Gpa2 detects the potato cyst nematode effector GpRBP-1 in the cytoplasm but requires a balanced nucleocytoplasmic distribution to trigger cell death

Abstract

The potato immune receptor Gpa2 confers host-specific resistance to the cyst nematode Globodera pallida. When transiently expressed in Nicotiana benthamiana it triggers cell death upon recognition of the matching effector GpRBP-1. Effector-triggered immunity by Gpa2 depends on the host factor RanGAP2, which is known to regulate the nucleocytoplasmic distribution and functioning of the highly homologous potato immune receptor Rx1. However, the subcellular localisation of Gpa2 and the role of RanGAP2 in determining the subcellular localisation of Gpa2 is not yet known. Moreover, the cellular mechanisms underlying detection of the nematode effector by Gpa2 and the subsequent activation of cell death also remain unknown. Here, we co-expressed Solanum tuberosum Gpa2 fused to nuclear localisation signals and its matching effector GpRBP-1 in N. benthamiana as a model to address these questions. The results indicated that both the nuclear and cytoplasmic pools of Gpa2 contribute to effector-triggered cell death and this depends on the formation of a complex with RanGAP2, which acts as a cytoplasmic retention and stabilising factor. However, using nuclear and cytoplasmic targeting signals, we found that detection of GpRBP-1 by Gpa2 occurs specifically in the cytoplasm. Based on these results, we propose that RanGAP2 retains Gpa2 in the cytoplasm to form a pre-activation complex that aids in the detection of GpRBP-1 and the activation of immune responses in a compartment-specific manner.

Keywords: Globodera pallida; Effector recognition; GpRBP-1; Gpa2; NLR; RanGAP2; hypersensitive response; immune receptor; plant immunity; potato cyst nematode; subcellular localisation.

Fig. 1.

Recognition of GpRBP-1 by Gpa2 occurs in the cytoplasm. (A) GFP-GpRBP-1 driven by the CaMV 35S promotor was fused to a nuclear localisation signal (NLS), a nuclear export signal (NES) or to mutated non-functional nls or nes constructs, and transiently expressed in N. benthamiana leaf cells. Confocal images taken at 2 days post infiltration (dpi) are shown, representative of three independent repeats. C, cytoplasm; N, nucleus. (B) Representative images of infiltrated leaf spots transiently co-expressing GPAII::Gpa2 with the GFP-GpRBP-1 constructs at 7 dpi. (C) Boxplots of the hypersensitive response induced by GpRBP-1 in the infiltrated spots at 7 dpi, quantified as chlorophyll loss relative to untreated areas of the leaf. Data are from n≥8 leaf spots from two independent experiments. Different letters indicate significant differences among means as determined using Wilcoxon signed-rank tests (P<0.05)

Fig. 2.

Cytoplasmic targeting of Gpa2 compromises cell death induced by GpRBP-1. (A) GFP-Gpa2 driven by the GPAII promotor was fused to either a nuclear localisation signal (NLS) or a nuclear export signal (NES) and transiently expressed in N. benthamiana leaf cells. Mutated non-functional constructs (nls, nes) were also used. Confocal images taken at 2 days post infiltration (dpi) are shown, representative of three independent repeats. (B) Representative images of infiltrated leaf spots transiently co-expressing GFP-Gpa2 constructs with or without the matching effector GpRBP-1 at 3 dpi. OD₆₀₀ for GpRBP-1 = 0.1, for GPAII::Gpa2 with GpRBP-1 = 0.3, and GPAII::Gpa2 without GpRBP-1 = 1.0. (C) Boxplots of the hypersensitive response induced by GpRBP-1 in the infiltrated spots at 3 dpi, as quantified by red light imaging relative to adjacent untreated areas of the leaf. Data are from n≥20 leaf spots from at least three independent experiments. (D) Boxplots of the relative fluorescence intensity ratios between the cytoplasmic pool (C) and nuclear pool (N) of the different constructs. Data are from 2–4 biological replicates. Different letters indicate significant differences among means as determined using Wilcoxon signed-rank tests (P<0.05).

