Keywords

Fusarium: An Overview

Fusarium is a filamentous fungi (Sordariomycetes: Hypocreales: Nectriaceae) containing phytopathogenic and toxigenic species. The genus Fusarium wasfirst described by Link in 1809 as Fusisporium and is presently known as Fusarium, referred to as Fusarium sensu Wollenweber (Wollenweber 1931; Wollenweber and Reinking 1935). The genus is highly diverse with twenty monophyletic species complex and outgroups of nine species. Infestation of Fusarium coincides with that of the flowering plants nearly 91.3 million years ago (Fig. 10.1). Fusarium species are distributed on the plants, in soil and in water either as parasites, endophytes or saprophytes. Plant pathogenic Fusarium species cause wilts, blights, rots and cankers affecting field, horticultural, ornamental and forest crops in both agricultural and natural ecosystems. Fusaria also produce diversified toxic secondary metabolites (such as trichothecenes and fumonisins that can contaminate agricultural product, making them unsuitable for food and feed; trichothecenes can also act as virulence factor in plant diseases. A few Fusarium species are opportunistic human pathogens also causing corneal infections (O’Donnell et al. 2004).

Fig. 10.1
figure 1

Evolutionary diversification of the 20 Fusarium species. (reproduced from Ma et al. 2013, which was modified with permission from O’Donnell et al. 2013)

Full size image

Diversity Amongst Fusarium Species

Fusarium pathogens have diverse life cycles, niche specialization, host adaptation and specificity. Fusarium graminearum (Fg) and Fusarium verticilloides (Fv) are a narrow host range pathogens infecting predominantly the cereals, whereas Fusarium oxysporum (Fo) has a broad host range and infects both monocotyledonous and dicotyledonous plants (Armstrong and Armstrong 1981); besides, it is also an emerging pathogen on immuno compromised patients (O’Donnell et al. 2004) and other mammals (Ortoneda et al. 2004). Fusarium species vary in reproduction strategy; Fo is asexual, others are both asexual and sexual with either self-fertility (homothalism) or obligate out-crossing (heterothalism). Fusarium species produce meiotic (sexual) spores and at least three types of mitotic (asexual) spores. However, all Fusarium species do not produce all type of spores: Also, less than 20% of Fusarium species reproduce sexually (Fig. 10.2).

Fig. 10.2
figure 2

Generalized life cycle of Fusarium depicting varying reproduction strategy (Source Ma et al. 2013) Abbreviations Fg, F. graminearum; Fol, F. oxysporum f. sp. lycopersici; Fp, F. pseudograminearum; Fs, F. solani f. sp. pisi; Fv, F. verticillioides

Full size image

Fusarium oxysporum

Fusarium oxysporum Schlechtend, Fr. emended by Snyder and Hansen (1940) is an anamorphic species within the genus Fusarium. It is genetically heterogenous, polytypic morphospecies (O’Donnell and Cigelnik 1997; Waalwijk et al. 1996) which represents most abundant and ubiquitous soil-borne fungus; few strains have been reported from tundra soils as well (Stoner 1981) which exist as saprophytes and pervasive plant root endophytes. During saprophytism mode, Fusarium species degrade lignin (Rodriguez et al. 1996; Sutherland et al. 1983) and complex carbohydrates associated with soil debris (Christakopoulos et al. 1995, 1996). Root endophytic Fusarium species may be pathogenic or beneficial. A few strains are also pathogenic on gymnosperms. Pathogenic species within F. oxysporum have been differentiated into opportunistic, true pathogens and obligate pathogens based on the level/specialization of fungal–plant interactions (Scheffer 1991). Opportunistic parasites colonize weakened host plants or enter through wounds, have broad host range and exhibit low virulence. True pathogens require living plants for their growth; however, it can survive outside their hosts also, but are highly virulent on few host species. Obligate pathogens essentially require living host plant to complete their life cycle. They utilize host plant metabolism for their own growth, and in process alters plant growth pattern and morphology (Jackson and Taylor 1996). Members of F. oxysporum species complex are capable of causing vascular wilt diseases in over one hundred agronomically important plant species. However, individual F. oxysporum isolates are characterized by a high degree of host specificity; isolates that are pathogenic on single host are grouped into a forma specialis, e.g. F. oxysporum forma specialis lycopersici for tomato pathogens. Several F.oxysporum formae speciales consist of multiple independent lineages that have evolved polyphyletically. Interestingly, substantial genetic diversity has been revealed by molecular phylogenetic studies amongst isolates, supporting the present view that F. oxysporum represents a species complex (FOSC).

Reproduction

Fusarium oxysporum reproduces asexually, and its sexual state has never been observed (Booth 1971); it produces chlamydospores, microconidia and macroconidia (Nelson et al. 1983). Microconidia are uninucleate which germinate poorly with germination efficiency ranging from 1 to 20% (Ebbole and Sachs 1990). Macroconidia are multinucleate and germinate rapidly. Chlamydospores are resulting from the structural modification of vegetative hyphae or a thick-walled conidial cell and accessory spores (Schippers and van Eck 1981); F. oxysporum is diversified on shape of macroconidia, structure of micro-conidiophores, formation of chlamydospores (Beckman 1987).

Formae Speciales

Pathogenic and non-pathogenic F. oxysporum species cannot be distinguished morphologically unless pathogenic tests are performed. Pathogenic isolates of F. oxysporum (Fo) exhibit high level of host specificity which is directly linked to its pathogenicity to various plant species (Fravel et al. 2003). There are over 150 described formae speciales for Fo (Gordon 1965; Michielse and Rep 2009). Single forma specialis consist of isolates with the ability to cause wilt on a unique host or group of plant host species (Table 10.1). Thus, formae speciales is defined as an informal rank in classification scheme assigned on ability to cause disease in a unique host. Preliminarily, a forma specialis may be assigned to a strain based on the host from which a F. oxysporum isolate was recovered. It has been assumed that members of a forma specialis (f. sp.) are closely related and may have arisen from a common ancestor (Correll 1991; Kistler 1997). However, considerable genetic diversity has been reported within representative isolates of a forma specialis based on sequence comparisons in conserved regions of mitochondrial and nuclear DNA (Guadet et al. 1989; O’Donnell 1993). Kim et al. (1992, 1993) have analysed mt DNA of five formae speciales within cucurbitaceae, f. sp. cucumerinum, f.sp. lagenaria, f.sp. luffae, f.sp. melonis and f.sp. niveum that are pathogens of cucumber, calabash gourd, vegetable sponge, muskmelon and watermelon, respectively, and identified, fourteen mt DNA haplotypes within each forma specialis, out of which a few were common across formae speciales. Thus, all five formae speciales within cucurbitaceae are monophyletic (Kim et al. 1993). Similarly, Fo f. sp. melonis isolates were placed in two separate clusters, one closely related to Fo f. sp. langenaria and other to Fo f. sp. melonis cluster. On the other hand, F. oxysporum isolates pathogenic to banana (f.sp. cubense) are more divergent than those of cucurbits. Koenig et al. (1997) identified 72 RFLP haplotypes in 165 isolates of f. sp. cubense and placed them into two major groups and seven lineages. Two lineages of Fo cubense were genetically similar to an isolate of f. sp. niveum, than to each other. Moreover, F. oxysporum f. sp. cubense strains from all over the world were placed in 10 clonal lineages. Thus, it is concluded that Fo cubense has polyphyletic origin, and pathogenicity to banana is acquired independently.

