The human genome is engaged in a similar evolutionary arms race… against itself.

The opponents are jumping genes called retrotransposons that can hop around the genome. They increase in number by copying themselves and pasting the duplicates into new locations. This mobile lifestyle is so successful that retrotransposons make up more than 40 percent of the human genome. Some have settled down, and are now static shadows of their once-active selves. Others are still on the move.

If the copies land in the right place, they could act as clay for building new adaptations. If they land in the wrong place, which is perhaps more likely, they could cause diseases by disrupting important genes. So genomes have ways of keeping these wandering sequences under control. One involves a gene called KAP1. It’s a kind of tranquiliser—it sticks to retrotransposons and stops them from activating.

KAP1 works differently in different species, targeting those retrotransposons that are active in that owner’s genome. Our KAP1 won’t keep a mouse’s jumping genes in line, and vice versa. Some scientists believe that this specificity is caused by another group of genes called KZNFs. They tell KAP1 where to go by searching for, and sticking to, specific retrotransposons. They’re like beat cops that patrol a neighbourhood, look for crime, and radio for back-up. Each KZNF targets a different type of retrotransposon and different species have their own set.

At least, that’s what happens in theory. In reality, we know that many KZNFs exist, but no one had matched any one of them to a specific target. Partly, that’s because these cops do such a good job that it’s hard to see jumping genes in action.

Frank Jacobs and David Greenberg from the University of California, Santa Cruz solved this problem by sticking the retrotransposons in mouse cells—a less policed environment. They filled the mouse stem cells with a single human chromosome. Mice are adapted to control their own retrotransposons, so they’re oblivious to ours. The jumping genes on the human chromosome, freed from their usual restraints, started spreading, much like an invasive species running amok on an island with no native predators. Now, the team could pit different human KZNFs against these restless genes to see if any could bring them to heel.

They found two that could—ZNF91 and ZNF93. Each of these represses a major class of retrotransposons—SVAs and L1s, respectively—that are still jumping about in the human genome today.

ZNF91 and ZNF93 are only found in primates, but they have changed a lot even without our narrow lineage. For example, the human version of ZNF91 has deluxe features that are shared by gorillas but not by monkeys. To understand the value of these changes, Ngan Nguyen and Benedict Paten took the modern genes and worked backwards, reconstructing their ancestral versions at different stages of their evolution.

They found that between 8 and 12 million years ago, ZNF91 gained features that dramatically improved its ability to keep retrotransposons in line. That’s the point in primate evolution before humans diverged from gorillas and chimps. ZNF93 went through similarly dramatic changes between 12 and 18 million years ago, before the we (and the other great apes) diverged from orang-utans.

These results suggest that ape KZNFs have rapidly evolved to keep jumping genes in check. Indeed, the KZNFs are one of the fastest growing families of primate genes. We have around 400 of them, and some 170 of these are primate-only innovations. This expanded police force reflects our ongoing genomic arms race.

And the jumping genes are starting to fight back. For example, the team found that ZNF93 represses L1 genes by recognising a short signature sequence that most of them have. But some L1s, especially the most recently evolved ones, have lost this signature entirely. They can jump unnoticed.

The missing sequence would normally makes the jumping genes better at jumping. But this booster rocket ended up as a wheel clamp, since ZNF93 evolved to recognise it. So some of the L1s lost the rocket. They jumped less effectively, but at least they could still jump.

This is a classic evolutionary arms race. The hosts thrusts, the parasite parries, and the duel continues. But unlike more familiar battles between snakes and toads, or hosts and viruses, this is a case where we’re waging war against our own DNA.

There’s a sense of futility about this. Much of our genome seems to be engaged in an ultimately pointless duel whether neither side can give or gain any ground. But these battles aren’t quite as fruitless as they might seem.

The team found that KZNFs partly suppress the genes around a retrotransposon too. When the cops finds their target, they tell all the bystanders to the lie on the ground too. This is important because it seriously affects the activity of many human genes, beyond retrotransposons. It means that KZNFs can eventually be used to control the activity of genes that jumping genes land next to. (“Excuse me, officer, but while you’re manhandling your suspect, would you mind also rescuing my cat?”) This arms race could have given rise to more complicated networks of genes, and perhaps more complicated bodies or behaviours.

Evolution in action?