OUP user menu

Botrytis cinerea virulence factors: new insights into a necrotrophic and polyphageous pathogen

Mathias Choquer, Elisabeth Fournier, Caroline Kunz, Caroline Levis, Jean-Marc Pradier, Adeline Simon, Muriel Viaud
DOI: http://dx.doi.org/10.1111/j.1574-6968.2007.00930.x 1-10 First published online: 1 December 2007


Botrytis cinerea is responsible for the gray mold disease on more than 200 host plants. This necrotrophic ascomycete displays the capacity to kill host cells through the production of toxins, reactive oxygen species and the induction of a plant-produced oxidative burst. Thanks to an arsenal of degrading enzymes, B. cinerea is then able to feed on different plant tissues. Recent molecular approaches, for example on characterizing components of signal transduction pathways, show that this fungus shares conserved virulence factors with other phytopathogens, but also highlight some Botrytis-specific features. The discovery of some first strain-specific virulence factors, together with population data, even suggests a possible host adaptation of the strains. The availability of the genome sequence now stimulates the development of high-throughput functional analysis to decipher the mechanisms involved in the large host range of this species.

  • Botrytis
  • gray mold
  • toxins
  • signal transduction
  • host adaptation

Phylogenetic introduction to the gray mold agent

Botrytis cinerea is an ascomycete responsible for gray mould on hundreds of dicot plants (Elad et al., 2004). In grapevine, conidia can contaminate leaves or inflorescences, but the fungus develops mainly in the autumn on ripe grape berries. The wide variety of symptoms on different organs and plants may suggest that B. cinerea has a large ‘arsenal of weapons’ to attack its host plants. Among the necrotrophic and polyphageous fungi, the gray mold agent is one of the most-studied models (van Kan, 2006). One challenge is to elucidate whether B. cinerea has specific virulence factors compared with other fungal pathogens with a narrower host range. Another question to address is whether all strains of B. cinerea share a common arsenal of weapons or whether there are strain-specific adaptations. This review focuses on some of the latest inputs in the knowledge of B. cinerea biology and highlights the interest of combining population genetics data, gene knock-out approaches and high-throughput genomic tools. Recent data confirm that the gray mold fungus shares conserved virulence factors with other fungal plant pathogens, but they also reveal unique features and a high variability between strains that may be linked to its necrotrophic and polyphagous behaviour.

Recent phylogenetic studies yielded a better knowledge of the B. cinerea (teleomorph Botryotinia fuckeliana) species. The genus Botrytis Persoon (Ascomyceta, Leotiomycetes, Sordariaceae) comprises two clades (Staats et al., 2005): one including four species that attack only Eudicot plants, the other including the 18 remaining species that attack mainly Monocot plants. Apart from the gray mold agent, most Botrytis species have a very narrow host range. Recently, multiloci gene genealogies showed that the B. cinerea morphological species were subdivided in two genetically isolated groups, named Group I and Group II, that were therefore proposed to be phylogenetic species. Group I (also referred as ‘Botrytis pseudocinerea’) and Group II (B. cinerea sensu stricto) strains exhibit differences in their ecology and their resistance pattern to fungicides (Fournier et al., 2003, 2005). Moreover, they are unable to cross with each other (D. Fortini & R. Fritz, pers. commun.). Functional studies presented in this review were performed using B. cinerea Group II model strains.

Main characteristics of the infection process

The infection process of B. cinerea is usually described by the following stages: penetration of the host surface, killing of host tissue/primary lesion formation, lesion expansion/tissue maceration and sporulation (van Kan, 2006). Gene inactivation approaches, based on fungal protoplast transformation and homologous recombination at target locus (Tudzynski & Siewers, 2004), allowed the identification of genes participating in all the stages of the disease cycle (Table 1).

