A blast fungus zinc-finger fold effector binds to a hydrophobic pocket in host Exo70 proteins to modulate immune recognition in rice

Significance Plant diseases destroy ∼20 to 30% of annual crop production, contributing to global food insecurity. Discovering how pathogen effectors target host proteins to promote virulence is essential for understanding pathogenesis and can be used for developing disease-resistant crops. Here, we reveal the structural basis of how an effector from the blast pathogen (AVR-Pii) binds a specific host target (rice Exo70) and how this underpins immune recognition. This has implications for understanding the molecular mechanisms of blast disease and for engineering new recognition specificities in plant immune receptors to confer resistance to a major crop pathogen.


Gene cloning
For protein production in E. coli, codon optimised AVR-Pii (amino acid residues Leu20 to Asn70) was synthesized (GenScript) and subsequently cloned into the pOPINM vector (1) using the In-Fusion cloning kit (Takara Bio USA). AVR-Pii Trp64Arg and Phe65Glu were synthesized as PCR products (Gblocks, IDT) and cloned into pOPINM in the same way. Truncated versions of rice Exo70 alleles OsExo70B1 Δ91 , OsExo70F2 Δ83 and OsExo70F3 Δ93 were generated using standard molecular biology techniques from appropriate templates described by Fujisaki et al. (2) and cloned into pOPINS3C (1) using the In-Fusion cloning kit (Takara Bio USA).
For Y2H, wild-type and mutant AVR-Pii effectors (amino acid residues Leu20 to Asn70) were cloned in pGADT7 while full length CDS of rice Exo70 alleles OsExo70B1, OsExo70F2 and OsExo70F3 were cloned in pGBKT7. In both cases, plasmids were linearized by double digestion with EcoRI and BamHI (New England Biolabs) and genes of interest were introduced using the In-Fusion cloning kit (Takara Bio USA).
For random mutagenesis we used the Diversify PCR Random Mutagenesis Kit (Takara Bio USA) and subsequently cloned the mutagenized AVR-Pii PCR fragments in pGADT7 as described above.

