Standard Article

You have free access to this content

Avirulence Genes

  1. Thierry Rouxel,
  2. Marie-Hélène Balesdent

Published Online: 15 JAN 2010

DOI: 10.1002/9780470015902.a0021267



How to Cite

Rouxel, T. and Balesdent, M.-H. 2010. Avirulence Genes. eLS. .

Author Information

  1. INRA-Bioger, Thiverval-Grignon, France

Publication History

  1. Published Online: 15 JAN 2010

Introduction: Co-evolution of Molecular Plant–Pathogen Interactions

  1. Top of page
  2. Introduction: Co-evolution of Molecular Plant–Pathogen Interactions
  3. Avirulence Genes: A Definition
  4. Avirulence Genes Known to Date
  5. Adaptative Evolution
  6. Avirulence Genes and Fitness
  7. Concluding Remarks
  8. References
  9. Further Reading

As intuitively suggested more than 40 years ago by Van Der Planck 1968 when differentiating horizontal (i.e. general) resistance versus vertical (i.e. race-specific) resistance, it is now widely accepted that plants have two lines of defence against pathogens. This was recently nicknamed the ‘Central Dogma’ in plant pathology (Bent and Mackey, 2007). The first line (also referred to as basal immunity) provides defence against all potential pathogens and is based on recognition of basal microbial features termed pathogen-associated molecular patterns (PAMP) to result in PAMP-triggered immunity (PTI) and restrict pathogen development or kill it (De Wit, 2007; Jones and Dangl, 2006; Figure 1). Well-documented examples of PAMP include bacterial flagellins, lipopolysaccharides, fungal chitin or oomycete heptaglucosides (Zipfel and Felix, 2005). In a second step of plant–pathogen co-evolution, the pathogens developed genes encoding for effectors aiming at suppression of PTI and acting either on the signalling cascade or on the setting up of the defence response. Plants then counteracted with the development of a more specific and more sophisticated resistance (R) gene-mediated perception of effector proteins. This gene-for-gene interaction where one R gene product in the plant specifically perceives one matching effector protein, then behaving as an avirulence gene product, was established genetically in the mid-1940s as the gene-for-gene model (Flor, 1971; Figure 1) (for a recent review on current models of perception of avirulence gene products by resistance gene products). See also Plant Disease and Defence, and Pathogen Resistance Signalling in Plants

thumbnail image

Figure 1. Model for the evolution of bacterial resistance in plants. Left to right, recognition of pathogen-associated molecular pattern (PAMP) triggers basal immunity, which requires signalling through MAP kinase cascades and transcriptional reprogramming mediated by plant transcription factors. In a second step of the co-evolution, a cocktail of effector proteins is produced and delivered within plant cells via the Type III secretion system. The effectors target multiple host proteins to suppress basal immune responses. In a third step of co-evolution, plant resistance proteins (CC-NB-LRR and TIR-NB-LRR here) detect effector activity or the presence of the effector protein, and restore resistance through effector-triggered immune response. In a fourth step (not represented here), the bacteria alters its effectors to alter their functions or structure to eventually prevent the specific recognition by the resistance proteins. Reproduced from Chisholm et al. 2006 with permission of Elsevier.

Avirulence Genes: A Definition

  1. Top of page
  2. Introduction: Co-evolution of Molecular Plant–Pathogen Interactions
  3. Avirulence Genes: A Definition
  4. Avirulence Genes Known to Date
  5. Adaptative Evolution
  6. Avirulence Genes and Fitness
  7. Concluding Remarks
  8. References
  9. Further Reading

In its simplest acceptation, a phytopathogen avirulence gene is a gene encoding a protein which is specifically ‘recognized’ by genotypes of the host plant harbouring the matching resistance gene, regardless of its function or role in pathogenicity. In plant–eukaryote or plant–bacteria interactions, recognition is usually followed by a hypersensitive response (HR) that keeps the pathogen localized to the primary infection site. Of the large cocktail of effectors produced by eukaryotes and bacteria, only very few are currently matched by R gene in crops or other cultivated plants and are termed avirulence genes. However, most of them probably may become avirulence genes as soon as the corresponding R gene is found and used.

