Towards Control of Dutch Elm Disease: dsRNAs and the Regulation of Gene Expression in Ophiostoma novo-ulmi

(1)

Towards Control of Dutch Elm Disease:

dsRNAs and the Regulation of Gene Expression in Ophiostoma novo-ulmi by

Joyce Silva Carneiro

Bachelor of Science, Universidade Católica de Goiás, 2004 A Dissertation Submitted in Partial Fulfillment

of the Requirements for the Degree of DOCTOR OF PHILOSOPHY

in the Department of Biology

 Joyce Silva Carneiro, 2013 University of Victoria

(2)

Supervisory Committee

Towards Control of Dutch Elm Disease:

dsRNAs and the Regulation of Gene Expression in Ophiostoma novo-ulmi

by

Joyce Silva Carneiro

Bachelor of Science, Universidade Católica de Goiás, 2004

Supervisory Committee Dr. Will Hintz, Supervisor

(Department of Biology, Centre for Forest Biology) Dr. Barbara Hawkins, Departmental Member

(Department of Biology, Centre for Forest Biology) Dr. Patrick von Aderkas, Departmental Member (Department of Biology, Centre for Forest Biology) Dr. Delano James, Departmental Member

(Department of Biology)

Dr. Caroline Cameron, Outside Member

(3)

Abstract

Supervisory Committee Dr. Will Hintz, Supervisor

(Department of Biology, Centre for Forest Biology) Dr. Barbara Hawkins, Departmental Member

(Department of Biology, Centre for Forest Biology) Dr. Patrick von Aderkas, Departmental Member (Department of Biology, Centre for Forest Biology) Dr. Delano James, Departmental Member

(Department of Biology)

Dr. Caroline Cameron, Outside Member

(Department of Biochemistry and Microbiology)

Ophiostoma novo-ulmi is the causal agent of Dutch elm disease (DED) which has had a severe impact on the urban landscape in Canada. This research program focused on developing molecular genetic strategies to control this pathogenic fungus.

The first strategy involved the development of RNA interference (RNAi) for the down-regulation of genes involved in pathogenicity. An efficient RNAi cassette was developed to suppress the expression of the endopolygalacturonase (epg1) locus which encodes a cell-wall degrading enzyme. This epg1-RNAi cassette significantly reduced the amount of polygalacturonase activity in the fungus and resulted in almost complete degradation of epg1 mRNA. The need for a native promoter to selectively down-regulate specific gene loci was addressed by developing a carbon-catabolite regulated promoter (alcA) to drive the expression of the epg1-RNAi cassette. The expression of an alcA-driven epg1-RNAi cassette resulted in the down-regulation of epg expression under glucose starvation but normal levels of expression in high glucose. The expression could therefore be controlled by culture conditions.

The second strategy explored the potential of using dsRNA viruses to vector disruptive RNAi cassettes. An isolate of O. novo-ulmi strain 93-1224 collected in the city

(4)

of Winnipeg, was infected by two dsRNA mitoviruses which upon sequence characterization were named OnuMV1c and OnuMV7.

To assess the transmissibility of this dsRNA virus the infected isolate 93-1224 was paired with three naive isolates of the related fungi O. ulmi and O. himal-ulmi. Through the use of nuclear and mitochondrial markers it was determined that the virus OnuMV1c may not rely on mitochondrial fusion for transmission but may have a cytoplasmic transmission route.

This investigation of gene expression and manipulation has provided tools to help understand gene regulation in O. novo-ulmi. It has also added to our knowledge of mitoviruses, their transmission and potential use as a biological control. By enhancing our understanding of transmissible hypovirulence this work contributes to efforts to develop a new approach to target DED as well as a potential model for the control of other fungal diseases.

(5)

List of Tables

Table 1. PCR primers used in RNAi cassette construction ... 50 Table 2. Polygalacturonase activity of epg1¯ and RNAi transformants relative to the wt isolate. ... 64 Table 3. Measurement of epg1 gene expression as determined by q-PCR ... 67

(9)

List of Figures

Figure 1. A classification of the genera Ophiostoma, Ceratocystiopsis and Grosmannia. 9 Figure 2. Construction of the epg1 RNAi expression cassette ... 51 Figure 3. Linear map of the final RNAi vectors 200i and pAN7-gpd-epg-409i ... 54 Figure 4. Polygalacturonase acid plate assay. ... 62 Figure 5. Measurement of enzyme activity, as determined in the polygalacturonic acid plate assay for the epg1- mutant, the RNAi transformants 409-4, 409-6 and 409-7, and the wt isolate. ... 65 Figure 6. The alcA-driven expression cassettes.. ... 84 Figure 7. Flourescence of alcA-driven YFP in O. novo-ulmi ... 87 Figure 8. Relative fluorescent intensity (RFI) of O. novo-ulmi cultures expressing YFP at 24 to 72 hours of growth ... 89 Figure 9. Measurement of EPG enzyme activity of RNAi transformants relative to the wt isolate. ... 92 Figure 10. Schematic representation of alignment of a series of SPAT and partial cDNA clones derived from the dsRNAs of O. novo-ulmi isolate 93-1224. ... 110 Figure 11. Unrooted maximum likelihood tree for the ORFs of dsRNA 01 and dsRNA 02 from O. novo-ulmi isolate 93-1224 with all RNA-dependent RNA polymerases (RdRPs) encoded by Ophiostoma mitoviruses. ... 115 Figure 12. Phylogenetic identification of OnuMV1c and OnuMV7 ... 117 Figure 13. Alignment of the conserved amino acid motifs within RNA-dependent RNA polymerases of O. novo-ulmi encoded by mitochondrial viruses ... 119 Figure 14. Northern blot hybridization using strand-specific probes of O. novo-ulmi 93-1224 mitovirus OnuMV1c. ... 121 Figure 15. Potential secondary structures of the ends of dsRNA01 (OnuMV1c) and dsRNA03... 125 Figure 16. Attempt to cure 93-1224 O. novo-ulmi strain from OnuMV1c virus ... 127

(10)

Figure 17. Screening isolates of O. novo-ulmi collected in the City of Winnipeg for the presence of mitovirus. ... 148 Figure 18. Incidence of mitovirus in the Winnipeg population of O. novo-ulmi... 151 Figure 19. Variation in mitochondrial loci is due to the presence or absence of optional elements ... 154 Figure 20. Transfer of OnuMV1c from O. novo-ulmi to O. ulmi. A) The SSR patterns were distinct between O. novo-ulmi isolate 93-1224 and O. ulmi isolate R21 ... 157 Figure 21. Transfer of OnuMV1c from O. novo-ulmi to O. himal-ulmi ... 160

(11)

List of Abbreviations

3’ - Three prime – The asymmetric end of a DNA strand possessing a terminal hydroxyl group

5’ - Five prime – The asymmetric end of a DNA strand possessing a terminal phosphate group

- Degree Celsius

BLAST - Basic local alignment search tool bp - Base pairs

cDNA - complementary DNA CDS - Coding Regions

CeMV - Chelara elegans mitovirus

Cluster - A unique putative transcript derived from a single or group of EST sequences likely to be derived from the same gene or same RNA species

Codon - The sequence of three nucleotide bases (triplet) in mRNA that specifies an amino acid and specifies the position of that amino acid in a polypeptide chain through complementary base pairing with an anticodon in tRNA

