Is it possible that you are more closely related to a chimpanzee than to another human? Ok, that's a bit of a loaded question. It depends on how we define 'related', or, more specifically, what we are measuring. If, for instance, you compared the anatomy or physiology of a human and a chimp, the conclusion would be obvious: all humans are more closely related to each other than any human is to a chimp. The same goes for a comparison of the entire genome. But what if we look at a small portion of the genome? This week's evolgen Phylogeny Friday deals with this question. More after the jump.
The major histocompatibility complex (MHC) consists of approximately 100 genes that code for proteins involved in immune response. The human versions of MHC genes are known as human leukocyte antigens (HLA). The antigens encoded by the class I HLA genes interact directly with infectious parasites. There is an evolutionary arms race between humans and their parasites. The human immune system wants to recognize and eliminate any invaders, and the parasites are trying to evade capture by the immune system. The class I antigens are constantly evolving in order to keep up with the changing landscape of human parasites.
The three class I genes are extremely polymorphic. It is thought that this polymorphism is maintained by natural selection because the allelic diversity allows for the recognition of a wide array of parasites. The polymorphism at one of the class I genes is so ancient that it extends beyond the speciation event between humans and chimpanzees. In the phylogeny below (taken from this paper by Austin Hughes and colleagues) I have underlined sequences from humans in red and sequences from chimps in blue.
The clade on the upper half of the image contains only human sequences, indicating that those alleles arose after the human chimp split (or the alleles had yet to be sampled in chimps, or they were lost in the chimpanzee lineage). The lower half of the phylogeny, however, contains a mixture of human and chimpanzee genes. If this gene tree corresponded perfectly with the species tree, we would expect all of the human sequences to cluster together and all of the chimp sequences to cluster together. The mixture of human and chimpanzee sequences indicates that some of the polymorphism in this gene existed prior to the speciation event that gave rise to the human and chimpanzee lineages and was maintain. Further statistical analyses have revealed that the polymorphism is maintained by balancing selection.
So, if you pick the right sequence to examine, you may find that you are more closely related to a chimpanzee that you are to me. This example shows why it is important to consider the evolutionary context of the gene from which you are creating a phylogeny. The common adage that gene trees do not necessarily resemble species trees holds very true in this case.
Hughes, AL, MK Hughes and DI Watkins. 1993. Contrasting Roles of Interallelic Recombination at the HLA-A and HLA-B Loci. Genetics 133: 669-680.
- Log in to post comments
beautiful and simple illustration about a basic point. Although I bet you and I take two drastically different interpretations of this. Just like we do on systematics, despite agreeing about the importance of molecular data.
Personally, I feel data like this shows that genes make a poor 'unit of selection' if we're going to be talking about things in their ecological context.
Either an individual is an aggregate of 1000's of 'units of selection' that happen to reside in the same body for a few decades, or a person is a 'unit of selection' comprised of the 1000's of genes that make him him.
*shrug*
I think the ultimate unit of selection is the entire (diploid) genome. Everything outside of that (from the intracellular world to the things outside of the organism) make up the environment. But we can also simplify our model (to make it easier to work with) and consider a portion of the genome as a unit of selection. In this case, the rest of the genome becomes part of the environment (due to epistasis, genome structure, etc).
I think this example shows that genes *are* a good unit of selection. It even jives well when put into the context of the environment. In this case, we see that selection has maintained variation at this locus across species boundries. If we use other loci to come with a consensus phylogenetic relationship between taxa, we can then use the non-consensus loci to tell interesting stories.
Ah, we're pretty much in agreement then.
My problem is with the Dawkinsian overly-simplistic 'selfish gene' concept where the gene is everything and the aggregate is nothing.
The modern synthesis gets a lot right, but kinda has a 'can't see the forest for the trees' sorta thing going on.
Sean Carrol's closer to a more realistic view in Endless Forms
And Niles Eldridge was pretty insightful in Rethinking Darwin as well.
Thank you for highlighting and discussing this topic!