WO2019018895A1

WO2019018895A1 - Horizontal transfer of fungal genes into plants

Info

Publication number: WO2019018895A1
Application number: PCT/AU2018/050777
Authority: WO
Inventors: Hiroshi Shinozuka; Maiko Shinozuka; Inoka Kumari HETTIARACHCHIGE; German Carlos Spangenberg; Noel COGAN; John White Forster; Timothy Ivor Sawbridge; Benjamin Cocks
Original assignee: Agriculture Victoria Services Pty Ltd
Priority date: 2017-07-28
Filing date: 2018-07-27
Publication date: 2019-01-31
Also published as: AU2018305727A1

Abstract

The present invention relates to methods of creating or enhancing symbiotic relationships between plants and symbionts, in particular symbionts carrying genes encoding fungal cell wall degrading enzymes. The present invention also relates to methods for selecting plants for symbiosis with symbionts and methods of modifying pathogen resistance in plants. The present invention also relates to new organisms and symbiota developed by the methods of the invention and to related nucleic acids, polypeptides and constructs including vectors.

Description

HORIZONTAL TRANSFER OF FUNGAL GENES INTO PLANTS

Field of the Invention The present invention relates to methods of selecting and breeding organisms, in particular organisms which exhibit symbiotic behaviour with symbionts such as fungal endophytes or epiphytes or bacterial microbiome in plants, methods of creating or enhancing symbiotic relationships between plants and symbionts, and new organisms and symbiota developed thereby.

The invention also relates to nucleic acids involved in the methods, vectors including the nucleic acids, plant cells, plants, seeds and other plant parts transformed with the nucleic acids and vectors, and methods of using the nucleic acids and vectors. Background of the Invention

Important plants, including forage grasses, legumes, trees, shrubs, and vines are commonly found in association with endophytes including fungi, bacteria, viruses and microbes. Both beneficial and detrimental horticultural and agronomic properties result from such associations, including improved tolerance to water and nutrient stress and resistance to insect pests. For example, in grasses, insect resistance may be provided by specific metabolites produced by the endophyte, in particular loline alkaloids and peramine. Other metabolites produced by the endophyte, for example lolitrems and ergot alkaloids, may be toxic to grazing animals and reduce herbivore feeding. Considerable variation is known to exist in the metabolite profile of endophytes. Endophyte strains that lack either or both of the animal toxins have been introduced into commercial cultivars.

However, there remains a need for methods of using organisms which exhibit symbiotic behaviour with endophytes. Difficulties in artificially breeding of these symbiota limit their usefulness. For example, many of the endophytes known to be beneficial to pasture-based agriculture exhibit low inoculation frequencies and are less stable in elite germplasm.

Moreover, in traditional breeding techniques, for example in forage grasses such as perennial ryegrass and tall fescue, grass varieties are bred using classic cross-breeding techniques and grass genotypes are selected for their superior characteristics, after monitoring their performance over a period of multiple years. The selected grass genotypes that form the experimental variety are then inoculated with a single endophyte and the resulting grass-endophyte associations are evaluated for any favourable characteristics such as insect resistance. The individual experimental synthetic varieties deploying a single endophyte in them are then evaluated for agronomic performance and resulting animal performance by grazing animals over a period of years. This evaluation process may reveal that the single endophyte being deployed in the different experimental synthetic varieties may not show vegetative and/or intergenerational stability in some of these varieties or the desired alkaloid profile conferred by the single endophyte may vary between different synthetic varieties failing to confer appropriate levels of insect resistance or causing animal toxicoses. It would be a significant development in the art if this time- consuming process could be accelerated or otherwise improved.

The Poeae tribe of the Poaceae family is composed of a range of cool-season turf and forage grass species, including those of sub-tribes Loliinae and Dactylidinae. Perennial ryegrass {Lolium perenne L; sub-tribe Loliinae) is one of the most important pasture crop species for the dairy industry, and it has consequently been a primary target for improvement using molecular biology and genetic technologies. Asexual fungal endophyte species of the genus Epichloe (syn. Neotyphodium) are symbionts of species belonging to the Poeae tribe genera Lolium and Festuca.

The fungal endophyte species rely on the plant host for nutrition, reproduction, and protection from abiotic and biotic stress. Benefits to the host plant include enhanced competitive abilities, tolerance to pathogens, and resistance to animal and insect herbivory. Due to its agronomic importance, the molecular basis of the symbiosis has been investigated, and deterrence of insect herbivory is largely due to the production of bioactive alkaloids, as well as makes caterpillars floppy-like {mcf-like) gene products.

Horizontal gene transfer (HGT) has been a source of evolutionary novelty in both prokaryotes and eukaryotes. In flowering plant species, organelle genomes have served as both donors and recipients of gene transfer events. In contrast to organellar genes, transfer of nuclear genes to angiosperms appears to have been rare, and has been confined to date to genes originating from prokaryotes or other plant species such as green algae, mosses and other angiosperms. A previous systematic in silico study, from investigation of four completely sequenced angiosperm genomes (those of Arabidopsis thaliana L , rice [Oryza sativa L], sorghum [Sorghum bicolor L], and poplar [Populus trichocarpa Torr. & A.Gray ex. Hook.]), found no evidence for HGT from fungal species, despite two and three highly reliable events for moss and lycophyte lineages, respectively [Richards, T. A. et a/.]. It was consequently concluded that gene transfer from fungi into angiosperms must be exceedingly infrequent. It is accordingly an object of the present invention to overcome, or at least alleviate, one or more of the difficulties or deficiencies associated with the prior art.

Summary of the Invention The applicant has found evidence for ancient horizontal transfer of a number of fungal genes into angiosperm lineages. In particular the applicant has found evidence for multiple events of horizontal gene transfer (HGT) from ancestral species of fungal endophytes (of the Epichloe genus) into grass species. In perennial ryegrass {Lolium perenne L), three fungus-originating genes were identified. One was specific to the Poeae subtribes Loliinae, Festucae and Dactylidinae. The gene has an enzymatic activity for degradation of components commonly found in cell walls of fungi. The gene may be previously considered to be specific to fungi. For example, the gene may be a β-1 , 6-glucanase gene. The β-1 ,6- glucanase genes may be isolated from fungal species such as Epichloe festucae, for example strains such as Leuchtm., Schardl and M.R. Siegel (the sexual counterpart to the perennial ryegrass endophyte), Hypocrea lixii Pat., and Trichoderma harzianum Rifai. Another gene was specific to the Loliinae subtribe, and the other was conserved within the Triticeae and Poeae tribes, including common wheat {Triticum aestivum L), barley (Hordeum vulgare L), and oat (Avena sativa L). This is the first report of horizontal transfer of a nuclear gene from a taxonomically distant eukaryote to modern flowering plants and provides evidence for a novel adaptation mechanism in angiosperms.

In a first aspect, the present invention provides a substantially purified or isolated nucleic acid or nucleic acid fragment encoding a fungal cell wall degrading enzyme, preferably a glucanase, more preferably a β-1 , 6-glucanase (BGNL), wherein the nucleic acid or nucleic acid fragment is isolated from a plant species.

In a preferred embodiment, the plant is a Loliinae or Dactylidnae species. In a more preferred embodiment the plant is from a Lolium, Festuca or Dactylis species. The Lolium or Festuca species may be of any suitable type, including Italian or annual ryegrass, perennial ryegrass, tall fescue, meadow fescue and red fescue. Preferably the ryegrass or fescue species is a Lolium species such as Lolium perenne or Lolium arundinaceum which is otherwise known as Festuca arundinacea.

The Dactylis species may be of any suitable type. Preferably the Dactylis species is a Dactylis marina or Dactylis glomerata.

In a second aspect, the present invention provides a substantially purified or isolated nucleic acid or nucleic acid fragment of a domain-of-unknown-function (DUF) gene, wherein the nucleic acid or nucleic acid fragment is isolated from a plant species.

In a preferred embodiment, the plant is a Loliinae species. In a more preferred embodiment the plant is from a Lolium or Festuca species.

The Lolium or Festuca species may be of any suitable type, including Italian or annual ryegrass, perennial ryegrass, tall fescue, meadow fescue and red fescue. Preferably the ryegrass or fescue species is a Lolium species such as Lolium perenne or Lolium multiflorum which is otherwise known as Festuca perennis.

In a third aspect, the present invention provides a substantially purified or isolated nucleic acid or nucleic acid fragment encoding a fungal transcriptional regulatory-like (FTR) protein, wherein the nucleic acid or nucleic acid fragment is isolated from a plant species.

In a preferred embodiment, the plant is a Triticeae and Poeae species. In a more preferred embodiment the plant is from a Triticeae or Aveneae species.

The Triticeae or Aveneae species may be of any suitable type, including barley, wheat and oat. Preferably the ryegrass or fescue species is a Triticum, Hordeum or Avena species such as Triticum aestivum, Hordeum vulgare and Avena sativa. Thus, the present invention provides a substantially purified or isolated nucleic acid or nucleic acid fragment:

encoding a fungal cell wall degrading enzyme;

of a DUF gene; or

encoding a FTR protein,

wherein the nucleic acid or nucleic acid fragment is isolated from a plant species. In other words, the present invention provides a substantially purified or isolated nucleic acid or nucleic acid fragment from a plant species, the nucleic acid or nucleic acid fragment being horizontally transferred from a fungal species, wherein the nucleic acid or nucleic acid fragment encodes a fungal cell wall degrading enzyme, is a DUF gene, or encodes a FTR protein.

By 'nucleic acid' is meant a chain of nucleotides capable of carrying genetic information. The term generally refers to genes or functionally active fragments or variants thereof and or other sequences in the genome of the organism that influence its phenotype. The term 'nucleic acid' includes DNA (such as cDNA or genomic DNA) and RNA (such as mRNA or microRNA) that is single- or double-stranded, optionally containing synthetic, non-natural or altered nucleotide bases, synthetic nucleic acids and combinations thereof.

The nucleic acid or nucleic acid fragment may be of any suitable type and includes DNA (such as cDNA or genomic DNA) and RNA (such as mRNA) that is single- or double- stranded, optionally containing synthetic, non-natural or altered nucleotide bases, and combinations thereof.

By 'substantially purified' is meant that the nucleic acid or nucleic acid fragment is free of the genes, which, in the naturally-occurring genome of the organism from which the nucleic acid or nucleic acid fragment of the invention is derived, flank the nucleic acid or nucleic acid fragment. The term therefore includes, for example, a nucleic acid or nucleic acid fragment which is incorporated into a vector; into an autonomously replicating plasmid or virus; or into the genomic DNA of a prokaryote or eukaryote; or which exists as a separate molecule (e.g. a cDNA or a genomic or cDNA fragment produced by PCR or restriction endonuclease digestion) independent of other sequences. It also includes a nucleic acid or nucleic acid fragment which is part of a hybrid gene. Preferably, the substantially purified nucleic acid or nucleic acid fragment is at least approximately 90% pure, more preferably at least approximately 95% pure, even more preferably at least approximately 98% pure.

The term "isolated" means that the material is removed from its original environment (e.g. the natural environment if it is naturally occurring). For example, a naturally occurring nucleic acid present in a living plant is not isolated, but the same nucleic acid separated from some or all of the coexisting materials in the natural system, is isolated. Such nucleic acids could be part of a vector and/or such nucleic acids could be part of a composition, and still be isolated in that such a vector or composition is not part of its natural environment.

In preferred embodiments, the substantially purified or isolated nucleic acid or nucleic acid fragment encoding a fungal cell wall degrading enzyme includes a nucleotide sequence selected from the group consisting of the sequences shown in Sequence ID Nos: 1 to 4; and functionally active fragments and variants thereof.

