Evolution Icon Evolution

“Irremediable Complexity”

An intriguing “hypothesis” paper entitled “How a neutral evolutionary ratchet can build cellular complexity”1, where the authors speculate about a possible solution to a possible problem, recently appeared in the journal IUBMB Life. It is an expanded version of a short essay called “Irremediable Complexity?”2 published last year in Science. The authors of the manuscripts include the prominent evolutionary biologist W. Ford Doolittle.
The gist of the paper is this. The authors think that over evolutionary time, neutral processes would tend to “complexify” the cell. They call that theoretical process “constructive neutral evolution” (CNE). In an amusing analogy they liken cells in this respect to human institutions:

Organisms, like human institutions, will become ever more ”bureaucratic,” in the sense of needlessly onerous and complex, if we see complexity as related to the number of necessarily interacting parts required to perform a function, as did Darwin. Once established, such complexity can be maintained by negative selection: the point of CNE is that complexity was not created by positive selection.2

In brief, the idea is that neutral interactions evolve serendipitously in the cell, spread in a population by drift, get folded into a system, and then can’t be removed because their tentacles are too interconnected. It would be kind of like trying to circumvent the associate director of licensing delays in the Department of Motor Vehicles — can’t be done.
The possible problem the authors are trying to address is that they think many systems in the cell are needlessly complex. For example, the spliceosome, which “splices” some RNAs (cuts a piece out of the middle of a longer RNA and stitches the remaining pieces together), is a huge conglomerate containing “five small RNAs (snRNAs) and >300 proteins, which must be assembled de novo and then disassembled at each of the many introns interrupting the typical nascent mRNA.”1 What’s more, some RNAs don’t need the spliceosome — they can splice themselves, without any assistance from proteins. So why use such an ungainly assemblage if a simpler system would do?

The authors think the evolution of such a complex is well beyond the powers of positive natural selection: “Even Darwin might be reluctant to advance a claim that eukaryotic spliceosomal introns remove themselves more efficiently or accurately from mRNAs than did their self-splicing group II antecedents, or that they achieved this by ‘numerous, successive, slight modifications’ each driven by selection to this end.”1

Well, I can certainly agree with them about the unlikelihood of Darwinian processes putting together something as complex as the spliceosome. However, leaving aside the few RNAs involved in the splicesome, I think their hypothesis of CNE as the cause for the interaction of hundreds of proteins — or even a handful — is quite implausible. (An essay skeptical of large claims for CNE, written from a Darwinian-selectionist viewpoint, has appeared recently3 along with a response from the authors4).

The authors’ rationale for how a protein drifts into becoming part of a larger complex is illustrated by Figure 1 of their recent paper (similar to the single figure in their Science essay). A hypothetical “Protein A” is imagined to be working just fine on its own, when hypothetical “Protein B” serendipitously mutates to bind to it. This interaction, postulate the authors, is neutral, neither helping nor harming the ability of Protein A to do its job. Over the generations Protein A eventually suffers a mutation which would have decreased or eliminated its activity. However, because of the fact that Protein B is bound to it, the mutation does not harm the activity of Protein A. This is still envisioned to be a neutral interaction by the authors, and organisms containing the Protein A-Protein B complex drift to fixation in the population. Then other mutations come along, co-adapting the structures of Protein A and Protein B to each other. At this point the AB complex is necessary for the activity of Protein A. Repeat this process several hundred more times with other proteins, and you’ve built up a protein aggregate with complexity of the order of the spliceosome.

