Calcium signalling and where immunity is executed

Calcium is the shared output of every functional receptor in this corpus. Where the calcium comes from is not shared at all, and one paper here breaks the assumption that it comes from outside the cell.

The standing model, and the exception

Activated CC-NLRs were understood to insert into the plasma membrane and conduct calcium inward. NRG1 does something else. On activation it forms puncta on the chloroplast outer envelope, colocalising with an outer-membrane marker and encircling a stromal one, and stromal calcium falls as those puncta appear. Time-lapse imaging over an hour ties the two events to the same chloroplast (ibrahim-2026-nrg1-chloroplast).

The specificity is functional, which is the part that makes it more than a localisation result. Trapping NRC4 at the chloroplast with a nanobody abolishes its cell-death activity; trapping NRG1 in the same place leaves it working and still draining the stroma. Trapping either at the plasma membrane affects neither (ibrahim-2026-nrg1-chloroplast).

An architectural explanation is offered: AlphaFold models give the NRG1 coiled-coil funnel a length of about 117 A against 66-69 A for the canonical receptors, roughly the extra reach needed to cross a double membrane. That model scores ipTM 0.43. It is the weakest structural claim in the corpus and the paper says so — no activated CCR-NLR resistosome has been resolved experimentally. The mutagenesis carries the argument instead: a double substitution in the extended fourth helix retains the receptor at the plasma membrane and reduces cell death (ibrahim-2026-nrg1-chloroplast).

The kinetics are not one shape either

Extended recording of a CCG10 receptor shows a multi-phasic cytosolic calcium rise, an initial peak within 12 hours followed by a delayed second phase, where earlier studies of ZAR1, Sr35, NRC4, NRG1.1 and ADR1 reported a single peak of about four hours. The authors attribute the difference to how long anyone watched (guo-2026-ccg10-resistosome).

Pore geometry differs to match. The octamer's ion pathway has a minimum radius of 12.7 A at Lys140; the NRC4 hexamer narrows to 4.3 A at Gln126. Both figures come from the resolved portions only, since the alpha1 helix that forms the pore is unresolved in each (guo-2026-ccg10-resistosome).

What is required, and unidentified

Neither receptor works in a human cell line. The CCG10 receptor produced no calcium influx and no detectable currents in HEK293T, while NRG1.1DV did both in the same system, so some plant-specific factor is needed and has not been found (guo-2026-ccg10-resistosome). The same requirement was previously reported for NRC resistosomes.

Calcium is also the readout that separates competent from blocked receptors in the splicing work: the intron-retained isoform gives weakened influx, and a variant locked in that state gives none (gao-2026-vnt1-splicing).

Compartment as a strategy

If a receptor can act at the chloroplast, mitochondria or ER as well as the plasma membrane, it can reach calcium stores that a plasma-membrane channel cannot, and it can act where an effector is not waiting for it. That is proposed rather than tested (ibrahim-2026-nrg1-chloroplast), and the effector papers here supply the other half of the argument by showing what suppression at the plasma membrane looks like.

Compartment recurs at a coarser scale too: legume NLRs are more than twice as tissue-specific as genes generally, and lean towards roots (marques-2026-legume-tissue). Organelle and organ are the same question asked at different magnifications.

Only the chloroplast was followed up. Mitochondria, ER and Golgi were seen and left.