MATE Transporter-Dependent Export of Hydroxycinnamic Acid Amides

Abstract

The ability of Arabidopsis thaliana to successfully prevent colonization by Phytophthora infestans, the causal agent of late blight disease of potato (Solanum tuberosum), depends on multilayered defense responses. To address the role of surface-localized secondary metabolites for entry control, droplets of a P. infestans zoospore suspension, incubated on Arabidopsis leaves, were subjected to untargeted metabolite profiling. The hydroxycinnamic acid amide coumaroylagmatine was among the metabolites secreted into the inoculum. In vitro assays revealed an inhibitory activity of coumaroylagmatine on P. infestans spore germination. Mutant analyses suggested a requirement of the p-coumaroyl-CoA:agmatine N4-p-coumaroyl transferase ACT for the biosynthesis and of the MATE transporter DTX18 for the extracellular accumulation of coumaroylagmatine. The host plant potato is not able to efficiently secrete coumaroylagmatine. This inability is overcome in transgenic potato plants expressing the two Arabidopsis genes ACT and DTX18. These plants secrete agmatine and putrescine conjugates to high levels, indicating that DTX18 is a hydroxycinnamic acid amide transporter with a distinct specificity. The export of hydroxycinnamic acid amides correlates with a decreased ability of P. infestans spores to germinate, suggesting a contribution of secreted antimicrobial compounds to pathogen defense at the leaf surface.

Figure 1.

Coumaroylagmatine Accumulates Extracellularly in Arabidopsis, But Not in Potato, and Inhibits P. infestans Spore Germination in Vitro. (A) and (B) Coumaroylagmatine content of leaves (A) and in droplets of the inoculum (B) from Arabidopsis and potato plants 24 h after inoculation with P. infestans. Data shown were obtained in at least five independent experiments (Arabidopsis, n ≥ 107; potato, n = 56). Error bars represent se. Significance analysis of differences was performed by t test, ***P < 0.001. fw, fresh weight. (C) Mycelial growth inhibition assays. Coumaroylagmatine was added to GFP-expressing P. infestans (isolate Cra208m2) mycelium at the indicated concentrations. GFP fluorescence was determined at the time points indicated (control, n = 5; 1 mM, n = 10; 10 mM, n = 9). Data shown are representative for at least three independent experiments. (D) Spore germination inhibition assay. Coumaroylagmatine was added to a P. infestans zoospore suspension (10⁵/mL⁻¹) at the indicated concentrations. After 24 h, the samples were evaluated microscopically for spore germination. Data are derived from two independent experiments (n = 4). Significance analysis of differences was performed by t test, *P < 0.05 and ***P < 0.001.

Figure 2.

P. infestans-Induced Expression of ACT and DTX18 in Arabidopsis. (A) and (B) Microarray analyses. ACT (A) and DTX18 (B) expression in response to inoculation with P. infestans (black bars) or water (white bars). Data were obtained from AtGenExpress (www.arabidopsis.org; n = 3). hpi, hours after inoculation. (C) and (D) Time course of ACT (C) and DTX18 (D) expression in response to inoculation with P. infestans. qRT-PCR was performed with RNA isolated from P. infestans-inoculated Arabidopsis (black bars) or water-treated plants (white bars) at the time points indicated. At-PP2A transcript levels were used for standardization. Data are derived from three independent experiments (n > 5). Error bars represent se. Significance analysis of differences was performed by t test, *P < 0.05, **P < 0.01, and ***P < 0.001.

Figure 3.

Reduced Levels of HCAAs in the act Mutant. Coumaroylagmatine (A) and feruloylagmatine (B) levels were determined by UPLC-ESI-QTOF-MS in methanolic extracts of leaves of wild-type and act lines 24 h after inoculation with P. infestans or water. Data are derived from four independent experiments (wild type, water, n = 27; P. infestans, n = 95; act, water, n = 12; P. infestans, n = 23). Error bars represent se. Letters indicate statistically different values (two-way ANOVA, P < 0.05).

Figure 4.

dtx18 Knockout Lines Are Unable to Secrete p-Coumaroylagmatine. (A) and (B) Characterization of T-DNA insertion lines of DTX18. DNA was isolated from leaves of wild-type plants or the T-DNA insertion lines SALK_062231 and GK_411D06. PCR was performed using specific primers for the alleles from the wild type, SALK_062231 (S), or GK_411D06 (G). C, water control. (C) Loss of DTX18 gene expression. RT-PCR using DTX18-specific primers was performed using RNA from leaves of wild-type plants and the T-DNA insertion lines SALK_062231 (S) and GK_411D06 (G). Genomic DNA (g) was used as a control. (D) and (E) Reduced extracellular levels of coumaroylagmatine in dtx18 mutant lines. Coumaroylagmatine levels were determined in leaves (D) and in droplets of the inoculum (E) from Arabidopsis wild-type and mutant plants (SALK_062231 and GK_411D06) 24 h after inoculation with P. infestans. Data shown were obtained in at least three independent experiments (wild type, n ≥ 106; act, n = 21; SALK_062231, n = 47; GK_411D06, n = 8). Significance analysis of differences was performed by t test, *P < 0.05, **P < 0.01, and ***P < 0.001. fw, fresh weight.

