (13) 37 (16) 33 (21) 35 (18) 37 (17) 66 (12) 25 (18) A. nidulans NsdD 41 (15) 54 (27) 41 (19) 37 (20) 35 (19) 31 (17) 43 (18) S. cerevisiae Srd1 Pre-rRNA processing 34 (6) 41 (16) 36 (19) 26 (14) 28 (9) 24 (7) 32 (15) 22 SI I. R. Lee et al.

C. neoformans Gat201 Anti-phagocytosis 33 (12) 100 (100) 45 (19) 43 (16) 33 (15) 39 (10) 41 (17)

S. pombe Ams2 Cell cycle 29 (18) 25 (18) 27 (17) 31 (18) 31 (18) 31 (17) 25 (21)

S. cerevisiae Ash1 Pseudohyphal growth 38 (17) 29 (19) 38 (21) 38 (19) 22 (18) 36 (18) 32 (19) S. cerevisiae Srd2 34 (4) 35 (15) 34 (18) 26 (12) 32 (8) 22 (4) 40 (13) C. albicans Ash1p 36 (12) 29 (19) 36 (17) 38 (20) 22 (18) 32 (16) 26 (21)

Protein sequences of fungal GATA factors retrieved from GenBank were individually aligned with each C. neoformans strain H99 GATA factor candidate using ClustalW to obtain the percentage (%) DNA-binding domain/protein identity. Values highlighted in bold represent the highest % identity with DNA-binding domain identity given priority over protein identity, due to the established fact of high conservation at each function-specific GATA DNA-binding domain.

CHAPTER 4

Results Characterization of an Nmr Homolog That Modulates GATA Factor-Mediated Nitrogen Metabolite Repression in Cryptococcus neoformans

56 4.1 Preface

The ability of Cryptococcus neoformans to sense nitrogen availability and respond with appropriate metabolic remodeling is crucial for adaptation, proliferation and colonisation in an infected host. As the previous chapter has concluded, Gat1/Are1 represents a nitrogen-responsive transcription factor that integrates metabolic functions (nitrogen acquisition) with elaboration of virulence factors (capsule, melanin and high temperature growth). This chapter extends the investigation from Chapter 3 to dissect the molecular mechanisms by which Gat1/Are1 activity is regulated in response to changes in nitrogen availability, as would occur when C. neoformans leaves its purine-rich ecological niche of pigeon guano to infect a mammalian host, which in comparison is nutrient limiting.

57 Characterization of an Nmr Homolog That Modulates GATA Factor-Mediated Nitrogen Metabolite Repression in Cryptococcus neoformans

I. Russel Lee1,2, Jonathan W. C. Lim1,2, Kate L. Ormerod1,2, Carl A. Morrow1,2, James A. Fraser1,2* 1 Australian Infectious Diseases Research Centre, University of Queensland, Brisbane, Queensland, Australia, 2 School of Chemistry and Molecular Biosciences, University of Queensland, Brisbane, Queensland, Australia

Abstract Nitrogen source utilization plays a critical role in fungal development, secondary metabolite production and pathogenesis. In both the Ascomycota and Basidiomycota, GATA transcription factors globally activate the expression of catabolic enzyme- encoding genes required to degrade complex nitrogenous compounds. However, in the presence of preferred nitrogen sources such as ammonium, GATA factor activity is inhibited in some species through interaction with co-repressor Nmr proteins. This regulatory phenomenon, nitrogen metabolite repression, enables preferential utilization of readily assimilated nitrogen sources. In the basidiomycete pathogen Cryptococcus neoformans, the GATA factor Gat1/Are1 has been co-opted into regulating multiple key virulence traits in addition to nitrogen catabolism. Here, we further characterize Gat1/Are1 function and investigate the regulatory role of the predicted Nmr homolog Tar1. While GAT1/ARE1 expression is induced during nitrogen limitation, TAR1 transcription is unaffected by nitrogen availability. Deletion of TAR1 leads to inappropriate derepression of non-preferred nitrogen catabolic pathways in the simultaneous presence of favoured sources. In addition to exhibiting its evolutionary conserved role of inhibiting GATA factor activity under repressing conditions, Tar1 also positively regulates GAT1/ARE1 transcription under non-repressing conditions. The molecular mechanism by which Tar1 modulates nitrogen metabolite repression, however, remains open to speculation. Interaction between Tar1 and Gat1/Are1 was undetectable in a yeast two-hybrid assay, consistent with Tar1 and Gat1/Are1 each lacking the conserved C-terminus regions present in ascomycete Nmr proteins and GATA factors that are known to interact with each other. Importantly, both Tar1 and Gat1/Are1 are suppressors of C. neoformans virulence, reiterating and highlighting the paradigm of nitrogen regulation of pathogenesis.

Citation: Lee IR, Lim JWC, Ormerod KL, Morrow CA, Fraser JA (2012) Characterization of an Nmr Homolog That Modulates GATA Factor-Mediated Nitrogen Metabolite Repression in Cryptococcus neoformans. PLoS ONE 7(3): e32585. doi:10.1371/journal.pone.0032585 Editor: Kirsten Nielsen, University of Minnesota, United States of America Received January 3, 2012; Accepted February 1, 2012; Published March 28, 2012 Copyright: ß 2012 Lee et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was supported by the University of Queensland International Research Tuition Award (UQIRTA) and University of Queensland Research Scholarship (UQRS) awarded to RL, and the National Health and Medical Research Council (NHMRC) CDA 569673 grant awarded to JF. No additional external funding was received for this study. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: [email protected]

Introduction both ammonium and glutamine are metabolites that likely trigger nitrogen metabolite/catabolite repression, resulting in the gener- With over 1,500,000 species estimated to be present in the ation of signals that antagonize activation of secondary (non- Earth’s biosphere, the kingdom Fungi consists of a spectacularly preferred) nitrogen gene expression by GATA factors [10,11,12]. large group of eukaryotes [1]. To successfully thrive in their broad Although GATA factors are employed to globally control range of niches, many fungal species have evolved mechanisms nitrogen metabolism in S. cerevisiae, N. crassa and A. nidulans, there that enable them to utilize a wide variety of nitrogen sources. The are compelling differences in key aspects of the molecular circuitry expression of permease and catabolic enzyme-encoding genes that underlies the regulation of these transcription factors in needed for the scavenging of most nitrogenous compounds different fungal species. Studies thus far have shown that GATA requires activation by global transcription factors belonging to factor activity is regulated in response to nitrogen availability in the GATA family. These nitrogen regulatory GATA factors are the surroundings and/or the nitrogen status of the cell. In S. conserved throughout the phyla Ascomycota and Basidiomycota. In the cerevisiae, transcriptional activation of GAT1 is cross-regulated by model ascomycetes, two positively acting GATA factors GLN3 and both the positively acting factor Gln3 and negatively acting Dal80 GAT1 are encoded in the genome of the yeast Saccharomyces cerevisiae [13]. The functions of Gln3 and Gat1 are also regulated while single positively acting factors nit-2 and areA are encoded in posttranscriptionally via protein subcellular localization, through the genomes of the moulds Neurospora crassa and Aspergillus nidulans, phosphorylation and interaction with the prion-forming glutathi- respectively [2,3,4,5,6,7,8]. Loss-of-function mutations in these one S-transferase Ure2 [14,15,16,17]. ascomycete GATA genes result in an inability to utilize a diverse In N. crassa, the function of Nit2 is instead regulated via array of nitrogen sources apart from the readily assimilated and interaction with a co-repressor protein known as Nmr1 (Nitrogen hence generally preferred ammonium or glutamine [9]. Thus, metabolic regulation 1) [18]. Loss-of-function mutations in nmr-1

PLoS ONE | www.plosone.org 1 March 2012 | Volume 7 | Issue 3 | e32585 Nitrogen Metabolite Repression in Cryptococcus result in derepression of a range of secondary nitrogen catabolic virulence factor expression and pathogenesis. Overall, our study genes under normally repressing conditions (nitrogen sufficient provides evidence of divergence between different fungal species in conditions, e.g. in the presence of ammonium and/or glutamine) the evolution of Nmr-associated proteins, and demonstrates the [19]. Unlike S. cerevisiae Ure2, there is no evidence suggesting that importance of Tar1 in modulating C. neoformans virulence. the structurally unrelated N. crassa Nmr1 affects Nit2 subcellular localization. Rather, Nmr1 likely exerts its effect by modulating Materials and Methods the trans-activation function of Nit2 by interfering with its DNA binding activity [20]. Like N. crassa Nit2, the function of A. nidulans Strains and media AreA is also regulated by an Nmr ortholog, NmrA [12,21,22,23]. All fungal strains used in this study are listed in Table S1, and In addition, AreA activity is further controlled by autogenous were grown in YPD (1% yeast extract, 2% Bacto-peptone, 2% regulation and control of transcript stability mediated through an glucose) or YNB (0.45% yeast nitrogen base w/o amino acids and element in the 39 untranslated region of the areA mRNA ammonium sulfate, 2% glucose, 10 mM nitrogen source) unless [23,24,25]. These pioneering studies of nitrogen metabolism in specified otherwise. C. neoformans biolistic transformants were S. cerevisiae, N. crassa and A. nidulans have played a crucial role in selected on YPD medium supplemented with 200 mg/mL G418 our current understanding of gene regulation in eukaryotes. (Sigma) or 100 mg/mL nourseothricin (Werner BioAgents). Over the past decade, interest in nitrogen regulation has Melanin-inducing media using L-3,4-dihydroxyphenylalanine (L- expanded to encompass a number of important human fungal DOPA), norepinephrine or caffeic acid as the laccase substrate pathogens, implicating gene regulation by nitrogen availability in supplemented with 10 mM of the specific nitrogen source were virulence. For example, during nitrogen limitation, Candida albicans prepared as described previously [32,38,44,45,46]. Unfiltered 1% dimorphic transition from the budding yeast to filamentous growth pigeon guano medium was also prepared as described previously form that facilitates tissue invasion is dependent on the Mep2 [34,47]. Escherichia coli Mach-1 cells served as the host strain for ammonium permease, whose expression is regulated by the GATA transformation and propagation of all plasmids using lysogeny factors Gln3 and Gat1 [26,27]. Both gln3D and gat1D mutants broth supplemented with either 100 mg/mL ampicillin (Sigma) or exhibit reduced virulence in a murine model of disseminated 50 mg/mL kanamycin (Sigma) [48]. Caenorhabditis elegans strain N2 candidiasis [28,29]. Likewise, the GATA factor AreA enables was maintained at 15uC and propagated on its normal laboratory nutritional versatility that is a key attribute influencing the ability food source E. coli OP50 cells [49,50,51]. Nematode growth of Aspergillus fumigatus to cause disease [30]. Consistent with this medium (NGM) was prepared as described previously [49]. notion, the AreA ortholog in Penicillium marneffei has recently been proposed to contribute to pathogenicity by regulating the Bioinformatic analyses production of extracellular proteases that are potential virulence C. neoformans genes were identified using annotation from the factors [31]. Impact of nitrogen regulation on virulence is not H99 genome sequence from the Broad Institute (http://www. merely limited to members of the Ascomycota but is also observed in broadinstitute.org/annotation/genome/cryptococcus_neoformans/ the basidiomycete Cryptococcus neoformans, a species that causes life- MultiHome.html). Gene annotations from the Broad are designated threatening meningoencephalitis predominantly in immunocom- by their nomenclature ‘‘CNAG#####.#’’. Sequence analyses promised individuals [32]. For instance, apart from nitrogen were performed using BLAST and MacVector 9.5 (MacVector Inc, catabolism, the GATA factor Gat1/Are1 regulates multiple Cary NC) [52]. Sequence alignments were created using ClustalW virulence attributes including infectious basidiospore production, v1.4 within MacVector [53]. Sequence traces generated at the capsule biosynthesis, high temperature growth and melanin Australian Genome Research Facility (Brisbane, Queensland) were pigment formation [33,34]. analysed using Sequencher 4.7 (Gene Codes Corporation, Ann Gaining insights into the molecular mechanism governing the Arbor MI). regulation of GATA factor activity is therefore important for the understanding of how these pathogenic fungi establish disease in a Construction and complementation of C. neoformans mammalian host. In C. albicans, Gln3 is proposed to be regulated mutant strains by an Ure2-like mechanism, similar to other closely related yeasts All primers and plasmids used in this study are listed in Table [35,36]. In contrast, AreA of both A. fumigatus and P. marneffei are S2 and S3, respectively. Gene deletion mutants were created using likely to be regulated in an Nmr-like fashion, similar to the related overlap PCR and biolistic transformation as described previously moulds [31,37]. A recent study by Jiang et al. suggests that the [54]. Briefly, to construct the tar1D mutant strain in the H99 yeast C. neoformans may follow this previously mould-specific background, the 1,067 bp TAR1 (CNAG04934.2) coding sequence paradigm [38]. The potential Nmr homolog Tar1 (Temperature was replaced with the neomycin phosphotransferase II-encoding associated repressor 1) was first identified in C. neoformans through selectable marker NEO using a construct created by overlap PCR the overproduction of melanin at 37uC in a clone from a random combining a ,1 kb fragment upstream the TAR1 start codon, the insertional mutagenesis library [38]. The expression of TAR1 was NEO marker and a ,1 kb fragment downstream the TAR1 stop induced at high temperature (37uC), and Tar1 was found to codon. Strain H99 genomic DNA and plasmid pJAF1 were used negatively regulate LAC1-encoded laccase that catalyses the as PCR templates [55]. The construct was transformed into C. formation of melanin in C. neoformans [38]. Like N. crassa Nmr1 neoformans cells via particle bombardment and transformants and A. nidulans NmrA, Tar1 harbours a predicted canonical selected on YPD plates supplemented with G418. A similar Rossmann fold motif that is also found in other co-repressors such approach was adopted to delete TAR1 in the H99 gat1/are1::NEO as S. cerevisiae Gal80 and mammalian CtBP [21,38,39,40,41, and wild-type KN99a strains using the NAT selectable marker 42,43]. from pCH233, with transformants selected on YPD supplemented As part of an ongoing effort to elucidate the global nitrogen with nourseothricin. Deletion of TAR1 was confirmed by regulatory circuit in the most clinically prevalent form of C. diagnostic PCR and Southern blot [56]. To complement the neoformans (var. grubii strain H99), we have investigated the genetic H99 tar1D mutant, the TAR1 gene including ,1 kb promoter and and physical interactions between Tar1 and the GATA factor terminator was amplified from genomic DNA using high fidelity Gat1/Are1. In addition, we have examined the role of Tar1 in PCR, cloned into pCR2.1-TOPO (Invitrogen) to give pIRL25,

PLoS ONE | www.plosone.org 2 March 2012 | Volume 7 | Issue 3 | e32585 Nitrogen Metabolite Repression in Cryptococcus and sequenced. The TAR1 fragment of pIRL25 was then (nucleotides 2,563–3,834) GAT1/ARE1 fragments, as well as full- subcloned into pCH233, creating the complementation construct length GAT1/ARE1 (3,834 nucleotides) and TAR1 (867 nucleo- pIRL26. pIRL26 was subsequently linearised and biolistically tides), were amplified from either pIRL27 or pIRL21 using transformed into the tar1D mutant. Stable transformants were restriction site-incorporating primers, subcloned into both selected on YPD supplemented with nourseothricin and comple- pGBKT7 and pGADT7 to give pIRL28-37, and again sequenced mented strains containing a single copy of the wild-type TAR1 to ensure correct reading frames obtained. The generated prey gene were identified by Southern blot. and bait constructs with different inserts, as well as the relevant positive (pGBKT7-53, pGADT7-T) and negative (pGBKT7, Quantitative real-time PCR pGADT7) controls, were subsequently co-transformed into the S. C. neoformans strains were grown in YNB or L-DOPA cerevisiae reporter strain AH109 using the lithium acetate/heat suppflemented with 10 mM of the specified nitrogen source shock method [59]. Transformants were selected on YNB medium and shaken at 30 or 37uC for 16 hr. For nitrogen starvation lacking leucine and tryptophan. Interaction was assessed by cultures, ammonium-grown cells were gently centrifuged at growth in the absence of adenine and histidine (+5 mM 3-Amino- 1,000 rpm for 3 min, the loose pellets washed with YNB, and 1,2,4-triazole), and b-galactosidase activity [60]. cells transferred to fresh YNB medium lacking a nitrogen source for an additional 4 hr. Overnight cultures were harvested, cell C. elegans killing assays pellets frozen and lyophilized, total RNA isolated using TRIzol Starter cultures of C. neoformans strains were prepared by growth reagent (Invitrogen) and cDNA generated using the SuperscriptIII in YPD at 30uC overnight with shaking. 10 mL overnight cultures First-Strand Synthesis System (Invitrogen). Primers for genes were spread onto both brain-heart infusion (BHI) (Becton PUT1 (CNAG02049.2), GAT1/ARE1 (CNAG00193.2), TAR1 Dickinson) and pigeon guano agar plates (35 mm), and incubated (CNAG04934.2), LAC1 (CNAG03465.2) and LAC2 (CNAG03464.2) at 25uC overnight. ,50 young adult C. elegans worms were then were designed to span exon-exon boundaries and tested to verify transferred from a lawn of E. coli OP50 on NGM to BHI and that they bind specifically to cDNA but not genomic DNA. pigeon guano medium-grown C. neoformans [34,61]. Plates were Quantitative real-time PCR (qRT-PCR) was performed using incubated at 25uC and worms examined for viability at 24 hr SYBR Green Supermix (Applied Biosystems) and an Applied intervals using a dissecting microscope, with worms that did not Biosystems 7900HT Fast Real Time PCR System with the respond to a touch with a platinum wire pick considered dead. following cycling conditions: denaturation at 95uC for 10 min, Each experimental condition was performed in triplicate. Survival followed by amplification and quantification in 45 cycles at 95uC was plotted against time, and P values were calculated by plotting for 15 sec and 60uC for 1 min, with melting curve profiling at a Kaplan-Meier survival curve and performing a log-rank (Mantel- 95uC for 2 min, 60uC for 15 sec and 95uC for 15 sec. Dissociation Cox) test using Graphpad Prism Version 5.0c. P values of ,0.05 analysis confirmed the amplification of a single PCR product for were considered statistically significant. each primer pair and an absence of primer dimer formation. Relative gene expression was quantified using SDS software 1.3.1 Murine inhalation model of cryptococcosis (Applied Biosystems) based on the 22DDCT method [57]. The For murine virulence assays, C. neoformans were used to infect 6 housekeeping actin-encoding gene ACT1 was used as a control for weeks old female BALB/c mice by nasal inhalation [62]. For every normalization. One-way analysis of variance was performed using tested strain, ten mice were each inoculated with a 50 mLdrop 5 the unpaired, two-tailed t test in GraphPad Prism Version 5.0c. containing 5610 cells. Mice were weighed before infection and P values of ,0.05 were considered statistically significant. daily thereafter; animals were sacrificed using CO2 inhalation once their body weight had decreased to 80% of the pre-infection weight. Nitrogen utilization, toxic analog sensitivity, Survival was plotted against time, and P values were calculated by melanization, high temperature growth and capsule plotting a Kaplan-Meier survival curve and performing a log-rank (Mantel-Cox) test using Graphpad Prism Version 5.0c. P values of assays ,0.05 were considered statistically significant. Starter C. neoformans cultures were prepared by growth in YPD at 30uC overnight with shaking, diluted to OD595 nm = 0.05 in Ethics statement water, then further diluted tenfold in series. Each diluted cell This study was carried out in strict accordance with the suspension was then spotted onto YPD or YNB, L-DOPA, recommendations in the Australian Code of Practice for the Care norepinephrine and caffeic acid medium supplemented with the and Use of Animals for Scientific Purposes by the National Health specified nitrogen source or toxic analog. Results were imaged and Medical Research Council. The protocol was approved by the after 2–3 days incubation at 30uC (nitrogen utilization and toxic Molecular Biosciences Animal Ethics Committee of The Univer- analog sensitivity assays), or both 30 and 37uC (melanization and sity of Queensland (AEC approval number: SCMB/008/11/UQ/ high temperature growth assays). For capsule assays, cells were NHMRC). Infection was performed under methoxyflurane scrapped off the YNB plates, stained with India ink (Becton anaesthesia, and all efforts were made to minimize suffering Dickinson), and visualized under a ZEISS Axioplan 2 epifluor- through adherence to the Guidelines to Promote the Wellbeing of escent/light microscope. Animals Used for Scientific Purposes as put forward by the National Health and Medical Research Council. Yeast two-hybrid assay Yeast two-hybrid experiments were conducted as described Results previously [20,58]. Briefly, full-length GAT1/ARE1 and TAR1 ORFs were amplified from H99 cDNA template using high fidelity Tar1 is not required for nitrogen source utilization PCR, and cloned into pCR2.1-TOPO to generate pIRL27 and Although the genome of VNI strain H99 of C. neoformans contains pIRL21, respectively. The inserted products were sequenced to multiple open reading frames encoding putative GATA factors, verify the absence of errors. cDNAs of first-third (nucleotides 1– only Gat1/Are1 regulates nitrogen catabolism [34]. An equivalent 1,284), second-third (nucleotides 1,285–2,562) and final-third nitrogen regulatory system that is dependent on a single positively

PLoS ONE | www.plosone.org 3 March 2012 | Volume 7 | Issue 3 | e32585 Nitrogen Metabolite Repression in Cryptococcus acting GATA factor also occurs in A. nidulans and N. crassa where glutamine and found that the tar1D mutant grew to the same extent as AreA and Nit2, respectively, have been extensively characterized wild-type, consistent with C. neoformans lacking nitrate reductase [2,3,8]. Like the ascomycete moulds, a predicted Nmr homolog needed to catabolize potassium chlorate (Figure 2A) [9]. Tar1 has been identified in C. neoformans but curiously, Jiang et al. The S. cerevisiae Dal80 and Gzf3, and A. nidulans AreB proteins are have suggested that Tar1 plays a negative role in nitrogen negatively acting GATA factors involved in nitrogen regulation metabolism as indicated by the tar1D mutants faster growth on [66,67]. In C. neoformans, six additional GATA factors have been potassium nitrate as compared to wild-type [38]. This observation is identified: Gat201 and Gat204 work in concert to prevent counterintuitive as the inability of C. neoformans to utilize nitrate is a phagocytosis by macrophages, Bwc2 regulates cell fusion and hyphal classic trait that has long been used as a diagnostic tool to identify development in response to blue light, Cir1 controls iron acquisition, this fungus; our recent bioinformatic and phenotypic analyses also while CNAG04263.2 and CNAG03401.2 have unknown functions supported this lack of nitrate utilization [34,63,64]. [34,68,69,70,71]. To determine if any of these GATA factors play a A set of deletion mutants lacking TAR1, GAT1/ARE1 or both negative role in nitrogen catabolism, we tested the gat201D, gat204D, was therefore created via homologous recombination in strain bwc2D,cir1D, CNAG04263.2D and CNAG03401.2D mutants sensitiv- H99 to investigate if Tar1 plays a regulatory role in nitrogen ity to thiourea in the presence of ammonium or glutamine (not utilization (Figure 1). All three generated mutants were viable and shown). The six GATA deletion mutants exhibited wild-type had a growth rate indistinguishable from wild-type on rich sensitivity, suggesting that negatively acting nitrogen regulatory undefined YPD medium. However, on YNB defined medium GATA factors are likely to be a unique evolutionary feature of the supplemented with a panel of different sole nitrogen sources ascomycetes. Consistent with this notion, all seven C. neoformans including ammonium, amino acids, purines, nitrate or nitrite, both GATA factors including Gat1/Are1, lack putative leucine zippers the gat1/are1D and double gat1/are1D tar1D mutants only had the that are present in the negative acting S. cerevisiae Dal80 and Gzf3, or ability to proliferate on proline. This result is consistent with A. nidulans AreB. Overall, our analysis thus far revealed that only Tar1 previous studies indicating that the gat1/are1D mutant is unable to plays a negative role in secondary nitrogen catabolism. utilize ammonium or glutamine but is able to consume proline, a marked difference in phenotype compared to well-characterized Tar1 modulation of nitrogen metabolite repression at the ascomycete GATA factor mutants [33,34]. On the other hand, the transcriptional level tar1D mutant displayed wild-type growth on all tested nitrogenous We sought to provide support for our sensitivity assays and prove compounds except nitrate and nitrite, which all strains were that Tar1 is implicated for the response to nitrogen metabolite unable to grow on. Together, our results indicate that unlike repression in a range of metabolic pathways. We therefore analysed Gat1/Are1, Tar1 is not required for nitrogen source utilization. the transcriptional regulation of the proline oxidase-encoding gene Additionally, the similar nitrogen utilization phenotypes displayed PUT1 using qRT-PCR on RNA extracted from the wild-type, tar1D, by the gat1/are1D and double gat1/are1D tar1D mutants suggest gat1/are1D and double gat1/are1D tar1D mutant strains grown in YNB that if Tar1 does indeed play a regulatory role in these nitrogen supplemented with proline or proline plus ammonium (Figure 2B). catabolic pathways, Gat1/Are1 functions downstream of Tar1. The proline catabolic pathway was chosen to be examined as this traditionally non-preferred nitrogen source supports the most robust Tar1 modulation of nitrogen metabolite repression at the growth in the gat1/are1D and double gat1/are1D tar1D mutants, thus phenotypic level enabling the isolation of RNA from a defined medium. Furthermore, Gat1/Are1 has been shown to play a role in nitrogen metabolite PUT1 has previously been shown to be sensitive to nitrogen metabolite repression, a regulatory mechanism that enables preferential repression, providing an ideal condition to test how Tar1 and Gat1/ utilization of readily assimilated nitrogen sources such as ammonium Are1 are affiliated to this regulatory control mechanism [34]. and glutamine [34]. We sought to determine if Tar1 is also As expected, high levels of PUT1 expression were observed functionally associated with this metabolic response. We note that when the wild-type strain was cultured in proline as the sole Jiang et al. have previously assessed the existence of inappropriate nitrogen source, and this upregulation was largely abolished (,30- derepression of secondary nitrogen catabolism in the tar1D mutant by fold difference) with the simultaneous presence of ammonium. In growing the strain solely on potassium chlorate (a toxic analog of the tar1D mutant, intermediate levels of PUT1 mRNA were potassium nitrate) in the absence of a preferred nitrogen source [38]. observed in the presence of proline, and this upregulation was only However, such sensitivity assays used to determine constitutive partially reduced (,5-fold difference) when ammonium was expression of non-preferred nitrogen structural genes that are simultaneously present. While the 2–3-fold more PUT1 transcript subjected to nitrogen metabolite repression could only be interpreted produced by the tar1D mutant relative to wild-type when grown in if a repressing nitrogen source is simultaneously present [21,23]. proline plus ammonium was anticipated, the 2–3-fold less PUT1 We therefore tested the tar1D mutants sensitivity to thiourea (a expression of the tar1D mutant compared to wild-type in the sole toxic analog of urea) in the presence of ammonium or glutamine presence of proline was surprising (WT nitrogen metabolite (Figure 2A). Like ascomycete nmrD mutants that exhibit partially repression sensitivity vs tar1D nitrogen metabolite repression derepressed phenotypes due to derepression of the urea permease, the sensitivity, P = 0.0195). The former indicates that PUT1 transcrip- tar1D mutant displayed a slight increase in sensitivity relative to wild- tion in the tar1D mutant was not completely repressed in the type while the negative control gat1/are1D and double gat1/are1D presence of a preferred nitrogen source, and reiterates the fact that tar1D mutants were unable to proliferate due to their inability to Tar1 prevents the expression of secondary nitrogen structural utilizethesenitrogensources[65].Thisresultsuggeststhatthetar1D genes under repressing conditions. The latter, intriguingly, mutant is metabolizing thiourea to a greater extent than wild-type, suggests that Tar1 also plays a positive role in the activation of despite the simultaneous presence of preferred nitrogen sources. secondary nitrogen structural genes under non-repressing condi- Complementation of the tar1D mutant subsequently restored the tions, despite the fact that the tar1D mutant exhibited indistin- wild-type phenotype (Figure S1). Together, these results indicate that guishable phenotypic growth rate relative to wild-type on various Tar1 plays a negative role in the regulation of genes involved in nitrogen sources (Figure 1). Tar1 therefore contributes to the utilization of more complex nitrogen sources. We also tested expression of nitrogen catabolism associated genes, however, it sensitivity to potassium chlorate in the presence of ammonium or does not overcome the requirement for Gat1/Are1.

PLoS ONE | www.plosone.org 4 March 2012 | Volume 7 | Issue 3 | e32585 Nitrogen Metabolite Repression in Cryptococcus

Figure 1. Tar1 is not required for nitrogen source utilization. Tenfold spot dilution assays for nitrogen utilization showed that the tar1D mutant exhibited wild-type growth on YNB supplemented with ammonium, glutamine, glutamate, asparagine, proline, glycine, phenylalanine, leucine, uric acid, urea and creatinine (10 mM each nitrogen source). In contrast, both the gat1/are1D and double gat1/are1D tar1D mutants were only able to proliferate on proline as the sole nitrogen source. All tested strains were unable to grow on 7% potassium nitrate or sodium nitrite. doi:10.1371/journal.pone.0032585.g001

The expression of GAT1/ARE1, but not TAR1, is regulated were detected in the gat1/are1D and tar1D mutants, respectively, in response to nitrogen availability validating our diagnostic PCR and Southern blot analyses for In A. nidulans, NmrA modulation of AreA activity is finely tuned; confirmation of gene deletion. Consistent with nitrogen indepen- this well-coordinated control mechanism begins at the transcrip- dence of TAR1 expression, the gat1/are1D mutant produced wild- tional level as expression of areA and nmrA are inversely regulated type levels of the TAR1 transcript. On the other hand, the tar1D in response to the quality of the nitrogen source available [37]. To mutant expressed ,3-fold less GAT1/ARE1 mRNA compared to determine if a similar phenomenon occurs in C. neoformans,we wild-type (WT vs tar1D, P = 0.0008). This result provides a analysed the transcriptional regulation of both GAT1/ARE1 and plausible explanation for why the tar1D mutant transcribed less TAR1 using qRT-PCR on RNA extracted from wild-type H99 PUT1 mRNA relative to wild-type in the sole presence of proline grown in YNB supplemented with different nitrogen sources (Figure 2B), and indicates that Tar1 positively regulates GAT1/ (Figure 3). During growth in ammonium, a low level of GAT1/ ARE1 expression under non-repressing conditions. ARE1 mRNA was observed, while an elevated level (,7-fold) of To provide support for the model that Tar1 plays a dual function GAT1/ARE1 transcript was present when grown in proline (Pro vs in positive and negative regulation of nitrogen catabolism, we NH4, P = 0.0003). A marked increase (.10-fold) in GAT1/ARE1 analysed the transcriptional regulation of GAT1/ARE1 using qRT- transcript level was seen when cells were starved for nitrogen (N PCR on RNA extracted from the wild-type and tar1D mutant free vs NH4, P = 0.003). In stark contrast to the expression profile strains grown in YNB supplemented with ammonium (Figure 4B). of GAT1/ARE1, the level of TAR1 transcript was relatively As expected, the wild-type strain produced low levels of GAT1/ unaltered when cultured under these different nitrogen conditions. ARE1 transcripts under such nitrogen sufficient conditions, while Overall, our qRT-PCR data indicates that GAT1/ARE1 expres- the tar1D mutant transcribed ,3-fold more GAT1/ARE1 mRNA sion is transcriptionally induced in response to nitrogen limitation, relative to wild-type (WT vs tar1D, P = 0.0481). This result indicates whereas TAR1 transcription is unaffected by nitrogen availability. that Tar1 negatively regulates GAT1/ARE1 expression under repressing conditions. Collectively, Tar1 can function to elicit both Tar1 plays a dual function in positive and negative activation and repression of GAT1/ARE1 autoregulation according regulation of GAT1/ARE1 expression according to the to the quality of the nitrogen source present. nitrogen source present Given that Gat1/Are1 is involved in the utilization of most Gat1/Are1 lacks the highly conserved extreme nitrogen sources, we sought to determine if Tar1 regulates GAT1/ C-terminus domain of A. nidulans AreA and N. crassa ARE1 expression, and vice versa. We first analysed the transcrip- Nit2 known to be involved in Nmr recognition tional regulation of both GAT1/ARE1 and TAR1 using qRT-PCR Our phenotypic and transcriptional data suggest that the on RNA extracted from the wild-type, tar1D and gat1/are1D molecular mechanism of action of Tar1 may operate in a fashion mutant strains grown in YNB supplemented with proline similar to ascomycete Nmr homologs. In A. nidulans and N. crassa, (Figure 4A). As expected, no GAT1/ARE1 or TAR1 transcripts certain mutations of the GATA DNA-binding domain or extreme

PLoS ONE | www.plosone.org 5 March 2012 | Volume 7 | Issue 3 | e32585 Nitrogen Metabolite Repression in Cryptococcus

Figure 2. Tar1 plays a role in modulating nitrogen metabolite repression. (A) Tenfold spot dilution assays for nitrogen utilization showed that the tar1D mutant exhibited slight sensitivity to 5 mM thiourea when compared to wild-type in the simultaneous presence of 10 mM ammonium or glutamine. (B) cDNA from wild-type H99 and the mutant strains grown in YNB supplemented with proline or proline plus ammonium (10 mM each nitrogen source) were amplified via qRT-PCR using primers against the proline oxidase-encoding gene PUT1 and the control gene ACT1. In the wild- type strain, PUT1 expression was significantly increased in the presence of proline but upregulation was largely abolished when ammonium was also present. In the tar1D mutant, PUT1 transcription was intermediate in the presence of proline and upregulation was only partially reduced when ammonium was simultaneously present. Nitrogen metabolite repression sensitivity of PUT1 in the wild-type strain statistically differs from that in the tar1D mutant (* denotes P,0.05). In the gat1/are1D and double gat1/are1D tar1D mutants, PUT1 expression was low under both tested growth conditions. Error bars represent standard errors across three biological replicates. doi:10.1371/journal.pone.0032585.g002

PLoS ONE | www.plosone.org 6 March 2012 | Volume 7 | Issue 3 | e32585 Nitrogen Metabolite Repression in Cryptococcus

Gat1/Are1, AreA and Nit2 [2,3]. In summary, one of the two regions of ascomycete GATA factors known to be important for interaction with Nmr proteins is absent in Gat1/Are1: the Cryptococcus GATA DNA-binding domain is conserved but the extreme C-terminus has diverged.

Tar1 lacks the conserved C-terminus domain of A. nidulans NmrA known to be important for interaction with AreA We next analysed the sequences of the Nmr homologs of A. nidulans, N. crassa and C. neoformans.InA. nidulans, the conserved a- helix at the C-terminus of NmrA is also known to be critical for interaction with AreA [40,41]. Given that the sequence of N. crassa Nmr1 is highly conserved with A. nidulans NmrA, it is likely that both these proteins share similar structural features [21,43]. In contrast, global alignment of C. neoformans Tar1 revealed only moderate overall sequence conservation to both NmrA and Nmr1 (19 and 16% identity, respectively) (Figure S3). Tar1 harbours the residues GlyXXGlyXXGly that are predicted to form a Rossmann fold in the N-terminus, while NmrA and Nmr1 contain a slightly diverged canonical Rossmann motif AsnXXGlyXXAla [38,41,73]. Howev- er, Tar1 lacks the long C-terminus region of NmrA and Nmr1 that is predicted to play a role in interaction with GATA factors. Based on bioinformatics conducted on these GATA factors and Nmr proteins, it was not apparent whether Tar1 is likely to physically interact with and inhibit the function of Gat1/Are1. Figure 3. The expression of GAT1/ARE1,butnotTAR1,is regulated in response to nitrogen availability. cDNA from wild- Interaction between Tar1 and Gat1/Are1 was not type H99 grown in YNB supplemented with ammonium, proline (10 mM each nitrogen source) or no nitrogen source were amplified via detectable in vivo qRT-PCR using primers against GAT1/ARE1, TAR1 and the control gene The direct interaction of N. crassa Nit2 and Nmr1 was originally ACT1. In the presence of a preferred nitrogen source (ammonium), the shown via yeast two-hybrid experiments. The DNA-binding expression of GAT1/ARE1 was low but in the presence of a traditionally domain or extreme C-terminus of Nit2 fused to the Gal4 non-preferred nitrogen source (proline), or under nitrogen starvation conditions, GAT1/ARE1 transcription was significantly upregulated activation domain (AD) interacted with an Nmr1-Gal4 DNA- (** denotes P,0.01, *** denotes P,0.001). In contrast, TAR1 expression binding domain (BD) fusion protein, with stronger interaction was relatively unaltered under these different nitrogen growth detected when both the DNA-binding domain and C-terminus conditions. Error bars represent standard errors across three biological regions of Nit2 were present [20]. To investigate if a similar replicates. situation occurs in C. neoformans, we also performed a yeast two- doi:10.1371/journal.pone.0032585.g003 hybrid assay to test if protein-protein interaction occurs between Gat1/Are1 and Tar1 (Table S4). We fused the first-third C-terminus of AreA and Nit2 lead to partially derepressed (nucleotides 1–1,284), second-third (nucleotides 1,285–2,562; phenotypes under nitrogen sufficient conditions similar to that contains the first conserved motif of unknown function) and observed for nmrD mutants, suggesting that these regions are final-third (nucleotides 2,563–3,834; contains the second con- critical for interaction with Nmr proteins and modulation of served motif of unknown function, GATA DNA-binding domain GATA factor transcriptional activity [18,23]. and C-terminus) GAT1/ARE1 fragments, as well as full-length However, global alignment of C. neoformans Gat1/Are1 showed GAT1/ARE1 and TAR1 cDNA, to both the GAL4 AD and BD. little overall sequence conservation to AreA or Nit2 (15 and 16% Every possible combination of the clones were then expressed in S. identity, respectively) (Figure S2). Importantly, the extreme C- cerevisiae AH109 in which the GAL promoter regulates ADE2, HIS3 terminus of AreA and Nit2 known to be critical for modulating and lacZ reporter genes. If any protein-protein interaction did activities by the co-repressors NmrA and Nmr1, respectively, is exist, it would likely be the final-third Gat1/Are1 fragment and/or absent in Gat1/Are1 (Figure 5A and E) [18,21,23]. In fact, the last full-length Gat1/Are1 (both contains the conserved GATA DNA- nine and 12 C-terminus residues of A. nidulans AreA are identical binding domain) that may interact with Tar1. to N. crassa Nit2 and Penicillium chrysogenum GATA factor Nre, In AH109 strains expressing the first-third fragment or full- respectively (Figure 5E and not shown) [2,23,72]. Nonetheless, length Gat1/Are1-Gal4 BD fusion protein, auto-activation of the three notable blocks of conservation do exist between Gat1/Are1 bait occurred indicating that an activation domain is present and AreA or Nit2: two regions of unknown function (in AreA and within the nucleotides 1–1,284 of GAT1/ARE1. However, in Nit2, these two regions lie near the N-terminus while in Gat1/ AH109 strains expressing any of the three individual fragments or Are1, they lie further towards the middle or C-terminus region of full-length Gat1/Are1-Gal4 AD fusion protein and Tar1-Gal4 BD the protein) (Figure 5A, B and C), and the GATA DNA-binding fusion protein, cells failed to grow in the absence of adenine and domain consisting of a zinc finger motif followed by an adjacent histidine, and did not produce b-galactosidase activity. Our basic region (Figure 5A and D). The zinc finger domain has been bioinformatics coupled with yeast two-hybrid data therefore shown to play a role in the DNA-binding activity of both AreA and suggest that no protein-protein interaction is likely occurring Nit2, and the same scenario is likely occurring in C. neoformans between Gat1/Are1 and Tar1 when expressed in the nucleus of given the high level of conservation seen among the zinc fingers of the hemiascomycete S. cerevisiae. Additionally, our two-hybrid

PLoS ONE | www.plosone.org 7 March 2012 | Volume 7 | Issue 3 | e32585 Nitrogen Metabolite Repression in Cryptococcus

bars represent standard errors across three biological replicates. (B) cDNA from wild-type H99 and tar1D mutant strains grown in YNB supplemented with 10 mM ammonium were amplified via qRT-PCR using primers against GAT1/ARE1 and ACT1. The expression of GAT1/ ARE1 was significantly higher in the tar1D mutant compared to wild- type (* denotes P,0.05). Error bars represent standard errors across three biological replicates. doi:10.1371/journal.pone.0032585.g004

analysis provides no indication of Gat1/Are1 or Tar1 homo- dimerization.

Gat1/Are1 negatively regulates laccase-encoding LAC1 expression at 37uC Jiang et al. originally identified Tar1 through its negative role in antioxidant melanin production at 37uC on norepinephrine- containing asparagine medium, which eventually led to their finding that Tar1 regulates laccase-encoding LAC1 expression [38]. The two paralogs LAC1 and LAC2 encode laccases that catalyse the formation of melanin in C. neoformans, with Lac1 being the main enzyme responsible [45,74,75]. Interestingly, our group has also observed that Gat1/Are1 negatively regulates melanin production at 37uConL-DOPA-containing proline medium [34]. We therefore investigated the genetic epistasis between TAR1 and GAT1/ARE1- regulated production of melanin and expression of laccases. Given that the gat1/are1D and double gat1/are1D tar1D mutants are unable to utilize asparagine as a nitrogen source, we analysed melanin production of the deletion mutants on L-DOPA (proline) agar grown at both 30 and 37uC (Figure 6A). All strains melanized robustly when cultivated at 30uC. However, at 37uC, the gat1/ are1D and double gat1/are1D tar1D mutants produced more melanin than the wild-type and tar1D mutant strains, reiterating the observation that the gat1/are1 mutation is dominant over the tar1 mutation but disagreeing with those seen by Jiang et al. [38]. We then analysed the transcriptional regulation of LAC1 and LAC2 using qRT-PCR on RNA extracted from the wild-type and mutant strains grown in L-DOPA (proline) at both 30 and 37uC (Figure 6B). As previously reported, transcription of LAC1 was severely decreased (50–100-fold) at 37uC compared to 30uC [76]. Consistent with our melanin phenotypic plate assays, both the gat1/are1D and double gat1/are1D tar1D mutants transcribed more (,2-fold) LAC1 mRNA as compared to the wild-type or tar1D mutant strains at 37uC (WT vs gat1/are1D, P = 0.0003; WT vs double gat1/are1D tar1D, P = 0.0131). No significant difference in the expression of LAC1 at 30uCorLAC2 at 30 and 37uC was observed among the tested strains. Altogether, our data indicates that Gat1/Are1, but not Tar1, represses melanin production and LAC1 expression at human body temperature when L-DOPA is used as the laccase substrate. The different substrate used for melanin production may have resulted in conflicting observations made by Jiang et al. and our group [38]. We therefore tested melanin production of wild-type H99 and three independent H99 tar1D mutants, as well as the isogenic wild-type MATa strain KN99a and six independent KN99a tar1D mutants, at both 30 and 37uC on additional laccase substrates: norepinephrine and caffeic acid, supplemented with asparagine as the nitrogen source (Figure S4). Under all tested growth conditions, no visual difference in melanin production was Figure 4. Tar1 plays both positive and negative roles in observed between the wild-type and tar1D mutant strains. regulating GAT1/ARE1 transcription according to the nitrogen Therefore, in our hands, we are unable to reproduce the increased source available. (A) cDNA from wild-type H99, gat1/are1D and tar1D melanin production and LAC1 expression phenotypes of the tar1D mutant strains grown in YNB supplemented with 10 mM proline were amplified via qRT-PCR using primers against GAT1/ARE1, TAR1 and the mutant as described by Jiang et al. [38]. We speculate that the control gene ACT1. The expression of GAT1/ARE1 was significantly lower passage of H99 during subculture between laboratories may have in the tar1D mutant compared to wild-type (*** denotes P,0.001). Error possibly caused the phenotypic variation. In support of this notion,

PLoS ONE | www.plosone.org 8 March 2012 | Volume 7 | Issue 3 | e32585 Nitrogen Metabolite Repression in Cryptococcus

Figure 5. Representative domain architecture and domain sequences of A. nidulans AreA, N. crassa Nit2 and C. neoformans Gat1/Are1. (A) The two conserved regions of unknown function are represented by red ovals and orange diamonds, while the GATA DNA-binding domain is represented by blue squares. The extreme C-terminus domain of AreA and Nit2 (dark green rectangles) is highly conserved but that of Gat1/Are1 has diverged (light green rectangle). Amino acid sequences of all four domains are shown in (B, C, D and E). In addition, the Gat1/Are1 domain architecture is marked with dotted lines to indicate the boundary of the Gat1/Are1 fragments that were tested for protein-protein interaction with Tar1 (see yeast two-hybrid assay section). The universally conserved cysteines involved in zinc ion chelation of the zinc finger are highlighted in yellow. doi:10.1371/journal.pone.0032585.g005

Morrow et al. have recently reported that various H99 subcultures investigated the role of Tar1 in pathogenesis by performing both in have differing melanization abilities [77]. Additionally, the vitro and in vivo virulence assays. First, we conducted C. elegans indistinguishable growth rate of the tar1D mutants in comparison killing assays using two different media: the standard BHI medium to their wild-type H99 or KN99a counterparts on YPD medium at for nematode killing experiments, as well as pigeon guano medium 37uC suggests that Tar1 does not play a role in high temperature to mimic the C. neoformans ecological niche (Figure 7A). Under both growth. growth conditions, killing of C. elegans by the tar1D [LT50 (time for Expanding our analysis on virulence factor expression, we half of the worms to die) = 5 and 3 days, for BHI and pigeon guano examined the polysaccharide capsule that has an antiphagocytic medium, respectively] and double gat1/are1D tar1D mutants function [78]. We grew the wild-type and tar1D mutant strains on (LT50 = 5 and 4 days, for BHI and pigeon guano medium, YNB supplemented with creatinine, uric acid or urea that are respectively) was not significantly different to that observed for known inducers of capsule formation, and found that both the wild-type (LT50 = 6 and 3 days, for BHI and pigeon guano wild-type and tar1D cells possess equally large capsule on each of medium, respectively). The undiminished pathogenicity of the these individual medium (not shown) [34]. Given that nitrogen double gat1/are1D tar1D mutant is consistent with our previous metabolite repression plays a role in the regulation of capsule work showing that Gat1/Are1 does not affect killing of C. elegans production, we also examined the capsule of both strains when [34]. grown on the same capsule-inducing media but with the Since the C. elegans virulence assays were conducted at 25uC, we simultaneous presence of ammonium (not shown) [34]. Under questioned if Tar1 is required for infection of a host at mammalian these growth conditions, the wild-type cells should possess body temperature given that the TAR1 promoter contains putative completely repressed (small) capsule since ammonium is a poor heat shock sequence elements [38]. To test this hypothesis, we inducer of capsule formation, while we were predicting that the performed a murine inhalation model of cryptococcosis (Figure 7B). tar1D cells would possess partially repressed (intermediate) capsule Mice infected with the tar1D mutant succumbed to infection slightly due to inappropriate depression of secondary nitrogen (creatinine, faster (between 16 and 19 days post infection, median survival of 18 uric acid, urea) catabolism [34]. Unexpectedly, the capsule size days) than mice infected with the wild-type strain (between 18 and was equally small and indistinguishable between the wild-type and 24 days post infection, median survival of 21.5 days) (WT vs tar1D, tar1D mutant strains under such nitrogen metabolite repression P = 0.0021). It should be noted that mice infected with the conditions. The subtle role of Tar1 in modulating nitrogen complemented tar1D+TAR1 strain did not progress to morbidity metabolite repression may explain why the tar1D mutants capsule at the same rate as wild-type (between 17 and 21 days post infection, was not detectably derepressed, as can be seen in the toxic median survival of 18 days). However, the difficulty associated with thiourea assays where the tar1D mutant only exhibited slight the restoration of wild-type phenotype of the tar1D mutant in certain sensitivity relative to wild-type (Figure 2A). Taken together, Tar1 physiological aspect has previously been reported [38]. Alternative- does not appear to affect well-established virulence factors ly, the role of Tar1 in virulence may be dependent on its expression levels, which could have been adversely affected by integration of including, but not limited to, melanin production, growth at TAR1 into a non-native locus. Mice infected with the double gat1/ human body temperature and capsule biosynthesis. are1D tar1D mutant also succumbed to infection slightly faster than mice infected with the wild-type strain, consistent with our previous Tar1 is not required for killing of C. elegans but modestly work demonstrating that Gat1/Are1 modestly represses virulence represses virulence in a murine inhalation model of during murine infection (between 15 and 22 days post infection, cryptococcosis median survival of 19 days) (WT vs gat1/are1D tar1D, P = 0.0342) Our overarching interest in C. neoformans lies in better [34]. Together, these results suggest that both the tar1D and double understanding its pathogenicity in an animal host. We therefore gat1/are1D tar1D mutants are slightly more virulent than wild-type

PLoS ONE | www.plosone.org 9 March 2012 | Volume 7 | Issue 3 | e32585 Nitrogen Metabolite Repression in Cryptococcus

Figure 6. Gat1/Are1 negatively regulates melanin production and LAC1 expression at 376C. (A) Tenfold spot dilution assays for melanization showed that the gat1/are1D and double gat1/are1D tar1D mutants produce more melanin than the wild-type and tar1D mutant strains on L-DOPA medium supplemented with 10 mM proline at 37uC. (B) cDNA from wild-type H99 and the mutant strains grown in L-DOPA (10 mM proline) at 30 and 37uC were amplified via qRT-PCR using primers against the laccase-encoding genes LAC1 and LAC2, and the control gene ACT1. LAC1 expression was significantly higher in the gat1/are1D and double gat1/are1D tar1D mutants compared to wild-type or the tar1D mutant when strains were grown at 37uC (* denotes P,0.05, *** denotes P,0.001). Error bars represent standard errors across three biological replicates. doi:10.1371/journal.pone.0032585.g006 during in vivo infection. Hence, in addition to fine-tuning regulation interaction with the prion-forming glutathione S-transferase of nitrogen catabolism, Tar1 also modulates pathogenesis in a Ure2 [14,15,16,17]. As in the system of the filamentous mammalian host. ascomycete A. nidulans, there is no genetic or molecular evidence for the existence of an Ure2 homolog in C. neoformans [82,83]. Discussion Although we have identified a putative protein sequence (CNAG04110.2) in the H99 genome which contains a glutathione Nitrogen starvation stimulates the initiation of the infection S-transferase domain that shows weak overall sequence similarity cycle of the ubiquitous C. neoformans, whereby sexual reproduction to Ure2, CNAG04110.2 lacks several features unique to Ure2 occurs, leading to the production of potentially infectious including an N-terminus extension and loop region [84,85]. basidiospores [69,79,80]. This mating process is controlled by In A. nidulans and N. crassa, the function of the GATA factors multiple signalling cascades including the high affinity ammonium AreA and Nit2 are negatively regulated by interaction with Nmr permease Amt2, which is in turn regulated by the GATA factor proteins that are unrelated to glutathione S-transferases Gat1/Are1 [34,81]. In addition to regulating nitrogen catabolism, [12,18,20,21,22,23,82,83]. Recently, the group of Jiang et al. Gat1/Are1 also impacts multiple virulence attributes such as identified the C. neoformans potential Nmr homolog Tar1 during melanin production, high temperature growth and capsule their search for gene disruption mutants that hypermelanize at biosynthesis [33,34]. Therefore, the underlying mechanism 37uC [38]. In this study, we further characterized Tar1 function regulating Gat1/Are1 activity in response to nitrogen availability from a different perspective based on our interest in regulation of is important for the understanding of the C. neoformans life cycle in nitrogen metabolism. Our findings indicate that Tar1 possesses the environment and successful proliferation during infection of a several unique regulatory and functional features in comparison to host. A. nidulans NmrA and N. crassa Nmr1. In the model hemiascomycete S. cerevisiae, the function of the In A. nidulans, nmrA expression is low during nitrogen limitation GATA factors Gln3 and Gat1 are negatively regulated by but increases during nitrogen sufficiency [37]. The bZIP

PLoS ONE | www.plosone.org 10 March 2012 | Volume 7 | Issue 3 | e32585 Nitrogen Metabolite Repression in Cryptococcus

Figure 7. The tar1D mutant kills C. elegans as efficiently as wild-type, but exhibits modestly enhanced virulence in a murine host. (A) C. elegans infection: ,50 nematode worms were transferred to a lawn of wild-type H99, tar1D, tar1D+TAR1 or double gat1/are1D tar1D cells as the sole food source on both BHI and 1% pigeon guano medium, and survival was monitored at 24 hr intervals. There was no observable difference in C. elegans killing by all four strains on both media. (B) Mus musculus infection: 10 mice were each intranasally infected with either 56105 cells of wild- type H99, tar1D, tar1D+TAR1 or double gat1/are1D tar1D strains, and survival was monitored daily. Mice infected with the tar1D, tar1D+TAR1 and double gat1/are1D tar1D strains progress to morbidity slightly more rapidly than mice infected with the wild-type strain. doi:10.1371/journal.pone.0032585.g007

PLoS ONE | www.plosone.org 11 March 2012 | Volume 7 | Issue 3 | e32585 Nitrogen Metabolite Repression in Cryptococcus transcription factor MeaB activates the expression of nmrA, and Such research into nitrogen regulation is of immense interest as MeaB homologs are found in all the available filamentous nitrogen source utilization influences key aspects of fungal biology ascomycetes genome databases [37]. In contrast, TAR1 transcrip- including development, secondary metabolite production and tion is unaffected by nitrogen availability and our bioinformatic pathogenesis. In the context of virulence, nitrogen regulated analyses indicate that the C. neoformans H99 genome does not pathogenesis has been documented in clinically prevalent human encode an obvious homolog of MeaB. In addition, the promoter pathogens such as C. albicans and A. fumigatus [28,29,30]. Our region of TAR1 lacks the conserved element, TTGCACCAT; in analysis of TAR1 and GAT1/ARE1 reiterated this dogma; both vitro studies have shown that A. nidulans MeaB binds to genes are suppressors of C. neoformans virulence. A parsimonious TTGCACCAT and this binding site is also present in the explanation as to why no opposing regulatory effects on promoters of NmrA homologs in other filamentous ascomycetes pathogenesis were observed between the tar1D and gat1/are1D or [37]. Instead, Jiang et al. have reported that TAR1 expression is double gat1/are1D tar1D mutants may relate to the scarce nutrient upregulated at high temperature (37uC), consistent with the fact availability during in vivo infection, a condition in which Tar1 is that its promoter contains three TTC/GAA repeats that are predicted to positively regulate GAT1/ARE1 activity [34]. putative binding sites for the heat shock factor [38]. Nonetheless, we cannot rule out the possibility that Tar1 is not However, Tar1 does contain the characteristic NAD/NADH exclusively involved in regulation of nitrogen catabolism but may dinucleotide binding motif, GlyXXGlyXXGly, found in members of also control targets involved in a broader spectrum of metabolic the short-chain dehydrogenase/reductase family such as the processes. The fact that the TAR1 promoter contains putative heat negatively acting Gal80 of S. cerevisiae, which acts to block the shock elements supports this notion [38]. Using whole transcrip- activation domain of Gal4 required for transcription of galactose- tome sequencing, we are currently attempting to identify novel inducible genes [42,86,87,88]. Like the negative transcription virulence and stress adaptation associated genes that are regulated regulators A. nidulans NmrA and N. crassa Nmr1, Tar1 retains the by both Tar1 and Gat1/Are1 during infection of a mammalian evolutionary conserved function of Nmr proteins of preventing host. Indeed, Kronstad et al. have recently highlighted transcrip- activation of secondary nitrogen catabolism when preferred nitrog- tional profiling and genetic studies as invaluable research tools, enous compounds are present. However, the role of Tar1 in this with the means to provide insights into C. neoformans adaptation to regulatory aspect is quite subtle, indicative that additional factors are key features within a mammalian host environment including likely operating to modulate nitrogen metabolite repression or to nitrogen availability [92]. repress Gat1/Are1 activity in C. neoformans.Forexample,inA. In summary, there is a pressing need to combat infection from nidulans, the activity of AreA is also controlled through autogenous the killer fungus C. neoformans that is responsible for .625,000 regulation and differential transcript stability [23,24,25]. deaths annually in predominantly AIDS patients [93]. Our study Intriguingly, we also discovered a novel role of Tar1 in has provided deeper insights into the molecular mechanisms that positively regulating the expression of the proline oxidase- wire the Cryptococcus global nitrogen regulatory circuit. Specifically, encoding gene PUT1 under non-repressing conditions. The dual we provide an indication of how Tar1 controls Gat1/Are1 activity function of Tar1 in fine-tuning nitrogen catabolism appears to be in response to changes in nitrogen availability, as would occur mediated through modulation of GAT1/ARE1 transcription in when C. neoformans leaves its purine rich ecological niche of pigeon response to the quality of the nitrogen source available (Figure S5). guano to infect a human host which in comparison is nutrient Given that Tar1 is not required for nitrogen source utilization, we limiting [94]. Understanding the intricate details of how Gat1/ speculate that only basal-to-intermediate levels of GAT1/ARE1 Are1 function is pivotal, as candidate genes in several nitrogen transcripts are needed to efficiently activate secondary nitrogen catabolic pathways represent potential drug targets for therapeutic catabolism under non-repressing conditions. The exact mecha- intervention in this important pathogen of humans. nism by which Tar1 regulates Gat1/Are1 activity remains unclear as no physical interaction could be detected amongst these two Supporting Information proteins. Since Tar1 does not appear to have a DNA-binding Figure S1 Complementation of toxic analog sensitivity domain, this protein may possibly control the activity of an phenotype to wild-type levels upon the re-introduction of unknown co-factor(s)/accessory transcription apparatus that in TAR1 into the tar1D mutant. Tenfold spot dilution assays for turn influences GAT1/ARE1 autoregulation in response to nitrogen utilization showed that the tar1D+TAR1 strain exhibited nitrogen availability. Further work will be required to dissect the wild-type growth on 5 mM thiourea plus 10 mM ammonium. nature and origin of intracellular signals that govern the activity of Gat1/Are1. (DOC) Notwithstanding, although a plethora of information about Figure S2 ClustalW multiple sequence alignment of A. nitrogen metabolism is known in A. nidulans and N. crassa, studies nidulans AreA (XP_681936.1), N. crassa Nit2 (P19212.2) into the regulation of GATA factor and Nmr protein activity have and C. neoformans Gat1/Are1 (CNAG00193.2). Identical had their fair share of controversies. For example, while Lamb et amino acid residues are shaded dark grey while similar residues al. have previously suggested that the N-terminus is required for are shaded light grey. Gat1/Are1 shows little overall sequence modulating derepression activity of A. nidulans AreA, Caddick and conservation to AreA and Nit2 at the entire protein level. Arst later disagreed with this theory [89,90]. In more recent times, However, three notable blocks of conversation do exist between Lamb et al. demonstrated that the extreme nine C-terminus Gat1/Are1 and AreA or Nit2: two regions of unknown function residues of AreA that are highly conserved across a range of (boxed in red and orange) and the GATA DNA-binding domain filamentous ascomycete homologs does not affect the affinity for (boxed in blue). The second conserved motif of unknown function NmrA binding, contradicting the findings of Pan et al. and Xiao et (boxed in orange) was identified through Pustell dot plot protein al. [18,20,40]. Most recently, Wagner et al. asserted that nmrA matrix. The extreme C-terminus domain of AreA and Nit2 (boxed expression in A. nidulans is not dependent on MeaB, an observation in green) is highly conserved but that of Gat1/Are1 has diverged. that contrasts those of Wong et al. [37,91]. Hence, the mechanism Enlarged versions of all four domains are shown in Figure 5B, C, of action governing the activity of GATA factors and Nmr proteins D and E. in model ascomycetes, too, remain elusive. (DOC)

PLoS ONE | www.plosone.org 12 March 2012 | Volume 7 | Issue 3 | e32585 Nitrogen Metabolite Repression in Cryptococcus

Figure S3 ClustalW multiple sequence alignment of A. Table S1 Fungal strains used in this study. nidulans NmrA (AAC39442.1), N. crassa Nmr1 (DOC) (P23762.2) and C. neoformans Tar1 (CNAG04934.2). Table S2 Primers used in this study. Identical amino acid residues are shaded dark grey while similar residues are shaded light grey. Tar1 shows moderate overall (DOC) sequence conservation to NmrA and Nmr1. The predicted Table S3 Plasmids used in this study. Rossmann fold motif at the N-terminus is boxed in red. The long (DOC) C-terminus region of NmrA and Nmr1 is absent in Tar1. (DOC) Table S4 Interaction between Tar1 and Gat1/Are1 could not be detected in a yeast two-hybrid assay. S. Figure S4 Multiple independently generated tar1D mu- cerevisiae AH109 was co-transformed with both the bait and prey tants all produced equal amount of pigment melanin in constructs, and selected for growth on double dropout medium comparison to their wild-type H99 or KN99a counter- (-Leu -Trp). Interaction (denoted by + symbol) was assessed by parts. The tenfold spot dilution assays for melanization on L- growth on quadruple dropout medium (-Leu -Trp -His -Ade), and DOPA, norepinephrine and caffeic acid agar (supplemented with b-galactosidase activity. 2 symbol denotes no protein-protein 10 mM asparagine as the nitrogen source) were conducted at both interaction. A.A. denotes auto-activation. N.T. denotes not tested. 30 and 37uC. (DOC) (DOC) Figure S5 Scheme representing the dual roles of Tar1 in Acknowledgments modulating GAT1/ARE1 transcription that in turn We thank Bob Simpson for technical advice with qRT-PCR, Joseph influences PUT1 expression according to the nitrogen Heitman for the provision of the H99 and KN99a strains as well as the source available. In the presence of the traditionally non- plasmid pCH233, Jim Kronstad for the provision of the H99 cir1D strain, preferred proline, the Nmr homolog Tar1 positively regulates the and Gary Newell for the provision of the pigeon guano. transcription of the GATA factor-encoding gene GAT1/ARE1 that is required for induction of the proline oxidase-encoding gene Author Contributions PUT1. In the presence of the preferred ammonium, Tar1 negatively regulates GAT1/ARE1 transcription leading to reduce Conceived and designed the experiments: IRL JF. Performed the levels of PUT1 expression. experiments: IRL JL KO CM. Analyzed the data: IRL JL JF. Contributed (DOC) reagents/materials/analysis tools: JF. Wrote the paper: IRL JF.

