The Case Against a Darwinian Origin of Protein Folds
Abstract
Four decades ago, several scientists suggested that the impossibility of any evolutionary process sampling anything but a miniscule fraction of the possible protein sequences posed a problem for the evolution of new proteins. This potential problem-the sampling problem-was largely ignored, in part because those who raised it had to rely on guesswork to fill some key gaps in their understanding of proteins. The huge advances since that time call for a careful reassessment of the issue they raised. Focusing specifically on the origin of new protein folds, I argue here that the sampling problem remains. The difficulty stems from the fact that new protein functions, when analyzed at the level of new beneficial phenotypes, typically require multiple new protein folds, which in turn require long stretches of new protein sequence. Two conceivable ways for this not to pose an insurmountable barrier to Darwinian searches exist. One is that protein function might generally be largely indifferent to protein sequence. The other is that relatively simple manipulations of existing genes, such as shuffling of genetic modules, might be able to produce the necessary new folds. I argue that these ideas now stand at odds both with known principles of protein structure and with direct experimental evidence. If this is correct, the sampling problem is here to stay, and we should be looking well outside the Darwinian framework for an adequate explanation of fold origins.
Downloads
Figures/Supplementary Files
Published
Issue
Section
License
Authors who publish in BIO-Complexity agree to the following terms:
Authors retain copyright and grant BIO-Complexity the right of first publication with the work simultaneously licensed under a Creative Commons Attribution License, which allows others to share the work, provided they acknowledgement the work's authorship and initial publication in BIO-Complexity.
Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of BIO-Complexity's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in BIO-Complexity.
Because you the author (including any co-authors) will claim copyrights for the final version of your work published in BIO-Complexity, you are responsible for ensuring that you are entitled to make this claim. In particular, you here certify that no one other than you or your co-author(s) holds the copyright to any figure or figure component included in this submission.
Comments on this article
These comments were posted between 2010 and 2011, when the journal accepted online comments. Commenting is now closed.
Director, The Gene Emergence Project, Department of ProtoBioCybernetics & ProtoBioSemiotics
Generic versus Non-Generic Interfaces
I have studied – as a non-specialist in the molecular „proteinology“ – your article „The case against…“ very carefully. However, I am not sure if I understand some passages of your very interesting paper adequately. Should the „non-generic“ nature of the interface between domains mean that the backbone of the polypeptide composed of, e.g, two domains is not formed exclusively by peptidic bonds? This would be – generally speaking - probably not possible at all? And why could not a module extend the existing structure by a peptidic bond at the „generic“ interface of say an alfa-helix?
Jiri Vacha
Re: Generic versus Non-Generic Interfaces
You've asked first whether my discussion of the non-generic nature of binding interfaces, particularly the interfaces between protein domains, implies that there should be no backbone connections between domains (if I've understood you correctly).
No, I didn't mean to imply that. As you can see in Figure 9 of the paper, there are in this case two backbone connections between the two domains. When we speak of the binding interface between protein domains we're typically referring to the non-covalent interactions (as with protein-protein binding). You're right that there are backbone connections in the case of multi-domain proteins, but these typically account for only a small fraction of the whole interface.
And in the experiment described with Figure 9, those covalent backbone connections are preserved. The problem is that the non-covalent interactions, being highly sequence dependent, are badly messed up when domains are interchanged in this way.
Your second question is, I think, answered in the paper: "the only possible generic additions at these interfaces are those that extend the regular structure, either by elongating the helix or by adding a strand to the edge of the sheet. Although examples of proteins accommodating these structural changes certainly exist [47, 48], the fact that both simply extend existing structure makes them unhelpful for explaining wholesale structural reorganization."
In other words, while these generic extension are possible in many cases, they are modest in scale, and they don't give you a new fold.
Doug
Modularity principle once again
I understand now that the chain of peptidic bonds as such in a polypeptide (as its backbone) is actually a trivial matter which says nothing specific about its function etc. Your experiment related to Fig. 9 demonstrates that the non-covalent components of the interface between two domains may be decisive for the function of the whole. Still, the structure and function of a polypeptide generally is a systemic effect of its primary structure (an „emergent“, „unpredictable“ phenomenon – see, e.g., just the cited work of Alexander et al.) and it seems to me to be quite possible that a combination of two originally independent (say) domains may show adaptively favorable features and be fixed during generations even when their non-covalent interface (determining the function) originated by chance. It may happen only rarely, but neo-Darwinism needs nothing more. Anyway, it is pertinently pointed out that the number of functioning proteins is much higher than the number of protein domains.
Jiří
Re: Modularity principle once again
Referring to adaptation by means of fortuitous association of previously unassociated protein domains, you say, "It may happen only rarely, but neo-Darwinism needs nothing more."
I certainly agree with the first part of this statement. Rare doesn't mean impossible, and adaptive changes do occur. As for the second part, I think it depends on our expectations for the theory. If we're content for neo-Darwinism to explain only some things, possibly minor things, then I agree.
But most proponents of the theory have much higher expectations. So the correct approach, I think, is to do the science to find out precisely what neo-Darwinism does explain and what it doesn't explain. Your final sentence points to the kind of question that needs to be answered: We know that some protein folds are put to a variety of functional uses, so to what extent does neo-Darwinism explain this functional diversity within fold families?
The paper under discussion here doesn't attempt to address this, but I and others at Biologic are involved in a number of projects that do address it.
Doug
Domain combinations vs. probabilistic resources
As I understand it now, a haphazard combination of two independent protein domains has only a very small chance to be functional ("adaptive"). It seems to me that the question whether the evolution of the realm of existing proteins could be explained in the selectionist´s manner could perhaps be solved in a quite analogical way as you are describing in connection with the primary structure of protein molecules. I.e., to compare the number of the theoretically possible variations of any two existing domains (may be, structural motifs) with the "probabilistic resources" which were at the disposal. Only when the possible variations outweighed massively the probabilistic resources, could the selectionist´s explanation be regarded falsified. As a non-expert in the field I do not dare to guess if such an endeavor is operational.
Jiří
functional possibilities vs. theoretical possibilities
It's not clear to me what you mean by "theoretically possible variations of any two existing domains," but I certainly agree that the possibilities are sparsely sampled whenever they greatly outnumber the events that sample them. As argued in the paper, though, we also have to examine the proportion of the possibilities that are functional before we can conclude that the sparse sampling is a problem for the selectionist (Darwinist) explanation.
If that proportion were high enough, then sparse sampling could still produce functional combinations. But the experiment described in connection with Figure 9 and the accompanying discussion suggests that the functional proportion is apt to be quite low. When you combine this with the rarity of the combining events themselves, there does appear to be a problem with domain combination, on top of the considerable problem of domain origination.
Doug