Abscisic acid deficiency increases defence responses against Myzus persicae in Arabidopsis

Abstract

Comparison of Arabidopsis thaliana (Arabidopsis) gene expression induced by Myzus persicae (green peach aphid) feeding, aphid saliva infiltration and abscisic acid (ABA) treatment showed a significant positive correlation. In particular, ABA-regulated genes are over-represented among genes that are induced by M. persicae saliva infiltration into Arabidopsis leaves. This suggests that the induction of ABA-related gene expression could be an important component of the Arabidopsis-aphid interaction. Consistent with this hypothesis, M. persicae populations induced ABA production in wild-type plants. Furthermore, aphid populations were smaller on Arabidopsis aba1-1 mutants, which cannot synthesize ABA, and showed a significant preference for wild-type plants compared with the mutant. Total free amino acids, which play an important role in aphid nutrition, were not altered in the aba1-1 mutant line, but the levels of isoleucine (Ile) and tryptophan (Trp) were differentially affected by aphids in wild-type and mutant plants. Recently, indole glucosinolates have been shown to promote aphid resistance in Arabidopsis. In this study, 4-methoxyindol-3-ylmethylglucosinolate was more abundant in the aba1-1 mutant than in wild-type Arabidopsis, suggesting that the induction of ABA signals that decrease the accumulation of defence compounds may be beneficial for aphids.

Keywords: Arabidopsis; Myzus persicae; abscisic acid; amino acids; aphid choice; glucosinolates.

Figure 1

Comparison of the effects of aphid saliva infiltration, abscisic acid (ABA) treatments and aphid feeding on the Arabidopsis transcriptome. (a) Diagram showing all the differentially expressed genes, either induced (red) or repressed (green), in response to Myzus persicae saliva in Arabidopsis, and the regulation of these genes by ABA treatment and aphid feeding. Different ABA treatments are summarized in one column. Results for aphid feeding for 48 and 72 h are shown individually. Yellow boxes indicate variable expression in response to ABA in different reports. Black boxes indicate no effect of the particular treatment on gene expression. (b) Summary of the results from the comparison of aphid saliva treatment versus ABA treatment presented in (a).

Figure 2

Aphid feeding increases Arabidopsis abscisic acid (ABA) content. Myzus persicae were caged on individual Arabidopsis leaves for 3 days before leaves were harvested for ABA analysis. Control leaves received empty cages. Mean ± standard error (SE) of n = 6 (control) and n = 5 (M. persicae). *P < 0.05, t‐test.

Figure 3

Aphid performance on abscisic acid (ABA) mutants. Four‐week‐old wild‐type and mutant Arabidopsis plants were infested with 10 aphids (adult apterae) and the number of aphids per plant was quantified after 1 week. Statistical analysis was performed using a t‐test assuming unequal variance. Results are the average of eight independent experiments, with at least 12 plants per treatment in each experiment. Error bars indicate standard error (SE).

Figure 4

Aphids choose between wild‐type and aba1‐1  Arabidopsis plants. Wild‐type and aba1‐1 mutant plants were grown in the same pot. When plants were 4 weeks old, one plant of each pair was infested with 10 aphids. Aphids were allowed to move freely between plants in each pot. Controls had two plants of the same genotype. Aphids on each plant were quantified after 1 week of infestation. The number of aphids on the originally infested plant is shown to the left and the number of aphids on the originally non‐infested plant is shown to the right. Statistical analysis was performed using paired t‐tests. Results are the average of three independent experiments. Error bars indicate standard error (SE).

Figure 5

Aphid choice is not ecotype dependent. Wild‐type and aba mutant plants were grown in different pots. When plants were 4 weeks old, pots were arranged as indicated for setup 1 or setup 2 (see Fig. S1). Thirty aphids were placed on filter paper connected to the pots, at the same distance from each plant. Aphids were allowed to move freely among plants in each arrangement. The number of aphids on each plant was determined after 48 h. Only adult apterae were quantified to eliminate any difference caused by variable reproduction on wild‐type or mutant plants. Statistical analysis was performed using paired t‐tests. Results are the average of five independent experiments. Error bars indicate standard error (SE).

Figure 6

Free amino acid levels in wild‐type and aba1‐1 plants. Four‐week‐old wild‐type Ler and aba1‐1 mutant plants were mock treated or infested with aphids. After 1 week of infestation, aphids were removed, leaves were collected and amino acids were analysed by gas chromatography‐flame ionization detection (GC‐FID). Only amino acids for which significant differences were found in response to aphid feeding are shown. Statistical analysis was performed using a t‐test assuming unequal variance. Only wild‐type plants showed significant differences in response to aphids. Results are the average of two independent experiments with four replicates in each experiment. Error bars indicate standard error (SE). ILE, isoleucine; TRP, tryptophan; WT, wild‐type.

Figure 7

Glucosinolate content in wild‐type and aba1‐1 plants, with and without aphids. Four‐week‐old wild‐type and aba1‐1 mutant plants were mock treated or infested with aphids. After 1 week of infestation, aphids were removed and stems and flowers were collected for glucosinolate analysis by high‐performance liquid chromatography (HPLC). Only glucosinolates with significant changes are shown. (a) I3M in stems. (b) 4MI3M in stems. (c) I3M in flowers. (d) 4MI3M in flowers. I3M, indol‐3‐ylmethylglucosinolate; 4MI3M, 4‐methoxyindol‐3‐ylmethylglucosinolate. Mean ± standard error (SE) of n = 10. Different letters indicate significant differences, P < 0.05, Tukey–Kramer honestly significant difference (HSD) test.