Everyone in Seattle is Seahawks-crazy now, so the football reference is well timed, at least. A reader points out the following from a post by biochemist Larry Moran, "Ann Gauger moves the goalposts":

What she and Doug Axe were really trying to do was to intelligently design an entirely new enzyme.

It seems Moran himself is the one who did the shifting, given the content of my previous post, but he may have inadvertently said something true.

The subject under discussion was our recently published paper, "Enzyme families — Shared evolutionary history or shared design," and our previous one, "The evolutionary accessibility of new enzyme functions."

In those papers, we did not set out to design an "entirely new enzyme." So in this sense Moran is wrong. The two enzymeswe were working with (Kbl₂ and BioF₂) look alike, share the same structure, and even share the particular amino acids in their active sites, so changing one into the other would not be making an "entirely new enzyme" by standard definitions, especially if it could be done with just a few mutations.

Getting that shift is beyond the reach of unguided Darwinian processes, and working from ancestral proteins won’t help. It is a red herring to say that we should have started with the hypothetical ancestral form. Epistatic interactions all along the way make the unguided evolution from ancestors highly unlikely. For all modern enzymes to have threaded very rare paths from their ancestral states, particularly when new chemistries were required, makes the whole process extremely fine-tuned.

In any event, because they have different chemistries, it turns out we were trying to make a new enzyme, in a sense that I will outline below.

It is possible to shift an enzyme to favor a "new" function when it already can carry out that reaction, but very weakly. This shift is often reported as the evolution of a new function in the literature, but almost invariably careful reading reveals that the starting enzyme began with a small amount of the target activity, and so it cannot be counted as genuinely new.

It’s only making a reaction that does not already exist that counts as making a genuinely new function.

Protein engineers acknowledge that getting this to happen is very hard. Romero and Arnold put it this way:

Evolution by the accumulation of single mutations has proven to be very effective at optimizing a function or property that already exists or can be reached through a series of intermediate steps. Some functions, however, simply can not be reached through a series of small uphill steps and instead require longer jumps that include mutations that would be neutral or even deleterious when made individually. Examples of functions that might require multiple simultaneous mutations include the appearance of a new catalytic activity or an activity on a substrate for which the parent and its single mutants show no measurable activity. [Emphasis added.]

Though many have tried to evolve new functions, there have been very few successes. Here’s a passage from a paper on just this subject.

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? Indeed, such speculation began as soon as the enolase superfamily was discovered [12]. Anecdotally, many attempts at interchanging activities in mechanistically diverse superfamilies have since been attempted, but few successes have been realized. [Emphasis added.]

Successes typically require multiple mutations at once, or are so weak in activity that they require over-expression to provide any selectable advantage to the cell. Together with the necessity of gene duplication to allow the extra copy to evolve, the requirements for mutation and overexpression put the evolution of the genuinely new function beyond reach.

If getting genuinely new functions is so hard, where do new functions come from? Clearly we have thousands of enzymes with distinct functions to explain.

Another way of thinking about the arrival of new functions would be more fruitful. Even though the two enzymes (Kbl₂ and BioF₂) we worked with look alike, the way they are put together is distinct. The particular amino acids that cause them each to fold into the same structure are unique in sequence and holistic in their interactions. Each requires high-level, top-down design.

As we discuss in our recent paper:

It may be that our prior attempts to convert Kbl2 to perform the function of BioF₂ failed not because we made the wrong alterations but rather because it is misguided even to think of this as an exercise in alteration.

… They use similar structures not because they are both adjusted versions of some older enzyme, but instead because the purposes they serve happen to call for similar structures. As we found in this work, it is not that Kbl has amino acid residues that are incompatible with the function of BioF₂, but rather that Kbl₂ is comprehensively suited to one function, while BioF₂ is comprehensively suited to another.

So in a sense Moran is right, if that is what he meant. Kbl₂ and BioF₂ represent two distinct ideas or concepts requiring holistic design.

Image credit: clappstar/Flickr.