Expression and purification of proteins for X-ray crystallography and in vitro binding studies
To enable the study of the OsExo70/AVR-Pii interactions in vitro, we produced stable rice OsExo70 proteins in E. coli. SUMO-tagged OsExo70 alleles with the predicted N-terminal α-helix truncated encoded in pOPINS3C were produced in E. coli Rosetta™ (DE3). Cell cultures were grown in autoinduction media (5) at 37°C for 5-7 hr and then at 16°C overnight. Cells were harvested by centrifugation and re-suspended in 50 mM Tris-HCl (pH 7.5), 500 mM NaCl, 50 mM glycine, 5% (vol/vol) glycerol, and 20 mM imidazole supplemented with EDTA-free protease inhibitor tablets (Roche). Cells were sonicated and, following centrifugation at 40,000xg for 30 min, the clarified lysate was applied to a HisTrap™ Ni2+-NTA column connected to an AKTA Xpress purification system (GE Healthcare). Proteins were step-eluted with elution buffer (50 mM Tris-HCl (pH7.5), 500 mM NaCl, 50 mM glycine, 5% (vol/vol) glycerol, and 500 mM imidazole) and directly injected onto a Superdex 200 26/60 gel filtration column pre-equilibrated with 20 mM HEPES (pH 7.5), 150 mM NaCl and 5% (vol/vol) glycerol supplemented with 1mM TCEP. Elution fractions were collected and evaluated by SDS-PAGE, revealing a band close to 70 kDa ( Figure S1a). Fractions were combined and incubated overnight with 3C protease (10 μg/mg fusion protein). Rice Exo70 proteins were separated from the SUMO tag by passing the protein mixture solution through a HisTrap™ Ni2+-NTA column equilibrated with 50 mM Tris-HCl (pH 7.5), 500 mM NaCl, 50 mM glycine, 5% (vol/vol) glycerol, and 20 mM imidazole ( Figure S1b). Exo70 proteins were mainly present in the first and second wash-through ( Figure S1b) from the column, whilst the SUMO tag was retained until the final elution with elution buffer ( Figure S1b). Fractions containing Exo70 proteins were pooled together and concentrated for further purification and buffer exchange by gel filtration onto a Superdex 200 16/60 column pre-equilibrated with 20 mM HEPES (pH 7.5), 150 mM NaCl and 5% (vol/vol) glycerol supplemented with 1mM TCEP (Figure S1c). Fractions containing purified Exo70 proteins were combined and concentrated for structural and biophysical studies.
For OsExo70 gel filtration analysis, a volume of 110 μl of each sample was separated at 4 °C on a Superdex 200 10/300 size exclusion column (GE Healthcare), pre-equilibrated in 20 mM HEPES (pH 7.5), 150 mM NaCl, 1 mM TCEP and 5% (vol/vol) glycerol and at a flow rate of 0.5 ml/min. Fractions of 0.5 ml were collected for analysis by SDS-PAGE.
MBP-tagged effector domain (amino acid residues 20 to 70) for wild-type AVR-Pii, Trp64Arg and Phe65Glu encoded by the pOPINM constructs were produced in E. coli SHuffle cells (6). Cell cultures were grown in autoinduction media (5) at 30°C for 5-7 hr and then at 16°C overnight. After harvest by centrifugation, cells were resuspended and disrupted as described above for OsExo70 expression.
The soluble fusion protein 6xHis:MBP:AVR-Pii was be purified from E. coli cell lysates by IMAC on a HisTrap™ Ni2+-NTA column connected to an AKTA Xpress purification system (GE Healthcare) coupled with gel filtration onto a Superdex 75 26/60 gel filtration column pre-equilibrated with 20 mM HEPES (pH 7.5), 150 mM NaCl and 5% (vol/vol) glycerol ( Figure S3a). The fractions containing the eluted protein were subsequently treated with 3C protease as before to remove the MBP tag. AVR-Pii was purified from the MBP solubility tag using HisTrap™ and MBPTrap™ (GE Healthcare) columns attached in tandem ( Figure S3b). Purified AVR-Pii was commonly present as a double band in the flow-through (FT) and wash-through (WT) from the columns ( Figure S3b).
The relevant fractions were concentrated and loaded onto a Superdex 75 16/60 gel filtration column for final purification and buffer exchange into 20 mM HEPES (pH 7.5), 150 mM NaCl and 5% (vol/vol) glycerol ( Figure S3c). Relevant fractions with purified AVR-Pii were concentrated as appropriate and used for structural and biophysical characterization.
The state of the protein was assessed by intact mass spectrometry, revealing a main peak with a molecular weight of 5677.68 Da, identical to that calculated for AVR-Pii ( Figure S4).
All protein concentrations were determined using a Direct Detect Infrared Spectrometer (Merck).

Crystallization, data collection and structure solution
For crystallization, OsExo70F2 (residues 84 to 689) in complex with AVR-Pii (residues 20 to 70) was concentrated to 6 mg/ml following gel filtration. Sitting drop vapor diffusion crystallization trials were set up in 96 well plates, using an Oryx nano robot (Douglas Instruments, United Kingdom), with drops comprised of 0.3 μl precipitant solution and 0.3 µl of the sample, and incubated at 20°C. After four to six days, protein crystals for the complex between OsExo70F2 and AVR-Pii appeared in the 0.3 M Ammonium iodide; 30% v/v PEG3350 condition of the PEG Suite screen (Qiagen). For data collection, all crystals were harvested using Litholoops (Molecular Dimensions) and flashcooled in liquid nitrogen.
X-ray data were collected from a single crystal at the Diamond Light Source using beamline i03 (Oxford, UK) at 100 K and recorded on a Pilatus3 6M hybrid photon counting detector (Dectris). The data were processed using the xia2 pipeline (7) and CCP4 (8). To solve the structure of OsExo70F2/AVR-Pii complex, we used the Arabidopsis Exo70A1 (PDB ID: 4L5R) divided into three ensembles (Ensemble 1 residues 76 to 325; Ensemble 2 residues 326 to 474; Ensemble 3 residues 475 to 593) as a template for molecular replacement using PHASER (9). Once we obtained a solution, automated model building using BUCANNEER (10) was able to identify and build the AVR-Pii effector. The asymmetric unit of the crystal contains only a single copy of the OsExo70F2/AVR-Pii complex with a stoichiometry of 1:1. The final structure was obtained through iterative cycles of model building and refinement using COOT (11), REFMAC5 (12), and ISOLDE (13) as implemented in the CCP4 suite (8) and ChimeraX (14). Structures were validated using the tools provided in COOT and MOLPROBITY (15).

