Taking on new functions, protein-wise

I have to note this commentary by Fyodor Kondrashov that I ran across. The commentary discusses this 2004 paper by Amir Aharoni and colleagues:

The 'evolvability' of promiscuous protein functions
How proteins with new functions (e.g., drug or antibiotic resistance or degradation of man-made chemicals) evolve in a matter of months or years is still unclear. This ability is dependent on the induction of new phenotypic traits by a small number of mutations (plasticity). But mutations often have deleterious effects on functions that are essential for survival. How are these seemingly conflicting demands met at the single-protein level? Results from directed laboratory evolution experiments indicate that the evolution of a new function is driven by mutations that have little effect on the native function but large effects on the promiscuous functions that serve as starting point. Thus, an evolving protein can initially acquire increased fitness for a new function without losing its original function. Gene duplication and the divergence of a completely new protein may then follow.

This can be understood through a broader consideration of protein structure and function. Enzymes catalyze reactions in large part due to their structures, which include binding domains for different biomolecules. These binding domains evolve specificity for particular target molecules, so that they have a lock-and-key configuration. But the system is not a perfect one-key-fits-one-lock arrangement. Instead, an enzyme may catalyze additional reactions inefficiently, over and above the one it has been adapted to catalyze well. Small changes in the amino acid sequence can sometimes add significantly to the efficiency in catalyzing these other reactions, without detracting from the primary adaptation.

Aharoni et al. (2004) demonstrated that this kind of change readily occurs for some proteins. In such cases, it is fair to say that there is no adaptive valley of reduced fitness between one enzymatic form (ancestral) that catalyzes one reaction efficiently, and another form (derived) that catalyzes two reactions efficiently. For the current discussion, I'm going to call such a protein overloaded -- it is doing twice the jobs we might expect it to do.

The paper raises a natural question from this observation: If it is often this easy for proteins to evolve additional functions, then why don't we see more multifunction proteins? Why aren't more proteins overloaded?

I think that the answer from a fitness perspective is not so simple. We see many gene duplications that have subsequently diverged in function. If a single gene could take on two functions with no fitness cost, we shouldn't probably see so many instances where the functional differentiation has followed gene duplication. Clearly there is a fitness cost; otherwise there would be no advantage to having duplicate genes instead of a single gene. It is even possible to imagine that overloading a protein with different functions might make subsequent duplication more likely. Overloading might be a temporary solution that provides a transition between a low-fitness one-function protein and a two high-fitness proteins with different functions.

Kondrashov (2005) (and this is weird; Nature Genetics has given them different dates even though they're in the same issue. It's because of their online publication dates. Get it together, Nature!) discusses what we know about the distribution of multi-function proteins:

There are many examples of proteins that have multiple functions: crystallins, antifreeze proteins, the p53 tumor-suppressor protein and a broad class of enzymes that recognize different substrates all come to mind. On the other hand, many other proteins can be described with the broad generalization of one geneone function. Why do proteins not take advantage of the opportunity to carry out multiple functions more often? It is true that evolution cannot produce the optimal solution, especially if doing so requires radical changes; think of a bilaterally symmetrical flatfish larva struggling to mold itself into an adult. But the results of Aharoni et al. show that the lack of stepping stones is probably not an obstacle for the evolution of new enzymatic functions, because even a small number of simple amino-acid substitutions are enough to improve promiscuous functions.
It is possible that broadening the range of functions of a molecule may be unnecessary, so that natural selection maintains the native function but never acts to improve the promiscuous one. This is analogous to mutational explanations of aging, which argue that aging occurs because natural selection does not act against mutations that impair performance only after reproductive age. This hypothesis, however, cannot explain why the evolution of new functions, when they are actually needed, does not generally proceed through the acquisition of multiple functions by a single enzyme.

He posits that multiple enzymes with distinct functions must in general be favored at the cellular level. He suggests that the evolution of multiple functions for single proteins may spur duplications as a way to maintain one-protein one-function systems. If he's right in this suggestion, which seems like the most likely alternative to me, then most gene duplications that become fixed probably are not neutral. Particularly if the different functions occur in different cell types; in which case simple regulatory differences between duplicate genes might provide functional differentation more-or-less automatically.


Kondrashov FA. 2005. In search of the limits of evolution. Nat Genet 37:9-10. doi:10.1038/ng0105-9

Aharoni A, Gaidukov L, Khersonsky O, Gould SM, Roodvelt C, Tawfik DS. 2004. The 'evolvability' of promiscuous protein functions. Nat Genet 37:73-76. doi:10.1038/ng1482