Multiple indole glucosinolates and myrosinases defend Arabidopsis against Tetranychus urticae herbivory

Abstract

Arabidopsis (Arabidopsis thaliana) defenses against herbivores are regulated by the jasmonate (JA) hormonal signaling pathway, which leads to the production of a plethora of defense compounds. Arabidopsis defense compounds include tryptophan-derived metabolites, which limit Arabidopsis infestation by the generalist herbivore two-spotted spider mite, Tetranychus urticae. However, the phytochemicals responsible for Arabidopsis protection against T. urticae are unknown. Here, we used Arabidopsis mutants disrupted in the synthesis of tryptophan-derived secondary metabolites to identify phytochemicals involved in the defense against T. urticae. We show that of the three tryptophan-dependent pathways found in Arabidopsis, the indole glucosinolate (IG) pathway is necessary and sufficient to assure tryptophan-mediated defense against T. urticae. We demonstrate that all three IGs can limit T. urticae herbivory, but that they must be processed by myrosinases to hinder T. urticae oviposition. Putative IG breakdown products were detected in mite-infested leaves, suggesting in planta processing by myrosinases. Finally, we demonstrate that besides IGs, there are additional JA-regulated defenses that control T. urticae herbivory. Together, our results reveal the complexity of Arabidopsis defenses against T. urticae that rely on multiple IGs, specific myrosinases, and additional JA-dependent defenses.

Figure 1

Trp-derived metabolites suppress mite feeding and fecundity. A, Experimental set-up for the mite feeding experiment where mites were given the choice to feed on Col-0 or cyp79b2 cyp79b3 (cyp79b2b3) leaves. To track leaves mites fed on, one leaf was supplemented with blue dye and the other remained unstained. Mites feeding on the blue leaf produced blue feces. B, The total number of blue and nonstained feces excreted by 10 mites after 48 h. Asterisks indicate a deviation from a 1:1 ratio of noncolored:blue feces (P ≤ 0.05). C, The effectiveness of Trp-derived metabolites upon mite transfer between Col-0 and cyp79b2 cyp79b3 (cyp79b2b3) leaves. Ten mites were added to the first leaf and 18 h later, they were transferred to the second leaf. The total numbers of feces and eggs were scored on the second leaf 24 h after transfer. Significant differences (P ≤ 0.05) are indicated by different letters. B and C, Experiments were performed in five biological replicates/trial and in three independent trials (n = 15). Data represent the mean ± se.

Figure 2

The contribution of individual Trp-derived metabolic pathways to Arabidopsis defenses against mites. A, Spider mite fecundity upon feeding on Col-0, cyp79b2 cyp79b3 (cyp79b2b3), cyp71a12, cyp71a13, and cyp71a12 cyp71a13 (cyp71a12a13) leaves. The total number of eggs per leaf was recorded 48 h after the addition of 10 mites/leaf. B, Levels of I3M in Col-0 and cyp79b2 cyp79b3 (cyp79b2b3) leaves supplemented with I3M and infested with 10 mites. −I3M/−mite, untreated leaves were immediately frozen after being cut from intact plant; −I3M/+mite, leaves kept in water only for 22 h and then challenged with mites for 48 h; +I3M/+mite, leaves supplemented with solution of 2.4 mM I3M for 6 h and kept in water for 16 h, and then challenged with mites for 48 h. C, The effect of I3M supplementation to Col-0 and cyp79b2 cyp79b3 (cyp79b2b3) leaves on total number of eggs laid by 10 mites 48 h after the infestation. −I3M, leaves kept in water; +I3M, leaves supplemented with solution of 2.4 mM I3M for 6 h and kept in water for 16 h before mite addition. A–C, Experiments were performed in at least five biological replicates/trial and in three independent trials (n ≥ 15). Data represent the mean ± se. Significant differences (P ≤ 0.05) are indicated by different letters.

Figure 3

Supplementation with I3M, 1MO-I3M, or 4MO-I3M IGs fully restores defenses in cyp79b2 cyp79b3 leaves. A, Levels of I3M, 1MO-I3M, and 4MO-I3M in Col-0 and cyp79b2 cyp79b3 (cyp79b2b3) leaves supplemented with IGs and infested with mites. water/−mite, untreated leaves immediately frozen after being cut from intact plant; +mite, leaves challenged with mites for 48 h. Leaves were supplemented with 2.4 mM I3M, 1MO-I3M, or 4MO-I3M. Values were log2 transformed for statistical analysis. B, Mite fitness upon feeding on Col-0 and cyp79b2 cyp79b3 (cyp79b2b3) leaves supplemented with 2.4 mM I3M, 1MO-I3M, or 4MO-I3M. The total numbers of deposited feces (top) and eggs (bottom) were recorded 48 h after the addition of 10 mites per leaf. Experiments were performed in five biological replicates/trial and in four independent trials (n = 20). A and B, Data represent the mean ± se. Significant differences (P ≤ 0.05) are indicated by different letters.

