UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The cloning and characterization of a THN reductase gene from ophiostoma floccosum 387n Eagen, Rebecca 2000

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

Item Metadata


831-ubc_2000-486311.pdf [ 16.93MB ]
JSON: 831-1.0089838.json
JSON-LD: 831-1.0089838-ld.json
RDF/XML (Pretty): 831-1.0089838-rdf.xml
RDF/JSON: 831-1.0089838-rdf.json
Turtle: 831-1.0089838-turtle.txt
N-Triples: 831-1.0089838-rdf-ntriples.txt
Original Record: 831-1.0089838-source.json
Full Text

Full Text

THE CLONING AND CHARACTERIZATION OF A THN REDUCTASE GENE FROM OPHIOSTOMA FLOCCOSUM 387N by REBECCA EAGEN B.Sc, University of Western Ontario, 1988 M.Sc, Simon Fraser University, 1991 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Department of Wood Science We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December 1999 ©Rebecca Eagen, 1999 In p resen t i ng this thesis in partial fu l f i lment of the requ i rements fo r an a d v a n c e d d e g r e e at the Univers i ty of Brit ish C o l u m b i a , I agree that the Library shall m a k e it f reely avai lable fo r re fe rence and s tudy. I fur ther agree that p e r m i s s i o n fo r ex tens ive c o p y i n g o f th is thesis fo r scho lar ly p u r p o s e s may b e gran ted by the h e a d of my depa r tmen t o r by his o r her representat ives. It is u n d e r s t o o d that c o p y i n g o r pub l i ca t i on of this thesis for f inancial gain shal l n o t b e a l l o w e d w i t h o u t m y wr i t t en p e r m i s s i o n . D e p a r t m e n t T h e Un ivers i ty of Brit ish C o l u m b i a V a n c o u v e r , C a n a d a D E - 6 (2/88) A B S T R A C T Sapstaining fungi, in particular Ophiostoma sp., cause significant economic problems for Canada's lumber industry and current control methods are less than ideal. The visible stain found on lumber is caused by melanin produced in the mycelia of sapstaining fungi as they grow through the sapwood. O. floccosom 387N is frequently isolated from stained lumber but knowledge of this fungus is lacking. A molecular biological approach, in addition to elucidating important data about this organism, may provide information that could prove valuable in designing specific melanin inhibitors or creating pigment deficient strains that could be used for biocontrol. The novel application and adaptations of molecular biological techniques to O. floccosom 387N were used to isolate and characterize a gene from the melanin biosynthetic pathway. Culture conditions were defined during which pigmentation and biomass yields were significant. Application of dihydroxynaphthalene (DHN) pathway inhibitors caused the reduction of visible pigment production in 387N and allowed us to speculate that the DHN pathway of melanin biosynthesis was important in 387N. Hybridization data indicated that 387N possessed sequences that were very similar to other filamentous fungal DHN genes. A genomic library of 387N was screened and we recovered a gene with an 877 nucleotide open reading frame that coded for a protein of 268 amino acids. The protein shared similarity with other fungal tri- or tetra-hydroxynaphthalene (THN) reductases and shared determinative characteristics of the family of short chain alcohol dehydrogenases. To prove the function of the gene that was isolated and characterized, the 387N THN reductase gene was used to complement THN reductase deficient buf mutants of Magnaporthe grisea. Several transformants were recovered that had wild type pigmentation. Hybridization studies and reverse transcriptase - polymerase chain reactions (RT-PCR) indicated that these transformants had integrated copies of the 387N gene and that gene transcripts were present. This evidence led us to conclude that the gene isolated from 387N functions as a THN reductase. n T A B L E O F C O N T E N T S Page Abstract ii Table of Contents iii List of Figures iv List of Tables vii Abbreviations viii Acknowledgements. xi Chapter 1 1 Pigment production and the melanin biosynthetic pathway in O. floccosum 387N Chapter 2 :. 34 Influence of nutrients on pigment production in O. floccosum 387N and characterization of the melanin biosynthesis pathway using DHN specific inhibitors Chapter 3 59 Isolation and characterization of a putative THN reductase gene from Ophiostoma floccosum 387N Chapter 4 104 Functional characterization of the THN reductase gene of O. floccosum 387N Chapter 5 148 Conclusions and future directions Literature Cited 154 Appendix 166 in L I S T O F F I G U R E S Figure 1.1. Canadian softwood lumber exports 1 Figure 1.2. Destinations of B.C. exported wood products 2 Figure 1.3. Western Canada isolates of sapstaining fungi 8 Figure 1.4. Pigmented fungal hyphae growing through softwood 11 Figure 1.5. Electron micrographs of 387N grown in culture and on lodgepole pine 15 Figure 1.6. DOPA melanin biosynthesis 18 Figure 1.7. Catechol melanin biosynthesis 19 Figure 1.8. GDHB melanin biosynthesis 20 Figure 1.9. DHN melanin biosynthesis 22 Figure 1.10. The chemical structures of DHN inhibitors 29 Figure 2.1. The L*, a*, b* colour space 38 Figure 2.2. Fungal biomass of 387N after grown in liquid culture 44 Figure 2.3. Dry weight and pigment intensity of 387N grown in liquid culture 46 Figure 2.4. The colony diameters of 387N treated with the DHN inhibitors 48 Figure 2.5. The colour of dried fungal mycelium when treated with DHN inhibitors 49 Figure 2.6. Hybridization of 387N with DHN genes from A. alternata 51 Figure 3.1. Amino acid alignment of several THN reductase genes 66 Figure 3.2. PCR products of genomic 387N and A. alternata DHN plasmids 68 Figure 3.3. PCR amplification of partial THN reductase gene 69 Figure 3.4. Transformants screened by Sphl/PstI digestion of plasmids 71 iv Figure 3.5. Nucleotide sequence of pGT2 insert 73 Figure 3.6. Nucleotide sequence of the pGT5 insert 74 Figure 3.7. Blots of 387N hybridized with either pGT2 or pGT5 ...76 Figure 3.8. Agarose gel separation of Mbolpartially digested genomic 387N DNA 77 Figure 3.9. Sucrose gradient fractions of Mbol partially digested genomic 387N DNA 78 Figure 3.10. PCR products of phage or genomic 81 Figure 3.11. Hindlll digested phage DNA from six positive phages 83 Figure 3.12. Hindlll digested phage DNA hybridized with a fragment of pGT2 84 Figure 3.13. The construction of the vector pBSA4.2 85 Figure 3.14. The location and orientation of the primers 88 Figure 3.15. Complete nucleotide sequence of 387N THN reductase gene 89 Figure 3.16. The organization of introns and exons in the reductase THNR genes 90 Figure 3.17. An alignment of the amino acid sequences of THNR 93 Figure 3.18. Unrooted neighbour joining consensus tree 94 Figure 3.19. Taxonomic organization of fungi 95 Figure 4.1. The construction of the disruption vector pAN7-A476 112 Figure 4.2. Predicted homologous integration event 114 Figure 4.3. PCR amplification products of putative disruptant DNA 115 Figure 4.4. Hybridization of the five putative disruptants; C5, B25, G27, G29 and G32 with a 250 bp portion of the THN reductase gene .' 117 Figure 4.5. Hybridization of PstIrestricted genomic DNA from disruptants with a 250 bp portion of the THN reductase gene 118 v Figure 4.6. Colony diameter of the five putative disruptants grown on MEA with and without hygromycin 121 Figure 4.7. The growth of the putative disruptants on wood as evaluated by ergosterol concentration measured by HPLC 122 Figure 4.8. The colour of dried fungal mycelium after 14 days of growth in NA supplemented with glycerol and asparagine 124 Figure 4.9. The construction of pCB1.9X and pAN1.9X 126 Figure 4.10. The colour of liquid cultures of Guy 11 and buf mutant IC4-1 and the transformats 4, 15, 16, and 17 130 Figure 4.11. The colour of liquid cultures of the control transformants 130 Figure 4.12. Hybridization of genomic 387N, IC4-1, IC4-5 and Guy 11 digested with either PstI or PvuII with a portion of the THN reductase gene 132 Figure 4.13. M. grisea buf mutants transformed with pCB1.9X or pAN1.9X and hybridized with a portion of the 387N THN reductase gene 133 Figure 4.14. PCR amplification with T22 and 20F primers using cDNA as template 136 Figure 4.15. Second round PCR amplification of 20F and 1301R primers using first round PCR products as template 137 Figure 4.16. Aatl/BstUI digestion 443 bp PCR product 138 Figure 4.17. The Aatlland BstUIdigestion schematic 138 vi L I S T O F T A B L E S Table 2.1. Visual assessment of fungal hyphal colour after growth in liquid 43 Table 3.1. List of oligonucleotides used in sequencing and as PCR primers 61 Table 3.2. Nucleotide sequence of degenerate oligonucleotides used as PCR primers 65 Table 3.3. Specific oligonucleotides for PCR screening of positive library clones 80 Table 3.4. Oligonucleotides primers to sequence pBSA4.2 88 Table 4.1. Summary of the colour of putative knockouts 119 Table 4.2. Putative transformants and their origin and colour in liquid culture 128 vn L I S T O F A B B R E V I A T I O N S amp ampicillin Arg arginine Asn asparagine bp base pair cAMP cyclic adenosine monophosphate cDNA copy deoxyribonucleic acid CTAB cetyltrimethylammonium bromide dH20 distilled water DNA deoxyribonucleic acid EDTA ethylene diamine tetra acetic acid DDAC didecyl dimethyl ammonium chloride DHN dihydroxynaphthalene DOPA dopaquinone or dopachrome GDHB glutaminyldihydroxybenzene GHB glutaminylhydroxybenzene HPLC high pressure liquid chromatography 2HJ 2-hydroxyjugalone 3HJ 3 -hydroxyjugalone HPH hygromycin phosphotransferase hyg hygromycin 4HR 4 hexyl resorcinol He or I isoleucine IR infrared kb kilobase kDa kilodalton K lysine nm nanometers NMR nuclear magnetic resonance Met methionine mm millimeters MS mass spectroscopy ORF open reading frame OD6oo optical density at 600 nm PCP polychlorophenates PCR polymerase chain reaction rpm revolutions per minute RT room temperature S serine SADH short chain alcohol dehydrogenase SDS sodium dodecyl sulphate SHAM salicylhydroxamin acid SSC sodium chloride/sodium citrate tet tetracycline T4HN 1,3,6,8-tetrahydroxynaphthalene ix TCMTB 2-(2-thiocyanomethylthio) benzothiozole TLC thin layer chromatography TBTC tributyltin chloride T3HN 1,3,8 -trihydroxynaphthalene Tris tris-(hydroxymethyl)-aminoethane Y tyrosine UV ultraviolet w/v weight by volume X A C K N O W L E D G E M E N T S I would like to thank the following people and organizations, in no particular order for their assistance and contributions. The B.C. Science Council for bestowing a GREAT Award that provided three years of welcomed financial support. Forintek Canada Corporation for their industrial contribution to this scholarship and specifically Tony Byrne and Dave Plackett for orchestrating this support. Takashi Tsuge for A. alternata DHN plasmids. Marc Orbach and Mark Farman for kindly sending several M. grisea mutants. Michael Wheeler for his advice regarding the extraction of intermediates and the gift of several standards. Adnan Uzunovic for the taxonomic characterizaion of 387N. Alex Serreqi for his biochemical advice and assistance with TLC, GC and HPLC. Seon Hwan Kim for his input and suggestions while reading my thesis. Louis Bernier and his lab for instruction in CHEF gel analysis. PAPRICAN for the use of their facilities. Colette Breuil for her constant support during this endeavour. Jim Kronstad for his sage advice and support and his students and postdocs for their willingness to listen. Thanks also to committee members Will Hintz and Jack Saddler for their input and suggestions. SFU and UVic co-op students Ron Janse and Sarah Riecken for their contributions to this project. And last but by no means the least I wish to thank my family and friends without whom my life would have been much poorer: Cyrla, Sandra, Linda, Paul, Monica, Jean, Farahad, Ron, Loralee, Don, Emi, Zarina, Serena, Claire, and of course Alvie. xi CHAPTER 1 Pigment production and the melanin biosynthetic pathway in O. floccosum 387N 1.1 Sapstain in Canada According to statistics from the lumber industry, Canada has roughly 17,800 million cubic meters of softwood forest, 45% of which is located within British Columbia (B.C.) (COFI, 1997). Canada is the world's prime exporter of softwood lumber products (Figure 1.1). In 1996, these exports contributed $34.2 billion to the Canadian trade balance (COFI, 1997). Offshore exports to the U.S and Japan account for roughly 80% of total B.C. coastal industry shipments (Figure 1.2). These markets pay premium prices for unseasoned, defect free wood. 1.1.1 Economic significance E.U 13% Scandinavia 20% British Columbia 31% 1 U S A " 5% Others 4% Former USSR 7% Rest of Canada 20% Figure 1.1. Canadian softwood lumber exports (COFI, 1997). 1 Other Figure 1.2. Destinations of B.C. exported wood products (COFI, 1997). 1.1.2 Sapstain Control Wood discolorations are unsightly and can lead to product devaluation and insurance claims. Currently in Canada, sapstain on lumber is being controlled by either kiln drying or anti-stain chemical treatments. During the summer months, when fungal spores are prevalent, modified logging practices, such as more rapid handling of logs and lumber accompanied by reduced storage times, can reduce the incidence of stain. Other practices such as the under-water storage of wood or sprinkling of log storage piles can also help reduce the occurrence of stain (Findlay, 1959). The Canadian lumber industry relies heavily on export markets and must spend millions of dollars annually to control the lumber stain problem. An ideal method for controlling stain should have the following attributes: a 6-12 month period of effectiveness, low environmental toxicity, specificity 2 towards sapstaining fungi and low corrosive action. The chemicals used must be non-leachable but have some degree of water solubility for easy application (Zabel and Morrell, 1993). Kiln Drying Kiln drying is a process that reduces the moisture content of the wood to less than 20%. The low moisture content is inhibitory to fungal growth since without water many fungal biochemical and enzymatic reactions are prevented (Zabel and Morrell, 1993). It is important to kiln dry the lumber immediately after processing prior to the visible appearance of any stain. Kiln drying is feasible where species and market factors permit. In British Columbia kiln drying is not always appropriate since some offshore markets demand green lumber or large dimension timber. During ocean freighter shipments to offshore markets accidental re-wetting of the lumber can occur. In 1996, 3.5 billion board feet of softwood lumber from BC was transported via water to world markets (COFI, 1997). Re-wetting may be accompanied by renewed fungal growth and associated staining. Large dimension lumber is not always suitable for drying since it may suffer drying defects. Kiln drying is an expensive procedure. The cost of kiln drying the entire volume of lumber exported to the E U alone would be fifty to one hundred times more expensive than'the current cost of anti-stain treatments (Smith, 1991). 3 A n t i - s t a i n t r e a t m e n t s Anti-stain chemicals inhibit fungal growth usually by the formation of a thin barrier that prevents the germination of fungal spores on the wood substrate (Zabel and Morrell, 1993). These chemicals are applied to the surface of lumber by dipping into tanks or by spraying. Due to the prevalence of sapstain in B.C., all offshore green lumber exports require treatment with anti-sapstain chemicals. In 1988, four billion board feet of softwood lumber, with an export value of over $2 billion, were treated with these chemicals (COFI, 1997). Historically, polychlorophenates (PCP) were highly efficacious in controlling sapstain but PCP's are environmentally hazardous and have been banned in Canada and many other countries (Smith, 1991). In Canada and the USA the following compounds are temporarily registered for use as pesticides to control sapstain: 2-(2-thiocyanomethylthio) benzothiozole (TCMTB), 3-iodo-2-propynyl-butyl-carbamate (IPBC), sodium borate, copper-8-quinolinolate and quaternary ammonium compounds such as didecyl dimethyl ammonium chloride (DDAC). None of these formulations have proven as effective for controlling stain as PCP has in the past. There are other problems associated with the use of these alternative anti-stain treatments such as: increased costs, higher required effective concentrations, unwanted discolouration of wood, corrosion of equipment, worker sensitivity, cold weather handling complications and high aquatic toxicity (Seifert, 1993). There is also a possibility that the temporary status of these compounds as acceptable pesticides could be subject to review. Unfortunately, many anti-stain 4 treatments have a broad spectrum of action against organisms other than wood staining fungi. There is an urgent need for a new class of anti-fungal agents that demonstrate low environmental toxicity, known environmental fate, and high target specificity. Potential alternative treatments Biocontrol methods could serve as an alternative approach to controlling sapstain. Biocontrol strategies involve exploiting our knowledge of a specific fungus and how it utilizes wood in order to somehow block growth or stain production. Many biochemical and enzymatic pathways could be targeted by this approach. Since sapstaining fungi cause minimal structural damage to timber, a possible method of control could be to simply block the synthesis of the pigment responsible for discoloration. Colourless mutants of O. piliferum have been shown to be as equally robust as wild type strains at quickly colonizing wood and preventing secondary wood rot infections (Blanchette et al., 1992). The pathway of pigment production and the genes involved in pigment biosynthesis will be explored with this future application in mind. 1.2 Sapstaining Fungi 1.2.1 Types of staining fungi Freshly felled wood and lumber, stored while awaiting processing or exportation are susceptible to microbial infection. Sapstain fungi can colonize and produce permanent blue to black stains in sapwood known as blue stain or sapstain. Other types of wood 5 stains can occur but are thought to be primarily oxidative in nature and commonly appear following drying or sawing (Zabel and Morrell, 1993). Wood contact with iron compounds can also cause stains which are usually superficial (Findlay, 1959). In general there are three groups of sapstain fungi. The first group is comprised of fungi and their anamorpbs' belonging to the following three ascomyceteous genera: Ophiostoma, Ceratocystis and Ceratocystiopsis. This group will be the focus of all following discussions. The second group are commonly known as black yeasts and include fungi such as Hormonema dematioides, Aureobasidium pullulans and Rhinocladiella atrovirens. The third group are molds that produce masses of conidia on the wood surface and can also cause deeper stain. Alternaria alternata, Cladosporium sphaerospermum and C. cladosporioides are examples of this third group (Seifert, 1993). Some molds such as Penicillium sp. and Trichoderma sp. can cause superficial discoloration on the surface of wood, primarily due to pigmented spores, this can be removed by planing or sanding (Zabel and Morrell, 1993). From an economic perspective, only the first category of saprobic fungi are of interest because the fungi can colonize the wood anytime during the lumber production process while heavily stained wood entering the mill can be easily discarded prior to processing. Forintek Canada Corporation conducted a Canada wide survey of sapstain fungi in 1991 (Seifert and Grylls, 1991). Untreated, fresh samples of several softwood species showing visible stain were randomly collected from mills across the country. The most frequently isolated sapstaining fungi, from Eastern and Western Canada were members of Ophiostomatales and their various anamorphs (Figure 1.3). Even when one species of 6 this family is isolated individuals can have a broad range of genotypes and colony morphology. In order to more precisely identify the populations of Ophiostomatales the 1991 survey was expanded upon in 1997-98. The purpose of the expanded study was a more systematic geographic collection, isolation and characterization of sapstain species and the study set out to determine if there were differences in the fungal colonization of the log versus sawn lumber and if there were differences between tree species colonized. The results from this study support the distribution and prevalence patterns of the previous study and verify the position of Ophiostomatales as the primary causative sapstaining organisms in British Columbia (Uzunovi et al., 1999). 7 Black yeasts Green Moulds 32% Figure 1.3. Western Canada isolates of sapstaining fungi (Seifert and Grylls, 1991). 1.2.2 Biology This section will focus on the first group of staining fungi, the Pyrenomycetes -Ophiostoma, Ceratocystis and Ceratocystiopsis and their anamorphs. In general, sapstaining fungi are thought to be the primary fungal colonizers of wood since they can tolerate high moisture levels (Seifert, 1993; Subramanian, 1983; Zabel and Morrell, 1993). The slimy wet conidia of the synnemata are well suited to the wet Pacific 8 Northwest and can be dispersed by mist laden air or splash droplets of rain (Dowding, 1969). Ascospores, found in perithecia, may also be disseminated by splash droplets of rain (Dowding, 1969). Our knowledge of the biology of O. floccosum is not extensive and any associations with insect vectors have not been specifically characterized. O. floccosum is not often found in living wood but like other species of Ophiostoma, may be introduced into diseased trees and newly felled wood by bark beetles or mites (Bridges and Moser, 1983; Bridges and Moser, 1986; Gibbs, 1993). Bark beetles have been closely associated with outbreaks of O. polonicum and O. penicillatum on Norway spruce and many other species of Ophiostoma have also been found in association with bark beetles (Bridges and Moser, 1986; Gibbs, 1993; Paine et al, 1997). Some bark beetles have very specialized structures called mycangium that can be used to transport fungal spores. It is thought that the fungi help to reduce the moisture content of the wood making it more suitable for bark beetle reproduction (Bridges and Moser, 1983). Mites have been shown to transport ascospores of C. minor from the phloem and xylem to the outer bark, increasing opportunities for beetle contact and broader dispersion (Bridges and Moser 1983). Further research needs to be conducted to elucidate the biology of this important sapstaining organism. Analysis of the life cycle of 387N could reveal potential techniques to control sapstain. 9 1.2.3 Growth on Wood Once on the wood, the spores of sapstaining fungi germinate and the fungal hyphae grow radially through the ray parenchyma cells of the sapwood (Figure 1.4) (Ballard and Walsh, 1982). The rays function as nutrient reservoirs for the tree and provide the growing fungal hyphae with easily assimilable substrates (Subramanian, 1983). Most softwood species contain 0.02-1% (wood dry weight) soluble sugars, 0.01-0.3%) nitrogen and 1.5-10% lipids (Gao et al, 1993; Merrill and Cowling, 1966; Terziev et al, 1993). Sapstaining fungi can utilize most of the soluble sugars and nitrogen in the form of protein or peptide. Approximately 60% of the lodgepole pine lipids are triglycerides. O. floccosum 387N can hydrolyze triglycerides into free fatty acids and glycerol. Both of these compounds can be used as carbon sources (Gao and Breuil, 1995b; Gao et al, 1993). More research regarding the biochemistry and metabolism of 387N would allow one to determine specifically what is required for growth and pigmentation on wood. Sapstaining fungi are unable to utilize the structural components of the wood cell walls, such as cellulose, hemicellulose and lignin but can be found in the tracheids and resin canals. It is thought that the fungi move from tracheid to tracheid via bordered pits or direct mechanical penetration of the cell wall using appressoria (Wingfield et al, 1993). As the hyphae within the ray parenchyma age, they become pigmented (Ballard and Walsh, 1982). On a cross section of a log a characteristic wedge shaped stain appears as fungal hyphae in the rays pigments (Gibbs, 1993). Pigmented hyphae and conidiophores 10 contribute to the appearance of stain. In many sapstain fungi, and plant pathogens, this pigment is melanin (Wheeler, 1983; Zink and Fengel, 1988). Figure 1.4. Pigmented fungal hyphae (F) growing through the ray parenchyma (P) and tracheids (T) of softwood (photo courtesy of C. Breuil). 11 1.3 Funga l pigment Melanins are derived from the products of secondary metabolism. Secondary metabolites comprise a vast number of compounds, with diverse chemical structures and often as yet unknown functions (O'Hagan, 1991). They are usually produced after the active growth of the organism has ceased and many are products of the polyketide pathway (Bhatnagar et al, 1992). In fungi and plants, secondary metabolites tend to be derived from acetate and are often aromatic and bioactive (Bennett, 1995). 1.3.1 Structure Melanins are structurally heterogeneous polymers, generally described as black pigments of high molecular weight formed by. oxidative polymerization of phenolic monomers (Wheeler and Bell, 1988). These dark pigments are found in many plants, animals and microorganisms and are often associated with carbohydrates, proteins and lipids (Zink and Fengel, 1988). The insolubility of melanin makes it difficult to characterize biochemically (Seraglia et al, 1993). Melanins are found in many fungi and can be synthesized by a variety of metabolic pathways. Ultraviolet (UV), visible, infrared (IR), and nuclear magnetic resonance (NMR) spectrographic data indicate that the pigment produced by O. floccosum 387N is melanin and resembles the DHN melanin of other ascomycetes (Brisson et al, 1996). A closely related species, O. piliferum has been shown to synthesize melanin via the dihydroxynaphthalene (DHN) pathway (Zimmerman etal, 1995). 12 1.3.2 Function The purported functions of fungal melanins are diverse. In many cases melanin is not essential for fungal growth but can influence survival and longevity. Melanin has been reported to protect fungal hyphae and spores from UV light and desiccation (Bloomfield and Alexander, 1967; Wheeler, 1983). Melanin can contribute to the biosorption of metal ions and other toxic compounds like tributyltin chloride (TBTC). Melanin is thought to sequester the metal or toxic substance, which would remove the substance from contact with the actively growing segment of the organism (Fogarty and Tobin, 1996). Melanin can be produced in response to wounding and/or stress and may enable fungi to withstand predation by mites or resist lytic enzymes (Braiser, 1978; Gibbs, 1993). In Aspergillus fumigatus, a human pathogen, melanin has been shown to be important in protecting the conidia from detection by the host's immune system (Tsai et al, 1998). An important function of melanin in many phytopathogenic fungi such as Magnaporthe grisea, Pyricularia oryzae, Colletotrichum lagenarium and Cochliobolus miyabeanus is the mechanical strengthening of the cell walls in appressoria (Kubo et al, 1987; Kubo et al, 1982a; Kubo et al, 1982b; Kubo et al, 1985; Money and Howard, 1996; Woloshuk et al, 1980). Appressorial melanin is thought to permit the build up of turgor pressure necessary for the appressoria to penetrate the host tissue (Chumley and Valent, 1990). Alternaria alternata is a phytopathogenic fungus that causes disease by means of the production of host-specific toxins (Kusaba and Tsuge, 1994). In this fungus melanin is located in cell walls of hyphae and conidia and may be important to the longevity of these structures (Kawamura et al, 1997). The role of melanin in O. floccosum 387N has not 13 yet been defined. Transmission electron micrographs indicate that electron dense melanin granules are located in the cell walls and in the extracellular matrix of 387N when grown on wood or in liquid culture (Figure 1.5). These granules are similar to the electron dense granules that have been documented in other ascomycetes (Brisson et al., 1996; Gadd, 1980). In other species of Ophiostoma melanin may play a role in the mechanical strengthening of modified appressoria used to penetrate wood cell walls (Ballard and Walsh, 1982; Gibbs, 1993). Alternatively melanin may play a structural role in the development of the perithecia since it has been shown that albino O. piliferum mutants are unable to produce mature perithecia unless supplemented with the melanin intermediate scytalone (Zimmerman et al, 1995). 14 Figure 1.5. Electron micrographs of 387N grown for 10 days in (a) liquid culture with asparagine and glycerol and on (b) lodgepole pine. Arrows indicate electron dense melanin granules associated with the cell wall and the extracellular matrix. Magnification: (a) 10 OOOx (b) 16 OOOx (photos courtesy of S. Gharibian). 