OUP user menu

Nutrition acquisition strategies during fungal infection of plants

Hege H. Divon, Robert Fluhr
DOI: http://dx.doi.org/10.1111/j.1574-6968.2006.00504.x 65-74 First published online: 1 January 2007

Abstract

In host–pathogen interactions, efficient pathogen nutrition is a prerequisite for successful colonization and fungal fitness. Filamentous fungi have a remarkable capability to adapt and exploit the external nutrient environment. For phytopathogenic fungi, this asset has developed within the context of host physiology and metabolism. The understanding of nutrient acquisition and pathogen primary metabolism is of great importance in the development of novel disease control strategies. In this review, we discuss the current knowledge on how plant nutrient supplies are utilized by phytopathogenic fungi, and how these activities are controlled. The generation and use of auxotrophic mutants have been elemental to the determination of essential and nonessential nutrient compounds from the plant. Considerable evidence indicates that pathogen entrainment of host metabolism is a widespread phenomenon and can be accomplished by rerouting of the plant's responses. Crucial fungal signalling components for nutrient-sensing pathways as well as their developmental dependency have now been identified, and were shown to operate in a coordinate cross-talk fashion that ensures proper nutrition-related behaviour during the infection process.

Keywords
  • plant pathogen
  • fungal nutrition
  • nitrogen and carbon regulation

Introduction

Parasitism defines the state in which one organism benefits at the expense of another. An inherent part of the success of the parasite in this misbalanced relationship is the securing of its nourishment. Much research has focused towards understanding pathogenicity and virulence factors. These include study of avirulence factors, toxins, cell wall-degrading enzymes (CWDE) and other mechanisms directly linked to the ability of the pathogen to obtain access to its host. A less-studied aspect of this interaction is the appreciation of how nutrients are acquired and how the parasite adapts to changing nutritional environments. Ultimately, a pathogenic lifestyle is characterized by metabolic dependency on the host. This is a fine-tuned association where host physiology is exploited and modified to optimize production of nutrients for the pathogen, and reciprocally pathogen metabolism is adapted to utilize what is made available. The final fitness of the fungus in this metabolic ‘tug-of-war’ will determine the success of colonization.

In this review, we examine in planta nutrition by plant–pathogenic fungi as established by biochemical and molecular tools. Many of the general concepts have been adopted from studies in saprophytes such as Saccharomyces cerevisiae, Aspergillus nidulans and Neurospora crassa. The central questions are as follows: what nutrients are acquired by the pathogen from the host, what metabolic adjustments have to be made during the pathogen's life cycle and what are the regulatory mechanisms that facilitate this adaptation?

Dynamic nutrient requirements throughout the infection cycle

Fungal plant pathogens have been traditionally classified by their lifestyle into biotrophic and necrotrophic types of parasitism. Biotrophic pathogens are defined by a total dependency on the host to complete their life cycle, deriving nutrients from living host cells by differentiation of specialized infection structures called haustoria (Mendgen & Hahn, 2002). In contrast, necrotrophs do not possess specialized infection structures and derive nutrients from sacrificed cells (Lewis, 1973). To accommodate further diversification of lifestyles, the term ‘hemibiotrophic’ has been coined to consider situations in which the pathogen renders its host largely alive while establishing itself within the host tissue. Only after this brief biotrophic-like phase does it switch to a necrotrophic lifestyle (Perfect & Green, 2001). The infection process itself can be separated into germination, proliferation and sporulation (Solomon et al., 2003). The utility of this division is used here by following the dynamic nutritional challenges facing the pathogen in the context of its respective infectious phases (Fig. 1).

Figure 1

Infectious phases in phytopathogenic fungi that are relevant to nutrition acquisition strategies during fungal infection of plants. The penetration and sporulation phases are common to all fungi. Hemibiotrophs are distinguished by going through successive biotrophic and necrotrophic phases. Necrotrophs do not have a biotrophic phase, whereas obligate biotrophs complete their life cycle in this phase, without proceeding to necrotrophic stage.

