An activated wheat CCG10-NLR immune receptor forms an octameric resistosome
An autoactive wheat CCG10-NLR assembles into an octameric resistosome, the largest plant NLR oligomer resolved so far, with its coiled-coil domain projecting away from the LRR rather than folding back against it.
Plant CC-NLR resistosomes were thought to converge on a small set of architectures: a pentamer for ZAR1 and Sr35, a hexamer for the NRC helpers, with the coiled-coil domain bent back to contact the LRR through the EDVID motif. WAI3, a wheat receptor from the poorly characterised CCG10 clade, does neither. A single L619F substitution in its LRR renders it autoactive, and cryo-EM of the purified protein resolves an octamer with C8 symmetry, plus two dual-octamer forms. Arabidopsis RPS2, a CCG10-NLR from the other side of the monocot-dicot split, also forms octamers. WAI3 lacks both the EDVID motif and the LRR arginine cluster, so the CC and LRR sit on opposite faces of the NBD, an arrangement closer to a TIR-NLR than to any CC-NLR described before. The resulting pore is wider than the NRC4 hexamer's and drives a prolonged, multi-phasic cytosolic calcium rise.
Key findings (9)
A single leucine-to-phenylalanine substitution at position 619 in the LRR domain of WAI3 (TraesCS2D01G380500, chromosome 2D) is the causal lesion behind the wheat M3405 autoimmune phenotype; the allele is semi-dominant and confers enhanced powdery mildew resistance.
Evidence forward genetics; map-based cloning; transgenic complementation in cv. Fielder
Source Figures 1G-1I; Figure S1E, S1F; Table S1
Activated WAI3 assembles into an octameric ring resistosome with C8 symmetry, roughly 200 A in diameter and 135 A high, resolved at 3.95 A. Two further states were resolved: an N-to-N dual octamer fused at the funnel ends (4.5 A, D8) and a back-to-back dual octamer joined through the LRR domains (3.6 A, D8). The B-to-B arrangement had not been seen in any plant resistosome.
Evidence cryo-EM single-particle reconstruction; 7,043 micrographs; BN-PAGE sizing at ~1,000 kDa
Source Figures 2B-2E, Figure S2, Table 1; PDB 9H2L / 9H4I / 9H73; EMDB EMD-51810 / EMD-51852 / EMD-51910
Arabidopsis RPS2, a CCG10-NLR from a dicot, also forms octameric rings, so the octamer is a property of the clade rather than of wheat. Purified RPS2 was heterogeneous with only a minority of particles oligomeric, which the authors attribute to incomplete activation without RIN4 fragments.
Evidence negative-stain EM of tandem-affinity-purified RPS2 delta-N26
Source Figure 2F; Figure S3C
WAI3GOF drives a prolonged, multi-phasic rise in cytosolic calcium in N. benthamiana, distinct from the single ~4 h peak reported for ZAR1, Sr35, NRC4, NRG1.1 and ADR1. LaCl3 abolishes both the calcium influx and the cell death. Mutations predicted to block membrane binding (L12E/L15E) or to disrupt the acidic inner lining of the alpha1 helix (D3K/E17K/D20K) abolish both.
Evidence GCaMP3 cytosolic calcium imaging over extended time course; channel-blocker and mutant series
Source Figures 4A-4H; Figures S3D-S3I
The WAI3 pore is geometrically distinct from the NRC4 hexamer pore: minimum radius 12.7 A at Lys140, against 4.3 A at Gln126 for NRC4DV.
Evidence PoreAnalyzer computational analysis of the resolved structures (alpha1 helix unresolved in both)
Source Figures 4I and 4J
WAI3 lacks the EDVID motif and the LRR arginine cluster, so the CC-LRR contact seen in ZAR1, Sr35, NRC2 and NRC4 is absent. The CC domain is rotated approximately 180 degrees about its hinge to the NBD, placing CC and LRR on opposite sides of the NBD and reversing membrane topology so the LRRs face the cytosol. The overall arrangement resembles the TIR-NLR RPP1 more than any previously characterised CC-NLR.
Evidence structural comparison against ZAR1 (PDB 6J5T), NRC4 (PDB 9CC8), RPP1 (PDB 7DFV); sequence alignment
Source Figures 5A-5E
AlphaFold3 monomer models of representative Arabidopsis CC-NLRs from the four CC sequence groups show that EDVID-lacking members (groups A and B, including CCR-NLRs such as NRG1 and CCG10-NLRs such as RPS2) adopt a WAI3-like CC arrangement, while EDVID-containing members (groups C and D, ZAR1 and RPP8) adopt a ZAR1-like one.