Avirulence Genes Known to Date

  1. Top of page
  2. Introduction: Co-evolution of Molecular Plant–Pathogen Interactions
  3. Avirulence Genes: A Definition
  4. Avirulence Genes Known to Date
  5. Adaptative Evolution
  6. Avirulence Genes and Fitness
  7. Concluding Remarks
  8. References
  9. Further Reading

Bacterial avirulence genes

Bacteria were the first phytopathogens in which an avirulence gene was cloned (AvrA from Pseudomonas syringae; Staskawicz et al., 1984). Since then, numerous avirulence genes (more than 40) were cloned, with most of the information originating from genera Pseudomonas and Xanthomonas (Vivian and Arnold, 2000; Bonas and Lahaye, 2002), and bacteria are the plant pathogens in which the larger number of avirulence genes is currently known (Table 1). Most of the effectors/Avr genes in genera Pseudomonas, Xanthomonas, Ralstonia and Erwinia are delivered within the host plant cell via a specific secretory and injection pathway, the Hrp (hypersensitive reaction and pathogenicity) (Mudgett, 2005). The Hrp is a type III secretory pathway (comprising multiple components in the inner and outer membranes that form a syringe-type export system that injects effectors and avirulence proteins directly into plant cells) active in plant and animal Gram-negative bacterial pathogens. One exception, however, regards the AvrXa21 avirulence protein of Xanthomonas oryzae pv. oryzae which seems to be secreted via a type I secretion system (Lee et al., 2006). In spite of the large number of cloned genes, there is mostly no significant sequence identity between bacterial avirulence genes (Bonas and Lahaye, 2002). The main exceptions, however, regard Xanthomonas campestris/axonopodis multigene families AvrBs3 and AvrRxv/yopJ (Bonas and Lahaye, 2002). The AvrBs3 family genes are present in a variety of Xanthomonas species and share 90–97% homology between one another. The main characteristic of the AvrBS3-family proteins is the presence of a central repeated motif of 34 amino acids responsible for avirulence range and pathogenicity (White et al., 2000). These also contain a nucleotide localization signal (NLS) and an acidic transcriptional activation domain (AD). Following dimerization in the plant cell cytoplasm, they are imported in the plant cell nucleus and act as a plant transcription factor inducing a cell size regulator (Kay et al., 2007). This leads to an increase in mesophyll cell size and protein production, providing the bacteria with a favourable ecological niche (Kay et al., 2007). In resistant plants, the Bs3 resistance gene promoter has a high binding affinity for AvrBs3, and the avirulence protein directly activates the apoptosis-causing Bs3 gene (Römer et al., 2007). In many cases of bacterial avirulence genes, an effector effect has been demonstrated, and usually corresponds to interference with plant defence responses (Table 1). Bacterial effectors can interfere with resistance protein activation, repress the salicylic acid pathway by activating the jasmonic acid pathway, modulate plant transcription and suppress cell wall–based defences as well as the execution of programmed cell death. In addition, some bacterial effectors interfere with the host proteins by mimicking the activity of plant proteases and phosphatases (Mudgett, 2005).

Table 1. Avirulence proteins from bacteria and their biochemical/effector functions
Avr proteinR geneHost plantPlant cell localizationBiochemical and/or effector function
Source: From Chisholm et al. 2006; Mudgett and White et al. (2005, 2000).
Pseudomonas syringae pv. glycinea
AvrARpg2SoybeanCytoplasmYopJ/AvrRxv family: interaction with plant MAP-kinases?
AvrBRpg1-bSoybeanUnknownKinase: phosphorylation of Rin4; manipulation of host jasmonic acid (JA) pathway
 Rpm1Arabidopsis thalianaUnknown
Pseudomonas syringae pv. maculicola
AvrRpm1Rpm1A. thalianaInner membrane?Kinase: phosphorylation of Rin4; inhibition of plant basal defence
Pseudomonas syringae pv. pisi
AvrPpiA1Rpm1A. thalianaUnknownUnknown
AvrRps4 (AvrPpiE)Rps4A. thalianaUnknownUnknown
Pseudomonas syringae pv. tomato
AvrPtoPto (and PRF)TomatoInner membraneKinase binding
AvrPtoBPto (and PRF)TomatoUnknownE3 ubiquitine ligase
AvrRpt2Rps2SoybeanUnknownCysteine protease: cleavage of Rin4; manipulation of host JA pathway; interferes with R-mediated defence
 Rps2A. thalianaCytoplasm
AvrRps4Rps4A. thalianaUnknownUnknown
AvrDRpg4SoybeanUnknownEnzyme involved in syringolide synthesis
Pseudomonas syringae pv. phaseolicola
HopAR1 (AvrPphB)Rps5 (and Pbs1)A. thalianaUnknownCysteine protease: cleaves Pbs1 and manipulates host JA pathway
HopX (AvrPphE)R2BeanUnknownUnknown
Xanthomonas axonopodis pv. vesicatoria
AvrBs1Bs1PepperUnknownYopJ/AvrRxv family
AvrBs2Bs2PepperCytoplasmAgrocipine synthase-like
AvrBs3Bs3PepperNucleusTranscription activator
AvrBs4Bs4TomatoUnknownAvrBs3 family
AvrBsTBsTA. thalianaUnknownYopJ/AvrRxv family; cysteine protease
AvrXv3Xv3TomatoNucleusTranscription activator
AvrXv4Xv4TomatoUnknownYopJ/AvrRxv family; cysteine protease
AvrRxvRxvBeanUnknownYopJ/AvrRxv family
Xanthomonas oryzae pv. oryzae
AvrXa3Xa3RiceUnknownAvrBs3 family
AvrXa5Xa5RiceUnknownAvrBs3 family
AvrXa7Xa7RiceNucleusAvrBs3 family; transcription activator
AvrXa10Xa10RiceProbably nucleusAvrBs3 family; transcription activator
AvrXa27Xa27RiceProbably nucleusTranscription activator
Xanthomonas campestris pv. armoraciae
Hax3Bs4TomatoUnknownAvrBs3 family
Hax4Bs4TomatoUnknownAvrBs3 family
Xanthomonas campestris pv. malvacearum
AvrB6B1CottonNucleusAvrBs3 family
Xanthomonas oryzae pv. oryzicola
Ralstonia solanacearum
PopP2Rrs1-RA. thalianaUnknownYopJ/AvrRxv family