Also in preferred embodiments, the substantially purified or isolated nucleic acid or nucleic acid fragment of the DUF gene includes a nucleotide sequence selected from the group consisting of the sequence shown in Sequence ID No: 28 and sequences encoding the polypeptide shown in Sequence ID No: 29; and functionally active fragments and variants thereof. Also in preferred embodiments, the substantially purified or isolated nucleic acid or nucleic acid fragment encoding a FTR protein includes a nucleotide sequence selected from the group consisting of the sequence shown in Sequence ID No: 30 and sequences encoding the polypeptide shown in Sequence ID No: 31 ; and functionally active fragments and variants thereof.

The substantially purified or isolated nucleic acid or nucleic acid of the present invention also includes sequences complementary or antisense to a sequence encoding a fungal cell wall degrading enzyme, to a DUF gene or to a sequence encoding a FTR protein, and preferably sequences selected from the group consisting of sequences complementary or antisense to Sequence ID Nos: 1-4, 28 and 30; or complementary or antisense to the sequences encoding the polypeptides shown in Sequence ID Nos: 29 and 31.

By "functionally active" is meant that the fragment or variant (such as an analogue, derivative or mutant) is capable of modifying pathogen resistance in a plant, or creating or enhancing a symbiotic relationship between a plant and a symbiont. Such variants include naturally occurring allelic variants and non-naturally occurring variants. Additions, deletions, substitutions and derivatizations of one or more of the nucleotides are contemplated so long as the modifications do not result in loss of functional activity of the fragment or variant. Preferably the functionally active fragment or variant has at least approximately 80% identity to the relevant part of the above-mentioned sequence to which the fragment or variant corresponds, more preferably at least approximately 90% identity, even more preferably at least approximately 95% identity, most preferably at least approximately 98% identity. Such functionally active variants and fragments include, for example, those having conservative nucleic acid changes. By 'conservative nucleic acid changes' is meant nucleic acid substitutions that result in conservation of the amino acid in the encoded protein, due to the degeneracy of the genetic code. Such functionally active variants and fragments also include, for example, those having nucleic acid changes which result in conservative amino acid substitutions of one or more residues in the corresponding amino acid sequence. By 'conservative amino acid substitutions' is meant the substitution of an amino acid by another one of the same class, the classes being as follows:

Nonpolar: Ala, Val, Leu, lie, Pro, Met Phe, Trp

Uncharged polar: Gly, Ser, Thr, Cys, Tyr, Asn, Gin

Acidic: Asp, Glu

Basic: Lys, Arg, His

Other conservative amino acid substitutions may also be made as follows:

Aromatic: Phe, Tyr, His

Proton Donor: Asn, Gin, Lys, Arg, His, Trp

Proton Acceptor: Glu, Asp, Thr, Ser, Tyr, Asn, Gin

Preferably the fragment has a size of at least 20 nucleotides, more preferably at least 50 nucleotides, more preferably at least 100 nucleotides, more preferably at least 200 nucleotides, more preferably at least 500 nucleotides.

In a fourth aspect of the present invention there is provided a genetic construct or a vector including a nucleic acid or nucleic acid fragment according to the present invention. In other words, the present invention provides a genetic construct or a vector including a nucleic acid or nucleic acid fragment according to the present invention, or sequences complementary or antisense thereto. Preferably the nucleic acid or nucleic acid fragment has a sequence selected from the group consisting of Sequence ID Nos: 1-4, 28, 30, and nucleic acids encoding the polypeptides shown in Sequence ID Nos: 29 and 31. In a preferred embodiment of this aspect of the invention, the vector may include a regulatory element such as a promoter, a nucleic acid or nucleic acid fragment according to the present invention and a terminator; said regulatory element, nucleic acid or nucleic acid fragment and terminator being operatively linked.

By 'genetic construct' is meant a recombinant nucleic acid molecule.

By a 'vector' is meant a genetic construct used to transfer genetic material to a target cell. By Operatively linked' is meant that the nucleic acid(s) and a regulatory sequence, such as a promoter, are linked in such a way as to permit expression of said nucleic acid under appropriate conditions, for example when appropriate molecules such as transcriptional activator proteins are bound to the regulatory sequence. Preferably an operatively linked promoter is upstream of the associated nucleic acid.

The vector may be of any suitable type and may be viral or non-viral. The vector may be an expression vector. Such vectors include chromosomal, non-chromosomal and synthetic nucleic acid sequences, e.g. derivatives of plant viruses; bacterial plasmids; derivatives of the Ti plasmid from Agrobacterium tumefaciens; derivatives of the Ri plasmid from Agrobacterium rhizogenes; phage DNA; yeast artificial chromosomes; bacterial artificial chromosomes; binary bacterial artificial chromosomes; vectors derived from combinations of plasmids and phage DNA. However, any other vector may be used as long as it is replicable or integrative or viable in the plant cell. The regulatory element and terminator may be of any suitable type and may be endogenous to the target plant cell or may be exogenous, provided that they are functional in the target plant cell.

Preferably the regulatory element is a promoter. A variety of promoters which may be employed in the vectors of the present invention are well known to those skilled in the art. Factors influencing the choice of promoter include the desired tissue specificity of the vector, and whether constitutive or inducible expression is desired and the nature of the plant cell to be transformed (e.g. monocotyledon or dicotyledon). Particularly suitable promoters include the Cauliflower Mosaic Virus 35S (CaMV 35S) promoter, the maize Ubiquitin promoter, the rice Actin promoter, and ryegrass endogenous OMT, 4CL, CCR or CAD promoters. A variety of terminators which may be employed in the vectors of the present invention are also well known to those skilled in the art. The terminator may be from the same gene as the promoter sequence or a different gene. Particularly suitable terminators are polyadenylation signals, such as the CaMV 35S polyA and other terminators from the nopaline synthase (nos) and the octopine synthase (ocs) genes.

The vector, in addition to the regulatory element, the nucleic acid or nucleic acid fragment of the present invention and the terminator, may include further elements necessary for expression of the nucleic acid or nucleic acid fragment, in different combinations, for example vector backbone, origin of replication (ori), multiple cloning sites, spacer sequences, enhancers, introns (such as the maize Ubiquitin Ubi intron), antibiotic resistance genes and other selectable marker genes [such as the neomycin phosphotransferase (npt2) gene, the hygromycin phosphotransferase (hph) gene, the phosphinothricin acetyltransferase {bar or pat) gene], and reporter genes (such as beta- glucuronidase (GUS) gene {gusA)]. The vector may also contain a ribosome binding site for translation initiation. The vector may also include appropriate sequences for amplifying expression. As an alternative to use of a selectable marker gene to provide a phenotypic trait for selection of transformed host cells, the presence of the vector in transformed cells may be determined by other techniques well known in the art, such as PCR (polymerase chain reaction), Southern blot hybridisation analysis, histochemical GUS assays, northern and Western blot hybridisation analyses.

Those skilled in the art will appreciate that the various components of the vector are operatively linked, so as to result in expression of said nucleic acid or nucleic acid fragment. Techniques for operatively linking the components of the vector of the present invention are well known to those skilled in the art. Such techniques include the use of linkers, such as synthetic linkers, for example including one or more restriction enzyme sites.

The nucleic acids and vectors of the present invention may be used to up-regulate or down- regulate expression of corresponding genes in plants. By 'up-regulating' expression of said gene is meant increasing expression of said gene and, as a result, the protein encoded by the gene, in a plant relative to a control plant. By 'down-regulating' expression of said gene is meant decreasing expression of said gene and, as a result, the protein encoded by the gene, in a plant relative to a control plant. The up-regulation or down-regulation may be carried out by methods known to those skilled in the art. For example, a gene may be up-regulated by incorporating additional copies of a sense copy of the gene. A gene may be down-regulated, for example, by incorporating an antisense nucleic acid, a frame-shifted or otherwise modified sense copy of the gene, or a nucleic acid encoding interfering RNA (RNAi).

The up- or down-regulation may be carried out by introducing into said plant an effective amount of a genetic construct including the gene or a modified form thereof, such as an antisense nucleic acid, a frame shifted copy of the gene or a nucleic acid encoding RNAi. The vectors of the present invention may be incorporated into a variety of plants, including monocotyledons (such as grasses from the genera Lolium, Festuca, Dactylis, Cynodon, Bracharia, Paspalum, Panicum, Miscanthus, Pennisetum, Phalaris, and other forage, turf and bioenergy grasses, corn, oat, sugarcane, rice, wheat and barley), dicotyledons (such as Arabidopsis, tobacco, legumes, Alfalfa, oak, Eucalyptus, maple, Populus, canola, soybean and chickpea) and gymnosperms (such as Pinus). In a preferred embodiment, the vectors are used to transform monocotyledons, preferably grass species such as Lolium, Festuca, Dactylis, Cynodon, Bracharia, Paspalum, Panicum, Miscanthus, Pennisetum, Phalaris, and other forage, turf and bioenergy grasses, more preferably a Lolium species such as Lolium perenne or Lolium arundinaceum, including cultivars for forage and turf applications, or a Dactylis species.

Techniques for incorporating the vectors of the present invention into plant cells (for example by transduction, transfection or transformation) are well known to those skilled in the art. Such techniques include Agrobacterium mediated introduction, electroporation to tissues, cells and protoplasts, protoplast fusion, injection into reproductive organs, injection into immature embryos and high velocity projectile introduction to cells, tissues, calli, immature and mature embryos. The choice of technique will depend largely on the type of plant to be transformed. Cells incorporating the vector of the present invention may be selected, as described above, and then cultured in an appropriate medium to regenerate transformed plants, using techniques well known in the art. The culture conditions, such as temperature, pH and the like, will be apparent to the person skilled in the art. The resulting plants may be reproduced, either sexually or asexually, using methods well known in the art, to produce successive generations of transformed plants.

In a fifth aspect of the present invention there is provided a transformed plant cell, plant, plant seed or other plant part, or plant biomass, including digestible biomass such as hay, including, e.g. transformed with, a nucleic acid, genetic construct or vector of the present invention. Preferably the transgenic plant cell, plant, plant seed or other plant part is produced by a method according to the present invention.

The present invention also provides a transgenic plant, plant seed or other plant part, or plant biomass, derived from a plant cell of the present invention and including a nucleic acid, genetic construct or vector of the present invention.

The present invention also provides a transgenic plant, plant seed or other plant, part or plant biomass, derived from a plant of the present invention and including a nucleic acid, genetic construct or vector of the present invention. The nucleic acid, genetic construct or vector of the present invention may be stably integrated into the genome of the plant, plant seed, other plant part or plant biomass.

The plant cell, plant, plant seed or other plant part may be from any suitable species, including monocotyledons (such as grasses from the genera Lolium, Festuca, Dactylis, Cynodon, Bracharia, Paspalum, Panicum, Miscanthus, Pennisetum, Phalaris, and other forage, turf and bioenergy grasses, corn, oat, sugarcane, rice, wheat and barley), dicotyledons (such as Arabidopsis, tobacco, legumes, Alfalfa, oak, Eucalyptus, maple, Populus, canola, soybean and chickpea) and gymnosperms (such as Pinus). In a preferred embodiment the plant cell, plant, plant seed or other plant part may be from a monocotyledon, preferably a grass species, such as Lolium, Festuca, Dactylis, Cynodon, Bracharia, Paspalum, Panicum, Miscanthus, Pennisetum, Phalaris, and other forage, turf and bioenergy grasses, more preferably a Lolium species such as Lolium perenne or Lolium arundinaceum or a Dactylis species. In a sixth aspect of the present invention there is provided a method of modifying pathogen resistance in a plant, said method including introducing into said plant an effective amount of a nucleic acid or nucleic acid fragment, genetic construct and/or a vector according to the present invention.

By "modifying pathogen resistance" is meant modifying a plant's ability to protect itself from, or reduce the extent of, pathogen growth on or in the plant. Such pathogen resistance may subsequently protect the plant from, or reduce the extent of, fungal diseases generated by such fungal pathogens. Fungal diseases may include leaf spot, stalk rot, root rot, dying-off, choke disease and seedling wilt. Modifying pathogen resistance includes creating resistance to a pathogen in a plant, or increasing or decreasing existing levels of resistance in a plant.