Spore germination and penetration

Pathogen penetration of the host is facilitated by direct local mechanical pressure during appressorium development, enzymatic digestion of plant barriers or by the more serendipitous availability of stomata, wounds and cracks at the plant environment interface. It is generally assumed that during spore germination and the penetration phase, the phytopathogenic fungi are in a state of starvation and are completely reliant on nutrients derived from internal stores. The major components of fungal spore stores include glycogen, trehalose, polyols such as mannitol and lipids (Thevelein, 1984; Thines et al., 2000; Voegele et al., 2005). Recent evidence from several plant–pathogen systems has focused on the role of fatty acid mobilization during the penetration process (Table 1). Mobilization of lipid stores occurs through lipolysis and β-oxidation cycles to form acetyl-CoA, which is further assimilated into the tricarboxylic acid cycle (TCA cycle) via the glyoxylate cycle. The glyoxisomal enzyme isocitrate lyase (ICL), a marker for lipolytic activity, is induced in the absence of glucose and during growth on gluconeogenic carbon sources such as acetate and ethanol, and on fatty acids (Bowyer et al., 2000). During infection, ICL promoter activity was detected predominantly in the prepenetration stage on the plant surface and declined after appressorial penetration, indicating a switch from gluconeogenic to glycolytic metabolism at this stage. Several mutants in which virulence is reduced were found to be defective in peroxisomal development (Kimura et al., 2001), or disrupted in the glyoxylate cycle enzymes ICL (Idnurm & Howlett, 2002; Wang et al., 2003) and malate synthase (MLS; Solomon et al., 2004a). The findings indicate a universal role for fatty acid metabolism and glyoxylate cycle during plant penetration, regardless of pathogenic lifestyle. Interestingly, in some studies, glucose was able to complement the loss of pathogenesis (Kimura et al., 2001; Idnurm & Howlett, 2002; Solomon et al., 2004a). The implications are that a critical role for the fatty acid metabolic activity is the support of the TCA cycle through acetyl-CoA production in order to generate ATP under the poor nutrient conditions that exist at the plant surface. An additional role for fatty acid metabolism in glycerol production was demonstrated in the ICL mutant in Magnaporthe grisea, which shows delayed appressorium turgor generation (Wang et al., 2003).

View this table:
Table 1

Fungal nutrition-related genes and ESTs expressed during plant infection. Signalling components are not included