Contig - A contiguous stretch of DNA sequence assembled from partial, overlapping sequences

CT - Cycle threshold. The cycle in qPCR when sample fluorescence exceeds a chosen threshold above calculated background fluorescence

ddH2O: Double distilled water DED: Dutch elm disease DEPC - Diethylpyrocarbonate DIG - Digoxigenin

DNA - Deoxyribonucleic acid

(12)

dsRNA: Double-stranded Ribonucleic acid DTT - Dithiothreotol

EAN - Eurasian aggressive race of O. novo-ulmi EST - Expressed Sequence Tag

EM - Electron microscopy

Fragment - An EST nucleotide sequence used in a consensus alignment

Genome: The complete set of genetic information contined in the DNA of an organism HE - Homing Endonuclease

HEG - Homing Endonuclease Gene kb - Kilobase kDA - Kilodalton L - Litre mg - Milligram ml - Millilitre min(s) - Minute(s)

mRNA - Messenger RNA mt-RNA – Mitochondria RNA MS - Mass spectroscopy

MS/MS - Mass spectroscopy tandem

Mycelium (pl. mycelia) - The entire mass of hyphae that constitutes the vegetative body or thallus of a fungus

mt DNA - Mitochondrial DNA

NAN - North America aggressive race of O. novo-ulmi NCBI - National Center for Biotechnology Information

(13)

OCM - Ophiostoma complete medium OMM - Ophiostoma mininal medium OMV - Ophiostoma Mitovirus

ORF: Open Reading Frame PTA - Phosphotungstic acid PCR - Polymerase chain reaction

PTGS - Post-transcriptional gene silencing

RT-PCR - Reverse-Transcribed Polymerase Chain Reaction RACE - Rapid Amplification of cDNA Ends

RdRp - RNA-dependent RNA polymerase RISC – RNA-induced silencing complex RNA - Ribonucleic acid

RNAi – RNA interference RT - Reverse transcription rpm - Revolutions per minute SDS - sodium dodecyl sulphate

SPAT - single primer amplification technique s - Seconds

TAE - Tris Acetate EDTA

TEM - transmission electron microscope Tm: Melting temperature

µl - Microlitre µg - Micrograms

(14)

µm - Micrometer UV - Ultraviolet UA - Uranyl acetate

UPT - Unique Putative Transcript UTR - Untranscribed region VLP - Virus-like particle VC - vegetative compatibility w/v – Weight-to-Volume

(15)

Acknowledgments

My friend Alexa once told me that it takes a “village” to achieve a PhD, she was not very far off with her observation...

First and foremost, I would like to thank Dr. William E. Hintz for accepting me as his graduate student, for contributing to this research and to my development as a

scientist in so many ways; my deepest appreciation for all the support and guidance throughout this journey. You made this dream possible and I will always be grateful.

I would also like to acknowledge and thank all my committee members: Dr. Barbara Hawkins, Dr. Caroline Cameron, Dr. Delano James and Dr. Patrick von Aderkas, for their time and advice during this project.

My special thanks to Dr. James and Aniko Varga from the CFIA for sharing your knowledge, protocols, time and frustrations over this project! Working with you has been an absolute delight!

Thank you to all those who have given me technical help and guidance throughout my research, in particular Dr. Paul de la Bastide, Dr. Kris von Schalburg, Glenn Cooper, Sarah Cockburn, Finn Hamilton, Meghan Chabout, Dr. Cornelia Brown and Dr. John Taylor.

To all my lab mates over the years; Jonathan, Amy, Irina, Webby, Cayla, Kevin, Alex, Joel, Tory, Razelin and many more who have passed through the doors, thank you for laughs, smiles and encouragement.

Thank you to all my friends, old and new, especially Roberta, Alexa and Heliana, you have kept me sane throughout this quest. Thank you Matt Austin, Beti Oughtred,

(16)

Erik Fleischer and Malu Souza for all your editing in the beginning and the end. Thank you Bruno de Oliveira, Dr. Peixoto Cruz and Dr. Kátia Pellegrino for helping me to initiate my scientific career both in Brazil and Canada.

Thank you to my new family, Noah, Matt, Deborah and Frank Austin, for your patience, unfailing love and outstanding support that has made everything possible, nothing insurmountable and all of it worthwhile. Matt, without your encouragement, editing and absolute faith in me I may never have reached the end of this journey. Thank you very much!

To my “geographical parents” Elisabeth and John Oughtred, thank you, thank you, thank you for all you have done to help me! Your welcome, support,

encouragement and faith has been amazing and has made everything achievable! Finally, I wish to thank my family in Brazil, Mum Ilza, Daniela and Isabela for teaching me the meaning of unconditional love and for whom I will always feel lucky! It has been a very long road. You have never doubted me. You have always been there for me and without any of you none of this would even have been imaginable, let alone possible!

(17)

Dedication

To my family – both the Brazilian and Canadian branches of that very healthy tree! Para minha família – tanto o ramo brasileiro quanto o canadense de uma árvore muito saudável.

(18)

Chapter 1: General Introduction and Literature Review

1.1 Elm trees

The American elm (Ulmus americana) is an ornamental tree species that has been widely used in urban landscapes by city planners as they are fast-growing and extremely hardy given that they are tolerant to pollution, drought, and extreme temperatures (as high and low as +/- 4 C). They are deciduous and semi-deciduous trees, and are believed to have originated in central Asia in the Miocene period (about 40 million years ago) (Collin, 2006).

Currently approximately 30 species of elm exist. Six are endemic to North America, twenty to Asia and three to Europe (Collin 2006). It has been estimated that a mature elm tree can increase residential property values by approximately $3,600 per tree. With approximately 650,000 elms in cities across Canada, this suggests their worth is roughly C$2.3 billion, while the estimated 7 million elms in the US are valued at US$17.5 billion (Hubbes, 1999; Westwood, 1991; Urban Forestry Branch city of Winnipeg 2011 survey). In Winnipeg, Canada, it is estimated that the city spends approximately C$2.5 million a year on elm sanitation and pruning programs.

In North America, American elm occurs from east Nova Scotia to west Alberta and Montana, south to Florida and central Texas. They are also encountered in British Columbia but are not as abundant (Plotnic and Arboretum, 2000). British Columbia is considered a quarantine for Dutch elm disease (DED) and therefore has the most elm tree nurseries in the country (Urban Forestry Branch city of Winnipeg personal

(19)

In addition to their cosmetic value in residential and urban settings, American elms have further economic value as they are hardwood trees that are used in the

production of furniture and ecological value as their leaves serve as food for the larvae of various Lepidoptera. Unfortunately, in North America and Europe, American elms have virtually disappeared as a favoured street tree after the introduction of DED in the earlier 20th century. This disease has been responsible for the destruction of vast populations of American elm and has severely impacted the ecosystems and economies associated with them after two major pandemics (Campana and Stipes, 1981; Hubbes, 1999; Brasier, 1991).

