UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Analysis of the hgl1 gene of Ustilago maydis Laidlaw, Robert David 2000

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

Item Metadata


831-ubc_2000-0455.pdf [ 3.83MB ]
JSON: 831-1.0089569.json
JSON-LD: 831-1.0089569-ld.json
RDF/XML (Pretty): 831-1.0089569-rdf.xml
RDF/JSON: 831-1.0089569-rdf.json
Turtle: 831-1.0089569-turtle.txt
N-Triples: 831-1.0089569-rdf-ntriples.txt
Original Record: 831-1.0089569-source.json
Full Text

Full Text

A N A L Y S I S O F T H E hgll G E N E O F Ustilago maydis  by R. DAVID L A I D L A W  B. Sc. (Microbiology and Immunology) University of B. C.  A THESIS S U B M I T T E D IN P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S FOR T H E D E G R E E OF  Master of Science  in  THE F A C U L T Y OF G R A D U A T E STUDIES M I C R O B I O L O G Y and I M M U N O L O G Y and THE B I O T E C H N O L O G Y L A B O R A T O R Y  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH C O L U M B I A  JUNE 2000 ©David Laidlaw, 2000  In  presenting  degree freely  at  this  the  available  copying  of  department publication  of  in  partial  fulfilment  of  the  University  of  British  Columbia,  I  agree  for  this or  thesis  reference  thesis by  this  for  his thesis  and  scholarly  or for  her  Department  DE-6  (2/88)  Columbia  I  further  purposes  gain  shall  that  agree  may  representatives.  financial  permission.  T h e U n i v e r s i t y of British Vancouver, Canada  study.  requirements  It not  be is  that  the  Library  an  granted  by  allowed  advanced  shall  permission  understood be  for  the that  without  for head  make  it  extensive of  my  copying  or  my  written  ABSTRACT  The cAMP/protein kinase A signal transduction pathway plays an important role in morphogenesis in fungi.  Previous work identified a gene encoding a Ustilago maydis  homologue of the catalytic subunit of protein kinase A , adrl, involved in maintaining budding growth and virulence. The filamentous phenotype of an adrl' mutant was used in a genetic screen for suppressor mutations to identify components of the pathway downstream of protein kinase A . This screen resulted in the recovery of a mutation in hgll (hyphal growth locus 1). hgW mutants display a number of phenotypes including suppression of the filamentous growth of an adrl' mutant, the production of a yellow pigment, and a failure to complete the sexual phase of the life cycle. There are no similar proteins to Hgllp in current databases and it is therefore difficult to identify its function by comparison to known genes.  To identify a  functional role for hgll, additional characterization studies have been conducted including further analysis of the predicted amino acid sequence, phosphorylation experiments, and mutagenesis to isolate suppressors of hgll'. The work in this thesis demonstrates that a reversion to the adrl' phenotype could be obtained through the collection of hgll' suppressor mutants.  Additional studies demonstrated that Hgllp serves as a direct target for  phosphorylation by Ustilago maydis protein kinase A in vitro. Analysis of the Hgllp sequence revealed several interesting features of the protein which will be relevant for future experimentation.  u  T A B L E OF CONTENTS  Abstract  11  Table of Contents  iii  List of Tables  vi  List of Figures  v  ii  List of Abbreviations  viii  Acknowledgements  x  1. Introduction  1  1.1 Literature review  1  1.1.1 The dimorphic basidiomycete Ustilago maydis  1 Sexual regulation of the U. maydis dimorphic switch  3 Raslp appears to be involved in filamentous growth in U. maydis  4 c A M P / P K A signaling participates in U. maydis dimorphism  5  1.1.2 Pseudohyphal growth and mating in Saccharomyces cerevisiae  10 Pheromone-response mediated filamentation in S. cerevisiae  11 The pseudohyphal growth response in S. cerevisiae  14 c A M P / P K A signaling regulates pseudohyphal growth in S. cerevisiae  15  1.2 Research basis and objectives  20  2. Materials and methods  23  2.1.1 Strains and media  23  2.1.2 D N A manipulations  23  2.1.3 Hgll sequence analysis  24  in  2.1.4 Web-based algorithms and web-sites used in the analysis of the hgll predicted amino acid sequence  24  2.2.1 Construction of the Hgllp expression construct pEThgl  25  2.2.2 In vitro expression of Hgllp from the plasmid pEThgl  27  2.2.3 Production of recombinant Ukalp and Adrlp  27  2.2.4 Ni-Agarose Ubc::6xHIS affinity purification of P K A from Ustilago maydis  28  2.3.1 Disruption of hgll with the nourseothricin resistance marker natl  28  2.3.2 Suppressor mutagenesis of strains 3020 and 3034  29  3. Results  30  3.1 Sequence analysis of Hgllp  30  3.1.2 Conserved domains in the Hgllp amino acid sequence 3.2 P K A phosphorylation of Hgllp  36 38  3.2.1 Purification of U. maydis P K A from Ubc::6xHIS chromatographic columns 3.2.2 Expression of Hgllp from pEThgl  38 40  3.2.3 Phosphorylation of Hgllp with bovine heart and U. maydis P K A 3.3 Suppressor mutagenesis of hgll' strains  43 44  3.3.1 Disruption of hgll with the natl cassette  45  3.3.2 Isolation of filamentous suppressor mutants of the budding hgll' phenotype  47  3.3.3 Complementation of suppressor mutations in the 3020 and 3034 strains  52  4. Discussion  53  4.1 Summary of Results  53  4.2 Discussion of results and future experiments  55  4.2.1 Hgllp sequence analysis  55  iv  4.2.2 Phosphorylation of Hgllp by P K A  58  4.2.3 Suppressor analysis hgll' strains  61  4.2.4 Complementation of hgll' suppressor mutants  63  4.2.5 Summary of hgll' phenotypes  63  4.3 Conclusions  65  5. References  66  V  LIST O F T A B L E S  1. Proteins identified by P R O P S E A R C H using Hgllp as the amino acid query sequence  35  2. Growth of suppressor mutants characterized on a variety of common laboratory media  51  3. Summary of hgll phenotypes  64  vi  LIST O F F I G U R E S  1. The complex life cycle of Ustilago maydis  2  2. U. maydis signaling pathways involved in filamentous growth and virulence  5  3. Mutations in hgll suppress the filamentous phenotype of adrl' mutants  8  4. The hgll gene affects teliospore formation  9  5. S. cerevisiae Signaling Pathways Involved in Pseudohyphal Growth  12  6. Amino Acid sequence of Hgllp used in bioinformatics sequence analysis  24  7. Construction of the plasmid pEThgl  26  8. Alignment of Hgllp to con4 2801 a n d F l o l l p  31  9. Alignment of the translated C. albicans contig 2801 with the S. cerevisiae protein Flol lp. 34 10. Distribution of domains in the predicted amino acid sequence of Hgllp  37  11. In vitro production of Ukalp and Adrlp  39  12. Evaluation of P K A activity and the effects of PKI and phosphatase treatment on the peptag® kinase assay  40  13. Ni-agarose purification of Ubcl ::HIS and P K A  41  14. Wester hybidization analysis of the myc-tagged protein M c m l p  42  15. Phosphorylation of Hgllp by bovine and U. maydis protein kinase A  43  16. Wheat Germ Extract does not contain intrinsic kinase A activity  44  17. Construction of an allele of hgll containing the nourseothricin disruption cassette  46  18. Disruption of hgll with the natl cassette in the wild type 001 background  47  19. Categorization of different hgll' suppressor mutants  48  20. The pigmentation phenotype of an hgll' strain  49  21. Hgllp resembles the yeast transcription factors Flo8p and Sfllp  60  22. Hgllp suppressors represent a variety of functions  62  vii  LIST O F A B B R E V I A T I O N S  adrl  aromatic dicarboximide resistance gene, encodes the catalytic subunit of U. maydis P K A  adrr.phleo  adrl insertional disruption strain carrying the phleomycin resistance marker  bE/bW  U. maydis regulatory heterodimer formed by the b East and b West proteins  cAMP  cyclic adenosine monophosphate  FL08  Flol lp-regulating transcription factor in S. cerevisiae (regulated by Tpk2p).  FLOl 1  cell surface flocculin protein in S. cerevisiae  GPCR  G-protein Coupled Receptor  hgll  hyphal growth locus 1 gene  hgll::natl  insertional disruption of the hgll gene with the nourseothricin cassette  MAPK  mitogen activated protein kinase  MAPKK  M A P K kinase  MAPKKK  M A P K K kinase  PEST regions regions of a protein rich in Proline (P), Glutamate (E), Serine (S), and Threonine (T) residues PKA  protein kinase A  PKI  specific inhibitor of protein kinase A catalytic subunits  bPKA  bovine Protein kinase A  UPKA  U. maydis Protein kinase A  prfl  U. maydis gene encoding the pheromone response factor  rasl  a gene encoding a small GTP-binding protein  SFL1  transcription factor in S. cerevisiae (regulates flocculation).  viii  TPK2  gene encoding the catalytic subunit of P K A in S. cerevisiae  uacl  U. maydis adenylyl cyclase  ubcl  gene encoding the regulatory subunits of protein kinase A in U. maydis  Ubc::6xHIS  6xHistidine-tagged version of Ubc protein for purification over Nickel columns  URL  uniform resource locator  v|/-BLAST  Position-Specific-Iterated (PSI = v|/) B L A S T  ix  ACKNOWLEDGMENTS  I would like to thank Jim Kronstad for his guidance, patience, personality and encouragement during my studies as a grad student. There are numerous friends and family members that have helped me at various times in my studies. O f them all, my dad deserves special thanks, his unconditional support in all aspects of my life has always made decision making an easy task.  Introduction  1. Introduction  1.1 Literature review 1.1.1 The dimorphic basidiomycete Ustilago maydis  The basidiomycete fungal pathogen Ustilago  maydis has emerged as a valuable  organism for the study of fungal morphogenesis and dimorphic switching. U. maydis is an ideal organism for this type of research because of its relatively rapid growth rate, its simple growth requirements, its benign nature towards researchers and its ease of genetic manipulation. Genes for morphogenesis that are identified in U. maydis will likely be informative for other fungal systems due to the high levels of conservation among components of signaling pathways, especially among basidiomycete fungi (e.g., U. maydis and the human pathogen Cryptococcus neoformans).  U. maydis requires infection of the monocotyledon Zea mays (corn) for completion of the sexual phase of the life cycle (Banuett, 1995; Christensen, 1963). The complex life cycle of U. maydis includes saprophytic and parasitic stages, and infection of host tissues is essential for the formation of sexual spores (Figure 1). Saprophytic haploids (N) of compatible mating type are capable of cell fusion to form a filamentous and infectious dikaryon (N+N). Alternatively, under  conditions  of  stress  or  nutrient  starvation,  1  the  cells  can  form  chains  Introduction  Figure 1. The complex life cycle of Ustilago maydis. 1. Haploid budding cells; 2. Chlamydospores; 3. Filamentous, infectious dikaryon; 4. Dormant, diploid teliospores; 5. Germinating teliospores.  of rounded, desiccation-resistant, haploid chlamydospores (Kusch and Schauz, 1989).  The  parasitic dikaryon establishes itself in the host and initiates tumor formation; this cell type cannot survive outside of the host plant. Following invasion of host tissues and completion of the proliferative stage of infection, the dikaryon undergoes karyogamy (nuclear fusion) and additional morphological changes that result in formation of the highly melanized, dormant, diploid teliospores (2N) (Christensen, 1963; Holliday, 1974). The accumulation of teliospores within tumors in the plant produces the hallmark black, sooty 'smut' stage of the disease. Completion of the life cycle is signaled by teliospore release from ruptured tumors; these spores subsequently germinate and undergo meiosis to generate haploid sporidia.  The ability of U.  maydis to infect the host plant is tightly correlated with its ability to switch from a budding to a filamentous phenotype - as occurs during formation of the dikaryotic cell type. This switch is  2  Introduction  currently known to be controlled by two mechanisms: a mating response pathway and a cAMP/protein kinase A (PKA) pathway. Sexual regulation of the U. maydis dimorphic switch Mating compatibility between haploid cells is determined by possession of non-identical alleles at the a and b loci.  The a locus carries genes responsible for the production of  pheromones and pheromone receptors (Bolker et al., 1992; Spellig et al., 1994). The products of the a locus are necessary for events which occur prior to fusion between haploids, such as recognition of compatible mating types and subsequent conjugation tube extension (Snetselaar et al., 1996; Spellig and Reichmann, 1994). The b locus encodes two proteins named bE and bW (b East and b West) that associate with one another and form a novel regulatory protein termed the bE/bW heterodimer (Gillissen et al., 1992). Because bE and bW must originate from different b loci for dikaryon formation, conjugation between haploid cells of different matingtype is essential for dikaryon establishment. The bE/bW heterodimer initiates events that result in formation of the dikaryon, and is necessary for maintaining filamentous growth during the pathogenic portion of the life cycle (Gillissen et ai, 1992). The U. maydis pheromone response pathway governs gene transcription through a M A P kinase signaling cascade which activates the pheromone response factor Prflp. This HMG-box transcription factor regulates the transcription of genes at the a and b mating-type loci.  A  M A P K K and a M A P K have been identified, and these enzymes are encoded by the genes fuz7, and ubc3, respectively (Figure 2) (Banuett and Herskowitz, 1996; Mayorga and Gold, 1999). Deletion of either fuz7 or ubc3 affects morphological switching and pheromone response-related gene induction. These processes require Prflp. Epistasis experiments suggest that Fuz7p and Ubc3p may act in different M A P kinase cascades, but additional investigation is still required. ubc3 mutants do not demonstrate mating assay interactions (filament formation in response to  Introduction  compatible haploid cells) suggesting that they are compromised for dikaryon formation. However, the pathogenicity of these mutants is not abolished. Prflp contains M A P kinase phosphorylation sites and binds cis-acting D N A sequences to upregulate pheromone response genes (Hartmann et al., 1996; Urban et al., 1996).  Elimination of the M A P kinase target  residues in Prflp eliminates upregulation of pheromone response genes upon mating (Muller et al., 1999). Elevated levels of P K A in a ubc3 mutant have a detrimental effect on pheromone induced gene transcription. This result raises the possibility that the M A P kinase and cAMP pathways converge on Prflp (Muller et al., 1999). R a s l p appears to be involved in filamentous growth in U. maydis Recent work in our laboratory has identified a role for rasl in the process of establishing filamentous growth, although it is still unclear how the protein functions in each of the known pathways. A dominant active rasl allele {rasl ) VAL,g  is capable of causing mild filamentation in  a high P K A background, but is not epistatic to mutations in the M A P kinase pathway. Cells carrying fuzT, ubc3~, or prfl' mutations maintain their budding phenotype upon transformation with the dominant active rasl allele. Pathogenicity is abolished by deletion of rasl, indicating an essential role for this signaling molecule during the infectious cycle (N. Lee, pers comm.).  4  Introduction  External signals  Gpa3  Filamentous Growth, Virulence, Pathogenic Development  Figure 2. U. maydis signaling pathways involved in filamentous growth and virulence. The figure is a combination of models proposed by Durrenberger et al. (1998) and Kallmann et al. (2000). Question marks represent areas of uncertainty or areas currently under investigation. The involvement of the Ras-like protein has been established by N . Lee (pers comm.) See the accompanying text for gene names. c A M P / P K A signaling participates in U. maydis dimorphism  The c A M P / P K A pathway plays key signaling roles in many different fungi. Often P K A signaling has regulatory effects on the development of morphological structures or other events required for virulence. Organisms often have more than one gene encoding the P K A catalytic subunit.  For example, the yeast Saccharomyces cerevisiae has three genes and the highly  homologous subunits were once thought to act redundantly. This has been disproved by recent  5  Introduction  work described below. P K A exists as a complex of catalytic and regulatory subunits and these subunits dissociate upon binding of c A M P to the regulatory protein. The regulatory subunit inhibits the protein kinase activity of the catalytic subunits and hypomorphic mutations in the regulatory subunit will increase the activity of P K A . In contrast, mutations in adenylyl cyclase or an activator of adenylyl cyclase (often a G-protein a subunit or Ras protein) will greatly decrease P K A activity. The U. maydis G-protein a subunit Gpa3p signals through the c A M P pathway and (among other events) is involved in upregulation of expression of the pheromone-encoding gene mfal (Kruger et al., 1998).  