References 1. Perfect JR, Casadevall A (2006) Fungal Molecular Pathogenesis: What Can It 16. Cooper TG (2002) Transmitting the signal of excess nitrogen in Saccharomyces Do and Why Do We Need It? In: Heitman J, Filler SG, Edwards JE, cerevisiae from the Tor proteins to the GATA factors: connecting the dots. FEMS Mitchell AP, eds. Molecular Principles of Fungal Pathogenesis. Washington, Microbiol Rev 26: 223–238. DC: ASM Press. pp 3–11. 17. Cunningham TS, Andhare R, Cooper TG (2000) Nitrogen catabolite repression 2. Fu YH, Marzluf GA (1990) nit-2, the major positive-acting nitrogen regulatory of DAL80 expression depends on the relative levels of Gat1p and Ure2p gene of Neurospora crassa, encodes a sequence-specific DNA-binding protein. Proc production in Saccharomyces cerevisiae. J Biol Chem 275: 14408–14414. Natl Acad Sci U S A 87: 5331–5335. 18. Pan H, Feng B, Marzluf GA (1997) Two distinct protein-protein interactions 3. Kudla B, Caddick MX, Langdon T, Martinez-Rossi NM, Bennett CF, et al. between the NIT2 and NMR regulatory proteins are required to establish (1990) The regulatory gene areA mediating nitrogen metabolite repression in nitrogen metabolite repression in Neurospora crassa. Mol Microbiol 26: 721–729. Aspergillus nidulans. Mutations affecting specificity of gene activation alter a loop 19. Premakumar R, Sorger GJ, Gooden D (1980) Physiological characterization of a residue of a putative zinc finger. EMBO J 9: 1355–1364. Neurospora crassa mutant with impaired regulation of nitrate reductase. J Bacteriol 4. Minehart PL, Magasanik B (1991) Sequence and expression of GLN3, a positive 144: 542–551. nitrogen regulatory gene of Saccharomyces cerevisiae encoding a protein with a 20. Xiao X, Fu YH, Marzluf GA (1995) The negative-acting NMR regulatory putative zinc finger DNA-binding domain. Mol Cell Biol 11: 6216–6228. protein of Neurospora crassa binds to and inhibits the DNA-binding activity of the 5. Mitchell AP, Magasanik B (1984) Regulation of glutamine-repressible gene positive-acting nitrogen regulatory protein NIT2. Biochemistry 34: 8861–8868. products by the GLN3 function in Saccharomyces cerevisiae. Mol Cell Biol 4: 21. Andrianopoulos A, Kourambas S, Sharp JA, Davis MA, Hynes MJ (1998) 2758–2766. Characterization of the Aspergillus nidulans nmrA gene involved in nitrogen 6. Mitchell AP, Magasanik B (1984) Three regulatory systems control production metabolite repression. J Bacteriol 180: 1973–1977. of glutamine synthetase in Saccharomyces cerevisiae. Mol Cell Biol 4: 2767–2773. 22. Lamb HK, Leslie K, Dodds AL, Nutley M, Cooper A, et al. (2003) The negative 7. Stanbrough M, Rowen DW, Magasanik B (1995) Role of the GATA factors transcriptional regulator NmrA discriminates between oxidized and reduced Gln3p and Nil1p of Saccharomyces cerevisiae in the expression of nitrogen-regulated dinucleotides. J Biol Chem 278: 32107–32114. genes. Proc Natl Acad Sci U S A 92: 9450–9454. 23. Platt A, Langdon T, Arst HN, Jr., Kirk D, Tollervey D, et al. (1996) Nitrogen 8. Stewart V, Vollmer SJ (1986) Molecular cloning of nit-2, a regulatory gene required for nitrogen metabolite repression in Neurospora crassa. Gene 46: metabolite signalling involves the C-terminus and the GATA domain of the 9 291–295. Aspergillus transcription factor AREA and the 3 untranslated region of its 9. Arst HN, Jr., Cove DJ (1973) Nitrogen metabolite repression in Aspergillus mRNA. EMBO J 15: 2791–2801. nidulans. Mol Gen Genet 126: 111–141. 24. Langdon T, Sheerins A, Ravagnani A, Gielkens M, Caddick MX, et al. (1995) 10. Magasanik B, Kaiser CA (2002) Nitrogen regulation in Saccharomyces cerevisiae. Mutational analysis reveals dispensability of the N-terminal region of the Gene 290: 1–18. Aspergillus transcription factor mediating nitrogen metabolite repression. Mol 11. Marzluf GA (1997) Genetic regulation of nitrogen metabolism in the fungi. Microbiol 17: 877–888. Microbiol Mol Biol Rev 61: 17–32. 25. Morozov IY, Martinez MG, Jones MG, Caddick MX (2000) A defined sequence 12. Wong KH, Hynes MJ, Davis MA (2008) Recent advances in nitrogen within the 39 UTR of the areA transcript is sufficient to mediate nitrogen regulation: a comparison between Saccharomyces cerevisiae and filamentous fungi. metabolite signalling via accelerated deadenylation. Mol Microbiol 37: Eukaryot Cell 7: 917–925. 1248–1257. 13. Coffman JA, Rai R, Cunningham T, Svetlov V, Cooper TG (1996) Gat1p, a 26. Biswas K, Morschhauser J (2005) The Mep2p ammonium permease controls GATA family protein whose production is sensitive to nitrogen catabolite nitrogen starvation-induced filamentous growth in Candida albicans.Mol repression, participates in transcriptional activation of nitrogen-catabolic genes Microbiol 56: 649–669. in Saccharomyces cerevisiae. Mol Cell Biol 16: 847–858. 27. Dabas N, Morschhauser J (2007) Control of ammonium permease expression 14. Beck T, Hall MN (1999) The TOR signalling pathway controls nuclear and filamentous growth by the GATA transcription factors GLN3 and GAT1 in localization of nutrient-regulated transcription factors. Nature 402: 689–692. Candida albicans. Eukaryot Cell 6: 875–888. 15. Bertram PG, Choi JH, Carvalho J, Ai W, Zeng C, et al. (2000) Tripartite 28. Liao WL, Ramon AM, Fonzi WA (2008) GLN3 encodes a global regulator of regulation of Gln3p by TOR, Ure2p, and phosphatases. J Biol Chem 275: nitrogen metabolism and virulence of C. albicans. Fungal Genet Biol 45: 35727–35733. 514–526.

PLoS ONE | www.plosone.org 13 March 2012 | Volume 7 | Issue 3 | e32585 Nitrogen Metabolite Repression in Cryptococcus

29. Limjindaporn T, Khalaf RA, Fonzi WA (2003) Nitrogen metabolism and 57. Livak KJ, Schmittgen TD (2001) Analysis of relative gene expression data using virulence of Candida albicans require the GATA-type transcriptional activator real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25: encoded by GAT1. Mol Microbiol 50: 993–1004. 402–408. 30. Hensel M, Arst HN, Jr., Aufauvre-Brown A, Holden DW (1998) The role of the 58. James P, Halladay J, Craig EA (1996) Genomic libraries and a host strain Aspergillus fumigatus areA gene in invasive pulmonary aspergillosis. Mol Gen Genet designed for highly efficient two-hybrid selection in yeast. Genetics 144: 258: 553–557. 1425–1436. 31. Bugeja HE, Hynes MJ, Andrianopoulos A (2011) AreA controls nitrogen source 59. Gietz RD, Schiestl RH (2007) High-efficiency yeast transformation using the utilisation during both growth programs of the dimorphic fungus Penicillium LiAc/SS carrier DNA/PEG method. Nat Protoc 2: 31–34. marneffei. Fungal Biol. In press. 60. Cardenas ME, Hemenway C, Muir RS, Ye R, Fiorentino D, et al. (1994) 32. Kwon-Chung KJ, Bennett JE (1992) Cryptococcosis. In: Kwon-Chung KJ, Immunophilins interact with calcineurin in the absence of exogenous Bennett JE, eds. Medical mycology. 2nd ed. Philadelphia, P.A.: Lea & Febiger. immunosuppressive ligands. EMBO J 13: 5944–5957. pp 397–446. 61. Mylonakis E, Ausubel FM, Perfect JR, Heitman J, Calderwood SB (2002) 33. Kmetzsch L, Staats CC, Simon E, Fonseca FL, Oliveira DL, et al. (2011) The Killing of Caenorhabditis elegans by Cryptococcus neoformans as a model of yeast GATA-type transcriptional activator Gat1 regulates nitrogen uptake and pathogenesis. Proc Natl Acad Sci U S A 99: 15675–15680. metabolism in the human pathogen Cryptococcus neoformans. Fungal Genet Biol 62. Cox GM, Mukherjee J, Cole GT, Casadevall A, Perfect JR (2000) Urease as a 48: 192–199. virulence factor in experimental cryptococcosis. Infect Immun 68: 443–448. 34. Lee IR, Chow EW, Morrow CA, Djordjevic JT, Fraser JA (2011) Nitrogen 63. Rhodes JC, Roberts GD (1975) Comparison of four methods for determining Metabolite Repression of Metabolism and Virulence in the Human Fungal nitrate utilization by cryptococci. J Clin Microbiol 1: 9–10. Pathogen Cryptococcus neoformans. Genetics 188: 309–323. 64. Slot JC, Hibbett DS (2007) Horizontal transfer of a nitrate assimilation gene 35. Edskes HK, Engel A, McCann LM, Brachmann A, Tsai HF, et al. (2011) Prion- cluster and ecological transitions in fungi: a phylogenetic study. PLoS One 2: forming ability of Ure2 of yeasts is not evolutionarily conserved. Genetics 188: e1097. 81–90. 65. Pateman JA, Dunn E, Mackay EM (1982) Urea and thiourea transport in 36. Edskes HK, Wickner RB (2002) Conservation of a portion of the S. cerevisiae Aspergillus nidulans. Biochem Genet 20: 777–790. Ure2p prion domain that interacts with the full-length protein. Proc Natl Acad 66. Conlon H, Zadra I, Haas H, Arst HN, Jr., Jones MG, et al. (2001) The Aspergillus Sci U S A 99 Suppl 4: 16384–16391. nidulans GATA transcription factor gene areB encodes at least three proteins and 37. Wong KH, Hynes MJ, Todd RB, Davis MA (2007) Transcriptional control of features three classes of mutation. Mol Microbiol 40: 361–375. nmrA by the bZIP transcription factor MeaB reveals a new level of nitrogen 67. Copper TG (2004) GATA-factor-mediated transcription and Tor1/2 in regulation in Aspergillus nidulans. Mol Microbiol 66: 534–551. Saccharomyces cerevisiae from a physiological perspective. In: Brambl R, 38. Jiang N, Xiao D, Zhang D, Sun N, Yan B, et al. (2009) Negative roles of a novel Marzluf GA, eds. Biochemistry and molecular biology The Mycota III. 2 ed. nitrogen metabolite repression-related gene, TAR1, in laccase production and Berlin, Germany: Springer Verlag. pp 81–119. nitrate utilization by the basidiomycete Cryptococcus neoformans. Appl Environ 68. Chun CD, Brown JC, Madhani HD (2011) A major role for capsule- Microbiol 75: 6777–6782. independent phagocytosis-inhibitory mechanisms in mammalian infection by 39. Kumar V, Carlson JE, Ohgi KA, Edwards TA, Rose DW, et al. (2002) Cryptococcus neoformans. Cell Host Microbe 9: 243–251. Transcription corepressor CtBP is an NAD(+)-regulated dehydrogenase. Mol 69. Idnurm A, Bahn YS, Nielsen K, Lin XR, Fraser JA, et al. (2005) Deciphering Cell 10: 857–869. the model pathogenic fungus Cryptococcus neoformans. Nat Rev Microbiol 3: 753–764. 40. Lamb HK, Ren J, Park A, Johnson C, Leslie K, et al. (2004) Modulation of the ligand binding properties of the transcription repressor NmrA by GATA- 70. Jung WH, Sham A, White R, Kronstad JW (2006) Iron regulation of the major virulence factors in the AIDS-associated pathogen Cryptococcus neoformans. PLoS containing DNA and site-directed mutagenesis. Protein Sci 13: 3127–3138. Biol 4: e410. 41. Stammers DK, Ren J, Leslie K, Nichols CE, Lamb HK, et al. (2001) The 71. Liu OW, Chun CD, Chow ED, Chen C, Madhani HD, et al. (2008) Systematic structure of the negative transcriptional regulator NmrA reveals a structural genetic analysis of virulence in the human fungal pathogen Cryptococcus neoformans. superfamily which includes the short-chain dehydrogenase/reductases. EMBO J Cell 135: 174–188. 20: 6619–6626. 72. Haas H, Bauer B, Redl B, Stoffler G, Marzluf GA (1995) Molecular cloning and 42. Thoden JB, Sellick CA, Reece RJ, Holden HM (2007) Understanding a analysis of nre, the major nitrogen regulatory gene of Penicillium chrysogenum. Curr transcriptional paradigm at the molecular level. The structure of yeast Gal80p. Genet 27: 150–158. J Biol Chem 282: 1534–1538. 73. Zheng X, Dai X, Zhao Y, Chen Q, Lu F, et al. (2007) Restructuring of the 43. Young JL, Jarai G, Fu YH, Marzluf GA (1990) Nucleotide sequence and analysis dinucleotide-binding fold in an NADP(H) sensor protein. Proc Natl Acad NMR Neurospora of , a negative-acting regulatory gene in the nitrogen circuit of Sci U S A 104: 8809–8814. crassa. Mol Gen Genet 222: 120–128. 74. Missall TA, Moran JM, Corbett JA, Lodge JK (2005) Distinct stress responses of 44. D’Souza CA, Alspaugh JA, Yue C, Harashima T, Cox GM, et al. (2001) Cyclic two functional laccases in Cryptococcus neoformans are revealed in the absence of the AMP-dependent protein kinase controls virulence of the fungal pathogen thiol-specific antioxidant Tsa1. Eukaryot Cell 4: 202–208. Cryptococcus neoformans. Mol Cell Biol 21: 3179–3191. 75. Zhu X, Williamson PR (2004) Role of laccase in the biology and virulence of 45. Pukkila-Worley R, Gerrald QD, Kraus PR, Boily MJ, Davis MJ, et al. (2005) Cryptococcus neoformans. FEMS Yeast Res 5: 1–10. Transcriptional network of multiple capsule and melanin genes governed by the 76. Jacobson ES, Emery HS (1991) Temperature regulation of the cryptococcal Cryptococcus neoformans cyclic AMP cascade. Eukaryot Cell 4: 190–201. phenoloxidase. J Med Vet Mycol 29: 121–124. 46. Williamson PR (1994) Biochemical and molecular characterization of the 77. Morrow CA, Lee IR, Chow EW, Ormerod KO, Goldinger A, et al. (2012) A diphenol oxidase of Cryptococcus neoformans: identification as a laccase. J Bacteriol unique chromosomal rearrangement in the Cryptococcus neoformans var. grubii type 176: 656–664. strain enhances key phenotypes associated with virulence. mBioIn press. 47. Nielsen K, De Obaldia AL, Heitman J (2007) Cryptococcus neoformans mates on 78. Janbon G (2004) Cryptococcus neoformans capsule biosynthesis and regulation. pigeon guano: implications for the realized ecological niche and globalization. FEMS Yeast Res 4: 765–771. Eukaryot Cell 6: 949–959. 79. Kwon-Chung KJ (1976) Morphogenesis of Filobasidiella neoformans, the sexual 48. Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning : a laboratory state of Cryptococcus neoformans. Mycologia 68: 821–833. manual. Cold Spring Harbor, N.Y.: Cold Spring Harbor Laboratory. 80. Nielsen K, Cox GM, Wang P, Toffaletti DL, Perfect JR, et al. (2003) Sexual 49. Brenner S (1974) The genetics of Caenorhabditis elegans. Genetics 77: 71–94. cycle of Cryptococcus neoformans var. grubii and virulence of congenic a and alpha 50. Garsin DA, Sifri CD, Mylonakis E, Qin X, Singh KV, et al. (2001) A simple isolates. Infect Immun 71: 4831–4841. model host for identifying Gram-positive virulence factors. Proc Natl Acad 81. Rutherford JC, Lin X, Nielsen K, Heitman J (2008) Amt2 permease is required Sci U S A 98: 10892–10897. to induce ammonium-responsive invasive growth and mating in Cryptococcus 51. Honda S, Ishii N, Suzuki K, Matsuo M (1993) Oxygen-dependent perturbation neoformans. Eukaryot Cell 7: 237–246. of life span and aging rate in the nematode. J Gerontol 48: B57–61. 82. Fraser JA, Davis MA, Hynes MJ (2001) The formamidase gene of Aspergillus 52. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local nidulans: regulation by nitrogen metabolite repression and transcriptional alignment search tool. J Mol Biol 215: 403–410. interference by an overlapping upstream gene. Genetics 157: 119–131. 53. Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTAL W: improving the 83. Fraser JA, Davis MA, Hynes MJ (2002) A gene from Aspergillus nidulans with sensitivity of progressive multiple sequence alignment through sequence similarity to URE2 of Saccharomyces cerevisiae encodes a glutathione S- weighting, position-specific gap penalties and weight matrix choice. Nucleic transferase which contributes to heavy metal and xenobiotic resistance. Appl Acids Res 22: 4673–4680. Environ Microbiol 68: 2802–2808. 54. Davidson RC, Blankenship JR, Kraus PR, de Jesus Berrios M, Hull CM, et al. 84. Bousset L, Belrhali H, Janin J, Melki R, Morera S (2001) Structure of the (2002) A PCR-based strategy to generate integrative targeting alleles with large globular region of the prion protein Ure2 from the yeast Saccharomyces cerevisiae. regions of homology. Microbiology 148: 2607–2615. Structure 9: 39–46. 55. Fraser JA, Subaran RL, Nichols CB, Heitman J (2003) Recapitulation of the 85. Umland TC, Taylor KL, Rhee S, Wickner RB, Davies DR (2001) The crystal sexual cycle of the primary fungal pathogen Cryptococcus neoformans var. gattii: structure of the nitrogen regulation fragment of the yeast prion protein Ure2p. implications for an outbreak on Vancouver Island, Canada. Eukaryot Cell 2: Proc Natl Acad Sci U S A 98: 1459–1464. 1036–1045. 86. Leuther KK, Johnston SA (1992) Nondissociation of GAL4 and GAL80 in vivo 56. Southern E (2006) Southern blotting. Nat Protoc 1: 518–525. after galactose induction. Science 256: 1333–1335.

PLoS ONE | www.plosone.org 14 March 2012 | Volume 7 | Issue 3 | e32585 Nitrogen Metabolite Repression in Cryptococcus

87. Leuther KK, Salmeron JM, Johnston SA (1993) Genetic evidence that an 91. Wagner D, Schmeinck A, Mos M, Morozov IY, Caddick MX, et al. (2010) The activation domain of GAL4 does not require acidity and may form a beta sheet. bZIP transcription factor MeaB mediates nitrogen metabolite repression at Cell 72: 575–585. specific loci. Eukaryot Cell 9: 1588–1601. 88. Selleck SB, Majors JE (1987) In vivo DNA-binding properties of a yeast 92. Kronstad J, Saikia S, Nielson ED, Kretschmer M, Jung W, et al. (2011) transcription activator protein. Mol Cell Biol 7: 3260–3267. Adaptation of Cryptococcus neoformans to mammalian hosts: integrated regulation 89. Caddick MX, Arst HN, Jr. (1998) Deletion of the 389 N-terminal residues of the of metabolism and virulence. Eukaryot CellIn press. transcriptional activator AREA does not result in nitrogen metabolite 93. Park BJ, Wannemuehler KA, Marston BJ, Govender N, Pappas PG, et al. (2009) derepression in Aspergillus nidulans. J Bacteriol 180: 5762–5764. Estimation of the current global burden of cryptococcal meningitis among 90. Lamb HK, Dodds AL, Swatman DR, Cairns E, Hawkins AR (1997) Deletion of the persons living with HIV/AIDS. AIDS 23: 525–530. N-terminal region of the AREA protein is correlated with a derepressed phenotype 94. Staib F, Grave B, Altmann L, Mishra SK, Abel T, et al. (1978) Epidemiology of with respect to nitrogen metabolite repression. J Bacteriol 179: 6649–6656. Cryptococcus neoformans. Mycopathologia 65: 73–76.

PLoS ONE | www.plosone.org 15 March 2012 | Volume 7 | Issue 3 | e32585 PLoS ONE

Supporting Information http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone .0032585

Characterization of an Nmr Homolog That Modulates GATA Factor-Mediated Nitrogen Metabolite Repression in Cryptococcus neoformans

I. Russel Lee, Jonathan W. C. Lim, Kate L. Ormerod, Carl A. Morrow, James A. Fraser

DOI: 10.1371/journal.pone.0032585

Figure S1. Complementation of toxic analog sensitivity phenotype to wild-type levels upon the reintroduction of TAR1 into the tar1Δ mutant. Tenfold spot dilution assays for nitrogen utilization showed that the tar1Δ + TAR1 strain exhibited wild-type growth on 5 mM thiourea plus 10 mM ammonium.

2 SI

Figure S2. ClustalW multiple sequence alignment of A. nidulans AreA (XP_681936.1), N. crassa Nit2 (P19212.2) and C. neoformans Gat1/Are1 (CNAG00193.2). Identical amino acid residues are shaded dark grey while similar residues are shaded light grey. Gat1/Are1 shows little overall sequence conservation to AreA and Nit2 at the entire protein level. However, three notable blocks of conversation do exist between Gat1/Are1 and AreA or Nit2: two regions of unknown function (boxed in red and orange) and the GATA DNA-binding domain (boxed in blue). The second conserved motif of unknown function (boxed in orange) was identified through Pustell dot plot protein matrix. The

3 SI extreme C-terminus domain of AreA and Nit2 (boxed in green) is highly conserved but that of Gat1/Are1 has diverged. Enlarged versions of all four domains are shown in Figure 5B, C, D and E.

4 SI

Figure S3. ClustalW multiple sequence alignment of A. nidulans NmrA (AAC39442.1), N. crassa Nmr1 (P23762.2) and C. neoformans Tar1 (CNAG04934.2). Identical amino acid residues are shaded dark grey while similar residues are shaded light grey. Tar1 shows moderate overall sequence conservation to NmrA and Nmr1. The predicted Rossmann fold motif at the N-terminus is boxed in red. The long C-terminus region of NmrA and Nmr1 is absent in Tar1.

5 SI

Figure S4. Multiple independently generated tar1Δ mutants all produced equal amount of pigment melanin in comparison to their wild-type H99 or KN99a counterparts. The tenfold spot dilution assays for melanization on L-DOPA, norepinephrine and caffeic acid agar (supplemented with 10 mM asparagine as the nitrogen source) were conducted at both 30 and 37°C.

6 SI

Figure S5. Scheme representing the dual roles of Tar1 in modulating GAT1/ARE1 transcription that in turn influences PUT1 expression according to the nitrogen source available. In the presence of the traditionally non-preferred proline, the Nmr homolog Tar1 positively regulates the transcription of the GATA factor-encoding gene GAT1/ARE1 that is required for induction of the proline oxidase-encoding gene PUT1. In the presence of the preferred ammonium, Tar1 negatively regulates the transcription of GAT1/ARE1 leading to reduce levels of PUT1 expression.

7 SI Table S1. Fungal strains used in this study.

Strain Genotype Original source/reference H99 Cryptococcus neoformans laboratory strain John Perfect RL1 Cryptococcus neoformans H99 gat1/are1::NEO Lee et al. 2011 RL10 Cryptococcus neoformans H99 tar1::NEO This study RL9 Cryptococcus neoformans H99 gat1/are1::NEO, tar1::NAT This study RL14 Cryptococcus neoformans H99 tar1::NEO + TAR1 This study RL2 Cryptococcus neoformans H99 gat201::NEO Lee et al. 2011 RL3 Cryptococcus neoformans H99 bwc2::NEO Lee et al. 2011 RL4 Cryptococcus neoformans H99 CNAG04263.2::NEO Lee et al. 2011 RL5 Cryptococcus neoformans H99 gat204::NEO Lee et al. 2011 RL6 Cryptococcus neoformans H99 CNAG03401.2::NEO Lee et al. 2011 H9CIR4 Cryptococcus neoformans H99 cir1::NAT Won Hee Jung KN99a Cryptococcus neoformans congenic strain Kirsten Nielsen RL11 Cryptococcus neoformans KN99a tar1::NAT This study AH109 Saccharomyces cerevisiae matchmaker two-hybrid strain Clontech

8 SI Table S2. Primers used in this study.

Primer Purpose Sequence (5’-3’) UQ1070 PUT1 qRT-PCR AAAAGCTAAAGAGAACGACGTTGC UQ1071 PUT1 qRT-PCR TCTCACTCTTTGACTTTGGAGGTTT UQ1634 GAT1/ARE1 qRT-PCR CAATGCGTGTGGACTTTTCTATAAA UQ1635 GAT1/ARE1 qRT-PCR GACCCGCTCGGTTTCTTTTC UQ2067 TAR1 qRT-PCR CAGTGTATCTGCTGCGGACTTG UQ2068 TAR1 qRT-PCR GACATTCCAGAGGCCTGAAGAC UQ1521 LAC1 qRT-PCR TCAATGGACGTGGCCAAAC UQ1522 LAC1 qRT-PCR GCAATTGACTGGCACGTGAA UQ1523 LAC2 qRT-PCR CCTAGTGTATTCGTTCTTTCAAATGACT UQ1524 LAC2 qRT-PCR TGGCCAAAGCCTCAGTAAGG UQ856 GAT1/ARE1 deletion / genomic DNA generation GTTGCGGGATGCTGGTGAAAT UQ857 GAT1/ARE1 deletion AGCTCACATCCTCGCAGCCCTAGCCGCCGCTAGCTG UQ858 GAT1/ARE1 deletion GTGTTAATACAGATAAACCGCGAATGAATGGAAGCA UQ859 GAT1/ARE1 deletion / genomic DNA generation GCCCAGCAACTACCATCGTGT UQ946 GAT1/ARE1 deletion CAGCTAGCGGCGGCTAGGGCTGCGAGGATGTGAGCT UQ947 GAT1/ARE1 deletion TTGCTTCCATTCATTCGCGGTTTATCTGTATTAACA UQ1418 TAR1 deletion / genomic DNA generation TAAAGGTCTGGGCACAGGAAG UQ1419 TAR1 deletion AGCTCACATCCTCGCAGCATTGAATTGAAGAATGTA UQ1420 TAR1 deletion ACTACATTCTTCAATTCAATGCTGCGAGGATGTGAG UQ1421 TAR1 deletion TCTCCCCGGAAGTCTTTTGGTTTATCTGTATTAACA UQ1422 TAR1 deletion TGTTAATACAGATAAACCAAAAGACTTCCGGGGAGA UQ1423 TAR1 deletion / genomic DNA generation CGGCCAGAGAAAGAGTAGCTG UQ18 Sequencing (M13F pCR2.1-TOPO) GTAAAACGACGGCCAG UQ19 Sequencing (M13R pCR2.1-TOPO) CAGGAAACAGCTATGAC UQ1983 Sequencing (pGBKT7/pGADT7) TAATACGACTCACTATAGGGC UQ1984 Sequencing (pGBKT7) TTTTCGTTTTAAAACCTAAGAGTC UQ1985 Sequencing (pGADT7) AGATGGTGCACGATGCACAG UQ1027 GAT1/ARE1 genomic DNA sequencing GAAAGGCGAATAAAGGCTGTT UQ1028 GAT1/ARE1 genomic DNA / cDNA sequencing TCTGCCGCGTCTTCCCGGCCC UQ1029 GAT1/ARE1 genomic DNA / cDNA sequencing CCAAAAGCTGCTAGGGGCACC UQ1030 GAT1/ARE1 genomic DNA / cDNA sequencing TTGCCAAACGGTTTGTCTCTT UQ1031 GAT1/ARE1 genomic DNA / cDNA sequencing GCAGCCGCAGCAACGGCTGCG UQ1032 GAT1/ARE1 genomic DNA sequencing AAAAAAAAGCATAAATAAAAA UQ1033 GAT1/ARE1 genomic DNA / cDNA sequencing TGACCGCACAGACGTATTTTT UQ1034 GAT1/ARE1 genomic DNA / cDNA sequencing AGCTTTCGGTAATGGCCGGAT UQ1035 GAT1/ARE1 genomic DNA / cDNA sequencing GTGGTCGAAGGATGTCTTCCG UQ1443 TAR1 genomic DNA sequencing CACCAGGCAACACCAGGTTCT UQ1444 TAR1 genomic DNA sequencing CATTCTTCAATTCAATATGGC UQ1445 TAR1 genomic DNA sequencing ATGCTTGTCACAGACCTCTCA UQ1446 TAR1 genomic DNA sequencing GGGGAGAGGGAGGGGGAGACT UQ1735 TAR1 genomic DNA sequencing GCTTGAAAGTGAGTCACTGGC UQ1765 TAR1 cDNA sequencing GACTTCTGCCTTACGACTTCA UQ1766 TAR1 cDNA sequencing CTTGGCGAAGGATGTGCCGCA UQ1716 GAT1/ARE1 cDNA generation CGACAATACGCCTCTCCCCGA UQ1754 GAT1/ARE1 cDNA generation GTCCACTTCTGCCGAAACCAA UQ1755 TAR1 cDNA generation GAAATGTCTTCACATTCGACA UQ1767 TAR1 cDNA generation AGTCTCCCCCTCCCTCTCCCC UQ1971 Nucleotides 1–1,284 GAT1/ARE1 cDNA (MfeI site) GTTAGCCAATTGACTGCGACCAAACAGCAG UQ1973 Nucleotides 1–1,284 GAT1/ARE1 cDNA (BglII site) GGGGCTAGATCTGTCATGAGGAGCTTCAGA UQ1974 Nucleotides 1,285–2,562 GAT1/ARE1 cDNA (MfeI site) GAAGCTCAATTGCCTCATGAAGGCCCCATG UQ1975 Nucleotides 1,285–2,562 GAT1/ARE1 cDNA (BglII site) GAGTTCAGATCTTGGACCATTGATACCAGG UQ1976 Nucleotides 2,563–3,834 GAT1/ARE1 cDNA (MfeI site) GGTCCACAATTGGGCCTCTATACACATAGC UQ1972 Nucleotides 2,563–3,834 GAT1/ARE1 cDNA (BglII site) ATTCGCAGATCTTCATTCAACAAGAGCATC UQ1969 TAR1 cDNA (EcoRI site) TCGACAGAATTCATGGCGGCCACGACCATC UQ1970 TAR1 cDNA (BamHI site) TCTTTTGGATCCCTATAAATCTTGTCGGAA

9 SI Table S3. Plasmids used in this study.

Plasmid Detail Original source pJAF1 NEO selectable marker in pCR2.1-TOPO James Fraser pCH233 NAT selectable marker in pCR2.1-TOPO Christina Hull pGBKT7 Matchmaker two-hybrid GAL4 BD vector Clontech pGADT7 Matchmaker two-hybrid GAL4 AD vector Clontech pGBKT7-53 Murine p53 cDNA in pGBKT7 Clontech pGADT7-T SV40 large T antigen cDNA in pGADT7 Clontech pIRL25 TAR1 genomic DNA in pCR2.1-TOPO This study pIRL26 TAR1 genomic DNA in pCH233 This study pIRL27 Full-length GAT1/ARE1 cDNA in pCR2.1-TOPO This study pIRL21 Full-length TAR1 cDNA in pCR2.1-TOPO This study pIRL30 Nucleotides 1–1,284 GAT1/ARE1 cDNA in pGBKT7 This study pIRL31 Nucleotides 1,285–2,562 GAT1/ARE1 cDNA in pGBKT7 This study pIRL32 Nucleotides 2,563–3,834 GAT1/ARE1 cDNA in pGBKT7 This study pIRL29 Full-length GAT1/ARE1 cDNA in pGBKT7 This study pIRL28 Full-length TAR1 cDNA in pGBKT7 This study pIRL35 Nucleotides 1–1,284 GAT1/ARE1 cDNA in pGADT7 This study pIRL36 Nucleotides 1,285–2,562 GAT1/ARE1 cDNA in pGADT7 This study pIRL37 Nucleotides 2,563–3,834 GAT1/ARE1 cDNA in pGADT7 This study pIRL34 Full-length GAT1/ARE1 cDNA in pGADT7 This study pIRL33 Full-length TAR1 cDNA in pGADT7 This study

10 SI Table S4. Interaction between Tar1 and Gat1/Are1 could not be detected in a yeast two-hybrid assay.

GAL4 AD Full-length First-third Second-third Final-third Full-length SV40 large T Empty vector GAT1/ARE1 GAT1/ARE1 GAT1/ARE1 GAT1/ARE1 TAR1 (+ve control, (-ve control, (pIRL34) (pIRL35) (pIRL36) (pIRL37) (pIRL33) pGADT7-T) pGADT7)

GAL4 BD

Full-length GAT1/ARE1 A.A. A.A. A.A. A.A. A.A. N.T. A.A. (pIRL29)

First-third GAT1/ARE1 A.A. A.A. A.A. A.A. A.A. N.T. A.A. (pIRL30)

Second-third GAT1/ARE1 - - - - - N.T. - (pIRL31)

Final-third GAT1/ARE1 - - - - - N.T. - (pIRL32)

Full-length TAR1 - - - - - N.T. - (pIRL28)

Murine p53 (+ve control, N.T. N.T. N.T. N.T. N.T. + - pGBKT7-53)

Empty vector (-ve control, ------pGBKT7)

S. cerevisiae AH109 was co-transformed with both the bait and prey constructs, and selected for growth on double dropout medium (-Leu -Trp). Interaction (denoted by + symbol) was assessed by growth on quadruple dropout medium (-Leu -Trp -His - Ade), and β-galactosidase activity. - symbol denotes no protein-protein interaction. A.A. denotes auto-activation. N.T. denotes not tested.

11 SI

CHAPTER 5

Results Reactive Oxygen Species Homeostasis and Virulence of the Fungal Pathogen Cryptococcus neoformans Requires an Intact Proline Catabolism Pathway

84 5.1 Preface

As reviewed in Chapters 3 and 4, the Cryptococcus neoformans proline degradation pathway has proven to be a fertile area of research for the molecular dissection of the regulatory mechanism of nitrogen metabolite repression. While transcriptional regulation of the predicted proline catabolic enzyme-encoding genes was investigated, the enzymes of this pathway have not been formally characterised. Based on the brief examination of the putative catabolic genes of this imino acid, several interesting features surfaced that warrant further study. For example, there appears to be two paralogs predicted to encode proline oxidase in the H99 genome; gene duplications are believed to facilitate evolutionary innovation that often results in an increase in the overall fitness of an organism. Also, intriguingly, proline assimilation in the basidiomycete C. neoformans is partially independent of a functional GATA transcription factor, counterintuitive to the scenario observed in the ascomyctes where GATA factors are traditionally required to activate secondary nitrogen source utilisation. Further inspection of the literature on proline catabolic enzymes of various biological systems support the theory that these homologs in C. neoformans are likely to be important for virulence due to their anticipated role in oxidative stress. In this chapter, all the C. neoformans genes encoding the proline catabolic enzymes are functionally characterised at the molecular level.

85 INVESTIGATION

Reactive Oxygen Species Homeostasis and Virulence of the Fungal Pathogen Cryptococcus neoformans Requires an Intact Proline Catabolism Pathway

I. Russel Lee,1 Edmund Y. L. Lui,1 Eve W. L. Chow, Samantha D. M. Arras, Carl A. Morrow, and James A. Fraser2 Australian Infectious Diseases Research Centre, School of Chemistry and Molecular Biosciences, University of Queensland, Brisbane, Queensland 4072, Australia

ABSTRACT Degradation of the multifunctional amino acid proline is associated with mitochondrial oxidative respiration. The two-step oxidation of proline is catalyzed by proline oxidase and D1-pyrroline-5-carboxylate (P5C) dehydrogenase, which produce P5C and glutamate, respectively. In animal and plant cells, impairment of P5C dehydrogenase activity results in P5C-proline cycling when exogenous proline is supplied via the actions of proline oxidase and P5C reductase (the enzyme that converts P5C to proline). This proline is oxidized by the proline oxidase-FAD complex that delivers electrons to the electron transport chain and to O2, leading to mitochondrial reactive oxygen species (ROS) overproduction. Coupled activity of proline oxidase and P5C dehydrogenase is therefore important for maintaining ROS homeostasis. In the genome of the fungal pathogen Cryptococcus neoformans, there are two paralogs (PUT1 and PUT5) that encode proline oxidases and a single ortholog (PUT2) that encodes P5C dehydrogenase. Transcription of all three catabolic genes is inducible by the presence of proline. However, through the creation of deletion mutants, only Put5 and Put2 were found to be required for proline utilization. The put2D mutant also generates excessive mitochondrial superoxide when exposed to proline. Intracellular accumulation of ROS is a critical feature of cell death; consistent with this fact, the put2D mutant exhibits a slight, general growth defect. Furthermore, Put2 is required for optimal production of the major cryptococcal virulence factors. During murine infection, the put2D mutant was discovered to be avirulent; this is the ﬁrst report highlighting the importance of P5C dehydrogenase in enabling pathogenesis of a microorganism.

ROLINE is an important biological proteinogenic amino by which proline protects against stresses remains poorly Pacid that is necessary for primary metabolism in organisms understood. Intracellular proline must be present at appro- including animals, plants, fungi, and bacteria. This imino acid, priate levels to confer stress-protective effects, and an excess as it is more precisely known, has multiple physiological func- of free proline has been shown to be detrimental to cell growth tions. Apart from being an osmoprotectant, proline can serve as or protein functions in numerous organisms (Davis 2000; Morita a transient storage of organic nitrogen, stabilize proteins and et al. 2002; Deuschle et al. 2004; Maxwell and Nomura and membranes during stressful conditions, prevent protein aggre- Takagi 2004). The transport, anabolism, and catabolism of pro- gation during its folding or refolding, and scavenge reactive line therefore need to be tightly regulatedforpropercellular oxygen species (ROS) such as superoxide anions (Rudolph and defense, adaptation, and growth. Crowe 1985; Carpenter and Crowe 1988; Ignatova and Degradation of proline that takes place in eukaryotic mi- Gierasch 2006; Kaul et al. 2008). However, the exact mechanism tochondria or prokaryotic plasma membrane results in high- energy output and may provide amino nitrogen and reducing Copyright © 2013 by the Genetics Society of America power to post-stress recovering cells (Atkinson 1977; Peng doi: 10.1534/genetics.113.150326 et al. 1996; Verbruggen et al. 1996; Hare and Cress 1997). Manuscript received February 11, 2013; accepted for publication March 31, 2013 ﬁ Supporting information is available online at http://www.genetics.org/lookup/suppl/ Hence, the process of proline catabolism is also bene cial, in doi:10.1534/genetics.113.150326/-/DC1. addition to its intracellular accumulation. The four-electron oxi- 1 These authors contributed equally to this work. 1 2Corresponding author: 358 Molecular Biosciences Bldg., Cooper Rd., University of dation steps that convert proline to glutamate, with D -pyrroline- Queensland, Brisbane, QLD 4072, Australia. E-mail: [email protected] 5-carboxylate (P5C) as an intermediate, occur in all living

Genetics, Vol. 194, 421–433 June 2013 421 systems (Figure 1A). Oxidation of proline to P5C is first catalyzed research into dissecting the mechanisms of virulence in path- by the O2-dependent proline oxidase (also known as proline ogenic fungi. One of the leading fatal fungal infections is dehydrogenase). The resulting P5C is nonenzymatically cryptococcal meningoencephalitis. Cryptococcus neoformans transformed into glutamate-g-semialdehyde (GSA), which acts predominantly infects immunocompromised patients, while as a substrate for P5C dehydrogenase to generate glutamate. its sister species Cryptococcus gattii generally infects immuno- Glutamate is an important precursor molecule for all other competent individuals (Heitman 2010).Historically,differential cellular nitrogenous compounds. media were used to distinguish the different basidiomycete Conversely, uncoupled proline oxidase activity can also species, capitalizing on their unique ability to metabolize differ- be detrimental depending on the behavior of other catabolic ent substrates. For example, C. neoformans is unable to utilize and biosynthesis enzymes of this metabolism. In animals D-proline while C. gattii is among the rare few microorganisms and plants, when proline oxidase activation is not accom- that can assimilate this dextrorotatory form of proline (Dufait panied by P5C dehydrogenase activation, intracellular levels et al. 1987; Ngamskulrungroj et al. 2012). of the toxic intermediate P5C or GSA (thought to be more In the more clinically prevalent C. neoformans, regulation chemically reactive and causative) as well as ROS are in- of proline (L-proline, unless specified otherwise) utilization creased and redox homeostasis is altered, ultimately leading is particularly interesting because a functional GATA factor to cell death. Excessive ROS is generated from the P5C- (Gat1/Are1) is not required for assimilation of this traditionally proline cycle that involves the constant interconversion be- nonpreferred nitrogen source, an observation that is different tween P5C and proline, catalyzed by activities of the cytosolic from the assimilation seen in the ascomycetes (Kmetzsch et al. P5C reductase and the mitochondrial proline oxidase (Phang 2011; Lee et al. 2011, 2012). Initiation of C. neoformans in- et al. 2008; Miller et al. 2009). This seemingly futile cycle con- fection occurs via inhalation of airborne basidiospores or desic- trols the cellular accumulation of the proline-derived metabolite cated yeast cells, and one of the initial lines of host defense that P5C/GSA and provides an excess of electron flow to the this pathogen encounters is reactive oxygen, and nitrogen spe- mitochondrial electron transport chain (mtETC) that is partially cies burst from alveolar phagocytes. In addition, C. neoformans diverted to O2, resulting in superoxide anion formation. P5C/ has to endure various stresses found in different body locations GSA has high reactivity with various cellular compounds and of the host such as nutrient limitation for successful prolifera- may act as a signal molecule to elicit deleterious effects in tion, dissemination, and colonization. Thus, the observation of animal, plant, and yeast cells (Mezl and Knox 1976; Maxwell Gat1/Are1 independence in proline utilization, coupled with and Davis 2000; Deuschle et al. 2001, 2004; Donald et al. the supporting literature that proline catabolism is likely a key 2001; Nomura and Takagi 2004). Incomplete proline oxida- regulator of virulence, led us to investigate the role of the tion can also enhance the transference of reducing equiva- proline catabolic enzymes in C. neoformans pathobiology. lents to the mitochondria altering the NADP+/NADPH ratio In this study, we examined how the C. neoformans proline at the cytosol, potentially impacting redox-sensitive pathways catabolic genes are transcriptionally regulated and both ge- such as the defense-associated oxidative pentose phosphate netically and phenotypically characterized their enzymatic pathway (Hare and Cress 1997). functions in relation to nitrogen and carbon source utilization, In the model ascomycete Saccharomyces cerevisiae,expres- balance of mitochondrial ROS formation, as well as stress- sion of PUT1-encoded proline oxidase and PUT2-encoded P5C adaptation and virulence-associated mechanisms. dehydrogenase is activated by the pathway-specifictranscrip- tion factor Put3 in the presence of proline and the absence of easily assimilated nitrogen sources such as ammonium or glu- Materials and Methods tamine (Brandriss and Magasanik 1979; Axelrod et al. 1991; Strains and media Huang and Brandriss 2000). Transcriptional regulation of PUT1 and PUT2 is therefore subjected to nitrogen catabo- All fungal strains used in this study are listed in Supporting lite/metabolite repression, a regulatory phenomenon that ena- Information, Table S1 and were grown in YPD (1% yeast bles preferential utilization of metabolically favored nitrogen extract, 2% Bacto-peptone, 2% glucose) or YNB (0.45% sources. Proline oxidase and P5C dehydrogenase use FAD and yeast nitrogen base w/o aminoacidsandammoniumsul- + NAD as electron acceptors to generate FADH2 and NADH, fate, 2% glucose, 10 mM nitrogen source) unless specified respectively, delivering electrons for mitochondrial respiration. otherwise. C. neoformans biolistic transformants were se- Transport of proline into yeast cells occurs mainly through the lected on YPD medium supplemented with 200 mg/ml general amino acid permease Gap1 and high-affinity proline- G418 (Sigma) or 100 mg/ml nourseothricin (Werner Bio- specific permease Put4 (Vandenbol et al. 1989; Jauniaux and Agents). L-3,4-dihydroxyphenylalanine (L-DOPA) medium Grenson 1990). Activation of the nitrogen catabolite repres- was prepared as described previously (Missall et al. 2005). sion-sensitive GAP1 and PUT4 is induced by the globally acting Escherichia coli Mach-1 cells served as the host strain for GATA transcription factors Gln3 and/or Gat1 (Daugherty et al. transformation and propagation of all plasmids using lysog- 1993; Stanbrough and Magasanik 1996). eny broth supplemented with either 100 mg/ml ampicillin Knowledge gained from pioneering studies of gene regulation (Sigma) or 50 mg/ml kanamycin (Sigma) (Sambrook et al. in model fungi has steadily laid the foundation for modern-day 1989).

422 I. R. Lee et al. Figure 1 The C. neoformans genome contains two paralogs that encode proline oxidases and a single ortholog that encodes P5C dehydrogenase. (A) Scheme representing the proline degradation pathway. (B) Exon–intron structures of PUT1, PUT5,and PUT2 and their chromosomal locations.

Bioinformatic analyses strains containing a single copy of the wild-type PUT1 gene were identified by Southern blot. C. neoformans genes were identified using annotation from the H99 genome sequence from the Broad Institute (http:// Cross-species complementation www.broadinstitute.org/annotation/genome/cryptococcus_ Full-length C. neoformans PUT1, PUT5, and PUT2 ORFs am- neoformans/MultiHome.html). Gene annotations from the plified from strain H99 complementary DNA (cDNA) were Broad Institute are designated by their nomenclature fused with 1-kb promoter and terminator sequences of “CNAG_#####.#.” Sequence analyses were performed us- S. cerevisiae BY4741 PUT1 or PUT2 via high-fidelity overlap ing BLAST and MacVector 9.5 (MacVector Inc., Cary NC) PCR (Brachmann et al. 1998). These fusion genes were cloned (Altschul et al. 1990). Sequence alignments were created us- into pCR2.1-TOPO, sequenced, and subcloned into the yeast ing ClustalW v1.4 within MacVector (Thompson et al. 1994). shuttle vector YCplac111 (Gietz and Sugino 1988). Likewise, Sequence traces generated at the Australian Genome Re- the control S. cerevisiae PUT1 and PUT2 genes, including their search Facility (Brisbane, QLD, Australia) were analyzed using 1-kb promoter and terminator, were amplified from strain Sequencher 4.7 (Gene Codes, Ann Arbor MI). Putative mito- BY4741 genomic DNA, cloned into pCR2.1-TOPO, sequenced, chondrial localization signals were examined using MitoProt and subcloned into YCplac111. Plasmids were transformed (http://ihg.gsf.de/ihg/mitoprot.html). into S. cerevisiae put1D or put2D strains obtained from the BY4741 deletion library using the lithium acetate/heat-shock Construction and complementation of C. neoformans mutant strains method (Gietz and Schiestl 2007). Prior to transformation, thesepreviouslyauxotrophicstrainshadURA3, HIS3,and All primers and plasmids used in this study are listed in MET15 restored, but not LEU2, to remove background in- Table S2 and Table S3, respectively. Gene deletion mutants terference through supplementation of potential nitrogen were created using overlap PCR and biolistic transformation sources (uracil, histidine, methionine). Transformants were as described previously (Davidson et al. 2002). For example, first selected on YNB (2% glucose, 10 mM ammonium sulfate) to construct the put1D mutant strain in the H99 background, lacking leucine; thereafter, LEU2 was restored in the strains. the 2035-bp PUT1 (CNAG_02049.2) coding sequence was Likewise, URA3, HIS3, MET15,andLEU2 were restored in the replaced with the neomycin phosphotransferase II-encoding wild-type and mutant control strains. Functional complemen- selectable marker NEO using a construct created by overlap tation of proline oxidase and P5C dehydrogenase activity was PCR combining an 1-kb fragment upstream from the PUT1 verified through growth on YNB (2% glucose, 10 mM proline) start codon, the NEO marker, and an 1-kb fragment down- lacking leucine. stream from the PUT1 stop codon. Strain H99 genomic DNA Construction and screening of library for C. neoformans and plasmid pJAF1 were used as PCR templates (Fraser et al. proline catabolism defective mutants 2003). The construct was transformed into C. neoformans cells via particle bombardment and transformants selected An Agrobacterium-mediated mutagenesis library of C. neo- on YPD plates supplemented with G418. Deletion of PUT1 formans H99 was created as described previously (Idnurm was confirmed by diagnostic PCR and Southern blot (Southern et al. 2004). The majority of transformants harbored a single 2006). To complement the put1D mutant, the C. neoformans copy of randomly integrated T-DNA and were mitotically PUT1 gene, including an 1-kb promoter and terminator, was stable. To identify genes involved in proline catabolism, rep- amplified from genomic DNA using high-fidelity PCR, cloned lica spotting of the insertional library consisting of 12,000 into pCR2.1-TOPO (Invitrogen) to give pELYL1, and sequenced. clones was carried out on YNB supplemented with proline or The PUT1 fragment of pELYL1 was then subcloned into ammonium (10 mM each nitrogen source). Mutants that pCH233, creating the complementation construct pELYL5. grew on ammonium but failed to grow on proline-containing pELYL5 was subsequently linearized and biolistically transformed medium were selected for further analysis. The T-DNA insertion into the put1D mutant. Stable transformants were selected on site was mapped using GenomeWalker according to the YPD supplemented with nourseothricin, and complemented manufacturer’s instructions (Clontech). BLAST analysis of

Proline Catabolism in Cryptococcus 423 these sequences was performed to reveal which loci were at a density of 1 · 107 cells/ml. Twenty yeast cells of each affected. strain were randomly chosen for fluorescence microscopy analysis. Images were collected on a Zeiss Axioplan 2 imaging Quantitative real-time PCR confocal microscope using a 66· oil immersion Plan-Apochromat C. neoformans strains were grown in YNB supplemented objective. Fluorochromes were excited using a 514-nm Argon/2 with 10 mM of the specified nitrogen source and shaken laser for MitoSOX and a 543-nm HeNe1 laser for MitoTracker. at 30 for 16 hr. Overnight cultures were harvested, cell Differential interference contrast images were collected simul- pellets were frozen and lyophilized, total RNA was isolated taneously with the fluorescence images using the transmitted using TRIzol reagent (Invitrogen), and cDNA was generated light detector. Images were processed using AIM 4.2 LSM using the SuperscriptIII First-Strand Synthesis System (Invi- software. MitoSOX fluorescence was also analyzed using a trogen). Primers for PUT1 (CNAG_02049.2), PUT5 (CNAG_ BD Accuri C6 flow cytometry. The parameters for FACS were 02048.2), and PUT2 (CNAG_05602.2) were designed to set at excitation of 488 nm, and for emission the data were span exon–exon boundaries and tested to verify that they collected at FSC, SSC, and 675LP (FL3) channels. Cell debris bind specifically to cDNA but not to genomic DNA. Quanti- represented by distinct low forward and side scatter were tative real-time PCR (qRT-PCR) was performed using SYBR gated out for analysis. All experiments were done in tripli- Green Supermix (Applied Biosystems) and an Applied Bio- cate. One-way analysis of variance was performed using the systems 7900HT Fast Real Time PCR System with the follow- unpaired, two-tailed t-test in GraphPad Prism Version 5.0c. ing cycling conditions: denaturation at 95 for 10 min, P-values of ,0.05 were considered statistically significant. followed by amplification and quantification in 45 cycles at Growth and melanization assays 95 for 15 sec and 60 for 1 min, with melting-curve profiling at 95 for 2 min, 60 for 15 sec, and 95 for 15 sec. Dissoci- Starter C. neoformans cultures were prepared by growth in ation analysis confirmed the amplification of a single PCR YPD at 30 overnight with shaking, diluted to OD600nm =0.05 product for each primer pair and an absence of primer dimer inwater,andthenfurtherdiluted10-foldinseries.Eachdiluted formation. Relative gene expression was quantified using SDS cell suspension was spotted onto YPD and YNB supplemented software 1.3.1 (Applied Biosystems) based on the 22ΔΔCT with the specified nitrogen source or L-DOPA medium. Results method (Livak and Schmittgen 2001). The housekeeping ac- were imaged after 2–3 days incubation at 30 (nitrogen tin-encoding gene ACT1 was used as a control for normaliza- utilization assay), 30 and 37 (melanization assay), or 37 tion. One-way analysis of variance was performed using the (high-temperature growth assay). unpaired, two-tailed t-test in GraphPad Prism Version 5.0c. Capsule assays P-values of ,0.05 were considered statistically significant. C. neoformans strains were inoculated into RPMI supple- Oxidative and nitrosative stress assays mented with 10% fetal bovine serum (Gibco) and incubated Starter C. neoformans cultures were prepared by growth in at 37 for 2 days. To visualize the capsule, cells were stained YPD at 30 overnight with shaking and diluted to a concen- with India ink (Becton Dickinson) and observed under tration of 1 · 107 cells/ml in water. A 50-ml drop containing a ZEISS Axioplan 2 epifluorescent/light microscope. Pictures 5 · 105 cells was then inoculated into YNB [proline and glu- were taken with an Axiocam grayscale digital camera using tamate(10mMeachnitrogensource)]supplementedwiththe the AxioVision AC imaging software. stressors menadione (0– 0.003 mM), tert-Butyl hydroperoxide Murine inhalation model of cryptococcosis (0–0.3 mM), sodium nitrite (0–3mM),andS-nitrosoglutathione (GSNO) (0–9mM)andshakenat30 for an additional 16 hr. For murine virulence assays, C. neoformans were used to Growth density was measured by monitoring the turbidity infect 6-week-old female BALB/c mice by nasal inhalation with an optical density at a wavelength of 600 nm (OD600nm). (Cox et al. 2000). For every tested strain, 10 mice were each Each experimental condition was repeated in triplicate. One- inoculated with a 50-ml drop containing 5 · 105 cells. Mice way analysis of variance was performed using the unpaired, were weighed before infection and daily thereafter; animals two-tailed t-test in GraphPad Prism Version 5.0c. P-values of were killed using CO2 inhalation once their body weight had ,0.05 were considered statistically significant. decreased to 80% of the pre-infection weight or at 50 days post infection. In addition, brain, lungs, liver, and spleen Mitochondrial staining, confocal microscopy, and were harvested from killed put2D-infected mice; organs flow cytometry were weighed, homogenized, and plated onto YPD supple- C. neoformans strains were grown in YPD and shaken at 30 mented with 50 mg/ml chloramphenicol (Sigma) for selec- for 16 hr; thereafter, cells were washed and transferred to tive isolation of yeast colony forming units (CFUs). Survival YNB supplemented with 10 mM proline for a further 4 hr. was plotted against time, and P-values were calculated CellswerethenstainedwithMitoSOXRed(Invitrogen)atafi- by plotting a Kaplan–Meier survival curve and performing nal concentration of 5 mM or with MitoTracker Red CMXRos alog-rank(Mantel–Cox) test using Graphpad Prism Ver- (Invitrogen) at a final concentration of 25 nM for 15 min at sion 5.0c. P-values of ,0.05 were considered statistically 30. After staining, cells were washed and resuspended in PBS significant.

