Comment on “Does constructive neutral evolution play an important role in the origin of cellular complexity?” DOI 10.1002/bies.201100010

Abstract

Speijer [1] has provided a critique of constructive neutral evolution (CNE) and its role in the origin and evolution of cellular complexity [2, 3]. Not surprisingly, we disagree with his assertions. Because his description of the CNE model does not precisely conform to our view of CNE, as we [2, 4] and Stoltzfus [3] have elaborated it, we briefly re-state the model before addressing Speijer’s objections. The underlying premise of CNE is a pre-existing, essentially neutral interaction (RNA:RNA, RNA:protein, protein:protein) between component A, which has some activity, and component B. The activity of A is not dependent on the interaction with B, nor is A’s activity negatively influenced by this interaction. Thus, B could disappear from the scene without any effect on the ‘‘fitness’’ of A. We imagine that a mutation occurs in A that compromises its activity and that normally this mutation would be eliminated from the population by purifying selection. However, the pre-existing interaction with B fortuitously suppresses the effect of this mutation, so that selection pressure is relaxed and themutationmay be harmlessly fixed by drift. Thus, A becomes dependent on B for its activity by virtue of the neutral, ‘‘pre-suppressive’’ effect of the A:B interaction. What we submit does not happen is that the mutation occurs first, after which the interaction with B is positively selected for because it suppresses the deleterious effect of the mutation. Speijer’s ‘‘Think again’’ article [1] embraces several misunderstandings about this important process. First, we note that Speijer’s critique is considerably longer than was our Perspective in Science [3], giving him space to deconstruct several points that hemay consider components or at least entailments of our hypothesis, but that we would not. Let us call these misunderstandings of Type A. Type B misunderstandings reflect Speijer’s conflation of microand macroevolution, or ‘‘levels of selection’’. He fails to recognize that some features that are neutral or even disadvantageous to individuals – and thus not expected to be fixed by selection operating within populations of species – can nevertheless be sufficiently advantageous to species (fostering enhanced speciation or reduced extinction rates) to spread by species or clade selection. Eukaryotic sex might be one of these. Introns could be another. No one really thinks that the insertion of introns was selected for at the level of individuals – that is, that all introns that are currently fixed within a species were fixed because individuals that bore them were at a selective advantage compared to conspecifics that lacked them. Indeed, for individuals within species, intron addition could be slightly deleterious. It might nevertheless still be true that species in which many introns have become fixed do better than species with few introns (speciate more frequently or become extinct less often), because introns facilitate exon shuffling or the elaboration of multiple gene products through alternative splicing. There would thus come to be more intronbearing species (and in consequence more introns) in the world, thanks to species selection. We see many CNEestablished features as evolving like this, influencing however subtly the future evolutionary potential – the ‘‘evolvability’’ – of species, and yet their initial establishment was the consequence of the neutral ratchet we describe. Type C misunderstandings are of a diverse sort. It seems simplest to attempt to correct these as we read DOI 10.1002/bies.201100039