Strains carrying constitutively active Gpa3p have a yeast-like  colony phenotype similar to strains missing the regulatory subunit of P K A . Mutants carrying disrupted versions of gpa3 display filamentous growth in a haploid background similar to cells deficient in adenylyl cyclase, and are also non-pathogenic upon mating (Kruger et al., 1998). These findings provide further support for the involvement of the P K A pathway in filamentous growth and virulence. The cAMP-dependent P K A pathway was discovered in U. maydis during studies on the connection between cellular morphology and virulence (Barrett et ai, 1993).  Because the  infectious dikaryon is a filamentous cell type, mutants which were either defective, or constitutive, for filamentous growth were identified and investigated. Ustilago adenylyl cyclase or uacl was identified as a gene necessary for normal budding growth and pathogenicity (Barrett et al., 1993).  The filamentous phenotype of adenylyl cyclase mutants can be  suppressed by the addition of 6mM exogenous c A M P to culture medium (Gold et al, 1994). Complementation of a uacl suppressor mutation by Gold et al. (1994) recovered the gene encoding the regulatory subunit of cAMP-dependent protein kinase, or ubcl (Ustilago bypass cyclase),  ubcl mutants presumably have high P K A activity, grow with a multiple-budding 6  Introduction  morphology, and are unable to form the filamentous cell type'during laboratory mating assays (Durrenberger et al., 1998; Gold et al., 1994).  Surprisingly, compatible haploids carrying  mutations in ubcl are still pathogenic (with highly attenuated symptoms) despite their apparent inability to form filaments as haploids (Gold et al, 1994). These findings implicate cAMP pathway signaling in events besides filament formation during pathogenesis.  Durrenberger et  al. (1998) later identified the U. maydis P K A catalytic subunit Adrlp as being essential for pathogenesis. The gene encoding the catalytic subunit of P K A (adrl) was first identified by Orth et al. (1994) in a mutant strain demonstrating resistance to the dicarboximide fungicide vinclozolin. Later work by Durrenberger et al. (1998) revealed that strains defective for adrl constitutively filamentous, but (surprisingly) non-pathogenic  in the presence of  are adrl'  compatible mating partners. The filamentous phenotype of adrl' mutants complicates mating interaction assays, which normally result in filamentous mating colonies when compatible wildtype haploids are co-cultured on culture medium. Therefore it is difficult to determine whether adrl' mutants are unable to mate, or, alternatively, are compromised during the proliferative stage of infection within the host. Conceivably, compatible ,adrl~ haploids cannot form the dikaryon, and are therefore unable to form the bE/bW heterodimer.  However, diploid strains  carrying homozygous adrl' deletions (which do not require mating interactions to form bE/bW) are also unable to infect plants (Durrenberger et al., 1998).  There are likely other hurdles  encountered by U. maydis during infection that can not be overcome without functional P K A . In an effort to identify pathway components downstream of P K A and involved in U. maydis morphogenesis, a suppressor mutant was recovered which displayed a budding phenotype in the adrl' background (Figure 3). Complementation of the mutant recovered hgll,  7  Introduction  a previously unidentified gene containing no significant similarity to sequences in existing  A  B  C  ^  ^  •  1  ^x*^  Figure 3. Mutations in hgll suppress the filamentous phenotype of adrl' mutants. Panel A represents an adrl' mutant colony at 40x, panel C presents the corresponding filamentous cell type of adrl' mutant cells. Panel B shows colonies from an adrl' strain that also carries a mutation in hgll, and panel D represents the corresponding cell type to the adrl'/hgll' colonies of panel B. Photos courtesy Franz Diirrenberger.  databases. Recent work from our laboratory reveals that Hgllp functions during events after cell fusion between haploids that are necessary for successful teliospore formation (unpublished data), hgll' mutants cause normal symptoms in the host such as anthocyanin production and the formation of tumors, but do not form teliospores and as such are compromised in their ability to complete the sexual phase of the life cycle (Figure 4). The involvement of hgll in the mating  8  Introduction  Figure 4. The hgll gene affects teliospore formation. Panel A demonstrates ears o f corn infected with either compatible w i l d type haploids (right), or compatible haploids carrying disruptions in the hgll gene. Panel B demonstrates a tissue cross-section from a kernel infected with hgll' strains. Panel C represents a tissue cross-section from a kernel infected with compatible wild-type strains o f U. maydis. Black bars measure 20pm, panels B and C photos courtesy Franz Durrenberger.  response  (i.e., in the sexual phase o f the life cycle and mating assays), as well as its  involvement in the P K A pathway (as a suppressor o f the adrl'  phenotype), suggests that it  might act as a mediator between the two pathways. This thesis presents additional studies on the hgll gene and its product. To understand the potential activity o f H g l l p as a target o f P K A , it is useful to review the organization o f the  Introduction  c A M P / P K A pathway in the well characterized yeast Saccharomyces  cerevisiae.  Where  appropriate, the following description of signaling in yeast will point out parallels with cAMP signaling in U. maydis. 1.1.2 Pseudohyphal growth and mating in Saccharomyces cerevisiae  As a model organism, the budding yeast Saccharomyces cerevisiae illustrates both the complexity  and  the  conservation  of components  in  signal  transduction  pathways.  Morphogenesis in S. cerevisiae is controlled by components of the mitogen activated protein kinase pathway ( M A P kinase pathway) and the c A M P / P K A pathway.  The M A P kinase  pathway responds to signals sent through pheromone and environmental (osmotic pressure, ammonium) sensors and regulates cell elongation and invasion (Pan and Heitman, 1999). The cAMP/PKA  pathway  responds  specifically to nutrient  conditions in the  surrounding  environment. Current studies suggest that a convergence of the M A P kinase and c A M P / P K A pathways coordinately regulate the morphological response.  As described above there is  evidence for similar convergent pathway control of morphogenesis in U. maydis. examination of signaling in S. cerevisiae  An  provides a framework for understanding the  components expected in other fungi. A n evaluation of the M A P kinase and cAMP signaling pathways in S. cerevisiae is presented in the following text; links between the two transduction cascades and their similarities to the Ustilago system are highlighted where possible. M A P kinase signaling in S. cerevisiae can be stimulated by osmotic stress, pheromones, cell wall damage, and nutrient starvation (Banuett, 1998; Gustin et al, 1998; Herskowitz et al., 1995). Signals from the different stimuli are transduced by conserved components, yet can result in the activation of specific molecules for each pathway that cause the morphological responses.  M A P kinase activation during pheromone response and pseudohyphal growth will  be presented here. The conservation of components shared between the two pathways suggests 10  Introduction  there must be mechanisms to prevent 'crosstalk', or inappropriate signaling through shared effector molecules. The current hypothesis about the sorting of different signals postulates that 'scaffold' proteins associate with the specific components necessary for alternative signaling outcomes (Madden and Snyder, 1998; Madhani and Fink, 1998).  The final outcomes of  signaling are dictated largely by the specific M A P kinase molecules activated at the end of the cascade, and its target proteins. The outcome of signaling through the M A P kinase pathway for pheromone response is to regulate genes necessary for cell fusion and to arrest the cell cycle in preparation for mating. These combined processes culminate in 'schmoo' formation (Banuett, 1998; Dolan et al, 1989; Errede and Ammerer, 1989; Miyajima et al, 1987; Whiteway et al, 1989). Pheromone-response mediated filamcntation in S. cerevisiae  The pheromone response M A P kinase pathway is initiated by the binding of pheromones to the pheromone receptors encoded by STE2/STE3  (Figure 5).  The Py subunits of the  heterotrimeric G-protein (encoded by STE4, and STE18) transmit signal from Ste2p and Ste3p to the immediate downstream components Cdc42p, Ste20p and Ste5p (Akada et al, 1996; Hirschman et al, 1997; Leberer et al, 1997b; Madhani and Fink, 1997; Miyajima et al, 1987; Mosch et al, 1996). Pheromone signal is transduced by Ste20p (p21 activated kinase) to the M A P kinase cascade encoded by STE5 (scaffold protein) STE11 ( M A P K K K ) , STE7 (MAPKK) and FUS3 (MAPK) (Leberer et al, 1997a). The U. maydis M A P kinase cascade components Fuz7p ( M A P K K ) and Ubc3p (MAPK) are homologs of Ste7p and Fus3p respectively. Additional upstream and downstream components will likely be identified in U. maydis that are similar to many of the components of the S. cerevisiae system. The Ste5p scaffold protein interacts with many cascade proteins bringing separate components of the cascade into close association to transduce signal (Akada et al, 1996). Got mutations (gpal) resulting in free Py 11  Introduction  subunit are dependent on Ste20p and Ste5p to constitutively activate the M A P kinase cascade for pheromone response (Hirschman et al., 1997). Cdc42p involvement is not essential for pheromone-initiated signal transduction, but may be relevant for component localization to the site of morphogenesis because Cdc42p interacts with the actin cytoskeleton and other structural proteins (Akada et al, 1996; Leberer et al, 1997b; Leeuw et al., 1995). Syglp and Akrlp, two  Pseudohyphal Growth  Gpa2  >-  J  R a s 2  >Cdo42 j [Bmh1,2 ]  Unipolar Budding, Pseudohyphal Growth  Ro11 j  Pheromone Response Ste2,3 t Ste4,18^ j VCdc42 Ste2CTi J  Cell Elongation, • Pseudohyphal Growth  Cell Adhesion  Pseudohyphal Growth  Mating  Figure 5. S. cerevisiae Signaling Pathways Involved in Pseudohyphal Growth. Compiled from the pathway models presented by Pan and Heitman 1999, Fink et al. (1999), Robertson and Fink (1999), Banuett (1998), and Leberer, Thomas and Whiteway (1997). See the accompanying text for definitions of the gene names.  12  Introduction  proteins of unknown function in the M A P kinase pathway (not shown), also associate with the G-protein Py subunit. Mutations in Syglp cause no discernable phenotype, whereas mutations in Arklp inhibit M A P kinase signaling as well as suppress the gpal mutant phenotype, indicating A k r l p may also play an additional role in the c A M P pathway (Kao et ah, 1996; Pryciak and Hartwell, 1996). The regulation of transcription factors is a major role of M A P kinase signaling. The Stellp/Ste7p kinases activate Fus3p, the M A P kinase responsible for activating downstream proteins during the pheromone response. Fus3p phosphorylates the transcription factor Stel2p, which then binds a D N A pheromone response element (PRE) to regulate growth morphology and mating interactions (Elion et al, 1993; Hung et al., 1997; Kronstad et al, 1987). Stel2p can be phosphorylated by both Fus3p and the M A P kinase Ksslp of the pseudohyphal M A P cascade, but only Ksslp can phosphorylate Diglp and Dig2p, the negative regulators of Stel2p (Cook et al, 1996; Tedford et al, 1997). These subtleties in M A P kinase specificity likely account for some of the differences between the pheromone-induced and pseudohyphal response growth morphologies. The cyclin-dependent kinase inhibitor Farlp, which is also a target of  Fus3p  phosphorylation, regulates cell cycle arrest during the mating response. Farlp interacts with the SH3 domain protein Bemlp, and therefore may also function in localizing cascade components (Lew and Reed, 1995).  Bemlp independently associates with Ste5p, Ste20p, Cdc24p (not  shown), Cdc42p, and actin, and participates in polarized growth during mating to localize signaling components to the schmoo tip (Leeuw et al., 1995; Lyons et al., 1996). Protein localization studies related to signaling and morphogenesis are starting to be carried out in U. maydis, and a kinesin protein (Kin2p) related to Nkin of N. crassa has been identified as being required for filamentous growth and virulence (Lehmler et al., 1997). Kin2 13  Introduction  mutants have greatly reduced aerial hyphae in U. maydis mating reactions and demonstrate reduced virulence compared to wild-type cells. It is unclear how these motor proteins function during filamentous growth, but they may play a role in localizing necessary components to the site of morphogenesis. The pseudohyphal growth response in S. cerevisiae M A P kinase signaling for the control of pseudohyphal growth in S. cerevisiae utilizes many of the same components as the pheromone response pathway, yet is initiated by nitrogen starvation and results in elongated cells capable of invading solid media.  The upstream  initiators of the M A P kinase cascade during the pseudohyphal growth response are Ras2p, Cdc42p, and Ste20p (Figure 5) (Gimeno et al., 1992; Mosch and Fink, 1997; Mosch et ai, 1996; Zhao et al, 1995). The 14-3-3 proteins Bmhlp and Bmh2p interact with Cdc42p and Ste20, and are required for Ras2-dependent pathway activation (Leberer et al, 1997b; Peter et al, 1996).  Bmhlp and Bmh2p may also function in the c A M P / P K A pathway to prevent  glycogen accumulation (Roberts et al, 1997).  Signaling via the M A P kinase cascade for  pseudohyphal growth results in the activation of the M A P kinase Ksslp, which ultimately activates the transcription factors Stel2p and Teclp. Teclp forms heterodimers with Stel2p and these regulate TEC1 expression, cell elongation, and the production of additional proteins such as the cell surface flocculin Flo l i p (Baur et al, 1997; Gavrias et al, 1996; Lo and Dranginis, 1998; Madhani and Fink, 1997). Nitrogen availability also regulates FLO 11 expression in that a high abundance of nitrogen results in low FLO 11 transcript levels. Overexpression of FLO 11 can mimic nitrogen starvation conditions during growth on rich media and causes agar invasion by the cells. In contrast, deletion of FLO 11 can eliminate pseudohyphal growth and agar invasion.  14  The complex  Introduction  promoter region of FLOll  is also responsive to the transcription factor Flo8p, an effector of the  c A M P / P K A pathway; this finding represents an exciting connection between the two signaling pathways that will be discussed below (Rupp et al., 1999). c A M P / P K A signaling regulates pseudohyphal growth in S. cerevisiae The S. cerevisiae c A M P / P K A pathway has a positive effect on pseudohyphal growth. Cells with elevated P K A activity grow as long chains of budding cells and invade solid media. Additionally, elevated P K A in S. cerevisiae results in loss of carbohydrate reserves, a failure to arrest in the G l phase of the cell cycle, and increased sensitivity to stress (Tokiwa et al, 1994). Conversely, low P K A levels produce unbudded cells arrested in G l ; these cells accumulate carbohydrates and demonstrate increased resistance to stress (Thevelein and de Winde, 1999). The glucose sensitive G-protein coupled receptor (Gprlp) activates the G-protein asubunit Gpa2p to initiate the P K A pathway at the level of adenylyl cyclase (Cyrlp) (Figure 5) (Thevelein and de Winde, 1999; Xue et al, 1998). Gpa2p is a group III Got subunit similar to Gpa3p of U. maydis, a relationship identified during phylogenetic analysis of fungal G-proteins by M . Bolker.  Other members of this group are also Got subunits that stimulate adenylyl  cyclase , such as Gpa2p of S. pombe (Bolker, 1998). S. cerevisiae Gpa2p also activates the cAMP pathway in response to nitrogen starvation to cause pseudohyphal growth, and activated forms of Gpa2p initiate pseudohyphal growth in the absence of nitrogen starvation.  