PhD thesis - Københavns Universitet

FACULTY OF SCIENCEUNIVERSITY OF COPENHAGENPhD thesisSigyn JordeFunctional analysis of proteins involved in actincytoskeleton organization in Ashbya gossypiiand Candida albicansAcademic advisor: Carlsberg Laboratory and Department of Biology, University of CopenhagenSubmitted: 04/11/10

Table of Contents1 Preface ........................................................................................................................... 32 List of Abbreviations .................................................................................................... 43 Abstract ......................................................................................................................... 54 Resumé .......................................................................................................................... 65 Introduction .................................................................................................................. 75.1 Fungi as model organisms .......................................................................................... 75.1.1 Saccharomyces cerevisiae .......................................................................................... 75.1.2 Comparative genomics ............................................................................................... 85.1.3 Ashbya gossypii .......................................................................................................... 95.1.4 Candida albicans ...................................................................................................... 105.2 Penelope Research Training Network and foci of this thesis ................................... 125.3 Endocytosis .............................................................................................................. 135.4 Protein-protein interactions ...................................................................................... 145.4.1 SH3 domains ............................................................................................................ 155.5 The actin cytoskeleton .............................................................................................. 155.5.1 Actin polymerization ................................................................................................ 165.5.2 The Arp2/3 complex ................................................................................................. 175.6 Polarized growth ...................................................................................................... 185.6.1 Polarization via Rho-GTPases ................................................................................. 185.6.2 Hyphal growth .......................................................................................................... 195.6.3 The Spitzenkörper .................................................................................................... 205.7 The endocytic machinery ......................................................................................... 215.7.1 Clathrin-dependent endocytosis ............................................................................... 225.7.2 Sla2 and Sac6 are key components in endocytosis of S. cerevisiae ......................... 255.8 Eisosomes ................................................................................................................. 255.8.1 Lipid rafts ................................................................................................................. 266 Methods ....................................................................................................................... 286.1 Strains and media ..................................................................................................... 286.2 Transformation ......................................................................................................... 286.3 Generation of GFP tagged C. albicans strains ......................................................... 296.4 Yeast Two Hybrid assay .......................................................................................... 316.5 β-galactosidase assay ................................................................................................ 336.6 Generation of mutant and GFP strains in A. gossypii .............................................. 336.7 Plate assays and germination of spores .................................................................... 366.8 Microscopy and staining procedures ........................................................................ 367 Results ......................................................................................................................... 377.1 Generation of heterozygous C. albicans strains ....................................................... 377.2 Localization of C. albicans SH3 domain proteins ................................................... 387.3 Yeast two hybrid assay ............................................................................................. 397.4 SLA2 is ascent from the A. gossypii genome ............................................................ 417.5 Generation of Agsac6 ............................................................................................... 427.6 Characterization of Agsac6 ...................................................................................... 431

7.7 Deletion of LSP1, SUR7, PKH1 and YPK1 .............................................................. 467.8 Characterization of lsp1 and sur7 ............................................................................. 467.9 Deletion of PIL1 ....................................................................................................... 497.10 Characterization of the pil1 phenotype .................................................................... 507.11 Localization of AgPIL1-GFP ................................................................................... 547.12 PIL1-GFP does not colocalize with actin patches .................................................... 557.13 Heterologous expression of AgPIL1-GFP in S. cerevisiae ...................................... 568 Discussion .................................................................................................................... 588.1 Localization of C. albicans SH3 domain proteins ................................................... 588.2 Functional relation between genes regulating actin filamentation inS. cerevisiae, C. albicans and A. gossypii ................................................................ 598.3 Vrp1-Wal1-Myo5 complex in C. albicans ............................................................... 608.4 Ligand binding of SH3 domains is weak ................................................................. 618.5 SLA2 is absent from the A. gossypii genome............................................................ 618.6 Agsac6 has a similar phenotype to Agwal1 .............................................................. 638.7 A. gossypii pil1 germlings cease to grow before reaching maturation ..................... 638.8 PIL1-GFP does not localize with cortical actin patches .......................................... 648.9 Sur7 is not necessary for eisosome formation but affects vacuolar fusion .............. 658.10 A link between lipid rafts and endocytosis .............................................................. 659 Summary ..................................................................................................................... 6610 Prospects ..................................................................................................................... 6711 Acknowledgements ..................................................................................................... 6912 References ................................................................................................................... 70Appendix I – Verification PCR on C. albicans strains ........................................................ 78Appendix II – Verification PCR on A. gossypii strains ....................................................... 79Appendix III – Plasmids ........................................................................................................ 80Appendix IV – Strains ............................................................................................................ 82Appendix V – Primers ............................................................................................................ 83Scientific publications ............................................................................................................ 862

1 PrefaceThis thesis presents the results of my PhD internship at Carlsberg Denmark. The project wassupervised by Jürgen Wendland, Professor of Yeast Biology at Carlsberg Denmark, and SteenHolmberg at the University of Copenhagen and was funded by Marie Curie Programme. Mywork resulted in three publications which are listed at the end of this thesis. I begin with anintroduction of fungal biology and the topics of this thesis and continue with a detailedMaterial & Methods part. My research has focused on several topics and they are presented inthe same order in all parts. Results are followed by a discussion and a short summary andfinally I mention the prospects of what I would have done if I had only more time in the lab.3

2 List of AbbreviationsAFMAshbya Full MediumANTH AP180 N-Terminal homologybpbase pairCdCandida dubliniensisCFWCalcofluor WhiteclonNAT nourseothricinCmCandida maltosaCo-IPCo-Immuno PrecipitationCSMComplete Synthetic MediaDiOC6 3,3′-dihexyloxacarbocyanine iodideENTH Epsin N-Terminal homologyG-protein Guanine-nucleotide binding proteinGAPG-protein Activating ProteinGEFGuanine-nucleotide Exchange FactorGPI-anchored Glycosyl Phosphatidyl Inositol-anchoredLYLuciferase YellowNPFNucleation Promoting FactorMAPK Mitogen Activated Protein KinaseMCCMembrane Compartment occupied by Can1ONPG ortho-Nitrophenyl-β-galactosideONPortho-NitrophenolPCRPolymerase Chain ReactionPEG400 Polyethylene Glycol 400SDSSodium Dodecyl SulphateSH3 Src homology 3TETris EDTAYNBYeast Nitrogen BaseYPDYeast Peptone Dextrose4

3 AbstractThis thesis deals with some of the aspects of endocytosis in fungi. The human pathogenCandida albicans and the filamentous Ashbya gossypii were used as models wheninvestigating some of the core mechanisms in this process. The virulence of C. albicans isdependent on its ability to switch between yeast and hyphal growth, which is why thesedynamic processes are of special interest. A. gossypii has been used in comparison to studythe extended polarized growth in hyphae. First, a set of SH3-domain containing proteins inC. albicans were tagged with GFP to allow for visualization and localization. Three proteinswere successfully visualized with a C-terminal GFP, namely Bbc1, Sla1 and Cyk3. Cyk3localized at septal sites and Bbc1 and Sla1, proteins involved in endocytosis, were seen incortical patches. Second, this PhD thesis addressed interactions formed in C. albicans amongthe core proteins in actin filament nucleation. For this we used the yeast two-hybrid assay,which indicates physical interactions between Vrp1 (WIP homolog) and the Arp2/3 nucleatorsWal1 and Myo5 (WASP and myosin I homologs, respectively); a similar complex is formedin S. cerevisiae. Also, functional analysis of the A. gossypii fimbrin mutant sac6 wasperformed, a gene that is synthetic lethal with SLA2 in S. cerevisiae. SLA2 is essential forendocytosis in other studied fungi but the gene is missing in A. gossypii. Deletion of SAC6leads to impaired endocytosis, greatly affected organization of the actin cytoskeleton andswelling hyphae when grown at elevated temperature. These phenotypic characteristics aresimilar to the wal1 (WASP) mutant implying that the two proteins take part in the samepathway. Finally, proteins of the eisosomes, Pil1, Lsp1 and Sur7 in A. gossypii wereinvestigated. As in S. cerevisiae deletion of PIL1 produces the more severe phenotype. Agpil1is barely viable under the growth conditions tested in this assay but investigation of a fewgerminating mutants showed no obvious defects in endocytosis or actin organization. A GFPtagged AgPil1 localized in a punctate pattern beneath the plasma membrane but in separatestructures from cortical actin patches. The same localization was observed in all eisosomemutants as well as in sac6 and wal1. This implies that eisosomes are not directly involved inthe endocytosis process but might have a regulatory function.5

4 ResuméDenne afhandling omhandler nogle aspekter af endocytose i svampe. Ved undersøgelse af decentrale mekanismer i denne proces, blev det humane patogen Candida albicans og denfilamentdannende svamp Ashbya gossypii anvendt som modeller. Virulens af C. albicans erafhængig af dens evne til at skifte mellem gær- og hyfevækst, hvilket er grunden til at dissedynamiske processer er af stor interesse. A. gossypii har været anvendt til at undersøge denudvidede polariserede vækst i hyfer. Først blev et sæt af proteiner, som indeholder et SH3-domæn i C. albicans, markeret med GFP for at muliggøre visualisering og dermedlokalisering. Tre proteiner, nemlig Bbc1, Sla1 og Cyk3, blev visualiseret succesfuldt med etC-terminalt GFP. Cyk3 findes på septale tværvægge, og Bbc1 og Sla1 som er involveret iendocytose findes i aktin-klumper i hyfespidsen. For det andet behandler denne Ph.D.-afhandling interaktioner blandt de centrale proteiner i dannelsen af aktinfilamenter iC. albicans. Her blev der anvendt gær 2-hybrid assay, der viser den fysiske interaktionmellem Vrp1 (den WIP homolog) og de Arp2/3-initiativtager Wal1 og Myo5 (henholdsvisWASP og myosin I homologe). Et lignende kompleks dannes i S. cerevisiae. Derudover blevden A. gossypii fimbrinmutant sac6 undersøgt. SAC6 er et gen, der er syntetisk dødelig iforbindelse med SLA2 i S. cerevisiae. SLA2 er afgørende for endocytose i de øvrigeundersøgte svampe, men genet mangler i A. gossypii. Fjernelse af SAC6 fører til forringetendocytose, forandringer i actinorganisation især i hyfespidsen og hævelse af svampetrådeved forhøjet temperatur. Disse fænotypiske kendetegn, som er identiske i wal1- (WASP)mutant, antyder at de to proteiner deltager i den samme signalproces. Endelig bleveisosomproteiner, nemlig Pil1, Lsp1 og Sur7, undersøgt i A. gossypii. Fjernelse af PIL1medfører en meget alvorlig fænotype, som også ses i S. cerevisiae. Agpil1 er næsten ikkelevedygtig under de vækstbetingelser, som blev testet i denne analyse, men en undersøgelse afnogle spirer viste ingen åbenlyse mangler i endocytose eller aktinorganisation. Et GFPfusioneret AgPil1-protein findes i et punktformet mønster under plasmamembranen, men detblev aldrig lokaliseret samme sted som aktinklumperne. Det forholder sig på denne måde ialle eisosom-mutanter samt i sac6 og wal1. Dette indebærer, at eisosomer ikke er direkteinvolveret i endocytose-processen, men de kunne have en regulerende funktion.6

5 Introduction5.1 Fungi as model organismsFungi can be found in most environments; from the surface of sweet fruits and symbioticlifestyles as mycchorizas and lichens to pathogens of plants, mammals and even other fungi.They can grow as individual cells, yeasts, or as filamentous fungi producing mycelia,sometimes building up a fruit body, mushroom, that you can pick in the forest. Some fungi aredimorphic and upon environmental stimuli are able to switch between the yeast-like andfilamentous growth mode (reviewed by Wendland 2001). Many species produce industriallyimportant metabolites such as antibiotics, ethanol and enzymes that degrade complex organiccompounds (Duff and Murray 1988; von Nussbaum et al. 2006). Fungi has recently emergedas a major cause of human disease, especially among immunocompromised patients (Pfallerand Diekema 2007). The ability of some pathogenic fungi to produce very resilient biofilmson medical implants causes problems to get rid of the infection (Chandra et al. 2001). Animportant feature of fungi is that they are eukaryotic; they have a nucleus containingchromosomes just like in mammalian cells, this makes them especially fitted as modelorganisms.5.1.1 Saccharomyces cerevisiaeSaccharomyces cerevisiae, also known as Bakers’ or Brewers’ yeast, has been studied for along time in the field of molecular biology. Many useful molecular tools have been developedwithin this organism in order to explore the basic molecular biology of eukaryotes. Byexploiting the amazing power of yeast genetics, pathways that are conserved from yeast toman can be unveiled and lay the foundation for understanding the rules of life. The yeast cellspropagate by budding, i.e. by the emergence of a daughter cell that is finally pinched off by aprocess called cytokinesis, completely separating the progeny from the mother. This buddingyeast is easy to manipulate genetically and is generally not harmful to man. S. cerevisiaeoccurs as a diploid in nature; its nucleus contains a double set of chromosomes, as in humans.These cells harbour both mating types, a and α, and in response to starvation the yeast cellproduces four haploid spores through meiosis, two a and two α spores. These spores caneasily be separated and crossed, or mated, with a spore of the opposite mating type. The yeastgenome is spread on sixteen chromosomes with just over twelve million base pairs andcontains about 6000 genes. Sequencing of the yeast genome was completed in 1997 andseveral projects exist to determine the function of all of its genes (Zagulski et al. 1998). About7

100 million years ago, the S. cerevisiae ancestor underwent a whole genome duplication,WGD, followed by a massive gene loss (Fig. 1). Most of the extra copies of genes becameredundant and were lost rapidly but several duplications remained and were allowed to evolvedivergently. This event may have enabled new ways of surviving and adapting to theenvironment. For example, many of the gained genes in S. cerevisiae have a function inethanol production, these genes are lacking in species that diverged before the duplicationevent (Thomson et al. 2005; Gordon et al. 2009). Some genes are still found in duplicates in S.cerevisiae even though their sequences have deviated, these twin genes often haveoverlapping functions.Figure 1. Phylogenetic tree of the Candida and Saccharomyces clades,arrows, showing the whole genome duplication, WGD, and the point fromwhen CTG was used to encode serin insead of leucine. Modified picturefrom Butler et al. 2009.5.1.2 Comparative genomicsYeasts like S. cerevisiae and mammalian cells share many basic biological properties andmany pathways involve homologous genes. As a result, yeast can be exploited to investigatethe function of genes involved in very basic cellular processes, such as DNA repair, celldivision or gene regulation and to test new drugs. In more closely related species like amongvertebrates, the sequences are very similar and the gene order, synteny, is highly conserved.8

The gene map of one species can then be used to help find genes in a related species with apoorly developed map.The two fungi that have been used in this study are the filamentous fungus Ashbya gossypiiand the human pathogen Candida albicans. The genomes of these two species are sequencedand readily available in genome databases. S. cerevisiae, A. gossypii and C. albicans belong tothe family Saccharomycetaceae of the phylum Ascomycota but they represent differentgenera. They also present different ways of growth and comparative functional studies ofprocesses involved in for example endocytosis and polarized growth may aid in revealingimportant and evolutionary conserved mechanisms.5.1.3 Ashbya gossypiiAshbya gossypii (also known as Eremothecium gossypii) is a plant pathogen that growsexclusively as a filamentous fungus. It was isolated from cotton plants (Gossypium hirsutum)and described by Ashby and Nowell in 1926. The fungus was recognized as an overproducerof riboflavin (vitamin B 2 ) which is responsible for its yellow colour (Wickerham et al. 1946).A. gossypii is an attractive model to study filamentous growth since it has a small haploidgenome that is easy to manipulate. Homologous recombination occurs with a high frequencyand it can propagate plasmids (Wright and Philippsen 1991; Steiner et al. 1995). The ninemillion base pair genome of A. gossypii with 4700 protein encoding genes is divided on sevenchromosomes. The A. gossypii lineage diverged from S. cerevisiae before the WGD butdespite their different life styles they are very closely related. The sequencing and annotationof the genome revealed that 95% of the A. gossypii genes have orthologs in S. cerevisiae andthe two species share a high degree of synteny (Dietrich et al. 2004). This knowledge wasuseful in the annotation of the S. cerevisiae genome and to reconstruct the evolution historyfrom their common ancestor.A. gossypii grows as a multinucleated mycelia divided by septa and sustains a polarizedgrowth for almost all of its life cycle (Fig. 2). The life cycle starts with a germinating spore; ashort period of non-polarized growth produces a spherical germ cell. Switch to polarizedgrowth leads to formation of a germ tube. A second germ tube emerges on the opposite sideof the germ cell and the young mycelium continues to branch laterally. About twenty hoursafter germination the hyphae will start to branch dichotomously, producing V-shaped tips, andthe speed of growth will increase about twenty-fold. The growth rate of the mature mycelia9

eaches a maximum of 200 µm/h (Knechtle et al. 2003). Upon sporulation the compartmentsdefined by septa will swell and produce sporangia containing the needle shaped spores. Eachspore can germinate and start a new life cycle. To produce a null mutant of A. gossypii aheterokaryotic mycelium is first generated containing both wild type and mutant nuclei. Thisstrain will grow under a selective pressure and yet behave as the wild type. When sporulated,each spore will contain one nucleus and homokaryotic null mutants can be selected for duringgermination.AFBECFigure 2. Lifecycle of A. gossypii. A The germcell swells during a short phase of polarizedgrowth, B switch to polarized growth produces first one germtube and then, C, another on theopposite side of the germ cell. D The hyphae continues to grow and begins to branch laterally.E When the mycelium matures the tips starts to split dichotomously and growth speedincreases. F The septa defines the compartments that will enclose the spores and producesasci. Spores are set free and can start a new cycle.D5.1.4 Candida albicansThe yeast Candida albicans is one of the most important human fungal pathogens. It occursas a relatively harmless commensal organism in the nasal, digestive and vaginal tract ofmammals but a small set-back in the protective immune system can allow C. albicans tocause vaginitis and urinary tract infections. As an opportunistic fungus it can invade tissue,spread and cause life-threatening systemic infections when the immune system is severelycompromised (Odds 1988). Its ability to form highly resistant biofilms on implants likecatheters and heart valves propose difficulties to rid the pathogen permanently from the host10

(Chandra et al. 2001). C. albicans is termed a dimorphic fungus because it proliferates eitherin a yeast form or a hyphal form (Fig. 3) (Berman and Sudbery 2002). The invasiveness andvirulence of C. albicans is dependent on its ability to switch between the two growth modes(Lo et al. 1997). Addition of serum to the growth media and elevated temperature (37 C) willinduce production of hyphae which consist of continuous elongated uni-nucleated cellsseparated by septa. Formation of hyphae is followed by coexpression of other virulencefactors such as degrading enzymes and adherence factors in order to enhance the overallvirulence of C. albicans (Naglik et al. 2003). Further, hyphal growth is a response to nutrientdeprivation, especially low nitrogen, and filamentous growth enables the fungus to forage fornutrients more effectively. However, C. albicans can also form opaque cells required formating, pseudohyphal cells like those formed in S. cerevisiae (Fig. 3), and chlamydospores,all are distinct cell types that form in response to genetic or environmental conditions(Whiteway & Oberholzer 2004). Biofilms of C. albicans consist of layers of yeast and hyphalcells embedded within a polysaccharide matrix providing resistance to antifungal drugs(Baillie and Douglas 2000; Chandra et al. 2001).ABCFigure 3. A. C. albicans budding yeast cells, Bpseudo-hypha and C hyphal filaments.C. albicans was initially classified as asexual because no direct observation of mating ormeiosis had been reported. Sequencing of the genome revealed mating related genes and theexistens of a and α mating types (Hull and Johnson 1999). The yeast cells are always diploidcontaining both a and α mating type loci and only if one set of mating type locus is removed,by deletion or induced chromosome loss, the cells are able to mate with the opposite sex (Hullet al. 2000). Mating diploids produce tetraploid cells, following mitosis and chromosome lossthe cells return to a diploid state. This parasexual cycle of C. albicans involves white-opaque11

switching, where the white cells is the default yeast state and opaque cells is the matingcompetentform of C. albicans (Miller and Johnson 2002).C. albicans uses a non-canonical genetic code that translates CUG into a serine instead of aleucine. However, the translation of this codon is ambiguous and C. albicans allowsincorporation of leucine at this position, dramatically increasing the number of differentproteins encoded by the C. albicans genome (Ohama et al. 1993; Santos and Tuite 1995;Gomes et al. 2007).C. albicans is distantly related to S. cerevisiae, with approx 90 % of their genes in common,and gene products and pathways are sometimes used in different ways. Many of the genes inC. albicans that are lacking in S. cerevisiae are coding for secreted hydrolytic enzymes andadhesins implicated to be virulence factors (Naglik et al. 2003). C. albicans has a rather largegenome, fourteen million base pairs with 6200 protein-coding genes on eight chromosomes.The genome of C. albicans is highly dynamic and the occurrence of numeral chromosomalrearrangements and high heterozygosity increases the diversity and dynamics of theC. albicans populations. Due to C. albicans diploid nature and lack of a complete sexualcycle, gene disruption mutants must be constructed through knockout of both alleles. This,together with an unstable genome, can make it difficult to construct a homozygous mutant andto determine whether a gene is essential (Enloe et al. 2000).5.2 Penelope Research Training Network and foci of this thesisPenelope is a Research Training Network within Marie Curie Programs with an overall goalto gain understanding of the evolution of protein interaction networks in eukaryotes. The maintopic of Penelope is focused on the interplay between SH3 domains and their binding partnersin four different ascomycetous yeasts, Saccharomyces cerevisiae, Candida albicans, Ashbyagossypii and Schizosaccharomyces pombe. Penelope has aided in the funding of this studywith the aim on SH3 domain proteins in C. albicans and proteins involved in the endocytosispathway in A. gossypii. Localization of a set of SH3 domain-containing proteins fromC. albicans was investigated for comparison with the known homologs of S. cerevisiae(Tab. 1). Analysis of these genes had previously not been done in C. albicans which is whythey were selected for localization studies. The yeast two-hybrid method was used to explorethe physical interactions of the total set of SH3 domains in C. albicans against Vrp1, homologto human WIP. Also, association between Vrp1 and the different domains of the WASP-like12

protein Wal1 was assayed with the two-hybrid system. Wal1 and Vrp1 play a fundamentalrole in the endocytosis process and actin cytoskeleton organization and interactions betweenthese and SH3 domain proteins in S. cerevisiae are known to exist. Further, the functional roleof SAC6 in A. gossypii was explored in this study. The SAC6 gene encodes the yeast fimbrinhomolog that enforces and stabilizes the actin cytoskeleton. SAC6 in S. cerevisiae is syntheticlethal with SLA2, a gene that is absent from the A. gossypii genome. Sla2 is a cortical actinpatch component with essential functions for endocytosis in other fungi. The last part of thisthesis investigates the function of eisosomes in A. gossypii. These fungal specific structureswere first described in S. cerevisiae and are implied to be involved in endocytosis (Walther etal. 2006).Table 1. SH3 containing genes of C. albicans to be analysed in this studygenesystematic# of SH3namenamedomainsfunction in S. cerevisiaeBBC1orf19.27911involved in assembly of actinpatchesinvolved in polar growth and bud-siteBUD14orf19.35551selectionCYK3orf19.62401involved in cytokinesisinvolved in actin patch assembly andPIN3orf19.59561actin polymerisationNBP2orf19.65881involved in the HOG pathwayRVS167-2orf19.47421NOHBYtransmembrane osmosensor,SHO1orf19.47721participates in the HOG pathwaySLA1orf19.14743required for assembly of the corticalactin cytoskeletonlocalization inS. cerevisiaecortical actin patchesIncipident bud sites,bud neck and sitesof polarised growthbud necknucleus andcytoplasmnucleus andcytoplasmplasma membranenucleus and corticalactin patchesInformation about S. cerevisiae genes was derived from the Saccharomyces genomedatabase, www.yeastgenome.org (September 2010).5.3 EndocytosisAll heterotrophic cells need a constant supply of precursor material and energy rich moleculesto keep up an active metabolism. Water and dissolved salts and smaller particles can readilypass the cell membrane through pores or via transporters. Molecules that are too large todiffuse through the cell membrane need to be internalized by an active process calledendocytosis (from greek endo: within, and cyto; cell). There are several types of endocytosisbut they all function to recycle plasma membrane components and regulate cell surface13

expression of signalling receptors. It is a complex process that involves a large number ofproteins and requires the timing and regulation of many successive events (Kaksonen et al.2003). The cell wall invaginates and is pushed into the cell to finally be pinched off as a smallvesicle. These vesicles fuse to form endosomes that are transported to other cellcompartments like lysosomes, vacuoles and the Golgi network. The reverse process when avesicle from inside the cell fuses with the cell membrane and releases its content on theoutside is called exocytosis. In this way the cell deposits new cell wall material andincorporates receptors and other cell wall components. The cell also excretes waste products,extra cellular material and signalling molecules. Exocytosis is critical for fast growth offungal hyphae and the rate of delivery of exocytic vesicles greatly influences the rate ofgrowth. Cells may ingest dissolved molecules from its environment and membrane-boundreceptors via constitutive or ligand-induced endocytosis. Fluid-phase endocytosis is a nonspecificand constitutive process where the cell engulfs small portions of its environment andcell membrane and anything that is attached to it. Often, however, the cell needs to selectivelyingest a molecule and therefore produces specific cell surface receptors for each targetmolecule, or ligand. When the ligand has bound its receptor, a signal is conveyed to the insideof the cell and the membrane with receptor and target molecule is invaginated andinternalized. Light microscopy and electron microscopy have been the tools to study thesedynamic events, providing temporal and spatial resolution (Kaksonen et al. 2003; Idrissi et al.2008). The actin cytoskeleton plays a fundamental role in both endo- and exocytotic events,providing rigidity, mechanical forces and the tracks for transport of vesicles. This thesis willexplore the dynamics and organization of the actin cytoskeleton during endocytosis withemphasis on the similarities and differences between the two fungi A. gossypii andC. albicans and the well known model yeast S. cerevisiae.5.4 Protein-protein interactionsIn order for a cell to survive, propagate and to execute its function it needs to respond tochanges in its environment, to sense developmental cues and environmental stresses. Thesignal needs to be relayed from the source of input to a cell function, often the regulation of agene. Signal transduction is propagated through close interactions of proteins, inducingmodifications of a protein or enzyme to change its conformation and binding capacity or toactivate or inactivate it. Most processes in the cell involve scaffolds and protein-proteininteractions; complexes of proteins that work in concert or functions as inhibitors oractivators. Several binding motifs are known to promote these interactions, for example14

Src Homology 3 domains, SH3, that connect to other SH3 domains, associating differentproteins. A scaffold is a protein whose major function is to serve as a structural platformwhich recruits and connects multiple members of a signalling pathway, bringing them closetogether and orienting them in a preferred way.5.4.1 SH3 domainsThe SH3 domain is a well-characterized family of protein interaction modules involved in avariety of biological processes. SH3 domains are conserved from yeast to man. The humangenome contains nearly 300 different SH3 domains dispersed in about 200 proteins, someproteins containing up to six individual SH3 domains (Kärkkäinen et al. 2006). The genomesof C. albicans and S. cerevisiae encode a total of 29 and 28 SH3 domains respectively in 24genes. The domains are 50–70 amino acids long and often present in eukaryotic signaltransduction and cytoskeletal proteins. SH3 domains lack enzymatic activity but bind withmoderate affinity to proline rich motifs, the core consensus motif is PxxP. The basic fold ofthe SH3 domain is formed from five or six β-strands arranged in two anti-parallel sheetsproducing a characteristic β-barrel (Fig. 4). The strands are connected by short loops orhelices that together with a conserved hydrophobic binding site define the ligand-bindingproperties of the motif (Yu et al. 1992). Weak SH3 interactions are functionally important andcontacts between the SH3 loops and residues outside the PxxP motif of the target ligand cangreatly enhance specificity and affinity of binding (Mayer and Saksela 2005).Figure 4. Ribbon diagram of the SH3 domainof chicken alpha spectrin. A β-barrel isformed by five anti-parallel strandsconnected by loops. Made with MOLMOL.5.5 The actin cytoskeletonMany fundamental processes of eukaryotic cells, such as cell motility, organelle movement,cytokinesis and endocytosis, require reorganization and polarization of the cytoskeleton.15

Assembly and disassembly of microtubule and actin filaments is required to orchestrate thesechanges. In S. cerevisiae a large number of genes have been identified encoding proteins thatare involved in actin polymerization and an overview of the budding yeast functions will bedescribed here.5.5.1 Actin polymerizationThe actin cytoskeleton is one of the most dynamic and complex systems in eukaryotic cellsand rapid actin assembly and turnover are required for diverse cellular processes (Tang et al.2000). Monomeric actin units polymerize into chains of filamentous actin which in turn buildsup actin patches, cables and rings (Fig. 5). Actin patches are spots that localize in the cellcortex and whose distribution is polarized towards sites of growth and endocytosis. The actincables consist of thick bundled actin filaments attaching at their ends to the cortical patchesand rings localize to bud necks and septal sites (Adams and Pringle 1984; Amberg 1998).Initiation of polymerization requires nucleation factors such as the Arp2/3 complex and theformins Bni1 and Bnr1. Actin monomers can be added to a growing filament in the slowgrowing, pointed end and in the fast growing, barbed end. Polymerization is accompanied bycapping of the newly formed end in order to regulate the length of the filament. Actin patchesare nucleated by Arp2/3 complex and the barbed end of the growing filaments is capped bythe diheteromer Cap1/Cap2 in yeast. The tight capping of this complex stimulates nucleationof new actin branches and the formation of short filaments, producing an intensely branchedactin network in a small patch (Kim et al. 2004). The formins are leaky cappers of the barbedend that allows for elongation at the same time as it protects the end from tight cappers. Thiswill increase the duration of elongation and promote formation of long unbranched actincables (Zigmond et al. 2003).AB CFigure 5. Actin structures visualised with rhodamine staining. A Actin patches andcables in S. cerevisiae. B Actin patches clustered in the hyphal tips of A. gossypiiand actin cables extending from the tip. C Acting rings at septal sites andnonpolarized patches in A. gossypii hyphae.16

5.5.2 The Arp2/3 complexArp2/3 complex is an actin filament nucleation machine consisting of seven proteins, highlyconserved from yeast to mammals (Fig. 6). Two of the proteins, Arp2 and Arp3, are actinrelatedand build up the two first blocks of the new filament (Machesky et al. 1994). TheArp2/3 complex itself has a low intrinsic actin nucleation activity and need nucleationpromotingfactors, NPF’s (Winter et al. 1999). One of the strongest and best characterizednucleation promoting factors of Arp2/3 complex in yeast is Las17, the Wiscott-AldrichSyndrome (WASP) homolog. The C-terminus of Las17 contains an acidic domain that bindsArp2/3 complex and a verprolin homology (VH) domain that brings the first actin monomertogether with Arp2/3 complex. Actin filaments themselves act as promoters of actinnucleation as the Arp2/3 complex often associates to the sides of pre-existing actin filaments(Machesky and Insall 1998).Arp2Arp3actinFigure 6. Arp2/3 complex nucleating a branch on an actinfilament. Arp2 and Arp3 (light blue and green) initiates thenew filament, the actin monomers (dark blue) are broughtby the NPF’s and are added to the barbed end(Boczkowska et al. 2008).Other NPFs include the type I myosins, also strong activators, and the weaker Abp1 and Pan1,all of which bind Arp2/3 complex via an acidic domain similar to that of Las17 (Lee et al.2000) Abp1 contains two acidic domains and does not nucleate Arp2/3 complex by providingactin monomers like the other NPFs do. Instead Abp1 associates tightly with Arp2/3 complexand attach it to the side of a preexisting actin filament (Goode et al. 2001). Having only weaknucleation-promoting activity, Abp1 has a suggested role as an antagonist to other NPFs,17

serving as a competition mechanism that helps balance out NPF activities in space and time(D'Agostino and Goode 2005).5.6 Polarized growthThe yeast cells of S. cerevisiae and C. albicans grow isotropically to increase in size. At thisstage the actin patches are randomly distributed beneath the plasma membrane. In the initialphase of budding in yeast and in germ tube and branch formation in filamentous fungi actinpatches polarizes and aggregates at a confined site. This defines a growth zone and allows forunidirectional growth for as long as the polarization is sustained. In yeast cells this phase isvery brief and switching back to isotropic growth produces an ellipsoid shape. In hyphal cellsgrowth is locked in a polarized mode for longer periods. External signals promote polarizationof the actin cytoskeleton via G-protein coupled receptors and enables cells to move andrespond to their environment.5.6.1 Polarization via Rho-GTPasesProteins of the Rho family belong to the small G-protein (GTP-binding protein) superfamilyand regulate various cell functions through the reorganization of the actin cytoskeleton (Hall1994). These Rho-GTPases, and in particular Cdc42, are essential for guiding the polarizationof actin to sites of growth. The recruitment of Cdc42 to growth sites on the plasma membraneactivates effectors that signal to the actin cytoskeleton (Pruyne and Bretscher 2000). Fig. 7shows the G-protein induced pathways leading to polarization of the actin cytoskeleton inyeast. Like most small GTP-binding proteins, Cdc42 is active in the GTP-bound state andinactive in the GDP-bound state. The cycling between these two states is tightly regulated byGAPs (GTPase-activating proteins) and GEFs (guanine-nucleotide exchange factors) (Zhenget al. 1994). At the start of a new cell cycle a site of bud emergence is established byrecruitment of Cdc42 and actin patches cluster at the incipient budding site. During a briefperiod of polarized growth at this site, actin cables extend from the mother cell into the bud asit grows. Redistribution of Cdc42 over the bud surface disperses the actin patches into the budas it continues to grow isotropically into an ellipsoid shape (Pruyne and Bretscher 2000). Acytokinetic ring of filamentous actin will form at the bud neck, which contracts duringcytokinesis (Field et al. 1999). Then the actin patches appear at the bud neck on both sides,directing cell wall synthesis between the new cells to separate them completely. The mothercell resumes budding immediately but the daughter will grow isotropically and increase insize before entering a new cycle (Pruyne and Bretscher 2000). The establishment,maintenance and termination of cell polarity require a feedback regulation at each stage, to18

