The other morning in my Google News feed I got the following Nature “News and Views” article about evolutionary trees and rates of speciation and extinction:
I’m curious whether that is a coincidence since I do click on Google's sciencey or science-ish or even outlandish science fiction news (if you put “Lock Ness” in a title I will click on it) – or whether this is a specific instance showing how much Google knows about what I’m doing on my computer – because I’ve recently been playing with evolutionary (or phylogenetic) trees of my own - check out one I made here (click on the upper right corner of the image to magnify it):
Darwin first sketched out conceptually what an evolutionary tree might look like in one of his notebooks. Check out this classic image below. I love in particular the words he scrawled in the upper left of that page: “I think”.
What exactly is an evolutionary tree? I think of it as a graphical way to rank how similar or different something is on a framework built on Darwin’s theory of evolution. If some ancestral species has recently evolved into two species, those two species are initially very closely related, and any differences in their characteristics (size, color, DNA or protein sequence, etc) we might imagine are quite small: a squirrel with red fur compared to one with gray. Graphically, we represent that as two adjacent twigs – the two new species are the tips of the twigs and the ancestral squirrel is the branch point of the twigs. As time passes and each species gives rise to others, the measurable differences may become quite large: a lumbering saber-toothed carnivorous squirrel compared to a small one that still likes peanuts. Graphically we have a series of branch points and end up with long twigs that are separated by significant distances and branchings – like the lower right overhanging branch in Darwin’s sketch. There is a nice explanation of evolutionary trees by the Khan Academy here:
My tree above looks at just one protein (Cdc25) and not whole organisms because that is the protein that I am researching. (To be more precise, I am not studying the protein itself, but how the gene is regulated.)
Out of a totally irrelevant, social-distancing-enabled curiosity I wanted to know how this particular protein was related evolutionarily among various organisms.
Now before I go further, I want to mention before anyone yells at me – there is a carefully crafted convention in biology for naming genes and proteins with rules for capitalizing and italicizing and all that jazz – I’m going to ignore those conventions and call the protein and gene "Cdc25" among all the species. Sorry if you’re a stickler for those kinds of things.
So… I grabbed the amino acid sequences of Cdc25 proteins from a bunch of organisms that have been posted online on the NCBI website here:
I entered the protein sequences into a new (new for me) online tool called Clustal Omega. That tool looks at all the sequences that you enter and finds the best matched pair, then the next closest to that pair, and so on, to build that tree. You can access Clustal Omega here:
If you’re a math and coding geek and like to see how these things work under the hood, here are a couple papers that describe some of the inner workings of Clustal Omega:
Here’s a bit of background on the Cdc25 protein and why I think that tree is fascinating…
The gene encoding the Cdc25 protein was first discovered to control cell division in fission yeast. Paul Russell and Paul Nurse made this discovery and published it in 1986, which eventually earned Paul Nurse the Nobel Prize in 2001.
Here is Paul Nurse’s Nobel Prize lecture:
Here is something much more interesting than a Nobel Prize lecture - a Moth Story Hour story told by Paul that is quite entertaining:
And one more meandering side-track… Paul Russell was Paul Nurse’s post-doc. My boss was Paul Russell’s post-doc. And me… well I’m not a post-doc. ;)
Anyways…
Finding the genes that control cell division is important since it is one of the fundamental processes required for any cell to propagate and is key to understanding diseases like cancer. But finding it in yeast elicits an automatic “who cares” from most people, scientists included. And the Nobel Prize wasn’t even a consideration for Nurse until it was found that humans have Cdc25 as well. If you put the human Cdc25 gene into yeast which has a defective Cdc25 gene, the human version will “rescue” the defective yeast gene - so we know it is a highly conserved gene spanning a diversity of species from single-celled yeast to humans. Pretty cool huh?
Here’s where it gets fun from that evolutionary perspective.
Humans have three Cdc25 genes. We call them Cdc25A, Cdc25B, and Cdc25C. Just like if you had triplets, you’d name them: Thing A, Thing B, and Thing C (a la Dr. Suess).
And yeast have a single Cdc25 gene.
I wanted to know how, and how closely, these various Cdc25 genes were related among a variety of species crossing that evolutionary chasm from a single-celled yeast to a human.