Ammonium permease (Mep2p) mutants demonstrate a  budding growth phenotype instead of pseudohyphal growth and this can be suppressed by cAMP addition or by mutations in the gene encoding the regulatory subunit of P K A (BCY1) (Lorenz and Heitman, 1998). Similarly, mutation of RAS2 to the dominant active  RAS2  WaU9  allele causes c A M P production by adenylyl cyclase (CYR1) and activation of P K A (Field et al, 15  Introduction  1998; Toda et al, 1985). Constitutive activation of Ras2p can restore pseudohyphal growth in mep2 mutants, possibly due to Ras2-dependent activation of Cyrlp (Lorenz and Heitman, 1997). ras involvement in filamentous growth has been identified in U. maydis by Nancy Lee in our laboratory, and might be mediated by components of the P K A and M A P kinase pathways {pers. comm.). Overall, the S. cerevisiae data indicate that nitrogen sensing and pseudohyphal growth are mediated by the P K A pathway and involve GPA2 and RAS2. Activation of P K A by mutations in the gene (BCYI) encoding the P K A regulatory subunit in S. cerevisiae restores pseudohyphal growth in the absence of the transcription factors STE12 and TEC1, implying that there are alternative pathways to M A P kinase activation of filamentous growth (Lorenz and Heitman, 1998; Pan and Heitman, 1999). There are three S. cerevisiae P K A catalytic subunits (encoded by TPKI, TPK2, and TPK3), which were at one time assumed to be redundant in function. However, despite their highly conserved C-termini (which encode the catalytic domains), Tpklp, Tpk2p, and Tpk3p function differently upon activation. The C-terminal regions of the kinases determine target specificity while a function for the N-terminal regions remains unknown; the amino terminal region is not responsible for determining specificity (Pan and Heitman, 1999).  Studies have shown that individual  homozygous deletions of each TPK gene in diploids cause distinct phenotypes, whereas deletion of all three TPK's is lethal. Tpk2p acts as an activator of filamentous growth and Tpklp and Tpk3p appear to be inhibitory for filament formation.  Mutations in TPK2 are epistatic to  mutations in TPKI or TPK3 with regard to filamentation (Pan and Heitman, 1999).  TPK2  transcript levels are not controlled by Tpklp and Tpk3p, which likely act either upstream or downstream of Tpk2p and may involve a feedback loop inhibiting c A M P production (Pan and Heitman, 1999). Unlike tpk2 mutants, tpkl and tpk3 mutants are still responsive to ammonium sulfate but exhibit a reduced pseudohyphal phenotype. 16  U. maydis has two genes encoding P K A  Introduction  catalytic subunits, ukal and adrl.  Although adrl  and ukal demonstrate high sequence  similarity, the ukal allele does not greatly alter cell morphology or virulence when disrupted in the wild type background under the conditions tested (Durrenberger et al., 1998). The discovery of different functions for the three S. cerevisiae P K A genes suggests a role for the ukal gene of U. maydis may yet be revealed. Perhaps the function of ukal will become apparent as the Ustilago c A M P and M A P kinase pathways become better understood. Several targets of Tpk2p have been identified, including the stress regulators Msn2p and Msn4p, the protein kinase Riml5p, the transcription factors Flo8p and Sfllp, and the daughter cell-specific protein Ashlp. The transcription factors Msn2p and Msn4p are regulated by P K A dependent phosphorylation, which prevents their nuclear localization (Gorner et al., 1998; Marchler et al., 1993; Smith et al., 1998). Msn2p and Msn4p regulate stress response element (STRE) -controlled genes and are necessary for expression of the YAK1 protein kinase, which is antagonistic to growth under low P K A conditions (Boy-Marcotte et al., 1999; Martinez-Pastor et al., 1996). The mechanism of Yaklp kinase-dependent growth suppression is not known, but may involve the suppressor of kinase gene, (SOK2) discussed later. The protein kinase Riml5p is negatively regulated by P K A phosphorylation, and is required for meiotic and STRE gene expression (Vidan and Mitchell, 1997). This result, combined with riml5 mutant suppression of a tpkl/2/3 triple mutant, indicates that some genes required for entry into GO are inhibitory to growth (Reinders et al., 1998). P K A regulation of Flo8p governs transcription of the gene encoding the cell surface flocculin Flo l i p , the cell wall protein essential for pseudohyphal growth and agar invasion. Flo l i p also participates in M A P kinase-regulated pseudohyphal growth. The dual regulation of FLO 11 occurs in its promoter region, which has recognition sites for both Flo8p and the Stel2p/Teclp heterodimer (Rupp et ai, 1999). Over-expression of FLOll  can suppress mutations in the transcription factors FL08 and STE12, which normally  17  Introduction  abolish pseudohyphal growth (Lambrechts et al, 1996; L i u et al, 1996; Lo and Dranginis, 1998). Flo8p contains 5 P K A consensus sequences "[RK] (2)-x- [ST]" and FL08  transcript  levels are not affected by P K A , suggesting a direct protein-protein interaction is necessary for regulation (Pan and Heitman, 1999).  The helix-turn-helix transcription factor Sfllp also has 5  P K A consensus sequences, and interacts with Tpk2p by yeast two-hybrid analysis (J. Heitman, pers comm).  Mutations in SFL1 are epistatic to mutations in TPK2 and result in increased  flocculation, invasive growth, and filamentation (Robertson and Fink, 1998). The daughter-cell specific transcription factor Ashlp acts as a downstream effector of the P K A pathway, and physically interacts with Tpk2p (J. Heitman pers. comm.). ashl mutants are not capable of growing by the pseudohyphal phenotype and are also incapable of invasive growth. Epistasis analysis supports Ashlp as part of the P K A pathway, and suggests that it is not involved in the M A P kinase pathway (Chandarlapaty and Errede, 1998). As described, the transcription factors Flo8p and Sfllp each contain 5 P K A target sequences, and demonstrate genetic phenotypes in the P K A pathway of 5". cerevisiae.  By comparison, the Hgllp protein of U. maydis contains 8  P K A target sequences, and demonstrates genetic involvement in processes controlled by the M A P kinase and P K A pathways.  Indeed the Hgllp protein may require phosphorylation by  P K A to regulate budding growth in U. maydis. The SOK2 gene encodes a transcription factor required for STRE gene expression that may also act as an effector of cAMP-dependent P K A (Ward et al, 1995). Mutations in SOK2 produce a pseudohyphal phenotype, suggesting that it may act as a suppressor of filamentation (Ward et al, 1995). The SOK2 homologue PHD1 does not induce pseudohyphal growth in diploids when overexpressed in the absence of nitrogen starvation, but is required for the pseudohyphal growth of sok2 mutant strains.  Diploids carrying PHD1 homozygous deletions  exhibit pseudohyphal growth indicating it is not the only effector in this mechanism.  18  Introduction  The evolving understanding of regulation of mating and pseudohyphal growth in S. cerevisiae indicates that the responses are mediated by complex combinations of different pathways.  Convergence of the pathways at succinct points does, however, represent an  interesting trend in signal transduction, one that might be similar to signal transduction in U. maydis.  19  Introduction  1.2 Research basis and objectives  Phenotypically, hgll appears to participate in both MAP-kinase and PKA-regulated processes. Identification of the hgll gene as a suppressor of the filamentous growth phenotype of an adrl' mutant indicates it functions as a part of the cAMP-dependent protein kinase pathway during filamentous growth and virulence. Additionally, involvement of hgll in events governed by mating interactions such as the formation of teliospores near the end of the sexual phase of the life cycle, indicates either that P K A participates in these processes, or that hgll is also involved in processes other than those under P K A control. Studies to examine Hgllp participation in each of these processes would further develop our understanding of signal transduction and dimorphism in U. maydis. The objective of this study was to investigate Hgllp involvement in the U. maydis dimorphic switch. Three approaches were employed; 1. web-based bioinformatics analysis of the Hgllp amino acid sequence; 2. P K A phosphorylation of the hgll gene product; 3. Isolation of suppressors of hgll' strains to identify genes located downstream of hgll that may function in the dimorphic switch. The development of web-based molecular biology tools has provided a medium by which gene and protein sequence data can be compared to databases to assist in the prediction of gene function. The premise is that high similarity between sequences may indicate homology and suggest that the proteins perform a similar function. The recovery of novel genes with low overall similarity to the databases presents an interesting situation. That is, the excitement of working on a novel sequence is balanced by the frustration of not finding a clue to the function of the protein. One approach to circumventing low overall similarity to other sequences is to look for conserved motifs (or domains) within the sequence of interest. The predicted amino acid sequence of hgll did not share high similarity to sequences in existing databases under  20  Introduction normal search conditions. When low complexity regions were allowed during B L A S T searches, Hgllp demonstrated a very weak similarity to two sequences, a previously unidentified ORF from the Candida albicans genome sequencing project (named con4 2801), and the S. cerevisiae cell surface flocculin Flo 11 p. Comparison of the S. cerevisiae and C. albicans sequences with one another revealed that the sequences shared a high degree of similarity. Searching domain databases with the Hgllp amino acid sequence identified several putative functional motifs in the protein, and led to the phosphorylation experiments discussed in this thesis. The constitutive budding phenotype of strains containing increased levels of P K A (ubcl' mutants) and the constitutive filamentous phenotype of strains defective for P K A (adrl' mutants) suggests that P K A activity is responsible for controlling cellular morphology. Increased P K A activity may be responsible for either promoting budding growth, or inhibiting filamentous growth.  Adrlp has previously been demonstrated  as the catalytic subunit  responsible for the majority of P K A activity in U. maydis cells (Durrenberger et ah, 1998). A n examination of the Hgllp predicted amino acid sequence suggests there may be as many as 8 P K A phosphorylation sites.  hgll  Suppression of the adrl' filamentous phenotype by mutation in  suggested that Hgllp protein was responsible for preventing budding growth in the  absence of P K A activity, and may allow budding growth in the presence of high P K A activity. Is Hgllp a direct target of P K A ?  To demonstrate an interaction between Hgllp and P K A ,  Hgllp was used in phosphorylation experiments with U. maydis and bovine heart P K A . The c A M P / P K A pathway can be dissected with suppressor mutagenesis to recover genes involved in the morphological change from budding to filamentous growth in U. maydis (see Gold et al., 1999 for example). Colony growth morphology is a useful screen for components involved in the pathway, and is effected by mutations in uacl, ubcl, adrl and hgll. To further characterize the c A M P / P K A pathway, a suppressor analysis of filamentous mutants in the 21  Introduction  budding hgll' background was undertaken. Two mutagenesis experiments were conducted, one in a mutant carrying a single hgll' disruption, and the second in an adrl'hgll'  double disruption  background. The work in this thesis demonstrates that Hgllp served as a direct target of Ustilago maydis protein kinase A phosphorylation in vitro.  In addition, suppressor mutants which  reverted the hgll' phenotype to the adrl' morphology were obtained, suggesting that Hgllp was not required for filamentous growth, and may act to inhibit budding. These results also suggest that downstream targets of Adrlp and Hgllp remain to be identified. Analysis of the Hgllp amino acid sequence with web-based bioinformatics tools revealed several interesting features of the protein; these will be relevant for future experimentation.  22  Materials and Methods  2. Materials and methods  2.1.1 Strains and media  U. maydis strain 001 (a2b2) was obtained from R. Holliday (Commonwealth Scientific and Industrial Research Organization, Laboratory of Molecular Biology, Sydney Australia). A l l U. maydis strains were derived from the 001 wild-type strain. Media for growing U. maydis cells was PDB or P D A (potato dextrose broth or agar), C M , D C M , D C M S , (complete medium, double complete medium, double complete medium + 0.8 M sorbitol; (Holliday, 1974)). The medium for screening mutants was D C M + 1% activated charcoal (Day and Anagnostakis, 1971). E. coli strain DH5cc (Bethesda Research Laboratories) was used for cloning procedures, and was grown in L B (Luria Bertani) medium (Sambrook et al., 1989).  E. coli strain  BL21(A,DE3) was used for protein expression experiments.  2.1.2 D N A manipulations  Small scale plasmid preparations were accomplished with standard alkaline lysis in SDS as described in Sambrook et al. (1989). Large scale plasmid preparations for use in sequencing reactions, protein expression experiments, or transformations of U. maydis was prepared with a Qiagen midi kit (Hilden, Germany).  Transformation of U. maydis were accomplished as  described by Wang et al. (1989) and disruption of the hgll gene was confirmed by standard genomic southern hybridization analysis (Sambrook et al., 1989). The D N A fragment used for hybridization of the southern blot was isolated from an agarose gel and purified from the agarose with a Gene Clean kit from Biol01 (Vista, CA).  23  Materials and Methods  2.1.3 Hgllp sequence analysis  The Hgllp sequence used for database analysis was the predicted amino acid sequence from the open reading frame determined to be necessary for complementation of a mutant defective in hgll (Figure 6). The sub-cloned coding region for the peptide sequence in figure 6 is capable of reverting an adrl'hgll'  mutant (budding growth phenotype) to the adrl'  filamentous phenotype (data not shown).  1  MFVATRRPSPTPSSPAVSRRSSWVDLPVASAPSSRRGSRVDLSIQTSLANTSVGSDSHAG  60  61  GSHTKIAINQLLTNLSHVSQEQAI^SSPRSSHSPAIJNDSPALKRKFEADSDSSTAYKVAS  120  121  AHEIHLPTDPVRRASIINLATAAAAAVLQSQAGQRRMSNFSDPEQKRQRVEQLGSLIEQA  180  181  RKASVTSMSHDMSRHVEAAQQNVAIATVLANHLGLTPQSSAASTPSAGTPAPASPTKPYA  2 4 0  241  PSTLAIPQTPTTPRSPLSAAPLTASQISDHQSATQAVAGPSARATNDQAEFSRPKPLVAS  3 0 0  301  PPRTPSITLNYSTDSVEPSSSSKLISEPLRASPPTSQALFNEAKEAAQTYSRFYRFEKEW  3 6 0  361  AQKALELERRRSSIRIDPLFNPNPSLSPLPADPNSNSNLNSSPTHNREGTAPISPQQHSA  4 2 0  421  PESRIMSRRHSPFCDSPVAPSARMSNSGSSGSFSSTHRASSSGLANVLSNFAELIEHRQR  4 8 0  481  SCSGLEALAKQAKELPVKRLSQPNPQFRTTFGDFSWAKAPPSGTDAVNPSAPPLRATPHA  5 4 0  541  SRDANTDADRSSVDKHDTASASTSTVGVRSTHTLPNAQSTRRVSATAPDSDADADANTTA  6 0 0  601  TARS TMS IASML  6 1 2  Figure 6. Amino Acid sequence of Hgllp used in bioinformatics sequence analysis.  2.1.4 Web-based algorithms and web-sites used in the analysis of the hgll predicted amino acid sequence  A l l of the web-based research tools used in the analysis of hglVs predicted amino acid sequence can be found at one of the following URL's: NCBI, http://www.ncbi.nlm.nih.gov; R.E.W. Hancock's links page, http://cmdr.ubc.ca/bobh/links.htm; Pedro's biomolecular research tools,  http://www, public, iastates. edu/~pedro/research  tools, html;  Stanford  Universities  domains search algorithms, http://dna.Stanford, edu/projects. html; TIGR, http://www, tigr. org.  The sequence in figure 6 was used in searches employing the B L A S T , Gapped B L A S T and PSIB L A S T algorithms at NCBI and TIGR against all of the available genomic databases as of 10/03/00 (Altschul et al, 1997). In addition to similarity searches based on the complete amino  24  Materials and Methods  acid sequence, searches for conserved domains were completed with B L O C K S (Pedro's URL), eMOTIF (Stanford U R L ) , P A T - S C A N (Pedro's U R L ) , Pfam (Hancock U R L ) , PRODOM (Pedro's U R L ) , P R O P S E A R C H  (Pedro's U R L ) , PROSITE (Pedro's U R L ) , ProfileScan  (Hancock URL), PRINTS (Hancock URL), SBASE (Pedro's URL), and S M A R T (Hancock URL) (Bucher and Bairoch, 1994; Henikoff and Henikoff, 1994; Hofmann et al, 1999; Schultz et al, 1998; Wright et al, 1999). For algorithms that filter out low complexity sequences, two searches were completed, the primary search employed the supplied filters while the secondary search was completed with all filters removed. Non-filtered searches increase the sensitivity of the comparisons, but have the limitation of retrieving sequences that do not demonstrate significant complexity to be of relevance.  2.2.1 Construction of the Hgllp expression construct pEThgl  A 1.8 kb P C R fragment encoding the hgll open reading frame was amplified with the primers;  HG51,  GCGGATCCACCATGTTCGTCGCTACGCGCA;  and  HGJ31,  G T G C G A A G C T T G C A T G C T A G C . The BamHI restriction endonuclease site of HG51 and the Hindlll site in HGJ31 are underlined. PCR reactions were conducted with pB105 as template (Figure 7) (pB105 contains the coding region of hgll). The 1.8 kb PCR product was digested with BamHI and Hindlll and inserted into BamHI IHindlll digested pET21a (Figure 7) for recombinant expression experiments in E. coli strain BL21(XDE3) and in the Promega (Madison WI)  TNT® wheat germ extract expression system.  25  Successful cloning of the PCR  Materials and Methods  Figure 7. Construction of the plasmid pEThgl. The PCR primers HG51 and HGJ31 were used to amplify the 1.8 kb hgll coding region from pB105. This fragment was subsequently directionally cloned into the HindlHI BamHI sites of pET21a. The 1.2 kb PstI fragment of pB105 was used during southern hybridization analysis of natl disrupted hgll' strains.  