The Partnership Between Orchids, Fungi and Viruses

In bed with viruses: the partnership between orchids, fungi and viruses

Thesis presented by

Jamie Wan Ling Ong

For the degree of Doctor of Philosophy

School of Veterinary and Life Sciences

Murdoch University

2016

Declaration

I declare that this thesis is my own account of my research and contains as its main content, work which has not previously been previously submitted for a degree at any tertiary education institution.

Jamie Wan Ling Ong

Abstract

The Orchidaceae is the largest and most diverse angiosperm family comprising of five subfamilies, over 800 genera and over 26,000 species. In Western

Australia, there are over 450 indigenous orchid species across 40 genera, concentrated predominately within the South West Australian Floristic Region, but with a few species in the tropical Kimberley. The southern species are all terrestrial and most belong to the Diurideae tribe, which are primarily restricted to Australia and New

Zealand. To varying degrees, orchids rely on associations with other organisms, particularly fungi for nutrient provision and insects for pollination. The partnerships between the orchids, their fungal symbionts and insect pollinators are quite well studied in some cases. However, the ecological influence of viruses, in particular indigenous viruses, within these symbiotic partnerships remains largely unexplored.

Orchids cultivated for their flowers or vanilla are frequently infected by viruses, which are spread from plant to plant by vectors, husbandry tools and through vegetative propagation, and from place to place in infected propagules by trade. Only recently have wild orchids been shown to also harbour viruses.

In this research, we used a combination of high throughput sequencing approach, traditional techniques and informatics to examine the leaf tissues of indigenous terrestrial orchid plants growing in their natural habitats for virus infection.

Further, we isolated fungi that form mycorrhizal associations within cortical root cells of these plants and examined them for the presence of viruses. Terrestrial orchids and their fungal symbionts were sampled from 17 species across six genera (Caladenia,

Diuris, Drakaea, Microtis, Paraceleana and Pterostylis) during the winter (June to

iii

August) and spring (September to November) growing seasons. This study represents the first of viruses from the indigenous orchids and fungal species examined.

Thirty-two viruses, representing seven viral families and eight genera

(Alphapartitivirus, Betapartitivirus, Endornavirus, Goravirus, Hypovirus, Mitovirus,

Platypuvirus and Totivirus), were identified and characterised from wild plants of

Drakaea, Microtis and Pterostylis orchids and their fungal symbionts. Four of the viruses were identified from leaves of Drakaea species and Pterostylis sanguinea orchids and the remaining 28 viruses were from six isolates of orchid mycorrhizal fungi of the genus Ceratobasidium. All but one of the viruses found were novel, and most were from taxonomic groups not previously described in the Australian continent.

In three Ceratobasidium isolates studied, there were 5-13 virus species present in each. The presence of several closely-related bi-partite partitiviruses within the one host presented challenges in determining the numbers of species present and accurate pairing of virus segments. This study proposes solutions to address these problems, which will no doubt also arise in future metagenomics studies.

Two of the new viruses described formed the bases of new genera (Goravirus and Platypuvirus), while other viruses could be tentatively classified within known taxa, but were often genetically divergent from existing members. For example, two novel partitiviruses represent a lineage basal to existing members of Alphapartitivirus, pointing to Australia as an important location in partitivirus evolution. The richness and uniqueness of viruses found in this study are likely a reflection of the orchid and

fungal diversity of the region, itself a consequence of over 25 million years of relative geological and climatic stability. The surprisingly high numbers of mycoviruses detected from only a few fungal samples indicate that there is a rich virus association with fungal component of orchid biology and that orchid flora might represent a potentially enormous reservoir of novel viruses.

Table of contents

Declaration ...... ii

Abstract ...... iii

Table of contents ...... vi

Abbreviations ...... ix

Publications and presentations ...... xiii

Acknowledgements ...... xv

Chapter 1: Introduction ...... 1

1.1 Vulnerability of orchids ...... 1

1.2 Western Australian orchids ...... 3

1.3 W.A. orchids – plant/fungus/pollinator complex ...... 5

1.3.1 Orchid mycorrhizas ...... 5

1.3.2 Orchid pollination ...... 7

1.4 Orchid fungal and plant viruses ...... 8

1.5 Viruses identified from orchids of the south-west Australian floristic region ...... 9

1.6 Detection of plant viruses ...... 13

1.7 Next generation sequencing for virus discovery ...... 13

1.8 Aims of this research project ...... 15

Chapter 2: Characterization of the first two viruses described from wild populations of hammer orchids (Drakaea spp.) in Australia ...... 18

Chapter 3: The challenges of using high-throughput sequencing to track multiple new bi-partite viruses of wild orchid-fungus partnerships over consecutive years ...... 34

3.1 Abstract ...... 34

3.2 Introduction...... 34

3.3 Materials and methods ...... 36

3.3.1 Sample collection ...... 36

3.3.2 Fungal isolation from underground stems ...... 37

3.3.3 Nucleic acids extraction, cDNA synthesis and amplification ...... 38

3.3.4 Identification of fungi ...... 39

3.3.5 Sequencing data analysis...... 39

3.3.6 RT-PCR amplification of partitivirus segments ...... 40

3.3.7 5' UTRs alignments ...... 40

3.4 Results ...... 41

3.4.1 Partitiviruses ...... 41

3.4.1.1 Partitivirus CPs ...... 43

3.4.1.2 Partitivirus RdRps ...... 43

3.4.2 Most partitiviruses occurred in both years ...... 47

3.4.3 Matching partitivirus segments ...... 47

3.4.4 Other viruses and viral-like contigs ...... 52

3.5 Discussion ...... 52

3.5.1 Ceratobasidium as a virus host ...... 52

3.5.2 Australian partitiviruses in a world context ...... 54

3.5.3 The challenge of matching viral segments ...... 55

3.5.4 Virus composition of mycorrhizal strains ...... 57

3.6 References...... 59

Chapter 4: Australian terrestrial orchids and their fungal symbionts are hosts of novel and divergent viruses ...... 67

4.1 Abstract ...... 67

4.2 Introduction...... 67

4.3 Materials and methods ...... 69

4.4 Results ...... 69

4.4.1 De novo assembly ...... 69

4.4.2 Identity of fungi ...... 70

4.4.3 Viruses from orchid-associated mycorrhizal fungi ...... 70

4.4.3.1 Ceratobasidium mitovirus A (CbMVA): a proposed new mitovirus ...... 70

vii

4.4.3.2 Ceratobasidium virus A (CbVA): a proposed new mycovirus ...... 76

4.4.3.3 Ceratobasidium virus B (CbVB): a proposed new mycovirus ...... 77

4.4.3.4 Ceratobasidium hypovirus A (CbHVA): a proposed new hypovirus ...... 77

4.4.4 Virus-like sequences identified from leaf samples ...... 79

4.4.4.1 Pterostylis sanguinea virus A (PsVA), a myco-like virus from leaf tissue ...... 79

4.4.4.2 Pterostylis sanguinea totivirus A (PsTVA): three isolates of a proposed new totivirus

from orchid plants ...... 80

4.4.5 Partitiviruses and other virus-like sequences ...... 82

4.5 Discussion ...... 83

4.5.1 Classification of new viruses ...... 84

4.5.2 Host identification ...... 85

4.5.3 Viruses, fungi and orchids...... 85

4.6 References...... 88

Chapter 5: Novel Endorna-like viruses, including three with two open reading frames, challenge the taxonomy of the Endornaviridae ...... 96

Chapter 6: General discussion ...... 108

6.1 Plant and fungal viruses ...... 109

6.2 Diversity and uniqueness of new viruses ...... 113

6.3 Virus ecology and evolution ...... 115

6.4 Viruses and orchid biology ...... 117

6.5 Virus exchange between hosts? ...... 121

6.6 Importance of wild plant virology ...... 122

Appendix 1 ...... 125

References ...... 131

viii

Abbreviations

AP Alphapartitivirus

BaEV Basella alba endornavirus

Blast Basic local alignment search tool

BP Betapartitivirus

BPEV Bell pepper endornavirus

BRVF Black raspberry virus F

BSMV Barley stripe mosaic virus

BVQ Beet virus Q

BYMV Bean yellow mosaic virus

BYV Beet yellows virus

CbEVA Ceratabasidium endornavirus A

CbEVB Ceratabasidium endornavirus B

CbEVC Ceratabasidium endornavirus C

CbEVD Ceratabasidium endornavirus D

CbEVE Ceratabasidium endornavirus E

CbEVF Ceratabasidium endornavirus F

CbEVG Ceratabasidium endornavirus G

CbEVH Ceratabasidium endornavirus H

CbHVA Ceratobasidium hypovirus A

CbMVA Ceratobasidium mitovirus A

CbVA Ceratobasidium virus A

CbVB Ceratobasidium virus B

CCRSAPV Cherry chlorotic rusty spot associated partitivirus

CDD Conserved Domain Database

CHV1 Cryphonectria hypovirus 1

CHV2 Cryphonectria hypovirus 2

CHV3 Cryphonectria hypovirus 3

CHV4 Cryphonectria hypovirus 4

CMV Cucumber mosaic virus

CP Coat protein

CRP Cysteine-rich protein

CThTv Curvularia thermal tolerance virus

CymMV Cymbidium mosaic virus

DOSV Donkey orchid symptomless virus

DPCV Diuris pendunculata cryptic virus dsRNA Double-stranded RNA

DVA Drakaea virus A

ELISA Enzyme-linked immunosorbent assay;

FgHV1 Fusarium graminearum hypovirus 1

FIM Fungal isolation medium

GABrV-XL Gremmeniella abietina type B RNA virus XL

GLRaV1 Grapevine leafroll associated virus 1

GORV Gentian ovary ring-spot virus

GT Glucosyltransferase

HEL Helicase

HmEV1 Helicobasidium mompa endornavirus 1

ICRISAT International Crops Research Institute for the Semi-Arid Tropics

ICTV International Committee on Taxonomy of Viruses

IPVC Indian peanut clump virus

ITS Internal transcribed spacer

IUCN International Union for conservation of Nature

LeSV Lentinula edodes spherical virus

LeV Lentinula edodes mycovirus

MET/MTR Methyltransferase

ML Maximum likelihood

MP Movement protein

MyRV1-Cp9B21 Mycoreovirus 1-Cp9B21

NCBI National Center for Biotechnology Information

NNI Nearest-Neighbor-Interchange

OGSV Oat golden stripe virus

OrEV Oryza rufipogon endornavirus

ORF Open reading frame

OrMV Ornithogalum mosaic virus

ORSV Odontoglossum ringspot virus

OsEV Oryza sativa endornavirus

PaEV Persea americana endornavirus

PBNSPaV Plum bark necrosis and stem pitting-associated virus

PCV Peanut clump virus

PEV1 Phytophthora endornavirus 1

PgLV-1 Phlebiopsis gigantea large virus-1

PsTVA Pterostylis sanguinea totivirus A

PsVA Pterostylis sanguinea virus A

PMWaV-1 Pineapple mealybug wilt-associated virus 1

PvEV1 Phaseolus vulgaris endornavirus 1

PvEV2 Phaseolus vulgaris endornavirus 2

RcEV1 Rhizoctonia cerealis endornavirus 1

RcMV1-RF1 Rhizophagus clarus mitovirus 1

RdRp RNA dependent RNA polymerase

RMV-HR1 Rhizophagus sp. HR1 mitovirus

RsRV-HN008 Rhizoctonia solani RNA virus HN008

RT-PCR Reverse-transcription polymerase chain reaction

ScV-L-A Saccharomyces cerevisiae L-A virus

SsDRV Sclerotinia sclerotiorum debilitation-associated RNA virus

SsEV1 Sclerotinia sclerotiorum endornavirus 1

SsHV1 Sclerotinia sclerotiorum hypovirus 1 ssRNA Single-stranded RNA

TaEV Tuber aestivum endornavirus

TEM Transmission electron microscopy

TeMV Tuber excavatum mitovirus

TGBp Triple gene block protein

TMV Tobacco mosaic virus

Umv-H1 Ustilago maydis virus H1

UTR Untranslated region

VfEV Vicia faba endornavirus

W.A. Western Australia

YmEV Yerba mate endornavirus

YTMMV Yellow tailflower mild mottle virus

xii

Publications and presentations

Publications Ong, JWL, RD Phillips, KW Dixon, MGK Jones and SJ Wylie. 2016. Characterization of the first two viruses described from wild populations of hammer orchids (Drakaea spp.) in Australia. Plant Pathology 65 (1): 163-172. (Chapter 2)

Ong, JWL, H Li, K Sivasithamparam, KW Dixon, MGK Jones and SJ Wylie. 2016. The challenges of using high-throughput sequencing to track multiple new bi-partite viruses of wild orchid-fungus partnerships over consecutive years. (Chapter 3; Virology – provisionally accepted)

Ong, JWL, H Li, K Sivasithamparam, KW Dixon, MGK Jones and SJ Wylie. 2016. Novel Endorna-like viruses, including three with two open reading frames, challenge the taxonomy of the Endornaviridae. Virology 499: 203-211. (Chapter 5)

Li, H, C Zhang, H Luo, MGK Jones, K Sivasithamparam, SH Koh, JWL Ong and SJ Wylie. 2016. Yellow tailflower mild mottle virus and Pelargonium zonate spot virus co-infect a wild plant of red-striped tailflower in Australia. Plant Pathology 65 (3): 503-509.

Koh, SH, JWL Ong, R Admiraal, K Sivasithamparam, MGK Jones and SJ Wylie. 2016. A novel member of the Tombusviridae from a wild legume, Gompholobium preissii. Arch. Virol. 161(10): 2893-2898.

xiii

Presentations

MUPSA Multidisciplinary Conference – Perth, Australia, 2012; Royal Society’s Annual Postgraduate Symposium – Perth, Australia, 2012 Impact of viruses on the Drakaea/mycorrhiza/pollinator complex (Poster)

Australasian Plant Pathology Student Symposium – Perth, Australia, 2013 Impact of viruses on Western Australian terrestrial orchids

Murdoch VLS Poster Day – Perth, Australia, 2013 Viruses of Western Australian terrestrial orchids (Poster)

24th Combined Biological Sciences Meeting – Perth, Australia, 2014 Viruses associated with Drakaea orchids of Western Australia

11th Australasian Plant Virology Workshop – Brisbane, Australia, 2014 Viruses of Australian terrestrial orchids and associated mycorrhizal fungi

7th Next Generation Sequencing conference – Palmerston North, New Zealand, 2015 An abundance of viruses co-inhabit Australian indigenous terrestrial orchids and their fungal partners

12th Australasian Plant Virology Workshop – Perth, Australia, 2015 Viruses associated with Pterostylis vittata orchids and their fungal partner

Royal Society’s Annual Postgraduate Symposium – Perth, Australia, 2015 Use of NGS for characterisation of novel viruses associated with orchids

Murdoch VLS Poster Day – Perth, Australia, 2015 Next generation sequencing for virus discovery (Poster)

xiv

Acknowledgements

I would like to express my sincere gratitude to my principal supervisor, Dr. Steve

Wylie, for his encouragement and support. Thank you for making this experience both enjoyable and rewarding. Thanks also to my co-supervisors, Prof. Mike Jones and

Prof. Kingsley Dixon, for their advice and guidance.

I am extremely grateful to Dr. Ryan Phillips from Kings Park Botanic Gardens and

Parks Authority for his assistance in collection of orchid samples and for sharing his expertise.

To Dr. Hua Li and Prof. Krishnapillai Sivasithamparam, thank you for sharing your insights and for all your assistance and feedbacks. Many thanks also to all who have helped with field and lab work.

To my family and friends, thank you for the constant support and understanding.

I would like to acknowledge the financial support of Australian Orchid Foundation,

Australian Research council (Linkage Grant LP110200180) and Botanic Gardens and

Parks Authority.

Chapter 1: Introduction

The Orchidaceae is the largest and most diverse of all angiosperm families, with five subfamilies comprising over 800 genera and over 26,000 species (Govaerts et al., 2011, Hoffman and Brown, 2011). Their geographical habitats are wide-ranging with occurrence on all continents except Antarctica proper, but including some sub-

Antarctic islands (Dressler, 1981). The epiphytic (growing on other plants) orchids make up the majority of species in the family, and most of these are distributed in the tropics of South America and South-East Asia (Atwood, 1986; Cribb et al., 2003).

The non-epiphytic species are classified as either geophytic (terrestrial, soil-dwelling) or lithophytic (rock surfaces) types. Terrestrial orchids comprise about a third of all orchid species, with Indo-china and South-west Australia being regions of terrestrial orchid richness (Atwood, 1986; Cribb et al., 2003; Swarts and Dixon, 2009). They have perennating tubers or rhizomes (underground structures that survive for multiple growing seasons), which allow them to survive extreme and variable climates

(Rasmussen, 1995; Brundrett, 2014). In South-west Australia, all but one orchid,

Cryptostylis ovata, share a deciduous growth habit where the leaves and stems die down at the end of each growing season (Hoffman and Brown, 2011; Brundrett, 2014).

1.1 Vulnerability of orchids

Despite their diversity, many orchid species are vulnerable to threats of extinction (Cribb et al., 2003; Swarts and Dixon, 2009). More than 50% of orchid species listed in the International Union for conservation of Nature’s (IUCN) Red List of threatened species are categorised as threatened and over 25% of the listed genera contained threatened species (IUCN, 2013). Terrestrial orchids are particularly

vulnerable; they represent nearly half of the orchid extinctions despite only accounting for a third of orchid species (IUCN, 2013).

Leading threats to orchid species are often linked directly or indirectly to actions of Man. Although orchids have adapted to a wide range of habitats throughout the world, many are highly specialised and therefore are very sensitive to small habitat changes. Land clearance for developmental purposes, overgrazing and invasion of weeds are the leading threats (IUCN/SSC Orchid Specialist Group, 1996;

Koopowitz et al., 2003). The impact of these factors can be further compounded by

Man’s indirect influences on climate change and the spread of diseases and pests

(Swarts and Dixon, 2009). Harvesting of wild orchid populations for trade, medicine, food and personal collections are also contributing factors to the decline in wild orchid populations. Species most affected are those with desirable flowers or those that produce edible products such as salep and vanilla (IUCN/SSC Orchid Specialist

Group, 1996).

Intrinsic aspects of their biology, which include dependence on fungi and pollinators, also play a part in orchid vulnerability (Rasmussen, 1995; Zelmer et al.,

1996; Swarts and Dixon, 2009). All orchid species, each to a varying degree, rely on association with compatible mycorrhizal partners to provide them with the nutrients they require for germination and growth (Rasmussen, 1995; Zelmer et al., 1996). An orchid species that requires a specific mycorrhizal fungus may be more at risk than one that can form mycorrhizal associations with a range of fungi (Brundrett, 2007;

Swarts and Dixon, 2009). Most orchids are pollinated by insects, with many species utilising mimicry to deceive and attract the pollinating insects. Mimicry mechanisms

can sometimes be so specific that the species can be pollinated by only one species of insect (e.g. Drakaea orchids and Zaspilothynnus wasps; Peakall, 1990; Phillips et al.,

2014). Thus, any vulnerability of the pollinating insects can directly impact orchid reproduction (Tremblay et al., 2005; Jalal, 2012).

1.2 Western Australian orchids

The south-west Australian floristic region in Western Australia (W.A.) is one of only two flora biodiversity hotspots in Australia (Myers et al., 2000; Williams et al., 2011). The relatively wet region (302,627 km2) is bordered by ocean to its south and west, and by arid lands to its north and east (Hopper, 1979). Despite its ancient, weathered and seemingly unfavourable landscapes, the region has a species-rich flora.

More than 7000 native vascular plant species have been described from the region

(Hopper and Gioia, 2004). The region also represents one of the most diverse areas for terrestrial orchids (Cribb et al., 2003; Swarts and Dixon, 2009; Brundrett, 2014).

In W.A., there are over 450 wild orchid species across 40 genera; only one, Disa bracteata (South African orchid), is an alien species. Majority of these species can be found within the floristic region (Fig 1.1) (Hoffman and Brown, 2011; Brundrett,

2014; Western Australian Herbarium, 2015). Many of them belong to the Diurideae tribe, which is primarily limited to Australia and New Zealand (Kores et al., 2001).

This high level of floral species diversity has been primarily attributed to evolutionary responses of the plants to the area’s ancient stable landscapes and its Mediterranean- type climate (Cowling et al., 1996; Beard et al., 2000; Coates and Atkins, 2001;

Hopper and Gioia, 2004). Periodic minor disturbances such as drought, flood and fire are other possible contributing factors to the diversity (Cowling et al., 1996; Hopper and Gioia, 2004; Brundrett, 2007).

Figure 1.1. Distribution map of orchid species in Western

Australia and Australia (inset); https://spatial.ala.org.au/

W.A. terrestrial orchids have adapted to the south-west Australian

Mediterranean climate of cool wet winters followed by hot dry summers when they exist as dormant underground tubers (Brundrett, 2007). Approximately 25% of W.A. orchid species (103 species) are listed as critically endangered, endangered, vulnerable or extinct (State of Western Australia, 2015). Of these 103 orchid species,

41 are classified as declared rare flora while 62 are priority flora (State of Western

Australia, 2015; Western Australian Herbarium, 2015). The decline of W.A. floral populations, including orchids, is associated with the same anthropogenic processes that are threatening orchid species globally, with leading factors land clearing, changes to salinity and hydrology of habitats, weed and pathogen invasion, etc (Fig

1.2; Coates and Atkins, 2001; Swarts and Dixon, 2009; Brundrett, 2016).

Small populations Accidental destruction Climate extremes

Mining Salinity, hydrology

Dieback

Land clearing Feral animals

Invasive weeds

Figure 1.2. Processes associated with decline of plant populations in the

south-west Australian floristic region (adapted from Coates and Atkins, 2001;

Swarts and Dixon, 2009).

1.3 W.A. orchids – plant/fungus/pollinator complex

Orchids are dependent on symbiotic relationships they have developed with mycorrhizal fungi (nutrient provision) and pollinators (insects for pollination). The level of dependency on these partners varies between species, but without both of these partners, many orchids are unlikely to survive in the long term.

1.3.1 Orchid mycorrhizas

Mycorrhizal associations are symbiotic associations between fungi and their host plants. They are primarily responsible for the transfer of nutrients such as carbon, nitrogen, phosphorus and water (Brundrett, 2004). Mycorrhizal associations are generally mutualistic, with bi-directional nutrient exchange between fungi and plants

(Brundrett, 2004; Smith and Read, 2010). In general, the fungi extract nutrients from the surrounding soil, which are then transferred to the plants via their roots, and in

exchange, the plants provide the fungi with carbon (Smith and Read, 2010). This association allows plants to increase the available surface area from which they can extract nutrients from nutrient-poor soils (Brundrett, 2004).

The orchid-fungus association begins at seed germination. Orchid seeds are minute in size, ranging from 0.05 mm to 6 mm, and lack the nutrient storage required for independent germination (Arditti and Ghani, 2000). Therefore, an orchid seed in nature depends completely on its association with a compatible fungus to provide the required nutrients for germination, growth and protocorm development (Rasmussen,

1995). The mycorrhizal fungus colonises the seeds and forms pelotons, masses of undifferentiated hyphae, within the embryo (Zelmer et al., 1996). The external hyphae absorb nutrients and minerals from soil and surrounding plants, animals and microbial residues (Rasmussen, 1995; Brundrett, 2004; Smith and Read, 2010). Nutrients are then transferred to the internal hyphae within the root cortex, which are absorbed by the plant through ingestion of the pelotons (Zelmer et al., 1996).

Orchids and mycorrhizae generally share a higher level of specificity than most other plants (Brundrett and Abbott, 1991; Brundrett, 2004). Most orchids associate with fungi from a narrow phylogenetic range of basidiomycetes – part of the

Rhizoctonia alliance including those of the genera Ceratobasidium, Sebacina,

Thanatephorus and Tulasnella (Warcup, 1981; Bonnardeaux et al., 2007; Smith and

Read, 2010; Phillips et al, 2011). Some orchid species (e.g. Caladenia orchids) will only associate with specific fungal species while others (e.g. Microtis orchids) tolerate a broader range of fungal associations, forming associations with multiple and diverse fungal species (Brundrett, 2007). Mycorrhizal associations may change during the

orchid life cycle, as seen with Gastrodia elata (Tall Gastrodia) (Xu and Mu, 1990,

Dearnaley, 2007). Plant-fungus specificity can be a contributing factor to orchid rarity under circumstances that limit distribution of a specific fungus (Phillips et al, 2011).

W.A. terrestrial orchids are dependent on associated fungal partners because they effectively extend the nutrient-absorption ability of their small or non-existent root systems (Brundrett, 2007). There are five categories of fungal colonisation in terrestrial orchids – infection via the stem collar (base of leaf), stem tuber, underground stem, root and root stem (Ramsay et al., 1986). Stem collar infection is the most common category and can be found in genera such as Caladenia and

Drakaea (Ramsay et al., 1986). The position of the stem collar near the surface of the soil surface and in close proximity to most organic matter maximises the orchids’ chance of being infected by a compatible fungus (Ramsay et al., 1986).

1.3.2 Orchid pollination

In W.A., the Orchidaceae is the only large plant family that is exclusively pollinated by insects such as bees, beetles, fungus gnats and wasps (Brown et al.,

1997; Brundrett, 2007). Pollination of W.A. orchids can be categorised into five groups: (1) self-pollination (e.g. Microtis, Disa), (2) food reward – provide food rewards such as nectar (e.g. Cyrtostylis, Eriochilus), (3) food deception – mimic other food rewarding flower species (e.g. Caladenia, Diuris), (4) fungus deception – late- autumn and winter flowering orchids that grows in habitats preferred by fungi and mimic appearance of fungi or fungal oviposition sites (e.g. Corybas, Pterostylis) and

(5) sexual deception – mimic physical morphology and pheromones of female insects

(e.g. Drakaea, Paracaleana) (van der Cingel, 2001; Jersáková et al., 2006; Brundrett,

2007; Brundrett, 2014).

As with fungal compatibility, pollination of orchids can be highly specialised.

For example, each of the ten species of Drakaea orchid (hammer orchid) is pollinated by a different species of thynnine wasp (Zaspilothynnus sp.) (Peakall, 1990; Phillips et al., 2014). Such specialisation is hypothesised to promote genetic transfer between populations and thus, resulting in an increase level of outcrossing (Peakall and Beattie,

1996; Jersáková et al., 2006; Brundrett, 2007; Hopper, 2009). This specificity and specialisation of insect pollination can be both advantageous and disadvantageous. It can lead to speciation between orchid populations but can also lead to higher risks of extinction if the local pollinator population becomes limited (Tremblay et al., 2005).

Any habitat and environmental changes that influence the numbers of pollinators may have a flow-on impact on orchid reproduction.

1.4 Orchid fungal and plant viruses

Studies on orchid viruses have been predominately focused on viruses that are detrimental to commercially cultivated orchids and their spread via the international trade of orchid plants. Virology of wild native orchids remains a poorly understood area of orchid research, with far fewer studies being carried out on viruses that naturally infect wild orchids or their mycorrhizal symbionts.

Prior to this study, no definitive mycovirus has been characterised from orchid mycorrhizal fungi. However, virus-like double-stranded RNA (dsRNA) were detected from two fungal isolates isolated from orchids Dactylorhiza fuchsii and Encyclia

alata, and rod-shaped virus-like particles were extracted from Ceratobasidium cornigerum associated with the orchid, Spiranthes sirensis (James et al., 1998).

The majority of work on viruses of native orchids has been done in Australia.

Australian orchids are infected by both exotic and indigenous viruses (Mackenzie et al., 1998; Gibbs et al., 2000; Wylie et al., 2012; Wylie et al., 2013a; Wylie et al.,

2013b; Vincent et al., 2014). Gibbs et al. (2000) tested orchids across 72 genera, including Australian native species, and found 11 virus species representing five genera – Potexvirus, Potyvirus, Rhabdovirus, Tobamovirus, and Tospovirus.

Mackenzie et al. (1998) found a new virus, Ceratobium mosaic virus (genus Potyvirus, subgroup Bean common mosaic virus), infecting approximately one third of the captive orchid plants tested from 33 genera. Exotic viruses such as Odontoglossum ringspot virus (genus Tobamovirus) and Cymbidium mosaic virus (genus Potexvirus) were found in populations of indigenous orchids (Gibbs et al., 2000). Viruses infecting wild orchids, especially the exotic viruses, pose a potential threat to the viability of orchid populations by reducing longevity and fecundity of infected plants.

1.5 Viruses identified from orchids of the south-west Australian floristic region

Previous studies have identified both novel and well-known exotic viruses from terrestrial orchids in the south-west Australian floristic region (Table 1.1). Ten viruses have been described to date and belong to five viral families –

Alphaflexiviridae (genus Platypuvirus), Betaflexiviridae (genus Divavirus),

Luteoviridae (genus Polerovirus), Partitiviridae (genus Alphapartitivirus) and

Potyviridae (genera Poacevirus, Potyvirus). These viruses were identified from three species of Caladenia (C. arenicola, C. latifolia and C. paludosa), one species of

Cymbidium (C. canaliculatum, an exotic cultivated species), one unidentified species of Dendrobium (an exotic cultivated species), six species of Diuris (D. corymbosa, D. laxiflora, D. longifolia, D. magnifica, D. micrantha and D. pendunculata), one species of Drakaea (D. elastica), one species of Microtis and one species of

Thelymitra (Table 1.1).

All the novel viruses described from W.A. terrestrial orchids (Table 1.1) are proposed to be indigenous viruses that have co-evolved with their hosts in their natural environments. While exotic viruses have been shown to be detrimental to both cultivated and native orchids, causing decline in orchid populations (Mackenzie et al.,

1998; Gibbs et al., 2000; Wylie et al., 2013a), the ecological influence of indigenous viruses on indigenous terrestrial orchids remains unknown.

Table 1.1. Viruses isolated from Western Australian terrestrial orchids Virus species Virus classification Orchid species Reference (native or exotic) (family, genus) Caladenia arenicola Caladenia latifolia Caladenia virus A (native) Potyviridae, Poacevirus Wylie et al., 2012 Drakaea elastica

Bean yellow mosaic virus (exotic) Diuris corymbosa Blue squill virus A (native) Ornithogalum mosaic virus (exotic)

Donkey orchid virus A (native) Potyviridae, Potyvirus Diuris laxiflora Ornithogalum mosaic virus (exotic)

Wylie et al., 2013a Bean yellow mosaic virus (exotic) Diuris magnifica Ornithogalum mosaic virus (exotic)

Diuris pendunculata cryptic virus (native) Partitiviridae, Alphapartitivirus Diuris virus A (native) Diuris pendunculata Betaflexiviridae, Divavirus Diuris virus B (native) Turnip yellows virus (exotic) Luteoviridae, Polerovirus

Caladenia latifolia Donkey orchid symptomless virus (native) Alphaflexiviridae, Platypuvirus Wylie et al., 2013b Diuris longifolia

Caladenia paludosa Diuris longifolia Diuris micrantha Bean yellow mosaic virus (exotic) Microtis sp. Thelymitra sp.

Potyviridae, Potyvirus Vincent et al., 2014 Cymbidium canaliculatum Dendrobium sp. Diuris longifolia Potyvirusa Diuris micrantha Microtis sp. Thelymitra sp.

aUndetermined potyvirus

1.6 Detection of plant viruses

Traditionally, studies on plant viruses were done using classical (e.g. symptomology, transmission studies), visual (e.g. Transmission electron microscopy;

TEM), serological (e.g. enzyme-linked immunosorbent assay; ELISA) and molecular

(e.g. reverse-transcription polymerase chain reaction; RT-PCR) techniques (Adams et al., 2009a; Boonham et al., 2014). Some classical techniques are not of sufficient definition to identify viruses to the species level, and the serological and molecular ones usually require prior access to viral proteins or knowledge of virus nucleotide sequences (Adams et al., 2009a; Boonham et al., 2014).

High throughput sequencing was first introduced in 2000 and in combination with informatics, has been successfully used in the field of plant virology since about

2009 (Adams et al., 2009a; Al Rwahnih et al., 2009; Kreuze et al., 2009). Its main advantages over previous methods of virus identification are that it can be generic (no prior knowledge of the virus is required), price per nucleotide is greatly reduced and information on host response at the transcription level can be gathered simultaneously

(Adams et al., 2009a; Barba et al., 2014; Boonham et al., 2014).

1.7 Next generation sequencing for virus discovery

Illumina sequencers have been the most popular platform used in recent plant virus studies because it provides the depth of sequence coverage required, at a relatively low cost and with a low error rate, to identify the relatively small amounts of viral RNA from amongst host RNA species (Quail et al., 2012; Barba et al., 2014).

Illumina sequencers utilise sequencing by synthesis approach (Fig 1.3; Mardis, 2008;

Shendure and Ji, 2008). Fragmented single stranded templates are ligated to adaptors

and bound to the surface of a flow cell, followed by bridge amplification using DNA polymerase to produce multiple copies (clonal clusters). During each cycle, a single fluorescently labelled reversible terminator nucleotide is added and detected. This cycle is repeated at a base per cycle until sequences of the fragments are obtained.

Figure 1.3. Overview of Illumina sequencing by synthesis approach.

From Mardis (2008).

1.8 Aims of this research project

The study of viruses of endemic orchids and their fungal partners from a biologically important flora located in an isolated region of the planet should provide insights into the distribution, ecology and evolution of viruses. On a practical level, knowledge of indigenous and exotic virus infections of orchid populations will assist management programmes for endangered orchids, especially when plants are clonally propagated for re-establishment of wild populations in natural environments.

The aims of this research project are to:

(i) identify viruses associated with wild indigenous orchid populations (Fig 1.4;

Table A1) from the south-west Australian floristic region,

(ii) identify mycoviruses associated with fungal mycorrhizae associated with

terrestrial orchids,

(iii) assess diversity and evolutionary history of viruses, and

(iv) provide the basis for subsequent research to determine if virus infection might

influence the survival of wild orchid species.

(a) (b) (c) (d)

(e) (f) (g)

(h) (i) (j)

(k) (l) (m)

(n) (o) (p) (q)

Figure 1.4. Sampled terrestrial orchid species: (a) Caladenia flava (cowslip orchid),

(b) C. latifolia (pink fairy orchid), (c) Diuris magnifica (pansy orchid), (d) D. porrifolia (western wheatbelt donkey orchid), (e) Drakaea concolor (kneeling

hammer orchid; photo by N Hoffman and A Brown), (f) D. elastica (glossy-leafed hammer orchid; N Hoffman and A Brown), (g) D. glyptodon (king-in-his-carriage),

(h) D. gracilis (slender hammer orchid; N Hoffman and A Brown), (i) D. livida

(warty hammer orchid), (j) D. micrantha (dwarf hammer orchid; M Brundrett), (k) D. thynniphila (narrow-lipped Hammer Orchid; N Hoffman and A Brown), (l)

Paracaleana nigrita (flying duck orchid), (m) Pterostylis sp. (snail orchid), (n) P. recurva (jug orchid), (o) P. sanguinea (dark banded greenhood orchid), (p) Microtis media (common mignonette orchid) and (q) Thelymitra benthamiana (leopard orchid;

N Hoffman and A Brown) (Hoffman and Brown, 2011; Brundrett, 2014).

Chapter 2: Characterization of the first two viruses described from wild populations of hammer orchids (Drakaea spp.) in Australia

Plant Pathology (2016) 65, 163–172 Doi: 10.1111/ppa.12396

Characterization of the first two viruses described from wild populations of hammer orchids (Drakaea spp.) in Australia J. W. L. Onga*, R. D. Phillipsbcd, K. W. Dixoncd, M. G. K. Jonesa and S. J. a Wylie aPlant Biotechnology Group – Plant Virology, School of Veterinary and Life Sciences, Western Australian State Agricultural Biotechnology Centre, Murdoch University, Perth, Western Australia 6150; bEvolution, Ecology and Genetics, Research School of Biology, Australian National University, Canberra, Australian Capital Territory 0200; cKings Park and Botanic Garden, West Perth, Western Australia 6005; and dSchool of Plant Biology, University of Western Australia, Nedlands, Western Australia 6009, Australia

Sequences representing the genomes of two distinct virus isolates infecting wild plants of two members of the genus Drakaea (hammer orchids) in Western Australia are described. The virus isolated from Drakaea livida has a bipartite genome of 4490 nt (RNA1) and 2905 nt (RNA2) that shares closest sequence and structural similarity to members of the genus Pecluvirus, family Virgaviridae, described from legumes in the Indian subcontinent and West Africa. However, it differs from Pecluviruses by lacking a P39 protein on RNA2 and having a cysteine-rich protein gene located 3' of the triple gene block protein genes. It is the first peclu-like virus to be described from Australia. The name Drakaea virus A is proposed (DVA; proposed member of the family Virgaviridae, genus unassigned). The second virus isolate was identified from Drakaea elastica, a species classed as endangered under conservation legislation. The genome sequence of this virus shares closest identity with isolates of Donkey orchid symptomless virus (DOSV; proposed member of the order Tymovirales, family and genus unassigned), a species described previously from wild Caladenia and Diuris orchids in the same region. These viruses are the first to be isolated from wild Drakaea populations and are proposed to have an ancient association with their orchid hosts.

Keywords: conservation, Drakaea, orchid, Tymovirales, Virgaviridae, wild plant virus

Introduction ten members classified as threatened and protected under the Western Australian Wildlife Conservation The Orchidaceae is the largest and most diverse of Act 1950 and the Commonwealth Environment all angiosperm families, with five subfamilies, over Protection and Biodiversity Conservation Act 1999 800 genera and well over 26 000 species (Govaerts (Hopper & Brown, 2007). Members of Drakaea, et al., 2011; Brundrett, 2014). Habitat destruction, which is a genus endemic to southern Western the naturally small population sizes of many species Australia, are commonly referred to as ‘hammer and specialized ecological interactions are some of orchids’ because of the hinged, hammer-shaped the leading factors hypothesized to cause a decline labellum (Hopper & Brown, 2007). The threats of in abundance of Australian orchid species (Swarts & extinction faced by Drakaea have been attributed to Dixon, 2009; Phillips et al., 2011, 2014). The anthropogenic influences, which may disrupt the impact of viruses on the ecology and decline of wild specialized partnerships they have with a single orchids is largely unknown, although exotic viruses species of mycorrhizal fungus (Tulasnella sp.) and such as Bean yellow mosaic virus (BYMV) and pollinating thynnine wasps (Ramsay et al., 1986; Ornithogalum mosaic virus (OrMV) are known to Phillips et al., 2014). The provision of mineral be pathogenic in both natural and ex situ substrates for germination and protocorm populations (Wylie et al., 2013a). development by mycorrhizal fungi compensates for The terrestrial orchid flora of southern Australia the lack of nutrient storage in orchids’ minute seeds is diverse, with a high incidence of intrinsically rare (Ramsay et al., 1986). This association is species, which typically exhibit specialization of maintained into adulthood, with the fungus pollination strategy and/or habitat requirements reinfecting the orchid at each growing season (Phillips et al., 2011). Among southern Australian (Ramsay et al., 1986; Swarts & Dixon, 2009). The orchids, the genus Drakaea has one of the highest highly specific pollination process is achieved by incidences of rarity with five of its attracting male thynnine wasps to the flower through the release of chemicals that mimic sex *E-mail: [email protected] pheromones of female wasps (Peakall, 1990; Bohman et al., 2014; Phillips et al., 2014). Drakaea plants produce a solitary flower Published online 21 May 2015 annually on a slender stem of 10–45 cm in height and one small heart-shaped leaf of 1–2 cm diameter

that grows flat on the ground (Fig 1; Hopper & ª 2015 British Society for Plant Pathology Brown, 2007; Brundrett,

163

164 J. W. L. Ong et al.

(a) ble roles of viruses in Drakaea biology remain largely unexplored. The only virus identified from Drakaea orchids is the proposed poacevirus Caladenia virus A,

which was recently identified from an ex situ population of Drakaea elastica (Wylie et al., 2012). Identifying and understanding the impact of Drakaea-associated viruses as either pathogens in wild Drakaea populations or as long-term symbiotic partners is important fot orchid conservation. Here, an unbiased high-throughput sequencing approach was used to identify RNA viruses infecting wild plants of seven Drakaea species growing in natural populations. The characteristics and phylogenies of the genome sequences of two viruses found were determined and Figure 1 Drakaea livida (a) flower, (b) leaf. Scale bar = 1 cm. possible implications for the ecology of Drakaea are

discussed.

