A Fungus That Knows It Is Inside a Plant Because the Plant Cut Its Mail Open

A pathogen that has just broken into a plant faces a timing problem. Growing aggressively is expensive and exposes it to the plant's defences, so it should not start until it is actually inside. But a fungus has no eyes. It has to infer its location from chemistry, and the chemistry outside a leaf and inside one are not obviously different enough to be a reliable signal.

The corn smut fungus has solved this by getting the plant to confirm the break-in. A group at Heinrich-Heine University Düsseldorf has worked out that the fungus secretes a protein, Pit2, with a short peptide folded away inside it, and that the peptide only gets released when the plant's own defensive enzymes chop the protein up. That peptide then docks onto a receptor on the fungal surface, and the fungus starts proliferating. The plant's immune response is the doorbell.

Why it matters: Corn smut is a crop disease, and the receptor is the sort of thing a fungicide could target. The wider point is about strategy: the fungus has made its growth signal depend on an enzyme family the plant cannot switch off, because the plant needs those enzymes for immunity and development.

The study, by Mendoza-Rojas et al. at Heinrich-Heine University Düsseldorf, appears in Nature Communications. The work combines a yeast reporter assay, protein structure determination, mutations in both the peptide and the receptor, and infection tests in corn.

A signal the sender cannot release

The receptor has been hiding in plain sight. A gene called pit1, sitting next to pit2 in the genome, was known to be essential for infection but annotated only as a predicted membrane protein of unknown function. It is in fact a G-protein-coupled receptor, the same architecture as the receptors that handle most human sensing and are targeted by something approaching 40% of approved drugs. The authors rename it Gpe1.

What makes the system clever is where the ligand sits. Solving the structure of Pit2 showed the activating peptide buried inside a Pit2 pair, which the authors read as a scaffold that prevents the fungus from setting off its own receptor prematurely. Applying whole Pit2 to yeast carrying Gpe1 produced only a weak response, around a quarter of the maximum. Cutting Pit2 up first with a protease raised that to full activation at a tenfold lower concentration. The protein is inert until something external opens it.

That something is the plant. Papain-like cysteine proteases are broadly conserved across plants, strongly induced by infection and stress, and already known to cleave Pit2 into fragments. One fragment inhibits the proteases themselves, which is how Pit2 was already understood to work. The finding here is that a different fragment activates Gpe1, so the same host enzyme both gets blocked and unwittingly issues the growth order.

Evidence the coupling is real

The structural claim is backed by mutagenesis on both halves. Altering residues in the peptide changed how strongly it activated the receptor, with some substitutions nearly abolishing the response. On the receptor side, changing a conserved arginine or a pair of adjacent asparagines either weakened activation substantially or removed it entirely. The same receptor region resembles the activation machinery of rhodopsin-like receptors in animals, which is a notable piece of deep conservation.

There is also a co-evolution signal. Comparing relatives, a Pit2 from one smut species activated the receptor from another only weakly, at 30% and 45% of full activity. Sender and receiver have been tuned to each other, which is what you would expect if the pair has been evolving together rather than being borrowed.

What the study can't say yet

Most of the dose-response work is in yeast engineered to report receptor activation, not in the fungus on a plant. That is a sensitive and controllable readout, but a receptor's behaviour in a yeast reporter is not the same as its behaviour on a hypha in corn tissue.

The authors are careful about their own mutants. Where a change reduced activation, they note that reduced activity of individual variants may also reflect effects on receptor stability or intracellular trafficking competence. In other words, a weakened response does not prove you have hit the binding site; the protein may simply be less stable or not reaching the surface.

The biggest open question is one they pose themselves. Pit2 cleavage yields more than one peptide, and the two known products pull in opposite directions: one shuts down the plant's proteases, the other tells the fungus to grow. How that balance is struck during a real infection is unresolved, and they state that Future work has to investigate the individual contribution of these two peptide versions during infection. Until that is settled, the model is a plausible account of a mechanism rather than a measured description of what dominates in the field.

The suggestion that host-gated signalling of this kind may be widespread in other pathogenic and symbiotic fungi is explicitly framed as a possibility worth exploring, not a result.

Quick questions

Could this lead to a treatment for infected crops? Possibly in time. A receptor the fungus needs in order to expand is an attractive target, but nothing here tests a compound against it.

Why would a fungus depend on the plant to activate its own signal? Because it makes the signal trustworthy. A cue the fungus could generate by itself would not distinguish inside from outside; one that requires a plant enzyme can only appear once the fungus is genuinely in plant tissue.

What is the one-line takeaway? A crop fungus hides its growth signal inside a secreted protein and lets the plant's own defensive enzymes release it, so the plant's immune response doubles as the fungus's confirmation that it has got in.