fragment was confirmed by HindlHI BamHI and Aatll digestion of the clones. One plasmid that carried the 1.8 kb fragment in the correct orientation was chosen and analyzed by nucleotide sequencing reactions to ensure proper in-frame ligation of the PCR product into the vector (data not shown). This plasmid was named pEThgl (Figure 7).  26  Materials and Methods  2.2.2 In vitro expression of Hgllp from the plasmid pEThgl  Hgllp was produced in vitro from pEThgl with a TNT® coupled wheat germ extract system according to the manufacturer's instructions (Promega Madison, WI). pEThgl was linearized at the unique Apal site in the LacI region of pET21a (Figure 7), precipitated with isopropanol, washed with 70% ethanol in diethylpolycarbonate (DEPC) treated sdH20, and resuspended in DEPC -treated sdH20 to a final concentration of 0.5 p.g/p.1 prior to use in the translation reactions.  For each set of kinase experiments, a single translation reaction was  performed and sub-divided into equal volumes for subsequent treatments. Hgllp protein was labeled with S-methionine during the translation reaction. A negative control TNT® reaction 35  was performed with the pET21a vector; no proteins were produced (data not shown).  2.2.3 Production of recombinant Ukalp and A d r l p  Attempts were made to use the TNT® transcription / translation system to produce the PICA catalytic subunits Ukalp and Adrlp with Hgllp for use in Hgllp phosphorylation experiments. For Promega peptag® kinase activity experiments, the plasmids pET21b::adr#31 and pET21b::uka#20 (F. Durrenberger unpublished) were linearized with Apal and TNT® transcribed/translated as per manufacturers instructions. During dual expression experiments with pEThgl and pET21b::adr#31 or pET21b::uka#20, protocols specific for translating two plasmids in the same reaction mix were attempted. Overnight 9% SDS-polyacrylamide gels of the reaction products were electrophoresed at 100 Volts and were removed from the apparatus before drying under vacuum and examination by autoradiography.  27  Materials and Methods  2.2 A Ni-Agarose Ubc::6xHIS affinity purification of P K A from Ustilago maydis  Ustilago  P K A was prepared as described by Durrenberger et al. (1998) with the  following modification: kinase was eluted from Ni-Agarose columns with 300 m M imidazole. Kinase activity of the elution fraction was determined visually with the Promega peptag® kinase assay for a dilution series of the recovered sample. Equivalent activities of bovine and Ustilago protein kinase A catalytic subunits were added based on visual inspection of a peptag  kinase  assay using 0.04 units of bovine heart protein kinase A . Units of P K A activity describe the number of moles of phosphate transferred per minute to substrate. A 1000 fold excess of PKI (1 ixM) was used during P K A inhibition experiments to eliminate phosphorylation. 100 units of lambda phosphatase were used for phosphate removal (as per manufacturer's instructions N E B , Mississauga, ON). Kinase reactions were performed as per standard reaction protocols outlined in the Promega kinase assay literature.  Control reactions were performed with bovine P K A  supplied by the manufacturer.  2.3.1 Disruption of hgll with the nourseothricin resistance marker natl  Disruption of hgll with a nourseothricin (natl) resistance marker was accomplished with the 2.0 kb Bglll-Xbal phgl::nat\0\  fragment of pUGZ4 inserted into the BgUI site of p H B l O l .  was linearized with Hindlll/BamHI  The  and the 5.0 kb fragment was agarose gel  purified prior to transformation into U. maydis strain 001. The transformation reaction was plated onto D C M S containing 40 (J,g/ml nourseothricin. Colonies demonstrating typical hgll' phenotypes were screened by southern hybridization analysis using BssHII digested genomic D N A preparations.  The southern blot was hybridized with the 1.2 kb PstI fragment of the  plasmid pB105 (Figure 7).  28  Materials and Methods  2.3.2 Suppressor mutagenesis of strains 3020 and 3034 U. maydis strains used for mutagenesis were grown for 24 hours on P D A plates prior to inoculation into 5.0 mis of PDB and incubation overnight at 30°C. The cells in the overnight cultures were recovered by centrifugation and washed once in sdHiO before re-suspension in sdHbO. The cells were diluted to 1x10 cells/ml in a final volume of 1 ml and subject to ultras  violet (U.V.) irradiation in an uncovered small petri dish.  Exposure time was adjusted to  accomplish a 90% reduction in the viable cell count (approximately 2 mins 30 seconds based on plate count assays). Following irradiation and removal from the U.V. light source, the petri-dish lid was replaced and the dish wrapped in aluminum foil for 3 minutes to prevent photoreactivation (Thoma, 1999). Mutagenized cells were diluted to achieve approximately 10,000 colonies per plate and plated onto D C M + charcoal. Inoculated plates were incubated at 30°C for 2-5 days until colonies appeared. Filamentous mutant colonies were plated on C M to check for phenotype stability before characterization on a variety of laboratory media. Fresh overnight 5.0 ml PDB cultures of the mutants were used to inoculate the solid and liquid medium for the characterization experiments. Aliquots of overnight cultures were frozen at -70°C for long term storage of strains.  29  Results  3. Results  3.1 Sequence analysis of Hgllp  The amino acid sequence of Hgllp was subjected to bioinformatics searches to identify similar proteins, or domains of proteins, which may suggest a function for Hgllp. The analysis was conducted with the predicted amino acid sequence of Hgllp from figure 6.  3.1.1 Filtered and non-fdtered B L A S T , PSI B L A S T , and B L A S T 2 with the Hgllp amino acid sequence  B L A S T searches with the complete amino acid sequence of Hgllp resulted in very low probability matches with low complexity regions of proteins such as the proline rich merozite surface protein, or the serine rich protein mucin (completed 10/03/00, data not shown). These short regions of similarity score an E value of approximately 10" , and these results were not 2  sufficiently informative to guide further experiments with Hgllp. E values are an indicator of the likelihood that an alignment has occurred by chance. The lower the E value, the less likely it is that any observed sequence similarity has occurred fortuitously as a result of the database randomly matching the queried sequence.  A n E value of 10" suggests that an identified 2  similarity will occur one in one hundred times simply by chance when using the current database and a random string of amino acids. Sequences reported with low probability scores do not likely represent truly homologous alignments. Non-filtered searches of the NCBI databases using the Hgllp amino acid sequence independently identified the C. albicans contig 4 2801 sequence, and the F l o l l p protein of 5*. cerevisiae (Figure 8 A i , A i i ) . The C. albicans and S. cerevisiae sequences used in the comparison were identified in non-filtered B L A S T and i|/-BLAST sequence identity searches.  30  Results  8 A i demonstrates the non-filtered B L A S T alignment of the C. albicans contig 2801 and the predicted amino acid sequence for the Hgllp protein.  Secondary alignments with the Hgllp  peptide sequence can identify alternative alignments with the same sequence which only have a slightly reduced E value. Similarly, the Flol lp alignment also demonstrates similarity to Hgllp Figure 8. A . Alignment of the Hgllp amino acid sequence to the C. albicans translated con4 2801 (Ai), and the S. cerevisiae flocculin protein F l o l l p (Aii). The alignments score E values of 3x10" , and 2x10" respectively, but are aligned in 2BLAST using the non-filtered alignment mode and hence represent compositionally biased portions of the proteins. B. Alignments with filters reduces the significance of the apparent matches to levels expected for the alignment of random sequences. B i , filtered alignment of Hgllp to translated contig 2801, E = 1.1. Bii, filtered alignment of Hgllp to Flol lp, E = 4.3. 10  14  Ai.  H g l l : 50 2801:  1  Hgll:  110  2801:  52  Hgll:  170  2801:  112  Hgll:  229  2801:  171  Hgll:  275  2801:  231  Hgll:  332  2801:  287  Hgll:  392  2801:  338  Hgll:  441  2801:  397  Hgll:  489  2801:  457  Hgll:  547  2801:  514  Hgll:  599  2801:  574  NTSVGSDSHAGGSHTKLAINQLLTNLSHVSQEQARMSSPRSSHSPALNDSPALKRKFEAD N V S G + T ++S Q S + + + S + N S P NLEVNQPSTNGATSTGHFFGPSIPSTSTHQQTPGETSNNVNTKSSSQNQSP  109  SDSSTAYKVASAHEIHLPTDPVRRASIINLATAAAAAVLQSQAGQRRMSNFSDPEQKRQR S S T+ A+A P R AS QS + + + P STSPTSTVAAAAATSSSPVASTRPASTSEQKQQEETTARQSTSPATTATTSNTPPSPSTS  169  VEQLGSLIEQARKASVTSMSHDMSR-HVEAAQQNVAIATVLANHLGLTPQSSAASTPSAG E S Q A + S++ Q + + V + T P + + S+P+ KETPTSNTAQTSSANNNQQSSNTAAPSTSVIQPSTSEVHVQSQQTSTTP-NTPTSSPNTP  228  51  111  170  T PAPAS PTKPYAPSTL AIPQTPTTPRSP LSAAPLTASQISDHQSAT 274 T + A+PT AP+T +P TPTT +P S AP+T S + TTSEAAPTTSAAPTTSEAPVTPSTSEVVPNTPTTSEAPNTPTTSEAPVTPSTSEVVPNTP 230 QAVAGPSARATNDQAEFSRPKPLVASPPRTPSIT LNYSTDSVEPSSSSKLISEPLRA 331 P+ T++ + P + P TP+ + + +T V P+ + S + + + + TTSKAPNTPTTSE APATPTTSEAPNTPTTSEAPVTPTTSEVVPTTSTQGDAVSTSS 286 SPPTSQALFNEAKEAAQTYSRFYRFEKEWAQKALELERRRSSIRIDPLFNPNPSLSPLPA 391 + TQ + + T + + A + + S S I + PSS TSVTEQTTLTSSTQLPPTTALTTQTSTPEASDS PKPSSTSIE TPSTSTFEQ 337 DPNSNSNLNS S PTHNRE-GTAPIS PQQHSA PESRIMSRRHSPFCDSPVAP 440 DP + S++ + P T E SPQS P + + ++P +P DPTTTSSVGTPSSEQPQPTTTSELAVTSNSPTQESTSLVEPTTSSLESSNTP-TPNPSTS 396 SARMSNSGSSGSFSSTHRASSSGLANVLSNFAELIEH RQRSCSGLEAL 488 A+ S S S +T A + L++ + + F+ L+ H Q S EAQPSTSASQAPPDTTSSAPAPELSSSNADFSNLVLHSSETTSLVNPTDSQIDSSSTTDA 456 AKQAKELPVKRLSQPNPQFRTTFGDFSWAK—APPSGTDAVNPSAPPLRATPHASRDANT 546 QA P + P T D + A+ AP S DA S+P + + + T VSQATTEPTSE-NTPTAASSVTANDINSAQSSAPTSNADAETASSPV—SEQSLATGSQT 513 DADRSSVDKHDTASASTSTVGVRSTHTLPNAQST RRVSATAPDSDADADANT 598 D ++ + A+ + ST NA T S P + A D +T SLDTTAGASSTASKATAENLSTFSTDGSSNASQTIAETTSNSTDQSVVTPSASASTDVST 573 TAT-ARSTMSIAS 610 T + S+ S+ S LPTGSESSTSLVS 586  31  Results  AM.  Hgll  : 5  TRRPSPTPSSPAVSRRSSWDLPVASAPSSRRGSRVDLSIQTSLANTSVGSDSHAGGSHT T P PTPSS S+ V P +S S + ++S A S S S 310 T T T P V P T P S S S T T E S S S A P V P T P S S S T T E S S S A P V T S S T T E S S S A P V P T P S S S T T E S S S A  Floll: Hgll  : 65  KLAINQLLTNLSHVSQEQARMSSPRSSHSPALNDSPALKRKFEADSDSSTAYKVASAHEI  + 370 PVT  Floll:  64 369 124  ++ + S S+ SS +P S + A SST +S+ + S S T T E S S S A P V T S S T T E S S S A P V P T P S S S T T E S S S A P V T S S T T E — S S S A P V 422  Hgll  : 125 HLPTDPVRRASIINLATAAAAAVLQSQAGQRRMSNFSDPEQKRQRVEQLGSLIEQARKAS T A + + T +++A + S S E V S + + A F l o l l : 423 TSSTTESSSAPVTSSTTESSSAPVTS STTESSSAPVPTPSSSTTESSSAP  184  Hgll  244  : 185 VTSMSHDMSRHVEAAQQNVAIATVLANHLGLTPQSSAASTPSAGTPAPASPTKPYAPSTL VTS + + S 473 V T S S T T E S S  Floll: Hgll  A + +T ++ +T SS + SA P P+S T SAPVPTPSSSTTESSSAPVT—SSTTESSSAPVPTPSSST  S+ TESSS  : 245 AIPQTPTTPRSPLSAAPLTASQISDHQSATQAVAGPSARATNDQAE—FSRPKPLVASPP A TP++ + S+AP+T+S 525 APAPTPSSSTTESSSAPVTSSTT  Floll: Hgll  Hgll  Floll:  630  581 362  TPS + S+ + P+ SS +E A PT + E+ A T S 582 PTPSSSTTESSSAPVPTPSSS-TTESSSAPAPTPSSSTTESSSAPVTSS : 363 KALELERRRSSIRIDPLFNPNPSLSPLPADPNSNSNLNSSP E + + S +P+P +S + +S+P  524 302  +S++ V PS+ T + S ++P ESSSAPVPTPSSSTTESSSTPVTSSTTESSSAPV  : 303 RTPSITLNYSTDSVEPSSSSKLISEPLRASPPTSQALFNEAKEAAQTYSRFYRFEKEWAQ  Floll:  472  629  THNREGTAPI + +AP+  413  TTESSSAPVPTPSSSTTESSSAPVPTPSSSTTESSSAPVPTPSSSTTESSSAPVTSS  686  Hgll  : 414 SPQQHSAPESRIMSRRHSPFCDSPVAPSARMSNSGSSGSFSSTHRASSSGLANVLSNFAE 473 + + SAP + + S +P + + S++ SST +SS+ + S+ E F l o l l : 687 T T E S S S A P V T S S T T E S S S A P V P T P S S S T T E S S S A P V P T P S S S T T E S S S A P V P T P S S S T T E 746  Hgll  : 474 LIEHRQRSCSGLEALAKQAKELPVKRLSQ PNPQFRTTFGDFSWAKAPPSG S + + + ++ PV S PP T T + S A P F l o l l : 747 SSSAPVTSSTTESSSAPVPTPSSSTTESSSAPVPTPSSSTT—ESSSAPVPTPS  523 798  Hgll  : 524 TDAVNPSAPPLRATPHASRDANTDADRSSVDKHDTASAS TSTVGVRSTHTLPNAQS 579 + S P+ TP +S + + A S+ T S+S T + + + P + S F l o l l : 799 SSTTESSVAPV-PTPSSSSNITSSAPSSTPFSSSTESSSVPVPTPSSSTTESSSAPVSSS 857  Hgll  : 580 TRRVSATAPDSDADADANTTATARSTMSIAS 610 T S+ AP + +N T++A S++ +S F l o l l : 858 T T E - S S V A P V P T P S S S S N I T S S A P S S I P F S S 887  Bj.  Hgll:  238 PYAPSTLAIPQTPTTPRSPLSAAPLTASQISDHQSATQAVAGPSARATNDQAEFSRPKPL P P+T P TPTT +P++ T S++ S S+ + +Q +  2801:  245 PATPTTSEAPNTPTTSEAPVTP  Hgll:  298 VASPPRTPSITLNYSTDSVEPSSSS +  2801:  BH.  Hgll  302  +  TTSEVVPTTSTQGDAVSTSSTSVTEQTTLTSSTQL 322 326  : 242 STLAIPQTPTTPRSPLSAAPLTASQISDHQSATQAVAGPSARATNDQAE  Floll:  TP++  +  S+AP+T+S  +S++  522 S S S A P A P T P S S S T T E S S S A P V T S S T T  V  PS+  T  — FSRPKPLVA +  S  Floll:  TPS +  S+ +  P+  SS  +E  A  PT  +  E+  A  32  351  T S  579 A P V P T P S S S T T E S S S A P V P T P S S S - T T E S S S A P A P T P S S S T T E S S S A P V T S S  299  +  ESSSAPVPTPSSSTTESSSTPVTSSTTESSS  : 300 SPPRTPSITLNYSTDSVEPSSSSKLISEPLRASPPTSQALFNEAKEAAQTYS +P  301  ++DS +PSS+S  PPTTALTTQTSTPEASDSPKPSSTS  S+ A  Hgll  T  297 L  629  578  Results  (E = 2xl0" ) due mainly to the compositional bias of the Flol lp protein. Flol lp contained a 14  600 amino acid, low complexity region identified by the SEG program which consisted mainly of the repeated sequence "PVPTPSSSTTSSES" (Schultz et al, 1998). This region of Flol lp demonstrated the highest similarity to Hgllp, a result similar to the low overall complexity of the con4 2801 sequence.  The application of filters to BLAST2 alignments between the Hgllp  and the Candida or S. cerevisiae sequences reduced the statistical significance of the alignments to E values of 1.1 and 4.3 respectively. Such comparisons did not fall below the probability threshold that the alignment occurred purely by chance, indicating that the non-filtered alignment with Hgllp should be treated with caution. It was unlikely that Hgllp is functionally related to the two sequences based on the nature of the similarities which have been identified, however these alignments were the most significant that could be found. A B L A S T alignment between the Candida and Saccharomyces sequences demonstrates significant similarity between the two sequences. Interestingly, the similarity between contig 2801 and F l o l l p was high for both filtered and non-filtered comparisons, perhaps suggesting the comparison data between Hgllp and each of the others independently may have had significance beyond the E value.  In the absence of filters, the E value of the comparison  between C. albicans con4 2801 and S. cerevisiae F l o l l p was in the neighborhood of 4xl0" , 66  and even with the filters in place the E value remained at 5xl0" (Figure 9). Comparison of 8  con4 2801 to a second flocculin protein, Flo5p, again identified a high similarity when filters were removed from the program (E = 9xl0" ), but did not demonstrate the high similarity seen 29  with the Flol lp sequence when filters were re-implemented into the comparison (E = 2.3xl0" , 2  data not shown). Hgllp did not share significant similarity to Flo5p in filtered or unfiltered searches. Overall, these results did not provide a reliable clue to the identity of Hgllp or its  33  Results  function in U. maydis. However, these searches did suggest that the dimorphic pathogen C. albicans may have a homologue of the S. cerevisiae FLO  11 gene.  2801  : 420 LSSSNADFSNLVLHSS--ETTSLVNPXXXXXXXXXXXXAVSQATTEPTSENTPTAASSVT 477 ++SS + S+ + SS E++S P V+ +TTE +S PT +SS T F l o l l : 434 V T S S T T E S S S A P V T S S T T E S S S A P V P T P S S S T T E S S S A P V T S S T T E S S S A P V P T P S S S T T 493  2801  : 478 ANDINSAQSSAPTSNADAETASSPVSEQSLATGSQXXXXXXXXXXXXXXXXXXENLSTFS 537 + SSAP +++ E++S+PV T S E+ S F l o l l : 494 ES SSAPVTSSTTESSSAPV P T P S S S T T E S S S A P A P T P S S S T T E S S S A P V 542 2801  : 538 TDGSSNASQTIAETTSNSTDQSVVTPSASASTDVSTLPTGSESSTSLVS 586 T ++ +S T S+ST +S TP S++T+ S+ P + SS++ S F l o l l : 543 T S S T T E S S S A P V P T P S S S T T E S S S T P V T S S T T E S S S A P V P T P S S S T T E S ' 5 9 1  Figure 9. Filtered alignment of the translated C. albicans contig 2801 with the 5". cerevisiae protein Flol lp. The resulting E value is 5xlO" . 8  P R O P S E A R C H with Hgllp as a query sequence found similarity between Hgllp and transcriptional regulators of the family involved in nitrogen regulation in filamentous fungi, or to the transcriptional regulators of flocculation (SFL1 and FL08) in the budding yeast Saccharomyces cerevisiae. The P R O P S E A R C H algorithm looks for compositional similarities between related proteins, and groups them into protein families based on the analysis (Hobohm and Sander, 1995).  Examples of the criteria used in the algorithm are: shared amino acid  composition, amino acid bulkiness, overall pH, isoelectric point (pi), and hydrophobicity.  