Can AI modeling of protein structures distinguish between sensor and helper NLR immune receptors?

Nucleotide binding and leucine-rich repeat (NLR) proteins are intracellular immune receptors that occur across all kingdoms of life but are particularly highly diversified in plants (Barragan & Weigel, 2021). In plants, NLRs that carry a coiled-coil (CC) domain at their N termini are the most phylogenetically widespread class. Following pathogen recognition, CC-NLR proteins oligomerize into pentameric or hexameric pore-like complexes (Wang et al., 2019; Förderer et al., 2022; Zhao et al., 2022; Liu et al., 2024; Madhuprakash et al., 2024). These complexes, known as resistosomes, are a defining feature of NLRs that execute the immune response; some of them are known to translocate to cellular membranes and trigger immune responses such as calcium influx and hypersensitive cell death (Duggan et al., 2021; Contreras et al., 2022; Ibrahim et al., 2024). The prevailing model is that the funnel-shaped structure of CC-NLR resistosomes inserts into membranes and is required for executing the cell death and immune response (Wang et al., 2019; Adachi et al., 2019b; Förderer & Kourelis, 2023). This funnel-shaped structure is formed by the N-terminal α1 helix, which is a structurally dynamic region that is difficult to resolve using cryo-electron microscopy (cryo-EM) (Förderer et al., 2022; Zhao et al., 2022; Liu et al., 2024; Madhuprakash et al., 2024).