424 I. R. Lee et al. Ethics statement level, suggesting that the duplication event is not recent (Figure S2). In support of this hypothesis, analysis of the sequenced This study was carried out in strict accordance with the genomes of the C. neoformans var. neoformans strain JEC21 recommendations in the Australian Code of Practice for (estimated to have diverged from C. neoformans var. grubii the Care and Use of Animals for Scientific Purposes by the 24.5 million years ago) and the C. gattii strain R265 (es- National Health and Medical Research Council. The protocol timated to have diverged from C. neoformans var. grubii 49 was approved by the Molecular Biosciences Animal Ethics million years ago) revealed that the duplication is also present Committee (AEC) of The University of Queensland (AEC in these isolates, consistent with the duplication event preceding approval no. SCMB/008/11/UQ/NHMRC). Infection was per- the divergence of these species. formed under methoxyflurane anesthesia, and all efforts were made to minimize suffering through adherence to the Guide- Transcription of all the predicted proline catabolic lines to Promote the Wellbeing of Animals Used for Scientific genes of C. neoformans is regulated by Purposes as put forward by the National Health and Medical pathway-specific induction Research Council (Australia). To provide support for our bioinformatics-based prediction, we analyzed the transcriptional regulation of the putative Results proline catabolic genes using qRT-PCR on RNA extracted from wild-type H99 grown in minimal YNB medium supplemented Bioinformatic analyses reveal conserved and duplicated with a range of different sole nitrogen sources (Figure 2). genes predicted to be involved in proline utilization While only basal levels of expression of PUT1, PUT5, and in C. neoformans PUT2 were observed when cells were cultured in ammonium, While transcriptional studies relating to nitrogen metabolite glutamine, glutamate, asparagine, or uric acid, transcription repression-associated regulation of the C. neoformans proof these genes was significantly upregulated (.10-fold in- line oxidase and P5C dehydrogenase-encoding genes have crease) during growth in proline. This result indicates that been performed, these putative proline degradation path- the expression of PUT1, PUT5, and PUT2 is inducible only in way genes have not been properly characterized at the mo- the presence of proline, suggesting that these genes are indeed lecular level (Kmetzsch et al. 2011; Lee et al. 2011, 2012). proline metabolism-related genes. Nonetheless, the result pro- Unlike the genome of the model hemiascomycete S. cerevisiae vides no indication as to whether PUT1 or PUT5 is likely to be that contains a single proline oxidase-encoding gene, chro- the true functional proline oxidase ortholog required for proline mosome 6 of C. neoformans var. grubii strain H99 contains assimilation or whether both PUT1 and PUT5 are enzymatically two tandemly arrayed genes that are both predicted to en- active. code proline oxidases (Figure 1, A and B) (Lee et al. 2011). These two Cryptococcus paralogs were previously named Cross-species complementation provides evidence that PUT1 (CNAG_02049.2) and PUT5 (CNAG_02048.2), for pro- C. neoformans PUT2 is a true functional ortholog encoding P5C dehydrogenase, but gives no indication line utilization 1 and 5 (Kmetzsch et al. 2011; Lee et al. of the functionality of PUT1 and PUT5 2011). A similar scenario is observed in the model plant Arab- idopsis thaliana where two proline dehydrogenase isoforms We tested the enzymatic functions of C. neoformans PUT1, have been functionally characterized (Funck et al. 2010). In PUT5,andPUT2 by heterologous expression in S. cerevisiae contrast to what was seen for proline oxidase, the C. neofor- strains lacking either the proline oxidase or P5C dehydroge- mans H99 genome encodes only a single P5C dehydrogenase nase-encoding gene (Figure S3). These well-characterized on chromosome 14 (Figure 1, A and B). This Cryptococcus S. cerevisiae put1D and put2D mutant strains are unable to ortholog was previously designated PUT2 (CNAG_05602.2) utilize proline as a source of nitrogen for growth (Brandriss for proline utilization 2 (Lee et al. 2011). Thus, unlike the 1983; Wang and Brandriss 1986). Transformation of the model filamentous ascomycete Aspergillus nidulans,where S. cerevisiae put2D strain with a plasmid harboring the the proline catabolism-associated genes form a cluster within full-length cDNA of C. neoformans PUT2 enabled the strain a single chromosome, the C. neoformans proline oxidase and to catabolize proline, albeit to a poorer extent compared to P5C dehydrogenase genes are unlinked, like those of S. cerevisiae a full-length S. cerevisiae PUT2. This result indicates that (Figure S1) (Arst and MacDonald 1978). C. neoformans PUT2 encodes an enzyme with P5C dehydroge- The duplication of the putative proline oxidase gene is an nase activity. However, transformation of the S. cerevisiae interesting occurrence into which we sought to gain further put1D strain with plasmids harboring the full-length cDNA of insights. Both genomic DNA and cDNA of these genes were either C. neoformans PUT1 or PUT5 did not rescue the strains’ PCR-amplified and sequenced, verifying the predicted exon– proline utilization defect, whereas a full-length S. cerevisiae intron structure. This exercise demonstrated that the posi- PUT1 restored the proline oxidase function. Difficulties associ- tion of the three introns in PUT1 is precisely conserved in ated with complementing the S. cerevisiae put1D strain with PUT5, providing evidence that these genes are indeed paral- proline oxidase cDNA of other organisms have previously been ogs (Figure 1B). Using this sequence information, a compar- reported (Kiyosue et al. 1996). Hence, the enzymatic functions ison of Put1 and Put5 revealed 67% identity at the protein of C. neoformans PUT1 and PUT5 remain inconclusive.

Proline Catabolism in Cryptococcus 425 Figure 2 Transcription of all the proline catabolic enzyme-encoding genes is regulated by pathway-speciﬁc induction. cDNA from wild-type H99 grown in YNB supplemented with proline, ammonium, glutamine, glutamate, asparagine, or uric acid (10 mM each) was ampliﬁed via qRT-PCR using primers for the proline catabolic genes PUT1, PUT5, and PUT2, normalized against the control gene ACT1. Expression of PUT1, PUT5, and PUT2 was induced in the presence of proline but not in the presence of the other nitrogen sources (*P , 0.05, **P , 0.01, ***P , 0.001, ****P , 0.0001). Error bars represent standard errors across three biological replicates.

Reverse genetics approach reveals that PUT5-encoded of 92 and 98%, respectively, suggesting that both these pro- proline oxidase and PUT2-encoded P5C dehydrogenase line oxidases are likely in the correct subcellular localization are required for proline utilization in C. neoformans to support proline catabolism (Claros and Vincens 1996). We therefore adopted another approach to characterize the Like Put5, Put2 is also essential for proline assimilation. D D D functions of the C. neoformans putative proline catabolic Complementation of the put5 , double put1 put5 , and D genes. Deletion mutants each lacking PUT1, PUT5,orPUT2 put2 mutants subsequently restored the wild-type pheno- were created via homologous recombination in strain H99. In type, verifying the proline utilization functions of Put5 and addition, a double mutant lacking both PUT1 and PUT5 was Put2 (Figure S4). constructed as we reasoned that the single put1D or put5D mutant might still possess an active proline oxidase isoform. Forward genetics approach involving This is unlike the case for the put2D mutant where degra- Agrobacterium-mediated insertional mutagenesis did not reveal any additional proline catabolic enzymes dation of proline to glutamate, a prerequisite for recycling of in C. neoformans nitrogen, is expected to be completely blocked. The generated proline oxidase deletion mutants (put1D, put5D, dou- P5C is thought to form a pH-dependent, nonenzymatic ble put1D put5D) as well as the P5C dehydrogenase deletion equilibrium with GSA. Recent findings in uric acid catabolism mutant (put2D) were all viable; the proline oxidase deletion raise the question of whether this dogma should be accepted mutants’ growth appears indistinguishable from wild-type prima facie; in addition to urate oxidase, two new biological on rich undefined YPD medium, whereas the put2D mutant enzymes (HIU hydrolase and OHCU decarboxylase) were exhibited a mild growth defect (see Put2 is required for recently discovered to play a role in the conversion of uric optimal production of the major virulence factors). acid to allantoin (Ramazzina et al. 2006). To provide unbi- On YNB-defined medium supplemented with proline as ased support for the bioinformatics and phenotypic-based the sole nitrogen source or sole nitrogen and carbon source model of C. neoformans proline catabolism, a T-DNA inser- (i.e., with or without the addition of glucose as a source of tional library of strain H99 was created using Agrobacterium carbon), the put1D mutant displayed wild-type growth while tumefaciens-mediated transformation. Insertional mutagene- the put5D, double put1D put5D,andput2D mutants were un- sis is a powerful tool for identifying novel genes and their able to proliferate (Figure 3). This result indicates that, unlike functions, and the frequency of random integration is very S. cerevisiae that can assimilate only proline as a nitrogen high using this transkingdom DNA delivery approach. In an source, C. neoformans follows the ability of fungi such as attempt to locate as-yet-identified genetic loci involved in A. nidulans in being able to utilize proline as both a nitrogen proline utilization, we replica-spotted 12,000 purified individ- and a carbon source (Brandriss and Magasanik 1979; ual transformants onto YNB supplemented with proline or Sophianopoulou et al. 1993). Importantly, C. neoformans ammonium (control) as the sole nitrogen source. Only a single Put1 does not play a role in proline utilization under this mutant, 65B8, failed to grow on proline but grew on ammo- specific condition; instead, it is Put5 that appears to be the nium. We then mapped the T-DNA insertion site of the 65B8 major proline oxidase responsible for proline degradation. strain and found that its disrupted gene encodes P5C dehy- Nonetheless, the MitoProt computer-based prediction pro- drogenase. Hence, our sequence analysis did not reveal any gram clearly indicates a characteristic mitochondrial import additional proline catabolic genes beyond what has been bio- signal at the N terminus of Put1 and Put5 with a probability informatically predicted or phenotypically characterized.

426 I. R. Lee et al. were assayed as each compound acts on a different cellular target.

Spectrophotometric OD600nm readings revealed that tolerance against menadione and sodium nitrite is equivalent between the wild-type strain and proline oxidase deletion mutants (Figure S5). Interestingly, the put5D and double put1D put5D mutants exhibit decreased resistance against tert-Butyl hydroperoxide and GSNO as compared to the wild-type and put1D strains (Figure 4). This suggests that lack of proline oxidase activity may potentially impair the C. neoformans cellular antioxidant system in handling the stresses generated by ROS/RNS-producing reagents. Therefore, unlike the scenario in S. cerevisiae, proline accumulation in C. neoformans cells may not necessarily result in enhanced resistance against oxidative or nitrosative stresses. Instead, the predicted occurrence of unregulated proline accumulation is likely to be slightly inhibitory to general cell growth, as observed by the put5D and double put1D put5D mutants’ slower growth rate relative to the wild-type and put1D strains in control cultures (YNB plus proline and glutamate) devoid of any stressors.

C. neoformans put2D mutant accumulates excessive mitochondrial ROS in the presence of proline In S. cerevisiae put2D mutants, exposure to proline has been shown to result in growth inhibition and an increase in intracellular ROS levels (Deuschle et al. 2001; Nomura and PUT5 PUT2 Figure 3 -encoded proline oxidase and -encoded P5C dehy- Takagi 2004). Through external supplementation of proline, drogenase are required for proline assimilation. Tenfold spot dilution assays for nitrogen or nitrogen and carbon source utilization showed that we hypothesized that the predicted P5C-proline cycling the put1D mutant exhibited wild-type growth on YNB supplemented with would provide a constant substrate for proline oxidase-linked 10 mM proline (with or without 2% glucose as a carbon source). In FAD reduction in the C. neoformans put2D mutant, resulting contrast, the put5D, double put1D put5D, and put2D mutants were un- in an increased rate of electron transfer to the mtETC and to able to proliferate on proline as the sole nitrogen or carbon source. The O and leading to elevated superoxide production. Genera- ammonium-supplemented YNB plates served as a control to demonstrate 2 that, unlike proline, ammonium can be utilized by C. neoformans only as tion and accumulation of ROS is a critical feature of pro- a nitrogen source but not as a carbon source. grammed cell death in animals, plants, and fungi, and this ROS-induced oxidative stress may be a factor causing the general growth retardation phenotype in the C. neoformans D D D C. neoformans put5 and double put1 put5 mutants put2D mutant (Jabs 1999). display reduce tolerance against certain oxidative and To determine whether the put2D mutant has abnormal nitrosative stressors intrinsic ROS levels, C. neoformans strains were first grown In S. cerevisiae, metabolically engineered proline-accumulated overnight in rich YPD medium after which they were trans- cells generated through enhancing the biosynthesis pathway ferred to YNB supplemented with proline for an additional and/or concomitant silencing of the catabolic pathway are 4 hr. Strains were then stained with the fluorogenic MitoSOX more tolerant to a wide array of stresses (Takagi et al. Red probe used to selectively detect mitochondrial superoxide 2000, 2005; Morita et al. 2002). For example, S. cerevisiae ions (the predominant ROS in mitochondria) of live cells. In put1D mutants display increased resistance against oxidative a subset (1/3 of the entire population) of put2D cells, the damage (Chen et al. 2006). To promote intracellular proline extent of ROS formation was by far higher than that generated accumulation in C. neoformans, the proline oxidase deletion in the wild-type or proline oxidase deletion mutant strains as mutants were grown in YNB supplemented with proline and seen by a more intense red fluorescence caused by oxidation of glutamate as the sole nitrogen sources. Glutamate is the end the MitoSOX reagent by superoxide (Figure 5A). Because con- product of proline catabolism and the precursor for proline focal microscopic imaging only allows semiquantitative detec- biosynthesis; in S. cerevisiae, PRO3-encoded P5C reductase tion of superoxide generation in relatively few live cells, we that converts P5C to proline is constitutively expressed (Brandriss also simultaneously employed flow cytometry that enables and Falvey 1992). Concurrently, these cultures were subjected a quantitative measure of a larger population of live cells. to various oxidative (menadione, tert-Butyl hydroperoxide) and The flow cytometry analysis also revealed two major distinct nitrosative (sodium nitrite, GSNO) stresses; numerous reactive populations of put2D cells, with a combined 3.9 6 0.3-fold oxygen/nitrogen species (ROS/RNS)-generating chemicals increase of mean MitoSOX fluorescence intensity relative to

Proline Catabolism in Cryptococcus 427 Figure 4 The put5D and double put1D put5D mutants are less tolerant against the oxidative stressor tert-Butyl hydroperoxide and nitrosative stressor GSNO. Wild-type H99 and proline oxidase deletion mutant strains were grown in YNB [proline and glutamate (10 mM each)] supplemented with (A) 0–0.3 mM tert-Butyl hydroperoxide or (B) 0–9 mM GSNO. The put5D and double put1D put5D mutants exhibited a signiﬁcantly lower growth percentage relative to the unstressed control when subjected to these individual stressors as compared to the wild-type or put1D strains (**P , 0.01, ***P , 0.001). Growth percentage relative to unstressed control is deﬁned by dividing each stressed strain’s

growth in OD600 with the same strain’s growth in OD600 when unstressed. Error bars represent standard errors across three biological replicates. the wild-type strain (WT vs. put2D, P = 0.0002) (Figure 5B tested whether the proline catabolic enzymes influence the and Figure S6). It remains to be determined why a subpop- expression of these classical virulence traits (Figure 6 and ulation of put2D cells displayed wild-type fluorescence in- not shown). No changes in the ability to grow at high tem- tensity while another subpopulation exhibited significantly perature or in the production of melanin and capsule were enhanced fluorescence. A single population generating wild- observed in the put1D, put5D, and double put1D put5D type levels of superoxide production was restored in the mutants, indicating that proline oxidase is dispensable for complemented put2D + PUT2 strain, verifying the role of virulence factor expression. In contrast, the mild growth de- P5C dehydrogenase in controlling mitochondrial ROS levels fect of the put2D mutant that may be caused by general in C. neoformans (Figure S6). Uncoupled coordination of growth retardation or death of a subset of cells was slightly proline oxidase and P5C dehydrogenase activities therefore exacerbated at 37 on YPD medium. The put2D mutant also leads to an imbalance in ROS homeostasis. exhibited a subtle but reproducible decrease in melanin pro- Given that mitochondria are organelles known to play an duction at 30 on L-DOPA medium, with this melanization important role in apoptosis, we questioned whether this defect slightly more pronounced at 37. No change in cap- anomaly in ROS levels of the put2D mutant impairs mito- sule size, however, was observed in the put2D cells when chondrial function. Various aspects of mitochondrial dys- induced in serum. Virulence factor expression was restored function can be detected by examining for perturbation in to wild-type levels in the complemented put2D + PUT2 membrane potential using MitoTracker dyes such as CMXRos. strain (Figure S8). Collectively, P5C dehydrogenase is crucial Confocal microscopy showed that all the proline catabolic for optimal production of at least two of the well-known deletion mutants, including the put2D strain, were able to virulence components of C. neoformans. retain this fluorescence dye, suggesting that mitochondrial D membrane potential is unperturbed and that the respiratory The put2 mutant is severely compromised for virulence in a murine inhalation model of cryptococcosis system is still intact in all these mutants (Figure S7). A recent study in S. cerevisiae has demonstrated that P5C Put2 is required for optimal production of the major directly inhibits the mitochondrial respiration, reiterating virulence factors our hypothesis that proline catabolism is likely to be important The ability of C. neoformans to successfully infect a host is for C. neoformans virulence (Nishimura et al. 2012). Given largely dependent on several well-established virulence fac- that our overarching interest in C. neoformans lies in better tors, namely, ability to grow at mammalian body tempera- understanding its pathogenicity in a mammalian host, we ture as well as synthesis of the antioxidant pigment melanin investigated the role of the proline catabolic enzymes during and antiphagocytic polysaccharide capsule. Therefore, we infection in vivo (Figure 7). Mice intranasally infected with

428 I. R. Lee et al. Figure 5 The put2D mutant generates excessive superoxide ions in the mitochondria when exposed to proline. Wild-type H99 and proline catabolic deletion mutant strains were briefly subjected to culture in YNB supplemented with 10 mM proline and stained with MitoSOX, and the accumulation of ROS in the mitochondria was assessed by fluorescence microscopy and flow cytometry. (A) Representative confocal images showing a subpopulation of put2D cells generate enhanced MitoSOX fluorescence compared to the wild-type or proline oxidase deletion mutant strains. Bar, 10 mm. (B) Repre- sentative histograms of flow cytometry showing two major distinct populations of put2D cells with one subpopulation exhibiting wild-type MitoSOX fluorescence and another subpopulation producing enhanced fluorescence. Overall, the put2D mutant generated a significant increase in mean fluorescence intensity of oxidized MitoSOX relative to the wild-type or complemented put2D + PUT2 strains (Figure S6). Unstained C. neoformans cells were used as a negative control. the put1D, put5D, and double put1D put5D mutants suc- put2D, P , 0.0001). Infection of mice with the complemented cumbed to infection at approximately the same rate as mice put2D + PUT2 strain restored wild-type virulence (killing infected with the wild-type strain (killing occurred between occurred between 16 and 23 days post infection; median 16 and 23 days post infection; median survival of mice = survival of mice = 18.5 days), verifying the role of P5C de- 18.0 days), indicating that proline oxidase is not required hydrogenase in pathogenesis. for pathogenesis. In stark contrast, the put2D mutant was To more precisely characterize the effect of the loss of unable to kill any mice even after 50 days, indicating that P5C dehydrogenase on C. neoformans survival and dissemi- P5C dehydrogenase is critical for pathogenesis (WT vs. nation inside the host body, we examined the fungal burden

Proline Catabolism in Cryptococcus 429 that serves as a nitrogen donor for all other cellular nitrogen- containing compounds (for example, amino acid biosynthesis), or it can be converted to a-ketoglutarate by NAD-dependent glutamate dehydrogenase that feeds into the Krebs/TCA cycle for energy generation (Cooper 1982; Magasanik 1992). The proline oxidase gene duplication that occurred millions of years ago has been retained in both C. neoformans and C. gattii. From an evolutionary perspective, this implies that the selective pressure that Cryptococcus faces in the environment or during infection may have resulted in the retention of both paralogs. Such an advantage usually increases the overall fitness of an organism and has previously been described in the model plant A. thaliana where two functional proline dehydrogenases with nonredundant but partially overlapping functions exist; ProDH1 is the main proline utilization enzyme while ProDH2 is believed to play a role in proline homeostasis under salt stress (Funck et al. 2010). In C. neoformans, Put5 is the major proline oxidase responsible for proline assimilation as well as for protection against certain oxidative and nitrosative stresses. Sensitivity of microor- Figure 6 P5C dehydrogenase is needed for optimal growth at high temperature and for optimal production of melanin. (A) Tenfold spot dilution ganisms to the toxic activity of oxidants is of particular assays on YPD medium demonstrated that the put2D mutant exhibited importance, as one of the primary mechanisms of host defense a slightly slower growth rate compared to wild type at 30, and this against microbial infection is oxidative burst. Regardless of the growth defect was exacerbated at 37. (B) Tenfold spot dilution assays abated nutrient acquisition and stress-sensitive phenotypes of put2D on L-DOPA medium showed that the mutant produced slightly less the put5D mutant in vitro, Put5 is dispensable for virulence melanin than wild type at 30 and that this melanization defect was exacerbated at 37. in vivo. This suggests that, in the absence of proline catabolism, other nitrogen and carbon sources can still be readily of various organs at the endpoint of the experiment. Despite acquired by C. neoformans within the host and that the activ- showing no sign of disease at 50 days post infection, the ity of other existing stress response pathways of C. neoformans put2D-infected mice had a low number of CFUs in 9 of 10 is sufficient to confer protection against oxidative and nitro- tested lung samples and an even lower number of CFUs in 3 sative insults by host immune cells. of 10 tested brain samples when selected on YPD supple- The role of the other proline oxidase paralog, PUT1,remains mented with chloramphenicol. These CFUs were confirmed ambiguous, and our attempted cross-species complementation to express resistance to G418 and were unable to utilize experiment failed to provide any insights into its enzymatic proline as a nitrogen source, consistent with C. neoformans functionality. The failure of both the C. neoformans PUT1 put2::NEO cells being isolated from the mice. No cells were and PUT5 cDNA to complement the S. cerevisiae put1D strain detected in the liver or spleen of each of these 10 mice. could be due to multiple factors, including codon usage bias Thus, the majority of the killed put2D-inoculated mice did or divergence of the C. neoformans proline oxidases’ mitochon- not clear the cryptococcal infection; dormant-lying put2D drial targeting signal peptides, resulting in incorrect subcellular cells were primarily located in the lungs, the initial site of localization of the expressed proteins. While we created fusion infection. Nevertheless, compared to the robust proliferation genes to engineer the S. cerevisiae PUT1 promotor and ter- and widespread dissemination characteristic of a typical minator sequences to flank the C. neoformans PUT1 and PUT5 wild-type H99 infection (Ormerod et al. 2013), the put2D ORFs, the predicted mitochondrial transit peptide of S. cer- mutant is attenuated in these virulence capacities. evisiae Put1 was not incorporated into the N terminus of C. neoformans Put1 or Put5 as the exchange of sequences within an ORF may potentially interfere with the function- Discussion ality test. Therefore, it remains open to speculation why the C. neoformans has emerged as one of the most medically native C. neoformans proline oxidase genes do not enable the important pathogens of the immunocompromised population expression of sufficient amounts of active and/or correctly (Heitman 2010). To establish infection in a host, C. neoformans targeted protein in S. cerevisiae, whereas the C. neoformans needs to sense and adapt to low-nitrogen and carbon con- P5C dehydrogenase gene produced a functional protein with centrations in the human lungs, brain, and other anatomical its native mitochondrial transit peptide. sites. The multifunctional amino acid, proline, can act as C. neoformans PUT2-encoded P5C dehydrogenase is the a nitrogen or carbon source for C. neoformans and is a poten- second enzyme of the proline degradation pathway that is tial source of nutrients during infection. For this to occur, required for proline utilization as well as for control of mi- proline needs to be degraded into the more usable glutamate tochondrial ROS production. All aerobic organisms generate

430 I. R. Lee et al. Figure 7 The put2D mutant is avirulent during infection of a murine host. Ten mice were each intranasally infected with either 5 · 105 cells of (A) wild-type H99, put1D, put1D +PUT1, (B) put5D, put5D +PUT5, (C) double put1D put5D, put1D put5D + PUT1 PUT5, and (D) put2D or put2D +PUT2, and survival was monitored daily. Mice infected with the put1D, put1D +PUT1, put5D, put5D +PUT5, double put1D put5D, put1D put5D + PUT1 PUT5, and put2D +PUT2 strains progressed to morbidity as quickly as mice infected with the wild-type strain. In contrast, mice infected with the put2D strain remained healthy throughout the experiment.

ROS as metabolic by-products primarily through aerobic res- P5C/GSA (Krishnan and Becker 2006). In plants lacking piration; ROS can damage DNA, proteins, and lipids, caus- P5C dehydrogenase, the inability to oxidize P5C to glutamate ing cytotoxicity. Multiple lines of evidence suggest that ROS and, in turn, to a-ketoglutarate may also cause a reduction in can also regulate apoptosis, a morphologically distinct form mitochondrial NADPH production, which can affect the re- of programmed cell death. In fact, a growing list of fungi generation of reduced glutathione by NADPH-dependent have been demonstrated to exhibit stress-induced cell death glutathione reductase and slow down the ascorbate/gluta- such as in the model ascomycetes S. cerevisiae and A. nidulans, thione cycle, leading to enhanced mitochondrial ROS levels as well as in the fungal pathogens Candida albicans, Colleto- (Moller 2001). In C. neoformans, P5C dehydrogenase is ad- trichum trifolii,andC. neoformans (Madeo et al. 1997; Cheng ditionally required for proper expression of the major virulence et al. 2003; Phillips et al. 2003; Chen and Dickman 2005; factors. Accordingly, yeast mutants lacking P5C dehydrogenase Ikeda and Sawamura 2008; Semighini et al. 2011). are profoundly attenuated in virulence during a murine inha- In animals, plants, and probably in the fungus C. neofor- lation model of cryptococcosis. To our knowledge, this is the mans (as we have suggestive evidence here), impairment of first report recognizing P5C dehydrogenase as an important P5C dehydrogenase activity that causes intense P5C-proline mediator that enables pathogenesis of a microorganism. cycling or proline oxidase overexpression leads to ROS ac- In conclusion, proline oxidases have been shown to cumulation and induction of apoptosis in cells (Maxwell and potentiate ROS-mediated cell death in various kingdoms of Davis 2000; Donald et al. 2001; Yoon et al. 2004; Phang life, and this paradigm likely extends to the basidiomycete et al. 2008; Miller et al. 2009). In fact, the proline oxidase pathogen C. neoformans. The role of P5C dehydrogenase in gene is a target of the pro-apoptotic and tumor suppressor suppressing proline oxidase–FAD complex activity that, in protein p53 in humans (Polyak et al. 1997). While definitive turn, controls apoptosis might therefore be an underappreci- proof underlying the mechanism of ROS accumulation is ated ancient function of the enzyme that has been conserved lacking for most proline oxidases, mounting evidence points throughout evolution. to the ability of the enzyme to load electrons into the mtETC or directly into O2, causing elevated production of superox- Acknowledgments ide ions. P5C dehydrogenase therefore plays a vital role in eukaryotes by irreversibly removing P5C from potential P5C- We acknowledge a University of Queensland International proline cycling, resulting in an overall reduction of ROS pro- Research Tuition Award and a University of Queensland duction from electron runoff. Research Scholarship awarded to I.R.L. and the National Interestingly, certain prokaryotes have evolved strategies Health and Medical Research Council CDA 569673 Grant to specifically avoid P5C/GSA accumulation by coupling awarded to J.A.F. proline oxidase and P5C dehydrogenase activities. For example, in bacterial species such as Bradyrhizobium japonicum and Helicobacter pylori, the proline catabolic enzymatic activ- Literature Cited ities are fused into a unique bifunctional enzyme to channel Altschul,S.F.,W.Gish,W.Miller,E.W.Myers,andD.J.Lipman, the reaction intermediate and thus prevent the release of 1990 Basic local alignment search tool. J. Mol. Biol. 215: 403–410.

Proline Catabolism in Cryptococcus 431 Arst, H. N. Jr., and D. W. MacDonald, 1978 Reduced expression of Fraser,J.A.,R.L.Subaran,C.B.Nichols,andJ.Heitman, a distal gene of the prn gene cluster in deletion mutants of 2003 Recapitulation of the sexual cycle of the primary fungal path- Aspergillus nidulans: genetic evidence for a dicistronic messen- ogen Cryptococcus neoformans var. gattii: implications for an out- ger in an eukaryote. Mol. Gen. Genet. 163: 17–22. break on Vancouver Island, Canada. Eukaryot. Cell 2: 1036–1045. Atkinson, D. E., 1977 Cellular Energy Metabolism and Its Regula- Funck, D., S. Eckard, and G. Muller, 2010 Non-redundant function. Academic Press, New York. tions of two proline dehydrogenase isoforms in Arabidopsis. Axelrod,J.D.,J.Majors,andM.C.Brandriss,1991 Proline-independent BMC Plant Biol. 10: 70. binding of PUT3 transcriptional activator protein detected by footprint- Gietz, R. D., and R. H. Schiestl, 2007 High-efﬁciency yeast trans- – ing in vivo. Mol. Cell. Biol. 11: 564 567. formation using the LiAc/SS carrier DNA/PEG method. Nat. Brachmann, C. B., A. Davies, G. J. Cost, E. Caputo, J. Li et al., Protoc. 2: 31–34. 1998 Designer deletion strains derived from Saccharomyces cere- Gietz, R. D., and A. Sugino, 1988 New yeast-Escherichia coli shut- visiae S288C: a useful set of strains and plasmids for PCR-mediated tle vectors constructed with in vitro mutagenized yeast genes – gene disruption and other applications. Yeast 14: 115 132. lacking six-base pair restriction sites. Gene 74: 527–534. Brandriss, M. C., 1983 Proline utilization in Saccharomyces cerevisiae: Hare, P. D., and W. A. Cress, 1997 Metabolic implications of stess- analysis of the cloned PUT2 gene. Mol. Cell. Biol. 3: 1846–1856. induced proline accumulation in plants. Plant Growth Regul. 21: Brandriss, M. C., and D. A. Falvey, 1992 Proline biosynthesis in 79–102. Saccharomyces cerevisiae: analysis of the PRO3 gene, which enc- Heitman, J., 2010 Cryptococcus: From Human Pathogen to Model odes delta 1-pyrroline-5-carboxylate reductase. J. Bacteriol. Yeast. ASM Press, Washington, DC. 174: 3782–3788. Huang, H. L., and M. C. Brandriss, 2000 The regulator of the Brandriss, M. C., and B. Magasanik, 1979 Genetics and physiology yeast proline utilization pathway is differentially phosphory- of proline utilization in Saccharomyces cerevisiae:mutationcaus- ing constitutive enzyme expression. J. Bacteriol. 140: 504–507. lated in response to the quality of the nitrogen source. Mol. Cell. – Carpenter, J. F., and J. H. Crowe, 1988 Modes of stabilization of a pro- Biol. 20: 892 899. tein by organic solutes during desiccation. Cryobiology 25: 459–470. Idnurm, A., J. L. Reedy, J. C. Nussbaum, and J. Heitman, Chen, C., and M. B. Dickman, 2005 Proline suppresses apoptosis 2004 Cryptococcus neoformans virulence gene discovery – in the fungal pathogen Colletotrichum trifolii. Proc. Natl. Acad. through insertional mutagenesis. Eukaryot. Cell 3: 420 429. Sci. USA 102: 3459–3464. Ignatova, Z., and L. M. Gierasch, 2006 Inhibition of protein ag- Chen, C., S. Wanduragala, D. F. Becker, and M. B. Dickman, gregation in vitro and in vivo by a natural osmoprotectant. Proc. 2006 Tomato QM-like protein protects Saccharomyces cerevi- Natl. Acad. Sci. USA 103: 13357–13361. siae cells against oxidative stress by regulating intracellular pro- Ikeda, R., and K. Sawamura, 2008 Bacterial and H2O2 stress-in- line levels. Appl. Environ. Microbiol. 72: 4001–4006. duced apoptosis-like events in Cryptococcus neoformans. Res. Cheng, J., T. S. Park, L. C. Chio, A. S. Fischl, and X. S. Ye, Microbiol. 159: 628–634. 2003 Induction of apoptosis by sphingoid long-chain bases in Jabs, T., 1999 Reactive oxygen intermediates as mediators of pro- Aspergillus nidulans. Mol. Cell. Biol. 23: 163–177. grammed cell death in plants and animals. Biochem. Pharmacol. Claros, M. G., and P. Vincens, 1996 Computational method to 57: 231–245. predict mitochondrially imported proteins and their targeting Jauniaux, J. C., and M. Grenson, 1990 GAP1, the general amino sequences. Eur. J. Biochem. 241: 779–786. acid permease gene of Saccharomyces cerevisiae. Nucleotide se- Cooper, T. G., 1982 Nitrogen metabolism in Saccharomyces cere- quence, protein similarity with the other bakers yeast amino visiae, pp. 39–99 in The Molecular Biology of the Yeast Saccharo- acid permeases, and nitrogen catabolite repression. Eur. J. Bio- myces, edited by J. N. Strathern, E. W. Jones, and J. R. Broach. chem. 190: 39–44. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Kaul, S., S. S. Sharma, and I. K. Mehta, 2008 Free radical scav- Cox, G. M., J. Mukherjee, G. T. Cole, A. Casadevall, and J. R. enging potential of L-proline: evidence from in vitro assays. Perfect, 2000 Urease as a virulence factor in experimental Amino Acids 34: 315–320. – cryptococcosis. Infect. Immun. 68: 443 448. Kiyosue, T., Y. Yoshiba, K. Yamaguchi-Shinozaki, and K. Shinozaki, Daugherty, J. R., R. Rai, H. M. el Berry, and T. G. Cooper, 1996 A nuclear gene encoding mitochondrial proline dehydro- 1993 Regulatory circuit for responses of nitrogen catabolic genase, an enzyme involved in proline metabolism, is upregu- gene expression to the GLN3 and DAL80 proteins and nitrogen lated by proline but downregulated by dehydration in catabolite repression in Saccharomyces cerevisiae. J. Bacteriol. Arabidopsis. Plant Cell 8: 1323–1335. 175: 64–73. Kmetzsch, L., C. C. Staats, E. Simon, F. L. Fonseca, D. L. Oliveira Davidson, R. C., J. R. Blankenship, P. R. Kraus, M. de Jesus Berrios, et al., 2011 The GATA-type transcriptional activator Gat1 reg- C. M. Hull et al., 2002 A PCR-based strategy to generate in- ulates nitrogen uptake and metabolism in the human pathogen tegrative targeting alleles with large regions of homology. Mi- Cryptococcus neoformans. Fungal Genet. Biol. 48: 192–199. crobiology 148: 2607–2615. Krishnan, N., and D. F. Becker, 2006 Oxygen reactivity of PutA Deuschle, K., D. Funck, H. Hellmann, K. Daschner, S. Binder et al., 2001 A nuclear gene encoding mitochondrial Delta-pyrroline- from Helicobacter species and proline-linked oxidative stress. J. – 5-carboxylate dehydrogenase and its potential role in protection Bacteriol. 188: 1227 1235. from proline toxicity. Plant J. 27: 345–356. Lee, I. R., E. W. Chow, C. A. Morrow, J. T. Djordjevic, and J. A. Deuschle, K., D. Funck, G. Forlani, H. Stransky, A. Biehl et al., Fraser, 2011 Nitrogen metabolite repression of metabolism 2004 The role of [Delta]1-pyrroline-5-carboxylate dehydroge- and virulence in the human fungal pathogen Cryptococcus neo- nase in proline degradation. Plant Cell 16: 3413–3425. formans. Genetics 188: 309–323. Donald, S. P., X. Y. Sun, C. A. Hu, J. Yu, J. M. Mei et al., Lee, I. R., J. W. Lim, K. L. Ormerod, C. A. Morrow, and J. A. Fraser, 2001 Proline oxidase, encoded by p53-induced gene-6, cata- 2012 Characterization of an Nmr homolog that modulates lyzes the generation of proline-dependent reactive oxygen spe- GATA factor-mediated nitrogen metabolite repression in Crypto- cies. Cancer Res. 61: 1810–1815. coccus neoformans. PLoS ONE 7: e32585. Dufait, R., R. Velho, and C. De Vroey, 1987 Rapid identiﬁcation of Livak, K. J., and T. D. Schmittgen, 2001 Analysis of relative gene the two varieties of Cryptococcus neoformans by D-proline assim- expression data using real-time quantitative PCR and the 2 ilation. Mykosen 10: 483. (-Delta Delta C(T)). Methods 25: 402–408.

432 I. R. Lee et al. Madeo, F., E. Frohlich, and K. U. Frohlich, 1997 A yeast mutant Polyak, K., Y. Xia, J. L. Zweier, K. W. Kinzler, and B. Vogelstein, showing diagnostic markers of early and late apoptosis. J. Cell 1997 A model for p53-induced apoptosis. Nature 389: 300– Biol. 139: 729–734. 305. Magasanik, B., 1992 Regulation of nitrogen utilization, pp. 283–317 in Ramazzina, I., C. Folli, A. Secchi, R. Berni, and R. Percudani, The Molecular and Cellular Biology of the Yeast Saccharomyces: Gene 2006 Completing the uric acid degradation pathway through Expression,editedbyE.W.Jones,J.R.Pringle,andJ.R.Broach. phylogenetic comparison of whole genomes. Nat. Chem. Biol. 2: Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 144–148. Maxwell, S. A., and G. E. Davis, 2000 Differential gene expression Rudolph, A. S., and J. H. Crowe, 1985 Membrane stabilization in p53-mediated apoptosis-resistant vs. apoptosis-sensitive tu- during freezing: the role of two natural cryoprotectants, treha- mor cell lines. Proc. Natl. Acad. Sci. USA 97: 13009–13014. lose and proline. Cryobiology 22: 367–377. Mezl, V. A., and W. E. Knox, 1976 Properties and analysis of Sambrook, J., E. F. Fritsch, and T. Maniatis, 1989 Molecular Clon- a stable derivative of pyrroline-5-carboxylic acid for use in met- ing: A Laboratory Manual. Cold Spring Harbor Laboratory Press, abolic studies. Anal. Biochem. 74: 430–440. Cold Spring Harbor, NY. Miller, G., A. Honig, H. Stein, N. Suzuki, R. Mittler et al., Semighini, C. P., A. F. Averette, J. R. Perfect, and J. Heitman, 2009 Unraveling delta1-pyrroline-5-carboxylate-proline cycle 2011 Deletion of Cryptococcus neoformans AIF ortholog pro- in plants by uncoupled expression of proline oxidation enzymes. motes chromosome aneuploidy and fluconazole-resistance in J. Biol. Chem. 284: 26482–26492. a metacaspase-independent manner. PLoS Pathog. 7: e1002364. Missall, T. A., J. M. Moran, J. A. Corbett, and J. K. Lodge, Sophianopoulou, V., T. Suarez, G. Diallinas, and C. Scazzocchio, 2005 Distinct stress responses of two functional laccases in 1993 Operator derepressed mutations in the proline utilisa- Cryptococcus neoformans are revealed in the absence of the tion gene cluster of Aspergillus nidulans. Mol. Gen. Genet. thiol-specific antioxidant Tsa1. Eukaryot. Cell 4: 202–208. 236: 209–213. Moller, I. M., 2001 PLANT MITOCHONDRIA AND OXIDATIVE Southern, E., 2006 Southern blotting. Nat. Protoc. 1: 518–525. STRESS: electron transport, NADPH turnover, and metabolism Stanbrough, M., and B. Magasanik, 1996 Two transcription fac- of reactive oxygen species. Annu. Rev. Plant Physiol. Plant Mol. tors, Gln3p and Nil1p, use the same GATAAG sites to activate Biol. 52: 561–591. the expression of GAP1 of Saccharomyces cerevisiae. J. Bacteriol. Morita, Y., S. Nakamori, and H. Takagi, 2002 Effect of proline and 178: 2465–2468. arginine metabolism on freezing stress of Saccharomyces cerevi- Takagi, H., K. Sakai, K. Morida, and S. Nakamori, 2000 Proline siae. J. Biosci. Bioeng. 94: 390–394. accumulation by mutation or disruption of the proline oxidase Ngamskulrungroj, P., Y. Chang, J. Roh, and K. J. Kwon-Chung, gene improves resistance to freezing and desiccation stresses in 2012 Differences in nitrogen metabolism between Cryptococ- Saccharomyces cerevisiae. FEMS Microbiol. Lett. 184: 103–108. cus neoformans and C. gattii, the two etiologic agents of Cryp- Takagi, H., M. Takaoka, A. Kawaguchi, and Y. Kubo, 2005 Effect tococcosis. PLoS ONE 7: e34258. of L-proline on sake brewing and ethanol stress in Saccharomy- Nishimura, A., R. Nasuno, and H. Takagi, 2012 The proline me- ces cerevisiae. Appl. Environ. Microbiol. 71: 8656–8662. tabolism intermediate Delta1-pyrroline-5-carboxylate directly Thompson, J. D., D. G. Higgins, and T. J. Gibson, 1994 CLUSTAL inhibits the mitochondrial respiration in budding yeast. FEBS W: improving the sensitivity of progressive multiple sequence Lett. 586: 2411–2416. alignment through sequence weighting, position-specific gap Nomura, M., and H. Takagi, 2004 Role of the yeast acetyltrans- penalties and weight matrix choice. Nucleic Acids Res. 22: ferase Mpr1 in oxidative stress: regulation of oxygen reactive 4673–4680. species caused by a toxic proline catabolism intermediate. Proc. Vandenbol, M., J. C. Jauniaux, and M. Grenson, 1989 Nucleotide Natl. Acad. Sci. USA 101: 12616–12621. sequence of the Saccharomyces cerevisiae PUT4 proline-perme- Ormerod, K. L., C. A Morrow, E. W. Chow, I. R. Lee, S. D. Arras ase-encoding gene: similarities between CAN1, HIP1 and PUT4 et al., 2013 Comparative genomics of serial isolates of Crypto- permeases. Gene 83: 153–159. coccus neoformans reveals gene associated with carbon utiliza- Verbruggen, N., X. J. Hua, M. May, and M. Van Montagu, tion and virulence. G3:Genes, Genomes, Genetics 3: 675–686. 1996 Environmental and developmental signals modulate pro- Peng, Z., Q. Lu, and D. P. Verma, 1996 Reciprocal regulation of line homeostasis: evidence for a negative transcriptional regu- delta 1-pyrroline-5-carboxylate synthetase and proline dehydro- lator. Proc. Natl. Acad. Sci. USA 93: 8787–8791. genase genes controls proline levels during and after osmotic Wang, S. S., and M. C. Brandriss, 1986 Proline utilization in Sac- stress in plants. Mol. Gen. Genet. 253: 334–341. charomyces cerevisiae: analysis of the cloned PUT1 gene. Mol. Phang, J. M., S. P. Donald, J. Pandhare, and Y. Liu, 2008 The Cell. Biol. 6: 2638–2645. metabolism of proline, a stress substrate, modulates carcino- Yoon, K. A., Y. Nakamura, and H. Arakawa, 2004 Identification of genic pathways. Amino Acids 35: 681–690. ALDH4 as a p53-inducible gene and its protective role in cellular Phillips, A. J., I. Sudbery, and M. Ramsdale, 2003 Apoptosis in- stresses. J. Hum. Genet. 49: 134–140. duced by environmental stresses and amphotericin B in Candida albicans. Proc. Natl. Acad. Sci. USA 100: 14327–14332. Communicating editor: J. Heitman

Proline Catabolism in Cryptococcus 433 GENETICS

Supporting Information http://www.genetics.org/lookup/suppl/doi:10.1534/genetics.113.150326/-/DC1

Reactive Oxygen Species Homeostasis and Virulence of the Fungal Pathogen Cryptococcus neoformans Requires an Intact Proline Catabolism Pathway

I. Russel Lee, Edmund Y. L. Lui, Eve W. L. Chow, Samantha D. M. Arras, Carl A. Morrow, and James A. Fraser

Figure S1 The proline catabolism-associated genes are located in a cluster on a single chromosome in A. nidulans, however in S. cerevisiae and C. neoformans, the genes are unlinked and located on different chromosomes. prnA encodes the pathway-specific transcription factor of prn expression, prnD, PUT1 and PUT5 encode proline oxidase, prnB and PUT4 encode the proline-specific permease, while prnC and PUT2 encode P5C dehydrogenase. The multiple candidates of proline permease-encoding genes in C. neoformans are designated using the Broad Institute annotations.

2 SI I. M. R. C. Lee et al.

Figure S2 ClustalW multiple sequence alignment of C. neoformans Put1 and Put5. Identical amino acid residues are shaded dark grey while similar residues are shaded light grey.

I. M. R. C. Lee et al. 3 SI

Figure S3 Cross-species complementation using C. neoformans PUT2 partially restored the proline utilization ability of the S. cerevisiae put2∆ mutant. Tenfold spot dilution assays for nitrogen source utilization showed that C. neoformans PUT2 complemented the S. cerevisiae put2∆ mutant’s proline assimilation defect, albeit to a poorer extent than an S. cerevisiae PUT2. On the other hand, neither C. neoformans PUT1 or PUT5 restored the proline utilization defect of the S. cerevisiae put1∆ strain.

4 SI I. M. R. C. Lee et al.

Figure S4 Restoration of proline utilization ability upon the re-introduction of PUT5, PUT1 PUT5 and PUT2 into the put5∆, double put1∆ put5∆ and put2∆ mutants, respectively. Tenfold spot dilution assays for nitrogen or nitrogen and carbon source utilization showed that the complemented put5∆ + PUT5, put1∆ put5∆ + PUT1 PUT5 and put2∆ + PUT2 strains exhibited wild-type growth on YNB supplemented with 10 mM proline (with or without 2% glucose as a carbon source).

I. M. R. C. Lee et al. 5 SI

Figure S5 Tolerance against the oxidative stressor menadione and nitrosative stressor sodium nitrite is equivalent between the wild-type and proline oxidase(s) deletion mutant strains. Wild-type H99, put1∆, put5∆ and double put1∆ put5∆ strains were grown in YNB [proline and glutamate (10 mM each)] supplemented with (A) 0 – 0.003 mM menadione, or (B) 0 – 3 mM sodium nitrite. All strains displayed relatively similar growth percentage relative to the unstressed control when subjected to these individual stressors. Growth percentage relative to unstressed control is defined by dividing each stressed strain’s growth in OD600 with the same strain’s growth in OD600 when unstressed. Error bars represent standard errors across three biological replicates.

6 SI I. M. R. C. Lee et al.

Figure S6 Restoration of mitochondrial superoxide ions to wild-type levels upon the re-introduction of PUT2 into the put2∆ mutant. Wild-type H99, put2∆ and complemented put2∆ + PUT2 strains were briefly subjected to culture in YNB supplemented with 10 mM proline, stained with MitoSOX, and the accumulation of ROS in the mitochondria was assessed by flow cytometry. Bar chart presented as average ± standard error of triplicate experiments showed that the put2∆ mutant had significantly enhanced ∆ mean fluorescence intensity (∆ MFI) of oxidized MitoSOX relative to the wild-type or put2∆ + PUT2 strains (*** denotes P < 0.001). ∆ MFI is defined by subtracting the background MFI (unstained cells) from the MFI generated from MitoSOX stained cells.

I. M. R. C. Lee et al. 7 SI

Figure S7 None of the proline catabolic deletion mutants display a defect in mitochondrial membrane potential. Wild-type H99 and proline catabolic deletion mutant strains were briefly subjected to culture in YNB supplemented with 10 mM proline, stained with MitoTracker Red CMXRos, and cells viewed under a confocal microscope. The staining profile of the mutants was similar to that of wild-type showing accumulation of the fluorescence dye that is dependent on normal mitochondrial membrane permeability. Scale bar, 10 µm.

8 SI I. M. R. C. Lee et al.

Figure S8 Restoration of virulence factor expression to wild-type levels upon the re-introduction of PUT2 into the put2∆ mutant. (A) Tenfold spot dilution assays on YPD medium demonstrated that the complemented put2∆ + PUT2 strain exhibited wild-type growth at both 30 and 37°C. (B) Tenfold spot dilution assays on L-DOPA medium showed that the put2∆ + PUT2 strain produced equal amount of melanin as the wild-type strain at both 30 and 37°C.

I. M. R. C. Lee et al. 9 SI

Table S1 Fungal strains used in this study

Strain Genotype Original source H99 Cryptococcus neoformans MATα John Perfect EL13 H99 put1::NEO This study EL1 H99 put5::NEO This study EL9 H99 put1 put5::NEO This study EL18 H99 put2::NEO This study EL22 H99 put1::NEO + PUT1 NAT This study EL20 H99 put5::NEO + PUT5 NAT This study EL21 H99 put1 put5::NEO + PUT1 PUT5 NAT This study EL23 H99 put2::NEO + PUT2 NAT This study BY4741 Saccharomyces cerevisiae MATa his3∆1 leu2∆0 met15∆0 ura3∆0 Carrie Brachmann 4099B7 BY4741 put1::KanMX Thermo Scientific 1000H4 BY4741 put2::KanMX Thermo Scientific

10 SI I. M. R. C. Lee et al.

Table S2 Primers used in this study

Primer Purpose Sequence (5’-3’)

UQ1070 Cn PUT1 qRT-PCR F AAAAGCTAAAGAGAACGACGTTGC UQ1071 Cn PUT1 qRT-PCR R TCTCACTCTTTGACTTTGGAGGTTT UQ1072 Cn PUT5 qRT-PCR F AGAAGGCTAAGGAGAACAACATCATT UQ1073 Cn PUT5 qRT-PCR R GTCCAGATCTCCTCCTTGGAAG UQ1074 Cn PUT2 qRT-PCR F TCATTAACGGTGAGGAGGTCAAG UQ1075 Cn PUT2 qRT-PCR R GCAAGAGCACCGTCAATGG UQ1300 Cn PUT1 deletion 5’ F / genomic DNA generation F CTGCCATGTACAATCCATAAT UQ1301 Cn PUT1 deletion 5’ R CTCTCCAGCTCACATCCTCGCAGCGCGATTTCGTTA UQ1302 Cn PUT1 deletion 3’ F TGTTAATACAGATAAACCGTTGGTTCAAGATACTTT UQ1303 Cn PUT1 deletion 3’ R / genomic DNA generation R TCTCTACATTACGCAACATGG UQ1304 Cn PUT1 deletion NEO F TAACGAAATCGCGCTGCGAGGATGTGAGCTGGAGAG UQ1305 Cn PUT1 deletion NEO R AAAGTATCTTGAACCAACGGTTTATCTGTATTAACA UQ1306 Cn PUT5 deletion 5’ F / genomic DNA generation F GCACGCCTTCCGAAGTGAAAG UQ1307 Cn PUT5 deletion 5’ R AGCTCACATCCTCGCAGCCGTTGTGGATCGTTGGTT UQ1308 Cn PUT5 deletion 3’ F TGTTAATACAGATAAACCAACGCTTAAGTGGAGATG UQ1309 Cn PUT5 deletion 3’ R / genomic DNA generation R CCTCGTTCGGCTAAAGCTGAC UQ1310 Cn PUT5 deletion NEO F CAACCAACGATCCACAACGGCTGCGAGGATGTGAGC UQ1311 Cn PUT5 deletion NEO R CATCTCCACTTAAGCGTTGGTTTATCTGTATTAACA UQ1312 Cn PUT2 deletion 5’ F / genomic DNA generation F CTCCGGTGTCCATTATTCCTA UQ1313 Cn PUT2 deletion 5’ R AGCTCACATCCTCGCAGCTTTTGGTAAGGATGAAGA UQ1314 Cn PUT2 deletion 3’ F TGTTAATACAGATAAACCGTCAATGAGAATAGAAAA UQ1315 Cn PUT2 deletion 3’ R / genomic DNA generation R TTCTGACGGTGCAGGCGAAGA UQ1316 Cn PUT2 deletion NEO F TCTTCATCCTTACCAAAAGCTGCGAGGATGTGAGCT UQ1317 Cn PUT2 deletion NEO R TTTTCTATTCTCATTGACGGTTTATCTGTATTAACA UQ1036 Universal deletion diagnostic (NEO marker specific) R TGTGGATGCTGGCGGAGGATA UQ1410 Cn PUT1 deletion diagnostic F AAACGCTTAAGTGGAGATGAA UQ1411 Cn PUT5 deletion diagnostic F AGGCTGAAGATGATGCGAGAA UQ1412 Cn PUT2 deletion diagnostic F CGATGACAGCCGATACCTACA UQ18 Sequencing (M13F pCR2.1-TOPO) GTAAAACGACGGCCAG UQ19 Sequencing (M13R pCR2.1-TOPO) CAGGAAACAGCTATGAC UQ1594 Cn PUT1 genomic DNA sequencing GCTGTTCCGAAAGCTACCAGA UQ1595 Cn PUT1 genomic DNA sequencing CCTCCTCGTCCTCTGGTTCT UQ1596 Cn PUT1 genomic DNA sequencing TGGCGTTGGAAAAGAGCGGAG UQ1597 Cn PUT1 genomic DNA sequencing AAAGAGAACGAGTAGACACAT UQ1598 Cn PUT1 genomic DNA sequencing GCGTCATTTTCGGTACTCACA UQ1599 Cn PUT1 genomic DNA sequencing AACTTGATTTCATTCGTGCTC UQ1600 Cn PUT5 genomic DNA sequencing TCCCCTGCTATTTTTTATCGT UQ1601 Cn PUT5 genomic DNA sequencing GACAACTCCTTTTCGACAACC UQ1602 Cn PUT5 genomic DNA sequencing TGAAACCGTGGAGGGATGTAT UQ1603 Cn PUT5 genomic DNA sequencing GAAGGGCAAGTGGGATGACAT UQ1604 Cn PUT5 genomic DNA sequencing CATACAACACGCCGAAGCCAA UQ1605 Cn PUT5 genomic DNA sequencing GGCGGCAGAGATGAGGCGAGT UQ1606 Cn PUT2 genomic DNA sequencing TGGGTGTTGAAGATTTGGTCG

I. M. R. C. Lee et al. 11 SI

UQ1607 Cn PUT2 genomic DNA sequencing ACCAAAAATGTCTTCCCAACT UQ1608 Cn PUT2 genomic DNA sequencing GCCGAAATCGACGCTGCCGCT UQ1609 Cn PUT2 genomic DNA sequencing CCCAAAACCTCGACATCTACA UQ1610 Cn PUT2 genomic DNA sequencing TACGCTTTTCATTTCCGGCCT UQ1611 Cn PUT2 genomic DNA sequencing AGCGATTAGGGTTTACGACAT UQ1612 Cn PUT2 genomic DNA sequencing AGAAGAAGAAGTCAGTCAGCA UQ270 GenomeWalker adaptor GTAATACGACTCACTATAGGGC UQ271 GenomeWalker nested adaptor ACTATAGGGCACGCGTGGT UQ272 GenomeWalker 3’ gene specific CTGAATGGCGAATGAGCTTGAGCTTGG UQ273 GenomeWalker 3’ nested gene specific ATCAGATTGTCGTTTCCCGCCTTCAG UQ274 GenomeWalker 5’ gene specific GAGAGGCGGTTTGCGTATTGGCTAGAG UQ275 GenomeWalker 5’ nested gene specific ACGTCCGCAATGTGTTATTAAGTTGTC UQ1719 Sc PUT1 genomic DNA / promotor generation F AGACAAAACTGAGCGAGACAG UQ1756 Sc PUT1 genomic DNA / terminator generation R GCACCTACACTAACAGCTTCT UQ1729 Sc PUT2 genomic DNA / promotor generation F TCACCGCATCGTTCAACTGTC UQ1760 Sc PUT2 genomic DNA / terminator generation R TGCAATGCCAAAAGTGTCTGT UQ1720 Cn PUT1 cDNA generation F GGCATCCCATCCATACAACCTCTAGATTTTAGACA UQ1761 Cn PUT1 cDNA generation R GACCAAAGTATCTTGAACCAA UQ1725 Cn PUT5 cDNA generation F CACTAACGCACGCTAGAAATGTCAGCAATCCGACC UQ1762 Cn PUT5 cDNA generation R TTAATCAGAGCCTCCAAAGAA UQ1730 Cn PUT2 cDNA generation F AATTCACAGGAATTGCTTTTCTTTCTTCATCCTTA UQ1763 Cn PUT2 cDNA generation R GATACGAGACATTATGTATGA UQ1814 Sc PUT1 promotor generation R for Cn PUT1 GCGGGAGGCACGCATTTCTAGCGTGCGTTAGTGTT UQ1758 Sc PUT1 promotor generation R for Cn PUT5 GGTCGGATTGCTGACATTTCTAGCGTGCGTTAGTG UQ1815 Sc PUT1 terminator generation F for Cn PUT1 CAAGATACTTTGGTCGAGGACTATAATATATACTC UQ1728 Sc PUT1 terminator generation F for Cn PUT5 AGATTCTTTGGAGGCTCTGATTAAGAGGACTATAA UQ1759 Sc PUT2 promotor generation R for Cn PUT2 TAAGGATGAAGAAAGAAAAGCAATTCCTGTGAATT UQ1816 Sc PUT2 terminator generation F for Cn PUT2 ATGTCTCGTATCAAAAATTTTTGTGGAATAGAACC UQ1770 Sc PUT1 genomic DNA sequencing CAGCTATGACCATGATTACGC UQ1769 Sc PUT1 genomic DNA sequencing GTCGCTACACACGTACATTGC UQ1818 Sc PUT1 genomic DNA sequencing AGTGAAGGTGAAAGCGAGGCA UQ1771 Sc PUT1 genomic DNA sequencing CGCCACTCATCAATTGTTTAT UQ1772 Sc PUT1 genomic DNA sequencing TCCAAAAGCTCCTTATTAGTT UQ1819 Sc PUT1 genomic DNA sequencing ACAACATCCTACTGCCCAATA UQ1773 Sc PUT1 genomic DNA sequencing ATGAATATAAGCACCACGAAC UQ1774 Sc PUT1 genomic DNA sequencing TCACTTGGTTGTCGCCTCTCA UQ1775 Sc PUT1 genomic DNA sequencing AAGATTTCCGTTTGAACTAGC UQ1776 Sc PUT2 genomic DNA sequencing ATGACCATGATTACGCCAAGC UQ1777 Sc PUT2 genomic DNA sequencing AAGATGGCAAGGGATATGTTT UQ1778 Sc PUT2 genomic DNA sequencing CATATCCCTTGCCATCTTCAC UQ1820 Sc PUT2 genomic DNA sequencing GACTTTCTCAGTAGGGGTAGA UQ1779 Sc PUT2 genomic DNA sequencing AAACCTTCACAAACCGCTGCC UQ 1821 Sc PUT2 genomic DNA sequencing CCATGGATATACTTCTGGAGT UQ1780 Sc PUT2 genomic DNA sequencing AGCCTTATCGTCGGTACCACT UQ1822 Sc PUT2 genomic DNA sequencing GCCACTTTGTATTTTGCCATA UQ1781 Sc PUT2 genomic DNA sequencing GGCGAGCAGTCCCTGACAAGA

12 SI I. M. R. C. Lee et al.

UQ1782 Sc PUT2 genomic DNA sequencing TTCCACATGGCAAAGAATAAA UQ1595 Cn PUT1 cDNA sequencing CCTCCTCGTCCTCTGGTTCT UQ1596 Cn PUT1 cDNA sequencing TGGCGTTGGAAAAGAGCGGAG UQ1598 Cn PUT1 cDNA sequencing GCGTCATTTTCGGTACTCACA UQ1601 Cn PUT5 cDNA sequencing GACAACTCCTTTTCGACAACC UQ1602 Cn PUT5 cDNA sequencing TGAAACCGTGGAGGGATGTAT UQ1603 Cn PUT5 cDNA sequencing GAAGGGCAAGTGGGATGACAT UQ1604 Cn PUT5 cDNA sequencing CATACAACACGCCGAAGCCAA UQ1605 Cn PUT5 cDNA sequencing GGCGGCAGAGATGAGGCGAGT UQ1607 Cn PUT2 cDNA sequencing ACCAAAAATGTCTTCCCAACT UQ1608 Cn PUT2 cDNA sequencing GCCGAAATCGACGCTGCCGCT UQ1609 Cn PUT2 cDNA sequencing CCCAAAACCTCGACATCTACA UQ1611 Cn PUT2 cDNA sequencing AGCGATTAGGGTTTACGACAT

I. M. R. C. Lee et al. 13 SI

Table S3 Plasmids used in this study

Plasmid Detail Original source pJAF1 NEO selectable marker in pCR2.1-TOPO James Fraser pCH233 NAT selectable marker in pCR2.1-TOPO Christina Hull pELYL1 H99 PUT1 genomic DNA including promotor and terminator in pCR2.1-TOPO This study pELYL2 H99 PUT5 genomic DNA including promotor and terminator in pCR2.1-TOPO This study pELYL4 H99 PUT1 PUT5 genomic DNA including promotor and terminator in pCR2.1- This study TOPO pELYL3 H99 PUT2 genomic DNA including promotor and terminator in pCR2.1-TOPO This study pELYL5 H99 PUT1 genomic DNA including promotor and terminator in pCH233 This study pELYL6 H99 PUT5 genomic DNA including promotor and terminator in pCH233 This study pELYL8 H99 PUT1 PUT5 genomic DNA including promotor and terminator in pCH233 This study pELYL7 H99 PUT2 genomic DNA including promotor and terminator in pCH233 This study YCplac111 YC-type (centromeric) shuttle vector with LEU2 and ampR markers Daniel Gietz pEWC38 H99 PUT1 cDNA in pCR2.1-TOPO This study pEWC39 H99 PUT5 cDNA in pCR2.1-TOPO This study pEWC37 H99 PUT2 cDNA in pCR2.1-TOPO This study pEWC45 H99 PUT1 cDNA flanked by BY4741 PUT1 promotor and terminator in This study pCR2.1-TOPO pEWC46 H99 PUT5 cDNA flanked by BY4741 PUT1 promotor and terminator in This study pCR2.1-TOPO pEWC40 H99 PUT2 cDNA flanked by BY4741 PUT2 promotor and terminator in This study pCR2.1-TOPO pEWC35 BY4741 PUT1 genomic DNA including promotor and terminator in pCR2.1- This study TOPO pEWC36 BY4741 PUT2 genomic DNA including promotor and terminator in pCR2.1- This study TOPO pEWC48 H99 PUT1 cDNA flanked by BY4741 PUT1 promotor and terminator in This study YCplac111 pEWC49 H99 PUT5 cDNA flanked by BY4741 PUT1 promotor and terminator in This study YCplac111 pEWC47 H99 PUT2 cDNA flanked by BY4741 PUT2 promotor and terminator in This study YCplac111 pEWC50 BY4741 PUT1 genomic DNA including promotor and terminator in This study YCplac111 pEWC51 BY4741 PUT2 genomic DNA including promotor and terminator in This study YCplac111

14 SI I. M. R. C. Lee et al.

CHAPTER 6

Results Characterization of the Complete Uric Acid Degradation Pathway in the Fungal Pathogen Cryptococcus neoformans

113 6.1 Preface

Uric acid catabolism in Cryptococcus neoformans is largely unexplored but multiple lines of evidence, some of which have been discussed in Chapter 3, suggest it plays a role in survival in the environment and virulence during infection. C. neoformans is commonly isolated from purine-rich pigeon droppings, and the majority of the nitrogen content is in the form of uric acid, the particulate white matter seen in the excreta. Uric acid induces capsule formation that protects the fungus from engulfment by the environmental predator amoebae as well as immune phagocytes. Assimilation of uric acid necessitates several enzymes including urease, which is also known to promote C. neoformans invasion of the host central nervous system. In humans, uric acid is the end product of purine metabolism due to nonsense mutations in the uricase gene, and uric acid released from dying cells alerts the immune system to the presence of invading microorganisms and promotes inflammation. Degradation of uric acid therefore not only serves as a prime candidate for a nitrogen source used during infection, but may also represent a method to help the pathogen modulate host immune response. This chapter aims to develop an understanding of how uric acid is metabolised by C. neoformans, how its utilisation is regulated, and whether the enzymes in this catabolic pathway are undiscovered virulence factors themselves.

