Powdery mildew genomes in the crosshairs

Authors


2nd International Powdery Mildew Workshop and 3rd New Phytologist Workshop, in Zürich, Switzerland, February 2012

Powdery mildews are fungal phytopathogens that cause the so-called powdery mildew disease on many plant species (Bélanger et al., 2002). The disease is prevalent on cereals such as wheat and barley (Fig. 1), but also on fruit (e.g. grapevine, apple trees, strawberries), vegetable crops (e.g. tomato, cucurbits) and ornamentals such as roses (Glawe, 2008). The powdery mildew fungi belong to the ascomycetes and are obligate biotrophs, that is, their growth and reproduction is fully dependent on living hosts. As a consequence, powdery mildews cannot be cultured in vitro, are difficult to cryo-conserve, and a reliable transformation protocol has not been established as yet, posing serious challenges for using them as experimental systems. Recently, the genomes of three powdery mildew species have been sequenced, and it appears that the genomes bear hallmarks of the obligate biotrophic lifestyle. Essentially, powdery mildew genomes are very large (> 120 Mb) and full of retrotransposons, potentially allowing for high genomic flexibility. Notably they lack a considerable number of genes otherwise present in ascomycetes, which may explain why powdery mildews rely on living host plants for propagation. The genomes code for a significant number of species-specific candidate secreted effector proteins (CSEPs), which presumably represent the weapons of mildews for pathogenesis (Spanu et al., 2010). Despite this recent breakthrough, many questions in powdery mildew biology and genomics remain open. For example, it would be interesting to learn whether and how the repertoires of CSEPs differ in various powdery mildew isolates, and whether there are any genomic adaptations found in strains that colonize cultivated host plants compared to isolates found on wild species. Also, how do the CSEPs contribute to the establishment of the biotrophic life-style? Are there any CSEPs that function as avirulence (Avr) proteins, that is, they can be recognized by cognate host resistance proteins? Besides the CSEPs, a large number of genes with as yet unknown functions suggest that the powdery mildew genomes harbor a plethora of evolutionary innovations, and it will be one of the future challenges to untangle these novel protein functions.

Figure 1.

A single powdery mildew (Bgh) colony on a barley leaf. Note the conidiophores (spore carriers) that emerge vertically from the epiphytic mycelium.

‘One further contentious point raised was whether standard common strains and varieties should be employed, in order to render experimentation comparable across laboratories world-wide.’

The powdery mildew workshops

The 1st International Powdery Mildew Workshop, held last year at the Max-Planck Institute for Plant Breeding Research (MPIPZ) in Cologne (Germany), was the direct continuation of the very successful series of six-monthly consortium meetings many of us had been involved in during the coordination of the so-called BluGen (Blumeria graminis f. sp. hordei (Bgh) genome sequencing) project (Spanu et al., 2010). Initially there was a need to continue working together on aspects of the Bgh genome (e.g. further annotation, assembly issues, work on repetitive sequences, etc.). The 2nd International Powdery Mildew Workshop was held at Zürich Botanical Institute, high on a hill overlooking the Golden Coast of the ‘Züri-see’ with the postcard backdrop of snow-clad Alps emerging mystically on the horizon. This workshop was kindly supported by the New Phytologist Trust and local organization was conducted by Beat Keller and his co-workers. Thirty-four participants from 11 countries met for two days to present and discuss, sometimes very vigorously, details of current advances in genome-led biology of powdery mildews.

The wheat powdery mildew genome – key to understanding the forma specialis conundrum?

The main efforts in genomics are currently spent on the wheat powdery mildew pathogen, Blumeria graminis f. sp. tritici (Bgt), which belongs to the same genus and species as Bgh, yet the two mildews have a strict host specialization on either wheat (Bgt) or barley (Bgh). The separation into different ‘formae speciales’ of B. graminis raises the question what the molecular basis of this strict host specialization is – a still controversially discussed subject (Aghnoum & Niks, 2010; Schulze-Lefert & Panstruga, 2011) that might be better addressed once the genomic differences between Bgh and Bgt are known. It emerges that similar to the other powdery mildews sequenced to date, Bgt possesses a large (c. 180 Mb) repeat-rich genome containing c. 6400 recognizable genes, which include 394 effector-like candidates. From a technological point of view the assembly of the Bgt genome promises to be much more extensive and consistent than that obtained with Bgh, due to the reliance, in part, on the Bacterial Artificial Chromosome (BAC) scaffolds used in the early stages of the project.

Effectors and avirulence proteins – the hunt continues

Many of the presentations by delegates were devoted to the identification, classification and functional analysis of effector genes/proteins. Originally, 248 CSEPs were identified in the Bgh genome (Spanu et al., 2010). This number was recently brought up to 491 upon an extensive manual search for additional candidate genes via iterative BLAST searches and the thorough exploitation of RNAseq data. The detailed computational characterization of this extended CSEP set with regard to the assignment of effector families, genomic clustering, structural predictions, evidence for diversifying selection and conservation in other mildew species is currently underway. There is preliminary evidence that subsets of CSEPs are expressed in a concerted manner during particular stages of fungal pathogenesis and it will be exciting to disentangle the transcriptional networks that govern these presumed effector regulons. Apart from the identification of novel CSEPs, some effort in the powdery mildew community is devoted to the analysis of additional Avr proteins in Bgh and Bgt– a challenging task. Strategies followed include the map-based cloning of BgtAvrPm3 (Srichumpa et al., 2005) and high-throughput transient expression of Bgh cDNA clones in transgenic Nicotiana benthamiana plants that express either the MLA1, MLA6 or MLA12 resistance protein (Halterman et al., 2001; Zhou et al., 2001; Shen et al., 2003).

