Design Framework for Evolution

Hi Albert de Roos,I've read a couple of your papers in the past on the design-by-contract methodology and how it relates to biological systems and I found them quite interesting.I have a quick question. Your approach is based on a design framework but it does not invoke an intelligent designer, correct?

Albert de Roos has already stated what he considers life: a replicating genome. Do viruses have replicating genomes? Yes, but that replication is often contingent on the host. There is debate within the biology community on the subject of whether or not viruses should be considered life. I don't think that debate can be resolved here. So by de Roos's definition of life, it would seem that an independently replicating genome constitutes life. If I understand him correctly, viruses by themselves are not considered life, inasmuch as transposons are not.

Not in Dutch, and de Roos's PhD education was done in the Netherlands. Just saying Added in edit: never mind. This has already been pointed out. Edited by Genomicus, : No reason given.

Hi Albert de Roos, You have applied your design-by-contract approach to things like the origin of life and the origin of the eukaryotic cell. But what about the origin of molecular machines, like, say, the bacterial flagellum? I suggest that the design-by-contract approach would show a possible evolutionary pathway, but if the flagellum was engineered, then that evolutionary pathway would not be an accurate portrayal of what happened in evolutionary history. The two competing hypotheses could be tested using something like molecular clocks.This brings me to a second question. Your scenario for the origin of the eukaryotic cell seems mechanistically plausible, but what kind of experiments could be done to test your hypothesis?

That really depends on what part of life, though. The machine analogy works very nicely at the molecular level (e.g., molecular machines actually are machines), and the machine analogy extends into the domain of multicellular life. Does life get "messy and illogical"? I suppose so, but this is the result of billions of years of evolutionary tinkering. At its core, however, life is not illogical from an engineering standpoint. That is entirely contingent on what part of life is under consideration. How is the structure of the bacterial flagellum or the F-ATPase illogical? Yet one would be hard-pressed to find the molecular equivalent of the recurrent laryngeal nerve among the core machinery of life. Why is that? Some parts of life do not make much sense from an engineering standpoint; others do. I'm pretty sure Albert de Roos was using the baseball as an example of the methodology of the design-by-contract approach; he was not using the baseball as an analogy of how cells work.

By that argument, then at the heart of all of technologies is simple chemistry. This is indeed true, but it misses the point. From the perspective of the genome - that is, when we consider the core features of the genomes of various taxa - life is not "illogical" from an engineering standpoint. The core structures of the bacterial flagellum and the F-ATPase are perfectly suited to performing their respective functions. That is, there is a logic to the arrangement of the parts of these machines - a logic that is defied in, for example, the backward wiring of the mammalian eye. There are plenty of possible arrangement of protein parts that would produce motility. Yet the vast majority of these arrangements would not produce optimal function - that is, the vast majority of the possible arrangements would result in a system that is the molecular equivalent of the recurrent laryngeal nerve. There are many possible protein arrangements that would produce motility and thereby provide an evolutionary advantage. Only a small fraction of these arrangements would make sense from an engineering perspective.Consider the bacterial flagellum. It is composed of two rings (the MS-ring and the C-ring); the C-ring houses the export apparatus (the flagellum-specific ATPase). FliC monomers (from which the filament is constructed) are exported by the F-ATPase through the FliF pore; other flagellar parts are also transported in this manner (e.g., the hook and junction proteins). Yet during the evolution of this system (specifically, when the F-ATPase associated with the primitive FliF pore), it would have been easy enough for the F1 component of the ATPase to bind partially to the side of the FliF pore (instead of fitting neatly into the "hole" of the pore, as it does in actual flagella), thereby minimally clogging transport of filament proteins. Perfect binding of the ATPase to the FliF pore would not be expected anyway, since evolution is a tinkerer, not an engineer. If this had occurred, the co-opted ATPase would still provide a selectable advantage to the organism by energizing protein transport, so this configuration would be inherited. Subsequent addition of protein parts to the proto-flagellum would occur; these parts would be integrated around the faulty-but-functional ATPase configuration. Thus, if a mutation occurred to place the ATPase directly in conjunction with the FliF pore, the system could no longer function properly. This is because the arrangement of the other protein parts would be scaffolded on the sub-optimal placement of the ATPase.This illustrates my point: that one would be hard-pressed to find the molecular equivalent of the recurrent laryngeal nerve among the core machinery of life, despite the fact that it is perfectly acceptable on evolutionary grounds. Edited by Genomicus, : No reason given.Edited by Genomicus, : No reason given.

