Infection Strategies of Colletotrichum Species
An understanding of the mode of infection of individual Colletotrichum species is a prerequisite for developing effective control strategies, particularly those based on host plant resistance. Knowledge of the factors influencing infection processes also provides epidemiologists with information which can be developed into forecasting models, and aids agronomists developing appropriate agricultural practices, based on crop sanitation and removal of volunteer, reservoir or collateral hosts.

Spore Dispersal and Adhesion
In many Colletotrichum species, conidia and ascospores are embedded in a matrix of moist hydrophilic mucilaginous material (Griffiths and Campbell, 1972; Nicholson and Moraes, 1980; McRae and Stevens, 1990), comprised mostly of polysaccharides and high molecular weight glycoproteins (Ramadoss et al., 1985; Louis et al., 1988). These matrices are readily soluble in water, with spores being released and dispersed by the action of free water (usually rain). The matrix may have several roles. Firstly it maintains the viability of conidia under adverse conditions such as extreme temperature, ultraviolet light, and low humidity (Nicholson and Moraes, 1980; Nicholson et al., 1986; Louis et al., 1988), and protects spores from toxic material produced in the host tissues during lesion development (Nicholson and Moraes, 1980; Nicholson et al., 1986, 1989). Secondly, there is evidence that it can also inhibit conidial germination, thus ensuring that conidia do not germinate until they are dispersed from the acervulus (Louis and Cooke, 1985; Seebach et al., 1989; Leite and Nicholson, 1992).

The first essential feature of successful pathogenesis is the attachment of dispersed fungal propagules to the host plant surface (Hamer et al., 1988; Nicholson and Epstein, 1991). Studies have shown that Colletotrichum conidia will adhere rapidly to a wide range of plant and artificial surfaces, including cellophane, polystyrene, polycarbonate and glass (Young and Kauss, 1984; Mercure et al., 1994a, b 1995; Sela-Buurlage et al., 1991), suggesting that adhesion is non-specific. Furthermore studies by Mercure et al. (1994a, b, 1995) showed that the respiration inhibitor sodium azide and the transcription inhibitor, actinomycin D, had no effect on adhesion suggesting that adhesion of ungerminated conidia is, at least in part, a passive process.

The ultrastructure of the conidia of several species of Colletotrichum (C. lindemuthianum, C. truncatum, and C. graminicola), has been examined after preparation by cryofixation and freeze-substitution (Van Dyke and Mims, 1991; Mims et al., 1995; O’Connell et al., 1996). These studies showed that the walls of ungerminated and germinated conidia were coated with a layer of fibrillar material, the ‘spore coat’, which was more electron dense than the underlying cell wall. An extracellular matrix with similar ultrastructure has been observed on various yeasts (Walther et al., 1988; Viret et al., 1994). In Candida albicans, fibrillar mannoproteins are responsible for the hydrophobicity of the cell surface and control the adhesion of this pathogen to host cells (Hazen and Hazen, 1992). It has been suggested that the spore coat in Colletotrichum may have a similar function, since the initial rapid adhesion of ungerminated conidia is known to involve hydrophobic interactions between proteins on the conidial surface and substrata (Young and Kauss, 1984; Sela-Buurlage et al., 1991; Mercure et al., 1994). The spore coat appears to be a preformed structure, present on ungerminated, unimbibed conidia (Van Dyke and Mims, 1991; O’Connell and Carzaniga, personal communication), which is consistent with a role in the initial attachment of conidia to the plant surface. In some Colletotrichum species, the subsequent release of protein exudates, may consolidate the initial hydrophobic attachment of conidia (Jones et al., 1995; Mercure et al., 1995).