Fungal avirulence genes

At present, 26 avirulence genes have been cloned from fungi. However, these were identified in only seven species, the Basidiomycete Melampsora lini, and the six Ascomycetes, Blumeria graminis, Rhynchosporium secalis, Cladosporium fulvum, Fusarium oxysporum, Magnaporthe grisea and Leptosphaeria maculans (Stergiopoulos and de Wit, 2009; Table 2). Except for ACE1 (avirulence confering enzyme 1) (4034 amino acids) of M. grisea that encodes a hybrid polyketide synthase/nonribosomal peptide synthase (PKS-NRPS enzyme) involved in the biosynthesis of the effective avirulence signal (Böhnert et al., 2004), fungal Avr genes cloned to date encode small proteins (63–314 amino acids before being processed) that are known or expected to be secreted and are often rich in cysteines (Bos et al., 2003; Rep, 2005; Table 2). Most of these Avr genes show no sequence homology among each other. Biochemical or cellular functions have been described for a few of these Avr proteins (Table 2). Avr-Pita of M. grisea is predicted to encode a putative metalloprotease (Orbach et al., 2000). In C. fulvum, the Avr4 protein acts as a chitin-binding protein protecting fungal cell walls from plant chitinases, and the Avr2 protein acts as an inhibitor of the plant cysteine protease Rcr3 required for activation of Cf2 (Cladosporium fulvum 2) -dependent resistance (Van den Burg et al., 2003; Rooney et al., 2005). The Nip1 (necrosis inducing peptide 1) gene identified in R. secalis encodes a necrosis-inducing peptide and thus functions as a toxin (Rohe et al., 1995). Finally, the AvrP123 gene of M. lini encodes a putative serine protease inhibitor, suggesting that host proteases might be a target for this effector (Catanzariti et al., 2006).

Table 2. Fungal and oomycete avirulence genes known to-date
Avr proteinNb of AA residuesaNb of cysteinesSignal peptidePutative functionLocalization in plantsExpressionRole in virulence/fitnessR gene
  1. a

    Number of amino acid residues in the unprocessed protein.