average  P R O P S E A R C H can identify proteins that share similar amino acid  compositions, and may identify a relatedness between two sequences that cannot be identified by traditional sequence alignment algorithms (Hobohm and Sander, 1995). Protein groupings are assessed on a scale that ranks their similarity with each other compared to the identified sequence and an unrelated protein. The comparison is scored as the 'distance' (DIST) between the two sequences; a larger predicted distance between two sequences results in a lower chance 34  Results  they belong to the same family (reported as a percentage). Results of a P R O P S E A R C H with the Hgllp amino acid sequence identified its compositional similarity to a nitrogen regulatory protein family (members identified in bold) (Table 1). Hgllp scored a distance (DIST) from the nitrogen regulatory proteins of 10.26, or an approximately 80% chance that Hgllp was from that protein family. The 11.51 and 12.03 DIST's reported between Hgllp, Sfllp, and Flo8p suggested there was a 68% chance that Hgllp was a family member of S. cerevisiae  Table 1. Proteins identified by P R O P S E A R C H using Hgllp as the amino acid query sequence. Sequences 5,12, and 15 included unidentified orf s and are not presented in the table. ID = identified family member. DIST = relatedness distance from the recovered protein. The length of the reported matching family member is provided in the L E N column. DEFINITION provides a brief description of the protein. Rank ID  DIST  LEN  1 area aspng 3 area emeni 4 s f 11_ y e a s t 6 area pench 7 area aspor 8 flo8~ yeast 9 wet a emeni 10 h t f 4 human 11 s t b l " y e a s t 13 h t f 4" r a t 14 area penro 16 crep human 17 area f usmo 18 nrf a penur  10 10 11 11 11 12 12 12 12 12 12 12 12 13  882 9.78 876 8.15 766 9. 40 725 7 .40 866 7 .88 799 10 .12 555 7 .91 682 7 .02 459 10 .40 707 6. 90 860 7 .56 505 7 .90 970 9.76 865 7 .34  _  transcription  26 90 51 58 59 03 20 43 61 75 76 88 94 01  factors.  Pi  DEFINITION NITROGEN REGULATORY PROTEIN AREA. NITROGEN REGULATORY PROTEIN AREA. FLOCCULATION SUPPRESSION PROTEIN(SFL1) NITROGEN REGULATORY PROTEIN AREA (NRE) . NITROGEN REGULATORY PROTEIN AREA. TRANSCRIPTIONAL ACTIVATOR FL08(PDH5). REGULATORY PROTEIN WETA. TRANSCRIPTION FACTOR HTF4. STB1 PROTEIN. TRANSCRIPTION FACTOR HTF4. NITROGEN REGULATORY PROTEIN AREA. CAMP RESPONSE ELEMENT BINDING PROTEIN. NITROGEN REGULATION PROTEIN AREA. NITROGEN REGULATORY PROTEIN NRFA.  A BLAST2 comparison between Hgllp and AreA of Aspergillus  nidulans, or the yeast Sfllp and Flo8p proteins identified by P R O P S E A R C H did not identify additional similarity. Removal of the filters from the B L A S T searches did not increase the sequence similarity.  35  Results  3.1.2 Conserved domains in the Hgllp amino acid sequence  Given the nature of the similarity between the Con4 2801, Flol lp, and Hgllp sequences, it was necessary to look for common domains within each protein that might suggest a shared function.  S M A R T analysis of the Candida con4 2801, Saccharomyces F l o l l p , and Hgllp  sequences demonstrated that they did not share an overall domain conservation (data not shown).  Flocculin proteins share a similar organization of a signal peptide at the amino  terminal end, and seven transmembrane motifs in the carboxyl end of the protein. The Candida and Ustilago sequences did not share this same domain conservation, suggesting that they did not share functions with the flocculin proteins. The three proteins share portions of the atrophin PRINTS fingerprint, though the E values for the similarities were insignificant for each individual protein (between .05 and 66, data not shown). Additionally, eMOTIF searches did not identify broad motif similarities between the three proteins, so the Hgllp sequence was examined individually.  The Hgllp protein had  several interesting motifs which might suggest a signaling-molecule function. The eMOTIF database identified both a cdc24 guanine-nucleotide dissociation factor family sign, and a Gprotein coupled receptor (GPCR) transmembrane motif (Figure 10). Each of the identified motifs was outside regions proposed to have low complexity in Hgllp and both occured with a stringency of approximately 10" , suggesting they were not false positives . However, the cdc24 5  GDS family members typically share a second family signature that is absent in the Hgllp sequence. factor.  There were no other indications that Hgllp was a guanine nucleotide dissociation  Similarly, the transmembrane motif of the GPCR was in amino acids 23-32 in the  sequence but was the only one identified by eMOTIF, the GPCR family typically share from 2 to 7 additional characteristic motifs which were absent in the Hgllp sequence (Attwood and Findlay, 1994). Examination of a hydrophobicity plot from the Tmpred search program reveals  36  Results  two potential transmembrane domains. O f the suggested domains spanning amino acids 134154 (a length of 21 residues) was the strongly favored transmembrane domain predicted to run from the inside to the outside of the cell (Figure 10).  The significance score for such an  orientation of this transmembrane domain was 942, and scores of 500 or higher are to be  0 MFVRTIUtPSPTPSSPAVSRRSSWVT^PVASAPSSRRGSRVDLSIQTSLAN 50 51 TSVGSDSHAGGSHTKLAINQLLTNLSHVSQEQARMSSPRSSHSPALNDSP 100 101 ALKRKFEADSDSSTAYKVASAHEIHLPTDPVRRASIINLATAAAAAVLQS 150 151 QAGQRRMSN FS D PEQKRQRVEQLGS LIEQARKAS VT SMS H DMS RH VEAAQ 200 2 01 Q|I?»IATVIJ^lX3LTPQSSARSTPSA<3'rPM>ASPTBCPYM>STLftIPQTP 250 251 TTPRSPLSAAPLTASQISDHQSATQAVAGPSARATNDQAEFSRPKPLVAS 300 301 PPRTPSITUrgSTDSVEPSSSSKLISEPLRASPPTSQALFNEAKEAAQTY  350  351 SRFYRFEKEWAQKALELERRRSSIRIDPLFNPNPSLS^  4 00  401 SSPTHNREGTAPISPQQHSAPESRIMSRRHSPFCDSPVAPSARMSNSGSS 450 4 51 GSFSSTHRASSSGLANVLSNFAELIEHRQRSCSGLEALAKQAKELPVKRL 500 501 SQPNPQFRTTFGDFSWAKAPPSGTDAVNPSAPPLRATPHASRDANTDADR 550 551 SSVDKHDTASASTSTVGVRSTHTLPNAQSTRRVSATAro  600  601 TARSTMSIASML 612  Figure 10. Distribution of domains in the predicted amino acid sequence of Hgllp. Sequences in blue identify P K A target sites. High scoring PEST sequences are indicated in bold. Sequences with a single black line below are additional PEST sequences. Regions under dashed lines represent the low complexity regions as determined by the SEG program. Residues 4-11 (underlined in pink) share 100% similarity to a protein tyrosine phosphatase gamma precursor identified by PSI-BLAST. The region from 23-32 under the crimson line identifies the GPCR transmembrane motif, and the region from 164-173 under the blue line represents the GDS cdc24 motif. The sequence from amino acids 134-154 underlined with the yellow line represents the Tmpred identified transmembrane motif.  considered significant (Hofmann and Stoffel,  1993).  For example, when a seven  transmembrane protein such as Flol lp was scanned with the hydrophobicity program, scores of approximately 1500 were recovered for four putative membrane spanning regions. 37  The  Results  program did not however identify all seven regions, and was incorrect for two of the four predicted domains.  The results of such domain-identifying programs should be cautiously  applied and used in parallel with experimental evidence.  3.2 P K A phosphorylation of Hgllp  Analysis of the Hgllp amino acid sequence identified 8 putative P K A phosphorylation sites. To investigate possible interactions between Hgllp and P K A , Ustilago maydis P K A was used to phosphorylate Hgllp produced in vitro.  3.2.1 Purification of Ustilago maydis P K A from Ubc::6xHIS chromatographic columns  Attempts were made to produce catalytic subunits of U. maydis  P K A (Ukalp and  Adrlp) in an in vitro transcription/translation reaction for use in phosphorylation experiments. Proteins of the expected size for ukal and adrl gene products were recovered in the TNT® system, but these proteins did not demonstrate kinase activity in the peptag® kinase assay (Figure 11).  The peptag® assay quantitates P K A activity by the amount of flourescently  labelled peptag® that migrates towards the anode following phosphorylation by kinase. Subsequently,  Ustilago  P K A was  recovered  38  from  Ubc::6xHIS  affinity  columns  Results  Figure 11. Production of the P K A molecules Ukalp and Adrlp using the TNT wheat germ extract system. Panel A demonstrates the methionine labeled bands produced when the pET21b::adr#31, and pET21b::uka#20 (Franz Durrenberger, unpublished) plasmids were utilized in TNT® transcription/translation reactions and purified over Ni-agarose affinity columns. The autoradiogram is from a single gel that has been cropped to allow direct comparison. Lanes a and d contain Ukalp and Adrlp (respectively) un-separated from the reaction components. Lanes b,c, and e,f demonstrate the recovery o f Ukalp and A d r l p from Ni-agarose affinity columns in 300 m M immidazole elution fractions. Panel B represents a Promega peptag® kinase assay experiment. Lane a contains control bovine P K A (sample contained low kinase activity); lane b, expressed control luciferase protein (no kinase activity): lane c, TNT® produced Ukalp from the elution fraction of lane c in Panel A; lane d, Adrlp from the elution fraction of lane f, panel A recovered from the TNT transcription/translation reaction.  and diluted to achieve an activity equivalent to 0.04 units of bovine heart P K A (figure 12A). The addition of PKI had an inhibitory effect on the apparent molecular weight increase caused by the Ustilago and bovine P K A catalytic subunits, but did not completely eliminate the shift  39  Results  (Figure 12B). A loss of the apparent molecular weight shift was achieved with the addition of 100 units of lambda phosphatase (Figure 12B).  a b c d e f g h i  B  a  b  c  d  jgte  i  e  f  —-  Figure 12. Evaluation of P K A activity and the effects of PKI and phosphatase treatment on the peptag® kinase assay. Panel A represents peptag assays with bovine and Ustilago P K A . Lane a contains no kinase added. Lanes b-e contain dilutions of bovine P K A as follows: lane b contains 3.92xl0" units; lane c, 3.92xl0" units; lane d, 3.92X10" units; and lane e 3.92xl0" units. Lanes f-i represent Ustilago P K A reactions with: f, undiluted P K A elution fraction; g, 1/10 P K A dilution; h, 1/100 P K A dilution fraction; i , 1/1000 P K A dilution fraction. Panel B represents bovine and Ustilago P K A reactions treated with a 1000 fold excess of PKI (1 \xM) and 100 units lambda phosphatase. Lanes a-c contain: a, bovine P K A ; b, bovine P K A + 1 u M PKI; c, bovine P K A + 100 units lambda phosphatase. Lanes d-f contain: d, Ustilago P K A ; e, Ustilago P K A + 1 u.M PKI; f, Ustilago P K A + 100 units lambda phosphatase. 1 unit of P K A activity transfers 3.92 pmoles of phosphate to substrate per minute at 37 °C. 2  3  4  5  3.2.2 Expression of Hgllp from pEThgl  Attempts were made to express Hgllp in E.coli strain BL21(?JDE3) for use in P K A phosphorylation assays, several expression conditions were examined but the production of Hgllp was not successful.  For example, protein production was attempted in separate  expression cultures incubated at 37°C, 30°C, and 22°C. Each temperature-condition experiment was repeated with varying degrees of aeration, and induction with either 0.5mM or I m M IPTG. Induced cultures were incubated for 2hrs, 6hrs, or 12hrs for each of the experimental conditions. In all experiments, there was no specific protein of the Hgllp size observable by either  40  Results  coomassie brilliant blue staining, or by western blot analysis of both the soluble and insoluble fractions for all of the protein preparations. The controls for protein expression and expression of Hgllp fusion proteins included use of the plasmids pHhmycl2 (not shown), pET21b::ubcl#3, pET21b::adr#31 (constructed by F. Durrenberger), as well as a pGEX plasmid expressing the janus kinase JNK (gift from Rob Gerl), to ensure that induced cultures were producing protein.  Control cultures (excluding  cultures for Hgllp) demonstrated high levels of expressed protein.  Additionally, the HIS-  tagged version of the Ustilago P K A regulatory subunit used for isolation of Ustilago P K A (pET21b::ubcl#3) was purified over a Ni-agarose column and subjected to electrophoresis (SDS-PAGE) to demonstrate recovery of soluble protein from the chromatography column (Figure 13).  Figure 13. Purification of Ubcl::HIS and P K A from a Ni-Agarose chromatography column. Panel A is a photo of a Coomassie brilliant blue stained SDS-PAGE. Panel A lanes a and b represent flow-through eluate from 20 m M immidazole washes prior to protein elution with 300 m M immidazole. Lanes c, d, and e represent elution fractions from successive washes of the same column with 300 m M immidazole. Panel B, the protein elution fractions from lanes c, d, and e of panel A were tested for kinase activity with the promega peptag assay. Panel B lane a contains a sample from the elution fraction of Panel A lane a. Panel B lanes b, c, and d contain samples from the elution fractions from panel A lanes c, d, and e respectively. The elution fraction from Panel A lane a contained the highest amount of active P K A .  41  Results  Attempts were made to fuse the Hgllp protein to the myc and hemagglutinin (HA) epitope tags in an effort to recover purified Hgllp protein. Experiments with the myc-epitope tag involved the control protein mycMcmlp (gift from Chris Nelson), which was detectable following SDS-PAGE and western blot analysis with Ab-conjugated chemiluminesence detection procedures (Figure 14).  Immunoprecipitations with mAb 9E10 against c-myc  Anti-myc  -25 kDa >  Figure 14. Western hybidization analysis of the myc-tagged protein Mcmlp. Anti-myc mAb 9E10 was used to probe a western blot for the presence of the myc epitope. A band of consistent size with the M c m l p transcription factor from S. cerevisiae (courtesy of Chris Nelson, Ivan Sadowski's lab) was detected as the only myc-tagged protein present on the membrane.  from large (500ml) induced cultures did not detect any Hgllp protein. Hgllp may have been expressed at levels too low for detection. Immunoprecipitation was attempted from U. maydis cultures carrying the myc and H A tagged versions of Hgllp under control of the native hgll promoter in hopes of recovering protein; these attempts also were unsuccessful. The hglv.myc  42  Results  fusion complemented the original hgll mutant, but the protein was not detectable by the antimvcmAB 9E10.  3.2.3 Phosphorylation of Hgllp with bovine heart and U. maydis P K A  Successful phosphorylation of Hgllp by either Ustilago protein kinase A (UPKA) or bovine heart protein kinase A (bPKA) was expected to result in an increased apparent molecular weight during SDS-PAGE (Sadowski et al., 1991). Protein kinase reactions were performed with  S-Methionine labeled in vitro produced Hgllp and either U P K A or bPKA. Figure 15  demonstrates that a mobility shift was produced when bPKA (panel A , lane b) or U P K A (panel B, lane b) were used to phosphorylate Hgllp.  A  a  b  c  d  B  a  b  c  d  Figure 15. Phosphorylation of Hgllp by bovine and U. maydis protein kinase A. Panel A shows reactions with bovine heart P K A and Panel B shows reactions with U. maydis P K A . Lane a of each panel shows the position of S-methionine labeled Hgllp without the addition of PKA. Lane b of each panel demonstrates the shift produced upon reaction with the catalytic subunits of P K A . Lane c is the reduced shift observed in the presence of the kinase inhibitor PKI, and lane d demonstrates elimination of the mobility shift upon treatment of the kinase reactions with lambda phosphatase (BRL). 35  Ustilago and bovine heart P K A had similar effects on the translated Hgllp protein product, producing an increased apparent molecular weight shift of the S-methionine labeled 35  43  Results  band. The TNT® wheat germ extract used to produce Hgllp did not have protein kinase A activity as determined in the fluorescent peptag® kinase assay (Figure 16). Introduction of a  Figure 16. Wheat Germ Extract does not contain intrinsic kinase A activity. Lane a demonstrates migration of the peptag® in response to phosphorylation by bovine heart PKA. Lane b contains an amount of wheat germ extract equal to that present in the Hgllp phosphorylation reactions from figure 15, the charge state of the peptag was not altered, and it migrated towards the cathode.  specific protein kinase A inhibitor (PKI) decreased the shift due to phosphorylation. PKI addition (to 1000 fold excess) did not completely inhibit kinase activity on Hgllp, a result that was also be observed in the peptag® assay under the conditions tested (Figure 12). Finally, the addition of lambda phosphatase reduced the shift to the migration level of the control protein. The results of this experiment demonstrated  that in vitro produced Hgllp could be  phosphorylated by the P K A catalytic subunits from U. maydis and bovine heart tissue.  This  experiment was repeated 8 times and a similar reduction in mobility upon phosphorylation was observed in each experiment.  