View this table:
Table 1

Botrytis cinerea virulence genes characterized by gene inactivation approaches

Biological processFunctionGeneStrainIn vitro mutant defectsIn planta mutant phenotype (Tested host)References
Appressorium penetrationTetraspaninBcpls1T4NDNo penetration C (Rose petals, bean and tomato leaves)Gourgues (2004)
Fungal cell wall integrityChitin synthaseBcchsIBd90Chitin content reduction; cell wall weakeningReduced colonization M (Grapevine leaves)Soulie (2003)
BcchsIIIaBd90Chitin content reduction; radial growth reduction; increase in hyphal ramificationReduced colonization M (Grapevine and thale cress leaves)Soulie (2006)
Toxin biosynthesisP450 monooxygenaseBcbot1T4 SAS56No botrydial secretionReduced colonization C (Tomato fruits, bean and tomato leaves) in T4 mutant onlySiewers (2005)
H2O2 generationSuperoxide dismutaseBcsod1B05.10Increased sensitivity to oxidative stressReduced colonization C (Bean leaves)Rolke (2004)
DetoxificationABC transporterBcAtrBB05.10Increased sensitivity to resveratrol and fenpiclonilReduced colonization C (Grapevine leaves)Schoonbeek (2001)
Plant cell wall degradationPectin methylesteraseBcpme1Bd90 B05.10Growth reduction on pectin medium observed in Bd90 mutantReduced colonization M (Apple fruits, grapevine and thale cress leaves) in Bd90 mutant onlyValette-Collet (2003) Kars (2005b)
Endo-polygalacturonaseBcpg1B05.10Growth reduction on pectin mediumReduced colonization C (Apple fruits, tomato leaves and fruits)ten Have (1998)
Bcpg2B05.10NDDelay in secondary lesion formation; reduced colonization C (Tomato and bean leaves)Kars (2005a)
Endo-β-1,4-xylanaseXyn11AB05.10Reduced endo-β-1,4-xylanase activityDelay in primary lesion formation; Reduced colonization C (Grape berries and tomato leaves)Brito (2006)
Signal transductionGα subunits of G-proteinsBcg1B05.10Altered colony morphology; defect in botrydial and proteases secretionInfection stopped after primary lesions formation C (Bean leaves, tomato fruits and leaves)Schulze Gronover (2001)
Bcg2B05.10NDReduced colonization C (Bean leaves, tomato fruits and leaves)Schulze Gronover (2001)
Bcg3B05.10Defect in germination and sporulation; excessive clerotia formationDelay in primary lesion formationC (Tomato leaves)Doehlemann (2006a)
Adenylate cyclaseBacB05.10Reduced germination, growth and sporulationReduced colonization; no conidiation C (Bean leaves)Klimpel (2002); Doehlemann (2006a)
MAP kinaseBmp1B05.10Defect in germination on hydrophobic surfaces; reduced vegetative growthNo appressoria; no penetration; no colonization C (Tomato leaves, carnation flowers)Zheng (2000); Doehlemann (2006a)
Bmp3B05.10Reduced vegetative growth and conidiation; no sclerotia; Sensitivity to low osmolarity, to oxidative stress and to phenylpyrroleReduced penetration efficiency; reduced colonization C (Gerbera flower petals, tomato leaves)Rui & Hahn (2007)
Bcsak1B05.10Defect in conidiation; increased sclerotial formation; Sensitivity to osmotic and oxidative stressesNo penetration; reduction of secondary lesions expansion M (Bean leaves)Segmuller (2007)
Histidine kinaseBos1USW111Osmosensitivity; resistance to dicarboximides and phenylpyrroles; defect in conidiationReduced colonization M (Apple fruits, bean and tomato leaves)Viaud (2006)
Protein foldingCyclophilin ABcp1T4Resistance to cyclosporin AReduced colonization C (Tomato and bean leaves)Viaud (2003)
(Peptidyl-prolyl isomerisation)FK506 binding protein (FKBP)Bcpic5T4 B05.10Resistance to FK506 drugReduced colonization C (Bean and tomato leaves) in T4 mutant onlyGioti (2006) Mey et al., (pers. comm.)
  • Strain dependant virulence factor.

  • ND, no detected defect; C, conidia inoculum; M, mycelium inoculum.