Gene/EST nameAcc. No.Fungal speciesSimilarityFunctionEPReferences
Biotrophic fungi
Aat1 (Pig27)AJ308252U. fabaeAmino acid transporterHistidine uptakeNDStruck et al. (2002)
Aat2 (Pig2)U81794U. fabaeAmino acid transporterAmino acid transportNDHahn et al. (1997)
Ard1AJ809335U. fabaed-arabitol dehydrogenase 1d-arabitol biosynthesisNDLink et al. (2005)
Bgl1AJ575269U. fabaeβ-glucosidaseCellobiose degradationNDHaerter & Voegele (2004)
Gat1AF271266C. fulvumGABA transaminaseGABA metabolismNDSolomon & Oliver (2002)
Gltrn1AF376000B. graminisGlucose transporterGlucose uptakeNDZhang et al. (2005)
GS-1AF395112B. graminisGlutamine synthetaseGlutamine synthesisNDZhang et al. (2005)
Hxt1AJ310209U. fabaeHexose transporterUptake of d-glc, d-frcNDVoegele et al. (2001)
Met6AF226997C. fulvumMethionine synthaseMethionine biosynthesisYesSolomon et al. (2000)
Mad1 (Pig8)O00058U. fabaeMannitol dehydrogenaseMannitol metabolismNDVoegele et al. (2005)
PmaAF406807B. graminisPlasma membrane H+ ATPaseNutrient uptakeNDZhang et al. (2005)
pSI-9Y14555C. fulvumAldehyde dehydrogenaseCarbon metabolismNDColeman et al. (1997)
pSI-10Y14556C. fulvumAlcohol dehydrogenaseCarbon metabolismNDColeman et al. (1997)
Thi1 (Pig1)AJ250426U. fabaePyrimidine biosynthetic enzymeVitamin B1 biosynthesisNDSohn et al. (2000), Hahn & Mendgen (1997)
Thi2 (Pig4)AJ250427U. fabaeThiazole biosynthetic enzymeVitamin B1 biosynthesisNDSohn et al. (2000), Hahn & Mendgen (1997)
Tr4BU672643P. triticinaPyrimidine biosynthetic enzymeVitamin B1 biosynthesisNDThara et al. (2003)
Tr21BU672660P. triticinaThiazole biosynthetic enzymeVitamin B1 biosynthesisNDThara et al. (2003)
UfAat3AJ628939U. fabaeAmino acid transporterleu, met, cys uptakeNDStruck et al. (2004)
UfInv1AJ640083U. fabaeInvertaseSugar hydrolysisNDVoegele et al. (2006)
Uf-Pma1AJ003067U. fabaePlasma membrane H+ ATPaseProton-driven nutrient uptakeNDStruck et al. (1996), Struck et al. (1998)
Hemi-biotrophic fungi
Arg1AB045736F. oxysporumArgininosuccinate lyaseArginine biosynthesisYesNamiki et al. (2001)
Gap1 (Fol5B10)CK615491F. oxysporumGeneral amino acid permeaseAmino acid uptakeNDDivon et al. (2005)
GpdhAY331190C. gloeosporioidesGlycerol-3-phosphate dehydrogenaseCarbon utilizationNoWei et al. (2004)
Icl1AY118108L. maculansIsocitrate lyaseGlyoxylate cycleYesIdnurm & Howlett, 2002
Mtd1 (Fol8D6)CK615494F. oxysporumOligopeptide transporterPeptide uptakeNDDivon et al. (2005)
pCgGSL78067C. gloeosporioidesGlutamine synthetaseGlutamine synthesisNDStephenson et al. (1997)
Uricase (Fol6D12)CK615495F. oxysporumUricaseUric acid anabolismNDDivon et al. (2005)
Necrotrophic fungi
Als1DQ167577S. nodorumDelta-aminolaevulinic acid synthaseHeme biosynthesisYesSolomon et al. (2006)
Cbl1EAA68828F. graminearumCystathionine betalyaseMethionine biosynthesisYesSeong et al. (2005)
Glo1AY576607S. nodorumGlyoxylase IGlycolytic bypass pathwayNoSolomon & Oliver (2004)
Icl1AF540383M. griseaIsocitrate lyaseGlyoxylate cycleYesWang et al. (2003)
Mls1AY508881S. nodorumMalate synthaseGlyoxylate cycleYesSolomon et al. (2004a)
Mpd1AY587541S. nodorumMannitol 1-phosphate dehydrogenaseMannitol metabolismYesSolomon et al. (2005)
Msy1EAA75179F. graminearumMethionine synthaseMethionine biosynthesisYesSeong et al. (2005)
Odc1AJ249387S. nodorumOrnithine decarboxylasePolyamine biosynthesisYesBailey et al. (2000)
Pth2AF027979M. griseaCarnitine acetyltranferaseFatty acid transportYesSweigard et al. (1998)
Pth3O42621M. griseaImidazoleglycerol-P dehydrataseHistidine biosynthesisYesSweigard et al. (1998)
Ptr2AY187281S. nodorumdi/tripeptide transporterdi/tri peptide uptakeNoSolomon et al. (2004b)
BM135467F. graminearumNADP-specific glutamate dehydrogenaseGlutamate synthesisNDKruger et al. (2002)
BM137184F. graminearumHexose transporter proteinHexose uptakeNDKruger et al. (2002)
MG03533M. griseaFormamidasetrp catabolismNDDonofrio et al. (2006)
MG04385M. griseaUrea transporterUrea uptakeNDDonofrio et al. (2006)
MG05526M. griseaMethylammonium permeaseNitrogen sensing and uptakeNDDonofrio et al. (2006)
  • EP, effects of mutation on pathogenicity

  • The reader is referred to Guldener et al. (2006) and Jakupovic et al. (2006), for additional genes expressed inplanta in F. graminearum and U. fabae, respectively.

Nitrogen constitutes a smaller part of the storage compounds in the spore and might represent a rate-limiting factor during penetration. Nevertheless, elevated expression of a di/tri-peptide transporter (Ptr2) at the prepenetration stage in Stagonospora nodorum is not essential for pathogenicity (Solomon et al., 2004b), suggesting that recycling of internal stores can serve as the primary source of nitrogen essentials at this stage. Indeed, a recent transcriptome survey of barley powdery mildew revealed a large number of genes involved in amino acid metabolism, protein turnover and amino acid recycling, which were expressed during spore germination and appressorium formation (Thomas et al., 2001).