Yeast-2-hybrid
The OsExo70 proteins encoded in pGBKT7 plasmids were co-transformed with AVR-Pii variants or mutants in pGADT7 into chemically competent Y2HGold cells (Takara Bio, USA) using a Frozen-EZ Yeast Transformation Kit (Zymo research).
Single colonies grown on selection plates were inoculated in 5 ml of SD -Leu -Trp overnight at 30 °C. Saturated culture was then used to make serial dilutions of OD600 1, 0.1, 0.01, and 0.001. 5 μl of each dilution was spotted on a SD -Leu -Trp plate as a growth control, and on a SD -Leu -Trp -Ade -His plate containing X-α-gal and supplemented with 0.2 or 1 μg/ml Aureobasidin A (Takara Bio, USA). Plates were imaged after incubation for 60 -72 h at 30 °C. Each experiment was repeated a minimum of three times, with similar results.
To confirm protein expression, total yeast extracts from transformed colonies were produced by boiling the cells for 10 min in LDS Runblue TM sample buffer (Expedeon). Samples were centrifugated and the supernatant was subjected to SDS-PAGE gels and western blot. The membranes were probed with anti-GAL4 DNA-BD (Sigma) for the OsExo70 proteins in pGBKT7 and with the anti-GAL4 activation domain (Sigma) antibodies for the AVR-Pii wild type and mutant effectors in pGADT7.     Binding of AVR-Pii to rice OsExo70 proteins determined by isothermal titration calorimetry (ITC). Upper panels show heat differences upon injection of AVR-Pii into the cell containing the respective OsExo70 allele. Middle panels show integrated heats of injection (dots) and the best fit (solid line) using to a single site binding model calculated using AFFINImeter ITC analysis software (16). Bottom panels represent the difference between the fit to a single site binding model and the experimental data; the closer to zero indicates stronger agreement between the data and the fit. The thermodynamic parameters obtained in each experiment are presented in Table S1.

Fig. S10. Analysis of the binding interface between OsExo70F2 and AVR-Pii using qtPISA.
Interface analysis was performed using qtPISA (19). The key interface parameters in the analysis are represented as an interaction radar and the values are listed in the adjacent table.

Fig. S11. The amphipathic AVR-Pii binding interface is conserved in rice Exo70s.
Conservation profile of rice Exo70 residues with a close-up view at the AVR-Pii interface as calculated by ConSurf (20). Exo70 is represented with solid surface colored according to the conservation of their residues ranging from purple (highly conserved) through white (moderately conserved) to cyan (highly variable). Surface areas highlighted in yellow correspond to residues for which a meaningful conservation level could not be derived from the set of homologues used for the analysis. A close-up view of the effector interface is also shown with AVR-Pii residues 44 to 70 represented in ribbons and colored in yellow with the side chains of important residues displayed as cylinders. Conservation analysis was generated using the rice Exo70 protein sequences reported by Cvrckova et al. (21).