1.2 Dutch Elm Disease

Dutch elm disease (DED) is caused by Ophiostoma ulmi and O. novo ulmi. This disease was named by a Dutch phytopathologist named Bea Schwarz in 1921 who first reported in the Netherlands. This disease has ravaged American elm populations and it is believed that only approximately 1 in 100,000 American elm trees is DED-tolerant (USDA program 2012 database). American elm or white elm (U. americana), is predisposed to DED, and after being infected with this fungus dies within one to seven years depending on the aggressiveness of the pathogen, and the age of the tree. Older trees are more likely to succumb to the disease and often die sooner than younger trees (Brasier, 1991). American elm trees also display greater susceptibility to DED infection during spring when they produce new growth (Hubbes, 1999). The fungus spreads within the tree’s vascular system and, though the fungus usually is vectored by insects, it

(20)

is believed that elms growing within seven meters of infected trees have nearly a 100 % chance of being infected though their interlocking root systems (Schreiber et al., 1993)

The most common species of Ophiostoma fungus studied here (O. ulmi and O. novo-ulmi) have been separated on the basis of their aggressiveness (i.e. virulence), phenotypic characteristics and geographical distribution, with the former (O. ulmi) representing the less aggressive species and the latter (O. novo-ulmi) the more aggressive one (Brasier, 1991). They both grow and reproduce exclusively within elms.

Ophiostoma novo-ulmi displays an increased ability to degrade cell walls and also secretes more of the hydrophobic protein cerato-ulmin (Scala et al., 1997; Temple et al., 1997). In addition, it has been further differentiated by variations in phenotype and geographical distribution into the subspecies novo-ulmi and americana (Brasier and Kirk, 2001). As described by Brasier and Kirk (2001), these sub-species differ phenotypically in their perethicial form and dimensions, with the former typically being found in Europe and the latter in North America.

For long spans of evolutionary history, Dutch elm pathogens were limited to the continent of Asia, where it is believed to have originated. However, the importation of timber to Europe and North American, from Asia, during the 20th century permitted initial introductions of O. ulmi and later introductions of O. novo-ulmi to these regions which had devastating consequences to elm populations. The introduction of O. ulmi to Europe in early 1918 led to high and persistent rates of DED infection, although

compared to the second DED wave the first one was considered relatively mild. The disease reached North America around 1928, ten years after Europe. In 1940 the more aggressive O. novo-ulmi began to replace O. ulmi causing even more devastating elm

(21)

losses. As a result of these occurrences, DED was termed a pandemic (Brasier et al., 1998).

American elms historically comprised a large portion of the forests covering eastern Canada; however as a result of the two pandemics a large proportion of these forests has been lost and we are currently faced with further losses due to the continued infection of the remaining population (Brasier et al., 1998; Temple et al., 2006). Today the disease is spreading westward from Winnipeg. Currently, Winnipeg has the largest surviving urban forest of American elms in North America, and combating of the disease involves the use of pesticides, fungicides and cutting down infected trees (Martha

Barwinsky, Winnipeg City Forester, personal communication). The evidence left from these two pandemics has shown that DED is one of the most destructive fungal plant pathogens ever seen in the northern hemisphere (Brasier et al., 1998; Hintz et al., 2004) making its study of utmost importance. Our understanding of such pathogens will not only potentially contribute to a possible cure for DED disease, but also further prevent environmental losses due to other pathogenic fungi.

1.3 Fungi Supergroup

The start of the modern age for mycology began with the publication of Pier Antonio Micheli's book in Florence 1729, and his work laid the foundations for the systematic classification of grasses, mosses, and fungi (Ainsworth, 1976; San-Blas, 2008). In 1969, Whittaker introduced a five-kingdom taxonomy that granted fungi equal status with plants and animals (Whittaker, 1969; Sharma, 2005).

(22)

Whittaker’s (1969) five kingdom structure was, from a pedagogical perspective revolutionary, as it departed from the traditional division of life into plant and animals or protists that are “plant-like” or “animal like”. Fungi (true fungi, Eumycota, Mycota) were placed in the Kingdom fungi and can be defined as being eukaryotic organisms with cell walls containing chitin and β-glucans, are unicellular or filamentous, consisting of multicellular haploid hyphae. Hyphae can be homo or heterokaryotic. Their

mitochondria contain flattened cristae and peroxisomes are nearly always present as well as Golgi bodies or individual cisternae. They reproduce sexually as well as asexually in which case the diploid phase is generally short-lived. As for nutrition, they are

osmotrophic (absorptive) lacking an amoeboid pseudopodial phase, saprobic, mutualistic or parasitic (Whittaker, 1969)

More recently with the improvement of our knowledge about the relationships among eukaryotes, along with the recent advances in molecular phylogenetics and the diversity of organisms for which data is available, the five-kingdom system has been supplanted by a classification system in which species traditionally treated as fungi are now distributed across several kingdoms (Simpson and Roger, 2004). Eukaryotes (now divided into a number of major groups) share among them a monophyletic lineage and are assigned to the supergroup Opisthokonta (opísthios the Greek word= "rear, posterior" and kontós= "pole" i.e. "flagellum"), includes kingdom Eumycota (often called kingdom Fungi) (Simpson and Roger, 2004; Berney et al., 2004; Stechmann and Cavalier-Smith, 2003).

Opisthokonta represents animals, true fungi, and other several unicellular groups (Simpson and Roger, 2004) based on molecular phylogenies studies in the early 1990s that

(23)

demonstrated animals, fungi and choanoflagellates are very closely related. For the purpose of this thesis instead of using the old nomenclature “kingdom of fungi”, I will refer to the fungi as a supergroup within of the Opisthokonta super group.

The fungi supergroup is monophyletic and is comprised of a large group of organisms, the classification of which was controversial for many years. For instance, two significant spore-forming groups traditionally studied by mycologists (e.g.

myxomycetes and microsporidia) were often considered to be Protozoa, but recently through molecular phylogeny studies they have been reclassified as animals and fungi. On the other hand, the classification of Chromista (e.g. downy mildews and water

moulds) has not changed (Simpson and Roger, 2004; Berney et al., 2004; Stechmann and Cavalier-Smith, 2003; McKenzie, 2004).

Fungi are traditionally defined as eukaryotic organisms distinct from plants and animals and members of several other smaller groups/kingdoms as they produce spores, are achlorophyllous and can reproduce both sexually and asexually (Alexopoulos et al., 1996). Among this diverse group of organisms fungi are represented by mushrooms, conks, corals, jellies, puffballs, stinkhorns, morels, cups, truffles, lichens, yeasts, rusts, smuts, bread molds and mildews (McKenzie, 2004). Recent molecular studies indicate that choanoflagellates and ichthyosporea are more closely related to animals than to fungi, but the precise higher-level relationships within Opisthokonta are still under investigation.

Many fungi have filamentous, branched hyphae (sing. hypha, from the Greek word hyphe = web), they are tube-like filaments with either single multinucleate cells (coenocytes) that lack septa (cross-walls) separating nuclei, or many septate cells

(24)

containing one, two, or more nuclei. They are surrounded by cell walls containing cellulose and/or chitin (the compound also found in arthropod exoskeletons)

(Alexopoulos et al., 1996). The mass of hyphae, constituting the thallus (pl. thalli, from the Greek word thallos = green shoot) of a fungus, is called the vegetative mycelium (pl. mycelia from the Greek word mykes = mushroom). The mycelium branch and extend via tip elongation, although some groups (like yeasts) consist only of individual hyphae (singular, hypha) cells (Alexopoulos et al., 1996; McKenzie, 2004).

