Leghemoglobin in root nodules

Why does a legume build an oxygen-grabbing pigment to starve its bacterial guests of the very air they are sitting in?

Cut a healthy clover nodule in half and it is pink inside, from a plant protein that is close kin to the haemoglobin in your blood. The usual gloss is that the plant made a pigment to protect its bacteria from oxygen. But that reading contains a contradiction the moment you look at it: the bacteria inside are respiring hard, and respiring means consuming oxygen. Why would a host go to the trouble of building a high-affinity oxygen trap in order to withhold from its guests the one molecule they need most?

Two demands are in collision, and the whole structure is the resolution of that collision.

The first demand comes from the enzyme. Nitrogenase splits the triple bond in atmospheric nitrogen, and its iron-containing components are wrecked by oxygen — not slowed, but irreversibly damaged, the iron protein within moments of exposure. There is no known nitrogenase that tolerates air. So the reaction has to happen somewhere effectively anoxic.

The second demand comes from the bill. Breaking that bond is one of the most expensive things biology does: the standard stoichiometry runs at sixteen ATP per nitrogen molecule reduced, and the reaction obligatorily throws off one molecule of hydrogen gas alongside the two of ammonia — a quarter of the electrons diverted, unavoidably, by the chemistry itself. That is worth pausing on, because it is a constraint rather than a design: no amount of selection has removed the hydrogen leak, which tells you it is not an inefficiency but part of how the enzyme works.

Sixteen ATP per turnover cannot be paid by fermentation. It requires oxidative phosphorylation, running fast, which requires oxygen delivered fast. So the bacteroid needs a high oxygen flux and a near-zero oxygen concentration at the same time — and that pairing sounds impossible only because we habitually treat those as the same quantity.

They are not. Flux is concentration times how quickly the carrier moves and unloads. Fill the cell with a protein at around a millimolar concentration that binds oxygen tightly, and almost every oxygen molecule present is held rather than free. Free oxygen inside an infected nodule cell sits in the nanomolar range — roughly four orders of magnitude below air-saturated water, and both figures here are recalled rather than derived. Yet the total oxygen in that cytoplasm is large, and the loaded carrier diffuses and hands its cargo to the bacteroid's terminal oxidase. Concentration is held at a level nitrogenase survives; delivery rate is held at a level respiration needs.

That only works if the receiving end is matched to it. An ordinary respiratory oxidase, tuned to micromolar oxygen, would be idle at nanomolar. Bacteroids switch to a high-affinity oxidase whose half-saturation lies in the range leghemoglobin actually permits — so the carrier and the consumer were tuned to each other, and neither makes sense alone.

There is a third component that is easy to miss. The nodule's inner cortex maintains a variable barrier to oxygen diffusion, and the plant adjusts its permeability. That is the coarse control — how much oxygen enters at all — while the pigment is the fine control of what fraction of it is free. It is also the lever a legume pulls when it withdraws support from a nodule that is not delivering nitrogen, which is where this physiology touches the economics of the partnership: the same valve that makes fixation possible is the one that makes the arrangement enforceable.

Three corrections to the tidy version.

The pigment is not an invention. Plants across the kingdom carry non-symbiotic haemoglobins that long predate nodules, and leghemoglobins are recruited from that existing family. What changed is mostly regulatory — where the gene is expressed and how enormously — along with tuning of the binding kinetics. So "the plant built a pigment to solve this problem" is the wrong grammar. A general-purpose globin that was already there was co-opted and over-expressed, which is the ordinary way such things arise and is worth insisting on, because the alternative reading quietly implies foresight.

Nor is oxygen control the legume's special trick. Other nitrogen fixers meet the same constraint by other means: some free-living bacteria simply burn oxygen at prodigious rates to keep it away from the enzyme, paying in carbon; filamentous cyanobacteria build dedicated thick-walled cells that shut down the photosystem that would otherwise make oxygen on the spot; others fix only at night. One constraint, several unrelated solutions — which is the pattern to expect when the pressure is physical rather than historical.

And one detail I would flag rather than assert: where the haem group itself comes from, plant or bacteroid, has been argued both ways, and I do not think it is settled.

The account is unusually easy to break. If the carrier is what supports the flux, then removing it should collapse fixation without any other change — and silencing leghemoglobin in a model legume does exactly that: the nodules go white, free oxygen inside them rises, the bacteroids' energy status falls, and nitrogen fixation is effectively abolished. The refuting observation is correspondingly clean. A nodule with no leghemoglobin that fixed nitrogen at normal rates would show that the pigment is incidental and that something else carries the oxygen; equally, the discovery of an oxygen-tolerant nitrogenase would remove the reason for the entire apparatus.

• How the nodule cortex actually varies its permeability, and how fast it can respond.
• Why the obligatory hydrogen produced alongside ammonia has never been engineered away, and what recycling it costs.

Every term the collection defines is gathered in the glossary.