Ste5coordinate and reinforce the ordering of events. Cdc42 will be activated by its GEF and aseries of events will lead to its deactivation by its GAP, ending the polarized growth phaseand initiating the switch to isotropic growth. Locking Cdc42 in an active state results inhyperpolarization of cortical actin patches and an elongated bud morphology (Richman andJohnson 2000). Both A. gossypii and C. albicans are dependent of the Rho-like GTPasemodules for establishment and maintenance of polarized hyphal growth (Wendland andPhilippsen 2000).Ras1Cdc42Cdc35Wal1Arp2/3Bni1Ste20cAMPpolarisomeactin patchesactin cablesSte11Ste7Fus3Kss1Protein kinase AendocytosissecretionSte12Sok2Polarised growth/hyphal growth/matingFigure 7. Model of signal transduction pathways in yeast leading to polarized growth.Homologs to the S. cerevisiae proteins are present in A. gossypii and C. albicans.Activation by Ras1 of the Rho-GTPase Cdc42 affects via a MAPK cascade and viaregulation of the actin cytoskeleton. Ste5 acts as a scaffolding protein, tethering kinasesand substrates of the MAPK pathway in close proximity. Ras-GTPase also directlyregulates the cAMP pathway via the adenylate cyclase, Cdc35. Modified picture fromMartin et al. 2005.5.6.2 Hyphal growthThe yeast cells of diploid S. cerevisiae can undergo a morphological transition in response tonitrogen starvation. Cells become elongated and incomplete cytokinesis results in chains ofcells. This pseudohyphal growth allows yeast cells to forage for nutrients as they grow awayfrom the colony and invade the substrate (Gimeno et al. 1992). The morphology is madepossible by sustaining polarized growth for a longer time than in normally budding cells.A. gossypii represents a very different life style in that it is an obligatory filamentous fungus,19

sustaining polarized growth for most of its life cycle. Maintained polarized growth leads tothe production of long tubular cells that can extend considerable distances enabling the fungusto explore and penetrate its environment. Hyphal cells are not separated but septa define thecompartments of the hyphae. The cytoplasm is connected along the filaments through pores inthe septa allowing for transport and distribution of nutrients and organelles within themycelium (Alberti-Segui et al. 2001).C. albicans cells normally bud as a yeast but is able to form both pseudohyphae and truehyphae in response to its environment. Hyphal growth is even a requirement for thepathogenicity of the organism. The morphogenetic switch in C. albicans is of special interestwhen trying to unlock the key events that confer polarized and non-polarized growth.Budding and pseudohyphal formation in both C. albicans and S. cerevisiae is distinct from thehyphal phase of C. albicans why the filamentous A. gossypii may serve as a better model tounderstand the underlying mechanisms of constant polarized tip growth in fungi. Hyphalgrowth requires maintenance of tip directed growth and a constant delivery of vesicles to theextending surface, a process dependent on the actin cytoskeleton and microtubules (Horio andOakley 2005). Filamentous fungi manage to sustain multiple axis of polarized growth and amuch higher growth rate than yeast cells. The rate-limiting steps of hyphal elongation inmature mycelia are the transport of vesicles to the tip and the subsequent incorporation of newmaterial into the expanding membrane and cell wall. At constant growth conditions the rate ofsupply of exocytic vesicles to the apex will exceed their consumption. The build-up of excessvesicles will eventually trigger a branch formation generating two new hyphal tips withdifferent axis of growth (Trinci 1974; Watters and Griffiths 2001).Endocytosis is also required for fast hyphal growth suggesting that a constant flow of new cellwall and plasma membrane material from recycled membrane components is important.When secretory vesicles fuse with the plasma membrane, excess membrane and componentsneed to be recycled by endocytosis. In this way endocytosis and exocytosis are tightly coupledvia early endosomes (Wedlich-Söldner et al. 2000).5.6.3 The SpitzenkörperThe Spitzenkörper was first described as a phase-dark structure located in tips of growinghyphae of fungi (Fig. 8). Spitzenkörper (“apical body”) is a zone at the apex which isdominated by vesicles and nearly devoid of other cell components (Grove and Bracker 1970).20

This complex structure is associated with the direction of tip growth. Vesicles produced in thesubapical region become concentrated in the apex where they are incorporated at theexpanding surface. The Spitzenkörper disappears when hyphal extension ceases and reappearsas tip growth resumes (Grove and Bracker 1970).A complex called the polarisome, is required for apical actin organization and is only presentat actively growing sites. The complex comprises Spa2p, Pea2p, Bud6 and the formin Bni1pand is localized to the very tip of buds and hyphae as long as polarized growth is sustained.Bni1p plays a central role in linking polarisome components to RhoGTPases such as Cdc42and Rho1, to promote actin filament assembly in response to G-protein signaling (Kohno etal. 1996). Components of the polarisome were found to localize to the Spitzenkörper offilamentous fungi (Sharpless and Harris 2002) suggesting that the polarisome is a componentof the Spitzenkörper. Microtubular cytoskeleton plays a major role in the formation andpositioning of the Spitzenkörper by providing the tracks for supplying vesicles. The findingthat actin localizes to the Spitzenkörper suggests that vesicles might switch from microtubulebasedto actin filament-based transport within this structure (Harris et al. 2005). Themicrotubules would primarily be responsible for the long-distance transport of secretoryvesicles to the Spitzenkörper, while actin filaments control vesicle transport to the plasmamembrane.Figure 8. Electron micrograph of ahyphal tip of Aspergillus niger showinga dense zone of microvesicles formingthe Spitzenkörper. Larger vesicles (V)surrounds the area. Golgi (G) andmitochondria (M) are visible. Scale baris 1 µm (Grove and Bracker 1970).5.7 The endocytic machineryMany genes are known to be involved in the processes that orchestrate the endocytic events.An outline of some of the factors in S. cerevisiae will be shown here, focusing on the clathrinmediatedendocytosis pathway with some of the key components further elaborated.Homologous genes to most of these factors exist in A. gossypii and C. albicans and pathwaysare expected to function similarly. Naturally, these organisms have evolved divergently and21

found new ways of coping with their environment and both similarities and differences aretargets for investigations in this study.5.7.1 Clathrin-dependent endocytosisLigand-induced and fluid-phase endocytosis are also called clathrin-dependent endocytosisbecause the invaginating vesicle is coated and stabilized with a lattice of clathin required forproper recruitment of later endocytic proteins (Fig. 9) (Kaksonen et al. 2005). In both cases,ligand-activated or unbound receptors set off the initial steps of the internalization process atthe plasma membrane. Several early endocytosis factors assemble at the site in a fashionindependent of the actin polymerization machinery and initiate the invagination by bendingthe cell membrane inwards. Proteins are recruited and disassembled in a sequential mannerduring the whole process of internalising the forming vesicle, first in a slow-movement phaserepresenting the invagination, but later in a faster moving phase coinciding with the releaseand inward transportation of the endocytic vesicle. Actin plays a fundamental part in thismovement as its polymerization generates the mechanical force for pushing the membrane.The vesicles are transported into the cell where they fuse with endosomes and are thentargeted to other compartments. Fig. 10 shows a sketch of some of the key components in theendocytic process.Early (34-240s)Clathrin (60-120s)Coat (20-40s)Wasp/Myo (10-40s)Amphiphysin (10s)Actin (10-15s)Variable PhaseRegular phaseFigure 9. Temporal organization at the site of clathrin mediated endocytosis inS. cerevisiae. Clathrin and early endocytosis factors assemble at the incipient site ofinvagination. These proteins initiate bending of the membrane and aid the recruitmentof other coat proteins. At the arrival of actin and actin binding proteins the vesicleelongates slowly inwards the cell and is finally pinched off by amphiphysin-likeproteins. The coat module disassemble and the vesicle is transported into the cell.Modified picture from Stimpson et al. 2009.As a first step of forming the endocytic vesicle, clathrin assembles at the endocytic sitebeneath the plasma membrane. The clathrin lattice is made up of triskelions of heavy and lightchains with ability to self assemble into a cage like structure (Ungewickell and Branton 1981).22

The lattice confers an initial curvature to the membrane and also provides anchorage for theendocytic machinery. Clathrin is dependent on Ent1/2 and AP1801/2 for recruitment, adaptorproteins harboring plasma membrane binding ENTH/ANTH modules. These modules arelocated at the N-terminal of clathrin binding proteins involved in signaling and actinregulation. ENTH/ANTH domains are built up by several alpha helices that bind lipids in themembrane and could aid in the bending of the membrane at this initial stage of endocytosis(Stahelin et al. 2003). Las17, a strong promoter of Arp2/3 nucleated actin polymerization, isrecruited early in the endocytosis process, soon followed by other early coat proteins like Sla2and the Pan1-Sla1-End3 complex (Tang et al. 2000). Sla2, homologous to the human Hip1/R,contains an ANTH domain that may help the localization to the plasma membrane, a coiledcoiledcentral domain that binds directly to clathrin and Pan1, and an actin-binding domain(McCann and Craig 1997). Sla1 contains three SH3 domains and is crucial for actin patchregulation and Sla2 localization (Ayscough et al. 1999). Pan1 is another activator of theArp2/3 complex, while Sla1 functions as an inhibitor of Las17 activity (Rodal et al. 2003).Vrp1 and Bzz1 are next recruited to the site. Bzz1 relieves Sla1p inhibition of Las17 so thatactin may start to assemble in patches marking sites of endocytosis (Sun et al. 2006). Vrp1, akey regulator of cortical actin-patch distribution, contains several actin-binding modules andis required for Arp2/3 activation by Las17 and the type I myosins (Anderson et al. 1998). Inthe following stages, actin, Arp2/3 complex and the actin binding protein Abp1 are recruited,coinciding with a slow membrane invagination. The coated pit will extend into a tubularshape, about 50 nm wide and up to 180 nm long. Sla2 is an important factor during thismovement; it couples the vesicle via clathrin to the polymerising actin cables, providing theforce to push the invaginating vesicle inwards the cell (Kaksonen et al. 2003). Myo5, one ofthe type I myosin motor proteins and an NPF of the Arp2/3 complex, arrives and furtherpromotes actin polymerization to drive the vesicle rapidly inward (Sun et al. 2006). Vrp1 isvery rich in prolines and is therefore a likely target for SH3 domain binding. Myo3 and Myo5in S. cerevisiae interact with both Las17 and Vrp1 via their SH3 domain (Evangelista et al.2000). Bbc1, another SH3-protein, functions as a negative regulator of both Myo5 and Las17(Rodal et al. 2003). As the clathrin coated pit internalizes, the amphiphysins Rvs161 andRvs167 are recruited to the neck of the invagination contributing to the release of the formingvesicle (Kaksonen et al. 2005). Sla1, Pan1 and Sla2 disassemble and the vesicle willinternalize in a fast-moving step. While Las17 executes its function mainly during the slowmovement and stays at the plasma membrane Myo5 is the prime nucleator of actinpolymerization during the rapid inward movement and follows the tip of the invagination.23

Myo5 appear at cortical patches immediately proceeding the fast movement of the actinstructures away from the plasma membrane, correlating with the vesicle scission event(Jonsdottir and Li 2004). A pool of Myo5 remains at the membrane base of the invaginationand could aid in the constriction of the neck (Idrissi et al. 2008).clathrinPan1, Sla1, End3Sla2Las17Abp1actinArp2/3 complexVrp1, Bzz1Myo5, Bbc1Sac6Rvs167, Rvs161Figure 10. Components of the endocytic machinery in yeast, SH3 domain proteins are writtenin bold. Clathrin is recruited in the initial phase along with the Pan1 complex, Sla2 and Las17.Arrival of actin and other factors regulating actin polymerization results in a slow invaginatingmovement of the clathrin coated pit. Rvs161/Rvs167 aid in the scission of the vesicle, allowingfor a fast internalization and fusion with early endosomes.Abp1 and the actin bundling protein Sac6 contribute to the dynamics of the actin network inthe later stages and follow the patch as it moves inward. Sac6, or fimbrin, contains two sets ofan actin binding domain, each comprised of two calponin homology, CH, domains. Sac6bundles actin filaments into tight bundles as well as stabilizes them against depolymerization(Goode et al. 1999). Several factors aid in the uncoating of the vesicle after it is pinched off,among them the kinases Ark1 and Prk1, and Abp1. The SH3 domain of Abp1 associates thekinases with the actin patch, enabling phosphorylation, inactivation and disassembly of thePan1-Sla1-End3 complex. This step is pivotal for the recycling and reorganization of actin(Zeng et al. 2001). During internalization the small vesicles become associated with actincables, tight bundles of actin filaments. As the bundles elongates the associated vesicles aremoved further into the cell. It has been shown that early endosomes are attached to actincables and is transported along them towards sites of internalization at the membrane(Toshima et al. 2006). Newly formed endocytic vesicles and early endosomes are in this waydirected towards each other, facilitating their fusion.24

5.7.2 Sla2 and Sac6 are key components in endocytosis of S. cerevisiaeSla2 belongs to a conserved family of actin-binding proteins; the human homologs Hip1 andHip1R bind to huntingtin, the protein whose mutation results in Huntington’s disease. In themodel of S. cerevisiae Sla2 arrives next after clathrin at the endocytic site and provides adirect link between the clathrin coated pit and polymerizing actin filaments (Fig. 11). TheN-terminal ANTH domain binds to the plasma membrane and could together with the clathrinlattice aid in the initial bending of the invaginating plasma membrane. The central domainbinds, in addition to clathrin also Pan1, Rvs167, Sla1 and it self; Sla2 is found as a dimer invivo (Wesp et al. 1997). The C-terminus harbors an F-actin binding talin-like motif that isimportant for endocytosis only at elevated temperatures. This domain could regulate Sla2function by folding itself, occluding actin binding.ScSla2968 aaANTH coiled-coil ABDScSac6642 aaCH1 CH2 CH3 CH4ABD1ABD2Figure 11..Functional domains of yeast Sla2 and Sac6. Sla2 is composed of the lipid bindingANTH domain, a central coiled-coil region that contributes to dimerization and binding to severalother proteins, including clathrin, and a C-terminal actin binding domain. Self regulation of Sla2could be achieved by the ABD folding upon itself, preventing actin binding. The two actin bindingdomains of Sac6 constitute double calponin homology domains.The fimbrin Sac6, member of a family of actin-bundling proteins, is important in the actindependent movement of the endocytic patch and might also have a function in scission of thevesicle. Actin binding is mediated by two pairs of calponin homology domains (Fig. 11),conserved motifs that are found in a set of cytoskeleton organizing proteins. The two actinbind domains associates actin filaments closely together and Sac6 can be found in bothpatches and along cables (Drubin et al. 1988). The S. cerevisiae and A. gossypii Sac6homologs have a high degree of similarity, 80 % identity at the amino acid level. They shareconserved residues in the actin binding sites with the mammalian dystrophin and actinin,members of the same superfamily containing calponin homology domains.5.8 EisosomesEisosomes are newly described large immobile protein assemblies that localize in a punctatepattern beneath the plasma membrane of fungi (Fig. 12). These structures are mainly25

composed of two highly similar proteins, Pil1 and Lsp1, in equimolar proportions. Pil1 andLsp1 have no recognizable domains but colocalize with a plasma membrane protein, Sur7.Eisosomes seem to be static with no exchange with free cytoplasmic pools of Lsp1 or Pil1(Walther et al. 2006). Pil1 and Lsp1 are phosphorylated by Pkh1 and Pkh2 in S. cerevisiaeand this phosphorylation is critical for eisosome organization. In addition, Ypk1 and Ypk2kinases, involved in sphingolipid-mediated signalling, are necessary for maintainingeisosomes. Eisosomes are subsequently dynamic structures whose formation and turnover areregulated by the sphingolipid-Pkh1/2-Ypk1/2 signalling pathway (Luo et al 2008). Pil1expression in S. cerevisiae is regulated by the cell cycle and determines size and number ofeisosomes (Moreira et al. 2009). Deletion of PIL1 leads to clustering of eisosome remnants,and cause a reduction of the endocytic rate (Walther et al. 2006). The discovery of eisosomeslinked them to endocytosis for several reasons; deletion of PIL1 results in impairedendocytosis, eisosome components show synthetic lethality and interaction with severalproteins known to function in endocytosis, and the colocalization of eisosomes with earlyendocytic endosomes and a subpopulation of actin patches (Walther et al. 2006). Lsp1 is notrequired for localization of Pil1 and despite their similarity in sequence they have distinct andnon-redundant functions.PIL1-GFP LSP1-GFP SUR7-GFPFigure 12. Proteins in the eisosome complex localizein a punctate pattern beneath the plasma membrane.Z-stack image from Walter et al, 2006.Sur7 was originally identified as a multicopy suppressor of RVS161 and RVS167 mutants,actin patch components involved in endocytosis (Sivadon et al. 1997). Sur7 itself does notlocalize to actin patches but is involved in sporulation and sphingolipid content of the plasmamembrane (Young et al. 2002).5.8.1 Lipid raftsLipid rafts are detergent-resistant assemblies laterally distributed in the plasma membrane.They are involved in many cellular processes such as protein sorting and trafficking. In yeastthey are enriched in ergosterol and sphingolipids and they recruit specific membrane proteinsinto distinct domains (Bagnat et al. 2000). Typically, GPI-anchored proteins and transporters26

associate with these rafts. On of these domains called MCC, membrane domain occupied byCan1, colocalizes with Sur7, linking them to the static eisosome structures. The MCCs arevery stable and immobile; they are not dependent on the cytoskeleton for formation and donot colocalize with sites of active endocytosis (Malinska et al. 2004); Walther et al. 2006;(Grossmann et al. 2008). These structures are not turned over by endocytosis but couldinstead act as protective areas. As long as a protein is associated with the MCC the turnoverby endocytosis is avoided (Grossmann et al. 2007). Deletion of the eisosome component Pil1leads to clustering of both Sur7 and Can1 in a manner similar to Lsp1. Disruption of the corecomponents of MCCs, including Pil1, causes disassociation and faster degradation of markerproteins in the complex (Grossmann et al. 2008).27

6 Methods6.1 Strains and mediaPlasmids, strains and primers used in this study are listed in Appendices III-V.Saccharomyces cerevisiae BY4741 (his3Δ1; leu2Δ0; met15Δ0; ura3Δ0) was used for in vivorecombination and generation of GFP fusions. For the yeast two-hybrid assay the followingstrains were used: PJ69-4a: trp1-901; leu2-3,112; ura3-52; his3-200; gal4Δ; gal80Δ;lys2::GAL1-HIS3; GAL2p-ADE2; met2::GAL7-lacZ and PJ69-4α: trp1-901; leu2-3,112;ura3-52; his3-200; gal4Δ; gal80Δ; lys2::GAL1-HIS3; GAL2p-ADE2; met2::GAL7-lacZ.Candida albicans and Saccharomyces cerevisiae were grown in rich media (YPD; 1 % yeastextract, 2 % peptone, 2 % dextrose) or minimal media (6.7 g/L YNB w/w ammoniumsulphate w/o amino acids, 0.69 g/L CSM, 20 g/L glucose) in 30 C. Hyphal formation ofCandida albicans was induced in YPD with 10 % newborn calf serum in 37 C for 2-4 hours.Ashbya gossypii was grown in rich media (AFM; 1 % yeast extract, 1 % peptone, 2 %dextrose and sporulated in minimal media (1.7 g/L YNB w/o ammonium sulphate w/o aminoacids, 0.69 g/L CSM, 20 g/L glucose, 2 g/L Asparagin, 1 g/L Myo-Inositol). Efficientsporulation of Ashbya gossypii was achieved by inoculating 5 ml fresh mycelia from an AFMculture into 200 ml sporulation media. After 3-7 days of vigorous shaking in 30 C, sporeswere harvested by gentle centrifugation. The pellet was resuspended in 30 ml TE buffer andincubated with 5 mg zymolase in 37 C until spores were set free and remaining myceliumwas degraded, 1-2 hrs. Spores were washed repeatedly in water with 0.03 % Triton X-100 andfrozen down in 25 % glycerol at -80 C. Antibiotics where used for selection with additions of200 µg/ml G418/geneticin or 100 µg/ml clonNAT/nourseothricin to rich or minimal media.Escherichia coli strain DH5α was used for propagation of all plasmids generated andelectroporation was used for transformation.6.2 TransformationS. cerevisiae was transformed with the lithium acetate procedure (Gietz and Schiestl 2007).Transformation of C. albicans was preformed with a modified lithium acetate protocol(Walther and Wendland 2003) with o/n incubation at 30 C in PEG4000 followed by a 15 minheat shock at 44 C. Transformation of A. gossypii was achieved by using electroporationaccording to (Wendland et al. 2000). After transformation mycelia were plated on AFM platesand incubated for 6 hrs in 30 C to allow for expression of the resistance marker, then the28

mycelia were overlaid with 8 ml 0.5% agarose with 6.6 mg G418 or 3.3 mg clonNAT (finalconc. of 200 µg/ml resp. 100 µg/ml).6.3 Generation of GFP tagged C. albicans strainsThe Candida albicans SN148 strain (arg4; leu2; his1; ura3, Noble and Johnson 2005) wasused as the progenitor strain for all GFP fusions of specific genes in this study. First,heterozygous strains were generated were the complete ORF of one gene allele was deleted.Then, GFP was fused to the 3’ end of the second allel. To this end, specifically designedcassettes were constructed with the 3’ end of each gene cloned separately and later fused withGFP via in vivo recombination in yeast. The cassettes were constructed with long homologousflanks to C. albicans sequences for proper targeting. The procedure is shown in Fig. 13 andeliminates expression of any wild type copies of the gene. Eight genes of C. albicans wereGFP tagged following this procedure, BBC1, BUD14, CYK3, NBP2, PIN3, RVS167-2, SHO1and SLA1 (see Tab. 1).AS1G2CdHIS1G3S2G1ORFG4BORFA1A4~1kbC3’S1-GFPGFPCmLEU2S2D3’GFPCmLEU2ORFFigure 13. Generation of GFP-tagged C. albicans strains. A A deletion cassette with 100 bpflanking regions will exchange the complete ORF of one allele of the target gene with theCdHIS1 marker, generating a heterozygous strain for that locus. B A 1 kb fragment coveringthe 3’ end of the gene was cloned in pRS417. C A GFP cassette was amplified by PCR usingprimers with homologous flanks to sequences in the target gene in C. albicans. Cotransformationof S. cerevisiae BY4741 with the pRS417-3’-end plasmid and the GFP cassettegenerates an in frame fusion of the 3’-end of the gene to GFP by in vivo recombination. DTransformation of C. albicans heterozygous strains with the fusion cassette tags the remainingallele with GFP.29

Deletion cassettes for the first allele were created using PCR based methods. The markercassette was amplified from pFA-CdHIS1 (Candida dubliniensis HIS1), with S1/S2 primerpair for each gene. The primers contain 100 bases of homologous regions to the target geneUTR flanks, enough to ensure for integration into the target locus. C. albicans SN148 wastransformed with the cassettes and mutant colonies were selected on CSM-plates lackinghistidine, generating heterozygous knockout strains. Integration of the marker into the wrightlocus was verified by PCR of 5’ and 3’ ends with primer pairs G1/G2 and G3/G4,respectively. G1 and G4 binds to sequences outside of the replaced gene in the genome, G2and G3 binds to the HIS1 marker. Two heterozygous strains where produced for every gene.The GFP-fusion cassettes for C-terminal tagging of the corresponding proteins had to beconstructed in several steps. The 3’ end of all genes where cloned and fused to GFP, deletingthe stop codon of the gene, before integration into the genomic locus. A fragment of about500 base pairs up- and downstream of the stop codon was cloned in either of the standardvectors pDrive Cloning Vector, Promega, (BBC1, CYK3, NBP2, RVS167-2) or pGEM-T EasyVector, QIAGEN, (BUD14 ,PIN3, SHO1, SLA1) using gene specific primers A1 and A4. Thefragments were cut out with restriction enzymes, XhoI/BamHI for pDrive fragments andSacII/SalI for pGEM fragments, and re-cloned into pRS417, a plasmid which was made byexchanging the LEU2 marker of pRS415 to GEN3. The pRS plasmid with the cloned 3’ endwas co-transformed into S. cerevisiae BY4741 with a GFP-cassette amplified from pFA-GFP-CmLEU2 (Candida maltosa LEU2) with the primer pairs S1-GFP/S2. The S1-GFP primerwill exclude the stop codon and add 45 bases at the 5’ end of the cassette which is enough forhomologous recombination in yeast. The S2 primers are the same as for making the deletioncassette, adding 100 bases at the 3’ end of the cassette. The GFP is optimized for C. albicanswith S65A, V68L and S72A (Morschhäuser et al. 1998). The efficient homologousrecombination machinery of S. cerevisiae will integrate the GFP cassette immediately afterthe cloned 3’ end of the gene on the plasmid, excluding the stop codon. Yeast clones wereselected on both the plasmid (GEN3) and the inserted GFP-cassette (CmLEU2) on CSMmedia lacking leucine with G418. Plasmids were harvested and further propagated in E. coli.Sequencing of the GFP fusion plasmid with primer 392 ensures correct in-frame fusion of thegene to GFP after homologous recombination. The whole GFP fusion cassette was cut outfrom the plasmid, using the same enzymes as for the initial cloning of the 3’ end into pRS417,providing 500 base pairs flanks for homologous recombination in C. albicans. Transformationof the cassette into the corresponding heterozygous strain tagged the remaining allele with30

GFP, clones where selected on CSM lacking leucine and histidine. Integration of GFP wasconfirmed by PCR using a specific primer set G1-GFP/392 for each gene. G1-GFP bindsupstream the 5’ homologous region, in the ORF, and 392 binds to GFP itself. SLA1-GFP andCYK3-GFP cassettes where also transformed into the progenitor strain SN148, producingstrains with one wild type copy of the gene left in the genome. Before microscopy, o/ncultures where diluted in YPD and grown to exponential phase in 30 C. Hyphal formationwas induced as previously described.6.4 Yeast Two Hybrid assayA yeast two hybrid assay with various fragments of the ORF of WAL1 or VRP1 and a set ofSH3 domains of C. albicans genes was performed as described in Borth et al. 2010 (Fig. 14).Wal1 was truncated at its N- and C-termini and the central proline-rich region was removed.The different truncated versions of WAL1 and the full length gene were fused to the DNAbinding domain, DBD, of the galactose induced transcription factor GAL4 on plasmid pGBT9.VRP1 was divided in two parts, the C-terminal was fused to the activation domain, AD, ofGAL4 on pGAD424 and the N-terminal was cloned in both vectors (Fig. 15A). 24 sets of SH3domains from 23 C. albicans genes (originally isolated by Reijnst, unpublished) where fusedto the AD on pGAD424 (Fig. 14). The SH3 domains were in most cases isolated individuallywith ~100 nucleotides up- and downstream of the domain. The two SH3 domains of ABP1and BEM1 and the two first SH3 domains of SLA1, named SLA1-1, were cloned as onefragment. The third SH3 domain of SLA1 is referred to as SLA1-2.DBDbait prey ADpreyADWal1WH1 BP1 P2 P3 P4VCAWal1ΔN-termP3P4VCAWH2WH2LBDWal1ΔproWal1ΔC-termVrp1 N-termVrp1 N-termVrp1 C-termWH1WH2WH2HOTWH2WH2HOTWH2WH2LBDLBDLBDSH3ABP1SLA1-1SLA1-2BOI1CYK3HOTHSE1BBC1BEM1BEM1LFUS1HOF1PEX13LSB1/2BUD14RVS167-1LSB3SHO1CDC25RVS167-2Q59U90NBP2Q5AAN3MYO5CDC25LFigure 14. Truncated versions of Wal1 and Vrp1 and SH3 domains from C. albicans usedin the two-hybrid assay. Fragments are fused to either the DBD or the AD of thetranscription factor GAL4. WH1: WASP-homology 1; WH2: WASP-homology 2, P1-P2;proline-rich regions; VCA: verprolin-central-acidic domain; LBD: Las17-binding domain;HOT: Hof1-trap.31

The yeast strains used for the two hybrid assay (PJ69-4a/α) are deleted for the GAL4transcription factor involved in galactose metabolism. If the fusion proteins interact with oneanother they form an active transcription factor that induces the expression of several reportergenes (Fig. 15B). S. cerevisiae PJ69-4a or PJ69-4α were transformed with two plasmids, oneexpressing the DBD and one with an AD construct. pGBT9 and pGAD424 harbors theselective markers TRP1 and LEU2, respectively, and yeast transformants were selected onCSM lacking tryptophan and leucine for maintenance of both plasmids. Interactions betweenprotein fragments expressed from the plasmids were analyzed qualitatively and quantitativelyexploiting the reporter genes lacZ and ADE2. lacZ codes for β-galactosidase that can breakdown lactose substitutes like X-gal and ONPG to colourful products that are readily seen ormeasurable by their light absorption. Growth on plates lacking adenine and production of bluecolor on X-gal plates was indicative of an interaction. The conversion of ONPG in aβ-galactosidase assay was measured photometrically.AGAL4DBDMSCT ADH1GAL4ADMSCT ADH1P ADH1pGBT9P ADH1pGAD424ampTRP1ampLEU2BpreybaitADDBDtranscriptionalactivator (Gal4)promoterreporter gene (ADE2/lacZ)Figure 15. A Maps of the two plasmids that were used to clone fragments for the two hybridscreen. pGBT9 contains the DNA binding domain, DBD, which was fused to the bait protein.The prey protein was cloned in pGAD424 which harbours the activation domain, AD. Theplasmids can be selected for by the TRP1 and LEU2 markers. B The two fusion proteinswere expressed in a yeast strain deleted for a transcription factor involved in galactosemetabolism. If bait and prey proteins interact with one another they form an activetranscription factor that can activate the expression of several reporter genes. The reportergene lacZ produces the enzyme ß-galactosidase. The lactose substitutes X-gal and ONPGcan serve as substrates for ß-galactosidase. The enzyme cleaves X-gal into galactose andan indigo blue dye which is readily visible. Colour less ONPG will be degraded to galactoseand yellow ONP and the absorption can be measured at 420nm.32

6.5 β-galactosidase assayThe β-galactosidase assay exploits the enzymatic activity of β-galactosidase, the product ofthe lacZ gene. The enzyme degrades the colourless substrate ONPG into galactose and theyellow ONP (o-nitrophenol) which absorption can be measured at 420 nm. Yeast strainscontaining plasmids with a DBD fusion and an AD fusion were grown o/n in selective mediaat 30 C. 1 ml of the culture was diluted 1:4 in YPD and grown for another 4 hrs. OD 600 wasmeasured and cells from 2 ml culture were harvested. To obtain comparable data in the assayall strains should have reached the same cell density. Cells were washed and resuspended in300 µl Z-buffer (60 mM Na 2 HPO 4 , 40 mM NaH 2 PO 4 , 10 mM KCl, 1 mM MgSO 4 ), 100 µlwas used in duplicate tests. Cells were broken by freeze/thawing and then incubated with150 µl 6 mg/ml ONPG solution at 37 C for 30 minutes. The enzymatic reaction was stoppedwith 400 µl 1 M Na 2 CO 3 and cell residues were spun down. The exact timing of the reactionof all samples is of great importance for the calculated values of the assay. 500 µl ofsupernatant was mixed with 500 µl water and absorption at OD 420 (yellow dye) and OD 550(cell debris) was measured. Enzyme activity, A, in Miller units was calculated with t= 30 min,V= 0.1 ml, according to the formulaA = 1000 * OD 420 -(1.75*OD 550 )t * V * OD 6006.6 Generation of mutant and GFP strains in A. gossypiiDisruption of genes in A. gossypii was achieved either by long-flanking deletion cassettes(Noble and Johnson 2005), PCR generated S1/S2 cassettes (Walther and Wendland 2008) orby insertion of a cloned marker cassette into the ORF. In all cases the kanMX marker derivedfrom the plasmid pFA-kanMX, was used. It contains the kanamycin resistance ORF regulatedby the promoter and terminator of the AgTEF1 gene. Two strains were produced for eachgenotype and all were made in the Agleu2 background, the strain that also acts as wild typecontrol in all assays in this study. In a first step a heterokaryon was generated byelectroporation (1.5 kV, 100, 25 μF in the Equibio EasyjecT electroporator, Wolf lab) producinga mycelium containing both wild type and transformed nuclei. Upon sporulation, germinationunder selective pressure and micromanipulation homokaryotic strains were isolated in caseswhere the deletion was not lethal. An MSM Micromanipulator (Singer Instruments) was usedto pick spores that had produced only a germ cell, larger germlings with germ tubes and a fewlateral branches could be isolated with a Patchman NP2 micromanipulator (Eppendorf). Only33

germlings where the spore remnants and germ cell could be identified were lifted and moved to anisolated spot on the plate, thereby making sure that the mycelia produced will be originated fromone spore and thus homokaryotic.Long flanking deletion cassettes were generated by a two step PCR (Fig. 16) and were used todelete SAC6, LSP1, SUR7, PKH1 and YPK1. 400 base pairs long fragments of the upstreamand downstream flanks of the target gene and the kanMX marker were amplified by PCR. Theprimers for generation of the gene flanks added regions overlapping with the marker cassetteso that the flanking fragments could be fused to the marker in a second PCR, providing verylong regions for homologous recombination and reliable integration of the marker into thetarget locus. Flanks were generated with primer pairs 5’a/5’b-S1 and 3’a-S2/3’b, and themarker cassette with S1/S2 primers in the first PCR step. S1 and S2 regions are theoverlapping parts, 22 bases long. The whole deletion cassette was amplified in the second stepusing primers 5’a and 3’b together with both flanks and the marker. The deletion cassetteswere cloned into a pSK background (XhoI/EcoRI cut C351 for pkh1, XhoI/XbaI cut C424 forsac6 and XhoI/SacI cut C424 for the others). Cassettes were then cut out with the sameenzymes as were used for cloning and transformed into A. gossypii. Integration of the deletioncassette into target locus was confirmed by PCR with G1/G6 and G4/G5 primer pairs for thepkh1 mutant and G1/G2 and G3/G4 primer pairs for the other mutant strains, where G1 andG4 bind outside of the targeted locus and G2/G3/G5/G6 binds in the kanMX marker.Heterokaryotic strains were generated for pkh1 and ypk1 and they would also give a PCRproduct with primers I1/I2 annealing inside the ORF. The homokaryotic null strains of sac6,lsp1 and sur7 did not give an internal PCR product with I1/I2 primers.Atarget gene5’a 5’b-S13’a-S23’bkanMXS1S2B5’aS1S23’bFigure 16. Generation of a long flanking deletion cassette. A 5’ and 3’ flanks of thetarget gene are and a marker are amplified by PCR, producing fragments with regionsthat overlap (S1/S2). B In a second PCR step, the flanks are fused to the marker,generating a deletion cassette with long homologous flanks, ~400 base pairs.34

Disruption of PIL1 was achieved in two different ways and the gene was tagged with GFP, asshown in Fig. 17. PIL1-S1/ PIL1-S2 primers were used to amplify the kanMX marker todelete the PIL1 ORF. Since the promoter region of PIL1 could not be amplified, the S1 primerwas placed in the very beginning of the ORF, thereby leaving the first 55 nucleotides intact inthe genome of the pil1 strain ASJ24. Integration of the deletion cassette in pil1 heterokaryonswas confirmed by PCR using primers G5, annealing in the 3’ end of the marker, and G4,annealing downstream of the integration point. Disruption mutants of PIL1 were alsogenerated by inserting the marker in the middle of the ORF as follows, generating ASJ21. Apart of the PIL1 gene was first cloned in pGEM-T Easy Vector, QIAGEN, using primer pairI1/I2, producing pGEM-AgPIL1-A. The marker as a blunt ended PvuII/EcoRV fragment frompFA-kanMX was ligated into an EcoRV site in the PIL1 ORF of pGEM-AgPIL1-A. Thedisruption cassette was cut out with NotI and transformed into leu2. I1/G2 and G3/I2 primersconfirmed integration of the cassette. A GFP was fused to the 3’ end of PIL1 via in vivorecombination in S. cerevisiae. First the whole ORF of PIL1 was cloned in pGEM, producingpGEM-AgPIL1-B, with primers I3/G4. The fragment was cut out with PmeI, provided withthe I3 primer and NdeI, cleaving at an endogenous site just upstream of G4, and ligated to apRS417 backbone (from 651). The GEN3 marker was exchanged to NAT5 from pFA-NAT5with restriction enzymes BstZ171/PacI, generating pRS418-AgTEF1p-AgPIL1. Both GEN3and NAT5 contain the ScTEF2 promoter and terminator, BstZ171 and PacI cleaves in theseregions.I3EcoRVS1-GFP S2-GFPABS1I1kanMXI2S2G4C5’ PIL1kanMX3’ PIL1DAgTEF1pPIL1 GFPFigure 17. A Annealing sites of primers in PIL1 for generation of ORF disruption and 3’ end GFPfusion. B Almost all of the ORF was replaced with a marker cassette amplified with S1/S2, leaving thefirst 55 nt in the genome. C A part of PIL1 was cloned in pGEM with primers I1/I2 and the kanMXmarker was inserted into the EcoRV site. The whole disruption cassette was cleaved out andtransformed into A. gossypii. D PIL1 was cloned behind an AgTEF1 promoter on a pRS418 plasmidwith I3/G4. A GFP-marker cassette with homologous flanks to the end of PIL1 was amplified withprimer pair S1-GFP/S2-GFP. The PIL1 ORF and GFP were fused in an in vivo recombination event inyeast.35

pRS418-AgTEF1p-AgPIL1was co-transformed into yeast with a GFP cassette amplified frompFA-GFP-kanMX with primers PIL1-S1-GFP/PIL1-S2-GFP, producing the final GFP-taggedgene construct, pRS418-AgTEF1p-AgPIL1-GFP. This plasmid was used for complementationof the pil1 mutant but also transformed into leu2, lsp1, sur7, sac6 and wal1. GFP wasintegrated into the the PIL1 locus of leu2 with an AatII/NotI fragment of pRS418-AgTEF1p-AgPIL1-GFP, cutting inside the ORF of PIL1 and after the GFP cassette.6.7 Plate assays and germination of sporesTo see how the growth of deletion strains was affected, 5 µl drops of mycelia were inoculatedon AFM plates and colony diameters were compared to leu2 after a week at differenttemperatures. Germination efficiency of sac6 was investigated by spreading diluted sporesolutions on AFM plates and counting visible colonies that had formed after tree days. 1 µl ofa sac6 spore solution with OD 600 of 0.1 was spread on plates with 200 µl water. leu2 sporeswere diluted another 100x before plating.6.8 Microscopy and staining proceduresMicroscopy was performed with an Axio-Imager microscope (Zeiss, Jena and Göttingen,Germany) and images processed with Metamorph 7 software (Molecular Devices Corp.,Downington, PA). Fluorescent microscopy was performed with the appropriate filtercombinations for GFP, FM4-64, Alexa488/Rhodamin, Calcofluor White, Luciferase Yellowand DiOC6. Either single images of one plane were taken or a Z-stack generated from20-30 planes. Images were acquired on a MicroMax1024 CCD-camera (PrincetonInstruments, Trenton NJ, USA). O/n cultures of GFP-tagged C. albicans strains where dilutedin YPD and grown to exponential phase in 30 C before microscopy. Hyphal formation wasinduced as previously described. A. gossypii strains were usually taken from an exponentiallygrowing AFM culture. Calcofluor White staining and actin staining with Alexa488-Phallodinand Rhodamine-Phallodin was performed as described (Wendland & Philippsen, 2000).Visualization with FM4-64 and mitochondrial staining with DiOC6 was done according to(Walther and Wendland 2004a). Luciferase Yellow was added to a fresh AFM culture(5-10 mg/ml LY) and samples were taken out at different time points and washed with waterbefore microscopy.36