Is this a reasonable hypothesis? I don’t mean to be unkind, but I think that the idea seems reasonable only to the extent that it is vague and undeveloped; when examined critically it quickly loses plausibility. The first thing to note about the paper is that it contains absolutely no calculations to support the feasibility of the model. This is inexcusable. The mutation rates of various organisms — viral, prokaryotic, eukaryotic — are known to sufficient accuracy5 that estimates of how frequently the envisioned mutations arrive could have been provided. The neutral theory of evolution is also well-developed6, which would allow the authors to calculate how long it would take for the postulated neutral mutations to spread in a population. Yet no numbers — not even back-of-the-envelope calculations — are provided. Previous results by other workers7-9 have shown that the development of serendipitous specific binding sites between proteins would be expected to be quite rare, and to involve multiple mutations. Kimura6 showed that fixation of a mutation by neutral drift would be expected to take a looong time. Neither of these previous results bodes well for the authors’ hypothesis.

The second thing to notice about the paper is that there is no experimental support for its hypothesis. As the authors point out:

Development of in vitro experimental systems with which to test CNE will be an important step forward in distinguishing complex biology that arose due to adaptation versus nonadaptive complexity, as part of a larger view to understand the interplay between neutral and adaptive evolution, such as the intriguing long-term evolution experiments of Lenski and coworkers.1

Yet no such experimental evolutionary results have been reported to my knowledge, either by Lenski or by other workers.10
Besides the lack of support from calculations or experiments, the authors discuss no possible obstacles to the scheme. I certainly understand that workers want to accentuate the positive when putting a new model forward, but potential pitfalls should be pointed out, so that other researchers have a clearer idea of the promise of the model before they invest time in researching it.

The first possible pitfall comes at the first step of the model, where a second protein is postulated to bind in a neutral fashion to a working protein. How likely is that step to be neutral? At the very least, we now have two proteins, A and B, that now have a large part of their surfaces obstructed that weren’t before. Will this interfere with their activities? It seems there is a good chance. Second, simply by Le Chatelier’s principle the binding of the two proteins must affect the free energies of their folded states. What’s more, the flexibility of both proteins must be affected. Will these individual effects serendipitously cancel out so that the overall effect will be neutral? It seems like an awful lot to ask for without evidence.
In the next step of the model Protein A is supposed to suffer a mutation that would have caused it to lose activity, but, luckily, when it is bound to Protein B it is stabilized enough so that activity is retained. What fraction of possible mutations to Protein A would fall in that range? It seems like a very specialized subfraction. Looking at the flip side, what fraction of mutations to Protein A and/or Protein B which otherwise would not have caused A to lose activity will now do so because of its binding to Protein B?

The last step of the model is the “co-adaptation” of the two proteins, where other, complementary mutations occur in both proteins. Yet this implies that the protein complex must suffer deleterious mutations at least every other step, provoking the “co-adaptive” mutation to fix in the population. Wouldn’t these deleterious mutations be very unlikely to spread in the population?

Finally, multiply these problems all by a hundred to get a spliceosome. Or, rather, raise these problems to the hundredth power. But, then, why stop at a hundred? As the authors note approvingly:

Indeed, because CNE is a ratchet-like process that does not require positive selection, it will inevitably occur in self-replicating, error prone systems exhibiting sufficient diversity, unless some factor prevents it.1

Why shouldn’t the process continue, folding in more and more proteins, until the cell congeals? I suppose the authors would reply, “some factor prevents it.” But might not that factor kick in at the first or second step? The authors give us no reason to think it wouldn’t.

The CNE model (at least on the scale envisioned by the authors) faces other problems as well (for example, it would be a whole lot easier to develop binding sites for metal ions or metabolites that are present in the cell at much higher concentrations than most proteins), but I think this is enough to show it may not be as promising as the article would have one believe.

Besides the model itself, it is interesting to look at a professed aspect of the motivation of the authors in proposing it. It may not have escaped your notice, dear reader, that “irremediable complexity” sort of sounds like “irreducible complexity.” In fact, the authors put the model forward as their contribution to the good fight against “antievolutionists”:

… continued failure to consider CNE alternatives impoverishes evolutionary discourse and, by oversimplification, actually makes us more vulnerable to critiques by antievolutionists, who like to see such complexity as ”irreducible.”1

So there you have it. The authors don’t think Darwin can explain such complexity as is found in the proteasome, and they apparently rule out intelligent design. (By the way, when will these folks ever grasp the fact that intelligent design is not “antievolution”?) “Irremediable complexity” seems to be all that’s left, no matter how unsupported and problematic it may be.