2014). They are perennial geophytic herbs, with leaf emergence occurring in autumn. At the end of each Materials and methods flowering season (August to October depending on the species), the above-ground parts senesce and the Plant materials orchids produce new tubers, which enable persistence until the following growing season (Hopper & Brown, During winter and spring of 2012 and 2013, partial leaves or other plant material were collected from 162 plants of 22 wild 2007). populations of Drakaea representing 7 of the 10 species (Fig Drakaea have highly specialized above- and below- 2; Table S1). ground ecological interactions (Phillips et al., 2014) and, as such, might provide an interesting system for RNA extraction, cDNA synthesis and amplification the investigation of viruses and their transmission between interacting partners. Like many other southern Tissue from 2–13 plants of the same species and population Australian orchids, Drakaea belongs to the tribe were pooled and sequenced together. Samples of 80–100 mg Diurideae, which is primarily restricted to Australia of leaf or plant material were subjected to RNA extraction by and New Zealand (Kores et al., 2001). As such, the either of two methods. Total RNA was extracted from samples collected in 2012 (DR01–17) using an RNeasy kit viruses associated with Australian orchids might be (QIAGEN) in accordance with manufacturer’s protocol. For indigenous to the region and unique compared to those samples collected in 2013 (DR18–29), total RNA was recorded in other groups of orchids. Mixtures of both enriched for double-stranded RNA (dsRNA) using a exotic and indigenous viruses representing four cellulose-based method (Morris & Dodds, 1979). families have been identified from Australian native cDNA synthesis was carried out on heat-denatured RNA in orchid species (Gibbs et al., 2000; Wylie et al., 2012, a 20 lL volume containing 1 9 GoScript RT buffer (Promega), ' 2013a,b, 2014). Currently, the presence and possi- 3 mM MgCl2, 0 5 mM dNTPs, 0 5 mM random primer (5 -

Figure 2 Distribution map of sample collection sites generated using GPS VISUALIZER. Detailed information of collected samples is shown in Table S1

Plant Pathology (2016) 65, 163–172

Drakaea viruses 165

' CGTACAGTTAGCAGGCNNNNNNNNNNNN-3, where N Drakaea virus A (DVA) host range survey is any nucleotide), and 160 U M-MLV reverse transcriptase A survey of DVA host range was carried out by sampling leaf (Promega). cDNA synthesis incubation conditions were 5 min materials from the Drakaea livida population that was the at 25°C, 60 min at 42°C and 15 min at 70°C. source of the virus, and surrounding plants from 12 other PCR amplification was performed using individually species within 12 genera (eight families) at Canning Mills in ' tagged (barcoded) primers (5- the Darling Ranges (32°04'54.2''S, 116°05'27.6''E) in ' XXXXCGTACAGTTAGCAGGC-3) consisting of different September 2014 (Table S3). The selection comprised five combinations of 4-nt barcodes (XXXX, e.g. AGAG and species in the Orchidaceae and eight species chosen to ' AGAA) at the 5 end of a 16 nt adaptor sequence that represent the most abundant eight plant families at the study annealed to the complementary sequence of the cDNA site. dsRNA isolation from leaves, cDNA synthesis and synthesis primer. These primers added a unique barcode label amplification were carried out as above. Presence of DVA to each sample, enabling multiple samples to be pooled, was confirmed by amplification of a 781 bp band using sequenced and later sorted into individual samples. primers DVA-5 and DVA-6 (Table S2). Amplification was performed in a 20 μL volume containing DVA-infected leaf material from D. livida was macerated -1 -1 1x GoTaq Green Master Mix (Promega), 1 mM barcode with inoculation buffer (11.5 g L Na2HPO4, 2.96 g L primer and 2 μL of cDNA (approximately 10–50 ng). NaH2PO4, pH 7.2) and Celite (diatomaceous earth). The Reactions were carried out in a 2720 thermal cycler (Applied extract was then manually inoculated onto fresh leaves of Biosystems) and consisted of an initial cycle of 3 min at Drakaea glyptodon, Nicotiana benthamiana accession RA-4 95°C; 35 cycles of 30 s at 95°C, 30 s at 60°C and 1 min at and Chenopodium amaranticolor, with three replicates per 72°C; and a final extension at 72°C for 10 min. species. Two weeks after inoculation, inoculated and new The amount of each amplicon was estimated by running a 4 leaves from each plant were both tested for presence of DVA μL aliquot on an agarose gel and comparing fluorescence to a using primers DVA-5 and DVA-6 (Table S2). standard. The remaining amplicons were pooled in approximately equimolar amounts and purified using QIAquick PCR Purification kit (QIAGEN) prior to quantification using a Results ND-1000 spectrophotometer (ThermoFisher Scientific). Ten micrograms of pooled amplicons were submitted for library Three indexed sequence data sets of 153 582 198, 92 construction followed by high-throughput sequencing of 046 118 and 35 630 376 101-nt paired-end reads were paired ends over 100 cycles in a HiSeq2000 machine generated from three independent Illumina sequencing (Illumina) at either the Australian Genome Research Facility runs. From each respective data set, 25 791 170, 6 560 (Melbourne, Australia) or Macrogen Inc. (Seoul, South 986 and 9 031 108 of the reads were derived from Korea). Drakaea samples. The reads were separated into

sample bins by identifying indices, then index/adaptor Sequencing and analysis sequences were removed and de novo assembly carried De novo assembly of 100 nt paired reads was done using the out to generate contigs of >200 nt for BLAST analysis. de novo assembly application within CLC GENOMICS WORKBENCH v. 6.5.1 (QIAGEN). Contigs greater than 200 Sequence analysis of DVA nt in length were subjected to BLASTN and BLASTX (Altschul et al., 1990) analysis of NCBI GenBank databases Partial genome sequences were attained by Illumina (http://blast.ncbi.nlm.nih.gov/) to identify contigs with sequencing. Gaps predicted in the genome were filled nucleotide or amino acid sequence identity (e-value <1) with using RT-PCR with primers designed to flank the gaps, known viruses. Putative viral contigs identified this way were followed by direct Sanger sequencing of the PCR submitted to the NCBI conserved domain database (CDD) products as described above. In cases where (http://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) to ambiguous nucleotides were observed between identify domains with identity to those of known viruses Illumina and Sanger sequencing data, Sanger (Marchler-Bauer & Bryant, 2004). Open reading frames (ORFs), deduced encoded proteins and their domains were sequences were used when there was a consensus annotated using applications within GENEIOUS v. 7.0.6 between forward and reverse sequence reads. Six (Biomatters; Kearse et al., 2012). Contigs that did not match hundred and sixty-four raw sequence reads were known sequences from any source were analysed for the mapped to the RNA1 genomic sequence (putative presence of ORFs using GENEIOUS, and compared against replicase gene) with pairwise identity of 94.6% and the NCBI database in all six reading frames using BLASTX. 11.2-fold mean coverage across the genome. The Putative virus-derived sequences were compared to nucleotide composition of RNA1 was 28.8% adenine, genomes of predicted relatives to confirm the approximate 14.2% cytosine, 25.5% guanine and 31.5% uracil. order of ORFs and to identify possible gaps. Primers were designed on either side of gaps and reverse transcription (RT) Surprisingly, about five times more reads (3109) PCR performed using RNA from an infected plant to amplify mapped to the RNA2 sequence (putative coat protein, the missing sequences (Table S2). After Sanger sequencing movement proteins and cysteine-rich protein genes). using BigDye v. 3.1 terminator mix (Applied Biosystems), Pairwise identity amongst RNA2 reads was 85.2% and the sequences of the RT-PCR amplicons were used to mean coverage was 86.9-fold. Predicted nucleotide assemble the complete genome sequence. composition of RNA2 was 25.5% adenine, 20.5% Phylogenetic analyses of amino acid sequences were per- cytosine, 22.5% guanine and 31 5% uracil. formed with CLUSTALW using the default setting and ‘Find The virus represented by the sequences was best DNA/Protein models (Maximum Likelihood, ML)’ within MEGA v. 6.06 (http://www.megasoftware.net/) designated Drakaea virus A isolate Canning Mills (Tamura et al., 2013). Maximum likelihood (ML) trees with (GenBank accession nos. KP760461 and KP760462), 1000 bootstrap replications were constructed with nearest following the name of the original host plant species neighbour interchange (NNI) as the ML heuristic method. and the location in the

Plant Pathology (2016) 65, 163–172

166 J. W. L. Ong et al.

Darling Ranges from where it was isolated. The virus shared 47% aa (54% nt) identity to the homologous was identified in one of the two D. livida plants region of its closest known relative, Peanut clump (DR03) analysed from the Canning Mills population. virus (PCV; genus Pecluvirus) (Tables 1 & S4). A BLAST analysis of the complete viral sequence putative readthrough stop codon (UGA, arrow in Fig indicated that it shared greatest nucleotide (nt) and 3a) at 3105 nt of DVA RNA1 is also present in the amino acid (aa) identities with bipartite single-stranded replicase protein of the Pecluviruses PCV (UGA, 3567 (ss) RNA viruses within the family Virgaviridae nt) and Indian peanut clump virus (IPCV) (UGA, 3523 (Table 1). DVA RNA1 was 4490 nt in length, which nt). corresponded to a single ORF with three predicted The DVA RNA2, of 2905 nt, was predicted to domains: methyltransferase (MET; 3–1539 nt), encode five proteins in all three plus-sense reading helicase (HEL; 2202–3011 nt) and RNA-dependent frames (Fig 3a). The complete sequence of the coat RNA polymerase (RdRp; 3261–4490 nt; Fig 3a). The protein (CP) was not obtained; the partial CP shared core RdRp motifs V and VI (SG/TGx3 Tx3 NS/NTx22 43% aa (51% nt) identity with the CP of IPCV, 30– GDD) (Koonin, 1991) were present at 1360– 1395 aa 54% aa (49–57% nt) identity with triple gene block (4080–4187 nt) as SGx3 Tx3 NTx22 GDD. BLAST proteins (TGBp) 1, 2 and 3 of IPCV and Beet virus Q analyses revealed that the DVA replicase sequence (BVQ; genus Pomovirus), and 30% aa identity with a hypothetical protein of

Table 1 BLAST analysis of predicted gene products of Drakaea virus A isolate Canning Mills and Donkey orchid symptomless virus isolate Capel

Predicted Amino Location molecular GenBank Query acid Putative gene on genome weight accession coverage identity Virus product (nt) (kDa) Closest match using BLASTP of match (%) e-value (%)

Drakaea virus A Replicase (partial) 1–4490 172 Replicase (Peanut clump NP_620047 100 0.0 47 virus) [Virgaviridae, Pecluvirus] Coat protein 1–587 22 Coat protein (Peanut clump AAO15507 92 1e–51 51 (partial) virus N) [Virgaviridae, Pecluvirus] Triple gene block 682–1767 40 First triple gene block protein NP_620030 87 4e–77 44 protein 1 (Peanut clump virus) [Virgaviridae, Pecluvirus] Triple gene block 1751–2128 12 Triple gene block protein 2 AGG82480 82 1e–35 65 protein 2 (Potato mop-top virus) [Virgaviridae, Pomovirus] Triple gene block 1962–2411 17 P17 protein (Peanut clump AAO15516 97 5e–21 37 protein 3 virus M) [Virgaviridae, Pecluvirus] Cysteine-rich 2460–2879 16 Hypothetical protein NP_059487 66 0.004 30 protein Ogsvs2gp3 (Oat golden stripe virus) [Virgaviridae, Furovirus] Donkey orchid 68 kDa protein 108–1997 68 69 kDa protein (Donkey AHA56699 80 1e–94 49 symptomless orchid symptomless virus virus isolate isolate Mariginiup12) Capel Replicase 113–4300 157 Replicase (Donkey orchid YP_008828152 99 0.0 78 symptomless virus isolate Mariginiup11) 42 kDa protein 4325–5464 42 44 kDa protein (Donkey YP_008828153 100 0.0 73 orchid symptomless virus isolate Mariginiup11) Coat protein 5498–6106 22 Coat protein (Donkey orchid YP_008828154 100 8e–131 87 symptomless virus isolate Mariginiup11) 31 kDa protein 6136–6939 31 27 kDa protein (Donkey AHA56703 89 6e–111 67 orchid symptomless virus isolate Mariginiup12) 14 kDa proteina 6308–6712 14 – – – – – Movement protein 6953–7660 26 Movement protein (Donkey YP_008828156 98 1e–150 85 orchid symptomless virus isolate Mariginiup11) a14 kDa proteins of DOSV-Mariginiup11 and DOSV-Mariginiup12 are not illustrated in the NCBI database record.

Plant Pathology (2016) 65, 163–172

Drakaea viruses 167

(a)

Figure 3 Genome organization (a) and phylogenetic analysis (b) of Drakaea virus A. (a) Shaded boxes within the replicase open reading frame represent methyltransferase (MET), helicase (HEL) and RNA-dependent RNA polymerase (RdRp) domains. Nucleotide positions are shown. An arrow indicates the position of a proposed readthrough stop codon. CP, coat protein; TGBp, triple gene block protein; CRP, cysteine-rich protein. (b) Maximum-likelihood (b) tree of replicase proteins of viruses from the six genera within the family Virgaviridae. Genus names are shown on the right. Drakaea virus A is indicated with a dot. For a comparable analysis, the MET-HEL domain on RNA1 and RdRp domain on RNA3 of Barley stripe mosaic virus were combined to form the replicase protein. The tree was constructed with 1000 bootstrap replications and confidence values of less than 60% were omitted. Beet yellows virus (Closteroviridae) was used as the outgroup

Oat golden stripe virus (OGSV; genus Furovirus) The only other proposed member of the (Tables 1 & S4). The DVA CP of 22 kDa belongs to Virgaviridae identified from the region’s native flora is the same coat protein family as members of genera Yellow tailflower mild mottle virus (YTMMV; genus within the Virgaviridae family, including Tobamovirus Tobamovirus) (Wylie et al., 2014), which was isolated and Hordeivirus. Triple gene block proteins (involved from an Australian member of the family Solanaceae, in cell-to-cell movement) shared 27–54% aa (49–59% Anthocercis littoria. Comparison of DVA with nt) identity to homologues from members of the YTMMV showed they were only distantly related: genera Hordeivirus, Pecluvirus and Pomovirus (Table their respective replicases shared 23% aa (46% nt) S4). DVA TGBp1 has a predicted mass of 40 kDa and identity and their CPs shared 16% aa (43% nt) identity was located at 682–1767 nt. The presence of a helicase (Table S4). domain within TGBp1 at 1003–1680 nt is consistent with viruses in Hordeivirus, Pecluvirus and Pomovirus Drakaea virus A shares identical genome genera. The TGBp2 (12 kDa) and TGBp3 (17 kDa) organization to a recently identified virus, Gentian were located at 1751–2128 nt and 1962–2411 nt ovary ring-spot virus (GORV), reported from the respectively (Fig 3; Table 1). A CDD search showed ornamental plant Gentiana triflora (Atsumi et al., presence of a plant virus movement protein domain in 2015) that originates in China, eastern Russia, Japan TGBp2, which is shared amongst members of ssRNA and Korea. DVA and GORV shared 47% aa (55% nt) viral genera such as Potexvirus and Hordeivirus identity between replicases, 36% aa (50% nt) between (Marchler-Bauer et al., 2013). A ‘viral Beta C/D-like CPs, 29–46% aa (48–55% nt) between homologues of family’ domain which corresponds to TGBp3 of TGBps and 17% aa (43% nt) between CRPs (Table members of family Virgaviridae, was detected within S4). These percentage identities were similar to those TGBp3 at 1980–2321 nt. TGBp3 shared 27–32% aa between DVA and other viruses within Virgaviridae (49–53% nt) identity with members of Hordeivirus, (Table S4). Phylogenetic analysis of the putative Pecluvirus and Pomovirus (Table S4). The 16 kDa replicase protein placed DVA and GORV together in a cysteine-rich protein (CRP), located at 2460–2879 nt, sister group to a clade containing the Pecluviruses shared low (10–26%) aa identity with CRPs from PCV and IPCV (Fig 3b). Like DVA and GORV, some members of the Virgaviridae that are responsible Pecluviruses are bipartite ssRNA (+ sense) viruses, for viral suppression of RNA silencing (Adams et al., with one RNA segment encoding a replicase and the 2012b). In DVA, the common Virgaviridae CRP motif second segment encoding the CP and TGBps. of CGx2 H was present at 2637–2651 nt and 60–64 aa However, both DVA and GORV differ from members as CGEKH (Te et al., 2005). The CRP shared highest of Pecluvirus by not having a P39 protein, which is aa identity (26%) with CRP of PCV (Pecluvirus). believed to be involved in transmission by fungi, located 3' to their CP in the genome, and by encoding their CRP on RNA2 instead of RNA1 (Herzog et al., 1994; Adams et al., 2012b).

Plant Pathology (2016) 65, 163–172

168 J. W. L. Ong et al.

Incidence and transmission of DVA were seven predicted ORFs, four of which overlapped. The ORFs ranged from 405 nt (14 kDa protein) to In 2014, a survey of five orchid species and eight non- 4188 nt (replicase; Fig 4a). BLASTP analysis of deduced orchidaceous species growing at the site of collection amino acid sequences of each ORF revealed that each of the original DVA isolate revealed the presence of shared greatest identity with those of DOSV (Table 1). DVA infecting one D. livida plant (CM01), but not the Because of its identical genome organization and high other plants tested (CM02–12). Slight discolouration sequence identity (Tables S5 & S6) with isolates of was observed on the leaf of the infected D. livida plant DOSV, the new sequence was designated as Donkey (Fig 1b), but Koch’s postulates were not carried out to orchid symptomless virus isolate Capel (GenBank determine if the discolouration was associated with accession no. KP760463), the isolate name following the locality in which the infected host plant grew. DVA infection. Inoculation of DVA onto a single plant DOSV-Capel has a 5' UTR of 107 nt and the first AUG of D. glyptodon growing in a greenhouse resulted in began at nt 108, corresponding to the start of a 68 kDa systemic infection of the plant as determined by RT- protein. The replicase protein, which overlapped the PCR assay with DVA-specific primers, confirming putative 68 kDa protein, had a predicted mass of 157 that DVA was transmissible between Drakaea species. kDa and was located at nucleotide positions 113–4300 No symptoms typical of virus infection were observed (Tables 1 & S5). It shared 78% aa (72% nt) identity on the inoculated plant. DVA-inoculated plants of N. with the replicases of other DOSV isolates. The NCBI benthamiana and C. amaranticolor did not become CDD database was used to predict the locations of its locally or systemically infected. three domains: MET (nt 221–1075), HEL (nt 2000– 2677) and RdRp (nt 3233–4090) (Fig 4a). The RdRp Donkey orchid symptomless virus (DOSV) core motifs of TGx3 Tx3 NTx22 GDD (Koonin, 1991) were located at aa 1185–1220 (nt 3665–3772). The CP For sample DR26, which was derived from two D. (22 kDa) shared 87% aa (74–75% nt) identity with elastica plants, 88 contigs >200 nt in length were both previously sequenced DOSV isolates. The 26 kDa generated from 214 246 reads. Of these contigs, 14 had putative movement protein shared 85% aa (74–76% high sequence identity to sequences of DOSV nt) identity to both DOSV isolates (Table S6) and low (accession no. NC_022894 and KC923235; Wylie et identity (19%) to the next closest match, Sorghum chlorotic spot virus (SCSV; Virgaviridae, Furovirus). al., 2013b). The contigs were mapped to the published The movement protein (MP) was terminated by a sequences to generate a complete genome sequence. A UAA stop codon at 7658–7660 nt, followed by a 3' total of 12 432 reads were mapped to DOSV sequences, UTR of 110 nt. Four other predicted proteins of and there was a pairwise identity of 93.5% to the unknown function (68, 42, 31 and 14 kDa) shared consensus sequence of the new isolate following lower identities (12–73% aa and 41–72% nt) with sequencing of its complete genome. Mean sequence homologues in DOSV isolates Mariginiup 11 and 12 coverage of each nucleotide was 139.2-fold. (Table S6). The 14 kDa protein shared the least Analysis of the sequence revealed an RNA genome identity with other DOSV isolates, at less than 22% aa of 7770 nt, with nucleotide composition of 25.8% and 45% nt identity. adenine, 34.5% cytosine, 22.0% guanine and 17.7% uracil. There

(a)

Figure 4 Genome organization (a) and phylogenetic analysis (b) of Donkey orchid symptomless virus isolate Capel. (a) Shaded boxes within the replicase represent methyltransferase (MET), helicase (HEL) and RNA-dependent RNA (b) polymerase (RdRp) domains. CP, coat protein; MP, movement protein. (b) Maximum likelihood analysis of amino acid sequences of replicases of representative species of five genera within the family Alphaflexiviridae are shown; the new isolate Capel is indicated by a dot. The tree was constructed with 1000 bootstrap replications and confidence values above 60% are shown. Botrytis virus F (family Gammaflexiviridae) was used as the outgroup.

Plant Pathology (2016) 65, 163–172

Drakaea viruses 169

Amino acid sequence identities for the replicase and myxa graminis (Herzog et al., 1994; Adams et al., CP with the corresponding regions of the genomes of 2012a). The lack of the P39 protein in DVA suggests other DOSV isolates were 78 and 87%, respectively that it might not be transmissible by a fungal vector. (Table S6). These values are only marginally below With a mycorrhizal fungus being such an integral part (replicase) or above (CP) the species demarcation limit of the Drakaea life cycle, it is interesting to speculate (80% identity) set for viruses within the family that DVA evolved from an ancestral pecluvirus that Alphaflexiviridae (Adams et al., 2012a), to which these originally infected Drakaea mycorrhizae, an proteins appear most closely related. Although DOSV undescribed species of Tulasnella (Linde et al., 2014; clearly does not belong to the Alphaflexiviridae, it is Phillips et al., 2014), but subsequently lost the fungus- proposed that the demarcation limits set for this family transmission gene after DVA was transferred to its may be used to place the Capel isolate within the plant host. Similarly, GORV lacks the P39 gene. The DOSV species. Previously, phylogenetic analysis of common genome organization and apparent close the DOSV replicase and CP showed that, although they phylogeny (Fig 3) indicate that both DVA and GORV probably share a recent common ancestor with should be placed together. Therefore, the authors homologues from the alphaflexiviruses, other gene support the proposal by Atsumi et al. (2015) that products did not. Consequently, DOSV does not GORV be classified in a new genus within the family warrant inclusion within the Alphaflexiviridae or any Virgaviridae, together with DVA. It is surprising that of the other existing families within the order DVA and GORV, discovered 8400 km apart in Tymovirales (Wylie et al., 2013b; Fig 4b). Australia and Japan, respectively, and in herbaceous plants indigenous to their countries of origin, more Discussion closely resemble one another than any other known virus. Until further research is done on viruses of wild Three partial or near complete virus-like sequences herbaceous plants in eastern Asia between Australia were identified from RNA collected from two plants of and Japan, the existence of other members of this new two Drakaea species. The sequences are proposed to virus group can only be speculated upon. derive from isolates of two viruses: a previously This is the first record of a peclu-like virus in undescribed bipartite virus provisionally named Australia, with the related viruses PCV and IPCV Drakaea virus A (two sequences), and a proposed new having so far been detected only in West Africa (PCV) isolate of DOSV (one sequence). DOSV is a species and the Indian sub-continent (IPCV) despite the natural already described from other orchids from the same host (peanut, Arachis hypogea) originating in Paraguay region (Wylie et al., 2013b). These viruses are the first and Bolivia (Seijo et al., 2007; Adams et al., 2012b). If to be identified from wild Drakaea plants. Neither pecluviruses occur naturally in legumes in South virus generated obvious symptoms on the orchids in America, this geographic range disjunct could arise which they occurred naturally. This, together with the from vicariance following the breakup of uniqueness of their genomes, suggests that these Gondwanaland approximately 35.5 million years ago viruses may have been associated with these hosts over (McLoughlin, 2001). Alternatively, if this group of a long period (Malmstrom et al., 2011). viruses naturally infects a broad range of hosts, it could The only virus in the family Virgaviridae recorded be widespread geographically, reflecting the to infect orchids is Odontoglossum ringspot virus distribution of their host species across all vegetated (ORSV), an unusual recombinant tobamovirus with continents. Wider generic virus surveys of wild identity to both Brassica- and Solanaceae-infecting Drakaea populations and the surrounding orchidaceous tobamoviruses, identified from cultivated and native and non-orchidaceous flora will not only inform on orchids including species of Odontoglossum, DVA host distribution, but also reveal if related viruses Cymbidium and Cattleya (Gibbs et al., 2000; Adams et exist in other Drakaea species, perhaps having been al., 2009). DVA is proposed to be the second member transferred by rare hybridization events (if pollen- of this family to infect orchids and the first to be borne), or from having associated with Drakaea prior isolated from Drakaea orchids. to the radiation of the genus. Classification of genera within the Virgaviridae Members of both Hordeivirus and Pecluvirus are family is dependent on properties that include the known to be transmissible through seeds and pollen number of RNAs, type of movement protein (30K (Reddy et al., 1998; Adams et al., 2009). If DVA, like superfamily or triple gene block) and location of the GORV (Atsumi et al., 2015), is transmitted via pollen, RdRp domain within the replication protein (Adams et spread would typically involve a different specific al., 2009, 2012b). The bipartite nature of DVA and thynnine wasp species for each Drakaea species presence of the RdRp domain at the C-terminal end of (Phillips et al., 2014). Despite being successfully the replicase, after the MET and HEL domains, mechanically transmitted to D. glyptodon, if DVA indicate that DVA is closer to Pecluvirus than were pollen-transmitted it would probably not be Hordeivirus, which differ by having a tripartite readily spread between Drakaea species because of the genome with the RdRp domain encoded on a separate specificity of the plant-pollinator system (Phillips et al., RNA to the MET and HEL domains. However, DVA 2014). Like other orchids, the dust-like seeds of D. lacks the typical Pecluvirus P39 protein located 3' of its livida are dispersed by wind (Arditti & Ghani, 2000). CP and has a CRP 3' of the TGBps on RNA2. The Field testing is required to understand the transmission Pecluvirus P39 protein is believed to be involved in and dispersal of this virus, where experiments can be implemented to test if pollen transmission by the fungal vector Poly-

Plant Pathology (2016) 65, 163–172

170 J. W. L. Ong et al. and/or seeds can transfer the virus from known latifolia (two populations) (Wylie et al., 2013b) and infected plants. This will provide a better 16 D. elastica plants (five populations, current study). understanding of the potential geographical spread of • The orchids sampled are not the primary hosts of the the virus and its efficiency of transmission through virus. The virus is adapted to another host but can subsequent generations. occasionally be transmitted to these orchids via The presence of the MET, HEL and RdRp domains vectors or pollen, but is unable to spread efficiently within the DOSV replicase gene (Fig 4a) and its within the species. sequence identity with the replicase genes of viruses within the Alphaflexiviridae point to a shared Drakaea virus A and YTMMV are currently the evolutionary history with members of the only two apparently indigenous viruses from the Alphaflexiviridae family (Martelli et al., 2007; Wylie Virgaviridae to be isolated from indigenous plants in et al., 2013b). Phylogenetic analysis of the deduced Australia. DOSV is also connected to this family via amino acid sequence also placed the three described its movement protein – the closest match was to the DOSV isolates basal to the plant-infecting members of MP of SCSV (genus Furovirus, family Virgaviridae). the Alphaflexiviridae but not to the fungus-infecting These linkages with the Virgaviridae indicate this member Sclerotinia sclerotiorum debilitation- virus family is likely to have an ancient association associated RNA virus (SsDRV) (Fig 4b). SsDRV is with members of the Australian flora. thought to have evolved from a plant virus to become a Current known anthropogenic threats to Drakaea persistent mycovirus, losing its CP and MP during the include clearing of natural bushland, the spread of process (Martelli et al., 2007). DOSV, on the other introduced plants in small habitat remnants and hand, may retain these genes because it is a non- grazing from feral herbivores (Swarts & Dixon, 2009). persistent plant virus. The DOSV CP is also related to The formation of a specialized mycorrhizal fungus CPs from members of the Alpha- and Betaflexiviridae, association and the requirement of a particular wasp providing further evidence of its close association with pollinator (Swarts & Dixon, 2009; Phillips et al., these groups. Phylogenetic analysis showed that the 2014) could also influence viability of orchid CP differs from the replicase in that it is not basal to populations, particularly if these partners are adversely CPs of other members of the Alpha- and affected by altered habitats or landscape modification. Betaflexiviridae family (Wylie et al., 2013b). This The impact of indigenous viruses in natural systems is points to the DOSV CP being acquired more recently a neglected but important area of study. In the case of than the replicase, probably from an allexivirus-like or Drakaea, the small physical size of the plants, the botrexvirus-like ancestor (Wylie et al., 2013b). rarity of some species, their short vegetative and Deducing the evolutionary history of DOSV from reproductive phases above-ground each year, and the its other genes is more problematic. Two possible MPs difficulty of growing them under glass house exist in the DOSV genome. ORF1 is predicted to conditions (Hopper & Brown, 2007; Swarts & Dixon, encode a 68 kDa protein that is most similar to the MP 2009) make them a challenging group to study. of tymoviruses in terms of its size, and position within While the impact of viruses on fecundity, lifespan the genome overlapping the replicase. ORF7 encodes and other aspects of ecological fitness of Drakaea can the more probable MP because of its close sequence only be speculated at this stage, the detection of identity with P30-like MPs of furoviruses (family viruses is a first step in addressing these questions. Virgaviridae) and dianthoviruses (family Symptoms were not evident in plants infected with Tombusviridae), groups only distantly related to the either virus, but this conclusion was based on flexiviruses. It is proposed that DOSV be included as a observations of the leaf and flower of a small number member of the order Tymovirales, but, due to its of plants, and tubers were not examined or compared unique genome organization and identity with multiple with uninfected plants. Exotic broad host-range viruses viral families, it does not warrant inclusion within such as BYMV and OrMV can potentially widely existing families of the order. Thus, a new taxa at the infect wild orchid populations in southwestern family level may need to be created to accommodate Australia (Gibbs et al., 2000; Wylie et al., 2013a). DOSV. However, not all exotic orchid-infecting viruses are Like the two DOSV isolates previously identified in necessarily a threat to wild orchid populations. For Diuris longifolia and Caladenia latifolia (Wylie et al., example, Cymbidium mosaic virus (CymMV) and 2013b), DOSV occurred uncommonly in the plants ORSV, commonly found in horticultural orchids, have sampled. Three scenarios, individually or in not been detected in wild orchids and Cucumber combination, may explain the apparent rarity of mosaic virus (CMV), a virus widely distributed in a DOSV: large number of plant genera, was present only at very low concentrations in wild Calanthe orchids and • The virus is naturally rare, perhaps because it is induced no visible symptoms (Elliott et al., 1996; poorly transmitted between hosts or because the Kawakami et al., 2007). Turnip yellows virus (TuYV) vector is rare. The allexivirus-like CP suggests that was detected in a plant of Diuris pendunculata, a eriophyid mites may play a role in transmission. threatened Australian donkey orchid, yet no symptoms Allexiviruses are vectored by eriophyid mites and of infection were visible (Wylie et al., 2013a). With vector determinants are present in the CP (Adams et numerous orchid species typically co-occurring in the al., 2012a). wild, detrimental exotic viruses could potentially • Its low incidence is a reflection of the small sampling spread amongst species. Thus, while size and low numbers of sampled population in both studies: 264 D. longifolia (two populations), 129 C. Plant Pathology (2016) 65, 163–172 25

Drakaea viruses 171

preventing the spread of detrimental exotic viruses into Hopper SD, Brown AP, 2007. A revision of Australia’s hammer wild orchid populations would seem to be desirable for orchids (Drakaea: Orchidaceae), with some field data on species- the long-term viability of populations, the implications specific sexually deceived wasp pollinators. Australian Systematic Botany 20, 252–85. on plant health from indigenous viruses that may have Kawakami K, Fuji S, Miyoshi K, 2007. Endangered wild populations of co-existed with their host for long periods is less endemic Calanthe orchids on an isolated Japanese island tested for certain. The current study is a first step in viruses. Australian Journal of Botany 55, 831–6. understanding that viruses exist in some wild Drakaea Kearse M, Moir R, Wilson A et al., 2012. GENEIOUS BASIC: an integrated and extendable desktop software platform for the populations, and their possible presence should be organization and analysis of sequence data. Bioinformatics 28, considered before ex situ propagation and 1647–9. reintroduction programmes are undertaken to bolster Koonin EV, 1991. The phylogeny of RNA-dependent RNA polymerases wild populations. of positive-strand RNA viruses. Journal of General Virology 72, 2197– 206. Kores PJ, Molvray M, Weston PH et al., 2001. A phylogenetic Acknowledgements analysis of Diurideae (Orchidaceae) based on plastid DNA sequence data. J.W.L.O, S.J.W. and M.G.K.J. were supported in part American Journal of Botany 88, 1903–14. Linde CC, Phillips RD, Crisp MD, Peakall R, 2014. Congruent species by ARC Linkage grant LP110200180 in collaboration delineation of Tulasnella using multiple loci and methods. New with Botanic Gardens and Parks Authority and Phytologist 201, 6–12. Australian Orchid Foundation. Fieldwork undertaken Malmstrom CM, Melcher U, Bosque-Perez NA, 2011. The expanding by R.D.P. was supported by an ARC Linkage grant field of plant virus ecology: historical foundations, knowledge gaps, and research directions. Virus Research 159, 84–94. LP098338 awarded to Rod Peakall and K.W.D. Marchler-Bauer A, Bryant SH, 2004. CD-Search: protein domain annotations on the fly. Nucleic Acids Research 32, W327–31. Marchler-Bauer A, Zheng C, Chitsaz F et al., 2013. CDD: conserved References domains and protein three-dimensional structure. Nucleic Acids Research 41, D348–52. Adams MJ, Antoniw JF, Kreuze J, 2009. Virgaviridae: a new Martelli GP, Adams MJ, Kreuze JF, Dolja VV, 2007. Family family of rod-shaped plant viruses. Archives of Virology 154, Flexiviridae: a case study in virion and genome plasticity. Annual 1967–72. Review of Phytopathology 45, 73–100. Adams MJ, Candresse T, Hammond J et al., 2012a. Alphaflexiviridae. McLoughlin S, 2001. The breakup history of Gondwana and its impact In: King AMQ, Adams MJ, Carstens EB, Lefkowitz EJ, eds. Virus on pre-Cenozoic floristic provincialism. Australian Journal of Taxonomy: Classification and Nomenclature of Viruses: Ninth Botany 49, 271–300. Report of the International Committee on Taxonomy of Viruses. Morris TJ, Dodds JA, 1979. Isolation and analysis of double-stranded San Diego, USA: Elsevier Academic Press, 904–19. RNA from virus-infected plant and fungal tissue. Phytopathology 69, Adams MJ, Heinze C, Jackson AO, Kreuze JF, Macfarlane SA, 854–8. Torrance L, 2012b. Virgaviridae. In: King AMQ, Adams MJ, Peakall R, 1990. Responses of male Zaspilothynnus tribobatus Turner Carstens EB, Lefkowitz EJ, eds. Virus Taxonomy: Classification wasps to females and the sexually deceptive orchid it pollinates. and Nomenclature of Viruses: Ninth Report of the International Functional Ecology 4, 159–67. Committee on Taxonomy of Viruses. San Diego, USA: Elsevier Phillips RD, Brown AP, Dixon KW, Hopper SD, 2011. Orchid Academic Press, 1139–62. biogeography and the factors associated with rarity in a biodiversity Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ, 1990. hotspot: the Southwest Australian Floristic Region. Journal of Basic local alignment search tool. Journal of Molecular Biology Biogeography 38, 487–501. 215, 403–10. Phillips RD, Peakall R, Hutchinson MF et al., 2014. Specialized Arditti J, Ghani A, 2000. Numerical and physical properties of ecological interactions and plant species rarity: the role of pollinators orchid seeds and their biological implications. New Phytologist and mycorrhizal fungi across multiple spatial scales. Biological 145, 367– 421. Conservation 169, 285–95. Atsumi G, Tomita R, Yamashita T, Sekine K-T, 2015. A novel Ramsay RR, Dixon KW, Sivasithamparam K, 1986. Patterns of infection virus transmitted through pollination causes ring-spot disease and endophytes associated with Western Australian orchids. on gentian (Gentiana triflora). Journal of General Virology 96, Lindleyana 1, 203–14. 431–9. Reddy AS, Hobbs HA, Delfosse P, Murthy AK, Reddy DVR, 1998. Bohman B, Phillips RD, Menz MH et al., 2014. Discovery of Seed transmission of Indian peanut clump virus (IPCV) in peanut and pyrazines as pollinator sex pheromones and orchid millets. Plant Disease 82, 343–6. semiochemicals: implications for the evolution of sexual deception. Seijo G, Lavia GI, Fernandez A et al., 2007. Genomic relationships New Phytologist 203, 939–52. between the cultivated peanut (Arachis hypogaea, Leguminosae) Brundrett M, 2014. Identification and Ecology of Southwest and its close relatives revealed by double GISH. American Journal Australian Orchids. Perth, Australia: Western Australian of Botany 94, 1963–71. Naturalists’ Club. Swarts ND, Dixon KW, 2009. Terrestrial orchid conservation in the age Elliott MS, Zettler FW, Zimmerman MT Jr, Barnett OW, Legrande of extinction. Annals of Botany 104, 543–56. MD, 1996. Problems with interpretation of serological assays in a Tamura K, Stecher G, Peterson D, Filipski A, Kumar S, 2013. virus survey of orchid species from Puerto Rico, Ecuador, and MEGA6: molecular evolutionary genetics analysis version 6 0. Florida. Molecular Biology and Evolution 30, 2725–9. Plant Disease 80, 1160–4. Te J, Melcher U, Howard A, Verchot-Lubicz J, 2005. Soilborne wheat Gibbs A, Mackenzie A, Blanchfield A et al., 2000. Viruses of mosaic virus (SBWMV) 19K protein belongs to a class of cysteine orchids in Australia; their identification, biology and control. rich proteins that suppress RNA silencing. Virology Journal 2, 18. Australian Orchid Review 65, 10–21. Wylie SJ, Tan AJY, Li H, Dixon KW, Jones MGK, 2012. Caladenia Govaerts R, Bernet P, Kratochvil K et al., 2011. World Checklist virus A, an unusual new member of the family Potyviridae from of Orchidaceae. [http://apps.kew.org/wcsp/incfamilies.do]. terrestrial orchids in Western Australia. Archives of Virology 157, Accessed 5 January 2015. 2447–52. Herzog E, Guilley H, Manohar SK et al., 1994. Complete Wylie SJ, Li H, Dixon KW, Richards H, Jones MGK, 2013a. Exotic and nucleotide sequence of peanut clump virus RNA 1 and indigenous viruses infect wild populations and captive collections of relationships with other fungus-transmitted rod-shaped viruses. temperate terrestrial orchids (Diuris species) in Australia. Virus Journal of General Virology 75, 3147–55. Research 171, 22–32.

Plant Pathology (2016) 65, 163–172

172 J. W. L. Ong et al.

Wylie SJ, Li H, Jones MGK, 2013b. Donkey orchid symptomless Table S3. Plant samples collected from Canning Mills (32°04' 54.2″S, virus: a viral ‘platypus’ from Australian terrestrial orchids. PLoS 116°05' 27.6″E) to test for presence of Drakaea virus A. ONE 8, e79587. Table S4. CLUSTALW comparison of nucleotide and amino acid Wylie SJ, Li H, Jones MGK, 2014. Yellow tailflower mild mottle identity of Drakaea virus A genes with those of closely related viruses. virus: a new tobamovirus described from Anthoceris littorea Table S5. Comparison of deduced molecular masses of proteins (Solanaceae) in Western Australia. Archives of Virology 159, 791–5. (kilo-daltons) and lengths (nucleotides; shown in parentheses) of genes and untranslated regions (UTR) between Donkey orchid symptomless virus isolates. Supporting Information Table S6. Pairwise comparison of coding regions between genomes of three Donkey orchid symptomless virus isolates. Additional Supporting Information may be found in the online version of this article at the publisher’s website.

Table S1. List of Drakaea orchid samples tested for viruses.

Table S2. Primers used to sequence across gaps and ambiguous nucleotide bases in virus genomes.

Plant Pathology (2016) 65, 163–172

Supplementary information

Table S1. List of Drakaea plant samples collected and tested for viruses. Sample No. (No. Orchid species Common namea Location of collection Latitude/Longitudec Year of collection of individuals) North-West of D. concolor Kneeling Hammer Orchid DR01 (7) - 2012 Northampton -32o 0' 27.2'' D. gracilis Slender Hammer Orchid DR02 (10) Lesmurdie 2012 116o 4' 47.8'' -32o 4' 54.2'' D. livida Warty Hammer Orchid DR03 (2)b Canning Mills 2012 116o 5' 27.6'' -32o 5' 33.9'' D. glyptodon King-in-his-carriage DR04 (11) Wandoo National Park 2012 116o 34' 11.8'' -32o 7' 29.4'' D.gracilis Slender Hammer Orchid DR05 (9) Wandoo National Park 2012 116o 28' 17.3'' Carrabungup Nature -32o 38' 50.6'' D. livida Warty Hammer Orchid DR06 (4) 2012 Reserve 115o 42' 55.9'' Glossy-leafed Hammer Carrabungup Nature D. elastica DR07 (7) - 2012 Orchid Reserve Carrabungup Nature -32o 38' 50.6'' D. glyptodon King-in-his-carriage DR08 (2) 2012 Reserve 115o 42' 55.9''

D. micrantha Dwarf Hammer Orchid DR09 (2) East of Margaret River - 2012

D. livida Warty Hammer Orchid DR10 (5) East of Margaret River - 2012

D. micrantha Dwarf Hammer Orchid DR11 (3) Canebrake Nature Reserve - 2012

D. glyptodon King-in-his-carriage DR12 (6) Canebrake Nature Reserve - 2012

-33o 53' 27'' D. glyptodon King-in-his-carriage DR13 (7) South of Manjimup 2012 115o 16' 31.1'' -34o 23' 53.33'' D. glyptodon King-in-his-carriage DR14 (10) West of Pemberton 2012 115o 48' 19.64'' -34o 19' 12.7'' D. glyptodon King-in-his-carriage DR15 (13) Peerabeelup 2012 115o 46' 14.8'' Narrow-lipped Hammer -34o 19' 12.7'' D. thynniphila DR16 (10) Peerabeelup 2012 Orchid 115o 46' 14.8'' Narrow-lipped Hammer -34o 19' 12.7'' D. thynniphila DR17 (9) Peerabeelup 2012 Orchid 115o 46' 14.8''

-33o 38' 33.5'' D. glyptodon King-in-his-carriage DR18 (8) Ruabon NatureReserve 2013 115o 30' 19.71'' -33o 42' 24'' D. livida Warty Hammer Orchid DR19 (1) South Yallingup 2013 115o 01' 40'' -33o 42' 24'' Drakaea sp. - DR20 (4) South Yallingup 2013 115o 01' 40'' Glossy-leafed Hammer Carrabungup Nature D. elastica DR21 (4) - 2013 Orchid Reserve Carrabungup Nature D. livida Warty Hammer Orchid DR22 (2) - 2013 Reserve Glossy-leafed Hammer Serpentine River Nature D. elastica DR23 (2) - 2013 Orchid Reserve

D. micrantha Dwarf Hammer Orchid DR24 (2) East of Margaret River - 2013

D. micrantha Dwarf Hammer Orchid DR25 (3) East of Margaret River - 2013

Glossy-leafed Hammer D. elastica DR26 (2)b Capel - 2013 Orchid -33o 42' 24'' D. livida Warty Hammer Orchid DR27 (2) South of Yallingup 2013 115o 01' 40'' Glossy-leafed Hammer Serpentine River Nature D. elastica DR28 (3) - 2013 Orchid Reserve -34o 17' 54.2'' D. glyptodon King-in-his-carriage DR29 (12) Nannup 2013 115o 45' 58.1''

a Species names are given, if known at the time of collection. b Samples from which the viruses (DVA and DOSV) were isolated from. c GPS co-ordinates of locations with classified rare Drakaea species were not included to comply with guidelines with flora permit.

