Shiny Things/via flickr
What's the genetic basis for the many ways that humans and chimps are different from one another?
What's the genetic basis for the many ways that humans and chimps are different from one another? Shiny Things/via flickr
When the protein-encoding genes of the human are compared with the protein-encoding genes of the chimpanzee, they are about 99 percent the same. Moreover, the one percent that are distinctive aren't obviously interesting, being involved with such traits as sperm surface proteins and immune responses.
So, given this, what's the genetic basis for the many ways, notably the cognitive ways, that humans and chimps are different from one another?
A most ingenious approach to this question is being developed in the lab of Katherine Pollard at the University of California in San Francisco. To understand their experiments, we first need a crash course in genes and embryos. I'll try to make it quick.
Our single-celled ancestors who lived more than 1.5 billion years ago were already impressively gene-rich and sophisticated, as per last week's blog. Notably, their genomes encoded a rich toolkit of regulatory molecules that turn on or off the expression of genes as appropriate to the occasion. For example, in the presence of bacterial food, the ancient ancestors turned on the expression of genes that allowed them to crawl around and engulf their prey. When things got lean, they instead turned on genes that allowed them to swim off to find new food sources.
The way these switches work is pretty straightforward to explain, albeit exquisitely intricate in detail. Basically, proteins are encoded by sectors of DNA called genes. Contiguous to each gene is DNA that doesn't code for protein; instead, it functions as the gene's on-off switch. When regulatory molecules bind to this switch DNA, the contiguous gene is either expressed or prevented from being expressed. So, very crudely, the thing on the wall is the switch DNA, your finger is the regulator, and gene expression is the light turning on or turning off.
The common-ancestral selves were unicellular, whereas the animal lineage has elected to construct multicellular selves. In making this transition, animals hung on to the same switch arrangement and the same sets of regulators used by the ancestors, but they added a splendid additional idea. In addition to being responsive to signals from the environment, they also became responsive to signals coming from their very own cells. So, to highly oversimplify the situation, after a fertilized egg has divided into two and then four, then eight, then 16 cells (where the human has, gulp, ten trillion cells), cell #16 makes a regulator that acts to switch on a set of unique genes in cell #10, the outcome being that cell #10 and its progeny eventually give rise to nervous tissue. Meanwhile, cell #11 expresses a different suite of genes, poising its progeny to influence yet other cells to differentiate into muscle.
As animals, and hence animal embryos, complexified over time, these cell-to-cell interactions have become increasingly impressive. In the developing mammalian brain, for example, neurons migrate up into the cranium, using much the same kind of amoeboid movement that our deep ancestor employed to capture bacteria. Neurons that reach a particular destination switch on genes that allow them to secrete a nerve growth hormone. As the next phalanx of neurons migrates into the region, they follow the hormone gradient, akin to male moths moving up pheromone gradients to find females, avidly competing for hormones that will enable their proliferation. The first to arrive at the pulsating source proceed to form synaptic connections with their targets; any laggards, by contrast, fail to proliferate and instead degenerate.
Granted that this is an absurdly simplified account of brain development, it suffices to make a key point, which is that brains build themselves. Bottom up. When A happens, that allows B and C to happen; B allows D and E to happen; and so on.
Because brain development is so contingent on what has gone on before, it's pretty easy to alter what happens. For example, if the pioneer neurons in our example carried a switch mutation that prevented them from secreting the nerve growth hormone at the appropriate time, the next phalanx of neurons wouldn't move towards them and might, instead, pick up on a more distant hormonal signal from another brain region and move in that direction, forming synapses with a new set of neurons altogether. A brain is still constructed, but it will have different kinds of neural pathways and connections and hence, perhaps, different ways of doing things.
So now we can return to the chimp-human question. If the chimp and human protein-encoding genes are virtually all the same, then are there any interesting differences in their switch regions? Given the bottom-up nature of development, mutant switches could have large-scale consequences.
The identification of switch sequences is much more computationally challenging than the identification of genes, but the Pollard lab is leading the charge.
Basically, they compare the DNA sequences adjacent to genes that are found not only in humans and chimps but also in mice and rats, where the most recent common ancestor of these four mammals roamed the planet some 60 million years ago.
The Pollard logic is this:
1) If a given set of sequences isn't doing anything important, which is usually the case, then the rat, mouse, human, and chimp versions are expected to be very different from one another. That's because they aren't under selection, so they tend to accumulate mutations. In genomic lingo, the sequences are said to "drift."
2) By contrast, if the sequences function as switches, then they are expected to be very similar because they are under selection to maintain their gene-regulating function.
3) Of particular interest are cases where the mouse, rat, and chimp sequences are all identical, indicating intense selection to maintain them, whereas the human sequence is markedly different from the other three. The Pollard lab has thus far identified 202 such cases, where each is called a human accelerated regions or HAR.
Now the fun begins. A researcher picks out a HAR (e.g. HAR34), figures out what gene it's contiguous to, and then asks: Where and when is that gene switched on/off during embryological development? And then: Is its expression pattern different in the human than in mouse, rat, or chimp? If it is, then the novel pattern may prove to be relevant to an understanding of how humans are distinctive creatures.
Thus far there are three preliminary stories relevant to the brain. HAR1 proves to mark a genetic region that is expressed early in the development of the neocortex; HAR152 is near the gene encoding a protein called neurogenin-2 that is expressed in a region of the hippocampus with a central role in learning and memory; and HAR2 is near a gene with strong expression in the hand, perhaps playing some role in human-specific hand coordination.
The knee-jerk response to this account is to think "Aha — maybe some of those novel HAR sequences are running some new human-specific brain module or widget! Like my consciousness!"
But if we circle back to our core notion, that brains build themselves, and think about the HARs in this context, then we realize that we're not likely to be talking about new modules or widgets. Just as in our hypothetical example, where a group of neurons failed to secrete a hormone and the second phalanx of neurons wandered off to find new targets, a mutant HAR is more likely to result in some human-specific pattern of regional brain differentiation. Indeed, going back to the finch-song domesticated-ape story told here and here, some of the mutant HARs may have the effect of releasing constraints on ape-brain organization, opening things up to greater novelty and plasticity.