Although plant pathogens are often regarded as plant destroyers, many deploy a genetic program to maintain viability of their plant hosts during at least part of the infection cycle. This lifestyle is referred to as biotrophy (Spanu, 2012). The other end of the lifestyle spectrum is defined by pathogens that kill plant tissue (necrotrophy; Oliver & Ipcho, 2004). Some ‘hemi-biotrophic’ pathogens deploy both strategies sequentially, beginning with a biotrophic phase that enables the pathogen to secure a foothold in the plant host, followed by a transition to necrotrophy that fuels rapid growth and reproduction (Oliver & Ipcho, 2004). In this issue of New Phytologist, Gan et al. (pp. 1236–1249) describe genomic and transcriptomic comparisons of two species from the hemi-biotrophic, fungal genus Colletotrichum. This study, together with a similar comparison of two additional Colletotrichum species by O'Connell et al. (2012), provides the most complete picture to date of the transcriptomic underpinnings of infection stage shifts during hemi-biotrophic pathogenesis. These studies also delineate intriguing diversity in the pathogenicity gene arsenals deployed by Colletotrichum species, compared with other fungal phytopathogens, while comparisons between the different Colletotrichum species uncover leads towards understanding the molecular basis of host-specific adaptation.
‘These studies also delineate intriguing diversity in the pathogenicity gene arsenals deployed by Colletotrichum species …’
Colletotrichum fungi can be cultured apart from the host and are amenable to genetic manipulation, thus they represent tractable models for understanding the hemi-biotrophic stage shift and other aspects of fungal pathogenicity (Munch et al., 2008). The Colletotrichum genus contains over 600 species that collectively parasitize a wide range of plants and cause several important crop diseases. Gan et al. focused their analysis on C. gloeosporioides and C. orbiculare. The host range of different C. gloeosporioides strains encompasses hundreds of plants; the strain sequenced in this study causes a postharvest disease of strawberry. Colletotrichum orbiculare causes anthracnose in cucurbits. The strain sequenced in this study can also infect Nicotiana benthamiana and is therefore being developed as a laboratory model. The centerpieces of O'Connell et al. (2012) are strains of C. higginsianum and C. graminicola that, respectively, infect Arabidopsis and maize.
Pathogenicity gene inventories highlight differences and similarities between Colletotrichum and other fungal genera
The inventories of pathogenicity genes in the four Colletotrichum species revealed several interesting differences compared with other fungal phytopathogens. To begin with, Colletotrichum genomes maintain the largest degradome (proteases and carbohydrate-active enzymes) of any fungus. Subtilisins are the largest class of proteases, and several of these appear to have been acquired by horizontal transfer from plants. Chitin-binding LysM proteins are also expanded, most likely to interfere with perception by the host of chitin fragments that would otherwise trigger immunity (Thomma & de Jonge, 2009). Compared with other fungal pathogens, Colletotrichum genomes also contain a larger inventory of genes for synthesis of secondary metabolites such as polyketides and non-ribosomal peptides (Mobius & Hertweck, 2009), which have previously been documented as phytotoxins (but see later).
Effector biology is currently a very active topic in molecular plant–microbe research (de Jonge et al., 2011). Similar to other fungi that employ biotrophy (Spanu, 2012), Colletotrichum genomes contain hundreds of genes encoding small, secreted proteins (SSPs) that represent candidate effectors. These genes are distributed randomly in the genome and are not associated with repetitive DNA, in contrast to effector genes of oomycete plant pathogens (Raffaele & Kamoun, 2012). The majority of Colletotrichum SSP genes encode unique proteins, but there are homologs of several experimentally validated effectors from other fungi. The conservation of these genes implies broadly important roles in fungal virulence; it will be of great interest to compare the function of these homologs in their respective pathosystems.
Transcriptomic differences between infection stages
The genome analyses in Gan et al. and O'Connell et al. (2012) were complemented nicely by experiments that compared transcriptomes from the biotrophic and necrotrophic growth stages with each other and with in vitro grown material. Gan et al. used a custom microarray to assay material from infection of N. benthamiana, while O'Connell et al. (2012) utilized RNAseq with material from an Arabidopsis infection. Both studies revealed dynamic and pervasive differences in material grown in planta compared with in vitro. For example, O'Connell et al. (2012) examined transcriptomes in appressorial cells (attachment to host, Fig. 1) isolated from the plant, compared with morphologically identical appressorial cells grown in vitro. They documented over 1500 genes induced specifically by contact with the host. The authors interpreted this global difference as evidence for a state of heightened sensory awareness in which the appressorial cells are responsive to cues from the host. Interestingly, a separate study demonstrated that appressoria could secrete effectors, underscoring the functional complexity of this specialized cell type (Kleemann et al., 2012).
