Interaction between domains of a plant NBS-LRR protein in disease resistance-related cell death

Abstract

Many plant disease resistance (R) genes encode proteins predicted to have an N-terminal coiled-coil (CC) domain, a central nucleotide-binding site (NBS) domain and a C-terminal leucine-rich repeat (LRR) domain. These CC-NBS-LRR proteins recognize specific pathogen-derived products and initiate a resistance response that often includes a type of cell death known as the hypersensitive response (HR). Co-expression of the potato CC-NBS-LRR protein Rx and its elicitor, the PVX coat protein (CP), results in a rapid HR. Surprisingly, co-expression of the LRR and CC-NBS as separate domains also resulted in a CP-dependent HR. Likewise, the CC domain complemented a version of Rx lacking this domain (NBS- LRR). Correspondingly, the LRR domain interacted physically in planta with the CC-NBS, as did CC with NBS-LRR. Both interactions were disrupted in the presence of CP. However, the interaction between CC and NBS-LRR was dependent on a wild-type P-loop motif, whereas the interaction between CC-NBS and LRR was not. We propose that activation of Rx entails sequential disruption of at least two intramolecular interactions.

Fig. 1. Complementation of the Rx CC–NBS and LRR regions. Rx constructs are represented schematically. The CC region of Rx is represented by an open box. A black box represents the NBS domain (including the NB and ARC subdomains). The leucine-rich repeat (LRR) region is represented by bars. Versions of Rx were expressed via agroinfiltration (OD₆₀₀ = 0.2 each) in N.benthamiana leaves from the Rx genomic promoter together with either 35S-GUS (left side of leaf) or 35S-CP (right side of leaf). Similar results were obtained when Rx fragments were expressed from the 35S promoter.

Fig. 2. Effect of cis elements on trans complementation. Rx constructs are schematically represented as in Figure 1. The CC–NBS(D460V) construct is identical to CC–NBS shown in Figure 1 except that residue D460 is mutated to V, indicated by an asterisk. (A) The CC– NBS(D460V) or CC–NBS polypeptides were co-expressed with the Rx derivatives indicated on the left, in conjunction with either 35S-GUS or 35S-CP via agroinfiltration in N.benthamiana leaves. (B) Rx derivatives indicated on the left were agroinfiltrated with CP plus either Rx LRR, ARC–LRR or NBS–LRR. The occurrence of a HR or not (–) is indicated. Similar results were obtained using either the Rx or the 35S promoter. No HR was observed with any combination in the absence of CP. (C) Potato leaves (rx genotype) were agroinfiltrated with pB1-Rx derivatives in conjunction with either 35S-CP or GUS, as indicated. Constructs were infiltrated at a concentration of agrobacterium of OD₆₀₀ = 0.2 each.

Fig. 3. Association of the Rx CC–NBS and LRR. (A) Nicotiana benthamiana plants were silenced for SGT1. Leaves were agroinfiltrated with 35S-driven constructs expressing Rx LRR-HA and CC–NBS-MT or GFP (OD₆₀₀ = 0.2 each), as indicated. These were combined with either GFP or GFP–CP32 (OD₆₀₀ = 0.1 each) such that the final OD₆₀₀ = 0.5. Four days post-infiltration, protein extracts were subjected to immunoprecipitation with α-HA (3F10) antibody followed by immunoblotting with α-HA (3F10) antibody (top panel) or α-MT (9E10) antibody (lower panel). (B) Protein extracts were subjected to immunoprecipitation with α-MT (A-14) antibody followed by immunoblotting with α-MT (9E10) antibody (top panel) and α-HA (3F10) antibody (lower panel). (C) SGT1 silenced plants were infiltrated with CC–NBS-MT and LRR-HA (OD₆₀₀ = 0.2 each) plus either GFP, GFP– CP32TK or GFP–CP32KR (OD₆₀₀ = 0.1 each) as indicated. Protein extracts were subjected to immunoprecipitation with α-HA (3F10) antibody and samples immunoblotted with α-HA (3F10) antibody (top panel) and α-MT (A-14) antibody (lower panel). (D) Non- silenced N.benthamiana were agroinfiltrated with 35S-driven constructs expressing Rx CC–NBS-MT in combination with GUS, Bs2 LRR-HA, or Rx LRR-HA. Three days post-infiltration, protein extracts were subjected to immunoprecipitation with α-HA (3F10) antibody followed by immunoblotting with α-HA (3F10) antibody (top panel) and α-MT (A-14) antibody (lower panel).