Source: Dong et al. 2009; Kamoun 2006; Li et al. 2009; Parlange et al. 2009; Qutob et al. 2009; Stergiopoulos and de Wit 2009; van Poppel et al. 2008; Yoshida et al. 2009.
Cladosporium fulvum (host: tomato)
Avr278820Protease inhibitorApoplastIn plantaInhibits Rcr3 and other proteasesCf-2
Avr4135818Chitin-bindingApoplastIn plantaProtects against chitinasesCf-4
Avr4E121610UnknownApoplastIn plantaUnknownHcr9-4E
Avr963623Carboxypeptidase inhibitorApoplastIn plantaUnknownCf-9
Leptosphaeria maculans (host: oilseed rape)
AvrLm1205122UnknownProbably in cytoplasmOver-expressed in plantaUnknown; weakly involved in fitnessRlm1
AvrLm6144620UnknownProbably in apoplastIn plantaUnknownRlm6
AvrLm4-7143821UnknownProbably in apoplastStrongly over-expressed in plantaUnknown; involved in fitnessRlm4 and Rlm7
Fusarium oxysporum f. sp. lycopersici (host: tomato)
Avr1 (Six4)242617UnknownXylemIn plantaSuppression of I-1- and I-2-mediated resistanceI
Avr2163219UnknownXylem (translocated in cytoplasm?)In plantaUnknownI-2
Avr3 (Six1)284821UnknownXylemStimulated by living cellsAggressiveness determinantI-3
Magnaporthe oryzae (host: rice)
Avr-Pita224816MetalloproteaseCytoplasmIn plantaNone knownPi-ta
Pwl1 to Pwl3137 to 1470, 1 or 221 or 23Glycin-rich hydrophilic proteinProbably in apoplastUnknownUnknownAvirulence towards weeping lovegrass
Ace1403543nonePKS/NRPSFungal appressoriumExpressed in appressoriumUnknownPi33
AvrPia85219UnkwownCytoplasmOver-expression in plantaUnknownPia
AvrPii70319Similarity to the C2H2 zinc finger motifCytoplasmOver-expression in plantaUnknownPii
Avr-Pik/km/kp113321UnknownCytoplasmOver-expression in plantaUnknownPik, Pik-m and Pik-p
Rhynchosporium secalis (host: barley)
Nip1821022Nonspecific toxinProbably in apoplastIn vitroInduces cell necrosisRrs-1
Blumeria graminis f. sp. hordei (host: barley)
Avra102864noneUnknownProbably in cytoplasmUnknownUnknownMla10
Avrk11773noneUnknownProbably in cytoplasmUnknownUnknownMlk1
Melampsora lini (host: flax)
AvrL567150123UnknownCytoplasmExpressed in haustoriaUnknownL5, L6 and L7
AvrM314128UnknownCytoplasmExpressed in haustoriaUnknownM
AvrP1231171123Serine proteaseCytoplasmExpressed in haustoriaUnknownP, P1, P2 and P3
AvrP495728Cystine knotted peptideCytoplasmExpressed in haustoriaUnknownP4
Hyaloperonospora parasitica (host: Arabidopsis thaliana)
Atr1NdWsB311215RXLR domain for translocationCytoplasmUnknownUnknownRpp1Nd and Rpp1-WsB
Atr13187019RXLR domain for translocationCytoplasmOver-expressed in plantaUnknownRpp13
Phytophthora sojae (host: soybean)
Avr1b-1138121RXLR domain for translocationCytoplasmExpressed during early stages of plant infectionSuppresses plant deathRps1b
Avr1a122024RXLR–dEER domain for translocation; W-like domainCytoplasmIn plantaUnknownRps1a
Avr3a111–119020RXLR–dEER domain for translocation; W-like domainCytoplasmIn plantaUnknownRps3a
Avr3c220020RXLR-dEER domain for translocation; W-like domainCytoplasmExpressed during early stages of plant infectionUnknownRps3c
Phytophthora infestans (host: potato)
Avr3a147021RXLR–dEER domain for translocation; manipulation of plant ubiquitin proteasome system?CytoplasmOver-expressed in plantaSuppresses INF1-mediated plant deathR3a
Avr4287024RXLR–dEER domain for translocationCytoplasm ?UnknownUnknownR4

Oomycete avirulence genes

Oomycetes resemble fungi in many respects but have evolved independently and are classified in the Stramenopila. At least 13 oomycete Avr genes have been cloned to date from Phytophthora species, Phytophthora infestans, the potato and tomato late blight causal agent and Phytophthora sojae, the soybean root and stem rot agent, and from Hyaloperonospora parasitica, causing downy mildew on crucifers (Kamoun, 2006; B Tyler, personal communication) (some examples are displayed in Table 2). As for bacteria and fungi, the encoded proteins share little sequence similarity. However, they all harbour a signal peptide followed by a conserved RXLR motif downstream from the signal peptide. This motif is often associated with a dEER motif. The RXLR–dEER motif is similar to a host-targeting signal that is required for translocation of proteins from malaria parasites (Plasmodium species) into the cytoplasm of host cells, suggesting a common origin for these effector translocation motifs, and that the RXLR motif also mediates translocation within plant cells (Kamoun, 2006). Conservation of the RXLR motif allowed whole-genome search for putative homologues in the genomes of P. sojae and of the causal agent of Sudden Oak Death, Phytophthora ramorum (Jiang et al., 2008). The authors thus identified more than 370 candidate genes in each of the species. Using the Avr1b-1 sequence of P. sojae as the initial query, a single superfamily was identified as avirulence homologue (Avh) genes, which accounted for most of the putative effectors and showed extensive sequence divergence (Jiang et al., 2008). Although the RXLR–dEER motifs for translocation are conserved among a wide range of effectors, much more variable motifs, called K, L, W and Y were identified in oomycete effectors and suggested to be important functional elements (Dou et al., 2008; Jiang et al., 2008).

Finally, structural features of euykaryote avirulence genes are consistent with their location of secretion. For example, richness in cysteine residues of fungal avirulence proteins and effectors is usually linked with secretion in the plant apoplast or xylem vessels, whereas Avr proteins with few cysteine residues often correspond to those of haustoria-forming pathogens (biotrophic fungi or oomycetes) for which the protein will be directly translocated within the plant cell (Table 2).