114 Characterization of the Complete Uric Acid Degradation Pathway in the Fungal Pathogen Cryptococcus neoformans

I. Russel Lee1,2, Liting Yang1,2, Gaseene Sebetso1,2, Rebecca Allen1,2, Thi H. N. Doan1,2, Ross Blundell1,2, Edmund Y. L. Lui1,2, Carl A. Morrow1,2, James A. Fraser1,2* 1 Australian Infectious Diseases Research Centre, University of Queensland, Brisbane, Queensland, Australia, 2 School of Chemistry and Molecular Biosciences, University of Queensland, Brisbane, Queensland, Australia

Abstract Degradation of purines to uric acid is generally conserved among organisms, however, the end product of uric acid degradation varies from species to species depending on the presence of active catabolic enzymes. In humans, most higher primates and birds, the urate oxidase gene is non-functional and hence uric acid is not further broken down. Uric acid in human blood plasma serves as an antioxidant and an immune enhancer; conversely, excessive amounts cause the common affliction gout. In contrast, uric acid is completely degraded to ammonia in most fungi. Currently, relatively little is known about uric acid catabolism in the fungal pathogen Cryptococcus neoformans even though this yeast is commonly isolated from uric acid-rich pigeon guano. In addition, uric acid utilization enhances the production of the cryptococcal virulence factors capsule and urease, and may potentially modulate the host immune response during infection. Based on these important observations, we employed both Agrobacterium-mediated insertional mutagenesis and bioinformatics to predict all the uric acid catabolic enzyme-encoding genes in the H99 genome. The candidate C. neoformans uric acid catabolic genes identified were named: URO1 (urate oxidase), URO2 (HIU hydrolase), URO3 (OHCU decarboxylase), DAL1 (allantoinase), DAL2,3,3 (allantoicase-ureidoglycolate hydrolase fusion protein), and URE1 (urease). All six ORFs were then deleted via homologous recombination; assaying of the deletion mutants’ ability to assimilate uric acid and its pathway intermediates as the sole nitrogen source validated their enzymatic functions. While Uro1, Uro2, Uro3, Dal1 and Dal2,3,3 were demonstrated to be dispensable for virulence, the significance of using a modified animal model system of cryptococcosis for improved mimicking of human pathogenicity is discussed.

Citation: Lee IR, Yang L, Sebetso G, Allen R, Doan THN, et al. (2013) Characterization of the Complete Uric Acid Degradation Pathway in the Fungal Pathogen Cryptococcus neoformans. PLoS ONE 8(5): e64292. doi:10.1371/journal.pone.0064292 Editor: Kirsten Nielsen, University of Minnesota, United States of America Received February 6, 2013; Accepted April 10, 2013; Published May 7, 2013 Copyright: ß 2013 Lee et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was supported by the University of Queensland International Research Tuition Award (UQIRTA) and University of Queensland Research Scholarship (UQRS) awarded to IRL, and the National Health and Medical Research Council (NHMRC) CDA 569673 grant awarded to JF. No additional external funding was received for this study. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: [email protected]

Introduction Until recently, the conversion of uric acid to allantoin had long been thought to involve a single step catalyzed by urate oxidase Purine catabolism is generally well conserved throughout [2]. However, production of allantoin is in fact preceded by two organisms ranging from animals and fungi to plants and bacteria. 1 additional distinct unstable intermediates: 5-hydroxyisourate While the degradation of purines to uric acid is common to all (HIU)2 and 2-oxo-4-hydroxy-4-carboxy-5-ureidoimidazoline kingdoms of life, the resulting uric acid can either be excreted or (OHCU)3. HIU can spontaneously decompose to OHCU and further degraded in the peroxisomes by active uric acid catabolic subsequently racemic allantoin at a slow, non-enzymatic rate in enzymes (Figure 1). The various nitrogenous end products in vitro [5,6,7]. However, only dextrorotatory allantoin, (S)-allantoin4, different organisms are a result of the successive loss-of-function is usually found in living cells [2]. The two additional biological mutations in urate oxidase (also known as uricase), allantoinase enzymes, HIU hydrolase and OHCU decarboxylase, are respon- and allantoicase during evolution. For instance, humans and apes sible for the quicker in vivo hydrolysis of HIU to OHCU and the share the inability to breakdown uric acid with Dalmatian dogs, decarboxylation of OHCU to (S)-allantoin, respectively [8,9]. birds, reptiles and insects [1,2]. As uric acid is a potent antioxidant Most mammals, except higher primates, have lost both allantoi- in the human blood plasma and protects cells from damage by nase and allantoicase and hence excrete allantoin as the end scavenging reactive oxygen species, it has been proposed that the product. loss of urate oxidase may have been selected for due to it Some organisms however, do have the ability to further promoting longevity [3]. Despite these protective advantages, uric catabolize (S)-allantoin to exploit its stored nitrogen, carbon and acid is poorly soluble in the serum and can precipitate to cause energy. Allantoinase catalyzes the hydrolysis of (S)-allantoin to ailments such as gout or urate kidney stones [4]. allantoate5, allantoicase catalyzes the degradation of allantoate to

PLOS ONE | www.plosone.org 1 May 2013 | Volume 8 | Issue 5 | e64292 Uric Acid Degradation in Cryptococcus neoformans

Figure 1. The end product of purine metabolism varies from species to species. While degradation of purines to uric acid is generally conserved among organisms, the end product of uric acid catabolism varies among taxa often due to the loss of functional catabolic enzymes in the pathway. doi:10.1371/journal.pone.0064292.g001 urea6 and ureidoglycolate7, while ureidoglycolate hydrolase not begin until the step of allantoin degradation) generally possess degrades ureidoglycolate to glyoxylate8 and another molecule of all the necessary nitrogen catabolic enzymes to completely degrade urea. In most amphibians and fishes, purine degradation proceeds uric acid to ammonia. Hence, ammonium is the most readily to urea and glyoxylate as the end products [10]. assimilated nitrogen source that is preferentially utilized by most Interestingly, in plants and in some bacteria (enterobacteria and bacteria and fungi in order to conserve energy. In the presence of Bacillus spp.), an alternative allantoate metabolic route also exists: ammonium, the expression of permease and catabolic enzyme- apart from the urea-releasing route, the other results in the encoding genes required to assimilate non-preferred nitrogen production of ammonia and carbon dioxide sources is usually repressed, a regulatory mechanism known as [2,11,12,13,14,15,16,17,18,19]. Allantoate deiminase is responsi- nitrogen catabolite/metabolite repression [20,21,22]. ble for the catalysis of allantoate for the ammonia-releasing route, Understanding the uric acid degradation pathway in Cryptococcus while the urea-releasing route necessitates subsequent urea neoformans, a fungal pathogen that commonly causes fatal hydrolytic activity by urease or urea amidolyase for full catabolism meningitis in the immunocompromised population, is clinically to ammonia9. Most plants, bacteria and fungi (an exception occurs relevant [23]. The C. neoformans primary ecological niche is purine- in the model yeast Saccharomyces cerevisiae where the pathway does rich pigeon droppings, and the majority of the nitrogen present is

PLOS ONE | www.plosone.org 2 May 2013 | Volume 8 | Issue 5 | e64292 Uric Acid Degradation in Cryptococcus neoformans in the form of uric acid, with the rest consisting primarily of Bioinformatic analyses xanthine, urea and creatinine [24]. In addition, the antiphagocytic C. neoformans genes were identified using annotation from the polysaccharide capsule is a well-known virulence factor that is H99 genome sequence from the Broad Institute (http://www. induced in the presence of uric acid as the sole nitrogen source broadinstitute.org/annotation/genome/ [25,26]. Assimilation of uric acid necessitates numerous catabolic cryptococcus_neoformans/MultiHome.html). Gene annotations enzymes including another well-established virulence factor, from the Broad are designated by their nomenclature urease, which plays a role in central nervous system invasion ‘‘CNAG_#####.#’’. Sequence analyses were performed using during host infection [27,28,29]. Incidentally, both uric acid and BLAST and MacVector 9.5 (MacVector Inc, Cary NC) [42]. urea are common metabolites found at high concentration in the Sequence alignments were created using ClustalW v1.4 within human blood. In fact, necrotic mammalian cells release uric acid MacVector [43]. Sequence traces generated at the Australian as a danger signal to activate both innate and adaptive immune Genome Research Facility (Brisbane, Queensland) were analyzed effectors including neutrophils, cytotoxic T-cell and antibody using Sequencher 4.7 (Gene Codes Corporation, Ann Arbor MI). immunity [30]. Putative peroxisomal localization signals were examined using The association between uric acid availability in the niches Target Signal Predictor Tool (http://www.peroxisomedb.org/ occupied by C. neoformans and how its utilization influences Target_signal.php). virulence factor expression and potential modulation of host immune response, prompted us to characterize the uric acid Construction and screening of library for C. neoformans catabolic enzymes. Every year, up to a million immunocompro- uric acid catabolism defective mutants mised individuals are infected with C. neoformans, two-thirds of An Agrobacterium-mediated mutagenesis library of C. neoformans whom will die [31]. Due to the poor outcomes associated with the H99 was created as described previously [44]. The majority of treatment of cryptococcosis, the cost of treatment and the toxicity transformants harbored a single copy of randomly integrated T- of the pharmaceutical agents currently available, it is imperative DNA and was mitotically stable. To identify genes involved in uric that we identify new therapeutic agents to help combat this acid catabolism, replica spotting of the insertional library disease. Rational drug design was pioneered in the purine consisting of 12,000 clones was carried out on YNB supplemented metabolic pathway over half a century ago, and has been the with uric acid or ammonium (10 mM each nitrogen source). paradigm for targeted drug discovery since the development of Mutants that grew on ammonium but failed to grow on uric acid- allopurinol and thioguanine for the treatment of gout and containing medium were selected for further analysis. T-DNA leukaemia, respectively [32]. Humans have successively lost the insertion site was mapped using GenomeWalker according to the uric acid enzymes during species evolution; should the C. manufacturer’s instructions (Clontech). BLAST analysis of these neoformans uric acid degradation pathway regulate pathogenesis sequences was performed to reveal which loci were affected. and provide a viable drug target, there exist over 50 years worth of literature describing innumerable compounds designed to modu- Construction and complementation of C. neoformans late the activity of these enzymes. In this study, we genetically and mutant strains phenotypically characterized the functions of all six of the C. All primers and plasmids used in this study are listed in neoformans uric acid catabolic enzyme-encoding genes in relation to Table S2 and S3, respectively. Gene deletion mutants were nitrogen source utilization and virulence-associated mechanisms. created using overlap PCR and biolistic transformation as described previously [45]. For example, to construct the uro1D Materials and Methods mutant strain in the H99 background, the 1,389 bp URO1 Strains and media (CNAG_04307.2) coding sequence was replaced with the neomycin phosphotransferase II-encoding selectable marker NEO using a All fungal strains used in this study are listed in Table S1, and construct created by overlap PCR combining a ,1 kb fragment were grown in YPD (1% yeast extract, 2% Bacto-peptone, 2% upstream the URO1 start codon, the NEO marker and a ,1kb glucose) or YNB (0.45% yeast nitrogen base w/o amino acids and fragment downstream the URO1 stop codon. Strain H99 genomic ammonium sulfate, 2% glucose, 10 mM nitrogen source) unless DNA and plasmid pJAF1 were used as PCR templates [46]. The specified otherwise. C. neoformans biolistic transformants were construct was transformed into C. neoformans cells via particle selected on YPD medium supplemented with 200 mg/mL G418 bombardment and transformants were selected on YPD plates (Sigma) or 100 mg/mL nourseothricin (Werner BioAgents). The supplemented with G418. Deletion of URO1 was confirmed by 12% (wt/vol) pigeon guano medium was prepared as described diagnostic PCR and Southern blot [47]. To complement the uro1D previously, with the exclusion of the filtration step to ensure that mutant, the URO1 gene including ,1 kb promoter and terminator insoluble nitrogen sources such as uric acid would not be removed was amplified from genomic DNA using high fidelity PCR, cloned [33]. V8 (5% V8 juice, 3 mM KH2PO4, 0.1% myo-Inositol, 4% into pCR2.1-TOPO (Invitrogen) to give pIRL5, and sequenced. Bacto-agar) and Murashige-Skoog (MS) mating media (PhytoTech- The URO1 fragment of pIRL5 was then subcloned into pCH233, nology Laboratories) at pH 5.0, and L-3,4-dihydroxyphenylala- creating the complementation construct pIRL13. pIRL13 was nine (L-DOPA) medium at pH 5.6, were prepared as described subsequently linearized and biolistically transformed into the uro1D previously [34,35,36,37]. Escherichia coli Mach-1 cells served as the mutant. Stable transformants were selected on YPD supplemented host strain for transformation and propagation of all plasmids with nourseothricin and complemented strains containing a single using lysogeny broth supplemented with either 100 mg/mL copy of the wild-type URO1 gene were identified by Southern blot. ampicillin (Sigma) or 50 mg/mL kanamycin (Sigma) [38]. Caenorhabditis elegans strain N2 was maintained at 15uC and Mating assays propagated on its normal laboratory food source E. coli OP50 cells Mating assays were conducted as described previously [33]. [39,40,41]. Nematode growth medium (NGM) was prepared as Briefly, strains were suspended in PBS either alone or mixed in described previously [39]. equal proportions with a strain of the opposite mating-type (MAT). Droplets (5 mL) were then spotted onto V8, MS or pigeon guano

PLOS ONE | www.plosone.org 3 May 2013 | Volume 8 | Issue 5 | e64292 Uric Acid Degradation in Cryptococcus neoformans

Figure 2. Predicted pathway of uric acid degradation in C. neoformans. While the metabolic pathway of purine degradation is present in bacteria, fungi, plants and animals, there is hidden complexity between species. In addition to species in which only part of the pathway is present, these variations include at least three examples of convergent evolution (conversion of xanthine to uric acid, allantoin to allantoate, and urea to ammonia), offshoots of the pathway that are only present in some species (allantoin racemase, ureidoglycine production and hydrolysis) as well as points at which spontaneous conversion occur without the need for enzyme activity (dashed arrows). Steps predicted to be absent from C. neoformans are opaque. doi:10.1371/journal.pone.0064292.g002 medium [33,35,37]. Plates were incubated at room temperature in suspension was then spotted onto YPD, YNB supplemented with the dark for 1 week and assessed by light microscopy for the the specified nitrogen source or L-DOPA agar. Results were formation of filaments and basidia. imaged after 2–3 days incubation at 30uC (nitrogen utilization assays), 30 and 37uC (melanization assays), or 37–39uC (high Capsule assays temperature growth assays). Strains were inoculated into RPMI supplemented with 10% fetal calf serum (Gibco) and incubated at 37uC for 2 days. To C. elegans killing assays visualize capsule, cells were stained with India ink (Becton Starter cultures of C. neoformans strains were prepared by growth Dickinson) and observed under a ZEISS Axioplan 2 epifluor- in YPD at 30uC overnight with shaking. Overnight cultures escent/light microscope. Pictures were taken with an Axiocam (10 mL) were spread onto 35 mm brain-heart infusion (BHI) greyscale digital camera using the AxioVision AC imaging (Becton Dickinson) agar plates, and incubated at 25uC overnight. software. Approximately 50 young adult C. elegans worms were then transferred from a lawn of E. coli OP50 on NGM to BHI Growth and melanization assays medium-grown C. neoformans [48]. Plates were incubated at 25uC Starter C. neoformans cultures were prepared by growth in YPD and worms examined for viability at 24-hr intervals using a at 30uC overnight with shaking, diluted to OD595nm = 0.05 in dissecting microscope, with worms that did not respond to a touch water, then further diluted tenfold in series. Each diluted cell with a platinum wire pick considered dead. Each experimental

PLOS ONE | www.plosone.org 4 May 2013 | Volume 8 | Issue 5 | e64292 Uric Acid Degradation in Cryptococcus neoformans

Figure 3. Nitrogen assimilation phenotypes of the C. neoformans uric acid catabolic deletion mutants. (A) Based on the enzymatic activity of homologs in other systems, inactivation of each respective catabolic enzyme (in red) is predicted to disrupt the ability of C. neoformans to utilize one or more pathway intermediates as the sole nitrogen source. (B) Ten-fold spot dilution assays on YNB supplemented with 10 mM uric acid or its derivatives revealed that the uro1D, uro2D, uro3D, dal1D, dal2,3,3D and ure1D mutants showed a correlation with the anticipated nitrogen utilization abilities. doi:10.1371/journal.pone.0064292.g003 condition was performed in triplicate. Survival was plotted against weight. Survival was plotted against time, and P values were time, and P values were calculated by plotting a Kaplan-Meier calculated by plotting a Kaplan-Meier survival curve and survival curve and performing a log-rank (Mantel-Cox) test using performing a log-rank (Mantel-Cox) test using Graphpad Prism Graphpad Prism Version 5.0c. P values of ,0.05 were considered Version 5.0c. P values of ,0.05 were considered statistically statistically significant. significant.

Murine inhalation model of cryptococcosis Ethics statement For murine virulence assays, C. neoformans were used to infect 6- This study was carried out in strict accordance with the week old female BALB/c mice by nasal inhalation [27]. For every recommendations in the Australian Code of Practice for the Care tested strain, ten mice were each inoculated with a 50 mL drop and Use of Animals for Scientific Purposes by the National Health containing 56105 cells. Mice were weighed before infection and and Medical Research Council. The protocol was approved by the daily thereafter; animals were sacrificed using CO2 inhalation Molecular Biosciences Animal Ethics Committee of The Univer- once their body weight had decreased to 80% of the pre-infection sity of Queensland (AEC approval number: SCMB/008/11/UQ/

PLOS ONE | www.plosone.org 5 May 2013 | Volume 8 | Issue 5 | e64292 Uric Acid Degradation in Cryptococcus neoformans

NHMRC). Infection was performed under methoxyflurane revealed 26% identity and 43% similarity (Figure S3). Notably, anaesthesia, and all efforts were made to minimize suffering peroxisomal targeting signals were not detected in Uro1, Uro2 and through adherence to the Guidelines to Promote the Wellbeing of Uro3 using Target Signal Predictor, providing no indication on Animals Used for Scientific Purposes as put forward by the their subcellular localization. Thus, Uro1, Uro2 and Uro3 may National Health and Medical Research Council. possibly rely on another unidentified peroxisomal targeting sequence, as previously demonstrated in the A. nidulans glyoxylate Results cycle enzyme isocitrate lyase (AcuD) that was shown to be peroxisomal without any obvious targeting signal peptide [53]. Agrobacterium-mediated insertional mutagenesis reveals URO1, URO2 and URO3 homologs are absent in S. cerevisiae, several enzymes required for uric acid assimilation in C. however, the genes encoding proteins required to complete the neoformans remaining steps of this pathway are arrayed in the largest Insertional mutagenesis is a powerful tool for identifying novel metabolic cluster known in yeast [54,55]. Allantoin degradation genes and their functions, and the frequency of random is best studied in S. cerevisiae and genes in this DAL cluster include integration is very high if Agrobacterium tumefaciens-mediated DAL1 (allantoinase), DAL2 (allantoicase), DAL3 (ureidoglycolate transformation is used. A T-DNA insertional library of C. hydrolase), DAL4 (allantoin permease), DCG1 (allantoin racemase) neoformans strain H99 was therefore created using this transking- and DAL7 (malate synthase that reduces the toxicity of allantoin dom DNA delivery approach in an attempt to locate loci involved catabolism via disposal of excess glyoxylate). Analysis of the H99 in uric acid utilization. We replica spotted 12,000 purified genome revealed that homologs of all these DAL genes are present individual transformants onto minimal YNB medium supplement- but are not clustered in a single chromosome. ed with uric acid or ammonium (control) as the sole nitrogen The C. neoformans DAL1 ORF is predicted to encode the source, and mapped the T-DNA insertion site of 8 disrupted allantoinase protein of 478 amino acids, and an alignment with the mutants that grew on ammonium but exhibited impaired growth S. cerevisiae homolog Dal1 revealed 43% identity and 59% on uric acid (Table S4). Our sequence analysis revealed four similarity (Figure S4). Interestingly, homology to S. cerevisiae independent ORFs that play a role in uric acid catabolism. Three DAL2 and DAL3 in C. neoformans is represented by a single ORF of these genes were predicted to encode well-established catabolic encoding a fusion allantoicase-ureidoglycolate hydrolase protein of enzymes known in other systems to be necessary for uric acid 813 residues. In fact, closer inspection revealed two tandemly degradation (urate oxidase, allantoicase and urease), while one was arrayed Dal3 products (Figure S5A). The existence of this predicted to encode the less familiar HIU hydrolase only recently plausible fusion event, presumably encoding a bifunctional proposed to be part of this pathway. enzyme, was confirmed by the acquisition of cDNA that spanned the entire ORF, disproving the possibility of gene clustering. This Bioinformatic analyses predict the complete uric acid phenomenon is only present in the phylum Basidiomycota and is not degradation pathway of C. neoformans observed in the Ascomycota or Zygomycota, indicating that the fusion Using the existing knowledge of purine catabolism in model event occurred 500 million – 1 billion years ago [56]. An ascomycetes, bioinformatic analyses were performed to complete alignment of C. neoformans Dal2 with the S. cerevisiae homolog the entire predicted uric acid degradation pathway of the Dal2 revealed 36% identity and 52% similarity (Figure S5B). An basidiomycete C. neoformans. Protein sequences of characterized alignment of C. neoformans Dal3a (first paralog) with the S. cerevisiae Aspergillus nidulans or S. cerevisiae uric acid catabolic enzymes were homolog Dal3 revealed 24% identity and 38% similarity, while C. queried against the H99 genome using reciprocal BLASTp neoformans Dal3b (second paralog) showed 23% identity and 40% searches. The candidate C. neoformans uric acid catabolic similarity (Figure S5C). enzyme-encoding genes identified were named: URO1 Finally, the URE1 ORF that encodes the nickel-containing (CNAG_04307.2), URO2 (CNAG_06694.2), URO3 urease of 833 amino acids is the only enzyme of the C. neoformans (CNAG_00639.2), DAL1 (CNAG_00934.2), DAL2,3,3 uric acid degradation pathway that has been characterized, due to (CNAG_01108.2), and URE1 (CNAG_05540.2). Our previous its precedence in fungi and bacteria as an important pathogenic work has shown that the expression of at least two of these putative factor [27,28,29]. An alignment of C. neoformans Ure1 with the A. catabolic genes, URO1 and DAL1, is inducible by the presence of nidulans homolog UreB revealed 60% identity and 75% similarity uric acid (pathway-specific induction) and repressible by the (Figure S6). In S. cerevisiae, urea utilization occurs via the DUR1,2- presence of ammonium (nitrogen metabolite repression) [25]. encoded urea amidolyase which is absent in C. neoformans [57]. The URO1 ORF is predicted to encode the urate oxidase Altogether, based on random insertional mutagenic screen and protein of 305 amino acid residues, and comparative analysis of exhaustive bioinformatics, the predicted uric acid degradation Uro1 with the A. nidulans homolog UaZ revealed 42% identity and pathway of C. neoformans is deciphered (Figure 2). 54% similarity (Figure S1). Importantly, the conserved N-terminus of eukaryotic uricases that is critical for enzymatic activity is highly Verification of the catabolic functions of the putative C. similar between Uro1 and UaZ. The URO2 ORF is predicted to neoformans uric acid enzymes using reverse genetics encode the HIU hydrolase protein of 122 amino acids, and an To verify the functions of URO1, URO2, URO3, DAL1, alignment with the A. nidulans homolog UaX revealed 32% DAL2,3,3 and URE1, gene deletion mutants were created via identity and 47% similarity (Figure S2). Both Uro2 and UaX homologous recombination in strain H99. All the deletion strains contain the conserved C-terminal tetrapeptide YRGS, a signature were viable and their growth appeared indistinguishable from of the thyroid hormone transporter transthyretin which probably wild-type on rich YPD medium. Growth of the set of deletion originated from a duplication of an ancestral HIU hydrolase gene mutants was then tested on YNB medium supplemented with uric [49,50]. Additionally, both Uro2 and UaX show complete acid and its pathway intermediates as the sole nitrogen source conservation of all residues involved in the binding of the substrate (Figure 3). However, growth on HIU, OHCU and allantoate were in characterized HIU hydrolases [50,51,52]. The URO3 ORF is not tested either due to the compound’s chemical instability (HIU predicted to encode the OHCU decarboxylase protein of 204 and OHCU) or discontinued production by biotechnological residues, and an alignment with the A. nidulans homolog UaW companies (allantoate).

PLOS ONE | www.plosone.org 6 May 2013 | Volume 8 | Issue 5 | e64292 Uric Acid Degradation in Cryptococcus neoformans

Figure 4. Uro1, Uro2, Uro3, Dal1, Dal2,3,3 and Ure1 are not required for C. neoformans initiation of mating. Filamentation assays on V8, MS and pigeon guano (PG) medium showed that filament formation appeared indistinguishable between the crosses of H99 6 KN99a and each of the uric acid mutants 6 KN99a. (A) Periphery of each individual colony. (B) Under higher magnification (6400). doi:10.1371/journal.pone.0064292.g004

As expected, loss of each respective gene resulted in inability of mutants exhibited growth defects on uric acid and allantoin, but the mutant to retain the enzymatic activity required to degrade its displayed wild-type growth on urea and ammonium. Lastly, the related purine derivative. The uro1D, uro2D and uro3D mutants ure1D mutant exhibited growth defects on uric acid, allantoin and exhibited growth defects on uric acid, but displayed wild-type urea, but displayed wild-type growth on ammonium. Comple- growth on allantoin, urea and ammonium. While the uro2 and uro3 mentation of all the uric acid catabolic deletion mutants restored mutations resulted in impaired growth on uric acid after a two-day the wild-type nitrogen utilization phenotypes, verifying the incubation period, growth however catches up on further enzymatic functions of Uro1, Uro2, Uro3, Dal1, Dal2,3,3 and incubation, consistent with the fact that spontaneous conversion Ure1 (Figure S8). of HIU to OHCU and subsequently to allantoin can still occur at a slow, non-enzymatic rate in the absence of HIU hydrolase or OHCU decarboxylase (Figure S7). The dal1D and dal2,3,3D

PLOS ONE | www.plosone.org 7 May 2013 | Volume 8 | Issue 5 | e64292 Uric Acid Degradation in Cryptococcus neoformans

Figure 5. Uro1, Uro2, Uro3, Dal1, Dal2,3,3 and Ure1 are not required for expression of the three major virulence attributes. (A) India ink cell staining under light microscopy revealed that the uric acid catabolic deletion mutants produced characteristic halos around its cells representing enlarged capsule that were similar to wild-type when cultured under serum-induced growth conditions. Scale bar, 10 mm. (B) Ten-fold spot dilution assays on L-DOPA medium at both 30 and 37uC showed that the catabolic deletion mutants melanized to the same extent as wild-type. C. Ten-fold spot dilution assays on YPD medium at human body temperature (37 and 39uC) demonstrated that the deletion mutants exhibited wild- type growth. doi:10.1371/journal.pone.0064292.g005

None of the C. neoformans uric acid catabolic enzymes virulence factor production that is influenced by multiple play a role in the initiation of mating physiological conditions. Uric acid-rich pigeon guano is a common source for infectious propagules of C. neoformans and is postulated to play a key part in HIU hydrolase activity is required for C. neoformans killing transmission from the environment to a human host [23,24]. of nematode worms Quite recently, C. neoformans was shown to undergo robust sexual There have been many instances in the Cryptococcus literature reproduction on medium containing sterilized pigeon guano whereby various mutant strains show no reduction in traits extracts in the laboratory [33]. To address whether the uric acid normally associated with virulence in vitro, yet significantly impact catabolic enzymes play a role in this sexual development, we pathogenicity in vivo [58,59]. C. neoformans presumably interacts unilaterally crossed the wild-type and mutant strains of H99a with C. elegans in the environment and killing of C. elegans by C. background with the opposite mating-type, KN99a, on V8, MS neoformans has previously been validated as a model for studying and pigeon guano medium (Figure 4). Production of filaments and fungal pathogenesis [48]. We performed C. elegans virulence assays basidia at the edges of the mating spots were indistinguishable using standard BHI medium to evaluate if any of the uric acid between the crosses of H99 6 KN99a and each of the uric acid catabolic enzymes contribute to C. neoformans pathogenesis mutants 6KN99a on the various mating medium, suggesting that (Figure 6). Killing of C. elegans by the uro1D, uro3D, dal1D, the uric acid enzymes are not likely to play a role in the initiation dal2,3,3D and ure1D strains was not significantly different to that of morphological differentiation of the C. neoformans life cycle. observed for the wild-type strain, indicating that urate oxidase, Further bilateral mating crosses may be performed to support this OHCU decarboxylase, allantoinase allantoicase-ureidoglycolate conclusion. hydrolase and urease are not required for C. neoformans-mediated killing of invertebrate worms. However, the uro2D mutant killed C. None of the C. neoformans uric acid catabolic enzymes elegans slightly less efficient than the wild-type strain (WT vs uro2D, contribute to the production of capsule or melanin, or P,0.0001), implicating HIU hydrolase in C. neoformans virulence. the ability to grow at high temperature The median survival of the worms infected with wild-type in comparison to uro2D is 7 and 8 days, respectively. As the influence of uric acid on certain cryptococcal virulence factor expression has been established, we investigated whether any of the uric acid catabolic enzymes affect the production of the None of the C. neoformans uric acid catabolic enzymes, antiphagocytic capsule or antioxidant melanin, and the ability to with the exception of urease, is required for grow at human body temperature (Figure 5) [25,26]. The uro1D, pathogenesis during infection of a mammalian host uro2D, uro3D, dal1D, dal2,3,3D and ure1D mutants all displayed We next performed a murine inhalation model of cryptococ- these three classical virulence traits that were indistinguishable cosis, widely regarded as the gold standard test for virulence assays from wild-type, suggesting that the uric acid enzymes are of C. neoformans, to examine the roles of urate oxidase, HIU dispensable for virulence factor expression when alternative hydrolase, OHCU decarboxylase, allantoinase and allantoicase- nitrogen sources are available to compensate for the lack of uric ureidoglycolate hydrolase during infection of a mammalian host acid utilization. This result reiterates the complex regulation of (Figure 7). The virulence-associated function of urease has already

PLOS ONE | www.plosone.org 8 May 2013 | Volume 8 | Issue 5 | e64292 Uric Acid Degradation in Cryptococcus neoformans

Figure 6. HIU hydrolase plays a subtle role in C. neoformans-mediated killing of C. elegans. ,50 nematode worms were transferred to a lawn of (A) wild-type H99, uro1D, uro1D + URO1, (B) uro2D, uro2D + URO2, (C) uro3D, uro3D + URO3, (D) dal1D, dal1D + DAL1, (E) dal2,3,3D, dal2,3,3D + DAL2,3,3, (F) ure1D or ure1D + URE1, as the sole food source on BHI medium and survival was monitored at 24-hr intervals. There was no observable difference in C. elegans killing by the uro1D, uro1D + URO1, uro2D + URO2, uro3D, uro3D + URO3, dal1D, dal1D + DAL1, dal2,3,3D, dal2,3,3D + DAL2,3,3, ure1D and ure1D + URE1 strains compared to wild-type. In contrast, the uro2D strain killed C. elegans slightly slower than wild-type. All experiments were repeated three times with similar results. doi:10.1371/journal.pone.0064292.g006 been thoroughly characterized in several different animal models pathway is conserved in bacteria, fungi, plants and animals, it of C. neoformans infection and hence it is omitted from this study nevertheless displays wondrous variation not only between [27,28,29]. Surprisingly, mice intranasally infected with the uro1D, kingdoms of life, but also from species to species. In the soil uro2D, uro3D, dal1D and dal2,3,3D mutant strains all succumbed to bacterium Bacillus subtilis, eight genes [pucL and pucM (uricases), infection at approximately the same rate as mice infected with the pucJ and pucK (uric acid transporters), pucH (allantoinase), pucI wild-type strain (killing occurred between 18 and 31 days post (allantoin permease), pucF (allantoate amidohydrolase), pucR infection, median survival of mice = 21 days). In order to provide (transcriptional activator of puc gene expression)] whose products parsimonious explanations to the conflicting uro2D results in the C. act in concert to drive uric acid catabolism, are all located in a elegans killing assays and the experimental murine cryptococcosis, cluster with the exception of pucI [14]. In the model hemi- investigating the exact mechanism by which C. neoformans kills C. ascomycete S. cerevisiae, genomic rearrangements have led to the elegans might be beneficial; while accumulation of the fungus within assembly of six adjacent genes of the DAL cluster that enables the gastrointestinal tract of the worm is known to be associated preferential utilization of allantoin as a nitrogen source over with killing, the precise nature of the cause of death is unclear [48]. assimilation of uric acid that requires the oxygen-consuming urate oxidase [55]. Incidentally, baker’s yeast can grow vigorously in Discussion anaerobic conditions. Such gene clustering for purine metabolism is not observed in Nitrogen scavenging through the uric acid degradation pathway filamentous ascomycetes such as A. nidulans, where the entire uric is complex and it involves multiple enzymatic steps. While this acid degradation pathway has recently been dissected [8]. The

PLOS ONE | www.plosone.org 9 May 2013 | Volume 8 | Issue 5 | e64292 Uric Acid Degradation in Cryptococcus neoformans

Figure 7. Uro1, Uro2, Uro3, Dal1 and Dal2,3,3 are not required for infection of a murine host. 10 mice were each intranasally infected with either 56105 cells of (A) wild-type H99, uro1D, uro1D +URO1, (B) uro2D, uro2D +URO2, (C) uro3D, uro3D + URO3, (D) dal1D, dal1D +DAL1, (E) dal2,3,3D or dal2,3,3D + DAL2,3,3, and survival was monitored daily. Mice infected with the uro1D, uro1D +URO1, uro2D, uro2D +URO2, uro3D, uro3D + URO3, dal1D, dal2,3,3D and dal2,3,3D + DAL2,3,3 strains progressed to morbidity as quickly as mice infected with the wild-type strain. Unexpectedly, the complemented dal1D +DAL1 strain appeared to be significantly less virulent than wild-type; this conundrum is likely caused by integration of DAL1 into a non-native, virulence-associated (possibly high temperature-associated) locus. doi:10.1371/journal.pone.0064292.g007 discovery of a different form of gene clustering (the gene-fusion specific enzymatic activity; indeed, multiple lines of evidence have event involving DAL2,3,3)inC. neoformans introduces a new demonstrated that fusion events may even increase the functional paradigm to the fungal uric acid degradation pathway. It is not efficiency of the fusion protein [61,63]. uncommon for certain protein families in a particular species Recent metabolic engineering efforts have enabled synthetic comprised of fused domains to correspond to a single, full-length protein scaffolds to be built that spatially recruit successive protein of another species. For instance, the E. coli acetate CoA pathway enzymes in a designable manner, ultimately improving transferase a and b subunits has regions of homology to the N- and the production of the end product while lowering the overall C-terminus of human succinyl CoA transferase [60,61]. One of metabolic load on the chassis organism [64]. The principle behind the most fascinating questions that can be asked of these such metabolic pipelines is to assemble enzyme complexes that biologically related proteins is whether they originated from a bring the active sites close together to prevent loss of intermediates primordial fused protein separated through evolution or vice versa. or competition from other pathways. Therefore, while we were An example of the latter, evolved by gene fusion, is the Arom locus unable to adequately test the enzymatic function of C. neoformans of A. nidulans that encodes a pentafunctional polypepetide. The allantoicase-ureidoglycolate hydrolase, we propose that this fusion AROM complex consists of domains analogous to five bacterial may enable faster and more efficient degradation of allantoate to enzymes (AroA, AroB, AroE, AroK and AroL) responsible for urea and glyoxylate, via substrate channeling. catalyzing polyaromatic amino acid biosynthesis [61,62]. This A closer inspection of the literature revealed that the AROM example affirms the fact that fused domains can retain its phenomenon of fusion uric acid catabolic protein is also observed

PLOS ONE | www.plosone.org 10 May 2013 | Volume 8 | Issue 5 | e64292 Uric Acid Degradation in Cryptococcus neoformans in plants and legumes, where a fusion OHCU decarboxylase-HIU Figure S3 ClustalW sequence alignment of A. nidulans hydrolase is present [9,65]. This fusion OHCU decarboxylase- UaW and C. neoformans Uro3. Identical amino acid residues HIU hydrolase protein contains enzymes that act one after are shaded dark grey while similar residues are shaded light grey. another, just like allantoicase-ureidoglycolate hydrolase. An (DOC) intriguing consideration to come out of these observations is that Figure S4 ClustalW sequence alignment of S. cerevisiae these fusions may suggest protein-protein interactions between the Dal1 and C. neoformans Dal1. Identical amino acid residues multiple enzymes of this pathway. In support of this notion, are shaded dark grey while similar residues are shaded light grey. allantoinase and allantoicase have been shown to form a complex (DOC) in amphibians [10]. Further studies into the enzyme structures and protein-protein interactions of members of the uric acid pathway Figure S5 The C. neoformans DAL2,3,3 gene encodes a may provide insights into how C. neoformans has undergone fusion allantoicase-ureidoglycolate hydrolase protein. A. evolutionary biochemical adaptation for optimized utilization of Representative protein architecture of S. cerevisiae Dal2 and Dal3, this abundant nitrogen source found in pigeon excreta. and C. neoformans Dal2,3,3. B. ClustalW sequence alignment of S. In fact, uric acid is also a common metabolite found at cerevisiae Dal2 and C. neoformans Dal2. C. ClustalW multiple concentrations as high as 2,696 mM in the serum of healthy sequence alignment of S. cerevisiae Dal3 and C. neoformans Dal3a and individuals [66]. This occurrence is caused by the inactivation of Dal3b. Identical amino acid residues are shaded dark grey while the urate oxidase gene of hominoids some 15 million years ago in a similar residues are shaded light grey. primate ancestor [1]. The subsequent inactivation of other (DOC) catabolic genes of the pathway is a possible evolutionary scenario; Figure S6 ClustalW sequence alignment of A. nidulans the human HIU hydrolase gene has several inactivating mutations UreB and C. neoformans Ure1. Identical amino acid residues while the OHCU decarboxylase gene, although potentially are shaded dark grey while similar residues are shaded light grey. encoding a complete protein, does not appear to be transcription- (DOC) ally expressed [9]. Recently, necrotic mammalian cells have been shown to release uric acid as a danger signal to enhance immune Figure S7 HIU hydrolase and OHCU decarboxylase response against invading microbes [30]. Such immunomodulat- activities are not entirely essential, but are required ing properties of uric acid are not likely to be exhibited in the for efficient uric acid degradation in C. neoformans. immunocompetent BALB/c mice, commonly used as an animal Tenfold spot dilution assays for nitrogen source utilization showed model system for testing cryptococcosis, since their biological uric that the uro1D, uro2D and uro3D mutants exhibited obvious growth acid catabolic enzymes are intact. Hence, in order to better mimic defects on YNB supplemented with 10 mM uric acid after a two- the infection of humans, we are currently in the process of day incubation period, however, growth of the uro2D and uro3D breeding the C57BL/6J urate oxidase-deficient (hyperuricemia) strains catches up on further incubation. mice colonies previously created by Wu et al., to further investigate (DOC) the virulence capacities of the C. neoformans uric acid catabolic Figure S8 Restoration of the ability to utilize uric acid mutants [67]. In support of our proposition, studies have shown and its pathway intermediates upon complementation of that AIDS patients, particularly those undergoing antiretroviral the uric acid catabolic deletion mutants. Tenfold spot therapies, have a higher prevalence of hyperuricaemia [68]. dilution assays for nitrogen source utilization showed that the Overall, the results we report here represent the first definitive complemented uro1D + URO1, uro2D + URO2, uro3D + URO3, characterization of the complete uric acid degradation pathway in dal1D + DAL1, dal2,3,3D + DAL2,3,3 and ure1D + URE1 strains a second phylum of the kingdom fungi, the Basidiomycota, and we exhibited wild-type growth on YNB supplemented with uric acid, have laid the foundation for future research into investigating C. allantoin, urea or ammonium (10 mM each). neoformans virulence via purine metabolism. (DOC) Supporting Information Table S1 Fungal strains used in this study. (DOC) Figure S1 ClustalW sequence alignment of A. nidulans Table S2 Primers used in this study. UaZ and C. neoformans Uro1. Identical amino acid residues (DOC) are shaded dark grey while similar residues are shaded light grey. The conserved N-terminus sequence needed for enzymatic activity Table S3 Plasmids used in this study. is boxed in red. (DOC) (DOC) Table S4 C. neoformans uric acid catabolism defective Figure S2 ClustalW sequence alignment of A. nidulans mutants identified during the Agrobacterium-mediated UaX and C. neoformans Uro2. Identical amino acid residues insertional mutagenic screen. are shaded dark grey while similar residues are shaded light grey. (DOC) The conserved C-terminus YRGS motif that distinguishes members of the transthyretin family is boxed in red. The residues Author Contributions indicated with an asterisk below the sequence are conserved active Conceived and designed the experiments: IRL LY CM JF. Performed the site residues, as determined in the structures of the enzymes from experiments: IRL LY GS RA TD RB EL. Analyzed the data: IRL LY JF. B. subtilus, Salmonella dublin and Danio rerio. Contributed reagents/materials/analysis tools: IRL JF. Wrote the paper: (DOC) IRL LY JF.

References 1. Oda M, Satta Y, Takenaka O, Takahata N (2002) Loss of urate oxidase activity 2. Vogels GD, Van der Drift C (1976) Degradation of purines and pyrimidines by in hominoids and its evolutionary implications. Mol Biol Evol 19: 640–653. microorganisms. Bacteriol Rev 40: 403–468.

PLOS ONE | www.plosone.org 11 May 2013 | Volume 8 | Issue 5 | e64292 Uric Acid Degradation in Cryptococcus neoformans

3. Ames BN, Cathcart R, Schwiers E, Hochstein P (1981) Uric acid provides an 34. D’Souza CA, Alspaugh JA, Yue C, Harashima T, Cox GM, et al. (2001) Cyclic antioxidant defense in humans against oxidant- and radical-caused aging and AMP-dependent protein kinase controls virulence of the fungal pathogen cancer: a hypothesis. Proc Natl Acad Sci U S A 78: 6858–6862. Cryptococcus neoformans. Mol Cell Biol 21: 3179–3191. 4. Hall AP, Barry PE, Dawber TR, McNamara PM (1967) Epidemiology of gout 35. Kwon-Chung KJ, Bennett JE, Rhodes JC (1982) Taxonomic studies on and hyperuricemia. A long-term population study. Am J Med 42: 27–37. Filobasidiella species and their anamorphs. Antonie Van Leeuwenhoek 48: 25–38. 5. Kahn K, Serfozo P, Tipton PA (1997) Identification of the true product of the 36. Williamson PR (1994) Biochemical and molecular characterization of the urate oxidase reaction. J Am Chem Soc 274: 5435–5442. diphenol oxidase of Cryptococcus neoformans: identification as a laccase. J Bacteriol 6. Kahn K, Tipton PA (1998) Spectroscopic characterization of intermediates in 176: 656–664. the urate oxidase reaction. Biochemistry 37: 11651–11659. 37. Xue C, Tada Y, Dong X, Heitman J (2007) The human fungal pathogen 7. Modric N, Derome AE, Ashcroft SJH, Poje M (1992) Tracing and identification Cryptococcus can complete its sexual cycle during a pathogenic association with of uricase reaction intermediates. A direct 13C-NMR/isotope-labelling plants. Cell Host Microbe 1: 263–273. evidence. Tetrahedr Lett 33: 6691–6694. 38. Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning : a laboratory 8. Gournas C, Oestreicher N, Amillis S, Diallinas G, Scazzocchio C (2011) manual. Cold Spring Harbor, N.Y.: Cold Spring Harbor Laboratory. Completing the purine utilisation pathway of Aspergillus nidulans. Fungal Genet 39. Brenner S (1974) The genetics of Caenorhabditis elegans. Genetics 77: 71–94. Biol 48: 840–848. 40. Garsin DA, Sifri CD, Mylonakis E, Qin X, Singh KV, et al. (2001) A simple 9. Ramazzina I, Folli C, Secchi A, Berni R, Percudani R (2006) Completing the model host for identifying Gram-positive virulence factors. Proc Natl Acad uric acid degradation pathway through phylogenetic comparison of whole Sci U S A 98: 10892–10897. genomes. Nat Chem Biol 2: 144–148. 41. Honda S, Ishii N, Suzuki K, Matsuo M (1993) Oxygen-dependent perturbation of life span and aging rate in the nematode. J Gerontol 48: B57–61. 10. Hayashi S, Fujiwara S, Noguchi T (2000) Evolution of urate-degrading enzymes 42. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local in animal peroxisomes. Cell Biochem Biophys 32 Spring: 123–129. alignment search tool. J Mol Biol 215: 403–410. 11. Cusa E, Obradors N, Baldoma L, Badia J, Aguilar J (1999) Genetic analysis of a 43. Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTAL W: improving the chromosomal region containing genes required for assimilation of allantoin sensitivity of progressive multiple sequence alignment through sequence nitrogen and linked glyoxylate metabolism in Escherichia coli. J Bacteriol 181: weighting, position-specific gap penalties and weight matrix choice. Nucleic 7479–7484. Acids Res 22: 4673–4680. 12. Gravenmade EJ, Vogels GD, Van der Drift C (1970) Hydrolysis, racemization 44. Idnurm A, Reedy JL, Nussbaum JC, Heitman J (2004) Cryptococcus neoformans and absolute configuration of ureidoglycolate, a substrate of allantoicase. virulence gene discovery through insertional mutagenesis. Eukaryot Cell 3: 420– Biochim Biophys Acta 198: 569–582. 429. 13. Munoz A, Piedras P, Aguilar M, Pineda M (2001) Urea is a product of 45. Davidson RC, Blankenship JR, Kraus PR, de Jesus Berrios M, Hull CM, et al. ureidoglycolate degradation in chickpea. Purification and characterization of the (2002) A PCR-based strategy to generate integrative targeting alleles with large ureidoglycolate urea-lyase. Plant Physiol 125: 828–834. regions of homology. Microbiology 148: 2607–2615. 14. Schultz AC, Nygaard P, Saxild HH (2001) Functional analysis of 14 genes that 46. Fraser JA, Subaran RL, Nichols CB, Heitman J (2003) Recapitulation of the constitute the purine catabolic pathway in Bacillus subtilis and evidence for a sexual cycle of the primary fungal pathogen Cryptococcus neoformans var. gattii: novel regulon controlled by the PucR transcription activator. J Bacteriol 183: implications for an outbreak on Vancouver Island, Canada. Eukaryot Cell 2: 3293–3302. 1036–1045. 15. Todd CD, Tipton PA, Blevins DG, Piedras P, Pineda M, et al. (2006) Update on 47. Southern E (2006) Southern blotting. Nat Protoc 1: 518–525. ureide degradation in legumes. J Exp Bot 57: 5–12. 48. Mylonakis E, Ausubel FM, Perfect JR, Heitman J, Calderwood SB (2002) Killing 16. van der Drift C, de Windt FE, Vogels GD (1970) Allantoate hydrolysis by of Caenorhabditis elegans by Cryptococcus neoformans as a model of yeast pathogenesis. allantoate amidohydrolase. Arch Biochem Biophys 136: 273–279. Proc Natl Acad Sci U S A 99: 15675–15680. 17. Werner AK, Romeis T, Witte CP (2010) Ureide catabolism in Arabidopsis thaliana 49. Lee Y, Lee DH, Kho CW, Lee AY, Jang M, et al. (2005) Transthyretin-related and Escherichia coli. Nat Chem Biol 6: 19–21. proteins function to facilitate the hydrolysis of 5-hydroxyisourate, the end 18. Werner AK, Sparkes IA, Romeis T, Witte CP (2008) Identification, biochemical product of the uricase reaction. FEBS Lett 579: 4769–4774. characterization, and subcellular localization of allantoate amidohydrolases from 50. Zanotti G, Cendron L, Ramazzina I, Folli C, Percudani R, et al. (2006) Arabidopsis and soybean. Plant Physiol 146: 418–430. Structure of zebra fish HIUase: insights into evolution of an enzyme to a 19. Winkler RG, Blevins DG, Randall DD (1988) Ureide Catabolism in Soybeans : hormone transporter. J Mol Biol 363: 1–9. III. Ureidoglycolate Amidohydrolase and Allantoate Amidohydrolase Are 51. Hennebry SC, Law RH, Richardson SJ, Buckle AM, Whisstock JC (2006) The Activities of an Allantoate Degrading Enzyme Complex. Plant Physiol 86: crystal structure of the transthyretin-like protein from Salmonella dublin,a 1084–1088. prokaryote 5-hydroxyisourate hydrolase. J Mol Biol 359: 1389–1399. 20. Magasanik B (2000) Global regulation of gene expression. Proc Natl Acad 52. Jung DK, Lee Y, Park SG, Park BC, Kim GH, et al. (2006) Structural and Sci U S A 97: 14044–14045. functional analysis of PucM, a hydrolase in the ureide pathway and a member of 21. Marzluf GA (1997) Genetic regulation of nitrogen metabolism in the fungi. the transthyretin-related protein family. Proc Natl Acad Sci U S A 103: 9790– Microbiol Mol Biol Rev 61: 17–32. 9795. 22. Wong KH, Hynes MJ, Davis MA (2008) Recent advances in nitrogen 53. Hynes MJ, Murray SL, Khew GS, Davis MA (2008) Genetic analysis of the role regulation: a comparison between Saccharomyces cerevisiae and filamentous fungi. of peroxisomes in the utilization of acetate and fatty acids in Aspergillus nidulans. Eukaryot Cell 7: 917–925. Genetics 178: 1355–1369. 23. Casadevall A, Perfect JR (1998) Cryptococcus neoformans. Washington, DC: ASM 54. Lawther RP, Riemer E, Chojnacki B, Cooper TG (1974) Clustering of the genes Press. for allantoin degradation in Saccharomyces cerevisiae. J Bacteriol 119: 461–468. 24. Staib F, Grave B, Altmann L, Mishra SK, Abel T, et al. (1978) Epidemiology of 55. Wong S, Wolfe KH (2005) Birth of a metabolic gene cluster in yeast by adaptive Cryptococcus neoformans. Mycopathologia 65: 73–76. gene relocation. Nat Genet 37: 777–782. 56. Heitman J (2011) Microbial Pathogens in the Fungal Kingdom. Fungal Biol Rev 25. Lee IR, Chow EW, Morrow CA, Djordjevic JT, Fraser JA (2011) Nitrogen 25: 48–60. metabolite repression of metabolism and virulence in the human fungal 57. Navarathna DH, Harris SD, Roberts DD, Nickerson KW (2010) Evolutionary pathogen Cryptococcus neoformans. Genetics 188: 309–323. aspects of urea utilization by fungi. FEMS Yeast Res 10: 209–213. 26. Staib F, Mishra SK, Able T, Blisse A (1976) Growth of Cryptococcus neoformans on 58. Idnurm A, Heitman J (2005) Light controls growth and development via a uric acid agar. Zentralbl Bakteriol [Orig A] 236: 374–385. conserved pathway in the fungal kingdom. PLoS Biol 3: e95. 27. Cox GM, Mukherjee J, Cole GT, Casadevall A, Perfect JR (2000) Urease as a 59. Jung WH, Sham A, Lian T, Singh A, Kosman DJ, et al. (2008) Iron source virulence factor in experimental cryptococcosis. Infect Immun 68: 443–448. preference and regulation of iron uptake in Cryptococcus neoformans. PLoS Pathog 28. Olszewski MA, Noverr MC, Chen GH, Toews GB, Cox GM, et al. (2004) 4: e45. Urease expression by Cryptococcus neoformans promotes microvascular sequestra- 60. Enright AJ, Iliopoulos I, Kyrpides NC, Ouzounis CA (1999) Protein interaction tion, thereby enhancing central nervous system invasion. Am J Pathol 164: maps for complete genomes based on gene fusion events. Nature 402: 86–90. 1761–1771. 61. Marcotte EM, Pellegrini M, Thompson MJ, Yeates TO, Eisenberg D (1999) A 29. Shi M, Li SS, Zheng C, Jones GJ, Kim KS, et al. (2010) Real-time imaging of combined algorithm for genome-wide prediction of protein function. Nature trapping and urease-dependent transmigration of Cryptococcus neoformans in mouse 402: 83–86. brain. J Clin Invest 120: 1683–1693. 62. Hawkins AR (1987) The complex Arom locus of Aspergillus nidulans. Evidence for 30. Behrens MD, Wagner WM, Krco CJ, Erskine CL, Kalli KR, et al. (2008) The multiple gene fusions and convergent evolution. Curr Genet 11: 491–498. endogenous danger signal, crystalline uric acid, signals for enhanced antibody 63. Schwartz J, Pinilla-Ibarz J, Yuan RR, Scheinberg DA (2003) Novel targeted and immunity. Blood 111: 1472–1479. immunotherapeutic strategies in chronic myeloid leukemia. Semin Hematol 40: 31. Park BJ, Wannemuehler KA, Marston BJ, Govender N, Pappas PG, et al. (2009) 87–96. Estimation of the current global burden of cryptococcal meningitis among 64. Dueber JE, Wu GC, Malmirchegini GR, Moon TS, Petzold CJ, et al. (2009) persons living with HIV/AIDS. AIDS 23: 525–530. Synthetic protein scaffolds provide modular control over metabolic flux. Nat 32. Elion GB (1989) The purine path to chemotherapy. Science 244: 41–47. Biotechnol 27: 753–759. 33. Nielsen K, De Obaldia AL, Heitman J (2007) Cryptococcus neoformans mates on 65. Reumann S, Babujee L, Ma C, Wienkoop S, Siemsen T, et al. (2007) Proteome pigeon guano: implications for the realized ecological niche and globalization. analysis of Arabidopsis leaf peroxisomes reveals novel targeting peptides, Eukaryot Cell 6: 949–959. metabolic pathways, and defense mechanisms. Plant Cell 19: 3170–3193.

PLOS ONE | www.plosone.org 12 May 2013 | Volume 8 | Issue 5 | e64292 Uric Acid Degradation in Cryptococcus neoformans

66. Tavazzi B, Lazzarino G, Leone P, Amorini AM, Bellia F, et al. (2005) 67. Wu X, Wakamiya M, Vaishnav S, Geske R, Montgomery C Jr., et al. (1994) Simultaneous high performance liquid chromatographic separation of purines, Hyperuricemia and urate nephropathy in urate oxidase-deficient mice. Proc pyrimidines, N-acetylated amino acids, and dicarboxylic acids for the chemical Natl Acad Sci U S A 91: 742–746. diagnosis of inborn errors of metabolism. Clin Biochem 38: 997–1008. 68. Walker UA, Hoffmann C, Enters M, Thoden J, Behrens G, et al. (2006) High serum urate in HIV-infected persons: the choice of the antiretroviral drug matters. AIDS 20: 1556–1558.

PLOS ONE | www.plosone.org 13 May 2013 | Volume 8 | Issue 5 | e64292 PLoS ONE

Supporting Information http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone .0064292#s5

Characterization of the Complete Uric Acid Degradation Pathway in the Fungal Pathogen Cryptococcus neoformans

I. Russel Lee, Liting Yang, Gaseene Sebetso, Rebecca Allen, Thi H. N. Doan, Ross Blundell, Edmund Y. L. Lui, Carl A. Morrow, James A. Fraser

DOI: 10.1371/journal.pone.0064292

Figure S1. ClustalW sequence alignment of A. nidulans UaZ and C. neoformans Uro1. Identical amino acid residues are shaded dark grey while similar residues are shaded light grey. The conserved N-terminus sequence needed for enzymatic activity is boxed in red.

2 SI

Figure S2. ClustalW sequence alignment of A. nidulans UaX and C. neoformans Uro2. Identical amino acid residues are shaded dark grey while similar residues are shaded light grey. The conserved C-terminus YRGS motif that distinguishes members of the transthyretin family is boxed in red. The residues indicated with an asterisk below the sequence are conserved active site residues, as determined in the structures of the enzymes from B. subtilus, Salmonella dublin and Danio rerio.

3 SI

Figure S3. ClustalW sequence alignment of A. nidulans UaW and C. neoformans Uro3. Identical amino acid residues are shaded dark grey while similar residues are shaded light grey.

4 SI

Figure S4. ClustalW sequence alignment of S. cerevisiae Dal1 and C. neoformans Dal1. Identical amino acid residues are shaded dark grey while similar residues are shaded light grey.

5 SI

Figure S5. The C. neoformans DAL2,3,3 gene encodes a fusion allantoicase-ureidoglycolate hydrolase protein. A. Representative protein architecture of S. cerevisiae Dal2 and Dal3, and C. neoformans Dal2,3,3. B. ClustalW sequence alignment of S. cerevisiae Dal2 and C. neoformans Dal2. C. ClustalW multiple sequence alignment of S. cerevisiae Dal3 and C. neoformans Dal3a and Dal3b. Identical amino acid residues are shaded dark grey while similar residues are shaded light grey.

6 SI

Figure S6. ClustalW sequence alignment of A. nidulans UreB and C. neoformans Ure1. Identical amino acid residues are shaded dark grey while similar residues are shaded light grey.

7 SI

Figure S7. HIU hydrolase and OHCU decarboxylase activities are not entirely essential, but are required for efficient uric acid degradation in C. neoformans. Tenfold spot dilution assays for nitrogen source utilization showed that the uro1Δ, uro2Δ and uro3Δ mutants exhibited obvious growth defects on YNB supplemented with 10 mM uric acid after a two-day incubation period, however, growth of the uro2Δ and uro3Δ strains catches up on further incubation.