15 1.3.3 Biosynthesis Traditionally, the pathways of melanin biosynthesis in fungi have been elucidated by isolating strains with pigment deficiencies (Bell et al, 1976b; Chumley and Valent, 1990; Faure et al, 1994; Kawamura et al, 1997; Kimura and Tsuge, 1993; Kubo et al, 1991; Kubo et al, 1983; Kubo et al, 1989; Wheeler and Stipanovic, 1985). Pigment deficient strains accumulate intermediates from different points in a given pathway. Intermediates have been extracted from fungal cultures and have been characterized by thin layer chromatography (TLC), UV, IR, mass spectroscopy (MS) and NMR (Bell et al, 1976a; Kubo et al, 1983; Stipanovic and Bell, 1977; Wheeler and Stipanovic, 1979). The information obtained by these analyses can indicate the structure of the intermediates. Another technique used to order the intermediates within the pathway are cross feeding experiments (Woloshuk et al, 1980). Cross feeding involves the application of an intermediate from one pigment deficient strain to another. The restoration of pigmentation can indicate the order of intermediates in the pathway. Diploid formation of selected pigment deficient strains and any resulting complementation is another technique that has been used to order the intermediates in a pathway (Bell et al, 1976b). The collection of data from these types of experiments has allowed the elucidation of several different melanin biosynthetic pathways from many different fungi. In fungi there are four major pathways of melanin biosynthesis. The glutaminyl dihydroxybenzene (GDHB), dopaquinohe or dopachrome (DOPA) and catechol pathways derive their precursors from the shikimic acid pathway while the DHN pathway derives 16 precursors from the polyketide pathway. The shikimic acid pathway is responsible for the synthesis of aromatic amino acids from aliphatic precursors. Although derived from different precursors the melanins share similar physical and chemical properties. The final structure of melanin depends on the composition of the precursor subunits, the degree of polymerization and association with proteins and carbohydrates. DOPA The DOPA pathway has been shown to occur in a wide range of organisms. DOPA melanins contain nitrogen and are derived from tyrosine, a product of the shikimate pathway (Figure 1.6). DOPA melanin has been characterized by IR, high pressure liquid chromatography (HPLC) and MS (Hoti and Balaraman, 1993). DOPA melanin is thought to be a heteropolymer formed from the different tyrosine derivatives in the pathway (Bell and Wheeler, 1986). Mammals, plants, mushrooms, yeast and bacteria, all use different derivations of the reduction and dehydration of tyrosine for DOPA melanin production (Bouchard et al, 1994; Chaskes and Tyndall, 1979; Luo et al, 1994; Yokoyama et al, 1994). For example, the Raper Mason pathway in mammals is much more complex than that shown in Figure 1.6 and involves a series of enzymatic reactions with either a proximal or distal phase that is responsible for eumelanin or phaeomelanin production (Sanchez-Ferrer et al, 1995). In fungi a tyrosinase converts tyrosine to DOPA melanin and has been purified from several different species. Tyrosinase has a copper active site and can catalyze either the hydroxylation of monophenols or the oxidation of o-diphenols using 17 molecular oxygen (Sanchez-Ferrer et al, 1995). Tyrosinase inhibitors such as 4-hexylresorcinol (4HR), salicylhydroxamin acid (SHAM), and cetyltrimethylammonium bromide (CTAB) can be applied to fungal cultures to determine if this pathway is being utilized by the fungus (Dawley and Flurkey, 1993). C O O H Tyrosine Melanochrome Indole-5,6-quinone Figure 1.6. DOPA melanin biosynthesis (Wheeler and Bell, 1988). 18 Catechol Originally, catechol melanin was isolated from Ustilago maydis (Wheeler and Bell, 1988). Very little is known about this pathway but it has been studied in several fungi and plants (Gafoor and Heale, 1971; Takahashi and Akiyama, 1993). Catechol melanin lacks nitrogen but, like DOPA, uses precursors from the shikimate pathway (Figure 1.7). Enzymatic oxidation of catechol may be responsible for the production of this type of melanin (Wheeler and Bell, 1988). Nonindolic catechol melanin has been characterized by UV absorbance, IR spectroscopy, X-ray diffraction and chromatography (Takahashi and Akiyama, 1993). Figure 1.7. Catechol melanin biosynthesis (Wheeler and Bell, 1988). 19 GDHB GDHB melanin has been isolated from the basidiospores of the genus Agaricus (Figure 1.8). This melanin contains nitrogen and is synthesized from the precursor, y-Glutaminyl-3,4-hydroxybenzene (GHB) (Wheeler and Bell, 1988). Basidiomycotina commonly utilize GDHB melanin biosynthesis. Fungi of the Agaricus genus produce tyrosinases (with peroxidase activity) which convert GHB to GDHB (Bell and Wheeler, 1986; Wheeler and Bell, 1988). OH OH O y-glutamyltransferase Peroxidase/Phenolase Figure 1.8. GDHB melanin biosynthesis (Wheeler and Bell, 1988). 20 D H N In 1983, a survey of 26 species of fungi encompassing 16 genera, concluded that all of the ascomycetes and fungi imperfecti, with the exception of Aspergillus niger, produced melanin via the DHN pathway (Wheeler, 1983). The DHN pathway of melanin biosynthesis has been extensively characterized in filamentous fungi. DHN melanin biosynthesis begins with acetate, a fundamental product of primary metabolism (Figure 1.9). Acetate is obtained from the carboxylation of pyruvate or phosphoenolpyruvate during glycolysis. Two carbons from each acetate are cyclically joined into a linear chain by polyketide synthase (Juzlova et al, 1996). DHN melanin biosynthesis is initiated with the cyclization of this linear chain into a pentaketide that undergoes a condensation by pentaketide synthase (PKS) to form 1,3,6,8-tetrahydroxynaphthalene (T4HN). This intermediate is reduced by a reductase to generate scytalone that is dehydrated to form 1,3,8-trihydroxynaphthalene (T3HN). T3HN is reduced by a reductase to produce vermelone and finally dehydrated to form 1,8-DHN which is thought to be polymerized into melanin by a laccase (Bell and Wheeler, 1986). Other melanin biosynthetic pathway derivatives such as DOPA and catechol have been tested for their ability to restore pigmentation in albino cultures of the DHN melanin producers, T. basicola and V. dahliae (Wheeler and Stipanovic, 1979). Microscopic examination of hyphal cell walls and chlamydospores of pigment restored albinos revealed that the melanin was unlike the natural wild type melanin in appearance being lighter in colour and did not have an electron dense appearance. The restored pigment 21 may not be the true melanin found in wild type DHN melanin producing fungal cell walls (Bell and Wheeler, 1986). H 3 C C o A T Acetyl C o A Pentaketide Synthase Pentaketide 3 Synthase C O O H Pentaketide O H O H O H O Scytalone Dehydratase O H O H HO 1,3,6,8-THN Scytalone 1,3,8-TH O OH OH O H Reductase ^ ^ Dehydratase HO M E L A N I N Vermelone DHN Figure 1.9. DHN melanin biosynthesis (Bell et al, 1976a; Bell et al., 1976b; Stipanovic and Bell, 1976; Stipanovic and Bell, 1977). 1.3.4. Molecular biology of DHN melanin biosynthesis In many filamentous fungi three types of pigmentation deficient mutants are typically isolated that correspond to deficiencies in three of the genes involved in DHN melanin biosynthesis (Bell et al., 1976a; Chumley and Valent, 1990; Kimura and Tsuge, 1993; Kubo et al, 1983; Kubo et al, 1989; Lundqvist et al, 1993; Tajima et al, 1989; Wheeler and Stipanovic, 1979). 22 Pentaketide synthase The first pigmentation mutant has an albino phenotype that can be restored to wild type with the addition of scytalone or vermelone (Bell et al, 1976a). This result indicates a blockage upstream of the position in the pathway of these two intermediates. The albino phenotype is reported to be due to a deficiency in the PKS gene. This gene was first characterized in Colletotrichum lagenarium and then shortly afterwards in Alternaria alternata when genomic DNA sequences were found to restore pigmentation when transformed into albino mutants (Kimura and Tsuge, 1993; Kubo et al, 1991). The PKS1 gene from C. lagenarium was sequenced and a single (ORF) was found to code for a protein of 2187 amino acids. PKS1 shared homology with p-ketoacyl synthase, acetyl/malonyl transferase and two acyl carrier protein domains of other polyketide synthases (Takano et al, 1995). Dehydratase Unlike the albino mutants, mutants deficient in scytalone dehydratase (SD) can display a range of phenotypes depending on the organism studied. In V. dahliae SD deficient mutants are rosy coloured and accumulate scytalone. When these mutants were supplied with T3HN, wild type pigmentation was restored (Bell et al, 1976a; Bell et al, 1976b; Wheeler et al, 1976). The SD gene from C. lagenarium called SCD1 has an ORF which codes for a protein of 188 amino acids and a 0.7 kb RNA transcript (Kubo et al, 1996). The amino acid sequence shared similarity with the 23 kDa polypeptide isolated from Cochliobolus miyabeanus (Tajima et al, 1989). SDH1 from P. oryzae coded for a 23 polypeptide of 172 residues with a calculated mass of 20 kDa and had a RNA transcript of 1 kb (Motoyama et al, 1998). The amino acid sequence shared 62% identity with SCD1 protein. Crystallographic studies of the M. grisea SD polypeptide show that it is a 69 kDa trimer formed from identical monomers of 172 residues (Lundqvist et al, 1994). An early report suggested that scytalone dehydratase is responsible for the dehydration of both scytalone and vermelone (Butler et al, 1988). However, the A. alternata SD was unable to complement a SD deficient mutant of M. grisea (Kawamura et al, 1997). This suggests that perhaps two SD genes are required for dehydration reactions of DHN melanin biosynthesis in A. alternata. Reductase A third type of V. dahliae mutant was buff or light brown in colour and accumulated 2-hydroxyjugalone (2HJ), the auto-oxidation product of T3HN. This mutant was found to be deficient in the THN reductase and wild type pigmentation could be restored by the addition of vermelone. The M. grisea reductase gene (T4HN) was first characterized by (Vidal-Cros et al, 1994). The gene was isolated using a reverse genetics approach. Initially, the formation of scytalone was used to monitor the enzymatic activity of fractions during protein purification. The protein was used for antiserum production and the antibodies were used to screen a cDNA library to select for the clones that expressed the reductase. The cDNA 24 was 846 nucleotides and coded for a 282 amino acid peptide with a predicted size of 29.9 kD. On a native SDS-PAGE gel the reductase appears as a tetramer of identical subunits (Vidal-Cros et al, 1994). Both T3HN and T4HN can be utilized as substrates by this reductase which suggests that only one reductase is necessary for the biosynthesis of DHN melanin in this organism. The T4HN protein sequence shares 56% identity with the Ver-1 gene of Aspergillus parasiticus and additional identity with other proteins which are members of a short chain alcohol dehydrogenase (SADH) family (Persson et al, 1991). This protein also displays the strictly conserved motif of SX 1 0 . 1 3 YX 3 K, which has been proposed to be an important element within the active site of the SADH family. SADH share several characteristics such as, a subunit size of -250 amino acids, the domain order of coenzyme binding domain followed by a catalytic domain with a highly conserved tyrosine reactive residue at the active site (Persson et al, 1991). Crystallographic studies were conducted on a M grisea reductase (THNR) protein isolated by another researcher (Andersson et al, 1996). It was shown that this reductase was also a member of the SADH family and confirmed the previous prediction of a homotetrameric structure (Vidal-Cros et al, 1994). An arginine residue at position 39 confers NADPH specificity and the active site of the THNR polypeptide lies within a large pocket which can be closed by a flexible fold. Tricyclazole, a DHN melanin biosynthesis inhibitor, binds competitively with the naphthol substrate at the active site of the enzyme. The conserved triad of Ser-Tyr-Lys at the active site is responsible for the transfer of the hydrogen of the NADPH to the polyphenol substrate (Andersson et al, 1997b). 25 A THN reductase gene, THR1, has also been characterized in C. lagenarium (Perpetua et al, 1996). THR1 was isolated by screening a genomic library of C. lagenarium with the heterologous gene BRM2, from A. alternata (Kimura and Tsuge, 1993). A melanin deficient mutant of C. lagenarium was transformed with the THR1 gene and melanin production and pathogenicity were restored. The nucleotide sequence of THR1 reveals one ORF of approximately 1300 nucleotides coding for a 282 amino acid peptide. This sequence has four introns. THR1 shares 56% similarity with the Ver-1 gene of A. parasiticus and 83% identity with the M. grisea T4HN reductase gene. Expression studies that show that the THR1 gene has a 1.0 kb transcript which can be detected within one hour of spore incubation. In A. alternata the BRM2 gene is thought to code for a THN reductase. This gene has not been sequenced but the role and expression of the gene has been examined (Kimura and Tsuge, 1993). The BRM2 gene was used to restore pigmentation to melanin deficient mutants by complementation and transformation mediated disruption of pigmentation was also demonstrated (Kimura and Tsuge, 1993). The size of the mRNA transcript of the BRM2 gene is 0.9 kb and it was isolated from mycelia after four days of growth (Kimura and Tsuge, 1993). This transcript appears synchronously with mycelial melanization. Similar to the situation that exists for SD, it is not clear whether one or two reductases are involved in DHN melanin biosynthesis. The reductase purified from M. grisea was able to reduce the T4HN to scytalone and T3HN to vermelone (Vidal-Cros et al, 1994). However, the reductase had different affinities for the substrates. Alternatively, only 26 oxidative products of T3HN could be detected in mutants deficient in the THR1 reductase gene of C. lagenarium (Perpetua et al, 1996). In addition these researchers were never able to isolate mutants defective in the conversion of T4HN to scytalone. In V. dahliae a pigmentation deficient mutant was isolated that was deficient in the reduction of both napthols but another mutant was found to be deficient only in the reduction of T3HN to vermelone (Bell and Wheeler, 1986). The authors suggest that these differences only reflect differences in substrate affinity. At this point, based on published findings we cannot state equivocally whether one or two reductases are involved in DHN melanin biosynthesis. Polymerase One last type of pigmentation deficient mutant, infrequency isolated from V. dahliae, was found to accumulate monomers and 2'2'-dimers of DHN indicating a deficiency in the last step of DHN polymerization to melanin (Bell and Wheeler, 1986). A comparable mutant has been isolated from Cochliobolus heterostrophus and it was found to be deficient in p-diphenol oxidase (laccase) activity (Tanaka et al, 1992). However it is premature to draw conclusions regarding the contributions of laccases to DHN melanin biosynthesis since they are characteristically non-specific with respect to their reducing substrate and can have a wide assortment of functions in fungal physiology (Thurston, 1994). 27 1.3.5 D H N Inhibitors Several compounds are known to inhibit the DHN melanin biosynthetic pathway (Figure 1.10). Cerulenin, an antibiotic isolated from Cephalosporium caerulens, causes a reduction in the synthesis of fatty acids and sterols in yeast (Matsumae et al, 1964; Ohno et al, 1974; Ohno et al, 1975; Sano et al, 1967). Cerulenin binds irreversibly to P-ketoacyl ACP synthetase which catalyzes the malonyl-ACP:acyl-ACP condensation reaction in fatty acid synthesis (Giuliano et al, 1973). Since PKS shares homology with these enzymes it was not surprising that cerulenin was observed to inhibit the condensation of acetate and malonate, during pentaketide synthesis (Ohno et al, 1975; Omura, 1976). Cerulenin has since been used to investigate the synthesis of other polyketides such as aflatoxins and antibiotics (Hiltunen and Soderhall, 1992). 28 Carpropamid Figure 1.10. The chemical structures of the DHN inhibitors; tricyclazole, pyroquilion, cerulenin and carpropamid (Tsuji et al, 1997; Wheeler and Bell, 1988; Woloshuk^a/., 1981). Tricyclazole is another compound that inhibits DHN melanin biosynthesis. It was first tested as a fungicide for the treatment of rice blast disease caused by Pyricularia oryzae (an anamorph of Magnaporthe grisea) (Froyd et al, 1976). The mechanism of action of tricyclazole in disease • control was not elucidated until it was noticed that even low concentrations of tricyclazole suppressed normal pigmentation in Magnaporthe and other fungi (Tokousbalides and Sisler, 1979). Many pathogenic fungi rely on melanin for the mechanical strengthening of appressoria that are necessary for the penetration of host 29 plant cuticle. It was found that tricyclazole treated cultures of V. dahlia accumulated the DHN intermediate oxidation product 2HJ and were therefore comparable to a THN reductase deficient mutant (Bell et al., 1976b). At higher concentrations of tricyclazole flaviolin, an oxidative, product of T4HN, was detected in V. dahlia culture extracts. This suggested that at lower concentrations, tricyclazole was acting primarily on the conversion of T3HN to vermelone and at higher concentrations on the reduction of T4HN to scytalone. Although different in structure, pyroquilion was found to have similar biological activity as tricyclazole (Woloshuk et al., 1981). Once it was established that tricyclazole specifically inhibited the THN reductase it was used to investigate melanin biosynthetic pathways of other fungi and the role of melanin in appressorial development (Kubo et al, 1984; Kubo et al, 1982; Wheeler, 1983; Wheeler and Stipanovic, 1979; Woloshuk et al, 1980; Woloshuk et al, 1981). Tricyclazole and pyroquilion have also been found to block the synthesis of fungal pigments and aflatoxins that are products of polyketide biosynthesis (Wheeler and Bhatnagar, 1995; Wheeler and Klich, 1995). A recently reported fungicide, carpropamid, has been shown to inhibit melanin biosynthesis and to cause the accumulation of scytalone in Magnaporthe grisea and Colletotrichum lagenarium (Tsuji et al, 1997). Increasing concentrations of carpropamid causes a decrease in the activity of purified scytalone dehydratase from C. lagenarium (Tsuji et al, 1997). Kinetic studies and the co-crystallization of carpropamid with the Pyricularia oryzae scytalone dehydratase have shown that tight binding occurs between the inhibitor and the enzyme that effectively occupies the substrate binding site 30 (Nakasako et al, 1998). The use of DHN inhibitors has become widely accepted as a means to determine if melanin is involved in pigmentation (Butler and Day, 1998). 1.4 Research objectives In order to implement some method of biocontrol, detailed information regarding the mechanisms of growth and pigmentation of the sapstain fungi is required. However, specific knowledge concerning the biology, lifecycle, biochemistry, metabolism, genetics and molecular biology of O. floccosum 387N is scarce. The traditional biochemical approach of defining the pigmentation pathway, by characterizing the intermediates that accumulate in pigment deficient mutants, historically has proven to be essential to elucidating the pathways involved. More modern biochemical techniques such as protein purification allow one to isolate and characterize the enzymes in the pigmentation pathway. Some preliminary research has been done to identify potential targets for biocontrol in 387N but this fungus is still virtually uncharacterized. Several fungal enzymes, which play key roles in the utilization of the nutritional resources found in wood, were identified and characterized using standard protein biochemical methods (Abraham et al, 1993; Gao and Breuil, 1995a; Gao and Breuil, 1995b; Gao et al, 1993). An alternative target for the biocontrol of sapstaining fungi could involve blocking a step in the pathway of pigment production. This requires knowledge of the structure and function of the enzymes and regulation of their production and expression, something that can only be obtained from a molecular biological study. The techniques required for these explorations are readily available and have been 31 applied to other filamentous fungi (Appleyard et al, 1995; Baltz and Hosted, 1996; Bennett, 1995). In fact new techniques for improved ease of manipulation are constantly being developed (deGroot et al, 1998). Published sequences from other fungi are available for several of the DHN genes. These can serve as starting blocks to identifying comparable genes from sapstaining fungi. Basic knowledge can be obtained and in the future manipulated to produce organisms with reduced staining ability or even the inability to regulate or initiate production of pigment. This type of information would be invaluable for implementing many strategies of biocontrol. The overall goal of this research was to characterize melanin biosynthesis in O. floccosum 387N. Preliminary background research had to be conducted. No one has studied the production of pigment in 387N and it was important to determine what environmental factors affected pigmentation and specifically which biosynthetic pathway of pigment production was in use by 387N. The second chapter will examine studies involving the optimization and quantification of the pigment produced by 387N. Also discussed in this chapter will be DHN melanin inhibitor experiments designed to determine which pathway of melanin biosynthesis was being utilized by 387N. Hybridization studies were conducted to determine if genes are present in 387N that are similar to other fungal DHN melanin biosynthetic genes. Once these studies were completed the focus shifted to attempting to characterize a gene in the melanin biosynthetic pathway using molecular biological techniques. Chapter three will summarize the isolation and characterization of a melanin biosynthetic gene of 32 O. floccosum 387N. A genomic library was constructed, screened and the sequences retrieved were characterized by sequence analysis. The gene recovered was similar to other fungal THN reductases. In chapter four we present a discussion concerning the function and role of the THN reductase gene. The function of the THN reductase gene was tested by gene disruption. Results indicated that site specific integration was not occurring in 387N and disruptants were not recovered. As an alternative technique, the function of the gene was elucidated by intergenic complementation using the 387N THN reductase sequence to complement melanin deficient mutants of Magnaporthe grisea. Transformants were recovered that displayed wild type pigmentation and it was shown that the 387N THN reductase gene and transcripts of the gene were present. All major conclusions and their implications as well as ideas for future explorations that could contribute towards the elucidation of melanin biosynthesis in O. floccosum 387N are addressed in chapter five. 33 C H A P T E R 2 Influence of nutrients on pigment product ion in O. floccosum 387N and characterizat ion of the melanin biosynthetic pathway using DHN specific inhibitors 2.1 Introduct ion The primary goal of the work presented in this thesis was to study pigmentation in O. floccosum 387N. This chapter describes some experiments that were completed before the molecular work could be undertaken. Melanin production in 387N has not been extensively investigated and it was important to determine what factors influence its production. Specifically, the first objective of this project was to define culture conditions that maximize pigmentation. This information was then used to determine quantitatively whether pigmentation was being affected in subsequent manipulations. The second objective of this chapter was to determine if 387N utilized the DHN melanin biosynthetic pathway. The melanin from 387N had been previously characterized as lacking nitrogen (Brisson et al., 1996). Two methods were used to test this hypothesis. First, several known DHN inhibitors were added to fungal cultures to determine if they affected pigmentation. Secondly, hybridization studies with heterologous DHN genes from Alternaria alternata combined with O. floccosum 387N genomic DNA were used to determine if sequences similar to the DHN genes of yi. alternata were present in 387N. The growth of fungi as well as the production of secondary metabolic products, such as melanin, can be influenced by a wide variety of environmental factors such as pH, 34 temperature, light, humidity and available nutrients, particularly carbon, nitrogen and phosphorus (Hofsten and Hofsten, 1958; Hofsten, 1956; Zabel and Morrell, 1993). UV light can influence melanogenesis in Verticillum albo-atrum and V. dahliae and visible light has a dramatic influence on pigmentation in Microdochium bolleyi (Gadd, 1982; Tan, 1978). Nitrogen source and pH have been shown to affect pigment production in Monascus purpureus (Chen and Johns, 1993). In 1960, Kaarik documented the highly variable effects of 27 different carbon sources on mycelial colour in Ophiostoma sp. (Kaarik, 1960). Factors such as temperature, moisture content, and pH have been shown to influence the rate of sapstain development on wood (Gibbs, 1993). We tested whether different combinations of carbon and nitrogen sources had any affect on pigmentation and fungal biomass production. 2.2 Materials and methods 2.2.1 Fungal strains Strain 387N was originally identified as Ophiostoma piceae (Munch) H. & P. Sydow based on colony morphology. This strain was obtained from Forintek Canada Corporation (Ste. Foy, Quebec, Canada) and was isolated from softwood chips at the MacLaren Mill (Mason, Quebec, Canada). 387N was chosen because it darkly stained the sapwood of lodgepole pine. However, recent mating type data indicates that 387N is not O. piceae because it has been shown to mate with O. floccosum (Uzunovi et al, 1999). 387N is very similar to O. piceae in morphology and has always been isolated in close physical association with O. piceae from softwood sapwood. 35 An albino strain of O. piliferum was used as a control in pigment inhibition studies. This strain was a gift from Repligen Sandoz Research Corporation. 2.2.2 M e d i a for growth of fungi Fungal cultures were maintained in malt extract (ME) liquid or on agar (MEA) (2% malt extract, 1.5% agar) (Difco, Detroit, USA). For the nutrition studies, 387N was grown in a carbon and nitrogen solution adapted from Kaarik (1960). This media was supplemented with the micronutrients of Vogel (1956) and filter-sterilized vitamins of Montenecourt and Eveleigh (1977). The compounds used to provide a source of carbon were: rhamnose, glycerol, ethanol, starch and raffinose. Those were added to media to a final concentration of 2%. The compounds used to provide a source of nitrogen were: L-asparagine, L-tyrosine, NH4C1, NH 4OH and pumpkin globulins. Those were added at a final concentration of 0.02%) nitrogen. Liquid cultures were grown in 50 mL of media in 250 mL Erlenmeyer flasks, in the dark at 22°C on a rotary shaker at 250 rpm. Agar plates were incubated at 22°C in the dark. 2.2.3 Fungal Biomass Estimations Biomass was estimated by the replicate filtering of 2 mL of liquid culture through pre-weighed oven baked, glass fibre filter (Whatman). Each filter was rinsed thoroughly to remove excess media and then oven dried. The filters were then weighed and the weight of the filters subtracted to give an estimate of the dry weight of mycelia. 36 2.2.4 DHN inhibitors Three DHN inhibitors were added to MEA plates and inoculated with 6 mm plugs of 387N. The colony diameter was measured periodically to determine if the inhibitor was affecting growth. Colour measurements of the dried fungal mycelium were taken after day 19 to estimate the intensity of pigmentation. Tricyclazole (Biorad), pyroquilion (a gift from Dow Elanco) and cerulenin (Sigma) were solublized in ethanol and added to MEA in concentrations from 0.1-20 pg/mL. 2.2.5 Colour measurements (AE*) The colour of pigmented fungal cultures was measured using a Technidyne Micro TB-1C apparatus which is commonly used to characterize the colour of pulp, in the pulp and paper industry and also food colour in the food industry (Bristow, 1994; Giese, 1995). To determine colour, one hundred milligrams of freeze-dried fungal mycelium was ground into a fine powder with a mortar and pestle and mixed with lOg of BaS04 (Brisson et al, 1996; Eagen et al, 1997b). The mixture was pressed into a powder tablet using the powder press of the Elrepho photoelectric reflectance photometer (Zeiss Instruments). Diffuse light from two quartz/tungsten/halogen lamps illuminates the sample placed in a circular (30 mm diameter) sample area at the base of a 150mm sphere. Light reflected from the sample passes separately through red, green and blue filters and is detected by a photoreceptor located on a perpendicular plane to the sample. For each tablet, the instrument determines the colour co-ordinates X, Y and Z which are converted into L*, a* and b*, and then converted into a final colour value E*. The E* value of the 37 BaS04 tablet alone was subtracted from this to give a AE* value for the sample. The larger the AE* value the darker the colour of the sample. The following equations are used to calculate AE* (Popson, 1983). L* = 116(Y/100)1/3 -16 a* = 500 [(X/98.04)1/3 - (Y/100)1/3 ] b* = 200 [ (Y/100)1/3 - (Z/118.1)1/3 ] AE* = [ (AL*)2 + (Aa*)2 + (Ab*)2 ] % The parameters of these equations can be visualized in the following fashion (Figure 2.1) (Popson, 1983). Darker Lighter 0 Figure 2.1. The L*, a*, b* colour space (Popson, 1983). 