Biotrophic growth phase and the entrainment of plant metabolism

Once inside the host, and with internal stores likely exhausted, the pathogen needs to establish itself rapidly by mobilizing mechanisms that will ensure adequate nutrient uptake from the host. For the obligate biotrophs, this phase is initiated by the formation of a haustorium (Voegele, 2006, and references therein). For hemibiotrophs, deficient in such feeding structures, some recent evidence suggests that they have developed mechanisms that entrain host metabolism to establish new sink tissues to suit the needs of the pathogen. Invertases and hexose sugar transporters play a key role in defining a leaf as a ‘source’ or as a ‘sink’ of carbohydrates. The up-regulation of plant extracellular invertase appears to be a common response to various biotic and abiotic stress-related stimuli, serving to fuel plant defence responses in the infected tissue (reviewed in Biemelt & Sonnewald, 2006). This defence strategy is used by Arabidopsis; however, during powdery mildew infection, this physiological response of the plant is exploited by the fungus to ensure a ready supply of carbohydrates for its own use (Fig. 2; Fotopoulos et al., 2003).

Figure 2

Examples of the entrainment of plant metabolism by phytopathogenic fungi. Sensing of nitrogen or carbon availability (NS, CS; flash arrow), or the absence of preferable nutrient sources, triggers signal transduction pathways that initiate transcription of transporters and pathway-specific metabolic enzymes. The de-repression of these pathways is regulated by de-phosphorylation and binding of AREA to nitrogen catabolite-regulated gene promoters (Ravagnani et al., 1997; Beck & Hall, 1999), and by SNF1-dependent phosphorylation and release of CREA repressor from carbon catabolite-regulated gene promoters (DeVit et al., 1997; Gancedo, 1998). Plant resistance responses (red arrows), involving ornithine or ROS scavengers such as GABA, are taken up through amino acid (AA) permeases and metabolized by the fungus (Oliver & Solomon, 2004; Guldener et al., 2006). Uric acid, the product of disease-induced enhanced catabolic processes in the plant, can serve as a fungal nutrient by conversion to allantoin by uricase and subsequent recovery of ammonia (Divon et al., 2005). Plant invertases that are induced by the plant to fuel its cellular defence needs, as well as CWDE secreted by the pathogen, will produce depolymerized sugars that the fungus can utilize (Ospina-Giraldo et al., 2003; Biemelt & Sonnewald, 2006). PM, plasma membrane; CW, cell wall.

In a manner similar to the utilization of plant carbohydrates during infection, nitrogen acquisition can also be subverted for fungal benefit as illustrated by the metabolism of γ-amino butyric acid (GABA) during Cladosporium fulvum infection of tomato (Solomon & Oliver, 2001). GABA, normally a major amino-acid constituent of the tomato apoplast, accumulates to even higher (millimolar) levels during the compatible interaction. The accumulation was concomitant with the induction of tomato glutamate decarboxylase (GAD) responsible for GABA synthesis, and the fungal GABA transaminase (Gat1; Table 1), metabolizing GABA into succinic semialdehyde (Solomon & Oliver, 2002). The utilization of GABA as a nutrient has so far only been demonstrated for Cladosporium fulvum, and may be part of the adaptation to a biotrophic lifestyle without haustoria, which occurs exclusively in the apoplastic spaces of the leaf. Interestingly, the ramifications of modulation of GABA metabolism may be more subtle. The GABA shunt can also serve as part of a cellular scavenger system for oxidative stress signals. The role of reactive oxygen species (ROS) in plant defence signalling is well established, and one can speculate that as in the case of carbohydrate acquisition successful colonization will subvert plant signals for defence to provide a source of nutrients for pathogenic fungi (Fig. 2; Oliver & Solomon, 2004). A similar scenario may occur in the Fusarium oxysporum interaction with tomato, where uricase, an enzyme catalysing the conversion of uric acid to allantoin, was induced during the early nonnecrotic stages of infection (Table 1; Divon et al., 2005). Uric acid can act as a scavenger of ROS (Becker et al., 1989) and like GABA may serve the plant in that role. Hence, the fungus may utilize uric acid derived from host catabolism and defence reactions (Fig. 2). Additional evidence for the role of metabolites with scavenging capacity comes from a study of the rust fungus Uromyces fabae. Voegele (2005) suggested that the pathogen-derived compound mannitol, which the host, Vicia faba, can neither produce nor metabolize, accumulates in infected leaves to serve as a ROS scavenger to protect the pathogen in the infection process.