All fungi are heterotrophic and do not produce chlorophyll. They are

predominantly osmotrophs, especially in terrestrial systems, playing crucial roles as decomposers and as symbionts or parasites of plants and therefore they can either live saprobically or symbiotically. Symbiotic relationships vary from mutualistic to commensalistic to parasitic. Ophiostoma species have some restricted parasitic

relationships with plants while microsporidia infect a wide range of animals, including insects. They disseminate via spores that are produced either sexually or asexually, termed meiospores (from the Greek words meioun = lessen and spora = spore) and mitospores (from the Greek word mitos = thread), respectively (Alexopoulos et al., 1996; McKenzie, 2004).

1.4 Ascomycota

Members of the Ascomycota division/phylum are commonly known as the sac fungi. They are characterized mainly by the presence of sexually produced spores called ascospores that have their formation within an ascus but can also reproduce asexually by the formation of non-sexual spores termed conidia (Kirk et al., 2001). The ascomycetes

(25)

are a monophyletic group and are subdivided into three subphyla: the Pezizomycotina (largest of this subphylum) includes all ascomycetes that produce ascocarps (fruiting bodies), except for one genus Neolecta, in the Taphrinomycotina that are thought to grow mostly in the filamentous form. Other ascomycetes include the Saccharomycotina also called "true" yeasts, which are unicellular and reproduce vegetatively by budding rather than by the production of hyphae; and the Taphrinomycotina, including both hyphal fungi, fission yeasts, and mammalian lung parasite. This last subphylum has been determined by molecular analysis to be disparate and is generally considered more primitive than the other two (Deacon, 2005).

1.5 The superfamily Ophiostomataceae: Characteristics, classification and fungal nutrition

The superfamily Ophiostomataceae has undergone very rapid taxonomic changes since its discovery late 1800’s due to the insights provided from the increasing use of molecular tools. The subdivision of the phylogeny has long been a source of taxonomic controversy (Hausner et al., 1993a, 1993b). Currently it corresponds to a family of fungi in the Ascomycota phylum, class Sordariomycetes and includes the genera Ophiostoma (Syd, and P. Syd.), Ceratocystiopsis (Upadh and Kendr.), and Grosmannia (Goid., Boll. Stab. Patel. Veg) (Van Wyk and Wingfield 1992; Seifert and Okada 1993; Wingfield et al., 1993; Zipfel et al., 2006). Earlier descriptions of the family contained the genus Ceratocystis (Ell. & Halst. sensu lato) (Wright and Cain 1961, Griffin 1968, Olchowecki and Reid 1973, Upadhyay 1981) (Fig. 1). Ceratocystis is now an independent and separate genus in the order Microascales (Luttrell ex Benny and Kimbr) (Hausner et al.1993b; Spatafora and Blackwell, 1994; Paulin-Mahady et al. 2002).

(26)

Figure 1. A classification of the genera Ophiostoma, Ceratocystiopsis and Grosmannia. The Cladogram includes their respective anamorphs Sporothrix, Leptographium, Pesotum and Hyalorhinocladiella.

(27)

(28)

Much of the confusion surrounding this group has been due to the methods of classification used by mycologists that allow anamorphs to be named separately from the holomorph of which they form a part, and results in two different names (Thwaites, 2003). In addition, recent molecular characterization based on small subunit (SSU), large subunit (LSU) ribosomal DNA sequences, β-tubulin genes, as well as other morphologic differences, have shown that species of Ceratocystis were not only closely related to those of Ophiostoma, Grosmannia and Ceratocystiopsis, but also suggested that the genus Ophiostoma was separated into at least three groups representing separate genera (Harrington, 1981; de Hoog and Scheffer, 1984; Hausner et al., 1993a, 1993b; Samuels, 1993; Mouton et al., 1994; Spatafora and Blackwell, 1994; Hausner and Reid, 2003; Zipfel et al., 2006). Grosmannia and Ceratocystiopsis were previously thought to be synonyms of Ophiostoma and the groups were merged under Ophiostoma sensu lato (s.l.). The newly proposed classification of the genera Ophiostoma, Ceratocystiopsis and Grosmannia, including their respective anamorphs, is illustrated in Figure 1.

Species in this family have a widespread distribution, share morphological similarities across different niches and show adaptations for their insect-based dispersal (Zipfel et al., 2006). Although this group has polyphyletic origins and share structural similarities (Hausner et al. 1992, 1993b, c, Spatafora and Blackwell 1994) their

differences are in their reproductive stage (teleomorph, anamorph and holomorph) which is the method of classification most used by mycologists. Teleomorph is the sexual reproductive stage (morph), typically a fruiting body like formation. Anamorph is an asexual reproductive stage (morph), often mold-like, and when a single fungus produces multiple morphologically distinct anamorphs these are called synanamorphs. Holomorph

(29)

is the whole fungus, including anamorphs and teleomorphs (Melin and Nannfeldt 1934; Hunt; 1956; Davidson, 1958; Mathiesen-Käärik, 1960; Zipfel et al., 2006).

Anamorphs often are named separately from the holomorphs of which they form a part, often resulting in two different names (Thwaites, 2003). Although anamorph

morphology is very diverse and problematic (as a number of Ophiostoma species produce combinations of up to four possible anamorphs states), it is still the best and most used characteristic to group species in the genera (Okada et al., 1998; Zipfel et al., 2006). Teleomorph characteristics are applied in taxonomic studies of Ophiostoma as well. They include the shape and size of the ascomata and ascospores, and the presence or absence of sheaths surrounding the ascospores (Zipfel et al., 2006). Other important characteristics that have assisted mycologists in separating these two genera is their sensitivity to the antibiotic cycloheximide that Ceratocystis presents, but Ophiostoma does not. Application of these definitions to the Ophiostoma genus has resulted in more than 140 species, with a large variety of distinct teleomorph and anamorph features.

The Ophiostoma genus is tolerant to cycloheximide, contains rhamnose and cellulose in their cell walls, centrum development, ascosporogenesis with anamorph morphology, ascospores with unusual shapes and ascomata with long necks that develop masses of sticky ascospores adapted for dispersal by insects (Upadhyay, 1981; Zipfel et al., 2006).

The fungal growth and nutrition of Ophiostomatoid fungi is very diverse as they manifest several different life cycles. They are polymorphic (grow in both filamentous and yeast-like states), and absorb nutrients in three different ways: saprobic, mutualistic or parasitic. Virtually all fungi produce spores, both asexual and sexual. The spores may

(30)

germinate to form vegetative thalli from primary and secondary mycelia. Thalli may be haploid dominant, diploid dominant, or exhibit haplo-diploid alternation of generations. They often exhibit dikaryotamy, wherein hyphal cells contain two (usually haploid) nuclei that migrate, multiply, and divide together. The term ‘yeast-like cells’ comes from blastospores. Blastospores are asexual spores often called conidia. They are formed by enlargement of an initial recognizable conidium before the initial is delimited by a septum (Hawksworth et al., 1995; Hawksworth 1997). Some Ophiostoma species produce conidia on asexual fruiting structures called synnemata or in a sporothrix-stage where conidia are formed on hyphae.

