Fig 1.
Identification of sequences required for RPP1 TIR domain autoactivity.
(A) Schematic overview of the domain architecture of RPP1. Numbers indicate the amino acid position of predicted domain borders for the Niederzenz (NdA) allele of RPP1. (B,C) Determination of the minimal autoactive TIR domain from the NdA allele of RPP1. Both C-terminal (B) and N-terminal (C) truncations were examined for their ability to elicit an effector-independent hypersensitive response (HR). The specific amino acids comprising each construct are indicated in subscript. (D,E) HR phenotypes associated with chimeras or site-directed mutants of N-TIR domains from the NdA and Wassilewskija (WsB) alleles. Site-directed mutagenesis was guided by the amino acid alignment depicted in S2 Fig. Constructs were tested in Nicotiana tabacum via Agrobacterium-mediated transient expression and images were captured at 48 hours post-infiltration. HR phenotypes are scored as negative (-), weak (w), or strong (+). An α-Flag antibody was used to evaluate protein expression, while staining of RuBisCO with Ponceau S provided a loading control. The experiment was performed three times with similar results.
Fig 2.
The autoactivity of the RPP1 N-TIR domain is correlated with self-association in solution.
Purified (N-)TIR domain proteins from the Niederzenz (NdA) and Wassilewskija (WsB) alleles of RPP1 were analyzed by size-exclusion chromatography (SEC) coupled with multi-angle laser light scattering (MALS). For each sample, 175 μg of purified protein was separated on a Superdex Increase 200 5/150 GL SEC column and the molecular mass calculated across the elution peak. The colored solid line represents the normalized refractive index trace (arbitrary units) of the protein eluting from the SEC column. At the elution peak, the averaged molecular mass (kDa) of the proteins was calculated from the protein concentration (derived from the refractive index changes) and light scattering data. The averaged molecular masses across the elution peak are represented by dashed lines of the corresponding color. In planta hypersensitive response (HR) phenotypes are indicated for each construct by a “+” (autoactive) or “-”(non-autoactive). These phenotypes are documented in Fig 1, S2 and S4 Figs. (A) Comparison of the solution properties of RPP1 TIR domains with and without native RPP1 N termini. Numbers in the legend refer to the amino acids that comprise each protein sample. (B) Solution properties of the wild-type WsB N-TIR domain and the gain-of-autoactivity mutants, WsB R230C and WsB K98R I100F. (C) Solution properties of the wild-type NdA N-TIR domain and the loss-of-autoactivity mutants, NdA G229A Y230A and NdA R104A F106A. (D) Comparison of theoretical monomer molecular masses and the measured molecular masses for the proteins analyzed in (A-C).
Fig 3.
The NB-ARC domain influences N-TIR domain autoactivity.
(A) Hypersensitive response (HR) phenotypes are altered by the successive addition of NB, ARC1, and ARC2 subdomains to the N-TIR domains of the RPP1 alleles Niederzenz (NdA) or Wassilewskija (WsB). Constructs were tested in Nicotiana tabacum via Agrobacterium-mediated transient expression and images were captured at 48 hours post-infiltration (hpi). The presence or absence of HR is indicated by a “+” or “-“, respectively. (B) Detection of self-association between NdA NB-ARC subdomain truncations by co-immunoprecipitation. Differentially epitope-tagged versions of NB, NB-ARC1, or NB-ARC1-ARC2 proteins were transiently expressed in N. benthamiana and samples were collected at 48 hpi for co-immunoprecipitation using α-Flag agarose beads. Asterisks indicate non-specific bands. Staining of RuBisCO with Ponceau S provides a loading control. (C) Site-directed mutagenesis of the NB subdomain compromises HR induction in an NdA N-TIR-NB-ARC1 background. Constructs were tested as in (A); protein expression for constructs from (A) and (C) is documented in S9 Fig. (D) Detection of self-association of NB-ARC1 mutants by co-immunoprecipitation, tested as in (B). Experiments were performed at least three times with similar results. (E) Predicted structure of the RPP1 NB-ARC domain derived from homology modeling using the Drosophila Dark protein (PDB:4v4l) as a template. The K299 residue within the P-loop is denoted in blue, while the putative oligomerization interface is highlighted in red. The specific residues at the putative oligomerization interface that were analyzed by mutagenesis are depicted in S10 Fig.
Fig 4.
Effector-dependent oligomerization of full-length RPP1 is generally unaffected by mutations in the NB subdomain.