In a seventh aspect of the present invention there is provided use of a nucleic acid or nucleic acid fragment according to the present invention, and/or nucleotide sequence information thereof, and/or single nucleotide polymorphisms thereof, as a molecular genetic marker.

More particularly, nucleic acids or nucleic acid fragments according to the present invention, and/or nucleotide sequence information thereof, and/or single nucleotide polymorphisms thereof, may be used as a molecular genetic marker for qualitative trait loci (QTL) tagging, mapping, DNA fingerprinting and in marker assisted selection, and may be used as candidate genes or perfect markers, particularly in ryegrasses and fescues. Even more particularly, nucleic acids or nucleic acid fragments according to the present invention, and/or nucleotide sequence information thereof, may be used as molecular genetic markers in forage and turf grass improvement, e.g. tagging QTLs for dry matter digestibility, herbage quality, mechanical stress tolerance, disease resistance, insect pest resistance, plant stature and leaf and stem colour.

In an eighth aspect of the present invention there is provided a method of creating or enhancing a symbiotic relationship between a plant and a symbiont carrying:

a fungal cell wall degrading enzyme

a DUF gene; or

a FTR protein,

said method including introducing into said plant an effective amount of a nucleic acid or nucleic acid fragment encoding the fungal cell wall degrading enzyme, of the DUF gene or encoding the FTR protein, or a functionally active fragment or variant thereof, wherein the nucleic acid or nucleic acid fragment is isolated from a plant species. In a ninth aspect of the present invention there is provided a method of creating or enhancing a symbiotic relationship between a plant and a symbiont carrying:

a fungal cell wall degrading enzyme;

a DUF gene; or

a FTR protein,

said method including introducing into said plant an effective amount of a nucleic acid or nucleic acid fragment, genetic construct and/or a vector according to the present invention and introducing said symbiont into said plant.

The present invention also provides a plant endophyte symbiota produced by a method of the present invention.

Preferably, the fungal cell wall degrading enzyme is a glucanase, more preferably a BGNL as hereinbefore described.

By "symbiont" is meant one or more organisms that live within the body or cells of another organism, such as fungal endophytes or epiphytes or a bacterial microbiome in plants. By "creating or enhancing a symbiotic relationship" is meant enabling a plant which otherwise would not form a symbiotic relationship with a selected symbiont to form said symbiotic relationship or increasing the stability of the symbiotic relationship. For example, the nucleic acid or nucleic acid fragment, genetic construct and/or a vector according to the present invention may be introduced into plants such as rice, wheat and barley, which do not naturally contain an ortholog of the gene of the present invention, to faciliate establishment of symbiotic relationships between these plants and symbionts.

By "an effective amount" is meant an amount sufficient to result in an identifiable phenotypic change in said plant cells or a plant, plant seed or other plant part derived therefrom. Such amounts can be readily determined by an appropriately skilled person, taking into account the type of plant cell, the route of administration and other relevant factors. Such a person will readily be able to determine a suitable amount and method of administration. See, for example, Maniatis et al, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, the entire disclosure of which is incorporated herein by reference. In a tenth aspect, the present invention provides a method for selecting a plant for symbiosis with a symbiont carrying a fungal cell wall degrading gene, a DUF gene or a FTR gene, said method comprising

a) determining the presence of a fungal cell wall degrading gene, a DUF gene or a

FTR gene in said plant,

b) measuring the level of expression of the gene identified in a), and

c) using the expression level measured in b) to determine if the plant will form a symbiotic relationship with said symbiont. Preferably, the fungal cell wall degrading enzyme is a glucanase, more preferably a BGNL as hereinbefore described.

Techniques for detecting the presence of a specific gene in a plant are well known to those skilled in the art. Such techniques may involve one or more of polymerase chain reaction (PCR), sequencing of PCR products, sequencing of genomic and/or mitochondrial DNA, and performing sequence analysis and comparisons to assess genetic variation.

Techniques for measuring the level of gene expression in plant tissues are well known to those skilled in the art. Such techniques may involve one or more of reverse transcription polymerase chain reaction (RT-PCR), quantitative polymerase chain reaction (qPCR), northern blotting, DNA microarray, Western blotting, 2D-Gel Electrophoresis, and Immunoassays.

In an eleventh aspect of the present invention there is provided a substantially purified or isolated fungal cell wall degrading enzyme, DUF or FTR protein, wherein said fungal cell wall degrading enzyme, DUF or FTR protein is isolated from a plant species.

In a preferred embodiment the fungal cell wall degrading enzyme is a glucanase, more preferably a BGNL, and the plant is a Loliinae or Dactylidnae species. In a more preferred embodiment the plant is from the Lolium, Festuca or Dactylis species.

The Lolium or Festuca species may be of any suitable type, including Italian or annual ryegrass, perennial ryegrass, tall fescue, meadow fescue and red fescue. Preferably the ryegrass or fescue species is a Lolium species such as Lolium perenne or Lolium arundinaceum which is otherwise known as Festuca arundinacea. The Dactylis species may be of any suitable type. Preferably the Dactylis species is a Dactylis marina or Dactylis glomerata.

In a preferred embodiment, the plant from which the DUF protein is isolated is a Loliinae species. In a more preferred embodiment the plant is from a Lolium or Festuca species.

In a preferred embodiment, the plant from which the FTR protein is isolated is a Triticeae, or Aveneae species, preferably Triticum aestivum, Hordeum vulgare or Avena sativa. In a preferred embodiment, the substantially purified or isolated fungal cell wall degrading enzyme includes an amino acid sequence shown in Sequence ID No: 7; or a functionally active fragment or variant thereof.

In a preferred embodiment, the substantially purified or isolated DUF protein includes an amino acid sequence shown in Sequence ID No: 29; or a functionally active fragment or variant thereof.

In a preferred embodiment, the substantially purified or isolated FTR protein includes an amino acid sequence shown in Sequence ID No: 31 ; or a functionally active fragment or variant thereof.

By "functionally active" in this context is meant that the fragment or variant is capable of modifying pathogen resistance in a plant, or creating or enhancing a symbiotic relationship between a plant and a symbiont. Preferably, the fragment or variant has one or more of the biological properties of the fungal cell wall degrading enzyme, DUF or FTR protein. Additions, deletions, substitutions and derivatizations of one or more of the amino acids are contemplated so long as the modifications do not result in loss of functional activity of the fragment or variant. Preferably the fragment or variant has at least approximately 60% identity to the relevant part of the above-mentioned sequence, more preferably at least approximately 80% identity, most preferably at least approximately 90% identity. Such functionally active variants and fragments include, for example, those having conservative amino acid substitutions of one or more residues in the corresponding amino acid sequence. Preferably the fragment has a size of at least 10 amino acids, more preferably at least 15 amino acids, most preferably at least 20 amino acids. In a further embodiment of this aspect of the invention, there is provided a polypeptide recombinantly produced from a nucleic acid or nucleic acid fragment according to the present invention. Techniques for recombinantly producing polypeptides are well known to those skilled in the art. The present invention will now be more fully described with reference to the accompanying Examples and drawings. It should be understood, however, that the description following is illustrative only and should not be taken in any way as a restriction on the generality of the invention described above. Brief Description of the Drawings/Figures

Figure 1 : Amino acid sequence alignment of LpBGNL (β-1 ,6-glucanase gene; Sequence ID No 7) and fungus β-Ι ,θ-glucanase gene products [Neotyphodium sp. (Sequence ID No 5), Epichloe festucae (Sequence ID No 6), Trichoderma harzianum (Sequence ID No 8), Hypocrea lixii (Sequence ID No 9)]. The aryl-phospho-beta-D-glucosidase domain is shown in boxes (Sequence ID Nos 10-14). Dash (-) in the amino acid sequences shows a gap. Following the CLUSTAL W format, asterisk (*), colon (:) and dot (.) under the alignment denote 'conserved amino acid residues', 'including conserved substitution(s)' and 'including semi-conserved substitution(s)'. NCBI Ul is shown at the end of each sequence.

Figure 2: Genome structure of the LpBGNL and E. festucae β-Ι ,θ-glucanase genes and genetic linkage analysis for LpBGNL. (a) Alignment of genome sequences from perennial ryegrass and Epichloe species. The filled and striped boxes in the upper part of Fig 2 (a) show the location of LpBGNL and the plant genome-related sequence, respectively. The black-filled box represents the BAC vector. The filled and empty boxes in the lower part of Fig 2 (a) show the location of the β-Ι ,θ-glucanase gene and flanking gene, respectively. The transcription direction of the genes is indicated with the arrow. Corresponding gene sequences are connected with black dashed lines, (b) Alignment of coding regions of the LpBGNL and Epichloe β-Ι ,β-glucanase genes. The solid lines represent non-gene coding region of perennial ryegrass (upper part of Fig 2 (b)) and Epichloe (lower part of Fig 2 (b)) species. The location of the intron and aryl-phospho-beta-D-glucosidase domain (breaking line) are shown, (c) Genetic linkage map of perennial ryegrass LG5 with the LpBGNL- related locus. The LpBGNL-related marker locus is indicated with the arrow. Genetic distance (cM) is shown on the right side of the genetic markers. Figure 3: The expression levels of LpBGNL and the E. festucae var. lolii β-Ι

gene, (a) The expression level of LpBGNL in each tissue of perennial ryegrass. The y-axis shows normalised read count number (CPM). (b and c) The expression levels of the genes in E- and E+ perennial ryegrass genotypes. The x-axis shows time since the germination treatment, and y-axis shows CPM for LpBGNL (b) and the E. festucae var. lolii β-1 ,6- glucanase gene (c).

Figure 4. Phylogenetic tree of LpBGNL orthologues and fungal β-Ι ,δ-glucanase genes. The phylogram was generated based on amino acid sequence of the aryl-phospho-beta-D- glucosidase domain. Sequences from angiosperm species are indicated in a dashed box; the remainder are fungal species. Asterisk (*) denotes species from which gene products have been confirmed to have the β-Ι ,θ-glucanase activity. For cocksfoot, three contigs (haplotypes 1 -3) were generated, and amino acid sequences from those contigs were used. For Dactylis marina, the contig without a putative premature stop codon was used. Strain and sequence contig (scaffold) identifiers of the Genome Project at the University of Kentucky website are shown in brackets. For the other fungal sequences, NCBI Ul is shown in brackets. The clade including the sequences from Trichoderma and Hypocreaas species were selected as an outer group to obtain a root of the phylogenic tree.

Figure 5. PCR screening for the LpBGNL sequence, and taxonomic classification of plant species described in the current study (a) PCR amplicons from perennial ryegrass genotype Impact04, cooksfoot, Dactylis marina, coast tussock-grass and harding grass, and a strain of E. festucae are visualised on the 2200 TapeStation instrument (Agilent Technologies, CA, USA). The target fragments are indicated with the large arrows. Two genotypes of coast tussock-grass were subjected to the screening. The uppermost and lowermost bands show the position of upper and lower markers, respectively, of the D1000 Kit (Agilent). NTC denotes no-template control for the PCR assay. The small arrows indicate the positions of 400 and 500-bp fragments of the ladder, (b) Species classified into angiosperms, the tribe Poeae, and sub-tribes Loliinae and Dactylidinae are indicated by the boxes on the right side of the phylogenetic tree. The species which were positive and negative in the PCR screening step are indicated by the plus (+) and minus (-) signs, respectively. The divergent points of plant species from other species (>1 ,000 MYA) and the clade monocotyledon from other plant species (ca. 150 MY A) are indicated with dashed arrows. The divergent points of sub-tribes Loliinae and Dactylidinae from the remaining Poeae species, especially from genus Poa, which is a closely related taxon (ca. 13 MYA), and fine-leaved fescue (Sheep fescue) from broad-leaved fescues (tall fescue and meadow fescue) (ca. 9 MYA) are indicated with black arrows. The triangle represents the putative period of HGT in the evolutionary lineage.