A synnema consists of a group of conidiophores cemented together by their stalks (Wingfield, Seifert and Webber, 1993). The hyphal system of the Ophiostomatoid fungi seems to be uniquely very adapted to penetrate wood. Hyphae move freely from one cell to another by growing through the pit membranes or directly through the cell wall using specialised constricted hyphal structures called transpressoria (Thwaites, 2003). The hyphae may be superficial however, they can also penetrate deeply into the sapwood, as is the case of O. novo-ulmi and O. piceae. This process can result in the blockage of the tree vessels with the result that water are incapable of reaching the branches of the tree. This causes wilting and eventually death of the tree, or just sapwood stain or

discoloration, respectively (Brasier, 1991; Uzunovic et al., 1999). Some Ophiostoma species produce conidia on asexual fruiting structures called synnemata, which is a group of conidiophores cemented together by their stalks (Pereira et al., 2000) or in a

(31)

Fungi in general are absorbotrophs. Their main nutrition is done by absorption. The hyphal system of Ophiostomatoid fungi appears to be uniquely adapted to externally digest, absorb and metabolise wood. The nutrients and substances in wood are broken down via enzymatic action. Like animals, fungi lack chlorophyll and do not

photosynthesize, so they must obtain nutrients from organic sources and store energy as glycogen instead of starch located in the tissue of the wood (Zabel and Morrel, 1992).

The predominant types of nutritive substances found in wood are present in the cytoplasm of parenchyma cells, vessels in the resin canals of the sapwood and lumen of tracheids. Two types of substances are found, although soluble sugars can constitute a major percentage of the non-primary carbohydrate present (Cranswick et al., 1987). The types of these two substances are described as hydrophilic compounds or energy storage such as proteins, starch, soluble sugars and amino acids, and hydrophobic that includes resins and wood extractives (Zabel and Morrell, 1992).

Many Ophiostomatoid fungi present a parasitic relationship with their host such as O. novo-ulmi and elm trees. Parasitic fungi in general produce extracellular specific enzymes that hydrolyse the living tissue in wood into assimilable nitrogen and carbon, usually sapping the energy of the host and quite frequently causing its demise.

1.6 Ophiostoma novo-ulmi (Brasier, Mycopathologia 115: 155 (1991). Life-Cycle and Infestation

Group: Opisthokonta Phylum: Ascomycota

Sub Phylum: Pezizomycotina Class: Sordariomycetes

(32)

Order: Ophiostomatales Family: Ophiostomataceae Genus: Ophiostoma Species: novo-ulmi

Subspecies: novo ulmi and americana

Ophiostoma novo ulmi occurs widely around the globe. It is believed to have originated in Asia. It displays dimorphic growth morphology, alternating between budding (yeast) and filamentous (mycelia) growth under different conditions. Budding O. novo-ulmi cells are typically the first to infect elm trees and this is likely due to their increased ability to spread from one tree to another. A single spore of this fungus is enough to spread the disease to a new host and eventually decimate an entire elm population. Ophiostoma novo-ulmi requires the aid of the bark beetles Scoltus multistriatus and Hylurgopinus rufipes in order for its spores to spread between

individual elm trees. The larvae of both S. multistriatus and H. rufipes excavate galleries underneath the bark in the cambial tissue of American elms. Each gallery contains a main chamber in which the eggs of the beetles are laid. After hatching the individual larvae chew their way out of the tunnel sides carrying fungal spores with them. Both sexual spores and asexual oidia or yeast derived from O. novo-ulmi form inside the tunnels are transported by adhering to the bodies of the bark beetles upon their emergence from the breeding galleries (Lanier, 1989).

When a tree is infected with O. novo-ulmi the spores migrate into the sapwood where it is installed. At this stage the fungus begins to produce enzymes such as xylanases (Binz and Canevascini, 1996), cellulases (Przybył et al., 2006), and

(33)

polygalacturonases (Niture, 2008) that are transported up the conductive vessels and begin to breakdown the xylem vessels of the tree. Furthermore, the susceptibility of American elms to DED has been positively correlated to the diameter of the xylem vessels (Schreiber et al., 1979; Schumann 1991). The larger the xylem vessels are the more susceptible the tree is to infection, as smaller vessels restrict the natural movement of fluid and budding cells (Solla and Gill, 2003). Therefore older trees are more inclined to be infected than younger trees.

As part of their natural defences against pathogens, elm trees react to the presence of fungal enzymes by plugging the cambial tissue in an attempt to stop the fungus from spreading further (Jeng et al., 2007). They also produce tyloses and gels that occlude the vessels (Dimond, 1995). The blockages that occur as a result of these natural defences inhibit the flow of water from the roots to the leaves causing wilting of the foliage. Consequently, DED is often referred to as a vascular wilt disease as the tree dies from lack of nutrients (Jeng et al., 2007).

In addition to causing vascular wilt, O. novo-ulmi mycelia may also block the conducting elements of the host tree causing the plant tissue to die. As the fungus spreads throughout the sapwood and the conductive wood of the roots, the mycelia can cross from the roots of one elm to another, as the roots of neighboring elms are often fused together (Upadhay, 1993). Therefore, once the bark beetles have transported O. novo-ulmi spores to a healthy tree, DED is able to spread rapidly between closely neighbouring trees. Following infection of DED, infected trees can lose 60 to 100 % of their foliage and have little chance of recovery (Temple et al., 1997). Furthermore, the root system starved of sugars (normally produced in the leaves), eventually dies.

(34)

(Upadhay, 1993). Some studies have been done on the biochemical and genetic properties of Ophiostoma (Blanchette et al., 1992; Farrell et al., 1993; Wendler et al., 1992; Schirp et al., 2003a, b; Temple et al., 1997 and 2008). Based on a significant effort by different groups two expressed sequence tag (EST) datasets for O. novo-ulmi and O. clavigerum (Grosmannia clavigera) have been published (Hintz et al., 2011; DiGuistini et al., 2007) as well as the genome for O. novo ulmi (Forgetta et al., 2013).

1.7 Dutch Elm Disease Mitigation Strategies

1.7.1 Chemical Control Strategies

Historically, attempts to control DED have adopted integrated pest management strategies using chemical control measures. Chemicals for DED control for both the fungal pathogen and the insect vector have been researched since the 1940s (Zentmeyer et al., 1946; Dimond et al., 1949; Kielbaso, 1978; Kondo et al., 1982; Miller, 1991). Hubbes (1999) explored sanitation efforts to prevent the spread of inoculums from infected trees. Systemic fungicides such as Arbotect were used widely at one time, however, due to significant detrimental and environmental harm their use has been restricted more recently (French et al. 1980; Haugen and Stennes, 1999).

1.7.2 Breeding Programs

Selective breeding programs to develop hardy DED-resistant American elms and hybrids have been underway for more than 70 years (Smalley and Guries 1993; Merkle et al. 2007). All North American elm species are highly susceptible to DED (Hubbes 1999). The main sources for DED resistance genes are Asian elm species Ulmus

(35)

parvifolia, U. pumila, U. propinqua, U. japonica, U. wilsoniana, and others (Smalley and Guries 1993; Hubbes 1999). This DED resistant phenomenon in Asian species is

believed to be due to the fact that the DED fungus originated in the Himalayas (Brasier and Mehrotra, 1995), and therefore Asian elms may have had longer evolutionary time to develop resistance. A number of DED-resistant hybrid elms have been developed from genetic combinations of Asian and, to lesser extent, European gene pools (Smalley and Guries 1993).