Botrytis cinerea develops different infection structures

Conidia of B. cinerea are considered to be the main produced and dispersed inoculum (Holz et al., 2004). Their germination and adhesion on plant surfaces represent crucial steps preceding host penetration and colonization, whereby sensing and recognition of the host surface characteristics, including hydrophobicity and sugar sources, seem essential (Doehlemann et al., 2005, 2006b). Germlings secrete an extracellular matrix that serves in their attachment through hydrophobic interactions but also release several extracellular enzymes that might facilitate their penetration by breaching the physical barriers of the plant (Doss, 1999). Botrytis cinerea is thought to enter the host mainly by producing degrading enzymes and causing an oxidative burst rather than using physical pressure. Appressorium-like structures depicted as swollen tips have been found frequently in B. cinerea (Fourie & Holz, 1995; Tenberge, 2004). These structures (Fig. 1a) are not highly melanized and not always separated from the germ tube by a septum, which would be necessary for the generation of an osmotic pressure as high as the one observed in appressoria of the rice blast fungus Magnaporthe grisea (Howard & Valent, 1996). However, both fungi share a conserved membrane protein, tetraspanin, required for the appressorium-mediated penetration into host plants (Gourgues et al., 2004). When plants are inoculated with mycelium, a different form of penetration was observed, whereby hyphae growing on the plant surface heavily ramify into short bulbous cell aggregates (Fig. 1b and d) described earlier as ‘claw-like’ structures (Kunz et al., 2006). These structures are similar to the ‘infection cushions’ or ‘complex appressoria’ of the closely related fungus Sclerotinia sclerotiorum (Hegedus & Rimmer, 2005; Jurick & Rollins, 2007). The Importance of infection cushions in fungal penetration has not yet been molecularly investigated in B. cinerea, but they were observed on different organs and plants (Sharman & Heale, 1977; Garcia-Arenal & Sagasta, 1980; Backhouse & Willetts, 1987; Fullerton et al., 1999).

Figure 1

Botrytis cinerea develops different penetration structures that are all accumulating H2O2. Conidial suspensions from K1 strain were inoculated on bean leaves (a). Pictures were taken 12 h postinoculation (hpi) using scanning electron microscopy (Kindly provided by Brigitte Gelly, INRA). Mycelium plugs (b and d) or conidia (c) from Bd90 strain were inoculated on the upper side of onion epidermis. At 10 hpi, the epidermal strips were transferred onto a solution containing 1 mg L−1 of DAB and were left overnight. Pictures were taken 24 hpi. Inoculated conidia (indicated by c) form germ tubes (gt) and appressorium-like structures (a), allowing penetration of an infectious hypha (ih), whereas inoculated mycelium (indicated by eh for external hyphae) forms infectious cushions (ic) as a penetration structure.

Host cell death occurs through fungal toxins and an oxidative burst generated by both the pathogen and the host

As a necrotrophic and polyphageous pathogen, B. cinerea secretes nonspecific phytotoxins to kill cells from a large spectrum of plants. Among the numerous metabolites isolated from fermentation broths (Collado et al., 2000, 2007), the most well known is the sesquiterpen botrydial (Fig. 2). Botrydial is produced during plant infection (Deighton et al., 2001), and induces chlorosis and cell collapse, which seems to facilitate both penetration and colonization (Colmenares et al., 2002). Botrydial biosynthetic pathway genes are organized into a physical cluster, coregulated and overexpressed in planta as shown by macro-array studies (Viaud et al., 2003; Gioti et al., 2006). Inactivation of one of these genes (CND5/Bcbot1) in three different strains demonstrated that botrydial is a strain-dependant virulence factor (Siewers et al., 2005). In addition to botrydial and its derivatives, aggressive isolates of B. cinerea also produce a second class of toxins derived from botcinic acid (Reino et al., 2004; Tani et al., 2006) (Fig. 2), which may explain the strain-dependent effect of Bcbot1 inactivation. In addition to these toxins, B. cinerea produces reactive oxygen species (ROS) during infection (Schouten et al., 2002; Tenberge, 2004). The hydrogen peroxide (H2O2) accumulates in the early steps of infection, both in germinating spores (Fig. 1c) and in the infection cushions (Fig. 1d). Moreover, gene inactivation of the superoxide dismutase (SOD) indicated that this H2O2-generating enzyme is a virulence factor (Rolke et al., 2004) contributing to the accumulation of phytotoxic H2O2 levels.