Figure 6. DNA sequence alignment of the two haplotypes from the C3 genotype (Sequence ID Nos 15 and 16). Dash (-) in the DNA sequence shows a gap, and asterisk (*) under the alignment denotes 'conserved nucleotide'. DNA sequence corresponding to the PCR primers for the indel-based genotyping assay is underlined.

Figure 7. Result of the indel-based genotyping assay. The PCR amplicons were visualised on an agarose gel including the SYBR™ Safe DNA Gel stain (Thermo Fisher Scientific, Waltham, Massachusetts, USA). EasyLadder I (BIOLINE) was used as a size standard, and the 250 and 500 bp fragments are indicated with arrows. The genotype Ul of the genetic mapping population is indicated in each lane. C3 and NTC stand for 'the C3 genotype' and 'no-template control'. Figure 8. PCR screening for the LpBGNL sequence using the LpBGNL locus-specific primers (LpBGNL_short). (a) PCR amplicons from perennial ryegrass genotypes Impact04 and C3, and perennial ryegrass-associated endophyte were visualised on the 2200 TapeStation instrument using the High Sensitivity D1000 Kit (Aglient). The uppermost and lowermost bands show the position of upper and lower markers, respectively, of the High Sensitivity D1000 Kit. (b) PCR amplicons from Lolium and Festuca species (sub-tribe Loliinae) were visualised on an agarose gel containing the SYBR™ Safe stain. Tall fescue genotypes from two cultivars (Demeter and Quantum) were subjected to the screening. NTC denotes 'no-template control' for the PCR assay, (c) PCR amplicons from cocksfoot, Dactylis marina, coast tussock-grass and harding grass were visualised on the 2200 TapeStation instrument using the D1000 Kit. The uppermost and lowermost bands show the position of upper and lower markers, respectively, of the D1000 Kit. Two genotypes of coast tussock-grass were subjected, and the Impact04 genotype was used as a positive control. Figure 9. DNA sequences of the conserved ca. 750-bp region from plant species (Sequence ID Nos 17-27). Putative premature stop codons were found in Dactylis marina haplotypes 2 and 3. Figure 10. PCR assay for confirmation of a cross-species amplification capacity. PCR amplicons from cooksfoot, Dactylis marina, coast tussock-grass and harding grass were visualised on the 2200 TapeStation instrument using the D1000 Kit. The uppermost and lowermost bands show the position of upper and lower markers, respectively, of the D1000 Kit. The position of the target amplicons is indicated with a solid arrow. Size markers are indicated with outlined arrows. The Impact04 genotype was used as a positive control, and NTC denotes 'no-template control' for the PCR assay.

Figure 1 1. Genomic sequence of LpBGNL gene (Sequence ID No 1 ). The coding sequence is underlined (Sequence ID No 2), and the region in bold is a single intron (Sequence ID No 3).

Figure 12. LpBGNL cDNA sequence (Sequence ID No 4).

Figure 13. Taxonomic relationships of plant and fungal species and phylogenetic tree of Ep/c/7/oe-Triticeae/Poeae HGT candidates and corresponding fungal genes, (a) Taxonomic relationships of plant and fungal species and the presence/absence status of the LpFTRL and LpDUF3632-like sequences. The divergent point of plant species from other species (fungi and animals) is indicated with the filled hexagon (>1 ,000 MYA), and the filled circle indicate the divergent point of species belonging to the Triticeae and Poeae tribes from other Poaceae species. The triangles represent species belonging to the Loliinae subtribe, or Epichloe or Claviceps genera. The presence/absence status of putative homologous (orthologous) sequence for each gene is shown on the right side of species (subtribe/genius) name, in which Ίη s/7/co' and 'PCR' denote the database and PCR-based screening results, respectively. The plus (+) and minus (-) indicate presence and absence, respectively, and 'N/A' stands for 'not analyzed'. The ancient Ep/c/7/oe-Poeae/Triticeae HGT events are shown with curved arrows. The phylograms for LpFTRL(b) and LpDUF3632(c) generated based on predicted amino acid sequences. Sequences from plant species are Barley, Tausch's goatgrass and perennial ryegrass. For LpFTRL, the sequence from Trichoderma was selected as an outer group to obtain a root of the phylogenic tree (b). The DUF3632-like sequence was specific to Epichloe and Claviceps species, and no corresponding sequence was found in other related species of the Clavicipitaceae family, except for Periglandula ipomoeae, which is shown with an asterisk. No sequence was, therefore, selected as an outer group in the phylogram for LpDUF3632 (c). For fungus sequences, strain and sequence identifiers of the Genome Project at the University of Kentucky are shown in brackets. For plant sequences, NCBI Ul is shown in brackets.

Figure 14. The expression levels of HGT candidates and corresponding fungal endophyte genes in perennial ryegrass, (a) The expression levels of LpFTRL and Z-pDUF3632 in each tissues of the perennial ryegrass lmpact₀₄ genotype, (b) The expression levels of the HGT candidates and corresponding fungal genes in E⁺ and E^" perennial ryegrass genotypes, (c) A time course gene expression analysis for the HGT candidates and corresponding fungal genes in perennial ryegrass seeds and young seedlings. The x-axis shows tissue types (c and d) or period after a germination treatment (e and f), and normalised read count numbers [counts per million (CPM)] are indicated on the y-axis. The bars or lines show the expression levels of the HGT candidates in E^" and E⁺ plants, respectively, and the expression levels of corresponding fungal genes in E⁺ plants. Note that the scales of the y- axes are not uniformed.

Figure 15. PCR-based detection of the Ep/cfr/oe-Poeae/Triticeae HGT candidate from retail products. PCR assay results using the primers designed for LpFTRL. The PCR products were visualised on an agarose gel, using the SYBR™ Safe DNA Gel Stain (Thermo Fisher Scientific). The size (bp) of PCR amplicons is indicated on the right side of each image. As a control experiment, PCR primers for the florigen candidate gene (Hd3a in rice, and FT in wheat and barley) were used. For demonstration of absence of Epichloe and Claviceps species in the retail products, PCR primers specific to the fungal species were used. The gDNA samples from the perennial ryegrass genotype lmpact₀₄ and E. festuca were used, as positive controls for amplification with the plant and fungus-specific primers, respectively. With the PCR primers specific to Claviceps species, amplification from E. festucae gDNA template was observed, presumably due to sequence similarity between Epichloe and Claviceps species. 'NTC stands for 'no DNA template control'.

Figure 16. Nucleotide sequence of Lolium perenne DUF3632-like protein gene.

Figure 17. Amino acid sequence of Lolium perenne DUF3632-like protein gene. Figure 18. Nucleotide sequence of Lolium perenne fungal transcriptional regulatory (FTR) gene.

Figure 19. Amino acid sequence of Lolium perenne fungal transcriptional regulatory (FTR) gene.

Detailed Description of the Embodiments Exam pie 1 : Identification of a putative plant β-1 ,6-glucanase gene

A single genotype of perennial ryegrass (lmpact₀₄) was subjected to whole-genome shotgun and transcriptome sequencing using the lllumina HiSeq platform (NCBI BioProject Accession: PRJNA379202). De novo assembly of sequencing reads generated a 7.2 kb genomic DNA sequence contig [NCBI GenBank unique identifier (Ul): KY771 173], which contained a putative β-Ι ,θ-glucanase gene. The gene-like sequence was designated LpBGNL {Lolium perenne f½-1 ,6-GJucariase-Like). LpBGNL showed 74-90% and 72-82% identity at the DNA and amino acid sequence levels, respectively, to β-Ι ,θ-glucanase genes of fungal taxa, such as E. festucae, H. lixii, and T. harzianum (Fig. 1 , Table 1 ). The full genomic sequence of LpBGNL is shown in Figure 1 1. No matching sequence, however, was identified in the full genome sequences of plants such as A. thaliana, rice, Brachypodium distachyon (L.) P. Beauv., barley {Hordeum vulgare L.) or wheat {Triticum aestivum L.), based on database searches.

Table 1

Sequence identity (length of homologus sequence)

Species NCB Ul E. festucae Neotyphodium H. lixii T. harzianum

Perennial ryegrass KY771 173 (LpBGNL) 90% (1 140 bp) 89% (1 140 bp) 75% (1 101 bp) 74% (1 103 bp)

Epichloe festucae EF015481.1 98% (1 164 bp) 74% (1 101 bp) 73% (1 1 12 bp)

Neotyphodium sp. AF535131.1 - 74% (1283 bp) 73% (1288 bp) O Hypocrea lixii EU747838 - - 96% (1320 bp)

Trichoderma harzianum X79197.1 - - -

>

J

Example 2: Genomic and genetic characterisation of LpBGNL Methods Plant materials and DNA extraction

Details of plant genotypes are summarised in Table 2. Genomic DNA was extracted from young leaves of plants and fungal endophyte mycelium using the DNeasy plant mini kit (QIAGEN, Hilden, Germany). PCR amplification

Locus-specific primers were designed using the Sequencher software (GENECODE, Ml, USA) and the PCR primers are listed in Table 4. PCR amplification was performed with MyFi polymerase kit (BIOLINE, London, UK). PCR amplicons were visualised on the 2200 TapeStation instrument.

Short-read sequencing of BAC clones and amplicons

The BAC-based genomic library was screened through use of PCR. For the phylogenomic analysis, PCR primers were designed to obtain genomic fragments from Loliinae and Dactylidinae species (Table 4). Sequencing libraries for the MiSeq platform (lllumina, San Diego, California, USA) were prepared from the BAC clones and PCR amplicons, following the previously described MspJI-based method [Shinozuka et al]. The library was characterised with the TapeStation and Qubit instruments (Thermo Fisher Scientific). The outcome reads were assembled with the Sequencher and SOAPdenovo programs. Genetic linkage analysis

PCR primers were designed to detect the indel polymorphism within the LpBGNL sequence (Table 3). Genetic linkage analysis was performed through use of the p150/1 12 reference genetic mapping population of perennial ryegrass using the JoinMAP 3.0 application. PCR amplification

Locus-specific primers were designed using the Sequencher software (GENECODE, Ml, USA) and the PCR primers are listed in Table 4 PCR amplification was performed with MyFi polymerase kit (BIOLINE, London, UK). PCR amplicons were visualised on the 2200 TapeStation instrument. Short-read sequencing of BAC clones and amplicons

The BAC-based genomic library was screened through use of PCR. For the phylogenomic analysis, PCR primers were designed to obtain genomic fragments from Loliinae and Dactylidinae species (Table 4). Sequencing libraries for the MiSeq platform (lllumina, San Diego, California, USA) were prepared from the BAC clones and PCR amplicons, following the previously described MspJI-based method. The library was characterised with the TapeStation and Qubit instruments (Thermo Fisher Scientific). The outcome reads were assembled with the Sequencher and SOAPdenovo programs. Genetic linkage analysis

PCR primers were designed to detect the indel polymorphism within the LpBGNL sequence (Table 3). Genetic linkage analysis was performed through use of the p150/1 12 reference genetic mapping population of perennial ryegrass using the JoinMAP 3.0 application.

Table 2

Species

Common name Scientific name Genotype or cultivar Ul Reference

Shinozuka et al. ,

Perennial ryegrass Lolium perenne L. Impact 2017

Genotype from Aberystwyth (Great

Darnel Lolium temulentum L. Britain) IBERS: BA13157 Hand et al., 2010

Genotype from Tadham Moor (Great

Meadow fescue Festuca pratensis Huds. Britain) IBERS: BF1 199 Hand et al., 2010

Festuca arundinacea

Tall fescue Schreb. Demeter

Festuca arundinacea

Tall fescue Schreb. Quantum Hand et al., 2010 Sheep fescue Festuca ovina Genotype from Ponterwyd (Great Britain) IBERS: BL2643 Hand et al., 2010 Cocksfoot (Orchard

grass) Dactylis glomerata L. Currie SARDI: 778

{Dactylis marina) Dactylis marina Borrill Wild genotype from Algeria SARDI: 38013

Poa poiformis (Labill.)