38 2.2.6 Molecu la r materials All antibiotics, molecular reagents, restriction enzymes and DNA modification enzymes were purchased from Boehringer Mannheim, New England Biolabs, Sigma, Promega, USB, Fisher and Pharmacia and were used under conditions specified by the manufacturer. Film used for autoradiography was obtained from Kodak (X-Omat™). All other reagents Were of analytical grade and/or for biochemical use. 2.2.7 Fungal D N A extraction Large scale DNA extractions were performed according to the method of Zolan and Pukkila (1986). Briefly, freeze dried fungal mycelia was mixed with an extraction buffer (0.7 M sodium chloride, 50 mM Tris-HCl pH 8.0, 10 mM EDTA, 1% mercaptoethanol, and 1% CTAB) and incubated for 30 minutes at 55-65°C. The supernatant was extracted with phenol: chloroform: isoamyl alcohol and then the DNA was precipitated in isopropanol. Smaller scale DNA extraction required for PCR was prepared according to Raedar and Broda (1985). This method was the same as described above except that 1% SDS was used in place of the CTAB in the extraction buffer. 2.2.8 Heterologous plasmids Three plasmids, generously donated by T. Tsuge, were designated: pMBEl, pMBE8 and pMBEPIO (Kimura and Tsuge, 1993). The plasmids were constructed from digestion fragments of a 40 kb cosmid that was recovered from a Brml' A. alternata mutant that 39 was restored to melanin production when transformed with the cosmid (Kimura and Tsuge, 1993). Plasmid pMBEl is comprised of a -10 kb EcoRI fragment from the original cosmid subcloned into the EcoRI site of pBR322. When an A. alternata Aim' mutant was co-transformed with plasmid pMBEl a wild type phenotype was recovered. This plasmid contains the gene that codes for a DHN melanin biosynthetic enzyme, polyketide synthase. ' Plasmid pMBEPIO was constructed from a -4.2 kb EcoRI/PstI fragment from the original cosmid subcloned into the EcoRI/PstI site of pSP72. Plasmid pMBEPIO, when co-transformed, was able to restore wild type pigmentation to A. alternata Brm2' mutants. This plasmid carries the gene that codes for a DHN melanin biosynthetic enzyme, THN reductase. Plasmid pMBE8 was constructed from a -3.3 kb EcoRI fragment from the original cosmid subcloned into the EcoRI site of pBR322. Plasmid pMBE8 was never observed to restore wild type pigmentation to A. alternata Brml' mutants. However, when pMBE8 was used as a disruption vector with a wild type strain, melanin deficient transformants were recovered which had Brml' phenotype. Plasmid pMBE8 carries the gene that codes for a DHN melanin biosynthetic enzyme, scytalone dehydratase. These plasmids were used as heterologus probes for hybridization studies with 387N genomic DNA in order to determine if DHN-like genes were present. Probes for 40 hybridization were labelled with the Oligolabelling Kit from Pharmacia Biotech using radioactive a-P32 nucleotides (NEN). 2.2.9 Hybridizations The alkali transfer of genomic DNA to Hybond™ N+ membrane were conducted according to protocols specified by the manufacturer (Amersham). Hybridizations were conducted at 60°C in a TekStar Hybridization oven for 24 hours with agitation in 0.5 M Na2HP04, pH 7.2 and 7% SDS hybridization solution. After hybridization, the membrane was washed for 60 minutes at 60°C in 40 mM Na2HP04, pH 7.2 and 5% SDS followed by another 60 minute wash in 40 mM Na2HP04, pH 7.2 and 1% SDS. The membrane was then wrapped in clear plastic sheet and placed in a film cassette with Kodak Xray film. Exposure took place at -70°C for 24-72 hours. 41 2.3 Results 2.3.1 The effects of carbon and nitrogen sources on biomass and pigment production The purpose of this experiment was to determine if different nutrients present in the media had any affect on pigmentation. Combinations of nutrients that were shown to have an affect were then used for temporal pigment production studies. The colour of fungal mycelia was visually assessed after 8 days of growth in liquid culture supplemented with different carbon and nitrogen sources. Some of the nutrients that were chosen for this study are those thought to be available to the fungus when it grows on wood. These results are summarized in Table 2.1. A black pigment was produced when either rhamnose or ethanol was combined with asparagine or ammonium hydroxide. With respect to nitrogen source, brown hyphae were observed with globulins, while white, beige or grey hyphae were observed when supplemented with ammonium chloride or tyrosine. The colour of fungal hyphae when supplemented with asparagine or ammonium hydroxide ranged from white to black depending on the carbon source. Five combinations of different carbon and nitrogen sources (asparagine with glycerol, rhamnose or ethanol and ammonium hydroxide with rhamnose or ethanol) are of interest since dark pigmentation was observed. The biomass of fungal mycelia supplemented with different carbon and nitrogen sources was measured after eight days of growth (Figure 2.2). The organic nitrogen sources, 42 globulins and amino acids supported higher biomass yields than the inorganic nitrogen sources. The highest biomass yields occurred when glycerol was the carbon source, and when globulin, asparagine or ammonium hydroxide was the nitrogen source. Ethanol and starch gave comparable yields with asparagine and tyrosine. Rhamnose and raffinose gave consistently lower biomass yields regardless of the nitrogen source. Of the five previous nutrient combinations that pigmented darkly, only asparagine and glycerol or ethanol displayed high biomass yields. The conditions selected to monitor pigment production during growth were glycerol/asparagine and rhamnose/asparagine since they represented two extremes of biomass yields and were darkly pigmented. Table 2.1. Visual assessment of fungal hyphae colour after 8 days of growth in liquid culture supplemented with different carbon (2% w/v) and nitrogen (0.02% w/v) sources. CARBON NITROGEN SOURCE SOURCE Asparagine Tyrosine Globulins NH4C1 NH 4OH Rhamnose black beige pale brown grey black Glycerol dark brown pale brown brown grey brown Ethanol black beige brown grey black Starch light beige white brown white brown Raffinose beige beige brown white white 43 Globulins Asparagine Tyrosine NH4CI NH40H Nitrogen Source gure 2.2. Fungal biomass of 387N after 8 days growth in liquid media supplemented with different combinations of carbon (2%) and nitrogen sources (0.02%). Data were calculated from samples in triplicate and vertical bars represent standard deviation. 44 The growth of 387N in glycerol/asparagine media proceeds as a typical growth curve (right hand axis, Figure 2.3A). A two day lag phase was followed by two days of exponential growth which then stabilized after day four. Biomass yields approached 9 mg/mL after four days. In this medium, the pigmentation (AE*) of the fungal hyphae was detected after the second day (left hand axis, Figure 2.3A) and increased sharply until day three before levelling off. In the rhamnose/asparagine media, the biomass of 387N remains constant for three days before slowly increasing over the next four days (right hand axis, Figure 2.3B). Pigmentation in the hyphae was detected after day four and steadily increased over the next four days (left hand axis, Figure 2.3B). Biomass yields were much lower than those obtained for the nutrient combination in Figure 2.4A as indicated in Figure 2.3. Experiments with different carbon and nitrogen sources revealed that the glycerol/asparagine combination consistently produced dark brown pigmented hyphae and had the highest biomass yields. This nutrient combination was chosen as the most suitable for use in future experiments: 45 0 1 2 3 4 5 6 7 8 Time (days) Figure 2.3. Dry weight (right y axis) and pigment intensity (left y axis) of 387N after 7 days growth in liquid media supplemented with L-asparagine (0.2% w/v) and either (A) glycerol or (B) rhamnose (2% w/v). Points are the mean of triplicate samples and vertical bars represent standard deviation. 46 2.3.2 Influence of D H N inhibitors on fungal growth and pigmentation. In order to determine if the DHN pathway of melanin biosynthesis was responsible for pigmentation in O. floccosum 387N, inhibitors were tested for their influence on pigmentation and growth. DHN inhibitors, tricyclazole, pyroquilion and cerulenin were added to malt extract agar at concentrations ranging from 0.1-20 pg/mL. The diameter of the fungal culture was measured periodically during 19 days of growth. After seventeen days no inhibition of fungal growth was observed for cultures treated with tricyclazole and pyroquilion (Figure 2.4a,c). The cultures supplemented with cerulenin at 10 and 20 pg/mL concentrations exhibited marked inhibition of growth (Figure 2.4b). The fungal mycelia, grown on cellophane placed over the agar plate, was collected after 19 days and the colour was assessed by the standard AE* protocol. Two tailed t-tests indicate that mycelial pigmentation with concentrations of the inhibitors from 1.0-20 pg/mL differs significantly from the control at 92% confidence values. Increasing concentrations of tricyclazole from 0.1 to 10 pg/mL resulted in corresponding decreases in pigmentation (Figure 2.5). Pyroquilion lowered the amount of fungal pigmentation but there was no clear correlation between increasing concentration of the inhibitor and reduced pigmentation (Figure 2.5). Increasing concentrations of cerulenin from 0.1-1.0 pg/mL inhibited pigmentation. At concentrations of cerelunin from 10 and 20 pg/mL the minimum required biomass could not be collected for AE* measurements (Figure 2.5). The three DHN inhibitors tested appear to slightly decrease pigmentation in 387N when used at concentrations of 1.0 pg/mL or greater. 47 I • s 70 60 50 40 30 20 10 0 —•—control —o—TGAD.1 -&-TCA1.0 - T - T C A 1 0 -0 -TCA2O .-a . 5 •" , 8 / ,<>'•' 10 15 20 E E 8 70-60-50. 55 40 J .CO 30-| Q >> c o 20-I 10 0 -•—Control -o-CERO.1 - A - C E R 1 . 0 - • - C E R 1 0 - o - CER20 10 - 1 — 15 20 c 70 -, 60-50-—•—Control - • - P Y 0 . 1 - A - P Y 1 . 0 - • - P Y 1 0 - © - PY20 .... i 40-30-20- ,i>'' 10-i f 0- ' I ' I 1 1 Time (days) Figure 2.4. The colony diameters of 387N treated with 0.1-20 ug/mL (0.1, 1.0, 10, 20) of the DHN inhibitors, (A) tricyclazole (TCA), (B) cerulenin (CER) or (C) pyroquilion (PY). Data graphed is the mean of three samples and the vertical bars represent standard deviations. 48 BControl • !% Ethanol §0.1 ug/mL m 1.0 ug/mL D10 ug/mL H20 ug/mL TCA CER PYRO I n h i b i t o r Figure 2.5. The colour (AE*) of dried fungal mycelium after 19 days of growth when treated with 0.1-20 pg/mL of the D H N inhibitors, tricyclazole (TCA), cerulenin (CER) and pyroquilion (pyro). Data graphed is the mean of three samples and the vertical bars represent standard deviations. Two way t-tests conclude that pair wise data comparisons between inhibitor treatments of 1.0-20 pg/mL and control, differ significantly at 92% confidence values. 49 2.3.3 Hybridization with heterologous DHN genes In order to determine if heterologous DHN sequences were present in 387N we conducted genomic hybridizations with A. alternata DHN genes. Genomic DNA from 387N was digested with EcoRI/XhoI and separated by electrophoresis through a 1% agarose gel. The DNA in the gel was transferred to a membrane that was subsequently hybridized with the radiolabeled A. alternata 6.1 kb EcoRI/PstI fragment of pMBEl (PKS), the 3.3 kb EcoRl fragment of pMBE8 (scytalone dehydratase) and the 4.2 kb EcoRI/PstI fragment of pMBEPIO (THN reductase) (lanes 1-3, Figure 2.6). Several strong bands were visible in each lane indicating that DHN gene sequences similar to A. alternata exist in 387N. Some of the bands appear to be similar in size suggesting that in 387N these three genes may be linked. 50 1 2 3 Figure 2.6. Hybridization of 387N digested with Xho I/Eco i?/with 6.1 kb EcoRI/PstI fragment of p M B E l (PKS) in lane 1, the 4.2 kb EcoRI/PstI fragment of pMBEPIO (THN reductase) in lane 2 and the 3.3 kb EcoRl fragment of pMBE8 (scytalone dehydratase) in lane 3. Sizes are indicated to the right in kilobases. 51 2.4 Discussion The first objective discussed in this chapter was to determine if culture conditions influence pigmentation. Darkly pigmented hyphae where observed when O. floccosum 387N was grown in either rhamnose, ethanol or glycerol as a carbon source combined with either asparagine or ammonium hydroxide as a source of nitrogen. One of the defining characteristics of secondary metabolites is that they are generally not produced except under limited or suboptimal conditions. The type and severity of limitation as well as its intensity and character are thought to be important regulatory parameters for secondary metabolite biosynthesis (Bu'lock, 1975). Carbon could be a limiting factor which influences pigment production in 387N since rhamnose, ethanol and glycerol are relatively simple in structure and are quickly utilized. In Penicillium chrysogenum, the secondary metabolite, penicillin is only produced after simple carbon sources have been exhausted (Bu'lock, 1975). Earlier work with 387N showed that starch, a more complex substrate, when supplied as the sole carbon source (with slightly higher concentrations of nitrogen) was metabolized slowly and did not become limiting until after six days in culture (Abraham et al, 1993). In C. lagenarium complex nutrients in the growth media are thought to repress the expression of the DHN melanin biosynthetic genes (Takano et al, 1997b). It is equally plausible that nitrogen can be the limiting factor in 387N since previous work indicated that most of the ammonium was metabolized after two days and completely absent from the culture on the third day (Abraham et al, 1993). In O. pini, hyphal pigmentation occurs only when either asparagine or nitrates serve as the nitrogen source and not when NH4C1 or (NH4)2S04 are present (Kaarik, 1950). Correspondingly 52 we found that 387N produced only trace amounts of pigment when grown in NH4C1 as the sole nitrogen source. In Gibberella fujikuroi, low nitrogen levels are associated with the production of the polyketide pigment, bikaverin. When nitrogen levels reach critically low amounts, an additional secondary metabolite, the diterpenoid, gibberellin is produced (Bu'lock, 1975). Phosphate levels can also influence secondary metabolite production. In some species of Claviceps low levels of phosphate coincide with the production of the ergot alkaloid (Bu'lock, 1975). Additionally, environmental effects and culture age may also have a regulatory influence on secondary metabolite production. Several specific inducers of secondary metabolites have been elucidated. For example, tryptophan is an inducer of the first enzymatic step of alkaloid biosynthesis in some species of Claviceps and methionine induces the production of cephalosporin C by Cephalosporium acremonium (Drew and Demain, 1977). Based on the data presented here, we cannot draw any firm conclusions regarding a specific nutrient that may be responsible for inducing pigmentation. We can conclude that when rhamnose, ethanol or glycerol are combined with either asparagine or ammonium hydroxide, conditions are appropriate for stimulating visible pigment production in the hyphae of 387N. A more exhaustive survey would be necessary to further screen for specific inducers. When biomass was assessed after eight days of growth in liquid media with different nitrogen sources, we found that proteins such as pumpkin globulins and amino acids supported higher yields of biomass than the inorganic nitrogen of either ammonium compound (Eagen et al., 1997a). These data were also in agreement with Kaarik (1950) who found that NFLCl supported reduced biomass yields in O.pini. It appears that 387N 53 grows well and pigments darkly on nutrients in vitro which are comparable to those readily available in wood. Most of the nitrogen found in wood is in the form of proteins or amino acids and the carbohydrates are simple and soluble (Laidlaw and Smith, 1965; Langheinrich and Tischner, 1991). Alternatively it has been shown that glycerol can be obtained by the hydrolysis of triglycerides which are abundant in many softwoods (Gao et al, 1993). In order to examine growth kinetics and quantify the production of pigment, several nutrient combinations were chosen for further study. Asparagine was chosen as the nitrogen source because when 387N was cultured in this media, dark hyphae and high biomass yields were observed. Glycerol was chosen as a carbon source since biomass yields were among the highest when this source of carbon was combined with any of the nitrogen sources. When glycerol was supplied as a source of carbon pigmentation was consistently dark except when combined with ammonium chloride as a source of nitrogen. Rhamnose was chosen as second carbon source because darker pigmentation was observed despite lower biomass yields. In agreement with other species of Ophiostoma, 387N grows quickly with a short lag phase on glycerol (Hofsten, 1956). When grown in rhamnose 387N displays a more extended lag phase. Unlike primary metabolites, secondary metabolites do not provide structure or energy to the organism and often do not appear until later during growth (Hopwood and Sherman, 1990). The later appearance of some secondary metabolites, specifically antibiotics, may be linked to enhanced culture sensitivity to these antibiotics during logarithmic growth (Drew and Demain, 1977). However, melanin may confer immediate survival advantages and under 54 appropriate conditions, secondary metabolite production could occur. This could explain the more rapid appearance of pigment in 387N during the logarithmic phase of growth on either glycerol or rhamnose. In order to examine which pathway was responsible for pigmentation in 387N we examined the influence of three DHN specific inhibitors on growth and pigmentation. Cerulenin inhibits both the growth and pigmentation of 387N at concentrations greater than 1.0 pg/mL. This observation could be explained by the reported influence of cerulenin on the condensing enzyme of fatty acid synthase (FAS), a critical enzyme of primary metabolism, as well as pentaketide synthase of DHN melanin production (also a condensing enzyme) (Giuliano et al, 1973; Ohno et al, 1974). The data presented here are in agreement with the pigmentation and hyphal growth inhibition observed in A. alternata and C. lagenarium at concentrations of cerulenin greater than 0.5 pg/mL (Hiltunen and Soderhall, 1992; Kubo et-al., 1982). Higher concentrations of pyroquilion, from 10-20 pg/mL, had negligible inhibitory effects on the radial growth of 387N even after 17 days. Pyroquilion at all concentrations seemed to reduce pigmentation in 387N: Similar effects of pyroquilion were observed for the growth and pigmentation of eleven Penicillium and five Aspergillus species (Wheeler and Klich, 1995). All of the cultures of 387N treated with 0.1 pg/mL of pyroquilion or greater where not as darkly pigmented as the control. However, no correlation between the inhibition of pigmentation and increasing doses of pyroquilion were observed. 55 Tricyclazole treated cultures displalyed similar results as those reported for pyroquilion. However, a correlation of increasing dose with colour reduction was observed with this inhibitor which was not found using pyroquilion. When V. dahlia was treated with less than 10 pg/mL of tricyclazole, the cultures accumulated the DHN intermediate T3HN and 2-HJ, which is an oxidation product. At concentrations higher than 10 pg/mL the cultures accumulate T4HN and its oxidation product flaviolin (Bell et al, 1976a; Bell et al, 1976b; Froyd et al, 1976; Tokousbalides and Sisler, 1979; Wheeler and Stipanovic, 1979; Woloshuk et al, 1981). This observation holds true for Pyricularia oryzae as well (Woloshuk et al, 1980). The three DHN inhibitors tested appeared to cause a reduction in pigmentation in 387N and this observation suggests that the DHN pathway is important for pigmentation in this organism. However, even though the application of DHN inhibitors is regarded as sufficient proof that melanin is involved, tricyclazole can inhibit the production of other polyketide pigments. If this secondary inhibition occurred, it may have interfered with the observation that pigmentation reduction was occurring in tricyclazole treated cultures. It was important to use several different DHN inhibitors that would allow us to rule out the possibility that non-DHN pigmentation was involved in 387N. In order to obtain additional support for the theory that the DHN pathway of melanin biosynthesis is operating in 387N, we attempted to detect the presence of DHN-like genes in 387N by DNA hybridization. The hybridization of 387N DNA with heterologous A. alternata DHN genes at moderate stringency (see section 2.2.9) shows the presence of several bands in each lane of the blot. These bands indicate that similar sequences for 56 pentaketide synthase, scytalone dehydratase and THN reductase genes may be present in 387N. The similarity in size between the bands in different lanes suggests that these genes may be linked. Hybridization studies at low stringency using the same DHN genes from A. alternata and M. grisea genomic DNA did not reveal any homology yet the A. alternata DHN genes were shown to complement melanin deficient mutants of M. grisea (Kawamura et al., 1997). 57 2.5 Conclusions In this chapter we described the growth of O. floccosum 387N on various carbon and nitrogen sources in order to determine if a combination of nutrients would result in adequate biomass yields and display darkly pigmented hyphae. Several combinations of nutrients were chosen in order to quantify the amount of pigment produced so that differences in pigmentation could be monitored during inhibition studies. It was found that in vitro nutrients that could be correlated with dark pigmentation and high biomass yields were similar to those nutrients available to the fungi when growing on wood. The use of the DHN specific inhibitors, tricyclazole, pyroquilion and cerulenin were found to reduce pigmentation in 387N. This observation suggests that the DHN pathway of melanin biosynthesis is utilized by O. floccosum 387N and it may be similar or identical to the DHN pathway proposed for other filamentous fungi. .Southern blotting showed that hybridization was occurring between genomic DNA from 387N and the DHN genes from A. alternata. These results suggest that sequences were present in 387N that could code for similar DHN genes. These molecular data support the premise that DHN melanin biosynthesis may occur in this organism. 58 CHAPTER 3 Isolation and characterization of a putative THN reductase gene from Ophiostoma floccosum 387N 3.1 Introduction The goal of this thesis was to characterize melanin biosynthesis in O. floccosum 387N. The results discussed in chapter 2 suggest that the DHN pathway is involved in pigmentation in this organism. It has been observed that the DHN melanin biosynthetic pathway and the genes in the pathway appear to be very similar in several different ascomycetes. We can use this information to examine the melanin biosynthetic pathway in 387N. At the inception of this project, only one nucleotide sequence was available for the PKS gene and the SD gene involved in DHN melanin biosynthesis. However, the THN reductase gene from two fungi had been characterized. An alignment of the amino acid sequences deduced from the THN reductase genes of M. grisea and C. lagenarium revealed several highly conserved regions. The alignment of these known sequences could provide information for the construction of oligomers to be used as primers for PCR isolation of similar genes from 387N. Therefore, this chapter describes a PCR approach that was adopted in our attempt to isolate the O. floccosum 387N THN reductase gene of the DHN melanin biosynthetic pathway. 59 3.2 Mater ia ls and methods 3.2.1 Bacter ial strains and growth media The Escherichia coli strain used in this work was DH5oc (F", rec Al, end Al, hsd Rl 7, rjf mk+, supE44, Ml, rec Al, gyrA (NaF), rel Al, A(lacZYA-argF), U169((p80d lacA(lacZ)M15)). Bacteria were grown in Luria-Bertani (LB) media (1% tryptone, 0.5% yeast extract, 0.5%) NaCI) supplemented with ampicillin (amp) at 100 pg/mL for the selection of ampicillin resistant transformants. E. coli were transformed by the calcium chloride procedure described in (Sambrook et al, 1989) 3.2.2 Oligonucleotides Oligonucleotides for PCR and DNA sequencing were synthesized by the Nucleic Acid and Protein Service Unit at U.B.C. Table 1 lists the name, sequence and application of each oligonucleotide. 3.2.3 D N A sequencing All sequencing was performed by the Nucleic Acid and Protein Service Unit at U.B.C. Applied Biosystems apparatus and AmpliTaq FS DyeDeoxy™ terminator cycle sequencing was used. Commercially available universal primers or gene specific oligonucleotides were used as sequencing primers. 60 Table 3.1. List of oligonucleotides synthesized for use in sequencing and as PCR primers. N A M E S E Q U E N C E U S E T29F 5'GG(CTA)AA(AG)GT(GTC)GCI(CT)TIGT(GTC)ACIGG(TCA)GCIGG3' PCR T172R 5 'TTI(GC)(AT)ICCI(GC)(AT)(AG)TAIACIGC(AG)TG(CT)TTIGGIAC3' PCR T124F 5' AA(AG)GA(CT)GT(GCT)ACICCIGA(AG)GA(AG)TT(CT)GA3' PCR T141R 5 '(AG)TAIGC(CT)TCIC(GT)IGCIAC(AG)AA(AG)AA(CT)TGICC3' PCR 2R 5 'GGCTTGCTGATATCGGCCTGC3' PCR 5F 5 'GCTATTAAGGCCAACGTCTCC3' PCR 20F 5 'TCCATGTCGCCCAGACCAG3' Sequencing 414F 5'CCTTTCGGACTTCAGAATGC3' Sequencing 841F 5'CTACGGTAGCAGCAGTGCCG3' Sequencing 1048F 5'TGTCAAGACGGACATGTACGACG3' Sequencing 1756R 5' ATTGTCCGAGGTGGCAATA ACG3' Sequencing 1301R 5 'CGTAGTGCCACGAGTTCTCGTCG3' Sequencing 881R 5 'GGCTTGCTGATATCGGCCTGC3' Sequencing 501R 5' GGCTGAAAGTTAATGAAGCACAGC3' Sequencing DVF1 GGGGTTTAAGCTTATGTCTCCTGCAACTGTCAAGGACG3' PCR DVR157 5 'GGGGTTTGGATCCACCGCTGAGCGACTCAGCTACTCATGAGG3' PCR 61 3.2.4 Polymerase C h a i n Reaction Polymerase chain reactions (PCR) were performed in either 50 or 100 ul volumes. PCR primers were added to a final concentration of 0.5 pmol with 100 ng of template DNA and 20 pM of each dNTP. Boehringer Mannheim Taq polymerase was used at 0.5-1 U/reaction in lx concentration of the supplied buffer. The reaction mixture was overlaid with mineral oil, and the reaction was performed in either a Perkin-Elmer Cetus DNA Thermal Cycler® or Hybaid Omnigene Thermalcycler®. In general a five minute 94°C denaturation cycle was followed by 30 cycles of one minute denaturation at 94°C, followed by one minute primer annealing at an appropriate temperature corresponding to the Tm of the primers used (usually 50-60°C) and concluded with a one minute primer extension at 72°C. 3.2.5 Hybr id iza t ion studies Genomic DNA was separated by electrophoresis through agarose gels, transferred and fixed to a positively charged nylon membrane (Hybond™ N+ Amersham LifeScience) according to the protocols specified by the manufacturer. The same protocols as those summarized in Chapter 2 Section 2.2.9 were followed except that hybridizations and washings were conducted at 65°C. 62 3.2.6 Genomic library construction and screening To prepare the library, Mbol digestion trials were conducted to adjust conditions for partial digestion of genomic 387N DNA. The quantity of enzyme and time of incubation were manipulated. The genomic DNA was prepared by the CTAB method and the appropriate digestion products were size fractionated in a sucrose gradient (Zolan and Pukkila, 1986). Fractions containing fragments of approximately 9-23 kb were ligated to pre-digested BamHI arms of EMBL3 X replacement vector (Promega). Fragments of 10-20 kb are the size of insert required by the ^EMBL3 vector. The phage vector was then packaged in vitro using a Packagene system that contains phage head, enzymes and appropriate co-factors (Promega). Purification of lambda DNA was conducted according to the LambdaSorb® Phage Adsorbent protocol (Promega). The plaques were transferred to Biotrans ICN nylon membranes and screened with two different probes labelled with a-32P dATP using the Oligolabelling Kit from Pharmacia Biotech. Once packaged the phages were titered and amplified according to the manufacturers directions. The following formula was used to calculate the number of clones required to be screened in order to retrieve the sequence of interest. The formula: N=ln(l-P)/ln[l-(I/G)] where N is the number of clones to be screened, P is the probability of retrieving the unique sequence of interest, I is the size of the vector insert and G is the size of the genome. When P=0.99, 1=2x104 and G=2.3xl07 the formulae predicts that it will be necessary to screen approximately 7,000 clones in order to have a 99% chance of 63 isolating the desired sequences (Ausubel et al, 1987). For G, we used a crude estimate based on the published chromosome sizes of O. ulmi (20 Mb) plus a margin for error (Dewar and Bemier, 1993). Ten thousand clones were screened to ensure that all sequences were represented since it was possible that the 387N genome size varied significantly from that of O. ulmi. 64 3 . 