Last February when we looked into the claims of University of Oregon biochemist Kenneth Prehoda, we saw him pounding Darwin’s pulpit with righteous fervor. He practically shouted that you could get instant animals by chance. His team’s discovery of a mutation that seemed to allow proteins to interact more easily in a choanoflagellate became the springboard for a sermon envisioning all the marvels of multicellularity without intelligent design.

Indeed, a breathless reporter from the Washington Post gave him credit for explaining human beings with that one random accident: “Every example of cells collaborating that has arisen since — from the trilobites of 500 million years ago to the dinosaurs, woolly mammoths and you — probably relied on it or some other similar mutation.”

He has toned down the rhetoric a bit in the latest news from his lab. Maybe he didn’t want to face another Twitter storm by engaging “the ire of anti-evolutionists” the way he did last time. “We’ve witnessed evolution,” he had said. “Evolution is just a fact, hands down.” Even his reviewers had gotten on his case for overstating the implications of his findings. We showed that there were plenty of empirical and logical reasons, not religious reasons, for doubting the significance of his instant-animal mutation.

This time, the news item makes more modest claims:

Just as the course of a drift boat can be irreversibly altered by a log in its path, a single mutation can send life in an entirely new direction.

That scenario, says UO biochemist Ken Prehoda, provides a window on how one mutation sparked a huge jump in the evolutionary course of a protein important for the evolution of animals.

Earlier this year, Prehoda was on a team that found that a random mutation 600 million years ago in a single-celled organism created a new family of proteins that are important for multicellular life. In a new paper, now online ahead of print in the Journal of the American Chemical Society, Prehoda and colleagues describe what the mutation did to the original protein family.

Mutations happen randomly. Most are bad news. But occasionally a mutation is good, helping an organism adapt to environmental changes or advancing overall fitness. Understanding such changes better, Prehoda said, could potentially point to new treatments for human diseases such as cancer.

Ah, yes; evolutionists can score extra points for claiming their otherwise esoteric research “could potentially” lead to cures for cancer. But we don’t need to deduct those points; there are enough other vulnerable points at risk of lowering Darwin’s score.

Prehoda’s basic claim was that a point mutation in an enzyme called guanylate kinase gave it a new protein-interacting domain (PID), launching the GK_PID family of enzymes used by all animals. And coincidentally, this mutation happened right when multicellular organisms were first appearing 600 million years ago. Could human beings be far behind?

Prehoda now reveals that all he found was that the mutation “stiffened” the GK enzyme a bit. One might think this to be a disadvantage, but he weaves a story that the stiffening of the enzyme’s backbone actually was a good thing.

The mutation, which researchers labeled s36P, set off a cascade of events in which protein interactions took new routes and evolved into more complex multicellular organisms, Prehoda said. The mutation is still conserved in all animals today, he added.

“A lot of the proteins that do the work in our bodies can be thought of as molecular machines,” Prehoda said. “They move in a way that is coordinated with function. Each protein spins in a circle or motors along filaments. Our protein, before the mutation, was an enzyme that had certain flexible movements related to its function. This one mutation fixed the protein’s backbone, locking the molecule into a shape that is important for its new function.“

Incidentally, the spinning machine is ATP synthase, and the motor is most likely kinesin. We find that out in the new paper, published this time not in eLife but in the Journal of the American Chemical Society, which does not include reviewer’s comments. For obvious reasons, Prehoda does not try to evolve ATP synthase by single point mutations.

In the paper, Whitney, Volkman and Prehoda mere “suggest” that the mutation that stiffened the GK enzyme “might have been important” for instigating new functions by “tuning” its “conformational flexibility” in some way. Even so, they retain some epistemic modesty in this less audacious hypothesis: “Furthermore, even if flexibility was important in the functional transition from enzyme to PID, we do not know how it was altered or how doing so could lead to such a dramatic change in function.”

Unfortunately, a new paper just appeared in the Proceedings of the National Academy of Sciences that undercuts their premise. Prehoda’s team assumes that random mutations can be ranked as “good” and “bad” — as if you can sort them like marbles into green jars and red jars. In this view, good things add up, and bad things get tossed out by natural selection. That was Darwin’s view, too:

It may be said that natural selection is daily and hourly scrutinising, throughout the world, every variation, even the slightest; rejecting that which is bad, preserving and adding up all that is good; silently and insensibly working, whenever and wherever opportunity offers, at the improvement of each organic being in relation to its organic and in organic conditions of life.