3.3 Suppressor mutagenesis of hgll' strains  A suppressor analysis in the strains 3020 (hgllr.natl hgll::natl  in 001) and 3034  (adrl::phleo,  in 001) was undertaken to determine whether it was possible to isolate mutants  44  Results  which could suppress the hgll' colony phenotypes.  Mutants were identified (in each of the  backgrounds) that recovered a filamentous colony morphology similar to that of adrl' strains. This indicated that Hgllp was not essential for filamentous growth, and would support a model that Hgllp acts as a repressor of budding growth.  3.3.1 Disruption of hgll with the natl cassette  The hgll disruption strain 3034 was used for the mutagenesis experiments to identify suppressor mutants. This strain contained the hygromycin resistance marker that was used for disruption of the hgll gene and selection for successful integration of the disruption construct. The most reliable cosmid library in our laboratory, for the purpose of complementation, also relied on hygromycin as the basis of selection for successful transformation. Therefore, it was necessary to create an hgll' strain which did not carry a hygromycin resistance marker. A n hgll' strain which carried a nourseothricin cassette in place of the hygromycin cassette was created. The HindlH/BamHI  cassette of phgl::nat (Figure 17) was used to transform strain 001.  45  Results  Figure 17. Construction of an allele of hgll containing the nourseothricin disruption cassette. The Bglll fragment of pUGZ4 was gel purified and blunt ended with T4 D N A klenow fragment before ligation into Bglll digested p H B l O l . Loss of the 0.9 kb Bglll fragment from pHBlOl eliminates a diagnostic EcoRI fragment that was used for identification of successful disruption.  Transformants were chosen for genomic D N A preparation based on their similarity to the original hgll' phenotypes, e.g., yeast-like colony growth and yellow pigmentation of the media. Successful integration of the hgllr.natl  construct resulted in a 0.5 kb shift of the 7.0 kb  genomic BssHII fragment seen in the wild type strain (Figure 18).  Strains used for the  preparation of genomic D N A in lanes c-g were named "3020 - 3025" and frozen at -70°C for long term storage.  46  Results  a  8kb 7kb  b  c  d:  e  f  g  h i  j  * k  ^^^^^^  •  I  jn  ^ 7.5 kb  ^  Figure 18. Disruption of hgll with the natl cassette in the wild type 001 background. Genomic D N A from each strain was digested with BssHII, probed with the 1.2 kb PstI fragment of pB105 (Figure 17). Successful disruption results in a shift of the wild type 7.0 kb hybridization band to a 7.5 kb band. Lane a contains wild type genomic D N A , b contains genomic D N A of the hygromycin disruption strain 3001 (8.0 kb), while lanes c-m contain genomic D N A from successful integrants of the nourseothricin disruption construct (7.5 kb).  3.3.2 Isolation of filamentous suppressor mutants of the budding hgll' phenotype  A variety of suppressor phenotypes were obtained following mutagenesis of hgll' strains.  Each of the strains (3020 and 3034) was mutagenized in six different experiments.  Each experiment involved mutagenesis and plating of approximately 1.2 x 10 cells in total. 6  Therefore between the two strain backgrounds, approximately 1.44 x 10 mutagenized cells 7  were screened. For the 3020 background, 17 hgll' suppressor mutants were recovered, and for the 3034 background, 25 hgll' suppressor mutants were recovered. Phenotypes in each strain backgournd, ranged from a morphology similar to that of the highly filamentous adrl' strain, to 47  Results  a weak filamentous phenotype. Similar mutant phenotypes were observed between different isolates from each strain background (3020 and 3034). Mutant strains were examined for phenotype variability on a variety of common media used for the growth of U. maydis. A n overview of mutant phenotypes is provided in Table 2, which lists the mutants recovered and their phenotypes on solid media. For simplicity, the filamentous and pigmentation phenotypes were recorded as +, ++, or +++ in order of increasing magnitude. A score of '+' (Figure 19 panel c) indicated that the mutant weakly demonstrates the phenotype of that category, '++' (Figure 19 panel b) indicated a moderate level, and '+++' (Figure 19 panel a) indicated the mutant strongly demonstrates the described phenotype. The degree of colony filamentation was categorized into one of three classes (Figure 19). A n example of the pigment produced by the hgll' strains and the collected suppressor mutants is  Figure 19. Categories of different hgll' suppressor mutants. Each of the suppressor mutants described in Table 2 have been assigned as sharing similar phenotypes to the colonies in either panels a, b, or c. Panel a categorizes colonies which have a consistent, thick, covering of short, velvet-like filaments. The body of the colony is difficult to view due to the high degree of filamentation. Panel b categorizes colonies which share longer filaments than those of group a, that are evenly dispersed about the colony surface. The colonies may grow as 'doughnut' shape, a distinct colony body is identifiable. Panel c represents colonies that contain shorter filaments evenly dispersed about the colony surface. The filaments are short, and may appear granular. A distinct colony body is apparent, and the colonies are typically small and 'dome' shaped.  48  Results  provided in figure 20. When selecting mutants for complementation experiments, the desired phenotypes were maintenance of highly filamentous colony growth and pigmentation on a variety of different media (such as mutant 7.1).  This was done in order to minimize the  phenotypic variability involved in the complementation experiments.  Figure 20. The pigmentation phenotype of an hgll' strain. Panel A represents wild type 001 U. maydis colonies grown for 48 hours on PDA. Panel B represents strain 3001 (hgll~) grown for 48 hours on PDA. The pigmentation phenotype of hgll' strains can be enhanced (more pigment production) or reduced (less pigment production) by U . V . suppressor mutagenesis. Nutrient abundance played some role in filament growth or colony morphology for several of the mutants. During characterization of the mutants, colony growth was observed on media ranging from the low nutrient M M to the nutrient rich D C M . Generally, as the media contained less nutrients (low glucose, yeast extract, casamino acids), the filamentation phenotype was enhanced. Mutants were recovered which did not respond to changing media composition, indicating that there was more than one additional pathway involved in establishing or regulating filamentous growth. A filamentous mutant which still responded to nutrient levels with an altered colony morphology was likely affected in a component that did  49  Results  not respond to nutrient availability. Mutants which maintained the same degree of filamentation despite the composition of the growth medium may be mutated for a nutrient sensory-response pathway that participated in filamentous growth. This situation was likely to be mediated by the c A M P / P K A pathway due to the constitutive phenotypes components of the P K A pathway in U. maydis.  50  of strains affected for other  Results  Table 2. Growth of suppressor mutants was characterized on a variety of common laboratory media and summarized in table 2. +, ++, and +++ indicate the strength of the phenotype (+ = low, +++ = high). General colony morphology of the mutants is classified as a, b, or c, as shown by the examples in Figue 19. * colonies display a slow growth phenotype resulting in small colony size.  Mutant Growth on Solid Media Mutant hgH  5.1 5.2 5.3 6.2 7.1 8.1 8.4 8.7 9.1 9.2a 9.2b 14.6 15.1a 8.8 26.4.2 26.23.4* 26.14.1* 26.11.1 26.19.1 26.22.3 26.35.3 26.24.1 26.24.2 26.24.3  Degree of  Pigment  Nutrient  fi lamentation -  production  sensitivity  ++ +++ +++ ++ ++ +  ++ + ++ + ++ +++ + + ++ +++ +++ +++ ++ ++ +  ++  ++  ++  ++ ++ ++ ++  + +  ++ ++  ++ +  ++ ++  ++ ++ ++  + ++ +  ++ ++ ++  ++  +  ++  +  +  ++  + + ++ ++ +++ +++ + +  51  n/a  +++ ++ +++ ++ + +  ++ +++ +++ +++ +++ +++ ++ ++  Example colony 1 n/a c c b b a a c c b a a b b c b b b b b b b b b c  Results  3.3.3 Cosmid complementation of suppressor mutations in the 3020 and 3034 strains Cosmid complementation of one selected mutant for each strain (3020 and 3034) was attempted with a cosmid library in order to recover a gene which acts epistatically to hgll. The complementation experiments were unsuccessful because of unforeseen problems with the selected mutants.  Specifically, the original mutant collection in the adrl' hgll' background  appeared to be intrinsically resistant to the fungicide carboxin, the basis of selection for the vector of one of the available genomic libraries. Unsuccessful attempts were made to transform the original filamentous adrl'hgU'  mutants with the hgllr.natl  disruption cassette to replace the  hygromycin gene (5.0 kb HindlHI BamHI fragment of phgl::nat from figure 17). As a solution to this problem, hgll' suppressor mutants were recovered in a strain carrying the nourseothricin resistance marker (strain 3020).  Mutagenesis was repeated with strain 3020 and the  hygromycin library was used for complementation experiments. A complementing gene was not identified because of difficulties in obtaining a sufficient number of transformants to screen the library.  52  Discussion  4. Discussion  4.1 Summary of Results  The objectives of this study were to: 1. Conduct a web-based bioinformatics analysis of the Hgllp amino acid sequence. 2. Conduct P K A phosphorylation experiments involving the hgll gene product.  3. Complete suppressor mutagenesis of hgll' strains to identify genes  located downstream of hgll in the dimorphic switch. A bioinformatics analysis revealed several characteristics that pointed towards a signaling function for Hgllp, but sequence comparisons to the current databases did not identify any significant similarity to the documented sequences.  P R O P S E A R C H suggested that Hgllp  was a regulator of transcription, and was related to a family of proteins that respond to nitrogen availability in other fungi. The Hgllp predicted polypeptide sequence contained separate motifs common to a tyrosine phosphatase, a guanine nucleotide dissociation factor, and a G-protein coupled receptor molecule.  There were 8 putative P K A phosphorylation sites within the  sequence, as well as several PEST regions that could be involved in regulating a ubiquitinmediated turnover of the Hgllp protein. Phosphorylation of Hgllp by P K A was accomplished in vitro, and this result suggested that Hgllp serves as a target for P K A phosphorylation in vivo. demonstrated as an increase in apparent molecular weight of an  Phosphorylation was  S-methionine labeled Hgllp  band during SDS-PAGE. The mobility shift was altered by the addition of the P K A inhibitor PKI, and was eliminated by treatment of the kinase reaction samples with lambda phosphatase. In vivo phosphorylation experiments were planned but not completed because of problems with the epitope tagging of Hgllp.  53  Discussion  Suppressor analysis of hgll' strains revealed that a variety of phenotypes were possible in the hgll' background, and that U. maydis pigment production and filamentation may not be part of the same pathway. Mutants were identified that respond to nutrient availability with an altered phenotype and this result may implicate other signaling pathways in the control of morphogenesis in culture. production.  It was also possible to collect mutants that varied for pigment  Variability for pigmentation appeared to be independent from filamentation,  suggesting that the two processes may not be results of the same pathways. The ability to isolate strains that suppress the hgll' phenotypes, and mutant identification indicated that Hgllp was not required for filamentation in culture, and supported the proposed model of Hgllp as a negative regulator of budding growth.  54  Discussion  4.2 Discussion of results and future experiments 4.2.1 H g l l p sequence analysis Primary B L A S T comparisons with full length Hgllp as a query sequence did not reveal significant similarity to sequences presently in any of the databases. Much of the weak overall similarity that was observed was due to the presence of proline, serine or threonine residues and was the result of amino acid compositional bias in the polypeptides. When the filters of low complexity regions are removed from search algorithms, sequences are identified with much lower E values (lower probability of a random alignment). However the apparent similarity between the sequences is due to compositional bias in this situation. The calculation of the true similarity is grossly affected by the compositional bias of the sequence, and may not represent a true alignment of specific amino acids (Boguski et al., 1998). Data from non-filtered searches must be scrutinized prior to predictions about the function of a protein, and can not be relied upon without supporting experimentation.  P R O P S E A R C H with Hgllp suggested it was related to a family of transcription factors that regulate nitrogen metabolism in filamentous fungi, and may also be related to other regulators of transcription from Saccharomyces  cerevisiae.  The AreA gene of Aspergillus  nidulans was the most compositionally similar sequence to Hgllp in the database, with an approximately 80% chance that Hgllp was from the same family. AreA, a G A T A binding factor is responsible for upregulating genes required for the utilization of alternative nitrogen sources versus the preferred sources ammonium and glutamine; these compounds repress AreAinduced gene expression.  Mutations in AreA affect the ability of Aspergillus  to express  structural genes necessary for growth on N sources such as nitrate, nitrite, amino alcohols, and  55  Discussion  various purines (Wilson and Arst, 1998). Results from P R O P S E A R C H support the suppressor analysis described here. This work suggested that hgll suppressors involved nutrient sensing mechanisms or pathways. The transcriptional regulators Sfllp and Flo8p of S. cerevisiae are also compositionally similar to Hgllp, though the probability that the sequences belong to the same family was only 68%. Flo8p is the regulator of transcription for Flol lp; this protein has 5 consensus P K A sites, and is under regulatory control by S. cerevisiae P K A (Pan and Heitman, 1999).  Phosphorylated Flo8p induces the transcription of FLOll.  transcription factor Sfllp is a negative regulator of FLOll  Conversely, the T A T A  gene expression. Sfllp also has 5  consensus P K A target sites and interacts with P K A ; Sfllp represses the transcription of (J. Heitman, pers comrh).  FLOll  Identification of Hgllp's similar composition to regulatory  transcription factors may imply that Hgllp is involved in the transcriptional regulation of genes during U. maydis morphogenesis.  The Hgllp amino acid sequence did not share any other  similarity to the AreAp, Sfllp, or Flo8p polypeptides that could be identified by a BLAST2 comparison either in the presence or absence of filters. The conservation of a motif between two proteins places less emphasis on overall sequence similarity, and instead focuses attention on regions of a sequence which may be responsible for protein function.  The presence of described motifs may suggest similar  functions for regions of a protein irrespective of its overall function. eMOTIF, PRO-DOM, PROSITE, S M A R T , P A T - S C A N , and B L O C K S represent some of the programs used in the analysis of the Hgllp domain structure. Not all of the domain databases share the same domain or profile entries, though many are derivatives of the large PROSITE database.  Regions  identified as having similarity to known structural or functional motifs were examined for their location in the Hgllp sequence; these sequences were within regions of Hgllp predicted to have  56  Discussion  significant complexity. These regions were also examined for their conservation with other known motifs that commonly co-exist in proteins that share the domains. eMOTIF searches revealed several putative motifs of interest in Hgllp (Figure 10). The cdc25 GDS family motif might indicate a role in signaling, and might be supported by the identification of the GPCR motif in the N-terminus of the protein due to their similar roles in Gprotein mediated signaling. A signaling role for Hgllp could be significant given the extensive arrangement of putative phosphorylation sites and possible tyrosine phosphatase motif in the N terminus (Figure 10). The coupling of kinase and phosphatase sites is a common regulatory structure for signaling molecules (Millward et al., 1999). At present, a definitive function for Hgllp was not determined from database comparisons. The phenotypes of hgll' strains implied that it was involved in both mating and c A M P / P K A filamentation, but it remained unclear how hgll functions in these processes. The presence of PEST sequences has been demonstrated to affect protein stability in a number of different organisms (Rechsteiner and Rogers, 1996). PEST sequences are regions of a peptide enriched for proline (P), glutamate (E), serine (S), and threonine (T) residues, resulting in hydrophobic properties ubiquitination pathway.  