A similar argument was raised in the past. From: My response was as follows: I would be interested on hearing your thoughts on where exactly the Krebs Cycle is the molecular equivalent of the recurrent laryngeal nerve. Which step in the Krebs Cycle should be eliminated? How would you design a system that performed the same function as the Krebs Cycle?Edited by Genomicus, : No reason given.Edited by Genomicus, : No reason given.

Your motorcycle is not designed to do so. But your motorcycle will convert gasoline into usable energy "all on its own." But salt crystals are not considered life, and the discussion here is whether life is "just simple chemistry." It is, but only in the sense that a gasoline engine is "simple chemistry." From the perspective of the genome, we are not looking at salt crystals, since the genome encodes the information needed to construct molecular machinery. When I say "genome," I do not mean merely the nucleic acids, but the information as well. So, when we consider the core features of the genome - that is, the features of the genome that are found in virtually all taxa (e.g., the Krebs cycle) - where do we find "illogical" or "irrational" design? Then why do we have a system like the recurrent laryngeal nerve? Why wasn't it selected against, since there's obviously a more optimal way to produce the same function? Now, I'm pretty sure we both know the answer to that: the recurrent laryngeal nerve is the result of evolution from an aquatic ancestor that possessed a different anatomy. In other words, stages in evolution constrain the nature of possible adaptations. However, we must still wonder why there are no "recurrent laryngeal nerves" in the core machinery of life, while there are such things in the anatomy of multicellular organisms. But that's not a completely accurate picture of how evolution works. If something works, then evolution will not throw it out. Even clumsy function is better than no function. Thus, I don't see how your response addresses the problem of why the core structure of the bacterial flagellum displays no "irrational design," when it could have very probably displayed this on evolutionary grounds. Edited by Genomicus, : No reason given.

From where I'm reading, your argument above is that the Krebs Cycle appears to use proteins that would make sense from the perspective of evolutionary history. But this is not what is at issue. We may discuss the arguments from homology, etc., but at present the issue is whether the Krebs Cycle is the molecular equivalent of the recurrent laryngeal nerve. From a paper discussing the evolution of the Krebs Cycle (again, the discussion of the evolution of the Krebs Cycle is just a bit different than the issue of whether it is the molecular equivalent of something like the recurrent laryngeal nerve): Source: The Puzzle of the Krebs Citric Acid Cycle: Assembling the Pieces of Chemically Feasible Reactions, and Opportunism in the Design of Metabolic Pathways During Evolution, 1996. Edited by Genomicus, : No reason given.Edited by Genomicus, : No reason given.

Hi Panda, I agree that they do not explicitly defend the claim that "a chemical engineer could not have found a better design." The paper clearly explains how the Krebs Cycle is an optimal design based on the materials at hand, yet there is no reason to suppose the Krebs Cycle is not also optimal from an engineering standpoint. In other words, the statement that its design is a "very clear case of opportunism," does not, in itself, suggest that an engineer could construct a better Krebs Cycle. While one might argue that the signatures of opportunism are indicative of evolution, this does not show how the Krebs Cycle is a poor engineer solution. This is an another issue, IMHO.

I'd say the genome is not a machine, but part of a machine; namely, the machine that is the cell. But that's just me. Others could argue that the various genomic interactions (e.g., DNA methylation) suggest that the genome is a molecular machine.

On second thought, you're right. The provided quote doesn't address AZPaul's argument. Yet I still do not see where exactly the inefficiencies lie. Merely because the Krebs Cycle uses parts in other systems does not make it inefficient. What part of the Krebs Cycle is inefficient, or a sub-optimal design?

Yes. But the genome conveys the information needed to construct protein sequences. This information is not purely chemical; while mRNA translation operates according to the rules governing chemical interactions, there is no chemical law that states that, for example, the codon ATG must encode methionine. Edited by Genomicus, : No reason given.