Figure 2

Botrytis cinerea produces a wide range of secondary metabolites. Biochemical structures have been determined for the sesquiterpenes botrydial (a. PubChem compound: CID: 185781) and abscissic acid (b. CID: 5375200) and for the polyketide called botcinic acid (c. CID: 11509607). Gene inactivation studies demonstrated that the phytoxin botrydial (Collado et al., 2000, 2007) is a strain-dependent virulence factor (Siewers et al., 2005) but did allow one to determine whether the vegetal hormone abscissic acid has a role in virulence (Siewers et al., 2004). The genes coding for the biosynthesis of the phytotoxic botcinic acid and its botcinin derivatives (Tani et al., 2006) have not been identified yet, but a genomic approach provided a list of 20 candidate genes that are coding for polyketide synthases (Kroken et al., 2003).

Plant cell death also occurs through an oxidative burst produced by the host itself in reaction to B. cinerea attack (Govrin & Levine, 2000). This rapid production of ROS by the plant was originally known to be required for defense gene expression and a hypersensitive reaction (HR), a type of programmed cell death (PCD) thought to limit access of the pathogen to water and nutrients. Nevertheless, in some host–pathogen interactions, PCD has a clear role in promoting pathogen growth (Greenberg & Yao, 2004). Studies in Arabidopsis thaliana and tobacco suggested that B. cinerea may even need the HR to achieve full pathogenicity (Govrin & Levine, 2000; Dickman et al., 2001). Indeed, A. thaliana mutants with a delayed or reduced cell death response are generally more resistant to B. cinerea infection, whereas mutants in which cell death was accelerated were more susceptible (van Baarlen et al., 2007). There is also increasing evidence that fungi generate PCD-promoting molecules as virulence factors (Greenberg & Yao, 2004). For example, B. cinerea endo-polygalacturonase 1 (BcPG1), first known as a virulence factor (ten Have et al., 1998), elicits defense responses in grape including the production of ROS (Poinssot et al., 2003; Vandelle et al., 2006). Recently, a nonpathogenic mutant overproducing BcPG1 was shown to provoke an HR-like reaction on grape and bean leaves that correlates with a cessation of tissue colonization (Kunz et al., 2006). Independent studies also correlated low virulence strains with strong HR in bean leaf disks (Unger et al., 2005). Therefore, PCD has an important role in B. cinerea virulence but, depending on its timing and strength and depending on the host, PCD may also correlate with resistance.

Enzymatic degradation arsenal relies on multigenic families

The capacity for degrading the host cuticle and cell wall has been extensively studied in B. cinerea (ten Have et al., 2002; Kars & van Kan, 2004). As a necrotroph, this fungus is notably equipped with multiple cell wall-degrading enzymes (CWDEs) that allow plant tissue colonization and the release of carbohydrates for consumption. Pectin, the major host cell wall component, can be degraded by a set of fungal pectinases. Among them, the polygalacturonase genes (BcPG16) were shown to be differentially expressed, depending on the stage of infection and the host, which suggested specialized functions for these pectinolytic activities (ten Have et al., 2001). Gene inactivation indicated that BcPG2 activity is involved in the penetration step by breaching the pectin network localized at the anticlinal cell wall (Kars et al., 2005a), whereas BcPG1 activity is required during colonization to breach the pectin network through the middle lamella (ten Have et al., 1998). Pectin methylesterases are assumed to facilitate the action of PGs by demethylating pectin to pectate, and inactivation of the bcpme1 gene resulted in a strong reduction of virulence in several plant hosts (Valette-Collet et al., 2003). Nonpectinolytic CWDEs, like cellulases or hemicellulases, also facilitate B. cinerea infection as illustrated by inactivation of an endo-β-1,4-xylanase gene that led to a reduced virulence (Brito et al., 2006).