Table 10.1 List of Fusarium oxysporum formae speciales along with its host plant
Full size table

The isolates within forma specialis also have overlapping host ranges. Gerlagh and Blok (1988) reported that Fo causing wilt in cucumber was pathogenic to both muskmelon and watermelon and grouped it as f. sp. cucurbitacearum.

Pathogenic Races

Pathogenic races are sub-divisions of individual forma specialis based on differential virulence to various cultivars of the same host (Correll 1991). F. oxysporum forma specialis cucumis (Armstrong and Armstrong 1978) constitute, five races (0, 1, 2 and 1, 2), and their pathogenicity to different melon cultivars varies (Risser et al. 1976). All pathogenic races within a forma specialis might have a single ancient ancestor (monophyletic forma specialis) or affiliated to distinct clades (polyphyletic forma specialis).

Vegetative Compatibility Groups (VCGs)

Vegetative compatibility group (VCG) comprises of isolates that undergo somatic fusion and form stable heterokaryons. They are genetically similar and represent a clonal population (Puhalla 1985; Kistler 1997; Gordon and Martyn 1997). Moreover, a VCG could also be correlated with virulence (Katan et al. 1989; Manicom et al. 1990).

There is a correlation between forma specialis, VCG and pathogenic races. Sometimes all isolates of a forma specials correspond to a single VCG (Puhalla 1985). Within certain formae speciales, vegetative compatibility could be used as a method for identifying and differentiating pathogenic races. For example, a large collection of isolates each of Fo apii race 2, a pathogen of celery and Fo vasinfectum race 3, a pathogen of cotton from diverse geographical locations correspond to a single VCG. FOX isolates pathogenic to crucifers are placed in three distinct VCGs, each containing isolates pathogenic to a specific host. All Fo niveum isolates are placed in three VCGs. VCG1 consist of all race 2 isolates, VCG 2 comprises of race 1 isolates from USA (all areas except Florida), Australia and Taiwan. VCG3 includes isolates from Florida.

Vegetative compatibility cannot always be used to identify races within a forma specialis because more than one race has been reported in a single VCG; also, isolates of single race may as well belong to different VCGs, e.g. four VCGs have been identified in f.sp.pisi, races 1 and 6 constitute single VCG, race 5 another and race 2 isolates were placed in two VCGs. Eight VCGs are identified in Fomelonis, out of which one contains isolates of different pathogenic races (Jacobson and Gordon 1988). Eleven VCGs have been identified in Fo cubense isolates from all over the world (Ploetz 1990; Ploetz and Correll 1988); a single VCG comprises of multiple races and a given race may belong to multiple VCGs. Similarly, in Fo lycopersici, three known races form single VCG as well isolates of one race are placed in multiple VCGs viz., race1 isolates belong to 41 different VCGs (Elias and Schneider 1991) and 46 distinct VCGs have been identified amongst a collection of isolates pathogenic to asparagus during greenhouse assay for testing pathogenicity (Elmer and Stephens 1989).

The high degree of VCG diversity has been observed in pathogenic and non-pathogenic strains of F. oxysporum. The mutations amongst isolates of single VCG lead to changes in virulence, which could be strong, weak, or non-existent (Leslie 1993, 1996). In general, RFLP patterns of mitochondrial (mt) DNA are identical within a VCG but vary between different VCGs of same formae speciales. All 44 isolates of Fo albedinis recovered from entire geographical range of disease occurrence were represented by a single VCG because of similarity in mt and nuclear DNA (Tantaoui et al. 1996). Four different models have been proposed to explain evolutionary relationships between VCG-race diversity in F. oxysporum (Kistler and Momol 1990). The VCG-race diversity is supported by genomic DNA and mitochondrial DNA (mt DNA), restriction fragment length polymorphism (RFLP) profiles. In F. oxysporum, f. sp. melonisa single VCG is shown to be associated with multiple races, e.g. VCG 0134 is associated with all four known races race 0, race1 and race1, 2 within f.sp. melonis; these races have identical mt DNA (Jacobson and Gordon 1990a, b) and nuclear DNA haplotypes (Schroeder and Gordon 1993). Minor genetic variation results in one pathogenic race giving rise to another; for example, race 3 was first identified in a field where race 2 of f.sp.lycopersici was already present. Isolates of race 2 and 3 constitute a single VCG (Elias and Schneider 1991) with identical isozyme (Elias and Schneider 1992) and nuclear DNA profiles. Co-occuring pathogenic and non-pathogenic Fo strains have similar mt DNA haplotype (Gordon and Okamoto 1992) or IGS haplotype (Appel and Gordon 1995) and placed in single VCG.

At times, non-pathogenic and pathogenic isolates are vegetatively compatible owing to a coincidental sharing of alleles at the loci-governing vegetative compatibility. The inter-isolate transfer of mt DNA through hyphal anastomosis has been reported. For example, in California, eight non-pathogenic isolates from a single field exhibited identical mtDNA haplotype and varying nuclear DNA fingerprints. Similarly, Fo f.sp. vasinfectum (Katan and Katan 1988) and f.sp. spinaciae (Fiely et al. 1995) were different from root colonizing non-pathogenic F. oxysporum isolates based on vegetative compatibility, whereas non-pathogenic isolates of F. oxysporum associated with cyclamen were similar to pathogenic isolates based on polymorphisms in the intergeneric spacer region (IGS) of nuclear rDNA (Woudt et al. 1995). Hence, VCGs are not markers for pathogenicity.

Diversity of Plant–Fungal Interactions

Fungal–plant interactions are complex, diverse and give rise to morphological and physiological alterations in both partners. Fungi produce species-specific signals and employ species-specific mechanisms during interactions with plant host. The outcome of plant–fungal interactions can be saprophytic, symbiotic and pathogenic based on receptors and expression pattern of defence-related plant proteins which interact with specific fungus-derived molecules (Grigoriev 2013). Fungal endophytes become pathogenic if they are able to evade the plant’s innate immunity that comprises physical barriers, mechanism of programmed cell death and production of antimicrobial compounds (Dangl and Jones 2001; Brundrett 2004).

The interactions between plants and their pathogens are constantly evolving, wherein pathogens employ innovative strategies to cause vascular infection. The process of vascular infection by F. oxysporum is complex and requires a series of highly regulated processes, adhesion, penetration and colonization. Pathogenic fungi form feeding structures similar to symbiotic fungi to establish obligate relationships with plants (Corradi and Bonfante 2012). Fungal pathogens after adhesion gain access into plant interior through stomata and wounds in leaf and stem tissue. However, in several cases, cell wall degrading enzymes (CWDEs) and secondary metabolites secreted by fungus facilitate penetration. Once pathogen penetrates host, it secretes protein effectors that suppress plant defence responses and promotes invasion (Lo Presti et al. 2015). Moreover, several morphological and biochemical alterations occur so that pathogenic fungi take over and utilize host metabolic pathways for their growth and development (Zeilinger et al. 2015).