3 R e s u l t s 3 .3 .1 P C R p r i m e r d e s i g n The conserved regions the THN reductase genes of C. lagenarium and M. grisea were used to create primers for the PCR isolation of related THN reductase sequences from 387N. Four degenerate oligonucleotides were designed based on highly conserved regions of an amino acid alignment between the THN reductase genes of C. lagenarium and M. grisea (Figure 3.1) (Perpetua et al, 1996; Vidal-Cros et al, 1994). In order to minimize degeneracy, the homologous regions chosen were weighted for codon usage. That is, regions where there were amino acids with fewer codons (one or two) were favoured over those with three or four codons. A codon frequency table (based on several genes from various filamentous fungi) was also used to help determine which nucleotide to substitute in the third codon position. The inosine nucleotide (I) was used at this position if all four codons appeared in approximately equivalent frequency. The oligonucleotides: T29F, T124F, T141R and T172R were used as primers for PCR (Table 3.2). Table 3.2. Nucleotide sequence of degenerate oligonucleotides used as PCR primers. Name Sequence T29F 5' GG(CTA) AA(AG)GT(GTC)GCI(CT)TIGT(GTC) ACIGG(TC A)GCIGG3' T124F 5' AA(AG)GA(CT)GT(GCT) ACICCIGA(AG)GA(AG)TT(CT)GA3' T141R 5' (AG)TAIGC(CT)TCIC(GT)IGCI AC(AG) AA(AG) AA(CT)TGICC3' T172R 5 TTI(GC)(AT)ICCI(GC)(AT)(AG)TAIACIGC(AG)TG(CT)TTIGGIAC3' 65 T4HN MPAVTQPRGE SKYDAIPGPL GPQSASLEGK VALVTGAGRG IGREMAMELG THR1 MPGVTSQSAG SKYDAIPGPL GLASASLMGK VALVTGAGRG IGREMAMELG Con MP VT SKYDAIPGPL G • SASL GK VALVTGAGtp IGREMAMELG T29F T4HN RRGCKVIVNY ANSTESAEEV VAAIKKNGSD AACVKANVGV VEDIVAMFEE THR1 RRGAKVIVNY ANSAETAEEV VQAIKKSGSD AASIKANVSD VDQIVKMFGE Con RRG KVIVNY ANS E AEEV V AIKK GSD AA KANV V IV MF E 50 50 100 100 T4HN THR1 Con T4HN THR1 Con T4HN THR1 Con AVKIFGKLDI VCSNSGVVSF GHVKDVTPEE FDRVFTINTR GQFFVAREAY AKQIWGRLDI VCSNSGVVSF GHVKDVTPEE FDRVFRINTR GQFFVAREAY A I G LDI VCSNSGVVSF GHVKDVTPEE FQRVF INTR GQFFVAREAY T124F KHLEIGGRLI KHLEVGGRLI KHLE GGRLI ITVNWAPGG ITVNRVAPGG ITVN VAPGG LMGSITGWAK LMGSITGWAK LMGSITGWAK IKTDMYHAVC IKTDMYRDVC IKTDMY VC AVPKHAVYSG SKGAIETFRR GVPKHAVYSG SKGTIETFVR VPKHAVYSG SKG I E T F R ^ T172R REYIPNGENL SNEEVDEYAA REYIPNGGEL DDEGVDEFAA REYIPNG L E VDE AA T141R CMAIDMADKK CMAIDFGDKK CMAID DKK SPWSPLHRVG -GWSPMHRVG WSP HRVG 150 150 200 200 250 229 T4HN LPIDIARVVC FLASNDGGWV TGKVIGIDGG ACM THR1 LPIDIARVVC FLASQDG ESR L E L Con LPIDIARVVC FLAS DG 283 272 Figure 3.1. A amino acid sequence alignment of the THN reductase gene of C .lagenarium (THR1) (Perpetua et al, 1996) and M. grisea (T4HN) (Vidal-Cros et al, 1994). Conserved regions (Con) are indicated by the third sequence and locations of the degenerate oligonucleotides are indicated by the arrows. Dashes indicate absent amino acids necessary for alignment purposes. 66 3.3.2 PCR amplification of a portion of a THN reductase gene The degenerate oligonucleotides (Table 3.1) were used as primers in PCR reactions to isolate related THN reductase sequences from 387N. Genomic DNA of 387N was used as a template and the A. alternata plasmid pMBEPIO (which carries the sequence that restores function to Brm2 mutants) was used as a positive control. Three combinations of oligonucleotides were used in PCR reactions: T29F/T141R, T29F/T172R and T124F/T172R. The products of the PCR reactions (50°C annealing) were separated and visualized on an ethidium bromide stained 1.5% agarose gel (Figure 3.2). An intense band (approximately 380 bp) was observed using the first primer combination with 387N and A. alternata (Figure 3.2, lanes 1 and 4). One faint band of approximately 500 bp in 387N and multiple bands of varying intensity in A. alternata were observed with the second primer combination (Figure 3.2, lanes 2 and 5). The third primer combination produced one fainter band around 215 bp in both fungi (Figure 3.2, lanes 3 and 6). The most intense bands of each reaction with 387N template DNA were gel purified and used as target DNA with the corresponding primer combination for an additional round of PCR in order to obtain more product. The -380 bp band from the first reaction produced a product of the same size (Figure 3.3, lane 1) and a faint band similar to the 215 bp product from the third reaction was also present (Figure 3.3, lane 3). The most prominent PCR product of 380 bp was chosen for subcloning and sequence verification. 67 387N A. alternata M 1 2 3 M 4 5 6 < - 380 bp <+- 215 bp 201 bp Figure 3.2. PCR products of genomic 387N (lanes 1-3) and the A. alternata plasmid pMBEPIO (lanes 4-6) using either T29F/T141R (lanes 1 and 4), T29F/T172R (lanes 2 and 5) and T124F/T172R (lanes 3 and 6) as primers. Lanes labelled ' M ' are kilobase ladder (BRL). 68 1 2 3 M Figure 3.3. PCR amplification of partial THN reductase gene fragments using gel purified major bands from Figure 3.2 as template DNA. Lane 1 is the PCR product using the reaction from Figure 3.2, lane 1. Lane 2 is the PCR product using the reaction from Figure 3.2, lane 2. Lane 3 is the PCR product using the reaction from Figure 3.2, lane 3. Lane ' M ' is kilobase ladder (BRL) 69 3.3.3 Subcloning and sequence confirmation of the PCR products The 380 bp PCR product amplified with the primer combination T29F/T141R and 387N genomic DNA was of interest since it may contain THN reductase related sequences. Subcloning and sequence analysis will confirm if this had occurred. This PCR product was gel purified and cloned directly into the T-tailed vector pGEM-T (Promega). The ligation products were transformed into calcium chloride competent DH5a E.coli and white colonies were selected from LB/amp/Xgal plates. Plasmid was isolated from several white colonies and checked to ensure that they contained a 420 bp insert by restriction with Sphl/PstI (Figure 3.4, lanes 2 and 5). The size disparity between 380 and 420 bp can be attributed to forty nucleotides, which were derived from plasmid sequence that flanks the 380 bp cloned insert. The two positive clones, in lanes 2 and 5, were chosen and the plasmid DNA sequenced through the multiple cloning site in both directions using the T7 and SP6 promoter primers. The two positive clones were named pGT2 and pGT5. 70 M 1 2 3 4 5 6 420 bp Figure 3.4. Transformants screened by Sphl/PstI digestion of plasmids. The arrow to the right indicates the size in basepairs of the insert in lanes 2 and 5. 71 Nucleotide sequence of pGT2 Analysis of the nucleotide sequence of the inserts of pGT2 and pGT5 would reveal whether or not they were similar to other THN reductase sequences. The 365 bp sequence obtained from the pGT2 insert is shown in Figure 3.5. The positions of the original degenerate oligonucleotide primers are indicated. A FASTA search of EMBL nucleotide databases using this sequence resulted in several strong matches (Altschul et al, 1990). The pGT2 insert shared 60-65% nucleotide identity with various THN reductase sequences from Cochliobolus heterostrophus, M. grisea, C. lagenarium, A. alternata and several Bipolaris sp. This sequence also shared 60% identity with the verl ketoreductase gene from several Aspergillus sp. The sequenced insert of the plasmid pGT2 59% identity with that of pGT5. Nucleotide sequence of pGT5 The 365 bp sequence obtained from the pGT5 insert is shown in Figure 3.6. The positions of the original oligonucleotide primers are indicated. A FASTA search of EMBL nucleotide databases using the 365 bp sequence resulted in two matches (Altschul et al, 1990). The pGT5 insert shared 70-85% nucleotide identity with various THN reductase sequences from C. lagenarium, M. grisea, C. heterostrophus and many species of Bipolaris. This sequence also shared 67% identity with C. lunatus 17 (5-hydroxysteroid dehydrogenase and 61% identity with the verl ketoreductase gene from A. parasiticus. 72 T29F • GGCAAAGTGGCGCTGGTGACGGGAGCGGGCCGTGGTATTGGTCGCGGTAT CCGTTTCACCGCGACCACTGCCCTCGCCCGGCACCATAACCAGCGCCATA TGCCACCGAGCTTGGCCGTCGTGGCGCAAATGTTATTGTCAACTACGGTA ACGGTGGCTCGAACCGGCAGCACCGCGTTTACAATAACAGTTGATGCCAT GCAGCAGTGCCGCTGCTGAGGAAGTTGTTGCCGACCTCAAAGCTCTTGGC CGTCGTCACGGCGACGACTCCTTCAACAACGGCTGGAGTTTCGAGAACCG CTGACGCTGTCGCCATGCAGGCCGATATCAGCAAGCCCGATGAAGTTGTC GACTGCGACAGCGGTACGTCCGGCTATAGTCGTTCGGGCTACTTCAACAG AAGCTGTTCGACCGTGCAGTTGCCCACTTTGGCGGAATTGACATTGTCGT TTCGACAAGCTGGCACGTCAACGGGTGAAACCGCCTTAACTGTAACAGCA CTCCAACTCTGGCATGGAGGTCTGGTCCTCGGAGCTTGACGTCACCCAAG GAGGTTGAGACCGTACCTCCAGACCAGGAGCCTCGAACTGCAGTGGGTTC T141R < AGCTTTTTGACAAGGTCTTCAACCTGAACTGCCGGGGCCAGTTTTTCCG TCGAAAAACTGTTCCAGAAGTTGGACTTGACGGCCCCGGTCAAAAAGGC TCCCCCGCAACCTAC AGGGGGCGTTGGATG Figure 3.5. Nucleotide sequence of the pGT2 insert. Arrows indicate the location of the original degenerate primers. 73 T29F GGTAAAGTGGCGCTGGTGACGGGCGCGGGCCGCGGCATTGGCCGCGAGAT CCATTTCACCGCGACCACTGCCCGCGCCCGGCGCCGTAACCGGCGCTCTA GGCCCTGGAGCTCGGACGCCGCGGCGCCAAGGTCATTGTCAACTATGCCA CCGGGACCTCGAGCCTGCGGCGCCGCGGTTCCAGTAACAGTTGATACGGT ACAGCGACTCGTCGGCCCAGGAGGTTGTCGATGCCATCAAGGCGGCCGGC TGTCGCTGAGCAGCCGGGTCCTCCAACAGCTACGGTAGTTCCGCCGGCCG TCCGACGCCGCCGCTATTAAGGCCAACGTCTCCGACGTCGACCAGATTGT AGGCTGCGGCGGCGATAATTCCGGTTGCAGAGGCTGCAGCTGGTCTAACA CACCCTCTTTGAAAAGACCAAGCAGCAGTGGGGCAAGCTTGACATTGTGT GTGGGAGAAACTTTTCTGGTTCGTCGTCACCCCGTTCGAACTGTAACACA GCTCCAACTCGGGCGTCGTCAGCTTTGGCCATGTCAAGGATGTCACGCCC CGAGGTTGAGCCCGCAGCAGTCGAAACCGGTACAGTTCCTACAGTGCGGG T141R < GAGGAGTTTGACCGCGTCTTCTCCGTCAACACCCGCGGCCAGTTCTTCGT CTCCTCAAACTGGCGCAGAAGAGGCAGTTGTGGGCGCCGGTCAAGAAGCA CGCCCGCGAAGCCTA . r •• GCGGGCGCTTCGGAT Figure 3.6. Nucleotide sequence of the pGT5 insert. Arrows indicate the location of the original degenerate primers. 74 3.3.4 D N A hybridizat ion to determine the linkage of the sequences from p G T 2 and p G T 5 Since two different sequences were obtained by PCR amplification with THN reductase homologous primers, it was important to determine if they represented two different genes. Hybridization studies will allow us to determine if the pGT2 and 5 inserts hybridize to the same or different sized genomic restriction fragments. Two membranes were prepared in replicate which contained 387N DNA digested with Mbol, EcoRI/XhoI, BamHI/Hindlll, PstI and PvuII (lanes 1-5 respectively, Figure 3.7). The membranes were hybridized with a radiolabeled insert from the pGT2 plasmid (Figure 3.7a) or from the pGT5 (Figure 3.7b) plasmid. In Figure 3.7a lane 1, a 1300 bp band is visible, lane 2 shows two faint bands at 5100 and 3300 bp, lane 3 has one band at 4500 bp, lane 4 has one band at 4800 bp and lane 5 has a band at 7000 bp. In Figure 3.7b lane 1 a 850 bp band is visible, in lane 2 a 7000 bp band, lane 3 shows two bands at 4000 and 3300 bp, lane 4 shows a 3000 bp band and lane 5 a 2800 bp band. The inserts from pGT2 and pGT5 hybridize to different sized genomic restriction fragments which may indicate that they are two unique sequences. 75 A B 1 2 3 4 5 1 2 3 4 5 Figure 3.7. Identical blots of 387N genomic D N A digested with Mbol (lane 1), EcoRI/XhoI (lane 2), BamHI/Hindlll (lane 3), PstI (lane 4) and PvuII (lane 5) hybridized with either the insert from pGT2 (A) or pGT5 (B). Band sizes (bp) are indicated by the arrows. 76 3.3.5 Lambda EMBL3 genomic library construction In order to retrieve full length intact sequence of the putative THN reductase genes, a genomic library of 387N was constructed in bacteriophage XEMBL3 (Promega). A 3 pg aliquot of genomic D N A was partially digested with 0.005 U of Mbol for increasing periods of time (Figure 3.8). The forty minute digestion time was chosen and the reaction was scaled up, and the products were size fractionated in a sucrose gradient to select for fragments of 9-23 kb (Figure 3.9). Fractions 4-6 were pooled and ligated to X.EMBL3 BamHI arms and packaged into phage head using Packagene® (Promega). 0 10 20 30 40 50 60 70 80 240 Figure 3.8. Agarose gel separation of Mbol partially digested 3 87N genomic DNA. The first lane is uncut and the last lane is a 4 hour digestion. Numbers above lanes indicate digestion times (minutes). Arrows to the left indicate size in kilobases. 77 Figure 3.9. Sucrose gradient fractions of Mboldigested 387N genomic DNA. Fractions 4-6, ranging in size between 23 and 9.4 kb, were pooled and ligated to A.EMBL3 BamHl arms. Lane M is Hindill digested X phage used as marker D N A and the sizes are indicated on the left of the photo in kilobases. 78 3.3.6 Genomic library screening Once the genomic library had been assembled in the phage, screening the fragments with the inserts of pGT2 and pGT5 would enable us to retrieve the genes of interest. The library was titered between 6.2x105 to 1.2xl06 plaques/PFU/ml. Plaques were cultured on either E. coli strains LE392 or KW251. The plaques were transferred to Biotrans ICN nylon membranes and screened using the 403 bp Sphl/PstI insert of pGT2 and pGT5 radiolabeled with a-32P by an Oligolabelling Kit (Pharmacia Biotech). Six positive plaques were obtained after four rounds of screening and phage of each plaque were labelled A through H. These six positive plaques hybridized to both pGT2 and pGT5 probes. In order to further verify the six positive plaques, four oligonucleotide primers, 2R, 5F, 5F1, 5R1 based on sequence recovered from pGT2 and pGT5, were designed to be specific for pGT2 or pGT5 (Table 3.3). The primers were tested using either pGT2 or pGT5 as template DNA to ensure that the appropriately sized PCR products would be obtained. The primers were used in the following combinations with the previously constructed primers: T29F/2R, 5F1/T141R, 5F1/5R1 and T29F/T141R with each of the six different phages (A, D, E, F, G, H) or genomic 387N as template DNA (Figure 3.10). All of the phages and the genomic DNA's with the exception of phage E gave a predicted 184 bp band with the T29F/2R primers (Figure 3.10, lanes 1). Only phage E gave products of the expected sizes (202 and 239 bp) when the 5F1/ T141R and 5F1/5R1 primer combinations were used (lanes 2 and 3, Figure 3.10). All phage and genomic 79 DNA samples showed the expected original 370 bp band with the T29F/T141R primers (Figure 3.10, lanes 4). Therefore, we were able to obtain phages A, D, F, G, and H from the library screen, which contained sequences that were similar to the insert from pGT2. Only phage E contained sequences that were similar to the insert from pGT5. The recovery of two unique but related clones from the library screen complements the previous results obtained from DNA hybridization studies in section 3.3.4. It appears that two different THN reductase sequences are present in 387N. Table 3.3. Specific oligonucleotides designed to either pGT2 or pGT5 for PCR screening of positive library clones. Name/Specificity Sequence 2R/pGT2 5' GGCTTGCTGATATCGGCCTGC 3' 5F/pGT5 5' GCTATTAAGGCCAACGTCTCC 3' 5Fl/pGT5 5' GAGATGGCCCTGGAGCTCGGA 3' 5Rl/pGT5 5' GACATGGCCAAAGCTGACGAC 3' 80 A D E F G H 387N M l 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 M Figure 3.10. PCR products amplified with phage D N A (A, D, E, F, G, H) or 387N genomic D N A with each of the following primer combinations T29F/2R (lanes 1), 5F1/T141R (lanes 2), 5F1/5R1 (lanes 3), and T29F/T141R (lanes 4). The sizes of the unique products are shown on the right in base pairs. 81 3.3.7 Subcloning of a T H N reductase gene f rom a phage clone obtained f rom the genomic l ibrary screening To more easily manipulate the sequences contained within the phage clones, the genomic insert was cloned into a plasmid vector. Recombinant phage DNA was digested with Hindlll, separated on a 1% agarose gel (in duplicate) (Figure 3.11), transferred to a membrane and was hybridized with the radiolabeled Sphl/PstI insert of either pGT2 or pGT5. The blots were identical in appearance. Radiolabeled insert from pGT2 and pGT5 hybridized to the same Hindlll phage fragments from each phage (only the pGT2 hybridized blot is shown in Figure 3.12). The 4.2 kb Hindlll fragment of phage A that hybridized to the pGT2 insert was gel purified and subcloned into the Hindlll site of the plasmid pBSKSII+ and called pBSA4.2 (Figure 3.13). A 5.5 kb Hindlll fragment from phage E was similarly subcloned and called pBSE5.5. Both plasmids were checked by restriction with Hindlll to ensure that a fragment of the appropriate size was present. The Hindlll fragment from each phage could now be sequenced. 82 A D E F G H M 1 M 2 Figure 3.11. Hindlll digested phage D N A from six positive phages, (labeled A, D, E, F, G, and H) were separated on an 1% agarose gel. Lane M l is a kilobase ladder and lane M2 is a base pair ladder. 83 A D E F G H Figure 3.12. Hindill digested phage D N A shown in Figure 3.11 was transferred to a membrane and hybridized with the radiolabeled Sphl/PstI fragment of pGT2. Arrows indicate the size of bands in kilobases. 84 A HindUI HindJU Figure 3.13. The construction of the vector pBSA4.2. (A) The 4.2 kb Hindlll fragment of a genomic X phage clone showing the orientation and approximate location of the THN-R gene (b) The cloning vector pBSKSII+ used to accommodate the fragment. 85 3.3.8 Nucleic acid sequence analysis The genomic fragments recovered from the library, now contained within plasmids, were sequenced to determine if the recovered sequences shared homology with other THN reductase genes. The 4.2 kb Hindill insert in plasmid pBSA4.2 was sequenced using the primer 2R (constructed from pGT2 sequence). When a sequence was determined, nested primers were constructed in order to obtain flanking sequences and sequences in both directions (see Table 3.4 and Figure 3.14). The complete sequence is shown in Figure 3.15. The sequence shown is 1751 bp in length with one possible open reading frame from nucleotide position 1 to 877 that may potentially code for a protein composed of 268 amino acids. Plasmid pBSE5.5 was sequenced using the primer 5F (constructed from pGT5 sequence). Sequence obtained from pBSE5.5 was found to be 99% identical to phageA.. Subsequent characterization using PCR to check for the existence of appropriate sequence was negative (data not shown). Digestion of the plasmid with Hindill failed to reveal a 5.5 kb insert. Plasmid pBSE5.5 was not characterized further since subcloning the appropriate fragment from the original phage was unsuccessful. Within the 1.7 kb nucleotide sequence obtained from pBSA4.2 there was a putative intron of 76 bp in length located at the nucleotide position from 766 to 841 (Figure 3.16). The signature sequences in the putative intron match the consensus sequences for Neurospora crassa 5' splice sites GGT(A/G)(A/C)G(T/C), the internal splicing sequences (G/A)CT(A/G)AC and the 3' splice signal (A/T)(T/C)AGG (Bruchez et al, 1993b). The 86 organization of introns and exons in. A alternata and C. heterostrophus is almost identical with a 61 nucleotide exon followed by a 51/63 nucleotide intron then a 477 nucleotide exon followed by a 49/61 intron and a 263 nucleotide exon. The sequence of the putative THN reductase gene from 387N shows a 765 nucleotide exon followed by one predicted intron, 76 bp in length near the 3' end of the gene and then a 35 nucleotide exon. Using the Matlnspector program (available at the Canadian Bioinformatics Resource website), the 605 bp sequence upstream of the +1 ATG codon was screened for transcription factor binding sites (TFBS) (Quandt et al, 1995). Several sequences were indicated that shared significant homology with several fungal TFBS, specifically heat shock factors and ABF1. Many low scoring matches were also defined which may not bind themselves but could participate in co-operative binding. Pyrimidine-rich areas can also serve as possible promoter elements and within this gene, the region 45 bp upstream of+1 has a 53% C/T content that could qualify as pyrimidine-rich. A potential polyadenylation consensus sequence AATAAA is located within the putative THN reductase gene from 387N at nucleotide position 995. This motif was detected by the Grail ORNL sequence analysis program (available at the Canadian Bioinformatics Resource website). 87 Table 3.4. Synthetic oligonucleotides used as primers to sequence of the insert in pBSA4.2. Name Sequence 20F 5' TCC ATGTCGCCC AGACC AG3' 414F 5' CCTTTCGGACTTC AGAATGC3' 841F 5' CTACGGTAGCAGCAGTGCCG3' 1048F 5' TGTC AAGACGGAC ATGTACGACG3' 1756R 5' ATTGTCCGAGGTGGC AATA ACG3' 1301R 5' CGTAGTGCC ACGAGTTCTCGTCG3' 881R 5 'GGCTTGCTGATATCGGCCTGC3' 501R 5' GGCTG AAAGTTAATGA AGC AC AGC3' pBSA4.2 TFIN reductase L7kb 201^ i = > 414F 1048F 881R LSJIR 1756R 501R Figure 3.14. The approximate location and orientation of the primers used to sequence a 1.7 kb section of the 387N THN reductase gene. 88 TCATCTTTGAGTACGGTTTCTTTTCCATGTCGCCCAGACCAGATTTTGCATACAAATAAGAATATGCACG -605 GAAGCATTAATATCCAGTAGATCGTGCAATGGCCTCAGTGAACTCTGTACAGGTTATGGACGTTCTGGAA -535 CTCGAATTTAGGCAAGACTTTTAATCGAACTATCCAGTTACTACCTGGGTCCCGTACAAAAAAGCTCAAA -4 65 AGCAAAGCCAGGGCGTTTTCACATCTCTTCGCTTAACACAAGAAACCAACCCACAGGCATAAGACTAAGT -3 95 CCCTGTCATCAGATATATTACACAATACAGCGTAGAGAGGACTTCTCGACACTTGGATGTACGGACTTGT -325 AACTTTATCGCCTTCCGACCAACTTCCCCAACTAAAGAACCGTTTCATGAAAATACGGAGATTCCTTTCG -255 GACTTCAGAATGCTAGTATCAAGAATAGAGATTGGGGAAGTTGTAGGCACTTGAACGTAATCATTATCCT -185 CATATAAGAGATGCTGTGCTTCATTAACTTTCAGCCTATTCTTGTTTTCAAACGACCACCTCGTCTCCAT -115 CAACAAGCTAACCAATCGTTATTAATAAAACAACGAAAAGAATCAACACTCGGTCGTAACCTTTAGGACT -4 4 ATTACCAATTTTCTCACAAACTCTCACACCACCACAAACAAATATGTCTCCTGCAACTGTCAAGGACGCC 28 M S P A T V K D A 9 GCCCGCCCTCTTGCTGGCAAGGTTGCCATCATCACTGGTGCTGGCCGTGGTATTGGTCGCGGTATTGCCAC 98 A R P L A G K V A I I T G A G R G I G R G I A T 33 CCGAGCTTGGCCGTCGTGGCGCAAATGTTATTGTCAACTACGGTAGCAGCAGTGCCGCTGCTGAGGAAGT 168 E L G R R G A N V I V N Y G S S S A A A E E V 56 TGTTGCCGACCTCAAAGCTCTTGGCACTGACGCTGTCGCCATGCAGGCCGATATCAGCAAGCCCGATGAA 238 V A D L K A L G T D A V A M Q A D I S K P D E 79 GTTGTCAAGCTGTTCGACCGTGCAGTTGCCCACTTTGGCGGAATTGACATTGTCGTCTCCAACTCTGGCA 308 V V K L F D R A V A H F G G I D I V V S N S G M 103 TGGAGGTCTGGTCCTCGGAGCTTGACGTCACCCAGGAGCTTTTTGACAAGGTCTTCAACCTGAACTGCCG 37 8 E V W S S E L D V T Q E L F D K V F N L N C R 126 GGGCCAGTTCTTTGTTGCCCAGCAAGGCCTCAAGCACTGCCGTCGTGGTGGCAGCATCATCCTGACCTCA 4 4 8 G Q F F V A Q Q G L K H C R R G G S I I L T S 149 TCAGTCGCTGCGTC.GCTCAGCGGTATCCCCAACCACGCTCTATACGCAGGCTCAAAGGCTGCTGTCGAGG 518 S V A A S L S G I P N H A L Y A G S K A A V E G 173 GCTTCACGCGTGCCTTCTCCGTTGACTGCGGCGAGAAGGGCGTCACTGTCAATGCCATTGCCCCGGCGG 588 F T R A F S V D C G E K G V T V N A I A P G G 196 TGTCAAGACGGACATGTACGACGAGAACTCGTGGCACTACGTTCCCGGTGGCTACAAGGGCATGTCGCAA 650 V K T D M Y D E N S W H Y V P G G Y K G M S Q 219 GATGTCATCGACGAGGGCATTCTCAAGGCTTGCCCGCTCAAGCGCGTCGGTACCCCGAGCGACATCGGCA 7 20 D V I D E G I L K A C P L K R V G T P S D I G K 243 AGGCTGTTGCTCTGCTTGTTAGCGAGGAGGGAGAATGGATCAACGgtaagcaagaaacccacaaactttc 7 90 A V A L L V S E E G E W I N G 258 GggccggtcggaatgcgaatacttgaatgttaatactaacaaaaatattagGCCAGATTATCAAGCTGTC 8 60 Q I I K L S 264 TGGCGGTTCTGCCGTTTGAGTTGGCAGCTCCATGGTGATGCGGGAAGGGGTTATCAATGTTTGTCGAGAG 930 G G S A V stop 268 TATCCAAAAAGAAGAAAACATCCATAGCGAAGCACCATAAATGAAATAAAGCAATATTTTTGTCATTATA 1000 TTATATGAGTTGTTCTCTTTTAGGATGACTACGAATGCTGTCACATCCTAACTCGTTATTGCCACCTCGG 107 0 ACAATC 107 6 Figure 3.15. Complete nucleotide sequence of 387N THN reductase gene. Deduced amino acid sequence of the THN reductase protein is indicated below their respective codons. Small case nucleotides indicate the intron. The start and the stop codon are in bold. Numbers in the right margin indicate either nucleotide position relative to the A nucleotide of the start codon, or amino acid position. 89 387N Brm2 Brnl Thrl 61 112 63 JZ£5_ 477 121 61 76 15" 61 51 477 49 263 471 263 50,——.216 57 22 50 184 Figure 3.16. The organization of introns (lines) and exons (boxes) in the reductase genes of 387N (387N), C. lagenarium (Thrl), A. alternata (Brm2), and C. heterostrophus (Brnl) (Kawamura et al, 1999). Numbers indicate the length of the span in nucleotides. 90 3.3.9 Protein sequence analysis The open reading frame of 877 nucleotides from the 1.7 kb sequence of the putative 387N reductase gene encodes a putative protein of 268 amino acids. A sequence alignment with other reductases shows a high degree of conservation. The six boxed residues marked by asterisks are those strictly conserved in the short chain alcohol dehydrogenase (SADH) family (Figure 3.17) (Persson et al, 1991). The three strictly conserved polar residues; serine 150, tyrosine 164, and lysine 168 are thought to form the active site and be functionally important in the catalytic domain for the transfer of hydrogen from the coenzyme to the naphthol substrate (Figure 3.17) (Persson et al, 1991). When individually aligned with other fungal TFTN reductases, 387N shares 43-48% identity. There are nineteen key residues shared by 75% of other characterized SADH of which thirteen residues are present in all of the THN reductases shown in the alignment in Figure 3.17 (Persson et al, 1991). Potential protein motifs detected by Prosite (©Swiss Institute Bioinformatics) are as follows. One putative protein kinase C phosphorylation site is located at amino acid position 5-7 and the motif is [ST]-X-[RK], where either residue in the square brackets is allowed. Five casein kinase II phosphorylation sites exist at locations 5-8, 75-78, 101-104, 108-111, and 237-240. The consensus pattern for this motif is [ST]-X2-[DE]. Nine potential myristoylation sites occur at amino acid locations 22-27, 26-31, 47-52, 135-140, 166-171, 186-191, 195-200, 212-217, and 225-230. The pattern for this motif is G-{EDRKHPFYW}-X2-[STAGCN]-{P}, where the residues in round brackets are those 91 residues not permitted. One putative C-terminal amidation site exists at amino acid position 35-38 which conforms to the X-G-(RK)-(RK) consensus pattern. C-terminal amidation has not been shown to occur in unicellular organisms or plants. In addition the SADH family signature [LIVSPADNK]-X,2-Y-[PSTAGNCV]-[STAGNQCIVM]-[STAGC]-K-{PC}-[SAGFYR]-[LIVMSTAGD]-X2-[LIVMFYW]-X3-[LIVMFYWGA-PTHQ]-[GSACQRHM] was detected at amino acids 151-179. These motifs could indicate potential post-translational modifications which could provide clues regarding the function and location of the gene product. Using Phylip 3.5c (©Joseph Felenstein, University of Washington, 1986-95), a phylogenic inference program, an unrooted neighbour joining consensus tree was generated based on the pairwise comparison of amino acid sequences (Figure 3.18). From the parsimonious tree derived from comparisons of only this specific gene, one could infer that 387N may be closely related to C. lagenarium and M. grisea. 92 337NThnR CIThr 1 MgThnR RaBrm2 ChBrn1 Consensus PRTUKDRR R pGUTSQSRG SKVDRIPGPL FIUTQPRGE SKVDRIPGPL 1SIEQTW R N IEQTW . TQ PJLhBK Uflll l(t^NGRG" GLRSRS_13K URJJTGRGRG GPQSRS_EGK Ufl.UTGRGRG SLRDK Ufl SLRGK URUUrrGECRG JUTGHGRG PGIRTELp UtmFljED_JJiR. rfel. El >,E\' R REM R <RKR <RM=J 1EL3 1EL3 I EL R I EL R 36 50 50 33 33 50 387NThnR CI Thr 1 MgThnR RaErm2 ChBrn1 Consensus RRGRHU RRGRKU RRG3KU RBGRKLill DNV UNV KftGRKlUplJNV KPGRKi) R JMV pSSSRRREED RNSRETREEU RNSTESREEU RNRUEG RE 3 U RNRUEGRE3U RNS.E U R D L K H L G ] — TpflTjflMCfipl U3RI URfll <EI U<E KfiL K R L p H G IKRUGHG S D R R S I K R H U S Q R R C U K R H U S D R R R F K R H U S D R H R F K R H U GI— SOfiRfl. KHNU G . UE . . . K SKPD SDUDQIUKMF GUUEDIURMF GNUEESEkLM GNUEESEkLM 78 98 93 83 83 100 387NThnR CI Thr1 MgThnR RaBrm2 ChBrn1 Consensus -EWKLFDRfl GERKQIUGR-EEflUkIFGK-DDUURHFGK-DDWKHFGK-.EUUK.FGK— URHFGGI PI -LP U CENSGUUSFG UpHSGMEUWS SEI^UTpELFTKpF C J S H S G U U S F G -LDI -LDIt CfSNSGUUSFG -LDI3 qSNSGUUSFG LtUJj dSUSGJJUSFG H . m i i l F HUK DUT; E E F DRUFR HUK [JUT: E E F DFlUFj HFK DUT:E E F HFKDUT=EEF l iEl CH2RG NrRG r n r R G DPJUFri NTRG DRJJFHI ITRG ItUTBD. 127 141 141 126 126 150 387NThnR CI Thr 1 MgThnR RaBrm2 ChBrn1 Consensus pFFURpQGI_R|HCRF*3G pFFUR pFFURRERVK HLE pFFUflkRRVk QFFWR ERVK HLEUGG pFFUflkflflVK RMEMpGRI . fiVK_H. E 3G SI 3 MEMGGR G E F : I IL IL IL IL IL IrsgUflRSLSG TGQfiKG ITGQRKfl ITGQRKG ITGQRKG 1GS 1GB 1GS 1GS LI M G E I I TGQRKG t J f rkHRLjMJns K ^Hfl.^ RGS UP(HRJYEGS iJP<HfljyBGS U = <hlRjy = GS UP<hlflJv'3GS PTI pfl I fll pfl I E F F E F F * E T - r j R =. r - rp fliEirEtrBc 177 191 191 176 176 200 387NThnR CIThrl MgThnR RaBrm2 ChBrn1 Consensus FSUJD MR I MR I MR I MR I :GER G D K < I J1RDK<I R G E K R I R G E K O I rruN p-UNRURPGG h-UNJURPGG tu. GEkki rn.iH hi nruN h"UN BPGGU V RPGGl V RPGG ulaEGG kTDMVPENSW KTDMV kTDMVHRUCR KTDMVHRUCR KTDMVHRUCR IDUCR HMJP33v'KGM P--33ELD EV E v INLS 3PEKLS DQLS EMIEMGJ . .LS Ev 1 EV IPNG IDEGIL DEGpPEFflfl-JEVflflS DDOUDEVflC-DDQJUPEVRC-SQDD U NEE J LlDEVfl. -227 240 241 225 225 250 387NThnR CI Thr 1 MgThnR RaBrm2 ChBrn1 Consensus kflCP GUSpMHRUG FlWSp .WSfeJ KpUGP" P E C HRUGL TUSPHNRUGp TUSpHNRUGp pppl BUGI P I 3KF D| RRUUf RRUJtF RL :F :F _ i S 3D G DIJUM iPlRRULtF Ut idr luUN EE 3: 3D 3 10 3 EWIN —E HUT pDppWUN GQI IKLSGGS SR LEL— GkUIGIDGGfl GKUIGIDGflfl GKUIGIDGflR flU CM CM CM GKUIGIDG.R CM 259 272 283 267 267 292 Figure 3.17. An alignment of the amino acid sequence of 387N THNR (387NThnR) with that of C. lagenarium (CIThrl), M. grisea (MgThnR), A. alternata (AaBrm2), and C. heterostrophus (ChBrn 1). The consensus sequence below indicates strictly conserved residues. Dashes indicate spaces introduced to maintain the alignment. Boxed residues indicate identity and shaded residues are conservative amino acid substitutions while dots are non-conserved positions. 93 387NThnR CIThrl MgThnR ChBrnl AaBrm2 Figure 3.18. Unrooted neighbour joining consensus tree of the reductase proteins of 387N THNR (387NThnR) C.lagenarium (CIThrl), M. grisea (MgThnR), A. alternata (AaBrm2), and C. heterostrophus (ChBrnl). 94 Fungi Ascomycota Euascomycetes Loculoascomycetes Pleosporales Anamophic Pleosporales Alternaria alternata Pleosporaceae Cochliobolus heterostrophus Pyrenomycetes Ophiostomatales Ophiostomataceae Ophiostoma floccosum Phyllachorales Magnaporthaceae Magnaporthe grisea Plyllachoraceae Colletotrichum lagenarium Figure 3.19. Morphological and molecular based taxonomic organization of the fungi of interest (Genbank). 