Recently I started writing a post about selfish genes in yeasts, based on a very interesting new paper. In the course of reading the paper and thinking about yeast selfish genes, I discovered another fascinating story that I just had to write about. Hence part II of “How to live with a killer.”
The last post describes the “killer” system in yeasts and the devilishly clever way that the killer viruses force their hosts to retain them as unwelcome boarders. As you may know, viruses are little more (sometimes nothing more) than pieces of genetic material (DNA or RNA) that can move between cells and use those cells to make more copies of themselves. Some are famously deadly, others are almost completely benign. Some are even endogenous, which means they are “built in” to the host. The human genome, like most animal genomes, contains millions upon millions of remnants of viruses, many of them endogenous viruses.
One of the most interesting stories at the intersection of genetics and evolutionary biology is the detection and analysis of warfare between viruses and their hosts. By far the biggest new insight came with the revolutionary discovery of RNA interference. The whole story of that discovery is worth a read some other time, but its relevance to evolution is our focus here.
In brief, RNA interference (RNAi) is a cell biological phenomenon explained by the presence of an elaborate defence system in cells throughout the kingdoms of life. The system is common in single cell organisms such as yeasts, and ubiquitous in animals and plants. Its purpose is to detect small pieces of RNA (a tell-tale sign of viral infiltration) and destroy that RNA and its source. As is often seen in such intense evolutionary arms races, many viruses come with inhibitors of the RNAi system.
Viruses are not the only DNA/RNA-based parasites confronted by cells. Transposons (so-called “jumping genes”) are everywhere in animal genomes, and many/most are basically endogenous viruses. RNAi systems suppress these parasites. In animals, deletion of this defence apparatus unleashes transposons, with typically disastrous results.
This brings us to the mystery I mentioned before. Many yeasts lack RNAi systems, and for years it was thought that RNAi did not exist at all in some famous fungal families, most notably the baker’s yeast Saccharomyces cerevisiae and many of its relatives. (The yeasts in question are the budding yeasts.) But in 2009, David Bartel’s group reported that such yeasts sport a slightly unique RNAi system that functions quite well in suppressing transposons and other parasites. This brought the study of RNAi into the budding yeasts, a big breakthrough for scientists interested in dissecting the machinery. But there was one unanswered question.
In the 2009 paper, Bartel and colleagues showed that S. cerevisiae and some of its relatives did indeed lack a working RNAi system. While budding yeasts as a group had a working system, some subsets had apparently deleted it. The scientists could restore the system by inserting a couple of genes from the budding yeasts that do have the intact defence. But they couldn’t explain why so many of the budding yeasts had lost the system.
So, they added the RNAi system back to S. cerevisiae and asked whether the yeast could still live happy yeast lives. The findings were published in 2011. They looked at growth under lots of different conditions, and everything was fine. There was no indication that RNAi is bad for S. cerevisiae. But when they looked at the details of RNA molecules inside the cells, they saw a major and important change. The cells with RNAi systems were destroying one particular piece of RNA. It was killer virus RNA.
They then showed that the yeasts with RNAi restored had lost the killer trait. Look at the two culture plates pictured below (from Figure 1 of the 2011 paper), and focus on the orange blobs (those are yeast colonies). On the left, nearly every orange colony is surrounded by a “halo” of death. Who’s being killed? Another strain of yeast that is all over the plate. On the right, most of the orange colonies are peacefully coexisting with the other yeast. There are some halos of death, but they are far less common. The orange colonies on the right are yeasts with RNAi restored.
That observation makes sense out of just about everything known about RNAi in budding yeasts. These yeasts don’t tend to have as many of those annoying transposons to deal with, so the loss of RNAi protection isn’t an immediate death sentence. On the positive side, removal of the RNAi defence makes it much easier for the killer system to take up residence in the yeast. The killer virus, as I described in Part I, offers its host the advantage of being able to kill competitors, while requiring the host to keep the whole package (toxin and antidote). So, Bartel and colleagues concluded that the absence of RNAi in various families of budding yeast is explained by the benefit of hosting the killer virus. That benefit comes at the cost of living without the RNAi defence system, but in many of the budding yeasts, the transposon threat is less intense than in animals (for example), and so the deal is a good one.
That’s the basic story, but let me end by briefly discussing two really satisfying aspects of the findings. First, Bartel et al. note that the scatter-shot existence of RNAi in numerous yeast lineages had presented an evolutionary puzzle. Some families had it, but many didn’t, even though all were known to descend from a defined set of common ancestors. One possibility was that the groups with RNAi had all acquired it independently, perhaps through gene transfer events. But the killer virus connection provided a more coherent explanation: some families had jettisoned RNAi and acquired killer virus systems in return. At least by Occam’s razor, this is a cleaner explanation.
Second, Bartel and colleagues found that the deletion of the RNAi system, in each of the families that lack it, was relatively recent. They hypothesize that the benefit is a somewhat short-term one (evolutionarily speaking). Here is the last sentence from the 2011 paper, and a nice way to close the story:
All nine RNAi losses were relatively recent, suggesting that individuals that lost RNAi earlier left no living descendants. Thus, although compatibility with the killer system can explain the persistence of RNAi-deficient fungal lineages for many millions of years, lineages that lose this elegant transposon defense might be doomed to extinction over the longer term.
