Global Analysis of Cdk1 Substrate Phosphorylation Sites Provides Insights into Evolution (Science, Sept 2009): Phosphorylation constrains the evolution of amino acids sequences... but not by a whole lot

Some terms: Phosphorylation and kinases were described in the last post, so scroll down to learn all about them.

Cell cycle: Cells divide and multiply over time, and the rate at which they do so is tightly controlled by a network of proteins, the core group of which participate directly in regulating the cell cycle. This cycle describes a cell as it goes from resting to preparing itself for division (primarily by duplicating all of its DNA for splitting) to finally dividing into two daughter cells. Cdk1 is one of a family of kinases that positively regulate the cell cycle. In cancer, overactivation of Cdk2, Cdk4, Cdk6, and/or Cdk8 have been implicated in a good percentage of cases: they cause the cells to multiply too often. However, the Cdk1 this paper is concerned with is yeast Cdk1. It phosphorylates serines (S) and threonines (T) on hundreds of proteins. Part of how Cdk1 knows which S/T to phosphorylate is whether it is followed by a proline (P), and sometimes also by a lysine (K) or arginine (R) two amino acids further down from the proline. Biologists write this "consensus phosphorylation sequence" as S/T*-P-x-K/R, where * is the phosphorylation site and x is any amino acid. Different kinases have different consensus sequences.

The basic results: What they did here is first to determine all of the proteins that yeast Cdk1 phosphorylates and where. They did this by comparing normal-growing yeast to yeast that had Cdk1's normal control of the cell cycle interrupted by chemicals. They then used an insanely expensive machine called a tandem mass spectrometer to figure out which serines and threonines were differentially phosphorylated (if you really want to get technical, they made this comparison possible by growing one set of yeast normally and the other set using "heavy" carbon and nitrogen; this allows the spectrometer to distinguish a mixture of the two by mass which makes everything a lot simpler). In total, they found 547 serines and threonines that were phosphorylated by Cdk1.

Ok, now the evolution part. They took the yeast amino acid sequence of all 547 sites and went to the web to download the amino acid sequences of the same 547 sites from 27 different types of fungal organisms, including 7 other kinds of yeast. Then, for each protein, they aligned the sequences for the 28 total organisms. Now, normally, if the amino acid sequence for a particular stretch lines up across all these species, it means that that stretch was specially preserved through natural selection because it's important. They found that about 60 of the phosphorylation sites retained an almost-perfect sequence alignment across the 28 organisms: since the cell cycle is so fundamental to growth, it should have a strong hand in guiding phosphorylation site evolution.

What's surprising is how low that number is. To understand why this happened a little better, the researchers used a different way of looking at phosphorylation site conservation. Instead of looking for precise alignment of the sites, they instead looked to see if the protein had some other way of conserving these crucial serines and threonines. What they found is that over 300 of the sites had phosphorylatable serines and threonines (remember, this is defined by the consensus Cdk1 sequence S/T*-P-x-K/R) at least somewhere in the very close vicinity of the original yeast site. In other words, for most of the sites, it doesn't matter what the specific larger sequence context is, as long as there's a Cdk1 consensus phosphorylation sequence in about the right region. This kind of mushiness might help to integrate new and evolving redundant kinase signals without upsetting the overall balance.

What this means for evolution: The high flexibility of Cdk1 phosphorylation site evolution within just the relatively small cluster of fungal organisms strongly suggests that evolution can be a highly adaptable process. This might be particularly true of such fundamental processes as the cell cycle, which likely faces multiple sources of constant environmental selective pressure. It will be interesting to see if similar experiments are performed on higher organisms, which may provide further insight into whether the rates of phosphorylation site evolution correlates with kinase family complexity, to tie this into the last post (it's no coincidence that these papers were published back-to-back in the same issue). Once again, we get to see the power that kinases - and in this case one particular kinase - can have on protein sequence evolution: significant swaths of the protein pantheon conform to the mercurial nature of the kinase.

Positive Selection of Tyrosine Loss in Metazoan Evolution (Science, Sept 2009): Protein diversity can select for changes in DNA

Some terms:
Phosphorylation:  Proteins are made up of chains of amino acids. Though there are 20 essential amino acids, only three are likely to have a phosphate group added to it: serine (S), threonine (T), or tyrosine (Y). The addition of the phosphate group can have profound changes on a protein's structure, function, and interactions with other proteins. For example, certain phosphorylations will change the shape of a protein such that one part blocks the function of another part; this is called auto-inhibition. Other phosphorylations activate whole proteins, waking them up from an inactive slumber. Yet other phosphorylations are bound by phosphate-binding proteins, which can alter the location of the bound protein or bring it closer to other proteins that it can interact with.