Evidence AlphaFold3 structural modelling of 5 representatives per group; consistent with an independent study (ref 35)
Source Figure S5
The HD1-WHD interdomain angle tracks with oligomeric state: 80 degrees for the WAI3 octamer, 88 for the NRC2 hexamer, 92 for the ZAR1 pentamer, with larger angles giving fewer protomers. AlphaFold-predicted monomer angles reproduce the experimental values (WAI3 82, RPS2 84 predicting an octamer, NRG1 94 predicting a pentamer).
Evidence structural superposition and angle measurement across RPP1, ZAR1, NRC2 and WAI3 protomers; AlphaFold monomer models
Source Figures S6A-S6C
WAI3GOF produced neither calcium influx nor detectable channel currents in HEK293T cells, while NRG1.1DV did both in the same system. As with NRC resistosomes, WAI3 appears to need unidentified plant-specific factors.
Evidence heterologous expression in HEK293T; calcium imaging; patch-clamp electrophysiology
Source Figures S4B-S4I
NLRs
| Name | Class | Role in this paper |
|---|---|---|
| WAI3 Wheat Autoimmunity 3, TraesCS2D01G380500, C5, WAI3GOF (L619F allele) | singleton (autoactive; no helper requirement tested) subject | central subject; autoactive allele purified and resolved as an octameric resistosome |
| RPS2 RESISTANCE TO PSEUDOMONAS SYRINGAE 2 | singleton subject | second CCG10-NLR shown to form octamers, establishing the architecture across monocots and dicots |
| NRC4 NLR required for cell death 4, NRC4DV (autoactive) | helper comparator | hexameric comparator for pore geometry and calcium kinetics; PDB 9CC8 |
| NRC2 | helper comparator | hexameric comparator for HD1-WHD angle and protomer superposition |
| ZAR1 HOPZ-ACTIVATED RESISTANCE 1 | singleton comparator | pentameric reference architecture; PDB 6J5T |
| Sr35 stem rust resistance 35 | singleton comparator | pentameric comparator; source of the N-to-N precedent |
| NRG1 N requirement gene 1, NRG1.1, NRG1.1DV | helper comparator | EDVID-lacking comparator; predicted pentamer from HD1-WHD angle; positive control in HEK293T assays |
| ADR1 activated disease resistance 1 | helper cited-only | cited comparator for CCR-NLR calcium influx |
| RPP1 recognition of Peronospora parasitica 1 | singleton comparator | tetrameric TIR-NLR whose domain arrangement WAI3 unexpectedly resembles; PDB 7DFV |
| ROQ1 recognition of XopQ 1 | singleton cited-only | cited tetrameric TIR-NLR comparator |
| RPS5 | singleton cited-only | clade member contrasted with WAI3 and RPS2, which lack a myristoylation motif |
| SUMM2 | singleton cited-only | same contrast as RPS5 |
| MLA13 Mildew Locus A 13 | singleton cited-only | cited as an alternative activation mode (stable heterodimer with its effector) |
| RPP8 | singleton comparator | representative of CC sequence group D in the AlphaFold3 comparison |
Effectors
No effector is studied in this paper. WAI3 is analysed as a gain-of-function autoactive allele, so activation is genetic rather than effector-triggered. RIN4 is discussed as the negative regulator of RPS2 and as a probable missing activator in the RPS2 purification, but it is a host protein, not an effector.
Structures
| Entry | State | Description |
|---|---|---|
| PDB 9H2L (EMDB EMD-51810) | octamer | Octameric WAI3 resistosome, 3.95 A, C8 symmetry, 14,572 particles; residues 36-919 modelled, alpha1 helix and residues 811-830 unresolved; 8 ATP |
| PDB 9H4I (EMDB EMD-51852) | dual octamer (16-mer) | N-to-N dual-octameric WAI3 resistosome, 4.5 A, D8 symmetry, 13,948 particles; two octamers fused at the funnel ends; 16 ATP |
| PDB 9H73 (EMDB EMD-51910) | dual octamer (16-mer) | B-to-B dual-octameric WAI3 resistosome, 3.6 A, D8 symmetry, 27,667 particles; two octamers joined through outer LRR surfaces; first such arrangement in a plant resistosome; possible artefact per the authors |
| negative-stain EM, not deposited | octamer | RPS2 delta-N26 octameric ring particles |
| AlphaFold3 predicted | monomer | Monomer models of 20 Arabidopsis CC-NLRs across CC sequence groups A-D, and of WAI3, RPS2 and NRG1 monomers for HD1-WHD angle measurement |
Open questions
- The N-terminal alpha1 helix is unresolved in every reconstruction, so the membrane-inserted pore itself has not been visualised. A lipid-environment structure is needed.