Most of the degrading enzymes are encoded by multigenic families and some may have partly redundant functions. This would explain why inactivation of several genes encoding CWDEs (Espino et al., 2005; Kars et al., 2005b; van Kan, 2006) or cutinolytic enzymes, i.e. cutinase A and lipase 1 (van Kan et al., 1997; Reis et al., 2005), did not affect fungal virulence. Taken together, the multiplicity of the reported degrading activities and the reduction of virulence observed for several mutants impaired in degrading enzymes strongly support a major role in pathogenicity for the enzymatic degradation arsenal of B. cinerea.

Signal transduction regulating the infection process shows specific features in B. cinerea

Signal transduction cascades regulating fungal development and virulence are remarkably conserved between distantly related fungi, especially those involving the cAMP-dependant protein kinase and mitogen-activated protein kinases (MAPK) (Lengeler et al., 2000; Xu, 2000). Different regulations among pathogens could then be generated by differences in upstream inputs (environmental signals and receptors), connection between pathways and their downstream outputs (target genes). Recent studies highlight specific roles in the pathogenicity of some signaling components of B. cinerea when compared with other fungal pathogens.

Heterotrimeric GTP-binding proteins form a family of regulators that receive signals directly from receptors and are required for the infection process in many fungal models (Lengeler et al., 2000). In B. cinerea, inactivation of Bcg1, one of the three Gα subunit encoding genes, led to a severe pleitrophic phenotype (Schulze Gronover et al., 2001; Siewers et al., 2005). The mutant has an abnormal colony morphology, a defect in protease and botrydial secretion and is blocked after primary lesion development. Some of the defects (colony morphology) can be restored by the addition of cAMP, which demonstrated that BCG1 is involved in the cAMP-dependant pathway. The importance of this common pathway in vegetative growth and pathogenicity was confirmed by the inactivation of the adenylate cyclase gene, bac (Klimpel et al., 2002). Interestingly, the bcg1 mutant also has defects (proteases and botrydial secretion) that cannot be restored by cAMP, suggesting that BCG1 controls another signaling pathway besides the cAMP one. A suppression subtractive hybridization (SSH) approach confirmed that BCG1 controls the expression of secondary metabolism genes including botrydial (Bcbot1 gene), protease and CWDE genes (Schulze Gronover et al., 2004). Another expression study, using macroarrays, has shown that the botrydial gene cluster (including Bcbot1) is controlled by the calcineurin-dependent pathway (Viaud et al., 2003). Calcineurin phosphatase has an essential role in fungal morphogenesis and virulence (Fox & Heitman, 2002) and was shown to be involved in M. grisea and B. cinerea appressorium formation (Viaud et al., 2002, 2003). The observation that the botrydial gene Bcbot1 is regulated by both BCG1 and calcineurin suggests a common signaling pathway that regulates infection-related functions including the production of secondary metabolites (J. Schumacher, M. Viaud & B. Tudzynski, pers. commun.).

Signaling networks of filamentous fungi usually contain three MAPK cascades (Xu, 2000). The corresponding MAPK, BMP1, BcSAK1 and BMP3 have been genetically characterized in B. cinerea. BMP1 is the ortholog of the yeast FUS3/KSS1 and of the M. grisea virulence factor PMK1. bmp1 null mutants can germinate on plant surfaces, but they do not produce any appressoria. Even on wounded plants, they are unable to colonize the host tissues (Zheng et al., 2000; Doehlemann et al., 2006a). This phenotype shows that, as in other pathogenic fungi, the widely conserved BMP1 pathway is essential for host surface sensing, infection structure morphogenesis and colonization.

