Phytophthora cinnamomi Rands was first isolated from cinnamon trees in Sumatra in 1922. The pathogen is believed to have originated near Papua New Guinea but now has a worldwide distribution. P. cinnamomi is heterothallic with A1 and A2 mating types; however, even in areas in which both mating types are present, it appears that genetic diversity arises asexually rather than as a result of sexual recombination. P. cinnamomi can grow saprophytically in the soil for long periods, rapidly capitalizing on the advent of favourable conditions to sporulate and produce vast numbers of asexual, biflagellate zoospores. The motile zoospores are attracted to suitable infection sites, where they attach and invade the plant. Within a few days, hyphae ramify throughout the tissues of susceptible plants, forming sporangia on the plant surface and rapidly amplifying the disease inoculum. Over the last 10–15 years, molecular analyses have clarified details of the phylogeny of P. cinnamomi and other Oomycetes. Research on P. cinnamomi has given rise to a more comprehensive understanding of the structure and function of the motile zoospores. New methods have been developed for P. cinnamomi identification and diagnosis. Long-term studies of diseased sites, particular those in southern Australia, have produced a better understanding of the epidemiology of P. cinnamomi diseases. Research has also increased our knowledge of the mode of action and efficacy of inhibitors of P. cinnamomi diseases, especially the phosphonates. This review will present an overview of the advances these studies have made in our understanding of P. cinnamomi pathogenicity, epidemiology and control.
Taxonomy: Phytophthora cinnamomi Rands; kingdom Chromista; phylum Oomycota; order Peronosporales; family Peronosporaceae; genus Phytophthora.
Host range: Likely to infect in excess of 3000 species of plants including over 2500 Australian native species, and crops such as avocado, pineapple, peach, chestnut and macadamia.
Disease symptoms: A root pathogen which usually causes rotting of fine and fibrous roots but which can also cause stem cankers. Often causes dieback of young shoots and is thought to do so as a result of interference with transpiration from roots to shoots.
The rapid spread of Phytophthora cinnamomi in many countries throughout the world during the last century, often inadvertently in contaminated nursery stock and other plant material, its devastation of important food crops and its decimation of large areas of native forest (Fig. 1) has meant that it has been a major focus of Phytophthora research. Early studies culminated in the American Phytopathological Society publication of a 96-page monograph on P. cinnamomi (Zentmyer, 1980) and in P. cinnamomi being one of the species that was prominently discussed at an international symposium on Phytophthora held at the University of California at Riverside in 1981 (Erwin et al., 1983). Since that time, the number of plant species known to be susceptible to P. cinnamomi has increased considerably. P. cinnamomi has continued to cause extensive economic losses in agriculture, horticulture and forestry and is a major threat to natural ecosystems and biodiversity.
P. cinnamomi can grow saprophytically in the soil and persists in soil or infected plant material as chlamydospores and, to a lesser extent, as oospores (Weste, 1983; Zentmyer and Mircetich, 1966) (Fig. 2). When conditions favouring growth prevail, the pathogen enters the asexual sporulation cycle (Fig. 2). Somatic hyphae form multinucleate sporangia that cleave and release 20–30 uninucleate, biflagellate zoospores. The wall-less zoospores encyst, forming walled cysts that germinate and penetrate the plant. Within 2 or 3 days in a susceptible host, sporangia will form on the plant surface. The asexual cycle may be repeated many times in quick succession, rapidly amplifying the inoculum potential in the infected area.
P. cinnamomi is known to survive for as long as 6 years in moist soil (Zentmyer and Mircetich, 1966) and it is clear that moisture is a key factor in the establishment, spread and longevity of P. cinnamomi diseases. Asexual sporulation requires a liquid environment, both for the formation of sporangia and for the release and activity of motile zoospores. Disease development is enhanced after heavy rain and in waterlogged soils. The biflagellate zoospores produced during asexual sporulation are not only the main infective agents for P. cinnamomi but they also possess one of the structural features that characterizes the taxonomic and phylogenetic groups to which Phytophthora and other oomycetes belong, namely the tripartite hairs called mastigonemes on the anterior flagellum.
Traditional taxonomic discrimination of Phytophthora species has been based predominantly on morphology and cultural criteria. Important structural characters include the apical thickening and width of the exit pore of sporangia, the caducity and pedicel length of sporangia, whether antheridia are amphigynous or paragynous and whether the isolate is heterothallic or homothallic (Waterhouse et al., 1983). Comparisons of these characteristics led Waterhouse to divide the genus into six main groups and these have served as the basis of taxonomic keys (Stamps et al., 1990; Waterhouse, 1963). P. cinnamomi was placed in Group VI along with P. cambivora, P. cryptogea, P. drechsleri, P. erythroseptica and P. vignae.
