If your heart is right, then every creature is a mirror of life to you and a book of holy learning, for there is no creature - no matter how tiny or how lowly - that does not reveal God’s goodness.
Thomas a Kempis - Of the Imitation of Christ (c.1420)
In part 2 of this series, we explored how the Long Term Evolution Experiment (LTEE) performed by the Lenski research group on E. Coli, demonstrates some key features of how biological “complex, specified information” (i.e. “CSI” as the ID movement terms it) arises through mutation and natural selection. To briefly recap, we noted that:
CSI does not need to arise all at once, but can arise piecemeal through independent mutation events.
Separate mutations that later combine to form CSI do not need to confer a specific advantage on their own. In other words, mutations that are “neutral” with respect to the survival of the organism can later be co-opted into CSI that does have a distinct survival advantage.
Neutral mutations may open up new future paths. In the LTEE, the brand-new ability of one bacterial population to use citrate as a food source required that a neutral mutation appear several thousand generations before it combined with other mutations to provide the CSI for using citrate.
When CSI arises, it can be pretty poor at the beginning. Nascent CSI, though poor, provides a survival advantage because it is the “best game in town” at that time. Further mutation in, and natural selection on, the offspring of the original CSI-holder quickly refine the nascent information into ever-more “specified” CSI.
And, as we noted at the close of the first post in this series, understanding how natural processes create information is in no way a threat to God’s ordaining and sustaining of creation. Rather, it is an opportunity to explore some of the mechanisms by which He does so.
With these important principles in mind, we are ready to examine a second fascinating case of a novel function arising through mutation and selection: the evolutionary history of steroid hormones and their protein receptors in vertebrates.
Experimental Evolution as Textual Criticism
The elegant work by the Lenski group has one very distinct advantage that other researchers surely envy: when new structures and functions arise in the experiment, a trip to the freezer is all that is needed to resurrect and examine the relevant ancestors. For researchers who study other organisms less amenable to laboratory experimentation, or for evolutionary transitions that happened deep in the past, other methods are needed. One approach to this type of problem is to “resurrect” ancient proteins in the lab in order to study their properties.
Bringing an ancient gene back to life starts with determining what its DNA sequence was, (and thereby determining the sequence of amino acids that made its functional protein product). While researchers don’t have direct access to ancient DNA, we have the next best thing: many modern examples of genes copied from the ancestral one.
For those who are familiar with textual criticism, the principles are very similar. Textual criticism is the process of recovering the words of an ancient manuscript by comparing several very similar, but still imperfect, copies. In general, as more copies agree on a certain wording, the more likely it is that the original had that wording. Also, the more widespread and older a certain wording is, the more likely it is original. Groups of manuscripts that have similar copying errors or other variations can be grouped together as more closely related, and so on. Given enough manuscripts, it is possible to recreate a copy of an ancient text with a very high degree of accuracy. As Christians, we benefit from this type of analysis daily when we read the Bible: though no two Greek manuscripts of the New Testament are exactly alike, scholars have used these methods to recover the original text with a very high degree of confidence.
And so too, for ancient gene sequences. Consider a hypothetical amino acid sequence in six modern organisms:
Though none of the modern sequences are identical, it is easy to see that there is a “consensus” at each of the 12 amino acid positions. This consensus sequence is very likely to be the ancestral sequence: explaining the pattern in any other way requires many more changes, with many changes occurring in parallel after species separate.
Once the researchers determine the correct ancestral amino acid sequence, it’s a relatively small matter to engineer a DNA sequence that encodes it and give it to cells to make into protein. This protein can then be tested to see how it functions compared to the modern sequence.
What makes this type of analysis even more interesting is that sometimes related genes acquire new functions. In cases like these, bringing the ancestral gene back to life in the lab allows researchers to not only test its properties, but to test hypotheses about what the specific amino acid changes were that changed the protein’s function over time:
The laboratory of Joseph Thornton at the University of Oregon has used this method (with great success) to determine how certain hormone / protein receptor complexes arose during vertebrate evolution. Hormones are small molecules that act as signals by binding to a protein target, called a receptor. The receptor / hormone pair then goes on to effect a change in the target cell by regulating other genes.