Kinase: A kinase is an enzyme (which itself is a protein) that phosphorylates another protein. Kinases, particularly tyrosine kinases tend to play essential roles in cells, including ones that prevent or accelerate cancer. Now, the thing is, kinases are very specific: they don't just add phosphates to any old protein. Any one particular tyrosine kinase, say, Cdk1, only phosphorylates a small number of certain sites on certain proteins, even though the number of such target sites in total numbers in the many hundreds. This tends to be fairly well conserved between and among different species.

 

The basic results: The researchers looked through 16 different species, from yeast through worms and flies up to cows and humans, and counted the total number of tyrosines found in all of the different proteins encoded for by the genome. As a measure of organismal complexity, they also counted the number of different cell types in the organisms, with yeast at one and humans topping out at 160+. What they discovered was that the total number of tyrosines decreased (from 3.4% of all amino acids down to 2.6%) as multicellular complexity increased. In other words, the more complex the creature, the less it wants to have tyrosines around in its protein repertoire.

The reasoning goes like this: as evolution proceeds, new tyrosine kinases come about through gene duplication events and subsequent diversion. These new tyrosine kinases are a little bit different from their predecessors, so they're likely to phosphorylate a new set of tyrosines. At the same time, other new proteins are evolving as well, so both old and new tyrosine kinases have to decide, in a way, whether or not to phosphorylate the tyrosines on these new proteins as well. The whole thing can get a little confusing for the cell, not knowing ahead of time whether or not it's good to phosphorylate this tyrosine or that one.

So, it goes through natural selection on this microscopic scale. Most newly phosphorylated tyrosines will usually throw the cell's carefully structured phospho-network out of balance, and even if they don't directly detriment the cell, the chaotic nature of too many tyrosine phosphorylation sites can get messy. The answer that evolution came up with was to get rid of too many tyrosines. This decrease means a little more simplicity in the total system, which should mean better fitness. Indeed, the total number of tyrosines also decreases as the total number of different tyrosine kinases increases. Fewer targets, more controllers, better control.

What this means for evolution: This is evolution writ large at both the microscopic and macroscopic levels. The DNA of the genome changes in response to the proteins surrounding it - a dynamic back-and-forth played out over millions of years across many, if not all species. It's not simply that less complex systems have less complex controls. It's that we can follow the ups and downs of the tyrosine kinase family by examining the tracks it has left behind in the DNA which encodes the entire body of real and potential target proteins. Phosphorylation is such a fundamental and powerful force in normal eukaryotic cellular processes - the mutation of a single kinase such as Ras can lead to unbridled cancer - that its power is also felt through evolutionary time.

Massive Horizontal Gene Transfer in Bdelloid Rotifers (Science May 2008): Bacterial genes can jump into animal genomes

http://www.sciencemag.org/cgi/content/short/320/5880/1210

Some terms: I'll use the terms "bacteria" and "animal" to represent the proper scientific terms "prokaryote" and "metazoan."

Rotifer: a tiny freshwater invertebrate that spins as it swims. I remember they showed us a few under the microscope in high school bio.

Horizontal transfer: the movement of genes from one organism to another other than through direct inheritance. This can happen in bacteria by direct contact and shuttling of DNA through a thin tube, for example. However, horizonal transfer is very rarely seen in animals.

Introns: DNA normally codes for proteins by "spelling out" amino acids. DNA that doesn't code for proteins don't spell out anything. When coding DNA (exons) is interrupted by non-coding DNA (don't worry, they get cut out later), this non-coding DNA is called an intron:

The basic results: The researchers found that, inside a certain type of rotifer, clusters of genes exist that don't look similar to genes in any known animal, but rather look very much like bacterial genes. For example, the Alr gene makes an enzyme involved in building bacterial cell walls, something animals don't do. (Most likely this Alr gene doesn't do much of anything in the rotifer, according to the article, but some of the other bacterial genes function just fine). So these genes must somehow have jumped from existing bacteria to the rotifers, somewhere back in history.

Even more surprising, some of these genes of bacterial origin have acquired something that bacterial genes don't normally have: introns. Animal genes have introns, meaning that what likely happened is that the bacterial genes jumped, then were slowly changed over evolutionary time (thousands or millions of years) to become animal-like (ie possessing introns).

What this means for evolution: Genetic diversity within and between different animals is both conservative (many of the same genes appear across different animals) and wide (many if not most animal types have genes that are specific to it). How does this diversity occur? It happens many ways, but the one in this article is one of the most surprising. It may be especially enhanced in the rotifer, according to the authors, since rotifers can survive being dried out and reconstituted, which may cause nearby bacteria to shed their DNA into the rotifer in between. Nevertheless, this kind of bacterial-animal horizontal gene transfer - now that we know it exists - could point to one of the ways in the farther past that some of today's genetic diversity could have come about. Or it may just be a weird rotifer thing. Time will tell.