8 SI

Figure S8. Restoration of the ability to utilize uric acid and its pathway intermediates upon complementation of the uric acid catabolic deletion mutants. Tenfold spot dilution assays for nitrogen source utilization showed that the complemented uro1Δ + URO1, uro2Δ + URO2, uro3Δ + URO3, dal1Δ + DAL1, dal2,3,3Δ + DAL2,3,3 and ure1Δ + URE1 strains exhibited wild-type growth on YNB supplemented with uric acid, allantoin, urea or ammonium (10 mM each).

9 SI Table S1. Fungal strains used in this study.

Strain Genotype Source H99 Cryptococcus neoformans laboratory strain [1] YL1 H99 uro1::NEO This study GS1 H99 uro2::NEO This study GS5 H99 uro3::NEO This study EL24 H99 dal1::NEO This study YL4 H99 dal2,3,3::NEO This study YL5 H99 ure1::NEO This study RL22 H99 uro1::NEO + URO1 NAT This study RL23 H99 uro2::NEO + URO2 NAT This study RL15 H99 uro3::NEO + URO3 NAT This study RL13 H99 dal1::NEO + DAL1 NAT This study RL16 H99 dal2,3,3::NEO + DAL2,3,3 NAT This study RL17 H99 ure1::NEO + URE1 NAT This study

Reference

1. Toffaletti DL, Rude TH, Johnston SA, Durack DT, Perfect JR (1993) Gene transfer in Cryptococcus neoformans by use of biolistic delivery of DNA. J Bacteriol 175: 1405-1411.

10 SI Table S2. Primers used in this study.

Primer Purpose Sequence (5’-3’) UQ270 GenomeWalker adaptor GTAATACGACTCACTATAGGGC UQ271 GenomeWalker nested adaptor ACTATAGGGCACGCGTGGT UQ272 GenomeWalker 3’ gene specific CTGAATGGCGAATGAGCTTGAGCTTGG UQ273 GenomeWalker 3’ nested gene ATCAGATTGTCGTTTCCCGCCTTCAG specific UQ274 GenomeWalker 5’ gene specific GAGAGGCGGTTTGCGTATTGGCTAGAG UQ275 GenomeWalker 5’ nested gene ACGTCCGCAATGTGTTATTAAGTTGTC specific UQ248 URO1 deletion 5’ F ACATTGTTGGTAAAGCCTTCG UQ249 URO1 deletion 5’ R AGCTCACATCCTCGCAGCTATTGCTGATGGGTGTTGTTG UQ250 URO1 deletion 3’ F CCGTGTTAATACAGATAAACCTATAGATATGCAGGTGCAAAG UQ307 URO1 deletion 3’ R AGTAACAACCGTGGCTGTCAG UQ276 URO1 deletion NEO F ACAACACCCATCAGCAATAGCTGCGAGGATGTGAGCTGGAGAGCG UQ277 URO1 deletion NEO R CTGCATATCTATAGGTTTATCTGTATTAACACGGAAGAGATGTAG UQ670 URO2 deletion 5’ F TTCTGGAGTTTTCGATCCACG UQ671 URO2 deletion 5’ R AGCTCACATCCTCGCAGCACTGGAGACGCGATTGAT UQ672 URO2 deletion 3’ F TGTTAATACAGATAAACCGATCGATACTGTGTTATA UQ673 URO2 deletion 3’ R CCTTTGATGATACCTTTGTGA UQ750 URO2 deletion NEO F ATCAATCGCGTCTCCAGTGCTGCGAGGATGTGAGCTGGAGAGCG UQ751 URO2 deletion NEO R TATAACACAGTATCGATCGGTTTATCTGTATTAACACGGAAGAGATGTAG UQ674 URO3 deletion 5’ F GCCAAACCTCAAGGTTAGTCT UQ675 URO3 deletion 5’ R AGCTCACATCCTCGCAGCGATGAGCGTATGCGTGAT UQ676 URO3 deletion 3’ F TGTTAATACAGATAAACCAGCTTTAGATAGATGCTA UQ677 URO3 deletion 3’ R GGTTAGAAAACTGATCTGCCA UQ752 URO3 deletion NEO F ATCACGCATACGCTCATCGCTGCGAGGATGTGAGCTGGAGAGCG UQ753 URO3 deletion NEO R TAGCATCTATCTAAAGCTGGTTTATCTGTATTAACACGGAAGAGATGTAG UQ236 DAL1 deletion 5’ F ACCAAGCCCAAACAGAGCAGC UQ237 DAL1 deletion 5’ R AGCTCACATCCTCGCAGCGAACATAACGGATCAATCTCG UQ238 DAL1 deletion 3’ F CCGTGTTAATACAGATAAACCGCTGCACAGGAGACATTACGG UQ239 DAL1 deletion 3’ R GTCGGTCTTGGCGGAGAGGTG UQ754 DAL1 deletion NEO F TTTTTCTGCATACAAACAGCTGCGAGGATGTGAGCTGGAGAGCG UQ755 DAL1 deletion NEO R TAATGTCTCCTGTGCAGCGGTTTATCTGTATTAACACGGAAGAGATGTAG UQ240 DAL2,3,3 deletion 5’ F GCCACTGCCCAAGCTATTGAC UQ278 DAL2,3,3 deletion 5’ R CGCTCTCCAGCTCACATCCTCGCAGCTGATATAAGCTGTGGGGGACG UQ279 DAL2,3,3 deletion 3’ F CATCTCTTCCGTGTTAATACAGATAAACCATAAACGTGTTGAGAAAGCGT UQ243 DAL2,3,3 deletion 3’ R CGCCTAATAACACAGCAAATA UQ756 DAL2,3,3 deletion NEO F CCCCCACAGCTTATATCAGCTGCGAGGATGTGAGCTGGAGAGCG UQ757 DAL2,3,3 deletion NEO R CTTTCTCAACACGTTTATGGTTTATCTGTATTAACACGGAAGAGATGTAG UQ244 URE1 deletion 5’ F ACACCAACTACCGCCATTTCC UQ762 URE1 deletion 5’ R AGCTCACATCCTCGCAGCGTATCCGGGCTCGTGTAT UQ246 URE1 deletion 3’ F CCGTGTTAATACAGATAAACCGTATCAGTGCATTTCTCCATT UQ247 URE1 deletion 3’ R TGTGTTCCACAGCGTTGAGAG UQ758 URE1 deletion NEO F ATACACGAGCCCGGATACGCTGCGAGGATGTGAGCTGGAGAGCG UQ759 URE1 deletion NEO R GGAGAAATGCACTGATACGGTTTATCTGTATTAACACGGAAGAGATGTAG UQ1323 URO1 genomic DNA sequencing TCTTCCGCTTAACGACTGGAC

11 SI UQ1324 URO1 genomic DNA sequencing CGCACGAGTCGTCCGTGATGG UQ1325 URO1 genomic DNA sequencing ACGGCAAGCCTCACAAATGGT UQ1326 URO1 genomic DNA sequencing ATTTTGACTTCATGCCCTGCC UQ1327 URO2 genomic DNA sequencing TATAAAATCAAATAAAAGGTG UQ1328 URO2 genomic DNA sequencing CATCGCAGTGTTAGATTCGTC UQ1329 URO2 genomic DNA sequencing GCATTGTAATTTTAAGCAAGA UQ1330 URO3 genomic DNA sequencing CTGTTCTTTGCGACGCTTGTT UQ1331 URO3 genomic DNA sequencing CCATACTCTGCTTGTTCCATC UQ1332 URO3 genomic DNA sequencing CGTTTGGCTGATCGGAAGGGC UQ1333 DAL1 genomic DNA sequencing AGATCGTATTTCCAATCCGTC UQ1334 DAL1 genomic DNA sequencing GCGTTTTCGTGATTTTGGATT UQ1335 DAL1 genomic DNA sequencing GCCAAAGAGGTCGGGGTGAGC UQ1336 DAL1 genomic DNA sequencing CCTCTACTGCAATCTGCCTTT UQ1337 DAL1 genomic DNA sequencing CTTGTAAGTCGGGCGGTGTCT UQ1338 DAL2,3,3 genomic DNA sequencing GGCGTGGTTCTGCTCATTGGA UQ1339 DAL2,3,3 genomic DNA sequencing CGTCCCCCACAGCTTATATCA UQ1340 DAL2,3,3 genomic DNA sequencing CCTGGAGAGGCTTTCCTGTTA UQ1341 DAL2,3,3 genomic DNA sequencing TCTTCTCCCTGGTCGTGGAAT UQ1342 DAL2,3,3 genomic DNA sequencing TCCCGCTCTGCCTCTTACCCC UQ1343 DAL2,3,3 genomic DNA sequencing GCGGAGTGAGTGCGACACAAG UQ1344 DAL2,3,3 genomic DNA sequencing GGCACCCTTATACCAGTCAGA UQ1345 DAL2,3,3 genomic DNA sequencing CGAATGATTGAATACTACAGT UQ1346 URE1 genomic DNA sequencing ATTCATCTGCATTCTGGATTA UQ1347 URE1 genomic DNA sequencing ATTTTTTGCTTTCCCTTTCTT UQ1348 URE1 genomic DNA sequencing TTCCCTGCTGCACCCGAGCCA UQ1349 URE1 genomic DNA sequencing ACAACTGGCGATAAGATCAAA UQ1350 URE1 genomic DNA sequencing TCTCTAGTAGGATGGGGATCA UQ1351 URE1 genomic DNA sequencing TCGTTGTCTGTGAATATGAAA UQ1352 URE1 genomic DNA sequencing AGATGATCCTCAAGGGCGGTG UQ1353 URE1 genomic DNA sequencing AAACCTCTAGTTATCTTCCAA

12 SI Table S3. Plasmids used in this study.

Plasmid Detail Source pJAF1 NEO selectable marker in pCR2.1-TOPO [1] pCH233 NAT selectable marker in pCR2.1-TOPO A gift from Christina Hull pIRL5 H99 URO1 genomic DNA including promotor and This study terminator in pCR2.1-TOPO pIRL6 H99 URO2 genomic DNA including promotor and This study terminator in pCR2.1-TOPO pIRL7 H99 URO3 genomic DNA including promotor and This study terminator in pCR2.1-TOPO pIRL8 H99 DAL1 genomic DNA including promotor and This study terminator in pCR2.1-TOPO pIRL9 H99 DAL2,3,3 genomic DNA including promotor and This study terminator in pCR2.1-TOPO pIRL10 H99 URE1 genomic DNA including promotor and This study terminator in pCR2.1-TOPO pIRL13 H99 URO1 genomic DNA including promotor and This study terminator in pCH233 pIRL14 H99 URO2 genomic DNA including promotor and This study terminator in pCH233 pIRL15 H99 URO3 genomic DNA including promotor and This study terminator in pCH233 pIRL16 H99 DAL1 genomic DNA including promotor and This study terminator in pCH233 pIRL17 H99 DAL2,3,3 genomic DNA including promotor and This study terminator in pCH233 pIRL18 H99 URE1 genomic DNA including promotor and This study terminator in pCH233

Reference

1. Fraser JA, Subaran RL, Nichols CB, Heitman J (2003) Recapitulation of the sexual cycle of the primary fungal pathogen Cryptococcus neoformans var. gattii: implications for an outbreak on Vancouver Island, Canada. Eukaryot Cell 2: 1036-1045.

13 SI Table S4. C. neoformans uric acid catabolism defective mutants identified during the Agrobacterium- mediated insertional mutagenic screen.

Transformant Gene disrupted Insertion site 65H07 CNAG_04307.2 (URO1) Promoter 3H06 CNAG_04307.2 (URO1) CDS 28E09 CNAG_04307.2 (URO1) CDS 7H04 CNAG_06694.2 (URO2) Promoter 7H05 CNAG_06694.2 (URO2) Promoter 53C07 CNAG_01108.2 (DAL2,3,3) Intron 88A05 CNAG_05540.2 (URE1) Promoter 88A06 CNAG_05540.2 (URE1) Promoter

14 SI

CHAPTER 7

Conclusions

142 7.0 Conclusions

This thesis has presented work that involves the characterisation of key regulatory proteins and catabolic enzymes of nitrogen metabolism in Cryptococcus neoformans. The important isolate H99 was used as the foundation for this molecular research since it is the most intensively studied type strain that forms the basis for the Cryptococcus gene deletion collection and its virulence is well established in mammalian models [1,2]. Here, insights were shed into the global as well as pathway- specific nitrogen regulatory network at the genetic, transcriptional, protein and phenotypic levels. First, the mechanisms by which the GATA transcription factor Gat1/Are1 and Nmr-like protein Tar1 coordinate to regulate nitrogen metabolite repression was dissected using the secondary catabolic pathways of proline and uric acid as a reference. Second, the proline and uric acid enzymes were individually scrutinised for their catabolic functions. In addition, these regulatory proteins and enzymes were examined for their roles in virulence factor expression and pathogenesis. Together, an outline of the nutritional (nitrogen) requirements for C. neoformans growth during disease development is slowly emerging, along with the realisation that certain nitrogen regulatory proteins and catabolic enzymes are co- opted to control metabolic functions as well as elaboration of virulence attributes.

7.1 Concluding Remarks

7.1.1 Characterisation of the GATA factor Gat1/Are1

Nitrogen metabolite/catabolite repression was first observed in members of the phylum Ascomycota; this regulatory phenomenon serves to enable preferential utilisation of readily assimilated nitrogen sources such as ammonium and glutamine [3,4,5]. In Chapter 3, this physiological response was shown for the first time to be evolutionarily conserved in the basidiomycete C. neoformans. Using qRT-PCR, loss of transcriptional activation of a selected panel of catabolic enzyme-encoding genes from the uric acid and proline degradation pathways was observed whenever ammonium is present. Additionally, ammonium was demonstrated to be dominant over other nitrogen signals in regulating the production of the virulence factors capsule and melanin.

143 Given that transcription factors belonging to the GATA family control nitrogen source utilisation in the Ascomycota, comprehensive bioinformatics ranging from reciprocal BLASTp to Perl Script searches of the C. neoformans genome was employed to predict potential GATA-type ORFs. GATA factors are distinguished by

their highly conserved DNA-binding domain consisting of a Cys-X2-Cys-X17-20-Cys-

X2-Cys zinc finger motif followed by an adjacent 5-12 basic residues within the 25 beyond the most C-terminal cysteine. A total of seven of such putative ORFs were identified, all of which were subsequently deleted via homologous recombination, and only one (named GAT1/ARE1) was found to play a role in nitrogen regulation. Notably, Gat1/Are1 also impacts multiple virulence factor expression and consequently pathogenesis. Whether Gat1/Are1 directly or indirectly regulates virulence remains open to speculation. Given that it is a global transcription factor, there is a possibility that the observed gat1/are1Δ mutant’s virulence-associated phenotypes are a pleiotropic effect as a consequence of perturbation of related signaling pathways affecting pathogenesis. Virulence factor expression is therefore to a certain extent embedded in the regulation of various functions needed for adaptation and growth in a mammalian host.

7.1.2 Characterisation of the Nmr-related protein Tar1

Chapter 4 extended the investigation into the molecular mechanisms governing nitrogen metabolite repression in C. neoformans that was first begun in Chapter 3. Deletion of the NMR homolog (named TAR1) was shown to result in inappropriate derepression of secondary nitrogen catabolic pathways in the simultaneous presence of preferred nitrogen sources. In comparison to wild-type, 1) the tar1Δ mutant displayed increased sensitivity to thiourea, a toxic analog of urea, in the presence of ammonium or glutamine, 2) the tar1Δ mutant also expressed more proline oxidase transcripts when cultured in proline plus ammonium. Interestingly, in the sole presence of proline, the tar1Δ mutant expressed less proline oxidase transcripts. This novel dual role of an Nmr protein in both positive and negative regulation of nitrogen catabolism according to nitrogen quality/availability was verified through another series of qRT-PCR experiments, where the tar1Δ mutant was shown to produce less GAT1/ARE1 transcripts in the presence of proline but more GAT1/ARE1 transcripts in the presence of ammonium as compared to wild-type.

144 The mechanism by which Tar1 modulates nitrogen metabolite repression is atypical to the situation seen in the model filamentous ascomycetes Aspergillus nidulans and Neurospora crassa. Protein-protein interaction between Gat1/Are1 and Tar1 was undetectable in a yeast two-hybrid assay consistent with bioinformatics- based prediction. Like gat1/are1Δ, the tar1Δ mutant is modestly hypervirulent during murine infection. A parsimonious explanation as to why no opposing regulatory effects on pathogenesis was observed between these mutant strains may relate to the scarce nutrient availability during in vivo infection, a condition in which Tar1 is predicted to positively regulate Gat1/Are1 activity.

7.1.3 Characterisation of the proline degradation pathway

As mentioned in Chapters 3 and 4, transcriptional studies of the proline catabolic enzyme-encoding genes have played a crucial role in elucidating how Gat1/Are1 and Tar1 coordinate to regulate nitrogen metabolite repression. Chapter 5 expanded the investigation into the C. neoformans proline degradation pathway by characterising the functions of these catabolic enzymes. Expression of PUT1 and PUT5 [both encodes proline oxidase that is known to catalyse the degradation of proline to Δ1- pyrroline-5-carboxylate (P5C) / glutamate-γ-semialdehyde (GSA)] as well as PUT2 (encodes P5C dehydrogenase that is known to oxidise P5C/GSA to glutamate) was shown via qRT-PCR to be inducible in the presence of proline but not other nitrogen sources.

However, through the creation of gene deletion mutants, only Put5 and Put2 were found to be required for proline utilisation. In the presence of proline, confocal fluorescence microscopy and flow cytometry also revealed that the put2Δ mutant produced excessive mitochondrial superoxide ions that are known to induce apoptosis. This phenomenon is likely caused by P5C-proline cycling through the coupled actions of proline oxidase and P5C reductase (converts P5C to proline), allowing proline to be intensively oxidise by the proline oxidase-FAD complex that

delivers electrons to the mitochondrial electron transport chain and to O2, leading to reactive oxygen species overproduction. Additionally, Put2 is required for optimal production of major cryptococcal virulence factors such as melanin and high temperature growth. The put2Δ mutant is consequently attenuated in its ability to cause disease in mice, making this the first report demonstrating the importance of

145 P5C dehydrogenase in enabling virulence of a microorganism. While the functions of PUT5 and PUT2 are now apparent, the role of PUT1 remains unclear and open to debate whether it is a pseudogene.

7.1.4 Characterisation of the uric acid degradation pathway

In Chapter 3, transcriptional studies of a selected panel of uric acid catabolic enzyme- encoding genes led to the first definitive characterisation of nitrogen metabolite repression in a second phylum of the kingdom Fungi, the Basidiomycota. Chapter 6 expanded the investigation into C. neoformans uric acid degradation by characterising the function of every catabolic enzyme of this pathway. A dual approach of Agrobacterium-mediated insertional mutagenesis and bioinformatics led to the identification of URO1-encoded urate oxidase, URO2-encoded HIU hydrolase, URO3-encoded OHCU decarboxylase, DAL1-encoded allantoinase, DAL2,3,3- encoded allantoicase-ureidoglycolate hydrolase, and URE1-encoded urease. The enzymatic functions of all these candidates were verified through the deletion of each individual ORF, and assaying the mutants’ ability to utilise uric acid and its pathway intermediates as the sole nitrogen source. The identification of these genes confirms the complete absence of clustering of the genes involved in purine utilisation in C. neoformans, unlike those seen in many bacterial species.

7.1.5 Summary

Two distinct regulatory aspects of fungal nitrogen catabolism is required for de novo synthesis of catabolic enzymes and permeases of specific pathways: first, a global signal indicating nitrogen derepression when readily assimilated nitrogen sources are unavailable, and second, a pathway-specific signal indicating the presence of the substrate or intermediate of a particular pathway. In C. neoformans, the globally acting GATA transcription factor Gat1/Are1 is co-opted into regulating nitrogen source utilisation as well as virulence-associated phenotypes. In response to nitrogen quality/availability, GAT1/ARE1 transcription is tightly controlled by the activity of the Nmr-related protein Tar1. Gaining such insights into the process of nitrogen sensing is key to elucidating how C. neoformans adapts and survive its transition from nutrient-rich ecological niche of pigeon guano to nutrient-poor environment of a mammalian host during infection.

146 Potential nitrogen sources available in a human host include proline and uric acid. The enzymes PUT5-encoded proline oxidase and PUT2-encoded P5C dehydrogenase are required for complete degradation of proline to glutamate. Importantly, P5C dehydrogenase is also a critical regulator of mitochondrial redox homeostasis and virulence in C. neoformans. Complete uric acid degradation to ammonia is more complicated than proline catabolism as it involves numerous enzymes: URO1-encoded urate oxidase, URO2-encoded HIU hydrolase, URO3- encoded OHCU decarboxylase, DAL1-encoded allantoinase, DAL2,3,3-encoded allantoicase-ureidoglycolate hydrolase, and URE1-encoded urease. Among these uric acid enzymes, urease is the most well characterised nitrogen-scavenging enzyme that plays an important role during pathogenesis. Understanding the nitrogen requirements for C. neoformans growth and proliferation is pivotal as this may contribute to novel strategies for pathogen control.

7.2 Future Directions

This thesis has presented investigations into aspects of C. neoformans biology and pathogenesis through the context of nitrogen regulation. Due to the broad and exploratory nature of this work, exciting avenues of future research present themselves and are briefly discussed in accordance to each results chapter. Additionally, the significance of examining the arginine catabolic pathway that is connected to, and inter-links the uric acid and proline pathways, is discussed. Indeed, many of its disease features are still hidden from us; work to understand and control cryptococcosis must continue.

7.2.1 Chromatin immunoprecipitation to determine the DNA- binding sites of GATA factors in the genome

To date, five GATA factors (Gat1/Are1, Bwc2, Cir1, Gat201, Gat204) have been functionally characterised in C. neoformans, all of which play distinct roles in various growth, developmental and survival processes [2,6,7,8,9,10]. Intriguingly, a common theme shared by these GATA factors is their reciprocal influence on the production of well-recognised virulence factors and their impact on pathogenesis, a finding that reinforces the complex regulation of these factors. The coordinated regulation by these regulatory proteins of functions for adaptation, nutrient acquisition and

147 elaboration of virulence factors is therefore a fascinating area for future investigation. Gat1/Are1, Bwc2, Cir1, Gat201, Gat204 as well as the two other uncharacterised GATA factors (Broad Institute annotation: CNAG_03401.2 and CNAG_04263.2) are all predicted to bind to the sequence 5’-GATA-3’ [11,12,13]. Whether these GATA factors antagonise and compete for the same 5’-GATA-3’ binding sites in the promoter regions of genes, or act in synergy to activate transcription of genes, or act interchangeably, remains open to investigation. Given the likelihood that Gat1/Are1, Bwc2, Cir1, Gat201 and Gat204 are acting in parallel to regulate multiple virulence pathways in C. neoformans, investigating the DNA-binding specificities of GATA factors in the genome via chromatin immunoprecipitation may provide insights into how these factors coordinate to regulate pathogenesis during infection of a mammalian host.

7.2.2 Whole transcriptome sequencing to identify novel virulence- and stress adaptation-associated genes that are regulated by Tar1 and Gat1/Are1

The TAR1 promoter contains putative heat shock elements suggesting that Tar1 may not be exclusively involved in regulation of nitrogen catabolism but may also control targets involved in a broader spectrum of metabolic processes. Transcriptional profiling and genetic studies are invaluable research tools with the means to provide insights into C. neoformans adaptation to key features within a mammalian host environment including nitrogen availability. Using whole transcriptome sequencing, novel virulence- and stress adaptation-associated genes that are regulated by Tar1 (and also Gat1/Are1 given that it is a global transcription factor) can potentially be identified.

7.2.3 Metabolic profiling of the proline catabolic deletion mutants using gas chromatography coupled to mass spectrometry

In order to test our hypothesis, certain assumptions were made about the generated C. neoformans proline catabolic deletion mutants based on findings from other biological systems. For instance, intracellular proline was assumed to accumulate in the proline oxidase deletion mutant when subjected to external supply of proline, while an anticipated P5C-proline cycling was predicted to occur in the P5C

148 dehydrogenase deletion mutant. To validate these assumptions, gas chromatography coupled to mass spectrometry can be performed to metabolically profile the proline and P5C contents in the mutant strains [14,15].

7.2.4 Virulence assays of the uric acid catabolic deletion mutants using urate oxidase-deficient mice as an animal model system

In humans, the final end product of purine catabolism is uric acid due to a dysfunctional urate oxidase gene caused by deleterious mutations in our hominoid ancestors [16,17]. Intriguingly, necrotic mammalian cells have been shown to release uric acid as a danger signal to activate both innate and adaptive immune effectors including neutrophils, cytotoxic T-cell and antibody immunity [18]. To date, all conducted murine virulence assays of Cryptococcus employ mice with an intact urate oxidase enzyme. In order to better mimic the actual biological and immunological system of humans, urate oxidase-deficient (hyperuricemia) mice can be tested as a new animal model for investigating cryptococcosis [19]. This is particularly relevant for characterising the function of C. neoformans urate oxidase and its potential role in modulating host immune response.

7.2.5 Investigating the arginine degradation pathway

In the model yeast Saccharomyces cerevisiae, hydrolysis of arginine occurs via the enzymatic activity of CAR1-encoded arginase, producing urea and ornithine [20]. The ornithine is transaminated by CAR2-encoded ornithine amino transferase into GSA/P5C and glutamate [20]. P5C is then subsequently converted into proline by PRO3-encoded proline reductase [21]. Arginine catabolism takes place in the cytosol while proline degradation occurs in the mitochondria.

Bioinformatic analyses revealed that homologs of CAR1, CAR2 and PRO3 are all present in the C. neoformans genome. A scheme representing the predicted arginine, uric acid and proline catabolic pathways is depicted in Figure 7-1. Notably, two characterised enzymes acting on a byproduct and an intermediate of arginine catabolism are essential for C. neoformans virulence. First, urease is a bona fide virulence factor that catalyses the hydrolysis of urea and is known to play a role in host central nervous system invasion [1,22,23]. Second, P5C dehydrogenase acts on

149 GSA/P5C and is also required for pathogenesis. Therefore, arginine catabolism is worth exploring, as it may also be important for C. neoformans virulence.

Figure 7-1. The predicted pathways of uric acid (black arrows), arginine (grey arrows) and proline (blue arrows), as well as central nitrogen metabolism (green arrows), in C. neoformans. This scheme represents all the nitrogen catabolic enzymes that have been transcriptionally and/or genetically analysed, as reviewed in this thesis. Note that deletion studies on the arginase, ornithine amino transferase and proline reductase-encoding genes are currently underway.

In support of this hypothesis, genes for arginine metabolism in the pathogenic yeast Candida albicans have been shown to be upregulated following internalisation by macrophages [24]. Furthermore, in the presence of urea or arginine, dur1,2Δ mutants lacking urea amidolyase are defective in germ tube formation leading to their inability to escape from macrophages following phagocytosis [25]. In this regard, dimorphic transition from yeast to filamentous hyphal growth form is needed for

150 piercing the macrophages’ cell membrane to facilitate escape. Therefore, study of arginine catabolism in C. neoformans is likely to provide deeper insights into how this fungus employs nitrogen scavenging to modulate host immune response during

infection.

7.3 Final Conclusions

With the growth of immunocompromised patient populations due to the AIDS pandemic and increase in immune system-suppressing medical interventions, the incidence of secondary infection from pathogenic fungi has risen sharply to become a significant cause of morbidity and mortality. Every year, more than a million immunocompromised individuals are infected with the basidiomycete fungus C. neoformans, two-thirds of whom will eventually succumb to the infection [26]. This “sugar-coated” yeast is versatile, complex and ruthless, and it certainly deserves our respect and attention. Yet, fungal infections generally do not beget widespread public support, allure millionaire philanthropist benefactors or secure supranational organisation-coordinated global health campaigns as readily as more promotable infectious diseases such as malaria, dengue fever or tuberculosis.

This thesis is dedicated to researchers of fungal pathogenesis to continually strive to elucidate the molecular mechanisms that enable fungal virulence and to be innovative with identifying novel targets for therapeutic intervention. Multiple features enable successful establishment of disease in a host including the ability to metabolise the available nutrients such as nitrogen and carbon sources. A comprehensive understanding of the requirements for fungal growth and proliferation in vivo may contribute to new strategies for pathogen control. Building upon the solid foundation gained from these nitrogen catabolism studies, work on host-pathogen interactions should be adopted in the near future to dissect tissue-specific nutritional requirements in order to understand and hopefully block the striking neurotropism of C. neoformans.

7.4 References

1. Cox GM, Mukherjee J, Cole GT, Casadevall A, Perfect JR (2000) Urease as a virulence factor in experimental cryptococcosis. Infect Immun 68: 443-448.

151 2. Liu OW, Chun CD, Chow ED, Chen C, Madhani HD, et al. (2008) Systematic genetic analysis of virulence in the human fungal pathogen Cryptococcus neoformans. Cell 135: 174-188. 3. Magasanik B, Kaiser CA (2002) Nitrogen regulation in Saccharomyces cerevisiae. Gene 290: 1-18. 4. Marzluf GA (1997) Genetic regulation of nitrogen metabolism in the fungi. Microbiol Mol Biol Rev 61: 17-32. 5. Wong KH, Hynes MJ, Davis MA (2008) Recent advances in nitrogen regulation: a comparison between Saccharomyces cerevisiae and filamentous fungi. Eukaryot Cell 7: 917-925. 6. Chun CD, Brown JC, Madhani HD (2011) A major role for capsule-independent phagocytosis-inhibitory mechanisms in mammalian infection by Cryptococcus neoformans. Cell Host Microbe 9: 243-251. 7. Idnurm A, Heitman J (2005) Light controls growth and development via a conserved pathway in the fungal kingdom. PLoS Biol 3: e95. 8. Jung WH, Sham A, White R, Kronstad JW (2006) Iron regulation of the major virulence factors in the AIDS-associated pathogen Cryptococcus neoformans. PLoS Biol 4: e410. 9. Kmetzsch L, Staats CC, Simon E, Fonseca FL, Oliveira DL, et al. (2011) The GATA-type transcriptional activator Gat1 regulates nitrogen uptake and metabolism in the human pathogen Cryptococcus neoformans. Fungal Genet Biol 48: 192-199. 10. Lee IR, Chow EW, Morrow CA, Djordjevic JT, Fraser JA (2011) Nitrogen metabolite repression of metabolism and virulence in the human fungal pathogen Cryptococcus neoformans. Genetics 188: 309-323. 11. Ko LJ, Engel JD (1993) DNA-binding specificities of the GATA transcription factor family. Mol Cell Biol 13: 4011-4022. 12. Merika M, Orkin SH (1993) DNA-binding specificity of GATA family transcription factors. Mol Cell Biol 13: 3999-4010. 13. Ravagnani A, Gorfinkiel L, Langdon T, Diallinas G, Adjadj E, et al. (1997) Subtle hydrophobic interactions between the seventh residue of the zinc finger loop and the first base of an HGATAR sequence determine promoter-specific recognition by the Aspergillus nidulans GATA factor AreA. EMBO J 16: 3974-3986.

152 14. Brosche M, Vinocur B, Alatalo ER, Lamminmaki A, Teichmann T, et al. (2005) Gene expression and metabolite profiling of Populus euphratica growing in the Negev desert. Genome Biol 6: R101. 15. Miller G, Honig A, Stein H, Suzuki N, Mittler R, et al. (2009) Unraveling delta1- pyrroline-5-carboxylate-proline cycle in plants by uncoupled expression of proline oxidation enzymes. J Biol Chem 284: 26482-26492. 16. Oda M, Satta Y, Takenaka O, Takahata N (2002) Loss of urate oxidase activity in hominoids and its evolutionary implications. Mol Biol Evol 19: 640-653. 17. Vogels GD, Van der Drift C (1976) Degradation of purines and pyrimidines by microorganisms. Bacteriol Rev 40: 403-468. 18. Behrens MD, Wagner WM, Krco CJ, Erskine CL, Kalli KR, et al. (2008) The endogenous danger signal, crystalline uric acid, signals for enhanced antibody immunity. Blood 111: 1472-1479. 19. Wu X, Wakamiya M, Vaishnav S, Geske R, Montgomery C, Jr., et al. (1994) Hyperuricemia and urate nephropathy in urate oxidase-deficient mice. Proc Natl Acad Sci U S A 91: 742-746. 20. Middelhoven WJ (1964) The Pathway of Arginine Breakdown in Saccharomyces cerevisiae. Biochim Biophys Acta 93: 650-652. 21. Brandriss MC, Falvey DA (1992) Proline biosynthesis in Saccharomyces cerevisiae: analysis of the PRO3 gene, which encodes delta 1-pyrroline-5- carboxylate reductase. J Bacteriol 174: 3782-3788. 22. Olszewski MA, Noverr MC, Chen GH, Toews GB, Cox GM, et al. (2004) Urease expression by Cryptococcus neoformans promotes microvascular sequestration, thereby enhancing central nervous system invasion. Am J Pathol 164: 1761-1771. 23. Shi M, Li SS, Zheng C, Jones GJ, Kim KS, et al. (2010) Real-time imaging of trapping and urease-dependent transmigration of Cryptococcus neoformans in mouse brain. J Clin Invest 120: 1683-1693. 24. Lorenz MC, Bender JA, Fink GR (2004) Transcriptional response of Candida albicans upon internalization by macrophages. Eukaryot Cell 3: 1076-1087. 25. Ghosh S, Navarathna DH, Roberts DD, Cooper JT, Atkin AL, et al. (2009) Arginine-induced germ tube formation in Candida albicans is essential for escape from murine macrophage line RAW 264.7. Infect Immun 77: 1596- 1605.

153 26. Park BJ, Wannemuehler KA, Marston BJ, Govender N, Pappas PG, et al. (2009) Estimation of the current global burden of cryptococcal meningitis among persons living with HIV/AIDS. AIDS 23: 525-530.

154

Appendix A

A Unique Chromosomal Rearrangement in the Cryptococcus neoformans var. grubii Type Strain Enhances Key Phenotypes Associated with Virulence

155 Downloaded from

A Unique Chromosomal mbio.asm.org Rearrangement in the Cryptococcus neoformans var. grubii Type Strain

Enhances Key Phenotypes Associated on February 28, 2012 - Published by with Virulence

Carl A. Morrow, I. Russel Lee, Eve W. L. Chow, et al. 2012. A Unique Chromosomal Rearrangement in the Cryptococcus neoformans var. grubii Type Strain Enhances Key Phenotypes Associated with Virulence . mBio 3(2): . doi:10.1128/mBio.00310-11. mbio.asm.org Updated information and services can be found at: http://mbio.asm.org/content/3/2/e00310-11.full.html

SUPPLEMENTAL http://mbio.asm.org/content/3/2/e00310-11.full.html#SUPPLEMENTAL MATERIAL REFERENCES This article cites 19 articles, 11 of which can be accessed free at: http://mbio.asm.org/content/3/2/e00310-11.full.html#ref-list-1 CONTENT ALERTS Receive: RSS Feeds, eTOCs, free email alerts (when new articles cite this article), more>>

Information about commercial reprint orders: http://mbio.asm.org/misc/reprints.xhtml Information about Print on Demand and other content delivery options: http://mbio.asm.org/misc/contentdelivery.xhtml To subscribe to another ASM Journal go to: http://journals.asm.org/subscriptions/

Downloaded from

RESEARCH ARTICLE

A Unique Chromosomal Rearrangement in the Cryptococcus mbio.asm.org neoformans var. grubii Type Strain Enhances Key Phenotypes

Associated with Virulence on February 28, 2012 - Published by

Carl A. Morrow,a,b I. Russel Lee,a,b Eve W. L. Chow,a,b Kate L. Ormerod,a,b Anita Goldinger,b Edmond J. Byrnes III,c Kirsten Nielsen,d Joseph Heitman,e Horst Joachim Schirra,b,f and James A. Frasera,b Australian Infectious Diseases Research Centre, University of Queensland, Brisbane, Australiaa; School of Chemistry and Molecular Biosciences, University of Queensland, Brisbane, Australiab; Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland, USAc; Department of Microbiology, Medical School, University of Minnesota, Minneapolis, Minnesota, USAd; Department of Molecular Genetics and Microbiology, Duke University Medical Center, Durham, North Carolina, USAe; and Centre for Advanced Imaging, University of Queensland, Brisbane, Australiaf

ABSTRACT The accumulation of genomic structural variation between closely related populations over time can lead to reproductive isolation and speciation. The fungal pathogen Cryptococcus is thought to have recently diversified, forming a species complex containing members with distinct morphologies, distributions, and pathologies of infection. We have investigated structural changes in genomic architecture such as inversions and translocations that distinguish the most pathogenic variety, mbio.asm.org Cryptococcus neoformans var. grubii, from the less clinically prevalent Cryptococcus neoformans var. neoformans and Cryptococ- cus gattii. Synteny analysis between the genomes of the three Cryptococcus species/varieties (strains H99, JEC21, and R265) reveals that C. neoformans var. grubii possesses surprisingly few unique genomic rearrangements. All but one are relatively small and are shared by all molecular subtypes of C. neoformans var. grubii. In contrast, the large translocation peculiar to the C. neoformans var. grubii type strain is found in all tested subcultures from multiple laboratories, suggesting that it has possessed this rearrangement since its isolation from a human clinical sample. Furthermore, we find that the translocation directly disrupts two genes. The first of these encodes a novel protein involved in metabolism of glucose at human body temperature and affects intracellular levels of trehalose. The second encodes a homeodomain-containing transcription factor that modulates melanin production. Both mutations would be predicted to increase pathogenicity; however, when recreated in an alternate genetic background, these mutations do not affect virulence in animal models. The type strain of C. neoformans var. grubii in which the majority of molecular studies have been performed is therefore atypical for carbon metabolism and key virulence attributes. IMPORTANCE The fungal pathogen Cryptococcus is a major cause of mortality among the immunocompromised population, primarily in AIDS patients of sub-Saharan Africa. Most research into the particular variety of Cryptococcus responsible for the vast majority of infections, Cryptococcus neoformans var. grubii, is performed using the type strain isolated in 1978 from a Hodgkin’s disease patient from North Carolina. We have determined that this particular isolate contains a chromosomal translocation that directly interrupts two genes, which all descendants of this strain from various research laboratories appear to possess. Disrup- tion of these two genes affects multiple virulence factors of Cryptococcus, particularly the ability to grow at human body temperature, which could have wide-ranging implications for molecular genetic studies and virulence assays using this important strain.

Received 29 December 2011 Accepted 5 January 2012 Published 28 February 2012 Citation Morrow CA, et al. 2012. A unique chromosomal rearrangement in the Cryptococcus neoformans var. grubii type strain enhances key phenotypes associated with virulence. mBio 3(2):e00310-11. doi:10.1128/mBio.00310-11. Editor Judith Berman, University of Minnesota Copyright © 2012 Morrow et al. This is an open-access article distributed under the terms of the Creative Commons Attribution-Noncommercial-Share Alike 3.0 Unported License, which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original author and source are credited. Address correspondence to James A. Fraser, [email protected].

ncreasing numbers of completed genome sequences have made tion of mutations, reproductive isolation without gene ﬂow in Iapparent that large changes in genome structure play a much freely mixing populations, local adaptation, and specialization greater role in the generation of phenotypic diversity between spe- (2). Direct association of chromosomal rearrangements with local cies than previously appreciated. While small mutations such as adaptation and speciation has since been demonstrated for fungi single nucleotide polymorphisms have long been associated with (3), plants (4), and insects (5). Both the ubiquity of structural disease states and phenotypic variation in humans, larger struc- variants within individual human genomes and contributions to tural alterations have recently been revealed to be at least as im- various disease states have also been demonstrated (6). portant in generating variation (1). Furthermore, recent theories Structural variants or gross chromosomal rearrangements can propose that the suppression of recombination associated with have profound inﬂuences on phenotype and encompass large large chromosomal rearrangement events allows for accumula- changes such as intra- and interchromosomal translocations,

March/April 2012 Volume 3 Issue 2 e00310-11 ¨ mbio.asm.org 1 Downloaded from

Morrow et al. large duplications and deletions, copy number variations, inver- strain of C. neoformans var. grubii and is shared by all subcultures sions, and whole-chromosome aneuploidies. Seminal studies tested. We find that it directly interrupts two genes, one encoding mbio.asm.org from Drosophila species have shown that inversions can delineate a protein involved in glucose metabolism during high- closely related species (7), while inversions characterize the sex- temperature growth and the other encoding a homeodomain- determining chromosomes in a variety of species, including hu- containing transcription factor that represses melanization. De- mans (8, 9). Translocations can suppress recombination and se- spite the fact that these two genes impact multiple virulence on February 28, 2012 - Published by quester sets of genes in a manner similar to that of inversions and factors, pathogenicity is unaffected in animal models. As H99 is can alter gene expression around breakpoints and along their the most intensively studied strain of Cryptococcus worldwide, this length through tertiary chromosome positioning (10). Large du- could have significant impacts on genetic and virulence studies in plications of genes allow for specialization and sub- and neofunc- this important human pathogen. tionalization of duplicate copies, with the most famous being the whole-genome duplications of the Saccharomycetaceae yeasts and RESULTS the teleost fishes (11, 12). Synteny analysis reveals surprisingly few genome rearrange- The opportunistic fungal pathogen Cryptococcus neoformans is ments unique to C. neoformans var. grubii. To develop an under- a major cause of mortality in the immunocompromised popula- standing of the evolutionary significance and clinical conse- tion, causing an estimated 625,000 deaths per annum, primarily in quences of large-scale genomic rearrangements in C. neoformans areas where AIDS is endemic (13). The species consists of two var. grubii, we performed pairwise comparisons using dot plot varieties that diverged approximately 24.5 million years ago: Cryp- analysis on three of the available Cryptococcus pathogenic species tococcus neoformans var. grubii, which causes the vast majority of complex genomes: C. neoformans var. grubii (strain H99), C. neo- mbio.asm.org infections, and Cryptococcus neoformans var. neoformans, which is formans var. neoformans (strain JEC21), and C. gattii (strain far less clinically prevalent. The pathogenic Cryptococcus species R265). The widely distributed type strain H99 was isolated in 1978 complex is further demarcated into molecular subtypes: C. neo- from the cerebrospinal fluid of a 27-year-old Caucasian male be- formans var. grubii comprises subtypes VNI, VNII, and VNB while ing treated for Hodgkin’s disease in North Carolina by John Per- C. neoformans var. neoformans is subtype VNIV. The sister species fect (23). Strain H99 is an important isolate as it is the foundation Cryptococcus gattii, which diverged 45 million years ago, com- of molecular research in Cryptococcus. Pairwise analysis revealed prises the subtypes VGI, VGII, VGIII, and VGIV and is generally broad synteny across the 14 chromosomes of the genomes, with considered a primary pathogen of healthy individuals, although the notable exceptions being the centromeres, the subtelomeric both species have the capacity to infect immunocompetent and regions, and MAT. Large genomic rearrangements (greater than immunodeficient patients (14). The three groupings (C. neofor- 1 kb) specific to C. neoformans var. grubii numbered eight in total. mans var. grubii, C. neoformans var. neoformans, and C. gattii) Six inversions were identified, two on chromosome 1 and one display marked differences in distribution, ecology, morphology, each on chromosomes 3, 7, 9, and 14; in addition, a small dupli- epidemiology, and virulence. cation is associated with the inversion on chromosome 14 and a Whether the two C. neoformans varieties represent separate large reciprocal translocation is present involving chromosomes 3 species is currently under debate, as are proposals to raise the C. and 11 (Fig. 1). The ancestral state of chromosomes 3 and 11 is gattii molecular subtypes to the variety or perhaps species level likely represented by the chromosomal arrangement in R265; (15, 16). Recent comparisons of completed Cryptococcus genome JEC21 has undergone a pericentric inversion on chromosome 3, sequences have highlighted rearrangements between C. neofor- while H99 has undergone a translocation between chromosomes 3 mans var. neoformans and C. neoformans var. grubii (17, 18), be- and 11. Rearrangements peculiar to each genome were evident tween C. gattii VGI and C. gattii VGII, and between C. gattii through our analyses, including previously identified events in C. VGI/II and C. neoformans var. neoformans (16). As C. neoformans neoformans var. neoformans: the introgression of C. neoformans var. grubii is far more prevalent than the other two organisms in var. grubii-specific sequence (the identity island) and a duplica- cryptococcal infections (Ն90%) and causes the majority of deaths tion and translocation involving JEC21 chromosomes 8 and 12 worldwide, we were interested in structural variants at the chro- (17, 24). mosomal level that were unique to C. neoformans var. grubii. These findings are consistent with those of Sun and Xu (18), Changes in karyotype arise frequently in Cryptococcus during in- who examined chromosomal rearrangements between C. neofor- fection, suggesting that it may be a common mechanism to gen- mans var. grubii and C. neoformans var. neoformans and identified erate variation and adapt to the host environment (19, 20). Similar five translocations and 27 rearrangements. Our inclusion of C. genomic microevolution has been linked with success of infection gattii establishes that many of those changes are unique to C. neo- in fungal pathogens such as Candida albicans (21, 22). Interrogat- formans var. neoformans. The five translocations comprise the ing such changes may elucidate modes of speciation in the patho- telomere-telomere fusion and associated duplication identified by genic Cryptococcus species complex or suggest genetic mecha- Fraser et al. (24), the pericentric inversion on JEC21 chromosome nisms for the prevalence and virulence of C. neoformans var. 3, and the chromosome 3/11 translocation in H99. There are seven grubii. C. neoformans var. neoformans-specific inversions, two on chro- In this study, we compared the sequenced genome of the C. mosome 1, two on chromosome 4, and one each on chromosomes neoformans var. grubii type strain, H99, with those of the C. neo- 5, 7, and 14. Both C. neoformans var. grubii and C. neoformans var. formans var. neoformans strain JEC21 and the C. gattii strain R265 neoformans possess separate inversions on chromosome 7 that using synteny analysis to identify translocations, duplications, de- share a common breakpoint. The remaining rearrangements letions, and inversions. We found that C. neoformans var. grubii identified in C. neoformans var. neoformans comprise the centro- possesses very few unique genomic rearrangements; however, the meres and MAT. single large translocation identified appears restricted to the type For the six C. neoformans var. grubii-unique inversions, five

2 ¨ mbio.asm.org March/April 2012 Volume 3 Issue 2 e00310-11 Downloaded from

Genomic Rearrangements in C. neoformans var. grubii mbio.asm.org on February 28, 2012 - Published by mbio.asm.org

FIG 1 Inversions unique to the C. neoformans var. grubii genome. Green regions are syntenic with the C. neoformans var. neoformans and C. gattii genomes, while red regions are inverted and yellow indicates the duplication. Genes are represented by black arrows indicating direction of transcription, while white arrowheads represent transposable or repeated elements near inversion breakpoints. Images are not to scale. CoA, coenzyme A.

encompass only two to four genes, while one inversion on chro- Nucleotide dS rates are uniform across the chromosomal in- mosome 1 captures 34 genes (Fig. 1). Most of the inversion break- versions. Chromosomal inversions lead to local suppression of points are intergenic, and all except the inversion on chromosome recombination in affected chromosomal regions (25). For the C. 9 are associated with a repeated or transposable element. One each neoformans var. neoformans/C. neoformans var. grubii split, we de- of the breakpoints on chromosomes 1, 7, and 14 is within genes termined the dS (synonymous substitution) values for all genes and creates novel 3 ends for the genes encoding DNA topoisom- contained within the C. neoformans var. grubii-specific inversions erase II Top2, the= replication checkpoint component Tof1, and and compared them to those of flanking genes on both sides of the elongation factor 1A Tif11. The inversion on H99 chromosome 14 inversions. All six inversions had average dS values that were com- also has an associated inverted duplication of 5,567 bp, which parable to that of flanking regions (see Fig. S2 in the supplemental encompasses three genes. material). This suggests that the C. neoformans var. grubii com- The six inversions identified in H99 appear to be a synapo- mon ancestor may have been geographically isolated and not have morphy of C. neoformans var. grubii. Genomic inversions can frequently undergone sexual reproduction with the C. neoformans lead to altered expression of genes surrounding the breakpoints var. neoformans ancestor, as there is no evidence of suppressed and contribute to speciation via suppression of homologous re- recombination across these regions. Furthermore, all inversion- combination. To investigate whether the six inversions identified associated genes had average dN/dS ratios (ratios of nonsynony- in the C. neoformans var. grubii genome sequence are unique to mous to synonymous substitutions) of Ͻ1, where values of Ͼ1 strain H99 (a VNI isolate), we designed PCR assays to test for their represent positive selection and values of Ͻ1 represent purifying presence or absence. All subcultures and progeny of H99 pos- selection (Fig. S2). Again, these ratios were similar to those of sessed the duplication and six inversions (see Fig. S1 in the sup- flanking genes. plemental material). The seven known amplified fragment length The H99 translocation shows evidence of being produced via polymorphism (AFLP) subclades of molecular type VNI also pos- nonhomologous end joining and disrupts two genes. Of all the sessed these seven rearrangements, as do molecular types VNII changes identified, only the translocation between chromosomes and VNB. Both C. neoformans var. neoformans VNIV and C. gattii 3 and 11 appeared to be H99 specific. To investigate the origin of VGII lack these features, confirming all seven traits to be charac- this genomic event, we interrogated an array of 80 diverse isolates teristic of C. neoformans var. grubii. via diagnostic PCR. The translocation was present exclusively in

March/April 2012 Volume 3 Issue 2 e00310-11 ¨ mbio.asm.org 3 Downloaded from

Morrow et al. mbio.asm.org on February 28, 2012 - Published by mbio.asm.org

FIG 2 Identification and reconstruction of the H99 translocation. (A) Diagnostic PCR assay spanning the translocation boundary designed to produce an amplicon only when the translocation is present. Extended testing establishes that the translocation is present in all subcultures of H99 tested but absent from all other strains. (B) Alignment of the translocation breakpoints from H99, JEC21, R265, and VNI isolate 125.91. Note the presence of a microhomology, AGC, at the breakpoint, along with a single-base-pair insertion, A, before the microhomology on H99 chromosome 11. In silico reversal of the translocation and removal of the base pair insertion yields two complete ORFs without frameshift or nonsense mutations. subcultures of strain H99 and its derivatives (Fig. 2 and results not contains three introns, and yields a predicted protein of 300 amino shown), suggesting that it was present in the originally isolated acids. BLAST analysis reveals that the gene is widespread in the H99 clinical sample. To more precisely characterize the chromo- Agaricomycotina, appearing in 20 of 27 available genomes, but some 3/11 H99 translocation, the breakpoint regions were ampli- appears absent from all other fungal species. There are no charac- fied and sequenced from the VNI isolate 125.91 (Fig. 2). There is a terized homologs and no predicted protein domains or secondary 3-bp microhomology at the breakpoint on both chromosomes, structure. plus a single-base-pair insertion before the microhomology on A deletion mutant constructed in VNB isolate Bt63 was tested chromosome 11, hallmarks of the translocation arising from non- with a wide variety of media, conditions, and stressors, including homologous end joining (26). BLAST analysis of the breakpoint various carbon and nitrogen sources; extremes of pH and ionic suggests that the translocation interrupts two genes such that the and osmotic stress; oxidative, nitrosative, and sulfhydryl stress; 5 ends of both genes converge on chromosome 3 while the 3 ends low-iron and low-phosphate media; metal and metalloid stres- face= away from each other on chromosome 11 (Fig. 2). If= tran- sors; extremes of temperature; UV resistance; cell wall- and cell scribed, the 5 fragments of both genes on the H99 chromosome membrane-perturbing agents; and CO2 and O2 stress. All tests would form relatively= short truncated proteins appended with se- were performed at both 30°C and 37°C to discriminate pheno- quence from the antisense strands of the other gene. Careful rean- types that may be evident only during infection of a mammalian notation of these genes with reference to a number of fungal ge- host. Over 70 conditions were tested in total. The majority of these nomes suggests that both have been misannotated in the existing tests were uninformative. However, although the mutant with the Cryptococcus genomes. deletion of this first translocation gene in Bt63 had no phenotype One gene disrupted in the H99 translocation, TGR1, is in- at 30°C when cultured on minimal medium plus glucose, at 37°C volved in metabolism of glucose at high temperature. Owing to the mutant showed increased growth compared to that of the wild the translocation, the shorter of the two translocation-associated type. We therefore dubbed the gene TGR1, for temperature- genes is annotated as two separate open reading frames (ORFs) in sensitive glucose-specific growth repressor 1 (Fig. 3). The pheno- the unpublished Broad Institute H99 assembly, with the 3 end on type was specific to the presence of glucose and was abolished chromosome 11 (CNAG_01857) and the 5 end on chromosome= when the mutant was grown on other sole carbon sources includ- 3 (CNAG_02713). The same gene is located= on chromosome 3 in ing mono-, di-, and trisaccharides; acetate; ethanol; glycerol; JEC21 (XM_569672), chromosome 3 in WM276 (CGB_C5260C), amino acids; and fatty acids (Fig. 3 and results not shown). The and supercontig 8 of R265 (CNBG_3377). The gene is 1,063 bp, phenotype was also apparent when glucose was present and when

4 ¨ mbio.asm.org March/April 2012 Volume 3 Issue 2 e00310-11 Downloaded from

Genomic Rearrangements in C. neoformans var. grubii mbio.asm.org on February 28, 2012 - Published by mbio.asm.org

FIG 3 Reverse genetics in strain Bt63 and phenotypic characterization reveal roles for the two genes disrupted by the translocation observed in strain H99. (A) The two genes disrupted by the translocation, designated YHP2 and TGR1, were deleted via homologous recombination in the VNB isolate Bt63. Tenfold serial dilutions of each strain were spotted onto YNB plates supplemented with the indicated carbon source, stressor, or osmotic stabilizer. The complemented strain Bt63 tgr1⌬ TGR1 displayed the wild-type Bt63 phenotype for all assays where it is not shown. (B) Polysaccharide capsule production visualized as a clear halo around the cell in an India ink stain. Bar ϭ 10 ␮m. (C) Melanization of strains on L-DOPA-containing medium. Melanized Cryptococcus cells turn brown-black. other variables were altered, such as culture pH or nitrogen source the fungal cell wall by binding to the assembling ␤-1,3-glucan (results not shown). Reperforming all other phenotypic tests with ﬁbrils and preventing their close interactions, resulting in a loss of galactose substituted for glucose revealed no additional changes. rigidity (27). Diamide oxidizes cellular thiol groups, especially In the presence of both glucose and any other carbon source protein cysteine sulfhydryl bonds, and induces pore formation in tested, the glucose effect was dominant and growth was enhanced the cell wall (28). Consistent with these effects, supplementation in the mutant at 37°C (Fig. 3). Reintroduction of TGR1 into the of the medium with the osmotic stabilizer sorbitol, mannitol, or Bt63 tgr1⌬ strain yielded a wild-type phenotype under all condi- sucrose ablated high-temperature differences in growth between tions tested. To ensure that the truncated tgr1⌬-309 gene present the wild type and mutant on glucose and galactose, as well as in H99 was nonfunctional, we cloned the tgr1⌬-309/yhp2⌬-588 growth on Congo red, suggesting that the phenotype is related to fusion construct from H99 and transformed it into the Bt63 tgr1⌬ cellular integrity (Fig. 3). strain. The truncated version was unable to complement the phe- The second gene disrupted, YHP2, encodes a homeodomain- notype of the Bt63 tgr1⌬ strain on glucose at 37°C (results not containing transcription factor. The second and larger of the two shown). translocation-associated genes is located on chromosome 11 in As extremes of temperature can affect cell wall and cell mem- JEC21 (XM_567674), chromosome 11 in WM276 brane integrity, we tested the mutant on both glucose and galac- (CGB_K2020C), and supercontig 7 of R265 (CNBG_3081). The tose at 30°C and 37°C on compounds that perturb these struc- gene is annotated as a single ORF at the Broad Institute (3 end tures. No common changes were observed on glucose and only) on chromosome 11 (CNAG_01858). Reannotation revealed= galactose when supplemented with caffeine, calcoﬂuor white, or that the JEC21, WM276, and R265 annotations all excluded a SDS; however, the tgr1⌬ mutant grew better than did the wild type section containing a putative homeodomain motif. The gene is in the presence of Congo red irrespective of carbon source or 2,377 bp long, contains two introns, and yields a protein of 754 temperature and displayed poorer growth on diamide (Fig. 3). residues. Reciprocal best-hit BLAST analysis using the well- Congo red is thought to interfere with ␤-1,3-glucan assembly in characterized fungal homeodomain proteins Cup9, Pho2, Tos8,

March/April 2012 Volume 3 Issue 2 e00310-11 ¨ mbio.asm.org 5 Downloaded from

Morrow et al.

