Expanding the Brain. Literally.

  

BY DEREK LOWE

I have to begin by saying that this story gives me the shivers at the end of it. Read on and see if it does the same for you,

First, a bit of background. Most of our functional genes can be traced back to, well, other genes. A common route is some sort of genetic duplication event, where various copies then start to diverge on their own. That’s how generally how we end up with closely related subtypes of enzymes and receptors, and you can trace these things back evolutionarily to get an idea of when the original split might have occurred. Serotonin receptor subtypes, for example, go back to around the time that vertebrates diverged from invertebrates, and they have been useful in more than just nerve function. Muscarinic receptor subtypes are also a vertebrate thing, with increases during the known “tetraploidization” events in evolutionary history, but zebrafish for some reason doubled their muscarinic count one more time than the rest of us backboned creatures (at the 3R event when the teleost fish appeared) and have ten such receptors rather than five. And so on.

But there are a few genes that don’t fit this pattern, the so-called “de novo” sequences. As the name implies, these seem to have evolved from previously noncoding DNA, which on the face of it seems extremely unlikely. This is a very good overview of the subject, and it breaks down the how-can-that-ever-happen intuition into three parts. The first is sparsity: we tend to believe that only a very small number of possible sequences would produce some sort of usable biological effect. The second is the idea that nongenic sequences are a random sample of possible sequences with no particular biases in them. And the last is the belief that surely not very many such nongenic sequences have had a chance to show effects on the evolutionary time scale, anyway. But although intuitively appealing, none of these necessarily have to be the case, or at least not to the extent that we imagine.

For one thing, it’s become clear that there’s a lot of noncoding DNA in the genome, and that a lot of it gets transcribed, at some level, into RNA. So the last of those three suppositions isn’t quite right; a lot of sequences get a chance to do something. There are several of these de novo genes whose sequences overlap in some way with commonly expressed genes, too, so they have had a lot of chances. And as for the first two suppositions, it appears that structural motifs like alpha helices, beta-sheets, and lipophilic transmembrane domains are not that hard to generate, even from random sequence libraries. The reason these motifs are used so much in natural protein biology may well that they are just plain abundant. And the barriers to cellular function are lower than we tend to think they are: the undoubted functional importance of things like microRNAs show that you don’t need large amounts of well-formed tertiary structure (and indeed, some of the de novo genes seem to function by acting as “sponges” for particular miRNA species).

So at this point the evidence for such genes is very strong indeed: there aren’t that many of them, but they are real. There are quite a few of them that have emerged in humans versus other vertebrates (and in humans versus our evolutionary close ape relatives), but this might be because we have studied human sequences so intensely.  Another trend that shows up is for many of these genes to be involved in carcinogenesis or tumor maintenance. But a new paper (with commentary here at Science) suggests a more direct connection with the emerged-in-humans observation, and here’s where things get unnerving.

These authors have been looking for years now at de novo genes in a “transcription first” manner: they’re looking at long noncoding RNAs in other species (especially rhesus macaques) and seeing if these have similarities to the known de novo human genes. This would suggest that some of these lncRNAs became somehow “translatable” along the way. lncRNA species tend to be more concentrated in the nucleus than the cytoplasm, and this new paper identifies some structural elements (such as the U1 binding motif in the sequences) that are associated with more localization in the nucleus. Mutations in these features may be what allow these RNA species to escape the nucleus and move out to where the ribosomes are, and thus get translated into proteins. These localization effects are supported by recent work in other groups as well.

A search for human de novo genes that had both lncRNA homologues in monkeys and had such mutations that could lead to nuclear exit turned up 29 genes that are shared between humans and chimpanzees, and 45 that are exclusively human. A closer look at nine of these that are known to be active in central nervous system tissue showed that their expression affects the size of cortical organoids grown from stem cells in culture (their overexpression makes these cultured neuron clumps grow somewhat larger, and their absence makes them smaller). The authors then inserted these gene sequences into mice, which animals showed “significant cortical expansion” as they matured. The Science commentary linked to above features a statement from one of the paper’s authors that a paper is coming that shows that these animal do indeed perform better in tests of cognitive function and memory. They are smarter mice, with larger brains. And now the hair may be rising up on the back of your neck, as it did on mine.

At the same time, it’s important to realize that there are a number of genes involved in the expansion of our brains relative to other species. Similar experiments have been done with some of the non-de-novo ones, such as ARHGAP11B, NOTCH2NL, TBC1D3, and more (all three of those seem to have arisen by good ol’ gene duplication). But this latest work is a particularly direct example of a clear break between humans and other species, and suggests that these de novo genes managed to fit well into an existing network of brain-development signaling and to potentiate it even more. It has been a longstanding question, how human brains managed to expand so dramatically, but it appears that we’re on the way to answering that one. The “why” part of the question is answered the same way every other question in evolution is answered: because it worked. A larger cerebral cortex seems to have conferred survival advantages, and here we all are with our large brains, having overrun the earth.

There are a number of very obvious experiments suggested by this work that we are going to have to be very cautious with. What happens if you splice a suite of these human-brain-only genes into mice, instead of just one at a time? What happens if you increase expression of one or more of them, with a different promoter or through adding more than one copy? We are going to be having some very interesting debates about quantifying animal intelligence. And we will want to practice very good laboratory hygiene as well, because I think we can all agree that it is not in our best interest to allow the earth’s mice and rats to become any more intelligent and wily than they are already. I wish that I were joking about that, but I’m not.

Nor am I joking when I think about the human implications. As the world knows, we have already seen irresponsible attempts to do germline editing and produce altered human babies. When will someone try adding in more of the cortical-expansion genes, to see what happens? I’ll leave it there.

https://www.science.org/content/blog-post/expanding-brain-literally