BcSAK1 is the ortholog of the yeast HOG1 involved in osmoregulation. In filamentous fungi, the corresponding pathway allows responses to osmotic and oxidative stresses. Null mutants, obtained in the phytopathogens M. grisea and Colletotrichum lagenarium, were still fully pathogenic (Dixon et al., 1999; Kojima et al., 2002). In contrast, bcSak1 mutants are severely impaired in pathogenic development (Segmuller et al., 2007). The putative upstream sensor of the BcSAK1 pathway could be BOS1 (S. Liu & S. Fillinger, pers. commun.), a member of the class III histidine-kinases (HKs) that are involved in osmoregulation and resistance to dicarboximide and phenylpyrrole fungicides (Catlett et al., 2003). As for the bcSak1 mutants, bos1 null mutants are impaired in osmotic stress, conidiation and plant tissue colonization (Viaud et al., 2006). This was the first HK reported to be a virulence factor in a phytopathogenic fungus. In M. grisea and in the seed-borne necrotrophic pathogen Alternaria brassicicola, inactivation of a homologous HK gene did not impair pathogenicity (Motoyama et al., 2005). The differential functions of the class III HKs and HOG1 orthologs among different pathogens may be linked to different fungal strategies for the management of osmotic and oxidative stresses during plant infection, or to differences in the level of stress encountered. Indeed, B. cinerea needs to resist the oxidative burst that occurs in the early stages of infection, which could explain the high importance of the BOS1 and BcSAK1 pathway(s). Macro-array approaches are in progress to investigate the downstream-regulated genes (S. Fillinger, S. Liu & M. Viaud, pers. commun.).

The third MAPK characterized in B. cinerea is BMP3 that is the homologue of the yeast SLT2 involved in cell wall integrity. However, bmp3 null mutants do not seem to be affected in cell wall integrity as expected from the data on other fungi. The phenotype suggests that BMP3 is involved in saprophytic growth, response to low osmolarity, conidiation, surface sensing, host penetration and lesion formation. Therefore, as for BcSAK1, BMP3 presents unique features in B. cinerea.

The recent inputs into the complex signaling network-controlling pathogenicity in B. cinerea has revealed both conserved patterns with other fungi and specific features for this necrotrophic pathogen. Other signaling components not mentioned in this review and cross-talks between them are under investigation (Tudzynski & Schulze Gronover, 2004).

A possible host adaptation of B. cinerea strains?

Botrytis cinerea populations display a significant phenotypic variability in their level of aggressiveness, the oxidative burst occurring during infection (Unger et al., 2005) and toxin production (Reino et al., 2004). The genetic bases for this variability are not yet elucidated, but different nonexclusive hypotheses are under investigation. A first source of variability can be the occurrence of mycoviruses among wild populations. Indeed, a double-stranded RNA virus isolated from B. cinerea was shown to confer hypovirulence-associated traits i.e. a reduced conidiation rate and a defect in bean leaf colonization (Castro et al., 2003). A second source of variability could be differences in gene content and gene variability among B. cinerea strains. This hypothesis is supported by some of the functional studies mentioned above, showing that inactivation of the same gene in different B. cinerea Group II strains can lead to different virulence phenotypes. These strain-dependent virulence factors are, for example, the botrydial biosynthesis gene Bcbot1 and the CWDE gene Bcpme1 (Table 1). Depending on the genetic background, the null mutants were affected or not in their virulence, suggesting that differences in gene content, gene regulation and enzyme or toxin production might occur between strains.

These observed variability and strain-dependant virulence factors may suggest that B. cinerea populations have developed strain-dependent infection strategies in which each toxin or CWDE has a more or less important role. This raises the question: does the host plant play a role in the evolution of virulence factors within B. cinerea populations? To the authors’ knowledge, this research question has not yet been directly addressed, but some indirect evidence for the role of host plant in the adaptation of B. cinerea populations already exists. First, conidia from B. cinerea strains isolated from tomato adhered and germinated faster on tomato leaves than strains isolated from grape (Cotoras & Silva, 2005). Moreover, the genotyping with microsatellite markers (Fournier et al., 2002) of 10 French populations of B. cinerea Group II sampled on grapevine and bramble showed that gene flow was slightly but significantly reduced between populations from the two hosts, even in strict sympatry (E. Fournier & T. Giraud, pers. commun.). Hence, B. cinerea populations may be slightly but efficiently adapted to their different host plants, rather in terms of host preference than true host specialization. Indeed, hosts and their parasites are involved in an evolutionary arms race characterized by adaptation and counter-adaptation of mechanisms of host defense and parasite attack. The selective pressures exerted by plants may thus impact the evolution of virulence factors at the population level, and the strain-dependent effects observed in many functional studies of B. cinerea virulence factors may be a consequence of this process.