Modes of Host Penetration

1. Penetration without appressoria
   The mechanisms by which species of Colletotrichum penetrate their hosts have been debated for many years. Several modes of penetration are possible: through natural openings, e.g. stomata, through wounds and by direct penetration of the cuticle. Of these, direct penetration is the most common means of tissue penetration (Bailey et al., 1992). There are few examples of stomatal penetration in the literature, however, it has been reported for the infection of rubber (Hevea brasiliensis) leaves by C. gloeosporioides (Sénéchal et al., 1987; Zakaria, 1995). Infection through wounds is not common, and is not usually a prerequisite for infection. However, for some diseases, e.g. crown and finger stalk rot of banana, infection through wounds is essential (Krantz et al., 1978; Agrios, 1988). Van der Bruggen and Maraite (1987) and Van der Bruggen et al. (1990) showed that undamaged cassava stems were resistant to C. gloeosporioides, with the pathogen being restricted within the epidermal layer. However, after attack by an insect, the pathogen was able to colonise the damaged tissue, and sporulating lesions formed rapidly. Using a hot needle to simulate the effects of an insect proboscis had a similar effect. In neither of these examples was the wound itself the initial site of penetration. The pathogen initially gained entry to the tissue through the cuticle, but further growth was restricted to within the epidermal layer, and it was only able to colonise the tissues after the death of the underlying tissues. Anthracnose diseases of tropical fruits which involve a period of latency, may have close analogies with the events described above. In these cases the pathogen is restricted within the epidermal layer in a latent or dormant form until after fruit ripening, when physiological changes in the host stimulate further pathogen development (Brown, 1975; Prusky and Plumbley, 1992). Examples of penetration without the formation of appressoria have been observed in C. gloeosporioides on Aeschynomene virginica, Carica papaya and on Citrus spp., and C. musae on Musa spp. (Brown, 1975; TeBeest et al., 1978; Muirhead and Deverall, 1981; Chau and Alvarez, 1983; and Porto et al., 1988).

2. Appressoria Formation
   In the majority of Colletotrichum species appressoria are usually, but not always a prerequisite for penetration of host cuticles. Appressoria are often sessile, but may form at the end of distinct germ-tubes and are sometimes produced at the tips of mycelial branches (Parbery, 1981). Appressorium formation is non-specific and they will form readily in the absence of a host surface, for example when conidia germinate on a hard surface such as a glass slide or a nitrocellulose membrane (Emmett and Parbery, 1975; Lenné, 1978).

   Ultrastructural studies have shown that appressoria of different Colletotrichum species have many features in common. Usually the appressorial wall has a two or three-layered structure, with melanin preferentially deposited within one layer (Bailey et al., 1992). Melanisation of the appressorial cell wall has also been shown to be necessary for mechanical penetration of the host cuticle and underlying cell wall by C. lindemuthianum, C. lagenarium, and C. graminicola (Wolkow et al., 1983; Bonnen and Hammerschmidt, 1989a,b; Rasmussen and Hanau, 1989). Melanin deficient strains of Colletotrichum could not penetrate artificial membranes nor infect normally susceptible host plants (Suzuki et al., 1982; Katoh et al., 1988). Restoration of melanin biosynthesis, by addition of a specific precursor (scylactone), restored pathogenicity to melanin deficient-mutants of C. graminicola (Wolkow et al., 1983; Rasmussen and Hanau, 1989).

   During penetration of plant surfaces, appressoria of some species, e.g. C. gloeosporioides (Brown, 1977), C. lagenarium (Xuei et al., 1988), C. lindemuthianum (O’Connell and Bailey, 1990) and C. trifolii (Mould et al., 1991), form germ-pores in their walls, which are surrounded by a funnel-shaped structure, termed the appressorial cone (Landes and Hoffman, 1979a). This cone appears to be an extension of the infection peg wall and may act to focus hydrostatic pressure to the site of penetration (Mercer et al., 1971; Landes and Hoffman, 1979b; Wolkow et al., 1983). In other species, e.g. C. graminicola (Politis and Wheeler, 1973), C. truncatum (O’Connell et al., 1993) and C. destructivum (Latunde-Dada et al., 1996) such cones have not been observed, suggesting that different Colletotrichum species vary in this respect, and that these structural variations could reflect with other different mechanisms for penetrating plant surfaces.