Table S2. Primers used to sequence across gaps and ambiguous nucleotide bases in virus genomes.

Virus Position on genome Primer name Primer sequence (5'→3')

722-741 (RNA-1) DVA-1 (F) CATGAGCAAAATGTCGGATG

3696-3677 (RNA-1) DVA-2 (R) GTGGGCTACGGTCCAACTTA

1648-1668 (RNA-1) DVA-3 (F) CGGAAGTGATAGAGGTCAGCA

2928-2908 (RNA-1) DVA-4 (R) CGTTCTCCGTACTCTTCAACC

3554-3573 (RNA-1) DVA-5 (F) TGTGCAAAGATGGTGGGATA DVA 4353-4334 (RNA-1) DVA-6 (R) TCAAAGGATCGGGTGAAAAA

207-226 (RNA-2) DVA-7 (F) AATGCTGGTTCACGTTTTCC

1245-1226 (RNA-2) DVA-8 (R) CACTTTGCGTTGGAGCAGTA

1863-1882 (RNA-2) DVA-9 (F) CGACTGAATCGGGAGACAAT

2256-2237 (RNA-2) DVA-10 (R) TGGGGTTACCTGGAACACTT

432-451 DOSV-1 (F) CTCACACCGCACATGAAGTC

782-763 DOSV-2 (R) GCCAGGAGAGGCAGTTAAGA

1742-1761 DOSV-3 (F) AAAGCCGACATCCACATCTC DOSV 2091-2072 DOSV-4 (R) TTGGTTGGGACGATTACCTC

3706-3725 DOSV-5 (F) CATGGCGTACTTCTTCACGA

4059-4040 DOSV-6 (R) AGTCTAATTTCGCGCTCGTC

Table S3. Plant samples collected from Canning Mills (-32o 4' 54.2'', 116o 5' 27.6'') to test for presence of Drakaea virus A. No. of DVA Sample No. Plant species Common name individuals result

CM01 Drakaea livida (Orchidaceae) Warty Hammer Orchid 1 Positive

CM02 Caladenia flava (Orchidaceae) Cowslip Orchid 20 Negative

CM03 Pterostylis barbata (Orchidaceae) Bird Orchid 1 Negative

CM04 Elythranthera brunonis (Orchidaceae) Purple Enamel Orchid 1 Negative

CM05 Pyrochis nigricans (Orchidaceae) Red Beak Orchid 15 Negative

CM06 Anigozanthos manglesii (Haemodoraceae) Mangles Kangaroo Paw 15 Negative

CM07 Lechenaultia biloba (Goodeniaceae) Blue Leschenaultia 15 Negative

CM08 Gompholobium knightianum (Fabaceae) Handsome Wedge Pea 17 Negative

CM09 Stylidium brunonianum (Stylidiaceae) Pink Fountain Triggerplant 17 Negative

CM10 Allocasuarina fraseriana (Casuarinaceae) Sheoak 15 Negative

CM11 Conostylis sp. (Haemodoraceae) - 12 Negative

CM12 Gladiolus caryophyllaceus (Iridaceae) Wild Gladiolus 13 Negative

CM13 Eucalyptus marginata (Myrtaceae) Jarrah 16 Negative

Table S4. ClustalW comparison of nucleotide and amino acid identity of Drakaea virus A with closely related viruses. Nucleotide identity (%) Amino acid identity (%) Genus/species Replicase CP TGBp1 TGBp2 TGBp3 CRP Replicase CP TGBp1 TGBp2 TGBp3 CRP oat golden stripe 51.3 44.6 - - - 43.5 39.1 11.4 - - - 21.7 virus sorghum chlorotic Furovirusa 51.0 41.4 - - - 47.9 40.8 11.8 - - - 21.0 spot virus soil-borne wheat 51.0 43.4 - - - 45.3 37.5 14.5 - - - 16.8 mosaic virus barley stripe Hordeivirusb 54.5 47.7 49.0 52.8 48.6 44.6 28.5 32.7 33.2 49.6 26.5 21.2 mosaic virus peanut clump 54.2 48.0 52.4 55.6 48.7 44.3 47.0 41.7 36.8 52.8 30.7 25.7 virus Pecluvirus Indian peanut 54.8 50.5 50.6 57.1 51.0 44.0 46.9 43.3 38.3 53.6 30.1 25.0 clump virus potato mop-top 51.2 42.7 49.1 59.0 53.2 42.1 39.0 11.0 33.0 52.8 28.7 13.3 Pomovirus virus beet virus Q 50.3 45.3 48.9 57.2 51.8 40.4 39.0 10.8 30.6 50.0 32.0 9.5 tobacco mosaic 45.3 41.8 - - - - 22.8 18.7 - - - - Tobamovirusa,c virus yellow tailflower 45.7 43.2 - - - - 23.0 15.5 - - - - mild mottle virus tobacco rattle Tobravirusa 48.0 43.9 - - - 43.3 30.7 19.8 - - - 16.4 virus gentian ovary Unassigned 55.2 50.3 48.4 54.7 49.0 42.8 47.3 35.9 31.0 46.0 29.1 17.2 ring-spot virus a Members of Furovirus, Tobamovirus and Tobravirus have single cell-to-cell movement protein instead of the triple gene block proteins. b Partial replicase (RNA dependent RNA polymerase domain on RNA-3) of Hordeivirus (barley stripe mosaic virus) was used for comparison. c Cysteine-rich protein (CRP) is not present in members of Tobamovirus

Chapter 3: The challenges of using high-throughput sequencing to track multiple new bi-partite viruses of wild orchid-fungus partnerships over consecutive years

3.1 Abstract

The bipartite alpha- and betapartitiviruses are recorded from a wide range of fungi and plants. Using a combination of dsRNA-enriched extraction and high- throughput shotgun sequencing, we report the occurrence of multiple partitiviruses associated with mycorrhizal Ceratobasidium fungi isolated from one population of wild Pterostylis sanguinea orchids over two consecutive years. Twenty-one partial or near-complete sequences representing approximately 16 alpha- and betapartitiviruses were detected from two fungal isolates. The majority of partitiviruses occurred in fungal isolates from both years. Two of the partitiviruses represent genetically distinct forms of Alphapartitivirus, suggesting that Australia is a region of partitivirus evolution, or that they evolved under long geographical isolation there. We address the challenge of pairing the partitivirus segments when multiple species co-occur in a host.

3.2 Introduction

Members of the family Partitiviridae are classified into five genera:

Alphapartitivirus, Betapartitivirus, Deltapartitivirus, Gammapartitivirus and

Cryspovirus (Nibert et al., 2014). Their host ranges include plants, fungi and protozoa

(Ghabrial et al., 2012). Members of this family are characterised by having isometric particles ranging from 25-40 nm in diameter and a bipartite genome that encodes for

an RNA-dependent RNA polymerase (RdRp) on one segment and a coat protein (CP) on the second segment (Ghabrial et al., 2012; Nibert et al., 2014). Infection by these viruses is often persistent and latent (Roossinck, 2010; Ghabrial et al., 2012; Nibert et al., 2014).

Alphapartitivirus and Betapartitivirus contain plant-infecting and fungus- infecting species (Nibert et al., 2014). Their genetic relatedness suggests that partitiviruses have transmitted among and between plants and fungi (Roossinck, 2010;

Nibert et al., 2014). Orchids rely on partnerships with compatible mycorrhizal fungi, whose hyphae are ingested by the plants to provide nutrients required for germination and growth (Swarts and Dixon, 2009). Such close interactions may provide opportunities for partitiviruses to transmit between plants and fungi. Currently, Diuris pendunculata cryptic virus (DPCV), isolated from an ex-situ population of D. pendunculata is the only proposed partitivirus reported in Australia and from orchids

(Wylie et al., 2013). The only two plant viruses described from Pterostylis orchids have both been potyviruses (family Potyviridae, genus Potyvirus) – bean yellow mosaic virus and Ornithogalum mosaic virus (syn Pterostylis virus y) (Gibbs et al.,

2000). Seven mycorrhizae-derived endornaviruses were identified from fungal pelotons in the related orchid species Pterostylis sp. (Ong et al., 2016). In this study, a high-throughput sequencing approach was used to identify partitiviruses infecting mycorrhizal fungi associated with a small population of Pterostylis sanguinea orchids

(dark banded greenhood orchid) growing in a natural habitat. We discuss the challenges in identifying co-occurring, novel, and closely-related bipartite viruses.

3.3 Materials and methods

3.3.1 Sample collection

Leaves and underground stems (Fig 3.1) were collected from a small natural population of P. sanguinea orchid plants located on the Murdoch University campus,

Western Australia (GPS coordinates -32° 3' 54.9714", 115° 50' 26.448") in 2012 and

2013. The population consisted of three (in 2012) and four (in 2013) orchid shoots growing within a one square metre area in natural bushland. Because orchid tubers may germinate unevenly (Brundrett, 2014), it was impossible to definitively select leaf material from the same plants in both years of the study. Leaf material was combined from three plants in 2012 (sample P-2012) and four plants in 2013 (sample

P-2013) before nucleic acids extraction and sequencing. In each of the years, a fungal culture was established from one peloton isolated from the underground stem (fungal isolates F-2012 and F-2013) of one of the plants sampled. Collection of plant tissues, including the underground stem, did not cause the death of plants because the new tubers remained undisturbed.

(A) (B) (C)

(E)

(D)

Figure 3.1. Pterostylis sanguinea (A) whole plant (B) labella (C) leaves (D)

underground stem and (E) old (brown) and new (white) tubers. Scale bar: (A) 5 cm

(B-E) 2 cm.

3.3.2 Fungal isolation from underground stems

Each underground stem was surface-sterilised by immersion in 2% (w/v) sodium hypochlorite solution for 3 min, dipped in 70% ethanol for 10 s, followed by two rinses in sterile distilled water. The stem was then transferred to a 1.5 mL centrifuge tube with sterile water and ground with a pestle to produce a suspension of pelotons (fungal coils located within the underground stem) and plant debris. Under a compound microscope, individual pelotons were located and transferred onto fungal

-1 -1 -1 isolation medium (FIM) agar plates (0.3 g L NaNO3, 0.2 g L KH2PO4, 0.1 g L

-1 -1 -1 -1 MgSO4.7H2O, 0.1 g L KCl, 0.1 g L yeast extract, 2.5 g L sucrose and 8 g L agar;

100 mg L-1 filter-sterilised streptomycin sulphate) (Clements & Ellyard, 1979).

Fungal isolates were left to incubate in the dark at 24oC for 5-7 days. Mycelium was

subcultured onto fresh FIM plates and into 100 mL FIM liquid medium (FIM without agar). Liquid cultures were incubated on a shaker at 24oC in the dark until 80-100 mg fungal biomass could be harvested.

3.3.3 Nucleic acids extraction, cDNA synthesis and amplification

DNA and RNA extraction was from 80-100 mg of plant or fungal tissue using a cellulose-based method that enriched the sample for double-stranded RNA (dsRNA)

(Morris & Dodds, 1979). The aqueous phase following phenol-chloroform processing was mixed with Whatman CF-11 cellulose powder, centrifuged and resulting supernatant containing DNA was collected.

cDNA synthesis was carried out in a 20 µL volume containing 1X GoScriptTM

RT buffer (Promega), 3 mM MgCl2, 0.5 mM dNTPs, 0.5 mM of random primer

(5' CGTACAGTTAGCAGGCNNNNNNNNNNNN 3', where N is any nucleotide),

160 units of GoScript™ and 4 µL of heat-denatured RNA (50-100 ng). cDNA synthesis occurred after an initial incubation at 25oC for 5 min, incubation at 42oC for

60 min and enzyme denaturation at 70oC for 15 min.

PCR amplification was done in a 20 µL reaction volume consisting of 1X

GoTaq® Green Master Mix (Promega), 1 mM barcode primer

(5' XXXXCGTACAGTTAGCAGGC 3') and 2 µL of cDNA. Each barcode primer was tagged with a unique 4-nt barcode at the 5' terminus of a 16-nt adaptor sequence that was complementary to the 5' end of the cDNA synthesis primer. The cycling reaction was carried out with an initial incubation of 3 min at 95oC, followed by 35

cycles of 30 s at 95oC, 30 s at 60oC and 1 min at 72oC, and a final extension for 10 min at 72oC.

Amplicons were pooled in equimolar amounts and purified using a Qiagen

QIAquick PCR Purification Kit. Ten micrograms of pooled amplicons were submitted to the Australian Genome Research Facility (Melbourne, Australia) or Macrogen Inc

(Seoul, South Korea) for library construction and high-throughput sequencing of paired ends over 100 cycles on a HiSeq 2000 (Illumina).

3.3.4 Identification of fungi

The 5.8S ribosomal gene and flanking internally transcribed spacer (ITS) regions were amplified using fungal universal primers ITS1

(5' TCCGTAGGTGAACCTGCGG 3') and ITS4 (5' TCCTCCGCTTATTGATATGC

3') (White et al., 1990). Amplified PCR products were purified using QIAquick

(Qiagen) columns and sequenced using the Sanger method (BigDye® version 3.1 terminator mix; Applied Biosystems). Sequences were edited and pairwise aligned using the alignment tool in Geneious v7.0.6 (Biomatters). Blastn (Altschul et al.,

1990) searches identified the fungal matches.

3.3.5 Sequencing data analysis

CLC Genomic Workbench v6.5.1 (Qiagen) software was used for de novo assembly of reads to form contigs. Settings used for assembly were word size of

23/24, bubble size of 50, auto-detect paired distance and a minimum contig size of

200 nt. Assembled contigs were subjected to Blastn and Blastx analysis to identify virus-like contigs (e-value < 1). The NCBI Conserved Domain Database (CDD)

(http://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) was used to identify virus-like domains such as RdRp and CP (Marchler-Bauer & Bryant, 2004).

Viral sequences with two ORFs were examined for the presence of ribosomal frameshift, represented by a heptanucleotide slippery sequence of XXXYYYN (where

X = A, G, or U; Y = A or U; N = A, C, or U) and an adjacent mRNA secondary structure, usually an mRNA pseudoknot (Brieiley et al., 1992).

Deduced amino acid sequences of virus-like sequences were aligned using

ClustalW within MEGA v6.06 (http://www.megasoftware.net/) and subjected to

“Find best DNA/Protein models (Maximum likelihood, ML)”. Maximum likelihood

(ML) phylogenetic trees with 1000 bootstrap replications were constructed with

Nearest-Neighbor-Interchange (NNI) as the ML Heuristic method.

3.3.6 RT-PCR amplification of partitivirus segments

Specific primers were designed for the CP and RdRp sequences that were present in only one of the two Ceratobasidium isolates. The fungal isolates were then reciprocally tested by RT (reverse transcription)-PCR amplification (Promega

GoTaq®) to identify these CP and RdRp segments.

3.3.7 5' UTRs alignments

Pairwise alignments were carried out on the 5' untranslated regions (UTR) of

CPs and RdRps of known complete partitiviruses to determine the suitability of using

5' UTRs as a mean to pair the associated proteins. This was done using Geneious v7.0.6 under the settings of IUB as the matrix model, gap open cost of 15 and gap

extend cost of 6.66. The alignment was later applied to the CPs and RdRps of identified partitiviruses in an attempt to pair up the proteins and differentiate individual partitiviruses.

3.4 Results

3.4.1 Partitiviruses

Thirty-two virus-like sequences resembling segments of partitiviruses were detected from mycorrhizal fungi isolated from P. sanguinea plants collected at two time points a year apart, but not from leaf samples. Partitiviruses are bipartite viruses; their genomes are characterized by two unrelated dsRNA segments, each with a single

ORF, one encoding a replicase with an RdRp motif and the other a CP (Nibert et al.,

2014). After Blastp analysis, 16 of the 32 partitivirus-like contigs were identified (456 to 2466 nt) as RdRp-like segments, and the remainder (469 to 2266 nt) resembled CP segments (Tables 3.1 and S1).

Previously characterised partitivirus genome segments range in size from 1.4-

2.4 kbp, and the two segments of individual viruses are usually closely similar in size

(Nibert et al., 2014). In the current study, only partitivirus-like sequences >1.3 kbp were considered to definitively represent a partitivirus segment because two or more fragments <1.3 kbp could be parts of the same partitivirus segment. Thus, short (<1.3 kbp) sequences (Table S1) were not analysed in detail. Long CP and RdRp fragments

(>1.3 kbp) consisting of an ORF flanked on each end by an untranslated region (UTR) were assumed to be complete or near-complete genomic segments. Each long segment was assigned the name ‘Ceratobasidium partitivirus’ followed by its assumed function (CP or RdRp), and then a letter (for CPs) or number (for RdRps) (Table 3.1).

CPs that shared high identities (>97% aa) were considered isolates of the same species and were differentiated by the addition of a number (e.g. CP-a1 and CP-a2).

The short (<1.3 kbp) partitivirus-like fragments were assigned the name

Ceratobasidium partitivirus-like contig followed by a letter or number as above

(Table S1).

In 2012, 10 partial and near-complete partitiviruses were detected in the single fungal isolate tested, as indicated by the presence of 10 distinct long RdRp sequences.

Notably, only five long CP sequences were detected in that fungal isolate, suggesting that the sequence data was incomplete, or that RdRp segments share CPs (three short

CP and two short RdRp fragments were also identified in the same fungal isolate). In

2013, only short fragments of RdRp segments were obtained, yet six long CP sequences were obtained, which we consider their presence to be evidence of at least six partial and near-complete partitiviruses present. Also detected in 2013 were two short CPs and four short RdRps. It is not known why long RdRp sequences were not obtained from fungal isolate F-2013.

Of the 10 long RdRp sequences collected in 2012, seven were closest to members of the genus Alphapartitivirus and the other three to Betapartitivirus. Four of five long CP segments from 2012 and one of the six long CP segments from 2013 were identified as potential members of Alphapartitivirus, and the others of

Betapartitivirus (Fig 3.2).

3.4.1.1 Partitivirus CPs

Phylogenetic analysis placed two CPs (Ceratobasidium partitivirus CP-d and

Ceratobasidium partitivirus CP-e) within Alphapartitivirus, but with longer branch lengths, suggestive that they represent ancestral forms or evolved independently (Fig

3.2). Pairwise identities between the new CP sequences ranged from 7-99% (Table

S2). Notably, deduced amino acid (aa) sequences of CPs isolated from the same fungal host usually shared <50% aa identity, with the exception of CP-f and CP-i from 2013 (74.9% aa identity).

3.4.1.2 Partitivirus RdRps

The ORFs of the long RdRp segments shared 7-86% aa (41-86% nt) identity with one another (Table S3). The majority of the RdRps identified conformed to the genus demarcation values set for the Partitiviridae at >27% aa identity with one another (Nibert et al., 2014) (Table S3).

Table 3.1. List of partitivirus-like sequences (>1.3 kbp in length) derived from endophytic Ceratobasidium fungal isolates F-2012 and F-2013 associated with Pterostylis sanguinea underground stems. A. CP segments showing sequence lengths (nt), blastp match, estimated percentage of CP gene, proposed classification of each segment at the genus level, sequence lengths of the ORFs and sequence lengths of the 5' and 3' untranslated regions. B. As above for RdRp segments.

Length of (A) Best Blastp match ORF GenBank Virus host CP (nt) Proposed genus 5 UTR (nt) 3 UTR (nt) protein (aa), (accession no.; e-value) ' (nt) ' accession no. [estimated %] Ceratobasidium partitivirus Cucurbitaria piceae virus 1 Betapartitivirus 106 2019 46 672 [100] KU291902 CP-a1 (2171) (ALT08066; 3e-133) Ceratobasidium partitivirus Rhizoctonia fumigata partitivirus Alphapartitivirus 69 1527 44 508 [100] KU291903 CP-b1 (1640) (AJE25831; 6e-141) Ceratobasidium sp. Ceratobasidium partitivirus Diuris pendunculata cryptic virus Alphapartitivirus 38 >1467 - >489 [>90*] KU291904 (F-2012) CP-c1 (1505) (AFY23215; 3e-26)

Ceratobasidium partitivirus Soybean leaf-associated partitivirus 2 Alphapartitivirus 129 >1269 - >456 [>90*] KU291905 CP-d (1498) (ALM62248; 5e-75) Ceratobasidium partitivirus Soybean leaf-associated partitivirus 2

Alphapartitivirus 92 1113 183 370 [100] KU291906

CP-e (1388) (ALM62248; 6e-66)

Ceratobasidium partitivirus Heterobasidion partitivirus 8 Betapartitivirus 89 2058 119 685 [100] KU291907 CP-f (2266) (AFW17811; 2e-50) Ceratobasidium partitivirus Ustilaginoidea virens partitivirus 2 Betapartitivirus 6 >1898 - >632 [>90*] KU291908 CP-g1 (1904) (AHU88026; 3e-43) Ceratobasidium partitivirus Ustilaginoidea virens partitivirus 2 Betapartitivirus 59 1617 21 538 [100] KU291909 Ceratobasidium sp. CP-h (1697) (AHU88026; 8e-52) (F-2013) Ceratobasidium partitivirus Heterobasidion partitivirus 8 Betapartitivirus - >1604 40 >533 [84] KU291910 CP-i (1644) (AFW17811; 7e-38) Ceratobasidium partitivirus Diuris pendunculata cryptic virus Alphapartitivirus 106 >1488 - >496 [>90*] KU291911 CP-c2 (1594) (AFY23215; 2e-33) Ceratobasidium partitivirus Dill cryptic virus 2 Betapartitivirus 98 >1212 - >404 [>60*] KU291912 CP-a2 (1310) (YP_007891055; 2e-69) *Estimated percentage of protein was limited by lack of complete ORF

(B) Length of Best Blastp match GenBank Virus host RdRp (nt) Proposed genus 5' UTR (nt) ORF (nt) 3' UTR (nt) protein (aa), (accession no.; e-value) accession no. [estimated %] Ceratobasidium partitivirus Ustilaginoidea virens partitivirus 2 Betapartitivirus 91 2289 86 762 [100] KU291913 RdRp-1 (2466) (AHU88025; 0.0) Ceratobasidium partitivirus Rhizoctonia solani virus 717 Betapartitivirus 115 2175 4 524 [100] KU291914 RdRp-2 (2294) (NP_620659; 0.0) Ceratobasidium partitivirus Ustilaginoidea virens partitivirus 2 Betapartitivirus 217 >1898 - >632 [>90*] KU291915 RdRp-3 (2115) (AHU88025; 0.0) Ceratobasidium partitivirus Heterobasidion partitivirus 5 Alphapartitivirus 115 1872 19 623 [100] KU291916 RdRp-4 (2006) (ADV15444; 0.0) Ceratobasidium partitivirus Soybean leaf-associated partitivirus 2 Alphapartitivirus 61 1740 99 579 [100] KU291917 RdRp-5 (1900) (ALM62247; 0.0) Ceratobasidium sp. Cherry chlorotic rusty spot associated (F-2012) Ceratobasidium partitivirus partitivirus Alphapartitivirus 91 >1754 - >584 [>90*] KU291918 RdRp-6 (1845) (CAH03668; 0.0) Ceratobasidium partitivirus Sclerotinia sclerotiorum partitivirus S Alphapartitivirus 148 >1646 - >548 [>90*] KU291919 RdRp-7 (1794) (YP_003082248; 5e-172) Ceratobasidium partitivirus Fusarium solani partitivirus 2 Alphapartitivirus 22 1524 19 507 [100] KU291920 RdRp-8 (1565) (BAQ36631; 9e-166) Ceratobasidium partitivirus Soybean leaf-associated partitivirus 1 Alphapartitivirus 56 >1495 - >498 [>80*] KU291921 RdRp-9 (1551) (ALM62245; 0.0)

Ceratobasidium partitivirus Soybean leaf-associated partitivirus 1 Alphapartitivirus 106 >1261 - >420 [>60*] KU291922 RdRp-10 (1367) (ALM62245.1; 2e-127)

*Estimated percentage of protein was limited by lack of complete ORF

99 Red clover cryptic virus 2 AGJ83766 100 Carrot cryptic virus ACL93278 99 White clover cryptic virus 2 AGJ83764 64 Dill clover cryptic virus 1 AGY36136 Hop trefoil cryptic virus 2 AGJ83767 100 Beet cryptic virus 1 ACA81389 (A) Cannabis cryptic virus AET80949 (B) Vicia cryptic virus AAX39024 99 99 Dill cryptic virus 2 AGJ83772 Red clover cryptic virus 1 AGY36139 Crimson clover cryptic virus 2 AGJ83770 100 89 98 White clover cryptic virus 1 AAU14888 Primula malacoides virus 1 ABW82142 92 Ceratobasidium partitivirus RdRp-6 96 Sclerotinia sclerotiorum partitivirus 1 AFR78159 Cherry chlorotic rusty spot associated partitivirus CAH03668 100 Rosellinia necatrix partitivirus 1 BAD98238 76 Rhizoctonia solani dsRNA virus 2 AGY54938 Rhizoctonia solani virus 717 AAF40300 68 78 Ceratobasidium partitivirus RdRp-4 Ceratobasidium partitivirus CP-a1 100 Diuris pendunculata cryptic virus AFQ95555 100 Ceratobasidium partitivirus CP-a2 Betapartitivirus Heterobasidion partitivirus 1 ADV15441 Alphapartitivirus 62 Heterobasidion partitivirus 8 AFW17811 82 Ceratobasidium partitivirus RdRp-9 99 Pleurotus ostreatus virus 1 AAT06080 100 Ceratobasidium partitivirus RdRp-10 98 Fusarium poae virus 1 AAC98725 Chondrostereum purpureum cryptic virus 1 CAQ53729 Ceratobasidium partitivirus CP-f 99 Flammulina velutipes browning virus BAH56481 99 Ceratobasidium partitivirus CP-i 99 Heterobasidion partitivirus 3 ACO37245 99 Ceratobasidium partitivirus CP-g1 100 Ceratobasidium partitivirus CP-h Raphanus sativus cryptic virus 1 AAX51289 98 Atkinsonella hypoxylon virus AAA61830 Rosellinia necatrix partitivirus 2 BAM78602 Ceratocystis resinifera virus 1 AAU26068 95 Ceratobasidium partitivirus RdRp-8 98 Heterobasidion partitivirus 2 ADL66906 100 Ceratobasidium partitivirus RdRp-5 100 Heterobasidion partitivirus 7 AEX87908 90 Ceratobasidium partitivirus RdRp-7 Penicillium stoloniferum virus F AAU95759 99 100 Sclerotinia sclerotiorum partitivirus S ACT55329 Discula destructiva virus 1 AAK13165 100 Atkinsonella hypoxylon virus AAA61829 99 Aspergillus ochraceous virus ABV30676 87 Ceratocystis resinifera partitivirus 1 AAU26069 60 Gremmeniella abietina RNA virus MS1 AAM12241 Heterobasidion partitivirus 2 ADL66905 Fig cryptic virus CBW77437 100 Heterobasidion partitivirus 7 AEX87907 Pepper cryptic virus 2 AEJ07893 100 Ceratobasidium partitivirus RdRp-1 100 99 Beet cryptic virus 2 ADP24756 Ceratobasidium partitivirus RdRp-3 Pepper cryptic virus 1 AEJ07891 100 Ceratobasidium partitivirus RdRp-2 Fragaria chiloensis cryptic virus ABC73696 81 Rhizoctonia solani virus 717 AAF22160 Southern tomato virus YP 002321510 100 Fusarium poae virus 1 AAC98734 Colletotrichum acutatum RNA virus 1 AGL42313 93 Heterobasidion partitivirus 8 AFW17810 Betapartitivirus 99 Ceratobasidium partitivirus CP-d 100 Pleurotus ostreatus virus 1 AAT07072 Ceratobasidium partitivirus CP-e Rosellinia necatrix partitivirus 1 BAD98237 Sclerotinia sclerotiorum partitivirus S ACT55330 Sclerotinia sclerotiorum partitivirus 1 AFR78160 Rosellinia necatrix partitivirus 2 BAK53192 Cannabis cryptic virus AET80948 Chondrostereum purpureum cryptic virus 1 CAQ53730 Crimson clover cryptic virus 2 AGJ83769 61 Flammulina velutipes browning virus BAH56482 100 Primula malacoides virus 1 ABW82141 Heterobasidion partitivirus 3 ACO37246 81 Dill cryptic virus 2 AGJ83771 Raphanus sativus cryptic virus 1 ABA46819 Hop trefoil cryptic virus 2 AGJ83771 100 Ceratobasidium partitivirus CP-c1 79 Red clover cryptic virus 2 AGJ83765 Ceratobasidium partitivirus CP-c2 99 99 Diuris pendunculata cryptic virus CP AFY23215 Alphapartitivirus White clover cryptic virus 2 AGJ83763 61 Heterobasidion partitivirus 1 ADV15442 97 Pepper cryptic virus 2 AEJ07892 100 Rhizoctonia solani dsRNA virus 2 AGY54939 Beet cryptic virus 2 ADP24757 Deltapartitivirus 100 Ceratobasidium partitivirus CP-b1 100 Pepper cryptic virus 1 AEJ07890 Cherry chlorotic rusty spot associated partitivirus CAH03669 Fragaria chiloensis cryptic virus AAZ06131 Beet cryptic virus 1 ACA81390 Fig cryptic virus CBW77436 78 99 100 Carrot cryptic virus ACL93279 Penicillium stoloniferum virus F AAU95758 100 Dill clover cryptic virus 1 AGY36137 86 Colletotrichum acutatum RNA virus 1 AGL42312 Vicia cryptic virus AAX39024 91 Discula destructiva virus 1 AAG59816 Gammapartitivirus Red clover cryptic virus 1 AGY36139 100 Gremmeniella abietina RNA virus MS1 AAM12240 White clover cryptic virus 1 AAU14889 78 Aspergillus ochraceous virus ABV30675 Southern tomato virus YP 002321509 1

1 Figure 3.2. Maximum Likelihood tree of Ceratobasidium partitivirus (A) CP and (B) RdRp segment sequences derived from Ceratobasidium isolates F-2012 (indicated by a dot) and F-2013 (indicated by a triangle), compared with previously described members of Partitiviridae.

Confidence values were estimated from 1000 bootstrap replications and those under 60% were omitted. CPs of gammapartitiviruses and deltapartitiviruses are not labeled because they did not form a distinct cluster. Southern tomato virus (Partitiviridae) was used as the outgroup for the RdRp analysis. Branch lengths represent calculated evolutionary distance in units of amino acid substitutions per site.

3.4.2 Most partitiviruses occurred in both years

Three of the five CPs (CP-a, CP-b and CP-c) identified in 2012 shared >90% identity (97.0-99.4% aa and 96.9-99.1% nt) with those from 2013, indicating they represent isolates of the same species (Table S2). In addition, RT-PCR analysis indicated presence of more shared partitivirus segments between the two

Ceratobasidium strains (Table 3.2). With the exception of Cp-d, all partitivirus genomes (four CPs and 10 RdRps) initially detected in mycorrhizal isolate F-2012 were also present in F-2013. Two of the six CPs (CP-f and CP-i) detected in 2013 sampling were only detected in mycorrhizal strain F-2013.

3.4.3 Matching partitivirus segments

Close sequence identity between the 5' UTRs of CP and RdRp segments of distinct partitivirus species has been noted (Hacker et al., 2006; Lesker et al., 2013;

Nibert et al., 2014). The appropriateness of pairing CPs and RdRps of partitiviruses based on their 5' UTRs was tested here by comparing the 5' UTRs of segments of previously described partitiviruses. This analysis found that the 5' UTRs of CP and replicase segments of the same species within alpha- and betapartitiviruses shared on average 80% and 76% nt identity, respectively, but identities between 5' UTRs of segments of species classified within Delta- and Gammapartitivirus were much lower

(45% and 50%, respectively), which was similar to identity of 5' UTRs of species belonging to different genera (Table 3.3).

A comparison of the 5' UTR sequences of Ceratobasidium partitivirus CPs and

RdRps from our study revealed high sequence identity between some of them (Fig S1;

Table 3.4). Amongst the proposed alphapartitiviruses, there was 73% nt identity

between 5' UTRs of RdRp-4 and CP-b1, which is within the range of identities seen between segments of species of alphapartitviruses. This indicates that RdRp-4 and

CP-b1 may be segments of the same virus. The 5' UTRs of the other 10 putative alphapartitivirus segments (four CPs and six RdRps) shared only 43-57% nt identity, indicating that none are species pairs. Within the proposed betapartitivirus segments,

5' UTRs of RdRp-2 and CP-a1 shared 77% nt identity, above the mean nt identity for

5' UTRs of species of betapartitiviruses, indicating they may belong to the same species. The 5' UTRs of the other seven putative betapartitivirus segments (two

RdRps and five CPs) shared 37-52% nt identities, below the range of 67-80% identities observed between segments within species (Tables 3.3 and 3.4). These identities suggest that none of these seven segments may be species pairs. The 5'

UTRs of RdRp-9 and CP-c2, identified from different fungal isolates in 2012 and

2013, respectively, shared nt identities of 64% nt, which indicates the two fungal isolates may be infected with the same partitivirus. The percentage identity of the 5'

UTRs of RdRp-9 and CP-c2 is slightly outside the range of identities shared by 5'

UTRs of other species of alphapartitivirus (68-90%). It was surprising that most of the

RdRp and CP segments identified in 2012 were not readily identified as pairs from 5'

UTRs identity. This could be explained by the pairwise identities of the 5' UTRs of these segments being lower than those of other alphapartitivirus species, or because complete sequences of many of the segments were not obtained. This was certainly true for the sample collected in 2013. In the latter case, sequencing to greater depth would probably identify the missing segments. Phylogenetic analysis of deduced aa sequences of the ORFs of segments placed them all in Alphapartitivirus and

Betapartitivirus, supporting the hypothesis that at least some of the corresponding pairs were present.

Table 3.2. Presence of Ceratobasidium partitivirus (A) CPs and (B) RdRps in fungal isolates F-2012 and F-2013 associated with Pterostylis sanguinea underground stems. Method of detection is represented in parenthesis – HTS: high-throughput sequencing, RT-PCR: reverse transcription-PCR.

(A) (B) Partitivirus CP F-2012 F-2013 Partitivirus RdRp F-2012 F-2013

Ceratobasidium partitivirus CP-a + (HTS) + (HTS) Ceratobasidium partitivirus RdRp-1 + (HTS, RT-PCR) + (RT-PCT)

Ceratobasidium partitivirus CP-b + (HTS) + (HTS) Ceratobasidium partitivirus RdRp-2 + (HTS, RT-PCR) + (RT-PCT)

Ceratobasidium partitivirus CP-c + (HTS) + (HTS) Ceratobasidium partitivirus RdRp-3 + (HTS, RT-PCR) + (RT-PCT)

Ceratobasidium partitivirus CP-d + (HTS, RT-PCR) - Ceratobasidium partitivirus RdRp-4 + (HTS, RT-PCR) + (RT-PCT)

Ceratobasidium partitivirus CP-e + (HTS, RT-PCR) + (RT-PCR) Ceratobasidium partitivirus RdRp-5 + (HTS, RT-PCR) + (RT-PCT)

Ceratobasidium partitivirus CP-f - + (HTS, RT-PCR) Ceratobasidium partitivirus RdRp-6 + (HTS, RT-PCR) + (RT-PCT)

Ceratobasidium partitivirus CP-g + (HTS) + (HTS) Ceratobasidium partitivirus RdRp-7 + (HTS, RT-PCR) + (RT-PCT)

Ceratobasidium partitivirus CP-h + (RT-PCR) + (HTS, RT-PCR) Ceratobasidium partitivirus RdRp-8 + (HTS, RT-PCR) + (RT-PCT)

Ceratobasidium partitivirus CP-i - + (HTS, RT-PCR) Ceratobasidium partitivirus RdRp-9 + (HTS, RT-PCR) + (RT-PCT)

Ceratobasidium partitivirus RdRp-10 + (HTS, RT-PCR) + (RT-PCT)

Table 3.3. Mean pairwise identities between the 5' UTRs of the two genomic segments (CP and RdRp) of (A) same species and (B) different species within the genera Alphapartitivirus, Betapartitivirus, Deltapartitivirus and Gammapartitivirus.

(A) Mean % identity (identity range) Alphapartitivirus 80 (68-90) Betapartitivirus 76 (67-79) Deltapartitivirus 45 (41-48) Gammapartitivirus 50 (47-57)

(B) Mean % identity (identity range) RdRp

CP Alphapartitivirus Betapartitivirus Deltapartitivirus Gammapartitivirus

Alphapartitivirus 49 (39-85) 44 (27-50) 42 (30-49) 43 (35-50) Betapartitivirus 44 (37-55) 56 (42-81) 42 (34-50) 43 (34-49) Deltapartitivirus 44 (37-51) 44 (36-55) 45 (37-51) 41 (32-49) Gammapartitivirus 40 (33-48) 42 (38-50) 40 (31-50) 51 (48-55)

Table 3.4. Pairwise comparison of 5' UTR sequences (nt) of Ceratobasidium partitivirus RdRp and CP segments. Sequences estimated to have less than 50% of 5' UTR sequences were omitted (RdRp-8, CP-g1 and CP-i). Letters in parentheses represent the proposed generic classifications of each sequence – (AP) Alphapartitivirus and (BP) Betapartitivirus. Proposed pairings of CPs and RdRps based on pairwise identities of 5' UTR sequences (nt) are indicated by colour codes.

F-2012 F-2013 CP-a1 (BP) CP-b1 (AP) CP-c1 (AP) CP-d (AP) CP-e (AP) CP-j (AP) CP-f (BP) CP-h (BP) CP-c2 (AP) CP-a2 (BP) RdRp-1 (BP) 42.1 44.9 50.0 40.7 37.4 47.2 39.6 44.3 48.0 38.5 RdRp-2 (BP) 76.8 44.4 55.0 39.3 41.2 35.2 44.9 50.0 37.0 46.4 RdRp-3 (BP) 48.6 47.3 56.4 43.9 42.7 44.7 50.0 51.6 45.5 42.0 RdRp-4 (AP) 46.7 72.7 46.2 40.5 44.7 49.3 45.7 38.5 38.3 46.2

RdRp-5 (AP) 39.3 55.3 50.0 50.0 44.3 44.6 37.7 38.6 42.6 42.6

2012

- RdRp-6 (AP) 45.3 44.1 44.4 34.4 38.9 45.8 40.4 44.3 40.6 46.9

F RdRp-7 (AP) 46.2 43.5 55.3 37.5 40.2 42.3 46.9 48.4 43.2 47.1 RdRp-9 (AP) 51.7 47.2 52.6 44.4 42.9 44.3 37.5 38.2 64.4 51.7 RdRp-10 (AP) 51.4 56.9 50.0 43.5 43.6 45.3 39.6 47.8 59.4 53.0 RdRp-12 (AP) 44.7 40.6 52.6 44.4 50.0 42.9 45.9 44.3 48.6 47.7

3.4.4 Other viruses and viral-like contigs

In addition to partitiviruses, other viruses and viral contigs were identified from this P. sanguinea population over the two-year period (Chapter 4). Two mycovirus-like viruses (Pterostylis sanguinea virus A and Pterostylis sanguinea totivirus A; PsVA and PsTVA), were identified from orchid leaf tissues in 2012 and

2013, respectively. Four mycoviruses that were not partitiviruses, two isolated in each year, were present in Ceratobasidium isolates. Other short (estimated <50% of genome) virus-like sequences were also detected from the orchid leaf tissue and from the fungal isolates (Table S1; Table S1 in Chapter 4). These most closely matched species from six virus families and seven genera.

3.5 Discussion

High-throughout sequencing was used in this study to identify 16 partitiviruses

(and six other viruses) associated with a small population of P. sanguinea plants and their mycorrhizial fungi at two time points. Detection of partial viral genomes and missing long RdRp segments, notably from the 2013 mycorrhizal fungal culture, suggests that sequencing depth was insufficient to capture all of the viral genetic material present. This was verified when RT-PCR-based assays confirmed that the apparently missing partitivirus RdRp segments were indeed present.