Comparisons of sequence data from the rDNA ITS regions have also given rise to a rather different picture of the relationships between genera within the Oomycetes than previously envisaged (Barr, 1983). Although the close relationship between Phytophthora and Pythium is confirmed, the analysis reveals that species of Peronospora and Bremia cluster with those of Phytophthora and thus appear to be obligate biotrophic Phytophthora lineages that have lost the ability to produce zoospores (Cooke et al., 2000).
The Oomycetes are recognized by their mycelial growth habit, aseptate hyphae and their production of biflagellate, heterokont zoospores. Although their fungus-like hyphae and mode of nutrient acquisition have led to inclusion of the Oomycetes within the fungal kingdom, it has long been recognized that there are significant structural and biochemical differences between the Oomycetes and the true fungi. Features that differ include the structure of the flagellar apparatus and spindle pole, the morphology of mitochondrial cristae, the occurrence of cellulose rather than chitin in the cell walls, differences in lysine and tryptophan biosynthetic pathways and completion of meiosis immediately prior to gametogenesis leading to a diploid somatic thallus (reviewed in Hardham et al., 1994). Structural analyses grouped the Oomycetes with the chromophyte algae and other taxa that possess tubular flagellar hairs, called mastigonemes, in an assemblage called the Stramenopiles (Barr, 1992; Patterson and Sogin, 1992). Molecular studies have confirmed the close relationship between these organisms and have shown that the Stramenopiles form a monophyletic group (Gunderson et al., 1987; Van de Peer and De Wachter, 1997) (Fig. 3). The molecular data indicate a close phylogenetic relationship between the Stramenopiles and the other protist group with tubular mitochondrial cristae, namely the Alveolates (Sogin and Silberman, 1998; Van de Peer et al., 1996). The Alveolates are distinguished by possession of flattened membranous alveoli beneath their plasma membrane. The group includes the apicomplexan malarial parasites such as Plasmodium falciparum, Cryptosporidium parvum and Toxoplasma gondii.
CELL AND MOLECULAR BIOLOGY OF ASEXUAL DEVELOPMENT AND PATHOGENICITY IN P. CINNAMOMI
Asexual sporulation and zoosporogenesis
In P. cinnamomi, asexual sporulation, the development of multinucleate sporangia, can be induced synchronously in vitro by replacing the nutrient medium with a simple mineral salts solution (Chen and Zentmyer, 1970). Sporangia first appear approximately 7 h later at the hyphal tips as cell components flow from the subtending hyphae into the expanding apex (Fig. 4A,B). At maturity, the sporangial cytoplasm is sealed off by deposition of a basal septum and typically contains 20–30 regularly spaced nuclei (Fig. 4C). The majority of nuclei occur near the sporangial wall and are aligned so that the narrow apex of the pyriform nucleus points towards the wall (Hyde et al., 1991a). Nuclear shape and distribution are maintained by microtubules that extend along the nuclear surface and that form aster-like arrays radiating into the cytoplasm from the narrow pole of the nucleus (Fig. 4D,E) (Hyde and Hardham, 1992, 1993).
A number of modern approaches have been applied to studies of Phytophthora species, including P. cinnamomi, in order to identify changes in gene expression and protein synthesis that occur during sporulation. A cDNA library has been made from mRNA isolated from P. cinnamomi mycelia 4 h after the induction of sporulation (Weerakoon et al., 1998). In an expressed sequence tag (EST) study, more than 5000 clones from this library have been differentially screened with probes made from mRNA isolated from vegetative hyphae or from mycelium 4 h after induction of sporulation (R. Narayan, J. S. Marshall and A. R. Hardham, unpublished observations). Over 300 clones reacted more strongly with the sporulation probe than with the vegetative probe and represented more than 200 different genes that are preferentially expressed during the early stages of asexual sporulation. About 50% of these genes have homologues in the sequence databases and encode ribosomal, metabolic, structural and regulatory proteins. A number of these genes have been fully sequenced and their genomic organization and patterns of expression determined (Marshall et al., 2001a,b; Weerakoon et al., 1998).
Proteomic studies also indicate that there are extensive changes in the complement of proteins synthesized during Phytophthora sporulation and zoospore formation (Shepherd et al., 2003; L. M. Blackman and A. R. Hardham, unpublished observations). A powerful way in which to study the synthesis of sporangial and zoospore proteins and the ontogeny and function of the organelles with which they are associated is through the production of antibodies that react specifically with selected proteins. This approach has been applied extensively and effectively to investigations of sporulation and zoospore cell biology in P. cinnamomi. Immunocytochemical studies of sporulation in P. cinnamomi have documented the synthesis of mastigonemes and three types of vesicles destined for the zoospore cortical cytoplasm (Cope and Hardham, 1994; Dearnaley et al., 1996) (Fig. 4F,G), and the re-organization of these and cytoskeletal components during sporangial cleavage (zoosporogenesis).