(A) RPP1_NdA oligomerization is induced by co-expression with ATR1_Emoy2 (E) but not ATR1_Cala2 (C). Constructs were transiently expressed in Nicotiana benthamiana and samples were collected at 36 hours post-infiltration (hpi) for co-immunoprecipitation using α-Flag agarose beads. The expression of ATR1:citrine was detected with an α-GFP antibody. Staining of RuBisCO with Ponceau S provides a loading control. (B) Site-directed mutagenesis of the NB subdomain compromises effector-dependent HR induction by RPP1_NdA. Constructs were transiently co-expressed with ATR1_Emoy2 in N. tabacum and images were captured at 48 hpi. The presence or absence of HR is indicated by a “+” or “-“, respectively. Protein expression for these constructs is documented in S9 Fig. (C) Detection of self-association of RPP1_NdA NB subdomain mutants by co-immunoprecipitation, tested as in (A). Experiments were performed at least three times with similar results.
Fig 5.
The TIR domain inhibits RPP1 self-association in the absence of the effector ATR1.
(A) RPP1_NdA constructs lacking the TIR domain self-associate in an effector-independent manner. Differentially epitope-tagged versions of either NB-ARC-LRR or LRR proteins were transiently expressed in N. benthamiana and samples were collected at 48 hours post-infiltration (hpi) for co-immunoprecipitation using α-Flag agarose beads. Asterisks indicate non-specific bands. Staining of RuBisCO with Ponceau S provides a loading control. (B) The TIR domain alone is responsible for inhibiting RPP1 self-association, as N-terminal truncations up to the TIR domain retain effector-dependent self-association. Co-immunoprecipitation experiments were performed using differentially tagged proteins as in (A) except that tissue samples were collected at 36 hpi (E = ATR1_Emoy2, C = ATR1_Cala2). The expression of ATR1:citrine was detected with an α-GFP antibody. The specific amino acids comprising each construct are indicated in subscript. Experiments were performed three times with similar results.
Fig 6.
Evaluation of interactions between the N-TIR and NB-ARC1 domains of RPP1_NdA.
(A) Non-autoactive N-TIR domain mutants associate with the NB-ARC1 subdomain. (B) Putative oligomerization interface mutants weaken the N-TIR:NB-ARC1 interaction. Constructs were transiently expressed in Nicotiana benthamiana and samples were collected at 36 hours post-infiltration (hpi) for (A) and 48 hpi for (B). Co-immunoprecipitations were performed using α-Flag agarose beads. Staining of RuBisCO with Ponceau S provides a loading control. Experiments were performed three times with similar results.
Fig 7.
Evaluation of interactions between the LRR and NB(-ARC) domains of RPP1_NdA.
The association between the NB(-ARC1) subdomain and the LRR is not disrupted by the presence of ATR1 (A) or by mutations at the putative oligomerization interface (B). Constructs were transiently expressed in Nicotiana benthamiana and samples were collected at 48 hours post-infiltration (E = ATR1_Emoy2, C = ATR1_Cala2). Co-immunoprecipitations were performed using α-Flag agarose beads. The expression of ATR1:citrine was detected with an α-GFP antibody. Asterisks indicate non-specific bands. Staining of RuBisCO with Ponceau S provides a loading control. Experiments were performed three times with similar results.
Fig 8.
Non-autoactive N-TIR domain mutants associate with the LRR domain of RPP1_NdA.
Constructs were transiently expressed in Nicotiana benthamiana and samples were collected at 36 hours post-infiltration. Co-immunoprecipitations were performed using α-Flag agarose beads. Staining of RuBisCO with Ponceau S provides a loading control. Experiments were performed three times with similar results.
Fig 9.
Proposed model of cell death activation by RPP1.
In the absence of ATR1, RPP1 is likely maintained in a largely inactive state by a network of N-TIR:NB, NB:LRR, and N-TIR:LRR interactions. Note that the N-TIR:NB interaction occurs at the putative oligomerization interface to prevent effector-independent association. Binding of a recognized allele of ATR1 to the LRR domain may stabilize conformational transitions that reorient the N-TIR domain to expose the oligomerization interface. This allows RPP1 oligomerization via the NB domain, potentially stabilized by LRR:LRR interactions. The reorientation of the N-TIR domain also permits N-TIR domain self-association, which outcompetes N-TIR:NB interactions and ultimately triggers a cell death response (HR). While this model accounts for the effector-independent NB:LRR interaction, the specific interaction interface(s) could differ before and after RPP1 activation.