7 ResultsSeveral of the proteins involved in the conserved pathways of endocytosis containSH3 domains. A set of SH3 domain-containing genes of C. albicans was fused to GFP forlocalization studies. Three C. albicans proteins were successfully tagged and their localizationwas compared to the S. cerevisiae homologs. The isolated SH3 domains of C. albicans werealso tested in a two-hybrid assay for their interaction with Vrp1, a protein very rich in prolinesand therefore a likely target for SH3 domain-binding. Several truncated versions of Wal1, theWASP homolog in C. albicans, were used in the two-hybrid experiment with Vrp1 since thetwo homologs of S. cerevisiae are known to interact. Sla2 is an important endocytosis factorin S. cerevisiae but was found to be lacking in the genome of A. gossypii. To investigate thisfurther, SAC6 of A. gossypii was deleted, a gene that shows synthetic lethality with SLA2 inS. cerevisiae. Eisosomes have been implicated to be involved in endocytosis in S. cerevisiaeand to colocalize to actin. The A. gossypii homologs were deleted in order to investigate theirimportance for endocytosis and localization of eisosomes was compared to the actin patchdistribution.7.1 Generation of heterozygous C. albicans strainsHeterozygous strains for a set of genes encoding SH3 domain proteins were produced bydeleting one of the two alleles by PCR based gene targeting. The CdHIS1 marker wasamplified with a specific primer pair for each gene producing a deletion cassette with longhomology regions to both flanks of the gene, ensuring efficient integration into the targetlocus. Heterozygous C. albicans mutants were verified for eight genes, BBC1, BUD14, CYK3,PIN3, NBP2, RVS167-2, SHO1 and SLA1. Integration of the CdHIS1 marker into the rightlocus of all strains was confirmed at both flanks by PCR (App. I). Fusing of GFP to thesecond allele will eliminate expression of any wild type copies of the gene. To this end GFPfusion constructs on a pRS417 plasmid were first generated for all genes. A fragment withthe last 4-500 bases of each ORF and 4-500 bases of the downstream region was isolated andfused to a GFP cassette via in vivo recombination in S. cerevisiae. Primers for the GFP fusioncassette were designed to exclude the stop codon of the target gene and introduce the GFPORF in the same reading frame. This strategy allows for sequencing of the fusion point, andconfirmation that the GFP has integrated in the correct reading frame on the plasmid. Thefusion cassette was cut out of the plasmid and transformed into the respective heterozygous37

strain tagging the remaining allele with GFP. Genomic integration of the GFP downstream ofthe target gene was confirmed by PCR in all cases (App. I).7.2 Localization of C. albicans SH3 domain proteinsOut of eight C-terminally GFP-tagged SH3-domain proteins three produced a convincing GFPsignal in C. albicans cells, namely Cyk3, Sla1 and Bbc1. In all cases the localizations of thefusion proteins were corresponding to the localization of their homologs in S. cerevisiae(Fig. 18) (Korinek et al. 2000; Mochida et al. 2002; Warren et al. 2002). Sla1-GFP andBbc1-GFP can be seen in cortical patches in both yeast and hyphal cells. Cyk3-GFP shows upat septal sites, in yeast cells with large buds and in the distal septa of hyphae. Cyk3 inS. cerevisiae localizes to the bud neck only in the short cytokinesis phase and this wouldexplain why the GFP-signal of CaCyk3 could only be seen in a small population of the cells(~30 % of large budded cells, n=82).ASLA1-GFP BBC1-GFP CYK3-GFP30 C37 CBproteinSla1Bbc1Cyk3S. cerevisiaenucleus, corticalactin patchescortical actinpatchesbud neckC. albicanscortical actinpatchescortical actinpatchesbud neck, septaFigure 18. Localization of three SH3 domain proteins in C. albicans. A GFPsignals in yeast and hyphal cells of C. albicans. Sla1-GFP and Bbc1-GFPlocalizes in cortical patches in both daughter and mother cells. Cyk3-GFPlocalizes to bud necks of yeast cells with a large bud and in the most distalsepta of hyphae. Scale bar is 5 µm. B Localizations of homologs in C. albicansand S. cerevisiae.38

7.3 Yeast two hybrid assayIn S. cerevisiae Vrp1 interacts with Las17 and the SH3 domain of the type I myosin Myo5. Tosee whether the same complex also exists among the homologs in C. albicans, Vrp1-Wal1-Myo5, a yeast two-hybrid analysis was performed (Borth et al 2010) (Fig. 19). SH3 domainshave a preference for proline rich sequences and Vrp1 which is very rich in prolines wasdivided in two parts. Both the C-terminal and N-terminal parts were fused to the GAL4 DBDon plasmid pGBT9. The C-terminus of Vrp1 was also fused to the AD on pGAD424 to beassayed against its N-terminus. Full length and truncated versions of Wal1 were fused to theDBD and interactions against the C-terminus of Vrp1 were investigated. A β-galactosidaseassay exploiting the reporter gene lacZ showed that the WH1 domain of CaWal1 interactswith the C-terminal of CaVrp1 and a stronger interaction was seen when the central prolinerich region of Wal1 was removed. The central proline rich and C-terminal regions of Wal1and the N-terminal of Vrp1 do not show any interaction with Vrp1-C-terminal. The resultsfrom the ONPG test agreed with the spot assays; the strongest interaction between Wal1 without its proline rich region and Vrp1-C-terminus produced the bluest spot on X-gal media.BD-pGBT9AD-pGAD424-ade-trp-leuX-galβ-galactosidase assay [Millerunits]WH1 BP1 P2 P3 P4VCA31P3 P4 VCAHOT24WH2LBD115WH120WH2WH2HOTLBD15LBD21WH2WH2HOTSH3WH2 WH2HOT LBD8021Wal1 WH1 BP1 P2 P3 P4 VCA Vrp1 WH2WH2HOTLBD SH3-Myo5SH3Figure 19. Two hybrid analyses of fragments of Wal1, Vrp1 and Myo5, visualised qualitatively onplates and quantitatively in an ONPG assay. Fragments are fused to the binding domain (BD) or theactivation domain (AD) of Gal4 on plasmids pGBT9 and pGAD424. When fragments interact the ADand BD forms an active Gal4 and reporter genes ADE2 and lacZ becomes active. Growth on –adeplates (left row of spot assay) indicates that the reporter gene ADE2 is active, colonies turn red on lowadenine media (middle row) when ADE2 is off. Blue colour (right row) and Miller units in the ONPGassay is a result of an active lacZ reporter gene. Negative control constitutes the empty plasmidspGBT9/pGAD424.39

posnegCaABP1CaSLA1-1CaSLA1-2CaBOI1CaCYK3CaHSE1CaBBC1CaBEM1CaBEM1LCaFUS1CaHOF1CaPEX13CaPIN3CaBUD14CaRVS167-1CaLSB3CaSHO1CaCDC25CaRVS167-2CaQ59U90CaNBP2CaQ5AAN3CaMYO5CaCDC25LMiller unitsposnegCaABP1CaSLA1-1CaSLA1-2CaBOI2CaCYK3CaHSE1CaBBC1CaBEM1CaBEM1LCaFUS1CaHOF1CaPEX13CaPIN3CaBUD14CaRVS167-1CaLSB3CaSHO1CaCDC25CaRVS167-2CaQ59U90CaNBP2CaQ5AAN3CaMYO5CaCDC25LMiller unitsAlso, interactions resulting in higher Miller units than the negative control grew on medialacking adenine, showing that the reporter gene ADE2 was activated. Red colonies on lowadenine media (CSM-Trp-Leu) indicate lack of interactions between the expressed fragments,as shown for Vrp1-C-terminal against its N-terminal and the proline rich and C-terminus ofWal1.The SH3 domains of C. albicans were isolated and cloned in the pGAD424 plasmid, assingle domains or, when two SH3 domains resided closely together, as double domains.Interactions of the C- and N-termini of CaVrp1 were assayed against all the SH3 domains, butonly the SH3 domain of Myo5 would test positive (Fig. 20 and 21).A100CaVRP1 N-terminus806040200B100CaVRP1 C-terminus806040200Figure 20. ONPG-assay with yeast two hybrid strains. The DBD of GAL4 was fused to the CaVrp1N- or C-terminus, A and B. A single or double SH3 domain of each gene was cloned to the AD ofGAL4 as described. The dotted line corresponds to the negative control, a strain containing emptyplasmids The value of the positive control, containing pGBT-ScBUD3 and pGAD-AgCDC3 is veryhigh (700-1100) and cut off in the graph.Growth on plates lacking adenine and a high β-galactosidase activity indicated a stronginteraction of Myo5-SH3 against the C-terminal part of Vrp1. Two hybrid strains with Myo5-SH3 and the N-terminus of Vrp1 showed some growth on plates lacking adenine but a verypoor result in the β-galactosidase assay. Together, these results show that Vrp1 binds directly40

to both Wal1 and Myo5 in vivo suggesting a complex similar to the Vrp1-Las17-Myo5complex in S. cerevisiae.AVrp1-NtVrp1-Ct1-12 13-24 1-12 13-241 2 3 45 6 7 89 10 11 12+ -B 1 ABP1 13 PIN32 SLA1-1 14 BUD143 SLA1-2 15 RVS167-14 BOI1 16 LSB35 CYK3 17 SHO16 HSE1 18 CDC257 BBC1 19 RVS167-28 BEM1 20 Q59U909 BEM1L 21 NBP210 FUS1 22 Q5AAN311 HOF1 23 MYO512 PEX13 24 CDC25L+ - + - + -Figure 21. A yeast two hybrid screen of all SH3 domainsfrom C. albicans against CaVrp1 N-terminus andC-terminus. SH3 domains were fused to the AD of Gal4 onpGAD424 and Vrp1 fragments to the DBD of Gal4 onpGBD9. Physical interaction of AD and DBD in the yeaststrains will produce the active transcription factor Gal4 andinduce expression of reporter genes, ADE2 and lacZ.Positive control is a strain with pGBT-ScBUD3 and pGAD-AgCDC3, the negative control strain contains emptyplasmids. A Strains with the SH3 domain of Myo5 andeither N- or C‐terminus of Vrp1 where able to grow onselective plates lacking adenine (circles). B All genesproviding the SH3 domains are numbered according to theirposition in the spot grid, from left to right, top to bottom.SLA1-1 contains the two first SH3 domains of SLA1,SLA1-2 contains the third domain.7.4 SLA2 is ascent from the A. gossypii genomeSla2 is an essential part of the endocytic machinery in S. cerevisiae but the SLA2 gene can notbe found in the A. gossypii genome. A search using NCBI pBLAST with the S. cerevisiaeSla2 sequence generates no hits in A. gossypii and further investigation indicates that theSLA2 gene has been lost during evolution, possibly in the development of a mating type locus.The synteny around SLA2 in S. cerevisiae has not been conserved, the gene is found on thethird chromosome next to SUI1. In A. gossypii SUI1 is connected to of all of the three matingtype loci on chromosomes 4, 5 and 6 (Fig. 22). In other yeasts SLA2 is coupled to the matingtype loci as in Klyveromyces lactis where SLA2 is located between SUI1 and the mating typelocus.41

S. cerevisiaeChr XIVA. gossypiiChr IVYNL246W YNL245C YNL244C YNL243W YNL242WVPS75CWC25SUI1SLA2ATG2YBR072W YNL246W YNL244C YHR202WSUI1A1A2Chr VYBR072W YNL246W YNL244C YHR202WSUI1A1A2Chr VIYNL247W YNL246W YNL244C YJL204CSUI1A1A2K. lactisYNL245C YNL244C YNL243W YJL207CSUI1SLA2A1A2Figure 22. Syntenic position of SLA2 in yeast. SLA2 is located next to SUI1 in the S. cerevisiaegenome but can not be found in A. gossypii. SUI1 in A. gossypii is in synteny with the mating typelocus on all three chromosomes containing the mating type information (in this picture a MATastrain).In Klyveromyces lactis SLA2 is positioned between SUI1 and the mating type locus.7.5 Generation of Agsac6SLA2 and SAC6 were found to be essential in abp1 strains of S. cerevisiae and showedsynthetic lethality with each other. SLA2 is not present in the genome of A. gossypii so wewanted to investigate the impact of a SAC6 deletion in this fungus. To this end, twoindependent homokaryotic sac6 strains where generated by replacing the whole SAC6 ORFwith a deletion cassette. The cassette was generated in a two step PCR (Fig. 16). First, 500base pair long sequences of upstream and downstream regions of the SAC6 ORF and a markercassette were amplified by PCR. The primers used in these reactions would add shortsequences to the flank fragments and the marker that overlap. In a second PCR, the threeoverlapping fragments were used as a template with primers defining the very ends of thecassette. Accordingly, a deletion cassette with long homologous regions to the SAC6 locuswas produced. Transformation of the deletion cassette produced heterokaryotic strains;integration of both ends of the cassette was confirmed by PCR (App. II). Sporulation andmicromanipulation of germinated spores enabled isolation of two homokaryotic strains. Nointernal PCR fragments of SAC6 could be amplified from these strains confirming that theywere null mutants (App. II).42

7.6 Characterization of Agsac6Both sac6 strains generated by PCR based gene targeting displayed the same features andmutant phenotypes were restored by introducing a plasmid containing the whole SAC6 gene(Fig. 23B). This confirms that the observed phenotypes have arisen due to deletion of theSAC6 ORF. The sac6 strains are temperature sensitive with reduced growth rate on full media(Fig. 23A). Spores are unable to form visible colonies in 37 C and germination frequency is20-30 folds lower at 30 C and room temperature in sac6 than in leu2 (Fig. 24).A30 ºCleu2 sac6 sac6 + p-SAC6BSAC6SAC6pSAC6t37 ºCleu2sac6pRS418-SAC67,9 kbNAT5ampCEN/ARSFigure 23. Growth of sac6 and complementation. A A small mycelial inoculum of leu2 and sac6on AFM plates after 7 days. B A plasmid bearing the whole gene of SAC6 and the resistancegene for ClonNat is able to restore the growth rate of the mutant. Scale bar is 2 cm.RT 30°C 37°Cleu2300 180 186x100sac6700 900 0Figure 24. Germinated spores on AFM plates after three days.leu2 spores were diluted another 100x. Numbers showapproximation of germinated spores on the plates. Scale bar is2 cm.43

ABDEC100µM20µMFigure 25. Bursting hyphae of sac6 and wal1 on microscope slides. A sac6spores where germinated in 30 °C o/n and then transferred to 37 °C o/n. Arrowsindicate a few of the parts of hyphae that has blown up or burst, two areas aremagnified, B and C. For comparison, a germinated spore, D, and part of a hypha,E, from a wal1 strain grown in 30 °C.When the mycelia of sac6 were grown on plates in 37 C the hyphae often swelled and burstclose to the tips, a feature similar to the wal1 mutant (Fig. 25). An actin stain revealed that theactin patches of sac6 are mislocalized and not polarized to the tip as in wild type (Fig. 26).wal1 shows a similar subapical localization of actin but with fewer patches.sac6+leu2sac6wal1p-SAC630°C 37°C 30°C 37°C 30°C 30°CA B C D E FFigure 26. Actin stain with Alexa488. Actin patches are subapically localized in sac6, C-D,and not polarized to the very tip as in leu2, A-B. p-SAC6 fully complements thephenotype, E. The actin mislocalization is also pronounced in wal1, F. Scale bar is 10 µm.44

To see whether the endocytosis was affected in these two mutants, both sac6 and wal1 werestained with the lipophilic dye FM4-64 and the fluid phase endocytosis marker LuciferaseYellow, LY. Both assays indicate that sac6 has a slightly reduced endocytosis rate comparedto wild type and that wal1 is even more affected (Fig. 27 and 28). The endosomes becomestained with FM4-64 very fast in all strains but the dye remains in the cell membrane for alonger time in sac6 and wal1 compared to leu2. After 15 minutes vacuoles are seen in leu2but they are still only faintly stained in the mutants even after 30 minutes. The FM4-64 dye isfinally delivered to elongated vacuoles in sac6, indicating movement of the vacuolarmembranes. In wal1 the vacuoles are spherical, filling up the diameter of the hyphae but donot elongate. It takes longer for the LY to reach the vacuoles in sac6 and wal1 than in leu2.Within 15 minutes the dye lights up vacuoles in leu2 but it takes 30-60 minutes for vacuolesin sac6 to be readily visible and even longer in wal1.leu215min 30min 120minsac6wal1Figure 27. FM4-64 uptake in sac6 and wal1. Endosomes in the tipsare readily visualized by the dye. Eventually the dye ends up in largerfused vacuoles in the proximal areas of hyphae. Scale bar is 5µm.45

leu215min 30min 60minsac6wal1Figure 28. Luciferase Yellow is taken up by fluid phase endocytosisand stains the inside of the vacuoles. Scale bar is 5 µm.7.7 Deletion of LSP1, SUR7, PKH1 and YPK1Null mutant strains of lsp1 and sur7 where generated by PCR based gene targeting asdescribed for the sac6 strains (Fig. 16). Long flanking deletion cassettes generated in a twostep PCR replaced the whole ORF of the two genes. PCR confirmed integration of thecassette into the target locus at both ends in the heterokaryons and no internal PCR productscould be amplified from purified DNA from micromanipulated strains. Heterokaryotic strainsof pkh1 and ypk1 were generated using long flanking deletion cassettes and verified by PCR.All verification PCRs are shown in Appendix II. No homokaryotic mycelia with PHK1 orYPK1 deletions could be isolated, spores would germinate in an o/n culture at 30 C but aftermicromanipulation they seized to grow. Approximately 30 spores that had germinated andproduced one or two germ tubes were selected for each strain but none of them grew furtherafter isolation. The pkh1 and ypk1 strains where not investigated further.7.8 Characterization of lsp1 and sur7Null mutants of lsp1 and sur7 where investigated on the aspects of growth rate on full mediaplates, actin distribution and endocytosis visualized by uptake of FM4-64. Aglsp1 displays nophenotypes compared to the precursor strain leu2 in all these cases. Actin patches arepolarized in growing hyphal tips and actin rings marks septa (Fig. 30). FM4-64 stainsendosomes and is delivered to large vacuoles at a normal rate (Fig. 31)and the size andmorphology of lsp1 colonies after a week on AFM plates are similar to leu2 (Fig. 29).46

Deletion of SUR7 results in a weak phenotype and the sur7 colonies are somewhat smallerthan leu2 and lsp1 on full media plates after a week (Fig. 29). At this time point both leu2 andlsp1 are yellow in the centre of the colonies, due to riboflavin production, but sur7 is still palein colour. All of the colonies have produced spores in the central zone and sur7 sporulatesefficiently after a few days in CSM media. Actin is normally distributed in rings at septal sitesand in patches polarized to sites of growth (Fig. 30). Uptake of FM4-64 in hyphal tips of sur7is fast, endosomes are stained very soon, indicating normal endocytosis (Fig. 31). However,sur7 is affected in the fusion of vacuoles; it takes up to three hrs before the larger vacuoles arevisualized with the dye, in leu2 it takes less than 30 minutes. Chitin deposition is affected insur7, calcofluor staining showed irregularly shaped septa and sometimes clumps of chitinaggregated on the inside of the cell walls (Fig. 32). Sur7 is a membrane protein and to testwhether the plasma membrane was affected in the mutant sur7 was grown on full mediaplates containing SDS or Calcofluor White, CWF. The hyphae of sur7 were indistinguishablefrom leu2 under a microscope on any of these plates but after a week the colony diameters ofsur7 were 68 % and 73 %, respectively, compared to leu2 grown under the same conditions(n=4-5) (Fig. 33). On full media plates the colony size of sur7 was 86 % of leu2 after a week.30 ºCleu2 lsp1 sur737ºCFigure 29. Growth of lsp1 and sur7 on AFM after sevendays at 30 C. Scale bar is 2 cm.47

leu2 lsp1 sur7 leu2 lsp1 sur7Figure 30. Alexa488-Phalloidin stain ineisosome mutants showing polarized actinpatches and acting rings. Scale bar is 10µm.leu215min 30min 120minlsp1sur7180minFigure 31. FM4-64 staining of eisosome mutants. Endosomes becomevisible in the tips very fast and the dye is delivered to the large tubularvacuoles eventually. Scale bar is 5 µm.leu2sur7Figure 32. Calcofluor stain showing chitin depositions at septalsites. Scale bar is 10 µm.48

0,005% SDS 100µg/ml CFW AFM42644260410741034104424141094256419941104105424241064108Aleu2sur7Figure 33. A Growth of leu2 and sur7 on plates containing 100 µg/ml CFW or0,005 % SDS after one week at 30 C. Scale bar is 2 cm. B Graph showing sizes ofthe colonies, 4-5 colonies each where measured.7.9 Deletion of PIL1Deletion cassettes with long flanks could not be generated for PIL1. The upstream region ofthe PIL1 start codon could not be amplified by PCR even though different primers weredesigned for the purpose (Fig. 34). One primer was placed in the upstream gene ISR1 but theregion between the genes could not be amplified in order to clone and sequenced it.500 bpISR1_AEL330CPIL1Figure 34. The PIL1 locus and the neighbouring gene ISR1 according to the annotatedsequence in AGD. Lines represent PCR products defined by nine different primer paircombinations, primer numbers below. Whole lines correspond to successfully amplified regionsand dotted lines to regions that could not be amplified with the defined primers.Mutant strains of PIL1 were instead generated by deletion of almost the complete ORF and byinsertion of a cassette into the ORF. Homokaryotic pil1 mycelia could not be produced,although independent heterokaryons where confirmed by PCR in both disrupted and ORFdeleted versions. The heterokaryotic strains where sporulated and micromanipulated in orderto isolate homokaryotic strains. Heterokaryotic mycelium contains both transformed and wild49

type nuclei and upon sporulation they will separate into individual spores. The wild typespores will not germinate in selective media and every spore that germinates should thereforeconstitute a null mutant genotype. Spores from pil1 (both deletion- and disruption mutants)where incubated o/n in selective full media at both 30 C and room temperature. A largeportion of the spores germinated (Fig. 35) and in total 50-60 germlings at different sizes wereselected, from spores with a germ cell to small mycelia with a few lateral branches. None ofthem continued to grow after they where transferred to new spots on the plate and placed at30 C or room temperature.Figure 35. Germinated spores of pil1 grown o/n in AFM-G418 at 30 C. Aheterokaryotic mycelium was sporulated producing spores with either a wild typeor pil1 nucleus. The germ cells that have formed should all be pil1 null mutantssince wild type does not germinate in selective media. Scale bar is 10 µm.However, a small population of the pil1 germlings grew long enough in liquid culture to allowfor characterization of several features. The germlings eventually ceased to grow and theynever reached the tip splitting stage. For this reason, no mycelia could be collected to verifythe disruption of the PIL1 gene by PCR. However, both pil1 strains generated by disruptionand by ORF deletion displays exactly the same properties and a plasmid bearing anAgTEF1p driven PIL1-GFP could rescue the strains. Since all mutants of pil1 show the samephenotype in all regards they are not distinguished between in the characterization.7.10 Characterization of the pil1 phenotypeA small population of the germinated pil1 spores continued to grow and generated germ tubesand lateral branches. Eventually the cells went into a stationary phase where the surfacebecame full of kernels and they hyphae blew up, giving them features that easily distinguishedthem from the still vital germlings. These two sets of pil1 species were been investigated andcompared on several characteristics. Approximately ten germlings of each genotype (deletionor disruption mutant) reaching at least the state of two germ tubes have been used to definethe typical characteristics of the two phenotypes (vital and stationary) in each test. The vital50

subset of pil1 mutants has polarized actin patches (Fig. 36), normal septa (Fig. 38) andFM4-64 uptake after 45 minutes is comparable to leu2 germlings of the same size (Fig. 41).Luciferase Yellow is taken up by fluid phase endocytosis in pil1 and ends up in large vacuolesin the tip (Fig. 39). In leu2 only small vacuoles are seen in the tip as the larger vacuoles arelocated further away from the tip (not shown). The pil1 germlings cease to grow at differentstages of lateral branching, usually within 10 hrs post germination. None of the hyphae wereobserved to reach maturation where they would do dichotomous branching and the growthspeed would increase remarkably. Fig. 40 shows a rather large pil1 germling that brancheslaterally from several hyphae, all the tips have polarized actin patches. When the germlingshave ceased to grow the actin is no longer polarized to the tips (Fig. 36) and mitochondriahave been fragmented (Fig. 37). Endocytosis is halted in this stage, fluid phase uptake of LYis blocked (Fig. 39) and FM4-64 is not seen in small vacuoles as in the vital pil1 germlings(Fig. 41).leu2 pil1 pil1A B CFigure 36. F-actin staining of pil1. A, B Germlings that still grow have polarized and wildtype like distribution of actin patches. C The germlings with a dotty structure and swollenhyphae have depolarized actin. Calibration bar is 10µm.leu2 pil1 pil1A B CFigure 37. Mitokondrial staining with DiOC6 showing long and tubularmitokondria in leu2, A, and in the vital pil1 hyphae, B. Mitochondria havebecome defragmented in C. Scale bar is 10µm.51

leu2 pil1 pil1A B CFigure 38. Calcofluor staining of pil1. Scale bar is 10µm.leu2 pil1 pil1A B CFigure 39. Fluid phase uptake of LY in young germlings after 2 hours. The dye lights upsmall vacuoles in the tips of leu2, A (arrows), and large vacuoles in the viable pil1 hyphaeB. LY has not been internalized in pil1 hyphae in C. Scale bar is 5 µm.AB20µM10µMFigure 40. A Extensive lateral branching in a small pil1 mycelium. B Actinstain of one hyphae showing polarized patches in the leading tip, arrow.52

leu2Apil1Bpil1CFigure 41. FM4-64 uptake of pil1 after 45 minutes, parts of the hyphae are enlarged forgreater details. They dye is normally delivered to small vacuoles in the germlings, A and B. InC no vacuoles can be seen, the dye is dispersed in dots all over the hyphae. Scale bar is10µm.53

NAT5CEN/ARSampkanMX7.11 Localization of AgPIL1-GFPOverexpression of a PIL1-GFP construct could rescue the pil1 phenotype. Heterokaryoticmycelia with the complementing plasmid were sporulated and the growth rate of the resultinghomokaryotic strains bearing the plasmid was restored to growth rate of leu2 in 30 C(Fig. 42). PCR products from a wild type PIL1 copy could not be amplified from thecomplemented strain, only the GFP-tagged version was found to be present. The GFPlocalized in a punctate cortical pattern (Fig. 43), representing the eisosomes, often leaving thevery tip of hyphae devoid of spots (Fig. 45). Eisosomes formed already in the germ cells(Fig. 44). The PIL1-GFP plasmid was transformed into leu2 and the other eisosome mutants,lsp1, sur7, sac6 and wal1. Distribution of eisosomes was the same in these strains eventhough their genomes still contain untagged PIL1 (Fig. 45 and 46).A30ºCleu2pil1+p-PIL1-GFPBAgTEF1pPIL1GFP37ºCpRS418-AgTEF1p-PIL18,9 kbFigure 42. Complementation of pil1. A Colony morphology on AFM plates aftersix days. B. A plasmid with an AgTEF1p-driven PIL1-GFP construct was usedfor complementation of pil1. Scale bar is 2 cm.Figure 43 . Pil1-GFP in leu2. Image of one planeshow the cortical localisation of eisosomes. Scale baris 10 µm.54

Figure 44. PIL1-GFP representing theeisosomes appears already in thegermcells.leu2 lsp1 sur7 pil1Figure 45. PIL1-GFP overexpressed from a plasmid represents eisosomes in leu2 andthe lsp1 and sur7 mutants and rescues pil1, merged Z stacks. Scale bar is 10 µm.7.12 PIL1-GFP does not colocalize with actin patchesAn overlay of the GFP signal with an actin staining in leu2 shows that actin patches andeisosomes never colocalize, any overlap would be seen as yellow in Fig. 46. The same is truefor all the eisosome mutants as well for the sac6 and wal1 strains. The two latter strains aredeficient in endocytosis and have mislocalized actin but yet a normal eisosome distribution.PIL1-GFP was integrated into the genome of A. gossypii, thereby leaving the expression ofthe fusion protein to be regulated by its own promoter. Localization of the integratedPIL1-GFP is the same as of the overexpressed construct. The tips of growing hyphae lackeisosomes but when the growth slows down or stops the eisosomes catch up and can be seenall over the tip (Fig 47).55

PIL1-GFP actin mergeABlsp1 sur7 pil1 sac6 wal1Figure 46. Actin stains and PIL1-GFP overexpressed from the pRS418-AgTEF1p-PIL1plasmid, Merged Z-stack pictures of eisosome mutants and two mutants deficient inendocytosis, sac6 and wal1. Co-localisation is visualised in yellow. Scale bar is 10 µm.exponential growthPIL1-GFP actin mergeABo/n cultureFigure 47. Z-stacks of actin and PIL1-GFP, colocalization is shown in yellow. PIL1-GFPwas integrated into the genome of leu2 and is driven by its endogenous promoter. A Ina fast growing hypha actin is polarized to the tip but depraved of eisosomes. B PIL1-GFP covers the tip when the hypha have stopped growing and actin patches aredepolarized. Scale bar is 10 µm.7.13 Heterologous expression of AgPIL1-GFP in S. cerevisiaeThe four proteins Pil1 and Lsp1 of A. gossypii and S. cerevisiae are very similar at the aminoacid level (Fig. 48). Comparison of Lsp1 and Pil1 in A. gossypii shows that the veryC-terminal part of the proteins differs most in sequence (Fig. 49). The AgTEF1p drivenAgPIL1-GFP construct localizes to cortical patches in the S. cerevisiae strain BY4741(Fig. 50).56

AgPil175 %AgLsp1Figure 48. Pil1 and Lsp1 ofS. cerevisiae and A. gossypii. Aminoacid identity calculated with MegAlign.76 %88 %85 %75 %ScPil171 %ScLsp1AgLsp1.proMHRTYSLRNQKAPTASDLQSPPPPPSSTRSKFFGRAGIASSFRKNAAGNFGPELARKLSSFVKTEKGVLRALEVVANERR 80AgPil1.proMHRTYSLRNSRAPTASQLQNPPPPPSTTKNRFFGKGGLANTFRKNTAGAFGPELSRKLSQLVKIEKNVLRAIEVAANERR 80AgLsp1.proAAARQLSMWGMDNDDDVSDVTDKLGVLIYELGELQDQFIDKYDQYRVTVKSIRNIEASVQPSRDRKQKITDQIAHLKYKE 160AgPil1.proDAAKQLSLWGLENDDDVSDITDKLGVLIYETSELDDQFIDRYDQYRLTLKSIRDIEGSIQPSRDRKAKITDKIAYLKYKD KIAYLKYKD 160AgLsp1.proPQSPKIPVLEQELVRAEAESLVAEAQLSNITREKLKAAFNYQFDAIRELSEKFALIAGYGKALLELLDDSPVTPGETRPA KAAFNYQFDAIRELSEKFALIAGYGKALLELLDDSPVTPGETRPA 240AgPil1.proPQSPKIEVLEQELVRAEAESLVAEAQLSNITRSKLKAAFNYQFDSLIEHSEKLALIAGYGKALLELLDDSPVTPGETRPA 240AgLsp1.proYDGYEASRQIIMDAEQALEEWTLDAAAVKPNLSFHQTVDDVYDGEDGGEEHDWEGTQDETEQATK 305AgPil1.proYDGYEASKQIIIDAEAALNDWTLDTAAVKPSLSIRRDYDEEFEEGDDGEQWEQDATEEQVAA 302Figure 49. Sequence alignment of AgLsp1 and AgPil1 with MegAlign.Figure 50. Heterologous expression ofAgPIL1-GFP in S. cerevisiae, Z-stacks. TheAgTEF1p driven construct localizes in apatch like pattern in S. cerevisiae. Scale baris 5 µm.57