   Adhesion of appressoria to the plant surface is essential for successful penetration of the cuticle and underlying cell wall. Adhesion ensures that the pathogen remains in contact with the host for sufficient time for penetration to occur, as well as placing the infection hypha at a site where penetration, whether mechanical or enzymatic, can be achieved. Firm anchorage is also essential for appressoria to exert the mechanical force required for penetration (Bailey et al., 1992). Appressorium formation is often accompanied by the secretion of a mucilaginous matrix, which surrounds the appressorium (Jones et al., 1995; O’Connell et al., 1996; Pain et al., 1996). This matrix extends outwards over the host surface, and appears to attach the appressorium to it (O’Connell et al., 1985; O’Connell and Ride, 1990; Bailey et al., 1992). However, it is interesting to note that mucilage is absent where the appressorial wall is most closely appressed to the plant cuticle. It remains unclear whether the mucilage is involved in adhesion, or whether adhesion is a distinct phenomenon. Appressorial mucilage, like the spore matrix, may be involved in protecting appressoria from extremes of heat, cold and/or desiccation, rather than adhesion. There is little information about the chemical nature of either the adhesives or the mucilage. However, studies using lectins and monoclonal antibodies (O’Connell et al., 1996; Pain et al., 1996) have identified specific carbohydrates and glycoproteins within the mucilages surrounding Colletotrichum infection structures. Furthermore, the mucilage surrounding appressoria and germ-tubes was shown to differ markedly in structure and composition from those surrounding conidia. Recent studies have also shown that Colletotrichum species may even incoporate host epicuticular proteins into these extracellular mucilages (Hutchison et al., 1996).

   Three mechanisms have been proposed for cuticle penetration: a) mechanical force alone; b) the secretion of cutin degrading enzymes alone; c) a combination of both processes (Bailey et al., 1992). There is evidence that the appressoria of some species of Colletotrichum (e.g. C. graminicola, C. lagenarium, C. lindemuthianum) can exert sufficient mechanical force to penetrate cuticles (Mercer et al., 1971; Suzuki et al., 1982; Wolkow et al., 1983; Rasmussen and Hanau, 1989; O’Connell et al., 1992; Pascholati et al., 1993). Alternatively, penetration may require enzymes to dissolve or soften the host cuticle. Several species of Colletotrichum (e.g. C. gloeosporioides, C. lagenarium, and C. capsici) produce esterases capable of degrading cutin (Bailey et al., 1992). Assessments of the role of cutinase in infection are based on work with specific inhibitors of cutinase, e.g. di-isopropyl fluorophosphate (DFP) and with mutants lacking cutinase activity. Dickman and Patil (1986) showed that cutinase-deficient mutants of C. gloeosporioides were not pathogenic when placed on surfaces of intact papaya fruit, but when fruits were wounded or their surface treated with cutinase normal lesions were produced. Similarly, lesions were not produced when chemical inhibitors of cutinase or antibodies raised to cutin were mixed with conidia of C. gloeosporioides (Dickman et al., 1983). However, penetration of plant surfaces by C. lindemuthianum was not prevented by DFP (Wolkow et al., 1983). Studies by Bonnen and Hammerschmidt (1989a, b) on cutinase mutants of C. lagenarium also showed that inhibition of cutinase did not affect pathogenicity.

Back to the top of the page.

Modes of Nutrition and Tissue Colonisation
Colletotrichum species use a variety of strategies to successfully colonise host tissue and avoid host defense responses. Based on their initial infection strategies, these can be categorised as; a) intracellular hemibiotrophic; b) subcuticular intramural; c) both intracellular hemibiotrophic and subcuticular intramural (see Table 2). However, the final phase of all three infection strategies is always necrotrophic (Bailey et al., 1992).

1. Intracellular Hemibiotrophic Infections
   Many Colletotrichum species (see Table 2) exhibit a two-phase infection and colonisation process involving an initial symptomless or biotrophic phase, during which the pathogen establishes itself in the host tissues, followed by a final visibly destructive phase (O’Connell et al., 1985; Latunde-Dada et al., 1996, 1997; O’Connell et al., 1993). During the symptomless biotrophic phase, the pathogen invades host cells without killing them and feeds on living cells. Subsequently, the pathogen switches to a necrotrophic mode of nutrition, feeding on dead host tissues. It is on this basis, that these species were described as hemibiotrophic by Luttrell, (1974). O’Connell and Bailey (1991) referred to these species as ‘intracellular hemibiotrophic pathogens’, where emphasis is placed not only on their two-phase infection process, but also on their ability to penetrate cell walls and grow within cell lumena. The two-phase infection process was first described in C. lindemuthianum by Leach (1922), who distinguished between the production of an initial ‘primary’ mycelium in the symptomless phase, and the transition to a ‘secondary’ mycelium seen in the visibly destructive phase. These were later illustrated in more detail by Skipp and Deverall (1972), Allard (1974), Mercer et al. (1975), Landes and Hoffman (1979) and O’Connell et al. (1985).