In North America there have been three major traditional breeding programs whose aims were to create DED-resistant American elms undertaken by: the Wisconsin, the Morton Arboretum and the US Department of Agriculture (USDA) (Hubbes, 1999). The USDA program has hybridized Eurasian elm species U. pumila, U. parvifolia and U. Japonica with American elm trees. They have also delivered two DED-resistant pure American elm cultivars to nurseries: the ‘Valley Forge’ and the ‘New Harmony’. These cultivars do not present immunity to DED but have increased resistance against the more aggressive subtype O. novo-ulmi (http://www.usna.usda.gov/Newintro/ american.html). Also an independent screening and selection breeding program (‘Liberty Elm’)

administered by the Elm Research Institute (http://www.libertyelm.com) has resulted in two varieties of resistant elm trees (Merkle et al., 2007).

The issue with DED-resistant cultivars is that their tree architecture is more similar to Eurasian elms than American elms (Newhouse et al., 2007). Given that one of the main uses of American elms has been as an ornamental tree, traditionally planted along roadways, prized for their arching branches, the architecture of the

(36)

2007). Also, although these straight selections by inoculation screening programs have resulted in DED-resistant trees, some other major challenges they have faced to date are their high cost, the time to deliver these hybrids (approximately 20 or more years) (Sinclair, 2001) and the difficulty of successfully hybridizing American elms with other elm tree species in order to introduce all the DED-resistant genes (unknown at a

molecular level) (Merkle et al. 2007). Therefore for this approach to be considered fully successful a hybrid which not only displays resistance to DED, but also has the

architecture of wild-type American elm trees should be the ultimate goal. To achieve this it is speculated that breeding may need to be carried out over a number of generations in order to slowly introduce the DED-resistant genes, and at the same time maintain the architecture of the wild type (Merkle et al. 2007). This presumes that there is a gene-for-gene correspondence between virulence and resistance in this pathosystem. It may be that resistance relies on physical features of the Eurasian elms that are not easily separated from the tree architecture. Furthermore, any DED-resistant trees which are produced need to undergo clonal propagation which would lead to a lack of genetic diversity and make large-scale plantings of these hybrids even more susceptible to other environmental problems and stochastic events, especially over a number of generations (Newhouse et al., 2007). An example of this is discussed by Sinclair (2001), who shows that while some American elm hybrids have shown increased resistance to DED they have still been susceptible to other diseases such as elm yellows caused by phytoplasmas.

(37)

In addition to the use of traditional breeding programs, transgenic technology has been applied to combat DED by the introduction of genes for anti-microbial peptides. These technologies have improved our ability to develop disease and pest-resistant genotypes of threatened trees species by directly introducing either genes from other organisms, or a synthetic gene that may provide pathogen resistance from other plants (Newhouse et al., 2005; 2006; Merkle et al., 2007). Concerns about the approach have been raised, specifically the potential of transgenic genes spreading from crop species to wild-type species and eventually outcompeting wild-type populations. The major advantage of gene modification, particularly for forest trees, is that it is much faster compared to conventional breeding approaches. In addition, transferring only the gene of interest into a tree genotype is more desirable than transferring larger portions of the genome which often necessitates breeding out any undesirable traits and can take many years.

The use of novel transgenic technology in the fight against DED has recently been applied in studies involving the introduction of a synthetic antimicrobial peptide (ESF 39) and a constitutively-expressed selectable marker (npt2) to American elms (Newhouse et al., 2007). In this study Newhouse et al (2007) successfully introduced a gene that encoded a cationic antimicrobial peptide (ESF39) to American elm trees. Antimicrobial peptides are often naturally produced by plants as part of their pathogen defence system and provide broad-spectrum anti-bacterial, anti-fungal, anti-viral and anti-protozoan properties (Schwab et al. 1999; Handock and Dianmond, 2000; Osusky et al. 2000; Tossi et al. 2000; Gura 2001; Rajasekaran et al. 2001; Ballweber et al. 2002). Furthermore, the synthetic derivatives of these peptides, such as the cationic antimicrobial peptide ESF39

(38)

used by Newhouse et al. (2006), have been shown to be efficient anti-microbial agents against many plant and animal pathogens (Schwab et al. 1999; Osusky et al. 2000; Tossi et al. 2000; Gura 2001; Rajasekaran et al. 2001; Ballweber et al. 2002).

The ESF39 peptide cassette used by Newhouse et al. (2006) was similar to the second structure of magainins (Zaoff, 1987) except that it contains a unique amino acid sequence designed to be quickly digested in mammalian digestive systems. This peptide has very little activity impact on plant and animal cells, but has been shown to inhibit the growth of selected plant pathogens (Powell et al., 1995; Powell et al., 2000; Powel and Maynard, 1997). This is similar to several other designs of constitutively expressed cationic anti-microbial peptides that have been found to enhance pathogen resistance in transgenic poplar trees (Liang et al., 2002; Mentag et al., 2003) and apple trees (Norelli et al., 1998). Enhancement of DED-resistance in juvenile transgenic American elm trees expressing ESF39 was first demonstrated by reduced vascular tissue staining of O. novo-ulmi-inoculated tissues in the absence of the pathogen in 2005, by Newhouse et al (2007). While the success of ESF39 in increasing DED-resistance in American elms in the field is still unknown, localized field tests that will compare these transgenic American elms and one of the ‘Liberty’ elm clones to their wild-type counterparts have recently been established. If these tests prove successful it will be a major breakthrough in the fight against DED.

1.7.4 Gene Regulation Approaches and RNA interference

Given the ineffectiveness of chemical control strategies and the limited success thus far of both traditional breeding programs and transgenic approaches to fight DED, it

(39)

is critical to pursue innovative control strategies for this disease. To date, O. ulmi, O. novo-ulmi and O. piceae are the most well characterised Ophiostoma species in terms of genetic makeup, protein production, expressed sequence tag (EST) libraries and more recently, complete genome (Forgetta et al., 2013; Hintz et al., 2011; Plourde et al., 2008; DiGuistini et al., 2007; Bernier et al., 2004; Temple et al., 1997, 2009; Abraham et al., 1995, 1998; Gao and Breuil, 1995, 1998; ; Binz et al., 1996, 1997; Bowden et al., 1996, 1994; Sutherland, et al., 1995; Takai, 1974; Beckman, 1956).

The first step in the process of managing DED through gene regulation is the identification of fungal-specific gene loci that determine species fitness and/or pathogen virulence. With the recent completion of the genome project and an EST library for O. novo-ulmi (Forgetta et al., 2013; Hintz et al., 2011), we can now potentially have a better understanding of the relevance of some genes related to fungal pathogenicity,

metabolism, virulence and host-pathogen interactions. The functional analysis of gene targets relies upon an effective method to regulate their expression. In the case of O. novo-ulmi, fungal virulence factors can be identified by modifying the expression of candidate genes, followed by an assessment of the effect of these changes on interactions between the fungal pathogen and the host. At least two approaches to the modification of gene expression can be explored: a gene locus conferring a putative virulence factor can be permanently disabled through targeted gene disruption and the resultant genetic mutant can be assessed for phenotypic traits such as morphology, enzyme production, or aggressiveness toward elm saplings; or a gene can potentially be regulated by RNA silencing, also called RNA interference (RNAi).