3.5.1 Ceratobasidium as a virus host

Ceratobasidium spp., together with species of Sebacina, Thanatephorus and

Tulasnella form the Rhizoctonia (sensu lato) fungi, which are responsible for mycorrhizal associations with the majority of orchids (Warcup, 1981; Bonnardeaux et al., 2007; Smith and Read, 2010). As an orchid mycorrhizal fungus, Ceratobasidium

is associated with orchid species worldwide, including members of Calanthe,

Prasophyllum, Pterostylis and Pyrochis (Warcup, 1981; Dearnaley and Le Brocque,

2006; Bonnardeaux et al., 2007). Ceratobasidium isolates F-2012 and F-2013 shared high (51-99%) nucleotide identities in their ITS regions with other orchid-associated

Ceratobasidium fungi from Australia and worldwide. Ceratobasidium spp. also function as endophytic, pathogenic and saprophytic fungi (Brundrett et al., 2003;

Brundrett, 2006; Mosquera-Espinosa et al., 2013).

The influence that mycoviruses have on their mycorrhizal hosts is largely unknown. Recent studies have confirmed the presence of diverse mycoviruses from both ascomycetous (Stielow and Menzel, 2010; Stielow et al., 2011; Stielow et al.,

2012) and basidiomycetous mycorrhizal hosts (Ong et al., 2016; Petrik et al., 2016).

All the mycoviruses identified from Ceratobasidium species so far have been identified from isolates associated with orchids. Virus-like rod-shaped particles were present in an isolate of Ceratobasidium cornigerum from the orchid Spiranthes sinensis (James et al., 1998). More recently, eight endornaviruses (family

Endornaviridae, genus Endornavirus) were identified from four isolates of

Ceratobasidium sp., isolated from two Australian species orchids (Microtis media and

Pterostylis sp.; Ong et al., 2016). More viruses have been identified from the

Ceratobasidium anamorph Rhizoctonia, including endornaviruses (Das et al., 2014; Li et al., 2014), mitoviruses (Lakshman and Tavantzis, 1994; Lakshman et al., 1998) and partitiviruses (Strauss et al., 2000; Zheng et al., 2014). The presence of 28 diverse mycoviruses (four families, five genera and unclassified mycoviruses) in only six

Ceratobasidium isolates (current study; Chapter 4; Ong et al., 2016) suggests that these fungal taxa might be host to an abundance of mycoviruses. Given the diverse

ecological roles of Ceratobasidium spp. and their potential interactions with multiple organism groups, their mycoviruses might have significant roles within ecosystems.

3.5.2 Australian partitiviruses in a world context

Betapartitiviruses have not previously been described from Australia, but they have been described from other continents, including Asia, Europe and North

America. The only other partitivirus described from Australia, from the leaves of an orchid, is the alphapartitivirus DPCV (Wylie et al., 2013). The close relationship of

Australian Ceratobasidium partitiviruses to those distributed internationally (Fig 3.2) suggests a natural movement of partitiviruses between continents. We assume that partitiviruses are spread over long distances in wind-borne fungal inocula.

Although two Australian Ceratobasidium partitivirus CPs (CP-d and CP-e) are genetically distinct from other internationally-distributed partitiviral CPs, this was not reflected in the RdRps, which all grouped with internationally widespread forms (Fig

3.2). This suggests that CP and RdRp segments are subjected to differential rates of evolutionary change (although incomplete sequence data of RdRp molecules may also account for this). Shorter branch lengths in the RdRp-generated phylogeny (Fig 3.2) suggest that partitivirus RdRps evolve at a slower rate than CPs. Within the same fungal host, members of both partitivirus genera co-occurred. The cost/benefit tradeoffs of partitivirus infection to the fungus and/or plant remain unknown. In some legumes there is a clear benefit; the betapartitivirus white clover cryptic virus 2 regulates nodulation in the presence of atmospheric nitrogen (Nakatsukasa-Akume et al., 2005).

Some of the new partitiviruses were genetically closer to plant-sourced partitiviruses than to fungal ones (Fig 3.2). This is consistent with a hypothesis that partitiviruses can be transmitted between endophytic fungi and host plants (Roossinck,

2010; Roossinck, 2013). The relatively high sequence identities of the RdRp of the plant-derived partitiviruses DPCV (Wylie et al., 2013) and cherry chlorotic rusty spot associated partitivirus (CCRSAPV; Coutts et al., 2004) with the fungus-derived

Ceratobasidium partitivirus RdRp-4 (67% aa, 63% nt) and Ceratobasidium partitivirus RdRp-6 (65% aa, 65% nt), respectively, supports this hypothesis.

3.5.3 The challenge of matching viral segments

The large number of partitivirus genome segments identified from both mycorrhizal fungi isolates presented challenges in determining the number of species present, and in matching corresponding CP and RdRp segments to identify species. A possible method of matching segments is to assume that both genome segments of a species are of similar masses, and use this property to distinguish them. Limitations to this approach are:

(i) the validity of the underlying assumption that both segments of all

partitiviruses share closely similar masses, and

(ii) assuming that sufficient segment size differentials exist between species in

mixed infections.

In mixed infections of partitiviruses, it is unclear if each partitivirus RdRp replicates and is encapsidated by a specific CP, or if multiple RdRp segments can share a CP. In co-infections, molecules must distinguish their partners from others. In mixed begomovirus (ssDNA plant viruses) infections, a conserved 200 nt sequence

present in both the DNA-A and DNA-B components of the virus enables recognition of the appropriate segments (Briddon et al., 2010). We compared sequences of complete genomes of previously described partitiviruses. No conserved region was found within coding sequences, but as shown previously (Lesker et al., 2013; Nibert et al., 2014), higher identities were found between respective 5' UTRs. Stem-loop structures in partitivirus 5' UTRs are proposed to be involved in dsRNA replication and virion assembly, so it seemed reasonable to assume that this structure might be recognized by RdRps encoded by the virus. 5' UTR sequences of the CP and RdRp segments of Dill cryptic virus 2 (DCV2; Betapartitivirus) share 85% nt identity, whereas the coding regions of the two segments shared only 45% nt identity. When applied to CPs and RdRps of known partitiviruses, we showed the proposal of using 5'

UTRs to match the protein fragments would be effective for known members of

Alphapartitivirus and Betapartitivirus, but not for Deltapartitivirus and

Gammapartitivirus (Table 3.3). If 5' UTR identity is important for segment recognition in alpha- and betapartitviruses, presumably it is less important in delta- and gammapartitiviruses.

The 5' UTRs comparison of described Ceratobasidium partitivirus sequences remains as preliminary results due to the lack of stop codon upstream of the proposed start codon in some of the sequences. It is uncertain if the stated 5' UTRs in these sequences represent the actual 5' UTRs. Their proposed 5' UTRs and starting ‘Met’ were predicted based on alignment with known partitiviruses and their sequence lengths. The upstream stop codon was present in four of the six segments (Cp-a1, CP- b1, CP-c2 and RdRp-9) in the three proposed pairings – CP-a1 with RdRp-2, Cp-b1 with RdRp-4 and CP-c2 with RdRp-9 (Table 3.4). Despite this, the pairings showed

much higher pairwise identity than their ORFs and were supported when phylogenetic analyses placed proteins of each pairing in the same genera (Fig 3.2); evidence that support the accurate representation of the 5' UTRs and accuracy of the pairings.

3.5.4 Virus composition of mycorrhizal strains

The fungal isolates from 2012 and 2013 shared most of the identified partitiviruses (Table 3.2). Other (non-partitiviral) viruses were less consistent – different viruses from different virus genera were detected in the two years (Chapter

4). Pterostylis underground stems are re-colonised annually by fungi from surrounding soil (Ramsay et al., 1986). The difference in virus species in the two mycorrhizal fungal isolates over the two growing seasons suggests that either re- colonisaion of orchids each year resulted in different lineages of Ceratobasidium, or the mycovirus composition within the same strain changes between the years. The majority of the persistent partitiviruses appeared to remain within their host while allowing for accumulation of other mycoviruses, including other partitiviruses, between the two growing seasons. Although both mycorrhizal isolates, F-2012 and F-

2013, were collected from the same orchid population, it is uncertain if they were derived from the same plant, or if one plant population can be simultaneously colonised by two Ceratobasidium lineages. This study site is subjected to a

Mediterranean-type climate of cool wet winters and hot dry summers, thus it is likely that viruses remain in dormant hyphae over the summer before autumn rain reactivates hyphae that can colonise newly developing orchid tubers or root structures

(Sivasithamparam, 1993).

The majority of the mycoviruses in both mycorrhizal isolates were partitiviruses, which are assumed to persistently infect fungal hosts over long periods.

Transfer of a persistent virus to another strain of the same fungus species can occur only between Ceratobasidium strains of the same anastomosis group (Parmeter e al.

1969). Ramsay et al. (1987) demonstrated that 18 of 19 isolates of mycorrhizal

Ceratobasidium isolates from P. sanguinea belonged to anastomosis group 1.

3.6 References

Altschul, SF, W Gish, W Miller, EW Myers and DJ Lipman. 1990. Basic local alignment search tool. J. Mol. Biol. 216(3): 403-410.

Bonnardeaux, Y, M Brundrett, A Batty, K Dixon, J Koch and K Sivasithamparam. 2007. Diversity of mycorrhizal fungi of terrestrial orchids: compatibility webs, brief encounters, lasting relationships and alien invasions. Mycol. Res. 111(1): 51-61.

Bougoure, J, M Ludwig, M Brundrett and P Grierson. 2009. Identity and specificity of the fungi forming mycorrhizas with the rare mycoheterotrophic orchid Rhizanthella gardneri. Mycol. Res. 113(10): 1097-1106.

Briddon, RW, BL Patil, B Bagewadi, MS Nawaz-ul-Rehman and CM Fauquet. 2010. Distinct evolutionary histories of the DNA-A and DNA-B components of bipartite begomoviruses. BMC Evol. Biol. 10(1): 97.

Brierley, I, AJ Jenner and SC Inglis. 1992. Mutational analysis of the “slippery- sequence” component of a coronavirus ribosomal frameshifting signal. J. Mol. Biol. 227(2): 463-479.

Brundrett, MC. 2006. Understanding the Roles of Multifunctional Mycorrhizal and Endophytic Fungi. Microbial Root Endophytes. BJE Schulz, CJC Boyle and TN Sieber, 281-298. Germany: Springer Berlin Heidelberg.

Brundrett, MC. 2014. Identification and Ecology of Southwest Australian Orchids. Perth: Western Australian Naturalists’ Club.

Brundrett, MC, A Scade, AL Batty, KW Dixon and K Sivasithamparam. 2003. Development of in situ and ex situ seed baiting techniques to detect mycorrhizal fungi from terrestrial orchid habitats. Mycol. Res. 107(10): 1210-1220.

Clements, MA and RK Ellyard. 1979. The symbiotic germination of Australian terrestrial orchids. Amer. Orchid Soc. Bull. 48: 810-816.

Coutts, RH, L Covelli, F Di Serio, A Citir, S Açıkgöz, C Hernandez, A Ragozzino and R Flores. 2004. Cherry chlorotic rusty spot and Amasya cherry diseases are associated with a complex pattern of mycoviral-like double-stranded RNAs. II. Characterization of a new species in the genus Partitivirus. J. Gen. Virol. 85(11): 3399-3403.

Das, S, RE Falloon, A Stewart and AR Pitman. 2014. Molecular characterisation of an endornavirus from Rhizoctonia solani AG-3PT infecting potato. Fungal Biol. 118(11): 924-934.

Dearnaley, JD and AF Le Brocque. 2006. Endophytic fungi associated with Australian orchids. Australas. Plant Conserv. 15(2): 7-9.

Ghabrial, SA, ML Nibert, E Maiss, TS Baker and YJ Tao. 2012. Partitiviridae. Virus taxonomy: classification and nomenclature of viruses: Ninth Report of the International Committee on Taxonomy of Viruses. AMQ King, MJ Adams, EB Carstens and EJ Lefkowitz, 523-534. San Diego: Elsevier Academic Press.

Gibbs, A, A Mackenzie, A Blanchfield, P Cross, C Wilson, E Kitajima, M Nightingale and M Clements. 2000. Viruses of orchids in Australia: their identification, biology and control. Austral. Orchid Rev. 65(3): 10-21.

Gowland, KM, MM van der Merwe, CC Linde, MA Clements and AB Nicotra. 2013. The host bias of three epiphytic Aeridinae orchid species is reflected, but not explained, by mycorrhizal fungal associations. Am. J. Bot. 100(4): 764-777.

Hacker, CV, CM Brasier and KW Buck. 2006. Determination of the 5′- and 3′- terminal sequences completes the sequences of the two double-stranded RNAs of Penicillium stoloniferum virus S. Virus Genes 32(2): 137-138.

James, JD, GC Saunders and SJ Owens. 1998. Isolation and Partial Characterisation of Double-stranded RNA-containing Viruses of Orchid Mycorrhizal Fungi. Mycorrhiza Manual. A Varma, 425-436. Germany: Springer-Verlag.

Lakshman, D and S Tavantzis. 1994. Spontaneous appearance of genetically distinct double-stranded RNA elements in Rhizoctonia solani. Phytopathology 84(6): 633-639.

Lakshman, DK, J Jian and SM Tavantzis. 1998. A double-stranded RNA element rom a hypovirulent strain of Rhizoctonia solani occurs in DNA form and is genetically related to the pentafunctional AROM protein of the shikimate pathway. Proc. Natl. Acad. Sci. U.S.A. 95(11): 6425-6429.

Lesker, T, F Rabenstein and E Maiss. 2013. Molecular characterization of five betacryptoviruses infecting four clover species and dill. Arch. Virol. 158(9): 1943- 1952.

Li, W, T Zhang, H Sun, Y Deng, A Zhang, H Chen and K Wang. 2014. Complete genome sequence of a novel endornavirus in the wheat sharp eyespot pathogen Rhizoctonia cerealis. Arch. Virol. 159(5): 1213-1216.

Marchler-Bauer, A and SH Bryant. 2004. CD-Search: protein domain annotations on the fly. Nucleic Acids Res. 32(suppl 2): W327-331.

Morris, TJ and JA Dodds. 1979. Isolation and analysis of double-stranded RNA from virus-infected plant and fungal tissue. Phytopathology 69: 854-858. 60

Mosquera-Espinosa, AT, P Bayman, GA Prado, A Gómez-Carabalí and JT Otero. 2013. The double life of Ceratobasidium: orchid mycorrhizal fungi and their potential for biocontrol of Rhizoctonia solani sheath blight of rice. Mycologia 105(1): 141-150.

Nakatsukasa-Akune, M, K Yamashita, Y Shimoda, T Uchiumi, M Abe, T Aoki, A Kamizawa, S-i Ayabe, S Higashi and A Suzuki. 2005. Suppression of root nodule formation by artificial expression of the TrEnodDR1 (coat protein of White clover cryptic virus 1) gene in Lotus japonicus. Mol. Plant Microbe Interact. 18(10): 1069- 1080.

Nibert, ML, SA Ghabrial, E Maiss, T Lesker, EJ Vainio, D Jiang and N Suzuki. 2014. Taxonomic reorganization of family Partitiviridae and other recent progress in partitivirus research. Virus Res. 188: 128-141.

Ong, JW, H Li, K Sivasithamparam, KW Dixon, MG Jones and SJ Wylie. 2016. Novel Endorna-like viruses, including three with two open reading frames, challenge the membership criteria and taxonomy of the Endornaviridae. Virology 499: 203-211.

Parmeter, JR, RT Sherwood and WD Platt. 1969. Anastomosis grouping among isolates of Thanatephorus cucumeris. Phytopathology 59(9): 1270-1278.

Petrzik, K, T Sarkisova, J Starý, I Koloniuk, L Hrabáková and O Kubešová. 2016. Molecular characterization of a new monopartite dsRNA mycovirus from mycorrhizal Thelephora terrestris (Ehrh.) and its detection in soil oribatid mites (Acari: Oribatida). Virology 489: 12-19.

Ramsay, RR, KW Dixon and K Sivasithamparam. 1986. Patterns of infection and endophytes associated with Western Australian orchids. Lindleyana 1(3): 203-214.

Ramsay, RR, K Sivasithamparam and KW Dixon. 1987. Anastomosis groups among Rhizoctonia-like endophytic fungi in southwestern Australian Pterostylis species (Orchidaceae). Lindleyana 2(3): 161-166.

Roossinck, MJ. 2010. Lifestyles of plant viruses. Philos. Trans. R. Soc. Lond., B, Biol. Sci. 365(1548): 1899-1905.

Roossinck, MJ. 2013. Plant virus ecology. PLoS Pathog. 9(5): e1003304.

Sivasithamparam, K. 1993. Ecology of root infecting pathogenic fungi in Mediterranean environments. Advances in Plant Pathology. JH Andrews and IC Tommerup, Vol 10, 245-279. London: Academic Press.

Smith, SE and DJ Read. 2010. Mycorrhizal Symbiosis. London: Academic press.

Stielow, B, H-P Klenk and W Menzel. 2011. Complete genome sequence of the first endornavirus from the ascocarp of the ectomycorrhizal fungus Tuber aestivum Vittad. Arch. Virol. 156(2): 343-345.

Stielow, B and W Menzel. 2010. Complete nucleotide sequence of TaV1, a novel totivirus isolated from a black truffle ascocarp (Tuber aestivum Vittad.). Arch. Virol. 155(12): 2075-2078.

Stielow, JB, Z Bratek, H-P Klenk, S Winter and W Menzel. 2012. A novel mitovirus from the hypogeous ectomycorrhizal fungus Tuber excavatum. Arch. Virol. 157(4): 787-790.

Strauss, EE, DK Lakshman and SM Tavantzis. 2000. Molecular characterization of the genome of a partitivirus from the basidiomycete Rhizoctonia solani. J. Gen. Virol. 81(2): 549-555.

Swarts, ND and KW Dixon. 2009. Terrestrial orchid conservation in the age of extinction. Ann. Bot. 104(3): 543-556.

Warcup, JH. 1981. The mycorrhizal relationships of Australian orchids. New Phytol. 87(2): 371-381.

White, TJ, T Bruns, S Lee and JW Taylor. 1990. Amplification and Direct Sequencing of Fungal Ribosomal RNA Genes for Phylogenetics. PCR Protocols: A Guide to Methods and Applications. MA Innis, DH Gelfand, JJ Sninsky and TJ White, 315-322. San Diego: Academic Press.

Wylie, SJ, H Li, KW Dixon, H Richards and MGK Jones. 2013. Exotic and indigenous viruses infect wild populations and captive collections of temperate terrestrial orchids (Diuris species) in Australia. Virus Res. 171(1): 22-32.

Zheng, L, M Zhang, Q Chen, M Zhu and E Zhou. 2014. A novel mycovirus closely related to viruses in the genus Alphapartitivirus confers hypovirulence in the phytopathogenic fungus Rhizoctonia solani. Virology 456: 220-226.

Supplementary information

(A)

(B)

(C)

Figure S1. Alignment of 5' UTRs of matched Ceratobasidium partitivirus CP and RdRp fragments – (A) CP-a1 and RdRp-2, (B) CP-b1 and RdRp-4, and (C) CP-c2 and RdRp-9.

Table S1. Partitivirus-like sequences (<50% of estimated genome) identified in Pterostylis sanguinea-associated Ceratobasidium species. Virus host Sequence Estimated percentage GenBank Name Best blastp match # (Sample no.) length of genome accession no. Ceratobasidium Ceratobasidium sp. Rosellinia necatrix partitivirus 2 1132 62% of CP KU291956 partitivirus-like contig 1 CP-j (F-2012) (Partitiviridae, Alphapartitivirus) Ceratobasidium Ceratobasidium sp. Heterobasidion partitivirus 8 1017 46% of CP KU291957 partitivirus-like contig 2 CP-g2 (F-2012) (Partitiviridae, Betapartitivirus) Ceratobasidium Ceratobasidium sp. Rosellinia necatrix partitivirus 2 702 41% of RdRp KU291958 partitivirus-like contig 3 RdRP-11 (F-2012) (Partitiviridae, Alphapartitivirus)

Ceratobasidium Ceratobasidium sp. Rosellinia necatrix partitivirus 5 637 31% of RdRp KU291959 partitivirus-like contig 4 RdRP-12 (F-2012) (Partitiviridae, Alphapartitivirus) Ceratobasidium Ceratobasidium sp. Diuris pendunculata cryptic virus 513 28% of CP KU291960 partitivirus-like contig 5 CP-k (F-2012) (Partitiviridae, Alphapartitivirus)

Ceratobasidium Ceratobasidium sp. Rosellinia necatrix partitivirus 4 851 35% of RdRp KU291961 partitivirus-like contig 6 RdRp-13 (F-2013) (Partitiviridae, Betapartitivirus) Ceratobasidium Ceratobasidium sp. Hop trefoil cryptic virus 2 760 31% of RdRp KU291962 partitivirus-like contig 7 RdRp-14 (F-2013) (Partitiviridae, Betapartitivirus)

Cherry chlorotic rusty spot associated Ceratobasidium Ceratobasidium sp. 522 partitivirus 34% of CP KU291963 partitivirus-like contig 8 CP-b2 (F-2013) (Partitiviridae, Alphapartitivirus) Ceratobasidium Ceratobasidium sp. Carrot cryptic virus 486 25% of RdRp KU291964 partitivirus-like contig 9 RdRp-15 (F-2013) (Partitiviridae, Alphapartitivirus)

Ceratobasidium Ceratobasidium sp. Dill cryptic virus 2

469 23% of CP KU291965

partitivirus-like contig 10 CP-a3 (F-2013) (Partitiviridae, Betapartitivirus)

Ceratobasidium Ceratobasidium sp. Rhizoctonia solani virus 717

454 19% of RdRp KU291966

partitivirus-like contig 11 RdRp-16 (F-2013) (Partitiviridae, Betapartitivirus)

#Estimated genome size is based on size of closest virus match from blastp.

Table S2. ClustalW comparison of amino acid (aa) and nucleotide (nt) identity of identified Ceratobasidium partitivirus-like CP sequences. CPs labelled in blue fonts represent Ceratobasidium partitivirus-like CP contigs (sequences of <1.3 kbp in length; Table S1).

aa F-2012 F-2013 nt CP-a1 CP-b1 CP-c1 CP-d CP-e CP-j CP-g2 CP-k CP-f CP-g1 CP-h CP-i CP-c2 CP-a2 CP-b2 CP-a3 CP-a1 11.6 10.6 10.7 12.7 10.0 15.0 13.4 19.8 17.3 16.2 14.8 9.9 98.5 10.7 98.7

CP-b1 43.0 23.8 13.1 16.6 12.0 15.2 15.1 13.2 13.8 14.8 15.1 25.0 11.3 99.4 15.3

CP-c1 43.4 44.0 15.6 16.8 14.2 14.7 28.4 10.6 13.7 12.1 10.0 97.0 11.9 21.1 11.2

CP-d 42.9 43.1 42.0 29.2 11.6 15.3 12.1 12.1 12.5 13.4 12.4 16.8 12.8 9.0 11.7

1012

F CP-e 42.5 43.0 42.3 41.1 11.6 12.0 10.6 11.4 12.8 11.6 12.2 15.4 11.3 10.5 10.8

CP-j 43.0 42.6 41.8 44.9 43.7 10.7 16.1 10.6 12.3 13.6 10.7 14.3 13.1 15.5 9.8

CP-g2 44.9 44.9 43.6 44.2 45.0 44.6 12.6 16.2 98.8 58.4 15.0 12.6 13.0 15.5 18.4

CP-k 43.0 44.1 51.6 45.3 46.4 44.6 43.6 8.1 11.4 9.2 11.0 23.6 13.8 9.4 11.0

CP-f 45.8 45.2 43.1 42.8 43.3 43.7 45.0 46.4 15.7 16.0 74.9 10.4 11.8 8.6 7.4

CP-g1 45.1 43.9 44.6 43.8 42.7 44.0 99.1 43.8 43.9 50.1 14.7 14.3 14.5 15.5 11.3

CP-h 43.8 42.5 43.3 41.9 42.6 43.4 62.5 44.7 43.9 60.1 10.8 11.0 11.5 12.0 11.1

CP-i 43.7 43.2 43.2 43.6 42.3 42.9 47.1 43.8 81.2 44.3 41.5 12.5 19.6 7.1 10.0

2013

- CP-c2 44.4 45.0 96.9 42.9 42.0 41.8 43.4 52.9 44.3 45.3 43.0 43.4 10.6 21.6 11.3

F CP-a2 97.9 43.7 42.5 41.0 41.6 44.1 43.9 43.2 45.5 43.6 44.1 42.4 43.5 13.6 13.1

CP-b2 42.7 98.9 45.3 44.4 43.6 43.7 42.1 41.8 45.1 44.7 43.1 44.2 41.3 42.7 14.2

CP-a3 99.1 45.8 42.2 44.2 42.8 45.2 44.2 41.4 46.7 43.8 43.6 46.9 45.6 42.3 44.4

Table S3. ClustalW comparison of amino acid identity (aa) and nucleotide (nt) of identified Ceratobasidium partitivirus-like RdRp sequences. Ceratobasidium partitivirus-like RdRps-11-16 (Table S1) are sequences of <1.3 kbp in length.

F-2012 F-2013 aa RdRp RdRp RdRp RdRp RdRp RdRp RdRp RdRp RdRp RdRp RdRp RdRp RdRp RdRp RdRp RdRp nt 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 RdRp1 33.0 58.5 19.5 18.7 20.1 16.9 16.6 19.6 17.4 18.2 11.1 22.6 38.0 17.8 13.8

RdRp2 49.6 36.3 20.6 19.3 20.9 20.3 16.9 19.4 17.0 17.3 11.2 38.1 59.6 17.9 16.3

RdRp3 62.9 50.9 20.1 21.6 19.2 18.7 18.3 22.0 19.2 9.7 12.4 10.4 41.6 12.5 15.7

RdRp4 43.9 45.8 42.8 25.7 64.3 25.0 26.8 43.2 35.1 31.0 16.5 10.2 13.3 24.7 12.7

RdRp5 43.7 43.9 41.5 46.4 24.7 44.6 37.6 29.6 24.5 36.8 22.0 10.7 16.8 17.4 12.2

RdRp6 45.2 45.0 44.3 64.8 46.1 26.9 25.9 41.2 33.0 28.1 13.2 10.2 14.8 29.5 11.0

2012 - RdRp7 43.3 43.8 43.3 47.1 50.8 45.5 33.8 29.4 25.6 36.8 19.2 10.9 16.2 19.1 15.0

RdRp8 42.6 43.1 42.8 46.0 48.9 45.3 49.3 25.9 21.5 55.1 63.2 11.2 12.6 19.8 10.6

RdRp9 44.0 45.8 43.6 53.0 48.6 50.6 47.5 45.3 86.3 31.4 16.8 11.8 23.1 27.6 8.5

RdRp10 45.4 44.3 45.0 48.2 46.1 47.7 45.6 44.0 87.2 18.2 20.7 13.1 15.9 21.5 9.8

RdRp11 42.3 46.0 44.0 45.8 50.4 47.1 53.0 59.5 48.7 46.3 13.8 11.6 10.9 21.7 13.2

RdRp12 44.1 43.3 41.0 43.6 45.7 45.2 46.7 60.7 47.9 46.4 41.3 12.3 13.6 12.7 12.3

RdRp13 46.7 55.0 44.1 44.8 45.0 45.0 42.3 42.3 44.6 45.2 44.1 43.3 12.2 6.7 9.2

RdRp14 49.9 62.0 52.6 42.9 44.4 45.6 44.6 45.6 45.9 46.3 40.0 42.7 42.3 13.6 13.3

2013 - RdRp15 46.9 45.8 46.3 60.0 45.5 80.5 46.8 46.5 49.7 49.7 47.8 44.0 44.2 41.6 11.7

RdRp16 47.8 86.3 45.6 43.5 46.0 46.8 45.2 43.6 44.3 44.7 42.7 43.5 45.2 44.1 42.7

Chapter 4: Australian terrestrial orchids and their fungal symbionts are hosts of novel and divergent viruses

4.1 Abstract

Terrestrial orchids represent a symbiotic union between plants and mycorrhizal fungi. This study describes the occurrence and nature of viruses associated with one population of wild Pterostylis sanguinea orchids and their fungal symbionts over two consecutive years. A generic sequencing approach, which combined dsRNA-enriched extraction, random amplification and high throughput sequencing, was used to identify presence of novel viruses. The majority of the virus- like sequences identified represent partial genomes and are based solely on the assembly of sequencing data. In leaf tissues we found three isolates of a novel totivirus and an unclassified virus, both resembling fungus-infecting viruses. Two mycorrhizal Ceratobasidium isolates from orchid underground stems contained at least 20 viruses, 16 of which were partitiviruses. A novel hypovirus and a mitovirus were genetically distant from existing members of the genera and did not readily fit into recognised subgroups. The high numbers of viruses associated with the orchids, in particular with their fungal partners, suggests that native orchid flora might represent a rich reservoir of novel and diverse viruses.

4.2 Introduction

Pterostylis is a genus of terrestrial orchids comprising over 200 species indigenous to Australia, Indonesia, New Caledonia, New Zealand and Papua New

Guinea (Janes & Duretto, 2010; Janes et al., 2010; Brundrett, 2014). Pterostylis

orchids and other terrestrial orchid genera represent a symbiosis between a plant and a fungus. Pterostylis orchids have short roots ranging from 5-10 cm in length and they form obligate fungal associations to provide water and nutrients from beyond the rhizosphere (Ramsay et al., 1986; Ramsay et al., 1987). Orchids differ from other composite organisms such as lichens in that the relationship is broken annually when the plant partner enters its dormant phase, and it becomes re-established when the shoot emerges from the underground tuber, which may occur up to several years later

(Brundrett, 2014). Pterostylis plants always establish mycorrhizal relationships with species of Ceratobasidium fungi (Warcup, 1973; Bonnardeaux et al., 2007), but it is unclear if the same species or strain of fungus re-establishes the relationship each year.

The viruses of cultivated orchids are widely studied. The most common viruses are Cymbidium mosaic virus (CymMV) and Odontoglossum ringspot virus

(ORSV) from families Alphaflexiviridae and Virgaviridae, respectively (Zettler et al.,

1990). They are spread during vegetative propagation, by vectors, and through trade in infected plants (Jensen, 1952; Blanchfield et al., 2001). In contrast, viruses infecting wild orchids are not well known. Exotic and indigenous viruses from five genera (Potexvirus, Potyvirus, Rhabdovirus, Tobamovirus, and Tospovirus) were described from a mixture of captive and wild orchids in eastern Australia (Gibbs et al.,

2000). In Western Australia, wild orchids were infected with exotic and indigenous members of Alphapartitivirus, Divavirus, Goravirus, Platypuvirus, Polerovirus and

Potyvirus (Wylie et al., 2012; Wylie et al., 2013a; Wylie et al., 2013b; Ong et al.,

2016a). In Japan, wild Calanthe izu-insularis orchids were infected with cucumber mosaic virus (genus Cucumovirus) (Kawakami et al., 2008). In India, ORSV (genus

Tobamovirus), CymMV (genus Potexvirus) and a novel potyvirus infected wild

epiphytic orchids (Sherpa et al., 2006; Singh et al., 2007).

In the current study, a generic approach based on high throughput sequencing was used to identify RNA viruses infecting plants of Pterostylis sanguinea (dark banded greenhood orchid) and mycorrhizal fungi associated with them. The samples were collected from a small natural population over two consecutive years. These orchids generate one or more new underground tubers each year, and the parent tuber dies. Tubers typically germinate unevenly, with some remaining dormant from one to several years (Brundrett, 2014). We describe the novel viruses identified from them and discuss them in ecological and evolutionary contexts.

4.3 Materials and methods

Experiments were carried out as specified in Chapter 3.3.

4.4 Results

4.4.1 De novo assembly

Three datasets of 153,581,198 and 92,046,118 and 35,630,376 reads, each of

101 nt, were generated from three independent Illumina sequencing runs. De novo assembly of the datasets generated a total of 119,854 contigs (12,896; 72,206 and

34,752 from each respective dataset) ranging from 200 nt to 22,685 nt, with a N50 length of 403, 359 and 305 respectively. Of these, 90 contigs were identified as virus- like; 14 were derived from P. sanguinea plants (474-10,716 nt) and 76 were from mycorrhizal fungi (407-8227 nt).

4.4.2 Identity of fungi

ITS sequences of both fungal isolates shared 99.7% nt identity, indicative they were of the same taxon of Ceratobasidium (Genbank accessions KU239992 (F-2012) and KU239993 (F-2013)). The ITS of F-2012 was most closely matched to that of

Ceratobasidium sp. (GQ405561; e-value: 0.0, 91% coverage and 99% nucleotide identity) and Ceratobasidium-anamorph, Rhizoctonia sp. (JQ859901; e-value: 0.0;

100% coverage and 97% nucleotide identity). F-2013 shared highest identity with

Ceratobasidium sp. (KT601568; e-value: 0.0, 99% coverage and 99% identity).

4.4.3 Viruses from orchid-associated mycorrhizal fungi

4.4.3.1 Ceratobasidium mitovirus A (CbMVA): a proposed new mitovirus

A virus-like contig of 2850 nt was identified from Ceratobasidium in 2012

(Table 4.1). 82,998 sequence reads were mapped to it with pairwise identity of 84.8%, and the mean coverage of the proposed virus genome was 3573.9-fold. There was a single ORF (nt 235-2688) whose encoded protein product has a predicted mass of 92 kDa. An RdRp-like domain was identified at aa 228-543 (nt 916-1863) (Fig 4.1A), indicative of a replicase function. Further support that the single ORF encoded a replicase was the existence of six core RdRp motifs between aa 297-508 (nt 1123-

1758) (Hong et al., 1999). The deduced protein sequence shared closest pairwise identities with replicases of mitoviruses (family Narnaviridae, genus Mitovirus), which infect the mitochondria of fungi (Hillman & Esteban, 2012). Mitovirus genomes typically comprise a single non-encapsidated positive-strand RNA of 2.3–

2.9 kb, which encodes a single protein of 80–104 kDa believed to function as a replicase (Hillman & Esteban, 2012).

The genome sequence was closest related to the mycorrhizal fungus-infecting mitoviruses Rhizophagus clarus mitovirus 1 (RcMV1-RF1) from Japan (27% aa, 45% nt), Rhizophagus sp. HR1 mitovirus (RMV-HR1) from Japan (27% aa, 45% nt) and

Tuber excavatum mitovirus (TeMV) isolated from Germany (20% aa, 43% nt), which together formed groups distinct from currently proposed mitovirus clades I and II (Fig

4.2A) (Doherty et al., 2006; Hillman & Cai, 2013). The deduced protein sequence shared 18-25% aa identity with other mitoviruses, figures below the accepted species demarcation limit of <40% (Hillman & Esteban, 2012). Thus, we propose that the sequence represents the complete genome of a previously-undescribed member of genus Mitovirus that we designate Ceratobasidium mitovirus A isolate Murdoch-1

(CbMVA; GenBank accession KU291923), named after the virus host genus and the location of its discovery.

(A) Frame: 235 RdRp 2688

+1 Mito_RdR p 5' UTR 916-1863 3' UTR

(B)

* RdRp RdRp_4 +2 5' UTR

+3 ORF1 5264 6752-7522 8227 42 4823 (C)

* +1 RdRp RdRp_4 * 3604 7089 +2 ORF1 5392-6201 1 3340

(D)

90 3725 RdRp * +3 ORF1 ORF2 5' UTR 3888 5631-5918 7143

(E)

6959 9758-10585 * RdRp RdRp_4 +2 1683-2045 5264 10716 +3 5' UTR N CP RdRp_4

UTR 1155 6893 dRp_4 8227

(F) 42 4823 2172 * +1 CP RdRp 1 * +3 RdRp_4 2196 3000-4088 4631

Figure 4.1. Proposed genome organisations of (A) Ceratobasidium mitovirus A

(Mitovirus) (B) Ceratobasidium virus A (unclassified mycovirus) (C) Ceratobasidium virus B (unclassified mycovirus) (D) Ceratobasidium hypovirus A (Hypovirus) (E)

Pterostylis sanguinea virus A (unclassified mycovirus-like) and (F) Pterostylis sanguinea totivirus A (Totivirus). Asterisks indicate incomplete 5' and/or 3' ends.

Shaded boxes represent Nudix hydrolase (N) and RNA dependent RNA polymerase domains.

Table 4.1. Viruses other than partitiviruses identified in Pterostylis sanguinea orchids and associated mycorrhizal fungi. Viruses were identified from pooled leaf tissue of three and four P. sanguinea plants (P-2012 and P-2013 respectively), and from Ceratobasidium isolates F-2012 and F- 2013, each from the underground stem of the one P. sanguinea plant.

Sequence Estimated Proposed Estimated Isolate length (nt) Virus host GenBank accession no. percentage GenBank Virus name classification Best blastp match percentage name [protein(s) (Sample no.) (e-value, % aa identity) of accession no. Family, Genus of genome* length (aa)] protein(s) #

Ceratobasidium Narnaviridae, 2850 Ceratobasidium sp. Rhizophagus sp. HR1 Murdoch-1 BAN85985 (9e-83, 31%) 100% 100% KU291923 mitovirus A (CbMVA) Mitovirus [817] (F-2012) mitovirus like ssRNA

Desulfovibrio oxyclinae Ceratobasidium virus A 8227 Ceratobasidium sp. transposase (ORF1); WP_026167673 (2.2, 26%) >60% Murdoch-2 Unclassified 92% KU291947 (CbVA) [988, >1593] (F-2012) Rosellinia necatrix mycovirus BAT50987 (1e-172, 35%) >90% 1-W1032/S5 (ORF2) Desulfovibrio oxyclinae Ceratobasidium virus A 7161 Ceratobasidium sp. transposase (ORF1); WP_026167673 (0.75, 26%) 100% Murdoch-3 Unclassified 80% KU291948 (CbVA) [1593, >632] (F-2012) Rosellinia necatrix mycovirus BAT50987 (3e-121, 37%) >50% 1-W1032/S5 (ORF2)

Ceratobasidium virus A 4051 Ceratobasidium sp. Rosellinia necatrix mycovirus Murdoch-4 Unclassified BAT50987 (7e-176, 34%) >70% 45% KU291949 (CbVA) [>1123] (F-2012) 1-W1032/S5

Ceratobasidium virus B 7089 Ceratobasidium sp. Rhizoctonia solani RNA virus YP_009158859 (2e-04, 25%) >90% Murdoch-5 Unclassified 93% KU291938 (CbVB) [>1112, >1162] (F-2012) HN008 YP_009158860 (1e-166, 31%) >90%

Ceratobasidium Hypoviridae, 7143 Ceratobasidium sp. NP_041091 (1e-12, 30%) 100% Murdoch-6 Cryphonectria hypovirus 1 56% KU291924 hypovirus A (CbHVA) Hypovirus [1211, >1085] (F-2013) ABI64296 (8e-30, 26%) >30%

Pterostylis sanguinea 10,716 P. sanguinea Lentinula edodes mycovirus BAM34027 (1e-105, 26%) 100% Murdoch-7 Unclassified 95-100% KU291925 virus A (PsVA) [1912, >1252] (P-2012) HKA BAM34028 (5e-157, 35%) >90%

Pterostylis sanguinea Totiviridae, 4631 P. sanguinea YP_001497150 (6e-139, 36%) >90% Murdoch-8 Black raspberry virus F 92% KU291927 totivirus A (PsTVA) Totivirus [>723, >812] (P-2013) YP_001497151 (0.0, 56%) >90%

Pterostylis sanguinea Totiviridae, 3631 P. sanguinea YP_001497150 (2e-52, 32%) >50% Murdoch-9 Black raspberry virus F 72% KU291926 totivirus A (PsTVA) Totivirus [>382, >685] (P-2013) YP_001497151 (0.0, 56%) >80%

Pterostylis sanguinea Totiviridae, 3613 P. sanguinea YP_001497150 (2e-152, 37%) >90% Murdoch-10 Black raspberry virus F 72% KU291928 totivirus A (PsTVA) Totivirus [>696, >404] (P-2013) YP_001497151 (1e-169, 56%) >50%

*Calculation of genome and protein percentage was based on sequence length of closest blastp match and/or related virus isolates. # Estimated percentage of protein was limited by lack of complete ORF

(A) 100 Botrytis cinerea debilitation-related virus YP 002284334 (B) 100 Botrytis cinerea mitovirus 1 ABQ65153 Lentinula edodes mycovirus HKA BAM34028 Ophiostoma mitovirus 3a CAA06228 100 Lentinula edodes spherical virus AGH07919 91 Clade II 100 Sclerotinia sclerotiorum mitovirus 3 AGC24232 99 Lentinula edodes mycovirus HKB BAG71788 Thanatephorus cucumeris mitovirus AAD17381 Phlebiopsis gigantea mycovirus YP 003541123 78 Tuber aestivum mitovirus YP 004564622 Pterostylis sanguinea virus A 87 Cryphonectria parasitica mitovirus 1 AAA61703 Rhizoctonia fumigata mycovirus AJE29745 Helicobasidium mompa mitovirus BAD72871 Ceratobasidium virus B Ophiostoma mitovirus 6 CAB42654 99 Rhizoctonia solani RNA virus HN008 AKO82515 99 Clade I Gremmeniella mitovirus S1 AAN05635 Rhizophagus sp. RF1 medium virus BAJ23141 99 Rosellinia necatrix mycovirus 1-W1032/S5 BAT50987 98 Ophiostoma mitovirus 4 CAB42652 Ceratobasidium virus A Murdoch-4 83 Ophiostoma mitovirus 5 CAB42653 99 98 Ceratobasidium virus A Murdoch-2 83 Ceratobasidium mitovirus A 100 Ceratobasidium virus A Murdoch-3 Tuber excavatum mitovirus AEP83726 Rhizophagus sp. HR1 mitovirus BAN85985 0.5 100 Rhizophagus sp. RF1 mitovirus BAJ23143 Saccharomyces cerevisiae narnavirus 23S AAC98708 80 Pterostylis sanguinea totivirus A Murdoch-4 98 Pterostylis sanguinea totivirus A Murdoch-5 1 (D) 100 Pterostylis sanguinea totivirus A Murdoch-6 99 Black raspberry virus F RdRp YP 001497151 Totivirus (C) Tuber aestivum virus 1 ADQ54106 100 100 Saccharomyces cerevisiae virus L-A AAA50508 Cryphonectria hypovirus 3 AAF13604 99 Saccharomyces cerevisiae virus La AAB02146 98 Sclerotinia sclerotiorum hypovirus 1 YP 004782527 Betahypovirus 100 Leishmania RNA virus 1-4 AAB50028 Cryphonectria hypovirus 4 AAQ76546 100 Leishmania RNA virus 1-1 AAB50024 Leishmaniavirus Ceratobasidium hypovirus A Leishmania RNA virus 2-1 AAB50031 Fusarium graminearum hypovirus 1 AGC75065 100 Helicobasidium mompa no. 17 dsRNA virus BAC81754 100 Cryphonectria hypovirus 1 AAD13750 Alphahypovirus 100 Helminthosporium victoriae virus 190S AAB94791 Victorivirus 99 Cryphonectria hypovirus 2 AAA20137 100 Thielaviopsis basicola dsRNA virus 1 AAS68036 Plum pox virus NP 040807 Helminthosporium victoriae 145S virus YP 052858 Gardia lamblia virus AAB01579 Gardiavirus 0.5

1 Figure 4.2. Maximum likelihood phylogenetic trees constructed from RdRp deduced amino acid sequences of (A) proposed mitovirus

Ceratobasidium mitovirus A (indicated by a dot), (B) proposed Ceratobasidium virus A (indicated by dots), Ceratobasidium virus B (indicated by a square) and Pterostylis sanguinea virus A (indicated by a triangle), (C) proposed hypovirus Ceratobasidium hypovirus A (indicated by a dot), and (D) proposed totivirus Pterostylis sanguinea totivirus A (indicated by dots) with those of most closely related described viruses. 1000 bootstrap replications were carried out and branch confidence values below 60% were omitted. Branch lengths represent calculated evolutionary distance in units of amino acid substitutions per site.