While great strides have been made in understanding how bacterial pathogen effectors are delivered through the type III secretion systems found in plant–pathogenic bacteria interactions, less is known at the molecular level of how fungi entrain host metabolism. For example, in the Cladosporium fulvum interaction with tomato, an extracellular fungal protein was detected that modifies a plant cysteine protease and subsequent plant defence responses (Rooney et al., 2005). Interestingly, proteins from the rust fungi Melamspora lini and U. fabae are localized to the host cytoplasm (Dodds et al., 2004; Kemen et al., 2005). This may indicate that fungal effector proteins can gain intracellular access and thus target cellular metabolic crossroads.

Necrotrophic growth phase and entrainment of plant metabolism

Shortly after penetration, necrotrophic plant pathogenic fungi, aided by production of toxins, hydrolytic enzymes and necrosis-related proteins, cause cell lysis and death of host cells. One may surmise that the lysis of host cells would lead to an increase in accessible nutrients; however, as evidenced from examination of auxotrophic mutants, certain amino acids such as methionine and histidine are not supplied in sufficient quantity by the host tissue and their biosynthesis is carried out by the fungus (Sweigard et al., 1998; Balhadere et al., 1999; Seong et al., 2005). The coordinate expression of nutrient acquisition genes during infection that includes possible sources for amino acid backbones was revealed in the transcriptome analysis of Fusarium graminearum infection (Guldener et al., 2006). In this case, transporters of sugars and nitrogenous compounds were expressed in early infection, including an ornithine transporter expressed from day 2 postinfection. Ornithine is a nonprotein amino acid and an important intermediate in metabolic pathways such as arginine and proline biosynthesis, as well as the biosynthesis of putrescine, a precursor for other polyamines and pyrimidines. Ornithine uptake was followed by expression of arginine, proline and putrescine biosynthetic genes from day 3 to 6 postinfection (arginine: ornithine carbamoyltransferase, carbamoyl-P synthase and acetylglutamate kinase; proline: ornithine aminotransferase; putrescine: ornithine decarboxylase). In line with this observation, a putrescine auxotrophic mutant from the wheat pathogen Stagonospora nodorum, which was disrupted in the gene encoding ornithine decarboxylase, odc1, was found to have greatly reduced virulence on wheat (Table 1; Bailey et al., 2000). Ornithine, as a precursor in polyamine synthesis, is involved in plant stress responses (Walters, 2003). Hence, we encounter, as in the case of carbohydrates, GABA and uric acid metabolism, the concept that successful pathogens hijack potential plant defences for their nutritional benefit during colonization (Fig. 2).

Sporulation

Sporulation marks the completion of the pathogenic life cycle. True biotrophic fungi complete their life cycle by maintaining a biotrophic state, whereas hemibiotrophs and necrotrophs complete their life cycle in a necrotrophic state (Fig. 1). In vitro studies have shown that sporulation can be triggered by exhaustion of nitrogen and carbon sources (Griffin, 1994). Thus, in a manner analogous to the finding that fungi undergo nutrient limitation during the initial penetration, it may be anticipated that nutrient limitation triggers sporulation. However, experiments describing the in planta sporulation, from the hemibiotrophic interaction of Mycosphaerella graminicola with wheat, points to a different situation. cDNA microarray analysis during in planta growth revealed fungal expression of a subset of genes that resemble transcripts associated with in vitro growth in rich medium rather than poor medium (Keon et al., 2005). The results suggest an alternative trigger for the sporulation event such as, for example, a quorum-sensing-like mechanism.

