Sclerotinia sclerotiorum: When “to be or not to be” a pathogen?

Abstract

Sclerotinia sclerotiorum is unusual among necrotrophic pathogens in its requirement for senescent tissues to establish an infection and to complete the life cycle. A model for the infection process has emerged whereby the pathogenic phase is bounded by saprophytic phases; the distinction being that the dead tissues in the latter are generated by the actions of the pathogen. Initial colonization of dead tissue provides nutrients for pathogen establishment and resources to infect healthy plant tissue. The early pathogenicity stage involves production of oxalic acid and the expression of cell wall degrading enzymes, such as specific isoforms of polygalacturonase (SSPG1) and protease (ASPS), at the expanding edge of the lesion. Such activities release small molecules (oligo-galacturonides and peptides) that serve to induce the expression of a second wave of degradative enzymes that collectively bring about the total dissolution of the plant tissue. Oxalic acid and other metabolites and enzymes suppress host defences during the pathogenic phase, while other components initiate host cell death responses leading to the formation of necrotic tissue. The pathogenic phase is followed by a second saprophytic phase, the transition to which is effected by declining cAMP levels as glucose becomes available and further hydrolytic enzyme synthesis is repressed. Low cAMP levels and an acidic environment generated by the secretion of oxalic acid promote sclerotial development and completion of the life cycle. This review brings together histological, biochemical and molecular information gathered over the past several decades to develop this tri-phasic model for infection. In several instances, studies with Botrytis species are drawn upon for supplemental and supportive evidence for this model. In this process, we attempt to outline how the interplay between glucose levels, cAMP and ambient pH serves to coordinate the transition between these phases and dictate the biochemical and developmental events that define them.

Sclerotinia sclerotiorum, Necrotroph, Saprophyte, Pathogenesis, Oxalic acid, Polygalacturonase

1 Introduction

The ascomycete fungus, Sclerotinia sclerotiorum (Lib.) de Bary, is a hugely destructive pathogen of many economically important crops, including grain legumes (soybean, pea and bean) and oilseeds (canola and sunflower). This necrotrophic pathogen exhibits little host specificity and has a host range that includes more than 400, primarily dicotyledonous, plant species [1]. As its name indicates, the fungus produces sclerotia, which are long-lived melanized resting structures. Sclerotia can germinate in two ways, carpogenically to form apothecia from which ascospores are liberated or myceliogenically to produce hyphae. The fungus is homothallic and no asexual spores are produced. Populations of S. sclerotiorum are predominantly clonal in temperate regions of North America, Europe and New Zealand [2]. There is now evidence of extensive recombination in the Western US and South America, with some deviation from clonality in Australia [ [; Kohn, personal communication].

2 Saprophytic growth is required to initiate infection

Carpogenic germination of sclerotia is stimulated by periods of continuous soil moisture. Apothecia are formed on the soil surface from which ascospores are released into the air. Infection of most crop species is principally associated with ascospores but direct infection of healthy, intact plant tissue from germinating ascospores usually does not occur [4]. Instead, infection of leaf and stem tissue of healthy plants results only when germinating ascospores colonize dead or senescing tissues, usually flower parts such as abscised petals, prior to the formation of infection structures and penetration. Myceliogenic germination of sclerotia at the soil surface can also result in colonization of dead organic matter with subsequent infection of adjacent living plants. However, in some crops, for example sunflower [5], myceliogenic germination of sclerotia can directly initiate the infection process of the roots and basal stem resulting in wilt. The stimulus for myceliogenic germination and infection in sunflower is not known but likely depends on nutritional signals in the rhizosphere derived from host plants [6].

3 The infection process

Infection of healthy tissue depends on the formation of an appressorium, which may be simple or complex in structure depending on the host surface [7]. In most cases, penetration is directly through the cuticle and not through stomata. Appressoria develop from terminal dichotomous branching of hyphae growing on the host surface and consist of a pad of broad, multi-septate, short hyphae that are orientated perpendicular to the host surface to which they are attached by mucilage. Complex appressoria are often referred to as infection cushions. Although earlier workers [8,9] considered penetration of the cuticle to be a purely mechanical process there is strong evidence from ultra-structural studies that enzymatic digestion of the cuticle also plays a role in the penetration process [10]. Little is known about S. sclerotiorum cutinases, however, the genome encodes at least four cutinase-like enzymes (Hegedus unpublished). A large vesicle, formed at the appressorium tip prior to penetration, appears to be released into the host cuticle during penetration. After penetration of the cuticle, a subcuticular vesicle forms from which large hyphae fan out growing over and dissolving the subcuticular wall of the epidermis.