4.4.3.2 Ceratobasidium virus A (CbVA): a proposed new mycovirus

Three partial monopartite virus sequences of 8227 nt, 7161 nt and 4051 nt respectively were identified from mycorrhizal fungus in 2012 (Fig 4.1B; Table 4.1).

These sequences were designated Ceratobasidium virus A (CbVA) isolates Murdoch-

2, Murdoch-3 and Murdoch-4 (GenBank accessions KU291947, KU291948 and

KU291949 respectively). Isolate Murdoch-2 was mapped to 9593 raw sequence reads with pairwise identity of 82.2% and a 119.6-fold mean coverage per base across the genome. 6792 reads were mapped to isolate Murdoch-3 with pairwise identity of

82.0% and mean coverage of 101.2-fold per base. Isolate Murdoch-4 was generated from 2558 reads at pairwise identity of 80.5% and mean coverage of 71.6-fold per base. CbVA isolates Murdoch-2 and Murdoch-3 had two non-overlapping ORFs – a complete ORF1 (177 kDa) and partial ORF2 (RdRp; >74kDa and >115 kDa) while the sequence of Murdoch-4, which represented about 45% of its genome, encoded an incomplete RdRp (>130 kDa). Neither a ‘slippery sequence’ nor pseudoknot, typical of ribosomal frameshift sites (Brierley et al., 1992), was detected upstream of proposed ORF2. Comparison of the isolates showed 43-80% nt identity between genomes, 91% aa identity (80% nt) between the ORF1s, and 53-95% aa identity (59-

81% nt) between the respective RdRps.

Blastp analysis showed that 9% of the translated product of ORF1s of isolates

Murdoch-2 and Murdoch-3 most closely matched the transposase of Desulfovibrio oxyclinae (Table 4.1). The encoded CbVA RdRps shared highest identity with RdRp of Rosellinia necatrix mycovirus 1-W1032/S5 (also named Yado-nushi virus

W1032a; Zhang et al., 2016) at a pairwise identity of 31-33% aa and 47-50% nt.

CbVA and Rosellinia necatrix mycovirus 1-W1032/S5 were grouped together and distantly related to other unclassified mycoviruses (Fig 4.2B).

4.4.3.3 Ceratobasidium virus B (CbVB): a proposed new mycovirus

A contig of 7089 nt representing Ceratobasidium virus B (CbVB) Murdoch-5

(GenBank accession KU291938) was mapped to 8508 reads with mean coverage of

127.3-fold per base across the genome and pairwise identity of 46.1%. CbVB encoded two partial non-overlapping ORFs representing a hypothetical protein (nt 1-3340;

>117 kDa) and an RdRp (nt 3604-7089; >31 kDa) (Fig 4.1C). There was no evidence of ribosomal frameshift ‘slippery sequence’ sites in the sequence. An RdRp_4 domain

(pfam02123) was identified at aa 597-866 (nt 5392-6201) (Fig 4.1C) and the core

RdRp motifs V and IV of T/SGx3 Tx3 NS/Tx22 GDD (where x is any residue)

(Koonin, 1991) were represented at aa 758-800 (nt 5875-6003) as SGx3 Tx3 NTx29

GDD.

Blast and phylogenetic analyses showed that CbVB grouped most closely with an unclassified mycovirus Rhizoctonia solani RNA virus HN008 (RsRV-HN008;

Zhong et al., 2015) (Fig 4.2B, Table 4.1). The two viruses shared 45% nt identity between genomes, 16% aa (43% nt) identity between ORF1s and 31% aa (48% nt) identity between ORF2s.

4.4.3.4 Ceratobasidium hypovirus A (CbHVA): a proposed new hypovirus

A contig of 7134 nt was generated from 5051 Illumina reads with mean coverage of 74.7-fold and pairwise identity of 85.5%. The partial genome sequence had a 5' UTR of 89 nt and two predicted ORFs (Fig 4.1D). ORF1 was 3636 nt in

length and is predicted to encode a protein of 133 kDa. The 3' part of the genome was not obtained, and so ORF2 is incomplete (nt 3888-7134; >121 kDa). Based on its length compared to related hypoviruses and the position of RdRp domain, the partial

ORF2 sequence represents about 60% of its complete ORF. Elements resembling slippery sequences and pseudoknots (Brieiley et al., 1992) upstream of ORF2 were absent. The RdRp domain was located at aa 582-677 (nt 5631-5918) (Fig 4.1D) and the core RdRp motifs V and IV were represented at aa residues 590-634 (nt 5655-

5789) as TGx3 Tx3 NTx31 GDD.

The virus represented by this sequence was designated Ceratobasidium hypovirus A (CbHVA) isolate Murdoch-6 (GenBank accession KU291924). We propose Ceratobasidium hypovirus A as a new hypovirus. CbHVA shared highest identity with the four definitive members of Hypovirus (Cryphonectria hypovirus 1-4,

CHV1-4; family Hypoviridae) that infect Cryphonectria parasitica, the Chestnut blight fungus (Table 4.1; Shapira et al., 1991; Hillman et al., 1994; Smart et al., 1999;

Linder-Basso et al., 2005) and two other proposed members infecting Fusarium species (Fusarium graminearum hypovirus 1 (FgHV1) from China (Wang et al.,

2013)) and Sclerotinia species (Sclerotinia sclerotiorum hypovirus 1 (SsHV1) from

China (Xie et al., 2011)). Hypoviruses are proposed to be categorised into subgroups

Alphahypovirus and Betahypovirus, distinguished by having either two or one ORF, respectively (Nuss & Hillman, 2012; Yaegashi et al., 2012). Having two ORFs,

CbHVA was expected to share greater sequence identity with alphahypoviruses than betahypoviruses, but phylogenetic analysis positions CbHVA equidistant between members of each group (Fig 4.2C). Comparison of CbHVA with the hypoviruses showed low aa identity; 8-15% aa identity with alphahypoviruses and 11-12% aa

identity with betahypoviruses. This is consistent with the species demarcation limit of less than 60% aa identity between CHV-1 and CHV-2 and 50% aa identity between CHV-3 and CHV-4 (Nuss & Hillman, 2012).

4.4.4 Virus-like sequences identified from leaf samples

Four distantly related virus-like sequences were identified from P. sanguinea leaf tissue samples, one in 2012 and three in 2013. Because these viruses resembled mycoviruses, but not known plant viruses, attempts to amplify fungal sequences from the leaf samples using primers ITS1 and ITS4 were carried out. However this test did not detect fungi from the leaf. This suggested that these two myco-like viruses may indeed be plant viruses.

4.4.4.1 Pterostylis sanguinea virus A (PsVA), a myco-like virus from leaf tissue

A partial monopartite virus genome sequence of 10,716 nt was identified from leaf tissue in 2012 (Fig 4.1E; Table 4.1). 56,636 101 nt reads were mapped to the sequence, with a pairwise identity of 87.4%, and a mean coverage of 1444.8-fold across its partial genome. This sequence was named Pterostylis sanguinea virus A

(PsVA) isolate Murdoch-7 (GenBank accession KU291925).

The PsVA genome has two consecutive non-overlapping ORFs of 5739 nt and

3758 nt, respectively (Fig 4.1E) encoded on adjacent frames. Ribosomal frameshift was not detected in this sequence. PsVA is predicted to have an unusually long 5'

UTR of 1154 nt. The first putative translational start codon is positioned at nt 1155 corresponding to the start of the hypothetical CP, estimated to have a mass of 208 kDa

(Fig 4.1E). A region encoding a Nudix hydrolase-like domain, responsible for

hydrolysis of nucleoside diphosphate derivatives, was detected at aa residues 177-297

(nt 1683-2045) (Fig 4.1E). In ORF2, an RdRp domain, which corresponds to similar domains in viruses belonging of Chrysovirus, Luteovirus, Rotavirus and Totivirus, was detected at aa residues 934-1209 (nt 9758-10,585) (Marchler-Bauer & Bryant,

2004; Marchler-Bauer et al., 2013). The highly conserved core RdRp motifs V and VI

(S/TGx3 Tx3 NS/Tx22 GDD) (Koonin, 1991) were present at aa 1109-1151 (nt

10,283-10,411) as SGx3 Tx4 NTx28 GDD.

ORF1 shared identity with the CP-encoding ORF1 of the monopartite mycovirus Lentinula edodes spherical virus (LeSV; an unclassified virus), identified from Shiitake mushroom (Lentinula edodes) in South Korea (Won et al., 2013). The

PsVA RdRp showed highest identity to RdRps of Lentinula edodes mycovirus isolates HKA and HKB (LeV; unclassified), also from Shiitake mushroom but from

Japan (Ohta et al., 2008; Magae, 2012), and the saprophytic fungus virus Phlebiopsis gigantea large virus-1 (PgLV-1; unclassified; Kozlakidis et al., 2009) (Fig 4.2B;

Table 4.1). ClustalW comparison of PsVA, LeV and PgLV-1 isolates showed 42-43% nt identity across the genomes. Identity between the homologous proteins of PsVA,

LeV and PgLV-1 was 19-20% aa identity (41% nt) between CPs and 28-30% aa identity (41-42% nt) between RdRps.

4.4.4.2 Pterostylis sanguinea totivirus A (PsTVA): three isolates of a proposed new totivirus from orchid plants

Three related contigs of 4643 nt, 3631 nt and 3613 nt were identified from leaf tissue of P. sanguinea plants collected in 2013 (Table 4.1). The sequences resembled those of members of genus Totivirus. Totiviruses have dsRNA genomes consisting of

a single molecule 4.6-7.0 kbp in length that encodes two usually overlapping ORFs

(Wickner et al., 2012). The three putative totivirus isolates encoded by these three partial genome sequences were designated Pterostylis sanguinea totivirus A (PsTVA) isolates Murdoch-8, Murdoch-9 and Murdoch-10 (GenBank accessions KU291927,

KU291926 and KU291928 respectively). 2970 raw sequence reads were mapped to

PsTVA-Murdoch-8 with pairwise identity of 81.7% and 61.9-fold mean coverage per base across its genome. Isolate Murdoch-9 was assembled from 94,091 reads at pairwise identity of 81.9% and mean coverage of 3142.8-fold per base. 68,653 reads were mapped to isolate Murdoch-10 with pairwise identity of 81.9% and mean coverage per base of 3153.0-fold. Each of the three sequences encoded two non- overlapping partial ORFs representing a CP and RdRp (Fig 4.1F), with no evidence of a slippery sequence indicating ribosomal frameshifting. The CPs had a L-A CP domain, which is typical of the yeast-infecting Totivirus type species Saccharomyces cerevisiae L-A virus (ScV-L-A). The L-A CP domain was located at nt residues 40-

1212 (Murdoch-8), 1-245 (Murdoch-9) and 9-1142 (Murdoch-10). RdRp_4-like domains were detected at nt 3000-4088 (Murdoch-8), nt 2009-3097 (Murdoch-9) and nt 2948-3589 (Murdoch-10). Conserved RdRp core motifs V and VI (Koonin, 1991) were located on the genomes of isolate Murdoch-8 at aa residues 527-564 (nt 3774-

3887) and isolate Murdoch-9 at aa 403-440 (nt 2783-2892) as SGx3 Tx3 NTx24 GDD.

Comparison of the isolates showed 42-60% nt identity between genomes, 36-44% aa identity (48-56% nt) between CPs and 62-65% aa identity (61-64% nt) between

RdRps. These figures are slightly below or above the suggested species demarcation limit for totivirus species of <50% aa identity (Wickner et al., 2012), indicating they may be categorised as the same species.

Blastp analysis of the two proteins revealed that the closest matches to described species were to the CP (32-37% aa identity) and RdRp (56-58% aa identity) of black raspberry virus F (BRVF; GenBank accession NC_009890), a proposed totivirus described from leaf tissue, or fungi infecting leaf tissue, of Rubus occidentalis in the USA. PsTVA also grouped with totiviruses that infect fungi and yeast (Fig 4.2D).

4.4.5 Partitiviruses and other virus-like sequences

In addition to the six viruses described above, there were alphapartitiviruses and betapartitiviruses (family Partitiviridae) associated with the mycorrhizal fungi, and these are described in the accompanying article (Chapter 3). There were at least

10 partitiviruses – seven alphapartitiviruses and three betapartitiviruses – found in fungal isolate F-2012. From isolate F-2013, five alphapartitiviruses and one betapartitiviruses were identified. Majority of these partitiviruses were subsequently detected in both mycorrhizal strains.

There is evidence from 41 short sequence fragments (454-3119 nt) that a number of other viruses were also present (Table S1; Table S1 in Chapter 3). These are not described in detail because they were estimated to represent less than 50% of genomes, thereby making assignment to taxonomic groups speculative. Closest matches to known viruses were predicted using Blastp and this revealed they most closely matched viruses from six virus families and seven genera, and some unclassified viruses (Table S1; Table S1 in Chapter 3). From leaf samples, a megabirnavirus-like sequence (P-2012), four other totivirus-like sequences (P-2013) and two related to members of the family Amalgaviridae (P-2013) were identified

(Table S1). The totivirus-like contigs were closely related to the three PsTVA isolates and may represent more isolates of PsTVA, or belong to related species. They shared

41-95% nt identity between genomes, 44-95% nt (12-93% aa) identity between RdRp sequences and 48-57% nt (11-49% aa) identity between CP sequences.

From the two fungal isolates, 34 further virus-like contigs were identified

(454-3119 nt) that most closely resembled species within the genera Alphapartitivirus,

Betapartitivirus, Endornavirus, Hypovirus, Megabirnavirus and unclassified mycoviruses (Table S1; Table S1 in Chapter 3). The four endornavirus-like contigs

(454-1245 nt) detected in Ceratobasidium sp. (F-2013) were matched to endornaviruses recently identified from mycorrhizal fungi of other terrestrial orchids in the region (Ong et al., 2016b). RT-PCR was done using primers specific to

Ceratobasidium endornaviruses A-H, and they confirmed the presence of

Ceratobasidium endornaviruses G and H (data not shown) (Ong et al., 2016b).

4.5 Discussion

A small wild population of three and four P. sanguinea plants collected in

2012 and 2013 and mycorrhizal fungi associated with two of the plants were found to be colonised by numerous persistent viruses, none of which had been described previously. At least 22 definitive viruses, proposed as belonging to the genera

Alphapartitivirus, Betapartitivirus, Hypovirus, Mitovirus, Totivirus and unclassified mycoviruses were identified. All but two of the new viruses were identified from pure cultures of Ceratobasidium derived from pelotons isolated from two P. sanguinea plants. The findings extend the geographical range of probable members of

Betapartitivirus (Ceratobasidium partitiviruses; Chapter 3), Hypovirus (CbHVA;

Ceratobasidium hypovirus-like contig 1-5), Megabirnavirus (Ceratobasidium megabirnavirus-like contigs 1-2; Pterostylis megabirnavirus-like contig 1), Mitovirus

(CbMVA) and Totivirus (PsTVA; Ceratobasidium totivirus-like contigs 1-4), which had not previously been identified from Australia. Although tentatively assigned classification by pairwise sequence identity with members of known groups, the proposed classifications are by no means certain because many sequences represented partial genomes, and most were genetically distant from described species.

4.5.1 Classification of new viruses

Most of the new viruses were tentatively classified with existing higher order taxa, but assigning them to existing lower order taxa was often problematical. For example, CbMVA was proposed as a member of Mitovirus, but it does not fit easily into the two proposed subgroups within the genus (Doherty et al., 2006). Instead it groups with other unclassified mycorrhizae-derived mitoviruses (Fig 4.2A) that usually encode tryptophan with UGG rather than UGA, which confers on them the capability of replicating in the host cytoplasm as well as its mitochondria (Kitahara et al., 2014). Similarly, the proposed hypovirus CbHVA is phylogenetically closest to the hypoviruses, but features of its genome organisation and host type place it outside the existing two hypovirus subgroups (Yaegashi et al. 2012). The two mycoviruses,

CbVA and CbVB share sequence identity and genome organisation with other mycoviruses from different continents, but none are currently assigned taxa. Together, these findings indicate that the evolutionary history of mycoviruses is more complex than currently recognised.

PsTVA is the only new virus that shared a relatively close evolutionary history with a previously described virus – the proposed totivirus black raspberry virus F.

Sequence identities of these two viruses are marginally above the species demarcation threshold of 50% aa identity for totiviruses set by the ICTV (Wickner et al., 2012), but given that the host species are distinct and their locations are widely separated, we propose that they belong to different species.

4.5.2 Host identification

Viruses identified from plant materials are usually assumed to be plant viruses, with no distinction being made between viruses capable of replicating in plant cells or fungal cells that co-occur with plant tissues. It can be difficult to ascertain the true host. The current experiment was designed to detect most RNA viruses, and therefore total leaf RNA was enriched for dsRNA before sequencing. An alternative approach would be to sequence total leaf RNA (depleted of ribosomal RNA) to eliminate the bias towards detection of dsRNA viral genomes and provide a clearer representation of the entire biome, including the presence of fungal transcripts and other RNA viruses. This metatranscriptomic approach was used recently in a study by Marzano et al. (2015), which identified 22 putative mycoviruses (dsRNA, ssRNA and ssDNA) from soybean leaf samples.

4.5.3 Viruses, fungi and orchids

In orchid biology, the plant-fungus symbiotic partnership is critical, but the roles viruses may play in this relationship remain largely unknown. Most mycoviruses appear to have little influence on fungal pathogenicity (Seo et al., 2004; Vainio et al.,

2012), but some are demonstrated to influence their hosts. The most well known

example of a mycovirus that reduces fungal pathogenicity is CHV1 of Cryphonectria parasitica, the fungal pathogen that causes chestnut blight (Anagnostakis & Day,

1979). The term ‘Rhizoctonia decline’ was used to describe the decreased growth rate and lack of sclerotia production of an isolate of Rhizoctonia solani infected with a dsRNA virus (Castanho & Butler, 1978a, b). In contrast, the plant pathogenic fungus

Nectria radicicola became more virulent in the presence of a 6.0 kbp dsRNA virus

(Ahn & Lee, 2001).

Ecological roles of mycoviruses may be elucidated when methods are developed to cure Ceratobasidium cultures of persistent viruses and reintroduce them one by one. It has been reported that culturing of some fungi in vitro can cause them to lose mycoviruses (Bao & Roossinck, 2013; Roossinck, 2015). Because the

Ceratobasidium isolates analysed here were cultured in vitro, this study may provide an underestimate of the number of mycoviruses that exist under natural conditions.

Ecological factors, such as drought, fire, grazing, low seed set, salinity of habitats and weed invasion, are known to impact negatively on orchid populations, and this is the case with many of the threatened orchid species in Western Australia

(Coates and Atkins, 2001; Swarts and Dixon, 2009; Brundrett, 2016). Whether viruses play positive or negative roles in orchid biology remains unclear. The identification of these viruses is an essential step in on-going studies of the interplay between wild plants, fungi and viruses. Such studies may shed light on understanding why populations of many orchids in south-west Australia, and globally, are shrinking alarmingly, while other orchid species are thriving to the point of becoming weeds, e.g. Microtis media and Disa bracteata (Bonnardeaux et al., 2007; Swarts & Dixon,

2009; De Long et al., 2013). More broadly, it may also provide clues to improving the efficiency of agricultural production through understanding the roles, positive or negative, that mycoviruses play in fungal interactions with crops.

4.6 References

Ahn, I-P and Y-H Lee. 2001. A viral double-stranded RNA up regulates the fungal virulence of Nectria radicicola. Mol. Plant Microbe Interact. 14(4): 496-507.

Anagnostakis, SL and PR Day. 1979. Hypovirulence conversion in Endothia parasitica. Phytopathology 69(12): 1226-1229.

Bao, X and MJ Roossinck. 2013. Multiplexed interactions: viruses of endophytic fungi. Adv. Virus Res. 86: 37-58.

Blanchfield, AL, AM Mackenzie, A Gibbs, H Kondo, T Tamada and CR Wilson. 2001. Identification of Orchid fleck virus by reverse transcriptase-polymerase chain reaction and analysis of isolate relationships. J. Phytopathol. 149(11‐12): 713-718.

Brierley, I, AJ Jenner and SC Inglis. 1992. Mutational analysis of the “slippery- sequence” component of a coronavirus ribosomal frameshifting signal. J. Mol. Biol. 227(2): 463-479.

Brundrett, MC. 2014. Identification and Ecology of Southwest Australian Orchids. Perth: Western Australian Naturalists’ Club.

Brundrett, MC. 2016. Using vital statistics and core-habitat maps to manage critically endangered orchids in the Western Australian wheatbelt. Aust. J. Bot. 64(1): 51-64.

Castanho, B and EE Butler. 1978a. Rhizoctonia decline: a degenerative disease of Rhizoctonia solani. Phytopathology 68(10): 1505-1510.

Castanho, B and EE Butler. 1978b. Rhizoctonia decline: studies on hypovirulence and potential use in biological control. Phytopathology 68(10): 1511-1514.

Coates, DJ and KA Atkins. 2001. Priority setting and the conservation of Western Australia's diverse and highly endemic flora. Biol. Conserv. 97(2): 251-263.

De Long, JR, ND Swarts, KW Dixon and LM Egerton-Warburton. 2013. Mycorrhizal preference promotes habitat invasion by a native Australian orchid: Microtis media. Ann. Bot. 111(3): 409-418.

Doherty, M, RHA Coutts, CM Brasier and KW Buck. 2006. Sequence of RNA- dependent RNA polymerase genes provides evidence for three more distinct mitoviruses in Ophiostoma novo-ulmi isolate Ld. Virus Genes 33(1): 41-44.

Gibbs, MJ, R Koga, H Moriyama, P Pfeiffer and T Fukuhara. 2000. Phylogenetic analysis of some large double-stranded RNA replicons from plants suggests they evolved from a defective single-stranded RNA virus. J. Gen. Virol. 81(1): 227-233.

Hillman, BI and G Cai. 2013. The family Narnaviridae: simplest of RNA viruses. Adv. Virus. Res. 86: 149-176.

Hillman, BI and R Esteban. 2012. Narnaviridae. Virus Taxonomy: Classification and Nomenclature of Viruses: Ninth Report of the International Committee on Taxonomy of Viruses. AMQ King, MJ Adams, EB Carstens and EJ Lefkowitz, 1055-1060. San Diego: Elsevier Academic Press.

Hillman, BI, BT Halpern and MP Brown. 1994. A viral dsRNA element of the chestnut blight fungus with a distinct genetic organization. Virology 201(2): 241-250.

Hong, Y, SL Dover, TE Cole, CM Brasier and KW Buck. 1999. Multiple mitochondrial viruses in an isolate of the Dutch Elm disease fungus Ophiostoma novo-ulmi. Virology 258(1): 118-127.

Janes, JK and MF Duretto. 2010. A new classification for subtribe Pterostylidinae (Orchidaceae), reaffirming Pterostylis in the broad sense. Aust. Syst. Bot. 23(4): 260- 269.

Janes, JK, DA Steane, RE Vaillancourt and MF Duretto. 2010. A molecular phylogeny of the subtribe Pterostylidinae (Orchidaceae): resolving the taxonomic confusion. Aust. Syst. Bot. 23(4): 248-259.

Jensen, DD. 1952. Virus diseases of orchids: symptoms, spread, host range, distribution, and control practices determined by experimental transmissions. Calif. Agric. 6(1): 3-15.

Kawakami, K, S Fuji and K Miyoshi. 2007. Endangered wild populations of endemic Calanthe orchids on an isolated Japanese island tested for viruses. Aust. J. Bot. 55(8): 831-836.

Kitahara, R, Y Ikeda, H Shimura, C Masuta and T Ezawa. 2014. A unique mitovirus from Glomeromycota, the phylum of arbuscular mycorrhizal fungi. Arch. Virol. 159(8): 2157-2160.

Koonin, EV. 1991. The phylogeny of RNA-dependent RNA polymerases of positive- strand RNA viruses. J. Gen. Virol. 72: 2197-2206.

Kozlakidis, Z, CV Hacker, D Bradley, A Jamal, X Phoon, J Webber, CM Brasier, KW Buck and RHA Coutts. 2009. Molecular characterisation of two novel double- stranded RNA elements from Phlebiopsis gigantea. Virus Genes 39(1): 132-136. 89

Linder-Basso, D, JN Dynek and BI Hillman. 2005. Genome analysis of Cryphonectria hypovirus 4, the most common hypovirus species in North America. Virology 337(1): 192-203.

Magae, Y. 2012. Molecular characterization of a novel mycovirus in the cultivated mushroom, Lentinula edodes. Virol. J. 9: 60.

Marchler-Bauer, A and SH Bryant. 2004. CD-Search: protein domain annotations on the fly. Nucleic Acids Res. 32(Web Server issue): W327-331.

Marchler-Bauer, A, C Zheng, F Chitsaz, MK Derbyshire, LY Geer, RC Geer, NR Gonzales, M Gwadz, DI Hurwitz, CJ Lanczycki, F Lu, S Lu, GH Marchler, JS Song, N Thanki, RA Yamashita, D Zhang and SH Bryant. 2013. CDD: conserved domains and protein three-dimensional structure. Nucleic Acids Res. 41(D1): D348-352.

Marzano, S-YL and LL Domier. 2016. Novel mycoviruses discovered from metatranscriptomics survey of soybean phyllosphere phytobiomes. Virus Res. 213: 332-342.

Nuss, DL and BI Hillman. 2012. Hypoviridae. Virus Taxonomy: Classification and Nomenclature of Viruses: Ninth Report of the International Committee on Taxonomy of Viruses. AMQ King, MJ Adams, EB Carstens and EJ Lefkowitz, 1029-1033. San Diego: Elsevier Academic Press.

Ohta, C, T Taguchi, S Takahashi, K Ohtsuka, K Eda, S Ayusawa and Y Magae. 2008. Detection of double stranded RNA elements in cultivated Lentinula edodes. Mushroom Science and Biotechnology 16: 155-158.

Ong, JWL, H Li, K Sivasithamparam, KW Dixon, MGK Jones and SJ Wylie. 2016b. Novel Endorna-like viruses, including three with two open reading frames, challenge the membership criteria and taxonomy of the Endornaviridae. Virology 499: 203-211.

Ong, JWL, RD Phillips, KW Dixon, MGK Jones and SJ Wylie. 2016a. Characterization of the first two viruses described from wild populations of hammer orchids (Drakaea spp.) in Australia. Plant Pathology 65(1): 163-172.

Ramsay, RR, KW Dixon and K Sivasithamparam. 1986. Patterns of infection and endophytes associated with Western Australian orchids. Lindleyana 1(3): 203-214.

Ramsay, RR, K Sivasithamparam and KW Dixon. 1987. Anastomosis groups among Rhizoctonia-like endophytic fungi in southwestern Australian Pterostylis species (Orchidaceae). Lindleyana 2: 161-166.

Roossinck, MJ. 2015. Metagenomics of plant and fungal viruses reveals an abundance of persistent lifestyles. Front Microbiol. 5: 767. 90

Seo, J-K, W-S Lim, J-H Jeong, Y-B Yoo, S-W Yie and K-H Kim. 2004. Characterization and RT-PCR detection of dsRNA mycoviruses from the oyster mushroom, Pleurotus ostreatus. Plant Pathol. J. 20(3): 200-205.

Shapira, R and DL Nuss. 1991. Gene expression by a hypovirulence-associated virus of the chestnut blight fungus involves two papain-like protease activities. J. Biol. Chem. 266(29): 19419-19425.

Sherpa, A, V Hallan, P Pathak and A Zaidi. 2006. Characterization of the coat protein gene of Cymbidium mosaic virus isolates from India. J. Phytopathol. 154(5): 275-280.

Singh, MK, AR Sherpa, V Hallan and AA Zaidi. 2007. A potyvirus in Cymbidium spp. in northern India. Australas. Plant Dis. Notes 2(1): 11-13.

Smart, CD, W Yuan, R Foglia, DL Nuss, DW Fulbright and BI Hillman. 1999. Cryphonectria hypovirus 3, a virus species in the family Hypoviridae with a single open reading frame. Virology 265(1): 66-73.

Swarts, ND and KW Dixon. 2009. Terrestrial orchid conservation in the age of extinction. Ann. Bot. 104(3): 543-556.

Vainio, EJ, R Hyder, G Aday, E Hansen, T Piri, T Doğmuş-Lehtijärvi, A Lehtijärvi, K Korhonen and J Hantula. 2012. Population structure of a novel putative mycovirus infecting the conifer root-rot fungus Heterobasidion annosum sensu lato. Virology 422(2): 366-376.

Wang, S, H Kondo, L Liu, L Guo and D Qiu. 2013. A novel virus in the family Hypoviridae from the plant pathogenic fungus Fusarium graminearum. Virus Res. 174(1): 69-77.

Warcup, J. 1973. Symbiotic germination of some Australian terrestrial orchids. New Phytol. 72(2): 387-392.

Wickner, RB, SA Ghabrial, ML Nibert, JL Patterson and CC Wang. 2012. Totiviridae. Virus taxonomy: classification and nomenclature of viruses: Ninth Report of the International Committee on Taxonomy of Viruses. AMQ King, MJ Adams, EB Carstens and EJ Lefkowitz, 639-650. San Diego: Elsevier Academic Press.

Won, H-K, S-J Park, D-K Kim, MJ Shin, N Kim, S-H Lee, Y-C Kwon, HK Ko, H-S Ro and H-S Lee. 2013. Isolation and characterization of a mycovirus in Lentinula edodes. J. Microbiol. 51(1): 118-122.

Wylie, SJ, H Li, KW Dixon, H Richards and MGK Jones. 2013a. Exotic and indigenous viruses infect wild populations and captive collections of temperate terrestrial orchids (Diuris species) in Australia. Virus Res. 171(1): 22-32. 91

Wylie, SJ, H Li and MGK Jones. 2013b. Donkey orchid symptomless virus: a viral ‘platypus’ from Australian terrestrial orchids. PLoS One 8(11): e79587.

Wylie, SJ, AJY Tan, H Li, KW Dixon and MGK Jones. 2012. Caladenia virus A, an unusual new member of the family Potyviridae from terrestrial orchids in Western Australia. Arch. Virol. 157(12): 2447-2452.

Xie, J, X Xiao, Y Fu, H Liu, J Cheng, SA Ghabrial, G Li and D Jiang. 2011. A novel mycovirus closely related to hypoviruses that infects the plant pathogenic fungus Sclerotinia sclerotiorum. Virology 418(1): 49-56.

Yaegashi, H, S Kanematsu and T Ito. 2012. Molecular characterization of a new hypovirus infecting a phytopathogenic fungus, Valsa ceratosperma. Virus Res. 165(2): 143-150.

Zettler, F, N Ko, G Wisler, M Elliott and S Wong. 1990. Viruses of orchids and their control. Plant Dis. 74(9): 621-626.

Zhang, R, S Hisano, A Tani, H Kondo, S Kanematsu and N Suzuki. 2016. A capsidless ssRNA virus hosted by an unrelated dsRNA virus. Nat. Microbiol. 1: 15001.

Zhong, J, C-Y Chen and B-D Gao. 2015. Genome sequence of a novel mycovirus of Rhizoctonia solani, a plant pathogenic fungus. Virus Genes 51(1): 167-170.

Supplementary information

Table S1. Viral sequences (<50% of estimated genome) identified in Pterostylis sanguinea orchids and associated mycorrhizal fungi. Estimated Virus host Sequence GenBank Name Best blastp match percentage of (Sample no.) length accession no. genome# Pterostylis P. sanguinea Sclerotinia sclerotiorum megabirnavirus 1 1910 22% of RNA-1 KU291967 megabirnavirus-like contig 1 (P-2012) (Megabirnaviridae, Megabirnavirus)

Pterostylis P. sanguinea Rhododendron virus A 1464 43% KU291968 amalgavirus-like contig 1 (P-2013) (Amalgaviridae, Amalgavirus) Pterostylis P. sanguinea Blueberry latent virus 474 14% KU291969 amalgavirus-like contig 2 (P-2013) (Amalgaviridae, Amalgavirus) Pterostylis P. sanguinea Black raspberry virus F 2306 45% KU291970 totivirus-like contig 1 (P-2013) (Totiviridae, Totivirus) Pterostylis P. sanguinea Black raspberry virus F 1985 39% KU291971 totivirus-like contig 2 (P-2013) (Totiviridae, Totivirus) Pterostylis P. sanguinea Black raspberry virus F 584 12% KU291972 totivirus-like contig 3 (P-2013) (Totiviridae, Totivirus) Pterostylis P. sanguinea Black raspberry virus F 553 11% KU291973 totivirus-like contig 4 (P-2013) (Totiviridae, Totivirus)

Ceratobasidium Ceratobasidium sp. Cryphonectria hypovirus 1 2425 19% KU291933 hypovirus-like contig 1 (F-2012) (Hypoviridae, Hypovirus) Ceratobasidium Ceratobasidium sp. Cryphonectria hypovirus 1 2204 17% KU291934 hypovirus-like contig 2 (F-2012) (Hypoviridae, Hypovirus)

Ceratobasidium Ceratobasidium sp. Fusarium graminearum hypovirus 1 2168 17% KU291935 hypovirus-like contig 3 (F-2012) (Hypoviridae, Hypovirus) Ceratobasidium Ceratobasidium sp. Fusarium graminearum hypovirus 1 1395 11% KU291936 hypovirus-like contig 4 (F-2012) (Hypoviridae, Hypovirus) Ceratobasidium Ceratobasidium sp. Cryphonectria hypovirus 2 1234 10% KU291937 hypovirus-like contig 5 (F-2012) (Hypoviridae, Hypovirus) Ceratobasidium Ceratobasidium sp. Rhizoctonia solani RNA virus HN008 3119 41% KU291939 mycovirus-like contig 1 (F-2012) (Unclassified) Ceratobasidium Ceratobasidium sp. Rhizoctonia solani RNA virus HN008 1618 21% KU291940 mycovirus-like contig 2 (F-2012) (Unclassified) Ceratobasidium Ceratobasidium sp. Rhizoctonia solani RNA virus HN008 1507 20% KU291941 mycovirus-like contig 3 (F-2012) (Unclassified) Ceratobasidium Ceratobasidium sp. Rhizoctonia solani RNA virus HN008 1474 19% KU291942 mycovirus-like contig 4 (F-2012) (Unclassified) Ceratobasidium Ceratobasidium sp. Fusarium graminearium dsRNA mycovirus 4 1300 75% of RNA-1 KU291950 mycovirus-like contig 5 (F-2012) (Unclassified) Ceratobasidium Ceratobasidium sp. Rhizoctonia solani RNA virus HN008 1199 16% KU291943 mycovirus-like contig 6 (F-2012) (Unclassified) Ceratobasidium Ceratobasidium sp. Rhizoctonia solani RNA virus HN008 1186 16% KU291944 mycovirus-like contig 7 (F-2012) (Unclassified) Ceratobasidium Ceratobasidium sp. Ustilaginoidea virens RNA virus 1159 22% KU291951 mycovirus-like contig 8 (F-2012) (Unclassified) Ceratobasidium Ceratobasidium sp. Lentinula edodes mycovirus HKB 670 6% KU291952 mycovirus-like contig 9 (F-2012) (Unclassified) Ceratobasidium Ceratobasidium sp. Pleosporales megabirnavirus 1 1115 13% of RNA-1 KU291945 megabirnavirus-like contig 1 (F-2012) (Megabirnaviridae, Megabirnavirus)

Ceratobasidium Ceratobasidium sp. Rosellinia necatrix megabirnavirus 2-W8 1058 12% of RNA-1 KU291946 megabirnavirus-like contig 2 (F-2012) (Megabirnaviridae, Megabirnavirus)

Ceratobasidium Ceratobasidium sp. Helicobasidium mompa endornavirus 1 1245 7% KU291929 endornavirus-like contig 1 (F-2013) (Endornaviridae, Endornavirus) Ceratobasidium Ceratobasidium sp. Rhizoctonia cerealis endornavirus 1 772 4% KU291930 endornavirus-like contig 2 (F-2013) (Endornaviridae, Endornavirus) Phaseolus vulgaris Ceratobasidium Ceratobasidium sp. 576 endornavirus 1 4% KU291931 endornavirus-like contig 3 (F-2013) (Endornaviridae, Endornavirus) Ceratobasidium Ceratobasidium sp. Helicobasidium mompa endornavirus 1 454 3% KU291932 endornavirus-like contig 4 (F-2013) (Endornaviridae, Endornavirus) Cryphonectria parasitica bipartitie dsRNA Ceratobasidium Ceratobasidium sp. 1418 mycovirus 1 70% of RNA-2 KU291953 mycovirus-like contig 10 (F-2013) (Unclassified) Ceratobasidium Ceratobasidium sp. Fusarium graminearium dsRNA mycovirus 4 1313 76% of RNA-1 KU291954 mycovirus-like contig 11 (F-2013) (Unclassified) Ceratobasidium Ceratobasidium sp. Phlebiopsis gigantea mycovirus dsRNA 1 759 9% KU291955 mycovirus-like contig 12 (F-2013) (Unclassified)

#Estimated genome size is based on size of closest virus match from blastp

Chapter 5: Novel Endor na-like viruses, including t hree with two ope n reading frames, challenge the taxonomy of the Endornaviridae Virology 499 (2016) 203–211

Contents lists available at ScienceDirect

Virology

journal homepage: www.elsevier.com/locate/yviro

Novel Endorna-like viruses, including three with two open reading frames, challenge the membership criteria and taxonomy of the Endornaviridae

Jamie W.L. Ong a, Hua Li a, Krishnapillai Sivasithamparam a, Kingsley W. b a a,n Dixon , Michael G.K. Jones , Stephen J. Wylie

a Plant Biotechnology Group – Plant Virology, Western Australian State Agricultural Biotechnology Centre, School of Veterinary and Life Sciences, Murdoch University, Perth, Western Australia 6150, Australia b School of Science, Curtin University, Bentley, Western Australia 6102, Australia

a r t i c l e i n f o a b s t r a c t

Article history: Viruses associated with wild orchids and their mycorrhizal fungi are poorly studied. Using a shotgun sequencing Received 7 June 2016 approach, we identiﬁed eight novel endornavirus-like genome sequences from isolates of Ceratobasidium fungi Returned to author for revisions isolated from pelotons within root cortical cells of wild indigenous orchid species Microtis media, Pterostylis 11 August 2016 sanguinea and an undetermined species of Pterostylis in Western Australia. They represent the ﬁrst Accepted 19 August 2016 endornaviruses to be described from orchid mycorrhizal fungi and from the Australian continent. Five of the Available online 24 September 2016 novel endornaviruses were detected from one Ceratobasidium isolate collected from one Pterostylis plant. The Keywords: partial and complete viral replicases shared low (9–30%) identities with one another and with endornaviruses Ceratobasidium described from elsewhere. Four had genome lengths greater than those of previously described endornaviruses, Endornavirus two resembled ascomycete-infecting endornaviruses, and unlike currently described endornaviruses, three had two Indigenous virus open reading frames. The unusual features of these new viruses challenge current taxonomic criteria for membership Orchid mycorrhizae Mycovirus of the family Endornaviridae. Virus taxonomy Wild plant virology Crown Copyright & 2016 Published by Elsevier Inc. All rights reserved.