95 3.4 Discussion Degenerate oligonucleotides based on fungal THN reductase sequences was a technique that allowed us to retrieve 387N PCR products. Two 365 bp PCR amplification products from oligos and 387N target DNA were subcloned and characterized by DNA sequencing. The two sequences, carried in pGEM vectors were named pGT2 and pGT5. They shared 60-65% and 70-85% nucleotide identity with other fungal THN reductases and 59% identity with each other. These two products are not likely to be the result of Taq polymerase misincorporation error since one would expect higher identity (on the order of 99%) due to the low misincorporation frequency of one nucleotide in 1000 nucleotides. Hybridization studies indicated that the two PCR products showed homology to separate Hindlll digested genomic sequences of 387N. These two PCR products may represent two unique sequences that may be involved in DHN melanin biosynthesis in 387N. The PCR derived sequences were used to screen a genomic library in order to recover full length reductase genes from 387N. Six positive phage were retrieved from the screening which were further characterized by PCR and hybridization experiments. Hybridization studies indicated that the genomic inserts of all six phage carried sequence that hybridized to both PCR cloned inserts of pGT2 and pGT5. Additionally PCR data, using either pGT2 or pGT5 specific primers with the six phage, indicated that only phage E had the pGT5 primer binding sites required to allow the production of either the 202 or 239 96 bp pGT5 specific PCR amplification products. The remaining five clones produced pGT2 specific PCR amplification products. A possible explanation for the presence of two reductase-like sequences is that one of the sequences isolated may be involved in the biosynthesis of polyketide metabolites other than DHN melanin. Both reductase-like sequences shared homology with other fungal THN reductases as well as a ketoreductase gene (verl) involved in aflatoxin production in Aspergillus sp. The antithetical explanation is that two enzymes are required for the reduction of both DHN naphthalene substrates in 387N. This premise would have been interesting to explore and additional characterization and subcloning of phage E was performed. Repeated attempts were made to retrieve the genomic insert of phage E but due to time constraints and technical difficulties, this was not achieved. It is interesting that no one has yet cloned two reductase genes involved in DHN melanin biosynthesis from the same organism. Regarding the dual reductase theory, genetic evidence from other fungi is inconclusive. In V. dahliae, a mutant has been isolated and shown to lack both reductase activities whereas another mutant lacked only the T3HN activity (Bell and Wheeler, 1986). The mutants were characterized by intermediate accumulation but were not characterized by genetic crossing so it was unclear whether the mutations were within a single gene. In M. grisea, C. miyabeanus and Pyricularia oryzae only mutants which lacked the ability to convert T3HN to vermelone were isolated (Chumley and Valent, 1990; Kubo et al, 1989; Woloshuk et al, 1980). In M. grisea, the mutants were analyzed by genetic crosses and 97 found to be the result of a single gene mutation (Chumley and Valent, 1990). However, the inability to detect something cannot be taken as evidence for non existence. The theory that only a single reductase is operating and that it has different substrate affinities appears to be supported by most enzyme kinetic studies. At the same time, these studies do not disprove a two reductase hypothesis. A T4HN reductase was purified from M. grisea and was shown to reduce both the T3HN and the T4HN substrates at pH 5.5 and 30°C (Vidal-Cros et al, 1994). The authors state that this evidence supports the one reductase theory. However, different amounts of substrate were used in this assay, 30 pM of T3HN compared to 200 pM of T4HN. The six fold difference in substrate concentration could reflect differences in substrate affinity. Information regarding the cloning and expression of this polypeptide were not reported and it was therefore not clear if this was a product of the same gene reported by (Andersson et al, 1997a; Andersson et al, 1997b; Andersson et al, 1996). Similar results were obtained using whole cell extracts from V. dahliae and P. oryzae (Viviani et al, 1991). Both T4HN and T3HN substrates were reduced at an optimum pH of 5.5 and like in M. grisea a ten fold higher affinity for T3HN was observed (Viviani et al, 1991). It seems that both the genetic and the enzymatic evidence are not conclusive and can neither support nor refute the single or dual reductase hypotheses. It is interesting that a similar debate exists regarding whether there are one or two SD in the DHN biosynthetic pathway. 98 Nucleotide sequence was obtained from the genomic insert of phage A which was subcloned in the plasmid pBSA4.2. 800 bp of this sequence shared individual identities of 43-48% with the other fungal THN reductase nucleotide sequences. Within the nucleotide sequence of 387N THN reductase gene, a putative transcription start site (ATG) was located at the beginning of the largest ORF. This was assigned position +1. Sequence from the 387N reductase surrounding the translation start site (AAACAAATATGTC) shared 5 out 10 nucleotides with the vertebrate consensus sequence (GCCA/GCCATGG) and 7 out of 10 nucleotides with the consensus sequence from Neurospora crassa (CNNNCAA/CTATGGC) (Bruchez et al, 1993a; Bruchez et al, 1993b; Kozak, 1987). The -3 position of the 387N reductase sequence conforms to the filamentous fungal consensus of an adenine at this position (Gurr et al, 1987). Intron number and location are highly variable structural features in filamentous fungi (Gurr et al, 1987). Fungal introns are on average less that 100 bp in length (Gurr et al, 1987). C. lagenarium THR1 has four putative introns which vary in length from 50-216 nucleotides. In M. grisea only the cDNA of a THN reductase was sequenced so the existence and/or position of introns is unknown (Perpetua et al, 1996; Vidal-Cros et al, 1994). Both A. alternata and C. heterostrophus have two introns from 49-63 bp in length. The putative intron in the 387N reductase is located very near the 3' end of the gene. The intron in the 387N reductase gene conforms to filamentous fungal 5' consensus splice sequence of GTANGT at five out of six positions, the 3' splice signal PyAG and the internal splice sequence PyGCTAACN in 7 out of 8 positions (Gurr et al, 99 1987). The 387N THN reductase gene sequence possesses an exon of 765 nucleotides in length which was not similar to any of the other fungi. The putative gene product of the 387N THN reductase gene is 268 amino acids in length. This is shorter than either of the reductases from C. lagenarium and M. grisea which are 282 amino acids in length (Kimura and Tsuge, 1993; Vidal-Cros et al, 1994). However, the lengths of all three proteins are consistent with the average subunit size of other SADH (Persson et al, 1991). Thirteen of the fourteen absent amino acids in 387N sequence occur at the C-terminus of the protein between amino acid positions 15-28 in the alignment. When compared with other SADH, these differences and lack of homology at the C-terminus are not unusual since other SADH also lack similarity in this region. An alignment of 20 SADH proteins from a variety of organisms showed that gaps of 6-19 amino acids were common in this region. In the SADH family there are six strictly conserved residues. The putative sequence from 387N and the other four fungal THN reductases share these conserved residues and the following features. Three of the conserved residues are glycines and are thought to be important in coenzyme binding. These glycines occur at positions 15, 22 and 28 within the 387N THN reductase sequence. The other three conserved residues are serine, tyrosine and lysine. These residues lie within the highly conserved motif of SX 1 0 . 1 3 YX 3 K and are proposed to be part of the active site of SADH (Persson et al, 1991). In M. grisea, these residues has been shown to be involved in the transfer of hydrogen from NADPH to the naphthol substrate (Andersson et al, 1997a; Andersson et al, 1997b; Andersson et al, 1996). These three residues occur at the following amino acid positions within the 387N THN reductase 100 protein sequence: serine 150, tyrosine 164 and lysine 168 and they conform to the SX 1 0. 1 3 Y X 3 K motif. One other feature that SADH share is the domain order of a coenzyme binding domain followed by a catalytic domain (Andersson et al, 1996). Linkage of the three genes for pentaketide synthase, scytalone dehydratase and THN reductase is variable in filamentous fungi. In A. alternata, these three genes are closely linked within a 30 kb region (Kimura and Tsuge, 1993). In M. grisea the genes do not appear to be linked and in C. lagenarium the PKS gene is not linked to the SD gene (Chumley and Valent, 1990; Kubo et al, 1991). In C. heterostrophus the PKS gene is tightly linked to the THN reductase gene (Tanaka et al, 1994). Hybridization studies described in chapter 2 suggest that in 387N the DHN genes may be linked since bands of similar sizes appear when XhoI/EcoRI digested genomic DNA is hybridized to each of the A. alternata DHN genes. Some molecular taxonomy has been done with species of Ophiostoma but not specifically with O. floccosum 387N. ITS and 18S rDNA sequences have been used to differentiate aggressive and non-aggressive isolates of O. ulmi and O. piceae from O. quercus (Jeng et al, 1996; Kim et al, 1999; Pipe et al, 1995). Sequence comparisons of the small subunit rDNA of many different Pyrenomycetes has been used to illustrate differences between the genera Ophiostoma and Ceratocystis (Spatafora and Blackwell, 1994). The THN reductase nucleotide sequence is the only sequence available from this isolate of O. floccosum other than a small portion of a protein sequence from a serine proteinase. Therefore it is premature to deduce molecular phylogenetic relationships based on the alignments of one gene sequence. However, an unrooted neighbour joining consensus 101 tree derived from amino acid homology between the five fungal THN reductases, suggests that the 387N THN reductase sequence shares more homology with the THN reductase sequences of C. lagenarium and M. grisea than with C. heterostrophus and A. alternata. The phylogenetic relationships illustrated in this tree are mirrored by the standard taxonomic divisions from Genbank pictured in Figure 3.19. 3.5 Conclusions Using degenerate PCR primers we were able to retrieve two PCR amplification products having sequence homology with other THN reductase genes. These two products were used to screen a genomic library of 387N and two clones were recovered. Subsequent characterization revealed that the clone with pGT5 homology was incomplete. The characterization of this gene was not pursued due to time constraints and technical difficulties. The existence of two sequences recovered from 387N suggests that perhaps two reductases may exist. However, in other fungi, there is no conclusive evidence to either support or refute the existence of two THN reductases. The pGT2 specific clone was subcloned and sequenced. The nucleotide sequence and putative amino acid sequence would suggest that the gene recovered may code for a 268 amino acid protein which is highly similar to other fungal THN reductases which are members of a family of short chain alcohol dehydrogenases (SADH). An unrooted neighbour joining consensus tree based on amino acid homology of the THN reductase supports standard taxonomic relationships. The organization of the THN reductase gene 102 in 387N is unique, and does not share the same number or location of introns found in the reductase genes of other fungi, nor does it share similar exon length characteristics. 103 C H A P T E R 4 Funct iona l characterization of the T H N reductase gene of O. floccosum 387N 4.1 Introduct ion The final goal of this thesis was to characterize the function of the gene that was described in Chapter three. Although the gene that was isolated was similar to other fungal THN reductases this data alone is not sufficient proof that the sequence has that particular biological function in 387N. We attempted to characterize gene function using the techniques of gene disruption and complementation. Mutation analysis has always been important to study a gene function. To demonstrate gene function, researchers complement new or previously characterized mutations with DNA libraries or their gene of interest and look for restoration of the function. In many filamentous fungi, gene disruption has proven to be very effective for creating precisely engineered mutants (Osiewacz and Weber, 1989; Tsuge et al, 1990). Gene disruption involves the transformation of the wild type fungi with an incomplete or defective copy of the gene vectored in a plasmid that carries a selectable marker. Targeted integration of the defective copy will disrupt the wild type copy, resulting in loss of function. Complementation is a popular method used successfully to restore gene function in many filamentous fungi. This technique requires the existence of a mutant that can be transformed with a functional copy of the gene of interest. Complementation does not require targeted integration, but does require gene expression once transformation has 104 occurred. Two recent papers have summarized the success of using Alternaria alternata DHN genes to complement melanin deficient mutants of the closely related ascomycetes, Magnaporthe grisea and Colletotrichum lagenarium (Kawamura et al, 1997; Takano et al, 1997a). Melanin deficient mutants of M. grisea and C. lagenarium are well characterized and do not exhibit variable pigmentation like that encountered in 387N (see Appendix). In the absence of melanin deficient mutants of 387N, intergenic complementation could serve as an alternative method for determining the function of the THN reductase gene. The Southern blot presented in Figure 2.7 illustrates that genomic 387N DNA shares similarity with A. alternata DHN sequences. The 387N THN reductase gene product shared 76% identity with the reductase from M. grisea and lends support to the possibility that the 387N THN gene could complement M. grisea mutants. Several different M. grisea buf mutants (deficient in THN reductase gene) were available to us to attempt complementation with the 387N THN reductase gene. The buf mutants used for this study have 50 kb deletions which span the BUF1 locus near marker CH3-24H on chromosome 2 (Nitta et al, 1997). There are effectively no BUF1 sequences in these mutants (Mark Farman, personal communication). We attempted to create pigment deficient mutants using traditional methods. UV mutagenesis was conducted on the O. floccosum 387N but we were unable to definitively characterize appropriate pigment deficient mutants due to pigment variability (see Appendix). Instead we attempted gene disruption to obtain THN reductase deficient mutants. 105 4.2 Mater ia ls and Methods 4.2.1 O. floccosum 387N protoplast formation and transformation A slightly modified procedure based originally on one used for O. ulmi was used to protoplast and transform single spores of 387N (Royer et al, 1991). The modifications in the protoplasting procedure involved extended incubation periods of the yeast cells with lysing enzymes and several cheesecloth filtrations to separate protoplasted yeast cells from non-protoplasted cells and mycelia. Currently there is another student from the lab working out optimal protoplasting protocols. The transformation protocol was unmodified. Transformants were selected on malt extract agar (MEA) supplemented with 150-400 ug/ml hygromycin. 4.2.2 Growth of putative 387N disruptants in vitro and in vivo Plugs of putative 387N disruptants were placed on MEA plates overlaid with a sterile sheet of cellulose. After 14 days, the cellulose was peeled away from the media and the fungal mycelia freeze dried and ground. The colour of the O. floccosum fungal mycelia was measured as described in Section 2.2.5. Blocks of gamma irradiated lodgepole pine (Pinus contorta var. latifolia) sapwood (30 x 10 x 5 mm) were inoculated with 5 mm diameter hyphal front plugs of the putative 387N disruptants. The wood blocks (8 blocks/ 387N disruptant) were placed on sterile distilled water saturated filter paper (Whatman 3M) in parafilm sealed petri dishes. The dishes were incubated at 23°C in the dark for 14 days. 106 4.2.3 M. grisea mutants used for complementation studies In M. grisea, the BUF1 locus is on chromosome 2 near the marker CH3-24H (Nitta et al, 1997). The gene at this locus codes for a THN reductase. In M. grisea, genes coding for enzymes of the DHN pathway of melanin biosynthesis are unlinked. The buf mutants IC4-1, IC4-4, IC4-5, and IC4-6-1 were recovered from crosses between parental strains of Guy 11 and 2638. These mutants were received from Dr. M. Farman, Department of Plant Pathology, University of Kentucky. The buf I locus of strain 2638 is unstable due to the presence of many transposable elements. This locus is spontaneously deleted at high frequency during mating with Guy 11. The mutations were caused by deletions of approximately 50 kb that removed all of the buf sequences. The absence of buf sequence was confirmed by hybridization studies (see Figure 4.11). The sizes of the deletions were inferred by hybridization of cosmids with buf mutant DNA and comparison with a contig map (Mark Farman, personal communication). These strains were routinely cultured on oatmeal agar at 25-28°C in the dark. 4.2.4 M. grisea protoplast formation and transformation The protocol used for the transformation of M. grisea was modified by M. Orbach (personal communication) (Leung et al, 1990; Sweigard et al, 1992). In preparation for protoplasting the buf mutants were grown on oatmeal extract plates until confluent. A sterile spatula was used to cut out 25 cm2 of mycelia that was blended with 100 mL of CM for 2 pulses of 20 seconds each on high. The mixture was transferred to 500 mL flasks containing 100 ml CM and shaken for 36-48 hours at 23°C. Mycelia was collected 107 by centrifugation at 3K (4100 x g) in a conical 50 mL Falcon tube. Under aseptic conditions 3 g of the mycelia was resuspended in 10 mL of 1 M sorbitol in a sterile 250 mL flask. Lysing Enzymes (Sigma) at 20 mg/mL were added to the mycelia and incubated at 100 rpm at RT for 60-90 minutes. Protoplasts were harvested by filtration through sterile cheesecloth and Nytex (25 urn pore) (Tetko, lnc). The protoplasts were pelleted by centrifugation at 3 K (4100 x g) for 10 minutes. The pellet was resuspended in 10 mL of 20% sucrose then re-centrifuged and resuspended in 10 mL protoplast buffer (20%o sucrose, 50 mM Tris pH8, and 50 mM CaCb). The protoplasts were counted and resuspended at a concentration of 5xl07/mL in protoplast buffer. For the transformation 200 uL protoplasts were mixed with 1-5 p.g of DNA. This mixture was incubated for 15 minutes and then 1 mL of PEG buffer (40% PEG 8000, 20% sucrose, 50 mM Tris pH8, and 50 mM CaC^) was added and mixed by inversion. The mixture was incubated for 20 minutes before the addition of 3 mL of regeneration buffer (20% sucrose, 1%> glucose, 0.3% yeast extract and 0.3%> casamino acids). This was incubated overnight with shaking. The following morning 10-15 mL of molten (50°C) regeneration agarose (regeneration buffer with 1.7% agarose) was added and the mixture poured onto regeneration agarose plates with 200 ug/mL Hygromycin B. Once solid the plates were incubated at 28°C for 6-8 days. Colonies that had grown out of the overlay were chosen and transferred to Oatmeal agar with 200 pg/mL Hygromycin B. These colonies were allowed to sporulate and single conidia isolations were performed to ensure single nuclear origin. 108 4.2.5 M. grisea DNA and RNA extraction Genomic DNA was extracted from cultures of M. grisea according to the miniprep protocol of Sweigard et al, (1998). For RNA extraction, the transformants were inoculated into 50 mL liquid cultures of CM and incubated for four days on a rotary shaker. The cultures were then centrifuged to concentrate the mycelia, which was flash frozen in liquid nitrogen. Total RNA was extracted by homogenizing frozen tissue in a mortar and pestle combined with acid washed glass beads under liquid nitrogen. RNA was extracted from the ground tissue using the QIAGEN™ RNeasy Plant Extraction kit following the published methodology. The absorbency of the extracts was measured at 260 and 280 nm. The sample was not used unless this ratio 260/280 was greater than or equal to 1.7. 4.2.6 RT-PCR of M. grisea transformants Reverse transcription of transformant mRNA was carried out according to the manufacture's instructions of Superscript II™ RNase H" Reverse Transcriptase (GibcoBRL). First strand cDNA synthesis was completed using 5 pg of total RNA and a T22 oligo. A duplicate aliquot of RNA was treated with RNase A prior to first strand synthesis. This was performed as a control to ensure that future PCR reactions are products of RNA and not of contaminating DNA. PCR amplification was performed using 10% of the first strand synthesis reaction with the T22 oligo and the gene specific oligo 20F (see Table 3.4) as primers. Cycling parameters were 95°C for 5 minutes,, 35 cycles of 94°C for 30 seconds, 50°C for 30 109 seconds and 72°C for 1.5 minutes, followed a 10 minute elongation period at 72°C. A second round of PCR was performed using 2 ul from the first PCR reaction as a template and oligos, 20F and 1301R (see Table 3.4) as primers. The plasmid pBSA4.2H was used as a positive control during this amplification. The same PCR cycling conditions as the first round were applied. The PCR products were resolved by electrophoresis on a 1.5% agarose gel. 4.2.7 Ergosterol Extraction of putative 387N disruptants grown on wood Total fungal biomass of fungi grown on wood blocks for 20 days was estimated by measuring ergosterol (Gao and Breuil, 1993). Method alterations included the substitution of hexane for petroleum ether. The extracts were quantified with a Waters high pressure liquid chromatographic (HPLC) system as previously reported except that a longer column (300 mm) with a slower flow rate of 1.25 mL/min at <2700 psi was used (Gao and Breuil, 1993). Ergosterol samples containing 5-40 ug/mL concentrations of commercially prepared ergosterol dissolved in methanol were run as standards (Sigma). 110 4.3 Results 4.3.1 Construction of gene disruption vector p A N A 4 7 6 To disrupt the THN reductase gene in 387N we first had to construct a disruption vector. We created two oligonucleotides with deviations from wild type sequence and used them as PCR primers with 387N DNA. The primers were designed to amplify the nucleotide sequence of the THN reductase gene from nucleotide positions 1 to 476 (Figure 4.1a). This sequence should include the first 157 codons of the putative N-terminal coenzyme-binding domain of THN reductase. The forward primer, 2DVF1, had a Hindlll site incorporated just before the homology to 5' end of the gene sequence. The reverse primer, 2DVR157 had three nucleotide changes that would be translated as stop codons in any reading frame. This primer also has a 3' BamHI site (Figure 4.Id). Primers 2DVF1 and 2DVR157 were used in a PCR reaction with pBSA4.2 as template DNA (Figure 4.1b). A product with the predicted size of 476 bp was ethanol precipitated, digested with BamHIlHindlll, and gel purified. The fungal transformation vector pAN7-1 (Accession # Z32698), carrying the hygromycin phosphotransferase gene (HPH) with a fungal promoter region, was also digested with BamHIl Hindlll and. the 5973 bp fragment gel purified. These two BamHIl Hindlll fragments were ligated together and transformed into E.coli DH5ct (Figure 4.1c). The disruption vector called pANA476 was isolated and used to transform 387N. I l l A pBSA4.2 THN-R B 2DVF1 476 bp 2DVR157 ;— XXX • 476 pAN7-l Hindill BamHl D 2DVF1 Hindill 5' GGGGTTT AAGCTT ATG TCT CCT GCA ACT GTC AAG GAC G 3' 2DVR571 BamHl 5' GGGGTTT GGATCC ACC GCT GAG CGA CTC AGC TAC TCA TGA GG 3' Figure 4.1. The construction of the disruption vector pANA476. (A) The THN reductase gene in vector pBSA4.2 was used as template DNA with (B) disruption primers 2DVF1 and 2DVR157. The PCR product was digested with BamHl and Hindill and (C) ligated into the BamHI/Hind III site of vector pAN7-l to create pANA476. (D) The sequence of disruption primers showing the restriction sites and the location of the three stop codons as indicated by the asterisks. 112 4.3.2 Characterization of putative 387N disruptants PCR Screening PCR was chosen as the method to determine whether the disruption vector had integrated specifically within the THN gene of 387N. Two hundred and forty transformants were chosen that were resistance to 400 pg/ml hygromycin and displayed altered pigmentation on MEA. However, with this fungus it is difficult to visually determine colour due to variable pigmentation. Of these, eighty displaying reduced pigmentation were chosen for further screening using PCR. When the primers 414F and HPH-2 were used on 80 different putative knockout DNA's, the appearance of a novel product of 1.8 kb would indicate that site specific integration of the disruption vector into the wild type gene had occurred. Two PCR primers similar to portions of the HPH gene called HPH-1 and HPH-2 were also used for PCR with putative knockout DNA as template. A predicted lkb PCR product should indicate the presence of the hygromycin gene of the disruption vector (Figure 4.2). Five putative disruptants were chosen for further analysis because they displayed both the novel 1.8 kb and the expected 1 kb PCR products (Figure 4.3). The presence of the appropriate PCR products could indicate that site specific integration of the disruption vector had occurred at the THN locus but further characterization was required. These putative disruptants were: C5, B25, G27, G29 and G32. 113 pANA476 476 HPH genomic Figure 4.2. The predicted integration event by homologous recombination of the Hindlll linearized pANA476 into genomic 387N D N A (A). The location of PCR primers and the size of the predicted products if (B) occurs as in the model. 114 Figure 4.3. PCR reactions of putative disruptant template D N A with either primer combination (a) 414F/HPH1 or (b) HPH1/HPH2. The arrows to the left indicate size in kilobases. 115 Hybridization studies Verification that gene disruption had occurred would be provided by hybridization studies of genomic DNA from the five putative disruptants hybridized with a portion of the THN gene. The Mbol digested genomic DNA of the five PCR characterized putative disruptants and PstI digested DNA of 42 other transformants were separated and transferred to a membrane. The membrane was hybridized with a 250 bp portion of the THN reductase gene. In the case of the five putative disruptants digested with Mbol, hybridization would indicate that gene disruption had not occurred if a band of 1348 bp (corresponding to Mbol digested wild type THN gene) hybridized with the THN probe. If gene disruption had occurred, one would expect a much smaller band to appear on the autoradiograph. This is assuming that this fungus employs previously characterized mechanisms for site specific integration. In all five putative disruptants only the 1348 bp band was visible (Figure 4.4). The control lane with genomic 387N displayed the same size band (lane 6). Hybridization studies indicate that the five putative disruptants, which were the strongest candidates for a pigmentation deficiency as determined by hygromycin resistance, culture pigmentation and PCR screening, still had an intact copy of the wild type gene. 116 1 2 3 4 5 6 1348 Figure 4.4. Hybridization of the five putative disruptants; C5, B25, G27, G29, G32 and 387N (respectively in lanes 1-6) with a 250 bp portion of the T H N reductase gene. The arrow to the left indicates band size in base pairs. Hybridization studies were conducted on the PstI digested D N A of 42 other transformants (Figure 4.5). If gene disruption had not occurred in the 42 transformants that were PstI digested, a 4.8 kb band would hybridize to the THN gene probe (see lane 4, Figure 3.7A). This band was not present, which might suggest gene disruption was occurring. However, three bands were observed in each transformant. If site specific integration had occurred one would expect to see only one novel band. If random integration had occurred, each transformant would display a unique pattern of integration. However, i f integration were not occurring, the transforming vector would hybridize to 117 the probe. In all cases except for the control, 3 bands of 4 kb, 6.5 kb and 13 kb were visible. These three bands may correspond to the transforming vector. The control was genomic untransformed IC4-1 D N A and no bands were visible (see lanes marked '+' in Figure 4.5). The interpretation of the results of hybridization analysis of 42 other putative transformants is not definitive. One cannot conclude that disruption had occurred. A Figure 4.5. Hybridization of PstI digested genomic D N A from putative disruptants with a 250 bp portion of the THN reductase gene. Photo A are transformants 1-21 and photo B are transformants 22-42. The arrows to the left indicate band size in kilobases. The lanes marked '+' indicate control untransformed IC4-1 DNA. 118 Growth and mitotic stability of putative 387N disruptants In addition to the molecular characterization of the five putative disruptants, it was necessary to define the growth and. pigmentation of the disruptants by other means. These morphological studies were conducted in parallel with the molecular characterizations. The growth and pigmentation of C5, B25, G27, G29 and G32 on MEA plates and on wood blocks was assessed. The colour of the disruptants is described in Table 4.1. Table 4.1. The colour of putative disruptants grown on MEA and wood for 14 days. Strain Co lour on M E A Colour on wood 387N Dark brown Brown/grey B25 White White C5 Very light brown Light brown/grey G27 Light brown Light brown/grey G29 Light brown Light brown/grey G32 Very light brown Light brown/grey The radial growth of the putative disruptants was recorded every 4-9 days over a 21 day period after MEA plates were inoculated with a 7 mm diameter plug of hyphal front mycelia. Hygromycin B was incorporated into the medium at a concentration of 200 pg/ml. Figure 4.6 shows the increase in colony diameter of the five putative transformants grown on MEA with and without hygromycin. After 21 days most of the disruptants had grown to the edge of the petri dish. When 200 pg/ml hygromycin was added to the medium, the wild type nontransformed 387N did not grow and the 119 disruptants grew on average between 20-35 mm in diameter over the 21 day period. Transformant G29 grew more quickly than the other transformants. To assess mitotic stability, hyphal front transfers of the disruptants from MEA plates were conducted for over a period of three months. At monthly intervals the disruptants were transferred to MEA plates with 200 pg/ml hygromycin B and growth rates were compared to baseline data. On plates without hygromycin the growth of the disruptants was comparable to growth of the wild type. After 21 days of growth the diameter of all disruptants was between 60-80 mm. In the presence of hygromycin the wild type did not grow and the diameter of knockout colonies after 21 days of growth was only 15-35 mm. Even though the five putative disruptants were shown not to differ from the wild type according to hybridization studies, they displayed visibly reduced pigmentation when grown on MEA plates and on wood blocks. These transformants also possessed the ability to grow in the presence of hygromycin B, which was a mitotically stable trait detectable even three months after transformation occurred. 