Fig. S13. Comparison of the effector binding interface between OsExo70F2 and OsExo70B1. (A)
Sequence alignment of residues located at the OsExo70F2 and OsExo70B1 α-helices 7 and 8 generated with Clustal Omega (24). Secondary structure features of Exo70 fold are shown above, and important residues for the formation of the AVR-Pii binding pocket are highlighted in red. Comparison of (B) OsExo70F2 and (C) OsExo70B1 surface hydrophobicity at the interaction interface with AVR-Pii, residues are colored depending on their hydrophobicity from light blue (low) to yellow (high). Comparison of (D) OsExo70F2 and (E) OsExo70B1 surface electrostatic potentials at the interaction interface with AVR-Pii, residues are colored depending on their electrostatic potential from dark blue (positive) to red (negative). The OsExo70B1 structure used for comparison was generated using AlphaFold2 (22) (Figure S12).   (24), focusing on OsExo70F2 residues that interact with AVR-Pii. OsExo70 proteins studied here are highlighted in blue. Equivalent positions of OsExo70 residues contributing to the formation of the hydrophobic pocket in OsExo70F2 are highlighted in red.

Fig. S16. Random mutagenesis identified AVR-Pii residues that alter binding to OsExo70F3. (A)
Amino acid sequence of the AVR-Pii mutants obtained by random mutagenesis. Secondary structure features of the AVR-Pii fold are shown above, and the residues not observed in the crystal structure are highlighted in orange. (B) Yeast-Two-Hybrid assay of AVR-Pii mutants obtained by random mutagenesis with OsExo70F3. The control plate for yeast growth is on the left, with quadruple dropout media supplemented with X-α-gal and Aureobasidine A (Au A) on the right. Growth and development of blue coloration in the selection plate are both indicative of protein:protein interactions. Wild-type AVR-Pii is included as positive control. OsExo70F3 was fused to the GAL4 DNA binding domain, and AVR-Pii mutants to the GAL4 activator domain. Each experiment was repeated a minimum of three times, with similar results. (C) Accumulation of AVR-Pii mutants in Yeast-Two-Hybrid assays analyzed by Western blot. Yeast lysate was probed for the expression of OsExo70F3 and AVR-Pii mutants obtained by random mutagenesis using anti-GAL4 binding domain (BD) and anti-GAL4 DNA activation domain (AD) antibodies, respectively. Total protein extracts were colored with Coomassie Blue Stain (CBS). (D) Deconvolution of residues involved in AVR-Pii binding by Yeast-Two-Hybrid assay with rice Exo70F3. The control plate for yeast growth is on the left, with quadruple dropout media supplemented with X-α-gal and Aureobasidine A (Au A) on the right. Growth and development of blue coloration in the selection plate are both indicative of protein:protein interactions. Wild-type AVR-Pii is included as positive control. OsExo70F3 was fused to the GAL4 DNA binding domain, and AVR-Pii point mutants to the GAL4 activator domain. Each experiment was repeated a minimum of three times, with similar results. (E) Accumulation of AVR-Pii point mutants in yeast-two-hybrid assays analyzed by Western blot. Yeast lysate was probed for the expression of OsExo70F3 and AVR-Pii point mutants using anti-GAL4 binding domain (BD) and anti-GAL4 DNA activation domain (AD) antibodies, respectively. Total protein extracts were colored with Coomassie Blue Stain (CBS).  Upper panels show heat differences upon injection of AVR-Pii mutants into the cell containing OsExo70F2. Middle panels show integrated heats of injection (dots) and the best fit (solid line) using to a single site binding model calculated using AFFINImeter ITC analysis software (16). Bottom panels represent the difference between the fit to a single site binding model and the experimental data; the closer to zero indicates stronger agreement between the data and the fit. The thermodynamic parameters obtained in each experiment are presented in Table S3. Upper panels show heat differences upon injection of AVR-Pii mutants into the cell containing OsExo70F3. Middle panels show integrated heats of injection (dots) and the best fit (solid line) using to a single site binding model calculated using AFFINImeter ITC analysis software (16). Bottom panels represent the difference between the fit to a single site binding model and the experimental data; the closer to zero indicates stronger agreement between the data and the fit. The thermodynamic parameters obtained in each experiment are presented in Table S3.  Hitomebore (Pii+). Eight independent Sasa2 transformants harboring wild-type AVR-Pii were spotted in both cultivars. Line numbers colored in red indicate the transformants removed from further quantification because they did not express AVR-Pii effectors as tested by RT-PCR ( Figure  S20) or were not infective.