Plant pathogenic fungal species have been classified as biotrophs, hemibiotrophs and necrotrophs, each interacting differently with their host plants. Pathogenic fusarium employ various infection strategies like biotropic (pathogen that colonizes living plant tissue and obtains nutrient from them), necrotrophic (pathogen that kills host cell and obtains nutrient from dead cells) and hemibiotrophic (pathogens that are initially biotrophic and subsequently necrotrophic) varying in mode of interaction. Biotrophic pathogens interact with the host through specialized hyphae which secrete host-specific effectors that suppress host immunity at interfacial zone (Perfect and Green 2001; Yi and Valent 2013). For example, powdery mildews develop primary and appressorial germ tubes on the plant cuticle and breach the cell wall using a combination of mechanical force and CWDEs (Takahashi 1985; Pryce-Jones et al. 1999). After plant cell wall penetration, a close metabolic interaction between plant host and biotrophic pathogen is established (Horbach et al. 2011). Subsequently, the aim of fungus is to block host defence and utilizes host processes for feeding and growth (Giraldo et al. 2013; Yi and Valent 2013). In necrotrophic pathogens, virulence has been correlated with toxin synthesis (Wang et al. 2014). A combination of CWDEs, reactive oxygen species (ROS) and or toxins destroys host cells, their nutrients are released (Kistler and Momol 1990) which results in colonization of plant host (Wolpert et al. 2002). Hemibiotrophic pathogens are initially biotrophic later switching to a necrotrophic lifestyle (Struck 2006; Gardiner et al. 2013). The biotroph–necrotroph switch in hemibiotrophs depends on molecular and physiological factors. Several hemibiotrophs require extended periods to establish infection while for others, the switch to necrotrophy is rapid (Kabbage et al. 2015). From evolutionary perspective, biotrophy is primitive while necrotrophy is a recent phenomenon (Pieterse et al. 2009); hemibiotrophy is a transitional infection strategy for pathogenic fungi (Horbach et al. 2011). The infection strategy of necrotrophic fungi is less complex than that of obligate biotrophs. Necrotrophs exhibit restricted physiological interaction with plant host on account of poorly developed infection structures and smaller number of biochemical compounds required for host penetration.

A mutation in either, fungal pathogen or host receptor genes alters pathogen–plant interactions from resistant to susceptible or vice versa (Stracke et al. 2002; Giraldo and Valent 2013).

Fusarium Pathogenicity and Pathogenicity Factors

Pathogenesis is the complete process describing disease development in the host, from initial infection to production of symptoms (Lucas 1998), and pathogenicity is the ability for pathogenesis. F. oxysporum initially penetrates roots asymptomatically; subsequently, it colonizes vascular tissue and triggers massive wilting, necrosis and chlorosis of aerial produce. Certain species-producing toxin, fusaric acid initially infect floral tissue during anthesis, spreads to flower through central axis of inflorescence, eventually damaging and contaminating grains with toxins (Gardiner et al. 2013).

Fusarium pathogens use both general and specific pathogenicity factors/mechanisms to invade their hosts (Fig. 10.3). Hydrolytic enzymes involved in plant cell wall degradation and components of cellular signalling pathways, which are often required for systemic pathogen invasion, comprise general pathogenicity factors, whereas production and secretion of effectors and host-specific toxin are specific pathogenicity factors. The counter defence mechanism of plants plays significant role in pathogenesis and categorised as general and specific (Poppenberger et al. 2003). General defence mechanisms encompass production of antifungal proteins and activation of defence signalling pathways, whereas pathogen-specific include recognition of specific pathogen effectors by plant resistance gene products and detoxification of pathogen-specific toxins (Proctor et al. 2007).The specific properties that discriminate endophytic strains from closely related non-endophytic strains have been identified in several studies (Taghavi et al. 2010; Mitter et al. 2013; Amadou et al. 2008; Tisserant et al. 2013; Tian et al. 2012; Karpinets et al. 2014).

Fig. 10.3
figure 3

Pathogenicity factors and host defence mechanisms during Fusarium–plant interactions (reproduced from Ma et al. 2013)

Full size image

Comparative studies analysing genomic and metabolic network revealed major difference in cellular processes and metabolic capabilities of pathogenic (n = 36) and mutualistic (n = 28) plant microbes (Tian et al. 2012). Genes regulating biosynthetic processes and functions were enriched and more diversed amongst plant mutualists, while those controlling degradation and host invasion were detected in phytopathogens. Pathogens possess genes and regulons required for plant penetration and colonization (Wright et al. 2013). Moreover, mutualists utilize more stress-related compounds, whereas pathogens compound from plant cell wall. Genes encoding secretion systems, required to invade the host plant, were present in pathogen genome, while those encoding nitrogen fixation proteins and ribulose bisphosphate carboxylase/oxygenase (RuBisCO) proteins were specifically present in mutualistic bacteria (Karpinets et al. 2014). Bacteria with relatively large genomes often successfully colonize a wide range of unrelated plant hosts and soils, whereas strains with smaller genomes have a narrow host range (Mitter et al. 2013).

Fungi generally secrete a mixture of CWDEs to enter plant cells and secrete effectors, toxins or plant hormone-like compounds that manipulate the plants’ physiology for its invasion and growth.

Cell Wall Degrading Enzymes

Fungal cell wall degrading enzyme (CWDE) system comprises of peroxidases and laccases for the degradation of lignin and glycoside hydrolases. Fungus secretes cellulases, hemicellulases and pectinases for degradation of cellulose, hemicellulose and pectin, respectively (Kubicek 2013). Genomics analysis of 103 fungi revealed that a large number of carbohydrate-active enzymes (CAZymes) such as carbohydrate esterase and pectate lyases (PL) are present in fungal pathogen compared to saprophytic fungi (Liao et al. 2013). Thus there is an upregulation of genes encoding CWDEs in Fusarium graminearum, the hemibiotrophic pathogen and Magnaporthe oryzae during infection of plant hosts (Kawahara et al. 2012; Zhao et al. 2013). In contrast, biotroph genomes possess few plant CWDEs encoding genes which completely lack glycoside hydrolase family 6(GH6) endoglucanase and cellobiohydrolase genes (Zhao et al. 2013).

Effector Proteins

Small effector proteins deregulate plant immune responses and facilitate pathogen in colonizing plant host (Rovenich et al. 2014). Fungal effectors are either apoplastic, those secreted into the plant extracellular component and cytoplasmic, those accumulated in plant membrane rich structure associated with invasive fungal hyphae (Giraldo et al. 2013). Apoplastic effectors include protease inhibitors that destroy host proteases. Plant proteases protect fungal cell walls against plant chitinase and small molecules minimising ROS levels. Host plant resistance (R) proteins recognize cytoplasmic effectors, thereby triggering the hypersensitive response (HR), a reaction characterized by rapid cell death in local infection region and thus blocks pathogen growth and spread (Giraldo and Valent 2013). Avirulence proteins are a type of cytoplasmic effectors. The interaction between an avr gene of pathogens and cognate resistance (R) gene of host leads to HR-mediated activation of host defence mechanism which prevents the pathogen invasion. This is an effector-triggered immunity (ETI) and is exemplified by Cf 9 and avr 9 genes for the Cladiosporum fulvum-tomato pathosystem. The product of fungal race-specific avr 9 gene induces HR on tomato plants carrying the complementary resistance gene Cf 9. The fungal races virulent on Cf 9 tomato genotypes lack avr 9 gene. The genome analysis predicts that biotrophic maize pathogen Ustilago maydis encodes ~550 secreted proteins. Several of these are upregulated during host colonization (Djamei and Kahmann 2012). U. maydis secretes ‘core’ and organ-specific effectors. Core effectors suppress plant defence during the penetration stage and organ-specific effectors infect different plant tissue (Skibbe et al. 2010; Djamei and Kahmann 2012). U. maydis genome has effector-encoding gene clusters. There are 23 genes in the largest effector gene cluster, 19A. These are differentialy induced when different plant organs are colonized. It has been observed that deletion of complete 19A cluster abolished tumour formation in maize plants, whereas deletion of individual genes showed minor reduction in virulence (Kamper et al. 2006; Brefort et al. 2014).