4 Oxalic acid is central to pathogenesis and has a plethora of roles

The involvement of oxalic acid in pathogenicity of some necrotrophic fungi is well known [11]; however, despite its simple organic structure and limited chemical interactions, oxalic acid plays complex and diverse roles in the infection process (Fig. 1). That oxalic acid is a necessary pathogenicity factor was shown by Godoy et al. [12], who demonstrated that oxalic acid-deficient mutants were non-pathogenic. Bateman and Beer's [13] hypothesis that oxalic acid exerts a direct toxic effect through acidification of the environment within the middle lamellae, sequestering calcium in the form of insoluble oxalate crystals and acting in concert with cell wall depolymerizing enzymes to result in loss of plant cell wall integrity has held up remarkably well. Several fungal pectinolytic enzymes, such as the various endo- and exopolygalacturonases as well as pectin methylesterase [4,14–17] are most active in the acidic, low calcium environment afforded by the secretion of oxalic acid. As well, oxalic acid was also reported to directly enhance the activity of the endopolygalacturonase SSPG1 at very low pH values [18]. Polygalacturonase gene expression, specifically that of Sspg1, increases sharply when the ambient pH falls to less than 3.8 [19,20]. Low pH levels may also disrupt the molecular interaction between plant polygalacturonase inhibitor proteins and cognate pathogen-derived polygalacturonases allowing the enzymes to escape inactivation [18].

1

The multitude of roles played by oxalic acid during infection of plants by necrotrophic fungal pathogens.

Guimaraes and Stotz [21] demonstrated that oxalic acid interferes with stomatal closure, firstly by stimulating the accumulation of potassium and starch hydrolysis in guard cells (both requisite for stomatal opening) and secondly by disrupting the ABA-dependent process leading to stomatal closure. At least part of the resistance to S. sclerotiorum infection among lines of the scarlet runner bean (Phaseolus coccineus) could be attributed to differences in oxalate sensitivity [22]. Ferrar and Walker [23] found oxalic acid to affect the activity of o-diphenol oxidase by competitive inhibition of the enzyme in apple fruit and by reducing pH in bean pods to a level where o-diphenol oxidase was inactive. Oxalic acid has also been shown to suppress the oxidative burst which serves as a front line host defence mechanism through the formation of reactive oxygen species (ROS) and hydrogen peroxide [24].

5 Interplay between plant cell wall degrading enzymes during host infection

Oxalic acid works in concert with cell wall degrading enzymes, such as polygalacturonase (PG), to bring about the destruction of host tissue by creating an environment conducive for PG attack on pectin in the middle lamella. This in turn releases low molecular weight derivatives that induce the expression of additional PG genes. Indeed, overall PG activity is induced by pectin or pectin-derived monosaccharides, such as galacturonic acid, and is repressed by the presence of glucose [25–26]. Examination of the expression patterns of individual Sspg genes has revealed that the interplay among PGs and with the host during the various stages of infection is finely co-ordinated. For the sake of clarity and comparison, we have adopted the nomenclature used to describe PGs from Botrytis species [27]; a convention which is supported by phylogenetic analysis [28].