A particularly illuminating example of dependence on nutrient transport in the late growth stages was revealed in the glycerol-3-phosphate dehydrogenase (Gpdh) disruption mutant of the hemibiotrophic plant pathogen, Colletotrichum gloeosporioides f.sp. malvae (Table 1; Wei et al., 2004). The mutant contains reduced amounts of glycerol, and during heterotrophic growth on defined media, the mutant strain exhibited severe defects in carbon utilization and failed to conidiate. This defect could be complemented by exogenous addition of glycerol. Surprisingly, in spite of these profound growth defects during heterotrophic growth, the gpdh mutant strain displayed normal pathogenicity in planta, progressing normally through infection and conidiation. Importantly, the analysis of plant tissues showed significant depletion of glycerol in the infection zone, indicating glycerol transfer from the plant to the fungal pathogen.

Regulation of fungal nutrition gene expression during infection

Nutrient availability in the external environment dictates expression of fungal nutrition genes via general metabolic processes known as catabolite repression. This process implies that, as long as sufficient amounts of a favourable nutrient source are available, transcription of alternative uptake systems and catabolic enzymes is repressed. The preferred source for the carbon-regulatory system is glucose, and it is ammonia and glutamine for the nitrogen-regulatory system. Catabolite repression has been extensively studied in saprobes like Saccharomyces cerevisiae, N. crassa and A. nidulans (Marzluf, 1997; Carlson, 1998; Gancedo, 1998; Magasanik & Kaiser, 2002), and given the high degree of conservation in this process, these saprobes have served as models to guide similar research in phytopathogenic fungi.

Carbon catabolite repression (CCR)

CCR signifies the transcriptional regulation of genes involved in utilization of alternative carbon sources. It is driven by glucose repression and substrate induction. In Saccharomyces cerevisiae, the SNF1 (sucrose nonfermenting 1) gene is required for expression of catabolite-repressed genes (Fig. 2; Gancedo, 1998, and references therein). It encodes a protein kinase whose major function is to phosphorylate Mig1, a DNA-binding transcriptional repressor. When glucose is limiting, phosphorylation of Mig1 causes its dissociation from the promoters of repressed genes (DeVit et al., 1997), thus motivating the release from CCR. SNF1-orthologues have been cloned from the maize pathogen Cochliobolus carbonum (CcSnf1; Tonukari et al., 2000) and the vascular wilt pathogen F. oxysporum (FoSnf1; Ospina-Giraldo et al., 2003). Disruption mutants were compromised for growth on secondary carbon sources, and production of several secreted CWDE was reduced with a concomitant reduction in virulence. This phenomenon was observed at the penetration stage (Tonukari et al., 2000), as well as in later stages of plant colonization (Ospina-Giraldo et al., 2003). Thus, CCR is directly linked to pathogenicity. In addition, SNF1 might facilitate de-repression of hexose transporters for uptake and utilization of sugars released from degraded cell wall polymers (Fig. 2; Carlson, 1998). The orthologue of MIG1 in filamentous fungi is called CreA, encoding a repressor with two zinc fingers of the C2H2 class (Fig. 2; Ronne, 1995). CreA genes were isolated from the plant pathogenic fungi Gibberella fujikuroi and Botrytis cinerea (Tudzynski et al., 2000); however, their contribution to nutrient acquisition during infection was not determined. In another approach, random insertional mutagenesis in the hemibiotroph Colletotrichum lindemuthianum identified a mutant disrupted in the metabolic switch between the biotrophic and the necrotrophic phases (Dufresne et al., 2000). The corresponding gene, encoding a putative zinc finger GAL4-like transcriptional activator, was designated Clta1.

Nitrogen catabolite repression (NCR) and the ‘nitrogen starvation hypothesis’