Although, effectors are mostly proteins but a few are metabolites also. Fungal pathogens of genera Cochliobolus, Alternaria and Pyrenspora species secrete host-specific toxins (Tsuge et al. 2013), e.g. Fusarium verticillioides produces fumonisin (Arias et al. 2012) and M. oryzae pyrichalasin H and Ace 1 (avirulence conferring enzyme 1). Secondary metabolite-synthesized by ace 1 has not been identified yet (Collemare et al. 2008; Yi and Valent 2013). Pathogenic fungi deliver small non-coding RNAs into plant host cell to suppress plant immunity. Botrytis cinerea small RNAs silence genes confer immunity in Arabidopsis and tomato through hijacking host RNA machinery (Weiberg et al. 2013). Pathogens, e.g. Cladosporium fulvum and M. oryzae escape plant defence by secreting LysM effectors that bind to soluble chitin fragment and prevent them from detection by plant chitin receptors.

Signalling During Fungal–Plant Pathogen Interaction

The most critical step in fungal–plant interactions is the recognition of appropriate plant host. The process begins prior to direct contact between partners. Fungi detect chemical and physical signals and respond through differentiation, movement to an appropriate infection site, and/or formation of invasion-related structures (Kumamoto 2008; Bonfante and Genre 2010). The following section summarizes the current knowledge on the signals as well as signalling pathways involved in plant–fungal interactions with focus on fungal partner.

Signalling Mechanism

Root Exudates

Plant roots release both low and high molecular weight substances into rhizosphere. Amino acids, ion-free oxygen, sugars, phenolics and secondary metabolites are low molecular weight substances, whereas mucilage and proteins are high molecular weight substances (Bais et al. 2006). Root exudates can be produced both constitutively (so-called phytoanticipins) and in response to pathogen attack (so-called phytoalexins) (Baetz and Martinoia 2014). When soil-borne pathogen F.graminearum attacks barley, phenylpropanoids are released by its roots (Boddu et al. 2006). Similarly, the production of terpenes in barley roots is triggered by Cochliobolus sativus and Fusarium culmorum (Fiers et al. 2013).

Flavonoids also contribute towards signalling in plant–fungus interaction. They exert both positive and negative effect on fungal phytopathogens. On one hand, flavonoid inhibits spore germination and hyphal growth in several fungal pathogens, whereas on the other have a stimulatory effect. In case of F. solani f.sp. pisi, the isoflavonoid pisatin induces expression of pda1 encoding a pisatin demethylase, a virulence factor of this fungus (Khan et al. 2003). Recent studies revealed that class III peroxidase (POX) secreted by tomato roots function in chemotrophic sensing by F. oxysporum via a pheromone receptor homologue and MAPK signalling (Turrà et al. 2015).

Oxylipins

Oxylipins are an oxygenated lipid secondary metabolites produced by plant and fungi. They are implicated in pathogenicity and promote disease progression. Fungal oxylipins act as endogenous signalling molecule that manipulates host lipid metabolism and alter its defence response (Tsitsigiannis and Keller 2007; Brodhagen et al. 2008). In contrast, plant oxylipins (jasmonates, JA) directly influence survival of invasive structures (Calvo et al. 1999), reproduction and production of secondary metabolites (Burow et al. 1997) in fungi.

Oxylipins act by inducing JA-responsive genes (Thatcher et al. 2009). JA signalling mediated by protein, coronatine insensitive (COL1) is responsible for susceptibility of Arabidopsis thaliana to Fo wilt. Oxylipins bind to G protein-coupled receptors (GPCRs) which induce cAMP signalling. Recently, it has been reported that cAMP signalling stimulated by plant oxylipin was absent in Gpr D (GPCR-encoding) mutant Aspergillus nidulans (Affedt et al. 2012). It was also reported that in the soil-borne plant pathogen Aspergillus flavus, endogenous oxylipins mediate, spore and sclerotia production and the biosynthesis of aflatoxin, which are regulated by Gpr C and Gpr D; Gpr C and Gpr D could be thus important for fungal–plant interactions.

Reactive Oxygen Species

Oxypilin-mediated signalling is linked to ROS-stimulated cell signalling during plant–fungus interaction, ROS interacts with phosphorylation cascades and controls transcription factors. Thus, mediating defence gene expression or oxypilins are generated through non-enzymatic oxygenation by ROS (Reverberi et al. 2012). In invading fungi nicotinamide adenine dinucleotide phosphate (NADPH) oxidase enzymes mediate production of superoxide. ROS accumulates at the plant–fungus interface and acts as signals for triggering attack and counterattack responses. In M. grisea, NOX1 and NOX2, NADPH oxidases trigger a local oxidative burst during plant infection (Egan et al. 2007). NOX1 and NOX2 are associated with appressorium formation. To establish infection, fungal pathogen must overcome plant’s oxidative defense by employing ROS scavenging enzymes and modifying ROS accumulation in plant host, e.g. DES1 (defence suppressor 1) in M. oryzae (Chi et al. 2009), Leucine zipper (bzip) transcription factor and yes-associated protein (YAP1) in U. maydis (Molina and Kahmann 2007).

Plant Surface Signals

Plant signals, cutin monomers and leaf waxes trigger appresssorium formation in foliar rice pathogen M.oryzae (Liu et al. 2011; Perez-Nadales et al. 2014). Appressorium formation is mediated through multicopy suppression of budding defect2 (Msb2) signalling mucin and synthetic high osmolarity sensitive 1 (Sho1) tetraspanin protein (Lanver et al. 2014) present on fungus. Msb2 also plays an important role in non-appressorium forming root-infecting, F. oxysporum by regulating plant infection and invasive growth through phosphorylating the Fmk1 MAPK in response to plant surface signals (Perez-Nadales and Di Pietro 2011). In pathogenic fungi, MAPK regulates the mechanical and enzymatic penetration of the host plant while the plant uses MAPK signalling for activation of immunity. In M. oryzae, cAMP-PKA signalling pathway controls plant surface recognition (Zhao and Xu 2007; Li et al. 2012) while Pmk1 (pathogenicity MAPK) stimulates appressorium formation and fungal growth in plant tissues (Xu and Hamer 1996). The membrane protein pth11 (aGPCR), that recognizes surface hydrophobicity, functions upstream of the cAMP-PKA pathway. It has been shown that although appressorium formation continues in PTH11 gene deletion mutant, they have reduced virulence (DeZwaan et al. 1999). This gives clear evidence of the overlapping roles of the Pth11 receptor and the signalling mucin Msb2 (which acts upstream of the Pmk1 MAPK cascade) in sensing surface hydrophobicity and regulation of appressorium formation (Xu and Hamer 1996; Liu et al. 2011).

In U.maydis, Kpp2 MAPK mediates virulence-related processes as filamentation and appressorium formation (Mendoza-Mendoza et al. 2009). Hence, there is a complex cross-talk of MAPK signalling with cAMP pathway. The two pathways appear to be connected at the Gpa3G protein subunit and the perforin 1 (Prf1) transcription factor. The Prf1 carries sequence motifs specific for PKA and MAPK-dependent phosphorylation, essential for its function (Bolker 2001). Similarly, proteins involved in calcium signalling are required for appressorium formation, turgor generation and host penetration in M.oryzae (Liu and Kolattukudy 1999). Hence, rice blast fungus is a model for deciphering the interplay of various signalling pathways in development of pathogenic potential.

Fungal Metabolic Diversity

The metabolic diversity of fungus determines whether fungus-plant interaction is beneficial or harmful (Zeilinger et al. 2015). Secondary metabolites enable fungus to colonize plant host systematically, survive in its niche and determines its virulence (Keller et al. 2005). Fungal species produce plant–specific secondary metabolites. The environmental changes affect the production of such secondary metabolites. Secondary metabolites associated with iron uptake govern the virulence potential of A. brassicicola, C. heterostrophus, C. miyabeanus and F. graminearum on their specific host plants (Oide et al. 2006).