Not surprisingly, the transcript profiles of biotrophic and necrotrophic growth stages were quite different from each other. The biotrophic phase was distinguished by induction of a large number of SSPs. This characteristic mirrors other biotrophic pathosystems and likely reflects the need to reprogram the host (e.g. suppression of immune responses). More surprisingly, the majority of secondary metabolite gene clusters were also induced at early stages. This is intriguing because many of the best-characterized fungal secondary metabolites function as toxins and have been associated with necrotrophy (Mobius & Hertweck, 2009). In apparent contrast, the Colletotrichum metabolites might function as nondestructive, small molecule effectors. Genes for biosynthesis and transport of secondary metabolites are also induced during early infection by the rice blast pathogen Magnaporthe oryzae (Soanes et al., 2012), highlighting the potentially broad importance of this understudied aspect of fungal biotrophy.
Gan et al. reported that genes for uptake and metabolism of quinic acid were induced during early invasion. In plants, quinic acid is a metabolic precursor of key immune system molecules such as salicylic acid. Thus, it is conceivable that these fungi could siphon quinate from the plant to utilize as a carbon source and to interfere with immune activation. Genes for quinate metabolism are also induced in appressorial cells of Magnaporthe oryzae, suggestive of a conserved strategy that should be prioritized for further investigation (Soanes et al., 2012). Gan et al. also noted a gene encoding a secreted isochorismatase, which could further retard salicylate biosynthesis. This strategy also appears to be employed by M. oryzae (Soanes et al., 2012).
The shift from biotrophy to necrotrophy in Colletotrichum is defined by induction of the degradome, mirroring other necrotrophic pathosystems. Another notable pattern is large-scale induction of genes encoding putative transporters for uptake of nutrients, consistent with the rapid growth of secondary hyphae during the necrotrophic phase.
Thematic variations amongst different Colletotrichum species
In addition to the aforementioned differences in pathogenicity gene inventory and expression between Colletotrichum and other fungal genera, both studies revealed intriguing differences between the four Colletotrichum species that could be related to host specialization. For example, the C. graminicola genome contained expanded inventories, compared with the other species, of several classes of pathogenicity genes, perhaps to support its relatively broad host range. The four species exhibited substantial variation in the content and/or expression of secondary metabolism gene clusters and SSPs, perhaps reflecting tailoring for specific hosts to meet the imperative of host-cell manipulation while avoiding effector-triggered immunity during the biotrophic phase. Host specialization during the necrotrophic phase was also evident. For example, C. graminicola deployed a different inventory of CAZymes compared with the other three: C. graminicola induces a larger number of genes encoding enzymes that facilitate degradation of hemi-cellulose that is predominant in monocot cell walls. Conversely, the other three species maintain a larger complement of pectinases, to degrade the pectin that is predominant in dicot cell walls.
Altogether, Gan, O'Connell, and their colleagues paint the most complete picture yet of fungal hemi-biotrophic infection, which is summarized in Fig. 1. These studies reinforce emerging generalities that apply broadly to fungi and oomycetes: the biotrophic lifestyle is underpinned by secretion of effectors that manipulate the host while preserving cell viability. Contrastingly, the necrotrophic phase is geared for plant-cell destruction, via hydrolytic enzymes and toxins. At the genome level, obligate biotrophic pathogens contain relatively large complements of effector-encoding genes and a minimal arsenal of degradative enzymes, with the converse being true for dedicated necrotrophs. Contrastingly, each of these classes of genes is well represented in the genomes of hemi-biotrophic pathogens, qualifying them as perhaps the best-armed class of phytopathogens.
These studies also illustrate that there is still a lot to be learned from simple genomic and transcriptomic comparisons, particularly those that delve deeply at the intra-generic levels. At the very least, such comparisons are invaluable for prioritizing the most potentially fruitful genes for detailed functional characterization. In the case of Colletotrichum, these likely include the CAZymes and secondary metabolite gene clusters that could respectively be exploited for biofuels or other, novel biological activities that could have impact beyond plant–pathogen interactions.