Horizontal Gene Transfer and Intelligent Design Theory

AUTHOR: Graham Lawton

SOURCE: "Why Darwin was Wrong About the Tree of Life"

COMMENTARY: Allen MacNeill

A recent article in the New Scientist trumpeted the news that "Darwin was wrong", at least insofar as his "tree of life" was concerned. To be specific, the author stated that new discoveries in the field of horizontal gene transfer had invalidated Darwin's "tree of life", as illustrated by his diagram in the Origin of Species.

Horizontal gene transfer (especially as the result of viral transduction) has been known to occur for almost half a century. In my undergraduate genetics course at Cornell (which I took in the spring of 1972) we did a lab in which we used lambda bacteriophage to transfer genetic material from one bacterial colony to another. Ergo, none of the mechanisms of HGT described in the article in New Scientist are all that new.

Indeed, I have listed at least six mechanisms of HGT in my blogpost on the “engines of variation” located here. In that list, they are numbers 28, 29, 33, 36, 40, and 41. Most of these HGT mechanisms have been known for decades and are among the best understood mechanisms of increasing both genetic and phenotypic variation.

What is relatively new is the application of the information gained about HGT to phylogenetic reconstruction. HGT is the rule among bacteria, and apparently occurs fairly frequently among eukaryotes as well. Evolutionary biologists, and especially phylogeneticists and systematists have been using HGT data for phylogenetic reconstruction for over a decade, even among eukaryotes. So, once again this is not new.

However, there is currently a tread at Uncommon Descent to the effect that

(1) the New Scientist article is pointing out that "Darwinism" is a bankrupt theory, and

(2) that HGT is more easily explained as part of "intelligent design theory" (ID).

Does the increasing recognition of HGT and its use in phylogenetic reconstruction mean that the current theory of evolution is invalid, or that ID can explain these phenomena better? On the contrary, the more we learn about HGT the more it seems to be even more random and undirected than vertical gene transfer (i.e. genetic recombination and heredity via reproduction). To be specific, the overwhelming majority of identified HGTs are of non-coding DNA sequences that have no detectable effect on the phenotypes of the organisms in which it has occurred.

That is, almost all of the DNA sequences that have been unambiguously shown to be the result of HGT are sequences that do not code for proteins nor participate in the regulation of coding sequences. Rather, they are sequences that have “gone along for the ride”, especially as the result of RNA retroviral HGT. Such sequences are so common that they are routinely used to construct and modify genetic phylogenies, as well as to determine genetic homologies.

The vast majority of HGTs are essentially neutral genetic mutations, as first described by Motoo Kimura in his neutral theory of molecular evolution. As such, they produce an immense amount of genetic variation without producing a corresponding amount of phenotypic variation. Furthermore, when such phenotypic variation does occur, it is more often deleterious than beneficial (usually mildly deleterious, as pointed out by Tomoko Ohta in her nearly neutral theory of molecular evolution). Only very rarely are such HGTs beneficial, and then only in relatively restricted ecological and evolutionary settings.

But neutral or slightly deleterious genetic changes (such as those produced by the vast majority of HGTs) are exactly the opposite of what one would expect to see as the work of an “intelligent designer”. Such an entity would (as several of the commentators in this thread have suggested) tailor HGTs to produce adaptive (i.e. beneficial) changes in the phenotypes of the recipients of its HGTs. Either that, or the “intelligent designer” doesn’t “tailor” its HGTs at all, but rather produces them randomly, rather like a dealer in a card game. But in that case, the actions of a soi dissant “intelligent designer” would be indistinguishable from Darwinian evolution, and including any reference to its actions (and/or inferring its existence) would be unnecessary (and would therefore violate Occam’s razor).

One last point: although the vast majority of HGTs produce either no phenotypic effect or slightly deleterious phenotypic effects, a relatively small number produce phenotypic effects that are correlated with increased survival and/or reproductive success. Unlike the vast majority of HGTs, these beneficial HGTs rapidly proliferate in the populations in which they arise, in exactly the way Darwin proposed in 1859. That is, they are preserved and passed on (while deleterious HGTs are eliminated), and thereby become more common over time among the populations in which they occur.

As always, comments, criticisms, and suggestions are warmly welcomed!

--Allen