- How L619F causes autoactivation is unknown. AlphaFold always returns an activated-like conformation for wild-type WAI3, and no CCG10-NLR resting-state structure exists.
- The plant-specific factors required for WAI3 and NRC4 calcium influx are unidentified.
- Whether the back-to-back dual octamer is physiological or a purification artefact is unresolved; the authors say so explicitly.
- The functional consequence of the reversed membrane topology, with LRR domains facing the cytosol, is untested.
- Whether the multi-phasic calcium pattern reflects genuine tissue-level oscillation or the Agrobacterium expression system is not separated.
Connections in this corpus
| Paper | Why |
|---|---|
ibrahim-2026-nrg1-chloroplast | Both concern where and how CC-type resistosomes conduct calcium. NRG1 is EDVID-lacking, and Guo predicts a ZAR1-like pentamer for it from the HD1-WHD angle while placing it in the WAI3-like CC group; Ibrahim shows NRG1 acts at the chloroplast envelope rather than the plasma membrane. The two papers pull in different directions on NRG1 and should be read against each other. |
toghani-2026-sni-undecamer | Both break the pentamer/hexamer expectation for NRC-clade and CC-NLR assemblies, Guo by cryo-EM on an octamer and Toghani by AlphaFold3 screening plus negative-stain on an undecamer. Guo's HD1-WHD angle heuristic and Toghani's Structural Novelty Index are two independent attempts to predict oligomeric state from monomer models. |
seager-2026-avrcap1b-nrc3 | Complementary views of resistosome assembly: Guo captures three stable end states of one receptor, Seager captures stalled intermediates of another. Both use cryo-EM on NLRs purified from N. benthamiana. |
madhuprakash-2026-avrcap1b-tol9a | Guo cites Madhuprakash et al. 2024 (ref 12) for NRC helper oligomerisation; both papers work on activated helper NLR complexes expressed in N. benthamiana. |
toghani-2026-sensor-helper-interface | Shared reliance on AlphaFold3 to reason about NLR complexes that resist experimental capture, and shared conclusion that predicted models can be trusted enough to drive mutagenesis. |
gomezdelacruz-2026-mla3-mimicry | Both are cereal NLR papers using structure to explain recognition or assembly; Guo cites the MLA13-AVRA13 heterodimer (ref 26) as an alternative activation mode, which is the same MLA system Gomez De La Cruz works in. |
pai-2026-lettuce-nrc | Pai counts 42 CCG10-NLRs in the lettuce NLRome, a class this paper characterises structurally for the first time. Pai gives the clade its abundance in a campanulid genome; Guo gives it an architecture. Together they bear on whether CCG10-NLR expansion tracks with the octameric assembly mode. |
marques-2026-legume-tissue | Marques finds CCG10-NLRs are the third most abundant NLR class across 28 legume genomes and root-favouring in expression, overrepresented in early-diverging legumes. This paper gives that same clade its first resistosome structure. One supplies abundance and expression context for CCG10-NLRs, the other their architecture. |
gao-2026-vnt1-splicing | Both bear on what the N terminus does to resistosome assembly. This paper resolves a CC domain rotated away from the LRR in a receptor lacking the EDVID motif; Gao shows an N-terminal extension ahead of the MADA motif that prevents funnel formation altogether. Both use BN-PAGE oligomerisation to separate assembly-competent from assembly-blocked variants. |
Provenance and caveats
Overall resolutions 3.95 / 4.5 / 3.6 A are taken from the Results text. Table 1 as extracted lists 4 / 4.6 / 3.6 for 'Overall resolution' and 4.0 / 4.6 / 3.6 for 'Resolution cutoff'. The 3.95 vs 4.0 and 4.5 vs 4.6 differences are probably rounding plus a column-extraction artefact, but they are not identical in the source. [UNVERIFIED - resolve against the published Table 1]
The count of approximately 78 printed references is an estimate from the highest reference number seen in text; not counted directly.
Sources
- Full citation: Guo G, Zhao H, Bai K, Lu J, Wu Q, Lu L, Zhang Y, Dong L, Li G, Chen Y, Hou Y, Lu P, Li M, Zhang H, Wang G, Zhu K, Huang B, Cui X, Fu H, Hu C, Chu Z, Lyu X, Kamoun S, Wang C, Liu Z, Selvaraj M, Jones JDG (2026) An activated wheat CCG10-NLR immune receptor forms an octameric resistosome. Cell 189:2955-2970.
- DOI: 10.1016/j.cell.2026.02.024
- PDF on file:
papers/1-s2.0-S009286742600231X-main.pdf - Extraction:
data/papers/guo-2026-ccg10-resistosome.json(schema v1.0) - References: 71 works, retrieved from OpenAlex
- Licence: Open Access (Elsevier, 'The Authors')