1. Introduction Currently, there are seven fungus-infecting, nine plant-infecting and one oomycete-infecting endornaviruses described, of which 12 have been ratified by the International Committee on Endornavirus (family Endornaviridae) are non-encapsidated Taxonomy of Viruses (ICTV) (International Committee on viruses with double-stranded (ds) RNA genomes. The genomes Taxonomy of Viruses, 2015, 2016). Endornavirus clades Al- of described members range from 9 kb to 17.6 kb (Fukuhara phaendornavirus (Clade I) and Betaendornavirus (Clade II) and Gibbs, 2012), and there is always only one open reading are proposed (Khalifa and Pearson, 2014) but not ratified by the frame (ORF) encoding a replicase. Current species are distinguished ICTV. This classification reflects relationships of active domains on the basis of host, genome size and organization, and nucleotide within the ORF (Khalifa and Pearson, 2014), which can include sequence variations. The nucleotide identities of different two or more of the following: helicase (Hel), methyltransferase endornavirus species range from 30–75% identity (Fukuhara and (MTR), glucosyltransferase (GT) and RdRp (Roossinck et al., Gibbs, 2012). The first endornaviruses were described from broad 2011; Fukuhara and Gibbs, 2012). The number and combination bean (Vicia faba), where the occurrence of large dsRNAs was of domains differ between species, with only the RdRp linked to cytoplasmic male sterility (Grill and Garger,1981). common to all endornaviruses (Roossinck et al., 2011; Endornaviruses have since been identified from plants, e.g. rice Fukuhara and Gibbs, 2012). With the exception of Persea (Oryza sativa) (Moriyama et al., 1995) and capsicum (Capsicum americana endornavirus 1 (Villanueva et al., 2012), all annuum) (Valverde et al., 1990), fungi, e.g. Helicobasidium mompa endornaviruses also encode a helicase domain. Current members (Osaki et al., 2006) and Tuber aestivum (Stielow et al., 2011), of Alphaendornavirus have larger genomes ( > 13,000 bp) and and oomycetes – e.g. Phytophthora sp. (Hacker et al., 2005). include viruses from basidiomycetes, oomycetes, and plants.

Members of Betaendornavirus have smaller genomes that range from 9760 bp (TaEV) to 10,513 bp (SsEV1) that lack a GT n Corresponding author. domain, and they infect ascomycetes (Steilow et al., 2011; E-mail address: [email protected] (S.J. Wylie).

204 J.W.L. Ong et al. / Virology 499 (2016) 203–211

Khalifa and Pearson, 2014).

55"/

Members of the Endornaviridae persist in their hosts over 1

multiple generations (Roossinck, 2010; Roossinck et al., 2011). 49'

50'

5°

° Infection with the majority of endornaviruses does not appear to 1

3.848''

1 negatively inﬂuence the growth and development of the host 7",

1",

50'

(Grill and Garger, 1981; Pfeiffer, 1998; Ikeda et al., 2003; Osaki °

30.5

et al., 2006; Roossinck, 2015). There is no evidence to support 3.64

94'',

horizontal transmission of endornaviruses to other hosts; the lack of 4

32°

32°4'

2.5

movement protein indicates absence of ability to move from cell to

population

5" 3"/

38"/

cell (Roossinck et al., 2011; Fukuhara and Gibbs, 2012). In plants

32°4'

they rely on vertical transmission through infected pollen and ova plant

53.4

9.968"

2 36"/

2.4667" 54.22

3.848"

. (Valverde and Gutierrez, 2007; Okada et al., 2011, 2013). In fungal

49'

50'

49' °

50'

5 1 hosts, they transmit vertically via spores and horizontally via hyphal

5°50'

7",

anastomosis (Ikeda et al., 2003; Tuomivirta et al., 2009).

7",

5",

305",

Endornaviruses occur in all cells of studied hosts at copy numbers 7

4.0 7

29.432

54.5034", 1

2.5494",

55.70 2.334", of 20–100 genomes per cell (Fukuhara et al., 2006; Fukuhara and

3462"

°4' Gibbs, 2012). The cluster of basidiomycete-, oomycete- and plant- 7

32°4'

32°3' 32°4' 32

32°3' 32°4' 32°4'

Latitude/Longitude

52.

------

infecting endornaviruses within the alphaendornaviruses demonstrates that their evolution has involved horizontal transmission between host types, e.g. between fungi and plants (Gibbs et al., 2000; Roossinck et al., 2011; Khalifa and Pearson,

of of 2014), but how this occurred is unknown.

(

Terrestrial orchids rely on symbiotic associations with mycor-

no. rhizal fungi to provide nutrients and other molecules required for

)

a germination and growth. The fungi form pelotons in the cortex of the s

(1)

(1) (1) (5)

(1)

sample

root systems, which are digested by the orchids to acquire the 0

C C02 C03 C04 – C05

required nutrients (Swarts and Dixon, 2009; Smith and Read, ungal

F individual 2010). This process provides a possible route by which viruses are exchanged – either from fungus to plant or vice versa. Here, we

used a shotgun sequencing approach to identify endornaviruses from

- - - -

e e e e orchid leaves and from fungal cultures initiated from mycorrhizal t t t t

sola

isola i isola isola

fungal pelotons isolated from orchid root cells.

sp. sp. sp. sp.

sp.

fungus 2. Materials and methods

hizal

cor

Ceratobasidium 1 Ceratobasidium 2 Ceratobasidium 3 Ceratobasidium 4 – Ceratobasidium

2.1. Collection sites

y l

te Leaves and underground stem or root tissue was collected from a

separ

plants of the common mignonette orchid (Microtis media; 2 po-

pulations), an unidentiﬁed snail orchid (Pterostylis sp.; 3 t

alia.

pooled populations), and dark banded greenhood orchid (Pterostylis

not

ust

as sanguinea; 4 populations) from remnant native forest located on the

(No.

ern Murdoch University campus, Western Australia (W.A.) (Fig. S1, t

No.

)

Table 1). It is uncertain if the three populations of the snail orchid

a sample

Pterostylis (Pterostylis sp. isolates 1, 2 and 3) represented the

sample

Perth,

(

(5)

fungal

(20)

same genetic lineage because snail orchids exhibit variable

Plant individual

P – P02 ( P03 P04 morphology and interspecies hybridization is common (Brundrett, P05

campus,

2014).

ersity

populations;

population(s).

populations.

each

population

2.2. Fungus isolation from root pelotons enhood

doch

name r

hid hid hid

sp.

M. microtis

P. sanguinea P.

c c c

Mu or or or Each underground stem or root tissue sample (Fig. 1) was mignonette mignonette

two

four

the

pooled

surface-sterilized by immersion in 2% sodium hypochlorite banded

Common

Snail Snail Snail

chid chid chid

Pterostylis

r and solution, then in 70% ethanol for 10 s followed by washing in r r r

Common o Common o Da o

for

sterile water, before being ground in sterile water with a pestle. The

pooled

samples

resulting liquid mixture was viewed under a compound microscope sampled

sampled to identify fungal pelotons (undifferentiated hyphae; Fig. 1).

fungal

fungi sampled fungi

not

d Individual pelotons were transferred onto fungal isolation medium or

and

izal

- 1 - 1 es

samples

(FIM) agar plates (0.3 g L NaNO , 0.2 g L KH PO , 0.1 g L

t 3 2 4 a

population population population

plant

corr

- 1 - 1 - 1 r - 1 ia

erial

MgSO .7H O, 0.1 g L KCl, 0.1 g L yeast extract, 2.5 g L t

sp. sp. sp. 4 2 of

r -1 -1 ma and sucrose and 8 g L agar; 100 mg L filter sterilized streptomycin

and

species

umber

N Leaf Mixtu Leaf

1 2 3 sulfate i

(Clements and Ellyard, 1979). Plates were left to incubate in

a b c d

chids

Pterostylis Pterostylis Pterostylis med Microtis media Microtis sanguinea Pterostylis

able

the dark at 24 °C for 5–7 days. Growing mycelium was T O

Orch subcultured into liquid media (FIM minus agar) and left on a shaker 97 in the dark at 24 °C until 80–100 mg fungal biomass was obtained.

J.W.L. Ong et al. / Virology 499 (2016) 203–211 205

Fig. 1. Scanning electron micrographs of (A) cross-section of Microtis media root with pelotons in cells (arrows) and (B) enlarged showing single pelotons with hyphae within root cells.

2.3. dsRNA extraction, cDNA synthesis and amplification Extraction Kit (Qiagen). Sanger sequencing of both strands of amplicons was carried out on an Applied Biosystems 3730 DNA and RNA enriched for dsRNA was obtained from 80 48-capillary sequencer using BigDyes version 3.1 terminator to100 mg of mycelia or leaf tissue using a cellulose powder- mix (Applied Biosystems). Sequences were analyzed within the based extraction method (Morris and Dodds, 1979). Geneious v7.0.6 software package (Biomatters; Kearse et al., dsRNAs extracted from the fungal samples (C01–C05) were 2012) and subjected to Blastn (Altschul et al., 1990) searches of separated on 1% agarose gels (TAE buffer) at 40 V for three NCBI Genbank databases (http://blast.ncbi.nlm.nih.gov/) to hours to confirm that the derived viruses were not artefacts of identify matches. sequence assemblies. For cDNA synthesis, 4 mL of heat-denatured RNA was added to 2.5. Analysis of high-throughput sequencing data a reaction volume of 20 mL, comprising of 1X GoScript™ RT buffer (Promega), 3 mM MgCl2, 0.5 mM dNTPs, 0.5 mM of CLC genomic Workbench v6.5.1 (Qiagen) and Geneious random primer with a known 16-nucleotide (nt) adapter at the 5' v7.0.6 software packages were used to analyze high-throughput end, and 160 units of reverse transcriptase (GoScript™, Promega). sequencing data. De novo assembly was carried out on the 100 The reaction was carried out at 25 °C for 5 min, followed by nt paired reads to form contigs > 200 nt in length. The contigs incubation at 42 °C for 60 min to synthesis cDNA, followed by were subjected to both Blastn and Blastx (Altschul et al., 1990) incubation at 70 °C for15 min to inactivate the reverse analysis of GenBank databases to identify contigs with shared transcriptase. identity to known viruses. Domains within putative viral contigs PCR amplification was carried out in a 20 mL reaction were identified by identity with homologs from known volume, which consisted of 1X GoTaqs Green Master Mix endornaviruses, or by using the NCBI Conserved Domain (Promega), 1 mM individually tagged barcode primer (part of Database (CDD) (Marchler-Bauer and Bryant, 2004). Annotation which was complementary to the 16-nucleotide adapter sequence of of genomes was done using Geneious. ClustalW pairwise the random primer used for cDNA synthesis) and 2 uL of comparisons were carried out in Geneious v7.0.6 using models, synthesized cDNA. Different barcodes were used for each leaf IUB (nt) and BLOSUM (amino acid; aa). Settings of gap open and fungal sample, including those collected from the same plant. cost of 15 (nt) and 10 (aa) and gap extend cost of 6.66 (nt) and The reaction was carried out with an initial incubation at 95 °C 0.1 (aa) were used. for 3 min, followed by 35 cycles of 95 °C for 30 s, 60 °C for The amino acid sequences of putative viruses identified 30 s, and 72 °C for 1 min, followed by a final extension at 72 °C and were aligned with known reference virus sequences within for 10 min. MEGA v6.06 using Gonnet as the protein weight matrix. Gap Amplicons were purified using columns of a QIAquick PCR opening penalty of 10 was set for both pairwise and multiple Purification Kit (Qiagen), quantified, and pooled in approximately alignments with a gap extension penalty of 0.1 and 0.2 for equimolar amounts. 10 μg of pooled amplicons were submitted to pairwise and multiple alignments respectively. “Find best either the Australian Genome Research Facility (Melbourne, DNA/Protein models (Maximum likelihood, ML)” application Australia) or Macrogen Inc (Seoul, South Korea) for library within MEGA was used to determine the appropriate model for construction and high-throughput sequencing of paired ends construction of respective ML phylogenetic trees with 1000 over 100 cycles on the Illumina HiSeq2000 platform. bootstrap replications. Phylogenetic tree of the endornavirus polyproteins was constructed from 6848 aa using WAG with

Freqs model and gamma distribution of 2. Analysis of 2.4. Identiﬁcation of fungi using ITS sequences endornavirus domains were carried out with the following

settings – MTRs (no of sites: 425 aa; LG model with The internal transcribed spacer (ITS) regions of fungal isolates gamma distribution (LG + G): 6), GTs (438 aa; LG + G 5), were ampliﬁed using universal primers ITS1 (5' Hels(299 aa; LG + G: 2; had invariant sites) and RdRps (452 TCCGTAGGTGAACCTGCGG 3') and ITS4 (5' aa; LG + G: 2). Homologous GT domains (superfamily cl10013) TCCTCCGCTTATTGATATGC 3') (White et al., 1990). from hypoviruses and non-viral organisms such as bacteria, fungi Ampliﬁcation was carried out in a 20 uL reaction volume and plants were included in the Maximum likelihood analysis. containing 1X GoTaqs mastermix (Promega), 0.5 mM of each

primer, ITS1 and ITS4, and 60–80 ng of extracted DNA. Cy- 2.6. Sequence confirmation of ORF2 cling conditions were an initial denaturation step at 95 °C for3 min, followed by 35 cycles of 30 s at 95 °C, 1 min at 52 PCR and Sanger sequencing were used to confirm the presence °C and 1 min at 72 °C, and a final extension at 72 °C for 10 of open reading frame (ORF) 2 in endornaviruses. 98 min. PCR amplicons were purified using columns of a QIAquick Gel

206 J.W.L. Ong et al. / Virology 499 (2016) 203–211

Genome fragments surrounding the two stop codons corresponding endornaviruses. The sequences of CbEVE, CbEVF and CbEVH to the end of ORFs 1 and 2 were amplified using specific primers are incomplete, but based on the genome sizes of relatives, and sequenced using the Sanger method. probably represent >75% of their complete genomes. Fifteen short endornavirus-like genome fragments ranged from 499 nt to 3. Results 5221 nt. Even the largest of these probably represents less than half of a genome, so it is uncertain how many viruses are Two Illumina sequencing runs generated 92,046,118 and present. These short endornavirus-like sequences were 35,630,376 101-nt reads. A total of 15,048,256 reads were gener- designated ‘Endornavirus-like contig’ and given a number (Table ated from the sampled plants (P01–P05) and fungi (C01–C05), 2). so the raw sequences were a mixture of viral and host sequences. The remaining reads were from other barcoded samples not related to CbEVB, CbEVC and CbEVG and Endornavirus-like contig 9 (3524 bp) encode a second ORF, a feature not previously this project that were pooled for sequencing then filtered out. associated with endornaviruses. Translation start codons of These are not discussed here. After de novo assembly of ORF2 of CbEVC and Endornavirus-like contig 9 were in a contigs and Blast analysis, 19 endornavirus-like contigs were Kozak context, but those of CbEVB and CbEVG were not. A identified ranging in size from 499 bp to 23,625 bp (GenBank Kozak-like sequence was present downstream of the first apparent accessions KX355142–KX355164; Table 2). All endornavirus-like start codon of CbEVG at 15,336-15,342 nt, indicating that this sequences described were limited to fungi, although three short may be the actual site of translation initiation. CbEVB had an sequences (endornavirus-like contigs 9–11) were detected in a intergenic spacer of 711 nt between ORFs 1 and 2, while ORF2s mixed plant sample (P04) of leaf and root-associated fungal of CbEVC and CbEVG overlaps ORF1 by 4 nt and 92 nt, samples of M. microtis (Table 2). Fungal isolates from orchid roots respectively. Sequences of endornavirus-like contig 9 encoded were identified as members of the genus Ceratobasidium. In this a partial ORF2 at 2046–3524 nt and had an intergenic spacer of study, the amplified ITS regions of these fungal hosts were 134 nt between the two ORFs. The presence of this second ORF insufficient for identification at the species level. The ITS region in CbEVB, CbEVC and CbEVG was confirmed through Sanger was used because it is the most likely to successfully identify the sequencing of the regions surrounding the 3' end of ORF1 and the broadest range of fungi (Schoch et al., 2012) and is the most 5' end of ORF2. 100% nt identity was obtained between the commonly amplified region in fungal identification studies and in sequences obtained from Sanger and Illumina sequencing. the NCBI database. ITS nucleotide identities between isolates were Complete ORF2s were 1452 aa to 1857 aa (4359–5652 nt) in 91.4–94.7%, which are below the 97% species demarcation value length and are predicted to encode proteins ranging from 165 kDa used for fungi (Izzo et al., 2005; O’Brien et al., 2005), indicative to 215 kDa. The predicted proteins shared low identities (9– that each of the five isolates was potentially of a distinct species. 12% aa, 42–43% nt) with one another, and with the exception However, ClustalW alignment of the ITS primers-amplified region of endornavirus-like contig 9, did not match proteins above the between species from the same genus showed much lower identities. threshold of 10 listed on the NCBI protein database using all For example, C. cornigerum and C. cereale shared only 77% nt blast options (protein-protein blast, position-specific iterated identity while Sebacina vermifera and S. allantiodea shared only blast, pattern hit initiated blast and domain enhanced lookup 72% nt identity Thus, we are reluctant to label them as distinct time accelerated blast; http:// blast.ncbi.nlm.nih.gov/). species based solely on the amplified region of approximately 600 Pairwise sequence comparison of complete and partial bp. Instead, we labeled as Ceratobasidium isolates 1–4 (each ORF1s showed there was 9–30% aa (41–49% nt) identity between isolate from a single peloton of a different plant) (Table 1). the new viruses, and a similar level of identity (9–22% aa, 41– Fungal sample C05 was not given an isolate number as it 45% nt) between the new endornaviruses and previously represented a combination of four fungal isolates from four P. identified endornaviruses (Table S1). The majority of genomes sanguinea populations. shared only 10–15% aa identity, but higher nt identity (40–45%) Eight of the larger endornavirus-like sequences, ranging in size (Table S1). New endornaviruses from the same fungal isolate did from 7367 bp to 23,625 bp were estimated to represent 50% or not share greater sequence identities than those from different more of the virus genome, as based on genome sizes of the closest fungal isolates or from different collection sites. All five known relative. Each genome sequence represents a distinct virus, viruses identified from fungal isolate C02 were genetically which were designated Ceratobasidium endornaviruses A-H closer to viruses from mycorrhizal fungi associated with other (CbEVA-H). Five of the proposed new endornaviruses (CbEVB- orchid populations than they were with one another (Fig. 2, Table F) were identified from Ceratobasidium isolate C02 from a S1). For example, CbEVD from fungal isolate C02 associated Pterostylis plant (Table 2). The three remaining endornaviruses with Pterostylis sp. population P02 shared 30% aa (49% nt) (CbEVA, CbEVG and CbEVH) were identified from three sequence identity with CbEVH from fungal isolate C04 associated different Ceratobasidium isolates, each from a different population with M. microtis population P03, but less than 14% aa ( < 43% of Pterostylis or Microtis. Presence of these endornaviruses nt) identity with co-infecting endornaviruses (Fig. 2; Table S1). was confirmed through detection of dsRNAs on agarose gel, which showed bands of size between 10,000 bp and 20,000 bp (Fig. S2). Analysis using the CDD indicated presence of four 3.1. Taxonomy protein domains, in different combinations, encoded by the endornavirus-like sequences – MTR (cl03298), Hel (pfam01443 The genome sequences of the new viruses were closer to the and smart00487), GT (cl10013) and RdRp (cl03049) (Table 2). genomes of endornaviruses described from fungi (9–22% aa, These domains shared the same motifs and belonged to the same 41–45% nt), oomycetes (9–14% aa, 42–43% nt) and plants superfamilies as other known endornaviruses (Roossinck et al., (9–15% aa, 41–45% nt) (Tables 3 and S1) than to any other 2011). known viruses. Species demarcation within Endornaviridae is The genomes of CbEVA, CbEVB, CbEVC, CbEVD and CbEVG dependent on both host range and sequence differentiation each consisted of one complete ORF, as determined by the (Fukuhara and Gibbs, 2012). The 41–49% nt sequence identity presence of stop codons before and after the ORF, and presence of 5' between these new viruses and those previously described (Table UTRs. The context of the proposed start codons of these five S1) fits within the broad species demarcation range of 30–75% endornaviruses is Kozak-like (RxxAUGR, where R represents either nt between genomes of known endornaviruses (Fukuhara and A or G and x is any base; Kozak, 1986), as seen in the Gibbs, 2012). Based on the differences in genome majority of known organization, sequence phylogeny and hosts, we propose the 99 new endornavirus genome sequences

Table 2 Molecular characteristics and blastp analysis of endornavirus-like genomes and genome fragments. Blast analyses were limited to genomes of >2000 bp. Open reading frames and locations of conserved domains are indicated. Bolded virus names represent probable complete genomes.

Proposed virus name Virus host Associated Length (nt) Blastp match [size nt; endornavirus Accession no. [e- % coverage, % 5' UTR Location of domains from ORF1 3' UTR [Genbank accession orchid [coding re- group] value] identity of a no.] species gion (s) ] nearest b b b b match MTR Hel GT RdRp

Ceratobasidium sp. Ceratobasidium en- Pterostylis sp. 15,207 Bell pepper endornavirus YP_004765011 26%, 34% 184 – nt 3839- – nt 14018- 116 dornavirus A isolate-1 (C01) 4599 14722 [KX355142] (P01) [14,907] [14,728; Alphaendornavirus] [3e–66] aa 1217-1472 aa 4612- 4846 Ceratobasidium en- Ceratobasidium sp. Pterostylis sp. 23,625 Helicobasidium mompa endornavirus YP_003280846 [0] 74%, 26% 13 – nt 6134-6922 nt 10394- nt 16040- 24 dornavirus B isolate-2 (C02) population 2 [17,235, 5652] 1 [16,614; Alphaendornavirus] aa 2041- 11554 aa 16846 aa [KX355143] 2303 3461- 5343-5611 3847 Ceratobasidium en- Ceratobasidium sp. Pterostylis sp. 21,004 Rhizoctonia cerealis endornavirus 1 YP_008719905 28%, 41% 265 – nt 4703-5488 nt 8837- nt 15215- 70

dornavirus C isolate-2 (C02) population 2 [16,224, 4449] [17,486; Alphaendornavirus] [2e–92] aa 1480-1741 9823 aa 16024 aa W

[KX355164] .L. 2858- 4984-5253

3186 Ong Ceratobasidium en- Ceratobasidium sp. Pterostylis sp. 19,406 Rhizoctonia cerealis endornavirus 1 YP_008719905 49%, 36% 262 nt 893- nt 4139-4876 – nt 18431- 64 dornavirus D isolate-2 (C02) population 2 [19,080] [17,486; Alphaendornavirus] [1e–180] 1312 aa aa 1293-1538 et 18868 aa [KX355144] 211-350 al.

6057-6202

Ceratobasidium en- Ceratobasidium sp. Pterostylis sp. 9271 [9271] Tuber aestivum endornavirus [9760; YP_004123950 48%, 27% – nt Hel1: nt – – – Vi

dornavirus E isolate-2 (C02) population 2 Betaendornavirus] [6e–105] r 1385- 5543-5943 ology [KX355145] 2119 aa aa 1848-1984

462- Hel2: nt 499 706 7544-8272

aa 2105-2757 (2

Ceratobasidium en- Ceratobasidium sp. Pterostylis sp. 7367 [7321] Tuber aestivum endornavirus [9760; YP_004123950 29,%, 31% – – – – nt 5921- 48 1 6) dornavirus F isolate-2 (C02) population 2 Betaendornavirus] [2e–75] 6922 aa [KX355146] 20

1958-2291 3

– Ceratobasidium en- Ceratobasidium sp. Pterostylis sp. 19,293 Helicobasidium mompa endornavirus YP_003280846 38%, 35% 271 – nt 3653-4441 – nt 13643- 37 2

dornavirus G isolate-3 (C03) (P02) [14,718, 4359] 1 [16,614; Alphaendornavirus] [2e–103] aa 1128-1390 14506 aa 1 [KX355147] 4458-4745 Endornavirus-like Ceratobasidium sp. Pterostylis sp. 5221 [5221] Rhizoctonia solani endornavirus 2 AMM45288 [0] 88%, 34% – – nt 3016-3771 – – – contig 1 [KX355149] isolate-3 (C03) (P02) [15,850; Alphaendornavirus] aa 1006-1257 Endornavirus-like Ceratobasidium sp. Pterostylis sp. 2887 [2851] Rhizoctonia solani endornavirus 2 AMM45288 98%, 38% – – – – nt 1895- 36 contig 2 [KX355150] isolate-3 (C03) (P02) [15,850; Alphaendornavirus] [1e–176] 2398 aa 632-799 Endornavirus-like Ceratobasidium sp. Pterostylis sp. 2037 [2037] Gremmeniella abietina type B RNA YP_529670 [1e–10] 80%, 23% – – nt 970-1380 – – – contig 3 [KX355151] isolate-3 (C03) (P02) aa 324-460 virus XL1 [10,375; Betaendornavirus] Endornavirus-like Ceratobasidium sp. Pterostylis sp. 1693 [1693] – – – – – – – – – contig 4 [KX355152] isolate-3 (C03) (P02) Endornavirus-like Ceratobasidium sp. Pterostylis sp. 685 [685] – – – – nt 67- – – – – contig 5 [KX355153] isolate-3 (C03) (P02) 600 aa 23-200 Endornavirus-like Ceratobasidium sp. Pterostylis sp. 619 [416] – – – 203 – – – – – contig 6 [KX355154] isolate-3 (C03) (P02) Endornavirus-like Ceratobasidium sp. Pterostylis sp. 499 [499] – – – – – – – – – contig 7 [KX355155] isolate-3 (C03) (P02) Ceratobasidium en- Ceratobasidium sp. Microtis media 14,266 Rhizoctonia cerealis endornavirus 1 YP_008719905 47%, 37% – – nt 3-557 aa – nt 13,227- 38 dornavirus H isolate-4 (C04) (P03) [14,228] [17,486; Alphaendornavirus] [2e–178] 1-185 13754 aa [KX355148] 4409-4584 Endornavirus-like Ceratobasidium sp. Microtis media 4947 [4947] Rhizoctonia cerealis endornavirus 1 YP_008719905 83%, 34% – nt – – – – contig 8 [KX355156] isolate-4 (C04) (P03) [17,486; Alphaendornavirus] [2e–121] 1585- 100 1920 aa 529- 207

640

Table 2 (continued ) 208

Proposed virus name Virus host Associated Length (nt) Blastp match [size nt; endornavirus Accession no. [e- % coverage, % 5' UTR Location of domains from ORF1 3' UTR [Genbank accession orchid [coding re- group] value] identity of a b b b b no.] species gion (s) ] nearest MTR Hel GT RdRp match

Endornavirus-like P04 Microtis media 3524 ORF1: Rhizoctonia cerealis en- YP_008719905 67%, 40% – – – – nt 895- – c contig 9 [KX355157] (P04) [1479,1911] dornavirus 1 [17,486; Alphaendorna- [1e–90] 17%,32% 1467 aa virus] ORF2: leukocyte im- XP_010330262 [1.4] 299-489 munoglobulin-like receptor sub- family A member 4 (Saimiri boliviensis boliviensis) Endornavirus-like P04 Microtis media 2493 [2493] Rhizoctonia solani endornavirus AHL25285 [9e–16] 34%, 32% – – nt 616-819 aa – – – c

contig 10 (P04) RS006-2 – partial [1946] 206-273 J.

W [KX355158] .L.

Endornavirus-like P04 Microtis media 1559 [1559] – – – – – – – – –

Ong c contig 11 (P04)

[KX355159] et

Endornavirus-like Ceratobasidium sp. Pterostylis 827 [827] – – – – – – – – – al.

contig 12 (C05) sanguinea Vi

[KX355160] (P05) r ology Endornavirus-like Ceratobasidium sp. Pterostylis 634 [634] – – – – – – – – – d contig 13 (C05) sanguinea

499 [KX355161] (P05)

Endornavirus-like Ceratobasidium sp. Pterostylis 602 [602] – – – – – – – – – (2

d 0 contig 14 (C05) sanguinea 1 6)

[KX355162] (P05) 20 Endornavirus-like Ceratobasidium sp. Pterostylis 571 [571] – – – – – – – nt 37-540 –

d –

contig 15 (C05) sanguinea aa 13-180 2

1 [KX355163] (P05) 1

a Endornaviruses with two ORFs (B, C and G). b Polyprotein domains: MTR (Methyltransferase), Hel (Helicase), GT (Glycosyltransferase) and RdRp (RNA-dependent RNA polymerase). c A mixture of fungal and plant materials from two populations. d A mixture of fungal materials from four populations.

101

J.W.L. Ong et al. / Virology 499 (2016) 203–211 209

Fig. 2. Maximum likelihood trees of (A) aa sequences of complete and partial polyprotein, (B) methyltranferase, (C) helicase, (D) Glycosyltransferase and (E) RNA dependent RNA polymerase domains of proposed Ceratobasidium endornaviruses (CbEVA to CbEVH) compared to related proteins of other viruses and non-viral organisms. Trees were constructed with 1000 bootstrap replications and statistical confidence values of <60% were omitted. Symbols represent endornaviruses infecting mycorrhizal fungi (Ceratobasidium spp.) – ∙ (C01), ■ (C02), ▲ (C03), ◆ (C04) associated with different orchid populations. Clades I (alphaendornavirus) and II (betaendornavirus) represent the two recognized clades within Endornaviridae based on host types - (I) basidiomycetes, oomycete and plants, (II) ascomycetes. Ampeloviruses (family Closteroviridae) PMWaV-1 and PBNSPaV were used as outgroups for complete polyproteins. The appropriate homologous domain of GLRaV1 was used as the outgroup for endornavirus MTR, Hel and RdRp domains. Abbreviations used for viruses: Bell pepper endornavirus (BPEV), Barley stripe mosaic virus (BMSV), Beet yellows virus (BYV), Cryphonectria hypovirus 3 (CHV3), Cryphonectria hypovirus 4 (CHV4), Grapevine leafroll associated virus 1 (GLRaV1), Gremmeniella abietina type B RNA virus XL (GABrV-XL), Helicobasidium mompa endornavirus 1 (HmEV1), Phaseolus vulgaris endornavirus (PvEV1), Phytophthora endornavirus 1 (PEV1), Pineapple mealybug wilt-associated virus 1 (PMWaV-1), Plum bark necrosis and stem pitting-associated virus (PBNSPaV), Oryza rufipogon endornavirus (OrEV), Oryza sativa endornavirus (OsEV), Rhizoctonia cerealis endornavirus 1 (RcEV1), Tobacco mosaic virus (TMV), Tuber aestivum endornavirus (TaEV) and Vicia faba endornavirus (VfEV). Bacteria, fungi and plants: Chlorophytum borivilianum (C. borivilianum), Lechevalieria aerocolonigenes (L. aerocolonigenes), Saccharomyces cerevisiae (S. cerevisiae), Streptomyces viridochromogenes (S. viridochromogenes), Tulasnella calospora (T. calospora). described here represent eight distinct species of endornavirus. from orchid leaves. The fungal hosts represent isolates of a distinct The new endornaviruses were closest to other fungus-derived species of the basidiomycete Ceratobasidium. The new viruses endornaviruses, GABrV-XL, HmEV-1, RcEV1 and TaEV (Fig. 2). are the first endornaviruses isolated from orchid mycorrhizal The proposal of two subgroups within Endornavirus was based fungi and the first of the family Endornaviridae identified on length of the genome, phylogeny of the RdRp from the continent of Australia. Other endornaviruses were (superfamily cl03049), and host type (Khalifa and Pearson, 2014). reported from Ceratobasidium anamorphs – Rhizoctonia spp. Phylogeny placed six of the new viruses (CbEVA, CbEVB, (Das et al., 2014; Li et al., 2014). CbEVC, CbEVD, CbEVG and CbEVH) within Clade I. This classification was supported by the phylogenies of individual 4.1. Challenges to criteria and taxonomy of the replicase domains – MTR, GT and/or Hel (Fig. 2). CbEVE was Endornaviridae excluded from phylogenetic analysis because its sequence lacked the conserved RdRp domain, but its domains MTR and Hel placed Many of the Ceratobasidium endornaviruses identified it with members of Clade II (Fig. 2). With the exception of CbEVE challenge the currently accepted criteria for membership of the and CbEVF (partial genomes), the genome sizes of the large En- dornaviridae and its proposed subgroups. CbEVB, Ceratobasidium endornaviruses ( > 14,266 bp) also placed them CbEVC and CbEVG represent the first endornaviruses reported to with members of Clade I whose genomes are all greater than encode two ORFs. RT-PCR and Sanger sequencing of the 13,000 bp. The placement of basidiomycete-infecting CbEVE and regions surrounding the 3' end of ORF1 and the 5' end of ORF2, CbEVF in Clade II with ascomycete-infecting viruses challenges a including the intergenic regions confirmed that the two ORFs exist justification for the formation of clades within Endornavirus and are not artefacts of high-throughput sequencing or software based on the Ascomycete/Basidio-mycete host division. Based assembly. Despite the difference in genome organization, we on complete polyprotein sequences, the three endornaviruses propose that the new virus sequences be tentatively assigned to CbEVB, CbEVC and CbEVG that possess two ORFs were placed genus Endornavirus, family Endornaviridae because they share with HmEV-1 in a clade closest to, but separate from, many of the same genome characteristics as members of the endornaviruses in Clade II. family such as having one large polyprotein encoded by ORF1 and encoding domains belonging to the same superfamilies of other members of the family. 4. Discussion Another challenge to the proposed taxonomic subgroup classification within Endornaviridae (Khalifa and Pearson, 2014) is A study of viruses associated with the symbiotic relationship of grouping basidiomycete-infecting endornaviruses, CbEVE and wild orchid plant and mycorrhizal fungus revealed eight new en- CbEVF, in Clade II with ascomycetes-derived dornaviruses from fungal pelotons isolated within the roots of endornaviruses (Fig. 2). Clues to the role of ORF2 proteins could wild terrestrial orchid plants. No endornaviruses were identified not be determined by similarity with known protein motifs. The origin of the three ORF2s remains unclear but their lack of identity with one another 102

210 J.W.L. Ong et al. / Virology 499 (2016) 203–211 indicates they may relate to specific functions in the host. The Mycoreovirus) was observed after co-infection with Cryphonectria presence of these new endornavirus genes implies that the full hypovirus 1-EP713 (CHV1-EP713; Hypovirus), while replication genetic diversity of endornavirus genomes remains to be de- and transmission of CHV1-EP713 remained unaffected in co-in- scribed. In addition, ten of the eleven endornaviruses currently fection with MyRV1-Cp9B21 (Sun et al., 2006). Plant classified in Clade I have a GT domain, but only two of the potyviruses are known to increase the fitness of co-infecting non- proposed six new Clade I endornaviruses, CbEVB and CbEVC, have potyviruses (e.g. Potato virus X; Vance, 1991), while their own this domain (Fig. 2D). The lack of GT domain is more consistent fitness remains unaffected (Pruss et al., 1997; Wang et al., 2009), with members of Clade II. probably because the helper component-protease (HC-Pro) of Based on the phylogenies of the polyprotein and domains, we potyviruses suppresses host-encoded RNA interference, enabling suggest a modification to the host range of the two sub-groups other viruses to replicate to higher levels (Wang et al., 2009; Lim proposed by Khalifa and Pearson (2014). Clade I (Alphaendornavirus) et al., 2011). will remain as proposed and consist of endornaviruses derived from Glycosyltransferase (GT) domains are uncommon in plant and basidiomycetes, oomycete and plants. Clade II (Betapartitivirus) will fungal viruses, and have been identified only in some endornaviruses be updated from ascomycete-derived endornaviruses to also include and hypoviruses (Hypoviridae) (Smart et al., 1999; Linder-Basso et basidiomycete-derived endornaviruses. al., 2005; Roossinck et al., 2011). It has been suggested that the viral GT domain was acquired from hosts during evolution, possibly prior to separation of kingdoms, to allow endornaviruses to protect them- 4.2. Diversity of Australian endornaviruses selves against host cellular enzymes by enforcing the membrane Endornaviruses isolated from one fungal isolate were less ge- surrounding their capsid-less dsRNAs (Markine-Goriaynoff et al., netically similar to one another than endornaviruses isolated from a 2004; Hacker et al., 2005; Roossinck et al., 2011; Chen and different fungal isolate, from different orchid species, and from Punja,2014). The GT domain of Phytophthora endornavirus 1 shares different collection sites (Fig. 2; Table S1). We observed the similar identity with homologous domains in Phytophthora species, same pattern with partitiviruses that multiply-infected in bacteria, and in fungi and plants. If the virus acquired its GT Ceratobasidium mycorrhizal fungi associated with P. vittata plants domain from its Phytophthora host, they would share greater identity (unpublished). Partitiviruses from the same host were more than with homologous domains from distantly related hosts like genetically more divergent than those infecting other hosts. bacteria, fungi and plants. Instead, it may have come from a source Many of the endornaviruses described here were genetically predating the separation of kingdoms (Hacker et al., 2005). In the closer to viruses described from other hosts in other continents only other report of co-infecting endornaviruses (PvEV1 and than to viruses inhabiting the same fungal host. The rate of PvEV2), both viruses encode a GT domain (Okada et al., 2013). genetic change of endornaviruses is not known, so estimations of However, with the multiple endornaviruses co-infecting the time separating populations of endornaviruses existing in Ceratobasidium C02, the GT domain was present in some but not in Australia with those from other continents cannot be determined. others (CbEVB, CbEVC ( + GT) and CbEVD (-GT)). No GT domain Based on their association with Australian indigenous plants was detected in CbEVE and CbEVF but their genomes were too growing in situ, and their distinct genetic features, it seems likely incomplete to conclusively determine the presence or absence of a GT that these viruses share an evolutionary history with the domain. With the exception of VfEV, the only accepted Australian fungal flora. However, the existence of related viruses endornaviruses without a GT domain are those in Clade II, the from other hosts on other continents indicates that the group is ascomycete-derived endornaviruses. This suggests a possible link naturally mobile, probably traveling with wind-borne fungal spores. between lack of requirement for GT with host type (ascomycetes) and/or genome size ( > 11,000 bp). However, similar to VfEV (plant), 4.3. Co-infection of endornaviruses four of the Ceratobasidium endornaviruses (A, D, G and H) classified in Clade I contradict this link by not having GT domain Co-infection of endornaviruses has been previously despite deriving from basidiomycetes and having larger genome size ( > reported only for common bean (Phaseolus vulgaris), where 14,000 bp). The lack of GT domain in some endornaviruses sug- Phaseolus vulgaris endornaviruses 1 and 2 (PvEV1 and PvEV2) gests that the domain is not always essential but if GT is indeed part of can co-occur (Okada et al., 2013; Khankhum et al., 2015). In the viral defensive mechanism, perhaps in the event of co-in- contrast, Oryza rufipogon virus (OrEV) and Oryza sativa virus fecting multiple endornaviruses, the role of GT may be shared (OsEV), related endornaviruses of Oryza (rice) species which share amongst the viruses. 80% aa sequence identity could not be made to co-infect a single host (Moriyama et al., 1999). In the current study, co-infection by five distinct endornaviruses (CbEVB-F) occurred in Ceratobasidium Acknowledgments isolate C02, and partial virus genomes suggestive of more than one endornavirus occurred in isolates C03 and C04. Co-infection of This study was funded by Australian Research Council endornaviruses may be reliant on their compatibility – presumably Linkage Grant LP110200180 in collaboration with Botanic a function of their ability to tolerate competition for the same cellular Gardens and Parks Authority and Australian Orchid Foundation. resources, and/or by utilising different cellular resources. Perhaps the low sequence identities of PvEV1 and PvEV2 related to differences in function, and so they are able to maintain stable Appendix A. Supporting information co-infection (Okada et al., 2013). It is possible that Supplementary data associated with this article can be found in endornaviruses have co-evolved to co-infect single hosts by the online version at http://dx.doi.org/10.1016/j.virol.2016.08.019. occupying different roles.

