Resurrection of plant disease resistance proteins via helper NLR bioengineering

Parasites counteract host immunity by suppressing helper nucleotide binding and leucine-rich repeat (NLR) proteins that function as central nodes in immune receptor networks. Understanding the mechanisms of immunosuppression can lead to strategies for bioengineering disease resistance. Here, we show that a cyst nematode virulence effector binds and inhibits oligomerization of the helper NLR protein NRC2 by physically preventing intramolecular rearrangements required for activation. An amino acid polymorphism at the binding interface between NRC2 and the inhibitor is sufficient for this helper NLR to evade immune suppression, thereby restoring the activity of multiple disease resistance genes. This points to a potential strategy for resurrecting disease resistance in crop genomes.

: NRC4 HD1-1 chimera is susceptible to inhibition by SS15. HR scores accompanying Fig.  2D. In all cases, Rx/CP was used to activate the system. HR was scored based on a modified 0-7 scale (49) between 5-7 days post-infiltration. HR scores are presented as dot plots, where the size of each dot is proportional to the number of samples with the same score (Count). Results are based on 3 biological replicates. Statistical tests were implemented using the besthr R library (50). We performed bootstrap resampling tests using a lower significance cut-off of 0.025 and an upper cut-off of 0.975. Mean ranks of test samples falling outside of these cut-offs in the control samples bootstrap population were considered significant. Significant differences between the conditions are indicated with an asterisk (*).

Fig. S5
: Crystal structure of SS15 in complex with SlNRC1 NB-ARC . Electron density map showing the relative orientation and arrangement of SS15 (orange) and SlNRC1 NB-ARC (violet) within an asymmetric unit. 2Fo-Fc map countered at 1 (B) Two possible interfaces between SS15 and SlNRC1 NB-ARC revealed from the crystal packing. Both interfaces (Interface 1 and Interface 2) are outlined (Left). Modelling of both potential binding interfaces for SS15 complex with full length SlNRC1 (magenta) reveals a steric clash between the CC-domain of SlNRC1 and SS15, making interface 2 unlikely to be biologically relevant in the full-length context (Right). (C) Close up view of interaction between SS15-SlNRC1 NB-ARC interaction interface relative to the ATPbinding site within the NB-ARC domain of SlNRC1. The pyrophosphate moiety of ADP is oriented facing opposite the SS15 binding interface (shown as ball and sticks), suggesting that SS15 is unlikely to displace bound nucleotide or prevent ATP hydrolysis. (D) Structure of SS15-SlNRC1 NB-ARC (yellow, PDB 8BV0) is superimposed over the NB-ARC domain of AtZAR1 in its inactive (green, PDB 6J5W), intermediate (cyan, PDB 6J5V), and active resistosome (magenta, 6J5T) conformations. Visualizing these three states reveals the trajectory of the NB domain as it moves relative to the HD1 and WHD domains while changing from inactive to activated states. The binding of SS15 at the critical hinge region between the NB and HD1-WHD domains likely immobilizes this loop, preventing these critical intramolecular rearrangements and therefore preventing NLR activation. See Movie S1.

Fig. S6: Out of 13 NRC2 variants tested, only E316P and D317K mutations abolish SS15 association and HR suppression. (A) Photo of representative leaves from N. benthamiana nrc2/3/4 KO plants showing HR after co-expression of Rx and PVX CP with NRC2
, or the different NRC2 variants generated. These effector-sensor-helper combinations were co-expressed with a free mCherry-6xHA fusion protein (EV) or with N-terminally 4xHA-tagged SS15. (B) Co-Immunoprecipitation (Co-IP) assays between SS15 and NRC2 variants. C-terminally 4xMyctagged NRC2 variants were transiently co-expressed with N-terminally 4xHA-tagged SS15. IPs were performed with agarose beads conjugated to Myc antibodies (Myc IP). Total protein extracts were immunoblotted with appropriate antisera labelled on the left. Approximate molecular weights (kDa) of the proteins are shown on the right. Rubisco loading control was carried out using Ponceau stain (PS). The experiment was repeated three times with similar results. HR scores accompanying Fig. 3C. In all cases, Rx/CP was used to activate the system. HR was scored based on a modified 0-7 scale (49) between 5-7 days post-infiltration. HR scores are presented as dot plots, where the size of each dot is proportional to the number of samples with the same score (Count). Results are based on 3 biological replicates. Statistical tests were implemented using the besthr R library (50). We performed bootstrap resampling tests using a lower significance cut-off of 0.025 and an upper cut-off of 0.975. Mean ranks of test samples falling outside of these cut-offs in the control samples bootstrap population were considered significant. Significant differences between the conditions are indicated with an asterisk (*).

Fig. S8: NRC2 D317K abolishes SS15-mediated suppression of Rx, Gpa2 and Prf.
HR scores accompanying Fig. 4A. NRCs were activated using Rx/CP (A), Pto/AVRPto (B) or Gpa2/RBP1 (C). HR was scored based on a modified 0-7 scale (49) between 5-7 days postinfiltration. HR scores are presented as dot plots, where the size of each dot is proportional to the number of samples with the same score (count). Results are based on 3 biological replicates. Statistical tests were implemented using the besthr R library (50). We performed bootstrap resampling tests using a lower significance cut-off of 0.025 and an upper cut-off of 0.975. Mean ranks of test samples falling outside of these cut-offs in the control samples bootstrap population were considered significant. Significant differences between the conditions are indicated with an asterisk (*).

Fig. S9: NRC2 D317K abolishes SS15-mediated suppression of all NRC2-dependent sensors tested and restores NRC2 resistosome formation. Photo of representative leaves from N. benthamiana nrc2/3/4 KO plants showing HR after co-expression of NRC2, or different NRC2
variants generated with various sensor/effector pairs. These effector-sensor-helper combinations were co-expressed with a free mCherry-6xHA fusion protein (EV) or with N-terminally 4xHAtagged SS15. (B) SDS-PAGE accompanying BN-PAGE shown in Fig. 4B. Total protein extracts were immunoblotted with the appropriate antisera labelled on the left. Approximate molecular weights (kDa) of the proteins are shown on the right. Rubisco loading control was carried out using Ponceau stain (PS). The experiment was repeated three times with similar results.   c CC½ is the correlation coefficient between symmetry equivalent intensities from random halves of the dataset.
d The data set was split into "working" and "free" sets consisting of 90 and 10% of the data respectively. The free set was not used for refinement. and Fcalc are the observed and calculated structure factor amplitudes, respectively.
g Real Space Correlation Coefficient as calculated by the PDB validation server.
Movie S1: SS15 immobilizes hinge region between NB and HD1-WHD domains of NRCs to prevent NLR activation.