8 DiscussionEndocytosis and the elaborate network of pathways during this process are well known in thebudding yeast model Saccharomyces cerevisiae. Here, the related ascomycetous fungiCandida albicans and Ashbya gossypii were used to study some of the key proteins involvedin actin organization that confer endocytosis and growth. SH3 domain proteins promoteprotein-protein interactions and are frequently involved in processes like endocytosis andsome of these proteins were studied in C. albicans. This human pathogen displays a widearray of morphological states and specifically the switch between yeast and hyphal growth ispivotal for its pathogenicity. C. albicans has evolved pathways that differ from S. cerevisiaebut core mechanisms of the actin polymerization machinery and localization of severalSH3 domain-containing proteins are conserved. In the yeast model, components such as theWASP homolog Las17, Vrp1 and the SH3 domain myosin I proteins Myo3 and Myo5 arecrucial for proper actin organization, polarity establishment and vacuolar morphology. Strainsof S. cerevisiae and C. albicans with deletions in the genes corresponding to these proteinshave similar phenotypes suggesting that they have related functions in the actin filamentnucleation. A. gossypii is very closely related to the model yeast but is strictly filamentous.Deletion of the WASP homolog WAL1 affects actin polarization, vacuolar fusion andseptation. A. gossypii sustains unidirectional growth at a high rate even though SLA2 wasfound to be absent from its genome. SAC6, which is synthetic lethal with SLA2 inS. cerevisiae, displays genetic interaction with WAL1 in A. gossypii placing SAC6 and WAL1in related pathways. Eisosomes, fungal specific structures implicated in endocytosis, arepresent in A. gossypii and are distributed in a punctate pattern beneath the cell membrane as inS. cerevisiae but do not colocalize with actin.8.1 Localization of C. albicans SH3 domain proteinsC. albicans is a complex organism and many rearrangements in its genome have resulted indifferent lifestyles and new functions compared to S. cerevisiae. Genetics are more tediousdue to its diploidy and less efficient homologous recombination machinery. Out of eight GFPconstructs in this study only three produced a convincing fluorescent signal when the GFPwas attached to the 3’ end of the gene. The localization of these C. albicans proteins, Bbc1,Cyk3 and Sla1, is the same as for their homologs in S. cerevisiae, even though the hyphalcells of C. albicans adds an extra dimension to the picture. Cyk3 localizes, in addition to budnecks in yeast cells and pseudohypae, to septal sites in C. albicans hyphae, a structure that is58

absent from S. cerevisiae. Correct fusion of GFP to the 3’ end of all eight genes on plasmidswas confirmed and incorporation into the genome was proven successful in all cases.Addition of the bulky GFP might have disturbed the folding or the 3D-structure of the nativeprotein, abolishing its localization. Fusing the GFP to the N-terminal of the remaining sevenproteins might have been more successful in some of the cases but that was not attempted inthis study. The fluorescence signal might have been too weak to observe if expression of theGFP fused gene was driven by a weak promoter, a double or triple GFP could have enhancedthe fluorescence. It is even possible that the gene was turned off under given circumstances.C. albicans can not propagate plasmids so an overexpression construct would have to begenomically integrated.8.2 Functional relation between genes regulating actin filamentation in S. cerevisiae,C. albicans and A. gossypiiSeveral studies have demonstrated that WASP family proteins, type I myosins and the Arp 2/3complex are key factors in the nucleation of actin filaments in diverse eukaryotic organisms.Defects in this pathway often lead to compromised endocytosis and problems to initiate orsustain polarized growth, resulting in round and swelling cells and slow growth (Holtzman etal. 1993; Walther and Wendland 2004b). Vacuolar fusion is altered by mutations affectingactin regulatory factors such as Vrp1, type I myosins, components of the Arp2/3 complex andthe fimbrin Sac6. Deletion of the acidic domain of S. cerevisiae Las17 that interacts withArp2/3 complex also results in fragmented vacuoles (Eitzen et al. 2002). Processes involvingcytoskeleton organization are highly conserved and due to the relatively close relationshipbetween A. gossypii and C. albicans homologs to the key proteins involved could be expectedto be functional equivalents. The S. cerevisiae type I myosins Myo3 and Myo5 have a highsimilarity and display functional redundancy, C. albicans has only one homolog, Myo5. Theyeast double myo3/myo5 mutant and the C. albicans myo5 have severe defects in actinpolarization, impairing hyphal growth in C. albicans (Goodson et al. 1996; Oberholzer et al.2002). Deletion of the WASP homolog in S. cerevisiae, C. albicans and A. gossypii disturbsactin patch formation, generating temperature sensitive strains (Li 1997; Walther andWendland 2004a; Walther and Wendland 2004b). Defects in polarized growth in Cawal1causes pseudohyphal growth under hypha-inducing conditions and the mislocalized corticalactin in Agwal1 is a likely reason for swellings of hyphae in subapical regions. These mutantsalso present defects in endocytosis of FM4-64 and in vacuolar morphology. This study addsthe notion that deletion of VRP1 in C. albicans produces defects in vacuolar fusion, actin59

polarization and endocytosis (Borth et al. 2010), thought the defects are not as severe as inCawal1.8.3 Vrp1-Wal1-Myo5 complex in C. albicansThe type I myosins, Myo3 and Myo5, in S. cerevisiae form a complex with Las17 and Vrp1that controls Arp2/3 mediated actin assembly (Evangelista et al. 2000) (Fig. 51). Part of thisstudy was assigned to investigate whether this complex also is formed in C. albicans. A twohybrid assay shows that Vrp1 of C. albicans binds strongly to the WH1 (WASP homology 1)domain in the N-terminus of Wal1 and the binding is more efficient when the downstreamproline rich stretch is removed (Borth et al. 2010). The assay also indicates a direct binding ofthe SH3 domain of Myo5 to Vrp1. The interaction with the C-terminal part of Vrp1 is strongbut, being proline-rich, Vrp1 probably provides multiple docking sites for SH3 domainbinding, as in S. cerevisiae (Anderson et al. 1998). Vrp1, Wal1 and Myo5 in C. albicans arethus likely to execute the same functions as their homologs in S. cerevisiae, forming an actinfilament nucleating complex in both yeasts. This could provide an explanation for theobserved similarities of phenotypes in the wal1 and myo5 mutants and the less severephenotype of vrp1. Both Wal1 and Myo5 are activators of the Arp2/3 complex, and loss ofeither of the corresponding these genes may therefore be more detrimental to cells than loss ofVRP1.Figure 51. Model of the myosin-I-Vrp1-Las17 complex in Arp2/3 mediated actin assembly.The SH3 domain of Myo3/5p (circle) binds to Vrp1p and Las17. Myo3/5p and Las17 bindand activate the Arp2/3 complex through their C-terminal acidic domains (arrow). TheWH2 domains of Las17 and Vrp1 (dashed box) bind and couple actin monomers to theArp2/3 complex for efficient actin nucleation. Activation of the Arp2/3 complex leads to itsincorporation into the actin cytoskeleton, either as a cap on the pointed end of an actinfilament or attached to the side of a filament, creating cross-links and branches. Modifiedpicture from Evangelista et al. 2000.60

8.4 Ligand binding of SH3 domains is weakOther SH3 domains where expected to interact with Vrp1 in the two hybrid experiment in thisassay. The SH3 binding motif PXXP is abundant in verproline but interactions are potentiallytoo weak for this experiment to discover additional interaction partners. The SH3 domain ofHof1 in S. cerevisiae binds to the HOT (Hof one trap) domain of ScVrp1 (Ren et al. 2005).However, the HOT domain is missing in CaVrp1 and the two-hybrid assay did not indicateany interaction between the Hof1-SH3 and the Vrp1 in C. albicans. In general the bindingproperties of SH3 domains are promiscuous and weak and other attempts to fish out ligandswith e.g. Co-IP have proven tedious and hard to interpret (unpublished results within thePenelope group). Other reasons for low affinity could be that the cloned fragment is too smallor not flexible enough to fold properly, or that the cloned fragment excludes part of theprotein that confers stability and specificity to the ligand binding of the SH3 domain.8.5 SLA2 is absent from the A. gossypii genomeWhen searching the A. gossypii genome database for homologs to endocytic factors itbecomes evident that A. gossypii lacks a homolog to Sla2/Hip1, Huntingtin InteractingProtein 1. Sla2 is one of the earliest factors to arrive at the endocytic site and provides anessential function in all investigated fungi. The N-terminal ANTH domain targets Sla2 to theplasma membrane, the central coiled-coil interacts with clathrin and the C-terminal talin-likedomain binds actin filaments. Localization of Sla2 to the marked endocytic site is actinindependent but the protein has a possible role to physically link the clathrin coated pit to theforces of actin polymerization that will push the invaginating vesicle inwards. S. cerevisiaesla2 displays stalled actin patches and a severe temperature sensitive and slow growingphenotype (Holtzman et al. 1993). Sla2 shows redundancy as a clathrin adaptor with otherANTH/ENTH containing proteins but has additional roles in endocytosis (Baggett et al,2003). The ENTH domain of the epsins, Ent1/2, is about half the size of the ANTH domainfound in Ap1801/2 and Sla2. The two domains bind to phosphatidyl inositol phosphates butusing different mechanisms. The binding of phosphatidyl inositol phosphates to ENTHdomains causes the folding of a helix that can insert into the membrane and drive membranecurvature; this does not happen in ANTH domains (Stahelin et al. 2003). Sla2 mightconsequently not be critically for induction of membrane curvature but uses its ANTH domainfor membrane targeting. Sla2p is required for viability in abp1 and sac6 mutants and theN-terminal and central domains of Sla2 where shown to be indispensable in these strains61

(Holtzman et al. 1993). Sla2 is essential for hyphal growth in C. albicans (Asleson et al.2001).In S. cerevisiae SLA2 is genetically linked to SUI1 but this gene appears besides all threemating type loci in A. gossypii. SLA2 is positioned next to the MAT locus in many yeasts, alocation that often is subjected to rearrangements during mating type switching. It is possiblethat SLA2 in A. gossypii was lost during the evolution of its mating-type locus. Sla2 binds tothe N-terminal part of Clc1 in yeast and a closer look at AgClc1 shows that the Sla2interacting domain is missing in the A. gossypii homolog (Fig. 52). The missing 5’ part ofAgCLC1 is not due to a sequence mistake or a minor insertion/deletion in the genome; theupstream UTR of AgCLC1 shows no resemblance to the beginning of the ScCLC1 ORF. This,together with the noted absence of SLA2 strongly indicates a directed evolution against theneed for SLA2.AgClc1.pro-----------------------------------------------------------------------MANDYSSSD SD 9KlClc1.pro MADKFPELEDDLVQDGFVTGDGDETEFLRREAEILGDEFKTEQDSELLSKDD---SDSKTLGTSVNEADAIAAASYQPQD 77ScClc1.pro MSEKFPPLEDQNIDFTPNDKKDDDTDFLKREAEILGDEFKTEQDDILETEASPAKDDDEIRDFEEQFPDINSANGAVSSD 80CaClc1.proMADKFPEIDTPAAGG----DDDYEGDFLSREKELVGDEFTTDQDKQVFQDDE----DEEINEFKEQFPEVDTKAQPSGIS 72AgClc1.pro E--LSDRQSSGAERSHAG-------GAHRTGAGATSDSSEPIRKWQERREAEIAERDESEAAATQRLQAEAIKHIDDFYE 80KlClc1.pro VEGVSNPVEEEEDDDEFG------EPQSSSAEPVVRGKSEALENWKARRELEISERDQAEDKAKADLQEEAAKHIDDFYE 151ScClc1.proQN-GSATVSSGNDNGEADDDFSTFEGANQSTESVKEDRSEVVDQWKQRRAVEIHEKDLKDEELKKELQDEAIKHIDDFYD QSTESVKEDRSEVVDQWKQRRAVEIHEKDLKDEELKKELQDEAIKHIDDFYD 159CaClc1.pro VTKGADKYDDDDDEFEGFE------SSNGAAKELNLSESQAIKEWKQRRDLEIEEREKLNSKKKEEIIEKAKSTIDDFYE 146AgClc1.proVYSKKKQQQVEQARREAEEFLQQRDTFFDQDNTVWDRVLQLINTE-DADVLGDRDRSKFKDILLRLKGQEHVPGAARG LRLKGQEHVPGAARG 157KlClc1.pro NYNIKKQQGIDQTQKEAEEFLAKTHAFASQDLTVWDKALQLINLE-DADIVDGRDRSKFKEILQRLKGNGSAPGATGQK 229ScClc1.pro SYNKKKEQQLEDAAKEAEAFLKKRDEFFGQDNTTWDRALQLINQD-DADIIGGRDRSKLKEILLRLKGNAKAPGA 233CaClc1.pro NYNSKRDNHQKEILSEQEKFISKRDDFLKRG-TLWDRVNELVTEVGELPGDESRDKTRFKELLTKLKGKENVPGAGGYQE 225Figure 52. Sequence alignment of Clc1 from A. gossypii, C. albicans, S. cerevisiae andKlyveromyces lactis. A. gossypii is the only fungus among these closely related fungi that lacks theN-terminal Sla2 binding domain. Alignment made with MegAlign.Efficient endocytosis is a requirement for fast growing filamentous fungi such as A. gossypii.Though SLA2 is found in all studied ascomycetes and performs an essential task inendocytosis A. gossypii has found a way to bypass the need for the protein or developedanother pathway that performs the same task. SLA2 was found in a synthetic lethal screenwith ABP1 in yeast and SAC6 was found to be a requirement in abp1 strains as well. Inaddition, SLA2 and SAC6 showed synthetic lethality interactions with each other (Holtzman etal. 1993). This prompted for functional analysis of a SAC6 deletion in A. gossypii.62

8.6 Agsac6 has a similar phenotype to Agwal1The yeast fimbrin, SAC6, participates in the endocytosis process, binding to actin in bothpatches and cables, aiding in the organization and stabilization of the actin network. Deletionof this actin bundler in S. cerevisiae disrupts the actin cytoskeleton and null mutants aredefective for internalization of α-factor (Kübler and Riezman 1993). In Aspergillus nidulansthe fimbrin homolog FimA is needed for polarity establishment and endocytosis (Upadhyayand Shaw 2008). It was suggested that filamentous actin have to be bundled into strongenough structures to aid the scission of vesicles and drive membrane invaginations from thesurface. Sac6 have a redundant function with Scp1, transgelin, in the organization andstabilization of actin (Goodman et al. 2003). Scp1 contains one CH domain and produces aloose meshwork of actin filaments, while Sac6 generates tight bundles. A two- to threefoldoverexpression of Scp1 partially suppresses defects in sac6 cells.The phenotype of Agsac6 shows similarities to Agwal1 suggesting that they take part in thesame pathway. Agsac6 have accumulations of cortical actin subapically and when grown in37 C the hyphae swell and burst. Vacuolar morphology seems unaffected in Agsac6, inopposition to Agwal1, the former produces large elongated vacuoles, suggesting vividmovement. Vacuoles in wal1 are round-shaped and immobile indicating that Wal1 activatedArp2/3 dependent actin polymerization but not filament bundling by Sac6 is required forvacuolar movement. Agwal1 is deficient in septation and sporulation but Agsac6 is normal inthose aspects. Deletion of WAL1 in the sac6 background might be possible, even though bothdeletions cause severe damage on its own, to further study their involvement in the samepathway. Additionally, investigation of any phenotypic rescue by overexpression of SAC6 inthe wal1 deletion background, or vice verse, could be valuable for the discussion. SAC6 inA. gossypii is not necessary for actin patch formation but for their polarization to hyphal tips.Aggregated actin patches behind the growing tip will lead endocytosis from the tip and slowdown growth. Loss of polarization might also be the cause for swelling of hyphae in therestrictive temperature 37 C. In S. cerevisiae the combined deletion of SLA2 and SAC6 islethal, but A. gossypii manages without Sla2 and is viable in the absence of Sac6. HowA. gossypii compensates for the lack of Sla2 is still unclear.8.7 A. gossypii pil1 germlings cease to grow before reaching maturationEisosomes are recently discovered fungi-specific structures that have been argued to marksites of endocytosis in S. cerevisiae. Their main components PIL1 and LSP1 share a high63

degree of similarity but displays distinct non overlapping functions, with only PIL1 beingnecessary for regulation of their size and number. LSP1 is not essential in S. cerevisiae but itsdeletion decreases the rate of endocytosis and suppresses the rvs161 endocytosis phenotype(Walther et al. 2006). Recently, eisosome proteins were described in Aspergillus nidulanswhere PilA and PilB are closely related to the Pil1 and Lsp1 proteins in S. cerevisiae andSurG is the Sur7 homolog (Vangelatos et al. 2010). The distribution and expression pattern ofthe A. nidulans eisosomal proteins are different from S. cerevisiae, all components are presentin the periphery of ungerminated conidia but only PilA is localized in patch like structures inhyphal cells. No phenotypes are generated by deletion of any of the A. nidulans eisosomeproteins. Deletion of LSP1 has no observable effect on A. gossypii, as was shown inC. albicans too (Reijnst et al, unpublished), whereas disruption of PIL1 leads to severe defectsand is barely viable in this study. The few mutant pil1 spores that do germinate seize to growbefore they reach the fast growing mature state. The young germlings have no obviousphenotype regarding actin patch polarization, FM4-64 uptake, septation or mitochondrialmorphology until they suddenly go into apoptosis. Perhaps the initial slow growth isindependent of the eisosome function or, residual copies of PIL1 in the spore are substantialto sustain growth for a while. Endocytosis mutants are in many cases temperature sensitive.Characterization of pil1 was performed in 30 C in liquid culture but an even lowertemperature might have allowed the mutant germlings to persist for longer. Spores of pil1were also germinated and grown on plates in room temperature but even so no colonies wereformed.8.8 PIL1-GFP does not localize with cortical actin patchesCortical actin is known to associate directly with invaginating endocytic pits, following thevesicle as it internalizes. This study shows that eisosomes in A. gossypii, defined byPIL1-GFP, never colocalizes with cortical actin patches. In fast growing hyphae with highlypolarized actin patches the very tip is completely deprived of eisosomes, although this couldbe due to a slow folding of GFP. Even mutants with grossly mislocalized actin patchesdisplay a normal PIL1-GFP pattern. However, being associated to the lipid rich MCCs,eisosomes can still play an important role in the regulation of endocytosis. These lipid raftsform isolated spots where membrane bound transporters are spared from recycling byendocytosis and PIL1, but not LSP1, is a core component (Grossmann et al. 2008).64

8.9 Sur7 is not necessary for eisosome formation but affects vacuolar fusionThe membrane protein Sur7 in S. cerevisiae was previously shown to localize to MCCs andeisosomes but not actin patches. Sur7 and its paralogs FMP45, YNL194C and YLR414C arenot necessary for actin function but affects sphingolipid content in the plasma membrane andsporulation (Young et al. 2002). Deletion of SUR7 in A. gossypii leads to minor defects ingrowth rate, chitin deposition and vacuolar fusion but PIL1-GFP localization and sporulationappears normal. Elevated chitin production could be a compensatory response to cell walldamage as has been reported in other yeasts (Smits et al. 2001). Cell wall disturbing agentsslowed down the growth further but did not result in more frequent cell lysis than in wildtype. The C. albicans sur7 mutant displays an even more pronounced phenotype with defectsin chitin distribution and abnormal actin polarization and cell wall growth (Alvarez et al.2008). Trafficking of vesicles to the vacuoles was delayed in Agsur7 as has been reported inC. albicans, but since SUR7 is localized to the membrane this is likely an indirect effect. InCasur7 this could be due to mislocalized actin but not in A. gossypii since the deletion mutanthave wild type like actin polarization.8.10 A link between lipid rafts and endocytosisBoth LSP1 and SUR7 of S. cerevisiae are genetically linked to RVS161 and RVS167, actinpatch components with a function in the scission event of internalized clathrin coated pits(Walther et al. 2006). rvs161 and rvs167 mutants are suppressed by mutations in thesphingolipid synthesis, e. g. overexpression of the transmembrane protein Sur7 (Desfarges etal. 1993). Sur7 localizes to MCCs which are involved in the regulation of turn over of severalmembrane proteins by endocytosis. Furthermore, deletion of PIL1 disrupts MCCs and itscomponents Can1, Sur7 and Lsp1 cluster in large chunks (Malinska et al. 2004). Theeisosome components Pil1 and Lsp1 are regulated by the sphingolipid-mediated signallingpathway via the kinases Pkh1/2 and Ypk1/2 (Luo et al. 2008). In wild type C. albicans hyphalcells lipid rafts and Rvs167 are polarized to the growing tip. Some Rvs167 patches colocalizeto actin patches but the other subset might interact with lipid rafts. Deletion of SLA2 or MYO5lead to depolymerized lipid rafts and cytoplasmic Rvs167 (Oberholzer et al. 2006). Deletionof RVS167 in C. albicans did not affect the distribution of eisosomes (unpublished results byReijnst et al). Overall this shows that eisosomes are linked to endocytosis even though they donot appear at the specific site of endocytosis, the cortical actin patch. Eisosome componentsmight not take part in the actual internalization process but nonetheless have a regulativefunction via the lipid metabolism.65

9 SummaryC. albicans displays a wide array of morphological states and has acquired skills to penetratethe host immune response. It does so with a genetic setup sharing many treats with the modelsystem S. cerevisiae. In line with this fact, the GFP-tagged C. albicans proteins Bbc1, Cyk3and Sla1 localized in a way similar to the S. cerevisiae homologs. The strictly filamentousA. gossypii is a very close relative to S. cerevisiae and is likely to follow conserved pathwaysas well. It was previously known that mutants of the actin filament nucleation machinery inthese species showed equivalent defects in regard to actin polarization, hyphal growth,endocytosis and vacuolar morphology. This study adds a few more connective lines in theC. albicans endocytosis pathway map showing that Vrp1, Myo5 and Wal1 physically interact.The closer to the core machinery a mutation occurs the more sever the defect is; deletion ofWAL1 in C. albicans generates a more pronounced phenotype than deletion of VRP1. InA. gossypii the sac6 strain is similar to but has a weaker phenotype compared to Agwal1,indicating that Sac6 participates in the same pathway but has a less important mechanism thanWal1 in the actin filament organization. Synthetic lethality between SLA2 and SAC6 inS. cerevisiae indicates a functional redundancy in the pathway. The genome of A. gossypii ismissing any homolog to SLA2 and deletion of SAC6 causes a temperature sensitive andslow-growing phenotype with gross actin patches anomalies. How or why A. gossypii hasdeveloped a pathway around Sla2 to manage the crucial steps of endocytosis is still to beunveiled. And finally, eisosomes that were previously said to mark sites of endocytosis wereshown not to localize to actin patches and thereby do not participate in the internalizationprocess. However, according to many studies they are more likely to regulate endocytosis byassociating to lipid structures in the plasma membrane.66

10 ProspectsSH3 domain proteins are likely to interact with the proline rich Vrp1 but to find interactionpartners in a two-hybrid assay the whole protein might have to be used and not only theisolated SH3 domain as was done in this study. Misstranslation of C. albicans proteins in ayeast system due to the ambiguous codon usage could also present problems. To circumventthe need for heterologous expression the two-hybrid system recently developed by Stynen etal. 2010 in C. albicans itself could be a solution.The fact that SLA2 is absent from A. gossypii is intriguing because it performs a conservedand important function in the endocytosis process of fungi. The Hip1 and Hip1R homologs inhigher eukaryotes share the yeast SLA2 function to associate with clathrin coated pits andfilamentous actin. The N-terminal part of Clc1 (Clathrin light chain 1) that interacts with Sla2in other fungi is not present in the A. gossypii homolog of Clc1 agreeing with the idea thatA. gossypii has evolved a pathway around the need for Sla2. It would be interesting to see ifheterologous expression of the S. cerevisiae Sla2 would affect the actin organization and if theprotein would localize to cortical actin patches in A. gossypii. And would the ScClc1 workproperly in A. gossypii if the endogenous CLC1 was deleted? This study shows thatA. gossypii can survive and perform endocytosis when the fimbrin-encoding SAC6 is deleted.This gene is synthetic lethal with SLA2 and ABP1 in S. cerevisiae. Deletion of ABP1 aloneresults in no obvious phenotype in S. cerevisiae but how would A. gossypii cope with thedeletion? If A. gossypii manage without ABP1 as well this fungus might have evolvedredundant pathways which do not exist in S. cerevisiae or it compensates otherwise.Overexpression of Scp1, transgelin, in S. cerevisiae can partially suppress the disrupted actincytoskeleton phenotype of sac6. A higher expression of the homologous gene might do thesame in A. gossypii. Deletion of SAC6 and WAL1 in A. gossypii produces phenotypes thatsuggest they participate in the same pathway. Overexpression of one gene in the deletionbackground of the other could perhaps partially rescue the deletion phenotype. A doubledeletion of these genetically interacting genes would also be interesting to study.The eisosome proteins Pil1 and Lsp1 in A. gossypii are very similar to each other and to thehomologs in S. cerevisiae, only the last stretch of 30 amino acids is substantially different.Even so, deletion of the two genes produces very different phenotypes in both fungi. Pil1 is67

esponsible for localization of the eisosome components in S. cerevisiae. Lsp1 and Sur7 needsto be GFP-tagged and localized in the A. gossypii pil1 mutant to see if the same is true in thisfungi. A more thorough investigation of the Agpil1 mutants should be done to test theirviability under different circumstance, e g lower temperature. The upstream region of thePIL1 ORF should be sequenced so that the promoter function can be tested. Generation ofchimeras between the closely related Pil1 and Lsp1 could indicate which part of the sequenceis responsible for their non-redundant functions.68

11 AcknowledgementsDuring my PhD I have met many wonderful people that I am grateful to and wish to thank.Jürgen Wendland for providing this opportunity to work in the fine surroundings ofCarlsberg and for guiding me during this whole time.Andrea who has been a support in every possible way, in the lab, at the microscope, asa tutor and as an inspiration.Janine and Alex for being very helpful in the lab at the beginning.Patrick for being a steady colleague for my whole internship who always took time tohelp me out on several matters and with whom I travelled a lot during this time.Anke who I shared office and many discussions with and who would have been theideal German teacher had I only given it some more effort.Sidsel for helping out in the lab with so many things, saving me a lot of precious time.Colleagues at Carlsberg for great chit-chat in the hallways and for pulling off manyentertaining parties.All students, PIs and coordinators in the Penelope consortium that attributed so muchmore value to this training period, providing long lasting connections all over the EU.Steen for helping me fulfil all the requirements for a PhD from KøbenhavnsUniversitet.My furry Tristan, who kept me company for hours at my desk and on my keyboard, Ifelt your warmth and support during my writing.And not least, Glenn, you have been a tremendous support and showed real interestinto my work. I am grateful for the firm but gentle boosts that finally got me to pullthis off.69

12 ReferencesAdams, A. E. and Pringle, J. R. (1984). "Relationship of actin and tubulin distribution to budgrowth in wild-type and morphogenetic-mutant Saccharomyces cerevisiae." J Cell Biol98(3): 934-945.Alberti-Segui, C., Dietrich, F., et al. (2001). "Cytoplasmic dynein is required to oppose theforce that moves nuclei towards the hyphal tip in the filamentous ascomycete Ashbyagossypii." J Cell Sci 114(Pt 5): 975-986.Alvarez, F. J., Douglas, L. M., et al. (2008). "The Sur7 protein regulates plasma membraneorganization and prevents intracellular cell wall growth in Candida albicans." Mol BiolCell 19(12): 5214-5225.Amberg, D. C. (1998). "Three-dimensional imaging of the yeast actin cytoskeleton throughthe budding cell cycle." Mol Biol Cell 9(12): 3259-3262.Anderson, B. L., Boldogh, I., et al. (1998). "The Src homology domain 3 (SH3) of a yeasttype I myosin, Myo5p, binds to verprolin and is required for targeting to sites of actinpolarization." J Cell Biol 141(6): 1357-1370.Ashby, S. F. and Nowell, W. (1926). "The Fungi of Stigmatomycosis." Annals of Botany40(157).Asleson, C. M., Bensen, E. S., et al. (2001). "Candida albicans INT1-induced filamentation inSaccharomyces cerevisiae depends on Sla2p." Mol Cell Biol 21(4): 1272-1284.Ayscough, K. R., Eby, J. J., et al. (1999). "Sla1p is a functionally modular component of theyeast cortical actin cytoskeleton required for correct localization of both Rho1p-GTPaseand Sla2p, a protein with talin homology." Mol Biol Cell 10(4): 1061-1075.Bagnat, M., Keranen, S., et al. (2000). "Lipid rafts function in biosynthetic delivery ofproteins to the cell surface in yeast." Proc Natl Acad Sci U S A 97(7): 3254-3259.Baillie, G. S. and Douglas, L. J. (2000). "Matrix polymers of Candida biofilms and theirpossible role in biofilm resistance to antifungal agents." J Antimicrob Chemother 46(3):397-403.Berman, J. and Sudbery, P. E. (2002). "Candida Albicans: a molecular revolution built onlessons from budding yeast." Nat Rev Genet 3(12): 918-930.Boczkowska, M., Rebowski, G., et al. (2008). "X-ray scattering study of activated Arp2/3complex with bound actin-WCA." Structure 16(5): 695-704.70

Borth, N., Walther, A., et al. (2010). "Candida albicans Vrp1 is required for polarizedmorphogenesis and interacts with Wal1 and Myo5." Microbiology.Butler, G., Rasmussen, M. D., et al. (2009). "Evolution of pathogenicity and sexualreproduction in eight Candida genomes." Nature 459(7247): 657-662.Chandra, J., Kuhn, D. M., et al. (2001). "Biofilm formation by the fungal pathogen Candidaalbicans: development, architecture, and drug resistance." J Bacteriol 183(18): 5385-5394.D'Agostino, J. L. and Goode, B. L. (2005). "Dissection of Arp2/3 complex actin nucleationmechanism and distinct roles for its nucleation-promoting factors in Saccharomycescerevisiae." Genetics 171(1): 35-47.Desfarges, L., Durrens, P., et al. (1993). "Yeast mutants affected in viability upon starvationhave a modified phospholipid composition." Yeast 9(3): 267-277.Dietrich, F. S., Voegeli, S., et al. (2004). "The Ashbya gossypii genome as a tool for mappingthe ancient Saccharomyces cerevisiae genome." Science 304(5668): 304-307.Drubin, D. G., Miller, K. G., et al. (1988). "Yeast actin-binding proteins: evidence for a rolein morphogenesis." J Cell Biol 107(6 Pt 2): 2551-2561.Duff, S. J. and Murray, W. D. (1988). "Production and application of methylotrophic yeastPichia pastoris." Biotechnol Bioeng 31(1): 44-49.Eitzen, G., Wang, L., et al. (2002). "Remodeling of organelle-bound actin is required for yeastvacuole fusion." J Cell Biol 158(4): 669-679.Enloe, B., Diamond, A., et al. (2000). "A single-transformation gene function test in diploidCandida albicans." J Bacteriol 182(20): 5730-5736.Evangelista, M., Klebl, B. M., et al. (2000). "A role for myosin-I in actin assembly throughinteractions with Vrp1p, Bee1p, and the Arp2/3 complex." J Cell Biol 148(2): 353-362.Field, C., Li, R., et al. (1999). "Cytokinesis in eukaryotes: a mechanistic comparison." CurrOpin Cell Biol 11(1): 68-80.Gietz, R. D. and Schiestl, R. H. (2007). "High-efficiency yeast transformation using theLiAc/SS carrier DNA/PEG method." Nat Protoc 2(1): 31-34.Gimeno, C. J., Ljungdahl, P. O., et al. (1992). "Unipolar cell divisions in the yeast S.cerevisiae lead to filamentous growth: regulation by starvation and RAS." Cell 68(6):1077-1090.Gomes, A. C., Miranda, I., et al. (2007). "A genetic code alteration generates a proteome ofhigh diversity in the human pathogen Candida albicans." Genome Biol 8(10): R206.Goode, B. L., Rodal, A. A., et al. (2001). "Activation of the Arp2/3 complex by the actinfilament binding protein Abp1p." J Cell Biol 153(3): 627-634.71

Goode, B. L., Wong, J. J., et al. (1999). "Coronin promotes the rapid assembly and crosslinkingof actin filaments and may link the actin and microtubule cytoskeletons inyeast." J Cell Biol 144(1): 83-98.Goodman, A., Goode, B. L., et al. (2003). "The Saccharomyces cerevisiae calponin/transgelinhomolog Scp1 functions with fimbrin to regulate stability and organization of the actincytoskeleton." Mol Biol Cell 14(7): 2617-2629.Gordon, J. L., Byrne, K. P., et al. (2009). "Additions, losses, and rearrangements on theevolutionary route from a reconstructed ancestor to the modern Saccharomycescerevisiae genome." PLoS Genet 5(5): e1000485.Grossmann, G., Malinsky, J., et al. (2008). "Plasma membrane microdomains regulateturnover of transport proteins in yeast." J Cell Biol 183(6): 1075-1088.Grossmann, G., Opekarova, M., et al. (2007). "Membrane potential governs lateralsegregation of plasma membrane proteins and lipids in yeast." EMBO J 26(1): 1-8.Grove, S. N. and Bracker, C. E. (1970). "Protoplasmic organization of hyphal tips amongfungi: vesicles and Spitzenkorper." J Bacteriol 104(2): 989-1009.Hall, A. (1994). "Small GTP-binding proteins and the regulation of the actin cytoskeleton."Annu Rev Cell Biol 10: 31-54.Harris, S. D., Read, N. D., et al. (2005). "Polarisome meets spitzenkorper: microscopy,genetics, and genomics converge." Eukaryot Cell 4(2): 225-229.Holtzman, D. A., Yang, S., et al. (1993). "Synthetic-lethal interactions identify two novelgenes, SLA1 and SLA2, that control membrane cytoskeleton assembly inSaccharomyces cerevisiae." J Cell Biol 122(3): 635-644.Horio, T. and Oakley, B. R. (2005). "The role of microtubules in rapid hyphal tip growth ofAspergillus nidulans." Mol Biol Cell 16(2): 918-926.Hull, C. M. and Johnson, A. D. (1999). "Identification of a mating type-like locus in theasexual pathogenic yeast Candida albicans." Science 285(5431): 1271-1275.Hull, C. M., Raisner, R. M., et al. (2000). "Evidence for mating of the "asexual" yeastCandida albicans in a mammalian host." Science 289(5477): 307-310.Idrissi, F. Z., Grotsch, H., et al. (2008). "Distinct acto/myosin-I structures associate withendocytic profiles at the plasma membrane." J Cell Biol 180(6): 1219-1232.Jonsdottir, G. A. and Li, R. (2004). "Dynamics of yeast Myosin I: evidence for a possible rolein scission of endocytic vesicles." Curr Biol 14(17): 1604-1609.Kaksonen, M., Sun, Y., et al. (2003). "A pathway for association of receptors, adaptors, andactin during endocytic internalization." Cell 115(4): 475-487.72

Kaksonen, M., Toret, C. P., et al. (2005). "A modular design for the clathrin- and actinmediatedendocytosis machinery." Cell 123(2): 305-320.Kim, K., Yamashita, A., et al. (2004). "Capping protein binding to actin in yeast: biochemicalmechanism and physiological relevance." J Cell Biol 164(4): 567-580.Knechtle, P., Dietrich, F., et al. (2003). "Maximal polar growth potential depends on thepolarisome component AgSpa2 in the filamentous fungus Ashbya gossypii." Mol BiolCell 14(10): 4140-4154.Kohno, H., Tanaka, K., et al. (1996). "Bni1p implicated in cytoskeletal control is a putativetarget of Rho1p small GTP binding protein in Saccharomyces cerevisiae." EMBO J15(22): 6060-6068.Korinek, W. S., Bi, E., et al. (2000). "Cyk3, a novel SH3-domain protein, affects cytokinesisin yeast." Curr Biol 10(15): 947-950.Kübler, E. and Riezman, H. (1993). "Actin and fimbrin are required for the internalizationstep of endocytosis in yeast." EMBO J 12(7): 2855-2862.Kärkkäinen, S., Hiipakka, M., et al. (2006). "Identification of preferred protein interactions byphage-display of the human Src homology-3 proteome." EMBO Rep 7(2): 186-191.Lee, W. L., Bezanilla, M., et al. (2000). "Fission yeast myosin-I, Myo1p, stimulates actinassembly by Arp2/3 complex and shares functions with WASp." J Cell Biol 151(4):789-800.Li, R. (1997). "Bee1, a yeast protein with homology to Wiscott-Aldrich syndrome protein, iscritical for the assembly of cortical actin cytoskeleton." J Cell Biol 136(3): 649-658.Lo, H. J., Kohler, J. R., et al. (1997). "Nonfilamentous C. albicans mutants are avirulent." Cell90(5): 939-949.Luo, G., Gruhler, A., et al. (2008). "The sphingolipid long-chain base-Pkh1/2-Ypk1/2signaling pathway regulates eisosome assembly and turnover." J Biol Chem 283(16):10433-10444.Machesky, L. M., Atkinson, S. J., et al. (1994). "Purification of a cortical complex containingtwo unconventional actins from Acanthamoeba by affinity chromatography on profilinagarose."J Cell Biol 127(1): 107-115.Machesky, L. M. and Insall, R. H. (1998). "Scar1 and the related Wiskott-Aldrich syndromeprotein, WASP, regulate the actin cytoskeleton through the Arp2/3 complex." Curr Biol8(25): 1347-1356.73