P. cinnamomi sporangia persist in sporulating cultures until they receive a signal (e.g. a decrease in temperature) that triggers cleavage of the sporangial cytoplasm into zoospores (Fig. 4H–K). Little is known of the receptors that perceive this signal but there is evidence that signal transduction involves increases in cytoplasmic Ca2+ concentration and pH (Jackson and Hardham, 1996; Suzaki et al., 1996). Four main structural changes occur during zoosporogenesis: the multinucleate sporangial cytoplasm becomes subdivided into uninucleate compartments, the zoospores; flagella are assembled from the two basal bodies at the apex of each nucleus; a number of organelles, including mitochondria and vesicles, become polarized within the cytoplasm; and a water expulsion vacuole forms and begins to function in each zoospore.
Subdivision of the sporangial cytoplasm is initiated through fusion of small, moderately electron-dense vesicles clustered near the narrow nuclear pole in uncleaved sporangia (Hyde et al., 1991a,b). As cleavage continues, these partitioning membranes extend progressively, most likely through addition of further membrane material from Golgi vesicles. Concurrently, the flagellar axonemes assemble within the plasma membrane-bound compartment and the mastigonemes become arranged in two rows along the anterior flagellum (Cope and Hardham, 1994). Immunolabelling and ultrastructural studies reveal that during cleavage the initially randomly distributed, large peripheral, dorsal and ventral vesicles (Fig. 4G) become localized to the dorsal or ventral surface of the future zoospores (Fig. 4K) (Hyde and Hardham, 1993). Movement of these vesicles to the proximity of the partitioning membranes occurs independently of cytoskeletal elements, but the development of polarity in vesicle distribution within the zoospore cortex is dependent on an intact microtubule array (Hyde and Hardham, 1993). Both movement and polarization of mitochondria require microtubules. During assembly of the partitioning membranes, arrays of actin microfilaments become associated with the zoospore surface (Fig. 4J) (Jackson and Hardham, 1998) and peripheral cisternae become positioned beneath the plasma membrane throughout the zoospore except within the groove (Fig. 4H). Towards the end of cytokinesis, the water expulsion vacuole forms and begins to function, pumping water out of the cytoplasm (Fig. 4H).
P. cinnamomi zoospores are released from the sporangium as the wall material at the apex of the sporangium suddenly expands to form an evanescent vesicle. Usually, the majority of zoospores are rapidly squeezed out of the sporangium, a process believed to be driven by the build up of hydrostatic pressure within the sporangium (Gisi et al., 1979; Gisi, 1983). This hydrostatic pressure could result from the accumulation of gel-like material or extracellular solutes in the space between the zoospores and the sporangial wall (Gisi and Zentmyer, 1980; Gisi, 1983; MacDonald and Duniway, 1978; Money and Webster, 1985; Money et al., 1988). In either case, the sporangial cell wall must act as a semi-permeable barrier that allows the influx of water, but does not permit efflux of the gel or solutes from the periplasm. The accumulation of osmotically active molecules within the periplasm could lead to dehydration of the zoospore cytosol. This hyper-osmotic extracellular sporangial lumen may be counterbalanced by the generation of high concentrations of proline within the zoospore cytosol (Ambikapathy et al., 2002; Grenville-Briggs et al., 2005). It is likely that the proline must be rapidly expelled after zoospore release into the hypo-osmotic solution in the environment. The consequent need to re-establish stores of proline in zoospores is consistent with the observation of enhanced levels of expression of genes encoding proline biosynthetic enzymes in the zoospores (Ambikapathy et al., 2002). Although zoosporogenesis is rapid, changes in gene expression during this process have recently been documented in P. infestans (Tani et al., 2004) and are likely to occur also in P. cinnamomi.
Zoospore motility, encystment and adhesion
Motile zoospores (Fig. 5A,C) released from the sporangia are the key to P. cinnamomi disease establishment and dissemination. The main functions of the zoospores are to move in a directed manner towards favourable infection sites and to become firmly attached at those locations on the plant surface. Research over the last 15 years has helped generate a much better understanding of the molecular and cellular basis of both these functions.
Zoospore motility is achieved through the action of the two flagella that emerge from basal bodies underlying the centre of the longitudinal ventral groove (Fig. 5A). Single zoospores may cover distances of several centimetres, at speeds of approximately 200 µm/s. The internal structure of Phytophthora flagella is typical of other eukaryotic flagella and, undoubtedly, so is their mode of action. The zoospores are described as being heterokont because the two flagella are morphologically different. The anterior flagellum is shorter than the posterior flagellum and possesses two parallel rows of mastigonemes along its length (Fig. 5B,D). The mastigonemes, the defining character of the Stramenopiles (meaning straw-like hairs), are tripartite with a basal section, a tubular shaft about 1 µm in length and a tuft of filamentous terminal hairs (Hardham, 1987). Both flagella propagate sinusoidal waves that commence at the base and propagate to the tip. The mastigonemes on the anterior flagellum reverse the thrust of flagellar beat and pull the zoospore forward (Cahill et al., 1996). The posterior flagellum acts like a rudder to turn the spore periodically.