I expected the three human Cdc25 genes to cluster together and be closely related to each other, which would be next closest to, say, three closely related mouse Cdc25 genes, which would in turn be close to reptile Cdc25, then to frog, then fish…. all the way down to yeast.
But that’s not what it looks like at all.
Look at that tree again.
I color coded groups of organisms and labeled it on the right-hand side with big bold letters in each colored region as “Yeast” or “Mammal” etc.
You can see there is one region at the top labeled “Yeast” and it has a bunch of yeast species, including the one I work with, Schizosaccharomyces pombe. That makes sense. All the yeast Cdc25 proteins are closely related to each other and their sequences are algorithmically clustered together into that tree. So far it seems to match that pattern of closely related species being close together on the branches of the tree.
But let’s look a little closer.
Look at the “Mammals” group - you see three reddish color groups with that label - and each one contains a human and a mouse Cdc25 gene. That’s odd.
One human and mouse Cdc25 gene is more closely related than the three human Cdc25 genes are related to each other.
In fact, it is even weirder.
Human Cdc25C is more closely related to frog (Xenopus laevis and Xenopus tropicalis - in light blue) than they are related to human Cdc25A or B.
Human Cdc25A is more closely related to lizard (Podarcis muralis - in yellow) than to human Cdc25B or C.
And human Cdc25B is more closely related to fish (Danio rerio, Nothobranchius furzeri, etc - in purple) than to human Cdc25A or C.
Ain’t that weird?
From a broader perspective, a couple aspects of this evolutionary relationship illustrated in that tree are well known.
For example… we humans share a basic developmental program with lizards, frogs, fish, lamprey, on down to segmented worms…. We all have a segmented developmental program - those segments are easy to see in things like worms and insects and other organisms which have - well - segments or repeated units making up their bodies (think of a lobster’s tail). Our own segmentation is harder to see - it is internal - and evidenced by our vertebra. When you fillet a fish, the beautiful segmented, repeating pattern becomes obviously clear. The gene, the protein molecule(s) that control the segmental development of everything from segmented worms to humans are called Hox genes. There is a similar high degree of evolutionary conservation in Hox genes as for Cdc25. Furthermore, humans and many vertebrates have multiple (four) clusters of Hox genes (Hox A, B, C and D), and each cluster is composed of a series of 13 individual Hox genes (named in true Suessian fashion as HoxA1 through HoxA13, etc).
Which means that the whole cluster of thirteen Hox genes (or the whole genome carrying them) was duplicated twice in the evolutionary distance between segmented worms and humans.
One of the ways that genes and organisms can evolve rapidly is through duplication events - parts of genes, whole genes, parts or whole chromosomes, even whole genomes. Most often these duplication events are damaging or deadly events. For example, some human diseases are associated with extra chromosomes (like Down syndrome which is known to be caused by an extra copy of chromosome 21). But on occasion this duplication appears to have had beneficial evolutionary advantages, and we currently know of several times when whole genomes were duplicated in various organisms. The duplicated genes or chromosomes or whole genomes can in some cases lead to dramatic evolution of new functions or features and eventually species.
The Hox genes that once simply controlled the sequence of segments in a simple worm hundreds of millions of years ago, now not only control the sequence of our vertebrae, but also controls the development of our limbs down to our fingers and toes. Hox mutations are often lethal, but some minor Hox mutations will cause things like polydactyly (extra fingers or toes).
That’s a peek at the evolutionary power of duplication events.
I think the Cdc25 evolutionary tree shows that somewhere between fish and frogs the Cdc25 gene (if not the whole genome) duplicated once, whereupon the 2nd copy began to drift and pick up different functions and characteristics from the 1st copy… and somewhere between frogs and lizards a third copy of Cdc25 (if not the whole chromosome) was made and it also began to drift evolutionarily from the other two. So today, most vertebrates from lizards to mammals including humans have three copies of Cdc25 which retain a sequence similarity that traces back to approximately when the duplication happened (the branch point on the evolutionary tree).
Just like we have traces of a tailbone that are only visible on x-ray… what we’re seeing in those phylogenetic trees are invisible traces of our evolutionary past locked in our protein sequences. Which I think is some seriously cool geewizery.