120 0 7 14 21 Time (days) Figure 4.6. Colony diameter of the five putative disruptants grown on M E A without (A) and with (B) 200 pg/mL hygromycin. 121 Ergosterol Measurements In addition to assessing the coloration of the five putative disruptants when grown on wood, we also examined how well they were able to utilize wood as a substrate for growth. The five putative disruptants were grown for 20 days on wood blocks. The biomass of 387N can be correlated to the amount of ergosterol extracted from the wood block as detected by HPLC analysis (Gao and Breuil, 1993). However, this correlation may be invalid since transformation could have caused mutations that alter ergosterol production in the disrupted fungi. The albino strain B25 had the most ergosterol followed by C5 and G32 (very light brown) then G29 and G27 (light brown). All of the disruptants had higher levels of ergosterol than 387N (Figure 4.7). 387N G29 C5 B25 G32 G27 Fungal Disruptants Figure 4.7. The growth of the putative disruptants on wood for 20 days, as evaluated by ergosterol yield measured, by HPLC. Data graphed is the mean of three samples and the vertical bars represent standard deviations. Two way t-tests conclude that all pairwise data comparisons differ significantly at 95% confidence values. 122 Co lou r measurements (AE*) In addition to the visual colour assessments, we also measured pigmentation using the Technidyne apparatus to obtain quantitative data. The colour of the freeze-dried fungal mycelium from the putative disruptants was measured according to the methods described in Section 2.2.5 and is graphed in Figure 4.8. All putative disruptants were lighter in colour than the wild type. The putative disruptant B25, was albino in appearance and was the lightest in colour followed closely by G32, C5, G29 and G27. These colour measurements support the visual observations summarized in Table 4.1. It is interesting to note that disruptants lighter in colour had higher ergosterol levels when grown on wood. 123 Darker Lighter °-_X_ 387N G29 C 5 G32 B25 Fungal Disruptants G27 Figure 4.8. The colour (AE*) of dried fungal mycelium after 14 days of growth in NA supplemented with glycerol and asparagine. Data graphed is the mean of three samples and the vertical bars represent standard deviations. Two way t-tests conclude that all pair wise data comparisons differ significantly at 95% confidence values. 124 4.3.3 Construction of vectors for M. grisea complementation It was necessary to find another method of analyzing gene function since we were not able to show that the transformants created by gene disruption were deficient in the THN reductase gene. Complementation of buf mutants of M. grisea with the putative THN reductase gene of 387N was attempted in order to examine gene function. We used two different fungal transformation vectors to carry the complete 387N putative THN reductase gene sequence. A 1.9 kb Xbal fragment from the 4.2 kb Hindlll insert in pBSA4.2 was cloned into the Xbal site of pCB1004 (Carroll et al, 1994) or pAN7-l and named pCB1.9X and pAN1.9X (Figure 4.9). The 1.9 kb Xbal fragment contained the entire sequence of the putative THN reductase gene as shown in Figure 3.15. The two vectors pCB1.9X and pAN1.9X were used to transform the M. grisea buf mutants IC4-1 and IC4-5. Four transformation experiments were conducted from which -200 putative transformants were collected for further characterization. As a control, the parental vectors without inserts, pCB1004 and pAN7-l were used to transform the M. grisea buf mutants IC4-1 and IC4-5. Fifteen control transformants were collected for further characterization. 125 A . p B S A 4 . 2 H THN-R H 1 m I - r — Hindlll Xbal Xbal Hindlll 1.9 kb Figure 4.9. The construction of (B) pCB1.9X and (C) pAN1.9X. (A) shows the 1.9 kb Xbal fragment from pBSA4.2H which was inserted into the Xbal site of either pCB 1004 or pAN7-1. MCS-multiple cloning site, HPH-hygromycin phosphotransferase gene, CAM-chloramphenicol resistance gene, AMP-ampicillin resistance gene. 126 4.3.4 Character izat ion of M.grisea transformants Hygromyc in resistance Transformants from experiments using both parental transformation vectors without THN reductase gene inserts and engineered complementation vectors were selected from media containing 200 pg/mL Hygromycin B. Transformants were transferred to either CM agar for ongoing propagation or liquid CM for DNA extraction. Table 4.2 lists the transformants recovered and indicates which buf mutant was transformed, which vector was used and the colour of mycelia in CM after 7 days. In table 4.2 the ' C before the transformant number designates those transformants selected from control experiments. 127 Table 4.2. Phenotypic features of hygromycin B resistant transformants of Buf mutants grown in liquid culture for 7 days. T r a n s f o r m a n t P a r e n t a l T r a n s f o r m i n g M y c e l i a l C o l o u r s t r a i n V e c t o r i n C M Guy 11 None Brown IC4-1 None Buf IC4-5 None Buf 6 IC4-1 PAN1.9X Brown 15 IC4-1 PAN1.9X Light Brown 4 IC4-1 PCB1.9X Brown 5 IC4-1 PCB1.9X Buf 16 IC4-1 PCB1.9X Brown 17 IC4-1 PCB1.9X Brown 18 IC4-1 PCB1.9X Buf 19 IC4-1 PCB1.9X Buf 20 IC4-1 PCB1.9X Buf 7 IC4-5 PAN1.9X Buf 8 IC4-5 PAN1.9X Buf 21 IC4-5 PAN1.9X Buf 1 IC4-5 PCB1.9X Buf 2 IC4-5 PCB1.9X Buf 3 IC4-5 PCB1.9X Buf 9 IC4-5 PCB1.9X Buf 11 IC4-5 PCB1.9X Buf 12 IC4-5 PCB1.9X Buf 13 IC4-5 PCB1.9X Buf 14 IC4-5 PCB1.9X Buf Cl IC4-1 PAN7-1 Buf C2 IC4-1 PAN7-1 Buf C3 IC4-1 PAN7-1 Buf C4 IC4-1 PAN7-1 Buf C5 IC4-1 PCB1004 Buf C6 IC4-1 PCB1004 Buf C7 IC4-1 PCB1004 Buf C8 IC4-5 PAN7-1 Buf C9 IC4-5 PAN7-1 Buf CIO IC4-5 PAN7-1 Buf C l l IC4-5 PAN7-1 Buf C12 IC4-5 PCB1004 Buf C13 IC4-5 PCB1004 Buf C14 IC4-5 PCB1004 Buf 128 Culture pigmentation In addition to hygromycin B tolerance, the colour of the transformant may be an indication that complementation had occurred. The colour of the transformants grown in liquid CM was recorded and is summarized in Table 4.2. Only transformants 4, 6, 16 and 17 had wild type pigmentation. Transformant 15 did not resemble the wild type but was darker than the parental buf strains. The remainder of the transformants including the controls were identical in colour to the non-transformed buf strains IC4-1 or IC4-5. The colour of mycelia from liquid cultures of transformants 4, 15, 16 and 17 as well as the wild type Guy 11 and the buf mutant IC4-1 is shown in Figure 4.10. The colour of mycelia from liquid cultures of the control transformants is shown in Figure 4.11. The resemblance of transformants 4, 15, 16 and 17 to wild type or near wild type suggests that complementation may have occurred. 129 15 16 IC4-1 Guy 11 Figure 4.10. The colour of liquid cultures of Guy 11 and buf 'mutant IC4-1 and the transformants 4, 15, 16, and 17. IC4-1 C l C2 C3 C4 C5 C6 C7 A IC4-5 C8 C9 CIO C l l C12 C13 C14 HUB 41 K M B H Figure 4.11. The colour of liquid cultures of the control transformants. The headings correspond to those in Table 4.2. Photograph (A) are IC4-1 parental control transforments while (B) are IC4-5 parental control transformants. 130 Hybridization studies In order to determine if the complementation vectors were present in the transformant genome we performed hybridization studies. The blot in Figure 4.12 was performed to determine if buf parental strains had any sequence similarity to the 387N THN reductase gene sequence. Genomic DNA from IC4-1, IC4-5, Guy 11 and 387N were digested with PstI or PvuII, separated and transferred to a membrane. The membrane was hybridized to a 548 bp EcoRV/Mbol fragment from within the 1.9 kb Xbal fragment of the THN reductase gene. A 4.8 kb and 4 kb band were seen in the control, but no bands were visible in either IC4-1 or IC4-5. One band of 4 kb appears in the PvuII lane of Guy 11. The results of this experiment suggest that both of the buf parental strains lack similar 387N THN reductase gene sequences while these sequences may be present in Guy 11. Genomic DNA from the transformants listed in Table 4.2, was BamHI digested, separated and transferred to a membrane. The membrane was hybridized with the radiolabeled 548 bp EcoRV/Mbol fragment from within the 1.9 kb Xbal fragment of the 387N THN reductase gene (Figure 4.13). With the exception of the two parental controls, non-transformed IC4-1 (lane 1) and non-transformed IC4-5 (lane 22), from one to eight bands can be seen in each lane. Transformants 4 and 16 appear to have identical patterns of integration. There are also common bands visible in many transformants; a 2 kb band is shared by transformants 4, 16, and 21; a 3.5 kb band is visible in transformants 4, 16 and 2; transformants 17 and 2 share a 4.2 kb band and transformants 4, 16, 1 and 11 share a 6.5 kb band. These data indicate that randomly integrated complementation vector 131 containing the 387N THN reductase gene is present in the genomes of all the transformants (excluding control transformants). 387N IC4-1 IC4-5 Guy 11 a b a b a b a b 4.8 4.0 i§;: Figure 4.12. Hybridization of (a) PstI or (b) PvuII digested genomic 3 87N, IC4-1, IC4-5 and Guy 11 with a portion of the 387N THN reductase gene. Arrows to the left indicate band size in kilobases. 132 Figure 4.13. M. grisea buf mutants transformed with pCB 1.9X or pAN 1.9X and hybridized with a portion of the 387N THN reductase gene. Lanes 4-20 are IC4-1 in origin and lanes 14-21 are IC4-5 in origin. Lanes are numbered with transformant designation number listed in Table 4.2. Buf strain IC4-1 in the lane on the far left and IC4-5 is in the lane on the far right. Transformants 15, 6, 6, 7, and 21 are those transformed with pAN1.9X and the remainder are transformed with pCB1.9X. Arrows on the right indicate band size in kilobases. 133 RNA transcripts Hybridization studies suggest that the THN reductase gene is present in the genomes of the transformants. It was now important to determine if the THN reductase gene of 387N was being expressed in M. grisea transformants. RT-PCR was used to determine if mRNA transcripts of the 387N THN reductase gene were present in the M. grisea transformants. Figures 4.14 'a' and 'b' show first round PCR products of cDNA template (+/- RNase A treatment) with 841F/T22 primers. The gel photo shown in Figure 4.14a is the positive control (RNase A treated) designed to indicate that contaminating DNA, which could potentially serve as PCR template, was not present. In this photo, a smear of product is seen of approximately 0.7-1 kb in the lanes containing control IC4-1, transformants 4,15 and 16 from the RNase A treated cDNA template. The first lane is a control without cDNA template. In Figure 4.14b, a continuous smear is seen in all the lanes with the exception of the negative control lane without RNA. The products of the second round of PCR, using the first round products as template with 841F/1301R oligos as primers, are shown in Figures 4.15 'a' and 'b'. In the RNase A treated samples (Figure 4.15a) no product is seen except in the PCR control lane 7 (pBSA4.2) indicating that only RNA (not DNA) was available as template for amplification. In the untreated sample (without RNase A, Figure 4.15b) , a 443bp product is visible in the lanes with transformants 4, 15, 16, and 17 but not in lane with untransformed IC4-1. This 443 bp PCR product was purified and digested with Aatll and BstUI. The products of this restriction are shown in Figure 4.16. The schematic shown 134 in Figure 4.17 illustrates the predicted product size of such a digestion. All three predicted products are seen in each of the transformants and in the 387N positive control. Using RT-PCR we were able to detect the presence of 387N TITN reductase transcripts in the transformants 4, 15, 16 and 17. 135 A M 1 2 3 4 5 6 7 1 2 3 4 5 6 7 M Figure 4.14. PCR amplification with T22 and 84IF primers using cDNA as template. In gel A , the RNA was treated with RNase A prior to first strand synthesis. Lane ' M ' is 100 bp D N A ladder. Lane 1 is a control without RNA, lane 2 is IC4-1, lane 3 is transformant 4, lane 4 is transformant 15, lane 5 is transformant 16, lane 6 is transformant 17 and lane 7 is control R N A from 387N. Gel B shows PCR products that were not subjected to RNase A treatment. Lane 1 is IC4-1, lane 2 is transformant 4, lane 3 is transformant 15, lane 4 is transformant 16, lane 5 is transformant 17 and lane 6 is control R N A from 387N and lane 7 is control without RNA. Lane ' M ' is a kilobase D N A ladder. Arrows to the left and right indicate band sizes in kilobases. 136 A B 1 2 3 4 5 6 7 M M 1 2 3 4 5 6 7 0.5 - W Figure 4.15. Second round PCR amplification of 841F and 1301R primers using first round PCR products as template. In gel A , the RNA was treated with RNase A prior to first strand synthesis. Lane 1 is IC4-1, lane 2 is transformant 4, lane 3 is transformant 15, lane 4 is transformant 16, lane 5 is transformant 17, lane 6 is 387N and lane 7 was the product of pBSA4.2 as template DNA. Lane ' M ' is 1 kilobase marker D N A ladder. Gel B was not treated with RNase A. Lane ' M ' is 1 kilobase marker D N A ladder. Lane 1 is IC4-1, lane 2 is transformant 4, lane 3 is transformant 15, lane 4 is transformant 16, lane 5 is transformant 17, lane 6 is 387N and lane 7 was the product of pBSA4.2 as template DNA. The arrows on the left and right indicate marker D N A of 0.5 kilobases. 137 M IC4-1 4 15 16 17 387N a b a b a b a b a b a b m ^ ^ ^ ^ Mitttt i i i i i ^ f : i H P <sgaae* w w 1000 • 500 • 396 • 220 154 • Figure 4.16. Aatl/BstUI digestion of the 443 bp PCR product shown in Figure 4.14. Arrows indicate band sizes in basepairs. 841F 1301R 443 bp ^ ~ 1 1 Aatll BstUl 146 bp 193 bp 104 bp Figure 4.17 The Aatll and BstUI digestion products of the 443 bp PCR product shown in Figure 4.14. 138 4.4 Discussion 4.4.1 THN reductase gene disruption in O. floccosum 387N A transformation system has been developed for the Ophiostoma species that causes Dutch elm disease (Royer et al, 1991). In O. ulmi, it has been reported that heterologous transforming vectors randomly integrate while vectors containing some homologous sequences can direct homologous recombination (Royer et al, 1991). This transformation system used for O. ulmi was applied to O. floccosum 387N. A gene disruption vector was constructed using a portion of the sequence that codes for the first 157 codons of the THN reductase gene. We engineered three stop codons, one in each frame, into the sequence. This vector was transformed into the wild type 387N and disruptants were screened for reduced pigmentation. The transformants with reduced pigmentation would theoretically have integrated the transforming DNA into the reductase gene and this could result in the disruption of reductase function. Complementation of the transformants with an intact copy of the reductase gene could result in the restoration of pigmentation and would be evidence to support the hypothesis that the cloned gene has a THN reductase function. This disruption attempt was probably premature for several reasons. Homologous integration of exogenous DNA has not been shown to occur in all filamentous fungi (Shiotani and Tsuge, 1995). Secondly, data from Chapter 3 suggests that in this organism there may exist two THN reductase-like sequences. If both reductase-like sequences function in the DHN melanin pathway, then disruption of only one of these sequences may not have resulted in the recovery of transformants with the expected phenotype. 139 Six transformation attempts, using the gene disruption vector pANA476, were conducted in 387N. The selection of appropriate pigmentation deficient mutants was difficult since the pigmentation of the putative disruptants was highly variable. Blue to green polyketide-derived pigments have been isolated from the conidia of Penicillium sp., Aspergillus sp., and Trichoderma sp. (Wheeler and Klich, 1995). It is possible that non-melanin polyketide-derived pigments exist in 387N and could contribute to the observed variability. The primary screening of putative disruptants was conducted using PCR. The appearance of a novel 1.8 kb PCR product should indicate the presence of a site specific integration event. A 1.8 kb product was detected in five of the disruptants; C5, B25, G27, G29, and G32. However, additional PCR amplification products of different sizes were also present. This suggests that perhaps genomic rearrangements other than site specific integration events were occurring during transformation that may or may not have involved the other THN reductase-like sequence. The method of integration of the transforming DNA can occur by several mechanisms and some fungi appear to lack the machinery for targeted integration (Shiotani and Tsuge, 1995). Hybridization studies were used to screen the five putative disruptants. In Mbol digested wild type 387N genomic DNA, a 1340 bp band will hybridize with the THN gene. Only one band of 1340 bp was visible in the putative disruptants, which suggested that site specific integration was not occurring. The apparent lack of site specific integration within the five PCR positive transformants is not unexpected since this type of recombination does not necessarily occur in all filamentous fungi. A colleague has attempted to disrupt the scytalone dehydratase gene of 387N. After screening -2000 140 transformants he was not able to demonstrate that homologous integration was occurring (H. Wang, personal communication). However, if transformation was occurring one would expect to at least observe random integration and this was not the case for these five disruptants. Random integration was observed during transformation of this fungus by another researcher (H. Wang, personal communication). Three bands of 4, 6.5 and 13 kb were present on the hybridization membranes of forty two other putative disruptants which were examined. An explanation for the presence of these three bands could be that re-ligation of the Hindill linearized plasmid during or after transformation was occurring. The PstI digestion of complete circular pANA476 would produce a fragment of 3967 bp that would contain the THN gene fragment and could correspond to the 4 kb band. Incomplete PstI digestion would leave pANA476 intact as a 6463 bp plasmid that could correspond to the 6.5 kb band and the re-ligation of the plasmid could also produce concatemers which could account for two plasmids being ligated and correspond to the 13 kb band. The presence of these three bands and the absence of other bands could indicate that integration was not occurring. This explanation requires further investigation. The apparent lack of a 4.8 kb band representing the wild type THN reductase gene is interesting since ordinarily this would suggest that disruption had occurred. Conflicting interpretations for the data from these hybridization studies suggests that a more extensive investigation into the transformation of this particular fungus, particularly regarding the fate of transformed DNA, is warranted. The interpretation of molecular data suggests that integration of the gene disruption vectors was not occurring in the five putative disruptants. However, the results of the 141 phenotypic characterization of the putative disruptants suggests that they are significantly different from the control. We found that the five putative disruptants displayed a heritable resistance to hygromycin B even after 3 months in culture without the antibiotic. Compared to the control the putative disruptants had reduced pigmentation and higher ergosterol levels when grown on wood. A possible explanation for these differences may be that the transformation event, while not appearing to affect the THN reductase gene directly, may have caused mutations in other genes. Perhaps the genes affected are those involved in the regulation of melanin biosynthesis or the production of ergosterol. Further molecular characterization is required to explain our observations. Since we could not recover a 387N mutant deficient in the THN reductase it was necessary to explore gene function using other techniques. 4.4.2 Complementation of M. grisea buf mutants wi th the 387N putative T H N reductase gene Complementation of M. grisea THN reductase deficient buf mutants with the putative THN reductase gene of O. floccosum 387N was attempted in an effort to prove function. Since the buf mutants of M. grisea used in this study have a 50 kb deletion of the THN reductase locus, the presence of the THN reductase gene and gene transcripts should indicate that transformation and expression of the exogenous gene was occurring. If wild type pigmentation was also observed in the transformed buf mutants, it should indicate that complementation had occurred, and the transforming DNA had THN reductase activity. It has been shown that integration does occur during the transformation of M. grisea with non-homologous DNA (Leung et al, 1990). 142 M. grisea transformants were selected if they were tolerant to high concentrations of hygromycin and exhibited wild type pigmentation. Visually different degrees of pigmentation were observed between restored buf transformant 15 and transformants 4, 16 and 17. The recovery of transformants with different degrees of pigmentation are in agreement with other reported complementation experiments where the colour of transformants ranged from wild type appearance to lighter than wild type (Kawamura et al, 1997; Takano et al, 1997a). There are several possible explanations for the different degrees of pigmentation observed between transformants. It has been reported by other researchers that copy number and integration location can influence expression (Kubo et al, 1991; Leung et al, 1990). We found that transformant 15 had only one integrated copy of the 387N THN reductase gene while the other transformants had between five and seven integrations. Copy number could account for our observations. Variability in pigmentation could also be attributed to interspecific differences between cis-acting elements of 387N gene and trans-acting regulators of M. grisea. Takano et al., (1997a) found that when the A. alternata ALM gene was used to complement PKS deficient C. lagenarium, lighter than wild type transformants were recovered. Disparity between cis elements and trans-acting regulators could lead to alterations in the temporal pattern of transcription. In wild type C. lagenarium transcripts of the PKS gene are not detectable until after conidial germination had begun but in melanin restored transformants, the A L M transcript was detected at low levels at all times (Takano et al, 1997a). One other explanation for the differences in pigmentation could be due to interspecific differences in the activity of the enzyme and its cellular location. 143 When using the A. alternata DHN genes to complement M. grisea, the frequency of melanin restored buf transformants compared with hygromycin resistant transformants was 25% (Kawamura et al, 1997). A similar low frequency of restored transformants (30%) was observed when the 387N THN reductase gene was transformed into M. grisea buf mutants. Transforming DNA may undergo rearrangements such as double stranded breaks or recombination within the gene sequence during integration which results in the disruption of function of the transforming DNA (Parsons et al, 1987). DNA hybridization data from Kawamura et al., indicated that M. grisea transformants which did not display restored pigmentation had only partial fragments of the transforming A. alternata gene present (Kawamura et al, 1997). It is possible that rearrangements during integration are responsible for the low frequency of melanin restored transformants which were recovered during our experiments. DNA hybridization was used to check the transformants for the presence of the complementation vector. Every one of the transformants tested by hybridization displayed at least one band indicating the presence of one integrated copy of the 387N THN gene. The variance between band intensities within each transformant suggests that tandem integrations may be occurring. Transformants 4 and 16 appear to have identical patterns of multiple integration while transformant 17 was unique. Shared integration patterns suggest that there may have been either some homology with the vector or there were hotspots within the genome that were more susceptible to integration. In addition to hybridization studies used to detect the presence of the gene, RT-PCR was used to detect the presence of THN reductase mRNA transcripts that would indicate expression of the 387N THN reductase gene in melanin restored transformants. Four of 144 the buf M. grisea 387N THN reductase restored transformants, 4, 15, 16 and 17, were tested for the presence of THN reductase RNA transcripts and were positive. The RT-PCR amplification products when digested with Aatll/BstUl produced fragments of sizes appropriate to the presence of 387N gene sequence. Evidence supports that the O floccosum 387N THN reductase gene is present in the genome of transformed M: grisea buf mutants and is expressed in those transformants and that the gene can restore pigmentation to transformed M. grisea buf mutants. Therefore one can conclude that the gene isolated from 387N functions as a THN reductase in M. grisea by restoring mycelial melanization and it is possible that the gene may have the same function in O. floccosum. We did not investigate whether pigmentation of M. grisea appressoria and consequently pathogenicity had been restored in the transformed buf mutants. Historically a correlation between visual pigmentation and biological function of melanin has been repeatedly demonstrated either by the application of DHN intermediates or by complementation studies (Chida and Sisler, 1987; Chumley and Valent, 1990; Kawamura et al., 1997; Woloshuk et al., 1980; Woloshuk et al., 1983). Melanin has different roles in different fungi. In A. alternata, melanin is found in the conidia and functions.in the longevity and survival of the organism (Kawamura et al, 1999). This is in contrast to the function of melanin in M. grisea and C. lagenarium where the pigmentation occurs in appressoria and is required for pathogenicity. In O. piliferum, melanin is required for the development of perithecia (Zimmerman et al, 1995). Melanin may play a similar role in 387N. 145 Despite the different roles of melanin, it has been shown that A. alternata DHN melanin biosynthetic genes can restore pigmentation in M. grisea and C. lagenarium melanin deficient mutants. Our results suggest that a similar situation has occurred where a O. floccosum THN reductase gene can restore function in M. grisea. Even though the genes share similarity, there appear to be differences in transcription factors or the regulation of gene expression between fungi. Since one sequence from 387N was enough to restore pigmentation to M. grisea mutants, perhaps only one reductase is required for DHN melanin biosynthesis in M. grisea and 387N. 146 4.5 Conclusions The purpose discussed in this chapter was to explore the function of the THN reductase gene from O. floccosum 387N. We were not able to investigate the function of this gene in 387N but we were able to restore pigmentation in the mycelia of THN reductase deficient buf mutants of M. grisea. Complementation studies with the 387N.THN reductase gene and M. grisea IC4-1 buf mutants resulted in the recovery of four hygromycin resistant transformants that had close to wild type pigmentation phenotype. When tested by hybridization, the transformants were shown to have from one to eight copies of the 387N THN gene integrated in either single or tandem fashion. Using RT-PCR we were able to confirm that 387N THN reductase gene transcripts were present and had a restriction digestion pattern which corresponded to the 387N gene. We were able to restore pigmentation in buf mutants of M. grisea using the THN reductase gene from 387N. The presence of the THN reductase gene and transcripts in the transformants, which lack an endogenous copy of this gene, suggest that the 387N gene may have THN reductase function. 