During in vitro saprophytic growth, the increase in PG activity was found to result from the sequential production of at least 16 PG isoforms, spanning a broad pI range, with successive replacement of enzymatic variants [29,30]. To date, genes encoding only four types of S. sclerotiorum endo-PGs (SSPG1, SSPG3, SSPG5, SSPG6) and two exo-PGs (SSXPG1 and SSXPG2) have been isolated [31] and it is plausible that post-translational (glycosylation) and post-secretional (proteolytic) modifications lead to the numerous isoforms observed. In planta, the systematic appearance of individual Botrytis cinerea PGs attests to the role each plays in the infection process [32] and recent evidence suggests that SSPG1 plays a role in both initiation of the infection and subsequent lesion expansion [31]. In B. cinerea, Bcpg1 and Bcpg2 are expressed at the earliest stages of the infection [32] and mutants unable to produce BCPG1 exhibit reduced rates of lesion expansion [33]. For S. sclerotiorum and many other pathogenic fungi, interaction with solid surfaces induces pre-penetration events, such as appressorium formation [7,34] and PG expression [31,35]. Expression of Sspg1 is induced by thigmotrophic interactions with hard, hydrophobic surfaces such as polystyrene, wax film and presumably plant cuticle [31] and appears to be confined to the expanding margin of the necrotic lesion [18,20].

Expression of Sspg1 precedes that of Sspg3, Sspg5 and Sspg6 during infection [18,28]. Pectin and galacturonic acid induce the expression of Sspg1, Sspg3, Sspg5 and Ssxpg1 and each is subject to catabolite repression by glucose to varying degrees [31]. This observation is in accordance with the notion that PGs are involved in the breakdown of cell wall polymers that provide the pathogen with nutrients prior to the utilization of host glucose [36]. cDNAs encoding enzymes required for mobilization and utilization of host glucose reserves have also been identified [27]. The regulatory region of Sspg1 possesses several DNA binding sites for the CRE1 repressor, a factor involved in carbon catabolite repression and believed to restrict Sspg1 expression under saprophytic conditions [37]; however, a basal level of Sspg1 expression is always observed even in the presence of glucose [31]. This is advantageous for the rapid colonization by germinating ascospores of senescing or dead organic matter that is essential for infection of healthy tissue. As well, constitutive expression of at least some SSPG1 is necessary since the contact zone between the pathogen and host provides few inductive signals and the pectin monomers resulting from SSPG1 activity can serve to induce additional endo- and exoPGs.

Structural proteins, such as extensins, are also vital for the maintenance of plant cell wall integrity and genes encoding an acid protease (ACP1) and an aspartyl protease (ASPS) have been identified from S. sclerotiorum [38,39]. AspS is expressed at the very early stages of the infection [39] and may function in concert with SSPG1 to promote the advancement of invading hyphae. Conversely, Acp1 is expressed during saprophytic growth on plant extracts and is responsive to glucose and nitrogen starvation as well as acidification [38].

6 Selective activation and inactivation of host defenses

Plant defences against fungal pathogens consist of both a localized response that is often associated with an oxidative burst and a more generalized systemic response mediated by signaling molecules. The oxidative burst results in the formation of reactive oxygen species (ROS) that form a toxic barrier to pathogen invasion. To combat this defence, fungal pathogens have evolved several strategies to limit the formation of or to directly overcome the effects of ROS. As noted above, oxalic acid can suppress the oxidative burst [24]. The mechanism underlying this inhibition is unknown but was not due to a reduction in apoplastic calcium levels which would interfere with calcium signalling. Other compounds derived from necrotrophic pathogens, such as 2-methyl-succinate from B. cinerea, may also limit the generation of ROS [40]. This metabolite had no effect on the initial oxidative burst but suppressed the later secondary burst that normally occurs as the pathogen spreads. ROS toxicity is at least partly due to the induction of programmed cell death (PCD) in the invading pathogen. Proline was shown to be an effective scavenger of intracellular ROS in Colletotrichum trifolii [41], but whether this amino acid can protect the pathogen from ROS released by affected plant cells has not been examined. Necrotrophic fungi also produce enzymes, such as superoxide dismutase and catalase, that serve to deplete ROS [42] and a cDNA encoding a zinc-superoxide dismutase was isolated from S. sclerotiorum [27]. Enzymes capable of detoxifying various plant phytoalexins, such as polyphenols [43] and brassinin [44] have also been identified from B. cinerea and S. sclerotiorum, respectively.