(40)

Common approaches to targeted gene disruption include replacing the wild-type gene with either a truncated version containing multiple stop codons, or an antibiotic resistance marker (Paddison et al., 2002; Mouyana et al., 2004). Unfortunately, one of the difficulties of targeted gene disruption in O. novo-ulmi is that it is a rather inefficient process. RNAi is a relatively new method also called post-transcriptional gene silencing (PTGS) that uses the natural ability of cells to destroy foreign RNA. It was first

demonstrated by Fire et al. (1998) who used it to block the expression of endogenous genes in a sequence specific manner in the nematode Caenorhabditis elegans. It is known that the ribonuclease Dicer enzyme binds and cleaves short dsRNA molecules to produce double-stranded fragments of 21-23bp with two base pair single-stranded overhangs on each end (Rose et al., 2005). The short dsRNA produced by Dicer, termed interfering RNAs (siRNAs), are believed to be separated by an enzyme with helicase activity and integrated into a multiprotein complex called the RNA-induced silencing complex (RISC). After integration into RISC, siRNAs direct this complex to the target messenger RNA (mRNA) through the complementary base pairing. This paring induces the RISC argonaute protein component of RISC to cleave the mRNA, thus preventing it from being used as a translation template (Campbell and Choy, 2005). This technique has been used to regulate gene expression in an array of organisms and under a number of various scenarios including O. novo-ulmi, O. floccosum and O. piceae (Liu et al., 2001; Kadotani et al., 2003; Agrawal et al, 2003; Cottrell and Doering, 2003; Mouyna et al., 2004; Moriwaki et al., 2007; Tanguay et al., 2006; Yamada et al., 2007; Carneiro et al., 2010). It has the potential for elucidating gene functions more efficiently than gene disruption and the interference effect is maintained in the progeny following repeated

(41)

sexual crosses as RNAi is transferred from one generation to the next (Fire., 1999; Liu et al., 2002; Nicolás et al., 2003; Milhavet et al., 2003).

Ophiostomataceae fungi typically produce extracellular enzymes that hydrolyse the macromolecules in wood into assimilable nitrogen and carbon (Abraham et al., 1998). Some of the extracellular enzymes produced by Ophiostoma species include pectinase, cellulase, β-galactosidase and xylanase (Beckman, 1956; Binz et al., 1996, 1997). One of the important secretory enzymes utilized by Ophiostoma species related to host invasion is the pectinase enzyme polygalacturonase. The fungi produce this enzyme to break down the middle lamella in plants so that it can extract nutrients from the plant tissue and insert fungal hyphae (Bussink et al. 1990, 1991; Abraham et al., 1998; Ranveer et al., 2005). Therefore one such target implicated in pathogenicity of O. novo-ulmi is the endopolygalacturonase (ePG) gene known as epg1. This gene locus belongs to the poly- galacturonase (PG) family of enzymes that catalyze the hydrolysis of pectin compounds that comprise approximately 30% of the primary cell wall in plants, and contribute to the integrity of plant tissues (Pilnik and Voragen, 1991; Juge, 2006; Mertens et al., 2008; Niture, 2008). The production of PG by pathogenic fungi that lack specialized

penetration structures is critical for their success and survival during host infection (De Lorenzo and Ferrari, 2002). To further assess the importance of this gene in strain pathogenicity and to use it as a model, an RNAi cassette was constructed using the epg1gene (Carneiro et al., 2010).

1.8 Viruses in fungi and the role of 93-1224 O. novo-ulmi strain against DED.

(42)

Temple et al (2006) described the presence of a dsRNA virus infecting O. novo-ulmi isolate strain 93-1224 from Winnipeg. It is speculated that the presence of dsRNA elements in fungi, specifically in Ophiostoma species, drives a high level of genetic variability in the pathogen (Brasier 1983a). Genetic variability and infection by dsRNA elements in O. novo-ulmi (also known as d factors) have been surveyed in both Europe and North America (Rogers et al., 1986; Charter et al., 1993; Temple et al 2006; Hintz et al., in preparation). In Winnipeg, Manitoba, Canada, only two isolates were found harboring dsRNA elements (Temple et al 2006), while in Europe a diversity of these elements have been found and genomes are available from NCBI Genome Data Bank. It is also known that there is lack of genetic diversity in the North American Ophiostoma population in contrast to Europe where genetically diverse and interbreeding populations are common. In an effort to explain this phenomenon, there has been renewed interest in extra-chromosomal genetic elements such as plasmids, transposable elements, introns and mycoviruses that might be modulators of fungal gene expression.

Mycoviruses (viruses infecting fungi) are quite common (Hollings 1962). The symptoms caused by the presence of a virus in the host organism may lead to attenuation of pathogenicity (hypovirulence) or enhanced aggression towards the host

(hypervirulence) (Hillman et al., 1990; Boland, 1992; Zhou and Boland, 1997; Deng et al., 2003; Deng and Boland, 2004). Observed effects of hypovirulence include various type of phenotypic changes resulting in sick morphology, slow mycelial growth, loss of female fertility in sexual crosses, reduction in asexual sporulation, changes to the mitochondria, reduced yield and others (Hillman et al. 1990, Nuss 2005, Ghabrial and Suzuki 2009, Pearson et al. 2009). Enhanced aggression may take the form of increased

(43)

susceptibility of the host to disease symptoms caused by the virus infected pathogenic fungus.

The minimal impact seen for many hypoviruses is thought to be the result of a long association and co-evolution between the virus and its host, resulting from selection against virulence in the viral parasite or selection for tolerance and resistance by the fungal host (Choi and Nuss et al. 1999). It is speculated that mycoviruses first evolved in plant hosts as a mechanism for defence against plant pathogens (Pearson et al. 2009).

Reduced pathogenicity as a result of viral infection of a fungal host is known as hypovirulence. The term hypovirulence was first used in the 1960s to describe reduced virulence in the Chestnut Blight pathogen d’Endothia parasitica now known as

Cryphonectria pasasitica (Gentre 1965). Viruses infecting C. parasitica strains resulted in a persistent set of phenotypic changes in the fungus diminishing its aggressiveness. Whereas virulent strains of the fungus penetrated and killed bark cambial tissue causing wilt and death, hypovirulent strains produced superficial cankers that were eventually detected by the defence response mechanism of the tree, and the reduction in virulence was notable (Fulbright 1984, Choi and Nuss 1992).

Mycoviruses have been identified in all major fungal families. Ophiostoma is one of several ascomycetous fungi that are known to harbor dsRNA (Temple et al., 2006, Buck and Brasier, 2002, Hong et al., 1999; Cole et al., 1998; Sutherland and Brasier, 1995). Their genomes can vary from dsRNA to single-stranded RNA (ssRNA). Often mycoviruses have more than one dsRNA present per virus particle (Bozarth 1972). For a virus to be considered a true mycovirus it must demonstrate an ability to be transmitted and infect other healthy fungi. Many dsRNA elements that have been described in fungi

(44)

do not fit this description, and in these cases they are referred to as virus-like particles (VLPs) (Nuss, 2005. Pearson et al., 2009; Ghabrial and Suzuki, 2009).

Transfer of the dsRNA, located in the cell cytoplasm or in the mitochondria, can occur during anatomosis of cells, or as a consequence of other events where cytoplasmic transfer occurs. Factors that could limit the dispersal of dsRNA in natural populations of this species may include the transmission of dsRNA through ascospores (vertical

transmission) and the restricted transmission among fungal mycelia as a consequence of vegetative compatibility groups (VCGs) (horizontal transmission) (Milgroom and Brasier,1997; Milgroom and Cortesi, 2004 ).

