PMs establish long term relationships with their host forcing the maintenance of compatibility conditions
. The large and diverse collections of effectors encoded in the PM genomes
 might indicate redundant functions of these compounds on their target molecules in the plant. This may explain why gene-for-gene resistance has negligible effect on the control of PM diseases in the field, and why plant resistance is overcome without apparent loss of fitness by the pathogen
. The identification of such effectors and their targets will be key to understand PM diseases, and design new alternatives to generate long-lasting passive resistance. To move in this direction, new experimental systems allowing the evaluation of compatibility as well as resistance by loss of susceptibility are required. In relation to the later condition, we have characterized the interaction between A. thaliana Te-0 plants and G. cichoracearum, where the host offers compatibility conditions during the initial, but not later stages of infection. In this plant, conidial germination is partially reduced and fungal growth is slightly delayed. Both features have been described in resistant interactions involving PTI or ETI/HR activation
[8, 20, 30, 34]. However, Te-0 infected tissues showed no signs of active defences (ROS accumulation, callose deposition, or cell death), in a response clearly distinguishable from that of infected Kas-1 tissues. Therefore, Te-0 resistance seems to be different from RPW8-mediated immunity, which confers broad-spectrum protection to powdery mildews in many different accessions of A. thaliana by activation of HR-like responses
[8, 27, 49–51]. The SA-dependent pathway mediating resistance to biotrophic pathogens was neither induced under basal conditions, nor over-stimulated in Te-0 fungal-infected tissues. In addition, the JA signaling cascade did not seem to increase resistance against G. cichoracearum in Te-0 plants. The LOX2 and PDF1.2 gene markers had mild basal expression without activation after infection, whereas VSP2 showed no basal expression and weak induction by pathogen, suggesting that resistance to the fungus was not due to major enhancement of JA-dependent defences. Curiously, the effect of the JA pathway on the interaction of A. thaliana with PMs is still unclear. The JA levels transiently increase in A. thaliana tissues infected with G. cichoracearum. Resistance to PMs is enhanced in wild type plants treated with JA, as well as in cev1 mutants that constitutively activate the JA pathway
. However, the coi1-1 and jar1-1 mutants, impaired in the JA-pathway, are not hyper-susceptible to PMs
. Therefore, in natural infection conditions, activation of the JA/ET pathway may not be sufficient to confer resistance to PMs
Te-0 plants did not show signs of premature senescence, alterations in size, or developmental-induced callose accumulation (http://www.arabidopsis.org/ABRC and Additional file 7). Neither did these plants stimulate strong defence features upon fungal infection. These observations argue against the possibility that resistance of Te-0 plants might originate from mutations in genes encoding PMR2, PMR4, EDR1, EDR2 and PUX2, previously described as PM-susceptibility factors, causing hyper-activation of defence responses
[22, 25, 34, 38, 54].
In A. thaliana, PMR5 and the pectate lyase PMR6 have been recognized as factors that potentiate susceptibility to PMs. PMR5 is involved in cell expansion whereas PMR6 functions in cell wall modeling
[23, 24], with both proteins being required for normal composition of plant cell wall pectin, and establishment of compatibility with PMs. Resistance of pmr5/pmr6 mutants does not involve the SA or JA/ET pathways, resembling resistance of Te-0 plants. However, these mutants are stunted, show reduction in cell size, constitutive sporadic mesophyll cell death, and deposition of auto-fluorescent compounds along the veins. As none of these features were observed in healthy or G. cichoracearum-infected Te-0 tissues, it is likely that resistance does not result from deficient PMR5/6 function in these plants.
To date, the two best characterized limiting stages for development of PMs are penetration of epidermal cells
[20, 55] and development of haustoria
[56, 57]. Interestingly, G. cichoracearum was not prevented from reaching any of these stages in Te-0 plants. More than 50% of inoculated conidia germinated in Te-0 tissues further developing penetration hyphae and functional haustoria. In contrast, fungal asexual reproduction was the main limiting stage for proliferation of the pathogen in this plant. Limitations in this developmental stage have been previously described for PM diseases in A. thaliana[34, 38]. In ascomycetes, coniditiation is regulated by several factors including inoculum density, light, temperature, humidity, and nutrient availability
[58, 59]. While significant progress has been made in elucidating this developmental program in model fungi (Neurospora spp, Aspergillus spp), its genetic basis and sensitivity to host clues remain elusive for obligate fungal biotrophs. In this sense, our results suggest that conidiation of G. cichoracearum in A. thaliana requires particular host conditions or signals which are present in Col-0, but not in Te-0 plants. Alternatively, Te-0 might offer limitations for optimal fungal development affecting for instance nutrient uptake, which negatively impact on Cph maturation, even though we found that the haustoria placed in these plants display normal assimilation of ArgC14.
Interestingly, Te-0 supports abundant mycelial development and moderate conidiation of the G. cichoracearum relative Erysiphe cruciferarum and is susceptible to field isolates of the oomycete Albugo spp. (E Kemen, JD Jones, unpublished results). Thus, the following evidences support the possibility that Te-0 is altered in components necessary for compatibility with G. cichoracearum: i) genetic nature of plant resistance (recessive), ii) susceptibility of the plant to other biotrophic pathogens, including a close relative species; iii) absence of active defences (markers of PTI, ETI, SA and JA pathways) in this interaction.
To characterize gene expression changes occurring under compatibility conditions, we looked for transcriptional differences in Te-0- and Col-0 infected tissues at the pre-conidiation stage. By SSH, we identified 8 differentially expressed genes, including classical defence-related genes (PR1, oxidoreductase), and genes affecting photosynthesis (chlorophyll a/b binding protein), which have been characterized elsewhere
[36, 37]. The remaining 5 candidates (MT1a, HSP70.1, END, GRF7 and STP4) showed increased expression in response to pathogen in susceptible Col-0 plants, but not in resistant Kas-1 plants, indicating they are sensitive to compatibility conditions, but not to ETI activation.
Previous studies reported activation of some of these genes in response to fungal infections. This has been observed for STP4 during the interaction of A. thaliana with G. cichoracearum, and for MT1a and its rice homologue in response to other haustoria-forming fungi
[61, 62]. In turn, HSP70.1 is induced by the obligate biotrophic oomycete Hyaloperonospora arabidopsidis (ex. parasitica)
, whereas the END-, and GRF7-homologues in barley are up-regulated by PMs
Depletion of genes encoding host compatibility factors is predicted to reduce pathogen growth. However, the functional redundancy of these genes might hinder the effect of null mutations on single genes. Conversely, an increased expression of one of these genes may generate tractable effects on the invader’s growth. Consistent with this prediction, A. thaliana plants over-expressing HSP70.1 are hyper-susceptible to H. arabidopsidis, and rice plants activating GRFs show negative effects on defence-gene induction and cell death
. Furthermore, activation of MT1a leads to increased susceptibility to bacterial pathogens and reduced ROS accumulation in Casuarina glauca. Curiously, plants over-expressing END are more sensitive to salt and drought stresses, although their responsiveness to pathogens has not been evaluated
. Therefore, taking these results together, the genes selected by our studies appear to be involved in responses to stress, with some of them acting in cell viability pathways.