Malinska, K., Malinsky, J., et al. (2004). "Distribution of Can1p into stable domains reflectslateral protein segregation within the plasma membrane of living S. cerevisiae cells." JCell Sci 117(Pt 25): 6031-6041.Martin, R., Walther, A., et al. (2005). "Ras1-induced hyphal development in Candida albicansrequires the formin Bni1." Eukaryot Cell 4(10): 1712-1724.Mayer, B. J. and Saksela, K., Eds. (2005). Modular Protein Domains. SH3 Domains,Wiley-VCH.McCann, R. O. and Craig, S. W. (1997). "The I/LWEQ module: a conserved sequence thatsignifies F-actin binding in functionally diverse proteins from yeast to mammals." ProcNatl Acad Sci U S A 94(11): 5679-5684.Miller, M. G. and Johnson, A. D. (2002). "White-opaque switching in Candida albicans iscontrolled by mating-type locus homeodomain proteins and allows efficient mating."Cell 110(3): 293-302.Mochida, J., Yamamoto, T., et al. (2002). "The novel adaptor protein, Mti1p, and Vrp1p, ahomolog of Wiskott-Aldrich syndrome protein-interacting protein (WIP), mayantagonistically regulate type I myosins in Saccharomyces cerevisiae." Genetics 160(3):923-934.Moreira, K. E., Walther, T. C., et al. (2009). "Pil1 controls eisosome biogenesis." Mol BiolCell 20(3): 809-818.Morschhäuser, J., Michel, S., et al. (1998). "Expression of a chromosomally integrated,single-copy GFP gene in Candida albicans, and its use as a reporter of gene regulation."Mol Gen Genet 257(4): 412-420.Naglik, J. R., Challacombe, S. J., et al. (2003). "Candida albicans secreted aspartylproteinases in virulence and pathogenesis." Microbiol Mol Biol Rev 67(3): 400-428,table of contents.Noble, S. M. and Johnson, A. D. (2005). "Strains and strategies for large-scale gene deletionstudies of the diploid human fungal pathogen Candida albicans." Eukaryot Cell 4(2):298-309.Oberholzer, U., Nantel, A., et al. (2006). "Transcript profiles of Candida albicans corticalactin patch mutants reflect their cellular defects: contribution of the Hog1p and Mkc1psignaling pathways." Eukaryot Cell 5(8): 1252-1265.Odds, F. C., Ed. (1988). Candida and Candidosis: a Review and Bibliography, Bailliere Tindall.Ohama, T., Suzuki, T., et al. (1993). "Non-universal decoding of the leucine codon CUG inseveral Candida species." Nucleic Acids Res 21(17): 4039-4045.74

Pfaller, M. A. and Diekema, D. J. (2007). "Epidemiology of invasive candidiasis: a persistentpublic health problem." Clin Microbiol Rev 20(1): 133-163.Pruyne, D. and Bretscher, A. (2000). "Polarization of cell growth in yeast." J Cell Sci 113 ( Pt4): 571-585.Ren, G., Wang, J., et al. (2005). "Verprolin cytokinesis function mediated by the Hof one trapdomain." Traffic 6(7): 575-593.Richman, T. J. and Johnson, D. I. (2000). "Saccharomyces cerevisiae cdc42p GTPase isinvolved in preventing the recurrence of bud emergence during the cell cycle." Mol CellBiol 20(22): 8548-8559.Rodal, A. A., Manning, A. L., et al. (2003). "Negative regulation of yeast WASp by two SH3domain-containing proteins." Curr Biol 13(12): 1000-1008.Santos, M. A. and Tuite, M. F. (1995). "The CUG codon is decoded in vivo as serine and notleucine in Candida albicans." Nucleic Acids Res 23(9): 1481-1486.Sharpless, K. E. and Harris, S. D. (2002). "Functional characterization and localization of theAspergillus nidulans formin SEPA." Mol Biol Cell 13(2): 469-479.Sivadon, P., Peypouquet, M. F., et al. (1997). "Cloning of the multicopy suppressor geneSUR7: evidence for a functional relationship between the yeast actin-binding proteinRvs167 and a putative membranous protein." Yeast 13(8): 747-761.Smits, G. J., van den Ende, H., et al. (2001). "Differential regulation of cell wall biogenesisduring growth and development in yeast." Microbiology 147(Pt 4): 781-794.Stahelin, R. V., Long, F., et al. (2003). "Contrasting membrane interaction mechanisms ofAP180 N-terminal homology (ANTH) and epsin N-terminal homology (ENTH)domains." J Biol Chem 278(31): 28993-28999.Steiner, S., Wendland, J., et al. (1995). "Homologous recombination as the main mechanismfor DNA integration and cause of rearrangements in the filamentous ascomyceteAshbya gossypii." Genetics 140(3): 973-987.Stimpson, H. E., Toret, C. P., et al. (2009). "Early-arriving Syp1p and Ede1p function in endocyticsite placement and formation in budding yeast." Mol Biol Cell 20(22): 4640-4651.Stynen, B., Van Dijck, P., et al. (2010). "A CUG codon adapted two-hybrid system for thepathogenic fungus Candida albicans." Nucleic Acids Res.Sun, Y., Martin, A. C., et al. (2006). "Endocytic internalization in budding yeast requirescoordinated actin nucleation and myosin motor activity." Dev Cell 11(1): 33-46.75

Tang, H. Y., Xu, J., et al. (2000). "Pan1p, End3p, and S1a1p, three yeast proteins required fornormal cortical actin cytoskeleton organization, associate with each other and playessential roles in cell wall morphogenesis." Mol Cell Biol 20(1): 12-25.Thomson, J. M., Gaucher, E. A., et al. (2005). "Resurrecting ancestral alcohol dehydrogenasesfrom yeast." Nat Genet 37(6): 630-635.Toshima, J. Y., Toshima, J., et al. (2006). "Spatial dynamics of receptor-mediated endocytictrafficking in budding yeast revealed by using fluorescent alpha-factor derivatives."Proc Natl Acad Sci U S A 103(15): 5793-5798.Trinci, A. P. (1974). "A study of the kinetics of hyphal extension and branch initiation offungal mycelia." J Gen Microbiol 81(1): 225-236.Ungewickell, E. and Branton, D. (1981). "Assembly units of clathrin coats." Nature289(5796): 420-422.Upadhyay, S. and Shaw, B. D. (2008). "The role of actin, fimbrin and endocytosis in growthof hyphae in Aspergillus nidulans." Mol Microbiol 68(3): 690-705.Walther, A. and Wendland, J. (2003). "An improved transformation protocol for the humanfungal pathogen Candida albicans." Curr Genet 42(6): 339-343.Walther, A. and Wendland, J. (2004a). "Apical localization of actin patches and vacuolardynamics in Ashbya gossypii depend on the WASP homolog Wal1p." J Cell Sci 117(Pt21): 4947-4958.Walther, A. and Wendland, J. (2004b). "Polarized hyphal growth in Candida albicans requiresthe Wiskott-Aldrich Syndrome protein homolog Wal1p." Eukaryot Cell 3(2): 471-482.Walther, A. and Wendland, J. (2008). "PCR-based gene targeting in Candida albicans." NatProtoc 3(9): 1414-1421.Walther, T. C., Brickner, J. H., et al. (2006). "Eisosomes mark static sites of endocytosis."Nature 439(7079): 998-1003.Vangelatos, I., Roumelioti, K., et al. (2010). "Eisosome Organization in the FilamentousAscomyceteAspergillus nidulans." Eukaryot Cell 9(10): 1441-1454.Warren, D. T., Andrews, P. D., et al. (2002). "Sla1p couples the yeast endocytic machinery toproteins regulating actin dynamics." J Cell Sci 115(Pt 8): 1703-1715.Watters, M. K. and Griffiths, A. J. (2001). "Tests of a cellular model for constant branchdistribution in the filamentous fungus Neurospora crassa." Appl Environ Microbiol67(4): 1788-1792.Wedlich-Söldner, R., Bolker, M., et al. (2000). "A putative endosomal t-SNARE links exo- andendocytosis in the phytopathogenic fungus Ustilago maydis." EMBO J 19(9): 1974-1986.76

Wendland, J. (2001). "Comparison of morphogenetic networks of filamentous fungi andyeast." Fungal Genet Biol 34(2): 63-82.Wendland, J., Ayad-Durieux, Y., et al. (2000). "PCR-based gene targeting in the filamentousfungus Ashbya gossypii." Gene 242(1-2): 381-391.Wendland, J. and Philippsen, P. (2000). "Determination of cell polarity in germinated sporesand hyphal tips of the filamentous ascomycete Ashbya gossypii requires a rhoGAPhomolog." J Cell Sci 113 ( Pt 9): 1611-1621.Wesp, A., Hicke, L., et al. (1997). "End4p/Sla2p interacts with actin-associated proteins forendocytosis in Saccharomyces cerevisiae." Mol Biol Cell 8(11): 2291-2306.Wickerham, L. J., Flickinger, M. H., et al. (1946). "The production of riboflavin by Ashbyagossypii." Archives of Biochemistry and Biophysics(9): 95-98.Winter, D., Lechler, T., et al. (1999). "Activation of the yeast Arp2/3 complex by Bee1p, aWASP-family protein." Curr Biol 9(9): 501-504.von Nussbaum, F., Brands, M., et al. (2006). "Antibacterial natural products in medicinalchemistry--exodus or revival?" Angew Chem Int Ed Engl 45(31): 5072-5129.Wright, M. C. and Philippsen, P. (1991). "Replicative transformation of the filamentousfungus Ashbya gossypii with plasmids containing Saccharomyces cerevisiae ARSelements." Gene 109(1): 99-105.Young, M. E., Karpova, T. S., et al. (2002). "The Sur7p family defines novel cortical domainsin Saccharomyces cerevisiae, affects sphingolipid metabolism, and is involved insporulation." Mol Cell Biol 22(3): 927-934.Yu, H., Rosen, M. K., et al. (1992). "Solution structure of the SH3 domain of Src andidentification of its ligand-binding site." Science 258(5088): 1665-1668.Zagulski, M., Herbert, C. J., et al. (1998). "Sequencing and functional analysis of the yeastgenome." Acta Biochim Pol 45(3): 627-643.Zeng, G., Yu, X., et al. (2001). "Regulation of yeast actin cytoskeleton-regulatory complexPan1p/Sla1p/End3p by serine/threonine kinase Prk1p." Mol Biol Cell 12(12): 3759-3772.Zheng, Y., Cerione, R., et al. (1994). "Control of the yeast bud-site assembly GTPase Cdc42.Catalysis of guanine nucleotide exchange by Cdc24 and stimulation of GTPase activityby Bem3." J Biol Chem 269(4): 2369-2372.Zigmond, S. H., Evangelista, M., et al. (2003). "Formin leaky cap allows elongation in thepresence of tight capping proteins." Curr Biol 13(20): 1820-1823.77

Appendix I – Verification PCR on C. albicans strainsHeterozygous strains – Integration of CdHIS1 markerbbc1 bud14 cyk3 nbp25’ 3’ 5’ 3’ 5’ 3’ 5’ 3’800bp500bp5’ G1+G23’ G3+G4pin2 rvs167-2 sho2 sla15’ 3’ 5’ 3’ 5’ 3’ 5’ 3’800bp500bpIntegration of GFP at the C-terminusBBC1 BUD14 CYK3 NBP2 PIN3 RVS167-2 SHO1 SLA21000bp500bpG1-GFP+ 392PCR verifying heterozygous C. albicans strains and C-terminal genomicintegration of GFP. G1/G2 cover the 5’ end of integration of CdHIS1 intotarget locus and G3/G4 the 3’ end. G1-GFP anneals in the target ORFand 392 in the GFP, the primer pair spans the integration point of theGFP-cassette. Arrows mark relevant reference bands.78

Appendix II – Verification PCR on A. gossypii strains1000bp700bp500bpsac61 2 31. 5’ sac6 G1+G2 0.7 kb2. 3’ sac6 G3+G4 0.6 kb3. iSAC6 I1+I2 0.7 kbpkh1/PKH1ypk1/YPK11 2 3 4 5 6sur7lsp11 2 3 4 5 6disruption mutantdeletion mutantpil1/PIL1pil1/PIL11 2 3 41. 5’ pkh1 G1+G6 1.2 kb2. 3’ pkh1 G4+G5 1.0 kb3. iPKH1 I1+I2 0.7 kb4. 5’ ypk1 G1+G2 0.7 kb5. 3’ ypk1 G3+G4 0.6 kb6. iYPK1 I1+I2 0.7 kb1. 5’ sur7 G1+G2 0.9 kb2. 3’ sur7 G3+G4 0.6 kb3. iSUR7 I1+I2 0.7 kb4. 5’ lsp1 G1+ G2 1.1 kb5. 3’ lsp1 G3+ G4 0.9 kb6. iLSP1 I1+I2 0.6 kb1. 5’ pil1 I1+G2 0.6 kb2. 3’ pil1 I2+G3 0.6 kb3. 3’ pil1 G4+G5 0.5 kb4. iPIL1 I1+I2 0.7 kbPCR verifying mutant strains, homokaryons of sac6, sur7, lsp1 and heterokaryonsof pkh1, ypk1 and disrupted and deleted pil1. Primers pairs amplify regionscovering the 5’ and 3’ points of integration of the kanMX marker to confirm targetinginto the correct locus. G2/G3/G5 and G6 anneal in the kanMX, G1/G4 anneal in thegenome outside the target ORF or next to the point of integration in the case of pil1disruption. Internal PCR fragments from target ORF can only be amplified fromheterokaryotic strains; homokaryotic strains lack the internal band. Arrows mark1000, 700 and 500 base pairs reference bands./79

Appendix III – PlasmidsPlasmids used in this studysourceGFP-tags in C. albicans627 pFA-CdHIS1 Schaub et al 2006697 pFA-GFP-CmLEU2 Schaub et al 2006C177 pRS417, GEN3 This studyCAGQS55 BBC1-UAU1-cassetteMitchellCAGCJ50 CYK3-UAU1-cassetteMitchellCAGFY04 SLA1-UAU1-cassetteMitchellC195 pDRIVE-3'BBC1 This studyC199 pRS417-3'BBC1 This studyC255 pRS417- 3'BBC1-GFP This studyC196 pDRIVE-3'CYK3 This studyC200 pRS417-3'CYK3 This studyC257 pRS417- 3'CYK3-GFP This studyC182 pGEM-3'SLA1 This studyC201 pRS417-3'SLA1 This studyC256 pRS417- 3'SLA1-GFP This studyYeast Two Hybrid291 pGAD424 lab collection292 pGBT9 lab collection339 pGAD424-AgCDC3 lab collection423 pGBT9-CaWAL1-ΔN-term Borth et al, 2010481 pGBT9-CaWAL1 Borth et al, 2010639 pGBT9-ScBUD4-C-term lab collectionC99 pGBT9-CaWAL1-ΔC-term Borth et al, 2010C100 pGBT9-CaWAL1-Δpro Borth et al, 2010C103 pGBT9-CaVRP1-N-Term Borth et al, 2010C104 pGAD-CaVRP1-C-Term Borth et al, 2010C113 pGBT9-CaVRP1-C-Term Borth et al, 2010C464 pGAD-SH3-CaABP1 This studyC476 pGAD-SH3-1-CaSLA1 This studyC477 pGAD-SH3-2-CaSLA1 This studyC491 pGAD-SH3-CaBOI1 This studyC492 pGAD-SH3-CaCYK3 This studyC493 pGAD-SH3-CaHSE1 This studyC494 pGAD-SH3-CaBBC1 This studyC495 pGAD-SH3-CaBEM1 This studyC496 pGAD-SH3-CaBEM1L This studyC497 pGAD-SH3-CaFUS1 This studyC498 pGAD-SH3-CaHOF1 This studyC499 pGAD-SH3-CaPEX13 This studyC500 pGAD-SH3-CaPIN3 This studyC501 pGAD-SH3-CaBUD14 This studyC509 pGAD-SH3-CaRVS167-1 This studyC510 pGAD-SH3-CaLSB3 This studyC511 pGAD-SH3-CaSHO1 This studyC512 pGAD-SH3-CaCDC25 This studyC516 pGAD-SH3-CaRVS167-2 This studyC517 pGAD-SH3-CaQ59U90 This studyC520 pGAD-SH3-CaQ5AAN3 This study80

C519 pGAD-SH3-CaNBP2 This studyC521 pGAD-SH3-CaMYO5 This studyC522 pGAD-SH3-CaCDC25L This studySAC6/Eisosomes in A. gossypii121 pFA-kanMX Philippsen651 pRS-AgTEF1p-lacZ Dünkler & Wendland, 2007C136 pFA-NAT5 lab collectionC469 pFA-GFP-kanMX lab collectionC548 pGEM-AgPIL1-A This studyC589 pGEM-AgPIL1-B This studyC595 pRS418-AgTEF1p-AgPIL1 This studyC608 pRS418-AgTEF1p-AgPIL1-GFP This study81

Appendix IV – StrainsStrains used in this studysourceGFP-tags in C. albicansarg4/arg4, leu2/leu2, his1/his1,SN148 ura3::imm434/ura3::imm434, iro1::imm434/iro1::imm434 Noble & Jonhsson, 2005CAS001 BBC1/bbc1::CdHIS1, arg4, leu2, ura3 This studyCAS034 BBC1-GFP-CmLEU2/bbc1::CdHIS1, arg4, ura3 This studyCAP038 CYK3/cyk3::CdHIS1, arg4, leu2, ura3 Reijnst, this studyCAS030 CYK3-GFP-CmLEU2/cyk3::CdHIS1, arg4, ura3 This studyCAS024 SLA1/sla1::CdHIS1, arg4, leu2, ura3 This studyCAS027 SLA1-GFP-CmLEU2/sla1::CdHIS1, arg4, ura3 This studySAC6/Eisosomes in A. gossypiiAgleu2 leu2 Mohr & PhilippsenASJ18 SAC6/sac6:kanMX, leu2 This studyASJ22 sac6:kanMX, leu2 This studyASJ06 lsp1::kanMX/LSP1, leu2 This studyASJ12 lsp1::kanMX, leu2 This studyASJ10 sur7::kanMX/SUR7, leu2 This studyASJ16 sur7::kanMX, leu2 This studyASJ08 pkh1::kanMX/PKH1, leu2 This studyASJ11 ypk1::kanMX/YPK1, leu2 This studyASJ21 PIL1/pil1::kanMX, leu2 This studyASJ24 PIL1/pil1::kanMX, leu2 This studyASJ31 PIL1-GFP, leu2 This studyAWE37 wal1::GEN3, leu2 Walther & Wendland, 2004Cd - Candida dubliniensis, Cm - Candida maltosa82

Appendix V – PrimersPrimers used in this studyGFP-tags in C. albicansCTTTACGTAGTTCTTTTGTTACCCCCAATTGATTGCTCGATTATCCGACACTTCAAAACTCCA3593 BBC1-S1CAATTATTAATAATTATCTTTTCCTGTTTTCAAATTACgaagcttcgtacgctgcaggtc3310 BBC1-S1-GFP AGATTAAGAGTATTTAGACCAGTTGGAAGACAATTTGTTGGTTGGggtgctggcgcaggtgcttcGAAATAATAAATGTGGTGATCTTCTTTCTCTCATCACCCCACACACTCAAAGAGTTTAACAAT3311 BBC1-S2GATGGTTACGTTTAAAACAATACTTCTTCTTCGTTAAtctgatatcatcgatgaattcgag3561 BBC1-G1 GGTACGTCAATGCCGGTTAG3259 BBC1-G1-GFP GGTAATTGAAGTTGCTTACGACG3216 BBC1-A1 CACCATTGGGGACTGGAC3217 BBC1-A4/G4 ATGGCGAATACTCTGGACCACACACAGACCACTTATTTTTAACAACACACACTTCAACACAGTACACCCTTCCCCCCTTCC3548 BUD14-S1CAATACCAATCTATCCACATCTACCTAATTTGAACAGCgaagcttcgtacgctgcaggtc3571 BUD14-S1-GFP TTTGATGAACTTGCTGAAAAGTTGGCAGAATTGGATGATATACTTggtgctggcgcaggtgcttcATTTACACAATTTTACACTACCAAATACTTTCCACTATCATTATTAACAGATTCGTATTATCCTT3549 BUD14-S2TATTATAAAATTTATGAATTATTATTAATACAAATtctgatatcatcgatgaattcgag3553 BUD14-G1 CAACCACTACTACCATTAAC3271 BUD14-G1-GFP CACAACTGATTGATCAGCACG3220 BUD14-A1 ACTCCAGAGACTGTTGAG3221 BUD14-A4/G4 CTTGCGGATCAGGTGTCC4019 CYK3-S1CCTTTCATTAATTACAAAGAAAAAAATAAGAACATCAACTATCTTTTCACTCTTTTTGAACAAATTTGTATCATACTAAAAGAATTAAATAATAAATAATgaagcttcgtacgctgcaggtc3312 CYK3-S1-GFP TATGTTTTCGCTCAGTGGGAGTGCATAGGTAGCACAGTTGCAAATggtgctggcgcaggtgcttcAATGTACAAATGGCAAAAAGAAGTAGTAGCAGAAGAGGTAATCTATAAAGAATTTAAAACTAA3313 CYK3-S2ATAATACCCACTCTGTTTCCCTCTTTATATATATATAtctgatatcatcgatgaattcgag4020 CYK3-G1 GCACACTTGATGATTTCATC3539 CYK3-G1-GFP GACTGCAAGGGCAACCAC3222 CYK3-A1 GCTAAGATCAAGGCAGTG3223 CYK3-A4/G4 GCAACTGCTGCAGTAGACGTCTTGTTTGTCCTGTGTGTGTGTGTGTGTGTGTTGATAAATCACCTGAAACATATACTATTT3589 NBP2-S1AATCATTTGTTATTCATCATTATTGTCCATTTTGAATAGgaagcttcgtacgctgcaggtc3578 NBP2-S1-GFP GAAAAGATCAATGCTATTGAAAAGAAATTAAATGATGTTGAAATAggtgctggcgcaggtgcttcCACATACACTCTGTTGGTATGAAAGTATAAAAACATTTGATAAAATTCGTAATCAACATTAATA3579 NBP2-S2TAACTTAATTGTCCCTATAAGCTGGCTAATATTGGAtctgatatcatcgatgaattcgag3557 NBP2-G1 GGTGTTTCACATTATTCTCCG3245 NBP2-G1-GFP GACAAGTCATTTCCCACC3230 NBP2-A1 CGTCATGGTCAAGGTTGG3309 NBP2-A4/G4 TGGCCGAACCCTTCCTGGGAAATTGTGGACTAAGGTCAACGCCAGTGTTTAATAATCGGAATGTTGTAAATCTTTGCCTT3588 PIN3-S1GACAACAATTTACCTCTTAACACGCAAGAATTACCGCATTGGgaagcttcgtacgctgcaggtc3576 PIN3-S1-GFP GGTGCTGGTGCGCTGATAGGTAGTAACATTGTCAATTCTATTTTTggtgctggcgcaggtgcttcCTTAGTAAAAAACTCATTTCATCTCAGATAATTGTACACCAAGAAATTCAAATGCCTTTTGGC3577 PIN3-S2TATACAACATTACTCCCATATATATGTATATTAAATTtctgatatcatcgatgaattcgag3556 PIN3-G1 CGGTGTGTGTGGCCACTAATG3253 PIN3-G1-GFP GTCAGCTGCCGATGTATTAG3228 PIN3-A1 AAGGCACAACAGGCTGGC3229 PIN3-A4/G4 CTTGGCTCCGCGTATGTCTACCTGGGTTGCAAAAAGTATAAGATACAACAAATAATTACTCCTCCACAAAACACACAAAAA3320 RVS167-2-S1 TACTAAATGATCTACTAGTAAAAAGTTACACTTCATCgaagcttcgtacgctgcaggtc3580 RVS167-2-S1-GFP CTTGTTGGGAATGGAACTGGGTGGGAAAGGCAATTAAACGGAAAAggtgctggcgcaggtgcttcAGAGAATCAATATACATATTCATTCTATTTTTCACTCCTGTAGTACTTTTAATGCATTTAACAA3321 RVS167-2-S2 ACCTGATAAAGAGTGTAAAACAATGGAATATCCTTGtctgatatcatcgatgaattcgag3322 RVS167-2-G1 GGTTTCTGGTTTATTGACCTG3543 RVS167-2-G1-GFP AGAGCAAAGCGAGCTCAC3234 RVS167-2-A1 AGGTAATGGGGTACGTCC3235 RVS167-2-A4/G4 ATCGGTGGGTGCCATTGG83

3316 SHO1-S1CAGTGTATCGATCTCCAATAGATTAGTGTTTATTGATAAACTTCCCAACACTACTACTACTATAGACAGAGATAAACTGTATTAAAATATTAAAGATTGAGgaagcttcgtacgctgcaggtc3586 SHO1-S1-GFP CAAGTTGGTATTTGTCCTTCAAATTATGTTAAATTATTAGATACTggtgctggcgcaggtgcttcCAAATCAAATTAACTCTTCATTTGGGGAAATATAATAATAGTGATAATAATAGTGATAATAAAC3317 SHO1-S2AGTAACAAATAACAAATAACATCAAACCAAAATATACtctgatatcatcgatgaattcgag3318 SHO1-G1 CTTCCTTCCTTCTATATCG3538 SHO1-G1-GFP ACCAGGTAGTGGAACTGG3552 SHO1-A1 GAGTTGGTGCCGGAAGAG3319 SHO1-A4/G4 GAATTCAATCAAGTGGAGG3594 SLA1-S1CAACTCCTATGTTAGAGCTAGTCGTGCTCAACACAAAACCTGATGTGAAACAATGAAACTTTCGACGATTCTACAAAAGTGCGGAAATTGCTTGAAATCAAAGgaagcttcgtacgctgcaggtc3314 SLA1-S1-GFP AGAGCTAATCTACAAGCAGCAACACCAGATAATCCCTTTGGATTCggtgctggcgcaggtgcttc3315 SLA1-S23562 SLA1-G1 CGGTAGAGATGATGTTGTG3243 SLA1-G1-GFP CACAACAACAACCGCCACC3236 SLA1-A1 TGGTGGAGCACCACAGACAGCATTACAAACTATGAAAGGAATAAGAAATAATGAATAATATTTTGTTTGATATACAATTATAAAATAAAAGAGTTAATAAAGGTTCAAAATGCACTTTtctgatatcatcgatgaattcgag3237 SLA1-A4/G4 CGGCTTTGCAACATCAAGAC1432 G2-CdHIS1 TCTAAACTGTATATCGGCACCGCTC1433 G3-CdHIS1 GCTGGCGCAACAGATATATTGGTGC392 GFPup CATAACCTTCGGGCATGGCACTCSAC6/Eisosomes in A. gossypii3982 SAC6-5'a GAGACGGCTACCGTAGACCG3983 SAC6-5'b-S1 GACCTGCAGCGTACGAAGCTTCcgaaatgcacgtgaccaacgtg4076 SAC6-3'a-S2 CGATACTAACGCCGCCATCCAGggaggtatacataccaggcgctg4077 SAC6-3'b attatttctagaCAGTATGTTACACACCCTGGC3982 SAC6-G1 GAGACGGCTACCGTAGACCG3986 SAC6-G4 GGTTGACCTTCTACACTTGGCC304 SAC6-I1 CGACACCAGAGTGCTCAAC305 SAC6-I2 CTGATTCAAAAGAATGGTATAG4111 LSP1-5'a attattctcgagCCTTGCTGTCCCGAAGCACC4112 LSP1-5'b-S1 GACCTGCAGCGTACGAAGCTTCcgtgcttagactggtctggc4113 LSP1-3'a-S2 CGATACTAACGCCGCCATCCAGgtcacacgtgtcttgtctatcgc4114 LSP1-3'b attattgagctcCTCGGGATGATACACGTTGGAC4115 LSP1-G1 CCCAATAACACAGACGTGCG4116 LSP1-G4 CACGTCTTCCGCCTGCTGGG4117 LSP1-I1 CGTTCCGCAAGAATGCAGCG4118 LSP1-I2 CTGTCTAGACGCCTCGTAGC4119 YPK1-5'a attattctcgagCTTGATTGGCACCGAGGAGC4120 YPK1-5'b-S1 GACCTGCAGCGTACGAAGCTTCggcctacaagtagatgctagcg4121 YPK1-3'a-S2 CGATACTAACGCCGCCATCCAGggtcagagcagcttggaagc4122 YPK1-3'b attattgagctcCTGGGAGTAACCGAGATGCGC4123 YPK1-G1 CTGCCTACTACGCGCAGACC4124 YPK1-G4 GCAGATCCTGGCTAGATGATCG4125 YPK1-I1 CTTCCGTGGAGCAGGTGACG4126 YPK1-I2 GCGAGCACAGTGCGTTCAGC4127 PKH1-5'a attattctcgagCCTGTCTACTACCGTCCTCTGC4128 PKH1-5'b-S1 GACCTGCAGCGTACGAAGCTTCgttgcacagctagaggctcg4129 PKH1-3'a-S2 CGATACTAACGCCGCCATCCAGctgaccgttccctgcgtcgg4130 PKH1-3'b attattgaattcCTGATGACAGTGCCAGTGCC4131 PKH1-G1 CTGCTTTGTGTCCGCCAGCC4132 PKH1-G4 CAGCCGAGGACATGCGCAGC4133 PKH1-I1 GCTCATCGGCAGGGAGGACG4134 PKH1-I2 GCAGCAGCTGCGGCAGAAGC4135 SUR7-5'a attattctcgagCGTGACCGCGACCGTGCACC84

4136 SUR7-5'b-S1 GACCTGCAGCGTACGAAGCTTCgtcccgctactctagccacg4137 SUR7-3'a-S2 CGATACTAACGCCGCCATCCAGctacggcagattggtcacgc4138 SUR7-3'b attattgagctcCGTGAACTCTAGCAACCCTGGC4139 SUR7-G1 GTGTGCTGTGCGACGCCACG4140 SUR7-G4 GTCTAGTCTCCTGGAGTGCG4141 SUR7-I1 GGTCGAAGTGGACGTTCTGG4142 SUR7-I2 CGCTGAGGCTCTGCGACAGG4199 PIL1-A1 GCTGAACGACTGGACGCTGG4103 PIL1-5’a attattggtaccCATCTGGTCCTGTCCGGTCC4104 PIL1-5’b gacctgcagcgtacgaagcttcCACTTGCCAGAGTAGCCTCGC4105 PIL1-3’a cgatactaacgccgccatccagCTTCGGCCTTGGCCGAGTGC4106 PIL1-3’b attattgagctcGAAGTCTGAACGGCCTCGTGG4107 PIL1-G1 CACCGTTTCGGACTCGCTGC4108 PIL1-G6 GTCGCTGATCGTGTGGAGCG4109 PIL1-I1 GTCGCAGCTGGTGAAGATCG4110 PIL1-I2 CATCCTGCTCCCACTGCTCG4199 PIL1-A1 GCTGAACGACTGGACGCTGG4200 PIL1-S1-GFP GGCGAGCAGTGGGAGCAGGATGCCACTGAGGAGCAAGTCGCAGCCggtgctggcgcaggtgcttc4201 PIL1-S2-GFP ACGAATAGAAATTAAAGATAGAAAAAGCAGCACTCGGCCAAGGCCtctgatatcatcgatgaattcgagc4241 PIL1-I3 GCACCGGACATACTCCCTAAGG4242 PIL1-G4 GATTCAGACCTGCCTGAAGCC4256 PIL1-I4 attattggcgcgcCGCTCGTTCGCCGCCACCTCG4260 PIL1-G CCGGCTCCTGCTGCCGTGCC4264 ISR1-I2 gttcgcggtcttgatgtccccgtg4270 PIL1-I3 attattgtttaaacATGCACCGGACATACTCCCTAAGG4758 PIL1-S1ATGCACCGGACATACTCCCTAAGGAACTCGAGGGCGCCCACGGCGTCGCAGCTTCgaagcttcgtacgctgcaggtc4759 PIL1-S2CACTCGGCCAAGGCCGAAGCGGGGGCGCGCAGGCGCGCCGGCGCTTTCCTCGCGCtctgatatcatcgatgaattcgag3725 S1 GAAGCTTCGTACGCTGCAGGTC3726 S2 CTGGATGGCGGCGTTAGTATCG1202 G2-kanMX GCGTTTCCCTGCTCGCAGGTC1198 G3-kanMX CGCCTCGACATCATCTGCCC474 G5-kanMX TCGCAGACCGATACCAGGATC473 G6-kanMX GTTTAGTCTGACCATCTCATCTGPrimers were ordered from Eurofins MWG, Germany. Capital letters correspond to annealing sites ingenes, and lowercase letters to sites on pFA-plasmids and added restriction sites.85

Scientific publicationsJorde, S., Walther, A., Wendland, J., (2010). "The Ashbya gossypii fimbrin SAC6 is requiredfor fast polarized hyphal tip growth and endocytosis.” Microbiological Research.Borth, N., Walther, A., Reijnst, P., Jorde, S., Schaub, Y., Wendland, J., (2010). “Candidaalbicans Vrp1 is required for polarized morphogenesis and interacts with Wal1 and Myo5.”Microbiology.Reijnst, P., Jorde, S., Wendland, J.,(2010). “Candida albicans SH3-domain proteins involvedin hyphal growth, cytokinesis, and vacuolar morphology.” Current Genetics.86

1+ModelMICRES 25382 1—9Microbiological Research xxx (2010) xxx—xxxARTICLE IN PRESSAvailable online at www.sciencedirect.comwww.elsevier.de/micres234The Ashbya gossypii fimbrin SAC6 is required forfast polarized hyphal tip growth and endocytosisSigyn Jorde, Andrea Walther, Jürgen Wendland ∗5Carlsberg Laboratory, Yeast Biology, Gamle Carlsberg Vej 10, DK-2500 Valby, Copenhagen, Denmark6Received 19 August 2010 ; received in revised form 15 September 2010; accepted 25 September 20101234567891011121314151617KEYWORDSEndocytosis;Polarized hyphalgrowth;Morphogensis;Actin cytoskeletonAbstractAshbya gossypii has been an ideal system to study filamentous hyphal growth.Previously, we identified a link between polarized hyphal growth, the organizationof the actin cytoskeleton and endocytosis with our analysis of the A. gossypiiWiskott—Aldrich Syndrome Protein (WASP)-homolog encoded by the AgWAL1 gene.Here, we studied the role of AgSAC6, encoding a fimbrin in polarized hyphal growthand endocytosis, and based on our functional analysis identified genetic interactionsbetween AgSAC6 and AgWAL1. SAC6 mutants show severely reduced polarizedgrowth. This growth phenotype is temperature dependent and sac6 spores do not germinateat elevated temperatures. Spores germinated at 30 ◦ C generate slow growingmycelia without displaying polarity establishment defects at the hyphal tip. Severalphenotypic characteristics of sac6 hyphae resemble those found in wal1 mutants.First, tips of sac6 hyphae shifted to 37 ◦ C swell and produce subapical bulges. Second,actin patches are mislocalized subapically. And third, the rate of endocytotic uptakeof the vital dye FM4-64 was reduced. This indicates that actin filament bundling, aconserved function of fimbrins, is required for fast polarized hyphal growth, polaritymaintenance, and endocytosis in filamentous fungi.© 2010 Published by Elsevier GmbH.1234IntroductionEndocytosis describes a process enabling theremodelling of cell surfaces, in the uptakeof nutrients, and in cellular signalling (Munn,∗ Corresponding author. Tel.: +45 3327 5230; fax: +45 3327 4708.E-mail address: jww@crc.dk (J. Wendland).2001). In recent years a strong genetic link 5between the actin cytoskeleton and endocytosis 6has been established, e.g. by using the actin- 7depolymerising drug Latrunculin-A it could be 8shown that loss of filamentous actin prevents 9endocytosis (Aghamohammadzadeh and Ayscough, 102009). Detailed studies using live cell imaging of 11fluorescently tagged proteins allowed the study of 12protein dynamics and thus the timing of events 130944-5013/$ – see front matter © 2010 Published by Elsevier GmbH.doi:10.1016/j.micres.2010.09.003Please cite this article in press as: Jorde S, et al. The Ashbya gossypii fimbrin SAC6 is required forMICRES fast polarized 25382 1—9hyphal tip growth and endocytosis. Microbiol Res (2010), doi:10.1016/j.micres.2010.09.003