   The intracellular hemibiotrophic Colletotrichum species can be further subdivided into two groups: a) those in which the biotrophic phase extends to many host cells, for example C. lindemuthianum on Phaseolus vulgaris (O’Connell et al., 1985), and those in which the biotrophic phase occurs only inside the initially infected epidermal cell, for example C. destructivum on Medicago sativa (Latunde-dada, 1997) and C. truncatum on Pisum sativum (O’Connell et al., 1993). In species where the biotrophic phase extends to many host cells, primary hyphae branch out from the infection vesicle within the initially infected epidermal cell and colonise adjacent cells. As the primary hyphae advance from cell to cell a biotrophic relationship is established in each successive cell colonised, while the previously infected cells gradually senesce and lose their viability (Plate). This is characterised by the cessation of cytoplasmic streaming, loss of membrane functional integrity (inability of the tonoplast and plasma membrane to plasmolyse and accumulate vital dyes), and cytoplasmic disorganisation. As a consequence, the most recently colonised cells are intact and viable, while at the same time earlier infected cells are dead or dying (O’Connell et al., 1985). This form of biotrophy is transient, lasting only 24 - 48 h in each infected cell, and differs from the more stable biotrophic relationships established by rusts, powdery mildews and other obligate fungi. In the group of Colletotrichum species which exhibit a biotrophic phase only in the initially infected epidermal cell, the fungus produces a large, multilobed, multiseptate infection vesicle, which remains confined within the cell lumen. After approximately 48 h, necrotrophic secondary hyphae emerge from the vesicle, and rapidly colonise surrounding cells (O’Connell et al., 1993; Latunde-dada, 1996, 1997).

   Electron microscopy of the infection processes of these intracellular hemibiotrophic pathogens shows that that the membranes of infected epidermal cells are invaginated around the infection vesicles, and that the cytoplasm of the infected cells initially shows no structural abnormalities (O’Connell and Bailey, 1991; O’Connell et al., 1993). At this stage the vesicle resembles the haustoria of obligate biotrophic pathogens and, like haustoria, it is separated from the host plasma membrane by a matrix layer (Brown, 1977; O’Connell 1987). The matrix layer around vesicles and primary hyphae of C. lindemuthianum arises from both the host and the pathogen, and contains glycoproteins and polysaccharides, but not chitin (O’Connell et al., 1986; O’Connell, 1987; O’Connell and Ride, 1990). The function of the matrix is not known. However, recent work using monoclonal antibodies has shown that it differs in composition from the matrices surrounding appressoria, germ-tubes and conidia and contains a multimeric complex composed of 32kD N-linked glycoprotein subunits (Pain et al., 1994). The fungal gene encoding this protein was recently cloned and sequenced (Perfect, Green and O’Connell, unpublished). The deduced amino acid sequence contains a proline-rich domain but shows no homology to any known protein. It has been suggested that the interfacial matrix may function in establishing basic compatibility, perhaps interfering with the plant-pathogen recognition event(s) that determine resistance (Siegrist and Kauss, 1990), or by immobilising extracellular fungal toxins or enzymes (Bailey, 1992).

   The invaginated plant plasma membrane formed around the haustoria of obligate biotrophs is modified to facilitate the required, and apparently specialised, nutrient uptake of these fungi (Gay, 1984). Modifications include the formation of a ‘neck band’, which fuses the host plasma membrane to the neck wall of the haustorium, the absence of cytochemically detectable adenosine triphosphatase activity and changes to the topography of the membrane. However, none of these modifications occur around the vesicles or primary hyphae of the intracellular hemibiotrophic Colletotrichum species, implying that biotrophic nutrition is less specialised in these fungi (O’Connell, 1987).

   Table .2. Initial infection strategies of Colletotrichum species (adapted from Bailey et al., 1992)