In vertebrates, two hormone – receptor pairs were of interest to the Thornton group: the mineralocorticoid receptor (MR), which binds a steroid hormone called aldosterone, and the glucocorticoid receptor (GR), which binds a second steroid hormone called cortisol (see diagram above). Cortisol can also activate MRs, but an enzyme that breaks down cortisol is present in tissues where MR is used so cortisol cannot accumulate. Aldosterone, on the other hand, cannot activate GR – it is specific to its binding partner MR. Even though these two hormone / receptor pairs regulate different processes in modern organisms, the two receptors are the result of an ancient gene duplication that occurred early in vertebrate evolution, around 450 MYA (million years ago). As time has gone by, the derivatives of the original gene have picked up distinct binding partners and physiological roles. Thornton and colleagues wanted to tease out the details of these important changes.
They started out by determining the ancestral sequence of the original receptor gene, prior to the duplication, and recreating it in the lab. When they tested this lab-designed protein, they found that it, like modern MRs, (but not GR’s)could bind either cortisol or aldosterone indicating that the ancestral protein must have been able to bind both. This result suggested that somewhere along the line the GR lost its ability to bind aldosterone and became specific to cortisol. This is interesting, because at the time the ancestral receptor was present,aldosterone didn’t exist. Aldosterone is a relative newcomer on the scene: it is present only in four-limbed vertebrates (tetrapods), which arose around 390 MYA. So, the ancestral receptor present prior to 450 MYA already had the ability to bind a hormone that wouldn’t evolve for tens of millions of years. Of course, the ancestral receptor “didn’t mind” – it had its own binding partner - a steroid hormone closely related to cortisol and aldosterone. It wasn’t sitting around doing nothing in the meantime.
This finding strongly suggested that the reason aldosterone binds only to MRs is because modern GRs, in contrast to the ancestral protein, have lost the ability to bind it. By comparing the amino acid differences between MRs and GRs, the Thornton group was able to test different combinations to see what the key changes likely were. They also did the (difficult) work of determining the precise new shape of the receptor for each of the changes that had an effect. All in all, it is an impressive body of scientific work.
Through these techniques, the Thornton group demonstrated that the loss of aldosterone sensitivity in GR occurred in a series of mutational steps that progressively remodeled the portion of the GR that binds the hormone molecule:
First, a mutation occurred that altered one of the amino acids near the hormone binding site. This change had no effect on its own (it was a neutral mutation).
Second, a change in an amino acid outside the binding pocket bent one side of the binding site into a new shape. Now the amino acid from the first neutral mutation in step #1 was thrust up against the hormone binding site. This amino acid can interact appropriately with cortisol, but not very well with aldosterone. The receptor was now strongly biased towards cortisol.
Later, several more mutations accrue that “tune” the receptor to its new specificity. Some of the mutations are neutral at first (like step #1) and then combine with later mutations to refine the receptor into its modern cortisol-specific role.
As the GR and MR lineages were becoming functionally distinct, other changes in other genes accumulated that refined their ability to regulate different processes (such as the enzyme that breaks down cortisol where MR is present, or the target genes that the two hormones regulate). While many of those details remain to be worked out, this work is an elegant demonstration of how a new function arose: gene duplication; sequence divergence with a neutral mutation that opened up a new possible trajectory; a second mutation that altered function in one of the gene copies; further mutations that refined this nascent difference; and the final result of new structures and functions that act as key regulators of important physiological processes in tetrapods, including humans.
In other words, new CSI.
Over and against these lines of evidence, however, the Intelligent Design Movement claims that such novelty is inaccessible to random mutation and natural selection. Rather, they claim that functional protein shapes are incredibly rare and therefore so isolated from each other that random mutation and natural selection cannot bridge the vast gulfs between them. Though Thornton’s work (and the work of Lenski that we examined previously) refutes this claim with detailed, concrete examples, new comparative genomics tools have addressed this issue with greater power and breadth than ever before. In the next post in this series, we’ll explore the question: are these examples rare, isolated cases, or indicative of a wider pattern?