Queen ants do gene therapy to control their brood

One genome produces ants of many different types, but how?
08 September 2025

Interview with 

Laurent Keller, University of Lausanne

Harvester_Ant_Laurent_Keller.jpg

A harvester ant

Share

Who invented gene therapy? You’re probably scratching your head in search of the name of some or other biologist, but the answer is, it definitely wasn’t a human! Because, scientists have discovered that, by feeding pieces of genetic material called short RNA molecules to their nest mates, some ant species can use this “diet” to manipulate how these social insects produce a repertoire of different ants with a range of different appearances and roles - like workers, soldiers and queens - from just a single genome. As he explains to Chris Smith, Laurent Keller has been looking at harvester ants, which were either allowed or denied access to a specialised form of ant egg used as a foodstuff and known as “trophic eggs”. These are devoid of genomic DNA, so they’re reproductively inviable, but they are packed with calories and, as it turns out, pieces of genetic material that can be absorbed to alter gene expression and the development of the larvae that eat them…

Laurent - We were interested in understanding how there were variations in size between workers and ants. Typically, you can have smaller or larger workers, and we did cross-fostering experiments. And while doing these experiments, we found something bizarre, which was that when we did transfer and remove the trophic eggs, trophic eggs are a type of eggs without genetic information, which cannot develop, produced by queens or workers, and which were thought to be a source of food for the larvae and maybe the queens.
And so when we did transfer eggs between colonies, we found that when there were no trophic eggs, there were fewer queens produced in these colonies.

Chris - So just to summarise that for a minute then, so in an ant colony, you have queens that lay eggs, but also workers can lay eggs, and they can lay one form of egg, a trophic egg, which doesn't have any genetic material in it, so it's never going to turn into an ant. So it looks like an egg, behaves like an egg, but doesn't ever turn into anything, but they can eat it. And when they eat it, it appears to change the outcome for the ants.

Laurent - Exactly. But should we not explain that within the colonies, you have individuals, females, which can be different morphologically, like the queens and different types of workers, and that usually this is due to the environment. So depending on what is fed to the larvae, you have development of the larvae into a queen, a worker or a soldier, for example.

Chris - Do you therefore believe that there's something, some signal, which is embedded in those eggs, which is affecting the appearance or phenotype of the ants when they eat them?

Laurent - Yeah, it must be so, because in the trophic eggs, they look a bit the same as viable eggs, but their content is different. They have less food inside, which was surprising when we did analysis, than in viable eggs. And what we found on the contrast that there was more small RNA, so small RNAs are short pieces of genetic material, which have genetic information. And so there was more small RNAs and more long RNAs. And so this was quite surprising, because trophic eggs should not have more genetic information if it's just a source of food.

Chris - Do you believe then that those short pieces of genetic material are being absorbed into the ants that consume the trophic eggs and that they're influencing gene expression and therefore changing the appearance of the ants that consume the eggs?

Laurent - Yes, so small RNAs are known to be, easily go from the food to the body, and they can influence different things. So this has been shown in other contexts, that they can influence the development or the behaviour of organism. So it's very likely that small RNAs and long RNAs in trophic eggs influence the development of larvae.

Chris - So these ants are effectively doing gene therapy on their nest?

Laurent - Exactly, yes, they do so. And so the queens can modulate development of the other members in the colony. So whether larvae will develop into a queen or a worker by producing variable amounts of those trophic eggs.

Chris - Have you been able to prove that though? Can you show that these RNAs are functional, and when they go into the ant that consumes them, gene expression is modulated?

Laurent - No, that will be the next stage to be done, to really demonstrate mechanistically that those small RNAs, so we identify some of them, and now the next stage will be to inject small RNAs into larvae, and be able to manipulate development with only this factor being different between larvae.

Chris - But is the recipe for a trophic egg always the same? So if you take a trophic egg, is the different relative proportions of these short RNAs that do this manipulation, is it always the same? Or at different stages of the nest life cycle, do the relative levels get changed? So the queen can enrich for some of these signals and suppress others to change the outcome when the ants consume these eggs?

Laurent - This is a very good question. And we don't know. And it's true because now we know that trophic eggs can influence the development of larvae. One could imagine that queens could manipulate or influence the development of larvae or other things in the colonies in many ways by producing different types of trophic eggs. But this will be something to be studied.

Chris - It's extraordinary this because previously we thought this was all down to pheromones and secreted chemicals, didn't we? And so this is a really dramatic change to our view as to what might be going on.

Laurent - This is a strong belief, but nobody has identified those pheromones. The only thing which has been demonstrated is that the presence of the queens tend to suppress production of new queens. People have said it's pheromones, but there's not a single study which demonstrates that. So this is a strong belief in the field, which is based on no data.

Chris - So you might have a really disruptive discovery here?

Laurent - Yeah, I think so. And it's very likely, I think it's really worth to test our communities in other ant species and maybe other social insects.

Comments

Add a comment