Figure 5.

Expression of ACT and DTX18 in Transgenic Potato Plants. (A) and (B) qRT-PCR analysis for ACT (black bars) or DTX18 expression (gray bars) was performed with RNA from leaves of wild-type and transgenic potato plants carrying an empty vector construct (ev) or expressing 35S-ACT (lines B, H, M, and Z) (A) or 35S-ACT-35S-DTX18 (lines D, K, N, and X) (B). Expression of EF1a was used as a reference. Data are derived from two independent experiments ([A]: wild type, n = 7; empty vector, n = 5; B, M, H, and Z, n = 3; [B]: wild type, n = 8; empty vector, n = 6; D, K, N, and X, n = 3). Error bars represent se. nd, not detectable. (C) Phenotype of 35S-ACT-35S-DTX18 potato plants.

Figure 6.

Characterization of ACT- and ACT-DTX18-Expressing Potato Plants. Coumaroylagmatine levels were determined in leaves ([A] and [B]) and in droplets of the inoculum ([C] and [D]) from wild-type and transgenic potato plants carrying an empty vector construct (ev) or 35S-ACT (lines B, H, M, and Z) ([A] and [C]) or 35S-ACT-35S-DTX18 (lines D, K, N, and X) ([B] and [D]). Significance analysis of differences was performed by t test, *P < 0.05, **P < 0.01, and ***P < 0.001. (A) Two independent experiments; wild type, empty vector, n = 10; B, n = 5; H, M, and Z, n = 6. (B) Three independent experiments; wild type, empty vector, n = 18; D2, n = 7; K3, n = 6; N3, n = 7; X2, n = 8. (C) Two independent experiments; wild type, empty vector, n = 10; lines B, H, M, and Z, n = 6. (D) Three independent experiments; wild type, empty vector, n = 18; D2, n = 7; K3, n = 5; N3, n = 6; X2, n = 7.

Figure 7.

Expression of ACT Leads to Alterations in HCAA Patterns in P. infestans-Infected Potato Leaves. HCAA levels were determined in leaves (A) and in the inoculum (B) 24 h after infection of leaves of control (wild-type and empty vector-carrying plants) or 35S-ACT plants (lines B, H, M, and Z). Data are derived from three independent experiments ([A]: n ≥ 32, except for spermidine derivatives, n ≥ 15; [B]: n ≥ 32, except for feruloyltyramine, n ≥ 20, and caffeoylagmatine, n ≥ 8). Significance analysis of differences was performed by t test, *P < 0.05, **P < 0.01, and ***P < 0.001.

Figure 8.

Expression of ACT and DTX18 Leads to Extracellular Accumulation of HCAAs in P. infestans-Infected Potato Leaves. HCAA levels were determined in leaves (A) and in the inoculum (B) 24 h after infection of leaves of control (wild-type and empty vector-carrying plants) or 35S-ACT-35S-DTX18-expressing plants (lines D, K, N, and X). Data are derived from three independent experiments ([A]: n ≥ 44; [B]: n ≥ 36, except for feruloyltyramine and caffeoylputrescine, n ≥ 24). Significance analysis of differences was performed by t test, *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 9.

Defense of ACT and ACT-DTX18-Expressing Potato Plants against P. Infestans. (A) and (B) Determination of pathogen biomass. Wild-type and transgenic potato plants carrying an empty vector construct (ev), 35S-ACT (lines B, H, M, and Z) (A), or 35S-ACT-35S-DTX18 (lines D, K, N, and X) (B) were inoculated with P. infestans. Pathogen biomass was determined 3 d after inoculation. Error bars represent se. Significance analysis of differences was performed by t test, *P < 0.05 and ***P < 0.001. P values for comparison of the ACT-DTX18-expressing lines: D-K, 0.098; D-N, 0.17; D-X, 0.57; K-N, 0.84; K-X, 0.06; N-X, 0.13. (A) Three independent experiments; wild type, empty vector, n = 24; B, H, M, and Z, n = 12. (B) Five independent experiments; wild type, n = 35; empty vector, n = 43; D, n = 17; K, n = 13; N, n = 10; X, n = 18. (C) Spore germination inhibition assay. P. infestans inoculum incubated on wild-type and transgenic potato plants carrying an empty vector construct (ev) or expressing 35S-ACT-35S-DTX18 (lines D, K, N, and X) were collected after 24 h, sterile filtrated, and added to a zoospore suspension of P. infestans. Inhibitory activity was determined by microscopy analyses. Data were obtained in three independent experiments (n = 9). At least 1300 spores were evaluated for each line. Significance analysis of differences was performed by t test, *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 10.

Hypothetical Model of DTX18-Mediated Coumaroylagmatine Secretion. Potato responds to infection by P. infestans with the biosynthesis of coumaroyl- and feruloyltyramine by THT (Schmidt et al., 1998) and their export by an unknown transport system. The extracellular accumulation of novel HCAAs in transgenic potato plants expressing DTX18 from Arabidopsis, predicted to encode a plasma membrane-localized MATE transporter, suggests that DTX18 is able to export coumaroyl, caffeoyl, and feruloyl conjugates of both agmatine and putrescine.