The ability of Phytophthora zoospores to respond to chemical and electrical gradients (Cahill and Hardham, 1994b; Gow et al., 1992; Morris and Gow, 1993; Morris and Ward, 1992; Morris et al., 1992) greatly enhances their chances of successfully reaching a favourable infection site, a factor that may be especially important for root pathogens such as P. cinnamomi. As yet, we have little information on the identity of zoospore proteins involved in the reception of chemotactic or electrotactic signals; however, a recent study of Phytophthora G-proteins has elegantly demonstrated a role for a G-protein α-subunit in motility and chemotaxis of P. infestans zoospores (Latijnhouwers et al., 2004). Consistent with the likelihood that regulation of flagellar activity also involves variations in Ca2+ concentration is the immunolocalization of two calcium-binding proteins, calmodulin and centrin, within the flagella apparatus of P. cinnamomi zoospores centrin (Gubler et al., 1990; Harper et al., 1995) (Fig. 5E). Genes encoding centrin, a dynein light chain protein and a radial spoke protein have recently been cloned from P. cinnamomi and are currently being further characterized (R. Narayan, L. M. Blackman and A. R. Hardham, unpublished results).
The behaviour of P. cinnamomi zoospores and the identity and fate of vesicles in their cortical cytoplasm have been revealed through immunocytochemical studies of the three types of peripheral vesicles (Fig. 5F–J). At the root surface, P. cinnamomi zoospores orientate so that their ventral surface faces the root before they encyst (Fig. 6A–D) (Hardham and Gubler, 1990). Within the first minute of encystment, the peripheral cisternae vesiculate and apparently fuse with the zoospore plasma membrane, thus potentially effecting a rapid and extensive change in spore plasma membrane properties (Hardham, 1989). Within the first 2 min, the contents of the dorsal vesicles are secreted to form a mucilage-like coating over the surface of the cysts, which may protect the cells from desiccation (Fig. 6C,E) (Gubler and Hardham, 1988). At the same time, the contents of the ventral vesicles are also secreted to a form a pad of material between the cyst and the root surface (Fig. 6D) (Hardham and Gubler, 1990). Measurement of the adhesiveness of the spores during encystment indicates that the cells become sticky about 2 min after the induction of encystment (Gubler et al., 1989). There is thus a good correlation between the spatial and temporal characteristics of ventral vesicle secretion and adhesive pad formation. By 5–10 min after the induction of encystment, the cysts have deposited a cell wall that is strong enough to allow turgor pressure to build up in the cells. The large peripheral vesicles move from their cortical location in the zoospores and become randomly distributed throughout the cyst cytoplasm (Fig. 6E) (Gubler and Hardham, 1988).
The material within and secreted from the ventral vesicles is labelled by Vsv antibodies, which in immunoblots react with a P. cinnamomi protein of approximately 220 kDa relative molecular weight (Hardham and Gubler, 1990). Immunoscreening of a P. cinnamomi cDNA library made from mRNA isolated from sporulating hyphae with Vsv antibodies has led to the identification of a gene, designated PcVsv1, that encodes the Vsv protein (Robold and Hardham, 2005). The PcVsv1 gene is 7356 nucleotides in length, contains no introns and encodes an inferred protein of 2452 amino acids with a molecular mass of 262 kDa and a pI of 5.52. Apart from short N- and C-terminal sequences, the bulk of the PcVsv1 protein is composed of 47 copies of a domain approximately 50 amino acids in length that shows homology to thrombospondin type 1 repeats (TSR1s) found in a large number of adhesive extracellular matrix proteins in animals (Adams and Tucker, 2000) and secreted adhesins in apicomplexan malarial parasites (Tomley and Soldati, 2001).
PcVsv1 expression is up-regulated after the induction of sporulation, consistent with the appearance of ventral vesicles at this time. Homologues of PcVsv1 can be found in P. sojae, P. ramorum, P. infestans and P. nicotianae (Robold and Hardham, 2005). Immunolabelling with Vsv antibodies shows that the Vsv adhesin also occurs in Pythium, Albugo and Plasmopara, suggesting that it is widespread across the Oomycetes.
Outside the Oomycetes, the closest homologue of PcVsv1 in GenBank is currently a TSR1-containing protein from the apicomplexan malarial parasite, Cryptosporidium parvum (AA039046) (Deng et al., 2002). This homology between the P. cinnamomi and apicomplexan adhesins draws attention to other similarities between Phytophthora zoospores and apicomplexan zoites. As discussed above, there is now good molecular confirmation of the phylogenetic affinities between the Stramenopiles and the Alveolates. Characterization of the P. cinnamomi TSR1-containing adhesin highlights common features in the infection strategies of these two parasitic groups. In both, onset of infection is marked by rapid, regulated secretion of adhesive-containing vesicles from a localized region of the pathogen aligned to face the host (Hardham and Gubler, 1990; Joiner and Roos, 2002). In P. cinnamomi, adhesive-containing ventral vesicles are confined to the ventral surface (Hardham and Gubler, 1990); in apicomplexans, adhesin-containing micronemes are part of the apical complex that gives rise to the name of the group (Tomley and Soldati, 2001). Remarkably, the cortical region at which secretion occurs is, in both cases, free of an underlying system of flattened membranes—the taxon-defining alveoli in the apicomplexans (typically called the inner membrane complex), and the zoospore peripheral cisternae in Phytophthora and related Oomycetes.