Signalling Pathways

Plant distinguishes whether fungus is friend or a foe at multiple levels (Fig. 10.4). The first level is regulated by the receptor protein, pattern recognition receptors (PRRs) located in the plasma membrane. PRRs recognise microbial-associated molecular patterns (MAMPs) and pathogen-associated molecular patterns (PAMPs). These lead to activation of PAMP-triggered immunity (PT1) via calcium signalling and mitogen-activated protein kinase (MAPK) cascades. MAPK is two-component system that regulates pathogenicity in fungal pathogens. It comprises of a membrane-bound histidine kinase which sense specific environmental stimuli and a response regulator that transmits the signal to a downstream pathway (Catlet et al. 2003). MAPK regulates stress responses and virulence in C. heterotrophus and F. graminearum (Oide et al. 2010) and in Alternaria brassicicola (Cho et al. 2009). Pathogens induce effector-triggered susceptibility (ETS) by blocking PTI response through effector proteins (Kazan and Lyons 2014). Calcium signalling is common cascade for plants to open a dialogue with their fungal partners because intracellular calcium levels are elevated during pathogenic as well as beneficial interactions (Navazio et al. 2007).

Fig. 10.4
figure 4

Signalling in plant–fungal pathogen interactions. (reproduced from Zeilinger et al. 2015)

Full size image

MAPK Cascade and Its Role in Virulence

The MAPK cascades of both partners help establish a molecular dialogue between plant and fungus (Hamel et al. 2012). MAPKs are organised as cascades consisting of three interlinked protein kinases, MAPK kinase (MAP3K), MAPK kinase (MAP2K) and MAPK, sequentially activated by phosphorylation (Widmann et al. 1999). The MAPK FmK1, an orthologue of the yeast Fus3/KSSI MAPKs, is essential for virulence of F. oxysporum on tamoto plants (Di Pietro et al. 2001). It is widely conserved and determines pathogenicity in all plant pathogenic fungi (Rispail et al. 2009). It is essentially required for all infection-related processes, invasive growth, fusion of vegetative hyphae and root adhesion (Di Pietro et al. 2001; Prados Rosales and Di Pietro 2008). Upstream and downstream components of this signalling cascade have been elucidated. The transcription factor Ste12 functions downstream of Fmk1 and regulates invasive growth of pathogen during plant infection (Rispail and Di Pietro 2009), and Msb2, a transmembrane protein is an upstream component of this cascade (Perez-Nadales and Di Pietro 2011). In addition, Saccharomyces cerevisae high osmolarity (Hog 1) and cell integrity Mpk1 gene orthologues have been identified in F. oxysporum also. The GTPase Rho1, which function upstream of Mpk1, was essential for morphogenesis and pathogenicity (Martinez-Rocha et al. 2008).

Fusarium Genomics

The genomes of three economically important and phylogenetically diverse species, Fusarium graminearum (Fg) strain PH-1, Fusarium verticilloides (Fv) strain 7600 and Fusarium oxysporum f.sp. lycopersici (Fol) strain 4287 were compared and analysed. Fg strain H-1 causes head scab disease of small grain cereals, Fv strain 7600 is a maize pathogen-producing mycotoxin, fumonisin that contaminate grain and Fol strain 4287 is a tomato pathogen. The fully completed genome of F. graminearum PH-1 and its manually curated annotation is available at ensemble databank (King et al. 2015). Whole shot gun genome of Fol strain 4287 and Fv strain 7600 is available at BROAD Institute Website (Ma et al. 2010). The Fol genome (60 megabase) is about 44% larger to Fv (42 Mb), and 65% larger to Fg (36 Mb). Fol genome has a greater number of protein-encoding genes. 28% of the F. oxysporum genome corresponds to short interspersed elements (SINES) and class II transposable elements (Table 10.2). Fusarium genome is compartmentalized into core and accessory genomes. Core genome is identical in all Fusarium species and encodes for growth and survival, whereas accessory genome varies amongst formae speciales and characterizes host specialization, virulence and production of secondary metabolites. In Fusarium solani f. sp. pisi, the three dispensible chromosomes have been linked to habitat specialization (Coleman et al. 2009) and pathogenicity towards pea (Han et al. 2001). In Fusarium oxysporum f. sp. lycopersici, chromosome 14 has been shown to convert a non-virulent strain into a virulent towards tomato via acquisition of the entire chromosome (Ma et al. 2010). In F. graminearum, regions of high single nucleotide polymorphism (SNP) density were found at the ends of chromosomes and in interstitial regions on three of the four chromosomes (Cuomo et al. 2007).

Table 10.2 Genome comparison amongst different Fusarium species
Full size table

A total of over 9000 conserved syntenic orthologues were identified amongst Fol, Fv and Fg genomes (Ma et al. 2010). Fol and Fv orthologues display 91% nucleotide sequence identity within themselves and 85% with Fg orthologues. The orthologues of three species, Fol, Fv and Fg are enriched for predicted transcription factors (P = 2.6 × 10ˉ6), lytic enzymes (p = 001) and transmembrane transporters (p = 7 × 10ˉ9) when compared to other ascomycete genomes. In all three genomes, a total of 46 secondary metabolite synthesis (SMB) gene clusters have been identified. Microarray analysis confirmed that the genes in 14 of 18 Fg and 10 of 16 Fv SMB gene clusters were co-expressed. Ten out 14 Fg and eight out of the 10 Fv SMB gene clusters are co-expressed and novel (Ma et al. 2010).

Lineage-Specific (LS) Genomic Region

Lineage-specific (LS) genomic region, also known as supernumerary chromosome, constitutes accessory genome. They are usually small (<2 MB) and specific for forma specialis. They acquire foreign genes (i.e. xenologs) through horizontal transfer of an entire plasmid or chromosome from other Fusarium species and subsequent integration into the core chromosome. LS chromosomes are characterized by, (a) lack of housekeeping genes, (b) G + C content different than core chromosomal complement, (c) varying within related species and (d) 95% of transposable elements (TEs) present in an entire genome. The LS region harbours genes putatively related to host–pathogen interaction or pathogenicity (Ma et al. 2010). In all, 20% LS genes have been identified functionally. They encode for secreted effectors, transcription factors and virulence factors, involved in signal transduction. Analysis of genome sequence data suggests that F. oxysporum LS region differs considerably in strains with varying host specificities.

Comparisons amongst Fol, Fg and Fv genomes revealed the presence of four lineage-specific (LS) chromosomes. The genome assembly of Fol, Fv and Fg has 15, 11 and 4 chromosomes, respectively (Table 10.2).The number of chromosomes in Fg are less as compared to Fv and Fo due to chromosome fusion. The fusion occurs in high diversity regions (Cuomo et al. 2007). The genomic region in Fol is larger due to the presence of additional and unique sequences in extra chromosomes. All 11-mapped chromosomes, except for their telomere-proximal regions in the Fv assembly (41.1 Mb), correspond to 11 chromosomes in Fol (41.8 Mb). Syntenic region in Fol, Fg and Fv are ‘core’ region of genome. The core region of Fol has 80% similarity with that of Fg and 90% with that of Fv. About 40% of the Fol genome assembly is designated as Fol lineage-specific (Fol LS) region. The Fol LS regions include four entire chromosome (3, 6, 14, and 15), parts of chromosome 1 and 2 (scaffold 27 and scaffold 31, respectively), and most of the small scaffolds not adhered to the optical map. The Fol LS region is 19 Mb, 28% of which is transposable elements (TEs). These are long interspersed nuclear elements (LINEs), retro elements copia-like and gypsy-like LTR retrotransposons, short interspersed nuclear elements (SINEs) and DNA transposons. DNA transposon classes like Pogo, hAT-like elements and MITEs are well represented in Fol. All in, about 74% TEs in Fol LS region are identifiable.

