When Enzymes Don't Lie, Part Two
The results of Axe and Gauger's paper do not bode well for a neo-Darwinian explanation for enzyme conversion. Evolution News and Views asked Dr. Gauger what she thought of the implications of their research:
- What is the minimum number of mutations needed to convert one enzyme to a functionally different (but structurally similar) enzyme?
- Are these results feasible based on the Neo-Darwinian model?
ENV: Your paper calls into question some assumptions regarding enzymes under the neo-Darwinian paradigm, however, not everyone is aware of the assumptions -- enzyme evolution 101. Can you elaborate on the typical neo-Darwinian theory regarding enzymes and evolutionary relatedness? I gather from your paper that the idea is the more similar two enzymes are, the more likely they are related. Is the theory of relatedness based on similarity in structure or function or does it depend?
AG: Much of evolutionary theory is based on the assumption that similarity of form, structure or sequence is due to shared evolutionary history. If different organisms share a common structure, it is generally assumed that they inherited it from a common ancestor. Notice that this is an assumption, not a proven fact. This reasoning has been applied to enzymes, to anatomical structures, to body plans, to all levels of biological form. What follows, then, is another assumption. If enzymes with similar structures are found to have different functions, it is most commonly assumed that at some point a gene duplication occurred, followed by divergence of the two copies over time, leading to genes encoding enzymes with similar structure but different functions. But for this to be true, it must be possible for enzymes to switch their chemistry easily, with just a few mutations. That is what we set out to test.
ENV: Isn't it usually assumed that the active site is where all of the chemistry happens? Your paper seems to indicate that parts of the scaffold might actually affect enzyme activity. I'm thinking of your findings with the histidine residue. What does this mean for assumptions about the "inactivity" of scaffolds? Do you think we have been missing some important findings because of assumptions about the scaffold?
AG: With the advent of genetic engineering, it became possible for scientists to begin to study how protein sequence, structure, and function are correlated. It has become apparent that a protein's sequence can affect its structure and function in unpredictable ways. I doubt many protein engineers would now hold to a simple model where the scaffold serves only to hold the active site amino acids in place. Although the chemistry may be carried out by just a few amino acids in the active site, that activity is supported by many other amino acid positions in the enzyme. For example, some amino acids serve to position the substrate(s) in the right way, and others permit conformational changes upon substrate binding that are necessary for the reaction to proceed. These kinds of interactions can include second or third order interactions of amino acids some distance away from the active site. Our results with histidine 152 suggest that something like that is happening in BioF. And our results also suggest that scaffolds, even though they look the same, may not be equivalent in function. We could not swap out Kbl's active site for BioF's and get a functioning protein. Something more was needed.
ENV: In your conclusions, you point out that there is not enough time for the kinds of mutational changes to occur for a minimal number of modifications to two functionally different and structurally similar enzymes. How is your research exploring the limits of what mutations coupled with natural selection can actually do? Where do you think those limits are?
AG: Our results indicate that changes in enzyme chemistry are difficult to achieve. This is actually in line with what other protein engineers are finding. A recent review reported that changing an enzyme's chemistry may require multiple neutral or deleterious mutations (Romero PA, Arnold FH (2009) Exploring protein fitness landscapes by directed evolution. Nat Rev Mol Cell Bio 10:866-876. doi:10.1038/nrm2805). Another review stated,
Interchanging reactions catalyzed by members of mechanistically diverse superfamilies might be envisioned as "easy" exercises in (re)design: if Nature did it, why can't we? [...] Anecdotally, many attempts at interchanging activities in mechanistically diverse superfamilies have since been attempted, but few successes have been realized. (Gerlt JA, Babbitt PC (2009) Enzyme (re)design: lessons from natural evolution and computation. Curr Opin Chem Biol 13(1):10-18. doi:10.1016/j.cbpa.2009.01.014)
With our study, we have put a number on the difficulty. Switching Kbl to BioF function requires at least 7 mutations, and probably many more, putting such a transition beyond the reach of undirected natural processes.
More test cases must be examined to see how general the problem is, but if the difficulty of enzyme conversion is pervasive, a few general conclusions can be drawn. First, our result suggests that caution is needed when using similarity of sequence or structure to infer common ancestry. Before such a claim can be made, it should be demonstrated that an adaptive path between the two forms actually exists.
Second, all possible explanations for the origin of enzyme families need to be examined. Some have postulated that ancient enzymes were promiscuous, carrying out many different kinds of reactions very poorly. Over time the genes for these promiscuous enzymes might have duplicated and diverged into the specialized enzymes with high efficiency we typically find today. But it seems unlikely that such ancestral enzymes could have provided the kind of chemical diversity we see today, which means that new functions would still need to evolve. In addition, any metabolic network produced by such non-specific enzymes would be inefficient at best. Still, it is much easier to enhance a weak promiscuous function than to develop a new one.
Alternatively, what if what we see now is the fortuitous product of one-way evolution? Ancestral proteins diverged into modern enzymes in a contingent fashion, following the paths natural selection and random mutation laid down, but epistatic interactions acquired over time now prevent any simple adaptive path between modern forms. That might be why it is so difficult to convert modern enzymes from one function to the other. But if protein evolution is so highly constrained, either ancestral proteins would have had to be remarkably designed indeed, or we must have been extremely lucky that just the right variants appeared whenever needed to provide new functions. More studies that examine the constraints on adaptive evolution of proteins are needed.
Last, proteins are the building blocks of life. The diversity of proteins in living things build higher level structures and allow organisms to adapt to different environments. This is the baseline for evolution. If undirected processes can't modify proteins to new functions, they can't do much at all.
ENV: Thanks, Dr. Gauger, for taking the time to discuss the implications of your recent paper in Bio-Complexity.