Cyst germination and plant penetration
Within 20–30 min after encystment, P. cinnamomi cysts germinate. The germ tube emerges from the centre of what was the zoospore ventral surface and grows towards, into or along the root surface (Fig. 7). The germ tubes usually grow and penetrate the root surface along the anticlinal wall between the epidermal cells but they are also capable of growing directly through the outer periclinal wall (Hardham, 2001). An appressorium-like swelling of the germ tube is sometimes formed, especially during penetration of the periclinal wall (Fig. 7). Roots of susceptible plants are rapidly colonized and sporangia may appear on the root surface within 2–3 days in susceptible plants. Sporulation is preceded in planta by the production of peripheral vesicles (Chambers et al., 1995).
Host penetration by Phytophthora, as is the case for other plant pathogens, involves the production and secretion of a range of enzymes that digest and degrade the underlying plant cell wall. Some of the first cell-wall-degrading enzymes to be secreted during infection are those that break down pectin (Cooper, 1983). These enzymes include the polygalacturonases which, in P. cinnamomi, are encoded by a large multigene family consisting of more than 20 genes (Götesson et al., 2002). The P. cinnamomi polygalacturonases differ in the number of potential glycosylation sites and in the structure of the N- and C-termini. Analysis of the polygalacturonase gene sequences revealed that the P. cinnamomi enzymes are more similar to fungal polygalacturonases than they are to plant polygalacturonases.
Traditionally, identification of Phytophthora species has been based on morphological and cultural criteria, and has required considerable experience and expertise in order to differentiate different species reliably (Waterhouse et al., 1983). Phytophthora pathogens in infected soil or plant samples are usually first captured (and thus concentrated) by baiting and then cultured on media containing a range of fungal and bacterial inhibitors that favour the growth of Phytophthora species (Eden et al., 2000; Greenhalgh, 1978; Tsao, 1983). Species identification is then made through observations of various morphological characteristics, in particular structural aspects of sporangia, antheridia, oogonia and mycelia (Waterhouse et al., 1983). Problems with this approach include failure to isolate the pathogen from the infected sample (Eden et al., 2000; Hüberli et al., 2000) and wide variations in morphological characters in different isolates of a single species or under different growth conditions (Daniel et al., 2003; Waterhouse et al., 1983). The need for alternative methods of diagnosis has long been recognized. One alternative approach investigated for P. cinnamomi compared electrophoretic patterns of isozymes isolated from cultured samples (Old et al., 1984; Oudemans and Coffey, 1991). These studies showed that intraspecies variations in the patterns of certain enzymes were low while interspecies differences were large and gave good separation of P. cinnamomi from other Phytophthora species. This method is useful for P. cinnamomi identification from pure cultures but is not applicable to situations in which more than one organism is present, such as in infected plant or soil samples.
After the introduction of antibodies as research tools, Phytophthora researchers were quick to investigate the potential use of antibodies in the detection and diagnosis of Phytophthora species. A number of studies of P. cinnamomi, like those of other Phytophthora species, raised polyclonal antibodies against mycelial extracts, but they failed to give species-specific reactions (Ferraris et al., 2004; Hahn and Werres, 1997; MacDonald and Duniway, 1979; Malajczuk et al., 1975). However, in the mid-1980s, two groups raised monoclonal antibodies against Phytophthora components. In one case, the antibodies formed the basis of detection kits that were produced commercially by Agri-Diagnostics Associates (Cinnaminson, NJ) using both membrane-based and ELISA formats. However, because the Agri-Diagnotics kits either did not react with P. cinnamomi or were not specific for individual Phytophthora species or for the genus, their usefulness has been limited (Ali-Shtayeh et al., 1991; Benson, 1991; Pscheidt et al., 1992; Werres et al., 1997). In the second case, monoclonal antibodies were raised against P. cinnamomi spore components, and antibodies were identified that were specific for selected P. cinnamomi isolates, for the species P. cinnamomi or for the genus (Hardham et al., 1986). These antibodies subsequently formed the basis of a series of diagnostic tests for use in P. cinnamomi identification, including immunofluorescence assays, ELISAs and a dipstick assay (Fig. 8A–C) (Cahill and Hardham, 1994a,b; Gabor et al., 1993).