2 S. Jorde et al.1415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869at sites of endocytosis (Kaksonen et al., 2005).With these studies it has been established thatactin, the Arp2/3 complex and its regulators aswell as several endocytic proteins co-localize in adynamic yet sequential manner at sites of clathrincoatedpits (Merrifield, 2004; Yarar et al., 2005).The role of the actin cytoskeleton in endocytosisis found in the requirement of an actin meshworkand actin polymerization at sites of endocytosis. Asa part of this, the Arp2/3 complex is required forbranched actin filament generation. Actin nucleatorslike the Wiskott—Aldrich Syndrome Protein(WASP) and the type I myosin Myo5 activate theArp2/3 complex resulting in the generation of actinfilaments (Machesky and Gould, 1999; Higgs andPollard, 1999; Goley et al., 2010). Defects in proteinsinteracting with actin or the Arp2/3 complexand its regulators typically impair endocytosis, cellpolarity or morphogenesis (Pruyne and Bretscher,2000). Polarized hyphal growth as well as polarizedpollen tube growth in plants not only depends onthe targeted delivery of secretory vesicles to thegrowing hyphal tip but also requires endocytosis,e.g. for the compensatory recycling of membranesand tip-localized proteins (Voigt et al., 2005; Fuchset al., 2006; Araujo-Bazan et al., 2008; Taheri-Talesh et al., 2008).Ashbya gossypii is a filamentous ascomycetebelonging to the pre-whole genome duplication Saccharomycetes(Dietrich et al., 2004). It has becomean excellent system to study polarized hyphalmorphogenesis, septation and nuclear divisionbased on its ease of molecular genetic manipulation(Wendland and Walther, 2005). In particularRho-protein modules, polarisome and exocyst componentsand their role for polarity establishment,maintenance and tip-growth speed have been characterizedin more detail (Wendland and Philippsen,2001; Schmitz et al., 2006; Kohli et al., 2008).These studies indicated a strong dependence on theactin cytoskeleton for polarized hyphal growth inA. gossypii. A link between endocytosis and polarizedgrowth was established based on the analysisof the A. gossypii WASP-homolog WAL1 (Walther andWendland, 2004).AgSAC6 encodes the only Ashbya homolog offimbrin. It belongs to the conserved family ofactin-bundling proteins whose binding to actin ismediated by two pairs of calponin homology (CH)domains (Nakano et al., 2001; Klein et al., 2004).In Saccharomyces cerevisiae, SAC6 is required forthe establishment and maintenance of cell polarity,colocalizes with actin patches, and bundles actinfilaments (Bretscher, 1981; Adams et al., 1989).Similarly, in Schizosaccharomyces pombe the fimbrinFim1 colocalizes with actin patches and thecytokinetic ring (Nakano et al., 2001; Wu et al., 702001). Studies in Aspergillus nidulans showed that 71FimA-GFP localizes in a patch-like pattern at the 72cell cortex. Deletion of fimA resulted in polarity 73defects, particularly during spore germination, as 74well as in endocytosis defects (Upadhyay and Shaw, 752008). 76Here we present the functional analysis of 77AgSAC6. Our data suggest a genetic link between 78AgSAC6 and AgWAL1 based on the common mutant 79phenotypes regarding aberrant hyphal growth and 80endocytosis, which further underscores the link and 81the importance of endocytosis for polarized hyphal 82growth. 83Materials and methods 84Strains and media 85The A. gossypii leu2 and wal1 strains were 86used in this study. The leu2 mutant is referred 87to as wild type in the text. Independent sac6 88transformants were generated using Agleu2 as 89parental strain. Strains were grown in yeast 90extract—peptone—dextrose (AFM; 1% yeast extract, 911% peptone, 2% dextrose) with the addition of 92the antibiotic G418 at 200 g/ml for mutant 93selection when required. For sporulation an A. 94gossypii culture grown overnight in AFM was fur- 95ther incubated in minimal medium (1.7 g/l YNB 96w/o ammonium sulphate w/o amino acids, 0.69 g/l 97CSM, 20 g/l glucose, 2 g/l asparagine, and 1 g/l 98myo-inositol) until the cultures were thoroughly 99sporulated. Escherichia coli strain DH5 was used 100for plasmid propagation. 101Transformation and strain construction 102A. gossypii was transformed by electroporation as 103described (Wendland and Philippsen, 2001). A SAC6 104disruption cassette was cloned and used for trans- 105formation. Primers were obtained from biomers.net 106GmbH (Ulm, Germany). Sequences will be made 107available upon request. 108Plasmid constructs 109To generate a SAC6 disruption cassette a fusion 110PCR approach was utilized as described (Noble 111and Johnson, 2005). Flanking homology regions 112upstream and downstream of the SAC6-ORF were 113amplified using primers #3982 and #3983 (384 bp 1145 ′ -flank) and primers #4076 and #4077 (392 bp 3 ′ - 115flank). These PCR products were fused to the 116Please cite this article in press as: Jorde S, et al. The Ashbya gossypii fimbrin SAC6 is required forMICRES fast polarized 25382 1—9hyphal tip growth and endocytosis. Microbiol Res (2010), doi:10.1016/j.micres.2010.09.003

The Ashbya gossypii fimbrin SAC6 is required for fast polarized hyphal tip growth and endocytosis 3572.3AgCdc24-CHScCdc24-CHAgCyk1-CHScCyk1-CHAgScp1-CHScScp1-CHAgSac6-CH1ScSac6-CH1AgSac6-CH3ScSac6-CH3AgSac6-CH4ScSac6-CH4AgBim1-CHScBim1-CHAgSac6-CH2ScSac6-CH2500400300200Amino Acid Substitutions (x100)1000Fig. 1. Tree comparing the relation ship of A. gossypii and S. cerevisiae calponin homology domains. The CH-domainsof the indicated proteins were identified using blastp searches at (http://blast.ncbi.nlm.nih.gov). The individual CHdomainshave a length of app 100 aa. Trees were constructed using Clustal W of the Lasergene 8 software package.117118119120121122123124125126127128129130131132133134135136137138139140141142143144145kanMX4 selection marker in a second PCR. The deletioncassettes were cloned into pBluescript usingthe terminal restriction sites XhoI/XbaI providedwith the primers. Plasmid DNA was amplified in E.coli and the transformation cassettes were cut fromtheir plasmid backbones prior to transformationinto A. gossypii. Correct integration of the deletioncassette into target locus and absence of theSAC6 ORF in the homokaryotic sac6 null mutant wasconfirmed by PCR using standard diagnostic primers(Walther and Wendland, 2008).Microscopy and staining proceduresMicroscopy was performed with an Axio-Imagermicroscope (Zeiss, Jena and Göttingen, Germany)and images processed with Metamorph 7software (Molecular Devices Corp., Downington,PA). Fluorescent microscopy was performed withthe appropriate filter combinations for eitherrhodamine—phalloidin or FM4-64. Either singleimages were taken or a Z-stack generated from20 to 30 planes. Images were acquired witha MicroMax1024 CCD-camera (Princeton Instruments,Trenton, NJ). Staining procedures were asdescribed (Walther and Wendland, 2004).ResultsAnalysis of A. gossypii calponin homologydomain proteinsWithin the A. gossypii proteome five proteinscan be identified harboring calponin homol-ogy (CH)-actin-binding domains. Four of these 146proteins, AAR024wp/Bim1p; ADR388cp/Cdc24, 147ADR409wp/Scp1, and AFL150cp/Cyk1 contain 148only a single CH-domain at their N-termini. 149AGR069cp/Sac6 on the other hand is composed 150of 4 CH-domains. CH-domains 1 and 3 as well as 151CH2 and CH4 are more similar towards each other 152indicating that AgSac6 also forms two actin-binding 153domains comprising the CH-domains 1/2 and 3/4. 154Individual CH-domains of A. gossypii are very simi- 155lar to the homologous CH-domains of S. cerevisiae 156proteins ranging from 60% amino acids identity 157in the Scp1 proteins and 92% identity for CH1 of 158the Sac6 homologs (Fig. 1). Interestingly, ScSAC6 159contains an intron that is absent from AgSAC6. 160Deletion of the A gossypii SAC6 gene and 161the analysis of growth defects 162To establish a function of AgSAC6 we generated 163a deletion cassette with long flanking homology 164regions positioned such as to result in a com- 165plete ORF deletion. After transformation of A. 166gossypii heterokaryotic transformants are obtained 167that show wild-type phenotype as they carry both 168wild type and mutant nuclei with hyphal com- 169partments. Clonal selection starting from spores 170that are uninucleate yields homokaryotic mutants. 171Verification of correct gene deletion was done 172by PCR (Fig. 2A). Phenotypic characterization of 173the sac6 mutants using plate assays indicated a 174severe reduction in filamentous growth rate. At 17530—37 ◦ C sac6 mycelia exhibited slow growth and 176only formed compact small colonies as compared 177to the wild type (Fig. 2B). We amplified the AgSAC6 178Please cite this article in press as: Jorde S, et al. The Ashbya gossypii fimbrin SAC6 is required forMICRES fast polarized 25382 1—9hyphal tip growth and endocytosis. Microbiol Res (2010), doi:10.1016/j.micres.2010.09.003

4 S. Jorde et al.Fig. 2. Slow growth phenotype of the Agsac6 mutant. (A) Diagnostic PCR on a homokaryotic sac6 strain. The expectedsizes at the novel joints of the 5 ′ and 3 ′ ends generated after integration of the kanMX marker into SAC6 locus wereobtained. Conversely, no internal PCR fragments could be amplified from the SAC6 ORF in the null mutant. (B) Theleu2 (herein referred to as wild type) and sac6 strains were grown on AFM plates for 7 days at 30 ◦ C and 37 ◦ C prior tophotography. Reintegration of a plasmid bearing the SAC6 gene (C) is able to restore the wild type like growth of thesac6 mutant.179180181182183184185186187188189190from genomic DNA and ligated it into a plasmidcarrying the clonat resistance marker NAT5, whichconsists of the resistance gene ORF and the S. cerevisiaeTEF2 promoter and terminator regulatorysequences (Fig. 2C). Using this construct we couldcomplement the slow growth phenotype of the sac6mutant indicating that the observed phenotype wasdue to deletion of SAC6.The slow growth phenotype was similar to thatof the Agwal1 mutant strain. However, wal1 mutanthyphae are temperature sensitive and do not growat 37 ◦ C(Walther and Wendland, 2004). To deter-mine if there is a temperature sensitive phenotype 191also with the sac6 mutant we germinated sac6 192spores at different temperatures and compared 193the colony forming ability with the parental strain 194(Fig. 3). Colony forming frequency was found to be 195two orders of magnitude lower in the sac6 mutant 196compared to the wild type. This assay also showed 197that sac6 mutants do not generate mycelia when 198germinated at 37 ◦ C due to cell lysis after germina- 199tion (Fig. 3). 200To monitor this phenotype more closely we ger- 201minated sac6 spores over night at 30 ◦ C and then 202Please cite this article in press as: Jorde S, et al. The Ashbya gossypii fimbrin SAC6 is required forMICRES fast polarized 25382 1—9hyphal tip growth and endocytosis. Microbiol Res (2010), doi:10.1016/j.micres.2010.09.003

The Ashbya gossypii fimbrin SAC6 is required for fast polarized hyphal tip growth and endocytosis 5transferred the cells to 37 ◦ C. Growth at 37 ◦ C was 203not completely abolished (see also Fig. 2B), yet 204under these conditions two distinct mutant pheno- 205types became apparent. First, hyphal tips began to 206swell and also lysed and, second, subapical hyphal 207segments formed bulges and swellings (Fig. 4A—C). 208Both phenotypes are very similar to those observed 209with the wal1 mutant strain (Fig. 4D and E). 210The Agsac6 mutant reveals defects in the 211organization of the actin cytoskeleton 212Fig. 3. Germinated frequency of sac6 spores. Spores ofthe leu2 and sac6 strains were plated on AFM plates andresulting colonies were counted and photographed after 3days. Representative plates are shown. Note that the germinationfrequency of leu2 spores is app. 100-fold higherthan that of sac6 spores at 30 ◦ C.The actin cytoskeleton is polarized towards the 213hyphal tip in filamentous fungi. Thus in the A. 214gossypii wild type cortical actin patches are con- 215centrated at the hyphal tip and defects in this 216organization will result in isotropic growth lead- 217ing to swellings of hyphal tips (Wendland and 218Philippsen, 2001). Therefore, we compared the 219organization of the cortical actin cytoskeleton of 220the sac6 mutant with the parental strain and that 221of wal1 mutant hyphae. Cortical actin patches are 222found along the hyphae but cluster in the tips of 223wild type hyphae (Fig. 5A and B). It was previously 224observed that this clustering does not occur in wal1 225hyphae (Walther and Wendland, 2004). Here we find 226that also in the sac6 hyphae intense staining of actin 227is found at subapical positions, which appears to 228be more pronounced than in wal1 cells (Fig. 5C, 229D, and F). This defect in patch localization could 230be complemented upon reintroduction of AgSAC6 231(Fig. 5E). 232Fig. 4. Polar growth defects of sac6 hyphae. Mutant sac6 spores where germinated on microscope slides at 30 ◦ C o/nand then transferred to 37 ◦ C. Growth defects in the sac6 mutant lead to tip cell lysis (A and B). Reinitiation of growthat subapical parts of sac6 hyphae led to bulge formation (C). Similar phenotypes were observed for wal1 germinatedspores and hyphae (D and E). Lysis and bulge formation is marked by arrows. Bar is 100 m.Please cite this article in press as: Jorde S, et al. The Ashbya gossypii fimbrin SAC6 is required forMICRES fast polarized 25382 1—9hyphal tip growth and endocytosis. Microbiol Res (2010), doi:10.1016/j.micres.2010.09.003

6 S. Jorde et al.Fig. 5. Comparison of actin cytoskeletal organization in the leu2, sac6 and wal1 mutants. Representative fluorescentimages are shown of hyphae of the indicated strains stained with rhodamine-phalloidin. Prior to fixation cells weregrown at either 30 ◦ C (A, C, E, and F) or 37 ◦ C (B and D). Bar is 10 m. Septation was analyzed in leu2 (G and H) andsac6 (I and J) strains using calcofluor white staining of hyphae grown at 30 ◦ C.233234235236237238239240241242243244245246247248249250251252253254Using calcofluor white we stained sac6 hyphaefor septal sites. Previously we noted a septationdefect and the absence of chitin rich septa in thewal1 strain. Concomitantly to the lack of septation,wal1 hyphae show a strong sporulation defect. Bothof these phenotypes could not be observed in sac6hyphae as septation did not seem to be affected inthe sac6 mutant (Fig. 5G—J).Agsac6 hyphae show delayed endocytosisTo monitor endocytosis we stained hyphae of thewild type, the sac6 and wal1 mutants with thevital dye FM4-64 and used in vivo fluorescenceat different time points to monitor the uptake ofthe dye (Fig. 6). In the wild type FM4-64 uptakeis very rapid and early endosomes became visiblevery shortly after the stain was applied. After15 min apical compartments showed a large numberof endosomes. In subapical compartments largervacuoles are formed and they also became visibleas dye-containing endosomes were deliveredto the vacuoles. After 2 h the cell membranehad essentially been cleared from the dye andFM4-64 had been internalized. At this stage large 255endosomes in the tip compartment and large vac- 256uoles in subapical compartments can be found. In 257wild type the shape of large vacuoles is round or 258elongated and previously we have shown by time- 259lapse microscopy how vacuolar movement results 260in shape changes. In the wal1 mutant endocyto- 261sis is very much delayed particularly in subapical 262parts of the hyphae. And, strikingly, vacuoles are 263not actively moved about due to lack of Wal1 264(Walther and Wendland, 2004). In the sac6 mutant 265endocytosis is also very slow in subapical parts, 266although after a 2 h period FM4-64 was observed 267in endosomes and large vacuoles. Interestingly, the 268vacuoles of sac6 hyphae did show an elongated 269morphology indicating that active movement of 270vacuoles is not affected by the absence of Sac6 271(Fig. 6). 272Discussion 273Yeast fimbrin was isolated in a screen for suppres- 274sors of a temperature sensitive mutation in the 275Please cite this article in press as: Jorde S, et al. The Ashbya gossypii fimbrin SAC6 is required forMICRES fast polarized 25382 1—9hyphal tip growth and endocytosis. Microbiol Res (2010), doi:10.1016/j.micres.2010.09.003

The Ashbya gossypii fimbrin SAC6 is required for fast polarized hyphal tip growth and endocytosis 7Fig. 6. Endocytosis is delayed in the sac6 mutant. Hyphae of the indicated strains were grown in AFM medium andstained with FM4-64. At the indicated time points samples were subjected to fluorescence microscopy. Representativeimages of apical and subapical hyphal segments are shown. Scale bar 5 m.276277278279280281282283284285286287288289290291292293294295296297298299300301302actin gene and named SAC6 (Adams and Botstein,1989). SAC6 encodes an actin bundling protein thatinteracts with actin through a pair of actin bindingdomains consisting itself of pairs of CH-domains.This organization is unique as other members ofthe calponin protein family only possess a singleCH-domain. ScSac6 was shown to be involved inactin organization, endocytosis, and cell polarity(Drubin et al., 1988; Adams et al., 1991; Kubler andRiezman, 1993).We have analyzed the A. gossypii proteome byreciprocal blast searches to identify the set of CHdomainproteins and identified four proteins withsingle CH-domains in their N-termini, homologs ofBim1, Cdc24, Cyk1, and Scp1, as well as Sac6.AgCYK1 and AgCDC24 have been analyzed previously.AgCYK1 is essential for actin ring formation atseptal sites. Deletion of AgCYK1 abolishes septationand the accumulation of chitin at septal sites withoutinhibiting polarized hyphal growth (Wendlandand Philippsen, 2002). AgCDC24 encodes the guaninenucleotide exchange factor for AgCdc42 andis essential for polarity establishment. Deletion ofAgCDC24 blocks germ tube formation at the germcell stage and is thus an essential gene (Wendlandand Philippsen, 2001). However, for both proteinsthe actual contribution of the CH-domains for theirfunction has not been elucidated. The BIM1 and 303SCP1 genes have not been analyzed in A. gossypii 304yet. 305We initiated this study to identify domains that 306when fused to GFP could be used to visualize 307actin cables in A. gossypii. However, using SAC6 308actin-binding domains resulted in unphysiological 309actin bundling (Hebecker and Wendland, our unpub- 310lished results). ABP140, which encodes another 311actin binding protein, has been used to analyze 312actin cables and movement of mitochondria along 313actin cables in yeast cells in vivo (Yang and Pon, 3142002; Fehrenbacher et al., 2004). Based on the 315a 17 amino acid peptide of this protein, termed 316lifeact, in vivo imaging of the actin cytoskeleton 317has overcome several technical limitations and has 318since been used in fungi and mice (Riedl et al., 3192008, 2010; Delgado-Alvarez et al., 2010). There- 320fore, introduction of lifeact also in A. gossypii will 321further our understanding of transport processes 322and organelle movement in this filamentous fungus. 323Deletion of AgSAC6 resulted in defects in endocy- 324tosis. The sac6 hyphae showed an intense staining of 325actin patches in subapical hyphal regions. In S. cere- 326visiae sac6 mutants were shown to abolish the fast 327phase of Abp1 patch movement (Kaksonen et al., 3282005). Thus the aberrant actin patch accumulation 329Please cite this article in press as: Jorde S, et al. The Ashbya gossypii fimbrin SAC6 is required forMICRES fast polarized 25382 1—9hyphal tip growth and endocytosis. Microbiol Res (2010), doi:10.1016/j.micres.2010.09.003

8 S. Jorde et al.330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380in A. gossypii could indicate delays in endocytosisat subapical sites similarly as found in S. cerevisiae.Endocytosis, measured by uptake and delivery ofFM4-64 to endosomes and vacuoles, was delayedin sac6 hyphae. This phenotype was less drastic asthe endocytosis delay observed in the wal1 mutant.However, polarized growth was similarly affectedin both mutants. Agwal1 hyphae are temperaturesensitive and cease growth at 37 ◦ C. At 30 ◦ C wal1hyphae form aberrant bulges in subapical regionsand show increased cell lysis as could be observedfor sac6 hyphae at elevated temperatures. It iscurrently unknown what triggers the aberrant cellgrowth at the sites of bulge formation. Lateralbranching events or remedial cell wall biosynthesiscould cause initiation of polarized growth at suchsites.Interestingly, there was a striking differencein vacuolar morphology when comparing the sac6hyphae with the wal1 mutant. In wal1 hyphaevacuoles are round shaped and largely immobileindicating that force generation via Wal1 activatedArp2/3 dependent actin polymerization is requiredfor this movement (Walther and Wendland, 2004).In sac6 hyphae vacuoles had wild type like tubularstructures suggesting that Sac6 is not requiredfor vacuolar motility in A. gossypii. This suggeststhat actin filament bundling via Sac6 is importantfor polarized hyphal growth and endocytosis butis dispensable for septation or vacuolar organellemovement.AcknowledgementThis study was co-funded by the EU-Marie CurieResearch Training Network “Penelope”.ReferencesAdams AE, Botstein D. Dominant suppressors of yeastactin mutations that are reciprocally suppressed.Genetics 1989;121:675—83.Adams AE, Botstein D, Drubin DG. A yeast actin-bindingprotein is encoded by SAC6, a gene found by suppressionof an actin mutation. Science 1989;243:231—3.Adams AE, Botstein D, Drubin DG. Requirement of yeastfimbrin for actin organization and morphogenesis invivo. Nature 1991;354:404—8.Aghamohammadzadeh S, Ayscough KR. Differentialrequirements for actin during yeast and mammalianendocytosis. Nat Cell Biol 2009;11:1039—42.Araujo-Bazan L, Penalva MA, Espeso EA. Preferentiallocalization of the endocytic internalization machineryto hyphal tips underlies polarization of the actincytoskeleton in Aspergillus nidulans. Mol Microbiol 3812008;67:891—905. 382Bretscher A. Fimbrin is a cytoskeletal protein that 383crosslinks F-actin in vitro. Proc Natl Acad Sci U S A 3841981;78:6849—53. 385Delgado-Alvarez DL, Callejas-Negrete OA, Gomez N, et 386al. Visualization of F-actin localization and dynamics 387with live cell markers in Neurospora crassa. Fungal 388Genet Biol 2010;47:573—86. 389Dietrich FS, Voegeli S, Brachat S, et al. The Ash- 390bya gossypii genome as a tool for mapping the 391ancient Saccharomyces cerevisiae genome. Science 3922004;304:304—7. 393Drubin DG, Miller KG, Botstein D. Yeast actin-binding pro- 394teins: evidence for a role in morphogenesis. J Cell Biol 3951988;107:2551—61. 396Fehrenbacher KL, Yang HC, Gay AC, Huckaba TM, 397Pon LA. Live cell imaging of mitochondrial move- 398ment along actin cables in budding yeast. Curr Biol 3992004;14:1996—2004. 400Fuchs U, Hause G, Schuchardt I, Steinberg G. Endo- 401cytosis is essential for pathogenic development in 402the corn smut fungus Ustilago maydis. Plant Cell 4032006;18:2066—81. 404Goley ED, Rammohan A, Znameroski EA, Firat-Karalar 405EN, Sept D, Welch MD. An actin-filament-binding 406interface on the Arp2/3 complex is critical for nucle- 407ation and branch stability. Proc Natl Acad Sci U S A 4082010;107:8159—64. 409Higgs HN, Pollard TD. Regulation of actin polymerization 410by Arp2/3 complex and WASp/Scar proteins. J Biol 411Chem 1999;274:32531—4. 412Kaksonen M, Toret CP, Drubin DG. A modular design for the 413clathrin- and actin-mediated endocytosis machinery. 414Cell 2005;123:305—20. 415Klein MG, Shi W, Ramagopal U, et al. Structure of 416the actin crosslinking core of fimbrin. Structure 4172004;12:999—1013. 418Kohli M, Galati V, Boudier K, Roberson RW, Philippsen P. 419Growth-speed-correlated localization of exocyst and 420polarisome components in growth zones of Ashbya 421gossypii hyphal tips. J Cell Sci 2008;121:3878—89. 422Kubler E, Riezman H. Actin fimbrin are required for the 423internalization step of endocytosis in yeast. EMBO J 4241993;12:2855—62. 425Machesky LM, Gould KL. The Arp2/3 complex: a mul- 426tifunctional actin organizer. Curr Opin Cell Biol 4271999;11:117—21. 428Merrifield CJ. Seeing is believing: imaging actin dynam- 429ics at single sites of endocytosis. Trends Cell Biol 4302004;14:352—8. 431Munn AL. Molecular requirements for the internalisation 432step of endocytosis: insights from yeast. Biochim Bio- 433phys Acta 2001;1535:236—57. 434Nakano K, Satoh K, Morimatsu A, Ohnuma M, Mabuchi 435I. Interactions among a fimbrin, a capping protein, 436and an actin-depolymerizing factor in organization 437of the fission yeast actin cytoskeleton. Mol Biol Cell 4382001;12:3515—26. 439Noble SM, Johnson AD. Strains and strategies for large- 440scale gene deletion studies of the diploid human 441Please cite this article in press as: Jorde S, et al. The Ashbya gossypii fimbrin SAC6 is required forMICRES fast polarized 25382 1—9hyphal tip growth and endocytosis. Microbiol Res (2010), doi:10.1016/j.micres.2010.09.003

The Ashbya gossypii fimbrin SAC6 is required for fast polarized hyphal tip growth and endocytosis 9442443444445446447448449450451452453454455456457458459460461462463464465fungal pathogen Candida albicans. Eukaryot Cell2005;4:298—309.Pruyne D, Bretscher A. Polarization of cell growth inyeast. J Cell Sci 2000;113(Pt 4):571—85.Riedl J, Crevenna AH, Kessenbrock K, et al. Lifeact:a versatile marker to visualize F-actin. Nat Methods2008;5:605—7.Riedl J, Flynn KC, Raducanu A, et al. Lifeactmice for studying F-actin dynamics. Nat Methods2010;7:168—9.Schmitz HP, Kaufmann A, Kohli M, Laissue PP, PhilippsenP. From function to shape: a novel role of a formin inmorphogenesis of the fungus Ashbya gossypii. Mol BiolCell 2006;17:130—45.Taheri-Talesh N, Horio T, Araujo-Bazan L, et al. The tipgrowth apparatus of Aspergillus nidulans. Mol Biol Cell2008;19:1439—49.Upadhyay S, Shaw BD. The role of actin, fimbrin and endocytosisin growth of hyphae in Aspergillus nidulans.Mol Microbiol 2008;68:690—705.Voigt B, Timmers AC, Samaj J, et al. Actin-based motilityof endosomes is linked to the polar tip growth of roothairs. Eur J Cell Biol 2005;84:609—21.Walther A, Wendland J. Apical localization of actinpatches and vacuolar dynamics in Ashbya gossypiidepend on the WASP homolog Wal1p. J Cell Sci 4662004;117:4947—58. 467Walther A, Wendland J. PCR-based gene targeting in Can- 468dida albicans. Nat Protoc 2008;3:1414—21. 469Wendland J, Philippsen P. Cell polarity and hyphal mor- 470phogenesis are controlled by multiple rho-protein 471modules in the filamentous ascomycete Ashbya 472gossypii. Genetics 2001;157:601—10. 473Wendland J, Philippsen P. An IQGAP-related protein, 474encoded by AgCYK1, is required for septation in the 475filamentous fungus Ashbya gossypii. Fungal Genet Biol 4762002;37:81—8. 477Wendland J, Walther A. Ashbya gossypii: a model 478for fungal developmental biology. Nat Rev Microbiol 4792005;3:421—9. 480Wu JQ, Bahler J, Pringle JR. Roles of a fimbrin and an 481alpha-actinin-like protein in fission yeast cell polar- 482ization and cytokinesis. Mol Biol Cell 2001;12:1061— 48377. 484Yang HC, Pon LA. Actin cable dynamics in budding yeast. 485Proc Natl Acad Sci USA2002;99:751—6. 486Yarar D, Waterman-Storer CM, Schmid SL. A dynamic actin 487cytoskeleton functions at multiple stages of clathrin- 488mediated endocytosis. Mol Biol Cell 2005;16:964— 48975. 490Please cite this article in press as: Jorde S, et al. The Ashbya gossypii fimbrin SAC6 is required forMICRES fast polarized 25382 1—9hyphal tip growth and endocytosis. Microbiol Res (2010), doi:10.1016/j.micres.2010.09.003

%paper no. mic041707charlesworth ref: mic041707&Cell and Molecular Biology of MicrobesMicrobiology (2010), 156, 000–000DOI 10.1099/mic.0.041707-0Candida albicans Vrp1 is required for polarizedmorphogenesis and interacts with Wal1 and Myo5Nicole Borth, 1,2 3 Andrea Walther, 1,2 Patrick Reijnst, 1 Sigyn Jorde, 1Yvonne Schaub 2 4 and Jürgen Wendland 1,2CorrespondenceJürgen Wendlandjww@crc.dk1 Carlsberg Laboratory, Yeast Biology, Gamle Carlsberg Vej 10, DK-2500 Valby, Denmark2 Junior Research Group: Growth Control of Fungal Pathogens, Leibniz Institute for Natural ProductResearch and Infection Biology – Hans Knöll Institute and Department of Microbiology,Friedrich Schiller University, D-07745 Jena, GermanyReceived 19 May 2010Revised 18 July 2010Accepted 19 July 2010Recently, a link between endocytosis and hyphal morphogenesis has been identified in Candidaalbicans via the Wiskott–Aldrich syndrome gene homologue WAL1. To get a more detailedmechanistic understanding of this link we have investigated a potentially conserved interactionbetween Wal1 and the C. albicans WASP-interacting protein (WIP)-homologue encoded byVRP1. Deletion of both alleles of VRP1 results in strong hyphal growth defects under seruminducing conditions but filamentation can be observed on spider medium. Mutant vrp1 cells showa delay in endocytosis – measured as the uptake and delivery of the lipophilic dye FM4-64 intosmall endocytic vesicles – compared to the wild-type. Vacuolar morphology was found to befragmented in a subset of cells and the cortical actin cytoskeleton was depolarized in vrp1daughter cells. The morphology of the vrp1 null mutant could be complemented by reintegration ofthe wild-type VRP1 gene at the BUD3 locus. Using the yeast two-hybrid system we coulddemonstrate an interaction between the C-terminal part of Vrp1 and the N-terminal part of Wal1,which contains the WH1 domain. Furthermore, we found that Myo5 has several potentialinteraction sites on Vrp1. This suggests that a Wal1–Vrp1–Myo5 complex plays an important rolein endocytosis and the polarized localization of the cortical actin cytoskeleton to promote polarizedhyphal growth in C. albicans.INTRODUCTIONCandida albicans is a pathogenic yeast that can respond tocertain environmental cues by forming hyphal filaments.This morphogenetic switch is regarded as one of severalattributes that enable C. albicans to cause disease (Sudberyet al., 2004; Whiteway & Oberholzer, 2004; Whiteway &Bachewich, 2007). Hyphal growth is an extreme form ofpolarized morphogenesis that requires constant delivery ofvesicles to support tip growth and remodelling of the cellwall at the tip. The actin cytoskeleton plays an importantrole by providing tracks for the delivery of vesicles to thetip along actin cables, and actin patches at sites ofendocytosis (Pruyne & Bretscher, 2000; Kaksonen et al.,2005). A balance between secretion and endocytosis is also3Present address: Cell and Molecular Biology, Leibniz Institute forNatural Product Research and Infection Biology – Hans Knöll Institute,D-07745 Jena, Germany.4Present address: Leibniz Institute for Age Research – Fritz LipmannInstitute, D-07745 Jena, Germany.Abbreviations: SH3, Src homology domain 3; WASP, Wiskott–Aldrichsyndrome protein; WH2, WASP homology 2; WIP, WASP-interactingprotein.important for the maintenance of polarized morphogenesis,although a mechanistic link has not yet been established(Aghamohammadzadeh & Ayscough, 2009). Twogenes that play a crucial role in endocytosis in C. albicansare CaMYO5, encoding myosin I, and the Wiskott–Aldrichsyndrome homologue CaWAL1 (Oberholzer et al., 2004;Walther & Wendland, 2004). In Saccharomyces cerevisiae,the corresponding homologues ScMYO3/5 and LAS17 havebeen shown to activate the Arp2/3 complex, promotingactin polarization at sites of endocytosis (Evangelista et al.,2000; Machesky, 2000; D’Agostino & Goode, 2005).Deletion of CaMYO5 leads to viable mutant strains thatcannot undergo hyphal development. Yeast cells of Camyo5mutants show depolarization of the actin cytoskeleton,which also affects their budding pattern (Oberholzer et al.,2002, 2004). Similarly, deletion of CaWAL1 results inmutant strains that are unable to generate hyphal filaments.During yeast growth of these mutants, depolarization of ;the actin cytoskeleton leads to the formation of round cellsthat show an increase in random budding. Additionally,loss of CaWAL1 leads to defects in the endocytosis of thelipophilic dye FM4-64 as well as defects in vacuolar fusion.Fragmented vacuoles have been observed in other mutant041707 G 2010 SGM Printed in Great Britain 1

