A mystery gene in human brain development: discovering ARHGAP11b

The human brain is often described using literally cosmic superlatives. Here is V.S. Ramachandran, a renowned neuroscientist:

The human brain, it has been said, is the most complexly organised structure in the universe and to appreciate this you just have to look at some numbers. The brain is made up of one hundred billion nerve cells or "neurons" which is the basic structural and functional units of the nervous system. Each neuron makes something like a thousand to ten thousand contacts with other neurons and these points of contact are called synapses where exchange of information occurs. And based on this information, someone has calculated that the number of possible permutations and combinations of brain activity, in other words the numbers of brain states, exceeds the number of elementary particles in the known universe.

That's some serious complexity there, yes, but hidden in the description is a mundane reality: the human brain is made of neurons which make synapses with other neurons, which means that it's made of the same stuff as the brain of a sloth or a goldfish or an earthworm. (Or even a parasitic mite on an earthworm.) Still, whether or not the human brain is really special or just big, something has caused it to grow in ways that differ from its predecessors. Especially at key junctures in the process, human brain development must depart from boring old mammalian brain development. And this should be reflected by — and perhaps explained by — patterns of gene expression.


Wieland Huttner's group (at the Max Planck Institute in Dresden) has been studying brain development for a couple of decades. Recently, they did a straightforward but fruitful experiment designed to detect human-specific patterns of gene expression during brain development. They looked specifically at genes that might underlie the expansion of the cerebral cortex in the human brain. The paper is "Human-specific gene ARHGAP11B promotes basal progenitor amplification and neocortex expansion," by Florio et al., published in Science in early 2015.

It's worth taking a moment to consider the human cerebral cortex. The cortex is the wrinkled stuff that covers nearly the whole brain. A glance at the human brain can give the impression that it's a big bowl of wrinkled gelatinous glop. But that wrinkly stuff is really just at the surface. A deeper look presents a weirder story, in which the wrinkly stuff has oozed out over the surface of the brain, growing out of control until the whole outside of the brain is wrinkly cerebral cortex.

That's pretty much what happened, specifically in mammals. Look at the picture on the right (click to enlarge). The wide range in brain size is obvious at a glance. If you look more closely, you can see that the wrinkly stuff dominates in many mammals, masking the fact that the basic organizational plan of a tetrapod brain is unchanged. The elements of the cortex are very old, surely older than tetrapods. But the expansion of the cortex is largely a mammalian thing. And the expansion in humans is very striking, to the point that at least 75% of the adult human brain is cortex.

So the mammalian cortex is big. But the human cortex is, well, super big. And we'd like to understand how that happens during human development. Well again, in some sense it's just the same old tissue, made of the same old cells, as a tiny fish brain. So maybe the main difference between a big brain and a little brain is the number of times that the relevant cells in the embryo underwent cell division. If you think about it, that could be sufficient to explain how any tissue in any animal gets to its final size. This is an approximation, but if every cell in a particular organ divides once, the organ doubles in size. And since tissues and organs are built by generations of cells, then it follows that increased division of early-generation cells can make a significant difference in how big a tissue or organ can become. Here's what I mean: if a single founder cell divides once, and then each cell in each subsequent generation divides once, then after 10 cell generations we have 2^^10 cells, which is about 1000. You can double that by simply using two founder cells instead of one. And so on. The point is that the awesome, massive human cerebral cortex owes its existence, at least in part, to the behavior of the founder cells. Somewhere along the line, probably early in the process of brain development, something changed about those founders. Either there are more of them, or they are more prolific in their division, or both.

So, back to the work by the Huttner group. They decided to compare the brain founder cells of humans to the same cells in mice. As far as we can tell, humans and mice build the cerebral cortex in the same basic fashion: some founder cells are established, then these cells divide to make subsequent generations of cells that also divide, then other really interesting things happen, and then you have this multi-layered computational marvel called the cerebral cortex. The mouse cortex is less obviously layered and of course it's far smaller, but otherwise it seems the same. A comparison, then, could reveal how the human process differs.

Comparisons like this have been done before, and have revealed some influences that can begin to explain the big mammalian cortex. But it's harder to do the comparison directly to humans, and the previous work had never looked directly at some of the most important founder cells. These are called neural progenitor cells, or NPCs. I will just call them founder cells.

So Huttner and colleagues started by developing a way to get a pure sample of founder cells from the relevant stage of development. Using a combination of fluorescent dye probes and a technology called cell sorting, they winnowed the founder cells out from among all the other cell types in the developing brain, to the point that they not only could collect founder cells in pure form, but even could subdivide them by type. Specifically, they collected a kind of founder cell called basal radial glial (bRG) cells. These cells are known from years of study to be important for cortex expansion, but no one had ever gotten human bRGs alone in a test tube. This is the technical feat achieved by the Huttner group, and was independently accomplished by another group (Chris Walsh and colleagues at Boston Children's Hospital and Harvard Medical School) at the same time (both papers were published at about the same time in 2015).

Once they had the special human founder cells purified, they could compare them with the same cells from mice to ask that straightforward question: what, if anything, is different or special about human founder cells in the cerebral cortex? This is done by identifying all of the genes that are turned on in those cells, and that's why they needed to have a pure sample of just those cell types.

First, they described a set of genes that aren't human-specific but are turned way up in the human founder cells. The results of that analysis are interesting but not mind-blowing: they found genes involved in cell-surface processes that were known to be a little different in human cortex. And more to the point, these are genes that are present in mice (and thus probably in all mammals). So, yes, these are interesting clues to how the human cortex is built. Evolutionarily speaking, however, the genes aren't new or unique.

But then Huttner and colleagues looked for human-specific genes among the ones activated in the human founder cells. I know I've mentioned this a few times already, but I think it is remarkable that they found human-specific genes at all. In fact, there were 56 genes that were both human-specific (compared to mouse) and turned on in founder cells more than in other cell types. Because they wanted genes that were truly specific to the founder cells and not to any other cell in the brain, they set some strict criteria that pointed to exactly one human-specific gene that was turned on strongly and specifically in the human founder cells. That gene was ARHGAP11b, a gene that had not been described before, in any experiment on any cell or tissue.

That's how ARHGAP11b was discovered — in a deliberate search for genes that are turned way up in the founder cells that generate the cosmically huge human cerebral cortex. The scientists' success relied on two main advances: their own technical achievement of getting pure samples of human founder cells, and the existence of technologies that allowed them to identify and measure the expression of every single gene in those cells. That technology is called RNA-seq, and it didn't even exist until 2008.

I still haven't told you what ARHGAP11b does. Patience!

---

Image credits:
Four primate brains, Figure 2 from Hofman (2014) Evolution of the human brain: when bigger is better, Frontiers in Neuroanatomy.
Mammalian brain assortment, Figure 3 from Herculano-Houzel (2009) The human brain in numbers: a linearly scaled-up primate brain, Frontiers in Human Neuroscience.
Labeled founder cells in developing mouse cortex (notice bRGs), Figure 1C from Florio et al., linked and cited above.