The production of ROS presents a dilemma for the host when defending itself against necrotrophic pathogens. On one side, ROS are rapidly produced to limit further disease progression. Alternatively, ROS have been implicated in the hypersensitive response (HR), a form of plant PCD, leading to the formation of necrotic tissue from which such pathogens obtain nutrients. Indeed, the accumulation of ROS was found to be directly proportional to the degree of infection [45]. The importance of the HR for successful infection by necrotrophic pathogens is exemplified by studies showing that plants unable to undergo PCD [46] or are incapable of generating a HR [45] are more resistant to such pathogens. This presents a dichotomy for the pathogen as it engages to suppress the oxidative burst or destroy the ROS that may serve to initiate the host HR. In an interesting evolutionary twist, necrotrophic fungi have developed ancillary factors that promote the host HR in the absence of ROS [47]. Botrytis elliptica secretes proteinaceous factors that initiate the HR via one or more cellular pathways including calcium flux, sphingolipid metabolism, nitrous oxide and oxidate signaling (Fig. 2). Recently, S. sclerotiorum PGs present early in the infection process were implicated in initiating the host HR in a calcium dependent manner [48]. This same phenomenon was observed with B. cinerea and was dependent upon PG activity as mutant forms did not induce necrosis, but was unrelated to the type of oligogalacturonide produced [49]. Therefore, necrotrophic fungi are capable of suppressing the oxidative burst and inactivating any residual ROS while still initiating a host HR.

2

Signalling pathways that may be involved in initiating cell death in response to factors secreted by Botrytis elliptica. Reprinted with permission [47].

The question as to whether the induction of a systemic defence response provides some measure of resistance remains unresolved. Induction of systemic resistance to Collectotrichum lagenarium in cucumber [50] and to S. sclerotiorum in oilseed rape [51] by oxalic acid has been demonstrated. In Arabidopsis thaliana, infection with B. cinerea, another oxalic acid producing fungus, led to the expression of various defence genes; however, this response was insufficient to protect against subsequent biotrophic pathogen attack [52]. The same study also showed that while induction of systemic acquired resistance with salicylic acid failed to inhibit fungal growth, reducing salicylic acid levels increased disease symptoms. PG activity releases oligo-galacturonides that are also potent inducers of plant defenses. Interestingly, B. cinerea BcPG1 induced a defense response in grape that was independent of its enzymatic activity and might therefore be considered a true avirulence entity [53]. A defense response may also be induced by other pathogen-derived factors such as β-1-3-glucan [54]. In Brassica napus, S. sclerotiorum infection induces the expression of polygalacturonase inhibitors [55], which unlike A. thaliana, possesses a large multi-gene family encoding more than 20 proteins (Hegedus unpublished).

7 cAMP is a key modulator of saprophytic growth, pathogenesis and development

The transition from the pathogenic to the necrotrophic–saprophytic phase appears to be related to the intracellular level of cAMP which is influenced by glucose levels and by ambient pH. The ambient pH in the immediate vicinity of the invading mycelium is determined by secretion of oxalic acid. The neutral (or alkaline) pH environment in host tissues at the onset of pathogenesis stimulates production of oxalic acid causing a reduction in ambient pH. This serves to self-limit further oxalic acid accumulation but induces expression of PG [56]. Such pH-dependent processes are regulated to some extent by the ambient pH sensor PAC1, which under alkaline conditions both activates gene expression, such as those involved in oxalic acid accumulation, and represses the expression of genes normally induced under acidic conditions, for example Sspg1 [19]. Interestingly, while pac1 mutants exhibited reduced oxalic acid levels and increased PG expression at neutral pH as expected, acidic ambient pH alone failed to derepress PG expression [57]. Low glucose levels during the pathogenic phase beget high intracellular levels of cAMP, which leads to de-repression of PG expression through a protein kinase A (PKA)-dependent phosphorylation and inactivation of the CRE1 repressor [58,59]. Glucose likely only becomes available during the later stages of the infection and could signal the transition to the necrotrophic/saprophytic phase. An even finer level of regulation is provided by exclusion of CRE1 from the nucleus when glucose levels are low [60,61]. Similarly, ACP1 protease expression is also stimulated by low glucose and increased cAMP levels [62].

