20 February 2017

When a GAP is not a GAP: ARHGAP11B, the mysterious human-specific gene

A truly human-specific gene, not merely a human-specific version of an animal gene or a mammal gene or a primate gene — that is something particularly interesting. Given that the human genome is 96% identical to that of our closest relatives (chimps and bonobos), and given that so much of those genomes is composed of mobile elements that are unlikely to end up being genes at all, I and perhaps others long thought that human-specific genes would be something pretty rare.

But there they are — genes by every definition, that code for protein and are expressed in human tissues, that are unique in humans. One of the most interesting is a gene that brings together some of my personal favorite topics in biology: brain development, cellular signaling systems, and of course evolution. The gene goes by the unfortunate "name" of ARHGAP11B.

I do consider ARHGAP11B to be a unique human gene, but its name betrays its evolutionary history and its membership in a family of genes, so it's not completely unique (specifics to come). That family is the family of GAPs, a group of proteins that were the focus of my postdoctoral research years ago. GAP stands for "GTPase-activating protein," and besides being a typical morsel of biochemical jargon, the phrase is a bit of an insult to the roles played by these proteins in cellular signaling systems.

A RhoGAP domain
The signaling system of interest here involves a superfamily of proteins called small G-proteins, or small GTPases. The 'small' refers to their actual size, and the rest describes a biochemical detail of their function. Both aspects of the name are a hindrance. Because while these proteins are indeed small, their roles in multiple biological processes are huge. And while they do in fact engage in the biochemical process of hydrolyzing GTP (it doesn't matter whether you understand that at all), this is not what they do. What they do is throw switches to activate major changes and activities within cells. When the small GTPase is tagged with GTP (again, the details are unimportant), it is on. And when the GTP is cleaved to GDP, it is off. To turn the switch on, the cell has to stick GTP onto the protein. This is done by proteins called GEFs (that stands for guanine nucleotide exchange factors). And when the cell wants the switch off, it uses GAPs. The GAP activates the GTPase, which cleaves the GTP to GDP.

The basic overview of that is: small GTPases are switches that are major players in signal systems inside cells. GEFs turn them on. GAPs turn them OFF.

It's hard to think of a basic cellular activity that isn't controlled by these switches. One famous system is the so-called MAP kinase pathway. This pathway is used in lots of contexts to do lots of things, but one common use is to stimulate cell division. (For this reason, chemicals targeting various players in that pathway are candidates for anti-cancer drugs.) One of the key focal points of the MAP kinase pathway is Ras, and it's possible that you have heard of this famous protein. It's a small GTPase, and in fact the superfamily of small GTPases is called the "Ras superfamily" in its honor. When the Ras switch gets stuck in the on position, the typical result is uncontrolled cell division. This means that Ras is a common oncogene, and whole families of cancer are defined in part by the presence of mutant Ras that is stuck in the on position.

One subfamily of small GTPases takes its name from the first known member: Rho. The members of this subfamily specialize in a set of cellular functions that almost always involve the cell's skeleton (aka the cytoskeleton). This means that these proteins are centrally involved in how cells acquire their distinctive shapes, how big or small they get, and notably how (or if) they move through the body.

This brings us to ARHGAP11B. This human-specific gene encodes a protein that looks a lot like a RhoGAP. It is basically identical to one part of another gene, found throughout the animal kingdom, called ARHGAP11A. That gene serves as a GAP for Rho proteins, and this means that it flips the switch on these important regulators to off. So it would be reasonable to hypothesize that this is what ARHGAP11B does.

But on the other hand, ARHGAP11B is a drastically shortened version of ARHGAP11A. ARHGAP11B is made of a piece of the business end of ARHGAP11A, with a string of other material stuck onto the end. That "string of other material" is a snippet of protein sequence, a string of 47 amino acids (green in the diagram below), that is not found in any other genome. Not in the mouse, not in chimps or bonobos or orangs. The protein made by the ARHGAP11B gene is truly unique to humans (including Neanderthals and Denisovans). Nevertheless, it starts off with about 230 amino acids that match the business end of ARHGAP11A, the part that makes ARHGAP11A a RhoGAP, the part that makes the protein an off switch for Rho proteins. That part, called the GAP domain, is about 250 amino acids long (it's in purple in the diagram below), and ARHGAP11B is only missing the last 26 amino acids of this domain. So still, it would be reasonable to hypothesize that ARHGAP11B functions as a RhoGAP.

Now, I haven't told you the whole story. The way the discovery happened has next to nothing to do with GAPs and domains and biochemical reactions. The scientists who discovered this new human gene were specifically searching for genes that were uniquely expressed in the developing human brain. The gene that really jumped out in their experiment was ARHGAP11B. So they knew that the gene was potentially important at a key stage of brain development. I mention this because once they saw that the gene was related to a RhoGAP, and that it contained a nearly-complete GAP domain, I suspect that they hypothesized that the gene encoded a RhoGAP.

So they tested the hypothesis. And the results explain why I think the story of ARHGAP11B is truly remarkable. The protein does in fact exert an important influence on brain development, which I will discuss in the next post. But it has no RhoGAP activity. The loss of those 26 amino acids from the GAP domain results in the erasure of the biochemical activity that the original gene (ARHGAP11A) always had.

And yeah, that means I made you read a bunch of stuff about RhoGAPs that ended up being irrelevant. But RhoGAPs are cool, so tough. And more seriously, because we don't yet know what ARHGAP11B does accomplish at the biochemical level, it remains possible that Rho proteins are somehow involved.

Why is this case so remarkable? It's this:

The protein encoded by the ARHGAP11B gene is a unique human gene with a new and thus far unknown biochemical function. It clearly arose from a RhoGAP gene, but the protein is not a RhoGAP. It's something else. Something new. And whatever the protein is doing at the level of biochemistry, the result is increased growth of a specific part of the human brain. That's for the next post.


Image credits: RhoGAP domain from the Structural Genomics Consortium, ARHGAP11 diagram from Figure 1D, Florio et al., Science Advances (2016).