Coast tussock-grass Druce Wild genotype from Australia SARDI: 41525

Harding grass

(Phalaris) Phalaris aquatica L. Landmaster

Table 3

Species Ul Instrument Data size Submitted by

Dactyl is glomerata L. lllumina HiSeq 1 run, 162.8M spots, 32.6G

SRX738187 Sichuan Agricultural University (orchardgrass) 2000 bases

SRX465632 lllumina HiSeq 1 run, 153.38M spots, 31 G KOREA POLAR RESEARCH

Deschampsia antarctica

2000 bases INSTITUTE

SRX745831 lllumina HiSeq 1 run, 129.3M spots, 25.9G

Poa annua Auburn University

2000 bases

SRX745855 lllumina HiSeq 1 run, 44.4M spots, 8.9G

Poa supina Auburn University

2000 bases

SRX745858 lllumina HiSeq 1 run, 46.3M spots, 9.3G

Poa infirma Auburn University

2000 bases

SRX669405 lllumina HiSeq 1 run, 50.4M spots, 10.2G

Phalaris aquatica Teagasc

2000 bases

Table 4

Sequence (5'- 3') Amplicon

Primer name Forward Reverse size^* DNA template

perennial ryegrass, darnel, meadow fescue, tall fescue, sheep fescue, cocksfoot, Dactylis marina,

LpBGNL_short GT C G G C ATG ATTG AG GTT CT ACTGCACATGGAGCTTGTTG 178 bp coast tussock-grass, Phalaris aquatica, perennial ryegrass-associated endophyte, perennial ryegrass genomic library

253/306

LpBGNLjndel AGGGCATCAACAAGATCAGG CGTCGCTCATCATCCATGGC p150/112 genetic mapping population

bp

LpBGNLJongl CGCGCCTAATCCTCTCCTCT CAGATATCTTGATACACATTCC 1452 bp perennial ryegrass, tall fescue, meadow fescue

LpBGNL_long2 GCCCGTCTGACGGGGCACAG CAGATATCTTGATACACATTCC 1590 bp darnel

LpBGNL_long3 CATCAACAAGATCAGGGGCG CTGCCCGTTCACGGTGCGAT 1067 bp sheep fescue

LpBGNL_long4 CACGACTTGGCTGCTTTCAA TTTGTCGTCCGGGCTCACGC 1147 bp cocksfoot, Dactylis marina

perennial ryegrass, cocksfoot, Dactylis marina, coast

LpBGNL_cons

CTGCCTCCGAGTTCGACTG TGGATGCGCYTCGTCATCC 415 bp tussock-grass, Phalaris

aquatica, perennial ryegrass-associated endophyte

LpHistone^**

TGCTTGCCCTTCAGGAGGCT ATCCTCCTGGCAAGCTGAATG 126 bp^** perennial ryegrass, cocksfoot, Dactylis marina, coast tussock-grass, Phalaris

LpABCG5 ATCAGGAAGGAGAGCCTCCA ATGATGGTGTTGGCCGCGTT 248 bp

aquatica

LpABCG6 TAACGCTCAACGGGGACG TCGTCGCCGATGATGGTGTT 235 bp

* Length of DNA fragment based on the perennial ryegrass genome sequence

^** Reference: Tu, Y. et al. 2010 Functional Analyses of Caffeic Acid O-Methyltransferase and Cinnamoyl-CoA-Reductase Genes from Perennial Ryegrass (Lolium perenne). The Plant Cell 22, 3357-3373. (doi : 10.1 105/tpc.109.072827)

Results

An in-house bacterial artificial chromosome (BAC)-based genomic library of perennial ryegrass had previously been constructed from endophyte-devoid (E^") individuals of the cultivar Grasslands Nui. PCR-based screening of the library identified two positive clones, designated LpBAC94-B20 and LpBAC125-N24. De novo sequence analysis and assembly identified the presence of LpBGNL in both clones. The gene was located within 39 kb- and 24 kb-contigs (NCBI Ul: KY771 171 and KY771 172) of LpBAC94-B20 and LpBAC125-N24, respectively, along with a sequence (ca. 2 kb in length) showing similarity at a DNA sequence identity of 82% to a Zea mays transposon-related gene (NCBI Ul: AF434192.1 ) (Fig. 2a). A 1 1 kb-contig of E. festucae genome sequence (NCBI Ul: EF015481 ), which includes the corresponding β-Ι ,θ-glucanase gene, was obtained from the NCBI database and was shown to contain three other genes located within a 5 kb distance from the β-1 ,6- glucanase gene. In the BAC clone-derived contigs, however, no sequences similar to these flanking genes were identified. Putative coding regions for the E. festucae β-Ι ,θ-glucanase and LpBGNL genes were identified (Fig. 2b). A single intron was found in both sequences, and comparison of the exonic and intronic regions identified 4 insertion-deletion (indel) variations between them. Although the position of the intron was conserved, it seemed that almost all intron sequence was replaced in perennial ryegrass, due to insertion and deletion events. Although the coding regions were relatively highly conserved, no sequence similarity was found in the flanking sequence of the coding regions. A BLAST search of 1.5-kb upstream and downstream sequences of LpBGNL identified partial sequence similarity to the genomes of wheat and rice, while the corresponding upstream and downstream sequences of the E. festucae β-Ι ,θ-glucanase gene included partial sequences of the flanking genes (Fig. 2b).

Sequencing of PCR products generated using LpBGNL-specific primers identified a 51 -bp intron-located polymorphism between haplotypes of the heterozygous parent (C3 genotype) of the perennial ryegrass p150/1 12 genetic linkage mapping population (Figure 6), which facilitated development of an indel-based DNA marker. From the p150/1 12 population, 48 individuals were genotyped (Figure 7), and the LpBGNL-related marker locus was assigned to a distal region of perennial ryegrass linkage group (LG) 5 (Fig. 2c). Example 3: LpBGNL gene expression analysis Methods Gene expression analysis

The transcriptome sequencing reads from Impact04 tissues were mapped against Impact04 genome contigs (>999 bp) for filtering. The number of reads which contained LpBGNL sequence (no sequence mismatch for 60 bp or longer) were counted as LpBGNL-derived reads. For gene expression in seedlings, E+ and E- seeds of perennial ryegrass cultivar Alto were subjected to germination treatment by placement on wet filter paper in the dark for 2 days followed by seedling growth under full-light conditions [44]. RNA was extracted with a CTAB extraction method, and sequencing libraries were prepared using the SureSelect strand-specific RNA library preparation kit (Agilent). Sequencing analysis was performed on the lllumina HiSeq 3000 platform. The full LpBGNL cDNA sequence is shown in Figure 12.

Results

Expression of LpBGNL was determined using data from the transcriptome sequence of the Impact genotype (NCBI BioProject Accession: PRJNA379202). Sequencing reads corresponding to LpBGNL were identified from leaf, root and flower samples, and higher expression levels [based on counts per million reads (CPM)] were detected in root and flower than in leaf (Fig. 3a, Table 6). Specificity of gene expression was examined using endophyte-infected (E⁺) and E^" perennial ryegrass seeds and seedlings. Due to sequence divergence, sequencing reads corresponding to the plant and fungal gene could be reliably discriminated (Table 7). Read counts were very low for both E⁺ and E^" seeds immediately after the germination treatment (Fig. 3b). Although no large-scale morphological change was observed during the following two days, the read counts substantially increased in both samples. The counts remained at relatively high levels in young seedlings at 5 and 10 days after treatment. As similar trends were observed for both E⁺ and E^" genotypes, presence of endophyte did not significantly affect LpBGNL expression pattern. Expression of the endogenous E. festucae var. lolii β-Ι ,θ-glucanase gene was observed in E⁺ seedlings, but the read count approach revealed relatively low levels throughout the 10 days, in contrast to LpBGNL (Fig. 3c, Table 7). Table 6

Library Total reads Counts CPM

Root tip 14,656,730 479 32.7

Root (middle) 14,515,287 798 55.0

Leaf middle 1 23,613,687 87 3.7

Leaf middle 2 19,292,712 49 2.5

Leaf middle 3 14,752,797 54 3.7

Leaf tip 1 22,525,848 272 12.1

Leaf tip 2 13,271 ,284 1 13 5.9

Leaf tip 3 16,738,932 154 9.2

Flower 1 1 ,536,030 767 66.5

Table 7

E- individual E+ individual

LpBGNL Epichloe β-1 ,6-glucanase gene LpBGNL Epichloe -1 ,6-glucanase gene

Total reads Counts CPM Counts CPM Total reads Counts CPM Counts CPM >

0 h 54, 109,844 19 0.35 0 0.00 36,389,285 4 0.11 5 0.14 J

4 h 62, 109,304 30 0.48 1 0.02 42,014,889 109 2.59 7 0.17

1 day 74,819,980 2751 36.77 3 0.04 64,295,493 802 12.47 28 0.44

2 days 71 ,779,944 6285 87.56 1 0.01 57,472,563 2838 49.38 25 0.43

5 days 59, 159,973 3391 57.32 1 0.02 57,259, 134 1847 32.26 9 0.16

10 days 102,085, 102 6565 64.31 0 0.00 53,748,829 3901 72.58 22 0.41

Example 4: Phylogenetic analysis of plant and fungal β-1 ,6-glucanase(-like) genes Method In silico analysis

The DNA sequences of fungal β-Ι ,δ-glucanase genes were obtained from the NCBI (http://www.ncbi.nlm.nih.gov/) database and the Genome Project at the University of Kentucky website. Putative orthologous sequences were sought in the NCBI, Brachypodium distachyon (http://www.brachypodium.org/) and Ensembl (http://plants.ensembl.org/index.html) databases. Non-synonymous and synonymous nucleotide substitution rates (Ka and Ks, respectively) were calculated using the Synonymous Non-synonymous Analysis Program (SNAP; http://www.hiv.lanl.gov/). Alignment of DNA sequences was performed with the CLUSTALW program (http://www.genome.jp/tools/clustalw/) with the default parameters. Phylogeny was generated with the MEGA7 program (http://www.megasoftware.net/).

The presence of LpBGNL orthologues in other Poeae species was determined by PCR- based screening. LpBGNL-specific primers were designed and short DNA fragments (178 bp in length) were amplified from genomic DNA templates of darnel (Lolium temulentum L), meadow fescue {Festuca pratensis Huds.), tall fescue (Festuca arundinacea Schreb.), sheep fescue (Festuca ovina) (Table 2, Figure 8). Products were also obtained from those of cocksfoot/orchard grass (Dactylis glomerata L.) and Dactylis marina Borrill, but not from those of coast tussock-grass [Poa poiformis (Labill.) Druce] or harding grass/phalaris [Phalaris aquatica L). DNA fragments were not amplified from genomic DNA templates of E. festucae var. lolii., confirming the specificity of the oligonucleotide primers that were used. Alignment of LpBGNL and the fungal genes identified a conserved region ca. 750 bp in length, corresponding to the aryl-phospho-beta-D-glucosidase domain. For a phylogenetic analysis, DNA fragments, including the aryl-phospho-beta-D-glucosidase domain, were amplified from the selected Loliinae and Dactylidinae species. De novo amplicon sequence analysis and assembly obtained a single sequence contig of the 750 bp region for each Lolium and Festuca species (Figure 9). For each Dactylis species, three contigs (haplotypes) were generated. A putative premature stop codon was found in two haplotypes of Dactylis marina, and the haplotypes with the premature stop codon were excluded from the further analysis. Fungal β-Ι ,θ-glucanase gene-like sequences were obtained from the NCBI database and the Genome Project at the University of Kentucky website. Phylogenetic analysis with the maximum likelihood method was performed, and the plant species-derived sequences were found to be clustered in the phylogram with those from Epichloe (Neotyphodium) species (Fig 4). The sequences from other fungi were more distantly related to the plant sequences. Synonymous and non-synonymous nucleotide substitution (K_s/K_a) ratios were calculated using the 750 bp sequences. Both K_s and K_a values between the Epichloe and plant species were substantially lower than those between the Epichloe species and other fungi (H. lixii and T. harzianum) (Table 2). The K_s/K_a ratios for candidate LpBGNL orthologues were between 0.027-0.221 , lower or equivalent to values from the fungal β-Ι ,θ-glucanase genes (0.057-0.275).