Yhp1, Yox1, MATalpha2, and HMRa1 from Saccharomyces cerevi- cyclic AMP (cAMP) pathway, although a variety of proteins and siae did not reveal any clear orthologs to the predicted protein. pathways are known to modulate activity (29). We therefore in- mbio.asm.org Further BLAST analysis using just the 60-amino-acid homeodo- vestigated potential targets of the Yhp2 transcription factor by main sequence revealed potential homologs in 25 of the 34 avail- examining changes in gene expression of components of the able basidiomycete genomes, although these display limited sim- cAMP and mitogen-activated protein kinase (MAPK)-protein ki- ilarity outside the homeodomain itself. We named the second nase C (PKC) pathway and the laccase genes LAC1 and LAC2. No on February 28, 2012 - Published by gene YHP2, for yeast homeodomain protein 2. significant changes were observed in any of the genes tested (re- Testing a Bt63 yhp2⌬ mutant on the same variety of nutrient sults not shown). This suggests that Yhp2 does not alter melaniza- sources, conditions, and stressors as those for the tgr1⌬ strain tion through changes in expression of known signaling pathway yielded no phenotypes under any conditions (Fig. 3 and results components that control melanization, nor through changes in not shown). No differences were observed between 30°C and levels of LAC1 or LAC2 transcript. 37°C. We also did not notice any epistatic effects on a tgr1⌬ yhp2⌬ Transcription of HXK1 and components of stress response double mutant, which was affected in the same fashion as was the pathways are affected in the tgr1⌬ mutant. Changes in growth on tgr1⌬ mutant on glucose at 37°C but did not show any novel glucose, the most easily assimilable carbon source, have been ob- phenotypes (Fig. 3). served previously in Cryptococcus mutants lacking peroxisome Melanin production is repressed by the YHP2 transcription components (30) or trehalose biosynthesis (31, 32). These factor. We next investigated effects of the single and double gene changes potentially involve cellular wall and membrane stresses deletions on classical Cryptococcus virulence components: mela- and a dysregulation of glycolytic flux. Initially, we determined if nin, capsule, extracellular protease, urease, and phospholipase B. TGR1 gene expression was regulated in response to increased tem- mbio.asm.org While no changes in production of extracellular proteins were perature using quantitative reverse transcription-PCR (qRT- observed for the tgr1⌬, yhp2⌬, and tgr1⌬ yhp2⌬ mutants (results PCR). Wild-type Bt63 grown on glucose and galactose at 30°C not shown), all three exhibit a slightly reduced capsule size, which showed no significant changes in expression, while strains grown reverts to wild-type levels in the tgr1⌬ TGR1 and yhp2⌬ YHP2 at 37°C showed an ~4-fold increase in TGR1 expression on glu- complemented strains (Fig. 3). While the tgr1⌬ mutant was indis- cose (Fig. 4A). tinguishable from wild-type strains in the melanin assays, both the To next investigate if Tgr1 modulates transcription of genes yhp2⌬ and tgr1⌬ yhp2⌬ mutants showed a subtle but reproducible involved in trehalose metabolism or cellular stress responses, we increase in melanin production at both 30°C and 37°C (Fig. 3). No analyzed transcript levels of both the wild type and mutant at the changes in mating proficiency, filamentation, or antifungal resis- growth-inhibiting temperature of 37°C on two carbon sources. In tance were observed (see Fig. S3 in the supplemental material; also addition to six trehalose metabolism genes (TPS1, TPS2, NTH1, results not shown). and NTH2; a putative trehalose synthase gene, CCG9; and a puta- Melanization and high-temperature growth on glucose are tive phosphotrehalase gene, PTH1), we examined hexokinase I multigenic traits. Owing to the reciprocal translocation in chro- gene HXK1 and hexokinase II gene HXK2, and components of the mosomes 3 and 11 of H99, at synapsis in meiosis I H99 chromo- calcineurin/calmodulin pathway, the PKC1-MAPK pathway and somes will form quadrivalents with the normal chromosomes 3 the cAMP pathway. Trehalose biosynthetic genes have previously and 11 of Bt63. Only parental H99-type or Bt63-type chromo- been shown in C. gattii to regulate components of the cell wall somes 3 and 11 should therefore be present together in any prog- integrity MAPK pathway (32), and yeast trehalose biosynthesis eny, as mixed parent chromosome 3/11 segregants will be unbal- mutants acquire glucose-sensitive (tps1⌬ and tps2⌬) and thermo- anced and contain large deletions and duplications that should be sensitive (tps2⌬) phenotypes. fatal. To test this, we analyzed chromosomes of 16 meiotic prog- Few significant differences in expression were found for genes eny of H99 ϫ Bt63 crosses using pulsed-field gel electrophoresis involved in trehalose metabolism, and those were primarily re- and probed Southern blots using probes to TGR1 and YHP2. Eight lated to induction or repression based on the carbon source pres- of the progeny displayed bands consistent with a Bt63 karyotype, ent (Fig. 4B). However, HXK1 expression was approximately dou- while the remaining eight displayed banding consistent with an ble wild-type levels in the tgr1⌬ mutant at 37°C on both glucose H99 genotype (see Fig. S4 in the supplemental material). No hy- and galactose and thus appears to be a general response of Cryp- brid combinations were observed, suggesting that the 50% of tococcus to loss of the Tgr1 protein. Hexokinase I was also 3-fold progeny bearing a nonparental set of chromosomes 3 and 11 are upregulated on galactose in both strains. Expression of stress re- inviable. Consistent with this, basidiospore germination frequen- sponse genes from a variety of pathways revealed six genes that cies for two separate H99 ϫ Bt63 crosses were below 50%, with the became unresponsive to galactose-based upregulation in the obtained values of 23.0% and 30.2% implying that further lethal tgr1⌬ strain (Fig. 4C). The protein kinase C gene PKC1, the cal- phenotypes may have been segregating among the progeny. When cineurin A gene CNA1, the MAPK gene HOG1, the Ras protein tested for melanization, capsule production, and growth at 37°C gene RAS1, the adenylyl cyclase gene CAC1, and the protein kinase in the presence of glucose, the phenotypes did not cosegregate R gene PKR1 were all upregulated 1.5- to 3-fold on galactose com- with their parental karyotype in the 16 progeny, and we instead pared to glucose in the wild-type strain; this expression change is observed a gradient of phenotypes for all three conditions abolished in the tgr1⌬ mutant. (Fig. S4). We hypothesize that these traits in Cryptococcus may be Deletion of HXK2 ablates the tgr1⌬ mutant phenotype. multigenic traits and that other contributing genes may obscure Hexokinase is the entry point for nongalactose hexose sugars into segregation of the deletion phenotypes. glycolysis and is tightly regulated. To determine if altered growth Transcription of genes involved in melanin production is on glucose in the tgr1⌬ mutant is effected through hexokinase, we unaffected in the yhp2⌬ mutant. Melanization in Cryptococcus is examined H99 hexokinase I and hexokinase II mutant strains, catalyzed by two laccases, Lac1 and Lac2, and is regulated by the H99 (tgr1⌬-309) hxk1⌬ and H99 (tgr1⌬-309) hxk2⌬, which, by

6 ¨ mbio.asm.org March/April 2012 Volume 3 Issue 2 e00310-11 Downloaded from

Genomic Rearrangements in C. neoformans var. grubii

growth on glucose but not galactose, while the R265 hxk2⌬ strain

displays normal growth on both sugars. Reintroduction of TGR1 mbio.asm.org into H99 increases the sensitivity of the strain to high temperature on glucose, confirming the phenotype observed in Bt63. Subse- quent reintroduction of TGR1 into the H99 hxk2⌬ mutant does not further increase the sensitivity of the strain, however. Deletion on February 28, 2012 - Published by of hexokinase II therefore suppresses the phenotype of the tgr1⌬ mutant in the H99 background. Metabolite profiling using 1D 1H NMR reveals increased trehalose levels in the tgr1⌬ mutant. The Bt63 tgr1⌬ mutant strain displays more robust growth on glucose at 37°C than does the wild type, suggesting a profound change in intracellular metabolism. To investigate how deletion of TGR1 affects the metabolome of Cryptococcus, we performed one-dimensional (1D) proton nuclear magnetic resonance (NMR) spectroscopy on metabolites extracted from strains Bt63 and Bt63 tgr1⌬ grown at 37°C on glucose-containing medium. Correlation of spectral peaks with existing metabolite databases and chemical shift standards revealed that the spectra were dominated primarily by signals from mbio.asm.org sugars, in particular, glucose and trehalose, as well as other metabolites such as succinate, acetate, and glycerol (Fig. 5). To identify sources of variation between the two strains, we performed a multivariate analysis employing principal component analysis (PCA). There was clear separation between the wild-type Bt63 and mutant Bt63 tgr1⌬ strains in the scores plot of principal components 1 and 2 (Fig. 5); separation along principal component 2 appeared to contain most of the biological variation between the two groups. In the corresponding bivariate loadings plot, which correlates specific metabolites with either the wild type or the mutant, variation in the mutant strain was significantly associated with higher levels of trehalose, fatty acids, and glycerophosphocholine and lower levels of glucose and glycerol than those of the wild type (Fig. 5). Deletion of TGR1 leading to increased trehalose levels is consistent with our phenotypic observations, as trehalose is accumulated in fungi exposed to a variety of stresses, particularly heat (32). Bt63 tgr1⌬ and yhp2⌬ mutants are unaffected for virulence in the nematode and murine models of cryptococcosis. Strain H99 is the type strain of C. neoformans var. grubii and was distributed to the community to ensure consistency and veracity of research into the organism’s biology and pathogenicity and is the strain used most widely around the world for virulence assays. Utilization of a strain altered for processes such as high- temperature growth and melanization as a standard for virulence FIG 4 Expression profiling of TGR1 and genes involved in high-temperature models could have ramifications across the Cryptococcus research growth and stress responses. Indicated strains were grown on YNB plus glu- field. To assess the impact on virulence of disruption of these cose or galactose plates for 2 days. (A) Gene expression analysis of TGR1 in Bt63 at 30°C or 37°C. (B) Expression analysis of trehalose biosynthesis genes TGR1 and YHP2 genes in strain H99, we first investigated the and the two hexokinases at 37°C. (C) Gene expression quantification of se- deletion strains in the nematode infection model. There were no lected genes implicated in high-temperature growth or cellular stress re- significant changes in survival times for Caenorhabditis elegans sponses at 37°C. Bars represent means Ϯ standard errors of gene expression cultured on the Bt63 tgr1⌬, yhp2⌬, or tgr1⌬ yhp2⌬ mutants from relative to that of ␤-actin (ACT1) from three biological replicates and three technical replicates. *, P Ͻ 0.05; **, P Ͻ 0.01; ***, P Ͻ 0.001; ****, P Ͻ 0.0001 the wild-type or complemented strains (Fig. 6). As Yhp2 influ- (Student’s two-tailed t test, two sample, equal variance). ences melanization, we repeated the assay in the presence of the melanizable substrate l-3,4-dihydroxyphenylalanine (L-DOPA); however, no change in virulence was apparent for the Bt63 yhp2⌬ virtue of being constructed in H99, also lack TGR1. At 37°C, H99 strain (results not shown). Neither Tgr1 nor Yhp2 appears critical hxk1⌬ displays normal growth on glucose but reduced growth on for virulence in the nematode model of cryptococcosis. galactose, fructose, and mannose, similar to the phenotype of the As the observed mutant phenotypes occur at 37°C, pathogenic- C. gattii R265 hxk1⌬ strain, which shows reduced growth on ga- ity of both mutants in a mammalian system was determined using lactose (see Fig. S5 in the supplemental material; also results not the murine inhalation model. Surprisingly, neither the single nor shown) (32). In contrast, the H99 hxk2⌬ mutant shows reduced double mutants were significantly altered for virulence, despite

March/April 2012 Volume 3 Issue 2 e00310-11 ¨ mbio.asm.org 7 Downloaded from

Morrow et al.

TGR1 and YHP2 affecting multiple cryptococcal virulence factors in vitro (Fig. 6). The mutants also displayed wild-type patterns of mbio.asm.org dissemination, with comparable burdens found in the lungs, brain, liver, spleen, and kidneys (see Fig. S6 in the supplemental material). If there are changes to virulence from these mutants, they are not detectable in this genetic background, although dif- on February 28, 2012 - Published by ferences may be apparent under certain conditions, such as within an immunocompromised host or by using a lower initial inoculum. Future molecular and pathogenesis studies utilizing H99 will carry an implicit caveat that the strain is atypical for virulence traits and carbon metabolism.

DISCUSSION Chromosomal rearrangements can contribute to reproductive isolation between closely related fungi and the generation of phenotypic diversity. Variations in karyotype are observed in both clinical and environmental isolates of pathogenic fungi and frequently arise during the infection process (33, 34). Genomic microevolution in Cryptococcus and Candida may potentially repre- mbio.asm.org sent an adaptive mechanism within the host environment (22), and changes in karyotype are associated with drug resistance, altered virulence components, and pathogenicity in animal models (35). Such changes are generally tolerated since pathogenic fungi can reproduce mitotically and sexual reproduction and meiosis may be infrequent (36). Elegant studies utilizing comparative genomics and parsimony analysis have detailed the structural evolution of the genome among the yeasts of the Saccharomycotina. The Saccharomyces sensu stricto complex yeasts Saccharomyces paradoxus, Saccharo- myces mikatae, and Saccharomyces bayanus are separated from S. cerevisiae by ~5 to 20 million years, similar to the C. neoformans var. grubii and C. neoformans var. neoformans estimated divergence time of ~24.5 million years (15), and are a useful comparison. Relative to the S. cerevisiae genome, S. paradoxus possesses four inversions and three duplications, S. mikatae possesses four translocations and 13 inversions, and S. bayanus possesses five translocations and three inversions (12). The 20 inversions (1 to 30 kb in size) were all flanked by tRNA genes, while one of the duplications (4 to 18 kb in size) and all the translocations were associated with the presence of transposable elements. Our analysis revealed that C. neoformans var. grubii possesses six inversions (~7 to 81 kb in size) and one duplication (~6 kb in size) that are unique. These were of a similar size as those observed in the Sac- charomyces sensu stricto complex yeasts, and all but one were associated with a repeated element at the breakpoints, suggesting that the rearrangements are associated with transposon movement and insertion in the genome. FIG 5 Metabolomics analysis of Bt63 and the tgr1⌬ mutant strain. (A) 1D proton NMR spectra of wild-type (top, blue) and tgr1⌬ mutant (bottom, Recently, D’Souza et al. (16) used synteny analysis to compare red) strains. Spectra of representative samples for each strain are shown. the genome of C. gattii VGI strain WM276 with both the C. gattii (B) Principal component analysis scores plot for the wild-type and tgr1⌬ VGII strain R265 and C. neoformans var. neoformans VNIV strain mutant strains. The distance between the points is an indicator of B3501A genomes. While the C. gattii genomes were broadly co- similarity between the samples. (C) Bivariate loadings line plot. The line plot displays on the loadings coefficient axis the correlation coefficients linear, there were widespread rearrangements between the two C. that relate individual integral regions of the NMR spectrum to the PC2 gattii genomes and the C. neoformans var. neoformans genome, axis of the scores plot. Individual peaks correspond to peaks in the 1D including multiple large inversions and translocations associated NMR spectra; here, peaks that are positive indicate metabolites that are with transposable elements. The presence of extensive chromo- increased in the wild-type Bt63 while negative peaks are increased in the somal rearrangements that differentiate C. gattii from C. neofor- tgr1⌬ mutant. The overlaid heat map relates the correlation of the peak to the PC2 axis of the scores plot when using Pareto scaling instead of center mans var. neoformans would be expected to significantly impair scaling. recombination and chromosome pairing at meiosis and is consistent with a contribution towards speciation (16). Our investigation revealed that only a small number of rearrangements

8 ¨ mbio.asm.org March/April 2012 Volume 3 Issue 2 e00310-11 Downloaded from

Genomic Rearrangements in C. neoformans var. grubii

melanin production, a key virulence com-

ponent of Cryptococcus. However, no mbio.asm.org changes in expression of melanization- related genes were observed in the Bt63 yhp2⌬ mutant, and since the perturbation of melanin production is very modest, on February 28, 2012 - Published by this may be a pleiotropic effect or Yhp2 may instead regulate proteins that potentiate melanin production through other mechanisms, such as uptake of diphenolic precursors or cell wall deposition. Despite the fact that certain subcultures of H99 mate poorly and that homeodomain proteins often regulate developmental processes or sexual reproduction (37), neither YHP2 nor TGR1 appears relevant to mating, as we did not observe any changes to hyphal growth or proliferation in unilateral and bilateral crosses using one or mbio.asm.org both disrupted genes. Future transcrip- tomic analysis of Bt63 and Bt63 yhp2⌬ may identify the regulatory targets of the Yhp2 protein. Trehalose is protective against heat stress and various other cellular stresses in fungi including cell wall stress. Trehalose biosynthesis mutants grow more poorly on glucose at elevated temperatures due to dysregulation of glycolysis and flux FIG 6 Virulence of translocation gene mutants in the nematode and murine inhalation models of cryptococcosis. (A) Survival times of N2 Bristol nematodes cultured on Bt63, tgr1⌬, tgr1⌬ TGR1, through hexokinase II (38); while os- yhp2⌬, yhp2⌬ YHP2, and tgr1⌬ yhp2⌬ strains. (B) Virulence of translocation mutant strains tested in motic stabilizers rescue the phenotype, vivo using a mouse model system. No nematodes or mice infected with mutant strains were significantly temperature sensitivity is not restricted to different from those infected with wild-type Bt63. glucose but is common to other hexose sugars and some disaccharides. Our tgr1⌬ phenotype, however, is restricted solely to uniquely distinguish C. neoformans var. grubii from C. neoformans glucose and is not observed on any other hexose or disaccharide, var. neoformans, and the profound differences in virulence char- implying that glucose may produce this effect extracellularly. acteristics and epidemiology may instead be conferred at the nu- Metabolomics analysis revealed that, primarily, trehalose levels cleotide level. Accordingly, the overall nucleotide sequence diver- increased and glucose levels decreased in the mutant strain, con- gence level between C. neoformans var. neoformans and C. sistent with our observed phenotype. The cAMP-protein kinase A neoformans var. grubii, between 10 and 15%, is substantial and (PKA) pathway in Saccharomyces responds to extracellular glu- suggests potential speciation (17). cose stimulus, and intriguingly, a low level of cAMP or reduced The widely distributed C. neoformans var. grubii type strain PKA pathway activity causes increased heat and stress resistance H99 forms the foundation for molecular and pathogenesis studies and accumulation of trehalose (39). As the phenotype is similar, in C. neoformans var. grubii. Unfortunately, it is not possible to Tgr1 in Cryptococcus may potentially modulate cAMP levels di- determine when the translocation occurred, though the absence rectly or indirectly in response to an extracellular glucose signal. of additional mutations in both disrupted genes suggests that it Disruption of TGR1 and YHP2 as a result of the H99 translocation was very recent. While it is tempting to suggest that a chromo- will require some caution to be adopted when interpreting genetic, somal rearrangement that enhanced growth at host body temper- biochemical, and virulence data for the C. neoformans var. grubii ge- ature and increased production of two key virulence components nome strain, particularly in regards to carbon metabolism and signal may have arisen during infection, in vivo virulence studies using transduction. As a corollary, the well-characterized S. cerevisiae strain nematodes and mice suggest that the gene disruptions did not S288c possesses a nonsense mutation in FLO8, encoding a transcrip- increase pathogenicity, though they may confer an advantage un- tion factor responsible for activation of FLO11 (40). Lack of Flo11 der specific host conditions or in a different strain background. If prevents flocculation and pseudohyphal growth in diploids and in- the translocation occurred during laboratory passage, it was prior vasive growth in haploids in S288c, which is not useful for such stud- to the distribution of the strain to the wider community as key ies but remains a good model for other processes. While virulence was subcultures all contain it. unaffected in C. neoformans var. grubii in two model systems, high- The only phenotypes that we could ascribe to the YHP2 gene lighting that many components beyond traditional virulence factors product were slight repression of melanization and capsule pro- contribute to successful infection of a host, undoubtedly there remain duction. We were excited at the prospect of a potential regulator of undiscovered epistatic interactions for both genes that potentially im-

March/April 2012 Volume 3 Issue 2 e00310-11 ¨ mbio.asm.org 9 Downloaded from

Morrow et al. pact pathogenesis. Crucially, each single-gene-deletion strain created of Ͻ0.05 being considered significant. Murine virulence assays were con- using H99, including the C. neoformans var. grubii gene deletion col- ducted in accordance with the guidelines in the Australian code of practice mbio.asm.org lection, is effectively a triple deletion strain. Strain H99 has a well- for the care and use of animals for scientific purposes from the National established virulence profile, however, and will remain an important Health and Medical Research Council and were approved by the Univer- model system for cryptococcal pathogenesis. Carbon metabolism has sity of Queensland Molecular Biosciences Ethics Committee (AEC# SCMB/008/11/UQ/NHMRC). not been extensively characterized in C. neoformans var. grubii to Quantitative real-time PCR. Gene expression analysis via qRT-PCR on February 28, 2012 - Published by date, and this perhaps should be conducted in a separate genetic was performed as described elsewhere (48). Strains were grown on yeast ni- background from H99. Signaling pathways, however, have been in- trogen base (YNB) plus 2% glucose and YNB plus 2% galactose plates at 37°C tensively investigated in Cryptococcus, and identification of a defini- or in liquid melanization medium (with or without L-DOPA) at 30°C and tive role for Tgr1 in signal transduction may require some reevalua- were harvested after 24 h. cDNA was synthesized using Superscript III (Life tion or reinterpretation of established interactions and cellular Technologies). Real-time PCR primers are found in Table S2 in the supple- responses. mental material. Relative gene expression was evaluated using the threshold cycle (2Ϫ⌬⌬CT) method normalized against ␤-actin. Statistical significance was determined using an unpaired Student’s t test in GraphPad Prism, where MATERIALS AND METHODS P Ͻ 0.05 was considered significant. Strains and media. Cryptococcus strains utilized in this study are listed in 1D 1H NMR spectroscopy. Bt63 and Bt63 tgr1⌬ strains were cultured Table S1 in the supplemental material. Strains representing VNI AFLP on YNB plus 2% glucose at 37°C, harvested after 48 h, and lyophilized. clades were selected from the work of Litvintseva et al. (41). For the trans- Metabolite extraction was performed using the yeast extraction protocol location PCRs, six H99 subcultures were assayed: from the Heitman lab, of Lewis et al. (49). For 1H NMR analysis, extracts were resuspended in mbio.asm.org H99 #1 (the closest subculture to the original Perfect lab isolate), H99 100 mM potassium phosphate buffer (pH 7.4), 10% D2O, 18 ␮M 2,2- Wimp, H99 Stud, and H99 #4413. A mixed H99 stock including H99 #1 dimethylsilapentane-5-sulfonic acid (DSS), and 18 ␮M 1,1-difluoro-1- was passaged through a rabbit and became Stud, and H99 #4413 was a trimethylsilanylphosphonic acid in 5-mm NMR tubes. NMR spectros- subculture of Stud. From the Lodge lab, we assayed H99 Eunuch, and copy was performed on a Bruker AV500 500-MHz spectrometer (Bruker from the Madhani lab, we assayed H99 CM018, which was derived from Biospin, Rheinstetten, Germany), equipped with a 5-mm self-shielded Eunuch and is the strain in which the Cryptococcus gene deletion collec- z-gradient triple resonance probe. Spectra were measured manually at tion is being constructed. Wimp and Eunuch were derived from H99 #1 298 K with 200 scans at 32k resolution with a spectral width of 14 ppm, and display attenuated virulence. H99 #1 melanizes poorly while Wimp, using the noesy1prd pulse program (Bruker pulse program library). Low- Eunuch, and CM018 do not melanize at 37°C. Eunuch and CM018 are power continuous wave irradiation was applied during the mixing time of sterile while Stud has increased mating proficiency. We also assayed the 100 ms and relaxation delay of 2.3 s to suppress the water signal. congenic mating pair KN99a and KN99␣, which were derived from H99 Metabolite profiling. 1D spectra were processed using Topspin 3.0 #4413; the VNI MATa isolate 125.91; and the VNII MAT␣ isolate 8-1 (42). (Bruker Biospin). Spectra were reduced to 0.001-ppm integral regions Cloning applications were performed in the Escherichia coli strain using AMIX 3.6.6 (Bruker Biospin); regions corresponding to the water MACH1 (Life Technologies, Carlsbad, CA). All chemicals were obtained signal were excluded (␦ ϭ 4.5 ppm to 5.05 ppm), and data were normal- from Sigma (St. Louis, MO), except nourseothricin (clonNAT), which ized to the total intensity of the spectrum. Principal component analysis was from Werner BioAgents (Jena, Germany). (PCA) with center scaling was performed in SIMCA-Pϩ 12.0 (Umetrics Bioinformatics. The unpublished C. neoformans var. grubii strain H99 AB, Sweden). The refined model comprised 15 samples and contained 2 genome is available at the Broad Institute (http://www.broadinstitute.org principal components, with an R2X value of 0.83 and a Q2 value (from /annotation/genome/cryptococcus_neoformans/MultiHome.html). Se- cross-validation) of 0.684. To gauge the contributions from weaker NMR quence manipulation and dot plot analysis were performed using signals more accurately, the 1D loading plot of principal component 2 of MacVector 11.0 (MacVector, Cary, NC). Synonymous and nonsynony- the center-scaled model was heat mapped with the corresponding load- mous substitution rates were determined using PAL2NAL (43). Se- ings coefficients from a Pareto-scaled PCA model (5 principal compo- quences for TGR1 and YHP2 from Bt63 are available from GenBank under nents; R2X ϭ 0.837, Q2 ϭ 0.47). Metabolites were identified using Che- nucleotide accession numbers JQ302192 and JQ302193, respectively. nomx NMR suite 7.1 (Chenomx, Edmonton, Canada) and the public Molecular techniques. Details of routine molecular techniques are given databases BioMagRes Data Bank (http://www.bmrb.wisc.edu) and Hu- in Text S1 in the supplemental material. Oligonucleotides utilized in this man Metabolome Data Bank (http://www.hmdb.ca). study are presented in Table S2. TGR1 and YHP2 were amplified from Bt63, cloned into pCR2.1-TOPO (Life Technologies), and subcloned into pJAF1 ACKNOWLEDGMENTS (44) and pPZP-NAT (45), respectively. The TGR1/YHP2 fusion gene was This work was supported by NHMRC project grant no. 455980 and amplified from H99, cloned into pCR2.1-TOPO, and subcloned into pJAF1. NHMRC CDA 569673 to J.A.F. C.A.M. was supported by an ANZ Trust- Phenotypic and virulence factor assays. Extensive details of pheno- ees PhD Scholarship in Medical Research. E.J.B. and J.H. were supported typic assays and virulence factor assays are given in Text S1 in the supple- by NIH/NIAID grant R37 AI39115 to J.H. K.N. was supported by NIH mental material. grant AI080275. Nematode and murine virulence assays. Virulence of the deletion strains was examined using the Cryptococcus nematode model as de- SUPPLEMENTAL MATERIAL scribed elsewhere (46) using Caenorhabditis elegans strain N2 Bristol. Mu- Supplemental material for this article may be found at http://mbio.asm.org rine infection was performed according to the protocol of Cox et al. (47) /lookup/suppl/doi:10.1128/mBio.00310-11/-/DCSupplemental. using 7-week-old BALB/c mice (Animal Resources Centre, Canning Vale, Text S1, DOCX file, 0.1 MB. Australia). Ten mice per strain were infected via nasal inhalation with 1 ϫ Figure S1, PDF file, 0.4 MB. 5 10 Cryptococcus cells. Mice were monitored for signs of distress or sick- Figure S2, PDF file, 0.2 MB. ness and sacrificed using CO2 once their body weight had decreased by Figure S3, PDF file, 1.1 MB. 20%. Organs were harvested from three mice per strain (the second, fifth, Figure S4, PDF file, 0.8 MB. and ninth mice to die), homogenized, and plated to determine fungal Figure S5, PDF file, 0.3 MB. CFU per gram organ weight. Kaplan-Meier survival analysis was per- Figure S6, PDF file, 0.2 MB. formed using GraphPad Prism 5.0 (GraphPad Software, La Jolla, CA). Table S1, DOCX file, 0.1 MB. Statistical significance was determined using a log-rank test, with a P value Table S2, DOCX file, 0.1 MB.

10 ¨ mbio.asm.org March/April 2012 Volume 3 Issue 2 e00310-11 Downloaded from

Genomic Rearrangements in C. neoformans var. grubii

REFERENCES 28. de Souza Pereira R, Geibel J. 1999. Direct observation of oxidative stress

1. Stranger BE, et al. 2007. Relative impact of nucleotide and copy number on the cell wall of Saccharomyces cerevisiae strains with atomic force mi- mbio.asm.org variation on gene expression phenotypes. Science 315:848–853. croscopy. Mol. Cell. Biochem. 201:17–24. 2. Rieseberg LH. 2001. Chromosomal rearrangements and speciation. 29. Alspaugh JA, et al. 2002. Adenylyl cyclase functions downstream of the Trends Ecol. Evol. 16:351–358. G␣ protein Gpa1 and controls mating and pathogenicity of Cryptococcus 3. Delneri D, et al. 2003. Engineering evolution to study speciation in yeasts. neoformans. Eukaryot. Cell 1:75–84. 30. Idnurm A, Giles SS, Perfect JR, Heitman J. 2007. Peroxisome function

Nature 422:68–72. on February 28, 2012 - Published by 4. Rieseberg LH, Willis JH. 2007. Plant speciation. Science 317:910–914. regulates growth on glucose in the basidiomycete fungus Cryptococcus 5. Navarro A, Barton NH. 2003. Accumulating postzygotic isolation genes neoformans. Eukaryot. Cell 6:60–72. in parapatry: a new twist on chromosomal speciation. Evolution 57: 31. Petzold EW, et al. 2006. Characterization and regulation of the trehalose 447–459. synthesis pathway and its importance in the pathogenicity of Cryptococcus 6. Wain LV, Armour JA, Tobin MD. 2009. Genomic copy number varia- neoformans. Infect. Immun. 74:5877–5887. tion, human health, and disease. Lancet 374:340–350. 32. Ngamskulrungroj P, et al. 2009. The trehalose synthesis pathway is an 7. Sturtevant AH, Dobzhansky T. 1936. Inversions in the third chromo- integral part of the virulence composite for Cryptococcus gattii. Infect. some of wild races of Drosophila pseudoobscura, and their use in the study Immun. 77:4584–4596. of the history of the species. Proc. Natl. Acad. Sci. U. S. A. 22:448–450. 33. Selmecki A, Gerami-Nejad M, Paulson C, Forche A, Berman J. 2008. An 8. Charlesworth B. 1991. The evolution of sex chromosomes. Science 251: isochromosome confers drug resistance in vivo by ampliﬁcation of two 1030–1033. genes, ERG11 and TAC1. Mol. Microbiol 68:624–641. 9. Fraser JA, et al. 2004. Convergent evolution of chromosomal sex- 34. Hu G, et al. 2008. Comparative hybridization reveals extensive genome determining regions in the animal and fungal kingdoms. PLoS Biol variation in the AIDS-associated pathogen Cryptococcus neoformans. Ge- 2:2243–2255. nome Biol 9:R41.

10. Harewood L, et al. 2010. The effect of translocation-induced nuclear 35. Fries BC, Casadevall A. 1998. Serial isolates of Cryptococcus neoformans mbio.asm.org reorganization on gene expression. Genome Res. 20:554–564. from patients with AIDS differ in virulence for mice. J. Infect. Dis. 178: 11. Taylor JS, Braasch I, Frickey T, Meyer A, Van de Peer Y. 2003. Genome 1761–1766. duplication, a trait shared by 22000 species of ray-finned fish. Genome 36. Morrow CA, Fraser JA. 2009. Sexual reproduction and dimorphism in Res. 13:382–390. the pathogenic basidiomycetes. FEMS Yeast Res. 9:161–177. 12. Kellis M, Birren BW, Lander ES. 2004. Proof and evolutionary analysis of 37. Hull CM, Boily MJ, Heitman J. 2005. Sex-specific homeodomain pro- ancient genome duplication in the yeast Saccharomyces cerevisiae. Nature teins Sxi1␣ and Sxi2a coordinately regulate sexual development in Cryp- 428:617–624. tococcus neoformans. Eukaryot. Cell 4:526–535. 13. Park BJ, et al. 2009. Estimation of the current global burden of crypto- 38. Hohmann S, Bell W, Neves MJ, Valckx D, Thevelein JM. 1996. Evidence coccal meningitis among persons living with HIV/AIDS. AIDS. 23: for trehalose-6-phosphate-dependent and -independent mechanisms in 525–530. the control of sugar influx into yeast glycolysis. Mol. Microbiol 20: 14. Heitman J, Kozel TR, Kwon-Chung KJ, Perfect JR, Casadevall A (ed). 981–991. 2010. Cryptococcus: from human pathogen to model yeast. ASM Press, 39. Versele M, de Winde JH, Thevelein JM. 1999. A novel regulator of G Washington, DC. protein signalling in yeast, RGS2, downregulates glucose-activation of the 15. Ngamskulrungroj P, et al. 2009. Genetic diversity of the Cryptococcus cAMP pathway through direct inhibition of Gpa2. EMBO J. 18: species complex suggests that Cryptococcus gattii deserves to have varieties. 5577–5591. PLoS One 4:e5862. 40. Liu H, Styles CA, Fink GR. 1996. Saccharomyces cerevisiae S288C has a 16. D’Souza CA, et al. 2011. Genome variation in Cryptococcus gattii, an mutation in FLO8, a gene required for filamentous growth. Genetics 144: emerging pathogen of immunocompetent hosts. mBio 2:e00342–10. 967–978. 17. Kavanaugh LA, Fraser JA, Dietrich FS. 2006. Recent evolution of the 41. Litvintseva AP, Thakur R, Vilgalys R, Mitchell TG. 2006. Multilocus human pathogen Cryptococcus neoformans by intervarietal transfer of a sequence typing reveals three genetic subpopulations of Cryptococcus neo- 14-gene fragment. Mol. Biol. Evol. 23:1879–1890. formans var. grubii (serotype A), including a unique population in Bo- 18. Sun S, Xu J. 2009. Chromosomal rearrangements between serotype A and tswana. Genetics 172:2223–2238. D strains in Cryptococcus neoformans. PLoS One 4:e5524. 42. Nielsen K, et al. 2003. Sexual cycle of Cryptococcus neoformans var. grubii 19. Fries BC, Chen F, Currie BP, Casadevall A. 1996. Karyotype instability ␣ in Cryptococcus neoformans infection. J. Clin. Microbiol 34:1531–1534. and virulence of congenic a and isolates. Infect. Immun. 71:4831–4841. 20. Spitzer ED, Spitzer SG, Freundlich LF, Casadevall A. 1993. Persistence 43. Suyama M, Torrents D, Bork P. 2006. PAL2NAL: robust conversion of of initial infection in recurrent Cryptococcus neoformans meningitis. Lan- protein sequence alignments into the corresponding codon alignments. cet 341:595–596. Nucleic Acids Res. 34:W609–W612. 21. Selmecki A, Forche A, Berman J. 2006. Aneuploidy and isochromosome 44. Fraser JA, Subaran RL, Nichols CB, Heitman J. 2003. Recapitulation of formation in drug-resistant Candida albicans. Science 313:367–370. the sexual cycle of the primary fungal pathogen Cryptococcus neoformans 22. Forche A, Magee PT, Selmecki A, Berman J, May G. 2009. Evolution in var. gattii: implications for an outbreak on Vancouver Island, Canada. Candida albicans populations during a single passage through a mouse Eukaryot. Cell 2:1036–1045. host. Genetics 182:799–811. 45. Idnurm A, Reedy JL, Nussbaum JC, Heitman J. 2004. Cryptococcus 23. Perfect JR, Casadevall A. 2010. The history of Cryptococcus and crypto- neoformans virulence gene discovery through insertional mutagenesis. Eu- coccosis, p. 17–26. In Heitman J, Kozel TR, Kwon-Chung KJ, Perfect JR, karyot. Cell 3:420–429. Casadevall A (ed), Cryptococcus: from human pathogen to model yeast. 46. Mylonakis E, Ausubel FM, Perfect JR, Heitman J, Calderwood SB. 2002. ASM Press, Washington, DC. Killing of Caenorhabditis elegans by Cryptococcus neoformans as a model of 24. Fraser JA, et al. 2005. Chromosomal translocation and segmental dupli- yeast pathogenesis. Proc. Natl. Acad. Sci. U. S. A. 99:15675–15680. cation in Cryptococcus neoformans. Eukaryot. Cell 4:401–406. 47. Cox GM, Mukherjee J, Cole GT, Casadevall A, Perfect JR. 2000. Urease 25. Kirkpatrick M, Barton N. 2006. Chromosome inversions, local adapta- as a virulence factor in experimental cryptococcosis. Infect. Immun. 68: tion and speciation. Genetics 173:419–434. 443–448. 26. Simsek D, Jasin M. 2010. Alternative end-joining is suppressed by the 48. Lee IR, Chow EW, Morrow CA, Djordjevic JT, Fraser JA. 2011. Nitro- canonical NHEJ component Xrcc4-ligase IV during chromosomal trans- gen metabolite repression of metabolism and virulence in the human fun- location formation. Nat. Struct. Mol. Biol 17:410–416. gal pathogen Cryptococcus neoformans. Genetics 188:309–323. 27. Kopecká M, Gabriel M. 1992. The influence of Congo red on the cell wall 49. Lewis IA, et al. 2007. Method for determining molar concentrations of and (1–3)-␤-D-glucan microfibril biogenesis in Saccharomyces cerevisiae. metabolites in complex solutions from two-dimensional 1H-13C NMR Arch. Microbiol 158:115–126. spectra. Anal. Chem. 79:9385–9390.

March/April 2012 Volume 3 Issue 2 e00310-11 ¨ mbio.asm.org 11

Appendix B

De novo GTP Biosynthesis Is Critical for Virulence of the Fungal Pathogen Cryptococcus neoformans

167 De novo GTP Biosynthesis Is Critical for Virulence of the Fungal Pathogen Cryptococcus neoformans

Carl A. Morrow1,2, Eugene Valkov2,3,4, Anna Stamp2, Eve W. L. Chow1,2, I. Russel Lee1,2, Ania Wronski2, Simon J. Williams1,2, Justine M. Hill2,5, Julianne T. Djordjevic6, Ulrike Kappler1,2, Bostjan Kobe1,2,4, James A. Fraser1,2* 1 Australian Infectious Diseases Research Centre, University of Queensland, Brisbane, Queensland, Australia, 2 School of Chemistry and Molecular Biosciences, University of Queensland, Brisbane, Queensland, Australia, 3 MRC Laboratory of Molecular Biology, Cambridge, United Kingdom, 4 Division of Chemistry and Structural Biology, Institute for Molecular Bioscience, University of Queensland, Brisbane, Queensland, Australia, 5 Centre for Advanced Imaging, University of Queensland, Brisbane, Queensland, Australia, 6 Centre for Infectious Diseases and Microbiology, Westmead Millennium Institute, University of Sydney at Westmead Hospital, Sydney, New South Wales, Australia

Abstract We have investigated the potential of the GTP synthesis pathways as chemotherapeutic targets in the human pathogen Cryptococcus neoformans, a common cause of fatal fungal meningoencephalitis. We find that de novo GTP biosynthesis, but not the alternate salvage pathway, is critical to cryptococcal dissemination and survival in vivo. Loss of inosine monophosphate dehydrogenase (IMPDH) in the de novo pathway results in slow growth and virulence factor defects, while loss of the cognate phosphoribosyltransferase in the salvage pathway yielded no phenotypes. Further, the Cryptococcus species complex displays variable sensitivity to the IMPDH inhibitor mycophenolic acid, and we uncover a rare drug- resistant subtype of C. gattii that suggests an adaptive response to microbial IMPDH inhibitors in its environmental niche. We report the structural and functional characterization of IMPDH from Cryptococcus, revealing insights into the basis for drug resistance and suggesting strategies for the development of fungal-specific inhibitors. The crystal structure reveals the position of the IMPDH moveable flap and catalytic arginine in the open conformation for the first time, plus unique, exploitable differences in the highly conserved active site. Treatment with mycophenolic acid led to significantly increased survival times in a nematode model, validating de novo GTP biosynthesis as an antifungal target in Cryptococcus.

Citation: Morrow CA, Valkov E, Stamp A, Chow EWL, Lee IR, et al. (2012) De novo GTP Biosynthesis Is Critical for Virulence of the Fungal Pathogen Cryptococcus neoformans. PLoS Pathog 8(10): e1002957. doi:10.1371/journal.ppat.1002957 Editor: Leah E. Cowen, University of Toronto, Canada Received June 3, 2012; Accepted August 26, 2012; Published October 11, 2012 Copyright: ß 2012 Morrow et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Funding: This work was supported by NHMRC Project Grant 455980 and NHMRC CDA 569673 to JAF. CAM was supported by an ANZ Trustees PhD Scholarship in Medical Research. BK is a National Health and Medical Research Council Research Fellow. JMH was supported by NHMRC RD Wright Fellowship 401748. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing Interests: The authors have declared that no competing interests exist. * E-mail: [email protected]

Introduction cases annually, causing ,624,000 deaths per year, mostly in areas with high HIV rates, such as sub-Saharan Africa. These are Fungal infections of humans are highly refractive to pharma- alarming numbers for a pathogen whose treatment regimen has cological intervention due to the similarities in eukaryotic cell not altered significantly in over a decade [8–10]. Treatment of physiology. The limited array of fungal cell-specific features has systemic fungal infections predominantly relies on a small group of therefore been the focus of antifungal drug research for many antifungals comprising azoles, polyenes, echinocandins and the years, with the fungal cell wall and cell membrane being primary antimetabolite flucytosine. Numerous problems exist with these targets. Recent studies exploring potential drug targets in fungal treatments however, including their notoriously variable efficacy genomes have found a surprisingly small number of essential across the limited spectrum of human fungal pathogens, high cost targets with little identity to a human homologue [1–4]. An and toxicity, a frequent requirement for hospitalization, and alternate approach to targeting fungal-specific components is emerging drug resistance [10,11]. The design of new classes of therefore to instead target shared proteins that are well charac- effective, readily available and affordable antifungals is therefore a terized in both the host and pathogen, and exploit more subtle matter of urgency. differences between the two. This approach is exemplified by the Rational drug design was pioneered in the purine metabolic novel antifungal sordarin and its derivatives [5,6]. pathway, a conserved series of processes responsible for providing One of the leading life-threatening fungal infections worldwide the cell with a ready supply of ATP and GTP as both an energy is cryptococcal meningitis caused by Cryptococcus neoformans,a source and for critical cellular processes including replication, pathogen that infects primarily immunocompromised individuals, transcription, translation and signal transduction. This pathway and its sister species Cryptococcus gattii, which generally infects has continued to serve as a fertile source of therapeutic agent immunocompetent individuals [7]. The estimated annual global development for over fifty years [12], and growing evidence incidence of cryptococcal meningitis is estimated to be 1.1 million supports it as a potential source of effective antifungal targets.

PLOS Pathogens | www.plospathogens.org 1 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus

Author Summary homologs of most genes of the canonical purine pathway (Figure 1A), with the exception of adenosine deaminase, adenine deaminase, and The fungal pathogen Cryptococcus neoformans is respon- GMP reductase. As previously reported [25], xanthine dehydrogenase sible for up to a million deaths annually, and the currently is also absent but a potential equivalent, an a-ketoglutarate-dependent available antifungal medicines used to combat this dioxygenase, is present. Each gene identified is present as a single copy, infection are limited and patently inadequate. We have including those encoding two key components of the GTP biosynthetic investigated the GTP biosynthesis pathway as a new drug pathway: IMPDH required for de novo GTP synthesis (IMD1)anda target, a pathway that has been a chemotherapy target in phosphoribosyltransferase (PRTase) required for the GTP salvage humans for many years. We found that inhibiting or pathway (HPT1). deleting one specific enzyme in this pathway, IMP dehydrogenase, was extremely effective at weakening Cryptococcus as a pathogen or killing it outright. Intrigu- IMPDH is essential for de novo GTP biosynthesis ingly, we isolated an extremely rare Cryptococcus isolate IMPDH performs the rate-limiting, first step in de novo GTP + that was naturally drug resistant, suggesting that microbial biosynthesis, the NAD -dependent conversion of inosine mono- competitors of Cryptococcus may also use this strategy to phosphate (IMP) to xanthosine monophosphate (XMP) via a two- out-compete it in the wild. By comparison of the subtype step oxidation and hydrolysis reaction. The reaction mechanism is responsible for the majority of infections worldwide with complex and involves a large conformational change mid-reaction, this rare drug resistant form, we have characterized the which a number of inhibitors exploit [26–30]. To determine if the enzyme’s structure and mechanism, revealing a number of C. neoformans IMD1 gene encodes a bona fide IMPDH, we deleted it strategies to develop more potent and specific anti-IMP in the well-characterized C. neoformans var. grubii strain H99. The dehydrogenase drugs for Cryptococcus. Treating nema- imd1D strain was a guanine auxotroph and could not grow on todes grown on Cryptococcus with the inhibitors led to minimal medium (Figure 1B). Supplementation of the medium longer survival times for the worms, validating our with exogenous guanine restored growth, which we later showed strategy. was via a salvage pathway. Introduction of the E. coli IMPDH guaB into the deletion mutant fully restored growth on minimal medium, confirming that the IMD1-encoded enzyme performs Disruption of de novo ATP or GTP biosynthesis genes in Candida the conversion of IMP to XMP (Figure 1B). Finally, the imd1D albicans and Aspergillus fumigatus leads to complete avirulence in phenotype could be mimicked by the addition of the IMPDH mammalian models [13–15]. In Cryptococcus, mutations that inhibitor MPA to the growth medium; both the wild-type and globally affect ATP and GTP biosynthesis lead to attenuated or IMD1 complemented strains were equally sensitive to MPA at low complete loss of virulence in vivo, as well as general growth defects concentrations (5 mg/mL). This phenotype is abolished when the and impaired virulence factor expression [16,17]. media is supplemented with guanine, as the salvage pathway is Two key enzymes supplying guanine nucleotides to a cell are able to bypass the block at IMPDH. inosine monophosphate (IMP) dehydrogenase (IMPDH), the rate- limiting catalyst and first committed step of de novo GTP imd1D rescue via purine salvage is mediated by biosynthesis, and hypoxanthine-xanthine-guanine phosphoribosyltransferase (HXGPRT), responsible for recycling purine nucleo- phosphoribosyltransferase Hpt1 bases into nucleoside monophosphates in the GTP and ATP Purines present in both the bird guano ecological niche of salvage pathways. As a key metabolic enzyme, IMPDH is highly Cryptococcus and within the human host are potential substrates of a expressed in proliferating cells and has become a major target of GTP salvage pathway. Purine salvage is mediated by PRTase immunosuppressive and antiviral chemotherapy, and has attracted enzymes, which catalyze the transfer of a ribose 5-phosphate- great interest as an anticancer, antiprotozoal, antibacterial and derived phosphoribosylpyrophosphate group to guanine or other antifungal target [18–21]. Four IMPDH inhibitors are currently purine nucleobases [31]. Upon deletion of the identified HPT1 gene approved for treatments: the immunosuppressants mycophenolic in strain H99, the mutant exhibited wild-type growth on rich and acid (MPA) and mizoribine, the anticancer agent tiazofurin, and minimal media (Figure 1B). MPA-mediated growth inhibition was the antiviral ribavirin. There are significant structural and abolished by the presence of guanine in all strains but the hpt1D functional differences between microbial and human IMPDHs, mutant, demonstrating that Hpt1 mediates salvage of guanine suggesting that species-specific inhibitors of key metabolic (Figure 1B). Heterologous expression of the E. coli PRTase gpt in the pathways hold considerable potential as novel therapeutics hpt1D mutant restored wild-type growth on guanine, confirming the [19,21–23]. In this study we have investigated the potential of role of the C. neoformans gene as a PRTase. Further testing revealed the GTP biosynthesis pathway and the enzymes IMPDH and that Hpt1 could accept hypoxanthine, xanthine and guanine as HXGPRT as candidate antifungal targets using genetic, structural substrates, defining the enzyme as a hypoxanthine-xanthine- and functional approaches to validate purine metabolism as a guanine phosphoribosyltransferase (HXGPRT) (Figure S1). viable chemotherapeutic target in C. neoformans. Loss of IMPDH delays synthesis of the cryptococcal Results polysaccharide capsule and melanin As GTP biosynthesis is crucial to many cellular processes, we Purine metabolism in Cryptococcus lacks several sought to determine if guanine auxotrophy affected pathogenicity canonical pathway elements via effects on virulence factor production. Despite addition of Unlike the purine-rich pigeon guano natural environment of C. guanine, the wild-type phenotype was not completely restored in neoformans,thehumancentralnervoussystem inhabited during systemic the imd1D mutant, which exhibited slow growth at 30uC and was infection is purine-poor [24], suggesting that during infection de novo unable to grow on rich media. Accordingly, in a growth curve purine synthesis could be important for cell survival. A bioinformatic assay the imd1D mutant had an extended lag phase of almost survey of the available C. neoformans and C. gattii genomes to identify 48 hours compared to wild-type and reached a much lower final components of the purine biosynthetic pathway identified Cryptococcus cell density (Figure 2A). In capsule-inducing conditions the imd1D

PLOS Pathogens | www.plospathogens.org 2 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus

Figure 1. Components of the purine metabolic pathway in Cryptococcus. (A) BLASTp analysis using S. cerevisiae orthologs reveals that the majority of the components of the canonical purine biosynthetic pathway are present in the genome of C. neoformans var. grubii. Enzymes or activities missing are greyed. (B) 10-fold serial dilutions of indicated strains were spotted onto YNB medium supplemented with specified purines (1 mM) and/or MPA (5 mg/mL) and incubated for two days at 30uC. The imd1D deletion strain is an auxotroph; MPA mimics the effect of an imd1D deletion, while E. coli guaB is highly resistant to MPA. A phenotype for the hpt1D strain is only observed in the presence of MPA. doi:10.1371/journal.ppat.1002957.g001 strain showed a significantly smaller capsule after 24 hours at both taining purine homeostasis. Although the hpt1D salvage deficient 30uC and 37uC (Figure 2B), approximately one-third and one- mutant showed no phenotype on rich or minimal media, we quarter smaller, respectively. Capsule size after extended growth hypothesized that as the infection process might pose a greater ($96 hours) was similar to wild-type (data not shown). On solid metabolic strain, this enzymatic function may play an important melanization media the imd1D mutant demonstrated reduced role. We cultured Caenorhabditis elegans, a natural predator of melanin production at both 30uC and human body temperature Cryptococcus, on the wild-type, hpt1D and hpt1D HPT1 strains. No of 37uC (Figure 2C), although this defect was not apparent in liquid significant difference was observed between survival times of the L-DOPA medium (Figure S2). Elaboration of hyphae was notably worms on the three strains on either rich or minimal media attenuated in matings involving an imd1D mutant created in an (Figure 3A). The addition of guanine to both the minimal and rich isogenic MATa mating-type background, while the imd1D mutant in media again produced no significant difference between the three the MATa mating-type background was unaffected (Figure 2D). In strains (data not shown), indicating that Hpt1-dependent salvage contrast to the pleiotropism of the imd1D mutation, the hpt1D of purines is dispensable in the nematode model. We also tested mutant showed no apparent defect in growth at human body the impact of loss of HPT1 on virulence in the murine inhalation temperature, capsule production, or melanization at either 30uC or model of cryptococcosis. Again, there was no significant difference 37uC, and loss of Hpt1 also appeared to have no effect on hyphal between the three strains (Figure 3B). Despite the extra stresses of filamentation during mating (Figure 2B, C, D and data not shown). increased temperature and the murine immune system, de novo purine biosynthesis alone proved sufficient for full virulence of The GTP salvage pathway is dispensable for C. Cryptococcus in a mammalian host. neoformans virulence in two animal models GTP salvage pathways are known to play an important role in recycling metabolically expensive guanine nucleobases and main-

PLOS Pathogens | www.plospathogens.org 3 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus

Figure 2. Contribution of purine biosynthesis genes to the virulence component of Cryptococcus. (A) Strains were grown for five days in RPMI 1640 media supplemented with 10% serum and monitored spectrophotometrically at 600 nm. (B) C. neoformans capsule biosynthesis was examined by growth in RPMI 1640 media plus 10% serum for 16 hours at 30uC and visualized using India ink; the halo of clearance denotes the capsule. Scale bar is 10 mm. Capsular diameter was measured relative to cellular diameter using CellProfiler image analysis software. Bars denote the average relative diameter from three biological replicates. Standard error bars are shown; p,0.05 *; p,0.01 **; p,0.001 ***; p,0.0001 ****. (C) Melanin production was determined on Cryptococcus L-DOPA melanization plates incubated for two days. (D) Mating proficiency was assayed by co- culturing strains on Murashige-Skoog or V8 medium in the dark for two weeks at 25uC. Magnification is 406. All assays involving the imd1D strain were supplemented with 1 mM guanine. doi:10.1371/journal.ppat.1002957.g002

De novo GTP biosynthesis is critical to cryptococcal guanine however, the imd1D strain was severely compromised for virulence in vivo virulence (p,0.0001) in the worm. Metabolic enzymes in Cryptococcus have been shown previously We next assessed the requirement for the GTP de novo pathway to play ‘moonlighting’ roles in virulence, most famously urease using the murine inhalation model of cryptococcosis. As in the [32,33], and some are found in extracellular vesicles of virulence nematode minimal media model, the imd1D mutant was unable to components secreted by Cryptococcus, including IMPDH [34,35]. In kill mice, even after 50 days (p,0.0001; Figure 3D). Certain the C. elegans virulence model on standard worm assay rich media functions of IMPDH are therefore critical for successful infection or rich media supplemented with exogenous guanine, the imd1D of a mammalian host by Cryptococcus. To more precisely examine strain did not display any significant difference in killing from the the effect of loss of IMPDH on Cryptococcus, we determined the wild-type strain H99 or the complemented strain imd1D IMD1 fungal burden for the imd1D strain for various organs at multiple (Figure 3C and data not shown). On minimal media lacking timepoints during murine infection. Compared to the proliferation and dissemination that occurs during a wild-type infection, imd1D cell counts dropped rapidly after initial inoculation and the strain

PLOS Pathogens | www.plospathogens.org 4 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus

Figure 3. Virulence of imd1D and hpt1D in invertebrate and mammalian host systems. (A) Survival of Bristol N2 strain nematodes co- cultured with indicated Cryptococcus strains cultured on rich Brain Heart Infusion medium or minimal YNB medium at 25uC. (B) Survival of female BALB/c mice infected with 56105 cells via nasal inhalation. (C) Virulence of indicated Cryptococcus strains co-cultured with nematodes on rich and minimal medium at 25uC. (D) Virulence of the imd1D mutant in BALB/c mice infected with 56105 cells via nasal inhalation. Virulence was determined using Kaplan-Meier survival analysis with statistical significance determined using a log-rank test. Survival times were not significantly different for the hpt1D strain in either model. imd1D was significantly attenuated in the nematode model on minimal medium (p,0.0001) and in the murine model (p,0.0001). doi:10.1371/journal.ppat.1002957.g003 was not found in the lungs after just seven days (Figure 4A). Cells demonstrated with fluconazole, which is toxic to the worm at were found in the brain on day one (CFU = 328/g), although after concentrations of ,100 mg/mL but is protective against C. albicans three days these were no longer present. No dissemination to the infection at up to 32 mg/mL [39]. Control experiments did not spleen, liver or kidneys was observed (data not shown). It is reveal significant changes in mortality in worms cultured on unusual for cells to be recovered from the brain at 3 days after 20 mg/mL MPA plus a food source versus worms on a food source intranasal infection, which typically take at least 7 days to alone (Figure S3). disseminate to the CNS from the lungs [36]. By comparison, at 14 days post infection when we observed total clearance, the wild- With the exception of a rare subtype, Cryptococcus is type strain is expected to have a log CFU/g of ,8 in the lungs and sensitive to the IMPDH inhibitor MPA ,5 in the brain [37]. Based on our virulence studies, Imd1 but not Hpt1 could serve as a potential antifungal target. To test this hypothesis, the IMPDH inhibitors protect C. elegans from infection in the sensitivity of the entire Cryptococcus species complex to two IMPDH nematode model of cryptococcosis inhibitors, MPA and mizoribine, was examined. The C. neofor- To validate IMPDH as an antifungal chemotherapeutic target mans/C. gattii species complex consists of eight haploid molecular in vivo, we utilized the C. elegans model of cryptococcosis to test the subtypes, VNI, VNII, VNIII and VNB of C. neoformans and VGI, effect of IMPDH inhibitors on survival times. Nematodes were VGII, VGIII and VGIV of C. gattii, which vary in their cultured on Cryptococcus in the presence of three concentrations of distribution, ecology and epidemiology [40,41]. While other MPA in the media (5, 10 and 20 mg/mL) under typical assay fungal pathogens such as C. albicans are sensitive to mizoribine conditions. As a control we simultaneously cultured nematodes in inhibition at very low concentrations, we did not observe any the presence of the fungistatic antifungal fluconazole (5 mg/mL). impact on Cryptococcus, even when high concentrations were used C. elegans cultured on Cryptococcus plus fluconazole displayed (100 mg/mL) in both MIC and serial dilution plate assays (data not significantly enhanced survival (p,0.01). The addition of 10 or shown) [21]. All subtypes were highly sensitive to low concentra- 20 mg/mL MPA also significantly enhanced survival times of tions of MPA (5 mg/mL) in defined minimal medium, with one worms (p,0.01; p,0.01) (Figure 4B). MPA is toxic to C. elegans exception: the rare C. gattii VGIV molecular type (Figure 5A). with a LD50 of ,90610 mM (28.8 mg/mL) [38], however, the Broth microdilution susceptibility assays supported this finding lower concentrations of MPA utilized appear tolerable and (Figure 5B). To confirm this observation, we tested a further 106 protective against C. neoformans infection; similar effects have been isolates and again found that the eight VGIV isolates in this

PLOS Pathogens | www.plospathogens.org 5 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus

Figure 4. IMPDH is essential for pathogenicity in murine and nematode hosts. (A) Fungal burden of imd1D cells recovered from the lungs and brain of BALB/c mice infected with 56105 cells via nasal inhalation. Bars represent mean colony-forming units (CFU)/g 6 standard error. (B) N2 Bristol young adult nematodes were co-cultured with C. neoformans wild-type strain H99 on minimal medium supplemented with 5, 10 or 20 mg/mL MPA or 5 mg/mL fluconazole. MPA significantly enhanced nematode survival over eight days, as did fluconazole treatment; p,0.05 *; p,0.01 **. doi:10.1371/journal.ppat.1002957.g004

Figure 5. The C. gattii VGIV molecular type is resistant to MPA. (A) Resistance to the IMPDH inhibitor MPA was assessed in all haploid molecular types of the pathogenic C. neoformans/C. gattii species complex using a serial dilution spotting assay on YNB plus 5 mg/mL MPA. (B) Precise sensitivity to MPA was determined via MIC broth microdilution assay. MIC50 and MIC90 are the concentrations at which growth was inhibited by 50% or 90%, respectively. Standard error bars are displayed for all strains. doi:10.1371/journal.ppat.1002957.g005

PLOS Pathogens | www.plospathogens.org 6 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus collection were easily identified by their MPA resistance, while all Spontaneous resistance to MPA does not arise readily in others were sensitive (data not shown). VGIV MPA resistance was Cryptococcus uniform and 100% reproducible, supporting this as an inherent To present an attractive target for antifungal chemotherapy, characteristic of the VGIV molecular type. spontaneous resistance to a drug should not arise readily through We reasoned that resistance to MPA could arise from mutations mechanisms such as single point mutations or rapid amplifications in the IMD1 ORF, overexpression of Imd1, increased IMD1 copy of target genes. In C. albicans, MPA-resistant strains were selected number, or changes in drug uptake or efflux from the cell. To by extended incubation on sub-lethal concentrations of the drug determine if MPA resistance could result from changes in resulting from a point mutation that changed the conformational expression level of IMPDH in C. gattii VGIV, we assessed the equilibrium of IMPDH [21]. We performed equivalent selection transcriptional response of C. neoformans (strain H99) and C. gattii using the VNI strain H99, but despite repeated attempts, we were (strain MMRL2651) IMD1 to challenge by guanine and MPA via unable to isolate colonies that were stably resistant to the drug qRT-PCR. After treatment with guanine, both C. neoformans and C. after transfer onto media containing the same or increasing gattii IMD1 were significantly downregulated, consistent with de concentrations of MPA (data not shown). novo GTP synthesis not being required when abundant salvageable guanine is present (Figure S4). Importantly, no real difference was Cryptococcus gattii VGIV IMPDH has multiple unique observed between the two strains’ responses to either substrate or polymorphisms inhibitor, suggesting that resistance to MPA may in fact be an To investigate the nature of the VGIV resistance to MPA, we intrinsic property of the VGIV IMPDH isoform. To determine if compared the IMD1 coding region between all known haploid either copy number or a feature of the C. gattii VGIV IMD1 allele molecular types. While there are 22 residue differences between (CgIMD1) confer MPA resistance, we introduced both the C. VNI (MPA-sensitive) and VGIV (MPA-resistant) IMPDH, there neoformans var. grubii VNI allele (CnIMD1) from the MPA-sensitive are only six residues unique to the VGIV subtype: three strain H99 and CgIMD1 from the MPA-resistant strain conservative and three nonconservative substitutions (Figure S5). MMRL2651 into wild-type H99 and the imd1D mutant. In the To establish the precise mechanism for MPA resistance, we wild-type H99 background, introduction of a second copy of introduced the six VGIV unique substitutions into the sensitive CnIMD1 resulted in only a very modest increase in MPA VNI enzyme. No single mutation was sufficient to confer VGIV- resistance, while introduction of the CgIMD1 allele was sufficient level resistance to MPA when introduced into the imd1D or the to enable robust growth and VGIV-level resistance to MPA wild-type background, however the simultaneous introduction of (Figure 6). In the imd1D deletion background, introduction of all six mutations into the CnIMD1 allele did confer MPA resistance CnIMD1 restores wild-type MPA sensitivity, while single-copy (Figure 6 and Figure S6). This finding confirms that a combination CgIMD1 successfully confers VGIV-level MPA resistance of these six residues is sufficient to confer MPA resistance, which (Figure 6). explains our inability to isolate spontaneous MPA-resistant isolates of VNI. We therefore generated VNI/VGIV IMPDH chimeras using convenient restriction sites that split the IMPDH gene into three parts to combine CnIMD1 and CgIMD1 into six possible

Figure 6. C. gattii IMPDH confers resistance to MPA, mediated by two amino acids. Serial dilution spotting assays of selected IMPDH mutants on YNB plus 5 mg/mL MPA. Robust growth on MPA is observed in the deletion mutant imd1D transformed with either the CgIMD1 allele or any IMD1 allele uniting the residues K336 and E446. Two copies of the MPA-sensitive CnIMD1 allele (in H99 CnIMD1) result in only a very modest increase in resistance to MPA and IMPDH heterotetramerization (in H99 CgIMD1) does not appear to produce novel phenotypes. The juxtaposed image depicts which portions of the IMPDH allele are present, with yellow representing CnIMD1 and blue representing CgIMD1, split into three thirds with the vertical bars depicting the six unique residues. Two proteins are depicted for the H99 background where the wild-type allele is also present. doi:10.1371/journal.ppat.1002957.g006

PLOS Pathogens | www.plospathogens.org 7 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus configurations. Only the construct comprising the first third of XMP*NNAD+ complex but meant that the data had to be fitted to CnIMD1 and the last two-thirds of CgIMD1 (bearing R446E, a modified version of the Michaelis-Menten equation that takes G450A and A500V) conferred VGIV-level resistance to MPA substrate inhibition into account. The Km values for both enzymes (Figure 6 and Figure S6). Based on this observation we employed were significantly different from the values for both human site-directed mutagenesis of the MPA sensitive VNI allele to isoforms of IMPDH (Table 1) while the turnover numbers were 21 21 introduce the conservative R336K substitution (the only unique similar (kcat = 1.560.1 s for CnImd1 and kcat = 2.260.2 s for residue in the middle fragment) in turn with each of the three CgImd1); IMPDH has a generally low turnover and the rate for residues from the last fragment; the combination of K336 and Cryptococcus is similar to those determined for IMPDHs from E446 was sufficient to confer MPA resistance (Figure 6 and Figure humans, other fungi and parasites [19,21,28]. S6). Again, transformation of these constructs into the wild-type H99 background yielded levels of MPA resistance comparable Inhibitor kinetics of CnIMPDH and CgIMPDH with the imd1D deletion strain, suggesting that simple gene MPA is an NAD+ analogue and an uncompetitive inhibitor that amplification may be insufficient to produce robust resistance, traps the E-XMP* enzyme-substrate covalent complex by and the resulting mixed IMPDH heterotetramers in these strains competing with the flap for the vacant NAD+ site [44]. To further did not appear to possess novel drug-resistant characteristics. investigate the mechanism behind increased MPA resistance, we examined the response of both enzymes in the presence of Steady-state kinetics of CnIMPDH and CgIMPDH inhibitor and varying amounts of both substrates. The concentra- The catalytic mechanism of IMPDH involves a transition from tion of MPA used was similar to the concentration of enzyme in a dehydrogenase to a hydrolase reaction; initially, the substrate the assays, which necessitated tight-binding inhibitor treatment, as IMP and cofactor NAD+ bind, IMP is then oxidized and forms a found in C. albicans and Aspergillus nidulans IMPDH [45]. MPA covalent transition state complex E-XMP* via a catalytic cysteine, dependent changes in kinetic parameters were studied for both while NAD+ is simultaneously reduced to NADH. The enzyme isoforms and were best described by an uncompetitive, tight then undergoes a conformational change whereby a large mobile binding inhibition mechanism versus NAD+, indicative of forma- flap folds into the vacant NAD+ site, carrying a catalytic arginine tion of E-XMP*NMPA complexes, as has been observed for residue that activates a water molecule to hydrolyze the E-XMP* multiple organisms (Figure S8). The Kii for the MPA-sensitive covalent complex, releasing XMP [26,42,43]. To investigate if CnIMPDH allele (240618 nM) is higher than the Kii observed for resistance to MPA is due to functional differences between the well-characterized human type I, human type II and C. albicans CnImd1 and CgImd1, we determined the steady-state kinetic IMPDHs, suggesting that the NAD+ site (Table 1), which is also parameters (Figure S7). Plots of velocity versus IMP concentrations the drug-binding site, is functionally different in the wild-type were best fit by the Michaelis-Menten equation Cryptococcus enzyme. Curiously, the inhibition constant for MPA is (Km(IMP) = 84612 mM, Km(NAD) = 634697 mM for CnImd1 and actually lower for the MPA-resistant CgImd1 enzyme Km(IMP) = 178615 mM, Km(NAD) = 532665 mM for CgImd1). How- (Kii = 9665 nM), although it displays a greater kcat/Km (NAD) ratio ever, both enzymes showed strong NAD+ substrate inhibition (4,135663 M21 s21) compared to CnImd1 (2,366639 M21 s21), above concentrations of 1,500 mM(Kii(NAD) = 4,4006900 mM for suggesting greater enzymatic efficiency. We note that fits for CnImd1 and Kii(NAD) = 4,0006600 mM for CgImd1), which is noncompetitive/mixed tight-binding inhibition for the MPA- common for IMPDH and is attributed to the formation of an E- resistant CgImd1 enzyme were extremely similar to those obtained

Table 1. Kinetic parameters for IMPDH from characterized species.