that increase  susceptibility to proteolysis by the  PEST sequences are evaluated based on a scale from +50 to -50.  Scores above zero are considered to be probable PEST regions, while scores above 5 are considered to be quite significant (Rechsteiner and Rogers, 1996). The predicted amino acid sequence for the Hgllp protein contained 4 potential PEST sequences scoring above zero, two of which had a score above 5 (Figure 10). There were two additional PEST regions in the Hgllp amino acid sequence which scored approximately -6, but were flanked P K A target sequences at one or both ends of the PEST region. Phosphorylation may act coordinately with nearby PEST regions resulting in increased hydrophobicity and susceptibility to proteolysis  57  Discussion  (Brown et al., 1995). These results may indicate that Hgllp has a rapid turnover rate, and may be targeted to the proteolytic pathway in a phosphorylation-dependent manner. A PROSITE analysis of the Hgllp amino acid sequence identified eight putative P K A target sequences "[RK](2)-x-[ST]" (Figure 10). Four of the proposed sites resided in the N terminus, and flanked the possible GPCR transmembrane motif, or the transmembrane motif identified by Tmpred.  The identification of hgll' as a suppressor of the adrl'  phenotype  suggested these phosphorylation sites may be important for Hgllp function in the presence of P K A . The absence of Hgllp caused a budding growth phenotype in the presence or absence of P K A , suggesting that it may function as a suppressor of budding growth.  Therefore, one  proposed model is that phosphorylated Hgllp may suppress budding growth. This prediction was supported by the adrl' phenotype, and by the filamentous mutants collected in both the 3020 and 3034 strains.  Presumably there was non-phosphorylated Hgllp present in adrl'  strains which suppressed the budding growth phenotype (ultimately causing a filamentous growth phenotype). Mutation in hgll restored budding growth to adrl' strains, and also caused budding growth in the wild type background.  The demonstration of Hgllp and Adrlp  interactions in vitro (Hgllp phosphorylation experiment) provided additional support for this model.  4.2.2 Phosphorylation of Hgllp by P K A  Attempts were made to examine the phosphorylation state of Hgllp in U. maydis strains using epitope-tagged versions of Hgllp. It was not possible to recover purified Hgllp protein expressed from the hglwmyc, or hgly.RlS constructs during experiments with U. maydis or E. coli. It is not clear why Hgllp protein could not be produced; several expression conditions  58  Discussion  were tested and all were unsuccessful. It could be that the product was extremely unstable following expression under native and recombinant conditions, or that there was some unknown mechanism or circumstance preventing its accumulation. It has been documented that proteins rich in arginine codons A G A and A G G are difficult to express in E. coli (Baneyx, 1999; Ivanov et al., 1997). A n Expasy compositional analysis identified the Hgllp sequence as arginine rich compared to conventional yeast proteins. Hgllp does have two such codons (AGA/AGG) early in its coding sequence. However, they do not comprise a significant proportion of the arginine codons present in the polypeptide, likely eliminating this possibility. The in vitro phosphorylation data demonstrated that Hgllp served as a target for both bovine and U. maydis P K A . The PKA-dependent phosphate transfer was partially inhibited by PKI, and phosphate group removal was seen with lambda phosphatase. PKI is a 20 amino acid peptide which specifically and competitively binds the catalytic subunit of P K A to inhibit its kinase activity (Demaille et al, 1977; Scott et al, 1985; Takio et al, 1980). Addition of PKI to the reactions was anticipated to eliminate the mobility shift, however, reactions with Hgllp and bovine or Ustilago P K A in the presence of PKI were not completely inhibited. The PKI partial inhibition result was consistent between the two kinase experiments.  Despite the reported  stability of PKI and the molar excess added to each experiment (1000 fold), long-term storage and multiple freeze-thaw cycles may have diminished the inhibitory potency or other unknown qualities of the peptide. Both panels in figure 15 show the same trend of increased apparent molecular weight when Hgllp is phosphorylated by P K A . While it was demonstrated that Hgllp served as a target of P K A in vitro, it was not possible to imply in vivo functional significance based on this result alone. It is interesting, however, that the predicted protein had several phosphorylation sites and served as a suppressor to the adrl' phenotype. Recent work in S. cerevisiae by Pan and Heitman (1999) has addressed 59  Discussion the possibility that the transcription factors Flo8p and Sfllp serve as targets of the S. cerevisiae P K A Tpk2p.  Their work is limited to epistasis and yeast two-hybrid analysis, but was  conducted under the premise that the 5 P K A sites in Flo8p and Sfllp may indicate a role in P K A signal transduction. Along the same lines, Hgllp contained 8 P K A target sites, and by epistasis analysis suppressed the low P K A phenotypes in U. maydis (Figure 21). The work presented here further implicated an enzymatic interaction between the Hgllp and U. maydis P K A proteins in vitro.  S.  cerevisiae  U.  Tpk2  0 8  maydis Adrl  Sfl1  (5 P K A sites each)  Pseudohyphal growth  Hgll  (8  P K A sites)  Filamentous growth  Figure 21. Comparison of the 5*. cerevisiae P K A target proteins Flo8p and Sfllp with the proposed U. maydis P K A target protein Hgllp. Each of the three proteins (Flo8p, Sfllp, and Hgllp) suppresses phenotypes of P K A mutants, and each contains several consensus P K A target sequences.  It will be necessary to conduct experiments which investigate the nature and consequences of Hgllp phosphorylation in vivo. Ideally it would be interesting to recover native Hgllp from wild-type, ubcl', and adrl' cell types to examine the phosphorylation state of Hgllp in each of these strains. One would expect wild-type cells to have an intermediary state of Hgllp phosphorylation dependent on the conditions examined, the ubcl' strain to have a  60  Discussion hyper-phosphorylated version of Hgllp, and for the adrl' cells to have a non-phosphorylated or weakly phosphorylated version of Hgllp. Diirrenberger et al. (1998) identified regulated P K A activity as necessary for pathogenesis and virulence; this may suggest that there is a regulatory function for P K A phosphorylation of Hgllp that is required in vivo for pathogenicity and virulence.  A site directed mutagenesis experiment in which each of the kinase target sites  individually (and in various combinations) are eliminated could examine the functional importance of phosphorylation.  A n alternative approach may be to attempt random  mutagenesis experiments with hgll to look for mutants which lose the ability to complement the hgll' phenotype. It may be possible to recover mutants which only partially complement the existing phenotypes, potentially identifying mutations in P K A phosphorylation sites or other motifs, and providing support for the importance of Hgllp phosphorylation in the P K A pathway.  4.2.3 Suppressor analysis hgll' strains U.V. induced mutants which either partially, or completely suppressed the hgll' budding growth phenotype were isolated.  This suggested that there were components of pathways  participating in filamentation that were epistatic to hgll. Recovery of filamentous cell types in the adrl' hgll' background, which strongly resembled hgll' suppressor mutants, suggested there were additional factors involved in the U. maydis dimorphic switch.  Suppression of the  budding hgll' phenotype in both backgrounds suggested that the mutations recovered were epistatic to hgll and possibly adrl. Perhaps additional targets of Adrlp exist which counteract or negatively regulate Hgllp and result in Hgllp serving as an activator of filamentous growth (Figure 22). The presence of Adrlp in these mutants might activate these genes or proteins at  61  Discussion  the same level as Hgllp in the pathway to cause a filamentous phenotype. It is not known whether the hgll' suppressor mutations also suppress the teliospore defect of the hgll' strains.  Adrl  B.  Adrl i  i  • Hgll  Hgll i  ?  *  i  Budding Growth  Filamentous Growth  Figure 22. Hgllp suppressors represent a variety of functions. Dashed lines represent interactions that may involve intermediate components. A . Experiments in this work suggest that Adrlp acts directly on Hgllp. As a direct target of Adrlp, Hgllp may serve as a repressor of budding growth. B . If there are intermediate proteins between Adrlp and Hgllp, then Hgllp may serve as an inhibitor of some other regulator of filamentous growth. For both models there are likely other proteins downstream of Hgllp that function to regulate filamentous growth in U. maydis.  The isolation of mutations which did not alter their phenotypes on different media suggests nutrient availability may affect filamentous growth. For mutants which do respond to nutrient availability, generally, filamentation became exaggerated as the nutrients were depleted. Similarly, for mutants with increased production of a pigment similar to the hgll' yellow pigment, pigment production appeared to increase at lower nutrient levels. It was possible to collect filamentous mutants which had decreased pigment production. A loss of pigment production in a filamentous suppressor mutant suggested that perhaps pigment productions was not be directly linked to the filamentous growth phenotype. The pigmentation response to nutrient availability appeared common to both the original hgll' knockout and the suppressor mutants alike.  Perhaps i f pigment production was somehow involved in teliospore 62  Discussion  formation (deficient in the hgll" mutants), the increased response to nutrient starvation may implicate nutrient sensing as part of the developmental program. Variability between different suppressor mutants and the amount of filamentation and black pigment production suggested interactions between these two phenotypes, Hgllp function, and nutrient sensing.  4.2.4 Complementation of hgll' suppressor mutants There were several difficulties encountered when selected mutants were used in complementation experiments and these prevented recovery of a complementing gene.  The  filamentous phenotype of mutant strains made protoplast formation extremely difficult, resulting in a low yield of cells with poor competency. In addition to this, the genomic cosmid libraries used for transformation had low transformation efficiencies. Taken together, the poor competency cells and low transformation efficiency made it unlikely that complementing genes would be identified with this approach. Several rounds of transformation were attempted with several different protoplast preparations for each of the hgll' disruption backgrounds and all of these trials were unsuccessful.  In future work, the insertional mutagenesis approach of  restriction enzyme mediated integration (REMI) should be used to isolate suppressor mutants. If these mutants can be obtained, the REMI procedure would facilitate subsequent cloning of the disrupted gene.  4.2.5 Summary of hgll' phenotypes The hgll' gene affects processes leading to filamentous growth in U. maydis.  A  summary of the phenotypes of hgll' mutants observed in different strain backgrounds supports the model of Hgllp being a suppressor of budding growth that is inactive in conditions of high P K A activity. Table 3 summarizes the phenotypes of strains defective for components of the 63  Discussion  P K A pathway in U. maydis, and includes the phenotypes caused by mutations in hgll.  PKA l e v e l s  HgMp  Morphology  Requlated A d r l  Variable ®-Hgl1  Variable morphology  ubcV  Hiqh A d r l  ®-Hgl1  Budding  adrl  NO A d r l  HgM  Filamentous  hgll  Regulated A d r l  NO HgM  Budding  adrl hgll  NO A d r l  NO HgM  Budding  hgll- S u p .  +/- A d r l  NO HgM  Filamentous  Strain Wild type  Table 3. Summary of hgll phenotypes. The 'strain' column lists the strains summarized in each of the P K A levels, Hgllp, and Morphology columns. Column ' P K A levels' describes the state of A d r l , the major catalyic subunit of P K A in U. maydis. Column ' H g l l p ' describes the probable state of the Hgllp protein in the strain, ©-Hgllp indicates that Hgllp is likely phosphorylated in the conditions of the background strain. The 'morphology' column presents the overall phenotype of colony growth, the variable morphology of the wild type strain can range from budding to filamentous growth. Row 'hgll' Sup.' represents hgll suppressors collected in this work.  From table 3, if the Hgllp protein was in the phosphorylated state in the ubcl background , then Hgllp may be a suppressor of budding growth that was inactivated by P K A .  In support of the  suppressor of budding growth model, the phenotypes of the adrl' or hgll' strains also suggested that Hgllp inhibited budding growth. The ubcl' strain shared similarities with hgll' mutants and the presence of Hgllp in the low P K A background was associated with a filamentous phenotype. This suggested that Hgllp may have been an activator of filamentous growth in the absence of P K A .  The collection of filamentous mutants in hgll' strains indicated that Hgllp  was not essential for filamentation, and is not likely an activator of filamentous growth.  64  Discussion  4.3 Conclusions Previous studies from our laboratory have identified many components of the cAMPdependent P K A signal transduction pathway. Many of these components (adenylyl cyclase, and the P K A regulatory and catalytic subunits) are well characterized proteins. This has aided in resolving their function in the U. maydis signal transduction pathway to establish filamentous growth. Recovery of a novel U. maydis gene (hgll) as a suppressor of the low P K A phenotype prompted experiments to examine how Hgllp participated during signal transduction and morphogenesis.  Sequence analysis of Hgllp revealed several regions of the polypeptide that  suggest it may function as a signaling molecule. In addition to possession of these regions, this thesis demonstrated phosphorylation of Hgllp by U. maydis P K A in vitro, providing biochemical evidence for its involvement in the c A M P / P K A signal transduction pathway. Finally, it was possible to suppress the hgll' colony and cellular morphologies in culture with mutations obtained by U . V . mutagenesis. Suppression of hgll' supported the current model that Hgllp serves as a repressor of budding growth in culture, and that Hgllp is not required for filamentous growth in culture. Whether the hgll' phenotypes were compromised in planta by the suppressor mutants remains to be investigated, but will provide exciting insights and information regarding events occurring during pathogenesis.  65  References  5. References  Akada, R., Kallal, L., Johnson, D J . and Kurjan, J. (1996) Genetic relationships between the G Protein Beta-Gamma complex, Ste5p, Ste20p and Cdc42p - investigation of effector roles in the yeast pheromone response pathway. Genetics, 143, 103-117. Altschul, S.F., Madden, T.L., Schaffer, A . A . , Zhang, J., Zhang, Z., Miller, W. and Lipman, D.J. (1997) Gapped B L A S T and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Research, 25, 3389-3402.  Attwood, T.K. and Findlay, J.B.C. (1994) Fingerprinting G-protein-coupled receptors. Protein Engineering, 1, 195-203. Baneyx, F. (1999) Recombinant protein expression in Escherichia coli. Current Opinion in Biotechnology, 10, 411-421. Banuett, F. (1995) Genetics of Ustialgo maydis, a fungal pathogen that induces tumors in maize. Annual Review of Genetics, 29, 179-208.  Banuett, F. (1998) Signalling in the Yeasts: A n informational cascade with links to the filamentous fungi. Microbiology  & Molecular Biology Reviews, 62, 249-274.  Banuett, F. and Herskowitz, I. (1996) Discrete Developmental Stages During Teliospore Formation in the Corn Smut Fungus, Ustilago maydis. Development, 122, 2965-2976. Barrett, K.J., Gold, S.E. and Kronstad, J.W. (1993) Identification and complementation of a mutation to constitutive filamentous growth in Ustilago maydis. Molecular Plant Microbe Interactions,  6, 274-283.  Baur, M . , Esch, R.K. and Errede, B. (1997) Cooperative binding interactions required for function of the T y l sterile responsive element. Molecular and Cellular Biology, 17, 4330-4337. Boguski, M . , Lewitter, F., Altschul, S., Brenner, S., Haussler, D., Eddy, S., Hofmann, K., Lake, J., Moore, J., Kanehisa, M . , Brownstein, M . , Trent, J., Boguski, M . and Thornton, J. (1998) Trends guide to bioinformatics. In Brenner, S., Lewitter, F., Patterson, M . and Brines, R. (eds.), Elsevier Trends Journals. Peter Desmond, Cambridge, U K . Bolker, M . (1998) Sex and crime: Heterotrimeric G proteins in fungal mating and pathogenesis. Fungal Genetics & Biology, 25, 143-156.  