Viral avirulence genes

Plant viruses have small genomes, and thus few genes and gene products. The availability of hundreds of viral genome sequences and sequence variation helped identify the viral avirulence genes, and the mechanisms they use to escape recognition by resistance genes. Studies thus identified as avirulence determinants viral coat proteins, viral RNA (ribonucleic acid) polymerases (and mainly the helicase domain of the polymerase), movement proteins or the VPg, cylindral inclusion protein, P3 protein or NIa protease (Table 3). This indicates that virtually all proteins of viruses can act as an avirulence determinant, either in typical HR-causing gene-for-gene interaction or when acting primarily as virulence factors (e.g. the VPg protein; Table 3). Moreover, recent evidence suggests that the viral genome itself can act as an avirulence determinant and an untranslatable 860 nt RNA sequence of cymbidium ringspot tombusvirus could elicit the HR in Datura stramonium (Harrison, 2002).

Table 3. List and intrinsic function of currently known viral avirulence genes
Viral avirulence geneVirus speciesVirus genusHost plantPlant resistance gene
Source: Adapted with permission from Janzac 2008.
Coat protein    
 Turnip crinckle virusCarmovirusArabidopsis thalianaHrt
 Cucumber mosaic virus (CMV)CucumovirusA. thalianaRcy1
 Potato virus X (PVX)PotexvirusPotatoNx
 Paprika & pepper mild mottle virusTobamovirusPepperL2
 Pepper mild mottle virusTobamovirusPepperL3
 Pepper mild mottle virusTobamovirusPepperL4
 Tobacco mosaic virus (TMV)TobamovirusTobaccoN
P25 protein    
 Beet necrotic yellow vein virusBenyvirusSugar beetRz-1
RNA-dependent RNA polymerase    
 CMVCucumovirusFrench beanCry
 CMVCucumovirusFrench beanRT4-4
Movement protein    
 Turnip mosaic virus (TuMV)PotyvirusOilseed rapeTurB01
 TuMVPotyvirusOilseed rapeTurB05
P3 protein    
 TuMVPotyvirusOilseed rapeTurB03
 TuMVPotyvirusOilseed rapeTurB04
 Zucchini yellow mosaic virusPotyvirusZucchiniZym
P3 & HcPro proteins    
 Soybean mosaic virusPotyvirusSoybeanRsv-1
NIa protease    
 Potato virus Y (PVY)PotyvirusPotatoRy
NSs protein    
 Tomato spotted wilt virusTospovirusPepperTsw
Viral protein genome-linked (VPg) (corresponding to a virulence factor)
 Barley yellow mosaic virusBymovirusBarleyRym-4
 Pea seed-borne mosaic virusPotyvirusPeaSbm-1
 Tobacco vein mottling virusPotyvirusTobaccova
 Bean yellow mosaic virusPotyvirusFrench beanwlv
 Rice yellow mottle virusSobemovirusRiceRymv-1
P3 protein (corresponding to a virulence factor)
 Pea seed-borne mosaic virusPotyvirusPeaSbm-2


Very few other gene-for-gene interactions are currently known to occur between plants and other phytopathogens or pests. The most notable exception regards the resistance conferred by the tomato Mi-1 gene. It confers resistance to three species of root-knot nematodes, Meloidogyne arenaria, Meloidogyne incognita, and Meloidogyne javanica and potato aphid Macrosiphum euphorbiae (Fuller et al., 2008). Both nematode resistance and aphid resistance seem to behave according to the gene-for-gene model, as field isolates of root-knot nematodes and potato aphids that overcome resistance mediated by Mi-1 have been identified. Although this remains very speculative for the aphid species, a secreted peptide, MAP-1, was found solely in the avirulent populations of Meloidogyne incognita and it was also found exclusively in the other Meloidogyne species against which Mi-1 is effective. MAP-1 may be involved in the early recognition stages of avirulent interactions as antibodies raised against the peptide labelled secretions from sense organs (termed amphids) and not from the salivary glands of the nematode (Fuller et al., 2008).

Adaptative Evolution

  1. Top of page
  2. Introduction: Co-evolution of Molecular Plant–Pathogen Interactions
  3. Avirulence Genes: A Definition
  4. Avirulence Genes Known to Date
  5. Adaptative Evolution
  6. Avirulence Genes and Fitness
  7. Concluding Remarks
  8. References
  9. Further Reading

Plants and their associated microbes have co-evolved for million years. Selection simultaneously promoted aggressiveness and overcoming of plant defences. Host and pathogen thus have undergone continuous antagonistic co-evolution with diversifying selection operating to promote protein diversity in both partners. In this respect two types of avirulence genes are present in genomes: those corresponding to highly conserved proteins in related (or less related) species, suggesting an important role in pathogenicity and, in contrast, genes submitted to an increased diversification (i.e. species-specific genes), which, when submitted to host-plant selection pressure, can also be isolate-specific. Examples of the first series of effectors encompass the conserved presence of a series of C. fulvum extracellular proteins (ECP) or Avr genes such as Avr4, Ecp2 and Ecp6 in Mycosphaerella and Cercospora species (Stergiopoulos et al., 2007, unpublished data). However, most of the currently known avirulence genes are species-specific and do not show matches in the ever-increasing genomic data available, including in whole genome sequences of related species. As illustrated with RXLR oomycete effectors (Jiang et al., 2008), this incredible diversity suggests an extremely rapid sequence diversification whose mechanisms are largely unknown.