Kii(MPA) (nM) vs. 21 + IMPDH kcat (s ) Km(IMP) (mM) Km(NAD) (mM) Kii(NAD) (mM) NAD Kii(MPA) (nM) vs. IMP

E. coli 13.0 61 2000 2.8 $10000 ND C. parvum 3.3 29 150 2.9 9300 ND T. foetus 1.9 1.7 150 6.8 9 ND H. sapiens type I 1.8 14 42 ND 11 ND H. sapiens type II 0.4 4 6 0.6 6 ND A. nidulans 0.7 10 170 1.5 25 (UC) 22 (UC) P. chrysogenum A0.8 40 290 2.4 ND ND P. chrysogenum B 0.0075 600 640 NA ND ND P. brevicompactum A 0.7 130 340 4.7 500 (UC) 450 (NC/mixed) P. brevicompactum B 0.4 1400 790 NA 14000 (UC) 20000 (NC/mixed) C. albicans 6 60 3500 1.5 11 ND C. neoformans VNI 1.560.1 84612 634697 4.460.9 240618 (UC) 198663 (NC/mixed) C. gattii VGIV 2.260.2 178615 532665 4.060.6 9665 (UC) 107631 (NC/mixed)

Steady-state and inhibition constants for species with well-characterized IMPDHs. UC, uncompetitive; NC, noncompetitive; ND, no data; NA, not applicable. Values are from [19,21,28,58]. doi:10.1371/journal.ppat.1002957.t001

PLOS Pathogens | www.plospathogens.org 8 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus for uncompetitive tight-binding inhibition and yield a much higher hamster IMPDH:IMP:MPA structure [26]. MPA makes far fewer Kii (293676 nM; data not shown), which may indicate some interactions in the cryptococcal IMPDH:IMP:MPA complex than affinity of the inhibitor for other enzyme forms. Surprisingly, fits of observed in the hamster IMPDH structure, perhaps accounting for MPA versus IMP for both enzymes were best described by some of the differences in Ki between the mammalian and fungal noncompetitive/mixed patterns of inhibition (Figure S8), with a enzyme. In particular, T347 (Cryptococcus numbering) faces away Kii = 198663 nM for CnImd1 and a Kii = 107631 nM for from the active site in the Cryptococcus enzyme and does not bind CgImd1. This suggests that in Cryptococcus IMPDH, MPA binds MPA, while several other residues including N317, D288 and to additional enzyme forms such as free enzyme or E-IMP. Such a G340 participate in fewer interactions with the ligands in the scenario is also found in MPA-resistant Penicillium brevicompactum cryptococcal complex. The six unique residues found in CgImd1 IMPDH, where the initial hydride transfer step (as opposed to the map to the amino terminal region near a monomer-monomer subsequent hydrolysis step) is either partially or completely rate- interface (V55M), the accessory cystathionine-b-synthase (CBS) + limiting, resulting in the predominate enzyme form being E-IMP- domain (A153T), adjacent to the active site loop in the NAD and NAD+ instead of E-XMP*, conferring MPA resistance [46]. MPA-binding cavity (R336K), close to a catalytic arginine in the movable flap (R446E and G450A), and near the carboxy terminus (A500V). R/K336 forms hydrogen bonds with IMP and MPA in Crystal structure of CnImd1 the hamster enzyme, and R/E446 and G/A450 are near the To provide a foundation for fungus-specific inhibitor develop- catalytic arginine (R458) that resides on the mobile flap. Of note, ment, we determined the crystal structure of the clinically relevant one of the two residues responsible for MPA resistance in C. gattii C. neoformans var. grubii IMPDH in complex with the substrate IMP VGIV IMPDH, R336 (K336 in the resistant enzyme), forms direct and the inhibitor MPA. CnImd1 forms a tetramer and in the final contacts with only IMP in the cryptococcal complex, suggesting a model each monomer contains 395 of the 544 residues in the possible mechanistic explanation for the failure of the R336K typical eight-stranded a/b barrel structure, plus one molecule each single mutant to exhibit any altered sensitivity to MPA. of IMP and MPA bound in the active site, consistent with our Substitution of R336 with K336 in silico increases the distance functional observations suggesting an uncompetitive mechanism of from the amine nitrogen of R336K to the IMP ribose O3 hydroxyl enzyme inhibition (Figure 7A). The large accessory subdomain from 3.1 A˚ to 3.9 A˚ , which would lead to the loss of contacts with between residues 130–244 and residues 438–449 from the mobile IMP. flap have no interpretable electron density, common to previous Comparison of the hamster and cryptococcal IMPDH active structural studies on this class of enzymes. site loops with IMP and MPA-bound states indicates differences in Many of the active site residues are invariant, and the purine conformations, despite considerable similarity at sequence level; in ring, the ribose group and the nucleotide phosphate group of IMP the human type I IMPDH structure (bound to the IMP analog 6- make mostly conserved interactions to those found in the Chinese chloropurine ribotide) this loop adopts yet another distinct

Figure 7. Structure of C. neoformans IMPDH. (A) Active site residues of C. neoformans IMPDH with IMP and MPA ligands in ball-and-stick representation, with hydrogen-bonding interactions indicated by dashed lines. Figure displays residues within 4 A˚ of the active site except for T347, which adopts a strikingly different conformation in the Cryptococcus enzyme. (B) Residues 339–356 of Cryptococcus IMPDH (green) form the active site loop and are highly conserved. Comparison of the C. neoformans ENIMPNMPA complex active site loop conformation with the hamster E- XMP*NMPA complex and the human type I IMPDH in complex with the substrate analogue 6-chloropurine ribotide (CPR) are shown. IMP, CPR and the catalytic cysteine are shown in stick representation, while MPA is omitted. (C) Surface representation of the active site pocket of C. neoformans and Chinese hamster IMPDH. The white arrow indicates the region around T347, which forms a unique, potentially exploitable cavity in the fungal enzyme. MPA is omitted for clarity. doi:10.1371/journal.ppat.1002957.g007

PLOS Pathogens | www.plospathogens.org 9 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus conformation (Figure 7B). The change in orientation of the loop in have examined and characterized two of the key enzymes from the CnImd1 creates a large, Cryptococcus-specific pocket in the active GTP de novo biosynthetic and salvage pathways, revealing the site (Figure 7C). There is considerable flexibility in the conforma- potential of this pathway as a target of anticryptococcal therapy. tion of the active site loop observed in structures of different We discovered that only the rare C. gattii VGIV molecular type is enzyme complexes, which suggests that it can mimic the changes resistant to a natural product inhibitor of IMPDH, presenting an during the reaction cycle where it may modulate closure of the flap opportunity to investigate the evolution of drug resistance in [47]. While IMP is covalently bound to the catalytic cysteine in the IMPDH in the Cryptococcus species complex, and whether this poses hamster and human structures, which may be responsible for the a risk to potential therapies. altered orientation, comparison with Tritrichomonas foetus IMPDH GTP synthesis appears to be a natural antimicrobial target, as bound to IMP or mizoribine phosphate (where the ligands are not both MPA and mizoribine are produced as secondary metabolites covalently bound in either structure) suggests the conformation by various mould species. The ubiquity of MPA resistance among observed in the cryptococcal complex to be unique, thus offering a C. gattii VGIV isolates suggests that it may be an adaptation, potential avenue for species-specific inhibition. IMPDH crystal perhaps to competitors in its environmental niche. Very little is structures have been demonstrated to provide a reliable indication known on the ecology of VGIV isolates in particular, and while it of the position of the loop at different parts of the reaction cycle, has been isolated from clinical samples from a small number of suggesting the pocket observed in the Cryptococcus structure to be a countries in South America, sub-Saharan Africa and the Indian legitimate target [47]. subcontinent, reports of environmental isolates are limited to The cryptococcal IMPDH structure is unique in having most of certain trees (primarily the Indian almond tree Terminalia catappa) the mobile flap, except for 12 residues that encompass a small from Colombia and Puerto Rico [49]. The VGIV subtype is insertion region unique to Cryptococcus, sufficiently ordered to nevertheless extremely rare, while the clinically-relevant VNI type permit all but three residues corresponding to the flap in human is widespread and exquisitely sensitive to MPA. IMPDH to be observed. The only comparable structural evidence The finding that the purine salvage deficient hpt1D mutant was for the flap is presented for the parasite T. foetus IMPDH fully virulent in both the worm and mouse model was somewhat complexed with mizoribine phosphate, in which the flap is in the surprising, given the intense nutritional and metabolic stresses and closed conformation within the cofactor site [29] and the Bacillus hostile immune system pressures Cryptococcus is under during the anthracis apoenzyme structure [48]. The cryptococcal structure infection process. Similar PRTase mutants in the apicomplexan reveals the flap in the ‘open’ conformation, as would be required parasites Leishmani donovani and Toxoplasma gondii are viable and for the initial dehydrogenation before the flap folds into the NAD+ virulent however, although many display certain fitness defects or site to hydrolyze the covalent enzyme intermediate. The flap is partial attenuation of infectivity [50,51]. Given that de novo purine folded back onto itself, with residues A455–A465 facing the biosynthesis is an energy intensive process, salvage of purines from following V466–V476, with the catalytic R458 positioned more a host could serve as an excellent way to conserve energy during than 13 A˚ away from the C2 carbon of the IMP purine ring infection; as the hpt1D mutant is not perturbed for virulence, this (Figure 8A). Comparison with both the T. foetus and B. anthracis may reflect a generally low concentration of salvageable purines enzymes suggests that the flap undergoes significant conforma- during the infection process. This model is supported by the tional rearrangement, swinging completely around and under- avirulence of the imd1D mutant; generating a ready supply of GTP neath itself to bring the catalytic arginine into close juxtaposition via the de novo pathway is obviously of critical importance to the to the substrate (Figure 8B). Despite being uncomplexed with fungus during infection. substrate or inhibitor, the B. anthracis structure also clearly shows These findings are consistent with other studies demonstrating the enzyme in the closed conformation, with the flap folded into that IMPDH and de novo GTP biosynthesis mutants from a range the active site (Figure 8C). Superposition of the bacterial enzyme of pathogenic bacteria are attenuated for virulence [52], and in onto the fungal enzyme reveals that the mobile portions of the flap pathogenic E. coli, Salmonella enterica and B. anthracis, de novo purine comprise residues H436-A465 (30 residues in Cryptococcus but only biosynthesis has been proposed to be the single most critical 15 residues in human), while the remaining residues of the flap, metabolic function for growth in human serum and blood [53]. Y414-E435 and V466-K479, appear stationary. Comparison of Purine de novo biosynthesis may potentially be rate-limiting to the open conformation of the C. neoformans enzyme (MPA-sensitive) growth of Cryptococcus during infection, and therefore present an with the closed conformation of the T. foetus and B. anthracis outstanding target for anti-cryptococcal chemotherapy. Further- enzymes (both MPA-resistant) readily demonstrates how MPA can more, in contrast to the simple guanine auxotrophy of S. cerevisiae more easily access the cofactor site in enzymes where the and E. coli IMPDH mutants [54,55], the imd1D mutant of conformational equilibrium favors the open state [44]. Unfortu- Cryptococcus exhibits a range of pleiotropic phenotypic defects even nately the second residue responsible for MPA resistance in C. gattii when supplemented with exogenous guanine, including slow VGIV, E446 (R446 in the sensitive enzyme), was not observed in growth, attenuated virulence factor expression, and impaired the electron density, however this is a strong indicator of dissemination and avirulence in two animal models of infection. considerable flexibility within the immediate structural context of Intriguingly, recent studies have revealed that both IMP dehy- this residue, indicating it unlikely to be proximal to the well- drogenase and GMP synthase, the first and second enzymes of de ordered active site in the open conformation. However, as the flap novo GTP biosynthesis, act as transcriptional repressors in undergoes structural rearrangement, the residue is likely to be Drosophila [35,56], while the accessory CBS domain of IMPDH brought into closer proximity to the active site pocket during the has been demonstrated to regulate purine homeostasis in E. coli reaction cycle. [54,57]. It would be valuable to investigate the potential regulatory roles of IMPDH in Cryptococcus and establish any contributions to Discussion pathogenicity. That IMPDH is absolutely critical for virulence of Cryptococcus, New drug targets and novel approaches to treatment are and the observation that the most clinically prevalent forms of urgently needed as the immunocompromised population grows Cryptococcus are exquisitely sensitive to the IMPDH-inhibiting drug and the burden of opportunistic fungal infections increases. We MPA, is encouraging for the investigation of IMPDH and de novo

PLOS Pathogens | www.plospathogens.org 10 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus

Figure 8. The open and closed conformations of the IMPDH mobile flap. (A) The Cryptococcus neoformans IMPDH structure indicates that the flap adopts an open conformation in the ligand-bound state, with the catalytic R458 folded above the active site. The flap is rainbow colored and the catalytic arginine and IMP in the active site are depicted in stick representation. (B) The Tritrichomonas foetus IMPDH structure in complex with mizoribine phosphate (MZP) is in the closed conformation, with the flap extending out over the active site pocket and the catalytic arginine positioned underneath. (C) The Bacillus anthracis structure shows the unbound apoenzyme also adopts the closed conformation, with the flap closed over the active site. doi:10.1371/journal.ppat.1002957.g008

GTP biosynthesis as an antifungal drug target. Kohler et al. [21] the cofactor site or other changes in the enzyme mechanism or previously established that the C. albicans form of the enzyme conformation. The dynamics of the flap movement are known to displays an altered kinetic profile relative to human IMPDH, be key to resistance to cofactor mimic drugs in IMPDH, and key suggesting that fungus-specific inhibitors could be developed that mutations can alter the preferred state from open and MPA- target the MPA binding site. Functionally, the two Cryptococcus sensitive (the state observed in our structure, with the flap poised enzymes appear quite similar to other fungal IMPDHs, including above the active site and the cofactor site exposed), to closed, the C. albicans and A. nidulans enzymes, and GTP biosynthesis is which restricts access of MPA to the cofactor site and confers critical for virulence of both C. albicans and A. fumigatus (closely resistance [44]. Residues close to the catalytic arginine are related to A. nidulans). Cryptococcal IMPDH also shows significant involved in hydrogen bonding interactions when the flap folds differences from the human enzyme, despite ,56% sequence into the NAD+ site in T. foetus and presumably other species [29]. + identity. The Km for both IMP and the cofactor NAD are The combination of residues observed in C. gattii VGIV is substantially different from the human forms, suggesting that reminiscent of the well-characterized MPA-resistant T. foetus there are exploitable aspects of the reaction cycle of the enzyme, which requires K310 (equivalent to Cryptococcus K336) cryptococcal enzyme. It is less clear why VGIV CgImd1 should and E431 (equivalent to Cryptococcus Q470, also on the flap) in have a lower Kii for MPA than the VNI variant. Residues K336 tandem to provide robust resistance [30]. Our data suggests that and E446 together appear to play a pivotal role in C. gattii VGIV while MPA affinity increased, a compensatory increase in enzyme resistance, as demonstrated by their introduction into the catalytic efficiency may have occurred through the combination sensitive form, perhaps by altering the affinity of the flap for of K336 and E446. Recent work also demonstrates that if the

PLOS Pathogens | www.plospathogens.org 11 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus initial hydride transfer step becomes rate-limiting, this reduces Medical Research Council. Infection and euthanasia were accumulation of the enzyme-substrate covalent complex and thus performed under methoxyfluorane anesthesia, and all efforts were reduces susceptibility to MPA inhibition. Notably, this has made to minimize suffering via strict adherence to the Guidelines occurred in MPA-resistant Penicillium brevicompactum IMPDH to promote the wellbeing of animals used for scientific purposes by without significant residue changes around the substrate/cofactor the National Health and Medical Research Council. binding sites [46,58]. Our work demonstrates that MPA binds to additional cryptococcal IMPDH forms (an uncompetitive inhib- Strains and media itor should bind only to the enzyme-substrate covalent complex), C. neoformans and C. gattii strains were cultured in YPD or YNB suggesting the enzyme to be fundamentally functionally different (Becton Dickinson, Franklin Lakes NJ) media at 30uC. imd1D from the human isoforms. mutants are guanine auxotrophs and were grown on minimal To provide a solid foundation for chemotherapeutic agent YNB media supplemented with 1 mM guanine for all manipula- development, we have determined the structure of C. neoformans tions; ade2D and hpt1D ade2D mutants are adenine auxotrophs and var. grubii IMPDH in complex with IMP and MPA. While the were maintained on rich media. Cloning and plasmid propagation accessory domain is presumably disordered and was not observed, was performed in E. coli strain Mach1 (Life Technologies, we were able to successfully model the residues comprising the Carlsbad CA). Genomic DNA for E. coli gene isolation was mobile flap domain, which to date had only been structurally prepared from strain K12. C. elegans strain N2 Bristol was characterized in the closed conformation in the T. foetus and the B. maintained at 20uC on lawns of the E. coli uracil auxotroph strain anthracis enzymes [29,48]. While Cryptococcus IMPDH possesses an OP50 on Nematode Growth Medium using standard procedures insertion within the flap that was mostly disordered and there is [63]. Recombinant Cryptococcus IMPDH was heterologously considerable flexibility associated with the flap region, the expressed in E. coli strain BL21(DE3)pLysS (Merck, Darmstadt, structural evidence suggests that the flap undergoes a drastic Germany). MPA, mizoribine, 5-fluoro-29-deoxyuridine, G418, conformational change during the reaction cycle, which has been NAD+, Tris, Bis-Tris, glycine, NaCl, KCl, DTT, EDTA, DMSO, demonstrated functionally but not structurally in previous studies uric acid, inosine, hypoxanthine, xanthine, xanthosine, guanine, [27,59,60]. In the open enzyme conformation, the catalytic guanosine, adenine, adenosine, IMP, AMP, XMP and GMP were arginine is positioned away from the active site, with the flap purchased from Sigma (St Louis, MO). Nourseothricin (clonNAT) undergoing a 270u swinging rotation to perform the hydrolysis was purchased from Werner BioAgents (Jena, Germany). MPA reaction. The near-complete flap structure in the open orientation was dissolved in DMSO and was used at a concentration of 5 mg/ will aid future drug design efforts, as the flap folds into the cofactor mL. Supplemental purines when required were added to a final site – a critical target site for inhibitors. The cofactor site, and the concentration of 1 mM [63]. adenosine subsite in particular, also present an opportunity for fungus-specific inhibitor development, as there are a number of Molecular techniques critical residues that vary in the Cryptococcus enzyme; for example, Standard molecular techniques were performed as described by the adenine ring of NAD+ likely stacks between R267 and Y296 in Sambrook et al. [64]. C. neoformans genomic DNA for Southern blot Cryptococcus IMPDH, and these two residues differ from the analysis was prepared as described [65]. Oligonucleotides used are corresponding H253 and F282 in the human enzyme. The eukaryotic parasite T. foetus shares these two substitutions, which given in Table S1. For Southern hybridizations, DNA was have been suggested as potential species-specific targets [61]. The digested with appropriate enzymes, electrophoresed on TAE- adenosine portion of the cofactor site is also a monomer-monomer agarose gels and blotted onto Hybond-XL membrane (GE interface, and one of the more variable parts of the enzyme Healthcare, Little Chalfont, United Kingdom) using standard between species. Recent strategies extend inhibitors from the procedures. Probes were generated using the Rediprime II Random Prime Labeling Kit (GE Healthcare) with PCR products nicotinamide subsite (where MPA binds) out to the adenosine 32 subsite to increase specificity [44,62]. The unique pocket in the from H99 or MMRL2651 template DNA, and a P dCTP active site loop region of Cryptococcus IMPDH will enable the design (PerkinElmer, Waltham MA). Blots were hybridized overnight at of inhibitors with moieties that occupy the cavity, which should be 65uC and membranes were exposed onto Fuji Super RX medical selective for the fungal enzyme over the human enzymes. X-ray film (Fujifilm, Tokyo, Japan). Splice overlap PCR, knockout Overall, we have demonstrated that de novo GTP biosynthesis is transformations and gene complementations were performed as crucial for cryptococcal virulence factor expression and patho- described [66–69], using particle delivery using a BioRad He-1000 genesis, and multiple exploitable avenues exist in IMPDH to Biolistic Device (Bio-Rad, Hercules CA). develop inhibitors that are specific for the fungal enzyme. This work lays the foundation for purine biosynthesis as a legitimate Bioinformatic analysis antifungal target, and the functional similarity between the The unpublished C. neoformans var. grubii genome is available from Cryptococcus and Candida enzymes presents an intriguing opportu- the Broad Institute of MIT and Harvard (http://www. nity to develop novel IMPDH inhibitors that could lead to broadinstitute.org/annotation/genome/cryptococcus_neoformans/ effective therapy against a spectrum of fungal pathogens. MultiHome.html). To amplify genes from C. gattii strain MMRL2651, primers were designed from the closely related C. Materials and Methods gattii strains WM276 and R265, which are also available from the Broad Institute (http://www.broadinstitute.org/annotation/ Ethics statement genome/cryptococcus_neoformans_b/MultiHome.html). Purine The Cryptococcus murine virulence assay protocol was approved biosynthetic genes were identified using reciprocal best-hit BLAST by the Molecular Biosciences Animal Ethics Committee of the analysis and annotated using MacVector 9.5 (MacVector, Cary NC) University of Queensland (AEC# SCMB/473/09/UQ/ [70]. Sequencing was performed at the Australian Genome NHMRC). Assays were conducted in accordance with the Research Facility (Brisbane, Australia) and sequence traces were guidelines in the Australian code of practice for the care and use analyzed using Sequencher 4.7 (Gene Codes, Ann Arbor MI). of animals for scientific purposes by the National Health and Sequences for Cryptococcus IMPDH from strains 8-1 (VNII), Bt33

PLOS Pathogens | www.plospathogens.org 12 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus

(VNB), NIH312 (VGIII) and MMRL2651 (VGIV) are available Gene expression analysis from GenBank under the accession numbers FJ418778–FJ418781. Quantitative reverse transcription PCR analysis was performed as described [79]. Overnight cultures of strains H99 and Phenotypic and virulence factor assays MMRL2651 were grown for 16 hours to mid-log phase, before For growth curve analysis, cells were cultured in RPMI 1640 addition of MPA (5 mg/mL) or guanine (1 mM). Cells were grown media (Life Technologies) plus 2% glucose, 10% fetal bovine for 15 minutes, 60 minutes or 240 minutes before harvesting and serum (Life Technologies) and 1 mM guanine. Cells were freshly flash freezing in liquid nitrogen. Total RNA was extracted using cultured on YNB or YNB plus guanine plates, washed, diluted to TriZOL reagent (Life Technologies) and cDNA was generated an OD600 of 0.05 and grown in 50 mL cultures with shaking at using Superscript III (Life Technologies). 180 rpm at 30uC for a total of five days and monitored spectrophotometrically. Melanization assays were performed on MIC determination and MPA resistance solid or liquid Cryptococcus melanization media containing L-DOPA MIC susceptibility testing for MPA was performed using the (L-3,4-dihydroxyphenylalanine) and 10 mM asparagine at 30uC broth microdilution method according to CLSI (CLSI M27-A2) and 37uC [71]. Spectrophotometric analysis was performed on modified for Cryptococcus neoformans [80]. Strains tested are found in 10 culture supernatant at OD475 using a NanoDrop spectrophotom- Table S2. For analysis of spontaneous MPA resistance, 1610 eter [72]. Capsule growth was induced via overnight growth in cells each of H99 overnight cultures were plated onto YNB with RPMI 1640 media plus 2% glucose and 10% fetal bovine serum, sub-lethal concentrations of MPA (1–5 mg/mL) as described [21]. and stained with India ink (Becton Dickinson) to visualize capsule. Plates were incubated at 30uC for two months. Putatively resistant Capsule sizes were determined from phase contrast images using colonies were subcultured onto fresh YNB plus MPA plates. CellProfiler image analysis software [73]. An algorithm was developed to identify the cryptococcal capsule and cell wall Cloning, chimera construction and site-directed boundaries by segmentation. Clumped cells, those outside size mutagenesis bounds and those touching the border of the image were excluded. All primers used in the study are listed in Table S1. IMPDH Geometric parameters were measured from the identified capsule genes were amplified from C. neoformans and C. gattii strains, cloned and cell wall boundaries, and the average of the major and minor into pCR2.1-TOPO (Life Technologies) and sequenced. Wild- axes from an ellipsoid fit used to estimate both capsule size and type strains utilized are listed in Table S2. IMD1 from C. neoformans cellular diameter. Capsule diameter was expressed as a relative strain H99 and from C. gattii strain were subcloned into percentage of cell size. At least 10 independent images were used pBluescript SK- (Agilent, Santa Clara CA), then pPZP-NEO from each sample with at least 50 and typically 100–200 cells (containing a G418 resistance cassette) and pPZP-NAT (contain- measured from each replicate; all measurements were performed ing a nourseothricin resistance cassette) to create complementation in biological triplicate [74]. Mating tests were performed on constructs [81]. Complementation constructs for C. neoformans Murashige-Skoog or V8 media plus 100 mg/mL myo-inositol bearing the E. coli xanthine-guanine phosphoribosyltransferase- (pH 5.0) using the MATa H99 isogenic strain KN99a, as encoding gpt and E. coli IMP dehydrogenase-encoding guaB genes described [75,76]. were generated via overlap PCR using the C. neoformans ACT1 promoter and TRP1 terminator, combined with the E. coli strain Nematode virulence assays K12 gpt or guaB ORF, cloned into pCR2.1-TOPO and sequenced. Assessment of cryptococcal virulence in the C. elegans model Site-directed mutagenesis was used to introduce new residues system was performed as described [77]. Cryptococcus strains were into wild-type CnIMD1. All six variants were subcloned into grown on either rich (Brain-Heart Infusion media, Becton pBluescript SK- and further subcloned into pPZP-NEO or directly Dickinson) or minimal (Nematode Growth Medium plus S. into pPZP-NAT. Double mutants and the sextuple mutant were cerevisiae Synthetic Complete Supplement Mixture) media with or created using successive rounds of mutagenesis. All mutagenesis without supplemental guanine. For C. elegans plus antifungal assays, products were sequenced to confirm identity and fidelity. IMPDH 5, 10 or 20 mg/mL MPA or 5 mg/mL fluconazole were added to gene chimeras were created by dividing the gene into thirds using the minimal media plates. For the control drug assays, worms were two common restriction sites in CnIMD1 and CgIMD1 (XhoI and cultured on NGM supplemented with 6 mM 5-fluoro-29-deoxyur- NcoI) and reconstructing all six possible combinations. The first idine to prevent egg laying, plus either OP50 bacteria or heat- fragment comprises 207 residues and contains the first two unique killed OP50 bacteria, with or without 20 mg/mL MPA. residues V/M55 and A/T153, the second fragment comprises 225 residues and contains only the third unique residue R/K336, and the final third comprises 111 residues and contains the final three Murine virulence assays and organ burden analysis unique residues R/E446, G/A450 and A/V500. Subsequently, all Assessment of cryptococcal virulence in the murine model chimeras were subcloned into pLITMUS28i (New England system was performed as described [78], with slight modification. Biolabs, Ipswich MA) before further subcloning into both pPZP- Due to the propensity for the imd1D strain to clump in liquid NEO and pPZP-NAT. All strains created are listed in Table S3. media, all strains were freshly grown on YNB or YNB plus guanine plates overnight before washing and counting for Expression, purification and crystallization of infection. Kaplan-Meier survival curves were plotted using the Cryptococcus IMPDH software package GraphPad Prism 5.0 (GraphPad Software, San In-frame cloning of the cDNA coding for CnImd1 and CgImd1 Diego CA), and statistical significance was assessed using a log- into expression vectors, heterologous expression in E. coli, rank test, with a p value ,0.05 considered significant. For organ purification of protein and crystallization parameters have been burden analysis, brain, lungs, liver, spleen and kidneys were described previously [82]. For this study, purified protein in harvested from three mice per time point, weighed, homogenized crystallization buffer was mixed with a 10-fold molar excess of the and plated in serial dilution to determine colony-forming units per substrate IMP plus 100 mM of the inhibitor MPA dissolved in gram organ weight. dimethylsulfoxide prior to crystallization. Large crystals were

PLOS Pathogens | www.plospathogens.org 13 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus obtained in wells containing 1.9 M lithium sulfate and 0.09 M PyMOL (http://www.pymol.org). Atomic coordinates and struc- imidazole/HCl pH 6.5 with either no additive or 3.0–12.0% ture factors have been deposited in the Protein Data Bank, www. pentaerythritol ethoxylate (15/04 EO/OH) for CnImd1. pdb.org (PDB ID: 4af0).

Steady-state and inhibitor kinetics Supporting Information All enzymatic assays contained standard IMPDH assay buffer Figure S1 Specificity of Cryptococcus Hpt1. (A) The (50 mM Tris-HCl (pH 8.0), 100 mM KCl, 3 mM EDTA, 1 mM nucleobase xanthine and the nucleosides guanosine and xantho- DTT, 500 mM IMP and 250 mM NAD+) and were performed as sine rescued the imd1D mutant, though with less robust growth described [21]. Enzymatic assays were performed at 25uC in triplicate by monitoring the production of NADH at 340 nm than guanine. Neither XMP nor GMP were able to complement (e = 6.22 mM21 cm21) using a Varian Cary 50 spectrophotometer the auxotrophy of the imd1D strain. Hypoxanthine also did not rescue the mutant, suggesting that the Cryptococcus a-ketoglutarate- (Agilent Technologies, Santa Clara CA). For Km determination, IMP and NAD+ concentrations were varied over a range from dependent dioxygenase is unable to perform the conversion of hypoxanthine to xanthine found in the Schizosaccharomyces pombe 62.5 mM to 5,000 mM, with fixed second substrate concentrations of 250 mM NAD+ and 500 mM IMP, respectively. Nonlinear homolog. (B) All xanthylic and guanylic nucleotides rescue the fitting of the data to the Michaelis-Menton equation (Equation 1) MPA phenotype in the wild-type and complemented strains. Only and the uncompetitive substrate inhibition equation (Equation 2) nucleoside monophosphates after the IMPDH blockage by MPA were performed using SigmaPlot 12.0 (SysStat Software, Chicago rescue the hpt1D phenotype. (C) The ade2D mutant, defective in de novo purine metabolism before the pathway branchpoint, can IL). IC50 values for both proteins were determined using fixed concentrations of IMP (850 mM or 1,800 mM) and NAD+ utilize either adenine or hypoxanthine for purine nucleotide (1,000 mM) while varying the concentration of MPA (0 nM– biosynthesis. Only adenine can rescue the ade2D hpt1D phenotype on minimal media. As hypoxanthine successfully supplements the 400 nM). The Km and kcat values were then determined in the presence of concentrations of MPA that produced 0%, 25%, 50% purine auxotrophy of the ade2D mutant, Hpt1 can salvage and 75% inhibition. Velocity data were fitted to the uncompetitive hypoxanthine. tight-binding inhibition equation (Equation 3) and the noncom- (TIF) petitive/mixed tight-binding inhibition equation (Equation 4) in Figure S2 Quantification of melanin production. Mela- SigmaPlot, nin production was measured in liquid L-DOPA medium from culture supernatant at OD475. (A) imd1D at 30uC. (B) imd1D at v~kcat E S =(Kmz S ) 1 37uC. (C) hpt1D at 30uC. (D) hpt1D at 37uC. Bars represent mean ½ ½ ½ ð Þ OD475 from three replicates with standard error shown. No significant differences were found between strains. (TIF) v~k E =(1zK = S z S =K ) 2 cat½ m ½ ½ ii ð Þ Figure S3 Effect of 20 mg/mL MPA on nematode survival. N2 Bristol young adult nematodes were cultivated for : 2 eight days on OP50 or heat-killed OP50 on standard NGM v~(v0=(2 E )) ( E { I {Kappz (( E { I {Kapp) z4 E Kapp)) ½ ½ ½ ½ ½ ½ 3 supplemented with 6 mM 5-fluoro-29-deoxyuridine to prevent egg qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð Þ Kapp~Kii(1zKm= S ) laying, plus or minus 20 mg/mL MPA, the highest concentration ½ of MPA used in the Cryptococcus/MPA nematode virulence assays. No significant differences were found between treatments.

: 2 (TIF) v~(v0=(2 E )) ( E { I {Kappz (( E { I {Kapp) z4 E Kapp)) ½ ½ ½ ½ ½ ½ 4 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð Þ Figure S4 Quantitative reverse transcription PCR anal- K ~( S zK )=((K zK )z( S =aK )) app ½ m m ii ½ ii ysis of IMD1 expression. (A) Strain H99 and strain MMRL2651 were grown in YNB media at 30 C overnight and where v is the initial velocity; K is the Michaelis constant; k is u m cat treated with 1 mM guanine before harvesting at three timepoints. the turnover number; [E] is enzyme concentration; [S] is substrate (B) Strains grown overnight and treated with 5 g/mL MPA concentration; [I] is inhibitor concentration; K and K are the m ii is before harvesting at three timepoints. Transcript levels of CnIMD1 intercept and slope inhibition constants. and CgIMD1 are expressed as a fold change relative to expression of the housekeeping gene b-tubulin. Standard error bars are Structure determination shown. p,0.05 *; p,0.01 **. Data were collected on the MX2 beamline of the Australian (TIF) Synchrotron, Melbourne, Australia, using the Blu-Ice software [83]. Diffraction data were processed and reduced using XDS and Figure S5 Alignment of IMPDH from all eight molecu- SCALA [84,85]. Initial crude molecular replacement phases were lar types of Cryptococcus. IMPDH was amplified and obtained using a search model based on the coordinates of the sequenced from the four molecular types for which there was no human type I IMPDH (PDB ID 1jcn) using the fast Fourier existing sequence data (VNII, VNB, VGIII & VGIV) and aligned transform molecular replacement method as implemented in with existing sequence data for VNI, VNIV, VGI and VGII. Phaser [86] as implemented in the PHENIX suite [87,88]. After MPA-resistant C. gattii VGIV IMPDH has 22 substitutions extensive manual re-building in Coot [89], combined with compared to MPA-sensitive C. neoformans VNI IMPDH, although maximum-likelihood-based restrained refinement in BUSTER- only six of these residues are not shared with another molecular TNT [90], the final model had Rwork/Rfree values of 0.17/0.20 type. Residues unique to VGIV are highlighted in red. The (Table S4). The stereochemistry of the structure was assessed and IMPDH accessory domain is highlighted in green, the active site validated with MolProbity [91]. Data collection and refinement loop in blue and the mobile flap in orange. statistics are listed in Table S4. Figures were generated with (TIF)

PLOS Pathogens | www.plospathogens.org 14 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus

Figure S6 Growth of all IMPDH mutants on MPA. Serial Initial velocity data was best described by the uncompetitive tight- dilution spotting assays of all IMPDH mutants on YNB plus 5 mg/ binding inhibition equation versus NAD+, while initial velocity data mL MPA. The juxtaposed images depict which portions of the of MPA versus IMP were best fit by a noncompetitive/mixed tight- IMPDH allele are present, with yellow representing CnIMD1 and binding model. (A) CnImd1. (B) CgImd1. blue representing CgIMD1, split into three thirds with the vertical (TIF) bars depicting the six unique residues. Two proteins are depicted for the H99 background where the wild-type allele is also present. Table S1 Primers used in this study. (A) Transformation of IMPDH variants (point mutants, double (DOC) mutants, sextuple mutants, chimeras plus the two wild-type alleles) Table S2 Wild-type strains used in this study. into the imd1D deletion background. (B) Transformation of (DOC) IMPDH variants (point mutants, double mutants, sextuple mutants, chimeras plus the two wild-type alleles) into the wild-type H99 Table S3 Strains created for this study. background, which possesses one copy of the MPA-sensitive allele. (DOC) (TIF) Table S4 Crystallographic data collection and refinement Figure S7 Steady-state kinetics of Cryptococcus IMPDH. statistics. Plots of velocity versus IMP and velocity versus NAD+ concentration (DOC) were generated by fixing one substrate (250 mM IMP and 500 mM NAD+) and varying the other. For both enzymes, velocity versus IMP Acknowledgments plots were best described by the Michaelis-Menton equation, while + The authors would like to thank Dr Drew Titmarsh for creating the velocity versus NAD plots were best fit by the uncompetitive CellProfiler analysis pipeline. substrate inhibition equation. (A) CnImd1. (B) CgImd1. (TIF) Author Contributions Figure S8 Inhibitor kinetics of Cryptococcus IMPDH. Conceived and designed the experiments: CAM UK BK JAF. Performed Inhibition by MPA was investigated at fixed concentration of IMP the experiments: CAM EV AS EWLC IRL AW SJW JMH JTD. Analyzed (850 mM for CnImd1 and 1,800 mM for CgImd1) and varying the data: CAM EV JAF. Wrote the paper: CAM JAF. NAD+ (62.5 mM–2,000 mM) and indicated MPA concentrations.

References 1. Liu M, Healy MD, Dougherty BA, Esposito KM, Maurice TC, et al. (2006) 16. Schiappa D, Gueyikian A, Kakar S, Alspaugh JA, Perfect JR, et al. (2002) An Conserved fungal genes as potential targets for broad-spectrum antifungal drug auxotrophic pigmented Cryptococcus neoformans strain causing infection of the bone discovery. Eukaryot Cell 5: 638–649. marrow. Med Mycol 40: 1–5. 2. Abadio AK, Kioshima ES, Teixeira MM, Martins NF, Maigret B, et al. (2011) 17. Perfect JR, Toffaletti DL, Rude TH (1993) The gene encoding phosphor- Comparative genomics allowed the identification of drug targets against human ibosylaminoimidazole carboxylase (ADE2) is essential for growth of Cryptococcus fungal pathogens. BMC Genomics 12: 75. neoformans in cerebrospinal fluid. Infect Immun 61: 4446–4451. 3. Roemer T, Jiang B, Davison J, Ketela T, Veillette K, et al. (2003) Large-scale 18. Chapuis AG, Paolo Rizzardi G, D’Agostino C, Attinger A, Knabenhans C, et al. essential gene identification in Candida albicans and applications to antifungal (2000) Effects of mycophenolic acid on human immunodeficiency virus infection drug discovery. Mol Microbiol 50: 167–181. in vitro and in vivo. Nat Med 6: 762–768. 4. Hu W, Sillaots S, Lemieux S, Davison J, Kauffman S, et al. (2007) Essential gene 19. Umejiego NN, Li C, Riera T, Hedstrom L, Striepen B (2004) Cryptosporidium identification and drug target prioritization in Aspergillus fumigatus.PLOSPathog parvum IMP dehydrogenase: identification of functional, structural, and dynamic 3: e24. properties that can be exploited for drug design. J Biol Chem 279: 40320–40327. 5. Justice MC, Hsu MJ, Tse B, Ku T, Balkovec J, et al. (1998) Elongation factor 2 20. Gollapalli DR, Macpherson IS, Liechti G, Gorla SK, Goldberg JB, et al. (2010) as a novel target for selective inhibition of fungal protein synthesis. J Biol Chem Structural determinants of inhibitor selectivity in prokaryotic IMP dehydroge- 273: 3148–3151. nases. Chem Biol 17: 1084–1091. 6. Botet J, Rodriguez-Mateos M, Ballesta JP, Revuelta JL, Remacha M (2008) A 21. Kohler GA, Gong X, Bentink S, Theiss S, Pagani GM, et al. (2005) The chemical genomic screen in Saccharomyces cerevisiae reveals a role for diphthamida- functional basis of mycophenolic acid resistance in Candida albicans IMP tion of translation elongation factor 2 in inhibition of protein synthesis by dehydrogenase. J Biol Chem 280: 11295–11302. sordarin. Antimicrob Agents Chemother 52: 1623–1629. 22. McMillan FM, Cahoon M, White A, Hedstrom L, Petsko GA, et al. (2000) 7. Heitman J (2010) Cryptococcus: From human pathogen to model yeast. Heitman J, Crystal structure at 2.4 A resolution of Borrelia burgdorferi inosine 59- Kozel TR, Kwon-Chung KJ, Perfect JR, Casadevall A, editors. Washington, monophosphate dehydrogenase: evidence of a substrate-induced hinged-lid DC: ASM Press. 646 p. motion by loop 6. Biochemistry 39: 4533–4542. 8. Saag MS, Graybill RJ, Larsen RA, Pappas PG, Perfect JR, et al. (2000) Practice 23. Umejiego NN, Gollapalli D, Sharling L, Volftsun A, Lu J, et al. (2008) Targeting guidelines for the management of cryptococcal disease. Infectious Diseases a prokaryotic protein in a eukaryotic pathogen: identification of lead compounds Society of America. Clin Infect Dis 30: 710–718. against cryptosporidiosis. Chem Biol 15: 70–77. 9. Park BJ, Wannemuehler KA, Marston BJ, Govender N, Pappas PG, et al. (2009) 24. Eells JT, Spector R (1983) Purine and pyrimidine base and nucleoside Estimation of the current global burden of cryptococcal meningitis among concentrations in human cerebrospinal fluid and plasma. Neurochem Res 8: persons living with HIV/AIDS. AIDS 23: 525–530. 1451–1457. 10. Perfect JR, Dismukes WE, Dromer F, Goldman DL, Graybill JR, et al. (2010) 25. Cultrone A, Scazzocchio C, Rochet M, Montero-Moran G, Drevet C, et al. Clinical practice guidelines for the management of cryptococcal disease: 2010 (2005) Convergent evolution of hydroxylation mechanisms in the fungal update by the Infectious Diseases Society of America. Clin Infect Dis 50: 291– kingdom: molybdenum cofactor-independent hydroxylation of xanthine via 322. alpha-ketoglutarate-dependent dioxygenases. Mol Microbiol 57: 276–290. 11. Sable CA, Strohmaier KM, Chodakewitz JA (2008) Advances in antifungal 26. Sintchak MD, Fleming MA, Futer O, Raybuck SA, Chambers SP, et al. (1996) therapy. Annu Rev Med 59: 361–379. Structure and mechanism of inosine monophosphate dehydrogenase in complex 12. Elion GB (1989) The purine path to chemotherapy. Science 244: 41–47. with the immunosuppressant mycophenolic acid. Cell 85: 921–930. 13. Kirsch DR, Whitney RR (1991) Pathogenicity of Candida albicans auxotrophic 27. Guillen Schlippe YV, Riera TV, Seyedsayamdost MR, Hedstrom L (2004) mutants in experimental infections. Infect Immun 59: 3297–3300. Substitution of the conserved Arg-Tyr dyad selectively disrupts the hydrolysis 14. Jiang L, Zhao J, Guo R, Li J, Yu L, et al. (2010) Functional characterization and phase of the IMP dehydrogenase reaction. Biochemistry 43: 4511–4521. virulence study of ADE8 and GUA1 genes involved in the de novo purine 28. Digits JA, Hedstrom L (1999) Kinetic mechanism of Tritrichomonas foetus inosine biosynthesis in Candida albicans. FEMS Yeast Res 10: 199–208. 59-monophosphate dehydrogenase. Biochemistry 38: 2295–2306. 15. Rodriguez-Suarez R, Xu D, Veillette K, Davison J, Sillaots S, et al. (2007) 29. Gan L, Seyedsayamdost MR, Shuto S, Matsuda A, Petsko GA, et al. (2003) The Mechanism-of-action determination of GMP synthase inhibitors and target immunosuppressive agent mizoribine monophosphate forms a transition state validation in Candida albicans and Aspergillus fumigatus. Chem Biol 14: 1163–1175. analogue complex with inosine monophosphate dehydrogenase. Biochemistry 42: 857–863.

PLOS Pathogens | www.plospathogens.org 15 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus

30. Digits JA, Hedstrom L (1999) Species-specific inhibition of inosine 59- 58. Hansen BG, Sun XE, Genee HJ, Kaas CS, Nielsen JB, et al. (2012) Adaptive monophosphate dehydrogenase by mycophenolic acid. Biochemistry 38: evolution of drug targets in producer and non-producer organisms. Biochem J 15388–15397. 441: 219–226. 31. Craig SP, 3rd, Eakin AE (2000) Purine phosphoribosyltransferases. J Biol Chem 59. Hedstrom L, Wang CC (1990) Mycophenolic acid and thiazole adenine 275: 20231–20234. dinucleotide inhibition of Tritrichomonas foetus inosine 59-monophosphate 32. Cox GM, Mukherjee J, Cole GT, Casadevall A, Perfect JR (2000) Urease as a dehydrogenase: implications on enzyme mechanism. Biochemistry 29: 849–854. virulence factor in experimental cryptococcosis. Infect Immun 68: 443–448. 60. Riera TV, Wang W, Josephine HR, Hedstrom L (2008) A kinetic alignment of 33. Olszewski MA, Noverr MC, Chen GH, Toews GB, Cox GM, et al. (2004) orthologous inosine-59-monophosphate dehydrogenases. Biochemistry 47: Urease expression by Cryptococcus neoformans promotes microvascular sequestra- 8689–8696. tion, thereby enhancing central nervous system invasion. Am J Pathol 164: 61. Gan L, Petsko GA, Hedstrom L (2002) Crystal structure of a ternary complex of 1761–1771. Tritrichomonas foetus inosine 59-monophosphate dehydrogenase: NAD+ orients the 34. Rodrigues ML, Nakayasu ES, Oliveira DL, Nimrichter L, Nosanchuk JD, et al. active site loop for catalysis. Biochemistry 41: 13309–13317. (2008) Extracellular vesicles produced by Cryptococcus neoformans contain protein 62. Pankiewicz KW, Patterson SE, Black PL, Jayaram HN, Risal D, et al. (2004) components associated with virulence. Eukaryot Cell 7: 58–67. Cofactor mimics as selective inhibitors of NAD-dependent inosine monophos- 35. Kozhevnikova EN, van der Knaap JA, Pindyurin AV, Ozgur Z, van Ijcken WF, phate dehydrogenase (IMPDH) - the major therapeutic target. Curr Med Chem et al. (2012) Metabolic enzyme IMPDH is also a transcription factor regulated 11: 887–900. by cellular state. Mol Cell 47: 133–139. 63. Brenner S (1974) The genetics of Caenorhabditis elegans. Genetics 77: 71–94. 36. Ngamskulrungroj P, Chang Y, Sionov E, Kwon-Chung KJ (2012) The primary 64. Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: A Laboratory nd target organ of Cryptococcus gattii is different from that of Cryptococcus neoformans in a Manual. 2 edition. Cold Spring Harbor (New York): Cold Spring Harbor murine model. M Bio 3: e00103–12. Laboratory Press. 37. Chayakulkeeree M, Johnston SA, Oei JB, Lev S, Williamson PR, et al. (2011) 65. Pitkin JW, Panaccione DG, Walton JD (1996) A putative cyclic peptide efflux SEC14 is a specific requirement for secretion of phospholipase B1 and pump encoded by the TOXA gene of the plant-pathogenic fungus Cochliobolus pathogenicity of Cryptococcus neoformans. Mol Microbiol 80: 1088–1101. carbonum. Microbiology 142: 1557–1565. 38. Baiga TJ, Guo H, Xing Y, O’Doherty GA, Dillin A, et al. (2008) Metabolite 66. Davidson RC, Cruz MC, Sia RA, Allen B, Alspaugh JA, et al. (2000) Gene induction of Caenorhabditis elegans dauer larvae arises via transport in the pharynx. disruption by biolistic transformation in serotype D strains of Cryptococcus ACS Chem Biol 3: 294–304. neoformans. Fungal Genet Biol 29: 38–48. 39. Breger J, Fuchs BB, Aperis G, Moy TI, Ausubel FM, et al. (2007) Antifungal 67. Toffaletti DL, Rude TH, Johnston SA, Durack DT, Perfect JR (1993) Gene chemical compounds identified using a C. elegans pathogenicity assay. PLOS transfer in Cryptococcus neoformans by use of biolistic delivery of DNA. J Bacteriol Pathog 3: e18. 175: 1405–1411. 40. Bovers M, Hagen F, Boekhout T (2008) Diversity of the Cryptococcus neoformans- 68. Fraser JA, Subaran RL, Nichols CB, Heitman J (2003) Recapitulation of the Cryptococcus gattii species complex. Rev Iberoam Micol 25: S4–12. sexual cycle of the primary fungal pathogen Cryptococcus neoformans var. gattii: 41. Ngamskulrungroj P, Gilgado F, Faganello J, Litvintseva AP, Leal AL, et al. implications for an outbreak on Vancouver Island, Canada. Eukaryot Cell 2: (2009) Genetic diversity of the Cryptococcus species complex suggests that 1036–1045. Cryptococcus gattii deserves to have varieties. PLOS One 4: e5862. 69. McDade HC, Cox GM (2001) A new dominant selectable marker for use in Cryptococcus neoformans. Med Mycol 39: 151–154. 42. Jackson RC, Morris HP, Weber G (1977) Partial purification, properties and 70. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local regulation of inosine 59-phosphate dehydrogenase in normal and malignant rat alignment search tool. J Mol Biol 215: 403–410. tissues. Biochem J 166: 1–10. 71. D’Souza CA, Alspaugh JA, Yue C, Harashima T, Cox GM, et al. (2001) Cyclic 43. Guillen Schlippe YV, Hedstrom L (2005) Guanidine derivatives rescue the AMP-dependent protein kinase controls virulence of the fungal pathogen Arg418Ala mutation of Tritrichomonas foetus IMP dehydrogenase. Biochemistry Cryptococcus neoformans. Mol Cell Biol 21: 3179–3191. 44: 16695–16700. 72. Hicks JK, D’Souza CA, Cox GM, Heitman J (2004) Cyclic AMP-dependent 44. Hedstrom L (2009) IMP dehydrogenase: structure, mechanism, and inhibition. protein kinase catalytic subunits have divergent roles in virulence factor Chem Rev 109: 2903–2928. production in two varieties of the fungal pathogen Cryptococcus neoformans. 45. Morrison JF (1969) Kinetics of the reversible inhibition of enzyme-catalysed Eukaryot Cell 3: 14–26. reactions by tight-binding inhibitors. Biochim Biophys Acta 185: 269–286. 73. Lamprecht MR, Sabatini DM, Carpenter AE (2007) CellProfiler: free, versatile 46. Sun XE, Hansen BG, Hedstrom L (2011) Kinetically controlled drug resistance: software for automated biological image analysis. Biotechniques 42: 71–75. how Penicillium brevicompactum survives mycophenolic acid. J Biol Chem 286: 74. Zaragoza O, Fries BC, Casadevall A (2003) Induction of capsule growth in 40595–40600. Cryptococcus neoformans by mammalian serum and CO(2). Infect Immun 71: 6155– 47. Josephine HR, Ravichandran KR, Hedstrom L (2010) The Cys319 loop 6164. modulates the transition between dehydrogenase and hydrolase conformations 75. Nielsen K, Cox GM, Wang P, Toffaletti DL, Perfect JR, et al. (2003) Sexual in inosine 59-monophosphate dehydrogenase. Biochemistry 49: 10674–10681. cycle of Cryptococcus neoformans var. grubii and virulence of congenic a and alpha 48. Makowska-Grzyska M, Kim Y, Wu R, Wilton R, Gollapalli DR, et al. (2012) isolates. Infect Immun 71: 4831–4841. Bacillus anthracis inosine 59-monophosphate dehydrogenase in action: The first 76. Xue C, Tada Y, Dong X, Heitman J (2007) The human fungal pathogen bacterial series of structures of phosphate ion-, substrate-, and product-bound Cryptococcus can complete its sexual cycle during a pathogenic association with complexes. Biochemistry 51: 6148–6163. plants. Cell Host Microbe 1: 263–273. 49. Escandon P, Sanchez A, Martinez M, Meyer W, Castaneda E (2006) Molecular 77. Mylonakis E, Ausubel FM, Perfect JR, Heitman J, Calderwood SB (2002) Killing epidemiology of clinical and environmental isolates of the Cryptococcus neoformans of Caenorhabditis elegans by Cryptococcus neoformans as a model of yeast pathogenesis. species complex reveals a high genetic diversity and the presence of the Proc Nat Acad Sci U S A 99: 15675–15680. molecular type VGII mating type a in Colombia. FEMS Yeast Res 6: 625–635. 78. Morrow CA, Lee IR, Chow EW, Ormerod KO, Goldinger A, et al. (2012) A 50. Chaudhary K, Darling JA, Fohl LM, Sullivan WJ, Jr., Donald RG, et al. (2004) unique chromosomal rearrangement in the Cryptococcus neoformans var. grubii type Purine salvage pathways in the apicomplexan parasite Toxoplasma gondii. J Biol strain enhances key phenotypes associated with virulence. M Bio 3: e:00310–11. Chem 279: 31221–31227. 79. Lee IR, Chow EW, Morrow CA, Djordjevic JT, Fraser JA (2011) Nitrogen 51. Boitz JM, Ullman B (2006) Leishmania donovani singly deficient in HGPRT, APRT metabolite repression of metabolism and virulence in the human fungal or XPRT are viable in vitro and within mammalian macrophages. Mol Biochem pathogen Cryptococcus neoformans. Genetics 188: 309–323. Parasitol 148: 24–30. 80. Clinical and Laboratory Standards Institute. (2005) Reference method for broth 52. Santiago AE, Cole LE, Franco A, Vogel SN, Levine MM, et al. (2009) dilution antifungal susceptibility testing of yeasts; approved standard-second Characterization of rationally attenuated Francisella tularensis vaccine strains that edition. CLSI document M27-A2. Wayne, PA: Clinical and Laboratory harbor deletions in the guaA and guaB genes. Vaccine 27: 2426–2436. Standards Institute. 53. Samant S, Lee H, Ghassemi M, Chen J, Cook JL, et al. (2008) Nucleotide 81. Idnurm A, Reedy JL, Nussbaum JC, Heitman J (2004) Cryptococcus neoformans biosynthesis is critical for growth of bacteria in human blood. PLOS Pathog 4: virulence gene discovery through insertional mutagenesis. Eukaryot Cell 3: 420– e37. 429. 54. Pimkin M, Markham GD (2008) The CBS subdomain of inosine 59- 82. Morrow CA, Stamp A, Valkov E, Kobe B, Fraser JA (2010) Crystallization and monophosphate dehydrogenase regulates purine nucleotide turnover. Mol preliminary X-ray analysis of mycophenolic acid-resistant and mycophenolic Microbiol 68: 342–359. acid-sensitive forms of IMP dehydrogenase from the human fungal pathogen 55. Hyle JW, Shaw RJ, Reines D (2003) Functional distinctions between IMP Cryptococcus. Acta Crystallogr Sect F Struct Biol Cryst Commun 66: 1104–1107. dehydrogenase genes in providing mycophenolate resistance and guanine 83. McPhillips TM, McPhillips SE, Chiu HJ, Cohen AE, Deacon AM, et al. (2002) prototrophy to yeast. J Biol Chem 278: 28470–28478. Blu-Ice and the Distributed Control System: software for data acquisition and 56. van der Knaap JA, Kozhevnikova E, Langenberg K, Moshkin YM, Verrijzer CP instrument control at macromolecular crystallography beamlines. J Synchrotron (2010) Biosynthetic enzyme GMP synthetase cooperates with ubiquitin-specific Radiat 9: 401–406. protease 7 in transcriptional regulation of ecdysteroid target genes. Mol Cell Biol 84. Kabsch W (2010) XDS. Acta Crystallogr D Biol Crystallogr 66: 125–132. 30: 736–744. 85. Evans P (2006) Scaling and assessment of data quality. Acta Crystallogr D Biol 57. Pimkin M, Pimkina J, Markham GD (2009) A regulatory role of the Bateman Crystallogr 62: 72–82. domain of IMP dehydrogenase in adenylate nucleotide biosynthesis. J Biol 86. McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, et al. Chem 284: 7960–7969. (2007) Phaser crystallographic software. J Appl Crystallogr 40: 658–674.

