Differential effector gene expression underpins epistasis in a plant fungal disease

Abstract

Fungal effector-host sensitivity gene interactions play a key role in determining the outcome of septoria nodorum blotch disease (SNB) caused by Parastagonospora nodorum on wheat. The pathosystem is complex and mediated by interaction of multiple fungal necrotrophic effector-host sensitivity gene systems. Three effector sensitivity gene systems are well characterized in this pathosystem; SnToxA-Tsn1, SnTox1-Snn1 and SnTox3-Snn3. We tested a wheat mapping population that segregated for Snn1 and Snn3 with SN15, an aggressive P. nodorum isolate that produces SnToxA, SnTox1 and SnTox3, to study the inheritance of sensitivity to SnTox1 and SnTox3 and disease susceptibility. Interval quantitative trait locus (QTL) mapping showed that the SnTox1-Snn1 interaction was paramount in SNB development on both seedlings and adult plants. No effect of the SnTox3-Snn3 interaction was observed under SN15 infection. The SnTox3-Snn3 interaction was however, detected in a strain of SN15 in which SnTox1 had been deleted (tox1-6). Gene expression analysis indicates increased SnTox3 expression in tox1-6 compared with SN15. This indicates that the failure to detect the SnTox3-Snn3 interaction in SN15 is due - at least in part - to suppressed expression of SnTox3 mediated by SnTox1. Furthermore, infection of the mapping population with a strain deleted in SnToxA, SnTox1 and SnTox3 (toxa13) unmasked a significant SNB QTL on 2DS where the SnTox2 effector sensitivity gene, Snn2, is located. This QTL was not observed in SN15 and tox1-6 infections and thus suggesting that SnToxA and/or SnTox3 were epistatic. Additional QTLs responding to SNB and effectors sensitivity were detected on 2AS1 and 3AL.

Keywords: NE; Parastagonospora nodorum; SnTox1; SnTox3; Triticum aestivum; epistasis; necrotrophic effector.

Figure 1

The SnTox1–Snn1 interaction is a major SNB determinant in the C × W population. (a) Visual assessment of effector sensitivity and disease symptom on C × W DH lines. SnTox1 and SnTox3 infiltrations and disease symptoms from conidial inoculation of P. nodorum SN15 on DH lines of four genotype groups: Snn1/Snn3, Snn1/snn3, snn1/Snn3, snn1/snn3. Representative DH lines for each genotype are shown. (b) The distribution of SN15, tox1–6 and toxa13 disease severity scores on the C × W population at the seedling stage and SN15 at the adult stage. (c) Statistical analysis was used to compare the average SNB seedling scores between SN15, tox1–6 and toxa13. An analysis of variance (anova) using the Tukey–Kramer test was used to compare all treatments. ^a,bLevels not connected by the same letter are significantly different. SN15–tox1–6 P < 0.001; SN15–toxa13 P = 0.7530; toxa13–tox1–6 P < 0.001. Standard error bars are shown.

Figure 2

Composite interval mapping of quantitative trait loci (QTL) associated with toxa13 CF sensitivity and SNB caused by SN15, tox1–6 and toxa13. Genetic maps of the chromosomes with genetic markers on the right and the centimorgan (cM) distances between loci are shown on the left. Snn1 (bold) was mapped as a trait marker in 1B1 in response to the SnTox1 effector.

Figure 3

The host genotype does not affect the level of SnTox3 expression in SN15 during infection. (a) SnTox1 expression was compared among snn1/Snn3, snn1/snn3, Snn1/Snn3 and Snn1/snn3 C × W DH genotype groups. No significant differences in SnTox1 expression was observed (P = 0.371). (b) No significant differences in SnTox3 expression was observed in all genotype groups (P = 0.396). Average gene expression was calculated from four lines per genotype performed in biological replicates (n = 4). Standard error bars are shown. ‘*’Indicates similar statistical grouping using anova.

Figure 4

SnTox3 expression is elevated in the SnTox1 deletion background. (a) Comparison of SnTox3 expression in P. nodorum tox1–6 during infection of snn1/Snn3, snn1/snn3, Snn1/Snn3 and Snn1/snn3 C × W DH groups. ‘*’Indicates similar statistical grouping using anova (P = 0.808). Average gene expression was calculated from four lines per genotype performed in biological replicates (n = 4). (b) A comparison of SnTox3 expression in SN15 and tox1–6 during infection on different genotype combinations. Average gene expression was calculated from four lines per genotype performed in biological replicates (n = 4). (c) SnTox3 expression in SN15 and tox1–6 during growth in vitro. ‘*’Denotes significant difference in a t‐test. The SnTox3 expression profile was also tested on two other tox1 isogenic strains in vitro (Figure S4). The experiment was performed in biological triplicates. Standard error bars are shown for all graphs.

Figure 5

SnTox3 infiltration increases the virulence of P. nodorum SN15 on Snn3 wheat cultivars. Total fungal DNA was measured from the amplification of Act1 normalised against total DNA. A t‐test was used to compare the amount of fungal DNA between SnTox3‐ and buffer‐infiltrated treatments infected with SN15. ^a,bLevels not connected by the same letter are significantly different (P < 0.05). Standard error bars are shown. The experiment was performed in biological triplicates.