The new PNAS paper by Bank et al., “On the (un)predictability of a large intragenic fitness landscape,” takes a serious look at the effects of mutational interactions. Mutations aren’t like isolated red and green marbles. They interact in complex ways. “Epistasis” is a term referring to the combinatorial effects of mutations. For instance, two neutral mutations might interact to produce a benefit; that would be a case of positive epistasis. On the other hand, a seemingly beneficial mutation could have negative effects elsewhere in the organism; that’s called negative epistasis. It won’t improve an organism’s fitness, for example, if a mutation for stronger muscles also produces heart attacks.

The point of the study is that epistatic interactions are profoundly unpredictable. By performing one of the largest-ever surveys of epistasis on engineered mutations to Hsp90, a well-known protein, they concluded that it is extremely difficult to predict what will happen. Because their conclusion has far-reaching implications for all evolutionary predictions, it bears quoting in full:

Originally introduced as a metaphor to describe adaptive evolution, fitness landscapes promise to become a powerful tool in biology to address complex questions regarding the predictability of evolution and the prevalence of epistasis within and between genomic regions. Due to the high-dimensional nature of fitness landscapes, however, the ability to extrapolate will be paramount to progress in this area, and the optimal quantitative and qualitative approaches to achieve this goal are yet to be determined.

Here, we have taken an important step toward addressing this question via the creation and analysis of a landscape comprising 640 engineered mutants of the Hsp90 protein in yeast. The unprecedented size of the fitness landscape, along with the multiallelic nature, allows us to test whether global features could be extrapolated from subsets of the data. Although the global pattern indicates a rather homogeneous landscape, smaller sublandscapes are a poor predictor of the overall global pattern because of “epistatic hotspots.”

In combination, our results highlight the inherent difficulty imposed by the duality of epistasis for predicting evolution. In the absence of epistasis (i.e., in a purely additive landscape), evolution is globally highly predictable because the population will eventually reach the single-fitness optimum, but the path taken is locally entirely unpredictable. Conversely, in the presence of (sign and reciprocal sign) epistasis evolution is globally unpredictable, because there are multiple optima and the probability to reach any one of them depends strongly on the starting genotype. At the same time, evolution may become locally predictable with the population following obligatory adaptive paths that are a direct result of the creation of fitness valleys owing to epistatic interactions.

The empirical fitness landscape studied here appears to be intermediate between these extremes. Although the global peak is within reach from almost any starting point, there is a local optimum that will be reached with appreciable probability, particular when starting from the parental genotype. From a practical standpoint, these results thus highlight the danger inherent to the common practice of constructing fitness landscapes from ascertained mutational combinations. However, this work also suggests that one promising way forward for increasing predictive power will be the utilization of multiple small landscapes used to gather information about the properties of individual mutations, combined with the integration of site-specific biophysical properties.

From this, we can see that Prehoda’s team has taken a leap to think that one mutation in one enzyme would start a path to animals. He has not taken into account the effects of epistasis. Bank et al. say that local fitness peaks would be more likely to strand the animal there, rather than let it progress. Walking right past the “danger” sign, Prehoda engaged in the “common practice of constructing fitness landscapes from ascertained mutational combinations.” At best, he should only investigate a “small landscape” around the mutation to see what might happen. Maybe it would help a certain choanoflagellate. Beyond that, he is on dangerous ground making predictions.

Prehoda might have this comeback argument. He could say that his work on “ancestral protein reconstruction” shows that the mutation occurred right at the time multicellularity took off. It’s a postdiction, therefore, not a prediction. This argument, however, commits the fallacy of “affirming the consequent” — i.e., “If P, then Q. Q occurs. Therefore, P.” You can’t say P caused Q. That overlooks multiple other possibilities for Q. In fact, there could be an infinite number of causes for Q besides P. Prehoda could only argue Q if and only if P: specifically, that the emergence of animals required a specific mutation to guanylate kinase. That would be unwarranted even within neo-Darwinian theory. The best he can say is that the mutation is “consistent with” a scenario in which a stiffer enzyme contributed to new functions useful to multicellular organisms, assuming it avoided negative epistasis in the process.

Such clarification, however, would be unlikely to yield headlines in the Washington Post. Empirically speaking, Bank et al.’s yeast remain yeast, and Prehoda et al.’s choanoflagellates remain choanoflagellates.

Image credit: University of Oregon.