Fol genome has one intra-chromosomal and two inter-chromosomal segmental duplications, totalling approximately 7 Mb. Overall, these regions share 99% sequence identity indicating recent duplication events. Proteins encoded by 20% of Fol LS region are known. These are related to pathogenicity and include secreted effectors, transcription factors and virulence factors, involved in signal transduction and ethylene induction (Qutob et al. 2006). The enzymes that degrade or modify plant or fungal cell walls (Ma et al. 2010) are related to pathogenicity. Many of these enzymes have been reported to be expressed during early stages of infection on tomato root. It also harbours genes that encode for lipid metabolism and lipid-derived secondary messengers (Ma et al. 2010). These genes play important role in fungal pathogenicity. Fol LS region also has transcription factor sequences related to FTF 1 and specifically involved during early stages of F. oxysporum f. sp. phaseoli infection to its host (Ramos et al. 2007). The core genome in all F. solani isolates is well conserved (Coleman et al. 2009). Its accessory genome contains three LS regions distinct from its ‘core’ genome. F. solani LS region is distinct from that of Fol.

Therefore in conclusion, Fusarium species have similar core region and distinct LS regions. The LS regions are distinct in genes related to host–pathogen interactions.

Secondary Metabolite Gene Clusters

Fusarium species produce an array of bioactive secondary metabolites of which polyketides and non-ribosomal peptides are most abundant (Table 10.3). The comparative analyses of non-ribosomal peptide synthetases (NRPSs) and polyketide synthases (PKSs) from ten different Fusarium species, such as F. avenaceum, F. acuminatum, F. culmorum, F. equiseti, F. fujikuroi, F. graminearum (two strains), F. oxysporum (12 strains), F. pseudograminearum, F. solani and F. verticillioides, led to the identification of 52 NRPS and 52 PKSs orthology groups (Hansen et al. 2015). All NRPS and PKS were not functional in the strains analysed. A total of eight NRPSs (NRPS2-4, 6, 10-13) and two PKSs (PKS3, PKS7) were conserved in all the strains analysed. However, the products of the majority of PKSs and NRPSs are unknown. For example, in F. graminearum, the products of, only 8/15 PKSs and 3/19 NRPSs are known (Hansen et al. 2012b; Jørgensen et al. 2014; Sørensen et al. 2014). Some of the RPSs and PKSs genes have not been linked to metabolite production and may possibly represent pseudogenes.

Table 10.3 Different protein families including secondary metabolites produced in pathogenic and saprophytic fungi
Full size table

Polyketide Synthases

Fungal pathogens synthesize polyketides from carboxylic acid derivatives (acetyl-CoA and malonyl-CoA) (Hopwood and Khosla 1992). Fungal PKSs contain five to eight functional domains and form two major groups based on their domain content: non-reducing PKSs (NR-PKSs) in which carbonyl groups are not reduced and reducing PKSs (R-PKS) in which carbonyl groups are partially or fully reduced. Fumonisin B, bikaverin, fusarubin and aurofusarin are examples of NR-PKS whereas fusaric acid and fusarielin are those of R-PKS. Fusarin C is NR-PKS with NRPS module. Zearalenone is both NR-PKS and R-PKS. Hansen et al. (2015) analysed the sequenced Fusarium strains through Blast P. The KS domains of PKS were extracted, phylogenetically analysed to identify PKS orthology group. Out of 52 different PKS genes, PKS3 and PKS7 were present in all the strains. In addition, PKS8 was also highly conserved amongst Fusarium strains although a part of KS domain was absent in F. culmorum, entire gene was absent in F. oxysporum f. sp. melonis. F. graminearum strain CS3005 shares 15 PKSs with Fg strain PH-1 which includes PKSs for biosynthesis of aurofusarin (PKS12), fusarubins PKS3, fusarins (PKS10), fusaristatin (PKS6 + NRPS7, and zearalenone (PKS4 + 13), (Sørensen et al. 2014) and a unique PKS (PKS52) which might have been acquired from Colletotrichum strains through horizontal gene transfer (HGT). F. graminearum and F. culmorum were closely related as there is overlap of 13 PKSs. PKS2 and PKS9 were absent in F. culmorum. Out of 14 putative PKSs identified in F. pseudograminearum, 13 have orthologue in F. graminearum, PKS40 was specific to F. pseudograminearum and encodes for W493A and B (a polyketide non-ribosomal peptide). Recently, PKSs and NRPS identified in F. fujikuroi earlier by Wiemann et al. (2013) have been renumbered (Hansen et al. 2015). Of the 17 PKSs identified in F. fujikuroi, 13 have orthologues in F. verticillioides and F. oxysporum, including bikaverin synthase (PKS16=BIKI) and hybrid PKS-NRPS (PKS18). PKSs ranging from 10 to 14 have been identified amongst completely sequenced 12 strains of different formae speciales. Six PKs were present in all 12 strains (PKS3, 7, 18, 20, 21 and 27), although the AT, KR, KS and MET domains are missing in F. oxysporum strain (F05176). The analysis shows that F. graminearum had the highest number of species-specific PKSs followed by F. solani, F. verticillioides and F. oxysporum, respectively (Hansen et al. 2015). The polyketide synthase gene clusters in Fol, Fv, Fg and Fs have been characterized (Ma et al. 2010).

Non-ribosomal Peptide Synthetases (NRPSs)

These play a role in fungal pathogenesis because the products of several NRPSs are proven virulence factors, e.g. enniatin is essential for virulence of Fusarium avenaceum on potatoes (Herrmann et al. 1996), AM-toxin for Alternaria alternata on apple (Johnson et al. 2000) and HC-toxin for Cochliobolus carbonum race1 on corn (Walton 1996). The NRPS genes are up to 63 kb in size and form products in conjunction with other mega synthetases (Straight et al. 2007).

These are large multi-modular enzyme assembly lines (NRPSs) that synthesize non-ribosomal peptides (NRPs). The NRPSs consist of modules, each possessing catalytic domains in a specific order facilitating the sequential initiation and modification of the growing peptide chain. It has the adenylation domain (A) which recognizes the specific amino acid substrate, which is then transferred by the peptide acyl carrier domain (T or PCP) to the condensation domain (C) where the formation of the peptide bond takes place. These core domains are often supported by tailoring domains such as: thioesterase domains (TE) for cleavage or cyclization of the final peptide, reductase domain (R) for reducing the final peptide, epimerization domains (E) which can change the epimeric form of the amino acid substrate, cyclization domain (Cy) for modification of serines, and threonines and cysteines and N-methylation domains (NM) (Strieker et al. 2010). NRPSs were identified through Blast P analysis using a selected panel of variable A domains. The A domains were extracted from each NRPS and used for phylogenetic analyses. In all, six NRPS (NRPS2, 3, 6, 10, 11 and 12) were detected in all sequenced strains. NRPS 2 and NRPS6 are responsible for production of the siderophores ferricrocin and fusarinine, respectively (Oide et al. 2006; Tobiasen et al. 2007). NRPS6-produced fusarinine acts as an extracellular siderophore and is important for plant infection. NRPS4 was identified in all strains except F. oxysporum (Fo5176) and is reported to be involved in surface hydrophobicity in F. graminearum (Hansen et al. 2012a). In F. graminearum, a total of 16 NRPS genes were identified, of which NRPS32 was species-specific. In F. culmorum, 18 NRPS were identified. In F. fujikuroi, all except one (NRPS31) was common to those present in other sequenced Fusarium species. Out of a total of the 12 F. oxysporum strains, two F. oxysporum f. sp. cubense strain NRRL54006 and F. oxysporum f. sp. pisi strain HDV247 had identical distribution of NRPS genes. All F. oxysporum strains possess nine common NRPSs. These included two siderophore synthetases, NRPS2 and NRPS6 and the enniantin/beauvericin synthetase NRPS22. NRPS1 which encodes malonichrome, a type of siderophore is absent in F. oxysporum f. sp. raphanin and F. oxysporum Fo5156 and F. fujikuroi. Further, NRPS39, an orthologue of ferrirhodin synthatase, FNR1 was present in seven F. oxysporum species. Malonichrome (NRPS1), ferricrocin (NRPS2) and ferrirhodin (NRPS39) has similar domain structure (ATC-ATC-ATCTCTC).