Fig. 4. Association of the Rx CC and NBS–LRR. (A) SGT1 silenced leaves were agroinfiltrated with 35S-driven constructs expressing Rx CC-HA and Rx NBS–LRR-MT polypeptides or GFP (OD₆₀₀ = 0.2 each). These were combined with either GFP or GFP–CPTK (OD₆₀₀ = 0.1 each) such that the final OD₆₀₀ was 0.5. Four days post-infiltration, protein extracts were subjected to immunoprecipitation with α-HA (3F10) antibody followed by immunoblotting with α-HA (3F10) antibody (top panel) and α-MT (A-14) antibody (middle panel). In the lower panel, protein extracts were subjected to immunoprecipitation with α-MT (9E10) antibody and subsequently immunoblotted with α-MT (A-14) antibody to demonstrate the presence of equal amounts of input NBS–LRR-MT protein. (B) SGT1 silenced plants were infiltrated with NBS–LRR-MT and CC-HA (OD₆₀₀ = 0.2 each) plus GFP, GFP–CP32TK or GFP–CP32KR (OD₆₀₀ = 0.1 each) as indicated. Protein extracts were subjected to immunoprecipitation with α-HA (3F10) antibody and samples immunoblotted with α-HA (3F10) antibody (top panel) and α-MT (A-14) antibody (lower panel). (C) Non- silenced N.benthamiana leaves were agroinfiltrated with 35S-driven constructs expressing Rx NBS–LRR-MT in combination with GUS, Bs2 CC-HA or Rx CC-HA. Three days post-infiltration, protein extracts were immunoprecipitated with α-HA (3F10) antibody and immunoblotted with α-HA (3F10) antibody (top panel) and α-MT (A-14) antibody (lower panel).

Fig. 5. Involvement of the NB domain in interdomain interactions. (A) Nicotiana benthamiana leaves were agroinfiltrated with 35S GUS (lanes 1 and 3) or 35S Rx CC-HA (lanes 2 and 4) along with wild-type Rx NBS–LRR-MT (lanes 1 and 2) or NBS–LRR-MT K1 (lanes 3 and 4). Two days post-infiltration, protein extracts were subjected to immunoprecipitation with α-HA (3F10) antibody and subsequently immunoblotted with α-HA (3F10) antibody (top panel) and α-MT (A-14) antibody (middle panel). In the lower panel, protein extracts were subjected to immunoprecipitation with α-MT (9E10) antibody and immunoblotted with α-MT (A-14) antibody to demonstrate the presence of equal amounts of input NBS–LRR-MT derivatives. (B) Nicotiana benthamiana leaves were agroinfiltrated with 35S GUS (lanes 1 and 3) or 35S Rx LRR-FH (lanes 2, 4 and 5) along with wild-type Rx CC–NBS-HA (lanes 1 and 2) or CC–NBS-HA K1 (lanes 3–5). These were combined with either GUS (lanes 1–4) or CP (lane 5). Two days post-infiltration, protein extracts were subjected to immunoprecipitation with α-FLAG (M2) antibody and subsequently immunoblotted with α-FLAG (M2) antibody (top panel) and α-HA (3F10) antibody (middle panel). In the lower panel, protein extracts were subjected to immunoprecipitation with α-HA (3F10) antibody and immunoblotted with α-HA (3F10) antibody to demonstrate the presence of equal amounts of input CC–NBS-HA derivatives.

Fig. 6. Effect of cis elements on trans interactions. (A) Nicotiana benthamiana leaves were agroinfiltrated with 35S-driven constructs expressing Rx CC–NBS-MT (all lanes) plus either GUS, Rx LRR-HA, Rx-HA or Rx NBS–LRR-HA. Two days post-infiltration, extracts were prepared and subjected to immunoprecipitation with α-HA antibody (3F10). Pellets were immunoblotted with α-HA (3F10) antibody (top panel) and α-MT (A-14) antibody (lower panel). (B) Nicotiana benthamiana leaves were agroinfiltrated with 35S-driven constructs expressing Rx-FH (both lanes) plus either Rx-HA or Rx NBS– LRR-HA. Two days post-infiltration, extracts were prepared and subjected to immunoprecipitation with α-HA antibody (3F10). Pellets were immunoblotted with with α-HA (3F10) antibody (top panel) and α-FLAG (M2) antibody (lower panel).

Fig. 7. Model of Rx activation. A schematic representation of a possible mechanism for Rx activation is presented with two alternatives for effector molecule action. In both models, the coat protein (grey square) interacts transiently with the CC–NBS–LRR protein and initiates a two step conformational change. In the first step, the LRR/NBS interaction is disrupted, resulting in a change to the nucleotide binding status of the NBS. This change then results in disruption of an interaction of the CC. In model 1, an effector molecule(s) (white circles) is constitutively associated with the CC–NBS–LRR protein and is released in an active form upon dissociation of the CC. In model 2, the inactive effector molecule(s) (grey circle) is recruited to Rx upon dissociation of the CC domain.