The dipstick assay, in particular, is a simple and easy test for P. cinnamomi which may be performed in the field with infected soil or plant samples (Cahill and Hardham, 1994a). As in traditional baiting techniques, the assay begins with incubation of soil or plant samples in water. Zoospores released from sporangia in the sample are negatively geotropic and swim to the surface where they concentrate and are attracted via chemotaxis or electrotaxis to the dipsticks floating on the water surface. For P. cinnamomi, dipsticks can be infused with chemoattractants such as glutamate or aspartate or can carry a positive charge (Cahill and Hardham, 1994b). At the dipstick, the zoospores encyst and glue themselves to the membrane. The dipstick and adherent spores is then incubated with the P. cinnamomi-specific monoclonal antibody that reacts with the cyst surface. Labelling can be visualized in a number of ways, the simplest being through deposition of an insoluble coloured product by an enyzme-conjugated probe (Fig. 8A,B). Labelled in this way, the presence of P. cinnamomi spores can be determined using a low-magnification hand lens or magnifer. Alternatively, the membranes can be processed through an ELISA and assessed using a microtitre plate reader as described for P. nicotianae (Gautam et al., 1999).
The other modern approach to the identification and diagnosis of P. cinnamomi and other Phytophthora species exploits differences in nucleic acid sequences between Phytophthora species. Nucleic acid probes based on species-specific sequences within the ITS regions between rDNA genes have been used to hybridize specifically to P. cinnamomi DNA that has been amplified by the polymerase chain reaction (PCR) from pure cultures or from mixed samples (Bailey et al., 2002; Lee et al., 1993). Alternatively, PCR primers have been designed from the rDNA ITS region and used to amplify Phytophthora DNA. The primers may be specific for P. cinnamomi or other Phytophthora species (Drenth and Irwin, 2001; Lévesque et al., 1998) or they may amplify DNA from a number of Phytophthora species, and subsequent species discrimination may be achieved by analysis of fragments produced by digestion of the PCR products with various restriction enzymes (Drenth et al., 2005; Ristaino et al., 1998). The primers designed by Drenth et al. (2005) have been used in the development of commercial diagnostic assays with components available from C-Quentec Diagnostics Pty Ltd (Epping, NSW, Australia).
Recently, variations within the ITS regions of rDNA genes have been used successfully to differentiate species of Phytophthora, including P. cinnamomi, using the technique of single-strand conformation polymorphism (SSCP) (Kong et al., 2003a). This technique is based on the fact that differences (as little as a single base) in the sequence of single-stranded DNA molecules of similar length cause the DNA molecules to fold into three-dimensional shapes that display differences in electrophoretic mobility (Sambrook and Russell, 2001). PCR primers were used to amplify DNA from 282 isolates of 29 Phytophthora species; the PCR products were then analysed by SSCP to identify all isolates correctly. Species-specific DNA probes or primers derived from the P. cinnamomi elicitin (cinnamomin) gene, from the Lpv putative storage protein genes or from unidentified regions of the genome have also been used to identify P. cinnamomi (Coelho et al., 1997; Dobrowolski and O’Brien, 1993; Judelson and Messenger-Routh, 1996; Kong et al., 2003b).
P. CINNAMOMI EPIDEMIOLOGY
Between the first description of P. cinnamomi in 1922 and Zentmyer's monograph on P. cinnamomi in 1980, it became clear that P. cinnamomi was a serious threat to a wide range of plant species throughout the world. Zentmyer (1980) listed approximately 950 host species including economically important crop plants such as avocado, pineapple, peach, chestnut and macadamia; horticultural plants such as rhododendron and camellia; and forest trees such as oak, pine and eucalyptus. In Australia, P. cinnamomi was a problem not only in agriculture and horticulture but also caused widespread damage in natural ecosystems in south-western Western Australia (Fig. 8D–F), Tasmania and Victoria (Fig. 1). In Western Australia, P. cinnamomi was introduced inadvertently in contaminated nursery stock in Perth in the 1920s from where it escaped into the jarrah (Eucalyptus marginata) forests and heathlands to the south. The disease it causes, Jarrah Dieback, spread rapidly, but it was not until after the causal agent, P. cinnamomi, was identified in the 1960s that it was realized that dispersal was being facilitated by the construction of roads throughout the forests (Podger et al., 1965; Podger, 1972). Many of the understorey species proved to be even more susceptible than the eucalypts (Fig. 8E,F) and the disease advanced at an alarming rate and with devastating results. Similar scenarios developed in a number of forests in Victoria (Weste, 1974; Weste and Kennedy, 1997).
