Bacterial dormancy

Why does a bacterium that could keep dividing stop growing and seal itself inside a spore instead?

A bacterium's whole advantage is speed. Doubling every twenty minutes, a lineage that keeps dividing while a rival stops will bury it. So sporulation looks like the worst move available: the cell spends hours building a structure, ceases all growth, and — this is the part that should stop us — the mother cell that built the spore then bursts and dies to release it.

Whatever a spore is, it is not that cell saving itself. So what problem is it solving, and for whom?

Start with the obvious answer and try to break it. The obvious answer is that starvation triggers sporulation. Plausible; but follow it through. Building a spore is heavy construction — a fresh chromosome copy, an asymmetric septum, a second membrane wrapped around the forespore, a cortex, several protein coats. In Bacillus subtilis the programme runs something on the order of eight hours, and every step needs energy. A cell that has genuinely run out cannot afford to start.

So the trigger cannot be starvation itself. It has to fire while reserves remain — which makes it a forecast, not a reaction. And a forecast can be wrong, which is expensive. Does the cell treat it as a bet rather than a certainty?

It does, in two ways. The decision stays reversible for a long while: if food returns before the asymmetric septum closes, the programme aborts and growth resumes. And some cells buy delay by killing their siblings — cells that have begun the pathway secrete the sporulation-delaying and sporulation-killing factors described in B. subtilis, which lyse genetically identical neighbours and feed on the remains. You do not evolve fratricide to defer an option unless the option is a last resort.

Now the sharper puzzle. Faced with the same failing medium, a clonal population does not do the same thing. Some cells sporulate, some sit in stationary phase, and the split follows no difference we can find between them — it comes out of noise in the phosphorelay that activates the master regulator Spo0A. Why would selection preserve noise in so consequential a decision?

Because the future is genuinely unknown, and the two errors are not symmetric. If the drought lifts in an hour, the sporulators lose a round. If it lasts a decade, the others lose everything, and a lineage that loses everything once has no later rounds to average over. What matters is growth compounded across generations, and against an uncertain, all-or-nothing environment the strategy that maximises that is not the best single guess but a split bet. The heterogeneity is the adaptation.

Then let me falsify my own account. If hedging against uncertain collapse were sufficient, every bacterium in a variable habitat should sporulate. Escherichia coli never does; it rides out bad times in a stationary-phase state governed by the sigma factor RpoS, far cheaper and far less durable. So bet-hedging explains why some cells commit, not why the hedge takes this extravagant form. The extra ingredient is the shape of the bad times: endospore-formers are largely soil and gut-to-soil organisms, whose lean periods are long, total, and of unpredictable length. Where the lean period is a few hours, a cheap hedge wins.

What is bought is worth naming. The spore core is extremely dehydrated and loaded with calcium dipicolinate — roughly a tenth of the spore's dry weight, though that varies — with its DNA wrapped in small acid-soluble proteins. Low water means almost no chemistry, and almost no chemistry means almost no ageing: heat resistance and longevity arrive together, which is why sterilising means an autoclave at 121 °C rather than boiling.

"Dormant" is not "slowed down". A spore is not running a low metabolism; by every measurement we can make it runs essentially none, which is why its survival is limited by chemical damage rather than by fuel.

Nor is a spore the same thing as a persister cell. Persisters are ordinary cells that happen to be non-growing and so survive antibiotics — tolerance without any structure, and not inherited the way the resistance in this collection's antibiotic-resistance explanation is. The endospore is a built object, and only a few bacterial lineages can build one.

The logic here is the one the weed-seed-bank explanation reaches for, but it settles at a different point. A seed bank hedges across seasons in one patch of soil and germinates on graded cues — light, temperature, disturbance — so it drains gradually and fairly predictably. A spore is nearer to indefinite and its waking is nearly uncued: it responds to a small set of germinants and otherwise waits. One hedge is tuned to a rhythm, the other to a shock of unknown length.

One honest caveat: claims of spores revived from amber tens of millions of years old, or from ancient salt crystals, are disputed and hard to separate from contamination. Survival over decades to centuries is solid; the headline numbers are not.

• Why germination is itself stochastic, so that a spore population never wakes all at once.
• How anaerobes such as Clostridioides difficile use the same machinery to cross open air between hosts.

Every term the collection defines is gathered in the glossary.