My last post on Perry Marshall

Once more unto the breach in Perry Marshall's cranium, dear friends. He is once again trying to claim that he alone has the one true understanding of Barbara McClintock's work, and he keeps getting it wrong. It's just embarrassing to watch.

He makes obvious statements like this:

Damage is random. Repair is not.

Well, duh. If the cell were to just go charging in and practice excision repair (a process that snips out a short piece of one strand of DNA and brings in polymerase to re-synthesize it) on random stretches of DNA, it would increase the frequency of errors. Polymerase proofreads as it goes; it checks to see if the nucleotide it just copied into a new strand properly complements the nucleotide on the other strand, and if it doesn't, it steps back, cuts out the error, and tries again.

This is familiar stuff. Students in our classes here at UMM get all this kind of material, in far greater detail, by their second year here. The problem is that Marshall carries it too far: he assumes that the cell "knows" the nature of the specific error made, and intelligently acts to directly repair it. It doesn't. The cell can invoke general mechanisms to attempt repair, but it doesn't in any way "know" what to do.

She is speaking from background knowledge that cells restructure their DNA in specific, programmed, predictable ways to common threats like heat and DNA damage.

He's strongly overstating the case. Cells can sense the presence of damaged DNA -- there's an extensive literature on DNA damage sensing and repair -- and activate enzymes that try to patch up the problem. There are common immediate responses, like arresting the cell cycle, or apoptosis, but there are also proteins that attempt to remove the blockage. But they are not sentient. They are chemicals.

For example, broken chromosomes leave ragged, dangling ends that are not supposed to be there; these activate proteins that attempt to bind and reanneal single stranded DNA together, to repair the break by splicing two broken ends together. They are not perfect. They can make mistakes. McClintock was well aware of this, and that's why what she studied was called genetic instability; the cell could, and often did, make quick patches that allowed the cell to continue on through division and make the same mistake over and over again. She said this herself in the Nobel speech that seems to be Marshall's sole source of information about her work.

A conclusion of basic significance could be drawn from these observations: broken ends of chromosomes will fuse, 2-by-2, and any broken end with any other broken end. This principle has been amply proved in a series of experiments conducted over the years. In all such instances the break must sever both strands of the DNA double helix. This is a “double-strand break” in modern terminology.

My emphasis. The DNA is broken! Quick! Activate enzymes to stitch it back together again! But all it can do is splice any broken end to another broken end, so you might get inversions, or translocations, or ring chromosomes. If DNA repair was as intelligent as Marshall claims it is, McClintock would have had nothing to do, because what she was studying was repeated anomalous instabilities of the chromosomes.

Let that sink in. All the phenomena McClintock studied were errors in repair; if repair mechanisms were so clever, she'd have been studying why cells never ever make mistakes.

That is a fundamental failure of comprehension on Marshall's part. He keeps reading this intelligence into the cell that just isn't there, and his evidence is always stuff we know about that was found because it has limits to what it can do.

He does this repeatedly with his other obsession, transposition. He really doesn't get it.

Notice this is also the wider context for transposition, the discovery for which she won the Nobel Prize – that cells re-arrange DNA in direct response to threats, stress, and shock. Transpositions are by their very description non-random patterns of DNA editing.

No. Transposons jump to mostly random locations in the genome. They do have a little bit of specificity: they recognize a short, sloppy pattern that is A-T rich and is fairly common. Its like having a long sequence of random digits, and tool that randomly cuts the sequence whereever a "0" is next to a "9". It's still random even if there are constraints on where it can cut.

Transposons are not specific, direct mechanisms of generating predetermined responses to environmental signals. They are sequences that generate random damage. Here's how one of my genetics texts (Genetic Analysis: An Integrated Approach
by Sanders and Bowman) describes them.

Transposable elements typically create mutations by their insertion into wild-type alleles. The insertion of new DNA into a functional gene is the equivalent of instering a random string of letters into a sentence. And just as the insertion of a random string of letters renders the senetene unintelligible, so too the consequece of DNA transpoition is to render the wild-type allele nonfunctional by making it unable to produce a wild-type gene product. This mutational process is known as insertional inactivation.

Our cell biology text, Essential Cell Biology
by Alberts et al., says something similar.

The insertion of a mobile genetic element into the coding sequence of a gene or into its regulatory region can cause the spontanious mutations that are obseved in many of today's organisms. Mobilie genetic elements can severely disrupt a gene's activity if they land directly within its coding squence. Such an insertion mutation destroys the gene's capacity to encode a useful protein -- as is the case for a number of mutations that cause hemophilia in humans, for example.

The activity of mobile genetic elements can also change the way existing genes are regulated. An insertion of an element into a regulatory DNA region, for often have a striking effect on where and when genes are expressed.

So I have to wonder who understand McClintock's work: the cell and molecular biologists who are doing research in the field, writing textbooks, and actually understand transposons as sources of random damage, or the self-taught electrical engineer who can't even get the basics right and wants to claim that transposons are cunning intelligent repair mechanisms at work in the cell?

There's not much point to arguing further with Marshall. What he desperately needs is to go back to school -- take the basic biology sequence at his local community college for instance. Nothing I've written goes beyond what you'll get in the first year or two of any college biology program -- it's all straightforward cell biology.

Unfortunately, he'd probably be an impossible student. After all, in the absence of any education on the subject at all, he's written a whole "science" book that overturns, he thinks, everything his professors would know about biology. He might be reasonably bright, but he's managed to build a giant wall of hubris and misinformation to stand between him and any real knowledge.

And that's just too sad.