A Blind Man Carrying a Legless Man Can Safely Cross the Street: Experimentally Confirming the Limits to Darwinian Evolution
I never thought it would happen but, in my estimation, Richard Lenski has acquired a challenger for the title of "Best Experimental Evolutionary Scientist." Lenski, of course, is the well-known fellow who has been growing E. coli in his lab at Michigan State for 50,000 generations in order to follow its evolutionary progress. His rival is Joseph Thornton of the University of Oregon who, by inferring the sequences of ancient proteins and then constructing (he calls it "resurrecting") their genes in his lab, is able to characterize the properties of the ancestral proteins and discern how they may have evolved into more modern versions with different properties.
I have written appreciatively about both Lenski and Thornton before, whose work indicates clear limits to Darwinian evolution (although they themselves operate within a Darwinian framework). Thornton's latest work is beginning to show a convergence with Lenski's that greatly boosts our confidence that they both are on the right track. In a recent review (Behe, 2010) I pointed out that all characterized advantageous mutations that Richard Lenski has observed in his twenty-year experiment have turned out to be degradative ones -- in which a gene or genetic control structure was either destroyed or rendered less effective. Random mutation is superb at degrading genetic material, which sometimes is helpful to an organism. In his latest work Thornton, too, shows evolution of a system by degradation, although he speculates that the changes were neutral rather than advantageous.
In Finnegan et al. (2012), "Evolution of increased complexity in a molecular machine," Thornton and colleagues study a ring of six proteins in a molecular machine (that also has many other parts) called a V-ATPase, which can pump protons (acid) across a membrane. The machine exists in all eukaryotes. In most eukaryotic species, however, the hexameric ring consists of five copies of one protein (let's call it protein 1) and one copy of another, related protein (call it protein 2). In fungi, however, the ring consists of four copies of protein 1, one copy of protein 2, and one copy of protein 3. Protein 3 is very similar in sequence to protein 1, so Finnegan et al. (2012) propose that proteins 1 & 3 are related by duplication of an ancestral gene and subsequent modification of the two, originally identical duplicated genes.
How did protein 3 insinuate itself into the ring? The original protein 1, present in five copies in most organisms, already had the ability to bind to itself, plus an ability to bind to one side of protein 2, plus a separate ability to bind to the opposite side of protein 2 (see Finnegan et al.'s Figure 3). Thornton's results are consistent with the idea that, by happenstance, the gene for protein 1 duplicated and spread in the population. These events apparently were neutral, the authors think, not affecting the organism's fitness.
Eventually one of the duplicates acquired a degradative mutation, losing the ability to bind one side of protein 2. This was not a problem because the second copy of the protein 1 gene was intact, and could bind both sides of protein 2, so a complete ring could still be formed. This also spread by neutral processes. As luck might have it, the second gene copy subsequently acquired its own degradative mutation, so that it could no longer bind the other side of protein 2. Again it's no problem, however, because the first mutant copy of protein 1 could bind to the first side of protein 2, bind a few more copies of itself, then bind a copy of protein 3, which still had the ability to bind the other side of protein 2. So a closed, six-member ring could still be formed. This apparently also spread by neutral processes until it took over the entire kingdom of fungi.
The work of Finnegan et al. (2012) strikes me as quite thorough and elegant. I have no reason to doubt that events could have unfolded that way. However, the implications of the work for unguided evolution appear very different to me than they've been spun in media reports. The most glaringly obvious point is that, like the results of Lenski's work, this is evolution by degradation. All of the functional parts of the system were already in place before random mutation began to degrade them. Thus it is of no help to Darwinists, who require a mechanism that will construct new, functional systems. What's more, unlike Lenski's results, the mutated system of Thornton and colleagues is not even advantageous; it is neutral, according to the authors. Perhaps sensing the disappointment for Darwinism in the results, the title of the paper and news reports emphasize that the "complexity" of the system has increased. But increased complexity by itself is no help to life -- rather, life requires functional complexity. One can say, if one wishes, that a congenitally blind man teaming up with a congenitally legless man to safely move around the environment is an increase in "complexity" over a sighted, ambulatory person. But it certainly is no improvement, nor does it give the slightest clue how vision and locomotion arose.
Finnegan et al.'s (2012) work intersects with several other concepts. First, their work is a perfect example of Michael Lynch's idea of "subfunctionalization," where a gene with several functions duplicates, and each duplicate loses a separate function of the original. (Force et al., 1999) Again, however, the question of how the multiple functions arose in the first place is begged. Second, it intersects somewhat with the recent paper by Austin Hughes (2011) in which he proposes a non-selective mechanism of evolution abbreviated "PRM" (plasticity-relaxation-mutation), where a "plastic" organism able to survive in many environments settles down in one and loses by degradative mutation and drift the primordial plasticity. But again, where did those primordial functions come from? It seems like some notable workers are converging on the idea that the information for life was all present at the beginning, and life diversifies by losing pieces of that information. That concept is quite compatible with intelligent design. Not so much with Darwinism.
Finally, Thornton and colleagues latest work points to strong limits on the sort of neutral evolution that their own work envisions. The steps needed for the scenario proposed by Finnegan et al. (2012) are few and simple: 1) a gene duplication; 2) a point mutation; 3) a second point mutation. No event is deleterious. Each event spreads in the population by neutral drift. Notice that the two point mutations do not have to happen together. They are independent, and can happen in either order. Nonetheless, this scenario is apparently exceedingly rare. It seems to have happened a total of one (that is, 1) time in the billion years since the divergence of fungi from other eukaryotes. It happened only once in the fungi, and a total of zero times in the other eukaryotic branches of life. If the scenario were in fact as easy to achieve in nature as it is to describe in writing, we should expect it to have happened many times independently in fungi and also to have happened in all other branches of eukaryotes.
It didn't. Thus it seems a good conclusion that such neutral scenarios are much rarer than some workers have proposed (Gray et al., 2010; Lukes et al., 2011), and that more complex neutral scenarios are unlikely to happen in the history of life.
Behe, M. J., 2010 Experimental Evolution, Loss-of-function Mutations, and "The First Rule of Adaptive Evolution." Quarterly Review of Biology 85: 1-27.
Finnigan, G. C., V. Hanson-Smith, T. H. Stevens, and J. W. Thornton, 2012 Evolution of increased complexity in a molecular machine. Nature doi: 10.1038/nature10724.
Force, A., M. Lynch, F. B. Pickett, A. Amores, Y. L. Yan et al. 1999 Preservation of duplicate genes by complementary, degenerative mutations. Genetics 151: 1531-1545.
Gray, M. W., J. Lukes, J. M. Archibald, P. J. Keeling, and W. F. Doolittle, 2010 Irremediable complexity? Science 330: 920-921.
Hughes, A. L., 2011 Evolution of adaptive phenotypic traits without positive Darwinian selection. Heredity (Edinb.) doi: 10.1038/hdy.2011.97.
Lukes, J., J. M. Archibald, P. J. Keeling, W. F. Doolittle, and M. W. Gray, 2011 How a neutral evolutionary ratchet can build cellular complexity. IUBMB Life 63: 528-537.