The provocative interplay between cAMP signalling and ambient pH sensing is more clearly demonstrated by their effect on sclerotial morphogenesis. High cAMP levels stimulate oxalic acid production and prevent the formation of sclerotia, while an acidic ambient pH environment, which is the consequence of oxalic acid production, promotes sclerotia development [56]. Sclerotia form during the final stages of host colonization within the necrotic region where the pH is acidic due to the production of oxalic acid during the pathogenic phase. Here, high glucose levels resulting from the utilization of starch cause cAMP levels to drop. This results in the loss of inhibition toward sclerotial development which is furthered by a pH environment already conducive to sclerotia formation. Recently, the inhibition of sclerotial development by cAMP was found to be mediated by interference with a mitogen-activated kinase (MAPK) pathway [63]. An acid pH environment normally leads to the activation of a Ras/MAPK and ultimately the downstream MAPK, Smk1. However, in the presence of cAMP a Rap-1 GTPase becomes active; this is another member of the Ras family which interferes with the interaction between Ras and an uncharacterized MAPKKK preventing Smk1 activation [64]. Therefore, high levels of cAMP during the pathogenic phase over-ride the stimulatory effect of low pH on sclerotial development thereby preventing precocious formation of sclerotia until the ultimate stages of the infection.

8 Summary

It is becoming abundantly clear that disease progression and destruction of host tissue by necrotrophic fungi is not a rampant and unrestrained process but rather a finely coordinated series of events. We have separated this process into three distinct phases (Fig. 3). The first is an opportunistic–saprophytic phase that allows for ascospore germination on senescent plant tissues and subsequent pathogen establishment. The second is a pathogenic phase having two spatially separate zones. The frontal zone at the leading edge of the lesion involves the constitutive expression of PG (SSPG1) and protease (ASPS), the release of oxalic acid and possibly the secretion of factors that initiate the plant HR. In the trailing zone immediately following this, low glucose (high cAMP) levels, low ambient pH and oligo-galacturonides and peptides released by the actions of enzymes in the frontal zone unite to maximally induce cell-wall degrading enzyme synthesis and tissue necrosis. This is followed by a second saprophytic phase whereby cAMP levels decline as host glucose becomes available and further hydrolytic enzyme synthesis is repressed. Low cAMP levels and an acidic ambient pH promote sclerotial development and completion of the life cycle.

3

Tri-phasic model for Sclerotinia sclerotiorum pathogenesis. In this model, an initial saprophytic phase is required for pathogen establishment on the host which followed by a pathogenic phase and a final saprophytic to complete the life cycle on the host. The transition between phases is governed by external levels of glucose, internal levels of cAMP and ambient pH. Legend: SSPG1, polygalacturonase 1; ASPS, aspartyl protease; OGA, oligo-galacturonides; PCD, programmed cell death; cAMP, cyclic adenosyl monophosphate; PKA, protein kinase A; CRE1, catabolite repressor 1; MAPK, mitogen-activated protein kinase. Arrows indicate increased or decreased levels of important metabolites and pH.

Despite the wealth of new information many key questions remain unanswered. So far, the role and relative contribution of individual PGs to the infection process is largely speculative. The availability of the entire S. sclerotiorum genome sequence will identify the entire suite of genes encoding such enzymes and allow for systematic disruption of each. It is not obvious how oxalic acid suppresses the oxidative burst or why the induction of host systemic defences provides measurable resistance to necrotrophic pathogens in some hosts but not in others. This latter phenomenon seems to imply that temporal and spatial patterns of defence gene expression are critical for resistance and that if defence gene expression is to be used as an indicator of potential resistance it must be examined on a much finer scale. Recent genetic studies have identified several A. thaliana mutants (bos) that are more susceptible to B. cinerea; however, these appear to affect disparate defence pathways [65,66]. Furthermore, the mechanisms that underlie the general intolerance of monocotyledonous plants to S. sclerotiorum infection are unknown. In such cases, it is possible that mechanisms other than defence responses, such as structural barriers or phytoalexins, may confer non-host resistance. Further elucidation of the signalling pathways that provide the interconnection between pathogen biochemistry and physiology as well as identification of the factors that induce host PCD will provide exciting new opportunities for research and should have profound effects on our understanding of host–pathogen interactions.