Darwin in the Genome

In doing a search on another topic, I came across the thread on this forum regarding molecular evolution. I have just written a book on this topic, discussing the effect of natural selection on the mechanisms that generate genome variation [which is affected by a whole range og biochemical activities, from polymerases to proofreading to mismatch and other repair, to the sizes of nucleotide pools, to recombination].
I have come to the conclusion that through natural selection, many mutations move away from being completely random.
It is important to note that it was not Darwin who suggested that variation was generated by completely random changes in DNA, and, in fact, the concept that the mechanisms that generate genome variation fall under natural selection not only is consistent with the Darwinian framework, but also gives us a deeper sense of the potential power of natural selection.
For those interested in more detail, there is a links page at http://www.DarwinGenome.info that includes other writing on this subject, including the conference volume "Molecular Strategies in Biological Evolution". There also will be a review in the 2003 volume of Annual Reviews of Microbiology.

I am sorry to have been so cryptic about how I came to this conclusion-- to discuss it fully took a whole book.
Here are three examples of nonrandom mutation:
1)generation of diversity in the immune response.
while this occurs within a generation, it suggests the availabilty of biochemical mechanisms that can focus mutation: recombination is focused on moving a selection of variable regions next to the constant region. Secondly, the variable region "hypermutates" in a way that is sequence dependent, i.e. sequences such as RGYW [shorthand for A or G, G, T or C, A or T] are hotspots of mutation through a focused biochemical mechanism that is just being worked out.
2) repeat sequences in bacteria: for example, tetranucleotide repeats tend to grow and shrink at a comparatively high rate as the 2 strands of DNA misalign. These sequences have become enriched in so-called "contingency genes" which tend to be involved in rapid adaptation to a host.
3) a hotspot of recombination in the mammalian germ line appears to be in the histocompatibility region; while high variation in this region is a problem for transplant surgeons, it would be of selective value in evolution for protection against pathogens.
"Molecular Strategies in Biological Evolution" is volume 870 of the Annals of the New York Academy of Sciences. My introductory chapter to the volume is available at http://www.amazon.com/...3311936/excerpt/103-6490997-6923024
It is important to emphasize again that non-random mutation is fully consistent with the Darwinian framework of variation followed by selection-- in this case variation of the mechanisms that generate variation.

Just to clarify an important point: Darwin and Wallace proposed that evolution takes place by selection acting on variation. They did not know the mechanisms that generate variation. Therefore, that some mutations may become more likely than others would not at all violate Darwin and Wallace's framework, but rather illustrate that variations in the mechanisms that generate variation can fall under natural selection, much as variations in wings and beaks can.

PaulK:
You have summarized the issues very well.
I agree with the concept that a biochemical innovation in the mechanisms that generate genome variation could lead to a great expansion, such as the Cambrian---the following example, on a less comprehensive scale than the Cambrian expansion, is discussed in the book:Cone snails appear to have evolved an efficient mechanism for exploring new toxin sequences [which they use eg to kill prey], facilitating the access to a wide range of food sources for diverse species of snail.

Peter-
When you say it will be confirmed at the molecular level soon: what kind of data at the molecular level would confirm your theory, and what kind of data would make it less likely to be true?
Lynn

Hi Peter-
Since you referred to our email exchange, I thought I'd include here my response to your question about implications for phylogenetic analysis to share with the others:
You are absolutely right that there are implications for phylogenetic analysis-- Lynn Ripley emphasized this to me when she was preparing her talk for the NYAS conference.
There has been some necessary work to incorporate recombination A Bayesian model for detecting past recombination events in DNA multiple alignments - PubMed
and variable mutation rates
The effects of variable mutation rates across sites on the phylogenetic estimation of effective population size or mutation rate of DNA sequences - PubMed
into phylogenetic analysis
Lynn

Percy-
Thank you for serving as a facilitator to stimulate discussion of the material in the book.One set of examples of non-random mutation, given in Chapter 6, are regions of the genome that encode repeat sequences such as GGGGG or CAATCAATCAAT.These repeats tend to mutate by changing in length at a rate orders of magnitude higher than the genome's overall mutation rate, and can throw the reading [which depends on a triplet code] out of frame.Because they can inactivate genes, such repeats would be expected to be selected against in much of the genome, and yet they accumulate in certain places, such as in genes in certain bacteria that are involved in host interaction.By facilitating surface variation, such repeats have the selective advantage of varying the bacteria's range of options, in terms of which tissues they can adhere to, and also by enabling the bacteria to hide from the immune response.Thus, selection can favor the emergence of non-random mutation [ie mutation in repeats become more probable than mutation elsewhere] and the alignment of non-random mutation with a biological role that provides a selective advantage [ie relative accumulation of "slippery" repeats in host-interaction genes].Lynn Caporale[This message has been edited by caporale, 01-14-2003]

Brad-
If you would like to examine the data for yourself, the book has extensive references to the original articles. Many of these are available on line, the rest in libraries.
Lynn