The effect of a host plant on pathogen evolution may also be investigated at an interspecific level: the reciprocal selective pressures leave signatures of positive or balancing selection at the molecular level that can be detected between closely related species specialized on different hosts. In the Botrytis genus, the evolutionary forces acting on two paralogous necrosis and ethylene-inducing like proteins (NLPs) have been studied (Staats et al., 2007). NLPs constitute a family of secreted phytotoxic proteins that elicit plant cell death and directly interact with plant molecules. Among the Botrytis species, the evolution of a small number of residues of both NEP paralogs were shown to be driven by positive selection whereas the remaining parts of the proteins seemed to be evolutionarily constrained. Although no correlation was found with host range or type of host attacked, these results show that genes involved in the infectious process may be the target of various selective pressures that may have driven the evolution of related species on different hosts.

Conclusion – Botrytis enters the genomic area

The number of sequenced fungal genomes from both saprophytic and pathogenic species is increasing significantly (Galagan et al., 2005; Xu et al., 2006), which provides new comparative and functional genomics approaches to investigation of virulence. The neighbor species B. cinerea and S. sclerotinium are the first Leotiomycetes, but also some of the first polyphageous and necrotrophic fungi, to be sequenced, by the Broad Institute (http://www.broad.mit.edu/annotation/fgi/) and the Genoscope (http://www.genoscope.cns.fr/) (Fillinger et al., 2007). Comparing the gene content and organization among them and with other sequenced fungi will highlight the differences that could be associated with their necrotrophic lifestyles and their broad host ranges. Moreover, two strains of B. cinerea were sequenced (B05.10 and T4), which may contribute to an understanding of the phenotypic and genotypic variability of this species and the occurrence of strain-dependent virulence factors.

These genomic resources will stimulate the development of high-throughput functional analyses such as transcriptomics and proteomics. In parallel, gene inactivation analysis will benefit from a genetically modified strain of B. cinerea (ku null mutant), which has a high rate of homologous recombination that should facilitate the inactivation of further candidate genes (M. Choquer & M. Viaud, pers. commun.).

DNA chips using the whole set of B. cinerea genes are under development. Such transcriptomic approaches have already been performed, to a lesser scale, by the use of expressed sequences tags (Viaud et al., 2005) spotted on high-density filters. This way, the expression of 3032 genes was studied in different genetic backgrounds or physiological conditions. Twenty-seven genes were shown to be overexpressed during the infection of Arabidopsis thaliana (Gioti et al., 2006). Among them, some had already been described as virulence genes (Bcpg1, Bcbot1) but most had unknown functions. For five of them, the corresponding null mutants were affected during colonization (A. Gioti & C. Levis, pers. commun.), which highlights the utility of genomic analysis for the identification of new virulence factors that may be specific to B. cinerea. The functional study of the interaction between the gray mould agent and its host plants should also greatly benefit from the sequencing of the grape genome (Jaillon et al., 2007). A kinetic view of both complete transcriptomes during infection may give a first integrative view of the interaction and insights into the molecular dialogue occurring between the partners.


A pathogen profile of Botrytis cinerea was published while this Minireview was in press: Williamson B., Tudzynski B., Tudzynski P., Van Kan I.A.L. (2007) Botrytis cinerea: the cause of grey mould disease. Mol Plant Pathol 8: 561–580.


The author's work is funded by an INRA ‘Jeune Equipe’ Grant.


  • Editor: Richard Staples


View Abstract