Next to the identification of CSEPs and Avr proteins, the assignment of biological and/or biochemical functions to these proteins is of chief importance. Accordingly, a considerable number of scientists in the powdery mildew community attempt to unravel the cellular activities of individual effectors. This typically comprises the ectopic expression/delivery of effectors in plant cells, gene silencing approaches of either host (transient induced gene silencing – TIGS; Douchkov et al., 2005) or fungal genes (host-induced gene silencing – HIGS; Nowara et al., 2010) and the identification of effector targets, for example, via yeast two-hybrid approaches – experimental procedures that might be subject to ample criticism (see later). Nevertheless, there is increasing experimental evidence that a number of powdery mildew effectors indeed play a role in fungal virulence.

Standards in public (mildew) life

One of the features of workshops is that they provide an opportunity for vigorous debate on themes that concern the community. This meeting was no exception and Paul Schulze-Lefert (MPIPZ, Cologne, Germany) raised some provocative points that gave the participants plenty of food for thought and fuelled further discussion. The two most contentious themes are, in his view, the need to define minimal standards for experimentation in effector research and the need to use reference strains and varieties across all laboratories.

In essence, the majority (if not all) the functional assays currently used to screen for effector activity, effector targets and Avr genes are based on (transient) over-expression of transgenes or over-expression of RNAi constructs to silence targets. This applies to techniques such as HIGS, TIGS, virus-induced gene silencing (VIGS; Senthil-Kumar & Mysore, 2011), effector protein delivery via bacterial type III secretion systems (Sohn et al., 2007), the expression of bimolecular fluorescence complementation (BiFC; Kerppola, 2008) constructs and yeast two-hybrid assays. All these techniques may be prone to artifacts that are difficult to monitor and control. Moreover, in many cases the assays only allow the survey of a limited ‘window’ or particular stage of pathogen development. Thus, it might be necessary to validate the methodologies by corroboration of some of these findings with complementary technologies. This may include the generation of stable transgenic plants and genetic approaches on the pathogen side. One further contentious point raised was whether standard common strains and varieties should be employed, in order to render experimentation comparable across laboratories world-wide.

In fact, these aspects are not only critical for research on powdery mildew effectors but are valid for the search of effector functions in general, irrespective of the plant–pathogen system used. With regard to powdery mildews, there was a lively debate on whether and how improvements could or should be implemented. While the benefits of standardization of experimental systems was generally appreciated, it was also argued that one of the beauties and strengths of the barley (and wheat) powdery mildew systems as a model with real life impact on the food and crop security stage is precisely the historical wealth of information gleaned from the analysis of variation. This is the foundation on which the power of genetics rests. A second concern was that even in communities where ‘standard’ strains have been used successfully, careful experimentation has often revealed that the standards are not actually as identical as generally believed. It became clear from this debate that a healthy tension between standardization and use of polymorphisms will continue to be felt.

The challenges ahead

All mildew genomes sequenced so far are significantly larger than those of close relative ascomycetes. This is a feature they share with some other biotrophic pathogens (Duplessis et al., 2011) and symbiotic fungi (Martin et al., 2010). The retrotransposons which make up the bulk of this repetitive ‘non-genic’ DNA are a universe in themselves. The question of what really are the factors permitting or driving such genome expansions still needs to be clarified, including the tantalizing relationships between these jumping genes, avirulence and the evolution of host specificity (Sacristán et al., 2009). Continuing research in this area is likely to yield fruitful insights as well as challenges to our current understanding of plant–pathogen interactions.

The genome sequences of three powdery mildew species are now available, a fourth (Bgt) will be completed soon. It is likely that the current analyses of these only scratched the surface of the molecular secrets hidden in these obligate biotroph genomes. Continued thorough data mining of the available genomes thus promises to uncover novel aspects of the obligate biotrophic life-style and the evolution of these plant parasites. This goal will be aided by the resequencing of additional powdery mildew strains isolated from cultivated and wild host habitats. Planned resequencing projects in addition include a Bgh isolate that is partially virulent on otherwise highly resistant barley mlo genotypes (Schwarzbach et al., 2002). In the pipeline is also the de novo sequencing of the pepper powdery mildew pathogen, Leveillula taurica, which follows an unusual pathogenic route: this mildew species does not penetrate epidermal cells and stays exclusively epiphytic, but penetrates through stomata, propagates inside the mesophyll and sporulates through stomatal openings (Kunoh et al., 1979).

Additional future challenges comprise the cloning of additional Avr genes and, ultimately, the establishment of a reliable transformation protocol. With many unresolved issues left and exciting biology ahead, a positive decision has been taken to continue these meetings with a 3rd International Powdery Mildew Genomics Workshop to be held in Denmark in the summer of 2013. This symposium will be coordinated by Hans Thordal-Christensen and his local colleagues at the University of Copenhagen.

Ancillary