In order to verify presence/absence boundaries, a set of PCR primers was designed to amplify 415 bp fragments within a region highly conserved between plant and fungal β-1 ,6- glucanase(-like) genes. This assay confirmed the absence of the gene in the Poa and Phalaris species samples (Fig. 5). As a control experiment for capacity to amplify cross- species, PCR was performed with primers specific to perennial ryegrass histone H3 and candidate plant architecture genes, such as the ATP-binding cassette protein sub-family G 5 and 6 genes ( _pABCG5 and _pABCG6, respectively). Although all primers were designed based on perennial ryegrass sequence, PCR amplification from coast tussock- grass and harding grass/phalaris was observed, except for the combination of ABCG 5 primers and one of the Poa genotypes (Figure 10). Database searches were performed using published short-read sequencing data on the NCBI Sequence Read Archive (SRA; https://www.ncbi.nlm.nih.gov/sra). Sequences significantly matching the LpBGNL and E. festucae β-Ι ,δ-glucanase genes were found from Dactylis species, but not from Poa or Phalaris species (Table 9, Figure 9). No significantly matching sequence was obtained from Antarctic hair grass {Deschampsia antarctica E. Desv), which is believed to be taxonomically closer than Poa species to members of the sub-tribes Loliinae and Dactylidinae (Fig 5b). As a control analysis, sequences similar to the LpABCG 5 and 6 genes were sought, leading to identification of matching sequences from all tested Poeae species, including those belonging to both Poa and Phalaris. Table 8

Ks, Ka (Ka/Ks) Candidate orthologue of LpBGNL β-Ι ,θ-glucanase gene

Darnel Meadow Sheep E. Neotyphodium T.

ryegrass Tall fescue Fescue fescue festucae sp. H. lixii harzianum

0.0622, 0.1302, 0.0991 , 0.3333, 0.1961 , 0.2539, 0.61 16, 0.6396,

Perennial

0.0017 0.0102 0.0102 0.0416 0.0501 0.0629 0.1507 0.1469 ryegrass

(0.0273) (0.0783) (0.1029) (0.1247) (0.2553) (0.2477) (0.2464) (0.2297)

0.1670, 0.1359, 0.3575, 0.1957, 0.2533, 0.6041 , 0.6383,

Darnel ryegrass - 0.0085 0.0085 0.0433 0.0518 0.0646 0.1491 0.1453

(0.0509) (0.0625) (0.121 1) (0.2647) (0.2550) (0.2468) (0.2276)

0.0308, 0.3313, 0.1887, 0.2369, 0.6285, 0.6563,

Tall fescue - - 0.0068 0.0467 0.0535 0.0655 0.1470 0.1432

(0.2213) (0.141 1) (0.2838) (0.2766) (0.2339) (0.2182)

0.3247, 0.1823, 0.2305, 0.6093, 0.6494,

Meadow Fescue - - - 0.045 0.0519 0.0638 0.1487 0.1449

(0.1387) (0.2845) (0.2770) (0.2441) (0.2231)

0.2207, 0.6598, 0.6805,

Sheep Fescue - - - - 0.0433 0.2969, 0.051 0.1442 0.1407

(0.1961) (0.1717) (0.2186) (0.2068)

0.0896, 0.6363, 0.6424,

Epichloe - - - - - 0.0247 0.1488 0.1492 festucae

(0.2751) (0.2339) (0.2323)

0.6546, 0.6670,

Neotyphodium - - - - - - 0.1533 0.1537 sp.

(0.2341) (0.2304)

0.1 186,

Hypocrea lixii - - - - - - - 0.0068

(0.0572)

Results

The high level of DNA sequence similarity to the fungal genes, and absence of LpBGNL- like sequences from other representative angiosperm species suggested that LpBGNL was obtained from a fungal species through HGT. Genomic and genetic characterisation was consequently performed in order to demonstrate that LpBGNL is located on a perennial ryegrass chromosome. Furthermore, LpBGNL orthologues were found to be present in other Loliinae and Dactylidinae species. It is hence unlikely that LpBGNL was an assembly or annotation artefact, even though LpBGNL shows unusually high DNA sequence similarity (ca. 90%) to the β-Ι

genes of contemporary species descended from the putative donor, when compared to other horizontally transferred genes in eukaryotes.

The PCR-based screening and database searches suggested that the β-1

gene is present in only a limited number of Poeae species including the genera Lolium, Festuca and Dactylis, which are confined to the sub-tribes Loliinae and Dactylidinae. The phylogenetic analysis suggested a common origin of the Epichloe-derived β-Ι ,δ-glucanase genes and LpBGNL orthologues, and the close relationship between the LpBGNL orthologues of contemporary Loliinae and Dactylidinae grasses suggests that the gene may have been introduced into the genome of a common ancestor of the sub-tribes by a single transfer event. The HGT event may consequently have occurred between ca. 9 to 13 million years ago (MYA), based on the predicted time of divergence of the two sub-tribes from other Poeae lineages (Fig. 5b).

The Ks/Ka ratios between plant β-1 ,6-glucanase-like genes (0.027-0.221) were not substantially different from those of the fungal β-Ι ,δ-glucanase genes, and those of LpABCG5 and LpABCG6 (0.166 and 0.238). This suggests that LpBGNL may have been subjected to selection pressures. Although similar Ks/Ka ratios were obtained from the angiosperm and fungus groups, DNA mutation rates between those two groups may not be equivalent. From cocksfoot/orchard grass and Dactylis marina, three haplotypes of the aryl-phospho-beta-D-glucosidase domain were identified, and two haplotypes obtained from Dactylis marina contained a putative premature stop codon. As these Dactylis species have autopolyploid genomes, the β-Ι

genes in these species may have been subjected to unique selection pressures due to genetic redundancy, when compared with genes in the other plant species. As asexual symbiotic Epichloe species grow as hyphae between cells of vegetative aerial tissues in Poeae species, it is likely that the close physical proximity of both partners in the symbiosis may have facilitated an HGT event, similar to the physical contacts between parasitic and host plants. The conservation of the intron position between the LpBGNL and Epichloe β-Ι ,β-glucanase genes suggests that a part of the endophyte genome including the β-Ι ,β-glucanase gene, rather than a reverse-transcription product of endophyte gene mRNA, was incorporated into the recipient genome. In prokaryotes, transformation is a prevalent mechanism of gene exchange, in which HGT occurs through uptake of exogenous double-stranded DNA by the recipient cell.

Table 9

Species Ul Source Data size LpBGNL £. festucae β-Ι ,θ-glucanase gene LpABCG5 LpABCG6

Dactylis glomerata L.

(orchardgrass) SRX738187 Transcriptome 32.6G bases 95% (8e-35) 99% (6e-41 ) 99% (3e-41) 94% (5e-3

Deschampsia antarctica SRX465632 Genome 31 G bases N.S. N.S. 99% (7e-42) 97% (1 e-3

Poa annua SRX745831 Transcriptome 25.9G bases N.S. N.S. 100% (5e-43) 95% (1 e-3

Poa supina SRX745855 Transcriptome 8.9G bases N.S. N.S. 99% (8e-42) 90% (6e-2

Poa infirma SRX745858 Transcriptome 9.3G bases N.S. N.S. 100% (2e-43) 92% (2e-2 0

> Phalaris aquatica SRX669405 Transcriptome 10.2G bases N.S. N.S. 97% (5e-39) 94% (4e-3

>

J

Example 5: Loss of function of LpBGNL

Given the role of plant-encoded β-1 , 6-glucanase enzyme in the establishment of a stable symbiotic relationship with an endophyte, a loss of function analysis is carried out. A LpBGNL-like gene is removed from the genome of a ryegrass plant which normally forms a stable relationship with an Epichloe endophyte. The stability of the plant-endophyte association is then evaluated in such plants compared with unmodified control plants.

Example 6: Gain of function of LpBGNL

Given the role of plant-encoded β-1 , 6-glucanase enzyme in the establishment of a stable symbiotic relationship with an endophyte, a gain of function analysis is carried out. A LpBGNL-like gene is introduced into plants such as rice and wheat that do not normally form stable symbiotic relationship with Epichloe endophytes. The stability of the plant- endophyte association is then evaluated in such plants compared with unmodified control plants.

Example 7: LpBGNL as a promoter of natural stable associations with asexual Epichloe

Fungal β-Ι ,θ-glucanases have been reported to be specifically secreted into plant apoplasts during endophyte infection, and may play a role in provision of nutrition to the infecting endophyte, control of branching of the endophyte hyphae, and protection of plant tissues from infection of other fungal pathogens. The plant-encoded enzyme may participate in one or more of these processes, and so contribute to establishment of a stable symbiotic relationship.

Although Epichloe endophytes do not colonise root tissues, a relatively high level of expression of LpBGNL was observed in root tissues. Hence, the plant-encoded enzyme may function to protect against infection by soil-borne fungal pathogens. Similarly, active expression in flowers may suggest the capacity to protect against fungal pathogens such as Epichloe typhina, which causes choke disease. However, some Festuca and Dactylis species are relatively susceptible to infection by E. typhina, even though those species presumably also possess the β-Ι ,δ-glucanase-like gene. As natural stable associations with asexual Epichloe endophytes are confined to the Poeae lineages that possess LpBGNL-like genes, the ancestral HGT event (which might have occurred from a sexual pathogenic Epichloe-like species) may have provided pre-adaptive conditions for the contemporary symbiosis. The evidence for selective pressure on the gene is suggestive.

As major symbionts, the asexual Epichloe species provide abiotic and biotic stress tolerance to grass species of the Poeae tribe. Tolerance to invertebrate herbivory is a well- characterised benefit to the host, partially attributable to the effects of a makes caterpillers floppy-like (mcf-like) gene. The mcf-like gene was horizontally transferred into the endophyte genome from a bacterial species 7.2-58.8 MYA. Hence, multiple horizontal transfer events, including transfer of the LpBGNL-like gene as described in the present study, may have been involved in the establishment of the current stable symbiotic relationship.

Given the role of plant-encoded LpBGNL enzyme in the establishment of a stable symbiotic relationship with an endophyte, a stable symbiotic relationship with a plant and an endophyte may be created, which otherwise may not occur. The introduction of BGNL into plants which do not contain BGNL and are thus not able to form a stable symbiotic relationship with endophytes, may enable the establishment of such a relationship.

Example 8: A role for LpBGNL in pathogenic resistance

The potential role of a β- Ι

gene in protection against other, pathogenic, fungal species is of particular interest. Species such as T. harzianum are mycoparasites of fungal phytopathogens, and this property is related to glucanase activity. Fungal-derived genes for anti-fungal enzymes such as endochitinases and glucanases have also been used for generation of transgenic plants with enhanced pathogen resistance.

To generate transgenic plants with enhanced pathogen resistance using the LpBGNL-like gene the LpBGNL-like gene is transferred into the genomes of crop plants such as rice and wheat and then the resistance of the transformed plants to a range of fungal diseases is evaluated.

Example 9: R-1 ,6-glucanase as an indicator of stable associations with asexual Epichloe Given the role of the plant-encoded β-1 , 6-glucanase enzyme in the establishment of a stable symbiotic relationship with the endophyte, the presence and level of expression of a plant-encoded β-1 ,6-glucanase may be used to predict the likelihood of the plant forming a stable association with an endophyte carrying the β-Ι, β-glucanase, such as Epichloe.