147 C H A P T E R 5 Conclusions and future directions Based on the experimental results and discussions presented in the four preceding chapters, it is now of interest to return to the original goals that motivated this research project. The purpose of this chapter is to first briefly summarize the results from each preceding chapter and consider the extent to which the original thesis goals have been fulfilled, the significance of the outcome and how that information may be applied. Secondly, this chapter also presents a short discussion of the obstacles that arose during the course of this investigation that could present future research challenges. In Chapter 1, the industrial and economic significance of the sapstaining problem was summarized and a review of the fungi responsible for causing sapstain was provided. It was found that Ophiostoma sp., such as O. floccosum 387N, are the frequently isolated casual fungi in British Columbia. However, detailed knowledge regarding lifecycle, physiology, ecology, biochemistry and genetics of this fungus is lacking. The role and production of melanin in different fungi was also discussed as well as the DHN pathway of melanin production and its importance to many ascomycetes. It became apparent that these subjects need to be addressed in 387N and that molecular biological studies could provide this information. Nutrient studies designed to explore what conditions affected pigmentation and how best to quantify the presence of melanin were addressed in Chapter 2. Our results indicate that 148 387N grew and pigmented differently on the variety of substrates tested. We also found that high biomass yields and intense pigmentation resulted when 387N was when grown in the presence of glycerol and asparagine, which are similar to substrates available in wood. Just precisely what is the biological significance of different degrees of pigmentation intensity? In O. floccosum like O. piliferum, pigmentation in perithecia may be necessary for reproduction. However, without exhaustive knowledge of the ecology of the fungus it would be difficult to assess the particular benefit of each incremental variation in pigmentation. Is there any connection between the types of nutrients utilized and the intensity of pigmentation? Limitations of primary metabolites are thought to trigger secondary metabolite production. It is possible that unique combinations of nutrients and therefore limitations could influence pigment production in a very specific manner. A portion of this chapter also dealt with whether the DHN pathway was active in 387N. Most basidiomycetes and oomycetes employ the DOPA, GHB or catechol pathways for fungal melanin biosynthesis. So far DHN melanin has only been reported to be associated with ascomycetes with one exception, the black yeast Phaeococcomyces sp., a basidiomycete (Butler et al., 1988). The results of the application of DHN inhibitors led us to speculate that the DHN pathway of melanin production is important to 387N and that the pathway may be similar to that characterized in other filamentous fungi. Hybridization studies confirmed that sequences similar to A. alternata DHN genes were present in the genome of 387N and this lends support to our hypothesis. 149 The objective of the research presented in Chapter 3 was to isolate and characterize a gene involved in DHN melanin biosynthesis. This was accomplished and we recovered a gene with an 877 nucleotide open reading frame that codes for a 268 amino acid that shared similarity with several other fungal THN reductases. It is interesting that a portion of a second THN reductase-like sequence was detected. More molecular experimentation is necessary to isolate, characterize and determine the role of the second gene. It is possible that in 387N melanin biosynthesis is more complex than in other ascomycetes or that the second reductase-like sequence may function in the production of other polyketide derived compounds. The final objective of this thesis was to determine the function of the gene that had been cloned and characterized in Chapter 3. Two different approaches that address this goal were attempted in Chapter 4. Gene disruption was used to create pigment deficient mutants that could be used to explore the function of the putative THN reductase. Two major obstacles were encountered. Variable pigmentation patterns in the transformants made the identification of disruptants difficult. Sectoring and pigmentation changes between hyphal front transfers were frequently observed. Perhaps other isolates of this species also need to be examined and strains procured that display more stable patterns of pigmentation. This would aid in the recovery of mutants that are obligatory for the study of any gene. There is a possibility that other polyketide-derived pigments could be the cause of the observed variation in which case further investigation of pigmentation in 387N is warrented. Secondly, the observation that hybridization results from the putative disruptants were inconclusive with respect to the fate of the transforming DNA. Many 150 questions arose during our gene disruption experiments reflecting that our knowledge of this fungus is very limited and perhaps these experiments were premature. What was happening during transformation in this fungus? Was integration even occurring? Were rearrangements occurring? Was the transforming DNA able to exist as an exogenously replicating plasmid? Does the form of transforming DNA affect the uptake or integration (circular, linear, single stranded vs. double stranded breaks)? What extent of homology is required for specific integration? And finally, is gene disruption possible is 387N? All of these questions will have to be addressed in order to ascertain whether gene disruption could prove a useful molecular tool for this organism since mutants are important for determining gene function. The later part of Chapter 4 deals with the complementation of THN reductase deficient mutants of M. grisea. Transformation of this fungus with the 387N THN reductase gene led to the recovery of several transformants with wild type phenotype. Transforming DNA was found to be integrated into the genome of the restored transformants and expression of the gene was detected by RT-PCR. Though the role of melanin differs in different fungi, these experiments provide a third example where the genes required for the production of melanin from one fungus were used to restore function in a recipient fungal host. Although restoration is not equivalent to wild type, these studies nevertheless demonstrate the common origins of the DHN genes and the importance of their function in many ascomycetes. In 387N, the function of melanin has not been defined, but melanin may lend structural support to perithecia. Experiments designed to 151 examine the regulation of melanin production could provide clues to the function of melanin in 387N. The appendix documents the attempt to create and characterize pigment deficient mutants of 387N created by UV mutagenesis. The purpose was to create mutants that could then be transformed with the characterized THN reductase gene and screened for restoration of pigmentation. Similar to our gene disruption studies, variable pigmentation in 387N made it difficult to unequivocally characterize the mutants. We found that GC proved to be a practical method for characterizing DHN intermediates that accumulate in pigment deficient mutants. The availability of additional DHN intermediate shunt products and perhaps comparable mutants of another well-characterized fungus would be helpful to the analysis of culture extractives. We recovered one potential 387N pigmentation mutant that could possibly be deficient in the THN reductase gene as indicated by the intermediates that accumulated. However, further molecular characterization would be required to establish that the THN reductase gene was affected. Despite difficulties, molecular biological techniques were applied to O. floccosum 387N. These tools were used to clone and characterize a THN reductase gene from 387N. The gene was then used to complement THN reductase deficient mutants of M. grisea. It was found that 387N THN reductase "gene could restore pigmentation in buf mutants of M. grisea. Now that one THN reductase gene has been cloned and characterized it would be of interest to investigate the other sequence that was detected. If gene disruption can be shown to operate in this fungus, it will be necessary to repeat early experiments and in addition to investigate the effects of dual THN reductase gene disruptions. It would be 152 interesting to study the expression and .regulation of the THN reductase genes in 387N and elucidate subtle factors involved in the regulation of gene expression. A comparison of how promoter signals or transcription factors differ between the pigment producing fungi that have different cellular locations and roles for melanin could contribute knowledge to gene regulation and insights into taxonomic relationships. More detailed information regarding pigment regulation in 387N could also be applied to controlling the growth of this organism on wood and limiting its cost to the lumber industry. This work is significant because we were able to transfer molecular fungal techniques to a previously unstudied organism that is a principal pest of the forestry industry. The data obtained provided many details about the DHN melanin biosynthetic pathway in this organism. The results indicate that future research should focus on an earlier step in the pathway if biocontrol applications are of interest. The characterization of how 387N responded to several DHN inhibitors provided practical knowledge that could be applied while investigating the potential application of other compounds as sapstain control agents in the lumber industry. Future experimentation and the refinement of some of these techniques could lead to their routine application with the possible result of discovering very precise methods to control fungal activity in lumber. For example, these techniques could now be used to characterize the nature of the albino mutation of O. piliferum. This knowledge could lead to the registration of this organism as an agent for biocontrol. 153 L I T E R A T U R E C I T E D Abraham, L., Roth, A., Saddler, J. and Breuil, C. 1993. Growth, nutrition, and proteolytic activity of the sap-staining fungus Ophiostoma piceae. Canadian Journal of Botany 71: 1224-1230. Altschul, S., Gish, W., Miller, W., Myers, E. and Lipman, D. 1990. Basic local alignment search tool. Journal of Molecular Biology 215: 403-410. Andersson, A., Jordan, D., Schneider, G. and Lindqvist, Y. 1997'a. Crystaline structure of the ternary complex of 1,3,8-trihydroxynaphthalene reductase from Magnaporthe grisea with NADPH and an active site inhibitor. Structure 4: 1161-1170. Andersson, A., Jordan, D., Schneider, G. and Lindqvist, Y. 1997b. A flexible lid controls access to the active site in 1,3,8-trihydroxynaphthalene reductase. FEBS Letters 400: 173-176. Andersson, A., Jordan, D., Schneider, G., Valent, B. and Lindqvist, Y. 1996. Crystallization and preliminary X-ray diffraction study of 1,3,8-trihydroxynaphthalene reductase from Magnaporthe grisea. Proteins Structure, Funtion and Genetics 24: 525-527. Appleyard, V., Unkles, S., Legg, M. and Kinghorn, J. 1995. Secondary metabolite production in filamentous fungi displayed. Molecular and General Genetics 247: 338-342. Ausubel, F., Brent, R., Kingston, R„ Moore, D., Seidman, J., Smith, J. and Struhl, K., Eds. 1987. Current Protocols in Molecular Biology. Boston, USA. John Wiley and Sons. Ballard, R. and Walsh, M. 1982. Blue stain fungi in xylem of lodgepole pine: a light microscope study on extent of hyphal distribution. Canadian Journal of Botany 60: 2334-2341. Baltz, R. and Hosted, T. 1996. Molecular genetic methods for improving secondary-metabolite production in actinomycetes. Trends in Biotechnology Techniques 14: 245-250. Bell, A., Puhalla, J., Tolmsoff, W. and Stipanovic, R. 1976a. Use of mutants to establish (+)-scytalone as an intermediate in melanin biosynthesis by Verticillium dahliae. Canadian Journal of Microbiology 22: 787-799. Bell, A., Stipanovic, R. and Puhalla, J. 1976b. Pentaketide metabolites of Verticillium dahliae. Tetrahedron 32: 1353-1356. 154 Bell, A. and Wheeler, M. 1986. Biosynthesis and functions of fungal melanins. Annual Review of Phytopathology 24: 411-451. Bennett, J. 1995. From molecular genetics and secondary metabolism to molecular metabolites and secondary genetics. Canadian Journal of Botany 73(suppl.l): 917-924. Bhatnagar, D., Ehrlich, C. and Cleveland, T. 1992. Oxidation reduction reactions in biosynthesis of secondary metabolites. Handbook of applied mycology-Mycotoxins in ecological systems. Eds, Bhatnagar, D., Lillehoj, E. B. and Arorz, D. K. . Marcel Dekker Inc. 5: 255-286. Blanchette, R., Farrell, R., Burnes, T., Wendler, P., Zimmermman, W., Brush, T. and Snyder, R. 1992. Biological control of pitch in pulp and paper production by Ophiostomapiliferum. Tappi Journal (Dec): 102-106. Bloomfield, B. and Alexander, M. 1967. Melanins and resistance of fungi to lysis. Journal of Bacteriology 93(4): 1276-1280. Bouchard, B., Vijayasaradhi, S. and Houghton, A. 1994. Production and characterization of antibodies against human tyrosinase. Journal of Investigative Dermatology 102(3): 291-295. Braiser, C. 1978. Mites and reproduction in Ceratocystis ulmi and other fungi. Transcripts of British Mycology Society 70: 81-89. Bridges, J. and Moser, J. 1983. Role of two phoretic mites in transmission of bluestain fungus, Ceratocystis minor. Ecological Entomology 8: 9-12. Bridges, J. and Moser, J. 1986. Relationship of phoretic mites (Acari: Tarsonemidae) to the bluestaining fungus, Ceratocystis minor in trees infested by southern pine beetle (Coleoptera: Scolitidae). Environ. Entomol. 15: 951-953. Brisson, A., Gharibian, S., Eagen, R., Leclerc, D. and Breuil, C. 1996. Localization and characterization of the melanin granules produced by the sap-staining fungus Ophiostoma piceae. Material und Organismen 30(1): 23-32. Bristow, J. 1994. What is ISO brightness? Tappi Journal 77: 174-178. Bruchez, J., Eberle, J. and Russo, V. 1993a. Regulatory sequences in the transcritption of Neurospora crassa genes: CAAT box, TATA box, introns, poly(A) tail formation sequences. Fungal Genetics Newsletter 40: 89-96. Bruchez, J., Eberle, J. and Russo, V. 1993b. Regulatory sequences involved in the translation of Neurospora crassa mRNA: Kozak sequences and stop codons. Fungal Genetics Newsletter 40: 85-88. 155 Bu'lock, J. 1975. Secondary metabolism in fungi and its relationships to growth and development. The filamentous fungi-industrial mycology. Eds, Smith, J., E and Berry, D.,R Butler, M., Lazarovits, G., Higgins, V. and Lachance, M. 1988. Partial purification and characterization of a dehydratase associated with the pentaketide melanogenesis pathway of Phaeococcomyces sp. and other fungi. Experimental Mycology 12: 367-376.London. Edward Arnold. 1: 33-57. Butler, M. and Day, A. 1998. Fungal melanins : a review. Canadian Journal of Microbiology AA: 1115-1136. Carroll, A., Sweigard, J. and Valent, B. 1994. Improved vectors for selecting resistance to hygromycin. Fungal Genetics Newsletter 41: 22. Chaskes, S. and Tyndall, R. 1979. Pigmentation and autofluorescence of Cryptococcus species after growth on tryptophan and anthranilic acid media. Mycopathologia 64(2): 105-112. Chen, M. and Johns, M. 1993. Effect of pH and nitrogen source on pigment production by Monascus purpureus. Applied Microbiology and Biotechnology 40: 132-138. Chida, T. and Sisler, H. 1987. Restoration of appressorial penetration ability by melanin precursors in Pyricularia oryzae treated with antipenetrants and in melanin-deficient mutants. Journal of Pesticide Science 12: 49-55. Chumley, F. and Valent, B. 1990. Genetic analysis of melanin deficient, nonpathogenic mutants of Magnoporthe grisea. Molecular Plant Microbe interactions 3(3): 135-143. COFI. 1997. British Columbia Forest Industry Fact Book-1997. Vancouver. Council of Forest Industries. Dawley, R. and Flurkey, W. 1993. Differentiation of tyrosinase and laccase using 4 hexyl resorcinol, a tyrosinase inhibitor. Phytochemistry 33(2): 281-284. deGroot, M., Bundock, P., Hooykaas, P. and Beijersbergen, A. 1998. Agrobacterium tumefaciens mediated transformation of filamentous fungi. Nature Biotechnology 16(September): 839-842. Dewar, K. and Bernier, L. 1993. Electrophoretic karyotypes of the elm tree pathogen Ophiostoma ulmi (sensu lato). Molecular and General Genetics 238: 43-48. Dowding, P. 1969. The dispersal arid survival of spores of fungi causing bluestain in pine. Transcripts of British Mycology Society 52(1): 125-137. Drew, S. and Demain, A. 1977. Effect of primary metabolites on secondary metabolism. Ann. Rev. Microbiol. 31: 343-356. 156 Eagen, R., Brisson, A. and Breuil, C. 1997a. The sapstaining fungus Ophiostoma piceae synthesizes different types of melanin in different growth media. Canadian Journal of Microbiology 43: 592-595. Eagen, R., Kronstad, J. and Breuil, C. 1997b. Isolation of a gene from the melanin pathway of the sapstaining fungi Ophiostoma piceae using PCR. International Research Group on Wood Preservation 10219. Faure, D., Bouillant, M. and Bally, R. 1994. Isolation of Azospirillum lipoferum 4T Tn5 mutants affected in melanization and laccase activity. Applied and Environmental Microbiology 60(9): 3413-3415. Fincham, J. 1989. Transformation in fungi. Microbiological Reviews 53(1): 148-170. Findlay, W. 1959. Sap stain of timber. Forestry Abstracts 20(1,2): 1-173. Fogarty, R. and Tobin, J. 1996. Fungal melanins and their interactions with metals. Enzyme and Microbial Technology 19: 311-317. Froyd, J., Paget, C , Guse, L., Dreikom, B. and Pafford, J. 1976. Tricyclazole: A new systemic fungicide for control of Pyricularia oryzae on rice. Phytopathology 66: 1135-1139. Gadd, G. 1980. Melanin production and differentiation in batch cultures of the polymorphic fungus Aureobasidiumpullulans. FEMS microbiology letters 9: 237-240. Gadd, G. 1982. Effects of media composition and light on colony differentiationa and melanin synthesis in Microdochium bolleyi. Transcripts of British Mycology Society 78(1): 115-122. Gafoor, A. and Heale, J. 1971. Melanin formation and peroxidase activity in Verticillium. Microbios 3(10-11): 87-95. Gao, Y. and Breuil, C. 1993. Ergosterol - A measure of fungal growth in wood for staining and pitch control fungi. Biotechnology Techniques 7(9): 621-626. Gao, Y., Breuil, C. and Chen, T. 1993. Utilization of triglycerides, fatty acids and resin acids in lodgepole pine wood by a sapstaining Ophiostoma piceae. Material und Organismen 28(2): 105-118. Gao, Y. and Breuil, C. 1995a. Extracellular lipase production by a sapwood-staining fungus, Ophiostoma piceae. World Journal of Microbiology and Biotechnology 11: 638-642. Gao, Y. and Breuil, C. 1995b. Wood extractives as carbon sources for staining fungi in the sapwood of lodgepole pine and trembling aspen. International Research Group on Wood Preservation 26: 2-26. 157 Gibbs, J. 1993. The biology of ophiostomatiod fungi causing sapstain in trees and freshly cut logs. Ceratosystis and Ophiostoma: Taxonomy, ecology and pathogenicity. Eds, Wingfield, M. J., Seifert, K. A. and Webber, J. F. St. Paul, Minnesota. American Phytopathological Society. 1: 153-160. Giese, J. 1995. Measuring physical properties of foods. Food Technology Feb: 54-63. Giuliano, D., Rosenfeld, I., Awaya, J., Omura, S. and Vagelos, P. 1973. Inhibition of fatty acid synthesis by the antibiotic cerulenin. Specific inactivation of p-ketoacyl carrier protein synthetase. Biochemical and Biophysical Acta. 326: 155-166. Gurr, S., Unkles, S. and Kinghorn, J. 1987. The structure and organization of nuclear genes of filamentous fungi. Gene structure in eukaryotic microbes. Ed., Kinghorn, J. Washington,D.C. IRL Press. 22: 93-162. Hiltunen, M. and Soderhall, K. 1992. Inhibition of Polyketide synthesis in Alternaria alternata by the fatty acid synthesis inhibitor cerulenin. Applied and Environmental Microbiology 58(3): 1043-1045. Hofsten, B. 1956. Growth and metabolism of Ophiostoma multiannulatum on different sources of carbon. Physiologia Plantarum 9: 624-639. Hofsten, A. and Hofsten, B. 1958. Factors influencing cell division and vegetative morphogenesis of Ophiostoma multiannulatum. Physiologia Plantarum 11: 106. Hopwood, D. and Sherman, D. 1990. Molecular genetics of polyketides and its comparison to fatty acid biosynthesis. Annual Review of Genetics 24: 37-66. Hoti, S. and Balaraman, K. 1993. Formation of melanin pigment by a mutant Bacillus thuringiensis H-14. Journal of General Microbiology 139: 2365-2369. Jeng, R., Hintz, W., Bowden, C , Horgen, P. and Hubbes, M. 1996. A comparison of the nucleotide sequence of the cerato-ulmin gene, and the rDNA ITS between aggressive and non-aggressive isolates of Ophiostoma ulmi sensu lato the causal agent of Dutch elm disease. Current Genetics 29: 168-173. Juzlova, P., Martinkova, L. and Kren, V. 1996. Secondary metabolites of the fungus Monascus: a review. Journal of Industrial Microbiology 16: 163-170. Kaarik, M. 1950. The nitrogen nutrition and vitamin requirement of Ophiostoma pini. Physiologia Plantarum 3: 93-102. Kaarik, A. 1960. Growth and sporulation of Ophiostoma and some other blueing fungi on synthetic media. Symb.Bot.Ups. 16(3): 1-168. Kawamura, C., Moriwaki, J., Kimura, N., Fujita, Y., Fuji, S., Hirano, T., Koizumi, S. and Tsuge, T. 1997.- The melanin biosynthesis genes of Alternaria alternata can restore 158 pathogenicity of the melanin deficient mutants of Magnaporthe grisea. Molecular Plant-Microbe Interactions 10(4): 446-453. Kawamura, C , Tsujimoto, T. and Tsuge, T. 1999. Targeted disruption of a melanin biosynthesis gene affects conidial development and UV tolerence in the Japanese pear pathotype of Alternaria alternata. Molecular and Plant Microbe Interactions 12(1): 59-63. Kim, S., Uzunovic, A. and Breuil, C. 1999. Rapid detection of Ophiostoma piceae and O. quercus in stained wood by PCR. Applied and Environmental Microbiology 65(1): 287-290. Kimura, N. and Tsuge, T. 1993. Gene cluster involved in melanin biosynthesis of the filamentous fungus Alternaria alternata. Journal of Bacteriology 175(14): 4427-4435. Kozak, M. 1987. An analysis of 5'-noncoding sequences from 699 vertebrate messenger RNAs. Nucleic Acids Research 15(20): 8125-8148. Kubo, Y., Suzuki, K., Furusawa, I., Ishida, N. and Yamamoto, M. 1982a. Relation of appressorium pigmentation and penetration of nitrocellulose membranes by Colletotrichum lagenarium. Phytopathology 72: 498-501. Kubo, Y., Suzuki, K., Furusawa, I. and Yamamoto, M. 1982b. Effect of tricyclazole on appressorial pigmentation and penetration from appressoria of Colletotrichum lagenarium. Phytopathology 72: 1198-1200. Kubo, Y., Suzuki, K., Furusawa, I. and Yamamoto, M. 1983. Scytalone as a natural intermediate of melanin biosynthesis in appressorial of Colletotrichum lagenarium. Experimental Mycology 7: 208-215. Kubo, Y., Furusawa, I. and Yamamoto, M. 1984. Regulation of melanin biosynthesis during appressorium formation in Colletotrichum lagenarium. Experimental Mycology 8: 364-369. Kubo, Y., Suzuki, K., Furusawa, I. and Yamamoto, M. 1985. Melanin biosynthesis as a prerequisite for penetration by appressoria of Colletotrichum lagenarium: Site of inhibition by melanin inhibiting fungicidies and their action on appressoria. Pesticide Biochemistry 23: 47-55. Kubo, Y., Furusawa, I. and Shishiyama, J. 1987. Relation between pigment intensity and penetrating ability in appressoria of Colletotrichum lagenarium. Canadian Journal of Microbiology 33:871 -873. Kubo, Y., Tsuda, M., Furusawa, I. and Shishiyam, J. 1989. Genetic analysis of genes involved in melanin biosynthesis of Cochliobolus miyabeanus. Experimental Mycology 13: 77-84. 159 Kubo, Y., Nakamura, H., Kobayashi, K., Okuno, T. and Furusawa, I. 1991. Cloning of a melanin biosynthetic gene essential for appressorial penetration of Colletotrichum lagenarium. Molecular Plant Microbe interactions 4(5): 440-445. Kubo, Y., Takano, Y., Endo, N., Yasuda, N., Tajima, S. and Furusawa, I. 1996. Cloning and structural analysis of the melanin biosynthesis gene SCD1 encoding scytalone dehydratase in Colletotrichum lagnarium. Applied and Environmental Microbiology 62(12): 4340-4344. Kusaba, M. and Tsuge, T. 1994. Nuclear ribosomal DNA variation and pathogenic specialization in Alternaria fungi known to produce host-specific toxins. Applied and Environmental Microbiology 60(9): 3055-3062. Laidlaw, R. and Smith, G. 1965. The proteins of the timber of Scots pine (Pinus sylvestris). Holzforschung 19: 129-143. Langheinrich, U. and Tischner, R. 1991. Vegetative storage proteins in poplar. Plant Physiology 97: 1017-1025. Leung, H., Lehtinen, U., Karjalainen, R., Skinner, D., Tooley, P., Leong, S. and Ellingboe, A. 1990. Transformation of the rice blast fungus Magnaportha grisea to hygromycin B resistance. Current Genetics 17: 409-411. Lundqvist, T., Weber, P., Hodge, C , Braswell, E., Rice, J. and Pierce, J. 1993. Preliminary crystallographic studies on scytalone dehydratase from Magnaporthe grisea. Journal of Molecular Biology 232: 999-1002. Lundqvist, T., Rice, J., Hodge, C , Basarab, G., Pierce, J. and Kindqvist, Y. 1994. Crystal structure of scytalone dehydratase-a disease determinant of the rice pathogen, Magnaporthe grisea. Structure 2: 937-944. Luo, D., Chen, H. and Jimbow, K. 1994. Co-transfection of genes encoding human tyrosinase related protein-1 prevents melanocyte death and enhances melanin pigmentation and gene expression of Lamp-1. Experimental Cell Research 213: 231-241. Matsumae, A., Nomura, S. and Hata, T. 1964. Studies on cerulenin IV. Biological characteristics of cerulenin. J. Antibiot. Ser. A 17: 1-7. Merrill, W. and Cowling, E. 1966. Role of nitrogen in wood deterioration:amounts and distribution of nitrogen in tree stems. Canadian Journal of Botany AA: 1555-1580. Money, N. and Howard, R. 1996. Confirmation of a link between fungal pigmentation, turgor pressure, and pathogenicity using a new method of turgor measurement. Fungal Genetics and Biology 20: 217-227. 160 Montenecourt, B. and Eveleigh, D. 1977. Semi-quantitative plate assay for determination of cellulase production by Trichoderma viride. Applied and Environmental Microbiology 33: 178-183. Motoyama, T., Imanishi, K. and Yamaguchi, I. 1998. cDNA cloning, expression and mutagenesis of scytalone dehydratase needed for pathogenicity of the rice blast fungus Pyricularia oryzae. Biosci. Biotechnol. Biochem. 62(3): 564-566. Nakasako, M., Motoyama, T., Kurahashi, Y. and Yamaguchi, I. 1998. Cryogenic X-ray crystal structure analysis for the complex of scytalone dehydratase of a rice blast fungus and its tight-binding inhibitor, carpropamid: The structural basis of tight-binding inhibition. Biochemistry 37(28): 9931-9939. Nitta, N., Farman, M. and Leong, S. 1997. Genome organization of Magnaportha grisea: integration of genetic maps, clustering of transposable elements and identification of genome duplications and rearrangements. Theoretical and Applied Genetics 95: 20-32. O'Hagan, D. 1991. The polyketide metabolites. Chichester. Ellis Horwood Limited. Ohno, T., Kesado, T., Awaya, J. and Omura, S. 1974. Target of Inhibition by the anti-lipogenic antibiotic cerulenin of sterol synthesis in yeast. Biochemical and Biophysical Communications 57(4): 1119-1124. Ohno, H., Ohno, T., Awaya, J. and Omura, S. 1975. Inhibition of 6-methylsalicylic acid synthesis by the antibiotic cerulenin. J. Biochem. 78: 1149-1152. Omura, S. 1976. The antibiotic cerulenin, a novel took for biochemistry as an inhibitor of fatty acid synthesis. Bacteriological Review 40: 681-697. Osiewacz, H. and Weber, A. 1989. DNA mediated transformation of the filamentous fungus Curvularia lunata using dominant selectable marker. Applied Microbiology and Biotechnology 30: 375-380. Paine, T., Raffa, K. and Harrington, T. 1997. Interactions among scolytid bark beetles, their associated fungi, and live host conifers. Annual Review of Entomology 42: 179-206. Parsons, K., Chumley, F. and Valent, B. 1987. Genetic transformation of the fungal pathogen responsible for rice blast disease. Proc.Natl.Acad.Sci. USA 84:4161-4165. Perpetua, N., Kubo, Y., Takano, Y. and Furusawa, I. 1996. Cloning and characterization of a melanin biosynthesis THR1 reductase gene essential for appressorial penetration of Colletotrichum lagenarium. Molecular Plant Microbe Interactions 9(5): 323-329. Persson, A., Krook, M. and Jornvall, H. 1991. Characteristics of short-chain alcohol dehydrogenases and related enzymes. European Journal of Biochemistry 200: 537-543. 