Gene-for-gene-based disease resistance is an economically important trait in crops due to its relative ease to use and breed for. However, its efficiency widely relies on Avr gene structure of the population of the pathogen faced to the R gene. Adaptation to R-gene selection pressure implicates that the avirulence gene will be submitted to an even faster evolution to adapt largely to grown resistance genes. In agricultural systems, this results in the so-called boom-and-bust cycles where a new resistance source is bred and successfully used on very large acreages due to its complete efficiency against pathogen populations which are mainly avirulent. This exerts a strong selection pressure on the corresponding avirulence allele of the pathogen population and, depending on the pathogen mode of life and the fitness penalty imposed by the loss of function of the Avr gene, in an evolution of the Avr gene so that its product is not recognized anymore by the plant-resistance genes. Pathogen populations then become largely virulent at this given Avr allele, and the resistance loses its efficiency and economic profitability. Numerous examples of this ‘boom-and-bust’ phenomenon have been described in all classes of pathogens and a risk assessment model taking into account the biological characteristics of the pathogen (mode of reproduction, effective size of populations, dissemination, mutation rate, etc.) has been proposed for fungi (McDonald and Linde, 2002) and viruses (Garcia-Arenal and McDonald, 2003). In the case of viruses, such models were improved by also taking into account the constraints imposed by amino-acid substitution effects on the viral fitness (Janzac et al., 2009). Pathogens with the highest ‘evolutionary potential’ (sensu McDonald and Linde, 2002) can overcome the corresponding resistance gene in only 3 years (e.g. the ‘breakdown’ of Rlm1 by L. maculans; Rouxel et al., 2003). The common adaptation of pathogens following recognition of one effector as an avirulence determinant either lies in generation of numerous highly divergent alleles, as illustrated by the AvrL567 locus in the Basidiomycete M. lini (Ellis et al., 2007) or the ATR1NdWsB loci of the oomycete H. parasitica (Rehmany et al., 2005), or in loss of gene function, including complete deletion of the gene. In laboratory mutants or collection isolates, pathogens were described to escape avirulence-mediated host recognition by different molecular mechanisms including nonsynonymous point mutations, frameshift mutations, deletion of the entire gene or transposon insertion (Fudal et al., 2009). However, the molecular mechanisms by which the pathogen will eventually modify its avirulence gene in the wild are largely unknown due the very few studies using large representative pathogen populations from widely diverse geographic origins. Examples of this recent approach encompass the works of Schürch et al. 2004 on evolution of Nip1 in R. secalis, Stukenbrock and McDonald 2007 on evolution of the host-selective effector SnToxA in Phaeosphaeria nodorum, Stergiopoulos et al. 2007 on a series of effectors and Avr genes of C. fulvum, Gout et al. 2007 on the evolution of AvrLm1 in L. maculans and Parlange et al. 2009 on the evolution of AvrLm4-7 in L. maculans. Contrasting information was obtained from these analyses. In the case of L. maculans, occurrence of AvrLm genes in a genome environment rich in repeated elements and strongly inhibiting meiotic recombination seems to be detrimental for diversification of alleles and complete deletion was the favoured mode of evolution towards virulence in AvrLm1 (Gout et al., 2007). A somewhat intermediate situation was observed for Nip1 of R. secalis, with deletion of the whole gene being the favoured mode of evolution under selection pressure in some geographic regions, whereas diversifying selection was prevalent in other geographic areas leading to unrecognizable forms of the protein (Schürch et al., 2004). This was hypothesized by the authors to be linked with heterogeneous use of the corresponding resistance Rrs1 by farmers. A diversified series of molecular events were observed for avirulence genes of field and greenhouse C. fulvum populations (Stergiopoulos et al., 2007). Avr9 and Avr4E mainly evolved via deletion, and only limited nucleotide polymorphisms were observed, some of them having no effect on the avirulence function. In contrast, no deletions were observed for Avr4 and Avr2, and their main mode of evolution was via numerous mutations/indels leading to amino acid substitutions or production of truncated proteins (Stergiopoulos et al., 2007). Of peculiar significance was the common substitution of cysteine residues involved in the building of disulfide bridges in Avr4, leading to the production of less stable Avr4 proteins (Stergiopoulos et al., 2007). The other extreme was observed in the AvrLm4-7 avirulence gene of L. maculans. The AvrLm4-7 gene shows a double recognition specificity by resistance genes Rlm4 and Rlm7 and, when submitted to the Rlm4 selection pressure, the most favourable event is a targeted single base nonsynonymous mutation, only suppressing recognition by the Rlm4 resistance gene without altering recognition by Rlm7 (Parlange et al., 2009).

