Most crop plants defend against soil bacteria. Their root receptors detect bacterial signals and activate the immune system — inflammation, cell wall reinforcement, the standard toolkit of rejection. Legumes do the opposite. When legumes detect the same class of signals from nitrogen-fixing rhizobia, their root receptors activate a symbiosis program — bacterial entry, nodule formation, nitrogen exchange. The plant opens the door instead of sealing it.
These two responses — immune rejection and symbiotic acceptance — were understood as fundamentally different systems. Defense and cooperation as separate pathways with separate evolutionary histories. The legume's ability to fix nitrogen appeared to require an entirely novel molecular apparatus built over tens of millions of years of co-evolution with rhizobia.
Andersen, Radutoiu, and colleagues at Aarhus University found the boundary between these systems (Nature, 2025). They identified a region in the root cell-surface receptor they call Symbiosis Determinant 1. This region functions as a decoder: the same bacterial signal arrives at the same receptor, and the decoder determines whether the plant reads it as threat or partner. The decoder is not a complex regulatory network. It is two amino acids.
By changing those two residues, the researchers converted a defensive receptor into one that supports nitrogen-fixing symbiosis. In Lotus japonicus, the modification reprogrammed the immune response into a cooperation program. In barley — a cereal that has never in its evolutionary history formed nitrogen-fixing partnerships — the same modification activated the symbiotic pathway. The machinery was already present. The decoder was set to defense.
The structural observation is about the distance between opposed biological outcomes. Immune rejection and symbiotic acceptance are not separated by a deep evolutionary gulf — different gene families, different signaling cascades, different receptor architectures. They are separated by a two-amino-acid setting in a shared receptor. The receptor reads the same input signal. The downstream pathways are available in both configurations. The system is not choosing between two separate machines. It is choosing between two settings on the same machine.
This matters because the assumption of deep separation shaped the research program. If defense and symbiosis were truly distinct systems, engineering nitrogen fixation into cereals would require transplanting the entire legume symbiosis pathway — dozens of genes, regulatory networks, signaling intermediaries. Decades of work have pursued this approach. What the Aarhus group found is that the recognition step — the molecular decision between “reject” and “accept” — is not the hard part. The downstream capacity is already there. The wall between hostility and cooperation is not a wall. It is a gate, and the gate's latch is two residues wide.
The general pattern: when two biological outcomes appear fundamentally opposed, the molecular distance between them may be vanishingly small. The opposition lives in the phenotype — the organism either attacks the bacterium or houses it — but the mechanism producing that opposition can be a minimal switch in a shared system. The drama of the outcome does not predict the complexity of the mechanism producing it.