Jasmonate-inducible plant enzymes degrade essential amino acids in the herbivore midgut

Abstract

The plant hormone jasmonic acid (JA) activates host defense responses against a broad spectrum of herbivores. Although it is well established that JA controls the expression of a large set of target genes in response to tissue damage, very few gene products have been shown to play a direct role in reducing herbivore performance. To test the hypothesis that JA-inducible proteins (JIPs) thwart attack by disrupting digestive processes in the insect gut, we used a MS-based approach to identify host proteins that accumulate in the midgut of Manduca sexta larvae reared on tomato (Solanum lycopersicum) plants. We show that two JIPs, arginase and threonine deaminase (TD), act in the M. sexta midgut to catabolize the essential amino acids Arg and Thr, respectively. Transgenic plants that overexpress arginase were more resistant to M. sexta larvae, and this effect was correlated with reduced levels of midgut Arg. We present evidence indicating that the ability of TD to degrade Thr in the midgut is enhanced by herbivore-induced proteolytic removal of the enzyme's C-terminal regulatory domain, which confers negative feedback regulation by isoleucine in planta. Our results demonstrate that the JA signaling pathway strongly influences the midgut protein content of phytophagous insects and support the hypothesis that catabolism of amino acids in the insect digestive tract by host enzymes plays a role in plant protection against herbivores.

Fig. 1.

The JA signaling pathway reduces herbivore performance on tomato. M. sexta larvae were grown on jai1, WT, and 35S::PS plants for 7 days, after which larval weights were determined. Data show the mean ± SD of at least 23 larvae per host genotype. Lowercase letters denote significant differences (unpaired Students's t test) at P < 0.001.

Fig. 2.

JA-regulated arginase and TD are active in the M. sexta digestive tract. (A) Arginase and (B) TD activity in midgut extracts from larvae that were grown on jai1, WT, and 35S::PS (PS) plants. Also shown in B is TD activity in frass (fr) from WT-reared larvae. (C) Arg (filled bar), Thr (open bar), and (D) NH₃ levels in midgut extracts from the same larvae used in A and B. Data show the mean ± SD of 10 larvae per plant genotype. Italicized letters denote significant differences (unpaired Students's t test) at P < 0.05.

Fig. 3.

Overexpression of arginase in tomato leaves depletes Arg availability in the M. sexta midgut and increases host resistance. (A) Arginase activity in leaves of unwounded control (Con) and M. sexta-damaged (10 days postchallenge) WT (filled bar) and 35S::ARG (open bar) plants. (B) Results of four (a-d) independent feeding trials of M. sexta on WT (filled bar) and 35S::ARG (open bar) plants. Numbers in parentheses indicate the duration (days) of each trial. Significant differences in larval weights (P < 0.01) were observed in all four trials. The number of larvae in each data set is indicated above the bar. Photograph of WT (C) and 35S::ARG (D) plants after feeding by M. sexta larvae for 10 days. (E) Arginase activity in midgut extracts from larvae reared on WT, 35S::ARG (ARG), and jai1 plants, or on artificial diet (diet). (F) Arg accumulation in midguts from larvae reared on WT and 35S::ARG plants. Data in E and F show the mean ± SD (n = 10 larvae). Italicized letters denote significant differences (unpaired Students's t test) at P < 0.05.

Fig. 4.

Ingested tomato TD is insensitive to negative feedback regulation by isoleucine. (A) TD activity in methyl-JA treated leaves (leaf), midguts of WT-reared larvae (midgut), and frass from WT-reared larvae (frass) was measured in the absence (0) or presence of different concentrations (mM) of Ile. Data were normalized to the amount of activity observed in the absence of Ile (100%) and show the mean ± SD of three independent experiments. (B) Complete amino acid sequence of tomato TD (gi 100257). Based on the 3D structure of Escherichia coli TD (30), amino acid sequences corresponding to the N-terminal catalytic (Cat) domain, C-terminal regulatory (Reg) domain, and the short “neck” region that connects the two domains of the homologous tomato enzyme are indicated. LC-MS/MS analysis of tomato flower TD identified amino acid sequences indicated by uppercase letters. Underlined letters (uppercase and lowercase) denote the sequence that was identified in the midgut enzyme. The chloroplast targeting peptide (Tp) (29) is shown by lowercase italicized letters.