Co-infection by two or more viruses may lead to synergistic

(enhances fitness of the viruses) or antagonistic (presence of one References virus lowers the fitness of others) interactions (Syller, 2012; Syller and Grupa, 2016), or more complex interactions that fall between Altschul, S.F., Gish, W., Miller, W., Myers, E.W., Lipman, D.J., 1990. Basic local these. In an infection of chestnut blight fungus, alignment search tool. J. Mol. Biol. 216, 403–410. Brundrett, M.C., 2014. Identification and Ecology of Southwest Australian Orchids, Cryphonectria parasitica, increased replication (2-fold increase) and Perth, Western Australian Naturalists’ Club. transmission (6-fold increase) of Mycoreovirus 1-Cp9B21 (MyRV1-Cp9B21; 103

J.W.L. Ong et al. / Virology 499 (2016) 203–211 211

Clements, M.A., Ellyard, R.K., 1979. The symbiotic germination of Australian ter- community analysis by large-scale sequencing of environmental samples. Appl. restrial orchids. Am. Orchid Soc. Bull. 48, 810–816. Environ. Microbiol. 71, 5544–5550. Das, S., Falloon, R.E., Stewart, A., Pitman, A.R., 2014. Molecular characterisation of an Okada, R., Kiyota, E., Sabanadzovic, S., Moriyama, H., Fukuhara, T., Saha, P., Roos- endornavirus from Rhizoctonia solani AG-3PT infecting potato. Fungal Biol 118, sinck, M.J., Severin, A., Valverde, R.A., 2011. Bell pepper endornavirus: mole- 924–934. cular and biological properties, and occurrence in the genus Capsicum. J. Gen. Fukuhara, T., Gibbs, M.J., 2012. Endornaviridae. In: King, A.M.Q., Adams, M.J., Virol. 92, 2664–2673. Carstens, E.B., Lefkowitz, E.J. (Eds.), In Virus Taxonomy: Classification and No- Okada, R., Yong, C.K., Valverde, R.A., Sabanadzovic, S., Aoki, N., Hotate, S., Kiyota, menclature of Viruses: Ninth Report of the International Committee on Taxonomy E., Moriyama, H., Fukuhara, T., 2013. Molecular characterization of two evolutio- of Viruses. Elsevier Academic Press, San Diego, pp. 519–521. narily distinct endornaviruses co-infecting common bean (Phaseolus vulgaris). J. Fukuhara, T., Koga, R., Aoki, N., Yuki, C., Yamamoto, N., Oyama, N., Udagawa, T., Gen. Virol. 94, 220–229. Horiuchi, H., Miyazaki, S., Higashi, Y., 2006. The wide distribution of en- Osaki, H., Nakamura, H., Sasaki, A., Matsumoto, N., Yoshida, K., 2006. An endornaviruses, large double-stranded RNA replicons with plasmid-like properties. dornavirus from a hypovirulent strain of the violet root rot fungus Helicobasi- Arch. Virol. 151, 995–1002. dium mompa. Virus Res. 118, 143–149. Gibbs, M.J., Koga, R., Moriyama, H., Pfeiffer, P., Fukuhara, T., 2000. Phylogenetic Pfeiffer, P., 1998. Nucleotide sequence, genetic organization and expression strategy analysis of some large double-stranded RNA replicons from plants suggests they of the double-stranded RNA associated with the '447' cytoplasmic male sterility evolved from a defective single-stranded RNA virus. J. Gen. Virol. 81, trait in Vicia faba. J. Gen. Virol. 79, 2349–2358. 227–233. Pruss, G., Ge, X., Shi, X.M., Carrington, J.C., Vance, V.B., 1997. Plant viral Grill, L.K., Garger, S.J., 1981. Identification and characterization of double-stranded synergism: the potyviral genome encodes a broad-range pathogenicity enhancer RNA associated with cytoplasmic male sterility in Vicia faba. Proc. Natl. Acad. Sci. that transactivates replication of heterologous viruses. Plant Cell 9, 859–868. USA 78, 7043–7046. Roossinck, M.J., 2010. Lifestyles of plant viruses. Philos. Trans. R. Soc. Lond. B Hacker, C.V., Brasier, C.M., Buck, K.W., 2005. A double-stranded RNA from a Phy- Biol. Sci. 365, 1899–1905. tophthora species is related to the plant endornaviruses and contains a putative Roossinck, M.J., 2015. Metagenomics of plant and fungal viruses reveals an abun- UDP glycosyltransferase gene. J. Gen. Virol. 86, 1561–1570. dance of persistent lifestyles. Front. Microbiol. 5, 767. Ikeda, K.-i, Nakamura, H., Matsumoto, N., 2003. Hypovirulent strain of the violet Roossinck, M.J., Sabanadzovic, S., Okada, R., Valverde, R.A., 2011. The remarkable root rot fungus Helicobasidium mompa. J. Gen. Plant Pathol. 69, 385–390. evolutionary history of endornaviruses. J. Gen. Virol. 92, 2674–2678. Schoch, C.L., Seifert, K.A., Huhndorf, S., Robert, V., Spouge, J.L., Levesque, C.A., International Committee on Taxonomy of Viruses, 2015. ICTV Master Species List Chen, W., Consortium, Fungal Barcoding, 2012. Nuclear ribosomal internal 2014, v4 (https://talk.ictvonline.org/files/master-species-lists/). transcribed spacer (ITS) region as a universal DNA barcode marker for fungi. International Committee on Taxonomy of Viruses. 2016. Taxonomic proposals for PNAS 109, 6241–6246. ratification. (https://talk.ictvonline.org/files/ratification/m/proposals-for- Smart, C.D., Yuan, W., Foglia, R., Nuss, D.L., Fulbright, D.W., Hillman, B.I., 1999. ratification). Cryphonectria hypovirus 3, a virus species in the family Hypoviridae with a Izzo, A., Agbowo, J., Bruns, T.D., 2005. Detection of plot-level changes in ectomy- single open reading frame. Virology 265, 66–73. corrhizal communities across years in an old-growth mixed-conifer forest. New Phytol. Smith, S.E., Read, D.J., 2010. Mycorrhizal Symbiosis, 3rd ed. Academic Press, London. 166, 619–630. Stielow, B., Klenk, H.-P., Menzel, W., 2011. Complete genome sequence of the first Kearse, M., Moir, R., Wilson, A., Stones-Havas, S., Cheung, M., Sturrock, S., Buxton, S., endornavirus from the ascocarp of the ectomycorrhizal fungus Tuber aestivum Cooper, A., Markowitz, S., Duran, C., 2012. Geneious basic: an integrated and Vittad. Arch. Virol. 156, 343–345. extendable desktop software platform for the organization and analysis of sequence Sun, L., Nuss, D.L., Suzuki, N., 2006. Synergism between a mycoreovirus and a hy- data. Bioinformatics 28, 1647–1649. povirus mediated by the papain-like protease p29 of the prototypic hypovirus Khalifa, M.E., Pearson, M.N., 2014. Molecular characterisation of an endornavirus CHV1-EP713. J. Gen. Virol. 87, 3703–3714. infecting the phytopathogen Sclerotinia sclerotiorum. Virus Res. 189, 303–309. Swarts, N.D., Dixon, K.W., 2009. Terrestrial orchid conservation in the age of ex- Khankhum, S., Valverde, R.A., Pastor-Corrales, M.A., Osorno, J.M., Sabanadzovic, S., tinction. Ann. Bot. 104, 543–556. 2015. Two endornaviruses show differential infection patterns between gene pools of Syller, J., 2012. Facilitative and antagonistic interactions between plant viruses in Phaseolus vulgaris. Arch. Virol. 160, 1131–1137. mixed infections. Mol. Plant. Pathol. 13, 204–216. Kozak, M., 1986. Point mutations define a sequence flanking the AUG initiator co- Syller, J., Grupa, A., 2016. Antagonistic within-host interactions between plant viruses: don that modulates translation by eukaryotic ribosomes. Cell 44, 283–292. molecular basis and impact on viral and host fitness. Mol. Plant. Pathol. Li, W., Zhang, T., Sun, H., Deng, Y., Zhang, A., Chen, H., Wang, K., 2014. 17, 769–782. Complete genome sequence of a novel endornavirus in the wheat sharp eyespot Tuomivirta, T.T., Kaitera, J., Hantula, J., 2009. A novel putative virus of Gremmeniella pathogen Rhizoctonia cerealis. Arch. Virol. 159, 1213–1216. abietina type B (Ascomycota: Helotiaceae) has a composite genome with en- Lim, H.-S., Jang, C.-Y., Bae, H.-H., Kim, J.-K., Lee, C.-H., Hong, J.-S., Ju, H.-J., dornavirus affinities. J. Gen. Virol. 90, 2299–2305. Kim, H.-G., Domier, L.L., 2011. Soybean mosaic virus infection and helper Valverde, R.A., Gutierrez, D.L., 2007. Transmission of a dsRNA in bell pepper and component-protease enhance accumulation of bean pod mottle virus-specific siRNAs. evidence that it consists of the genome of an endornavirus. Virus Genes 35, Plant Pathol. J. 27, 315–323. 399–403. Linder-Basso, D., Dynek, J.N., Hillman, B.I., 2005. Genome analysis of Cryphonectria Valverde, R.A., Nameth, S., Abdallha, O., Al-Musa, O., Desjardins, P., Dodds, A., hypovirus 4, the most common hypovirus species in North America. Virology 1990. Indigenous double-stranded RNA from pepper (Capsicum annuum). Plant 337, 192–203. Sci. 67, 195–201. Marchler-Bauer, A., Bryant, S.H., 2004. CD-Search: protein domain annotations on the Vance, V.B., 1991. Replication of potato virus X RNA is altered in coinfections with potato virus Y. Virology 182, 486–494. fly. Nucleic Acids Res. 32, W327–W331. Villanueva, F., Sabanadzovic, S., Valverde, R.A., Navas-Castillo, J., 2012. Complete Markine-Goriaynoff, N., Gillet, L., Van Etten, J.L., Korres, H., Verma, N., Vander- genome sequence of a double-stranded RNA virus from avocado. J. Virol. 86, plasschen, A., 2004. Glycosyltransferases encoded by viruses. J. Gen. Virol. 85, 1282–1283. 2741–2754. Wang, J., Turina, M., Medina, V., Falk, B.W., 2009. Synergistic interaction between Moriyama, H., Nitta, T., Fukuhara, T., 1995. Double-stranded RNA in rice: a novel the Potyvirus, Turnip mosaic virus and the Crinivirus, Lettuce infectious yellows RNA replicon in plants. Mol. Gen. Genet. 248, 364–369. virus in plants and protoplasts. Virus Res. 144, 163–170. Moriyama, H., Horiuchi, H., Nitta, T., Fukuhara, T., 1999. Unusual inheritance of White, T.J., Bruns, T., Lee, S., Taylor, J.W., 1990. Amplification and direct sequencing evolutionarily-related double-stranded RNAs in interspecific hybrid between rice of fungal ribosomal RNA genes for phylogenetics. In: Innis, M.A., Gelfand, plants Oryza sativa and Oryza rufipogon. Plant Mol. Biol. 39, 1127–1136. D.H., Sninsky, J.J., White, T.J. (Eds.), In PCR Protocols: A Guide to Methods Morris, T.J., Dodds, J.A., 1979. Isolation and analysis of double-stranded RNA from and Applications. Academic Press, San Diego, pp. 315–322. virus-infected plant and fungal tissue. Phytopathology 69, 854–858. O'Brien, H.E., Parrent, J.L., Jackson, J.A., Moncalvo, J.-M., Vilgalys, R., 2005. Fungal

104

Supplementary information

(A) (B) (C)

Figure S1. Terrestrial orchid species from which mycorrhizal fungi with endornaviruses were identified: (A) Microtis media (common mignonette orchid) (B)

Pterostylis sp. (snail orchid) and (C) Pterostylis sanguinea (dark banded greenhood orchid).

1 2 3 4 5 6 7

— 23.13 kb 10 kb —

3 kb — 2.322 kb —

1 kb —

— 0.564 kb

0.3 kb —

Figure S2. Agarose gel electrophoresis of dsRNAs extracted from orchid mycorrhizal fungi, Ceratobasidium. Arrows indicate the position of dsRNA bands. Lane 1: DNA ladder (Axygen 1 kb DNA ladder), Lane 2: Ceratobasidium C01 (Pterostylis sp.), Lane 3: Ceratobasidium C02 (Pterostylis sp.), Lane 4: Ceratobasidium C03 (Pterostylis sp.), Lane 5: Ceratobasidium C04 (Microtis media), Lane 6: Ceratobasidium C05 (Pterostylis sanguinea), Lane 7: Lambda DNA (HindIII cut).

105

Table S1. Pairwise identity (%) of amino acid and nucleotide sequences between (A) Ceratobasidium endornaviruses (CbEV) A-H and (B) proposed and previously described endornaviruses; numbers in parentheses represent nt identity (%). Abbreviations: Basella alba endornavirus (BaEV), Bell pepper endornavirus (BPEV), Gremmeniella abietina type B RNA virus XL (GABrV-XL), Helicobasidium mompa endornavirus 1 (HmEV1), Oryza rufipogon endornavirus (OrEV), Oryza sativa endornavirus (OsEV), Persea americana endornavirus (PaEV), Phaseolus vulgaris endornavirus 1 (PvEV1), Phaseolus vulgaris endornavirus 2 (PvEV2), Phytophthora endornavirus 1 (PEV1), Rhizoctonia cerealis endornavirus 1 (RcEV1), Sclerotinia sclerotiorum endornavirus 1 (SsEV1), Tuber aestivum endornavirus (TaEV), Vicia faba endornavirus (VfEV) and Yerba mate endornavirus (YmEV).

(A) aa CbEVA CbEVB CbEVC CbEVD CbEVE CbEVF CbEVG CbEVH nt

CbEVA 10.5 14.3 13.1 10.0 9.7 14.4 12.9

CbEVB 42.1 16.9 12.7 9.4 10.7 18.5 12.8

CbEVC 42.6 42.1 12.7 8.5 9.4 15.7 13.4

CbEVD 42.2 40.9 42.0 8.0 9.5 13.3 30.2

CbEVE 42.0 44.5 42.6 42.2 11.5 9.8 9.9

CbEVF 42.5 44.1 42.5 43.1 44.1 10.2 11.2

CbEVG 42.0 42.8 42.9 41.8 42.6 43.0 12.6 CbEVH 42.2 41.6 42.7 49.1 42.3 41.9 42.3

106

(B)

Host Plant Fungus Oomycete GABrV BaEV BPEV OrEV OsEV PaEV PvEV1 PvEV2 VfEV YmEV HmEV1 RcEV1 SsEV1 TaEV PEV1 -XL 14.4 14.9 12.8 13.3 13.6 13.2 14.9 13.7 13.2 11.5 12.9 14.0 10.3 10.8 12.9 CbEVA (42.3) (42.1) (42.8) (42.4) (43.3) (42.2) (42.3) (41.5) (42.6) (42.8) (42.6) (42.9) (42.3) (42.0) (42.5) 14.0 15.4 13.6 13.6 14.2 14.1 14.8 11.1 14.3 9.6 22.2 10.8 10.1 9.3 14.0 CbEVB (44.7) (44.7) (45.4) (44.9) (44.9) (44.7) (44.1) (42.6) (45.2) (43.9) (45.3) (43.5) (44.3) (43.1) (43.4) 13.5 14.5 12.8 13.1 13.6 13.0 14.2 11.7 13.8 9.7 16.2 14.6 10.8 10.4 13.9 CbEVC (42.4) (43.0) (42.7) (42.2) (43.3) (42.0) (43.0) (42.6) (42.5) (42.0) (42.7) (42.1) (42.1) (41.8) (42.3) 12.4 14.2 12.7 12.0 11.6 12.5 14.4 12.7 11.7 9.6 14.1 14.8 10.7 10.6 12.0 CbEVD (41.9) (42.3) (41.9) (42.0) (42.7) (42.5) (42.7) (42.0) (42.1) (41.8) (42.1) (43.2) (41.7) (41.4) (42.6) 9.9 10.2 10.1 9.6 11.0 9.5 9.5 9.3 9.9 11.3 10.0 10.2 11.1 16.7 9.4 CbEVE (43.3) (43.5) (43.6) (44.2) (43.3) (43.7) (43.3) (42.6) (43.7) (43.7) (42.7) (43.2) (43.4) (42.9) (42.3) 9.6 11.7 10.3 10.5 9.0 9.8 11.4 10.2 9.1 12.6 9.1 8.7 13.4 13.9 10.2 CbEVF (44.0) (44.2) (43.7) (44.1) (43.9) (44.1) (43.2) (42.5) (44.7) (44.5) (43.1) (43.6) (44.4) (44.1) (43.2) 13.2 12.6 12.3 12.2 13.3 12.5 12.0 12.6 12.7 11.4 14.1 14.0 11.5 11.3 13.2 CbEVG (42.8) (42.9) (42.5) (42.9) (43.1) (43.2) (42.4) (42.3) (42.8) (42.7) (42.8) (43.4) (42.8) (42.9) (43.3) 13.0 12.6 13.1 13.0 12.6 12.8 13.3 12.7 12.7 11.4 12.7 15.3 11.4 10.8 13.4 CbEVH (41.4) (42.7) (41.3) (40.7) (42.6) (41.8) (42.7) (42.5) (41.5) (41.9) (42.5) (43.0) (41.8) (41.7) (41.9)

107

Chapter 6: General discussion

Western Australian terrestrial orchids are intrinsically linked with mycorrhizal fungi and insect pollinators as part of their life cycles. The impact of viruses, in particular indigenous viruses, on these partnerships remains largely unexplored. The unusual features of Western Australia’s biological, environmental and geographical landscapes presented a unique study site for detection of viruses, using a high throughput sequencing strategy, associated with native Western Australian terrestrial orchids and their fungal partners in their natural environments.

Thirty-two viruses, of which 31 are proposed new species, were partially characterised from leaves and mycorrhizal fungi of wild plants of Drakaea, Microtis and Pterostylis orchids. In addition, other small virus-like sequences were detected but not analysed in depth. Of 215 plants from 34 orchid populations tested, four viruses were discovered from leaves of 11 plants, indicating that virus infection of the wild orchids tested was uncommon. In contrast, 28 viruses were identified from 10 mycorrhizal fungal isolates from nine orchid populations. Earlier studies reporting viruses of wild orchids growing under natural conditions were from our research group in Western Australia (Wylie et al., 2012; 2013a; 2013b), and also from eastern

Australia (Gibbs et al., 2000), India (Sherpa et al., 2006; Singh et al., 2007) and Japan

(Kawakami et al., 2007). There is one report of mycoviruses associated with orchids

(James et al., 1998), but no one had previously examined orchid plants and their mycorrhizal associates together.

In combination with traditional methods of virus detection, high-throughput sequencing has enabled efficient detection and molecular characterisation of novel

108

viruses. This approach was essential in identifying viruses infecting Western

Australian native orchids and their associated mycorrhizal partners. Until high- throughput shotgun sequencing became generally available and affordable in the early part of this decade, it would have been very difficult to undertake this study using traditional methodologies. For example, before high throughput sequencing technologies were used in the field of plant virology, only about one new plant virus was identified per year from the UK over the previous 30 years (Adams et al., 2009a).

6.1 Plant and fungal viruses

Four viruses were identified from the leaves of 215 terrestrial orchid plants – two from hammer orchids (Drakaea spp.; DVA and DOSV) and two from dark banded greenhood orchids (Pterostylis sanguinea; PsVA and PsTVA). Virus prevalence (1.9%) in this study is comparable to those found in some other studies of viruses in wild orchids. Incidence of DOSV ranged from 0.8% to 7.8% (detected in a pooled sample of 10 plants) in two populations of Caladenia and Diuris orchids in

Western Australia (Wylie et al., 2013b). In Japan, cucumber mosaic virus (CMV) was reported to naturally infect 3.8% of 104 wild Calanthe orchids (Kawakami et al.,

2007). In another study of wild Western Australian terrestrial orchids, four exotic and native viruses, viz. bean yellow mosaic virus (BYMV), blue squill virus A, donkey orchid virus A and Ornithogalum mosaic virus were detected from only 10 symptomatic wild Diuris (donkey orchid) plants (Wylie et al., 2013a). Studies of wild non-orchid plants reported higher rates of virus infection (MacClement and Richards,

1956; Prendeville et al., 2012). Prendeville et al. (2012) detected at least one virus in each of 12 of the 14 sampled wild Cucurbita pepo populations, with typical prevalence of less than 30% within populations. MacClement and Richards (1956)

109

surveyed more than 2000 plants representing 29 species from six American wild plant communities and reported an average virus infection rate of 10%. In addition, they reported variation in percentage of infection between seasons, but the rate was higher in perennial plants than annuals. Their calculated infection rate is certainly an underestimation because only viruses that induced symptoms on inoculated experimental host plants Nicotiana tabacum and Solanum lycopersicum were counted.

In the current study, DOSV was transmissible to N. benthamiana (Wylie et al.,

2013b), but transmission of DVA was restricted to Drakaea orchids (Ong et al.,

2016a), so transmission to experimental hosts is an unreliable method of detecting viruses.

An unexpected finding of this study was the abundance of mycoviruses, predominately endornaviruses and partitiviruses, discovered from fungal isolates in orchid roots. Twenty eight mycoviruses were identified from six isolates of

Ceratobasidium taxa at a much higher incidence than orchid plant viruses. Three of the six fungal isolates studied had more than five viruses co-infecting them, while remaining three isolates had one characterised virus. In contrast, Feldman et al.

(2012) found the incidence of mycoviruses to be comparable to that of wild plant viruses. Using presence of virus-indicative dsRNA banding patterns followed by high-throughput sequencing (Roche 454), they found 10% of the 225 samples of fungal mycelia contained 25 viral sequences representing 16 recognised viral taxa.

Higher incidence rates were detected in studies targeting specific viruses and fungi

(Peever et al., 1997; Voth et al., 2006). Peever et al. (1997) reported presence of dsRNA, including the hypoviruses Cryphonectria hypovirus 1-3 (CHV1-3) in 28% of

Cryphonectria parasitica isolates from eastern North America. Ustilago maydis virus

110

H1 (Umv-H1) was found to infect 34% and 100% of tested Ustilago maydis (corn smut fungus) isolates in USA and Mexico, respectively (Voth et al., 2006).

Actual numbers of mycoviruses present in the Ceratobasidium isolates studied may be even higher than described, for the reasons discussed below. We made a decision not to count viruses from which less than 50% of the estimated genome was obtained from shotgun sequencing. This cautious approach avoided overestimating virus numbers, but it might well have lead to an underestimation of numbers of viruses present. Why might there be even more viruses present in the mycorrhizal fungi studied? For unknown reasons, some mycoviruses are lost from fungal hosts during culture on artificial media (Márquez et al., 2007; Feldman et al., 2012;

Roossinck, 2015). The loss of mycoviruses during prolonged culture periods might explain the apparent absence of mycoviruses from slower growing fungi such as

Tulasnella sp. isolated from Drakaea orchids. Slow growing fungal isolates spent months on both solid and liquid media, which might have resulted in loss of viruses.

As high-throughput sequencing chemistry becomes more sensitive (e.g. Illumina’s

TruSeq Nano DNA library Prep kit designed for preparing libraries for sequencing from very low amounts of input DNA), it is expected that mycoviruses will be detected directly from pelotons, eliminating the need for fungal cultures. For other mycoviruses, their low titres or complex RNA structures may make it difficult to obtain their complete sequences.

Like other orchid mycorrhizal fungal species, Ceratobasidium species are not obligate symbionts. They can also adopt ectomycorrhizal, endophytic, plant pathogenic and saprophytic lifestyles (Brundrett et al., 2003; Brundrett, 2006;

111

Mosquera-Espinosa et al., 2013). The choice of mycorrhizal partner by an orchid is more likely to be influenced by the compatibility and availability of the fungus

(Warcup, 1981; Bonnardeaux et al., 2007; Brundrett, 2007) than by the number and types of viral taxa infecting it, although this has not been shown experimentally.

Plant-fungus-mycovirus ecology is a poorly examined field, but in two cases the viral component of this complex three-way relationship has been shown to positively influence the plant partner (Shapira and Nuss, 1991; Márquez et al., 2007).

Much remains unknown about the fungus-mycovirus relationship. It is not known if mycoviruses are transmitted horizontally by arthropod or other vectors as most plant viruses are. It is not known if fungi can rid themselves of mycoviruses under natural conditions as many plants do through seed generations. It is thought that horizontal transmission of mycoviruses occurs through inter- and intra-species associations that lead to their accumulation (Liu et al., 2003; Milgroom and Hillman,

2011; Vainio et al., 2011). Mycoviruses are thought to accumulate over long periods via anastomosis and are maintained during both asexual and sexual generations

(Campbell, 1996; Milgroom and Hillman, 2011; Vainio et al., 2015). Vertical transmission of mycoviruses in basidiomycetes is predominately through basidiospores (sexual) (Buck, 1998; Milgroom and Hillman, 2011), but has also been detected in conidial isolates (asexual) (Ihrmark et al., 2002). Spore transmission of

Ceratobasidium mycoviruses remains to be shown experimentally, but in related

Rhizoctonia, a virus-like dsRNA was demonstrated to transmit via basidiospores at a rate of 37-88% over multiple (4-6) generations (Castanho and Butler, 1978). If mycoviruses in orchid mycorrhizal fungi do transmit vertically, it is likely to be an uncommon event as sporulation by these fungal species has rarely been observed. Our

112

findings suggest that an abundance of viruses exists in the fungal flora. If this is the case their diversity is undoubtedly high, and the roles they play in ecosystems significant.

In contrast to fungi, orchids and many other plants have the ability shed some plant viruses during sexual reproduction by excluding them during seed development.

In agriculture, transmission of virus occurs predominantly via vectors, and seed transmission is relatively uncommon, accounting for only about 18% of plant viruses

(Johansen et al., 1994). Similarly, orchid-infecting viruses are typically transmitted via vectors or mechanical means but only rarely through their seed (Zettler et al.,

1990). Cymbidium mosaic virus was seed-transmitted at rates of 0.3-0.4% (Yuen et al.,

1979; Hu et al., 1993). On the other hand, vegetative generations of orchids that emerge from stolons or tubers of infected parent plants can accumulate viruses in a manner similar to fungi.

6.2 Diversity and uniqueness of new viruses

High virus diversity was identified in other studies where generic high- throughput sequencing approaches were used (e.g. Roossinck et al., 2010; Al

Rwahnih et al., 2011; Feldman et al., 2012; Marzano and Domier, 2016). Feldman et al. (2012) identified 18 mycoviruses from seven species of fungal endophytes isolated from Ambrosia psilostachya (Western ragweed) and its parasitic plant Cuscuta. These viruses belonged to the Chrysoviridae, Endornaviridae, Hypoviridae, Narnaviridae,

Partitiviridae and Totiviridae. These same virus families were also identified from fungi by Al Rwahnih et al. (2011), Roossinck et al. (2010), and, with the exception of

Chrysoviridae, from our study. To date, no member of the Chrysoviridae has been

113

identified from Australia, but that might be because of the paucity of studies done on mycoviruses there.

Three new virus genera were recently formally ratified by the ICTV to accommodate new Australian orchid viruses – viz. Divavirus (Diuris virus A and

Diuris virus B; family Betaflexiviridae; Wylie et al., 2013a), Goravirus (DVA; family

Virgaviridae; Ong et al., 2016a) and Platypuvirus (DOSV – three isolates; family

Alphaflexiviridae; Wylie et al., 2013b; Ong et al., 2016a). Other viruses described in this study challenge classification criteria of existing virus genera Endornavirus

(CbEVB, CbEVC and CbEVG; Ong et al., 2016b), Hypovirus (CbHVA) and

Mitovirus (CbMVA). The three new endornaviruses identified in Ceratobasidium isolates encode a second ORF, a feature not seen before in members of this genus.

The new hypovirus and mitovirus are clearly accommodated within existing genera, but differ in significant ways from other members of these genera.

The diversity and uniqueness of the viruses associated with Western

Australian orchids and their fungal partners reflect the unusual features of Western

Australia’s biological, environment and geographical landscapes. Since separation of the Australian continent from Antarctica and the rest of Gondwanaland, its biota is thought to have evolved predominantly in isolation (Crisp et al., 2004), but with influences from floras to the north and east (Hopper and Gioia, 2004). Australia’s flora and fauna has a high level of endemism, especially in South-western Australia,

Tasmania and the wet tropics of the north-east (Hopper, 1979; Crisp et al., 2001). The current Australian floral landscape, predominately of eucalypts, acacias and casuarinas, was influenced by the vegetation of Gondwanaland, its changing

114

latitudinal position as the Australian plate drifted northwards, and a climatic shift from tropical to one increasingly cooler and drier (Frakes, 1999; Crisp et al., 2004). In the south-western corner of Western Australia, where this study was done, the ancient land has been geologically stable for over 3 billion years. It has had no ice cover for

300 M years, it was covered in rainforest from 145 M to 65 M years ago, and it has experienced a Mediterranean climate for the last 20 M years. The soil is highly weathered and infertile. These are predominant among a host of factors that have stimulated high biological endemism within the region (Myers et al., 2000; Coates and Atkins, 2001; Cribb et al., 2003).

Overall, the variety of viruses observed in this study reflects: (1) the floral diversity and endemism in Western Australia, in particular the diversity of terrestrial orchids (Coates and Atkins, 2001; Crisp et al., 2001), (2) compatibility of Western

Australian orchids with diverse groups of fungi – e.g. Ceratobasidium, Tulasnella and

Sebacina (Bonnardeaux et al., 2007), (3) genetic isolation (in some cases), (4) gene flow to Australia from Asia and elsewhere, (5) long term occupation of the same area by the plants and fungi, and (6) long period of association between the viruses and their wild hosts.

6.3 Virus ecology and evolution

How do plant and fungal viruses move from region to region and globally?

Man has undoubtedly played a large role in virus movement by spreading viruses in crop plants, possibly to the extent of triggering massive speciation within the genus

Potyvirus when agriculture began in China over 7000 years ago (Gibbs and Ohshima,

2010). Exotic introductions to Australia by man occurred long before the main influx

115

of colonialists arrived from Europe 200 years ago. The Australian dingo (Canis lupus dingo) was introduced 6000 years ago (Savolainen et al., 2004) and tamarind

(Tamarindus indica) was introduced by Makassans from Indonesia to northern

Western Australia several hundred years before Europeans settled (Russell, 2004).

Since waves of colonialists have arrived on the continent in the last two centuries, they have brought with them a flood of exotic plants, many of which must have been sources of exotic viruses and their vectors (Cooper and Jones, 2006). Some of these exotic viruses have been identified infecting native plant species, including orchids

(Guy and Gibbs, 1985; McKirdy et al., 1994; Wylie et al., 2013a), which is indicative of their ability to colonise new hosts. There are also reports about introductions of soil-borne microorganisms such as fungi and their viruses (Maccarone et al., 2010a;

2010b; 2010c). While both plants and fungi have been shown to be capable of transporting viruses from other continents, it would be more likely for viruses to cross the oceans in infected fungal spores than in infected plants.

The viruses identified in this study are probably long-time residents of

Australia, not recent microbial invaders that were passively carried to the continent with recent human immigrants. All share phylogeny to a greater or lesser degree with higher order taxa described from other continents. The viruses identified in this study had a mixture of unusual and familiar features. Novel features were perhaps developed in response to challenges/opportunities faced over millions of years in the

Australian environment. The familiar features, as seen in viruses from other parts of the world, suggest a more recent shared ancestry. Such a mixture indicates there has been a natural flow of viruses into, and presumably out of, the Australian continent over evolutionary time. A good example is the totivirus PsTVA that shares almost

116

60% sequence identity with black raspberry virus F (BRVF) isolated from wild black raspberry in North America. PsTVA and BRVF are almost close enough to be considered isolates of the same species, so it is reasonable to assume they share a recent, internationally mobile, common ancestor. On the other hand, two of the

Ceratobasidium partitiviruses (CP-d and CP-e; Alphapartitivirus) appear to be ancestral to all other known alphapartitiviruses, indicative of long isolation in

Australia. In some cases progenitors may be Gondwanan in origin, becoming isolated on the Australian landmass when it separated from Antarctica 35.5 million years ago

(McLoughlin, 2001). Other partitiviruses characterised in this study are much closer to those identified from other continents, indicative of relatively recent dispersal into and/or out of Australia. These findings tell us that components of the virus-infected flora or fungi of the isolated south-western corner of the Australian continent have recent international connections and have not evolved in isolation for millions of years.

6.4 Viruses and orchid biology

The importance of the partnership between orchids, mycorrhizal fungi and insect pollinators is well established. However, our knowledge of the roles or impact of other microorganisms within this symbiosis is limited. No visible symptoms were evident on the virus-infected orchid plants studied, but this does not necessarily mean there is no impact, positive or negative, on the infected plants. While stunting, flower abortion, etc caused by viruses would be clear indicators of pathogenesis, there may be more subtle costs of infection that influence plant reproductive success over the short or long term. It is also possible that the influence of a virus on its host may change over its life cycle, or under the specific biotic and abiotic stresses the plant

117

encounters over its life. For example, some viruses have been reported to induce tolerance in their plant hosts to heat (Curvularia thermal tolerance virus (CThTv);

Márquez et al., 2007), cold (CMV; Xu et al., 2008) and drought (brome mosaic virus,

CMV, tobacco mosaic virus and tobacco rattle virus; Xu et al., 2008). If viruses are generally beneficial to the breeding success of wild orchids, one might expect to find virus-infected orchids commonly because infected plants would be more successful.

Conversely, if viruses have a negative influence on orchid fecundity and survival under normal conditions, one would expect to see infected plants occurring more rarely than non-infected plants. The relatively low number of virus-infected orchid plants found in this study, and the low proportion of infected orchid plants within populations in other studies (Kawakami et al., 2007; Wylie et al., 2013b) support the second scenario. These assumptions are based on native viruses infecting native plants growing in their natural environments. Much higher rates of infection in native plants, including orchids by recently-introduced exotic viruses are reported (Cox, 2004;

Jones and Baker, 2007; Wylie et al., 2013a; Vincent et al., 2014). A possible third scenario exists that superficially resembles the second scenario – viruses occur rarely because they generally decrease plant vigour, but under rare detrimental biotic or abiotic circumstances, infected plants display greater reproductive success than uninfected plants. Experimental support for the third scenario would be difficult to establish because of the need to impose all possible stressors at all possible life stages.

Six of the fungal isolates tested were infected by at least one persistent virus.

The presence of multiple viruses in mycorrhizal fungi does not necessarily indicate that these mycoviruses play an important role in the biology of either the fungus or the orchid, but the observed tolerance or receptivity of fungi to infection by multiple

118

viruses hints they may at least have a mutualistic role. While many mycoviruses showed no significant effects on their fungal hosts, some mycoviruses are clearly indirectly beneficial to plants, for example CHV1 (Shapira and Nuss, 1991). The presence of CHV1 reduces the virulence of Cryphonectria parasitica, the causative agent of chestnut blight, thereby reducing symptoms of its infection (Shapira and

Nuss, 1991; Chen and Nuss, 1999; Dawe and Nuss, 2001). In a relationship resembling that of orchids and mycorrhizal fungi, panic grass (Dichanthelium langinosum) plants that live in the hot soils of geothermal areas in the USA form a mutualistic relationship with the ascomycete Curvalaria protuberata, which is itself is infected with CThTV (Márquez et al., 2007). The plant is incapable of surviving the heat of its environment without the presence of both the fungus and its mycovirus, although the mechanism for this was not elucidated (Márquez et al., 2007).

Do Ceratobasidium isolates from orchid pelotons carry the same mycovirus infections as free living Ceratobasidium isolates of the same species? This question was not asked here, but it cannot be inferred that all Ceratobasidium strains in the soil are similarly infected with multiple viruses. However, it would be informative to address this question experimentally because it would clarify whether mycoviruses play a role in formation of mycorrhizal associations with plants. This could be addressed by curing Ceratobasidium isolates of successive numbers of mycoviruses and testing relative abilities to form stable mycorrhizal associations. Glasshouse inoculation experiments with mycovirus-infected and mycovirus-free mycorrhizal fungi to orchid plants may determine the physical (e.g. differences in the rate of growth and flowering, and longevity) and physiological (e.g. up or down regulation of metabolites) effects of mycoviruses on orchids. Eliminating mycoviruses from fungal

119

cultures is possible but treatments are not always effective (Martins et al., 1999;

Romo et al., 2007). Some successful cures include cyclohexamide treatments (Elias and Coty, 1996), dehydration combined with freeze-thawing (Márquez et al., 2007), single conidium subculture (Elias and Coty, 1996; Azevedo et al., 2000), temperature treatments (Romo et al., 2007) and long periods of growth on artificial media

(Márquez et al., 2007; Feldman et al., 2012; Roossinck, 2015).

If Ceratobasidium strains are universally and asymptomatically infected with viruses, it would infer that there exists a mutualistic equilibrium between the fungal hosts and viruses (Yamamura, 1996; Roossinck, 2010; Bao and Roossinck, 2013). A possible benefit of the virus in the fungus is that infection with a mild strain of virus can protect it against a more severe strain, as in cross-protection reported in plants

(Fulton, 1986; Fraser, 1998). Thus, such a role in maintaining fungal viruses would indirectly benefit the orchid with which it was associated.

The impact of vectors on transmission of viruses between native orchids has not been investigated, but in studies of exotic orchids such as Cymbidium,

Dendrobium, Masdevallia and Phalaenopsis, aphids and mites weere found to transmit BYMV (Hammond and Lawson, 1988; Zettler et al., 1990) and orchid fleck virus (Maeda et al., 1998) respectively. Viruses related to DVA, such as members of

Goravirus, Hordeivirus and Pecluvirus are transmitted via pollen grains from plant to plant (Reddy et al., 1998; Adams et al., 2009b; Atsumi et al., 2015). Thus, if transmission of DVA is indeed through pollen, specialist thynnid wasp pollinators of

Drakaea orchids are likely to have a role in virus transmission. This applies to any other viruses that are either contact or pollen transmissible because all Western

120

Australian terrestrial orchids are pollinated to a greater or lesser extent by insects

(Brundrett, 2014). The interdependence of plant, fungus, and insect might have facilitated viruses to specialise in orchids.

An important factor in the reproductive success of orchids is their ability to attract pollinating insects via physical and/or chemical mimicry, for example Drakaea orchids (and several other genera) use physical and sex pheromone mimicry to attract male wasps. Plants infected with viruses such as CMV and potato leafroll virus have been shown to alter insect behaviour to enhance rate of virus acquisition and transmission (Mauck et al., 2009; Ingwell et al., 2012; Rajabaskar et al., 2014).

Infection by viruses changed the concentration of emitted plant volatile compounds, which increased their attractiveness to non-viruliferous aphid vectors; while viruliferous aphids preferred non-infected hosts (Eigenbrode et al., 2002; Mauck et al.,

2009; Ingwell et al., 2012; Rajabaskar et al., 2013). Thus, it is important to determine if viruses have an effect on expression of pheromone-mimicking compounds that influence attractiveness of the orchids to pollinators, and therefore influence reproductive success. The relative rarity of plant viruses infecting the orchids studied suggests that viruses do not play a significant role in pollination success, and although it seems unlikely that mycoviruses might influence this process, the experiments proposed in preceding paragraphs (with mycovirus-free fungal partners) could be used as a basis to determining if mycoviruses influence pollination.

6.5 Virus exchange between hosts?

The two viruses detected from leaves of P. sanguinea, PsTVA (proposed totivirus) and PsVA (unclassified virus), were more closely related to mycoviruses

121

than to known plant viruses. The question remains as to whether these two viruses replicate in the cells of plants or in fungi. A PCR-based test of the leaves did not reveal the presence of fungus in the plant leaves, but other tests including leaf staining and fungal isolation from leaves are required to confirm this. If they are indeed plant viruses that resemble fungal viruses, how did they cross the species barrier from fungi? Plant-infecting partitiviruses are hypothesised to have been transmitted horizontally between plant and fungal hosts at some point during their evolution

(Roossinck, 2010). This is based on the incongruent grouping of plant- and fungus- infecting members in both Alphapartitivirus and Betapartitivirus (Roossinck, 2010;

Nibert et al., 2014). It must be noted that it is far from certain that all described plant- infecting partitiviruses are able to replicate in plant cells in the absence of a fungal host; indeed some may be mycoviruses from unidentified fungal endophytes within plants.

Many orchid mycorrhizal species, including Ceratobasidium, interacts with plants outside of the Orchidaceae family. For example, species of Sebacina can occur as orchid mycorrhizas (e.g. Caladenia), endophytic fungi (e.g. Phyllanthus) as well as ectomycorrhizas (e.g. Eucalyptus) (Warcup, 1988). Their interaction with members of multiple plant families suggests that these multifunctional fungi can potentially be important virus vectors, especially if the infecting viruses can transmit between the two host types.

6.6 Importance of wild plant virology

In the field of plant virology, most research has concentrated on disease- causing viruses of horticultural and agricultural crops, predominately in highly in-

122

bred and vegetatively-propagated cultivars, often growing in places far from where they evolved. Many of these studies relied on traditional methods of ELISA and RT-

PCR, which are targeted approaches to diagnosing known viruses. As a consequence, the number of viruses described from horticultural and agricultural crops is probably an under-representation of true virus diversity in these plants. The introduction of high throughput shotgun sequencing in combination with traditional methods has enabled detection of a surprising diversity of novel viruses from a wide range of organisms in both wild and human-managed environments (e.g. Roossinck et al.,

2010; Feldman et al., 2012, Wylie et al., 2012; Wylie et al., 2013a). Studies of natural ecosystems reveal they possess a rich array of both plant and fungal viruses. One important reason for studying the viruses associated with wild plants lies in their potential to spill over into agricultural crops. For example, turnip mosaic virus

(Potyvirus; Potyviridae), a highly widespread and damaging virus, probably spread from wild European orchids to brassicas (Nguyen et al., 2013). The virulence of

‘emerging viruses’ is dependent on the susceptibility of host species, presence of vectors, and ecological and environmental conditions (Elena et al., 2011; Hily et al.,

2016). Disease emergence is hypothesised to be partly attributable to factors such as disturbance to natural landscapes and reductions in biodiversity as a direct or indirect result of human activities (Keesing et al., 2010; Roossinck and García-Arenal, 2015).

Fragmentation of natural environments offers greater potential opportunity for virus spill over from wild to cultivated plants, and vice versa. The numbers of asymptomatic viruses found associated with wild orchids and fungi suggest that native flora is a rich reservoir of viruses, some of which have emerged to infect exotic new hosts (Webster et al., 2007; Luo et al., 2011; Kehoe et al., 2014; Li et al., 2016).

Thus, a justification for directing resources to understanding the ecology of viruses of

123

native plants is that such studies may protect agricultural and horticultural crops from epidemics because we will already have knowledge of the biology of these pathogens

(Li et al., 2016).

The global movement of plants and their viruses makes it difficult to make meaningful assumptions about the origins of many viruses isolated from cultivated plants. For example, DVA, a goravirus from Australian orchids is related to two pecluviruses, both from peanut crops in western Africa and the Indian subcontinent.