Figure 4

Intact I3M, 1MO-I3M, and 4MO-I3M are not efficient against mites. A, Mite fecundity upon direct application of 0.23, 2.3, or 4.6 mM I3M to mites. Mites were treated for 19 h with I3M solutions and were subsequently transferred to bean leaves. Mite fecundity was determined at 24 h and 48 h after treatment. B, The total numbers of feces and eggs deposited over 24 h of feeding on bean leaf disk treated with 2.4 mM of I3M, 1MO-I3M, or 4MO-I3M. Experiments were performed in three (in A) and five (in B) biological replicates per trial and in three independent trials (n = 9 (in A), and n = 15 (in B)). A and B, Data represent the mean ± se. Significant differences (P ≤ 0.05) are indicated by different letters.

Figure 5

The Arabidopsis myrosinases AtTGG1 and AtTGG2 are required for Arabidopsis defense against mites. A, Fecundity of 10 mites upon feeding on Col-0 and tgg1 tgg2 leaves for 48 h. B–D, Comparison of leaf damage and mite fitness upon feeding on Col-0, tgg1 tgg2, and cyp79b2 cyp79b3 (cyp79b2b3) leaves. B, Leaf damage resulting from feeding of 10 mites per plant over 3 d. C, Time required for larvae to become nymphs. D, Larval mortality. E, Fecundity of 10 mites upon feeding on Col-3 gl1 and pen2-1 leaves for 48 h. Experiments were performed in 5 (in A, C–E) and 10 (in B) biological replicates per trial and in three independent trials (n = 15 (in A, C–E), and n = 30 (in B)). A–E, Data represent the mean ± se. Significant differences (P ≤ 0.05) are indicated by different letters.

Figure 6

IGs levels decrease in mite-infested leaves at early time points. Levels of I3M, 1MO-I3M, and 4MO-I3M in attached Col-0 leaves infested with 10 mites for 3 h. −mite, unchallenged leaves; +mite, leaves challenged with mites for 3 h. Experiments were performed in three biological replicates per trial and in two independent trials (n = 6). Data represent the mean ± se. Significant difference (P ≤ 0.001) is indicated by asterisks.

Figure 7

Intact IGs and their putative breakdown products detected in mite-infested cyp79b2 cyp79b3 leaves supplemented with 2.4 mM of I3M, 1MO-I3M, or 4MO-I3M. Detached cyp79b2 cyp79b3 leaves were administered water or 2.4 mM solutions of I3M, 1MO-I3M, or 4MO-I3M, and infested with mites for 48 h. Metabolites were extracted and analyzed by HPLC-MS. Metabolite data were exported into MZmine for untargeted analysis. RT; sd, standard deviation; nd, not detected; tr, traces (relative area below 0.1); DHA, dihydroascorbigen.

Figure 8

AtMYC2 AtMYC3 AtMYC4 are required for IG-mediated and Trp-independent defenses against mites in Arabidopsis. A, Fecundity of 10 mites feeding for 48 h on Col-0, myc2 myc3 myc4 (myc234), and cyp79b2 cyp79b3 (cyp79b2b3) leaves. B, Levels of I3M, 1MO-I3M, and 4MO-I3M in Col-0 and myc2 myc3 myc4 (myc234) leaves supplemented with 2.4 mM I3M for 6 h and kept in water for 16 h before mite addition. −I3M/−mite, untreated leaves were immediately frozen after being cut from intact plant; −I3M/+mite, leaves kept in water only for 22 h and then challenged with mites for 48 h; +I3M/+mite, leaves supplemented with solution of 2.4 mM I3M for 6 h and kept in water for 16 h, and then challenged with mites for 48 h. C, Fecundity of 10 mites upon feeding for 48 h on Col-0, cyp79b2 cyp79b3 (cyp79b2b3), and myc2 myc3 myc4 (myc234) leaves. myc2 myc3 myc4 leaves were supplemented with 4.8 mM I3M or 1MO-I3M for 24 h before mite addition. A–C, Experiments were performed in five biological replicates per trial and in three independent trials (n = 15). Data represent the mean ± se. Significant differences (P ≤ 0.05) are indicated by different letters.