Nonhost resistance of barley to different fungal pathogens is associated with largely distinct, quantitative transcriptional responses

Abstract

Nonhost resistance protects plants against attack by the vast majority of potential pathogens, including phytopathogenic fungi. Despite its high biological importance, the molecular architecture of nonhost resistance has remained largely unexplored. Here, we describe the transcriptional responses of one particular genotype of barley (Hordeum vulgare subsp. vulgare 'Ingrid') to three different pairs of adapted (host) and nonadapted (nonhost) isolates of fungal pathogens, which belong to the genera Blumeria (powdery mildew), Puccinia (rust), and Magnaporthe (blast). Nonhost resistance against each of these pathogens was associated with changes in transcript abundance of distinct sets of nonhost-specific genes, although general (not nonhost-associated) transcriptional responses to the different pathogens overlapped considerably. The powdery mildew- and blast-induced differences in transcript abundance between host and nonhost interactions were significantly correlated with differences between a near-isogenic pair of barley lines that carry either the Mlo wild-type allele or the mutated mlo5 allele, which mediates basal resistance to powdery mildew. Moreover, during the interactions of barley with the different host or nonhost pathogens, similar patterns of overrepresented and underrepresented functional categories of genes were found. The results suggest that nonhost resistance and basal host defense of barley are functionally related and that nonhost resistance to different fungal pathogens is associated with more robust regulation of complex but largely nonoverlapping sets of pathogen-responsive genes involved in similar metabolic or signaling pathways.

Figure 1.

PCA of transcript profiles of different plant inoculations and tissues. Mean values from three to four biological replicates of the 1,667 unigenes corresponding to differentially accumulating transcripts upon pathogen attack were used for the analysis. The clusters of noninoculated samples (from mock-inoculated or nontreated, control plants) are highlighted by gray shading. PC1 and PC2 reflect tissue type and treatment (inoculation), respectively. h, Hours after inoculation. For abbreviations of pathogens, see Table I.

Figure 2.

Nonhost resistance of barley to fungal pathogens is associated with largely nonoverlapping sets of pathogen-responsive transcripts. A, Venn diagram of transcripts differentially regulated during the interaction of barley with PM and Mag (left panel) or with PM and Rust (right panel). Transcripts regulated by both pathogen genera are shown inside the thick frame. The set of transcripts regulated in a nonhost-specific manner by the two genera are highlighted by gray shading. For abbreviations of pathogens, see Table I. B, The overlap of transcripts (relative to one or the other pathogen as specified below the column) that were generally regulated during interactions with PM and Mag in the epidermis or with PM and Rust in entire leaf samples (white bars) was compared with the overlap found with transcripts regulated specifically during nonhost-resistant interactions.

Figure 3.

Differentially regulated transcripts between host and nonhost interactions. A hierarchical clustering of pathogen-regulated transcripts was performed that showed a significant (P < 0.05, q < 0.1) quantitative difference of expression in leaf epidermis between Bgh and Bgt or between TH and CD. Per time point and treatment, mean values of signal intensities from three to four biological replicates were calculated, and Pearson correlation with complete linkage of log₂-transformed, median-centered mean signal intensities was applied for the clustering. Color code is as follows: blue, down-regulation; yellow, up-regulation.

Figure 4.

Stronger regulation of transcript abundance in nonhost-resistant interactions. Distribution of the DI for the interaction-specific marker transcripts identified in barley epidermis (according to Supplemental Table S2) is shown. The transcripts were grouped according to the regulons highlighted in Figure 3. DI values of genes with up-regulated or down-regulated transcript abundance (black or white symbols, respectively) were sorted in decreasing order. Positive DI values of genes with up-regulated transcript abundance indicate a stronger transcript accumulation during nonhost interactions, whereas negative DI values of genes with down-regulated transcript abundance reflect stronger repression of gene expression during nonhost interactions. Absolute DI values of 0.2 and 0.3 correspond to 1.5- and 2-fold average differences, respectively, of transcript abundance between host or nonhost interactions.

Figure 5.

Different pathogens induce similar changes in signaling or metabolic pathways. Statistical significance of the overrepresentation or underrepresentation of regulated transcripts belonging to a specific functional transcript category was calculated relative to the entire cDNA array (χ² test). Superbin nomenclature was taken from Sreenivasulu et al. (2008). Pink and red bars indicate genes with up-regulated transcript abundance; light and dark green bars indicate genes with down-regulated transcript abundance. Statistical significance is indicated with asterisks: * P < 0.05, ** P < 0.005, *** P < 0.0005. AA, Amino acid; Asc, ascorbate; CHO, carbohydrate; epi, epidermis; Gluth, glutathione; n.a., not analyzed; OPP, oxidative pentose phosphate; PR, pathogenesis related. For abbreviations of pathogens, see Table I.

Figure 6.

Model of selective suppression of host defense genes by different pathogens. The observed small overlap of the nonhost-specific part of the transcriptional response of barley may be due to pathogen-specific suppression of basal defense, which is also under negative control by the Mlo gene, by the different adapted host pathogens. Despite differences in the sets of regulated transcripts responding to different adapted or nonadapted pathogens, pathways and cellular responses triggered by all types of pathogens were converging into similar defense reactions.