The method involves selecting a plant for screening for symbiosis with a symbiont such as an Epichloe endophyte. The first step requires determining whether the plant to be screened contains the BGNL gene, through genetic analysis. If the plant selected for screening contains the BGNL gene, the second step is to determine the level and location of expression of the gene. The expression level of BGNL measured is then used to determine if the said plant screened would form a symbiotic relationship with the Epichloe endophyte.

High levels of expression in the root tissues and in the flowers of the plant are indicative of the plant screened being able to form a stable association with the endophyte. Exam pie 10: Identification of putative plant FTRL and DUF genes

Using non-polyadenylated RNA from an endophyte-absent (E ) perennial ryegrass individual, a sequencing library was prepared. A single lllumina MiSeq run generated a total of 8,216,014 reads from the library. A dataset of unique RNA reads, including 1 ,424,274 sequences, was generated. A BLAST search of the perennial ryegrass (E ) transcriptome shotgun assembly (TSA) and unique RNA read datasets against the Epichloe festucae transcriptome data identified 88 and 123 sequence similarity hits, respectively. Sequences putatively derived from microbiome, highly conserved genes between fungi and plants, such as actin and ubiquitin genes, and low confident matches were excluded through a manual examination, and 2 novel HGT candidates were identified.

From the TSA data, a sequence [unique identifier (Ul): ID_150936_C1449060_17.0] showing a relatively high similarity (85%) to an E. festucae unknown protein gene (Ul: EfM3.066060. partial-2.mRNA-1 ) was identified. Due to a similarity of predicted amino acid sequence to a fungal transcriptional regulatory (FTR) protein (Genbank Ul: CRL18938; identity: 48%, e-value: 2e-150) of Penicillium camemberti (Ascomycota, Trichocomaceae), one identified gene was designed LpFTRL (FTR-like).

The other candidate, designated -pDUF3632, showing a higher similarity (96%) to the E. festucae DUF3632 (domain of unknown function 3632)-like gene (Ul: EfM3.028800. mRNA- 1 ) was identified from the unique reads (Sequence Ul: 734684-1 ). Genome sequence contigs containing LpFTRL and LpDUF3632 from an in-house shotgun sequencing assembly data of the perennial ryegrass lmpact₀4 genotype (E ) were subject to a BLASTN search against nucleotide sequences catalogued in the NCBI GenBank database. Relatively high sequence similarity matches against /.pFTRL were identified from Tausch's goatgrass (Aegilops tauschii L, Ul: XM_020322891.1 ) and barley (Ul: AK375773.1 ), followed by ascomycota species, while only fungal genes showed significant sequence similarity to -pDUF3632. Putative orthologues of LpFTRL were identified in barley chromosomes 1 H and 7H (Ensembl Ul: HORVU1 Hr1 G009870 and HORVU7Hr1 G108080, respectively), while the corresponding sequences were only found from chromosome 7 of each sub-genome of hexaploid wheat (Ensembl Ul: TRIAE_CS42_7AL_TGACv1_557369_AA1780460, _AA1881 100, and _AA1997520), and from cereal rye {Secale cereal L.) chromosome 7 (NCBI short read archive (SRA) Ul: ERX140518), suggesting that the fungal gene was transferred into the chromosome 7 of a common ancestor of Triticeae and Poeae species. No significant match was obtained from the genome sequences of Brachypodium [B. distachyon (L.) P.Beauv.], rice {Oryza sativa L), sorghum [Sorghum bicolor (L.) Conrad Moench], Zea Mays (L.) and Arabidopsis thaliana (L). The LpDUF3632 sequence was subjected to a BLASTN search against NCBI SRA data from cool-season grass species, to find significant hits from Italian ryegrass (L multiflorum L.) and tall fescue (Festuca arundinacea L.), but not from orchard grass {Dactylis glomerata L.) or Antarctic hairgrass {Deschampsia Antarctica E.Desv.). The gene presence/absence status of the DUF3632-\ ke sequence in Poeae species was confirmed with a PCR-based assay (Fig. 13a).

The presence of LpFTRL-like sequence in all tested Poeae species was confirmed. A phylogenetic analysis revealed close relationships of LpFTRL and .pDUF3632 between corresponding Epichloe sequences (Fig. 13b and c), suggesting that the two genes were transferred from the fungus lineage into plant species, but not vice versa, after diversification [59 million years ago (MYA)] of ancestral Epichloe species from the Claviceps lineage. Based on the divergence ages of the Brachypodieae tribe from other Pooideae species, the FTRL gene was predicted to have been transferred into plants after 32-39 MYA. An LpDUF3632-derived DNA-based marker was assigned into perennial ryegrass linkage group (LG) 3, which corresponds to the chromosome 3 of Triticeae species. The DLT3632-like sequence is likely to have been transferred independently of the

gene, of which DNA-based marker has been assigned into perennial ryegrass LG 5, indicating a possibility of at least three independent HGT events since 32-39 MYA. Comparison of genomic sequences including DUF3632-\ ke genes suggested transformation-like ancestral HGT events, due to conservation of the position of the putative intron between plant and Epichloe species, similar to LpBGNL. No intron was predicted for LpFTRL and corresponding sequences in Epichloe species.

From the E^" perennial ryegrass transcriptome data, a leaf tip-specific expression pattern of LpFTRL was proposed (Fig. 14a). Although the counts per million reads (CPM) values were relatively low, LpDUF3632 was ubiquitously expressed (Fig. 14b). An analysis of the SRA data from Massay University, which did not include a leaf tip sample, indicated that the expression levels of LpFTRL and the corresponding Epichloe gene (EfM3.066060) were below the limit of detection in most tested tissues, regardless of the endophyte-infected (E+) IE- plants (Fig. 14c). Expression patterns of LpDUF3632 were similar between E+ and E- plants, and the Epichloe DUF3632 gene (EfM3.028800) was expressed in all tested tissues of the E⁺ plant (Fig. 14d). From the E⁺ and E^" seeds/young seedlings, an increment of LpDUF3632 expression level was observed in E⁺ seeds after a germination-treatment, although the expression levels of LpFTRL, EfM3.066060, and EfM3.028800 stayed low throughout the tested 10 days (Fig. 14e and f). Although the LpDUF3632 sequence was identified from the non-polyadenylated RNA sequencing library, it is likely that the gene is also expressed as polyadenylated RNA, due to identification of corresponding reads from the SRA data.

Identification of LpFTRL provided a unique opportunity to compare nucleotide substitution rates between fungus and plant lineages, using the sequences diverged only less than 40 MYA. The synonymous (Ks) substitution rate of the plant FTRL sequences was not substantially different from those from fungus sequences. Although some variation was observed in the non-synonymous (Ka) substitution rates, the Ka/MYA rate from the plant sequences was between those from Claviceps-Epichloe and E. gansuensis-other Epichloe combinations (Table 1). Table 10.

Diverged

age (MYA) Synonymous Non-synonymous Ka/Ks

Ks (AVE) SDV Rate (Ks/MYA) Ka (average) SDV Rate (Ka/MYA)

Claviceps-plants 58 1.0612 0.2560 0.0183 0.2307 0.0064 0.0040 0.2174

Claviceps-Epichloe 58 0.8517 0.1745 0.0147 0.1412 0.0073 0.0024 0.1657

Triticeae-Poeae 21 0.2767 0.0231 0.0132 0.1041 0.0008 0.0050 0.3760

E. gansuensis-other Epichloe 7.2 0.1435 0.01 12 0.0199 0.0519 0.0055 0.0072 0.3619

Ka and Ks rates for LpFTRL orthologues and corresponding Epichloe and Claviceps genes. The ratios calculated from the Claviceps and Epichloe, Claviceps and Plants, Triticeae and Poeae (plants), and E. gansuensis and the rest of Epichloe combinations are shown as Claviceps-Epichloe, C/av/^'ceps-Plants, Triticeae- Poeae, and E. gansuensis-other Epichloe, respectively. As Epichloe and Claviceps were diverged around 58.8 MYA followed by horizontal transfer of the FTRL sequence into plant species after 39 MYA, the comparison of the Claviceps and Plants combination include the diversification period in the Epichloe lineage (from 58.8 MYA to the HGT event), while the Triticeae and Poeae combination merely represent diversification between plant species.

The Ka/Ks ratio for the plant FTRL and corresponding E. gansuensis-other Epichloe sequences (0.376 and 0.3619, respectively) were relatively close, suggesting that the plant genes have been subjected to selection pressures since HGT. A comparison with E. festucae DUF3632 suggested a similar level (Ka/Ks = 0.31 ) of selection pressures for /.pDUF3632. These results proposed that plant FTRL and DUF3632 genes retain functionalities at the protein level, and have contributed to natural adaptation of the plant species. Similar to LpBGNL, the intron sequences of .pDUF3632 were less conserved than the exon sequences, likely due to lower levels of the pressure, which implies that the selective pressures have been an important factor for retaining of sequence similarity to the corresponding Epichloe sequences. The close physical contacts between the host plants and symbiotic and/or pathogenic fungi could have facilitated the relatively frequent transfer events. As both Epichloe and Claviceps species infect to plant reproductive tissues, the presence close to germ cells might have been an important factor of the multiple HGT events. Comparison of codon usage of /.pFTRL orthologues from diploid plant species and corresponding Epichloe sequences revealed a similar usage ratio for each amino acid, except for the termination codons. Only TGA was used in the plant genes as termination codon, the ratio of TGA was only 38.5% in the Epichloe sequences.

Pearl barley, wheat plain flour, rolled oat, and rice flour products were obtained from retail shops, and DNA was extracted from the retail products. PCR primers for the LpFTRL sequence within a conserved region between plant and Epichloe species were designed. Through PCR, the sequences corresponding to LpFTRL were amplified from barley, wheat, oat, and E. festuca gDNA templates, but not from rice, confirming the presence of the 'natural' transgene in Triticeae and Poeae cereal species (Fig. 15). A control experiment with fungus-specific primers confirmed absence of gDNA of Epichloe and Claviceps species, especially the rye ergot fungus, C. purpurea, in the retail products.

Materials and Methods sRNA sequencing

Fresh young leaves of an individual of perennial ryegrass cultivar Trojan were harvested. Total RNA including small molecules was extracted using the RNeasy Plant mini kit (QIAGEN) following the modified protocol of related products (Purification of miRNA from animal cells using the RNeasy® Plus Mini Kit and RNeasy MinElute® Cleanup Kit). A sequencing library was prepared using the Small RNA sequencing library preparation kit (NEB), and a fraction containing non-coding RNA molecules (bp in length) was purified using the BluePippin platform (Sage Science). The products were characterised on the 2200 TapeStation instrument using the D1000 kit (Agilent). The library was loaded on the MiSeq platform (lllumina), following the manufacture's instruction, and a sequencing analysis was performed with the MiSeq Reagent Kit v3 (150-cycle) kit (lllumina). The outcome data was processed with the FastX-tool-kit package, to remove adapter sequence and generate a unique read dataset.

BLAST-based screening

The transcriptome dataset of E. festucae (file name: M3 transcript sequences) was obtained from the website of Kentucky university (http://www.endophyte.uky.edu). The TSA (GFSR01000001-GFSR01044773) and sRNA unique read datasets of perennial ryegrass were prepared for a DNA sequence homology search. The homology search was performed with the megablast function of the BLAST+ package, and the threshold E value was set at 1 e-10. The resulting data was imported into the Microsoft Excel software for a manual examination. The manual examination was performed using the NCBI BLAST tools.