161 Pipe, N., Buck, K. and Brasier, C. 1995. Genomic fingerprinting supports the separation of Ophiostomapiceae into two species. Mycological Research 99(10): 1182-1190. Popson, S. 1983. Measurement and control of the optical properties of paper. Pointe Claire,Que. Technidyne Corporation. . Quandt, K., Freeh, K., Karas, H., Wingender, E. and Werner, T. 1995. Matlnd and Matlnspector: new fast and versatile tools for detection of consensus matches in nucleotide sequence data. Nucleic Acid Research 23(23): 4878-4884. Raedar and Broda. 1985. A quick method for DNA isolation. Letters in Applied Microbiology 1: 17-20. Royer, J., Dewar, K., Hubbes, M. and Horgen, P. 1991. Analysis of high frequency transformation system for Ophiostoma ulmi, the causal agent of Dutch elm disease. Molecular and General Genetics 225: 168-176. Sambrook, J., Fritsch, E. and Maniatis, T. 1989. Molecular cloning - A laboratory manual. Cold Spring Harbour. Cold Spring Harbour Laboratory Press. Sanchez-Ferrer, A., Rodriguez-Lopez, J., Garcia-Canovas, F. and Carcia-Carmona, F. 1995. Tyrosinase: a comprehensive review of its mechanisms. Biochimica et Biophysica , Acta 1247: 1-11. Sano, Y., Nomura, S., Kamio, Y. and Omura, S. 1967. Studies on cerulenin III Isolation and physico-chemical properties of cerulenin. Journal of. Antibiotics (A) 20: 344-348. Seifert, K. and Grylls, B. 1991. A survey of the sapstaining fungi of Canada. Vancouver, Canada. Forintek Canada Corporation. Seifert, K. 1993. Sapstain of commercial lumber by species of Ophiostoma. and Ceratocystis. Ceratosystis and Ophiostoma: Taxonomy, ecology and pathogenicity. Eds, Wingfield, M. J., Seifert, K. A. and Webber, J. F. St. Paul, Minnesota. American Phytopathological Society. 1: 141-152. Seraglia, R., Traldi, P., Elli, G., Bertazzo, A., Costa, C. and Allegri, G. 1993. Laser desorption ionization mass spectrometry in the study of natural and synthetic melanins. I-Tyrosine melanins. Biological Mass Spectrometry 22: 687-697. Shiotani, H. and Tsuge, T. 1995. Efficient gene targeting in the filamentous fungus Alternaria alternata. Molecular and General Genetics 248: 142-150. Smith, R. 1991. Lumber protection in the 90's. Vancouver, Canada. Forintek Canada Corporation. Spatafora, J. and Blackwell, M. 1994. The polyphyletic origins of Ophiostomatoid fungi. Mycological Research 98(1): 1-9. 162 Stipanovic, R. and Bell, A. 1976. Pentaketide metabolites of Verticillium dahliae. Journal of Organic Chemistry 41(14): 2468-2469. Stipanovic, R. and Bell, A. 1977. Pentaketide metabolites of Verticillium dahliae. Accumulation of naphthol derivatives by the aberrant-melanin mutant brm-2. Mycologia 69: 164-172. Subramanian, C. 1983. Hyphomycetes on wood and bark. Hyphomycetes: Taxonomy and Biology. Ed., Subramanian, C. V. New York. Academic Press: 148-169. Sweigard, J., Chumley, F. and Valent, B. 1992. Disruption of a Magnaportha grisea cutinase gene. Molecular and General Genetics 232: 183-190. Sweigard, J., Orbach, M., Valent, B. and Chumley, F. 1998. A miniprep procedure for isolating genomic DNA from Magnaportha grisea. Fungal Genetics Newsletter 37: 4. Tajima, S., Kubo, Y., Furusawa, I. and Shishiyama, J. 1989. Purification of a melanin biosynthetic enzyme converting scytalone to 1,3,8-trihydroxynaphthalene from Cochliobolus miyabeanus. Experimental Mycology 13: 69-76. Takahashi, R. and Akiyama, T. 1993. Characterization of a melanin associated with low temperature induced browning in soybean seed coats. Phytochemistry 34(2): 587-588. Takano, Y., Kubo, Y., Shimizu, K., Mise, K., Okuno, T. and Furusawa, I. 1995. Structural analysis of PKS 1, a polyketide synthase gene involved in melanin biosynthesis in Colletotrichum lagenarium. Molecular and General Genetics 249: 162-167. Takano, Y., Kubo, Y., Kawamura, C , Tsuge, T. and Furusasa, I. 1997'a. The Alternaria alternata melanin biosynthesis gene restores appressorial melanization and penetration of cellulose membranes in the melanin deficient albino mutant of Colletotrichum lagenarium. Fungal Genetics and Biology 21: 131-140. Takano, Y., Kubo, Y., Kuroda, I. arid Furusawa, I. 1997b. Temporal transcriptional pattern of three melanin biosynthesis genes, PKS1, SCD1 and THR1 in appressorium differentiating and nondifferentiating conidia of Colletotrichum lagenarium. Applied and Environmental Microbiology 63(1): 351-354. Tan, K. 1978. Light induced fungal development. The filamentous fungi-Developmental mycology. Eds, Smith, J. and Berry, D. New York. John Wiley & Sons. 3: 334-357. Tanaka, C , Tajima, S., Furusawa, I. and Tsuda, M. 1992. The Pgrl mutant of Cochliobolus heterostrophus lacks a p-dipheol oxidase involved in naphthalenediol melanin synthesis. Mycological Research 96(11): 959-964. Tanaka, C , Taniguchi, N. and Tsuda, M. 1994. Genetic analysis of Cochliobolus heterostrophus beige mutant. Mycological Research 99(5): 621-622. 163 Terziev, N., Boutelje, J. and Soderstrom, O. 1993. The influence of drying schedules on the redistribution of low molecular sugars in Pinus sylvestris. Holzforschung 47: 3-8. Thurston, C. 1994. The structure and function of fungal laccases. Microbiology 140: 19-26. Tokousbalides, M. and Sisler, H. 1979. Site of inhibition by tricyclazole in the melanin biosynthetic pathway of Verticillium dahliae. Pesticide Biochemistry and Physiology 11: 64-73. Tsai, H., Chang, Y., Washburn, R., Wheeler, M. and Kwon-Chung, K. 1998. The developmentally regulated albl gene of Aspergillus fumigatus: Its role in modulation of conidial morphology and virulence. Journal of Bacteriology 180(12): 3031-3038. Tsuge, T., Nishimura, S. and Kobayashi, H. 1990. Efficient integrative transformation of the phytopathogenic fungus Alternaria alternata mediated by the repetitive rDNA sequences. Gene 90: 207-214. Tsuji, G., Takeda, T., Furusawa, I., Horino, O. and Kubo, Y. 1997. Carpropamid, an anti-rice blast fungicide, inhibits scytalone dehydratase activity and appressorial penetration in Colletotrichum lagenarium. Pesticide Biochemistry and Physiology 57: 211-219. Uzunovi, A., Yang, D., Gagne, P., Breuil, C , Bernier, C , Byrne, A., Gignac, M. and Kim, S. 1999. Fungi that cause sapstain in Canadian softwoods. Canadian Journal of Microbiology in press. Vidal-Cros, A., Viviani, F., Labesse, G., Boccara, M. and Gaugry, M. 1994. Polyhydroxynapthalene reductase involved in melanin biosynthesis in Magnaporthe grisea. European Journal of Biochemistry 219: 985-992. Viviani, F., Gaudry, M. and Marquet, A., Eds. 1991. Biosynthesis of melanin by Pyricularia oryzae: mechanism of reduction of polyhydroxynaphthalenes. Bioorganic Chemistry in Healthcare and Technology. New York. Plenum Press. Vogel, H. 1956. A convenient growth medium for Neurospora. Microbial Genetics Bulletin 13: 42-43. Wheeler, M., Tolmsoff, W. and Meola, S. 1976. Ultrastructure of melanin formation in Verticillium dahliae with (+)-scytalone as a biosynthetic intermediate. Canadian Journal of Microbiology 22: 702-711. Wheeler, M. and Stipanovic, R. 1979. Melanin biosynthesis in Thielaviopsis basicola. Experimental Mycology 3: 340-350. Wheeler, M. 1983. Comparisons of fungal melanin biosynthesis in ascomycetous, imperfect and basidiomycetous fungi. Transcripts of British Mycology Society 81(1): 29-36. 164 Wheeler, M. and Stipanovic, R. 1985. Melanin biosynthesis and the metabolism of flaviolin and 2-hydroxyjugalone in Wangiella dermatitidis. Archives of Microbiology 142: 234-241. Wheeler, M. and Bell, A. 1988. Melanins and their importance in pathogenic fungi. Current topics in medical mycology. Ed., McGinnis, M. R. New York. Springer-Verlag. 2: 338-387. Wheeler, M. and Bhatnagar, D. 1995. Inhibition of aflatoxin production by Aspergillus flavus with pentachlorobenzyl alcohol, phthalide and pyroquilon. Pesticide Biochemistry 52:109-115. Wheeler, M. and Klich, M. 1995. The effects of tricyclazole, pyroquilon,phthalide, and related fungicides on the production of conidial wall pigmnets by Penicillium and Aspergillus species. Pesticide Biochemistry and Physiology 52: 125-136. Wingfield, M., Seifert, K. and Webber, J., Eds. 1993. Ceratocystis and Ophiostoma: Taxonomy, ecology and pathogenicity.. Minnesota. American Phytopathological Society. Woloshuk, C , Sisler, H., Tokoubalides, M. and Dutky, S. 1980. Melanin biosynthesis in Pyricularia oryzae:site of tricyclazole inhibition and pathogenicity of melanin deficient mutants. Pesticide Biochemistry 14: 256-264. Woloshuk, C , Wolkow, P. and Sisler, H. 1981. The effect of three fungicides, specific for the control of rice blast disease on the growth and melanin biosynthesis by Pyricularia oryzae Cav. Pesticide Science 12: 86-90. Woloshuk, C , Sisler, H. and Vigil, E. 1983. Action of the antipenetrant, tricyclazole on appressoria of Pyricularia oryzae. Physiology and Plant Pathology 22: 245-259. Yokoyama, K., Yasumoto, K., Suzuki, H. and Shibahara, S. 1994. Cloning of the human DOPAchrome tartomerase/tyrosinase related protein 2 gene and identification of two regulatory regions required for its pigment cell specific expression. Journal of Biological Chemistry 269(43): 27080-27087. Zabel, R. and Morrell, J. 1993. Wood microbiology: Decay and its prevention. San Diego. Academic Press Inc. Zimmerman, W., Blanchette, R., Burnes, T. and Farrell, R. 1995. Melanin and perithecial development in Ophiostomapiliferum. Mycologia 87(6): 857-863. Zink, P. and Fengel, D. 1988. Studies on the colouring matter of blue-stain fungi. Part 1. General Characterization and the Associated Compounds. Holzforschung 42(4): 217-220. Zolan, M. and Pukkila, P. 1986. Inheritance of DNA methylation in Coprinus cinereus. Molecular and Cellular Biology 6(1): 195-200. 165 A P P E N D I X The purpose of this appendix is to provide information regarding our use of UV mutagenesis to isolate pigment deficient mutants of O. floccosum 387N. This was attempted with the anticipation that pigmentation deficient mutants would have distinctive phenotypes like those found in other filamentous fungi. The purpose of creating pigment deficient mutants was ultimately to use them to demonstrate restoration of function by complementation with the THN reductase gene that was isolated and sequenced in Chapter 3. A . l Materials and Methods A. 1.1 UV mutagenesis of 387N O. floccosum 387N was grown in liquid MEA for 3 days and then filtered through four layers of sterile cheesecloth to ensure that only single spores were harvested and mycelial masses removed. The yeast-like cells were counted and diluted to 3x104 cells/mL. Approximately 1500 cells were plated onto 15 cm diameter culture dish of MEA with 0.01% sodium deoxycholate and then exposed without lids for 60-90 seconds. The plates were exposed at a distance of 41 cm from a 30 Watt germicidal UV lamp (Sylvania). After mutagenesis, the plates were incubated at 23 °C in the dark to prevent excision repair of UV induced thiamine dimers. This treatment resulted in the death of 90-99%) of the cells. After 14 days, putative pigmentation mutants were visually assessed and plated 166 on MEA. Non-mutagenized 387N cultures were grown in parallel and half were treated with 10 mg/mL of tricyclazole in order to be able to comparatively identify the pigmentation effects of a putative THN reductase deficiency A . 1.2 Intermediate extraction and T L C analysis Nutrient broth was inoculated with 5mm hyphal front plugs of each of the putative mutants. After 2 weeks, the cultures were extracted according to the following protocol and then evaluated for the accumulation of any DHN intermediates. Intermediates were extracted by adding 2 volumes of acetone to the liquid culture and stirring overnight in the dark under nitrogen gas, on a magnetic stir plate. The mixture was then filtered through No. 5 Whatman filter paper to remove solids and the extract was rotorevaporated under a vacuum at 35°C and 200 rpm to remove the organic phase. The residual aqueous solution was adjusted to pH 5.0 with potassium phosphate or phosphoric acid and then saturated with sodium chloride. This mixture was extracted three times using a separately funnel and approximately equal volumes of ethyl acetate. Any residual aqueous liquid was removed by drying with the addition of anhydrous magnesium sulphate. The organic solution was then filtered through No. 1 Whatman filter paper and rotor-evaporated under a vacuum at 30°C and 150 rpm until only a residual film was left in the round bottom flask. This residue was then solublized in 500 pL of ethyl acetate for TLC analysis. The extraction method used was a consolidation of several extraction methods (Bell et al., 1976b; Kimura and Tsuge, 1993; Stipanovic and Bell, 1977; Tokousbalides and Sisler,1979). 167 The crude culture extracts were spotted on TLC plates using capillary tubes. TLC plates were Whatman AL SIL 256um G/UV254 and the solvent system used was either chloroform: methanol (9:1) or ether: hexane: formic acid (50:50:1). The extracts were run in duplicate, co-spotted with the standard and flanked by the standard. Standard intermediates used were, 1,8-DHN, T3HN, 2-HJ, 3-HJ and scytalone. These standards were a generous gift from Michael Wheeler. Under visible light, the naphthoquinones such as flaviolin, jugalone, 2-HJ or 3-HJ, should appear bright yellow or orange in colour. When the TLC plates are viewed under 254 nm UV light, both the tetralones such as scytalone and the naphthols such as DHN, T3HN and T4HN will quench. Under 366 nm UV light, the tetralones will fluoresce yellow and the naphthols will fluoresce light blue. Tetralone or naphthols present on the TLC plate will oxidize and appear as a dark grey stain when developed in a 5% solution of ammonium molybdate in sulphuric acid. A solution of 0.1 M ferric chloride in a 1:1 mixture of water and acetone was used to detect aromatic phenols on the TLC plates. This solution will cause any tetralones to form red-brown chelates. A . 1.3 Acetylation and gas chromatographic analysis of extractives DHN intermediates are unstable and need to be acetlyated prior to analysis by GC that operates under conditions of high temperature and pressure. The extracted intermediates were dissolved in 500 ul of anhydrous pyridine. Then 500 ul of acetic anhydride was added to the extract and acetlyation was allowed to progress overnight at room temperature. The reaction was stopped by the addition of 20 mL of methanol that was 168 used to extract the weakly azeotropic pyridine. The residue was rotoevaporated at 35°C and then dissolved in 500 pi ethyl acetate. A Hewlett-Packard 5890 series II gas chromatograph with a 30m x 0.25 mm J&W column and a 0.25 pm BD5 film was used to analyze the extracts. The carrier gas-.was helium at 13.1 psi and the detector was flame ionization at 290°C. The inlet temperature was 260°C and the run time was 66 minutes. A.1.4 P C R screening of mutants for changes in the T H N reductase gene Each mutant was cultured in 50 mL of MEA for 6 days before the mycelia was harvested for extracting the DNA by the method of Raeder and Broda (1985). PCR reactions with UV mutant template DNA and the primers 414F and 1756R were performed according to the conditions outlined in Section 3.2.4 with an annealing temperature of 50°C. The control DNA should yield a PCR product of 1342 bp. A .2 Results A.2.1 U V mutant characterization Two hundred putative UV mutants were selected and propagated on MEA for the characterization of any pigment aberration. In general six colour classes were observed; albino, very light brown, light brown, dark brown, light red and yellow. However, within these classes, sectoring was observed and it was difficult to definitively classify most mutants by colour due to the variable pigment morphology. The relative proportions of mutants in each class are summarized in Table A. 1. 169 Table A. 1. Summary of the types and prevalence of recovered UV mutants (in percent). Class Prevalence Albino 29.7% Very light brown 6.7% Light brown 21.9% Dark brown 5.1% Light red 5.1% Yellow 2.3% Wild type 29.2% 170 A . 2 .2 T L C evaluation of selected mutants Eight mutants which seemed to be stable in terms of their pigmentation, were chosen for TLC analysis after preliminary colour evaluation on MEA plates, these are listed in Table A.2. Included as controls were, tricyclazole treated and untreated non-mutagenized 387N. The standard DHN intermediates that were evaluated by TLC ran smoothly on TLC plates, the spots observed were clean and migrated tightly. All of the intermediate standards tested behaved as expected regarding their activity under different wavelengths of light and in their response to different development treatments. For the mutants and control extracts, 6-10 large smears were often present per lane and ran without discrete divisions. In the case of some of the mutants, some smears were present that exhibited conflicting behaviour with different treatments and did not share characteristics of the standards. No clear patterns could be discerned that could indicate if any intermediates were accumulating. Table A.2. Colour of UV mutants selected for TLC and GC analysis and grown on MEA for 12-14 days. Mutan t Co lour on M E A 34 light brown 312 very light brown 323 very light brown 324 albino 330 dark brown 47 reddish 568 brown 571 brown 171 A.2.3 GC evaluation of selected mutants The purpose of using GC was to determine whether DHN intermediates were accumulating in UV mutagenized cultures of isolate 387N. Peak height was used as the parameter to evaluate the relative proportions of individual intermediates since all the peaks were very similar in appearance being long and narrow with minimal width. The peaks of interest were identified by retention time when compared to the standards. The peaks were then measured by height (cm) from the baseline. The peak heights of all intermediates present for a given mutant extract were summed and the individual peak heights of each intermediate divided by this sum in order to calculate the relative proportion of the individual intermediates. An assumption made using this technique was that the intermediates had similar separation sensitivities. The purity and concentration of DHN intermediate standards was unknown since they were received as a gift from Michael Wheeler and are not available commercially. Therefore, this was not a quantitative GC analysis. All of the intermediate DHN standards that were resolved by GC, produced chromatograms that had clean major peaks. Retention times are summarized in Table A.3 and a chromatogram of the standards is illustrated in Figure A . l . T3HN and scytalone had the same retention times and could not be differentiated from one another. In the control, non-mutagenized 387N, all four intermediates were detected and DHN was present in the largest proportion (58%) followed by 2-HJ (22%), 3-HJ (14%) and T3HN/Scytalone (5%>) (Figure A.2). In the albino mutant 324 only two intermediates 172 were detected. DHN was present at approximately twice the concentration of 2-HJ (Figure A.2). In the dark brown mutant, 330 and in the light brown mutant, 312, only DHN was detected, whereas in the phenotypically similar light brown mutant 34, three intermediates were present. 3-HJ (68%) was the most prevalent followed by DHN and T3HN/Scytalone in roughly equivalent amounts (Figure A.2). The remaining light brown mutant, 323 and the brown mutant 568 had very similar profiles with the most prevalent intermediate being T3HN/Scytalone (87%) and a small proportion of DHN (13%) (Figure A.2). There were no peaks corresponding to those of the standard intermediates in brown mutant 571. In mutant 47, which had a reddish phenotype, DHN was present in the highest proportion (84%>) and there was some detectable T3HN/Scytalone (16%) (Figure A.2). Table A.3. The GC retention times for acetylated DHN intermediate standards. Intermediate Standard Retention Time (minutes) 2-HJ 17.7 3-HJ 18.6 1,8-DHN 24.3 Scytalone 30.7 1,3,8-THN 30.7 Ethyl Acetate 32.1 173 c (D TO (1) CL 10 20 30 40 Elution Time (minutes) 50 Figure A. l . A compilation of the GC chromatograms of all of the DF£N intermediate standards. Figure A.2. Analysis of GC peak height and corresponding relative proportions of the intermediates extracted from each UV mutant. 174 A.2.4 Growth and pigmentation on M E A and wood In addition to the mutants selected for TLC and GC analysis, many others were evaluated for their colour when grown on wood. Colour observations of these other mutants when grown on MEA and on wood are reported in Table A.4. A.2.5 P C R screening of mutants for changes in the T H N reductase gene PCR was used to try and detect differences in the sizes of PCR products of the recovered UV mutants. The 'PCR #' column in Table A.4 corresponds to the mutant DNA used as template in each PCR reaction and corresponds to the lane in one of the following three gel photographs in which that PCR reaction was resolved by gel electrophoresesis (Figure A3a-c). In most lanes, a 1342 bp product can be seen and in those cases where the product is missing (lanes 14, 15, and 27), the reactions were repeated and were run again in the far four right lanes of gel ' C . The '+' lanes indicate a positive control using the same primers with pBSA4.2 plasmid DNA as template. In the gel in Figure 4.1c a '-' symbol indicates a PCR reaction without template DNA. In all of the mutants tested, a 1342 bp PCR product was present. 175 Table A.4. Summary of pigment phenotypes of UV mutants grown on MEA and wood after a 2 week culture period. PCR# Mutant Phenotype on M E A Colour on wood 1 324 albino white 2 67 albino white 3 638 albino white 4 36 very light brown very light brown 5 39 very light brown very light brown 6 48 very light brown very light brown 7 65 very light brown white to gray 8 312 very light brown white 9 319 very light brown white 10 323 very light brown white 11 513 very light brown very light brown 12 519 very light brown very light brown 13 551 very light brown very light brown 14 556 very light brown very light brown 15 564 very light brown very light brown 16 647 very light brown white 17 651 very light brown very light brown 18 34 light brown white 19 52 light brown light brown 20 55 light brown light brown 21 61 light brown brown/black 22 64 light brown gray 23 510 light brown light brown/gray 24 512 light brown light brown 25 521 . light brown gray 26 543 light brown light brown 27 545 light brown light brown 28 563 light brown light brown 29 616 light brown light brown 30 617 light brown light brown 31 626 light brown light brown 32 629 light brown light brown 33 631 light brown white 34 643 light brown light brown 35 645 light brown white 36 649 light brown brown 37 568 . brown no growth 38 571 brown dark brown 39 330 dark brown no growth 40 528 dark brown gray/black 41 47 reddish no growth 176 M 1 2 3 4 5 6 7 8 9 10 11.12 13 14 15 16 17 18 19 20 + M M 21 22 23 24 25 26 27 28 29 30 31 + M M 32 33 34 35 36 37 38 39 40 41 1 14 15 27 + - M Figure A.3. PCR amplification of UV mutant template DNA with 414F/1756R primers . 177 A.3 Discussion The proportion of mutants of 387N recovered from each class had a different distribution than that of M. grisea mutants recovered from conidia treated in a similar fashion (Chumley and Valent, 1990). The buf mutant phenotypes from M. grisea occurred at 0.1-0.01%, and the albino phenotype was obtained in 0.05-0.01% of the survivors fashion (Chumley and Valent, 1990). In 387N almost 30%> of survivors had a light brown phenotype and an additional 30% of survivors had the albino phenotype. These differences probably reflect the difficulty in determining the true pigmentation status in 387N. ' Some of the UV induced mutants had altered pigmentation phenotype when grown on MEA and on wood. These observations warrant further investigation to characterize the nature of the mutations. PCR analysis of the UV induced mutants demonstrated that both primer sites were present and that there were no significant deletions within the intervening sequences such that a 1342 bp was detected for all of the mutants tested. However, since UV mutagenesis induces point mutations, PCR may not be the appropriate means to detect these differences. No conclusions can be drawn from the TLC analysis of culture extracts since it was difficult to interpret the data due to-the lack of resolving power or this technique. V. dahliae strains, used by other researchers to obtain the intermediate standards, would have been useful to determine if a modification of the extraction protocol was required in order to obtain better resolution. Culture conditions may also have to be modified since 178 the naphthols are prone to auto-oxidation under aerobic conditions. Acetylation of the extracts, as performed in this work, may be a way to stabilize the DHN intermediates and prevent auto-oxidation. The purpose of the GC analysis of culture extracts was to determine if mutants were accumulating DHN intermediates. Albino mutants of V. dahliae do not accumulate any DHN (Bell et al, 1976b). Contrary, to this, we found that the albino mutant 324 was accumulating DHN and 2-HJ. Scytalone and flaviolin were the intermediates found to accumulate in fungi which were mutant in the SD gene. These mutants are characteristically red or pink in colour. A reddish mutant, number 47, was found to accumulate large amounts of DHN and smaller amounts of T3HN/scytalone. It is apparent that our visual assessment of the mutant phenotype does not correspond to what is occurring at the biochemical level. Another possible explanation is that these mutants were not pure isolates. Light brown mutants of V. dahliae, deficient in THN reductase, were found to accumulate 2-HJ (Bell et al., 1976a). This intermediate was not detected in any of the light brown mutants of 387N. However, 3-HJ was accumulating in mutant 34 (light brown). An explanation that could account for this observation is that alternative branch pathways exist for the metabolism of accumulated intermediates in other fungi (Figure A.4). In general, flaviolin and 2-HJ are the products that accumulate in BRM mutants of tricyclazole treated DHN fungi. Some fungi such as Wangiella dermatitidis, V. dahliae, and P. oryzae have unique branch pathways that can convert 2-HJ to 3-HJ (Wheeler and 179 Stipanovic, 1985). It is possible that this mechanism may also be present in 387N and consequently mutant 34 could be deficient in THN reductase. A polymerization deficient mutant of V. dahliae, isolated only infrequently, accumulates monomers and dimers of DHN probably due to a defect in the oxidase that polymerizes the DHN into melanin (Bell and Wheeler, 1986). Two of the 387N UV mutants, 312 and 330, accumulated DHN. However, it was difficult to reconcile the appearance of mutant 312, (very light brown) with that of 330 (darker than wild type). It is possible that these two mutants share this polymerization defect. Two other UV mutants from 387N were brown/light brown in colour and shared identical accumulated intermediate profiles of T3HN/scytalone and DHN. One would expect that if T3HN were accumulating, it would be metabolized to either 2-HJ or 3-HJ as in mutant 34. If on the other hand, scytalone was accumulating the cultures should have had a reddish hue. Since neither of these situations were observed, it was difficult to draw any conclusions. Another possible explanation is that a DHN melanin regulatory gene was affected. 180 Acetate Melanin n .a T4HN l = > Scytalone l = > T3HN C = > Vermelone t=C> D H N Flaviolin 2-HJ 1,2,4,5,7-PHN l^ 4> 5-HS B§^> 1,2,4,5-THN Jugalone ( = > 3-HJ 4-HS 3,4,8-THT c z > 1,4,5-THN / — i 1,3,4,5-THN £ ft £ 4,8-DHT Lll 2,4,8-THT Figure A.4. DHN pathway of melanin biosynthesis showing branch pathways. The conversion from 1,2,4,5,7-PHN to 1,2,4,5-THN is a branch unique to W. dermatitidis while the conversion of 1,2,4,5,7-PHN to 4-HS is a branch common to V. dahliae and P. oryzae. The following chemical abbreviations are used: DHN-dihydroxynaphthalene, T3HN-trihydroxynaphthalene, T4HN-tetrahydroxynaphthalene, HJ-hydroxyjugalone, PHN-pentahydroxynaphthalene, HS-hydroxyscytalone, THT-tetra or tri-hydroxytetralone and DHT-dihydroxytetralone (Wheeler and Stipanovic, 1985). 181 A.4 Conclusions The pigmentation of 387N is variable, and it was difficult to ascertain whether we obtained true pigmentation deficient mutants based on visual colour assessments. PCR alone was unsuitable for differentiating point mutations but combined with sequencing could have been definitive. Sequencing would have been expensive but it would yield more precise information useful in future complementation experiments. TLC was unable to resolve the culture extractives and this was unfortunate since it could have been an inexpensive and rapid method for determining if intermediates were accumulating. A more effective method for analyzing extracts was GC, which had the same extraction protocol. The only problem with this method was the co-elution of T3HN and scytalone. If standards were available from the branch pathways, it would be easy to test whether any of these were present and form a clearer concept of what is occurring during pigmentation in O. floccosum 387N. Using GC we were able to determine that mutant 34 was accumulating 3-HJ and could have a mutation in the THN reductase. With this knowledge, further characterization of this mutant would be valuable. 182 


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