Modeling by researchers at the Institute of Science Tokyo suggests that yeast’s “latecomer killing” strategy survives only in environments that switch between abundance and starvation.

When food runs short, some yeast turn chemical warfare on their own kind. Cells that have adapted to starvation release toxins that they themselves can withstand, while genetically identical cells that arrive later, without that adaptation, are killed. The behavior, known as “latecomer killing,” has an evolutionary puzzle at its core: it is costly, and in theory it should not last.

Yet it has lasted a very long time. Budding yeast and fission yeast diverged roughly 300 to 600 million years ago, and both still carry the system and use the same toxins.

A team led by Tetsuhiro Hatakeyama at the Earth-Life Science Institute (ELSI) at the Institute of Science Tokyo now offers an explanation. In a study published in the Journal of the Royal Society Interface, the researchers combined population-dynamics theory with comparisons across yeast species and concluded that the strategy can remain evolutionarily stable only when environments alternate between nutrient-rich periods and starvation.

A system that should collapse

Budding yeast, Saccharomyces cerevisiae, is the yeast of bread and beer, and fission yeast, Schizosaccharomyces pombe, is another workhorse of laboratory biology. The two are so distantly related that finding the same survival strategy in both suggests it has been maintained over an enormous span of evolutionary time.

Yeast are already known for other forms of chemical warfare. So-called killer strains of budding yeast, first described in the 1960s, secrete protein toxins that kill susceptible yeast while remaining immune themselves. Those killers differ from their victims by carrying genetic elements that encode the toxin and immunity to it. Latecomer killing is stranger, because the cells doing the killing and the cells being killed can be genetically identical, differing only in whether they have adapted to starvation.

The difficulty is that both making toxins and resisting them carry costs. Evolutionary theory predicts that such systems are vulnerable to “cheaters”: in this case, resistant cells that stop paying to produce the toxin while still benefiting from its protection. Over time, cheaters would be expected to outcompete producers and erase the system.

Similar dilemmas appear throughout biology, wherever individuals pay a cost to produce something that others can exploit. Explaining why cooperative or costly traits persist in the face of cheaters is one of the classic problems of evolutionary theory.

Yeast have long served as a model for these problems. In a well-known set of experiments, budding yeast cells that produce an enzyme to break down sucrose share the resulting sugars with neighbors, and mutant “cheater” cells that skip making the enzyme can benefit from the work of others. Studies of such systems have shown how the balance between producers and cheaters can depend on the conditions they live in.

Modeling feast and famine

To explore the paradox, the researchers built a mathematical model that simulated yeast populations under different environmental conditions. They tested what happened when nutrients were always abundant, when cells were always starving, and when conditions switched between the two.

“The surprising result was that neither constant abundance nor constant starvation could maintain this system,” Hatakeyama said. “Only when the environment switched between the two did the toxin-adaptation strategy become evolutionarily stable.”

The details of the switching mattered. The system was stable when long nutrient-rich periods alternated with shorter episodes of starvation. According to the researchers, the two phases weed out different threats. The long periods of plenty remove cheaters, while the short starvation episodes eliminate toxin-sensitive cells. Neither phase alone is enough to keep the system intact.

Hatakeyama said the results suggest that “long-term environmental fluctuations can shape the survival of biological strategies that would otherwise disappear.” That points to a role for environmental change in evolution that may be larger than previously recognized, preserving strategies that pay off only under alternating conditions.

Clues in the family tree

The team also compared yeast species. Two species more closely related to budding yeast showed no clear evidence of latecomer killing, even though budding yeast and the far more distant fission yeast both retain it. The authors suggest that whether a species keeps the system may reflect its ecological history rather than simply its evolutionary relatedness.

If that is right, the presence or absence of such systems could serve as a kind of record of the conditions a lineage has lived through over long periods of time.

The model also makes testable predictions. Similar toxin-adaptation systems should be found in microbes that live in environments that are generally rich in nutrients but are occasionally starved. The researchers suggest that environmental surveys and laboratory evolution experiments could test this idea, including in bacteria and archaea.

From microbes to major transitions

The researchers propose that such systems may help illuminate major transitions in evolution, including the origin of multicellularity, when single cells began living and dividing labor together. Strategies in which cells sacrifice or eliminate genetically identical neighbors raise some of the same questions about cooperation and conflict that multicellular life had to solve.

The work is theoretical and comparative. It does not directly show that the environments of ancestral yeasts alternated in the way the model requires, and the predictions about other microbes remain to be tested.

The study’s co-authors include Kunihiko Kaneko of the Niels Bohr Institute at the University of Copenhagen and Kunihiro Ohta and Arisa H. Oda of the University of Tokyo. Funding came from sources including the Japan Society for the Promotion of Science, the Japan Science and Technology Agency’s CREST program, the Japan Agency for Medical Research and Development and the Novo Nordisk Foundation.

---

Notes

1. Hatakeyama, T. S., Kaneko, K., Ohta, K., and Oda, A. H. “Evolvability of the toxin-adaptation system in yeast.” Journal of the Royal Society Interface, 22 July 2026. DOI: 10.1098/rsif.2026.0345.
2. Institute of Science Tokyo. “How changing environments can preserve an ancient microbial survival strategy.” EurekAlert!, 27 July 2026. https://www.eurekalert.org/news-releases/1137047.

Featured image: File image of yeast. Credit: DICP