The hypothesis that plant–pathogenic fungi experience nitrogen nutrient limitation during in planta growth (Snoeijers et al., 2000) is based on observations that showed the induction of starvation-regulated genes in Cladosporium fulvum, Magnaporthe grisea and others during the infection process (Talbot et al., 1993; Van den Ackerveken et al., 1994; Coleman et al., 1997; Stephenson et al., 1997; Snoeijers et al., 2000). NCR embodies the transcriptional regulation of permeases and catabolic enzymes needed for the utilization of secondary nitrogen sources. In yeast, the gene GLN3 encodes a global activator of the GATA-type zinc finger family, recognizing the core consensus sequence 5′-(A/T/C)GATA(A/G)-3′ found in NCR-regulated promoters (Ravagnani et al., 1997; Starich et al., 1998). In the presence of favourable nitrogen sources such as ammonia and glutamine, Gln3, phosphorylated by the TOR (target of rapamycin) kinase (Beck & Hall, 1999; Fitzgibbon et al., 2005), is sequestered to the cytosol by inhibitory binding to the negative regulator Nil2. During nitrogen starvation, due to the lack of favourable nitrogen sources, de-phosphorylation of Gln3 relieves it from Nil2 repression, and it relocates to the nucleus where it activates transcription of NCR-controlled genes, either alone or in combination with pathway-specific activators (Marzluf, 1997; Magasanik & Kaiser, 2002). Once identified as a key regulator of nitrogen metabolism, orthologues of GLN3 have been extensively studied in saprobes like A. nidulans and N. crassa (AreA and nit-2, respectively; Marzluf, 1997), and in several phytopathogenic fungi. They act as central control devices aiding the fungus in adapting to different nutrient environments (Fig. 2).

As the nutrient environment varies with the pathogenic lifestyle, the dependency on AREA-like regulators during infection is expected to vary accordingly. Disruption of Clnr1 in the hemibiotrophic fungus Colletotrichum lindemuthianum severely compromised the infection cycle at the stage of secondary hyphae formation and almost completely eliminated the occurrence of anthracnose symptoms (Pellier et al., 2003). Disruption of Fnr1 in F. oxysporum caused a delay in disease symptom development, attributed in part to the abrogated expression of nutrition genes normally up-regulated during in planta growth (Divon et al., 2006). Conversely, disruption of Nut1 in the rice blast fungus, Magnaporthe grisea, had only a minor effect on virulence (Froeliger & Carpenter, 1996). In Cladosporium fulvum, recent evidence shows that deletion of the Nrf1 gene significantly reduces Cladosporium fulvum virulence on tomato; however, this reduction is not connected to the abolished expression of Avr9 (Thomma et al., 2006). Interestingly, GABA uptake via the GABA permease in A. nidulans is subject to NCR (Hutchings et al., 1999). Under the assumption that GABA constitutes a significant nutrient for Cladosporium fulvum during pathogenicity (Solomon & Oliver, 2002), it is likely that loss of virulence in the nrf1 mutant is, at least in part, caused by nutrient deficiency due to the inability to take up and metabolize GABA.

Based on the existing evidence, it is clear that the main impact of AREA on pathogenic ability is through the regulation of nutrition. Interestingly, in hemibiotrophs, the impact of AREA on regulation seems to somewhat correlate with the ‘degree’ of biotrophy, i.e. the time spent in the plant apoplast. Thus, in Magnaporthe grisea, which grows necrotrophically almost immediately after penetration, the effect of AreA disruption is minor (Froeliger & Carpenter, 1996). In contrast, in the hemibiotrophs F. oxysporum, Colletotrichum lindemuthianum and Cladosporium Fulvum, such a disruption shows a clear effect (Pellier et al., 2003; Divon et al., 2006; Thomma et al., 2006). Apparently, hemibiotrophs that have an extended growth period in the nutrient-poor apoplast will tend to rely on AREA-regulatory nutrient control. In comparison, this regulatory control will be less important in the obligate biotrophic pathogens that feed through specialized infection structures or necrotrophs to whom ample nutrients are available. Finally, during in planta growth, nitrogen starvation might be seen as a stage-specific or recurring condition, rather than a continuous condition. Transition periods from one infectious phase to the next, i.e. germination, sporulation and transition between the biotrophic and necrotrophic phase, are likely to increase the demand for energy and metabolite precursors and require precise regulation.