Finally, and regardless of the importance of the Avr gene for the pathogen fitness, the mode of RAvr interaction (direct or indirect interaction) is also postulated to influence the favoured mode of evolution of Avr genes under R gene selection pressure. Although the first models of Avr product recognition postulated direct protein–protein interaction (the so-called elicitor-receptor model, popularized by NT Keen in the early 1990s), this could only be experimentally demonstrated on very few models, and the lack of evidence for direct binding suggested that many ‘gene-for-gene’ interactions can involve more than two genes. The ‘guard model’ proposes that the R protein is not the direct target of the Avr factor produced by the pathogen. Instead, the R gene protein acts as a ‘guard’ of the actual pathogen target (the ‘guardee’). On these bases, Bent and Mackey 2007 postulated that maintenance of effector while escaping recognition is linked with direct recognition of effector protein by resistance gene product, whereas indirect recognition will rather favour gene inactivation or deletion. In the first case, direct recognition can be suppressed by only minute mutations changing the protein conformation, whereas in the second case, it is the functional consequences of the effector effect on the guardee that will be sensed by the plant machinery, thus necessitating complete inactivation of the effector. See also Pathogen Resistance Signalling in Plants

Avirulence Genes and Fitness

  1. Top of page
  2. Introduction: Co-evolution of Molecular Plant–Pathogen Interactions
  3. Avirulence Genes: A Definition
  4. Avirulence Genes Known to Date
  5. Adaptative Evolution
  6. Avirulence Genes and Fitness
  7. Concluding Remarks
  8. References
  9. Further Reading

Simply defined, fitness is the combined ability of an organism to survive and reproduce. Fitness is quantifiable, and can be measured in absolute and relative terms. For plant pathogens, several traits, such as reproductive rate, rate of multiplication, infection efficiency or amount of disease caused (aggressiveness), have been used to measure pathogen or parasitic fitness (Leach et al., 2001). By analogy with bacterial avirulence genes, it is accepted that fungal and oomycete avirulence genes are primarily effectors. However, in most of the cases where intrinsic function could not be easily deduced from the sequence data (or in the case of viruses, when function is known), their contribution to the aggressiveness and fitness of pathogens still has to be demonstrated (see Table 2). In addition, for eukaryotes, an intrinsic effector function is often difficult to identify due to complex life cycle of many species, some parts of it being not experimentally accessible, and probable redundancy of effects within the wide diversity of effectors produced. In many cases, only the observation of deficit of infectivity of mutants with altered avirulence genes in controlled conditions were used to deduce that one given avirulence gene is involved in the pathogen fitness (Leach et al., 2001), and the significance of this in natural populations is questionable. More comprehensive approaches to elucidate a role in the ‘fitness deficit’ linked with loss or modification of the avirulence gene were undertaken in very few viral, bacterial and fungal models. Such researches usually are very tedious and involve combination of controlled condition experiments targeting biological/phytopathological traits (growth and sporulation in axenic media, infectiveness of spores or bacteria, latent period, speed of growth within the plant tissues, in planta competition between virulent and avirulent isolates, etc.) and in-field experiments aiming at evaluating competitiveness of avirulent versus virulent isolates (maintenance of virulent populations in absence of selection pressure, in planta competitiveness of avirulent versus virulent populations) (Leach et al., 2001; Huang et al., 2006; Janzac et al., 2009). A few viral models were investigated successfully to establish fitness penalty linked with loss of the avirulence function (Janzac et al., 2009). In viruses, all genes are indispensable and only point or multiple mutations lead to gain of virulence. In these organisms, fitness penalty and reduced competitiveness is directly linked to the evolutionary constraints acting on amino acid substitution in avirulence factors. In other words, a very strong fitness cost is associated with amino acid substitutions in cases where the target protein is a highly constraint protein as is for example the case for the RNA-dependant RNA polymerase of PVY acting as an avirulence gene towards the Pvr4 gene of pepper (Janzac, 2008). In contrast to viruses, only one bacterial and one fungal model provide to date convincing evidence on the involvement of avirulence genes in pathogen fitness (Vera Cruz et al., 2000; Huang et al., 2006). In X. oryzae pv. oryzae, the bacterial blight pathogen of rice, combinations of mutations in known avr genes avrXa7 and avrXa10 along with population surveys showed that mutations in avrXa7 caused the largest reduction in pathogen fitness in controlled conditions (shorter lesions and reduced bacterial numbers compared to wild type), and that virulent populations did not persist in natural populations in the absence of selection pressure (Leach et al., 2001). In contrast, mutations in avrXa10 had no detectable effect (Leach et al., 2001). Comparison of virulent and avirulent isolates of L. maculans in the course of colonization of oilseed rape plants in field conditions, and comparison of near-isogenic isolates in controlled conditions both indicated a strong fitness penalty linked with the lack of avirulence function at the AvrLm4 locus, whereas the effect was much lower or not measurable at the AvrLm1 locus (Huang et al., 2006; unpublished data).