   Colletotrichum	Host	Reference
   A. Intracellular hemibiotrophic species
   *destructivum     gloeosporioides       graminicola *lindemuthianum           *malvarum lagenarium       *truncatum	Vigna unguiculata Medicago sativa   Stylosanthes guianensis Stylosanthes scabra Populus tremuloides Citrus spp. Zea mays Phaseolus vulgaris           Malva pusilla Cucumis melo       Pisum sativum	Bailey et al. (1990) Latunde-Dada et al. (1996) Latunde-Dada et al. (1997) Ogle et al. (1990) Trevorrow et al. (1988) Marks et al. (1965) Brown (1977) Politis and Wheeler (1973) Elliston et al. (1976) Landes and Hoffman (1979b) Mercer et al. (1975) O'Connell et al. (1985) O'Connell and Bailey (1986) Skipp and Deverall (1972) Wei et al. (1997) Anderson and Walker (1962) Dargent and Touzé (1974) Stumm and Gessler (1984) Xuei et al. (1988) Uronu (1989) O'Connell et al. (1993)
   B. Subcuticular intramural species
   capsici   circinans gloeosporioides   phomoides	Gossypium hirsutum Vigna unguiculata Allium cepa Carica papaya Muscadinia rotundifolia Lycopersicum esculentum	Roberts and Snow (1984) Pring et al. (1995) Walker (1921) Chau and Alvarez (1983) Daykin and Milholland (1984) Fulton (1948)
   C. Species that exhibit both intracellular hemibiotrophic and subcuticular intramural infections
   gloeosporioides           trifolii	Stylosanthes spp.     Citrus spp. Hevea brasiliensis   Medicago sativa	Irwin et al. (1984) Trevorrow et al. (1988) Vinijsanun et al. (1987) Brown (1977) Sénéchal et al. (1987) Zakaria (1995) Porto et al. (1988) Mould et al. (1991)

   *species for which existence of a biotrophic phase has been critically assessed by host cell plasmolysis

   As mentioned previously, when plant tissues have been successfully colonised by Colletotrichum the pathogen switches to a ‘classic’ necrotrophic behaviour. The necrotrophic phase causes the typical anthracnose and blight symptoms of Colletotrichum species. During this stage hyphae proliferate throughout host tissues, inside cells, in walls and through walls and in intercellular spaces. Several species (e.g. C. lindemuthianum, C. gloeosporioides) produce cell wall degrading enzymes and low molecular weight phytotoxins that may, by killing cells in advance of the invading hyphae, contribute to the necrotrophic growth of these pathogens (O’Connell, 1985; Bailey et al., 1992). Eventually conidiophores rupture through the host cuticle and form acervuli on the plant surface (Bailey et al., 1992).

    

2. Subcuticular Intramural Infections
   For Colletotrichum species exhibiting a subcuticular intramural mode of infection (see Table 1.2), penetration is followed by growth of hyphae beneath the cuticle and within the periclinal walls of epidermal cells, but initially not into the lumen of underlying epidermal cells (Bailey et al., 1992; Pring et al., 1995). Continued subcuticular hyphal growth causes considerable wall degradation. This mode of infection was first described by Walker (1921) for C. circinans attacking onion. In these species, there is no visible evidence of infection during the initial 24 hours after penetration, so that like the intracellular hemibiotrophs, the initial infection phase is asymptomatic. However, an extensive network of intramural hyphae is formed within a few days and water-soaked lesions appear. It is not clear from any study of these intramural pathogens whether, during the early symptomless stage, the underlying cells survive infection or are killed (Bailey et al., 1992). As with the intracellular hemibiotrophs, however, it appears that the host tissues do not express any resistance responses, and pathogen growth switches to a necrotrophic mode of nutrition and continues unchecked.

3. Combined Intracellular Hemibitrophic and Subcuticular Intramural Infections
   Although intracellular hemibiotrophic and subcuticular intramural strategies are common, in some species of Colletotrichum this distinction is less clear and these species appear to exhibit both strategies (see Table 1.2). Examples include the infection of some species of Stylosanthes, Citrus and Hevea by C. gloeosporioides (Irwin et al., 1984; Trevorrow et al., 1988; Vinijsanun et al., 1987, Brown, 1977; Sénéchal et al., 1987; Zakaria, 1995). Studies by Zakaria (1995) on the infection of Hevea brasiliensis by C. gloeosporioides showed that different isolates varied markedly in their initial infection strategies. All isolates penetrated the cuticle through production of appressoria. However, some were capable of penetrating the host through open stomata, sometimes after the production of appressoria but on most occasions without appressoria. Others produced extensive mycelial growth on the host surface and penetrated the cuticle without the formation of appressoria. In all cases, after penetration no special infection structures such as infection vesicles and large diameter primary hyphae were produced, and there was no evidence of a biotrophic phase. In some isolates penetration was followed by the production of intracellular hyphae of varying diameter which colonised the host tissue causing no extensive degradation of cell walls. In others, penetration was followed by the production of both intracellular and intramural hyphae which rapidly colonised the host tissue and caused extensive cell wall degradation.

Back to the top of the page.