Molecular Basis of Pathogenicity

Our understanding of molecular mechanisms involved in pathogenicity has improved through the genome sequencing and application of forward and reverse genetics. Michielse et al. (2009) identified more than 100 potential pathogenicity genes in Fusarium oxysporum f. sp. lycopersici upon analysis of 10,000 transformants for pathogenicity. With sequencing of more Fusarium genomes, similar genes have also been identified in Fusarium species other than f. sp. lycopersici (Kazan et al. 2012; Walter et al. 2010). A few with known functions have been listed (Table 10.4; Michielse et al. 2009; Sutherland et al. 2013). Functional characterization of putative genes indicates that those encoding for cell wall integrity, cell wall degrading enzymes, transcriptional regulators for carbon and nitrogen metabolism, cellular processes, such as amino acid and lipid metabolism, cell wall remodelling, protein translocation and degradation, seem to be important for complete pathogenicity of F. oxysporum. MAPK and cyclic AMP-protein kinase A (CAMP-PKA) cascade regulate virulence in Fo (Delgado-Jarana et al. 2005; Di Pietro et al. 2001; Jain et al. 2002, 2003, 2005). Cell wall integrity is necessary for invasive growth and resistance to plant defence compounds (Caracuel et al. 2005; Madrid et al. 2003; Martinez-Rocha et al. 2008; Martin-Udiroz et al. 2004, 2008). Cell wall degrading enzymes have been implicated in root penetration and colonization, but their role in infection process is not yet completely known. F. oxysporum f. sp. lycopersici virulence remains unaffected upon inactivation of individual genes, e.g. pectate lyase gene plt1, xylanase genes xyl3, xyl4 and xyl5, polygalacturase genes, pg1, pg5, pgx4 and subtilase gene prt1. Similarly, deletion of xlnR did not affect virulence. However, targeted disruption of SNF1 reduced virulence as well as expression of various CWDEs indicating that central carbon metabolism plays key role in pathogenicity (Ospina-Giraldo et al. 2003). It has been reported that inactivation of Fnr1 (global nitrogen regulator) abolished the expression of nutrition genes, normally induced during early phase of infection, leading to reduction in pathogenicity. In addition, several plant degrading genes, pH-responsive transcription factors and regulators also play important role in pathogenicity. Moreover, the genes for peroxisome biosynthesis, ion homeostasis and toxin biosynthesis are also related to virulence. All the pathogenicity genes are categorized into basic and specialized pathogenicity genes.

Table 10.4 Potential pathogenicity genes in Fusarium oxysporum f. sp. lycopersici
Full size table

Basic Pathogenicity Genes

The genes are common in Fusarium and other pathogenic fungi. These genes encode essential components of conserved pathways involved in sensing exogenous or endogenous signals. For example, mitogen-activated protein kinase (MAPK) signalling pathways in pathogens (Di Pietro et al. 2001; Hou et al. 2002; Urban et al. 2003), Ras protein (small GTPase) (Bluhm et al. 2007), G-protein signalling component and their downstream pathway (Jain et al. 2002; Park et al. 2012), components of the velvet (LacA/veA/VelB) complex (Lee et al. 2012; Weimann et al. 2010) and cAMP pathway (Garcia-Martinez et al. 2012).

Specialized Pathogenicity Genes

These genes determine the pathogenicity of individual Fusarium species on specific plant host. These include avr genes and six (secreted in xylem) genes, mycotoxin and gibberellins encoding genes. Of these, six are directly involved in host pathogen interaction and encode for hundreds of small secreted proteins that play significant role in determining host specificity. Mycotoxins are additional virulence factors and act in a host or pathogenic-specific manner. A few Fusarium species produce mycotoxin, trichothecene which promotes virulence towards wheat (Triticuma estivum) and maize (Zea mays) but not barley (Hordeum vulgare) (Jansen et al. 2005). The complete gibberellin gene cluster is present in all species within F. fujikuroi species complex but expressed/detected in only three, F. fujikuroi, F. sacchari and F. konzum. During disease development, gibberellins alter host tissues. The genes that encode for CWDEs and hydrolytic enzymes and are significant in gaining access to nutrition during infection are present in all Fusarium genomes. However, very few of these genes have been directly connected to pathogenicity. One exception is FGLI, a secreted lipase gene, which is responsible for virulence of F. graminearum of barley, maize and wheat (Ilgen et al. 2008, Voigt et al. 2005). Further, if FGL1 gene is overexpressed, the virulence of non-pathogenic MAPK mutant on wheat is restored (Salomon et al. 2012).

Six Genes

Host specificity between different races of F.oxysporum f. sp. lycopersici and tomato cultivar is determined by genes encoding small cysteine-rich effector protein termed six (secreted in xylem) (Rep et al. 2004). One of these proteins Avr1 triggers a resistance response in tomato plant carrying resistance (R) gene I-1. Interestingly, Avr1 also functions as a virulence effector by suppressing disease resistance conferred by two other R gene, I-2 and I-3 (Houterman et al. 2007). In Fo, all six genes are located on lineage-specific chromosome 14, also called the pathogenicity chromosome (Ma et al. 2010). Expression of the six genes requires a transcription factor Sge1 located on a core chromosome (Michielse and Rep 2009). Most of six genes are induced in planta.

Fusarium Oxysporum as a Model for Fungal Trans-Kingdom Pathogenicity

F.oxysporum f sp lycopersici isolate FGSC 9935 was the first fungal strain reported to cause disease both on plant (tomato) and mammalian host (immuno-depressed mice) (Ortoneda et al. 2004). The fungal genes encoding Fmk1, MAPK or small G protein Rho1 are essential for its pathogenicity on tomato but not on mice (Di Pietro et al. 2001; Martinez-Rocha et al. 2008), whereas pH response factor Pac C (Caracuel et al. 2003) or the secreted pathogenesis-related 1(PR1) like protein Fpr1 (Prados Rosales et al. 2012) are required for virulence in mice but not on plants. Recently, Hapx, a transcription factor that governs iron homeostasis, was identified in the fungal strain virulent on both plant and animal (Lopez-Berges et al. 2012). Similarly, the velvet protein complex contributes to infection of plants and mammals, in part by promoting the biosynthesis of beauverin, a mycotoxin (Lopez-Berges et al. 2013). The fungal pathogenicity on plants and animals has fundamentally distinct evolutionary origins despite involvement of common virulence proteins.