1.8.1 Diversity of viruses and Double-Stranded RNA Mycoviruses

Nucleotide sequence analyses of positive and negative-strand RNA, double-stranded RNA (dsRNA) viruses, and retroviruses have revealed an extensive diversity among each of these types of viruses. A summary of formally named and recognized mycoviruses is included in Gibbs et al. (2005) and Ghabrial and Suzuki (2009). Virus nomenclature is recognized by the International Committee for Taxonomy of Viruses (ICTV) which follows a hierarchical classification including species, genera and families. There are about 70 identified encapsidated mycovirus species that are included in 10 viral families and 20 that are still unassigned to a genus or in some cases even to a family. For many years mycoviruses were considered to be morphologically isometric particles constructed from monomers of a single capsid protein and dsRNA (Lemke and Nash, 1974) (families Totiviridae, Partiviridade, and Chrysoviridae). More recent evidence has shown however, that a variety of other particle morphologies may occur (Ghabrial and Suziki 2009; Choi and Nuss 1992; Howitt et al. 2001, 2006). The mycoviruses are

(45)

usually classified based on the number of genome segments into 8 major families: Totiviridae, Partiviridae, Hypoviridae, Narnavriridae, Barnaviridae, Pseudoviridae, Reoviridae and Chrysoviridae.

dsRNA mycoviruses are classically isometric particles, 25-50 nm in diameter, with their genomes encoding open reading frames (ORFs). Most have segmented genomes and present themselves as being encapsidated. Their icosahedral capsids range from concentric multi-shelled structures (cystoviruses and reoviruses), to simpler single-shelled structures (totiviruses, birnaviruses, partitiviruses, and chrysoviruses). The capsids consist of 60 asymmetric coat protein dimers in what is called T=2 organization. The capsid organizes the replicative complexes that are actively involved in genome transcription and replication (Ochoa et al., 2008; Person et al., 2009; Gibbs et al. 2005). They are classified according to the number of genome segments and their icosahedral capsid formation (i,e, Totiviridae, Birnaviridae, Partitiviridae, Cystoviridae,

Chrysoviridae, and Reoviridae) (Gibbs et al. 2005). Except for totiviruses, all have segmented genomes, with the segments numbering from 2-12 depending on the family or genus.

The unencapsidated dsRNA usually consist of multiple genome segments that are associated with cell membranes (Ghabrial and Suzuki 2009, Nuss and Koltin 1990). To explain why dsRNAs are so common in fungi, Buck (1986) proposed that dsRNA might be a more suitable genome for a persistently intracellular virus as it would not be directly competitive with its host dsDNA. A viral dsDNA genome could eventually result in a genetic load on the host system especially if the virus count is high and starts acting as a

(46)

sink for all of the cellular resources of the host (e.g. DNA polymerase and other enzymes, and even greater competition for RNA dependent RNA polymerase (RdRp).

Having a dsRNA genome does not preclude the replication process of the host genome by the mycovirus. For example: DNA topoisomerase I, encoded by the host, is essential for the replication of M dsRNA, a satellite dsRNA dependent on the totivirus ScV-L-A for replication and encapsidation (Ghabrial 1998). On the other hand, if the virus is preferentially dsRNA and it is directly encoding its own RdRp, the resulting drag on the cellular machinery is not as nearly as acute. Another speculated reason for the origin of dsRNA genomes in mycoviruses is that it resulted from the process of evolution within the hosts (Person et al., 2009; Gibbs et al. 2005). Organisms have evolved

accurate methods for recognizing and excising aberrant DNAs. As a result, it is more feasible that a RNA virus could evade the cellular defence mechanisms of an organism than a DNA virus. Although there are mechanisms that can respond to dsRNA viral infections such as RNAi, dsRNAs have also evolved a mechanisms to evade them Mahajan et al., 2009)

Mycoviruses are not strictly limited to infecting fungi and there is considerable potential for movement of viruses between groups of organisms. For example, the mycoviruses of the family Totiviridae have a nonsegmented dsRNA that codes for a capsid protein (CP) and RdRp, yet these mycoviruses can also infect protozoa. Members of the family Partiviridae have two segmented dsRNAs and have been found to infect plants. Viruses in the family Chrysoviridae are comprised of four dsRNA segments and primarily infect only fungi (Ghabrial and Suzuki 2009). Although phylogenetic analyses of the RdRps reveal that chrysoviruses are more closely related to totiviruses that are

(47)

known to infect protozoa, a similar virus causing disease in penaeid shrimp has been reported as well (Poulos et al. 2006). Other related dsRNA viruses can infect animals (bi- and tri-picoviridae) (Kazama and Schornstein 1972, Gibbs et al. 2005, Ghabrial and Suzuki 2009).

Despite the increasing understanding of mycoviruses, their characterization is often incomplete and their encoded proteins vary considerably depending on the genera, even within the same family. A common highly conserved protein among mycoviruses is RdRp. Phylogenetic analyses of RdRp sequences suggest that for dsRNA viruses, RdRps have a polyphyletic origin, but the positive-ssRNAs tend to be placed into a different super-group (Ghabrial and Suzuki 2009).

The RdRp plays an important feature in mycoviruses as it seems to be associated with initiating virus infectivity. The main function of RdRp is to catalyze the replication of RNA from an RNA template. This contrasts with the common or DNA-dependent RNA polymerase which catalyzes the transcription of RNA from a DNA template (Iyer, Koonin and Aravind 2003). The advantage that RdRp confers to viruses is that their replication lacks a DNA stage. There is no competition between the virus and the host cell for replicases. A disadvantage is that there is no 'back-up' DNA copy. In many eukaryotes RdRP is involved in the process of RNAi (Iyer et al., 2003). For the purpose of this thesis I will focus on the family Narnaviridae as it is the only family characterized for O. novo-ulmi mitovirus to date.

1.8.2 Family Narnaviridae

(48)

Genus: Narnavirus Genus: Mitovirus

The family Narnaviridae is comprised of viruses that have a single molecule of non-encapsidated positive-strand RNA of 2.3-3.5 kilobase-pairs (kb). They mainly encode a single protein of 80-104 kilo dalton (kDa) with amino acid sequence motifs characteristic of an RdRp (Solorzano et al., 2000). They are unique among other viruses because they lack a protein coat and therefore are known as capsid-less viruses. The loss of capsid is believed to be responsible for the loss of virus transmissibility via

extracellular routes, limiting these agents to vertical transmission only. The capsid‐less RNA viruses include the families Narnaviridae, Hypoviridae, and Endornaviridae as well as several viruses scattered among diverse viral lineages (King et al., 2011). The majority infect fungi and oomycetes but can also infect plants. Two genera have been described for this family: the genus Narnavirus where viruses remain within the

cytoplasm of the host and the genus Mitovirus that infect mitochondria (Cai et al., 2011).

1.8.2.1 Genus Mitovirus

Type species: Chryphonectria parasitica 1; Ophiostoma novo-ulmi mitovirus 6-Ld Members of the genus Mitovirus has been found associated with the pathogens of the Chestnut Blight fungus, C. parasitica), the DED fungus O. novo-ulmi, O. ulmi, and the dollar spot of turf grass fungus S.homeocarpta. They are viruses that have no true virions and are located in the mitochondria (Doherty et al., 2006; Yiguo et al., 1990). They show evolutionary affinity to RNA bacteriophages, suggesting a potential ancient origin from viruses of bacteria that gave rise to the mitochondrial endosymbiont of the