The homologues sequence in representative flowering plant species were performed on the Ensembl Plants website (http://plants.ensembl.org/index.html). The /.pFTRL sequence was subjected to BLAST-based search against cereal rye SRA dataset (SRA Ul: DRP000390). The LpDUF3632 sequence were subjected to the search against the following SRA datasets; Italian ryegrass (SRX1604870 and SRX1604871 ), tall fescue (SRX1056957), orchard grass (ERX1842528), and Antarctic hairgrass (SRX465632), from which significant homology hits against the perennial ryegrass architecture candidate genes, LpABCG5 and 6 (GenBank Uls: JN051254.1 and JN051255.1 ) were obtained, but no significantly similar sequence to an Ep/c/7/oe-specific gene, makes caterpillars floppy (mcf)-like gene (GenBank Ul: KJ502561.1 ), was identified. This screening method was validated through a BLAST search of the .pFTRL, _pDUF3632, and Epichloe mcf sequences against E+ and E- perennial ryegrass SRA datasets (SRA Uls: SRX1 167577-SRX1 167590).

DNA sample preparation

Plant seeds were obtained from the Genetic Resources Unit of Institute for Biological Environmental and Rural Studies (IBERS; Aberystwyth, Wales, UK) and the South Australian Research and Development Institute (SARDI). The plants were germinated on a filter paper in a petri dish. Total DNA was extracted from fresh leaf of each plant genotype using the DNeasy Plant mini kit (QIAGEN) following the manufacture's instruction. Total DNA was also extracted from barley (pearled grains), wheat (plain flour), oat (rolled grains) and rice (flour) using the DNeasy Plant mini kit.

PCR-based screening

PCR primers were designed using the Oligo Calc tools. For cross-species amplification, primers were designed to generate short DNA fragments (< 251 bp in length) within highly conserved regions of the target sequences. The PCR was performed on the CFX Real- Time PCR Detection Systems (Bio-Rad), with the Luna® Universal qPCR Master Mix (NEB), and data analysis was performed using the CFX Manager™ Software (Bio-Rad). Visualisation of PCR products was performed on the 2200 TapeStation instrument or on an agarose gel (2.0% w/v) stained with SYBR Green (Thermo Fisher) through electrophoresis.

Validation of the PCR-based screening method

PCR primers were designed based on perennial ryegrass sequences to amplify larger fragments (> 900 bp), for use as DNA templates of a standard curve assay (SCA). The PCR amplicons were amplified from the lmpact₀4 genotype, using the MyTaq™ DNA Polymerase kit, and the amplicons were cleaned using the Monarch® PCR & DNA Cleanup Kit (NEB). DNA concentration was adjusted to 1 pg/μΙ, and a dilution series was prepared. For fungus-specific primers, a dilution series of E. festucas gDNA was prepared. The SCA was performed on the CFX Real-Time PCR Detection Systems, with the Luna® Universal qPCR Master Mix, followed by data analysis with the CFX Manager™ Software.

In silico analysis

Amino acid and DNA sequences were prepared for in silico analysis. Gene structure prediction was performed using the FGENESH program of the Softberry website using the 'Monocot plants' parameter. Alignments of amino acid sequences were generated with the CLUSTALW program with the default parameters. Sequence homology search was performed with the NCBI and PredictProtein websites. Phylogenetic analysis was performed with the MEGA7 program. Non-synonymous and synonymous nucleotide substitution rate (Ka and Ks, respectively) was calculated using the Synonymous Non- synonymous Analysis Program (SNAP). For gene expression analysis, an in-house transcriptome read dataset was prepared through filtering of the transcriptome sequencing reads from lmpact₀4 tissues using lmpact₀4 genome contigs (>999 bp). The number of reads which contained LpFTRL or _pDUF3632 sequences (no sequence mismatch for 60 bp or longer) were counted. Finally, it is to be understood that various alterations, modifications and/or additions may be made without departing from the spirit of the present invention as outlined herein.

References

Hand, M. L, Cogan, N. O., Stewart, A. V. & Forster, J. W. Evolutionary history of tall fescue morphotypes inferred from molecular phylogenetics of the Lolium-Festuca species complex. BMC Evolutionary Biology 10, 303 (2010). Richards, T. A. et al. Phylogenomic analysis demonstrates a pattern of rare and ancient horizontal gene transfer between plants and fungi. The Plant Cell 21, 1897-191 1 (2009).

Shinozuka, H. et al. A simple method for semi-random DNA amplicon fragmentation using the methylation-dependent restriction enzyme MspJI. BMC Biotechnology 15, 25 (2015).

Claims

THE CLAIMS DEFINING THE INVENTION ARE AS FOLLOWS:

1. A method of creating or enhancing a symbiotic relationship between a plant and a symbiont carrying:

a fungal cell wall degrading enzyme;

a domain-of-unknown-function (DUF) gene; or

a FTR protein,

said method including introducing into said plant an effective amount of a nucleic acid or nucleic acid fragment encoding the fungal cell wall degrading enzyme, of the DUF gene or encoding the FTR protein, or a functionally active fragment or variant thereof, wherein the nucleic acid or nucleic acid fragment is isolated from a plant species.

2. A method according to claim 1 , wherein said fungal cell wall degrading enzyme is a glucanase.

3. A method according to claim 2, wherein said glucanase is β-1 , 6-glucanase (BGNL).

4. A method according to claims to claim 2 or 3, wherein said plant is from a Lolium, Festuca or Dactylis species.

5. A method according to claim 4, wherein said plant species is Lolium perenne.

6. A method according to claim 1 , wherein said nucleic acid or nucleic acid fragment is of a DUF gene and wherein said plant is from a from a Lolium,or Festuca species.

7. A method according to claim 6, wherein said plant species is lolium perenne or Lolium multiflorum.

8. A method according to claim 1 , wherein said nucleic acid or nucleic acid fragment is encoding a FTR protein and wherein said plant is from a from a Triticeae,or Aveneae species.

9. A method according to claim 8, wherein said plant species is selected from the group consisting of Triticum aestivum, Hordeum vulgare and Avena sativa.

10. A plant endophyte symbiota produced by a method according to any one of claims 1 to 9.

11. A method for selecting a plant for symbiosis with a symbiont carrying:

a fungal cell wall degrading gene;

a DUF gene; or

a FTR gene,

said method comprising

a) determining the presence of a fungal cell wall degrading gene, a DUF gene or a FTR gene in said plant,

b) measuring the level of expression of the gene identified in a), and

c) using the expression level measured in b) to determine if the plant will form a symbiotic relationship with said symbiont.

12. A method according to claim 1 1 , wherein the fungal cell wall degrading enzyme is a glucanase.

13. A method according to claim 12, wherein the glucanase is BGNL.

14. A method according to any one of claims 1 1 to 13, wherein the symbiont is an endophyte from an Epichloe species.

15. A substantially purified or isolated nucleic acid or nucleic acid fragment:

encoding a fungal cell wall degrading enzyme;

of a DUF gene; or

encoding a FTR protein,

wherein the nucleic acid or nucleic acid fragment is isolated from a plant species.

16. A nucleic acid or nucleic acid fragment according to claim 15, wherein said fungal cell wall degrading enzyme is a glucanase.

17. A nucleic acid or nucleic acid fragment according to claim 16, wherein said glucanase is BGNL.

18. A nucleic acid or nucleic acid fragment according to claim 15 or 17, wherein said plant is from a Lolium, Festuca or Dactylis species.

19. A nucleic acid or nucleic acid fragment according to claim 18, wherein said plant species is Lolium perenne.

20. A nucleic acid or nucleic acid fragment according to claim 15, wherein said nucleic acid or nucleic acid fragment is of a DUF gene and wherein said plant is from a from a Lolium or Festuca species.

21. A nucleic acid or nucleic acid fragment according to claim 20, wherein said plant species is Lolium perenne or Lolium multiflorum.

22. A nucleic acid or nucleic acid fragment according to claim 15, wherein said nucleic acid or nucleic acid fragment is encoding a FTR protein and wherein said plant is from a from a Triticeae or Aveneae species.

23. A nucleic acid or nucleic acid fragment according to claim 22, wherein said plant species is selected from the group consisting of Triticum aestivum, Hordeum vulgare and Avena sativa.

24. A substantially purified or isolated nucleic acid or nucleic acid fragment:

encoding BGNL;

of a DUF gene;

encoding a FTR protein; or

complementary or antisense to a nucleic acid or nucleic acid fragment encoding BGNL, of a DUF gene or encoding a FTR protein,

said nucleic acid or nucleic acid fragment including a nucleotide sequence selected from the group consisting of:

(a) the sequence shown in Sequence ID Nos: 1- 4, 28 or 30;

(b) complement of the sequence recited in (a);

(c) sequences antisense to the sequence recited in (a); and

(d) functionally active fragments and variants of the sequences recited in (a), (b) and (c).

25. A nucleic acid or nucleic acid fragment according to claim 24, wherein said functionally active fragment or variant has a size of at least 20 nucleotides and has at least approximately 90% identity to the relevant part of the sequence recited in (a), (b) or (c), upon which the fragment or variant is based.

26. A nucleic acid or nucleic acid fragment according to claim 25, wherein said functionally active fragment or variant has a size of at least 100 nucleotides and has at least approximately 95% identity to the relevant part of the sequence recited in (a), (b) or (c), upon which the fragment or variant is based.

27. A vector including a nucleic acid or nucleic acid fragment according to any one of claims 24 to 26.

28. A vector according to claim 27 further including a promoter and a terminator, said promoter, nucleic acid or nucleic acid fragment and terminator being operatively linked.

29. A plant cell, plant, plant seed or other plant part, including vector according to claim 27 or 28.

30. A method of modifying pathogen resistance in a plant, said method including introducing into said plant an effective amount of a nucleic acid or nucleic acid fragment according to any one of claims 24 to 26, or a vector according to claim 27 or 28.

31. Use of a nucleic acid or nucleic acid fragment according to any one of claims 24 to 26 for modifying pathogen resistance in a plant.

32. Use of a nucleic acid or nucleic acid fragment according to any one of claims 24 to 26, or nucleotide sequence information thereof, or single nucleotide polymorphisms thereof, as a molecular genetic marker.

33. A method of creating or enhancing a symbiotic relationship between a plant and a symbiont carrying:

a plant fungal wall degrading enzyme;

a DUF gene; or

a FTR protein, said method including introducing into said plant an effective amount of a nucleic acid or nucleic acid fragment any one of claims 24 to 26, or a vector according to claim 27 or 28.

34. A plant endophyte symbiota produced by a method according to claim 33.

35. A substantially purified or isolated fungal cell wall degrading enzyme, DUF or FTR protein, wherein said fungal cell wall degrading enzyme, DUF or FTR protein is isolated from a plant species.

36. An enzyme according to claim 35, wherein the enzyme is a fungal cell wall degrading enzyme and the fungal cell wall degrading enzyme is a gluconase.

37. An enzyme according to claim 36, wherein the glucanase is BGNL.

38. An enzyme according to claim 36 or 37, wherein the plant is a Loliinae or Dactylidnae species.

39. An enzyme according to claim 38, wherein said plant is from a Lolium, Festuca or Dactylis species.

40. An enzyme according to claim 39, wherein said plant species is Lolium perenne.

41. A DUF protein according to claim 35, wherein said plant from which the DUF protein is isolated is a Lolium or Festuca species.

42. A DUF protein according to claim 41 , wherein the plant species is Lolium perenne or Lolium multiflorum.

43. A FTR protein according to claim 35, wherein said plant from which the FTR protein is isolated is a Triticeae or Aveneae species.

44. A FTR protein according to claim 43, wherein said plant species is selected from the group consisting of Triticum aestivum, Hordeum vulgare and Avena sativa.

45. A substantially purified or isolated BGNL, DUF or FTR protein, said protein including an amino acid sequence selected from the group consisting of:

(a) the sequence shown in Sequence ID No: 7;

(b) the sequence shown in Sequence ID No: 29;

(c) the sequence shown in Sequence ID No: 31 ; and

46. A protein according to claim 45, wherein the functionally active fragment or variant has a size of at least 20 amino acids and has at least approximately 90% identity to the relevant part of the sequence recited in (a) upon which the fragment or variant is based.

47. A protein according to claim 46, wherein the functionally active fragment or variant has a size of at least 100 amino acids and has at least approximately 95% identity to the relevant part of the sequence recited in (a) upon which the fragment or variant is based.