Evolution of Pathogenicity

The genetic and evolutionary relationships within and amongst formae speciales of F. oxysporum are revealed by sequence analysis of DNA-directed RNA polymerase II subunit and elongation factor-1α (EF-1α) genes (O’Donnell et al. 1998; Baayen et al. 2000; Mbofung et al. 2007). It was revealed that F. oxysporum formae speciales viz.,lili and tulipae were monophyletic (Baayen et al. 2000) while f.sp. asparagi, cubense, dianthi, lycopersici and vasinfectum were polyphyletic (O’Donnell et al. 2000; Baayen et al. 2000; Skovgaard et al. 2001; Cai et al. 2003). The evolution of pathogenesis in FOX isolates can be traced by analysing formae speciales of closely related plant species assuming that pathogens of closely related plant species are also closely related. The phenomenon of gene duplication (GD) and horizontal gene transfer (HGT) are responsible for their constant/continuous diversification. Virulence genes are acquired through HGT and further diversify due to GD and gene loss (Joaramillo et al. 2015; Steindorff et al. 2015). Fusarium genome very well exemplifies the role of HGT in acquiring diversity. A non-pathogenic strain of F. oxysporum f.sp. lycopersici was transformed into a tomato pathogen subsequent to transfer of a pathogenicity chromosome to it (Ma et al. 2013).

Origin of LS Regions

LS region in Fol might have originated in three ways: (i) Fol LS region was present in the last common ancestor of four Fusarium species but lost in Fg, Fv and Fs during vertical transmission, (ii) LS regions arose from the core genome by duplication and divergence of Fol lineage and (iii) LS regions acquired as a result of horizontal transfer. In all, 90% of the Fol genes in core regions have homologues in Fg and Fv. About 50% of the genes on FOL LS regions lack homologues in either Fg or Fv. sequence divergence between Fol and Fv orthologues in core regions was less as compared to Fol and Fg orthologues. The LS genes that have homologues in the other Fusarium species are roughly equally distinct from both Fv and Fg genes indicating that the phylogenetic history of the LS genes differs from genes in the core region of the genome. The distinct evolutionary trend of the core and LS region is supported by the distinct codon usage in LS encoding genes compared to core/conserved genes. The most significant differences were observed for amino acids Ala, Cys, Gln, Glu, Gly, Thr and Val, with a preference for G and C over A and T amongst Fol LS regions. Their third codon positions have higher GC content. Nearly 93% of 1285 LS-encoded proteins are homologous to other ascomycetous fungi. Phylogenetic analysis of 362 proteins sharing homologues in seven ascomycete genomes—Fg, Fv, Fol, Fs, Magnaporthe grisea (Dean et al. 2005), Aspergillus nidulans (Galagan et al. 2005) and Neurospora crassa (Galagan et al. 2003) indicates that they originated within the genus Fusarium but were placed basal to the three most closely related Fusarium species Fg, Fol and Fv. Thus, it is concluded that Fol LS regions originated through horizontal acquisition of genes or gene regions from other Fusarium species.

Host Specificity: Variations in LS Regions

F. oxysporum is a species complex, comprising of several different asexual lineages that are non-pathogenic or pathogenic towards different hosts. Fo strains with varying host specificities have different LS regions which has been determined by comparing sequences of Fo strain 5176, pathogen of Arabidopsis (Thatcher et al. 2009) and Fo f.sp vasinfectum (Dowd et al. 2004), a pathogen of cotton. Although overall sequence divergence between common sequences of Fol and Fo5176 is less than 2%, sequences in the Fo LS region do not have homologues in Fo5176. Fov EST sequences (Dowd et al. 2004) have 99% sequence identity to the Fol genome in core region only. These are large-scale genome polymorphism within Fo as karyotypes between strains vary (Teunissen et al. 2003). Small polymorphic and conditionally dispensable chromosomes confer host-specific virulence in the fungi Nectria haematococca (Miao et al. 1991) and Alternaria alternata (Harimoto et al. 2007). Small (<2.3 Mb) and variable chromosomes are absent in non-pathogenic F. oxysporum isolates indicating that Fol LS chromosomes are responsible for pathogenicity towards specific host. Chromosome 14 of Fol is probably responsible for its pathogenicity towards tomato; its transfer rates could increase the overall pathogenesis. Proteome analysis revealed that small proteins Six1 (Avr3), Six3 (Avr2) and in-planta oxidoreductase (Oxi1) are secreted during colonization of Fol in tomato xylem system (Houterman et al. 2007; van der Does et al. 2008) and are involved in virulence (Houterman et al. 2009; Rep et al. 2004). It was further revealed that genes encoding above proteins are present on chromosome 14 present in strains causing tomato wilt, but are generally not present in other strains (van der Does et al. 2008). Further, genome analysis identified that six5, six6, six7 are also present on chromosome 14. It has been demonstrated that entire LS chromosome 14 could be transferred through simple co-incubation between two, otherwise genetically separated members of Fo leading to the emergence of new pathogenic lineages. Horizontal transfer of host specificity factors between otherwise distant and genetically separated lineages of Fo explains the host specialization originated polyphyletically (Gale et al. 2003). Fol LS regions are enriched for genes regulating host-pathogenic interactions. These chromosomes could transfer an entire set of genes required for host compatibility to a new genetic lineage in a single event. If transferred to the recipient lineage with an environmental adaptation different from the donor, the overall incidence of disease in a host increases because pathogenicity is introduced in the genetic background pre-adapted to a local environment.

Conclusion and Future Prospectives

The establishment of F. oxysporum as plant and animal infection model, the use of molecular genetics approaches in this species, and the genomic characterization of different Fusarium f. sp. has advanced our understanding of several key aspects related to fungal pathogenicity and its evolutionary origins.

The availability of sequenced genomes, gene annotations and genome expression data of various Fusarium spp., has conclusively shown that pathogens harbour conserved as well as specialized pathogenicity genes. With the analysis of genome expression data, several conserved pathogenicity genes, such as those encoding MAPKs have been characterized (Ma et al. 2013), whereas specialised pathogenicity genes linked to host adaptation or evasion have remained largely undefined except, secreted in xylem (SIX) effectors in F. oxysporum f. sp. lycopersici and several mycotoxins in other Fusarium species. The multiple processes, e.g. population diversity in specific genomic regions, horizontal acquisition of whole pathogenicity chromosomes (Ma et al. 2010) or a few pathogenicity-related genes (Gardiner et al. 2014) have been involved in evolution of pathogenicity and host specificity in Fusarium. Diversifying selection studies in the three Fusarium pathogens Fg, Fv and Fol with distinct pathogenicity profiles have revealed that all Fusarium species have core group of genes under purifying selection to preserve their function and specialised group of genes as those encoding proteins with a N-terminal [SG]-P-C-[KR]-P sequence motif and pathogen-associated proteins evolve at a faster rate. These rapidly evolving gene groups are functionally associated with pathogenicity (Sperschneider et al. 2015). Further, diversifying selection acts strongly on accessory-/lineage-specific chromosomes. Moreover, diversifying selection studies combined with in planta expression data are useful for identifying pathogenicity genes involved in competition between pathogen and host. Future developments will be in the form of improved gene annotations, greater sequencing depth in the genus so that genes that show weak signal for diversifying selection during pathogen–host interactions are identified and generation of in planta expression data for other Fusarium species to detect effector production at early stages during infection. In future, F. oxysporum is likely to provide valuable new insights into molecular mechanisms of host specificity and pathogenicity in evolutionarily distant hosts.