Linkage of nutritional regulation and development

Apart from the influence of the external environment, stage-specific requirements for nutrients are also linked to the development of specialized infection structures. This is illustrated by the fact that, in obligate biotrophic fungi, nutrition genes are insensitive to the exogenous nutrient availability, and instead their transcriptional control is linked to the development of haustoria. For example, thiamine biosynthetic genes, thi1 and thi2, in the rust fungus U. fabae do not respond with down-regulation to exogenously applied thiamine as do saprophytic fungi (Sohn et al., 2000). Similarly, appressorium turgor generation depends on accumulation of large quantities of glycerol, which is acquired from breakdown of internal stores of lipids and glycogen (Thines et al., 2000). Such processes are usually under CCR (Hardie et al., 1998); however, impairment of the CCR regulatory pathway in Cochliobolus carbonum by disruption of Snf1 did not affect appressorium formation and turgor generation (Tonukari et al., 2000; see below), indicating that other mechanisms override SNF1 regulation of fatty acid breakdown and peroxisomal function.

The complexity of metabolic and developmental interaction was investigated in the process of turgor generation in developmental, nonpathogenic mutants in Magnaporthe grisea (Thines et al., 2000). In the MAP kinase mutant, Δpmk1, appressorium formation is impaired and glycogen and lipid mobilization do not occur during germination. The same processes were also retarded markedly in a cAMP-dependent protein kinase A (PKA) mutant (ΔcpkA). Furthermore, in a Δmac1 sum1-99 mutant, carrying a mutation in the regulatory subunit of PKA that renders it cAMP-independent and nearly constitutively active, glycogen and lipid degradation were rapid and preceded appressorium morphogenesis (Thines et al., 2000). In another example, appressorium formation was also impaired in the nir1 mutant in Colletotrichum acutatum (Horowitz et al., 2006). Nir1, an orthologue of NirA in A. nidulans, is the pathway-specific activator of nitrate utilization genes, and the Nir1 mutation abolished activation of nitrate and nitrite reductase and was assumed to account partly for the penetration-defect phenotype. However, the nir1 mutant displayed additional appressorium developmental defects not present in nitrate reductase mutants (Horowitz et al., 2006). Together, these results indicate that considerable crosstalk occurs during development of the functional appressorium, involving both traditional catabolite-repressed regulatory pathways, MAP kinase pathways and cAMP-dependent PKA (Fig. 3).

Figure 3

Nutrition-related processes that take place during appressorium-mediated penetration. MAP kinase- and cAMP-dependent PKA signalling pathways are involved in mobilization of glycogen and lipid stores from the spore. Turgor generation is accomplished by compartmentalization and rapid degradation of lipid and glycogen reserves in the appressorium (Thines et al., 2000). The glyoxylate cycle (Glc) is involved in turgor generation, but also in replenishing intermediates to the TCA cycle from acetyl-CoA generated in fatty acid oxidation (Kimura et al., 2001; Idnurm & Howlett, 2002; Wang et al., 2003; Solomon et al., 2004a). An NirA-like transcription factor, Nir1, is required for transcription of nitrate reductase (NR), necessary for nitrate assimilation in Colletotrichum acutatum, but is also involved in additional processes related to appressorium formation (Horowitz et al., 2006). The grey/white arrow in the germ tube indicates transport of glycogen and lipid metabolic products; the flash arrow indicates sensing mechanism(s); and the small white arrows in the appressorium signify turgor generation.

Conclusions

An understanding of fungal metabolism and adaptation during host colonization is of critical importance to evaluate fully the complex orchestration of the activities that rule host–pathogen interactions. This knowledge will provide a rational basis for the development of new disease control strategies. Studies of gene expression and auxotrophic mutants have provided considerable information regarding those nutrients available from the plant and those that are not. Elaborate mechanisms have been unravelled that indicate how host and pathogen compete for nutrient supplies, and how plant defence mechanisms can be exploited by fungi for nutritional purposes. Besides amino acids, the nonprotein amino acids GABA and ornithine stand out as potential providers of nitrogen, and as precursors in biosynthetic pathways for compounds not adequately provided by the plant. More case studies will be needed to determine whether these represent specific or global mechanisms. Key regulatory factors involved in nutrient signalling and gene expression have been characterized. Although their impact on pathogenicity appears to be species specific, commonalties between lifestyle preference and regulatory control are now emerging. While important advances have been made in our ability to describe the metabolic events surrounding fungal ingress, little is known at the molecular level of how fungi entrain host metabolism. This accentuates the future challenge in understanding metabolite requisition by fungal parasites.

Footnotes

  • Editor: Richard Staples

References

View Abstract