PLOS Pathogens | www.plospathogens.org 16 October 2012 | Volume 8 | Issue 10 | e1002957 GTP Biosynthesis in Cryptococcus

87. DiMaio F, Terwilliger TC, Read RJ, Wlodawer A, Oberdorfer G, et al. (2011) 90. Blanc E, Roversi P, Vonrhein C, Flensburg C, Lea SM, et al. (2004) Refinement Improved molecular replacement by density- and energy-guided protein of severely incomplete structures with maximum likelihood in BUSTER-TNT. structure optimization. Nature 473: 540–543. Acta Crystallogr D Biol Crystallogr 60: 2210–2221. 88. Adams PD, Afonine PV, Bunkoczi G, Chen VB, Davis IW, et al. (2010) 91. Chen VB, Arendall WB, 3rd, Headd JJ, Keedy DA, Immormino RM, et al. PHENIX: a comprehensive Python-based system for macromolecular structure (2010) MolProbity: all-atom structure validation for macromolecular crystallog- solution. Acta Crystallogr D Biol Crystallogr 66: 213–221. raphy. Acta Crystallogr D Biol Crystallogr 66: 12–21. 89. Emsley P, Cowtan K (2004) Coot: model-building tools for molecular graphics. Acta Crystallogr D Biol Crystallogr 60: 2126–2132.

PLOS Pathogens | www.plospathogens.org 17 October 2012 | Volume 8 | Issue 10 | e1002957

Appendix C

Comparative Genomics of Serial Isolates of Cryptococcus neoformans Reveals Gene Associated with Carbon Utilization and Virulence

184 INVESTIGATION

Comparative Genomics of Serial Isolates of Cryptococcus neoformans Reveals Gene Associated With Carbon Utilization and Virulence

Kate L. Ormerod,*,† Carl A. Morrow,*,† Eve W. L. Chow,*,† I. Russel Lee,*,† Samantha D. M. Arras,*,† Horst Joachim Schirra,†,‡ Gary M. Cox,§ Bettina C. Fries,**,†† and James A. Fraser*,†,1 *Australian Infectious Diseases Research Centre, †School of Chemistry and Molecular Biosciences, and ‡Centre for § Advanced Imaging, University of Queensland, Brisbane, QLD 4072 Australia, Department of Medicine, Mycology Research Unit, Duke University, Durham, North Carolina 27710, and **Department of Microbiology & Immunology and ††Department of Medicine, Albert Einstein College of Medicine, Bronx, New York 10461

ABSTRACT The opportunistic fungal pathogen Cryptococcus neoformans is a leading cause of mortality KEYWORDS among the human immunodeficiency virus/acquired immunodeficiency syndrome population and is known microevolution for frequently causing life-threatening relapses. To investigate the potential contribution of in-host micro- fungal evolution to persistence and relapse, we have analyzed two serial isolates obtained from a patient with pathogenesis acquired immunodeficiency syndrome who suffered an initial and relapse episode of cryptococcal menin- cryptococcosis goencephalitis. Despite being identical by multilocus sequence typing, the isolates differ phenotypically, relapse exhibiting changes in key virulence factors, nutrient acquisition, metabolic profiles, and the ability to disseminate in an animal model. Whole-genome sequencing uncovered a clonal relationship, with only a few unique differences. Of these, two key changes are expected to explain the phenotypic differences observed in the relapse isolate: loss of a predicted AT-rich interaction domain protein and changes in copy number of the left and right arms of chromosome 12. Gene deletion of the predicted transcriptional regulator produced changes in melanin, capsule, carbon source use, and dissemination in the host, consistent with the phenotype of the relapse isolate. In addition, the deletion mutant displayed altered virulence in the murine model. The observed differences suggest the relapse isolate evolved subsequent to penetration of the central nervous system and may have gained dominance following the administration of antifungal therapy. These data reveal the first molecular insights into how the Cryptococcus neoformans genome changes during infection of humans and the manner in which microevolution progresses in this deadly fungal pathogen.

Microevolution enables rapid adaptation to selective pressures, permitting expansion into new niches and persistence in the face of ever-changing harsh environmental conditions. Pathogenic organisms encounter a wide Copyright © 2013 Ormerod et al. doi: 10.1534/g3.113.005660 array of challenges during infectionofthehumanhost,includingox- Manuscript received January 17, 2013; accepted for publication February 16, 2013 idative and nitrosative stresses from immune cells, nutrient limitation, This is an open-access article distributed under the terms of the Creative and high temperature. Accordingly, genomic changes that confer selec- Commons Attribution Unported License (http://creativecommons.org/licenses/ tive advantages enabling their survival and proliferation have been iden- by/3.0/), which permits unrestricted use, distribution, and reproduction in any fi medium, provided the original work is properly cited. ti ed in diverse species. The gastric pathogen Helicobacter pylori relies Supporting information is available online at http://www.g3journal.org/lookup/ on an elevated mutation rate combined with genetic exchange be- suppl/doi:10.1534/g3.113.005660/-/DC1. tween strains during mixed infection to generate genetic diversity Sequence data from this article have been deposited with the NCBI Short Read (Kennemann et al. 2011), and chronic Pseudomonas aeruginosa infec- Archive under BioProject PRJNA189304. tion associated with cystic fibrosis involves selection against virulence 1Corresponding author: 358 Molecular Biosciences Building, Cooper Road, The University of Queensland, Brisbane, QLD 4072 Australia. E-mail: jafraser@uq. factors required for an acute infection (Smith et al. 2006; Hoboth et al. edu.au 2009). Such flexibility is not limited to bacterial pathogens but is also

Volume 3 | April 2013 | 675 evident in eukaryotic parasites such as Leishmania (Ubeda et al. 2008) grubii. Two isolates were obtained from the cerebrospinal fluid of and Trypanosoma (Minning et al. 2011). This capacity for rapid geno- a woman positive for human immunodeficiency virus suffering from micchangerenderspathogensdifficult to counter as they can develop cryptococcal meningoencephalitis, the first during initial diagnosis and resistance to drug therapies and evade the host immune response. the second during a relapse episode 77 days later. Using whole-genome Pathogenic fungi also undergo microevolution during infection, sequencing of both isolates, we describe not only the shared ancestry most notably in the emergence of drug resistance, which has been but also key differences between the two, ranging from single-nucleotide described in the three most common systemically infecting fungal variation to chromosomal aneuploidy. A single mutation in an AT-rich genera: Candida, Aspergillus,andCryptococcus (Selmecki et al. 2006; interaction domain (ARID) protein appears responsible for multiple Sionov et al. 2010; Mortensen et al. 2011). However, beyond antifun- phenotypic differences observed between the isolated strains, and may gal resistance, the adaptive potential of mutation during infection also contribute to a dissemination defect observed in the later isolate predicts the emergence of fungal strains that are better suited to the supporting its evolution during infection of the central nervous system hostile host environment. With an estimated global burden of 625,000 (CNS). deaths per annum (Park et al. 2009), Cryptococcus neoformans is one such fungal pathogen for which this potential is of significant clinical MATERIALS AND METHODS relevance. A key feature of C. neoformans pathogenesis is the ability to Strains and phenotypic assays establish dormant infections, as immunocompetent hosts are thought to acquire an asymptomatic or subclinical initial infection that remains C. neoformans var. grubii strains used in this study were the type strain latent until immune suppression, at which time the infection can H99 (Perfect et al. 1980) and serial isolates F0 and F2 and G0, G1, and emerge (Garcia-Hermoso et al. 1999). Patients who appear to have G2. Fraser Lab H99 is a subculture of the Heitman Lab H99 #4413 been successfully treated still frequently relapse, with the majority of (Morrow et al. 2012). F0 and F2 were derived from the cerebrospinal fl fi relapses believed to be a consequence of persistence of the original uid (77 days apart) of a patient infected with human immunode - strain rather than reinfection with a new isolate (Desnos-Ollivier et al. ciency virus who presented with cryptococcal meningoencephalitis to 2010). The molecular mechanisms underlying latency and relapse re- Jacobi Medical Center in the Bronx, New York in 1994. The patient main unknown and although the patients’ compromised immune sys- was initially treated with amphotericin B and had also received at least fl tem is undoubtedly a contributing factor, changes in the fungal genome 7daysof uconazole prior to obtaining the F0 isolate. The patient was fl may aid the escape from dormancy at the opportune moment. discharged and presumably further treated with uconazole (compliance Previous analyses of serial isolates of C. neoformans have estab- unclear). She was readmitted 2 months later with relapsed/persistent fl lished the occurrence of in-host microevolution and attendant phe- cryptococcal meningoencephalitis. Amphotericin B and uconazole notypic variation, but the precise molecular events occurring in the treatment was reinstituted. The F2 isolate was cultured from the cere- fl genome are poorly understood, although changes in karyotype indi- brospinal uid 14 days later. Both F0 and F2 have been frozen contin- cating large-scale genomic rearrangements are frequently observed uously except during transfer from Albert Einstein College of Medicine. (Fries et al. 1996; Fraser et al. 2005). Experimental evolution studies Details of phenotypic assays are given in the Supporting Information, in Saccharomyces cerevisiae under nutrient-limiting conditions have File S1. demonstrated that large chromosomal mutations occur at a high frequency under such conditions and provide adaptive potential (Dunham Multilocus sequence typing (MLST) et al. 2002; Gresham et al. 2008). Similarly stressful conditions are likely and transposon mapping to be encountered during a pathogens’ progress through the host, pro- Strains were grown in yeast peptone dextrose (YPD) at 30 with viding ideal selective pressure for adaptive mutations to proliferate and shaking for 16 hr. Washed cell pellets were frozen and lyophilized spread throughout the population. and DNA extracted via the CTAB extraction method as described Recent studies of C. neoformans var. grubii,themostclinically (Pitkin et al. 1996). Sanger sequencing data were analyzed using prevalent form of this pathogen, have shown that it possesses a re- Sequencher (4.7, Gene Codes Corp, Ann Arbor, MI) and compared markably stable genome, with few rearrangements since its divergence with the Cryptococcus MLST sequence database (http://cneoformans. from the less common var. neoformans an estimated ~20 million years mlst.net). A search for known transposons of C. neoformans with ago (Morrow et al. 2012). However, there is evidence that the genome Basic Local Alignment Search Tool (BLAST) (Altschul et al. 1990) becomes more plastic when placed under the stresses encountered was performed to identify matches in the var. grubii strain H99 ge- while infecting a human host, with the appearance of remarkably nome. Primers were designed based on these matches to amplify probes different electrophoretic karyotypes not only between patients, but from H99 genomic DNA for detection of the DNA transposon Cnirt2 also between initial and relapse infections of an individual (Fries and the retrotransposons Tcn1, Tcn2, Tcn4, and Tcn6. Genomic DNA et al. 1996). Clinical isolates readily display profound changes in from each strain was digested, electrophoretically separated, and South- karyotype (Fries et al. 1996), and isolates placed under antifungal ern blot (Southern 2006) hybridization patterns for each selected ele- stress rapidly accumulate characteristic transient aneuploidies that ment were then compared. are lost when the drug pressure is removed (Sionov et al. 2010). A nascent theme from ongoing molecular analyses of the systemic fungal Pulsed-field gel electrophoresis pathogens is that although they appear to possess low genetic vari- Preparation of agarose-embedded intact Cryptococcus chromosomal ability and are genetically stable, once under the selective pressure of DNA was performed as described (Lengeler et al. 2000). Chromo- a host, microevolution and genetic variation rapidly increases. Elegant somes were separated in 1% pulsed-field certified agarose gels using studies in the commensal Candida albicans have also demonstrated aCHEF-DRIIIpulsed-field gel electrophoresis system (Bio-Rad, Rich- that passage through a host or exposure to various stressors increases mond, CA) in 0.5· Tris-borate-ethylenediaminetetraacetic acid run- the rate of genetic and phenotypic variation (Forche et al. 2009, 2011). ning buffer. Running conditions were as follows: Block 1: ramped Here we describe the first genome sequencing-based investigation switch time from 75 sec to 150 sec, 120,4V/cm,20hr;Block2: into microevolution between serial isolates of C. neoformans var. ramped switch time from 200 sec to 400 sec, 120,4V/cm,30hr.All

676 | K. L. Ormerod et al. separations were performed at 10214 using a Bio-Rad cooling mod- may involve more than one strain, and relapse may sometimes rep- ule. Chromosomes were stained and visualized with ethidium bromide. resent a new infection with a new isolate (Desnos-Ollivier et al. 2010). We therefore initiated our study of serial isolate strains F0 and F2 by 1D nuclear magnetic resonance (NMR) spectroscopy performing MLST analysis with 12 well-established polymorphic loci Metabolite extraction, 1D 1H NMR spectroscopy, analysis, and me- to determine their relationship (Litvintseva et al. 2006). The MLST tabolite profiling were performed as described (Morrow et al. 2012). profiles of F0 and F2 were identical, consistent with isolate F2 being Strains H99, F0, and F2 were cultured in YNB or YPD plus 2% glucose derived from the original F0 strain. This genotype (designated M1) is at 37 for 16 hr for metabolite extraction. The statistical characteristics widespread and in the United States accounted for greater than 25% of of the resulting models are listed in Table S3. isolates in a study by Litvintseva et al. (2006); it differs from the well- studied common laboratory strain H99 MLST genotype by only a sin- Sequencing and genomic analysis gle nucleotide in one (LAC1) of the 12 loci. Although the identical fi Using the Illumina Genome Analyzer II, the Australian Genome MLST pro les of F0 and F2 are consistent with a serial relationship, Research Facility (Brisbane, Australia) generated 75-bp paired end the ubiquity of this genotype and similarity to H99 highlights the reads (~17 million for F0 and F2, ~8 million for H99) with an average limited resolution of this technique in a species that is predominantly insert size of 200 bp. Reads for H99 were generated to provide a clonal. control against which variation in F0 and F2 could be normalized. A more powerful epidemiological tool to discriminate between fi Details of read trimming, mapping, and assembly are given in File S1. highly related isolates is an analysis of their transposon pro le. Transposable elements were identified using RepeatMasker (version Approximately 5% of the C. neoformans genome consists of trans- open-3.3.0) (Smit et al. 1996-2010). Structural variation was detected posons (Loftus et al. 2005), and the distribution of these has been using BreakDancer (Chen et al. 2009), CREST (Wang et al. 2011), and shown to be highly variable among strains (Jain et al. 2005). We used fi Dindel (Albers et al. 2011). Details are provided in File S1. a Southern blotting approach that included the use of probes to ve transposons (Cnirt2, Tcn1, Tcn2, Tcn4, and Tcn6) identified by fi Molecular techniques Goodwin & Poulter (2001) to compare their hybridization pro le against an array of clinical isolates and each other. Of these trans- Details of molecular techniques are given in File S1 in the supplementary posons, Cnirt2, Tcn1, Tcn2, and Tcn4 showed identical profiles be- information. AVC1 was amplified from H99, cloned into pCR2.1-TOPO tween F0 and F2, profiles that differed from all other isolates tested. (Life Technologies), and subsequently subcloned into the neomycin The Tcn6 profile was highly variable in all isolates, indicating that this resistance marker containing the plasmid pJAF1 (Fraser et al. 2003) and element is likely still functional and highly mobile (Figure S1). Com- the nourseothricin resistance marker containing pCH233 (McDade and bined with the MLST result, these data show F0 and F2 are clonally Cox 2001). Biolistic transformation was performed as described related and therefore represent an opportunity to characterize within- (Magditch et al. 2012). Details of primers used in this study are host microevolutionary events that occur during human infection. contained in Table S4. F0 and F2 exhibit altered virulence factor expression Nematode and murine survival assays but respond similarly to multiple stressors Survival assay of the nematode Caenorhabditis elegans was performed The theory that C. neoformans undergoes microevolutionary change according to the protocol of Mylonakis et al. (2002) using strain N2 in response to selective pressures encountered in the human host Bristol with Escherichia coli OP50 as the food source (Brenner 1974). predicts that strains F0 and F2 should exhibit differences in phenotype Assays were performed in triplicate. For murine assays, 7-wk-old fe- such as their virulence factor profile and response to stress. Testing male BALB/c mice were infected via nasal inhalation (Cox et al. 2000). revealed growth of the relapse strain F2 exceeded F0 and more closely Ten mice for each strain were inoculated by pipetting a 50-mLdrop resembled that of H99, the var. grubii type strain, at all temperatures, containing 2.5 · 106 cells (H99, F0, F2) or 5 · 105 (H99, avc1D, H99 + particularly at 37 and 39, potentially indicative of selection for opti- AVC1) onto their nares. Mice were killed when their body weight had mized growth at human body temperature (Figure 1). Unexpectedly, reduced by 20% from their preinfection weight. The brain, lungs, capsule size was reduced in F2, as was melanization at 37.Protease spleen, and liver from killed mice were subsequently removed, homog- production appeared greater in F2 and phospholipase B and urease enized, and plated to determine colony-forming units per gram organ levels were equivalent between the strains; however, these results are weight. For both survival assays, Kaplan-Meier survival curves were difficult to interpret due to the growth differential between the strains. plotted using GraphPad Prism 6.0 (GraphPad Software, La Jolla CA) Both strains were largely unaffected when exposed to oxidative, and significance was determined using a log-rank test, whereas organ nitrosative, ultraviolet rays, membrane and osmotic stress, nutrient burden significance was determined using the Student’s t-test. Pvalues limitation, and most cell wall stressors (Figure S2). F0 was severely of , 0.05 were considered significant. Murine virulence assays were impaired at pH 9 and greater and F2 at pH 11 (Figure S2). A slight conducted in accordance with the Australian code of practice for the increase in resistance to cell wall stress was observed in F2 grown on care and use of animals for scientific purposes by the National Health caffeine. F2 also displayed increased resistance to fluconazole, with an and Medical Research Council and were approved by the Molecular MIC (i.e., the minimum inhibitory concentration needed to inhibit Biosciences Animal Ethics Committee of The University of Queens- 90 the growth of 90% of organisms) of 25 mg/mL, twice that seen for land (AEC approval number: SCMB/008/11/NHMRC). strain F0 (12.5 mg/mL). The acquisition of increased fluconazole re- fl RESULTS sistance is consistent with the patient having received uconazole as part of their treatment regime. Serial isolates F0 and F2 share a recent common ancestor Serial isolates F0 and F2 therefore meet our criteria for strains that Although relapse of cryptococcal meningoencephalitis is generally have undergone microevolution: MLST and transposon studies support caused by a derivative of the original infecting strain (Spitzer et al. a common origin, whereas phenotypic analysis reveals emergent differ- 1993), recent work has shown that approximately 20% of infections ences in key phenotypes associated with proliferation and virulence. This

Volume 3 April 2013 | Comparative Genomics of C. neoformans | 677 Unique differences exist between F0 and F2 Of the 12,000 differences we detected in F0 and F2 in comparison with the reference strain H99, almost all were common to both isolates. F0 contains just three unique SNVs, one unique indel and a large inversion affecting the first ~210 kb on the left arm of chromosome 3, whereas F2 possesses one unique dinucleotide substitution (Table 1). Of the unique SNVs and indels in F0, the three SNVs are predicted to be silent mutations, whereas the indel alters the coding sequence of DNA polymerase g (CNAG_06769). This creates a deletion of serine 817, a residue conserved in Cryptococcus species, but within a large, 200-residue insertion found only in basidiomycete genomes. Polymer- ase g has been shown to be essential in the maintenance of the mitochondrial genome in both S. cerevisiae (Foury 1989) and Schizo- saccharomyces pombe (Chu et al. 2007). As we estimated the number of mitochondrial genomes per cell to be similar between F0 and F2 (11 and 9, respectively, based on read coverage), there is likely no phenotypic effect from this mutation. The unique inversion in F0 affects the first ~210 kb on the left arm of chromosome 3 and affects two genes: a small oligopeptide transporter (CNAG_03013) and a short hypothetical protein (CNAG_03012) likely to be an incomplete annotation. A BLAST search identifies the transporter as part of the OPT oligopeptide transporter superfamily. The F2 relapse genome contains one unique change, a substitution of a TA dinucleotide with a G within an exon of a hypothetical protein (CNAG_02579) on chromosome 3; the Figure 1 Strains F0 and F2 differ in their production of several ensuing frameshift results in a predicted truncation of the protein virulence factors. (A) Growth assays at 30, human body temperature from 1543 to 130 residues. The gene appears to be unique to Crypto- of 37, and febrile body temperature of 39 on YPD. F2 displays similar growth to H99 at 30 and 37 but diminished growth at 39. coccus, with high identity to homologs in C. neoformans var. neofor- (B) India ink staining under light microscopy reveals the capsule; F2 is mans and C. gattii but no other species. The predicted protein reduced compared with F0. Scale bar is 10 mM. (C) Melanization was contains an ARID found in proteins associated with transcriptional comparable at 30 on L-DOPA2containing media; however, F2 was regulation (Kortschak et al. 2000), which is lost after truncation of the not melanized at 37. (D) F0 produces lower levels of extracellular protein. The presence of genetic variations in F0 absent from F2 protease when grown on bovine serum albumin agar. (E) and (F) suggests the relationship between the isolates is not a simple chrono- Comparable levels of phospholipase and urease production were logical one and instead likely represents a clonal expansion of an ’ observed when strains were grown on egg yolk and Christensen s original infecting isolate to form multiple independent lineages, each agar, respectively. having evolved concurrently during infection. indicates microevolutionary events not observable through MLST Both F0 and F2 exhibit aneuploidies of chromosome 12 and transposon profile analysis have occurred during infection, Careful inspection of the karyotype of strain F2 reveals a minichro- presenting an ideal opportunity to utilize whole genome sequencing mosome of approximately 300 kb, only faintly visible in high quality technologies to uncover their causative mutations. pulsed-field gels (Figure 3A). Evaluation of read depth revealed a trip- lication of the left arm of chromosome 12 in F2 (Figure 3B). The size Substantial genomic variation exists between F0 and F2 of the minichromosome suggested it may represent an isochromo- and the reference strain H99 some comprising both the extra copies of the 140 kb left arm of Illumina Genome Analyzer II reads for each strain were mapped chromosome 12 (12L). Subsequent Southern hybridizations using tar- against the Broad Institute’s unpublished 19 Mb draft genome of geted probes confirmed the minichromosome contained 12L, and H99, which comprises 14 chromosomes ranging in size from ~700 digests of intact whole chromosomes on pulsed-field gels yielded kb to ~2.3 Mb. Alignment produced 99.5% coverage of the refer- bands that were larger than would be expected if each copy of 12L ence genome at 59·,58·,and30· read depth for F0, F2, and H99, were to be maintained individually (results not shown). Sequence read respectively. In our MLST analysis we had identified one single- depth analysis also uncovered an aneuploidy of chromosome 12 in nucleotide variant (SNV) between H99 and our serial clinical iso- strain F0, which possesses two complete copies of the chromosome. lates F0 and F2; when expanded to the whole genome, 12,119 SNVs Via parsimony, this suggests that the duplicate copy of the right arm were identified in F0 and F2 in comparison with H99, representing of chromosome 12 was lost during the infection process and the left 0.06% of the genome (Figure 2). In addition, 1279 shared inser- arm was then triplicated, potentially indicating a selective advantage tionsanddeletionslessthan50bpinlength(thoseaffectinggenes associated with multiple copies of this genomic region. with functional annotation are listed in Table S1A) and 33 larger Chromosome 12 contains 270 predicted genes, 3.9% of the total deletions (Table S1B) were detected in the alignments. Seven ad- genes currently annotated in the H99 genome; 56 of these are con- ditional insertions present in F0 and F2 were subsequently detected tained on the left arm (Table S2). We canvassed the genes on this arm using de novo assembly of reads that remained unmapped to the for those that may be potentially advantageous during infection. There reference sequence (Table S1B). In short, the strains differ consid- are four dehydrogenases present, including three aldehyde and aryl- erably from the control strain H99 despite their almost identical alcohol dehydrogenases potentially involved in oxidative stress resis- MLST profile. tance. Interestingly, both genes required for asparagine catabolism are

678 | K. L. Ormerod et al. n TABLE 1 Unique SNVs and indels in F0 and F2 Strain Chr H99 Change Gene function Effect of mutation F0 2 CTT — DNA polymerase g Deletion of S817 (of 1,433) 7 C A Conserved hypothetical protein (280 bp upstream) 10 T A Flavin-containing monooxygenase (30 bp downstream) 11 G C Dynamin Synonymous mutation F2 3 AT G Conserved hypothetical protein Q39R, truncation at 130 (of 1,543) present, asparaginase (CNAG_06008) and aspartate transaminase the Cryptococcus cluster also includes a putative trehalose permease (CNAG_06026). Asparagine is a preferred nitrogen source of Crypto- (Slot and Rokas 2010). We subsequently assayed growth of F0 and F2 coccus and produces robust melanization (Lee et al. 2011). However, on galactose and trehalose and found F2, which in previous assays our initial assays revealed F2 was unable to produce significant showed stronger growth, was severely impaired on galactose at both amounts of melanin on L-DOPA media containing asparagine as 30 and 37, whereas both strains were inhibited on trehalose (Figure the nitrogen source at 37. We extended our analysis of melanization 4). This finding raised the possibility that the growth differential seen to other nitrogen sources to discern any potential gene dosage effects between F0 and F2 was sensitive to the carbon source provided. from the aneuploidies. Melanin production remained low in F2 at 37 Growth assays on a range of other sugar and nonsugar carbon sources on all nitrogen sources tested (Figure S3), and this defect extended to revealed F2 was also diminished on lactose and maltose, while the 30 on most other nitrogen sources, although the extent of inhibition apparently weaker strain F0 outperformed F2 on most non-sugar was not as drastic. carbon sources tested, except ethanol (Figure S4). Growth on combi- We next examined genes on the right arm of chromosome 12 for nations of glucose plus a variety of nonsugar carbon sources resem- any that could potentially contribute to the phenotypic differences bled that of glucose alone, suggesting the phenotype is dominant (data between the two strains. We first noted the presence of the virulence not shown). factor gene phospholipase B (PLB1,CNAG_06085);however,increased copy number of this gene in F0 did not yield increased levels F0 and F2 exhibit differences in metabolic profiles of secreted phospholipase B as determined by egg yolk agar assay The observed differences in nutrient acquisition prompted an in- (Figure 1E). Also present on this arm is the galactose gene cluster, vestigation of the metabolic profiles of F0 and F2. We performed 1D consisting of UDP-glucose epimerase 2 (GAL10), galactokinase proton NMR spectroscopy on metabolites extracted after growth at (GAL1), and galactose-1-phosphate uridylyltransferase (GAL7). Un- 37 in rich or minimal media containing the most easily assimilated like the clusters of Saccharomyces, Candida,andSchizosaccharomyces, carbon source glucose. Correlation of spectral peaks with existing

Figure 2 Extensive SNVs and indels are observed in F0 and F2 compared with H99. Circos plot of the 14 Cryptococcus chromosomes (outermost rectangles) depicting the .12,000 SNVs and 1200 indels detected in serial isolates F0 and F2 compared with the reference strain H99. Inner rings represent (from outermost): SNVs common to F0 and F2, indels common to F0 and F2, SNVs and indels unique to F0; and SNVs and indels unique to F2.

Volume 3 April 2013 | Comparative Genomics of C. neoformans | 679 Figure 3 Karyotypic analysis reveals a minichromosome in F2 associated with aneuploidy observed via read depth analysis. (A) Pulsed-ﬁeld gel electrophoresis of reference strain H99 and serial isolates F0 and F2 reveals a minichromosome of approximately 300 kb. (B) Read depth analysis of serial isolates F0 and F2 in comparison with H99 indicates a duplication of chromosome 12 in F0 and a triplica- tion of the left arm of chromosome 12 in F2.

metabolite databases and chemical shift standards revealed primarily in vivo, we conducted virulence assays in two separate animal model sugars and amino acids were present, in particular glucose and treha- systems. In the nematode infection model, no significant change in lose, as well as other metabolites such as succinate, acetate, and glyc- survival time of Caenorhabditis elegans was observed when cultured erol. Principal components analysis (PCA) showed systematic on strains F0 and F2 over 8 d (Figure 6A). In the murine inhalation differences between all three strains in both growth media (Figure model, we initially observed apparently similar virulence between 5A) with the metabolic fingerprint of F2 being more similar to H99 strains F2 and H99, whereas F0 appeared attenuated (Figure 6B). than F0, in agreement with the growth pattern observed in other However, two mice cleared the infection with strain F2, whereas all glucose based assays. Pairwise comparisons between F0 and F2 F0-infected mice died. Overall, survival times did not differ signifi- in minimal media revealed clear separation (Figure 5B), with strain cantly between F0 and F2. Analysis of fungal burden within multiple F2 associated with greater levels of trehalose, choline, and lactate and organs, however, revealed significant differences in the progression of decreased levels of glucose, glycerol, glutamate, glutamine, formate, infection between the two strains: F2 cells disseminated from the lungs acetate, alanine, and glycerophosphocholine (Figure 5C). These differ- poorly, and failed to reach the brain in all but two mice (Figure 6C). ences in metabolite profiles were much smaller in rich media than Organ colony-forming unit per gram was significantly reduced for the in minimal media, with isolates not dividing into separate clusters and the corresponding model having only marginal predictive power, indicating compromised metabolic pathways may be present in F0 which could result in a greater sensitivity to environmental conditions (Figure S5,AandB,andTable S3). In addition, systematic differences were observed for each individual isolate when profiles were compared between rich and minimal media (Figure S5,C-F). The metabolites observed are in agreement with those identified in previous C. neoformans metabolomics studies, which have reported dominance of trehalose and lipids (Bubb et al. 1999; Himmelreich et al. 2001). Trehalose is also associated with capacity to grow at 37 (Petzold et al. 2006) and increased levels of trehalose have been associated with heat tolerant mutants of Cryptococcus (Morrow et al. 2012), in agreement with the increased capacity of F2 for growth at this temperature. The large number of genes affected by the duplication of chromosome 12 likely contributes to the metabolome differences observed between the strains; for example the right arm contains two glycerol-1-phosphatases, amplification of which may be responsible for increased glycerol levels in F0. However, more than half the genes on this chromosome have no functional annotation, leaving many potential contributors unknown. Figure 4 Strains F0 and F2 exhibit different growth on alternate F0 and F2 are equally virulent in nematode and murine carbon sources. Growth assays on minimal media supplemented with models but F2 displays compromised dissemination various carbon sources. F0 displays diminished growth on glucose and To determine whether the microevolutionary changes observed trehalose at 30 and 37. F2 displays poor growth on galactose and between strains F0 and F2 are associated with a selective advantage myo-inositol at both temperatures, and acetate at 37.

680 | K. L. Ormerod et al. Figure 5 Strains F0 and F2 exhibit different metabolic profiles. (A) Three-dimensional scores plot of the PCA analysis of F0, F2, and H99 under two different growth conditions. All six metabolic types can be distinguished. Red: F2 in YNB; orange: F0 in YNB; yellow: H99 in YNB; green: F2 in YPD; cyan: F0 in YPD; blue: H99 in YPD. Hotelling’s 95% confidence range of the PCA model is indicated by the ellipsoid. (B) PCA scores plot of principal components 1 and 2 for F0 (orange triangles) and F2 (red squares) in YNB. Distance between the points is an indicator of similarity between the samples. (C) Corresponding bivariate loadings line plot. The line plot displays on the loadings coefficient axis the (center scaled) correlation coefficients that relate individual integral regions of the NMR spectrum to the PC1 axis of the scores plot. Individual peaks correspond to peaks in the 1D NMR spectra; peaks that are positive indicate metabolites significantly increased in F2 whereas negative peaks are increased in F0. The overlaid heat map relates the relative contribution of the peak to the scores plot when using Pareto scaling instead of center scaling. brain, liver, and spleen of F2-infected mice, and fungal burden in the relatively small sample size, we discovered a second series in which lungs themselves was also reduced. Given it was isolated from the CNS, a mutation disrupts this same gene in a relapse isolate, this time in the this dissemination defect suggests the evolution of F2 after penetration third of three isolates denoted G0, G1, and G2. Probing of Southern of the blood brain barrier by the infecting ancestral strain. blotted G2 DNA revealed that a translocation interrupts exon 3 of the ARID-containing gene (results not shown), which would result in Mutation of the ARID protein occurs in multiple, a truncated protein of approximately 500 residues (of 1543). The independent serial isolates disruption of this gene in a second, unrelated relapse isolate suggests Given no clear association between the genes on chromosome 12 and a potential role in disease recurrence. Further investigation revealed the observed phenotypes, we decided to investigate the ARID- G2 exhibited similar growth inhibition on carbon sources as observed containing gene affected by the single unique mutation identified in in F2, as well as reduced capsule size. None of the strains in the series F2. We thus performed sequencing of this gene in our existing produced significant amounts of melanin at 37. Reintroduction of the collection of 14 clinical isolate series. To our surprise, despite the ARID-containing gene into both F2 and G2 mostly abolished the carbon

Volume 3 April 2013 | Comparative Genomics of C. neoformans | 681 source phenotype in both strains (Figure S6). Virulence factor testing revealed melanin production was not restored in F2 however capsule size increased in both complemented strains (Figure S7 and Figure S8). A deletion mutant was subsequently created in the H99 background and revealed similar phenotypic characteristics: reduced growth on a selection of carbon sources, capsule inhibition, and poor melanization at 37. We were able to rescue all three of these phenotypes through reintroduction of the gene (Figures 7 and 8). We therefore dubbed the gene AVC1 for ARID-containing regulator of virulence traits and carbon assimilation. The consistency of the carbon source phenotype between F2 and G2 and avc1Δ indicates the known glucose-related phenotype of H99 is not distorting these results (Morrow et al. 2012). We also tested the three strains in the presence of fluconazole and found slightly increased resistance when AVC1 is disrupted, which was ablated when the gene was reintroduced into avc1Δ and F2, but not G2, suggesting the effect is genotype dependent (Figure S9). Significantly, infection with avc1Δ resulted in a reduction in fungal burden in a similar manner to that observed in F2, although not as severe, as more isolates successfully penetrated the CNS (Figure 9). However, unlike in F2, growth in the lungs was not inhibited in avc1Δ. We also observed an overall significant reduction in virulence of avc1Δ which was not seen in F2.

DISCUSSION One of the challenges regarding the exploration of within-host microevolution is conclusively demonstrating the relationship between serial isolates. Traditionally, it was believed that most relapse was caused by the original infecting isolate and that these reinfections often corresponded with karyotypic change. More recently it has been shown that mixed infections occur in up to 20% of cases (Desnos- Ollivier et al. 2010) and karyotypic change described as microevolutionary may in fact reflect infection with other unrelated strains. Although MLST analysis has been used to distinguish sequential isolates from co-infection, the resolution of this technique limits the capacity for differentiation as evidenced by our MLST result comparing F0 and F2 with H99. Sequencing technology now permits complete sequencing of serial isolates, and our preliminary work in a larger collection of series has shown that strain relationships fall within two distinct groups: those that differ by thousands of SNVs, Figure 6 F0 and F2 exhibit significantly different dissemination which we propose are coinfections, and those that differ by only a few, patterns despite apparently similar virulence. (A) No significant which we propose are cases of microevolution, such as occurs in F0 difference in survival was seen between each of the strains when and F2. Significantly, our results indicate the evolution of multiple young adult C. elegans worms were cultured on indicated strains. (B) A divergent lineages during infection. Although we cannot conclusively murine virulence assay revealed no significant difference in virulence fi state the timeframe over which these mutations occurred, strain F0 between F0 and F2; H99 was signi cantly more virulent than F0 (P = fi had already accumulated unique changes from the common infecting 0.0001). (C) F0 and F2 exhibited signi cantly different patterns of organ burden: ÃP , 0.05, ÃÃP , 0.01, ÃÃÃP , 0.001, ÃÃÃÃP , 0.0001; ancestor at the time of initial diagnosis. This is in line with the Student’s two-tailed t-test, two sample, equal variance (liver) or un- observed prevalence of rapid and large-scale change in studies of equal variance (lungs, brain, spleen). S. cerevisiae (Dunham et al. 2002; Lynch et al. 2008) as well as other fungi such as C. glabrata (Polakova et al. 2009) and Mycosphaerella 2009). The associated gene duplications can have a profound effect graminicola (Stukenbrock et al. 2010). The dissemination defect seen on the proteome (Pavelka et al. 2010) and may also be a substrate for in F2 supports its evolution in-host, once the infection had reached neofunctionalization (Ohno 1970) through accelerated mutation in the CNS. These findings give further credence to the idea that fungal gene copies that are largely free of selection (Kellis et al. 2004). Al- pathogens may possess dedicated mechanisms for rapidly generating though we did not observe an increase in mutation rate on chromo- such large scale changes as a means of adaptation (Selmecki et al. some 12 using a crude measure of SNV per nucleotide, we do not 2010) especially given the stable genome inheritance of var. grubii since know whether this duplication event occurred before or during infec- divergence from var. neoformans 20 million years ago. tion; if it did occur during infection, the timeframe since its appearance may be too short to detect an increase in accumulation of mutations. Aneuploidy of chromosome 12 occurs in both F0 and F2 Tolerance of aneuploidy is contentious; aneuploidy is common The development of aneuploidy as an evolutionary process has been and well tolerated in C. albicans (Selmecki et al. 2009), but demon- extensively described in multiple yeast species, both in vitro and strated to be detrimental or advantageous in S. cerevisiae, presumably in vivo (Kellis et al. 2004; Selmecki et al. 2006; Polakova et al. depending on genetic context and growth conditions (Torres et al.

682 | K. L. Ormerod et al. Figure 7 The growth of the deletion mutant avc1D resembles that of relapse isolate F2 on alternate carbon sources. Growth assays on minimal media supplemented with various carbon sources. Growth inhibition of avc1D is abolished after reintroduction of the gene. Figure 8 Capsule and melanin production are inhibited in avc1D. (A) Growth assays at 30, human body temperature of 37 and febrile body temperature of 39 on YPD; no change was observed following 2007; Pavelka et al. 2010). Although certain disomic C. neoformans deletion of AVC1. (B) India ink staining under light microscopy reveals strains have displayed impaired growth and attenuated virulence in the capsule; capsule production is inhibited in avc1D and restored mice (Sionov et al. 2010), many are afforded increased azole resis- following reintroduction of the gene. Scale bar is 10 mM. (C) Melani- tance. Simultaneous amplification of multiple chromosomes has been zation on L-DOPA containing media is inhibited in avc1D and restored following reintroduction of the gene. (D2F) Comparable levels of pro- linked to changes in melanin production in C. neoformans (Hu et al. tease, phospholipase, and urease production were observed in all 2011). Accumulating examples of aneuploidies suggest chromosomal strains when grown on bovine serum albumin, egg yolk, and Christen- fi ampli cation as an important mechanism of variation in C. neoformans. sen’s agar, respectively. Our data support short-term tolerance of aneuploidy in C. neoformans with minimal phenotypic consequences. Swi1 and other members of the ARID1 subfamily, making ortholog Disruption of transcriptional regulator AVC1 in F2 identification difficult. Members of the ARID1 subfamily contain glu- The ARID is associated with chromatin remodeling, and loss of such tamine-rich regions and LXXLL motifs, although the position of these a gene may be pleiotropic and confer broad phenotypic deviation, as is not conserved between family members. Avc1 contains both gluta- observed in other C. neoformans regulatory gene mutants (Cheon mine-rich regions and a single LXXLL motif, as well as proline and et al. 2011; Lee et al. 2011). ARID-containing proteins are found in alanine-rich regions. The other uncharacterized ARID-containing diverse species, including humans, plants, insects, and fungi, and can protein in var. grubii, CNAG_00240, also contains glutamine-rich bind DNA in both a sequence-specific and nonsequence-specificman- regions but no LXXLL motifs. BLAST of both ARID regions yields ner (Kortschak et al. 2000). Seven subfamilies exist in mammalian and Swi1 orthologs for Avc1 only. Further work is needed to determine insect genomes: ARID1 to ARID5 and JARID1 and JARID2. Although whether either of these genes is the Cryptococcus homolog of SWI1. originally named for their interaction with AT-rich sequences, only two of the subfamilies, ARID3 and ARID5, show AT-rich site prefer- Multiple independent disruptions of AVC1 indicate ence (Patsialou et al. 2005). Orthologs of ARID1, JARID1 and poten- a role in relapse tially ARID2 and JARID2 have been identified in fungi (Wilsker et al. The significance of AVC1 is increased by its being identified in two 2005). Swi1, an ARID1 ortholog in S. cerevisiae, is a component of the relapse isolates, indicative of a role in promoting infection persistence. SWI/SNF chromatin remodeling complex that regulates a wide variety However, phenotypic analysis revealed decreased production of two of transcripts, including invertase (SUC2)andgalactokinase(GAL1) key virulence traits, capsule and melanin, and an overall decrease in (Sudarsanam and Winston 2000). S. cerevisiae has one other anno- virulence was observed in avc1D. Clinical isolates of C. neoformans tated ARID-containing protein (Ecm5), whereas Schizosaccharomyces produce inconsistent virulence profiles, however, and melanin in par- pombe contains four including Sol1, a SWI/SNF component. AVC1 is ticular is frequently variable in these isolates (Clancy et al. 2006; one of three annotated ARID-containing proteins in C. neoformans Pereira et al. 2009). Mutants lacking melanin production due to mu- var. grubii; CNAG_00240 on chromosome 1 has no known function tation in the laccase gene LAC1 have been shown to have compro- while CNAG_07411, within the mating-type locus on chromosome 5, mised ability to escape from the lungs, whereas growth in the brain is is the transcription factor Rum1. unaffected (Noverr et al. 2004). Although both F2 and avc1D Typically, although amino acid identity is high within the ARID exhibited a degree of dissemination defect, the liver was colonized region within a subfamily (70280%), between subfamilies identity in all mice, indicating the infection was not trapped in the lungs. drops to below 30%. Fungal ARID regions drop significantly in iden- However, often gene products affecting melanin also affect virulence tity, however, with less than 30% identity between Saccharomyces and/or dissemination independently (Chun et al. 2007; Waterman

Volume 3 April 2013 | Comparative Genomics of C. neoformans | 683 therapy. Disruption of a single gene, AVC1, in separate relapse isolates supports a role for this gene in disease persistence or relapse. These results illustrate the importance of population diversification in pathogenic organisms and the clonal expansion of the fittest member as selective pressures alter. Studies of more serial isolates may facilitate the identification of further genomic variation pertinent to clinical relapse and persistence associated with this important fungal pathogen.

ACKNOWLEDGMENTS This work was supported by the National Health and Medical Research Council (NHMRC project grant 455980 and NHMRC CDA 569673).

LITERATURE CITED Albers, C. A., G. Lunter, D. G. MacArthur, G. McVean, W. H. Ouwehand et al., 2011 Dindel: Accurate indel calls from short-read data. Genome Res. 21: 961–973. Altschul, S. F., W. Gish, W. Miller, E. W. Myers, and D. J. Lipman, 1990 Basic local alignment search tool. J. Mol. Biol. 215: 403–410. Brenner, S., 1974 Genetics of Caenorhabditis elegans. Genetics 77: 71–94. Bubb, W. A., L. C. Wright, M. Cagney, R. T. Santangelo, T. C. Sorrell et al., 1999 Heteronuclear NMR studies of metabolites produced by Crypto- coccus neoformans in culture media: Identification of possible virulence factors. Magn. Reson. Med. 42: 442–453. Chen, K., J. W. Wallis, M. D. McLellan, D. E. Larson, J. M. Kalicki et al., 2009 BreakDancer: an algorithm for high-resolution mapping of genomic structural variation. Nat. Methods 6: 677–681. Cheon, S. A., K.-W. Jung, Y.-L. Chen, J. Heitman, Y.-S. Bahn et al., 2011 Unique evolution of the UPR pathway with a novel bZIP tran- Figure 9 The deletion mutant avc1D exhibits reduced virulence and scription factor, Hxl1, for controlling pathogenicity of Cryptococcus altered dissemination. (A) A murine virulence assay revealed a signifi- neoformans. PLoS Pathog. 7: e1002177. cant difference in virulence between H99 and avc1D (P = 0.0077); no Chu, Z., J. Li, M. Eshaghi, R. K. M. Karuturi, K. Lin et al., 2007 Adaptive significant difference was observed between H99 and avc1D + AVC1. expression responses in the Pol-g null strain of S. pombe depleted of (B) H99 and avc1D exhibited significantly different patterns of organ mitochondrial genome. BMC Genomics 8: 323. burden: ÃP , 0.05, ÃÃP , 0.01; Student’s two-tailed t-test, two sample, Chun, C. D., O. W. Liu, and H. D. Madhani, 2007 A link between virulence equal variance [lungs, brain (avc1D), spleen (avc1D)] or unequal vari- and homeostatic responses to hypoxia during infection by the human – ance [brain (avc1D + AVC1), liver, spleen (avc1D + AVC1)]. fungal pathogen Cryptococcus neoformans. PLoS Pathog. 3: 225 238. Clancy, C. J., M. H. Nguyen, R. Alandoerffer, S. J. Cheng, K. Iczkowski et al., 2006 Cryptococcus neoformans var. grubii isolates recovered from per- et al. 2007). Changes in carbon source utilization as well as melanin sons with AIDS demonstrate a wide range of virulence during murine production at 37 have also been observed in snf1 mutants (Hu et al. meningoencephalitis that correlates with the expression of certain viru- 2008). Snf1, a serine/threonine protein kinase, was found to increase lence factors. Microbiology-Sgm 152: 2247–2255. the expression of LAC1 under low glucose conditions and mutants Cox, G. M., J. Mukherjee, G. T. Cole, A. Casadevall, and J. R. Perfect, were unable to persist in the brain, rendering them avirulent. Al- 2000 Urease as a virulence factor in experimental cryptococcosis. Infect. – though the majority of F2 infections failed to reach the brain, the titer Immun. 68: 443 448. Desnos-Ollivier, M., S. Patel, A. R. Spaulding, C. Charlier, D. Garcia-Hermoso when infection was established was high, indicating F2 was unable to et al., 2010 Mixed infections and in vivo evolution in the human fungal reach the brain rather than unable to persist. Capsule inhibition has pathogen Cryptococcus neoformans. Mbio 1: e00091–e10. been shown to produce strains both with limited growth in the lungs Dunham, M. J., H. Badrane, T. Ferea, J. Adams, P. O. Brown et al., and dissemination to the brain (Wilder et al. 2002). Although F2 2002 Characteristic genome rearrangements in experimental evolution showed lower colony-forming unit in the lungs, avc1D did not, in- of Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 99: 16144–16149. dicating this is unlikely to be contributing to the virulence phenotype Forche, A., D. Abbey, T. Pisithkul, M. A. Weinzierl, T. Ringstrom et al., of avc1D. Future work consisting of virulence studies utilizing the 2011 Stress alters rates and types of loss of heterozygosity in Candida intracranial infection model will help to clarify the role of AVC1 in albicans. Mbio 2: e00129–e11. infection without the confounding variable of CNS penetration. Forche, A., P. T. Magee, A. Selmecki, J. Berman, and G. May, This work reveals the simultaneous evolution of multiple lineages 2009 Evolution in Candida albicans populations during a single passage – derived from a very recent common ancestor during C. neoformans through a mouse host. Genetics 182: 799 811. Foury, F., 1989 Cloning and sequencing of the nuclear gene MIP1 encoding infection. The observed aneuploidies support via parsimony the pres- the catalytic subunit of the yeast mitochondrial DNA polymerase. J. Biol. ence of a chromosome 12 duplication in the common ancestor; how- Chem. 264: 20552–20560. ever, we cannot determine whether this arose during infection or was Fraser, J. A., J. C. Huang, R. Pukkila-Worley, J. A. Alspaugh, T. G. Mitchell present ex vivo. The CNS dissemination defect in F2 (coupled with its et al., 2005 Chromosomal translocation and segmental duplication in recovery 77 d after the initial diagnosis) suggests it arose after pene- Cryptococcus neoformans. Eukaryot. Cell 4: 401–406. tration of the CNS, and the associated increase in fluconazole resis- Fraser, J. A., R. L. Subaran, C. B. Nichols, and J. Heitman, tance may have given the isolate a selective advantage after antifungal 2003 Recapitulation of the sexual cycle of the primary fungal pathogen

684 | K. L. Ormerod et al. Cryptococcus neoformans var. gattii: Implications for an outbreak on Mortensen, K. L., R. H. Jensen, H. K. Johansen, M. Skov, T. Pressler et al., Vancouver Island, Canada. Eukaryot. Cell 2: 1036–1045. 2011 Aspergillus species and other molds in respiratory samples from Fries,B.C.,F.Y.Chen,B.P.Currie,andA.Casadevall,1996 Karyotypeinstability patients with cystic fibrosis: a laboratory-based study with focus on As- in Cryptococcus neoformans infection. J. Clin. Microbiol. 34: 1531–1534. pergillus fumigatus azole resistance. J. Clin. Microbiol. 49: 2243–2251. Garcia-Hermoso, D., G. Janbon, and F. Dromer, 1999 Epidemiological Mylonakis, E., F. M. Ausubel, J. R. Perfect, J. Heitman, and S. B. Calderwood, evidence for dormant Cryptococcus neoformans infection. J. Clin. Mi- 2002 Killing of Caenorhabditis elegans by Cryptococcus neoformans as crobiol. 37: 3204–3209. a model of yeast pathogenesis. Proc. Natl. Acad. Sci. USA 99: 15675– Goodwin, T. J. D., and R. T. M. Poulter, 2001 The diversity of retrotrans- 15680. posons in the yeast Cryptococcus neoformans. Yeast 18: 865–880. Noverr, M. C., P. R. Williamson, R. S. Fajardo, and G. B. Huffnagle, Gresham, D., M. M. Desai, C. M. Tucker, H. T. Jenq, D. A. Pai et al., 2004 CNLAC1 is required for extrapulmonary dissemination of Cryp- 2008 The repertoire and dynamics of evolutionary adaptations to con- tococcus neoformans but not pulmonary persistence. Infect. Immun. 72: trolled nutrient-limited environments in yeast. PLoS Genet. 4: e1000303. 1693–1699. Himmelreich, U., T. E. Dzendrowskyj, C. Allen, S. Dowd, R. Malik et al., Ohno, S., 1970 Evolution by Gene Duplication. Springer, New York. 2001 Cryptococcomas distinguished from gliomas with MR spectros- Park, B. J., K. A. Wannemuehler, B. J. Marston, N. Govender, P. G. Pappas – copy: An experimental rat and cell culture study. Radiology 220: 122 128. et al., 2009 Estimation of the current global burden of cryptococcal Hoboth, C., R. Hoffmann, A. Eichner, C. Henke, S. Schmoldt et al., meningitis among persons living with HIV/AIDS. AIDS 23: 525–530. 2009 Dynamics of adaptive microevolution of hypermutable Pseudo- Patsialou, A., D. Wilsker, and E. Moran, 2005 DNA-binding properties of monas aeruginosa during chronic pulmonary infection in patients with ARID family proteins. Nucleic Acids Res. 33: 66–80. fi – cystic brosis. J. Infect. Dis. 200: 118 130. Pavelka, N., G. Rancati, J. Zhu, W. D. Bradford, A. Saraf et al., Hu, G., P.-Y. Cheng, A. Sham, J. R. Perfect, and J. W. Kronstad, 2010 Aneuploidy confers quantitative proteome changes and pheno- 2008 Metabolic adaptation in Cryptococcus neoformans during early typic variation in budding yeast. Nature 468: 321–325. – murine pulmonary infection. Mol. Microbiol. 69: 1456 1475. Pereira, C. B., F. L. Bueno, A. L. T. Dias, M. Brigagao, C. R. Paula et al., Hu, G., J. Wang, J. Choi, W. Jung, I. Liu et al., 2011 Variation in chro- 2009 Evaluation of laccases and melanization in clinical and environ- fl mosome copy number in uences the virulence of Cryptococcus neofor- mental Cryptococcus neoformans samples by non-denaturing PAGE. J. mans and occurs in isolates from AIDS patients. BMC Genomics 12: 526. Med. Microbiol. 58: 563–566. Jain, N., B. L. Wickes, S. A. Keller, J. Fu, A. Casadevall et al., 2005 Molecular Perfect, J. R., S. D. R. Lang, and D. T. Durack, 1980 Chronic cryptococcal epidemiology of clinical Cryptococcus neoformans strains from India. J. meningitis: A new experimental model in rabbits. Am. J. Pathol. 101: Clin. Microbiol. 43: 5733–5742. 177–193. Kellis, M., B. W. Birren, and E. S. Lander, 2004 Proof and evolutionary Petzold, E. W., U. Himmelreich, E. Mylonakis, T. Rude, D. Toffaletti et al., analysis of ancient genome duplication in the yeast Saccharomyces cer- 2006 Characterization and regulation of the trehalose synthesis pathway evisiae. Nature 428: 617–624. and its importance in the pathogenicity of Cryptococcus neoformans. Kennemann, L., X. Didelot, T. Aebischer, S. Kuhn, B. Drescher et al., Infect. Immun. 74: 5877–5887. 2011 Helicobacter pylori genome evolution during human infection. Pitkin, J. W., D. G. Panaccione, and J. D. Walton, 1996 A putative cyclic Proc. Natl. Acad. Sci. USA 108: 5033–5038. peptide efflux pump encoded by the TOXA gene of the plant-pathogenic Kortschak, R. D., P. W. Tucker, and R. Saint, 2000 ARID proteins come in fungus Cochliobolus carbonum. Microbiology-Uk 142: 1557–1565. from the desert. Trends Biochem. Sci. 25: 294–299. Polakova, S., C. Blume, J. A. Zarate, M. Mentel, D. Jorck-Ramberg et al., Lee, I. R., E. W. L. Chow, C. A. Morrow, J. T. Djordjevic, and J. A. Fraser, 2009 Formation of new chromosomes as a virulence mechanism in 2011 Nitrogen metabolite repression of metabolism and virulence in the yeast Candida glabrata. Proc. Natl. Acad. Sci. USA 106: 2688–2693. human fungal pathogen Cryptococcus neoformans. Genetics 188: 309– Selmecki, A., A. Forche, and J. Berman, 2006 Aneuploidy and isochromo- 323. – Lengeler, K. B., P. Wang, G. M. Cox, J. R. Perfect, and J. Heitman, some formation in drug-resistant Candida albicans. Science 313: 367 2000 Identification of the MATa mating-type locus of Cryptococcus 370. neoformans reveals a serotype A MATa strain thought to have been Selmecki, A., A. Forche, and J. Berman, 2010 Genomic plasticity of the – extinct. Proc. Natl. Acad. Sci. USA 97: 14455–14460. human fungal pathogen Candida albicans. Eukaryot. Cell 9: 991 1008. Litvintseva, A. P., R. Thakur, R. Vilgalys, and T. G. Mitchell, Selmecki, A. M., K. Dulmage, L. E. Cowen, J. B. Anderson, and J. Berman, fi 2006 Multilocus sequence typing reveals three genetic subpopulations 2009 Acquisition of aneuploidy provides increased tness during the of Cryptococcus neoformans var. grubii (Serotype A), including a unique evolution of antifungal drug resistance. PLoS Genet. 5: e1000705. population in Botswana. Genetics 172: 2223–2238. Sionov, E., H. Lee, Y. C. Chang, and K. J. Kwon-Chung, 2010 Cryptococcus Loftus, B. J., E. Fung, P. Roncaglia, D. Rowley, P. Amedeo et al., 2005 The neoformans overcomes stress of azole drugs by formation of disomy in fi genome of the basidiomycetous yeast and human pathogen Cryptococcus speci c multiple chromosomes. PLoS Pathog. 6: e1000848. neoformans. Science 307: 1321–1324. Slot, J. C., and A. Rokas, 2010 Multiple GAL pathway gene clusters evolved Lynch, M., W. Sung, K. Morris, N. Coffey, C. R. Landry et al., 2008 A independently and by different mechanisms in fungi. Proc. Natl. Acad. – genome-wide view of the spectrum of spontaneous mutations in yeast. Sci. USA 107: 10136 10141. – Proc. Natl. Acad. Sci. USA 105: 9272–9277. Smit, A. F. A., R. Hubley, and P. Green, 1996 2010 RepeatMasker Open- Magditch, D. A., T.-B. Liu, C. Xue, and A. Idnurm, 2012 DNA mutations 3.0. Available at: http://www.repeatmasker.org/. mediate microevolution between host-adapted forms of the pathogenic Smith, E. E., D. G. Buckley, Z. N. Wu, C. Saenphimmachak, L. R. Hoffman fungus Cryptococcus neoformans. PLoS Pathog. 8: e1002936. et al., 2006 Genetic adaptation by Pseudomonas aeruginosa to the air- McDade, H. C., and G. M. Cox, 2001 A new dominant selectable marker for ways of cystic fibrosis patients. Proc. Natl. Acad. Sci. USA 103: 8487– use in Cryptococcus neoformans. Med. Mycol. 39: 151–154. 8492. Minning, T. A., D. B. Weatherly, S. Flibotte, and R. L. Tarleton, Southern, E., 2006 Southern blotting. Nat. Protoc. 1: 518–525. 2011 Widespread, focal copy number variations (CNV) and whole Spitzer, E. D., S. G. Spitzer, L. F. Freundlich, and A. Casadevall, chromosome aneuploidies in Trypanosoma cruzi strains revealed by array 1993 Persistence of initial infection in recurrent Cryptococcus neofor- comparative genomic hybridization. BMC Genomics 12: 139. mans meningitis. Lancet 341: 595–596. Morrow, C. A., I. R. Lee, E. W. L. Chow, K. L. Ormerod, A. Goldinger et al., Stukenbrock, E. H., F. G. Jorgensen, M. Zala, T. T. Hansen, B. A. McDonald 2012 A unique chromosomal rearrangement in the Cryptococcus neo- et al., 2010 Whole-genome and chromosome evolution associated with formans var. grubii type strain enhances key phenotypes associated with host adaptation and speciation of the wheat pathogen Mycosphaerella virulence. mBio 3: e00310–11. graminicola. PLoS Genet. 6: 12.

Volume 3 April 2013 | Comparative Genomics of C. neoformans | 685 Sudarsanam, P., and F. Winston, 2000 The Swi/Snf family: nucleosome-remodeling Waterman, S. R., M. Hacham, G. Hu, X. Zhu, Y.-D. Park et al., 2007 Role of complexes and transcriptional control. Trends Genet. 16: 345–351. a CUF1/CTR4 copper regulatory axis in the virulence of Cryptococcus Torres, E. M., T. Sokolsky, C. M. Tucker, L. Y. Chan, M. Boselli et al., neoformans. J. Clin. Invest. 117: 794–802. 2007 Effects of aneuploidy on cellular physiology and cell division in Wilder, J. A., G. K. Olson, Y. C. Chang, K. J. Kwon-Chung, and M. F. haploid yeast. Science 317: 916–924. Lipscomb, 2002 Complementation of a capsule deﬁcient Cryptococcus Ubeda, J.-M., D. Legare, F. Raymond, A. A. Ouameur, S. Boisvert et al., neoformans with CAP64 restores virulence in a murine lung infection. 2008 Modulation of gene expression in drug resistant Leishmania is Am. J. Respir. Cell Mol. Biol. 26: 306–314. associated with gene ampliﬁcation, gene deletion and chromosome an- Wilsker, D., L. Probst, H. M. Wain, L. Maltais, P. W. Tueker et al., euploidy. Genome Biol. 9: R115. 2005 Nomenclature of the ARID family of DNA-binding proteins. Wang, J. M., C. G. Mullighan, J. Easton, S. Roberts, S. L. Heatley et al., Genomics 86: 242–251. 2011 CREST maps somatic structural variation in cancer genomes with base-pair resolution. Nat. Methods 8: 652–654. Communicating editor: B. J. Andrews

686 | K. L. Ormerod et al.