Their only known host is peanut (Arachis hypogea), an allotetraploid originating from northern Argentina/southern Bolivia (Kochert et al., 1996; Seijo et al., 2007; Adams et al., 2012). This situation raises questions about whether the peanut pecluviruses are indigenous to the continents on which they were described, presumably as spill over from the indigenous flora, or if they are originally from peanuts in South America where they were subsequently transported to India in germplasm (perhaps to the

International Crops Research Institute for the Semi-Arid Tropics, ICRISAT), and subsequently to Africa. Both peanut-infecting pecluviruses are seed borne (Reddy et al., 1998; Adams et al., 2009b; Dieryck et al., 2009), supporting this hypothesis. If so, pecluviruses still exist undetected in South America, and the ancestors of the orchid goravirus and legume pecluviruses probably evolved in Gondwanaland and became separated during continental drift. This situation illustrates why the study of wild plant and fungal viruses is so important to understanding their ecology and evolution.

There is a far greater degree of certainty associated with the geographical and host origins of viruses isolated from indigenous plants living in natural systems than there is from cultivars of domesticated species that may have been traded and cultivated internationally for centuries.

124

Appendix 1

Table A1. List of orchid plant and mycorrhizal fungus samples tested Sample ID. Sample ID. Orchid (No. of Mycorrhizal fungi Location of collection Date of GPS Common namea [Chapter no.; species individuals) species in WA collection Co-ordinatesb,c ID] [Chapter no.; ID] Caladenia Cowslip Orchid - Sebacina sp. F-CA01 Murdoch 10/9/2012 - flava

Caladenia sp. - - Sebacina sp. F-CA02 Beeliar Regional Park 19/10/2012 -

-32o 04.497'' Caladenia sp. - - Sebacina sp. F-CA03 Beeliar Regional Park 1/11/2012 115o 49.906'' -32o 4' 15.0234'' Caladenia sp. - CA01 (4) - - Murdoch 27/06/2013 115o 50' 7.983'' -32o 4' 14.4192'' Caladenia sp. - CA02 (5) Tulasnella sp. F-CA04 Murdoch 27/06/2013 115o 50' 10.377'' -32o 4' 13.13009'' Caladenia sp. - CA03 (2) - - Beeliar Regional Park 14/07/2013 115o 50' 9.54947'' -32o 4' 13.39865'' Caladenia sp. - CA04 (5) Sebacina sp. F-CA05 Beeliar Regional Park 14/07/2013 115o 50' 8.44351'' -32o 4' 13.48453'' Caladenia sp. - CA05 (4) - - Beeliar Regional Park 14/07/2013 115o 50' 7.75212'' -32o 4' 29.43237'' Caladenia sp. - CA06 (3) - - Beeliar Regional Park 21/08/2013 115o 49' 53.43738'' -32o 4' 31.46553'' Caladenia sp. - CA07 (7) - - Beeliar Regional Park 21/08/2013 115o 49' 57.50694''

125

Caladenia -32o 4' 30.7218'' Cowslip Orchid CA08 (2) - - Beeliar Regional Park 21/08/2013 flava 115o 49' 56.64933'' Caladenia -32o 3' 55.44509'' Cowslip Orchid CA09 (3) - - Murdoch 4/9/2013 flava 115o 50' 27.68189'' Caladenia -32o 3' 0.36024'' Pink Fairy Orchid CA10 (10) - - Murdoch 4/9/2013 latifolia 115o 50' 28.70699'' Caladenia -32o 3' 55.12941'' Cowslip Orchid CA11 (5) - - Murdoch 4/9/2013 flava 115o 50' 24.96919''

Diuris -32o 04.485'' Pansy Orchid - Tulasnella sp. F-DI01 Beeliar Regional Park 1/11/2012 magnifica 115o 49.900'' Diuris -32o 4' 29.67241'' Pansy Orchid DI01 (9) - - Beeliar Regional Park 21/08/2013 magnifica 115o 49' 52.34223'' Diuris -32o 4' 30.89295'' Pansy Orchid DI02 (7) Tulasnella sp. F-DI02 Beeliar Regional Park 21/08/2013 magnifica 115o 49' 52.16102'' Diuris -32o 4' 29.36531'' Pansy Orchid DI03 (5) - - Beeliar Regional Park 21/08/2013 magnifica 115o 49' 58.45682'' Diuris -32o 4' 31.63556'' Pansy Orchid DI04 (5) - - Beeliar Regional Park 21/08/2013 magnifica 115o 49' 58.80717'' Diuris -32o 3' 55.25098'' Pansy Orchid DI05 (2) - - Murdoch 4/9/2013 magnifica 115o 50' 28.08919'' Diuris -32o 3' 55.14544'' Pansy Orchid DI06 (4) Tulasnella sp. F-DI03 Murdoch 4/9/2013 magnifica 115o 50' 25.24981'' Diuris Western Wheatbelt Monadnocks -32o 23' 05.9'' DI07 (1) Tulasnella sp. F-DI04 5/9/2013 porrifolia Donkey Orchid Conservation Park 116o 15' 05.4''

Drakaea Kneeling Hammer DR01 (7) Private property, - - 1/9/2012 - concolor* Orchid [2; DR01] North-West of Drakaea Slender Hammer DR02 (10) NorthamptonPomeroy Rd, -32o 0' 27.2'' Tulasnellaceae F-DR01 17/09/2012 gracilis** Orchid [2; DR02] Lesmurdie 116o 4' 47.8''

126

Drakaea Warty Hammer DR03 (2) Canning Mills Rd, -32o 4' 54.2'' - - 17/09/2012 livida** Orchid [2; DR03] Canning Mills 116o 5' 27.6'' Drakaea King-in-his- DR04 (11) Qualen Rd, Wandoo -32o 5' 33.9'' - - 17/09/2012 glyptodon** carriage orchid [2; DR04] National Park 116o 34' 11.8'' Drakaea Slender Hammer DR05 (9) Lightning Rd, Wandoo -32o 7' 29.4'' Tulasnellaceae F-DR02 17/09/2012 gracilis** Orchid [2; DR05] National Park 116o 28' 17.3'' Drakaea Warty Hammer DR06 (4) Carrabungup Nature -32o 38' 50.6'' - - 17/09/2012 livida** Orchid [2; DR06] Reserve 115o 42' 55.9'' Drakaea Glossy-leafed DR07 (7) Carrabungup Nature Tulasnellaceae F-DR03 17/09/2012 - elastica** Hammer Orchid [2; DR07] Reserve Drakaea King-in-his- DR08 (2) Carrabungup Nature -32o 38' 50.6'' - - 17/09/2012 glyptodon** carriage [2; DR08] Reserve 115o 42' 55.9'' Drakaea Dwarf Hammer DR09 (2) Mowen 22, East of Tulasnellaceae F-DR04 2/10/2012 - micrantha* Orchid [2; DR09] Margaret River Drakaea Warty Hammer DR10 (5) Mowen 22, East of -33o 55' 25.5'' - - 2/10/2012 livida* Orchid [2; DR10] Margaret River 115o 23' 46.4'' Drakaea Dwarf Hammer DR11 (3) Canebrake Nature Tulasnella sp. F-DR05 2/10/2012 - micrantha* Orchid [2; DR11] Reserve Drakaea King-in-his- DR12 (6) Canebrake Nature -33o 53' 27'' Tulasnella sp. F-DR06 2/10/2012 glyptodon* carriage [2; DR12] Reserve 115o 16' 31.1'' Drakaea King-in-his- DR13 (7) Grays Rd, South of -33o 53' 27'' - - 14/10/2012 glyptodon* carriage [2; DR13] Manjimup 115o 16' 31.1'' Drakaea King-in-his- DR14 (10) Scott River Rd, West -34o 23' 53.33'' - - 14/10/2012 glyptodon* carriage [2; DR14] of Pemberton 115o 48' 19.64'' Drakaea King-in-his- DR15 (13) -34o 19' 12.7'' - - Peerabeelup 14/10/2012 glyptodon* carriage [2; DR15] 115o 46' 14.8'' Drakaea Narrow-lipped DR16 (10) -34o 19' 12.7'' - - Peerabeelup 14/10/2012 thynniphila* Hammer Orchid [2; DR16] 115o 46' 14.8'' Drakaea Narrow-lipped DR17 (9) -34o 19' 12.7'' - - Peerabeelup 14/10/2012 thynniphila* Hammer Orchid [2; DR17] 115o 46' 14.8''

127

Drakaea King-in-his- DR18 (8) Ruabon National -33o 38' 33.5'' Tulasnella sp. F-DR07 30/06/2013 glyptodon carriage [2; DR18] Reserve 115o 30' 19.71'' Drakaea Warty Hammer DR19 (1) -33o 42' 24'' - - South Yallingup 30/06/2013 livida Orchid [2; DR19] 115o 01' 40'' DR20 (4) -33o 42' 24'' Drakaea sp. - - - South Yallingup 30/06/2013 [2; DR20] 115o 01' 40'' Drakaea Glossy-leafed DR21 (4) Carrabungup Nature - - 22/08/2013 - elastica* Hammer Orchid [2; DR21] Reserve Drakaea Warty Hammer DR22 (2) Carrabungup Nature -32o 38' 50.6'' - - 22/08/2013 livida* Orchid [2; DR22] Reserve 115o 42' 55.9'' Drakaea Glossy-leafed DR23 (2) Serpentine River - - 22/08/2013 - elastica* Hammer Orchid [2; DR23] Nature Reserve Drakaea Dwarf Hammer DR24 (2) Mowen Rd, East of - - 22/08/2013 - micrantha* Orchid [2; DR24] Margaret River Drakaea Dwarf Hammer DR25 (3) Mowen 22, East of - - 22/08/2013 - micrantha* Orchid [2; DR25] Margaret River Drakaea Glossy-leafed DR26 (2) - - Private property, Capel 22/08/2013 - elastica* Hammer Orchid [2; DR26] Drakaea Warty Hammer DR27 (2) Spencer Rd, South of -33o 42' 24'' - - 22/08/2013 livida* Orchid [2; DR27] Yallingup 115o 01' 40'' Drakaea Glossy-leafed DR28 (3) Serpentine River - - 22/08/2013 - elastica* Hammer Orchid [2; DR28] Nature Reserve Drakaea King-in-his- DR29 (12) -34o 17' 54.2'' - - Nannup 10/9/2013 glyptodon carriage [2; DR29] 115o 45' 58.1'' Drakaea Warty Hammer DR30 (1) Canning Mills Rd, -32o 4' 54.2'' - - 30/09/2013 livida Orchid [2; CM01] Canning Mills 116o 5' 27.6''

Microtis Common - Ceratobasidium sp. F-MI01 Beeliar Regional Park 19/10/2012 - media Mignonette Orchid

128

Microtis Common -32o 3' 54.9714'' - Ceratobasidium sp. F-MI02 Murdoch 30/10/2012 media Mignonette Orchid 115o 50' 17.736'' Microtis Common -32o 4' 2.5494'' - Ceratobasidium sp. F-MI03 Murdoch 30/10/2012 media Mignonette Orchid 115o 50' 13.848'' Microtis Common MI01 (5) -32o 3' 54.9714'' Rhizoctonia sp. F-MI04 Murdoch 13/09/2013 media Mignonette Orchid [5; P04] 115o 50' 17.736'' Microtis Common MI02 (5) F-MI05 -32o 4' 2.5494'' Rhizoctonia sp. Murdoch 13/09/2013 media Mignonette Orchid [5; P03 and P04] [5; C04] 115o 50' 13.848''

Paracaleana Flying Duck -34o 17' 54.2'' PA01 (3) Nannup 10/9/2013 nigrita Orchid 115o 45' 58.1''

PT01 (5) F-PT01 -32o 3' 54.5034'' Pterostylis sp. Snail Orchid Ceratobasidium sp. Murdoch 13/08/2012 [5; P01] [5; C01] 115o 50' 19.968'' Pterostylis Dark Banded PT02 (4) F-PT02 -32o 3' 55.59798'' Ceratobasidium sp. Murdoch 15/08/2012 sanguinea Greenhood [3 and 4; P-2012] [4; F-2012] 115o 50' 26.85752''

Pterostylis sp. Snail Orchid Ceratobasidium sp. F-PT03 Murdoch 28/08/2012 -

-32o 3' 54.9714'' Pterostylis sp. Snail Orchid - Ceratobasidium sp. F-PT04 Murdoch 10/9/2012 115o 50' 26.448'' F-PT05 -32o 4' 14.0515'' Pterostylis sp. Snail Orchid - Ceratobasidium sp. Murdoch 13/09/2012 [5; C02] 115o 50' 12.4667'' -32o 4' 2.1072'' Pterostylis sp. Snail Orchid - Ceratobasidium sp. F-PT06 Murdoch 13/09/2012 115o 50' 12.4667'' Pterostylis Dark Banded -32o 4' 16.1436'' PT03 (3) Ceratobasidium sp. F-PT07 Murdoch 27/06/2013 sanguinea Greenhood 115o 50' 11.9256'' Pterostylis Dark Banded PT04 (8) F-PT08 -32o 4' 27.87305'' Ceratobasidium sp. Beeliar Regional Park 14/07/2013 sanguinea Greenhood [5; P05] [5; C05] 115o 49' 54.22273'' Pterostylis Dark Banded PT05 (1) F-PT9 -32o 4' 13.64061'' Ceratobasidium sp. Beeliar Regional Park 14/07/2013 sanguinea Greenhood [5; P05] [5; C05] 115o 50' 8.14155''

129

Pterostylis Dark Banded PT06 (4) F-PT10 -32o 4' 29.43237'' Ceratobasidium sp. Beeliar Regional Park 21/08/2013 sanguinea Greenhood [5; P05] [5; C05] 115o 49' 53.43738'' Pterostylis Dark Banded PT07 (7) F-PT11 -32o 4' 30.55127'' Ceratobasidium sp. Beeliar Regional Park 21/08/2013 sanguinea Greenhood [5; P05] [5; C05] 115o 49' 52.73462'' Pterostylis Dark Banded -32o 4' 30.90817'' PT08 (4) Ceratobasidium sp. F-PT12 Beeliar Regional Park 21/08/2013 sanguinea Greenhood 115o 49' 51.48677'' -32o 4' 30.61748'' Pterostylis sp. Snail Orchid PT09 (10) Ceratobasidium sp. F-PT13 Beeliar Regional Park 21/08/2013 115o 49' 53.01233'' Pterostylis Dark Banded PT10 (4) F-PT14 -32o 3' 55.59798'' Ceratobasidium sp. Murdoch 4/9/2013 sanguinea Greenhood [3 and 4; P-2013] [4; F-2013] 115o 50' 26.85752'' PT11 (10) F-PT15 -32o 3' 55.70277'' Pterostylis sp. Snail Orchid Ceratobasidium sp. Murdoch 4/9/2013 [5; P02] [5; C03] 115o 50' 27.64415'' Pterostylis Monadnocks -32o 22' 57.84'' Jug Orchid PT12 (3) Ceratobasidium sp. F-PT16 5/9/2013 recurva Conservation Park 116o 15' 9.71'' Pterostylis Monadnocks -32o 23' 06.1'' Jug Orchid PT13 (1) Ceratobasidium sp. F-PT17 5/9/2013 recurva Conservation Park 116o 15' 05.2'' Pterostylis Monadnocks -32o 23' 10.94'' Jug Orchid PT14 (4) Ceratobasidium sp. F-PT18 10/10/2013 recurva Conservation Park 116o 14' 59.12'' Pterostylis Monadnocks -32o 23' 04.2'' Jug Orchid PT15 (1) Ceratobasidium sp. F-PT19 10/10/2013 recurva Conservation Park 116o 15' 07.2''

Thelymitra -32o 04.485'' Leopard Orchid - - F-TH01 Beeliar Regional Park 2/11/2012 benthamiana 115o 49.903''

a Common name given if known b GPS co-ordinates are given, if known. c GPS co-ordinates of locations with classified rare Drakaea species are not given in order to comply with guidelines on the flora permit. * GPS Samples provided by Dr. Ryan Phillips (Australian National University, Canberra; Kings Park Botanic Gardens and Parks Authority, Perth; University of Western Australia, Perth) ** Samples collected by Jamie W.L. Ong and Dr. Ryan Phillips

130

References

Adams, IP, RH Glover, WA Monger, R Mumford, E Jackeviciene, M Navalinskiene, M Samuitiene and N Boonham. 2009a. Next-generation sequencing and metagenomic analysis: a universal diagnostic tool in plant virology. Mol. Plant Pathol. 10(4): 537- 545.

Adams, MJ, JF Antoniw and J Kreuze. 2009b. Virgaviridae: a new family of rod- shaped plant viruses. Arch. Virol. 154(12): 1967-1972.

Adams, MJ, C Heinze, AO Jackson, JF Kreuze, SA Macfarlane and L Torrance. 2012. Virgaviridae. Virus Taxonomy: Classification and Nomenclature of Viruses: Ninth Report of the International Committee on Taxonomy of Viruses. AMQ King, MJ Adams, EB Carstens and EJ Lefkowitz, 1139-1162. San Diego: Elsevier Academic Press.

Al Rwahnih, M, S Daubert, D Golino and A Rowhani. 2009. Deep sequencing analysis of RNAs from a grapevine showing Syrah decline symptoms reveals a multiple virus infection that includes a novel virus. Virology 387(2): 395-401.

Al Rwahnih, M, S Daubert, JR Urbez-Torres, F Cordero and A Rowhani. 2011. Deep sequencing evidence from single grapevine plants reveals a virome dominated by mycoviruses. Arch. Virol. 156(3): 397-403.

Arditti, J and AKA Ghani. 2000. Numerical and physical properties of orchid seeds and their biological implications. New Phytol. 145(3): 367-421.

Atsumi, G, R Tomita, T Yamashita and K-T Sekine. 2015. A novel virus transmitted through pollination causes ring-spot disease on gentian (Gentiana triflora) ovaries. J. Gen. Virol. 96(2): 431-439.

Atwood, JT. 1986. The size of the Orchidaceae and the systematic distribution of epiphytic orchids. Selbyana 9(1): 171-186.

Azevedo, ACS, DR Sosa-Gómez, MR Faria and MHP Fungaro. 2000. Effects of double-stranded RNA on virulence of Paecilomyces fumosoroseus (Deuteromycotina: Hyphomycetes) against the silverleaf whitefly, Bemisia tabaci strain B (Homoptera: Aleyrodidae). Genet. Mol. Biol. 23(1): 61-63.

Bao, X and MJ Roossinck. 2013. Multiplexed interactions: viruses of endophytic fungi. Advances in Virus Research 86: 37-58.

Barba, M, H Czosnek and A Hadidi. 2014. Historical perspective, development and applications of Next-Generation Sequencing in plant virology. Viruses 6(1): 106-136.

131

Beard, JS, AR Chapman and P Gioia. 2000. Species richness and endemism in the Western Australian flora. J. Biogeogr. 27(6): 1257-1268.

Boonham, N, J Kreuze, S Winter, R van der Vlugt, J Bergervoet, J Tomlinson and R Mumford. 2014. Methods in virus diagnostics: from ELISA to next generation sequencing. Virus Res. 186: 20-31.

Brown, EM, AH Burbidge, J Dell, D Edinger, SD Hopper and RT Wills 1997. Pollination in Western Australia: A Database of Animals Visiting Flowers. Perth: Western Australian Naturalists' Club.

Brundrett, MC. 2004. Diversity and classification of mycorrhizal associations. Biol. Rev. 79(3): 473-495.

Brundrett, MC. 2007. Scientific approaches to Australian temperate terrestrial orchid conservation. Aust. J. Bot. 55(3): 293-307.

Brundrett, MC 2014. Identification and Ecology of Southwest Australian Orchids. Perth: Western Australian Naturalists’ Club.

Brundrett, MC. 2016. Using vital statistics and core-habitat maps to manage critically endangered orchids in the Western Australian wheatbelt. Aust. J. Bot. 64(1): 51-64.

Brundrett, MC and LK Abbott. 1991. Roots of jarrah forest plants .I. Mycorrhizal associations of shrubs and herbaceous plants. Aust. J. Bot. 39(5): 445-457.

Buck, KW. 1998. Molecular Variability of Viruses of Fungi. Molecular Variability of Fungal Pathogens. P Bridge, Y Couteaudier and J Clarkson, 53-72. UK: CAB International.

Campbell, RN. 1996. Fungal transmission of plant viruses. Annu. Rev. Phytopathol. 34(1): 87-108.

132

Castanho, B and EE Butler. 1978. Rhizoctonia decline: a degenerative disease of Rhizoctonia solani. Phytopathology 68(10): 1505-1510.

Chen, B and DL Nuss. 1999. Infectious cDNA clone of hypovirus CHV1-Euro7: a comparative virology approach to investigate virus-mediated hypovirulence of the chestnut blight fungus Cryphonectria parasitica. J. Virol. 73(2): 985-992.

Coates, DJ and KA Atkins. 2001. Priority setting and the conservation of Western Australia's diverse and highly endemic flora. Biol. Cons. 97(2): 251-263.

Cooper, I and RA Jones. 2006. Wild plants and viruses: under-investigated ecosystems. Advances in Virus Research 67: 1-47.

Cowling, RM, PW Rundel, BB Lamont, MK Arroyo and M Arianoutsou. 1996. Plant diversity in Mediterranean-climate regions. Trends Ecol. Evol. 11(9): 362-366.

Cox, GW 2004. Alien Species and Evolution: The Evolutionary Ecology of Exotic Plants, Animals, Microbes, and Interacting Native Species. Washington: Island Press.

Cribb, PJ, SP Kell, KW Dixon and RL Barrett. 2003. Orchid Conservation: A Global Perspective. Orchid Conservation. KW Dixon, SP Kell, RL Barrett and PJ Cribb, 1- 24. Kota Kinabalu, Sabah: Natural History Publications.

Crisp, M, L Cook and D Steane. 2004. Radiation of the Australian flora: what can comparisons of molecular phylogenies across multiple taxa tell us about the evolution of diversity in present–day communities? Philos. Trans. R. Soc. Lond. B Biol. Sci. 359(1450): 1551-1571.

Crisp, MD, S Laffan, HP Linder and A Monro. 2001. Endemism in the Australian flora. J. Biogeogr. 28(2): 183-198.

Dawe, AL and DL Nuss. 2001. Hypoviruses and chestnut blight: exploiting viruses to understand and modulate fungal pathogenesis. Annu. Rev. Genet. 35(1): 1-29.

Dearnaley, JDW. 2007. Further advances in orchid mycorrhizal research. Mycorrhiza 17(6): 475-486.

Dieryck, B, G Otto, D Doucet, A Legrève, P Delfosse and C Bragard. 2009. Seed, soil and vegetative transmission contribute to the spread of pecluviruses in Western Africa and the Indian sub-continent. Virus Res. 141(2): 184-189.

Dressler, RL 1981. The Orchids, Natural History and Classification. Cambridge: Harvard University Press.

133

Eigenbrode, SD, H Ding, P Shiel and PH Berger. 2002. Volatiles from potato plants infected with potato leafroll virus attract and arrest the virus vector, Myzus persicae (Homoptera: Aphididae). Proc. R. Soc. Lond. B Biol. Sci. 269(1490): 455-460.

Elena, SF, S Bedhomme, P Carrasco, JM Cuevas, F De la Iglesia, G Lafforgue, J Lalic, À Pròsper, N Tromas and MP Zwart. 2011. The evolutionary genetics of emerging plant RNA viruses. Mol. Plant Microbe Interact. 24(3): 287-293.

Elias, KS and PJ Cotty. 1996. Incidence and stability of infection by double-stranded RNA genetic elements in Aspergillus section flavi and effects on aflatoxigenicity. Can. J. Bot. 74(5): 716-725.

Feldman, TS, MR Morsy and MJ Roossinck. 2012. Are communities of microbial symbionts more diverse than communities of macrobial hosts? Fungal Biol. 116(4): 465-477.

Frakes, L. 1999. Evolution of Australian Environments. Flora of Australia Volume 1. AE Orchard, 163-203. Australia: Australian Biological Resources Study/CSIRO Publishing.

Fraser, RSS. 1998. Introduction to Classical Crossprotection. Plant Virology Protocols: From Virus Isolation to Transgenic Resistance. GD Foster and S Taylor, 13-24. Totowa: Humana Press.

Fulton, RW. 1986. Practices and precautions in the use of cross protection for plant virus disease control. Annu. Rev. Phytopathol. 24(1): 67-81.

Gibbs, A and K Ohshima. 2010. Potyviruses and the digital revolution. Annu. Rev. Phytopathol. 48: 205-223.

Govaerts, R, P Bernet, K Kratochvil, G Gerlach, G Carr, P Alrich, AM Pridgeon, J Pfahl, MA Campacci, DH Baptista, H Tigges, J Shaw, P Cribb, A George, K Kreuz and J Wood. 2011. World Checklist of Orchidaceae. http://apps.kew.org/wcsp/incfamilies.do.

Guy, PL and AJ Gibbs. 1985. Further studies on turnip yellow mosaic tymovirus isolates from an endemic Australian Cardamine. Plant Pathology 34(4): 532-544.

Hammond, J and RH Lawson. 1988. A strain of bean yellow mosaic virus is aphid- transmitted from orchid. Acta Horticulturae 234: 365-370.

134

Hily, JM, N Poulicard, MÁ Mora, I Pagán and F García‐Arenal. 2016. Environment and host genotype determine the outcome of a plant–virus interaction: from antagonism to mutualism. New Phytol. 209(2): 812-822.

Hoffman, N and A Brown 2011. Orchids of South-west Australia. Nedlands: University of Western Australia Press.

Hopper, SD. 1979. Biogeographical aspects of speciation in the Southwest Australian Flora. Annu. Rev. Ecol. Evol. Syst. 10: 399-422.

Hopper, SD. 2009. OCBIL theory: towards an integrated understanding of the evolution, ecology and conservation of biodiversity on old, climatically buffered, infertile landscapes. Plant Soil 322(1-2): 49-86.

Hopper, SD and P Gioia. 2004. The Southwest Australian Floristic Region: evolution and conservation of a global hot spot of biodiversity. Annu. Rev. Ecol. Evol. Syst. 35: 623-650.

Hu, JS, S Ferreira, M Wang and MQ Xu. 1993. Detection of Cymbidium mosaic virus, Odontoglossum ringspot virus, Tomato spotted wilt virus, and potyviruses infecting orchids in Hawaii. Plant Disease 77(5): 464-468.

Ihrmark, K, H Johannesson, E Stenström and J Stenlid. 2002. Transmission of double- stranded RNA in Heterobasidion annosum. Fungal Genet. Biol. 36(2): 147-154.

Ingwell, LL, SD Eigenbrode and NA Bosque-Pérez. 2012. Plant viruses alter insect behavior to enhance their spread. Sci. Rep. 2: 578.

IUCN. 2013. The IUCN Red List of Threatened Species. Version 2013.2. http://www.iucnredlist.org.

IUCN/SSC Orchid Specialist Group, 1996. Orchids - Status Survey and Conservation Action Plan. E Hágsater and V Dumont. Gland and Cambridge: IUCN.

Jalal, JS. 2012. Status, threats and conservation strategies for orchids of western Himalaya, India. J. Threat. Taxa 4(15): 3401-3409.

Jersáková, J, SD Johnson and P Kindlmann. 2006. Mechanisms and evolution of deceptive pollination in orchids. Biol. Rev. 81(2): 219-235.

135

Johansen, E, MC Edwards and RO Hampton. 1994. Seed transmission of viruses: current perspectives. Annu. Rev. Phytopathol. 32(1): 363-386.

Jones, DR and RHA Baker. 2007. Introductions of non-native plant pathogens into Great Britain, 1970–2004. Plant Pathology 56(5): 891-910.

Kawakami, K, S Fuji and K Miyoshi. 2007. Endangered wild populations of endemic Calanthe orchids on an isolated Japanese island tested for viruses. Aust. J. Bot. 55(8): 831-836.

Keesing, F, LK Belden, P Daszak, A Dobson, CD Harvell, RD Holt, P Hudson, A Jolles, KE Jones and CE Mitchell. 2010. Impacts of biodiversity on the emergence and transmission of infectious diseases. Nature 468(7324): 647-652.

Kehoe, MA, BA Coutts, BJ Buirchell and RAC Jones. 2014. Hardenbergia mosaic virus: crossing the barrier between native and introduced plant species. Virus Res. 184: 87-92.

Kochert, G, T Halward, WD Branch and CE Simpson. 1991. RFLP variability in peanut (Arachis hypogaea L.) cultivars and wild species. Theor. Appl. Genet. 81(5): 565-570.

Koopowitz, H, PS Lavarack and KW Dixon. 2003. The Nature of Threats to Orchid Conservation. Orchid Conservation. KW Dixon, SP Kell, RL Barrett and PJ Cribb, 25-42. Kota Kinabalu, Sabah: Natural History Publications.

Kores, PJ, M Molvray, PH Weston, SD Hopper, AP Brown, KM Cameron and MW Chase. 2001. A phylogenetic analysis of Diurideae (Orchidaceae) based on plastid DNA sequence data. Am. J. Bot. 88(10): 1903-1914.

Kreuze, JF, A Perez, M Untiveros, D Quispe, S Fuentes, I Barker and R Simon. 2009. Complete viral genome sequence and discovery of novel viruses by deep sequencing of small RNAs: a generic method for diagnosis, discovery and sequencing of viruses. Virology 388(1): 1-7.

Li, H, C Zhang, H Luo, MGK Jones, K Sivasithamparam, SH Koh, JWL Ong and SJ Wylie. 2016. Yellow tailflower mild mottle virus and Pelargonium zonate spot virus co-infect a wild plant of red‐striped tailflower in Australia. Plant Pathology 65(3): 503-509.

Liu, YC, D Linder‐Basso, BI Hillman, S Kaneko and MG Milgroom. 2003. Evidence for interspecies transmission of viruses in natural populations of filamentous fungi in the genus Cryphonectria. Mol. Ecol. 12(6): 1619-1628.

136

Luo, H, SJ Wylie, B Coutts, RAC Jones and MGK Jones. 2011. A virus of an isolated indigenous flora spreads naturally to an introduced crop species. Ann. Appl. Biol. 159(3): 339-347.

Maccarone, LD, MJ Barbetti, K Sivasithamparam and RAC Jones. 2010a. Molecular genetic characterization of Olpidium virulentus isolates associated with big-vein diseased lettuce plants. Plant Disease 94(5): 563-569.

Maccarone, LD, MJ Barbetti, K Sivasithamparam and RAC Jones. 2010b. Comparison of the coat protein genes of Mirafiori lettuce big-vein virus isolates from Australia with those of isolates from other continents. Arch. Virol. 155(9): 1519-1522.

Maccarone, LD, MJ Barbetti, K Sivasithamparam and RAC Jones. 2010c. Comparison of the coat protein genes of Lettuce big-vein associated virus isolates from Australia with those of isolates from other continents. Arch. Virol. 155(5): 765- 770.

MacClement, WD and MG Richards. 1956. Virus in wild plants. Can. J. Bot. 34(5): 793-799.

Mackenzie, AM, M Nolan, KJ Wei, MA Clements, D Gowanlock, BJ Wallace and AJ Gibbs. 1998. Ceratobium mosaic potyvirus: another virus from orchids. Arch. Virol. 143(5): 903-914.

Maeda, T, H Kondo, K Misuhata and T Tamada. 1998. Evidence that orchid fleck virus is efficiently transmitted in a persistent manner by the mite Brevipalpus californicus. Proceedings of the 7th International Congress of Plant Pathology, Edinburgh.

Mardis, ER. 2008. Next-generation DNA sequencing methods. Annu. Rev. Genomics Hum. Genet. 9: 387-402.

Márquez, LM, RS Redman, RJ Rodriguez and MJ Roossinck. 2007. A virus in a fungus in a plant: three-way symbiosis required for thermal tolerance. Science 315(5811): 513-515.

Martins, MK, MC Furlaneto, DR Sosa-Gomez, MR Faria and MHP Fungaro. 1999. Double-stranded RNA in the entomopathogenic fungus Metarhizium flavoviride. Curr. Genet. 36(1-2): 94-97.

Marzano, S-YL and LL Domier. 2016. Novel mycoviruses discovered from metatranscriptomics survey of soybean phyllosphere phytobiomes. Virus Res. 213: 332-342.

137

Mauck, KE, CM De Moraes and MC Mescher. 2010. Deceptive chemical signals induced by a plant virus attract insect vectors to inferior hosts. Proc. Natl. Acad. Sci. U.S.A. 107(8): 3600-3605.

McKirdy, SJ, BA Coutts and RAC Jones. 1994. Occurrence of bean yellow mosaic virus in subterranean clover pastures and perennial native legumes. Crop Pasture Sci. 45(1): 183-194.

McLoughlin, S. 2001. The breakup history of Gondwana and its impact on pre- Cenozoic floristic provincialism. Aust. J. Bot. 49(3): 271-300.

Milgroom, MG and BI Hillman. 2011. The Ecology and Evolution of Fungal Viruses. Studies in Viral Ecology: Microbial and Botanical Host Systems. CJ Hurst, Vol 1, 217-253. New Jersey: Wiley-Blackwell.

Myers, N, RA Mittermeier, CG Mittermeier, GAB Da Fonseca and J Kent. 2000. Biodiversity hotspots for conservation priorities. Nature 403(6772): 853-858.

Nguyen, HD, Y Tomitaka, SY Ho, S Duchêne, H-J Vetten, D Lesemann, JA Walsh, AJ Gibbs and K Ohshima. 2013. Turnip mosaic potyvirus probably first spread to Eurasian brassica crops from wild orchids about 1000 years ago. PLoS One 8(2): e55336.

Peakall, R. 1990. Responses of male Zaspilothynnus tribobatus turner wasps to females and the sexually deceptive orchid it pollinates. Funct. Ecol. 4(2): 159-167.

Peakall, R and AJ Beattie. 1996. Ecological and genetic consequences of pollination by sexual deception in the orchid Caladenia tentactulata. Evolution 50(6): 2207-2220.

138

Peever, TL, Y-C Liu and MG Milgroom. 1997. Diversity of hypoviruses and other double-stranded RNAs in Cryphonectria parasitica in North America. Phytopathology 87(10): 1026-1033.

Phillips, RD, MD Barrett, KW Dixon and SD Hopper. 2011. Do mycorrhizal symbioses cause rarity in orchids? J. Ecol. 99(3): 858-869.

Phillips, RD, R Peakall, MF Hutchinson, CC Linde, T Xu, KW Dixon and SD Hopper. 2014. Specialized ecological interactions and plant species rarity: the role of pollinators and mycorrhizal fungi across multiple spatial scales. Biol. Cons. 169: 285- 295.

Prendeville, HR, X Ye, TJ Morris and D Pilson. 2012. Virus infections in wild plant populations are both frequent and often unapparent. Am. J. Bot. 99(6): 1033-1042.

Quail, MA, M Smith, P Coupland, TD Otto, SR Harris, TR Connor, A Bertoni, HP Swerdlow and Y Gu. 2012. A tale of three next generation sequencing platforms: comparison of Ion Torrent, Pacific Biosciences and Illumina MiSeq sequencers. BMC Genomics 13: 341.

Rajabaskar, D, NA Bosque-Pérez and SD Eigenbrode. 2014. Preference by a virus vector for infected plants is reversed after virus acquisition. Virus Res. 186: 32-37.

Ramsay, RR, KW Dixon and K Sivasithamparam. 1986. Patterns of infection and endophytes associated with Western Australian orchids. Lindleyana 1(3): 203-214.

Rasmussen, HN 1995. Terrestrial Orchids, From Seed to Mycotrophic Plant. Cambridge: Cambrudge University Press.

Reddy, AS, HA Hobbs, P Delfosse, AK Murthy and DVR Reddy. 1998. Seed transmission of Indian peanut clump virus (IPCV) in peanut and millets. Plant Disease 82(3): 343-346.

Romo, M, A Leuchtmann, B García and I Zabalgogeazcoa. 2007. A totivirus infecting the mutualistic fungal endophyte Epichloë festucae. Virus Res. 124(1): 38-43.

Roossinck, MJ. 2010. Lifestyles of plant viruses. Philos. Trans. R. Soc. Lond. B Biol. Sci. 365(1548): 1899-1905.

Roossinck, MJ. 2015. Metagenomics of plant and fungal viruses reveals an abundance of persistent lifestyles. Front. Microbiol. 5: 767.

Roossinck, MJ and F García-Arenal. 2015. Ecosystem simplification, biodiversity loss and plant virus emergence. Curr. Opin. Virol. 10: 56-62.

139

Roossinck, MJ, P Saha, GB Wiley, J Quan, JD White, H Lai, F ChavarrÍA, G Shen and BA Roe. 2010. Ecogenomics: using massively parallel pyrosequencing to understand virus ecology. Mol. Ecol. 19(S1): 81-88.

Russell, D. 2004. Aboriginal-Makassan interactions in the eighteenth and nineteenth centuries in northern Australia and contemporary sea rights claims. Australian Aboriginal Studies(1): 3-17.

Savolainen, P, T Leitner, AN Wilton, E Matisoo-Smith and J Lundeberg. 2004. A detailed picture of the origin of the Australian dingo, obtained from the study of mitochondrial DNA. Proc. Natl. Acad. Sci. U.S.A. 101(33): 12387-12390.

Seijo, G, GI Lavia, A Fernández, A Krapovickas, DA Ducasse, DJ Bertioli and EA Moscone. 2007. Genomic relationships between the cultivated peanut (Arachis hypogaea, Leguminosae) and its close relatives revealed by double GISH. Am. J. Bot. 94(12): 1963-1971.

Shapira, R and DL Nuss. 1991. Gene expression by a hypovirulence-associated virus of the chestnut blight fungus involves two papain-like protease activities. J. Biol. Chem. 266(29): 19419-19425.

Shendure, J and H Ji. 2008. Next-generation DNA sequencing. Nat. Biotechnol. 26(10): 1135-1145.

Sherpa, AR, V Hallan, P Pathak and AA Zaidi. 2006. Characterization of the coat protein gene of Cymbidium mosaic virus isolates from India. J. Phytopathol. 154(5): 275-280.

Singh, MK, AR Sherpa, V Hallan and AA Zaidi. 2007. A potyvirus in Cymbidium spp. in northern India. Australas. Plant Dis. Notes 2(1): 11-13.

Smith, SE and DJ Read 2010. Mycorrhizal Symbiosis. London: Academic press.

State of Western Australia. 2015. Western Australian Government Gazette, No 166. Western Australia, State Law Publisher: 4525-4531.

Swarts, ND and KW Dixon. 2009. Terrestrial orchid conservation in the age of extinction. Ann. Bot. 104(3): 543-556.

Tremblay, RL, JD Ackerman, JK Zimmerman and RN Calvo. 2005. Variation in sexual reproduction in orchids and its evolutionary consequences: a spasmodic journey to diversification. Biol. J. Linnean Soc. 84(1): 1-54.

140

Vainio, EJ, J Hakanpää, Y-C Dai, E Hansen, K Korhonen and J Hantula. 2011. Species of Heterobasidion host a diverse pool of partitiviruses with global distribution and interspecies transmission. Fungal Biol. 115(12): 1234-1243.

Vainio, EJ, MM Müller, K Korhonen, T Piri and J Hantula. 2015. Viruses accumulate in aging infection centers of a fungal forest pathogen. ISME J. 9(2): 497-507. van der Cingel, NA 2001. Atlas of Orchid Pollination: Orchids of America, Africa, Asia and Australia. Rotterdam: A.A.Balkema.

Vincent, SJ, BA Coutts and RAC Jones. 2014. Effects of introduced and indigenous viruses on native plants: exploring their disease causing potential at the agro- ecological interface. PLoS One 9(3): e91224.

Voth, PD, L Mairura, BE Lockhart and G May. 2006. Phylogeography of Ustilago maydis virus H1 in the USA and Mexico. J. Gen. Virol. 87(11): 3433-3441.

Warcup, J. 1988. Mycorrhizal associations of isolates of Sebacina vermifera. New Phytol. 110(2): 227-231.

Warcup, JH. 1981. The mycorrhizal relationships of Australian orchids. New Phytol. 87(2): 371-381.

Webster, CG, BA Coutts, RAC Jones, MGK Jones and SJ Wylie. 2007. Virus impact at the interface of an ancient ecosystem and a recent agroecosystem: studies on three legume-infecting potyviruses in the southwest Australian floristic region. Plant Pathology 56(5): 729-742.

Western Australian Herbarium. 2015. FloraBase-the Western Australian Flora. Department of Parks and Wildlife. http://florabase.dpaw.wa.gov.au/.

Williams, KJ, A Ford, DF Rosauer, N De Silva, R Mittermeier, C Bruce, FW Larsen and C Margules. 2011. Forests of East Australia: The 35th Biodiversity Hotspot. Biodiversity Hotspots. FE Zachos and JC Habel, 295-310. Germany: Springer-Verlag.

Wylie, SJ, H Li and MGK Jones. 2013b. Donkey orchid symptomless virus: a viral ‘platypus’ from Australian terrestrial orchids. PLoS One 8(11): e79587.

141

Xu, JT and C Mu. 1990. The relation between growth of Gastrodia elata protocorms and fungi. Acta Botanica Sinica 32(1): 26-31.

Xu, P, F Chen, JP Mannas, T Feldman, LW Sumner and MJ Roossinck. 2008. Virus infection improves drought tolerance. New Phytol. 180(4): 911-921.

Yamamura, N. 1996. Evolution of mutualistic symbiosis: a differential equation model. Res. Popul. Ecol. 38(2): 211-218.

Yuen, CKKH, H Kamemoto and M Ishii. 1979. Transmission of Cymbidium mosaic virus through seed propagation in Dendrobium. Amer. Orchid Soc. Bull. 48(12): 1245-1247.

Zelmer, CD, L Cuthbertson and RS Currah. 1996. Fungi associated with terrestrial orchid mycorrhizas, seeds and protocorms. Mycoscience 37(4): 439-448.

Zettler, FW, NJ Ko, GC Wisler, MS Elliott and SM Wong. 1990. Viruses of orchids and their control. Plant Disease 74(9): 621-626.

142