%paper no. mic041707charlesworth ref: mic041707&N. Borth and othersstrains, e.g. vac1 or vps11. These strains were also shown tobe defective in hyphal morphogenesis (Palmer et al., 2003;Franke et al., 2006). Characteristically, during hyphalgrowth large vacuoles are formed in the germ cell and inthe rear parts of the hyphal filaments. An unequal distributionof vacuoles was also shown to influence thetiming of branch emergence (Veses et al., 2008); however,fragmented vacuoles, per se, do not abolish polarizedmorphogenesis, which was recently also shown in a Caboi2mutant (Reijnst et al., 2010).The functional overlap of C. albicans Myo5 and Wal1 andtheir rather similar mutant phenotypes suggests that thetwo proteins can function in a complex in C. albicans. Inmammalian cells, Wiskott–Aldrich syndrome protein(WASP) was shown to interact with the WASP-interactingprotein (WIP) (Ramesh et al., 1997; Thrasher & Burns,2010). WIP suppresses growth defects of the S. cerevisiaeend5/vrp1 mutant (Vaduva et al., 1999). ScEnd5/Vrp1 is avery proline-rich protein that is involved in cytoskeletalorganization and can interact with both Las17 and Myo5(Anderson et al., 1998; Evangelista et al., 2000; Munn &Thanabalu, 2009). The temperature sensitivity and loss ofviability of an end5-1/vrp1 mutant can be suppressed by theoverexpression of ScLAS17 (Naqvi et al., 1998). And finally,loss of ScEnd5/Vrp1 results in severe defects in cytokinesisand Hof1 cannot be recruited to the bud neck (Ren et al.,2005).Here we describe the analysis of the C. albicans VRP1homologue. The mutant strain shows defects in hyphalgrowth, endocytosis, organization of the actin cytoskeletonand budding pattern similar to, but less pronounced than,the wal1 and myo5 mutant strains. Two-hybrid studies in S.cerevisiae showed that Vrp1 interacts strongly with the N-terminal domain of Wal1 and also with Myo5. The datasuggest that a Wal1–Vrp1–Myo5 complex is crucial forendocytosis and polarized morphogenesis in C. albicans.METHODSStrains and media. C. albicans strain SN148 (Noble & Johnson,2005) was used to generate the vrp1 heterozygous and homozygousmutant strains. For the yeast two-hybrid experiment, the following< strains were used: PJ69-4a (MATa trp1-901 leu2-3,112 ura3-52 his3-200 gal4D gal80D LYS2 ::GAL1-HIS3 GAL2p-ADE2 met2 ::GAL7-lacZ) and PJ69-4alpha (MATa trp1-901 leu2-3,112 ura3-52 his3-200gal4D gal80D LYS2 ::GAL1-HIS3 GAL2p-ADE2 met2 ::GAL7-lacZ).Strains were grown either in yeast extract/peptone/dextrose medium[YPD; 1 % yeast extract, 2 % peptone, 2 % dextrose (glucose)], or in= minimal synthetic defined media [6.7 g l 21 yeast nitrogen base (YNB)with ammonium sulphate and without amino acids, 20 g l 21 glucose]supplemented with 0.69 g l 21 complete supplement mixture (CSM),or with the addition of required amino acids and uridine. Hyphalformation was induced with 10 % serum at 37 uC or by incubation onSpider medium. Escherichia coli strain DH5a served as a host forplasmid propagation.Transformation and strain construction. S. cerevisiae and C.albicans were transformed using the lithium acetate procedure(Walther & Wendland, 2003; Gietz & Schiestl, 2007). Independenthomozygous mutant strains were constructed and verified followingstandard PCR-based gene targeting methods based on the use of pFAplasmids for cassette generation (Walther & Wendland, 2008).Deleting both ORFs of VRP1 by sequential transformation ofSN148 with PCR-generated cassettes resulted in the heterozygousstrains CAB9 (VRP1/vrp1 ::ARG4) and CAB10 (VRP1/vrp1 ::URA3),and then the homozygous strains CAB12 (vrp1 ::ARG4/vrp1 ::CdHIS1) and CAB13 (vrp1 ::URA3/vrp1 ::CdHIS1). To complementthe Dvrp1 phenotype, VRP1 was amplified from genomic DNA andligated in to cloning vector pDrive, generating #C597. The insert wascloned in #C873, which contains the BUD3 locus for integration andCmLEU2 as selectable marker, using SalI and BamHI restriction sites.This generated #C598. The Dvrp1 homozygous mutant strain CAB13was transformed with SpeI-linearized #C598, generating CAP225.Strain CAT41 was generated by targeting a GFP-HIS1 cassette to theC. albicans TEF1 locus. All primers were obtained from biomers.netand their sequences will be made available upon request . >Plasmid constructs. For the yeast two-hybrid experiments, freelyreplicating plasmids were generated using pGAD424 and pGBT9(Clontech) as backbones. These plasmids contained the Gal4-transcription-factor-activation domain or the Gal4-DNA-bindingdomain, respectively. Restriction fragments of WAL1, VRP1 and theregion encompassing the SH3 domain of MYO5 were amplified fromgenomic DNA or plasmid clones and cloned into the correspondingrestriction sites of pGAD424 or pGBT9. Correct cloning was verifiedby sequencing (Eurofins MWG Operon).Microscopy and staining procedures. Microscopic analyses weredone with an Axio-Imager microscope (Zeiss) using Metamorph 7software tools (Molecular Devices) to drive the automated imageacquisitionprocedures. Images were acquired with a MicroMax1024CCD camera (Princeton Instruments). Fluorescence microscopy wasperformed using the appropriate filter combinations for FM4-64imaging and actin staining as described previously (Walther &Wendland, 2004; Martin et al., 2005). Samples were analysed bygenerating either single images or Z-stacks of up to 20 images thatwere processed into single-plane projections using Metamorphsoftware.Yeast two-hybrid analysis. S. cerevisiae was transformed with twoplasmids expressing constructs fused to either the Gal4-DNA-bindingdomain, based on plasmid pGBT9, or the Gal4-activation domain,based on plasmid pGAD424. Transformants were grown on mediaselecting for the maintenance of both plasmids (2Trp 2Leu). Whitecolonies revealed an interaction of the two expressed fusion proteins,which results in the expression of the ADE2 reporter gene, whereasred colonies appeared when the ADE2-reporter could not beactivated. For quantitative analysis, liquid-culture b-galactosidaseassays were performed. To this end, strains were incubated overnightat 30 uC. Cells were harvested by centrifugation, protein extracts wereprepared using a liquid nitrogen/glass-bead method and theconversion of ONPG (o-nitrophenyl b-D-galactopyranoside) wasmeasured photometrically (Rose & Botstein, 1983).RESULTSSequence comparisonsThe C. albicans homologue of S. cerevisiae END5/VRP1 hasbeen identified as orf19.2190. C. albicans VRP1 encodes avery proline-rich protein of 664 aa, of which 154 residuesare proline. Sequence comparisons with other fungalhomologues were done using the CLUSTAL W alignment2 Microbiology 156

%paper no. mic041707charlesworth ref: mic041707&C. albicans VRP1tool (Fig. 1). The N-terminal region of CaVrp1p contains aproline stretch present in most fungi and only annotated tobe absent in Ashbya gossypii. Reinspection of the VRP1-locus in A. gossypii, however, indicates that there is apolyproline region upstream of the annotated start codon.Furthermore, a Vrp1 homologue in the closely relatedspecies Eremothecium cymbalariae also contains thispolyproline region at the N terminus of Ecym_Vrp1 (ourunpublished results). Downstream of the polyprolineregion in Vrp1, two putative WASP homology 2 (WH2)domains are located. Here, the filamentous ascomyceteNeurospora crassa lacks the second putative WH2 domain.In S. cerevisiae, a short region after the second WH2domain has been characterized as a docking site for Hof1(Ren et al., 2005). This Hof1-trap (HOT)-domain seems tobe rather specific for S. cerevisiae as it is not found in theother fungal species analysed (Fig. 1a). The central part ofVrp1 orthologues exhibits only a low degree of amino acidsequence conservation, not regarding the many prolinerichstretches, whereas the C-terminal regions in fungalVrp1 proteins show better conservation. This domain hasbeen characterized as the Las17p-binding domain (Naqviet al., 1998; Madania et al., 1999; Fig. 1b).Generation of C. albicans vrp1 mutant strainsTo delete both alleles of C. albicans VRP1, a PCR-basedgene targeting approach was applied (Walther &Wendland, 2008). Initially, independent heterozygousmutant strains were generated in which the ORF of oneallele of VRP1 was deleted by either the C. albicans ARG4or URA3 marker gene. To generate homozygous mutantstrains based on these heterozygous strains, the remainingcopy of VRP1 was deleted using the Candida dubliniensisHIS1 gene. Verification of correct gene targeting and theabsence of the VRP1 ORF in the homozygous mutantstrains CAB11 and CAB13 was done by diagnostic PCR andreintegration of VRP1 at the BUD3 locus was used forcomplementation.Phenotypic assay of growth morphology ofmutant strainsThe heterozygous and homozygous VRP1-deletion strainswere compared to the SC5314 wild-type strain, the SN148strain used as a host for transformation and a wal1 mutantstrain deleted for the C. albicans homologue of the humanWASP, described previously (Walther & Wendland, 2004).Hyphal induction was tested on Spider medium, whichcontains mannitol as the primary carbon source. The wildtypeshowed strong filamentation at the edge of the colony,whereas the wal1 strain was afilamentous. The SN148precursor strain also showed a strong increase in colonywrinkling, which was also found in the heterozygous VRP1/vrp1 strain. The homozygous vrp1 strain did not show thiscolony-wrinkling phenotype; however, the colony edges ofthe mutant vrp1 strain did show invasive filamentousgrowth (Fig. 2a). Hyphal induction in liquid media wasdone using serum as an inducing cue. Here, the wild-typeshowed abundant filamentation. SN148 showed slightlyless filamentation (due to the ura3 deletion) and wal1 againshowed no hyphal formation. The heterozygous VRP1/vrp1strain showed filamentation similar to that of the wildtype,whereas the vrp1 mutant strain showed a strongreduction in filamentation and produced instead a largenumber of pseudohyphal cells and new yeast cells (Fig. 2a).The distribution of cell types after hyphal induction in thevarious strains used was quantified by counting .100 cellsfor each strain (Fig. 2c ). These analyses indicated a strong ?defect in hyphal formation of the vrp1 mutant, which was,Fig. 1. Alignment of fungal Vrp1 homologues.Amino acid residues corresponding to themajority of analysed sequences are shaded.(a) The N-terminal actin-binding sequencesare boxed in all Vrp1 proteins. S. cerevisiaeVrp1 harbours non-conserved regions with theconsensus sequence ‘PxPSS’ (boxed) thatwere shown to interact with ScHof1. (b)Alignment of the Vrp1 C-terminal regionsharbouring the conserved Las17-bindingdomain. Protein information was obtained fromthe following sources: A. gossypii, the AshbyaGenome Database (http://agd.vital-it.ch/index.html); C. albicans, the Candida GenomeDatabase (http://www.candidagenome.org);@Kluyveromyces lactis, XP_451805; Neurosporacrassa, XP_963859; and S. cerevisiae,AAB67263.http://mic.sgmjournals.org 3

%paper no. mic041707charlesworth ref: mic041707&N. Borth and othersFig. 2. Characterization of the growth defectsof the vrp1 mutant SN148. (a) The indicatedstrains were grown on Spider plates for 5 daysat 37 6C prior to photography (top row).Microscopic images of the colony edges showthe degree of filamentation (middle row).Hyphal formation in liquid medium wasinduced by using 10 % serum. Images weretaken after 6 h of induction (bottom row). (b)Cells induced by serum were counted after 6 hand classified according to their morphology.(c) Complementation of the vrp1 filamentationdefect by reintegration of VRP1 at the C.albicans BUD3 locus.however, slightly less severe than in the wal1 mutant. Thisresult is in line with the filamentation assay on Spidermedium. To demonstrate that these filamentation defectsare solely due to the deletion of the VRP1 gene, wereintegrated the VRP1 gene at the BUD3 locus in a vrp1/vrp1 mutant strain (see Methods). This reintegrant wasphenotypically like the wild-type, for example, whenassayed for germ tube production (Fig. 2b).Analysis of the actin cytoskeleton in the vrp1mutantHyphal growth defects may be associated with an alteredorganization of the actin cytoskeleton. Therefore, we usedrhodamine-phalloidin staining of fixed cells to analyse thedistribution and polarization of the cortical actin cytoskeleton.Wild-type cells show a polarization of cortical actin inthe emerging bud and at the hyphal tip. The actincytoskeleton of the wal1 mutant was shown to be largelydepolarized during all growth stages. In the vrp1 mutantsuch a depolarization could also be observed in yeast andpseudohyphal cells. Remarkably, in both yeast and hyphalstages the apical growth region showed more intensestaining indicating an accumulation of actin at sites ofpolarized growth (Fig. 3a). Analysis of the budding patternvia fluorescence microscopy of the bud scars showed thatthe vrp1 mutant is able to generate a bipolar buddingpattern, as found in the wild-type (Fig. 3b).VRP1 mutants show defects in vacuolar fusionand endocytosisAltered polarization of the actin cytoskeleton may also affectendocytosis. To study vacuolar morphology and endocytosis4 Microbiology 156

%paper no. mic041707charlesworth ref: mic041707&C. albicans VRP1Two-hybrid analyses reveal interaction of Vrp1with Myo5 and Wal1In S. cerevisiae Vrp1 was shown to interact with Srchomology domain 3 (SH3) of the yeast type I myosinMyo5p and the WASP homologue Las17 (Evangelista et al.,2000). To determine if CaVrp1 interacts with both Las17and Myo5 we performed yeast two-hybrid analyses (Fig. 6).In this assay we found that the C-terminal part of Vrp1interacts strongly with the N-terminal part of Wal1containing the WH1-B domains. Interaction of Vrp1 withthe full-length Wal1 protein was somewhat weaker.Highest b-galactosidase activity was obtained with a Wal1fragment in which the central part of WAL1, encodingseveral proline-rich regions, was removed for the assay. Toassay the interaction of the myosin I SH3 domain withVrp1 we used two fragments containing the N and Ctermini of Vrp1 and a fragment containing only the SH3-domain of Myo5. Here, the Myo5 SH3-domain interactedstrongly with the C-terminal part of Vrp1 (Fig. 6).Fig. 3. Analysis of the actin cytoskeleton and budding pattern. (a)Strains were grown to form yeast and hyphal stages, fixed byformaldehyde and stained using rhodamine-phalloidine. Bar,10 mm. (b) Representative images of calcofluor-stained cells.we employed the lipophilic dye FM4-64. Using time-lapsemicroscopy we compared, at the same time, uptake of FM4-64 between the vrp1 mutant and a wild-type strainexpressing cytoplasmic GFP. Within 20 min the dye hadbeen taken up via endocytosis and delivered to the vacuolein the wild-type. This staining of wild-type vacuoles furtherincreased over time. Compared to this the vrp1 mutantshowed a delayed uptake and only after more than 1 h didsmall endocytic vesicles, and possibly vacuoles, becamestained (Fig. 4). Quantification of the vacuoles showed aslight increase in the vrp1 mutant compared to the wildtype.This vrp1 phenotype is thus somewhat intermediatebetween the wal1 mutant and the wild-type (Fig. 5).DISCUSSIONIn this report we have characterized the function of the C.albicans VRP1 gene in the polarized morphogenesis andendocytosis of this dimorphic human pathogen. Cellpolarization in C. albicans is important for budding,filamentation and mating. The establishment of cellpolarity occurs either due to intrinsic factors (duringbudding) or in response to environmental stimuli (duringfilamentation and mating) (Whiteway & Bachewich, 2007).One result of this polarity is the polarized organization ofthe actin cytoskeleton, which results in the apicalpositioning of cortical actin patches and the generationof actin cables emanating from the cell apex (Smith et al.,2001). Generation of actin cables from the emerging bud orthe tip of the hyphae has been fairly well characterized. Acascade from locally activated Rho-type GTPases, mostnotably Cdc42, triggers downstream effector genes, such asthe formin Bni1, which nucleates actin filaments(Evangelista et al., 2002; Sagot et al., 2002). Bni1 is partof a complex termed the polarisome, which in S. cerevisiaealso contains Pea2, Spa2 and Bud6 (Sheu et al., 1998).Clustered assembly of actin patches occurs at sites ofpolarized growth. Mutants of S. cerevisiae and C. albicansthat are affected in the position of actin patches, e.g. in theLAS17/WAL1 or MYO3/5 genes, show defects in polarizedgrowth (Li, 1997; Lechler et al., 2000; Oberholzer et al.,2002; Walther & Wendland, 2004). Las17 and Myo3/5 havebeen shown to stimulate actin filament formation via theArp2/3 complex (Lechler et al., 2000). Mutants affected inthese genes show defects in the assembly and organizationof the actin cytoskeleton and since the actin cytoskeleton isessential for endocytosis in S. cerevisiae they also showdefects in clathrin-mediated endocytosis (Munn, 2001;Kaksonen et al., 2003). Most of the proteins known to beinvolved in endocytosis co-localize with actin patches.http://mic.sgmjournals.org 5

%paper no. mic041707charlesworth ref: mic041707&N. Borth and othersFig. 4. Time-lapse analysis of endocytosis ofthe lipophilic dye FM4-64. Wild-type cellscarrying a cytoplasmic GFP label (on the rightside of each panel) were mixed with vrp1/vrp1cells (on the left side of each panel).Microscopy slides with wells were filled with0.75 ml 0.5¾ YPD and 0.75 ml 3.4 % agarose.To this mixture 1 ml FM4-64 (200 mg ml ”1 inDMSO) was added. Image acquisition started10 min after preparation of the slide for a durationof 3 h with a frequency of 1 image min ”1 .Selected frames are shown at the indicatedtime points, starting with a GFP imageidentifying the wild-type cells followed by abright-field differential interference contrast(DIC) image of all cells. Bar, 10 mm.AFig. 5. Analysis of vacuolar morphology. Strains were grownovernight and then stained with FM4-64 for 2 h prior tophotography. Cells were counted and characterized according tothe number of vacuoles they contained. Representative images ofcells displaying one, two to three, or more than four vacuoles areshown.Thus actin patches are sites of endocytosis (Kaksonen et al.,2005).Deletion of S. cerevisiae VRP1 results in temperaturesensitivity and the depolarization of actin patches. Specifically,actin patches do not cluster in emerging buds(Lambert et al., 2007). We have also observed adepolarization of actin patches in both mother anddaughter cells of the C. albicans vrp1 mutant. Furthermore,hyphal morphogenesis in the vrp1 strains wasinhibited – although not abolished as in wal1 cells.Nevertheless, during growth of vrp1 germ tubes, actinappeared to accumulate in a cap-like structure.Our two-hybrid analysis suggests that a Wal1–Vrp1–Myo5complex may be formed in C. albicans similar to thatidentified in S. cerevisiae (Evangelista et al., 2000). Thiscould provide an explanation of the mechanism responsiblefor the similar phenotypes of the wal1 and myo5 mutantsobserved previously. Both of these genes are activators of theArp2/3 complex, and loss of either of these genes may bemore detrimental to cells than loss of VRP1. Consequently,the observed defects in endocytosis and vacuole formationwere less severe in the vrp1 strains compared to wal1.Interestingly, S. cerevisiae Vrp1 contains a region characterizedas the Hof1-trap domain, which is essential for bindingthe Hof1-SH3 domain (Ren et al., 2005). A Hof1-trapdomain could not be identified in C. albicans Vrp1. Ourattempts to identify a two-hybrid interaction of the C.albicans SH3-domain of Hof1 with Vrp1 were unsuccessful,which may suggest that this interaction is occurring either6 Microbiology 156

%paper no. mic041707charlesworth ref: mic041707&C. albicans VRP1COLOURFIGUREFig. 6. Two-hybrid analysis. PCR fragmentscarrying different domains were combined withthe DNA-binding domain (BD) or the DNAactivationdomain (AD) as indicated in theupper panel. For CaMYO5, only the core SH3domain was used (for details see Methods).Domains: WH1, WASP-homology 1; WH2,WASP-homology 2; P1–P4, proline-richregions in CaWal1; VCA, verprolin-centralacidicdomain; LBD, conserved Las17-bindingdomain; HOT, Hof1-trap domain. Plasmidscarrying the BD/AD pairs were transformedinto S. cerevisiae and selected on –Trp/–Leuplates (lower panel). Transformant colonieswere tested by X-Gal overlay and the b-galactosidase activity was tested in triplicate.with a low affinity or not at all (our unpublished results). Onthe other hand, it was shown that the SH3 domain of S.cerevisiae Hof1 also interacts with formins Bnr1 and Bni1,which could provide an alternative route for the localizationof Vrp1 to sites of polarized growth and septation(Evangelista et al., 2003). Formins and Vrp1 may shareanother feature: the binding of profilin. Bni1 binds via itsFH1 domain to profilin. This domain includes a polyprolinestretch similar to that found at the N terminus of Vrp1homologues, which may explain, mechanistically, how Vrp1contributes to F-actin formation.Thus, our analysis contributes to our understanding of themechanistic link between Wal1 and Myo5 in C. albicans.With the defects in hyphal morphogenesis and endocytosisof the vrp1 mutant strain we have identified another playerpartaking in the yeast-to-hyphal switch in C. albicans.ACKNOWLEDGEMENTSWe thank Alexander Johnson and Suzanne Noble for generouslyproviding the reagents used in this study. Parts of this study werefunded by the EU-Marie Curie Research Training Network‘Penelope’.REFERENCESAghamohammadzadeh, S. & Ayscough, K. R. (2009). Differentialrequirements for actin during yeast and mammalian endocytosis. NatCell Biol 11, 1039–1042.Anderson, B. L., Boldogh, I., Evangelista, M., Boone, C., Greene, L. A.& Pon, L. A. (1998). The Src homology domain 3 (SH3) of a yeast typeI myosin, Myo5p, binds to verprolin and is required for targeting tosites of actin polarization. J Cell Biol 141, 1357–1370.D’Agostino, J. L. & Goode, B. L. (2005). Dissection of Arp2/3 complexactin nucleation mechanism and distinct roles for its nucleationpromotingfactors in Saccharomyces cerevisiae. Genetics 171, 35–47.Evangelista, M., Klebl, B. M., Tong, A. H., Webb, B. A., Leeuw, T.,Leberer, E., Whiteway, M., Thomas, D. Y. & Boone, C. (2000). A rolefor myosin-I in actin assembly through interactions with Vrp1p,Bee1p, and the Arp2/3 complex. J Cell Biol 148, 353–362.Evangelista, M., Pruyne, D., Amberg, D. C., Boone, C. & Bretscher, A.(2002). Formins direct Arp2/3-independent actin filament assemblyto polarize cell growth in yeast. Nat Cell Biol 4, 32–41.http://mic.sgmjournals.org 7

%paper no. mic041707charlesworth ref: mic041707&N. Borth and othersEvangelista, M., Zigmond, S. & Boone, C. (2003). Formins: signalingeffectors for assembly and polarization of actin filaments. J Cell Sci116, 2603–2611.Franke, K., Nguyen, M., Hartl, A., Dahse, H. M., Vogl, G., Wurzner, R.,Zipfel, P. F., Kunkel, W. & Eck, R. (2006). The vesicle transport proteinVac1p is required for virulence of Candida albicans. Microbiology 152,3111–3121.Gietz, R. D. & Schiestl, R. H. (2007). High-efficiency yeasttransformation using the LiAc/SS carrier DNA/PEG method. NatProtoc 2, 31–34.Kaksonen, M., Sun, Y. & Drubin, D. G. (2003). A pathway forassociation of receptors, adaptors, and actin during endocyticinternalization. Cell 115, 475–487.Kaksonen, M., Toret, C. P. & Drubin, D. G. (2005). A modular designfor the clathrin- and actin-mediated endocytosis machinery. Cell 123,305–320.Lambert, A. A., Perron, M. P., Lavoie, E. & Pallotta, D. (2007). TheSaccharomyces cerevisiae Arf3 protein is involved in actin cable andcortical patch formation. FEMS Yeast Res 7, 782–795.Lechler, T., Shevchenko, A. & Li, R. (2000). Direct involvement ofyeast type I myosins in Cdc42-dependent actin polymerization. J CellBiol 148, 363–373.Li, R. (1997). Bee1, a yeast protein with homology to Wiscott-Aldrichsyndrome protein, is critical for the assembly of cortical actincytoskeleton. J Cell Biol 136, 649–658.Machesky, L. M. (2000). The tails of two myosins. J Cell Biol 148,219–221.Madania, A., Dumoulin, P., Grava, S., Kitamoto, H., Scharer-Brodbeck, C., Soulard, A., Moreau, V. & Winsor, B. (1999). TheSaccharomyces cerevisiae homologue of human Wiskott-Aldrichsyndrome protein Las17p interacts with the Arp2/3 complex. MolBiol Cell 10, 3521–3538.Martin, R., Walther, A. & Wendland, J. (2005). Ras1-induced hyphaldevelopment in Candida albicans requires the formin Bni1. EukaryotCell 4, 1712–1724.Munn, A. L. (2001). Molecular requirements for the internalisationstep of endocytosis: insights from yeast. Biochim Biophys Acta 1535,236–257.Munn, A. L. & Thanabalu, T. (2009). Verprolin: a cool set of actinbindingsites and some very HOT prolines. IUBMB Life 61, 707–712.Naqvi, S. N., Zahn, R., Mitchell, D. A., Stevenson, B. J. & Munn, A. L.(1998). The WASp homologue Las17p functions with the WIPhomologue End5p/verprolin and is essential for endocytosis in yeast.Curr Biol 8, 959–962.Noble, S. M. & Johnson, A. D. (2005). Strains and strategies for largescalegene deletion studies of the diploid human fungal pathogenCandida albicans. Eukaryot Cell 4, 298–309.Oberholzer, U., Marcil, A., Leberer, E., Thomas, D. Y. & Whiteway, M.(2002). Myosin I is required for hypha formation in Candida albicans.Eukaryot Cell 1, 213–228.Oberholzer, U., Iouk, T. L., Thomas, D. Y. & Whiteway, M. (2004).Functional characterization of myosin I tail regions in Candidaalbicans. Eukaryot Cell 3, 1272–1286.Palmer, G. E., Cashmore, A. & Sturtevant, J. (2003). Candida albicansVPS11 is required for vacuole biogenesis and germ tube formation.Eukaryot Cell 2, 411–421.Pruyne, D. & Bretscher, A. (2000). Polarization of cell growth inyeast. J Cell Sci 113, 571–585.Ramesh, N., Anton, I. M., Hartwig, J. H. & Geha, R. S. (1997). WIP, aprotein associated with Wiskott–Aldrich syndrome protein, inducesactin polymerization and redistribution in lymphoid cells. Proc NatlAcad Sci U S A 94, 14671–14676.Reijnst, P., Jorde, S. & Wendland, J. (2010). Candida albicans SH3-domain proteins involved in hyphal growth, cytokinesis, and vacuolarmorphology. Curr Genet 56, 309–319.Ren, G., Wang, J., Brinkworth, R., Winsor, B., Kobe, B. & Munn, A. L.(2005). Verprolin cytokinesis function mediated by the Hof one trapdomain. Traffic 6, 575–593.Rose, M. & Botstein, D. (1983). Construction and use of gene fusionsto lacZ (beta-galactosidase) that are expressed in yeast. MethodsEnzymol 101, 167–180.Sagot, I., Rodal, A. A., Moseley, J., Goode, B. L. & Pellman, D. (2002).An actin nucleation mechanism mediated by Bni1 and profilin. NatCell Biol 4, 626–631.Sheu, Y. J., Santos, B., Fortin, N., Costigan, C. & Snyder, M. (1998).Spa2p interacts with cell polarity proteins and signaling componentsinvolved in yeast cell morphogenesis. Mol Cell Biol 18, 4053–4069.Smith, M. G., Swamy, S. R. & Pon, L. A. (2001). The life cycle of actinpatches in mating yeast. J Cell Sci 114, 1505–1513.Sudbery, P., Gow, N. & Berman, J. (2004). The distinct morphogenicstates of Candida albicans. Trends Microbiol 12, 317–324.Thrasher, A. J. & Burns, S. O. (2010). WASP: a key immunologicalmultitasker. Nat Rev Immunol 10, 182–192.Vaduva, G., Martinez-Quiles, N., Anton, I. M., Martin, N. C., Geha, R. S.,Hopper, A. K. & Ramesh, N. (1999). The human WASP-interactingprotein, WIP, activates the cell polarity pathway in yeast. J Biol Chem274, 17103–17108.Veses, V., Richards, A. & Gow, N. A. (2008). Vacuoles and fungalbiology. Curr Opin Microbiol 11, 503–510.Walther, A. & Wendland, J. (2003). An improved transformationprotocol for the human fungal pathogen Candida albicans. Curr Genet42, 339–343.Walther, A. & Wendland, J. (2004). Polarized hyphal growth inCandida albicans requires the Wiskott–Aldrich Syndrome proteinhomolog Wal1p. Eukaryot Cell 3, 471–482.Walther, A. & Wendland, J. (2008). PCR-based gene targeting inCandida albicans. Nat Protoc 3, 1414–1421.Whiteway, M. & Bachewich, C. (2007). Morphogenesis in Candidaalbicans. Annu Rev Microbiol 61, 529–553.Whiteway, M. & Oberholzer, U. (2004). Candida morphogenesisand host–pathogen interactions. Curr Opin Microbiol 7, 350–357.Edited by: J. F. Ernst8 Microbiology 156

Curr GenetDOI 10.1007/s00294-010-0301-7RESEARCH ARTICLECandida albicans SH3-domain proteins involved in hyphal growth,cytokinesis, and vacuolar morphologyPatrick Reijnst · Sigyn Jorde · Jürgen WendlandReceived: 23 December 2009 / Revised: 22 March 2010 / Accepted: 29 March 2010© Springer-Verlag 2010Abstract This report describes the analyses of threeCandida albicans genes that encode Src Homology 3(SH3)-domain proteins. Homologs in Saccharomyces cerevisiaeare encoded by the SLA1, NBP2, and CYK3 genes.Deletion of CYK3 in C. albicans was not feasible, suggestingit is essential. Promoter shutdown experiments ofCaCYK3 revealed cytokinesis defects, which are in linewith the localization of GFP-tagged Cyk3 at septal sites.Deletion of SLA1 resulted in strains with decreased abilityto form hyphal Wlaments. The number of cortical actinpatches was strongly reduced in Δsla1 strains during allgrowth stages. Sla1-GFP localizes in patches that are foundconcentrated at the hyphal tip. Deletion of the Wrst twoSH3-domains of Sla1 still resulted in cortical localizationof the truncated protein. However, the actin cytoskeleton inthis strain was aberrant like in the Δsla1 deletion mutantindicating a function of these SH3 domains to recruit actinnucleation to sites of endocytosis. Deletion of NBP2resulted in a defect in vacuolar fusion in hyphae. Germcells of Δnbp2 strains lacked a large vacuole but initiatedseveral germ tubes. The mutant phenotypes of Δnbp2 andΔsla1 could be corrected by reintegration of the wild-typegenes.Keywords Candida albicans · PCR-based genetargeting · pFA-plasmids · Actin cytoskeleton · FM4-64Communicated by C. D'Enfert.P. Reijnst · S. Jorde · J. Wendland (&)Carlsberg Laboratory, Yeast Biology,Gamle Carlsberg Vej 10, 2500 Valby,Copenhagen, Denmarke-mail: jww@crc.dkIntroductionCandida albicans is one of the most important human fungalpathogens. It occurs as a commensal on epithelial surfacesin oropharyngeal tissue, the gastro-intestinal tract,and in the vagina. Particularly, vaginitis and urinary tractinfections caused by C. albicans are frequent in otherwisehealthy individuals. Immuno-compromised patients mayadditionally develop life-threatening systemic infections ofinner organs (Odds 1994; Calderone and Fonzi 2001).The morphological transition of C. albicans from yeastto hyphal growth has been recognized as an important virulenceattribute amongst others (Sudbery et al. 2004;Kumamoto and Vinces 2005). Filamenting germ cellscharacteristically generate large vacuoles. This compartmentalizesthe germ tube in an apical region that containsendosomes and small vacuoles, and subapical regionswhich harbor large vacuoles at the expense of cytoplasm.Septation in hyphal Wlaments further promotes this compartmentalization.The unequal distribution of vacuolar volumeinXuences the branching frequency during Wlamentousgrowth (Barelle et al. 2006; Veses et al. 2009). The actincytoskeleton is polarized at sites of polarized growth andcortical actin patches cluster in the hyphal tips. Defects inthe polarization of the actin cytoskeleton, e.g. interferingwith the function of several Rho-type GTPases generallylead to growth defects (Wendland 2001; Court and Sudbery2007; Zheng et al. 2007). Actin ring formation promotedvia Iqg1 at sites of septation is required for septum formation(Epp and Chant 1997; Wendland and Philippsen 2002).In S. cerevisiae, CYK3 can act as a multicopy suppressor ofan IQG deletion (Korinek et al. 2000). Processes like polarizedhyphal growth, endocytosis, and cytokinesis requireprotein networks and timely regulation within the cellcycle. SH3-domain encoding proteins are well suited to123

Curr Genetplay important roles in these processes since they can mediateprotein–protein interactions via their SH3-domains (amBusch et al. 2009). Several C. albicans SH3-protein encodinggenes have already been characterized including MYO5,BEM1, and CDC25 (Enloe et al. 2000; Michel et al. 2002;Oberholzer et al. 2002; Bassilana et al. 2003). BEM1 wasfound to be essential, while MYO5 plays an important roleduring endocytosis. Both Myo5 and Cdc25 are required forWlamentation under speciWc conditions. Therefore, otherSH3-domain encoding genes may also play important morphogeneticroles.Establishing the genome sequence of C. albicans hasopened the way for the functional analysis of the C. albicansgene set as has been elegantly achieved in S. cerevisiae(Winzeler etal. 1999). PCR-based gene targetingapproaches similar to those used in S. cerevisiae have beenestablished to generate homozygous mutant strains aftertwo successive rounds of transformation (Berman and Sudbery2002; Walther and Wendland 2008).To contribute further to the functional analysis ofC. albicans genes, we have functionally analyzed theC. albicans homologs of the S. cerevisiae SH3-domainencoding genes SLA1, NBP2, and CYK3.Materials and methodsStrains and mediaThe C. albicans strains used and generated in this study arelisted in Table 1. Generally, at least two independenttransformants were generated for each desired geneticmanipulation. Strains were grown either in yeast extract–peptone–dextrose (YPD; 1% yeast extract, 2% peptone,2% dextrose) or in deWned minimal media [CSM; completesupplement mixture; 6.7 g/l yeast nitrogen base(YNB) with ammonium sulfate and without amino acids;0.69 g/l CSM; 20 g/l glucose] with the addition ofrequired amino acids and uridine. Promoter shut down ofMET3-promoter or MAL2-promoter controlled geneexpression was done as described previously (Bauer andWendland 2007).Strains were generally grown at 30°C to keep them inthe yeast phase; hyphal induction of C. albicans cells wasdone at 37°C with the addition of 10% serum to the growthmedium. Escherichia coli strain DH5α was used for pFAplasmidpropagation.Transformation of C. albicansCompletely independent C. albicans homozygous completeORF-deletion strains were constructed startingfrom C. albicans strain SN148 (Noble and Johnson2005). PCR-generated disruption cassettes were used totarget both alleles of a gene, which were deleted bysequential transformation of Wrst SN148 and then theresulting heterozygous strains. PCR-products for transformationof C. albicans were ampliWed from pFA-vectors(Table 2) using S1- and S2-primers as described(Walther and Wendland 2008). Primers were purchasedfrom biomers.net GmbH (Ulm, Germany). S1- and S2-primers(see Table 3) harbor 100 nt of target homology at theirTable 1 C. albicans strainsused in this studyCm C. maltosa,Cd C. dubliniensisa All CAxxxx strainsare derivates of SN148Strain a Genotype SourceSC5314 C. albicans wild type Gillum et al. 1984SN148arg4/arg4, leu2/leu2, his1/his1ura3::imm 434 /ura3::imm 434 , iro1::imm 434 /iro1::imm 434Noble andJohnson 2005CAP046 NBP2/nbp2::CdHIS1, leu2, ura3, arg4 This studyCAP015 nbp2::CdHIS1/nbp2::URA3, leu2, arg4 This studyCAP191 nbp2::CdHIS1/nbp2::URA3, BUD3/bud3::NBP2-CmLEU2, arg4 This studyCAP147 CYK3/cyk3::CdHIS1, leu2, ura3, arg4 This studyCAP007 URA3-MET3p-CYK/cyk3::CdHIS1, arg4, leu2 This studyCAP054 SLA1/sla1::CdHIS1, leu2, ura3, arg4 This studyCAP024 sla1::CdHIS1/sla1::URA3, leu2, arg4 This studyCAP204 sla1::CdHIS1/sla1::URA3, BUD3/bud3::SLA1-CmLEU2, arg4 This studyCAP025 sla1::ARG4isla1::URA3, his1, leu2 This studyCAP026 sla1::ARG4/sla1::URA3, his1, leu2 This studyCAS024 SLA1/sla1::CdHIS1, leu2, ura3, arg4 This studyCAP206 URA3-MAL2p-sla1 ΔSH3#1,2 /sla1::CdHIS1, leu2, arg4 This studyCAP221 URA3-MAL2p-sla1 ΔSH3#1,2 -GFP-CmLEU2/sla1::CdHIS1, arg4 This studyCAS027 SLA1-GFP-CmLEU2/sla1::CdHIS1 This studyCAS030 CYK3-GFP-CmLEU2/cyk3::CdHIS1, ura3, arg4 This study123