Fig. 3.

Overexpression of RanGAP2 has no impact on Gpa2 localisation and cell death, but the nuclear-targeting of Gpa2 via its WPP domain reduces cell death induced by GpRBP-1. (A) Confocal images of co-localisation of different GFP-Gpa2 and RanGAP2 constructs in N. benthamiana leaf cells at 2 days post infiltration (dpi). mCH, mCherry; NLS, nuclear localisation signal; nls, non-functional mutated form of NLS. Representative images from n=10 cells in two independent repeats are shown. (B) Representative images of infiltrated leaf spots transiently co-expressing GpRBP-1 with the different constructs at 3 dpi. (C) Boxplots of the relative fluorescence intensity ratios between the cytoplasmic pool (C) and the nuclear pool (N) of the different constructs. Data are from 3–5 biological replicates. (D) Boxplots of the hypersensitive response induced by GpRBP-1 in the infiltrated spots at 3 dpi, as quantified by red light imaging relative to adjacent untreated areas of the leaf. Data are from n≥12 leaf spots from three independent experiments. Different letters indicate significant differences among means as determined using Wilcoxon signed-rank tests (P<0.05).

Fig. 4.

Co-expression of RanGAP2 and different Gpa2 subdomains results in an altered nucleocytoplasmic distribution. The Gpa2 N-terminal CC domain, the NB-ARC domain, and the C-terminal LRR-domain were fused with GFP and expressed in N. benthamiana leaf cells either with or without co-expression of the RanGAP2-mCherry (mCH) construct. (A) Confocal images taken at 2 days post infiltration (dpi), representative of five cells in two independent repeats. (B) Immunoblot of extracted proteins upon co-expression of the different Gpa2 subdomains and RanGAP2 at 3 dpi. The subdomains were captured with anti-GFP (α-GFP)-conjugated magnetic beads as bait and the bound proteins were analysed by immunoblotting. The presence of a protein in the infiltration combination is indicated, with expression of all input proteins shown on the left and the presence after pull-down shown on the right. Rubisco stained with Coomassie Brilliant Blue (CBB) served as a loading control.

Fig. 5.

Both RanGAP2 presence and activity are required for a balanced Gpa2 subcellular distribution and for its correct function. Virus-induced gene-silencing with Tobacco rattle virus (TRV) was used to silence RanGAP2 (TRV:: RanGAP2) in N. benthamiana, with TRV::GUS as a negative control. At 3 weeks after silencing, GFP-Gpa2 was transiently expressed in the leaves via agro-infiltration. (A) Confocal images of GFP-Gpa2 localisation in the leaves at 2 days post infiltration (dpi). The images are representative of a total 15 cells that were imaged in three independent experiments. (B) Boxplots of the relative fluorescence intensity ratios between the cytoplasmic pool (C) and the nuclear pool (N) of the control and silenced plants, calculated from the 15 cells that were imaged. (C) Immunoblot of the detection of GFP-Gpa2 at 2 dpi using the anti-GFP antibody (α-GFP). Rubisco stained with Coomassie Brilliant Blue (CBB) served as a loading control. (D) At 3 weeks after silencing, N. benthamiana leaves were complemented with either RanGAP2 (Rg2) or the RanGAP2(D335N) mutant [Rg2(D335N)]. These RanGAP2 constructs were transiently co-expressed with GpRBP-1 and Gpa2, and the resulting hypersensitive response was quantified by red light imaging at 3 dpi. The boxplots show the response in the infiltrated spots relative to adjacent untreated areas of the leaf. Data are from n≥12 leaf spots from three independent experiments. Different letters indicate significant differences among means as determined using Wilcoxon signed-rank tests (P<0.05). (E) Representative images of the cell-death response for the treatments shown in (D).