Historical Contingency and the Improbability of Protein Evolution, Part 1 (of 2)

Historical Contingency and the Improbability of Protein Evolution, Part 1 (of 2)

Can evolutionary processes produce biological innovation?

Critics of the evolutionary paradigm—including me—would say, “No.” However, the reasons for my skepticism differ from many of evolution’s chief detractors. One argument against the evolutionary paradigm that causes me discomfort has to do with the “improbability” of the evolutionary process. For example, one common version of this argument relates to the evolutionary emergence of proteins, with critics asserting that the evolution of novel proteins from preexisting proteins would have been so improbable that it defies an evolutionary explanation. To justify this position, these critics often point to studies such as the one published by scientists from the Universities of Oregon and Chicago that seemingly buttresses their point.1 But does it?

The Evolutionary Origin of a Protein Receptor

This research team hoped to gain insight into the role that chance historical events play in evolutionary processes. Working within the framework of the evolutionary paradigm, they determined what they believe to be the amino acid sequence and structure of the ancestral protein that evolved into the cellular receptor protein that binds the hormone cortisol. They claim to have resurrected an ancient protein they believe existed 450 million years ago, before the cortisol-specific glucocorticoid receptor evolved its specificity for this particular hormone.2

Today, the cortisol-specific glucocorticoid receptor assumes a key role in the endocrine system by regulating development and the stress response. The activity of this protein is mediated by cortisol binding. However, the researchers believe that in the past the ancestral protein was biochemically promiscuous, binding a number of hormones, and only later evolved its specificity for cortisol through amino acid changes mediated by the putative evolutionary process. Based on a reconstruction of the evolutionary pathway, they conclude that seven amino acid changes transformed the ancestral receptor protein into one that exclusively binds cortisol.

The researchers classified the changes into two categories: 1) functional; and 2) permissive. They deemed five of the changes as functional, meaning that these changes contributed to the receptor’s cortisol-binding specificity. They dubbed the other two changes as permissive. These changes do not contribute to the binding specificity of the glucocorticoid receptor, but must occur for the functional changes to take effect. In other words, if the functional changes took place independently of the permissive changes, the resulting hormone receptor would not bind cortisol. The researchers determined that the permissive changes help to stabilize the receptor protein’s structure so that it can tolerate the five functional changes.

Because cortisol binding depends upon the permissive mutations, the researchers reasoned that historical contingency must have played some role in the evolution of the cortisol-specific receptor protein. The permissive mutations must have appeared first, because if they didn’t, the functional changes would not have been selected (again) since they aren’t functional apart from the permissive changes.

The Improbability of Protein Evolution

The question then becomes, “How prominent is contingency in the evolutionary history of the cortisol-specific receptor protein?” To address this point, the investigators synthesized the ancestral receptor protein with the five functional amino acid changes (AP+5). Then, they subjected the AP+5 protein to random amino acid changes to try and determine the number of possible alternate permissive changes that could stabilize the receptor protein in the same way as the historical permissive changes.

They screened about 12,500 random variants of the AP+5 protein. These variants yielded an estimated 1,025 unique single amino acid replacements, 1,802 unique double amino acid replacements, and 825 unique higher order combinations of amino acid substitutions. That is, they examined about 3,650 variants of the AP+5 protein. (The other 8,850 variants were duplicates of the 3,650 variants.) They also engineered 10 additional AP+5 variants using rational design principles. To their surprise, none of the 3,660 variants (3,650 in the screened library, plus the additional 10 engineered double mutants) yielded a functional cortisol-specific receptor that would not disrupt the function of the ancestral protein. (Four of the AP+5 variants displayed cortisol-specific binding, but these four changes destroyed the function of the ancestral protein. From an evolutionary perspective, these alternate permissive substitutions would have been selected against because of their disruptive influence.)

This result indicates that it is highly improbable that the permissive amino acid changes necessary to support the evolution of a cortisol-specific receptor protein could ever occur (with an upper bound of 0.03 percent). The researchers conclude:

“The total frequency is probably far lower…The universe of possible variants containing two or more replacements is very large, so alternative permissive sets may exist; however, these genotypes would require multiple independent substitutions, and the joint probability of such events would be very low because they cannot be acquired deterministically by selection for the derived function.”3

Their probability assessment doesn’t even include the likelihood of the five functional changes occurring after the two permissive changes took place, meaning that the probabilities for the evolution of the cortisol-specific receptor protein from a promiscuous ancestral receptor are even more unlikely.

The Contingency of the Evolutionary Process

As a skeptic of the evolutionary paradigm, it is tempting to point to this study as evidence that evolutionary transformations are so improbable that these processes cannot account for biological innovation. But this would be an unfair conclusion that misrepresents the way evolutionary biologists interpret these results. Instead, these scientists argue that these results tell them something about the evolutionary process: Namely, that historical contingency plays a central role in evolutionary transformations.

The concept of historical contingency is the theme of the late Stephen Jay Gould’s book Wonderful Life.4 According to this idea, the mechanism that drives the evolutionary process consists of an extended sequence of unpredictable, chance events. To help clarify this concept, Gould used the metaphor of “replaying life’s tape.” If one were to push the rewind button, erase life’s history, and then let the tape run again, the results would be completely different each time.

According to the researchers:

“If evolutionary history could be replayed from the ancestral starting point, the same kind of permissive substitutions would be unlikely to occur. The transition to GR’s [glucocorticoid receptor’s] present form and function would likely be inaccessible, and different outcomes would almost certainly ensue. Cortisol-specific signaling might evolve by a different mechanism in the GR—or the vertebrate endocrine system more generally—would be substantially different.”5

A Flawed Argument

In other words, while evolutionary transformations are highly improbable, their unlikelihood cannot be used as a legitimate basis for skepticism about the evolutionary paradigm. To use them in this way would be to make a straw man argument against biological evolution. This probability argument assumes that evolutionary end points are fixed, but evolutionary biologists don’t see them that way at all—because of the historically contingent nature of the process.

Still, there are some legitimate reasons to be skeptical about the capacity of evolutionary mechanisms to account for the design and diversity of life. And one of those reasons is exposed by this study and the historically contingent nature of the evolutionary process.

I will elaborate in my next blog post.


  1. Michael Harms and Joseph Thornton, “Historical Contingency and Its Biophysical Basis in Glucocorticoid Receptor Evolution,” Nature 512 (August 2014): 203–7.
  2. For a Christian perspective on resurrected ancient proteins, see my article, Fazale Rana, “Resurrected Proteins and the Case for Biological Evolution,” Today’s New Reason to Believe (blog), Reasons to Believe, October 14, 2013, https://www.reasons.org/articles/resurrected-proteins-and-the-case-for-biological-evolution.
  3. Harms and Thornton, “Historical Contingency,” 204.
  4. Stephen Jay Gould, Wonderful Life: The Burgess Shale and the Nature of History (New York: W. W. Norton & Company, 1990).
  5. Harms and Thornton, “Historical Contingency,” 207.