Bolker, M . , Urban, M . and Kahmann, R. (1992) The a mating type locus of Ustilago maydis specifies cell signaling components. Cell, 68, 441-450.  66  References  Boy-Marcotte, E., Lagniel, G., Perrot, M . , Bussereau, F., Boudsocq, A., Jacquet, M . and Labarre, J. (1999) The heat shock response in yeast: differential regulations and contributions of the Msn2p/Msn4p and Hsflp regulons. Molecular Microbiology, 33, 274-283. Brown, K., Gerstberger, S., Carlson, L., Granzoso, G. and Siebenlist, U . (1995) Control of IkBalpha proteolysis by site-specific signal induced phosphorylation. Science, 267, 14851488. Bucher, P. and Bairoch, A. (1994) A gerneralized profile syntax for biomolecular sequences motifs and its function in automatic sequence interpretation. International Conference on Intelligent Systems for Molecular Biology. A A A I Press, Menlo Park, Vol. 2, pp. 5361. Chandarlapaty, S. and Errede, B . (1998) A s h l , a daughter cell-specific protein, is required for pseudohyphal growth of Saccharomyces  cerevisiae. Molecular & Cellular Biology, 18,  2884-2891. Christensen, J.J. (1963) Corn smut caused by Ustilago maydis. American Society Monograph, Saint Paul, Vol. 2, p. 41.  Phtyopathological  Cook, J.G., Bardwell, L., Kron, S.J. and Thorner, J. (1996) Two novel targets of the map kinase Kssl are negative regulators of invasive growth in the yeast Saccharomyces cerevisiae. Genes & Development, 10, 2831-2848.  Day, P.R. and Anagnostakis, S.L. (1971) Corn smut dikaryon in culture. Nature New Biology, 231, 19-20. Demaille, J.G., Peters, K . A . and Fischer, E.H. (1977) Isolation and properties of the rabbit skeletal muscle protein inhibitor of adenosine 3',5'-monophosphate dependent protein kinases. Journal of Biological  Chemistry, 16, 3080-3086.  Dolan, J.W., Kirkman, C. and Fields, S. (1989) The yeast STE12 protein binds to the D N A sequence mediating pheromone induction. Proceedings of the National Academy of Sciences, U. S. A., 86, 5703-5707. Diirrenberger, F., Wong, K . and Kronstad, J.W. (1998) Identification of a cAMP-dependent protein kinase catalytic subunit required for virulence and morphogenesis in Ustilago maydis. Proceedings  of the National Academy of Sciences of the U.S.A., 95, 5684-5689.  Elion, E.A., Satterberg, B. and Kranz, J.E. (1993) Fus3 phosphorylates multiple components of the mating signal transduction cascade: evidence for Stel2 and Farl. Molecular Biology of the Cell, 4, 495-510. Errede, B. and Ammerer, G. (1989) STE12, a protein involved in cell-type-specific transcription and signal transduction in yeast, is part of protein-DNA complexes. Genes and Development, 3, 1349-1361. 67  References  Field, J., Nikawa, J.-i., Broek, D., MacDonald, B., Rogers, L., Wilson, I.A., Lerner, R.A. and Wigler, M . (1998) Purification of a iMS-responsive adenylyl cylcase complex from Saccharomyces cerevisiae by use of an epitope addition method. Molecular and Cellular 8,2159-2165. Gavrias, V., Andrianopoulos, A., Gimeno, C.J. and Timberlake, W.W. (1996) Saccharomyces cerevisiae TEC1 is required for pseudohyphal growth. Molecular Microbiology, 19, 1255-1263. Gillissen, B., Bergemann, J., Sandmann, C , Schroeer, B., Bolker, M . and Kahmann, R. (1992) A Two-Component Regulatory System For Self/Non-Self Recognition in Ustilago maydis. Cell, 68, 647-657. Gimeno, C.J., Ljungahl, P.O., Styles, C A . and Fink, G.R. (1992) Unipolar cell divisions in the yeast S. cerevisiae lead to filamentous growth: regulation by starvation and RAS. Cell, 68, 1077-1090. Gold, S.E., Duncan, G.A., Barrett, K.J. and Kronstad, J.W. (1994) c A M P regulates morphogenesis in the fungal pathogen Ustilago maydis. Genes and Development, 8, 2805-2816. Gorner, W., Durchschlag, E., Martinezpastor, M.T., Estruch, F., Ammerer, G., Hamilton, B., Ruis, H . and Schuller, C. (1998) Nuclear Localization of the C2H2 Zinc Finger Protein Msn2p Is Regulated By Stress and Protein Kinase a Activity. Genes & Development, 12, 586-597. Gustin, M . C , Albertyn, J., Alexander, M . and Davenport, K. (1998) M A P kinase pathways in the yeast Saccharomyces cerevisiae [Review]. Microbiology & Molecular Biology Reviews, 62, 1264-1300. Hartmann, H.A., Kahmann, R. and Bolker, M . (1996) The Pheromone Response Factor Coordinates Filamentous Growth and Pathogenicity in Ustilago maydis. EMBO Journal, 15, 1632-1641. Henikoff, S. and Henikoff, J.G. (1994) Protein family classification based on searching a database of blocks. Genomics, 19, 97-107. Herskowitz, I., Park, H.O., Sanders, S., Valtz, N . and Peter, M . (1995) Programming of cell polarity in budding yeast by endogenous and exogenous signals. Cold Spring Harbor Symposia on Quantitative Biology, 60, 717-727'. Hirschman, J.E., De Zutter, G.S., Simonds, W.F. and Jenness, D.D. (1997) The G-beta-gamma complex of the yeast pheromone response pathway. Subcellular fraction and proteinprotein interactions. Journal of Biological Chemistry, 272, 240-248.  68  References Hobohm, U . and Sander, C. (1995) A sequence property approach to searching protein databases. Journal of Molecular Biology, 251, 390-399. Hofmann, K., Bucher, P., Falquet, L. and Bairoch, A. (1999) The PROSITE database, its status in 1999. Nucleic Acids Research, 27, 215-219. Hofmann, K. and Stoffel, W. (1993) TMbase - A database of membrane spanning proteins segments. Biological Chemistry, 166. Holliday, R. (1974) Ustilago maydis, in Handbook of Genetics. Plenum, New York. Hung, W., Olson, K . A . , Breitkreutz, A. and Sadowski, I. (1997) Characterization of the basal and pheromone-stimulated phosphorylation states of Stel2p. European Journal of Biochemistry, 245, 241 -251. Ivanov, I.G., Saraffova, A . A . and Abouhaidar, M . G . (1997) Unusual effect of clusters of rare arginine (AGG) codons on the expression of human interferon alpha 1 gene in Escherichia coli. International Journal of Biochemistry and Cell Biology, 29, 659-666. Kao, L.R., Peterson, J., Ji, R., Bender, L. and Bender, A. (1996) Interactions between the ankyrin repeat-containing protein A k r l p and the pheromone response pathway in Saccharomyces cerevisiae. Molecular & Cellular Biology, 16, 168-178. Kronstad, J.W., Holly, J.A. and MacKay, V . L . (1987) A yeast operator overlaps an upstream activation site. Cell, 50, 369-377. Kruger, J., Loubradou, G., Regenfelder, E., Hartmann, A. and Kahmann, R. (1998) Crosstalk between c A M P and pheromone signalling pathways in Ustilago maydis. Molecular & General Genetics, 260, 193-198. Kusch, G. and Schauz, K. (1989) Light and electron microscopic studies of chlamydospore development in Ustilago maydis (Ustilaginales, Basidiomycetes). Cryptogamic Botany, 1,230-235. Lambrechts, M . G . , Bauer, F.F., Marmur, J. and Pretorius, LS. (1996) M u c l , a mucin-like protien that is regulated by MsslO, is critical for pseudohyphal differentiation in yeast. Proceedings of the National Academy of Sciences U.S.A., 93, 8419-8424. Leberer, E., Thomas, D.Y. and Whiteway, M . (1997a) Pheromone signalling and polarized morphogenesis in yeast. Current Opinion in Genetics & Development, 7, 59-66. Leberer, E., Wu, C.L., Leeuw, T., Fourestlieuvin, A., Segall, J.E. and Thomas, D.Y. (1997b) Functional characterization of the Cdc42p binding domain of yeast Ste20p protein . kinase. EMBO Journal, 16, 83-97.  6 9  References  Leeuw, T., Fourest-Lieuvin, A., Wu, C , Chenevert, J., Clark, K., Whiteway, M . , Thomas, D., Y. and Leberer, E. (1995) Pheromone response in yeast: Association of Bern lp with proteins of the M A P kinase cascade and actin. Science, 270, 1210-1213. Lehmler, C , Steinberg, G., Snetselaar, K . M . , Schliwa, M . , Kahmann, R. and Bolker, M . (1997) Identification of a motor protein required for filamentous growth in Ustilago maydis. The EMBO Journal, 16, 3464-3473.  Lew, D.J. and Reed, S.I. (1995) Cell cycle control of morphogenesis in budding yeast. Current opinion in Genetics and Development, 5, 17-23.  Liu, H., Styles, C A . and Fink, G.R. (1996) Saccharomyces cerevisiae S288C has a mutation in FL08, a gene required for filamentous growth. Genetics, 144, 967-978. Lo, W.-S. and Dranginis, A . M . (1998) The cell surface flocculin Flol 1 is required for pseudohyphae formation and invasion by Saccharomyces cerevisiae. Mollecular and Cellular Biology, 9, 161-171.  Lorenz, M . C . and Heitman, J. (1997) Yeast pseudohyphal growth is regulated by Gpa2, a G protein alpha homolog. EMBO Journal, 16, 7008-7018. Lorenz, M . C . and Heitman, J. (1998) The Mep2 Ammonium Permease Regulates Pseudohyphal Differentiation in Saccharomyces  cerevisiae. EMBO Journal, 17, 1236-1247.  Lyons, D.M., Mahanty, S.K., Choi, K . Y . , Manandhar, M . and Elion, E.A. (1996) The SH3Domain Protein Beml Coordinates Mitogen-Activated Protein Kinase Cascade Activation With Cell Cycle Control in Saccharomyces cerevisiae. Molecular & Cellular Biology, 16, 4095-4106. Madden, K. and Snyder, M . (1998) Cell polarity and morphogenesis in budding yeast. Annual Review of Microbiology,  52, 687-744.  Madhani, H.D. and Fink, G.R. (1997) Combinatorial control required for the specificity of yeast M A P K Signaling. Science, 275, 1314-1317. Madhani, H.D. and Fink, G.R. (1998) The riddle of M A P kinase signaling specificity. Trends in Genetics, 14, 151-155. Marchler, G., Schuller, C , Adam, G. and Ruis, H. (1993) A Saccharomyces cerevisiae UAS element controlled by protein kinase A activates transcription in response to a variety of stress conditions. EMBO Journal, 12, 1997-2003. Martinez-Pastor, M.T., Marchler, G., Schuller, C , Marchler-Bauer, A., Ruis, H. and Estruch, F. (1996) The Saccharomyces cerevisiae zinc finger proteins Msn2p and Msn4p are required for transcriptional induction through the stress-response element (STRE). EMBO Journal, 15, 2227-2235.  70  References  Mayorga, M . E . and Gold, S.E. (1999) A M A P kinase encoded by the ubc3 gene of Ustilago maydis is required for filamentous growth and full virulence. Molecular Microbiology, 34, 485-497. Millward, T.T., Zolnierowicz, S. and Hemmings, B . A . (1999) Regulation of protein kinase cascades by protein phosphatase 2A. TIBS, 24, 186-191. Miyajima, I., Nakafuku, M . , Nakayama, N . , Brenner, C. and Miyajima, A. (1987) GPA1, a haploid-specific essential gene, encodes a yeast homolog of mammalian G protein which may beinvolved in mating factor signal transduction. Cell, 50, 1011-1019. Mosch, H.U. and Fink, G.R. (1997) Dissection of filamentous growth by transposon mutagenesis in Saccharomyces cerevisiae. Genetics, 145, 671-684. Mosch, H.U., Robers, R.L. and Fink, G.R. (1996) Ras2 signals via the Cdc42/Ste20/mitogenactivated protein kinase moducle to induce filamentous growth in Saccharomyces cerevisiae. Proceedings of the National Academy of Sciences USA, 93, 5352-5356. Muller, P., Aichinger, C., Feldbrugge, M . and Kahmann, R. (1999) The M A P kinase Kpp2 regulates mating and pathogenic development in Ustilago maydis. Molecular Microbiology, 34, 1007-1017. Pan, X . W . and Heitman, J. (1999) Cyclic AMP-dependent protein kinase regulates pseudohyphal differentiation in Saccharomyces cerevisiae. Molecular & Cellular Biology, 19, 4874-4887. Pryciak, P . M . and Hartwell, L . H . (1996) A k r l encodes a candidate effector of the G-BetaGamma complex in the Saccharomyces cerevisiae pheromone response pathway and contributes to control of both cell shape and signal transduction. Molecular & Cellular Biology, 16, 2614-2626. Rechsteiner, M . and Rogers, S.W. (1996) PEST sequences and regulation by proteolysis. in Biochemical Sciences, 21, 267-271.  Trends  Reinders, A., Burckert, N., Boiler, T., Wiemken, A . and De Virgilio, C. (1998) Saccharomyces cerevisiae cAMP-dependent protein kinase controls entry into stationary phase through the Riml5p protein kinase. Genes and Development, 12, 2943-2955. Roberts, R., Mosch, H.-U. and Rink, G.R. (1997) 14-3-3 proteins are esential for R A S / M A P K cascade signaling during pseudohyphal development in S. cerevisiae. Cell, 89, 10551065. Robertson, L.S. and Fink, G.R. (1998) The three yeast A kinases have specific signaling functions in pseudohyphal growth. Proceedings of the National Academy of Sciences the U.S.A., 95, 13783-13787.  71  of  References  Rupp, S., Summers, E., Lo, H.-J., Madhani, H. and Fink, G. (1999) M A P kinase and cAMP filamentation signaling pathways converge on the unusually large promoter of the yeast FLOll  gene. The EMBO Journal, 18, 1257-1269.  Sadowski, I., Niedbala, D., Wood, K. and Ptashne, M . (1991) G A L 4 is phosphorylated as a consequence of transcriptional activation. Proceedings of the National Academy of Sciences USA, 88, 10510-10514. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Cold Spring Harbor Laboratories, Plainview, N . Y .  Manual.  Schultz, J., Milpetz, F., Bork, P. and Ponting, C P . (1998) SMART, a simple modular architecture research tool: Identification of signalling domains. Proceedings of the National Academy of Sciences, U.S.A., 95, 5857-5864.  Scott, J.D., Fischer, E.H., Demaille, J.G. and Krebs, E.G. (1985) Identification of an inhibitory region of the heat-stable protein inhibitor of the cAMP-dependent protein kinase. Proceedings of the National Academy of Sciences, USA, 82, 4379-4383.  Smith, A., Ward, M.P. and Garrett, S. (1998) Yeast P K A represses Msn2p/Msn4p-dependent gene expression to regulate growth, stress response and glycogen accumulation. EMBO Journal, 17, 3556-3564. Snetselaar, K . M . , Bolker, M . and Kahmann, R. (1996) Ustilago maydis mating hyphae orient their growth toward pheromone sources. Fungal Genetics & Biology, 20, 299-312. Spellig, T., Boelker, M . , Lottspeich, F., Frank, R., W and Kahmann, R. (1994) Pheromones trigger filamentous growth in Ustilago maydis. Embo, 13, 1620-1627. Spellig, T.R., Erika. and Reichmann, M.S., Florian. Bohlmann, Ralph. Urban, Martin. Boelker, Michael. Kaemper, Joerg. Kahmann, Regine. (1994) Control of mating and development in Ustilago maydis. Antonie van Leeuwenhoek, 65, 191-197.  Takio, K., Walsh, K . A . , Neurath, H., Smith, S.B., Krebs, E.G. and Titani, K. (1980) The amino adic sequence of a hinge protein in the regulatory subunit of bovine cardiac muscle cylic Amp-dependent protein kinase II. FEBS letters, 114, 83-88. Tedford, K., Kim, S., Sa, D., Stevens, K. and Tyers, M . (1997) Regulation of the mating pheromone and invasive growth responses in yeast by two M A P kinase substrates. Current Biology, 7, 228-238.  Thevelein, J.M. and de Winde, J.H. (1999) Novel sensing mechanisms and targets for the cAMP-protein kinase A pathway in the yeast Saccharmyces cerevisiae. Molecular Microbiology, 33, 904-918. Thoma, F. (1999) Light and dark in chromatin repair: repair of UV-induced D N A lesions by photolyase and nucleotide excision repair. EMBO Journal, 18, 6585-6598. 72  References  Toda, T., Uno, I., Ishikawa, T., Powers, S., Kataoka, T., Broek, D., Cameron, S., Broach, J., Matsumoto, K. and Wigler, M . (1985) In yeast, RAS proteins are controlling elements of adenylate cyclase. Cell, 40, 27-36. Tokiwa, G., Tyers, M . , Volpe, T. and Futcher, B. (1994) Inhibition of G l cyclin activity by the Ras/cAMP pathway in yeast. Nature, 371, 342-345. Urban, M . , Kahmann, R. and Bolker, M . (1996) Identification of the pheromone response element in Ustilago maydis. Molecular & General Genetics, 251, 31-37.  Vidan, S. and Mitchell, A.P. (1997) Stimulation of yeast meiotic gene expression by the glucose-repressible protein kinase Riml5p. Molecular and Cellular Biology, 17, 26882697. Ward, M.P., Gimeno, C.J., Fink, G.R. and Garrett, S. (1995) SOK2 may regulate cyclic A M P dependent protein kinase-stimulated growth and pseudohyphal development by repressing transcription. Molecular and Cellular Biology, 15, 6854-6863. Whiteway, M . , Hougan, L., Dignard, D., Thomas, D.Y., Bell, L., Saari, G . C , Grant, F.J., O'Hara, P. and MacKay, V . L . (1989) The STE4 and STE18 genes of yeast encode potential beta and gamma subunits of the mating factor receptor-coupled G protein. Cell, 56, 467-477. Wilson, R.A. and Arst, H.N. (1998) Mutational analysis of A R E A , a transcriptional activator mediating nitrogen metabolite repression in Aspergillus nidulans and a member of the "Streetwise" G A T A family of transcription factors. Microbiology and Molecular Biology Reviews, 62, 586-596.  Wright, W., Scordis, P. and Attwood, T.K. (1999) B L A S T PRINTS - A n alternative prespective on sequence similarity. Bioinformatics, 15 - in press. Xue, Y., Batlle, M . and Hirsch, J.P. (1998) GPR1 encodes a putative G protein-coupled receptor that associates with the Gpa2p G subunit and functions in a Ras-independent pathway. EMBO Journal, 17, 1996-2007.  Zhao, Z.-S., Leung, T., Manser, E. and Lim, L. (1995) Pheromone signalling in Saccharomyces cerevisiae requires the small GTP-binding protein Cdc42p and its activator CDC24. Mollecular  and Cellular Biology, 15, 5246-5257.  73  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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


Related Items