The severity of P. cinnamomi epidemics in natural ecosystems in Australia has not abated in the intervening period since Zentmyer's publication (1980). Indeed, continued research has revealed that the threat is, if anything, worse than initially realized. Studies in Western Australia, Victoria and Tasmania have led to a dramatic increase in the number of Australian native species known to be susceptible to P. cinnamomi. In the Stirling Range National Park in Western Australia (Fig. 8D), 36% of the 330 species assessed were found to have individuals killed by P. cinnamomi (Wills, 1992). Several plant families had many susceptible species, with a notable example being the Proteaceae in which 85% of the species were susceptible. Extrapolation of the results of this study suggest that as many as 2000 species in the park are likely to be at risk. This means that once P. cinnamomi gets into an area, it can kill a large proportion of the plants present, leading to a drastic change in the floristic composition of the area as susceptible species are replaced by resistant plants, often herbacious perennials, rushes, sedges or introduced weeds (Fig. 8D). A recent assessment of the data collected by this and three other studies of the susceptibility of native flora of south-west Western Australia has concluded that all four studies arrive at similar figures and the estimates of the proportion of species that are susceptible to P. cinnamomi are likely to be realistic (Shearer et al., 2004a). Mean values of 40% susceptible and 14% highly susceptible equate to 2284 and 800 species, respectively, of the 5710 species of plants described from the study areas. Similarly high percentages of susceptible plants have been documented in forest, woodland and heathland communities in Victoria (Weste, 2003) and in moorland in Tasmania (Brown et al., 2002).
A key factor associated with the onset of P. cinnamomi epidemics in these areas is the occurrence of conditions that support asexual sporulation and zoospore production. In the Stirling Ranges, for example, there had been heavy rain (Wills, 1992). In Tasmania, loss of vegetation due to fire or clearing allowed soil temperatures to rise by a few degrees so that they fell within the range that allowed sporulation (Podger and Brown, 1989; Podger et al., 1990). In the context of this association of conditions favouring zoospore production and disease epidemics, it is of interest to note the current situation in three epidemic sites in Victoria, namely Wilsons Promontory, the Grampians and the Brisbane Ranges that have now been monitored for 30 years (Weste, 2003). The most recent surveys of these regions revealed evidence of regeneration of some susceptible species that had disappeared from the assessment sites. In these latest data, in quadrats at 10 of the 13 sites across the three parks, P. cinnamomi was absent or rare and 30–40 susceptible species had reappeared. It remains to be determined if this recovery is stable or not, but it is noteworthy that this regeneration has taken place after the five driest consecutive years on record, a feature that is consistent with the association of conditions that support zoospore production and the existence of disease epidemics.
The important role of environmental conditions such as rainfall and temperature on the development of P. cinnamomi diseases has also been investigated with respect to possible effects of global climate change on the prevalence of diseases caused by P. cinnamomi. Analysis of scenarios in which a 20% increase in rainfall is coupled with three temperature change regimes has led to a prediction that global warming could result in P. cinnamomi diseases of oak becoming more severe in regions of Europe where the pathogen is already present and to expansion of the disease to the north and east (Brasier & Scott 1994, cited in Chakraborty et al., 1998). The potential range of expansion could be up to a few hundred kilometres east from the Atlantic coast within the next hundred years (Bergot et al., 2004).
Population studies of P. cinnamomi indicated that there are three clonal lineages in Australia, South Africa and in many areas around the world (Dobrowolski et al., 2003; Linde et al., 1999). These results indicate a lack of frequent, if any, sexual reproduction in P. cinnamomi populations even though both A1 and A2 mating types are present, except in its region of likely origin in Papua New Guinea. Considerable variation can, however, occur within a single clonal lineage and there is evidence that this variation arises through mitotic recombination during asexual growth and development (Dobrowolski et al., 2003; Hüberli et al., 2001). Thus, despite the general absence of crossing between mating types in the field, P. cinnamomi is fully capable of adapting to new environmental conditions and of developing virulence on new hosts during asexual growth.
Losses due to P. cinnamomi diseases in agriculture, horticulture or forestry can be substantial and can be calculated in economic terms by considering such factors as expenditure on control measures, loss of yield, and reduction of product quality or value. Estimates of losses due to P. cinnamomi diseases include AU$1 million in pineapple crops in a wet year in Queensland (Zentmyer, 1980), US$30 million annually in avocado groves in California (Erwin and Ribeiro, 1996) and US$5 million annually in pine plantations in forests from Virginia to Mississippi (Zentmyer, 1980). Estimates of losses in natural ecosystems and conservation areas are also substantial but more difficult to evaluate in monetary terms. As well as any expenditure on control measures, P. cinnamomi diseases in native forests and heathlands in Australia, for example, result in extensive environmental degradation. P. cinnamomi has caused significant changes in plant community structure and floral composition, reduced plant cover and biomass, reduced biodiversity, threatened endangered plant species and put dependent biota at risk. The impact of P. cinnamomi in natural ecosystems in Australia has been recognized by inclusion of P. cinnamomi as a ‘Key Threatening Process’ in the Commonwealth Environmental Protection & Biodiversity Conservation Act 1999.
