Tuesday, September 15, 2009

Chromosomes, Drift, and Demes

One of the characteristics of evolution is change in chromosome number and organization. These large-scale changes are often associated with speciation events although it would be a mistake to assume that there's a causal relationship.

One particular chromosomal rearrangement has been getting a lot of press recently because it has been featured on blogs and in some recent trade books on evolution.

Humans (H) have only 23 pairs of chromosomes while most other apes, such as the chimpanzee (C), have 24 pairs. Evidence for a fusion of two of these ancestral chromosomes into a single chromosome 2 in humans has been well supported by genome sequence data. Our fusion chromosome contains remnants of telomeres at the fusion point and it has another centromere-like region at just the right position.

Intelligent design proponents have a hard time explaining this event. They don't propose an explanation based on their concept of intelligent design—that would be too ridiculous—instead they concentrate on raising questions about evolutionary explanations. One of the common objections is that the new fusion chromosome would screw up mitosis and meiosis because it would initially have two centromeres. According to them, the chromosomal rearrangement would be detrimental and could never be fixed by natural selection.

As it turns out, the latter part of this statement is correct. Natural selection is not responsible for this kind of evolution. But no serious scientist would suggest otherwise.

The first part of the statement isn't as serious as the IDiots would like to think. Rearrangements of this sort aren't much of a problem. Many species are heterozygous for such rearrangements and we can see that it has little effect on the viability of dividing cells. Nevertheless, a newly rearranged chromosome is unlikely to be completely neutral. It's probably slightly deleterious with respect to the ancestral chromosome(s).

So, how does a slightly deleterious mutation become fixed in a population? The answer, of course, is random genetic drift. But in order to understand the importance of random genetic drift you have to understand the substructure of species. Species are usually subdivided into many smaller, locally inbreeding, populations or "demes." Slightly deleterious (nearly neutral) mutations can easily become fixed in a deme by accident.

Over at Panda's Thumb, Dave Wisker Arthur Hunt has been writing about chromosomal rearrangements and how they become fixed in a species. His latest essay talks about how fixation within a deme can lead to fixation by random genetic drift in the entire population: The Rise of Human Chromosome 2: Beyond the Deme.

This is a nice, short, explanation of a very important mechanism of evolution. I urge everyone to get on over to Panda's Thumb and read it right now.


[Image Credit: This drawing is from: Chromosome Fusion: Chance or Design?. I don't know the original source.

7 comments :

  1. Dave Wisker wrote the articles, not Art Hunt.

    ReplyDelete
  2. Cool! This is the kind of example of drift I was looking for!

    I haven't visited the Panda's Thumb in so long...

    ReplyDelete
  3. Larry: "His latest essay talks about how fixation within a deme can lead to fixation by random genetic drift in the entire population"

    As near as I can figure out, a deme seems to be beneficial when the deleterious effect is in heterozygous individuals ("heterozygote disadvantage"). That is, (if I read it correctly) it's not that demes make fixation of a slightly deleterious allele more likely in all cases (e.g., an enzyme that is slightly less good than the wild type).

    In the case of a recently fused chromosome, individuals homozygous for the fused condition are just as fit as those who are homozygous for the separate condition, when each is in a deme of peers.

    ReplyDelete
  4. Anything that makes the effective population size smaller accelerates the rate (but not the probability) of allele fixation. Demes just seem to be a way of discussing this acceleration in subpopulations.

    ReplyDelete
  5. Genome doubling in plants is the only rearrangement that directly causes speciation, all the others are products of speciation, usually caused by periods of allopatry.

    Calling chromosome rearrangement purely random is not the whole story. There seem to be hotspots for breakage. Both gorillas and chimps have an inversion on chromsome 9 containing the same genes but their common ancestor does not. Seems these are two different events, the breakpoints are close yet not exact. This is not to say that selection has anything to do with it although there is evidence of consistent patterns of breakages along altitude and climate clines.

    Human chromosome 2 is very cool as you can see both the tandem fusion and a very small centromere region, you can find it easily with RepeatMaker and then view it in any genome browser. The dead centromere contains three SVA primate retrotransposons of which at least two are human specific. As retrotransposons can be very disruptive it would be interesting to speculate that these might have something to do with it. The only other SVA elements found so far in human centromeres is one in chromosome Y. Is there a way bioinformatically to date these rearrangements.

    There was some confusion about tandem fusions as they are as far as I know, unknown in insects. Robertsonian fusions, the model for most rearrangements, well these were first discovered in grasshoppers and don't include tandem whole chromosome fusion. Mice have Robersonian fusions of acentric chromosomes. Seems like other mammals also have tandem fusions, horse and ass differ by an fusion and then a percentric inversion of the remaining centromere. Mujac deer, a very specious mammal, has great diiference in karyotype and evidence points to multiple tandem fusions.

    ReplyDelete
  6. I'd like to point out that Kimura's Neutral Theory is about base change only. Rearrangements, repeats, insertions and deletions at all levels is a separate random process that has no general theory. Ohta's brilliant statement that neutral theory provides a null test for selection allows for its measurement. As there is no null hypothesis yet for rearrangements there is no way to tease out any signal of selection. Studies of patterns so far suggest two or more linked random processes of which one has a major effect on all the others, sort of like a gene of large effect.

    ReplyDelete
  7. Thak you, Larry, for the plug!

    ReplyDelete