On Somatic Variation of the Genome

I'm slowly working my way¹ through my complementary copy of What We Believe but Cannot Prove. I'm almost done -- at page 214 out of 252 -- and I can say that it is very diverse. The essays range from very thoughtful and interesting to way too specific to a particular discipline . . . to off the mark. Thankfully essays in the last category are few and far between. But when someone gets it so very wrong (or not even wrong), and they're writing about something you know about . . . well, it warrants a blog entry. This is that blog entry (with a maximal number of ellipses).

The someone in this instance is Kevin Kelly. Kevin helped start Wired magazine and is currently a "senior maverick" for that publication. I'm serious. That's his title. I wonder if that's synonymous for going way out of one's area of expertise and predicting something that has already happened.

Enough with the foreplay and cheap jabs (hmm, heavy petting and boxing together . . . sounds like fun), here's a bit of what Kevin Kelly wrote:

I believe, but cannot prove, that the DNA in your body (and in the bodies of all living organisms) varies from part to part.

If Kelly had done any background research, he would have discovered that we have already 'proven' (as much as we prove anything in biology) that there is intra-organismal genomic variation. Now, Kelly acknowledges that he is familiar with telomeres decreasing in size in any given lineage of cells. But he's interested in more than just the shrinkage of telomeres.

When I think of somatic mutations (ie, those that do not occur and the germ line and contribute to variation within an organism), I think of cancer. See this review (from 1998, so Kelly has no excuse) as an example. Many cancers are the result of somatic mutations that disrupt genes involved in pathways that regulate cell division. Cancer cells contain a different genome than non-cancer cells from the same individual. I'll stop here before I put my own foot in my mouth, but I'm sure Orac would be happy to tell you more.

If we have such striking evidence that there is intra-individual genomic variation, I'd say it's beyond belief . . . er, you can with some certainty that the DNA in your body varies from part to part. But that's not Kevin Kelly's entire essay. He decides to pull out some outdated models of animal development:

At the trivial end, if my belief is true, it would matter where in your body a sample of DNA is taken. And it would also matter when your DNA is sampled, as this variation could change over time. If my belief is true, this variation might have some effect on locating the correct seminal cells for growing replacement organs and tissues.

When I read this passage, I thought of the incorrect idea that is covered in most introductory genetics and development courses to contrast the correct model. The incorrect model -- which was considered a viable hypothesis for some time in the past -- posits that cells differentiate via changes in their genome. A more accurate picture is that the cells from different tissues have different expression patterns. In other words, the genomic content is (nearly) identical throughout your body², but different genes are turned on in different types of cells. It would be quite a revolution if the incorrect model were shown to be correct. But, considering that mountains of evidence for gene expression driving cellular differentiation, I highly doubt that such a revolution would occur.

And Kelly presents genetic mutations as if we know nothing about them:

We have a pretty good idea of how the selection in natural selection works: Less fit organisms die. But when it comes to how variation arises, all we can say is "random mutation," which is another way of saying, "We don't know exactly."

Wrong! Here is a really bad post I wrote on my old blog about the random nature of mutations. I plan to write a much better review (it's sitting in the pipeline), but that one will have to do for now. Anyway, we know a shitload about how mutations (substitutions, insertions, deletions, duplications, inversions, translocations, fusions, etc) occur at a molecular level. Kelly's model for natural selection ("Less fit organisms die") is just as unsatisfying as his explanation of mutation. I'd argue that we know less about how natural selection works than how genetic mutations occur. Natural selection is complex; it depends on the environment, the other genotypes in the population, and even mutations.

Those are the technical details. There is also one overarching flaw in Kevin Kelly's essay. It suffers from the same problem that multiple essays in the book suffer from: his belief is provable (or at least easily scientifically testable). Kelly's particular belief has been tested, but there are still more experiments can be done to study intra-individual genetic diversity.

Kelly presents the current understanding in a mottled, and often incorrect, manner. This is a potentially fascinating area of research. Studies of the genetic variation within an individual may reveal whether certain cell lineage leaves more derivatives based on their particular genomic sequence (sort of like intra-individual natural selection). But that essay should be written by someone other than Kevin Kelly. Maybe someone who has actually studied the molecular processes underlying mutations. Or possibly someone with a decent understanding of genetics.

1- The book sits in my bathroom, and I read it whenever I have finished everything else that happens to be near the porcelain throne. Too much information?

2- One fine point that I don't think Kelly was inferring because if he were he'd have said it explicitly: the genomes do differ at the epigenetic level.