CONTROL OF P. CINNAMOMI
The evolutionary distance of Phytophthora and other Oomycetes from the true fungi has major ramifications for the effectiveness of traditional fungicides in controlling Phytophthora diseases. It means that chemicals that are active inhibitors of fungi are usually not inhibitors of Oomycetes. Control of Phytophthora diseases has thus called for novel approaches. Root pathogens, such as P. cinnamomi with its ability to survive as chlamydospores in soil and in roots of symptomless plants, impose additional demands on inhibitory chemicals.
A number of classes of compounds have been found to inhibit Phytophthora growth and plant infection (Erwin and Ribeiro, 1996; Griffith et al., 1992; Schwinn and Staub, 1995). Two groups of chemicals that have proved to be the most effective for P. cinnamomi are the phenylamides (e.g. metalaxyl) and the phosphonates (e.g. fosteyl-Al). The phosphonates are salts or esters of phosphonic acid (Guest and Grant, 1991). Because the term phosphonate is also applied to compounds containing a carbon-phosphorus bond, in much of the recent literature the phosphonic acid salts used for Phytophthora control are referred to as phosphite (Hardy et al., 2001). Metalaxyl and phosphite are both systemic inhibitors, with metalaxyl being translocated in the xylem, and phosphite in the xylem and the phloem (Guest and Grant, 1991). Both treatments give effective control in agricultural and horticultural situations; phosphite has also been extensively tested in natural ecosystems (Hardy et al., 2001).
The use of phosphite has dominated studies of chemical control of P. cinnamomi diseases over the last 10–15 years, which much of the research being conducted in Western Australia where P. cinnamomi causes such widespread devastation. Phosphite's mode of action is still uncertain but it appears to be a combination of direct inhibition of pathogen growth and stimulation of the plant defence response, possibly via an increase in the production of pathogen-derived elicitors or a decrease in the production of pathogen-derived suppressors of the defence response (Guest and Grant, 1991; Jackson et al., 2000). In one study, phosphite was shown to reduce the production of zoospores in infected plant material (Wilkinson et al., 2001a), a factor that would help limit the amplification of inoculum even if it did not eradicate the pathogen from an infected area. Given the extensive knowledge of the biology of sporulation in P. cinnamomi, more detailed investigations of the effect of phosphite on this process may be illuminating and could help elucidate phosphite's mode of action.
Phosphite is a valuable inhibitor, but its effectiveness varies with different P. cinnamomi isolates (Hardy et al., 2001; Wilkinson et al., 2001b) and environmental conditions, such as the phosphorus levels in the soil (Guest and Grant, 1991). Although phosphite generally displays low phytotoxicity, recent research has shown that in some plants phosphite can be responsible for foliar damage, reduction of pollen viability and pollen tube growth, an increase in the frequency of abnormal mitotic and meiotic cell divisions and a decrease in root growth (Barrett et al., 2002; Fairbanks et al., 2001, 2002; Nartvaranant et al., 2004). Dosage rates and timing of application need to be monitored and adjusted to minimize phytotoxicity in sensitive plants. Comparisons have been made between different modes of application of phosphite, in particular between foliar sprays and trunk injections (Hardy et al., 2001; Shearer et al., 2004b). In Western Australia, ultra-low-volume mist-spraying from the air has been effective in plant communities of high conservation value (Hardy et al., 2001).
The impact of P. cinnamomi on crops, horticultural plants and natural ecosystems is likely to ensure that it remains an important species for research. How is P. cinnamomi able to evade or overcome the defence responses of so many plant species? Why are such large numbers of Australian native plants susceptible to P. cinnamomi? How quickly can P. cinnamomi adapt to changing environmental conditions and exploit new niches? Recognition of the important role of asexual sporulation and motile zoospores in the infection process means that the detailed understanding of fundamental aspects of zoospore cell biology will provide a valuable foundation for further studies that explore the molecular basis of P. cinnamomi pathogenicity.
As yet, a reliable method for the transformation of P. cinnamomi has not been established, a factor that will hamper direct molecular characterization of the function of P. cinnamomi genes until a suitable protocol is developed. However, in the meantime, studies of other Phytophthora species are likely to yield valuable information on many aspects of the P. cinnamomi infection process. The extensive synteny that has become evident between Phytophthora species following the release of sequence information from P. sojae, P. ramorum (http://genome.jgi-psf.org/physo00.info.html; http://phytophthora.vbi.vt.edu) and P. infestans (Randall et al., 2005) will aid the application of results from one species to another. It is likely that many of the key features of Phytophthora pathogenicity are conserved between species, and principles elucidated in other Phytophthora species may be extrapolated to P. cinnamomi.
I would like to thank all colleagues with whom I have worked on Phytophthora cinnamomi, and especially Drs D. M. Cahill, F. Gubler, S. L. Jackson and H. J. Mitchell for allowing me to reproduce their micrographs. Electron microscopy was conducted in the ANU Electron Microscopy Unit. My thanks are extended to the EMU staff for their assistance, and to C. Eadie and S. Wragg for preparing Fig. 2.