Although the authors seem not to notice, their entire model is built on a classic argument from ignorance, beginning with the definition of irremediable complexity:

”irremediable complexity”: the seemingly gratuitous, indeed bewildering, complexity that typifies many cellular subsystems and molecular machines, particularly in eukaryotes.1

“Seemingly gratuitous.” In other words, the authors don’t know of a function for the complexity of some eukaryotic subsystems; therefore, they don’t have functions. Well the history of arguments asserting that something or other in biology is functionless is pretty grim. More, the history of assertions that even “simple” things (like, say, DNA, pre-1930) in the cell either don’t have a function or are just supporting structures is abysmal. Overwhelmingly, progress in biology has consisted of finding new and ever-more-sophisticated properties of systems that had been thought simple. If apparently simple systems are much more complex than they initially seemed, I would bet heavily against the hypothesis that apparently complex systems are much simpler than they appear.


  1. 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.
  2. Gray, M. W., J. Lukes, J. M. Archibald, P. J. Keeling, and W. F. Doolittle, 2010 Cell biology. Irremediable complexity? Science 330: 920-921.
  3. Speijer, D., 2011 Does constructive neutral evolution play an important role in the origin of cellular complexity? Making sense of the origins and uses of biological complexity. Bioessays 33: 344-349.
  4. Doolittle, W. F., J. Lukes, J. M. Archibald, P. J. Keeling, and M. W. Gray, 2011 Comment on “Does constructive neutral evolution play an important role in the origin of cellular complexity?” Bioessays 33: 427-429.
  5. Drake, J. W., B. Charlesworth, D. Charlesworth, and J. F. Crow, 1998 Rates of spontaneous mutation. Genetics 148: 1667-1686.
  6. Kimura M., 1983 The neutral theory of molecular evolution. Cambridge University Press, Cambridge.
  7. Nissim, A., H. R. Hoogenboom, I. M. Tomlinson, G. Flynn, C. Midgley, D. Lane, and G. Winter, 1994 Antibody fragments from a ‘single pot’ phage display library as immunochemical reagents. EMBO Journal 13: 692-698.
  8. Griffiths, A. D., S. C. Williams, O. Hartley, I. M. Tomlinson, P. Waterhouse, W. L. Crosby, R. E. Kontermann, P. T. Jones, N. M. Low, T. J. Allison, and G. Winter, 1994 Isolation of high affinity human antibodies directly from large synthetic repertoires. EMBO Journal 13: 3245-3260.
  9. Smith, G. P., S. U. Patel, J. D. Windass, J. M. Thornton, G. Winter, and A. D. Griffiths, 1998 Small binding proteins selected from a combinatorial repertoire of knottins displayed on phage. Journal of Molecular Biology 277: 317-332.
  10. Behe, M. J., 2010 Experimental Evolution, Loss-of-function Mutations, and “The First Rule of Adaptive Evolution”. Quarterly Review of Biology 85: 1-27.

Michael J. Behe

Senior Fellow, Center for Science and Culture
Michael J. Behe is Professor of Biological Sciences at Lehigh University in Pennsylvania and a Senior Fellow at Discovery Institute’s Center for Science and Culture. He received his Ph.D. in Biochemistry from the University of Pennsylvania in 1978. Behe's current research involves delineation of design and natural selection in protein structures. In his career he has authored over 40 technical papers and three books, Darwin Devolves: The New Science About DNA that Challenges Evolution, Darwin’s Black Box: The Biochemical Challenge to Evolution, and The Edge of Evolution: The Search for the Limits of Darwinism, which argue that living system at the molecular level are best explained as being the result of deliberate intelligent design.