Concluding Remarks

  1. Top of page
  2. Introduction: Co-evolution of Molecular Plant–Pathogen Interactions
  3. Avirulence Genes: A Definition
  4. Avirulence Genes Known to Date
  5. Adaptative Evolution
  6. Avirulence Genes and Fitness
  7. Concluding Remarks
  8. References
  9. Further Reading

The advent of genomics shortly followed the development of the effector concept in bacteria and its generalization to all plant pathogens. In this respect, the numerous advances in the knowledge of bacterial, fungal and oomycetes avirulence genes cannot be dissociated from the even more in-depth understanding and advances on effectors. The recent identification of Avr genes from oomycetes and from a series of new fungal species raises several key questions that will drive future research in this area. For example, the intrinsic function of many Avr genes from eukaryotes, as well as the mode of entry within the host cells of translocated effectors remains to be elucidated. Similarly the actual target in the plant cell has only been found for very few avirulence genes, and mostly bacterial ones. More and more information suggest that one of the primary targets of bacterial effectors may be the host ubiquitin proteasome system which contributes significantly to the regulation of plant defences (Birch et al., 2009). In contrast not such effect has been identified for eukaryote effectors even though some of them were shown to actually interfere with plant defence mechanisms (see Table 2). However, the most intriguing point regards the origin and evolution of Avr genes. In bacteria, it is commonly accepted that many effectors were obtained via lateral gene transfer (LGT) between species, and are postulated to be ‘recent events’. This is strongly substantiated by their common occurrence on plasmids and/or their association with transposable elements or phage sequences. In some cases, their GC content is also different from that of the rest of the genome. In oomycetes, in contrast, the conservation of the RXLR along with rapid divergence of the rest of the sequence in multigenic family of Avh suggested all Avh were related and likely evolved from a common ancestor by rapid duplication and divergence (Jiang et al., 2008). There is currently no evidence of multiple Avh gene families within genomes of fungi, maybe because of extreme diversification of sequences. Similarly, indications of LGT are sparse. Some fungal Avr genes occur in dispensable regions of the genome, can be associated with transposable elements and show unusual GC content of codon usage (for a review, van der Does and Rep, 2007), but only one example of LGT is currently documented: the transfer of the host-selective toxin-encoding gene ToxA between P. nodorum and Pyrenophora tritici-repentis (Friesen et al., 2006). With neither LGT nor rapid duplication and divergence being firmly established as mode of birth and evolution of fungal Avr genes, this remains an open and fertile field for future research that will greatly benefit from the ever-increasing genome initiatives in fungi.

Aggressiveness (sensu Van der Planck, 1968)

A quantitative trait expressing the relative ability to colonize the plant tissues and the amount of damages one isolate/strain will cause if being virulent.


Unable to cause disease following specific recognition of an avirulence gene product by the corresponding major resistance gene in the plant.


Pathogen molecules that manipulate host cell structure and function thereby facilitating infection.


The combined ability of an organism to survive and reproduce.

Virulent (sensu Van der Planck, 1968)

A qualitative trait expressing the ability to cause disease in gene-for-gene systems.


  1. Top of page
  2. Introduction: Co-evolution of Molecular Plant–Pathogen Interactions
  3. Avirulence Genes: A Definition
  4. Avirulence Genes Known to Date
  5. Adaptative Evolution
  6. Avirulence Genes and Fitness
  7. Concluding Remarks
  8. References
  9. Further Reading

Further Reading

  1. Top of page
  2. Introduction: Co-evolution of Molecular Plant–Pathogen Interactions
  3. Avirulence Genes: A Definition
  4. Avirulence Genes Known to Date
  5. Adaptative Evolution
  6. Avirulence Genes and Fitness
  7. Concluding Remarks
  8. References
  9. Further Reading
  • Molecular Plant Pathology Special issue celebrating the 25th anniversary of the cloning of a type III effector gene Volume 10 Issue 6 (November 2009)