Curr GenetTable 2 Plasmids used in this studyPlasmid Description Source200 pFA-URA3 Gola et al. 2003230 pFA-URA3-MAL2p Gola et al. 2003627 pFA-CdHIS1 Schaub et al. 2006697 pFA-GFP-CmLEU2 Schaub et al. 2006873 pRS-CaBUD3-CmLEU2 WendlandC508 pDrive-SLA1 This studyC553 pRS-BUD3-SLA1-CmLEU2 This studyC527 pDrive-NBP2 This studyC530 pRS-BUD3-NBP2-CmLEU2 This studyC177 pRS417 (GEN3) This studyC196 pDRIVE-3-CYK3 This studyC200 pRS417-3-CYK3 This studyC257 pRS417-3-CYK3-GFP This studyC182 pGEM-3-SLA1 This studyC201 pRS417-3-SLA1 This studyC256 pRS417-3-SLA1-GFP This studyCAGFY04 SLA1-UAU1-cassette MitchellCAGO130 SLA1-UAU1-cassette MitchellCa C. albicans, Cm C. maltosa, Cd C. dubliniensis5-ends. Shorter primers were used for diagnostic PCR toverify the integration of the cassettes and absence of the targetgene in homozygous mutants. Transformation was doneas described (Walther and Wendland 2003).SLA1 was also disrupted using two gene-speciWcUAU1 cassettes kindly provided by Aaron Mitchell.Transformation with the SLA1 cassettes on plasmidsCAGFY04 and CAGO130 required linearization of theplasmid using NotI and transformation of C. albicanswith a Wrst selection on ¡Arg media. Restreaking of theprimary transformants on ¡Arg and ¡Ura media wasdone to select for recombinants in which both alleleshave been disrupted (see Nobile and Mitchell 2009 forfurther details).Reintegration of SLA1 and NBP2Reconstitution of the Δsla1 and Δnbp2 strains was done byreintegration of the wild-type gene at the BUD3 locus. Tothis end, SLA1 was ampliWed using primers #3562 and#3237 and cloned into pDrive (C508). From there SLA1was cloned as an XhoI/BamHI fragment into plasmid #873to yield plasmid C553. The Δsla1 strain CAP024 was transformedwith SpeI-linearized plasmid C553. Similarly,NBP2 was ampliWed using primers #3557 and #4196 andcloned into pDrive generating plasmid C527. NBP2 wasthen cloned as a XhoI/BamHI fragment into #873 generatingplasmid C530. Plasmid C530 was transformed aftercleavage with SpeI to generate strain CAP191.Construction of SLA1-GFP and CYK3-GFPTo generate chromosomally GFP-tagged strains, the followingprocedure was applied. The 3-ends of SLA1 andCYK3 were ampliWed using primer pairs #3236/#3237 and#3222/#3223, respectively, and cloned into pDrive (plasmidsC182 and C196). The fragments were recloned intopRS417, which is based on pRS415 but carries a GEN3marker instead of LEU2. This generated plasmids C200 andC201, which were used for in vivo recombination in S.cerevisiae to add the GFP-CmLEU2 cassettes ampliWedusing the primer pairs #3314/#3315 for SLA1 and #3312/#3313 for CYK3, respectively. The resulting plasmids C256and C257 were cleaved by XhoI/BamHI and SacII/XhoI,respectively, to release the targeting cassettes used fortransformation of C. albicans. Correct fusion was veriWedby sequencing and correct integration of the cassettes wasveriWed by diagnostic PCR.Construction of sla1 ΔSH3#1,2 -GFPTo generate a SLA1-allele which is expressed from the regulatableMAL2-promoter and lacks the Wrst two SH3-domains, a PCR-based gene targeting approach was used.The URA3-MAL2p-cassette was ampliWed from a pFAvector using primers #3594 and #4265. This cassette wastransformed into strain CAS023 (SLA1/sla1::CdHIS1).This generated strain CAP206 bearing a deletion of oneSLA1 allele and converting the remaining allele tosla1 ΔSH3#1,2 . To be able to record the localization of thetruncated protein in living cells, this SLA1 allele was taggedwith GFP. To this end, the SLA1-GFP-tagging cassette wasused and CAP206 was transformed with SpeI/SacIIdigested C256 (pRS417-3-SLA1-GFP). This resulted inthe addition of GFP to the C-terminus of sla1 ΔSH3#1,2 .Microscopy and staining proceduresFluorescence microscopy was done with an Axio-Imagermicroscope (Zeiss, Jena and Göttingen, Germany) usingMetamorph software tools (Molecular Devices Corp.,Downington, PA, USA) and a MicroMax1024 CCD-camera(Princeton Instruments, Trenton, NJ, USA). Imagingwas performed using the appropriate Wlter combinations forFM4-64-imaging, GFP-localization, and actin-staining asdescribed (Walther and Wendland 2004a, b). Quinacrinestaining was done according to (Weisman et al. 1987). Tothis end, strains were grown overnight in YPD, and thendiluted in YPD + Serum and grown for an additional 4 h.Quinacrine was added to a Wnal concentration of 200 μM.123

Curr GenetTable 3 Primers used in this studyGenesCaCYK3CaCYK3CaCYK3CaCYK3CaCYK3CaCYK3CaCYK3CaCYK3CaNBP2CaNBP2CaNBP2CaNBP2CaNBP2CaNBP2CaNBP2CaSLA1CaSLA1CaSLA1CaSLA1CaSLA1CaSLA1CaSLA1CaSLA1CaSLA1CaSLA1CaSLA1CaSLA1CaURA3CaURA3CdHIS1CdHIS1CmLEU2CaMAL2Primer names and sequences a#4019: S1-CaCYK3: CCTTTCATTAATTACAAAGAAAAAAATAAGAACATCAACTATCTTTTCACTCTTTTTGAACAAATTTGTATCATACTAAAAGAATTAAATAATAAATAATgaagcttcgtacgctgcaggtc#3313: S2-CaCYK3: AATGTACAAATGGCAAAAAGAAGTAGTAGCAGAAGAGGTAATCTATAAAGAATTTAAAACTAAATAATACCCACTCTGTTTCCCTCTTTATATATATATAtctgatatcatcgatgaattcgag#4020: G1-CaCYK3: GCACACTTGATGATTTCATC#3222: G3-CaCYK3: GCTAAGATCAAGGCAGTG#3223: G4-CaCYK3: GCAACTGCTGCAGTAGAC#3312: S1-GFP-CaCYK3: TATGTTTTCGCTCAGTGGGAGTGCATAGGTAGCACAGTTGCAAATggtgctggcgcaggtgcttc#3539: G1-CaCYK3-GFP: GACTGCAAGGGCAACCAC#3747: G2-CaCYK3: AGGATTTAaagcttttaCCCAAGTGGGGTTGTTCCAGC#3589: S1-CaNBP2: GTCTTGTTTGTCCTGTGTGTGTGTGTGTGTGTGTTGATAAATCACCTGAAACATATACTATTTAATCATTTGTTATTCATCATTATTGTCCATTTTGAATAGgaagcttcgtacgctgcaggtc#3579: S2-CaNBP2: CACATACACTCTGTTGGTATGAAAGTATAAAAACATTTGATAAAATTCGTAATCAACATTAATATAACTTAATTGTCCCTATAAGCTGGCTAATATTGGAtctgatatcatcgatgaattcgag#3557: G1-CaNBP2: GGTGTTTCACATTATTCTCCG#3309: G4-CaNBP2: TGGCCGAACCCTTCCTGG#3245: I1-CaNBP2: GACAAGTCATTTCCCACC#3246: I2-CaNBP2: CTTCAGCAACTAACCAACCTTG#4196: A4-CaNBP2: CACATACACTCTGTTGGTATG#3594: S1-CaSLA1: CAACTCCTATGTTAGAGCTAGTCGTGCTCAACACAAAACCTGATGTGAAACAATGAAACTTTCGACGATTCTACAAAAGTGCGGAAATTGCTTGAAATCAAAGgaagcttcgtacgctgcaggtc#3315: S2-CaSLA1: AGCATTACAAACTATGAAAGGAATAAGAAATAATGAATAATATTTTGTTTGATATACAATTATAAAATAAAAGAGTTAATAAAGGTTCAAAATGCACTTTtctgatatcatcgatgaattcgag#3562: G1-CaSLA1: CGGTAGAGATGATGTTGTG#3765: G2-SLA1: AGGATTTAaagcttttaAGGTGGTGCAGGGAAATCCG#3236: G3-CaSLA1: TGGTGGAGCACCACAGAC#3237: G4-CaSLA1: CGGCTTTGCAACATCAAGAC#3241: I1-CaSLA1: CATAGGGATAGATCACCAG#3242: I2-CaSLA1: CTTCTCTCAAACCATGGGC#3243: I3-CaSLA1: CACAACAACAACCGCCACC#3244: I4-CaSLA1: CCATACCAGTTGGTTGTGAC#3314: S1-GFP-CaSLA1: AGAGCTAATCTACAAGCAGCAACACCAGATAATCCCTTTGGATTCggtgctggcgcaggtgcttc#4265: S2-MALp-SLA1ΔSH3#1-2:GAATCTGAGTCTGTTGCTGTGGAATAGCCTGCTGTTGTTGTGGTGGTGGTTGGAAAACCTGTTGTGGTTGTTGCTGTTGATGCTGTGCTGGCTCTGCTGTcattgtagttgattattagttaaaccac#600: U2: GTGTTACGAATCAATGGCACTACAGC#599: U3: GGAGTTGGATTAGATGATAAAGGTGATGG#1432: H2: TCTAAACTGTATATCGGCACCGCTC#1433: H3: GCTGGCGCAACAGATATATTGGTGC#1743: L3: GCTGAAGCTTTAGAAGAAGCCGTG#4269: G3-CaMALp: GTACAACTAAACTGGGTGATGCa C. albicans, Cm C. maltosa, Cd C. dubliniensisa Upper case sequences correspond to C. albicans DNA sequences and lower case sequences correspond to 3-terminal annealing regions for theampliWcation of pFA-cassettes. All sequences are written from 5 to 3Cells were incubated at 37°C for 5 min, collected by centrifugationand resuspended in 200 μl YPD + Serum with50 mM NaH 2 PO 4 . Visualization was done with the GFPWlter. Samples were analyzed by generating either singleimages or stacks of 5–20 images that were processed intosingle plane projections using Metamorph software.ResultsSequence comparisonsThe C. albicans Sla1, Nbp2, and Cyk3 proteins share onefeature in the possession of SH3-domains, which were123

Curr GenetABNbp2SH3-domainCyk3Sla1CAP025CAP026Sla1 SH3#1-2MAL2-promFig. 1 Alignment and position of SH3-domains. a The Wve SH3-domainsof the three genes that were functionally analyzed in this studywere aligned using the MegAlign tool of the DNASTAR software (http://www.dnastar.com).Consensus sites are shaded in gray while identicalsites in all Wve domains are shaded in black. b The positions of theSH3-domain vary within the proteins and are for Nbp2 at amino acids127–184, for Cyk3 at position 11–68 and for Sla1 at positions 7–73,76–133, and 399–457. For reference, the complete protein length isdrawn to scale. No other domains were found using the SMART toolat (http://smart.embl.de/) also Sla1 has several repeats at its C-terminus.Deletion of the N-terminal SH3-domains results in a sla1 alleletermed Sla1 ΔSH3#1,2identiWed using the SMART tool at http://smart.embl.de/.An SH3-domain is composed of app. 70 amino acids andseveral conserved residues can be found (Fig. 1a). Theposition of a SH3-domain within a protein may vary andthere are also proteins like Sla1 that contain more than oneSH3-domain (Fig. 1b). Amino acid sequence identitiesbetween the C. albicans and, for example, Saccharomycescerevisiae proteins are overall not very high and rangebetween 27 and 37%. The C. albicans proteins are largerthan the S. cerevisiae homologs: Sla1 by only 13aa, Cyk3by 135aa, and Nbp2 by one-third (342aa compared to236aa). The characterizations of the yeast genes revealedthat SLA1 plays a role in actin cytoskeleton assembly andendocytosis; Nbp2 is required for mitotic growth at hightemperatures and for cell wall integrity and Cyk3 isinvolved in cytokinesis (Korinek et al. 2000; Warren et al.2002; Ohkuni et al. 2003).Generation of C. albicans mutant strainsIn this report, we have employed several strategies forgene function analysis relying on diVerent pFA-vectorsand also used a single transformation approach relyingon UAU1-cassettes as described below. Initially, we usedthe pFA-series to generate complete ORF-deletion strainsin the three genes. From two heterozygous mutantstrains, we went onto obtain two independent homozygousmutant strains thereof using the C. albicans URA3and the Candida dubliniensis HIS1 marker genes. Inorder to characterize SLA1 in more detail, we used insertionaldisruption cassettes based on SLA1-UAU1-cassettes.Since deletion of CYK3 was not feasible, we useda promoter shutdown approach to analyze the consequencesof Cyk3 depletion. Mutant phenotypes could beobtained with the homozygous mutants of sla1 and nbp2,which in both cases could be complemented by the reintegrationof the wild-type gene at the BUD3 locus. Localizationof Sla1 and Cyk3 was done by Xuorescencemicroscopy of GFP-tagged strains. Using this array oftools, we were able to achieve an initial characterizationof gene function for these genes, which will be describedin the following sections.Deletion of SLA1 and NBP2 results in hyphal growthphenotypesHomozygous mutant strains of sla1 and nbp2 were characterizedto reveal their growth potential under diVerentgrowth conditions. When grown on minimal media, nostrong defects during yeast growth were observable.Hyphal growth was monitored by inducing yeast cellseither on solid media or in liquid culture (Fig. 2). The wildtype strongly Wlaments after addition of serum or in spidermedium. Mutants in sla1 or nbp2 were also able to induceWlament formation. Hyphae of these mutant strains, however,were shorter than the wild type after several hours ofinduction. Interestingly, the nbp2 mutant showed frequentreinitiating of germ tube formation from the germ cell.123

Curr GenetA B CWTsla1nbp2Fig. 2 Characterization of growth phenotypes of the sla1 and nbp2mutants. The strains were grown overnight in liquid culture and theninoculated on minimal medium (a), on serum containing plates or inliquid medium supplemented with 10% serum (b), or on spider platesand in spider liquid medium (c). Plates were incubated for 3 days at30°C (a) or 37°C (b, c) prior to photography. Hyphal induction in liquidmedia was done for 3 h prior to microscopy. Bar 10 μm. Note theshort germ tube in the sla1 strain (middle row) and the multiple germtubes in the nbp2 mutant (bottom row)Thus, after 3 h, most of the germ cells had formed two orthree germ tubes (Fig. 2).SLA1-deletion leads to defects in actin patch assemblyand distributionASC5314 sla1::UAU1-CAP025 sla1::UAU1-CAP026The actin cytoskeleton plays a major role in polarizedgrowth. During hyphal growth, actin cables in C. albicansare formed by the tip-localized polarisome and actinpatches localize to the hyphal tip at sites of endocytosis (forreview see Pruyne and Bretscher 2000a, b). To analyze thegrowth defects of sla1 in detail, we used Xuorescencemicroscopy of rhodamine stained germ tubes. We analyzedthree diVerent sla1 mutant strains: two bearing UAU1 cassettes,CAP025 and CAP026, and a complete ORF-deletionstrain. In this way, we could analyze the eVect of truncatingSLA1 at positions that potentially leave all three SH3-domainsintact, truncate the protein after the Wrst two SH3-domainsor entirely delete SLA1 (see also Fig.1). The two UAU1insertions in SLA1 already showed diVerent phenotypes.The CAP026 strain basically showed no defect, grew likewild type and accumulated actin patches in the hyphal tips.The CAP025 strain in which the UAU1 insertion truncatesSLA1 downstream of the region coding for the second SH3-domain, shows decreased hyphal lengths when comparedwith the wild type. The number of actin patches in thismutant is decreased and the patches do not accumulate inthe tips of hyphae (Fig. 3a). This phenotype is even morepronounced in the complete ORF-deletion strain. Thisstrain has a more drastic growth defect and even fewer corticalactin patches. This demonstrates that Sla1 is involvedin the assembly and polarization of cortical actin patches inC. albicans. The assembly of actin cables in the hyphal Wlamentswas not impaired (Fig. 3b). The growth defect of thenull mutant could be complemented by reintegration of thewild-type SLA1 gene at the BUD3 locus. The assembly ofcortical actin patches and their polarized localization wasalso restored in both yeast and hyphal cells (Fig. 3b).B SC5314 sla1sla1,BUD3/bud3::SLA1SLA1-GFP/sla1Fig. 3 Deletion of SLA1 leads to defects in cortical actin patch assembly.Strains were grown overnight in YPD, diluted in new medium andgrown for 4 h at 30°C for yeast cell growth or at 37°C in the presenceof 10% serum to induce Wlament formation. Cells were then Wxed andstained with rhodamine-phalloidin. Bright-Weld and Xuorescence imagesshowing the actin cytoskeleton of the indicated strains are shown.a DiVerential eVect on cortical actin patch assembly of SLA1-UAU1insertions compared to the wild type. b Reintegration of SLA1 complementsthe actin patch defect of the sla1 complete ORF-deletion mutant.In vivo localization of Sla1-GFP was done without Wxation. Bars 5 μmTo visualize the localization of Sla1 in vivo, a chromosomallytagged SLA1-GFP strain was constructed based ona heterozygous mutant. Sla1 shows a patch-like localizationin yeast cells and hyphae (Fig. 3b). An increased number ofSla1-GFP patches can be found at sites of polarized growth,123

Curr GenetFig. 4 The two amino terminalSH3-domains of Sla1 are requiredfor the organization of theactin cytoskeleton. CAP221 wasgrown overnight in YPD for afull SLA1 shut-down. Then cellswere diluted in either YPD (repressed)or YPM (induced) withor without serum and grown for4 h. Afterwards, cells were Wxedand stained with rhodaminephalloidinprior to GFP and actinXuorescence microscopy. Bar5 μmDIC GFP DIC actin DIC GFP DIC actinMAL2p-sla1 SH3#1,2 -GFP repressedMAL2p-sla1 SH3#1,2 -GFP inducede.g. the hyphal tip. This localization pattern resembles, forexample, that of C. albicans Abp1 and other proteinsinvolved in actin patch assembly or endocytosis (Martinet al. 2007).The N-terminal SH3-domains of Sla1 are important foractin cytoskeleton assembly but not for localization of Sla1To demonstrate that the two N-terminally located SH3-domains of Sla1 contribute to the function of Sla1, we generateda truncated allele, sla1 ΔSH3#1,2 . This allele was placedunder control of the regulatable MAL2-promoter using aPCR-based gene targeting approach, which at the sametime eliminated the Wrst two SH3-domains. Furthermore, tobe able to localize the truncated protein, a C-terminal tagwas added to sla1 ΔSH3#1,2 . This strain was then used to visualizeboth the localization of Sla1 ΔSH3#1,2 -GFP and the organizationof actin cytoskeleton in yeast and hyphal cells(Fig. 4). When grown in glucose, sla1 ΔSH3#1,2 expressionwas turned down and the protein could not be detected.Actin organization resembled that of a sla1 mutant strain(compare Figs. 3, 4). Growth in maltose medium inducedthe expression of sla1 ΔSH3#1,2 -GFP. Hence Sla1 ΔSH3#1,2 -GFP could be detected as cortical patches in yeast andhyphal cells. Sla1 ΔSH3#1,2 -GFP was also found enriched inthe hyphal tips. Thus, Sla1 ΔSH3#1,2 -GFP localizes in a similarmanner as full length Sla1-GFP indicating that the N-terminalSH3 domains do not play a role in Sla1-targeting to thecortex. However, under inducing conditions the actin cytoskeletonassembly in the strain expressing Sla1 ΔSH3#1,2 -GFPwas still aberrant. Cortical actin patches that did notlocalize to the hyphal tip were reduced in number and thusresembled the situation in the sla1 deletion strain. Thisindicates that the N-terminal SH3 domains of Sla1 do playan important role in organization of the actin cytoskeletonat sites of endocytosis (Fig. 4).NBP2-deletion leads to multiple germ tube formationDeletion of NBP2 did not reveal any defects during theyeast growth phase. Yet, under hyphal inducing conditions,we observed that all germ cells developed multiple germtubes after 4 h and the length of the primary germ tube wasdecreased in nbp2 compared to that of the wild type(Figs. 2, 5). Previously, it was shown in C. albicans andAshbya gossypii that in hyphae large vacuoles are formed insubapical compartments (Walther and Wendland 2004a;Veses and Gow 2008). This inXuences the ratio of cytoplasmversus vacuole and inXuences the branching frequency(Veses et al. 2009). Therefore, we analyzed thevacuolar compartments in FM4-64 stained yeast cells andhyphae of the wild type and the nbp2 mutant (Fig. 5). Duringyeast growth, both strains accumulated a larger vacuolein mother cells and showed no observable diVerence. However,staining of germlings revealed the inability of nbp2germ cells to generate large vacuolar compartments. Fragmentedvacuoles were found throughout nbp2 hyphae.Thus, the altered ratio of cytoplasm versus vacuolar spacemay be the causal link to increased branching of germ cellsin the nbp2 mutant. To corroborate that the defect in vacuolarfusion was speciWc for the nbp2 deletion strain, we reintegratedthe NBP2 gene at the BUD3-locus. As expected,123

Curr GenetSC5314nbp2nbp2,BUD3/bud3::NBP2Fig. 5 Deletion of NBP2 results in vacuolar fragmentation in hyphae.Strains were grown under yeast or germ tube inducing conditions for4 h. FM4-64 (0.2 μg/ml) was added and samples were processed forbrightWeld and Xuorescence microscopy after 1 h to allow uptake of thedye. Bars 5 μmthe reintegrant showed wild type vacuolar phenotype. Wealso generated a chromosomally encoded NBP2-GFP,which, however, did not yield a Xuorescent signal. To analyzevacuolar acidiWcation in the nbp2 strain, we used quinacrinestaining and Xuorescence microscopy (Fig. 6).Quinacrine diVuses through membranes and accumulates inacidic compartments like the vacuole (Weisman et al.1987). The accumulation of the dye and staining of vacuolesof nbp2 hyphae indicated that vacuolar fusion but notthe function of the vacuoles was aVected in the nbp2 strain(Fig. 6, 7).Depletion of Cyk3 results in cytokinesis defectsWe were unable to generate homozygous cyk3 strains frominitial heterozygous mutants, without also generating sometriplication event that left a wild-type copy of the gene inthe genome. Thus, we conclude that CYK3 is an essentialgene. In S. cerevisiae, CYK3 is involved in cytokinesis andlocalizes to the bud neck in large budded cells (Korineket al. 2000). Cyk3 localization in C. albicans was determinedby producing a fusion between the chromosomalCYK3 gene with GFP in a heterozygous mutant. In C. albicans,CYK3 was found to localize at the bud neck in largebudded cells similar to Cyk3 in S. cerevisiae (Fig. 5a).In S. cerevisiae, deletion of CYK3 results in only mildcytokinesis defects, which contrasts the situation inC. albicans. To assess a phenotype upon depletion of CYK3transcript, we produced a strain which expressed CYK3 fromFig. 6 Vacuoles of an nbp2 mutant strain are acidic. Strains weregrown under yeast or germ tube inducing conditions for 4 h. Quinacrinestaining reveals acidiWed and functional vacuolesthe regulatable C. albicans MET3 promoter (Care etal.1999). Shutdown of CYK3 expression resulted in a severecytokinesis defect. CYK3-depleted cells were elongated ormisshapen and showed abnormal chitin deposition (Fig. 5b).DiscussionIn this study, we have generated C. albicans mutant strainsfor three SH3-domain encoding genes using PCR-basedgene targeting methodologies and a single-step transformationprotocol with UAU1 cassettes (Walther and Wendland2008; Nobile and Mitchell 2009). SH3-domains are smallprotein domains that promote protein–protein interactions,particularly by binding to proline-rich ligands with a PxxPmotif (Mayer 2001). The binding aYnity and binding speciWcityare inherently rather low. This may pose some diYcultieswhen trying to establish protein interactions usingthe yeast two-hybrid system. Sla1, on the other hand, containsthree SH3 domains, which may help to increase speciWcbinding of target proteins.Given the strong potential of SH3-domains to promotesignaling and morphogenesis, a large variety of SH3-domain123

Curr GenetFig. 7 CYK3 localization anddepletion after promoter shutdown.a Cyk3-GFP Xuorescencein large budded yeast cells wasobserved at the bud neck. b Cellsin which CYK3 expression iscontrolled by the regulatableMET3-promoter were grownovernight in YPD at 30°C with(repressed) or without (induced)the addition of 3.5 mM methionineand cysteine. Prior tomicroscopy, calcoXuor was addedto the medium to stain chitinrich regions. Bar 10 μmABMET3p-CYK3inducedMET3p-CYK3repressedproteins can be found in eukaryotic genomes ranging from20–30 in yeast-like ascomycetes to over 300 in humans(Karkkainen et al. 2006). In yeast-like ascomycetes, thereseems to be limited evolution of SH3-domain encodinggenes. For example, in S. cerevisiae, Abp1 contains oneSH3-domain, while in C. albicans, the Abp1 homolog hastwo adjacent SH3-domains, yet deletion of CaABP1showed no discernible phenotype (Martin et al. 2007).Deletion of the C. albicans SLA1 resulted in similar actinassembly defects compared to a SLA1 deletion in S. cerevisiae.The severe reduction in actin patches, however, didnot abolish the ability to generate germ tubes in the C. albicanssla1 mutants although a decreased polarized growthrate could be observed. A similar phenotype was observedin a Camyo5(S366D) allele, which mimics a phosphorylatedserine. A strain bearing this allele was found to Wlament,yet shows a largely delocalized actin cytoskeleton(Oberholzer et al. 2002). In this paper, we identiWed theN-terminal region of C. albicans Sla1 containing two SH3-domains to be required for correct organization of the actincytoskeleton. In S. cerevisiae, Sla1 localizes to the cortexvia an interaction of the Sla1 C-terminal repeat region withEnd3 (Tang et al. 2000; Warren et al. 2002). Furthermore,Sla1 interacts with Las17 and Abp1 as shown by immunoprecipitation(Warren et al. 2002). The elimination of twoSH3 domains from Sla1 resulted in profound disorganizationof the actin cytoskeleton indentifying Sla1 as a majorplayer linking early events of endocytosis with the actincytoskeleton. Nevertheless, sla1 mutants were able to generate,albeit short, hyphae.Surprisingly, sla1 did not show a defect in the formationof large subapical vacuoles (see also Fig. 2). This, on theother hand, was observed for the nbp2 mutant. In S. cerevisiae,nbp2 mutants are temperature sensitive and also sensitiveto cell wall stress (Ohkuni et al. 2003). Our C. albicansnbp2 mutants were not temperature sensitive and grew wellat 40°C even with the addition of 1 M sorbitol or 1.5 MNaCl (data not shown). The transformation frequency,which requires a heat shock, was also not aVected in nbp2cells. Thus, our results indicate some novel vacuolar functionsfor NBP2 which are more pronounced during hyphalgrowth stages and not apparent in yeast cells. Germ tubeformation in the nbp2 strain was altered in a way that germcells quickly generated multiple hyphae rather than onedominant germ tube as in the wild type. Thus, such a phenotypecould be useful in larger scale screenings of aC. albicans mutant collection once available.SH3-domain proteins in C. albicans are taking part in avariety of processes. In this study, we identiWed a key roleof the Wrst two Sla1 SH3-domains for the polarized assemblyof the actin cytoskeleton, which had not previouslybeen identiWed in other studies. We also revealed theinvolvement of Nbp2 in vacuolar fusion, and of Cyk3 incytokinesis. The promoter shutdown experiment using123

Curr GenetMET-promoter controlled CYK3 did not result in growtharrest. This may be due to the leakiness of the promoter.However, cells were found to be deWcient in cell separationproviding evidence that also in C. albicans Cyk3 isinvolved in this process. Our GFP-localization data ofCyk3-GFP provide further evidence for that. Similar toS. cerevisiae, C. albicans Cyk3 may, therefore, act at thelevel of actin ring formation or constriction.Due to the diploidy of C. albicans, gene function analysesstill require much more eVort to produce the correctdeletion strains. Using PCR-based gene targeting methods,detailed structure–function analyses are possible andreduce the time required to construct the desired strains.Thus, larger scale approaches can be undertaken also inC. albicans (Noble and Johnson 2005). Our study of threepreviously uncharacterized C. albicans genes, therefore,adds to the repository of functional analysis information forthis human fungal pathogen.Acknowledgments We thank Alexander Johnson, Suzanne Noble,and Aaron Mitchell for generously providing reagents used in thisstudy; Sidsel Ehlers for providing technical assistance and AndreaWalther for support on microscopy. This study was funded by the EU-Marie Curie Research Training Network “Penelope” and we thankmembers of this consortium for discussions.Referencesam Busch MS, Mignon D, Simonson T (2009) Computational proteindesign as a tool for fold recognition. Proteins 77:139–158Barelle CJ, Richard ML, Gaillardin C, Gow NA, Brown AJ (2006)Candida albicans VAC8 is required for vacuolar inheritance andnormal hyphal branching. Eukaryot Cell 5:359–367Bassilana M, Blyth J, Arkowitz RA (2003) Cdc24, the GDP-GTP exchangefactor for Cdc42, is required for invasive hyphal growthof Candida albicans. Eukaryot Cell 2:9–18Bauer J, Wendland J (2007) Candida albicans SX1 suppresses Xocculationand Wlamentation. Eukaryot Cell 6:1736–1744Berman J, Sudbery PE (2002) Candida albicans: a molecular revolutionbuilt on lessons from budding yeast. Nat Rev Genet 3:918–930Calderone RA, Fonzi WA (2001) Virulence factors of Candida albicans.Trends Microbiol 9:327–335Care RS, Trevethick J, Binley KM, Sudbery PE (1999) The MET3 promoter:a new tool for Candida albicans molecular genetics. MolMicrobiol 34:792–798Court H, Sudbery P (2007) Regulation of Cdc42 GTPase activity in theformation of hyphae in Candida albicans. Mol Biol Cell 18:265–281Enloe B, Diamond A, Mitchell AP (2000) A single-transformationgene function test in diploid Candida albicans. J Bacteriol182:5730–5736Epp JA, Chant J (1997) An IQGAP-related protein controls actin-ringformation and cytokinesis in yeast. Curr Biol 7:921–929Gillum AM, Tsay EY, Kirsch DR (1984) Isolation of the Candida albicansgene for orotidine-5-phosphate decarboxylase by complementationof S. cerevisiae ura3 and E. coli pyrF mutations. MolGen Genet 198:179–182Gola S, Martin R, Walther A, Dunkler A, Wendland J (2003) Newmodules for PCR-based gene targeting in Candida albicans: rapidand eYcient gene targeting using 100 bp of Xanking homology region.Yeast 20:1339–1347Karkkainen S, Hiipakka M, Wang JH, Kleino I, Vaha-Jaakkola M,Renkema GH, Liss M, Wagner R, Saksela K (2006) IdentiWcationof preferred protein interactions by phage-display of the humanSrc homology-3 proteome. EMBO Rep 7:186–191Korinek WS, Bi E, Epp JA, Wang L, Ho J, Chant J (2000) Cyk3, a novelSH3-domain protein, aVects cytokinesis in yeast. Curr Biol10:947–950Kumamoto CA, Vinces MD (2005) Contributions of hyphae and hypha-co-regulatedgenes to Candida albicans virulence. CellMicrobiol 7:1546–1554Martin R, Hellwig D, Schaub Y, Bauer J, Walther A, Wendland J(2007) Functional analysis of Candida albicans genes whose Saccharomycescerevisiae homologues are involved in endocytosis.Yeast 24:511–522Mayer BJ (2001) SH3 domains: complexity in moderation. J Cell Sci114:1253–1263Michel S, Ushinsky S, Klebl B, Leberer E, Thomas D, Whiteway M,Morschhauser J (2002) Generation of conditional lethal Candidaalbicans mutants by inducible deletion of essential genes. MolMicrobiol 46:269–280Nobile CJ, Mitchell AP (2009) Large-scale gene disruption using theUAU1 cassette. Methods Mol Biol 499:175–194Noble SM, Johnson AD (2005) Strains and strategies for large-scalegene deletion studies of the diploid human fungal pathogen Candidaalbicans. Eukaryot Cell 4:298–309Oberholzer U, Marcil A, Leberer E, Thomas DY, Whiteway M (2002)Myosin I is required for hypha formation in Candida albicans.Eukaryot Cell 1:213–228Odds FC (1994) Pathogenesis of Candida infections. J Am Acad Dermatol31:S2–S5Ohkuni K, Okuda A, Kikuchi A (2003) Yeast Nap1-binding proteinNbp2p is required for mitotic growth at high temperatures and forcell wall integrity. Genetics 165:517–529Pruyne D, Bretscher A (2000a) Polarization of cell growth in yeast.JCell Sci 113:571–585Pruyne D, Bretscher A (2000b) Polarization of cell growth in yeast. I.Establishment and maintenance of polarity states. J Cell Sci113:365–375Schaub Y, Dunkler A, Walther A, Wendland J (2006) New pFA-cassettesfor PCR-based gene manipulation in Candida albicans.J Basic Microbiol 46:416–429Sudbery P, Gow N, Berman J (2004) The distinct morphogenic statesof Candida albicans. Trends Microbiol 12:317–324Tang HY, Xu J, Cai M (2000) Pan1p, End3p, and S1a1p, three yeastproteins required for normal cortical actin cytoskeleton organization,associate with each other and play essential roles in cell wallmorphogenesis. Mol Cell Biol 20:12–25Veses V, Gow NA (2008) Vacuolar dynamics during the morphogenetictransition in Candida albicans. FEMS Yeast Res 8:1339–1348Veses V, Richards A, Gow NA (2009) Vacuole inheritance regulatescell size and branching frequency of Candida albicans hyphae.Mol Microbiol 71:505–519Walther A, Wendland J (2003) An improved transformation protocolfor the human fungal pathogen Candida albicans. Curr Genet42:339–343Walther A, Wendland J (2004a) Apical localization of actin patchesand vacuolar dynamics in Ashbya gossypii depend on the WASPhomolog Wal1p. J Cell Sci 117:4947–4958Walther A, Wendland J (2004b) Polarized hyphal growth in Candidaalbicans requires the Wiskott–Aldrich Syndrome protein homologWal1p. Eukaryot Cell 3:471–482Walther A, Wendland J (2008) PCR-based gene targeting in Candidaalbicans. Nat Protoc 3:1414–1421123

Curr GenetWarren DT, Andrews PD, Gourlay CW, Ayscough KR (2002) Sla1pcouples the yeast endocytic machinery to proteins regulating actindynamics. J Cell Sci 115:1703–1715Weisman LS, Bacallao R, Wickner W (1987) Multiple methods ofvisualizing the yeast vacuole permit evaluation of its morphologyand inheritance during the cell cycle. J Cell Biol 105:1539–1547Wendland J (2001) Comparison of morphogenetic networks of Wlamentousfungi and yeast. Fungal Genet Biol 34:63–82Wendland J, Philippsen P (2002) An IQGAP-related protein, encodedby AgCYK1, is required for septation in the Wlamentous fungusAshbya gossypii. Fungal Genet Biol 37:81–88Winzeler EA, Shoemaker DD, AstromoV A, Liang H, Anderson K,Andre B, Bangham R, Benito R, Boeke JD, Bussey H, Chu AM,Connelly C, Davis K, Dietrich F, Dow SW, El Bakkoury M,Foury F, Friend SH, Gentalen E, Giaever G, Hegemann JH, JonesT, Laub M, Liao H, Liebundguth N, Lockhart DJ, Lucau-Danila A,Lussier M, M’Rabet N, Menard P, Mittmann M, Pai C,Rebischung C, Revuelta JL, Riles L, Roberts CJ, Ross-MacDonald P,Scherens B, Snyder M, Sookhai-Mahadeo S, Storms RK, VeronneauS, Voet M, Volckaert G, Ward TR, Wysocki R, Yen GS, Yu K,Zimmermann K, Philippsen P, Johnston M, Davis RW (1999)Functional characterization of the S. cerevisiae genome by genedeletion and parallel analysis. Science 285:901–906Zheng XD, Lee RT, Wang YM, Lin QS, Wang Y (2007) Phosphorylationof Rga2, a Cdc42 GAP, by CDK/Hgc1 is crucial for Candidaalbicans hyphal growth. EMBO J 26:3760–3769123