Image credit: Drinnenberg et al., cited below
Drinnenberg, I., Fink, G., & Bartel, D. (2011). Compatibility with Killer Explains the Rise of RNAi-Deficient Fungi Science, 333 (6049), 1592-1592 DOI: 10.1126/science.1209575
The idea of the “selfish gene,” made famous by Richard Dawkins nearly 40 years ago, was instantaneously controversial. It invigorated a sometimes-rancorous discussion of the focus of natural selection, with Dawkins and others arguing that the gene is the thing on which natural selection ultimately acts. The debate is ongoing and always entertaining. I think the real strength of the selfish gene concept is not so much that it acts as a bona fide theory of evolutionary mechanisms, but that it presents a truly useful view of how evolution often works. The gene’s eye-view that Dawkins adopts is particularly helpful when we encounter genes and interactions for which the word ‘selfish’ seems clearly apt.
Biology has described many interesting examples of genes behaving badly, acting in ways that admit no other explanation than that provided by the metaphor of a “selfish replicator.” Dawkins provides some nice examples in The Selfish Gene and in its superior but more technical sequel, The Extended Phenotype. One really interesting example comes from the world of yeasts, in a story that has provided an explanation for one of the biggest mysteries of yeast biology. Part I is about the new example, and Part II is about how a selfish replicator explained that big mystery.
Part I: AT-AT = destruction
The story begins with the discovery (decades ago) of some virus-like parasites that plague various groups of fungi (yeasts). These parasites are virus-like in that they are small bits of DNA or RNA that rely on hosts for survival and propagation, and they’re often called viruses even though they don’t usually make a vehicle to travel around in. The parasites in this story are called “killer elements” for good reason. They are pieces of DNA that enable the host cell to make a toxin that can kill neighbouring cells. The package deal includes a gene for making the toxin and a gene that makes the host cell immune to the toxin. A yeast cell without the killer element is susceptible to the toxin, whereas a cell carrying the element is able to kill without being harmed itself.
At first, this looks like a pretty good deal for the host—as long as its children carry the element, it can kill competitors indiscriminately. But it does come at a cost, since the host must make at least two proteins (the toxin and the immunity protein), and it’s easy to picture how the yeasts would cheat: they would grab the gene for immunity and insert it permanently into their genetic endowment. If it were that simple, we would expect the immunity gene to spread rapidly through a population and become permanent in the species (or population). But this is uncommon. Why?
Amazingly, some of these killer elements have made it effectively impossible for a yeast cell to cheat by permanently installing the immunity gene. A new paper in PLOS Genetics explains how it works, and an accompanying piece discusses how the logic of the system could be widespread in biology.
The authors (Kast et al., a collaborative group from two institutions in Germany) were working on a DNA-based killer element that lives in the cytoplasm (meaning outside the nucleus, and that’s important here) and is passed on to other cells by normal reproduction. The toxin and the immunity protein are fairly well understood. But there is something strange about the immunity gene. It works fine, and generates immunity, as long as it lives in the cytoplasm. (This makes it a plasmid, in technical terms.) But if the gene is moved into the nucleus (by an experimenter), it doesn’t work at all. The strategy (on the part of the killer element) is brilliant and selfish—cells can’t separate the immunity gene from the rest of the killer element, which can only exist as a package deal. But how does this work? After all, DNA is DNA. How does this particular gene suddenly go silent when it’s put into the place where all normal yeast genes live?
Kast and colleagues found that an immunity gene put into the nucleus was transcribed with no problem, generating a messenger RNA (mRNA) just the way a good gene should. But the mRNA was quickly destroyed. So, the gene was working fine but the mRNA was somehow tagged for destruction. This led to a very interesting hypothesis based on one peculiarity of the parasite’s genes: they are A/T rich. What this means is that they have an unusually high ratio of A-T base pairs compared to G-C base pairs. That bit of technical jargon isn’t by itself important. This is: lots of A’s and T’s in an mRNA molecule can trick a cell into destroying that mRNA. Why? Because the signal for “add a poly-A tail here” is made of strings of A’s and T’s. For our purposes, it hardly matters what a poly-A tail is. What matters is that this normal process (adding tails) can go haywire if someone sprinkles “add tail here” messages in places they don’t belong. That leads to mRNAs with lots of tails. And that leads to destruction.
Here’s the experiment the authors did to test their hypothesis. They surmised that the high A-T content explained the inability of cells to make the immunity protein from a gene in the nucleus. So, they made a synthetic version of the gene, one that they knew would still generate the same protein, but one with a lot fewer A-T base pairs. That version of the gene completely protected cells from the toxin. In the graph below (Figure 6B from the paper), you can see the results. Blue and green show what happens when the gene for immunity to toxin B is put into cells exposed to toxin A. It doesn’t matter whether the gene has fewer A-T base pairs (green); the antidote must match the toxin, so the cells are killed. The gene for immunity to toxin A is also ineffective at rescuing cells from toxin A (black), unless the gene has been scrubbed of excess A-T’s (red).
That’s really cool.
It’s brilliant, actually. The sneaky killer element offers the yeast cell a package deal: you can have the poison and the antidote, but you can’t have one without the other. If you have just the poison, you die. If you try to put the gene for the antidote into your genome, you discover that the gene has been wired for self-destruction if it finds itself in the wrong place (the nucleus). It’s like a device made by a Bond villain.
Stay tuned for Part II, in which we look at an extraordinary story of evolutionary conflict, in which some yeasts appear to have disarmed themselves in self-defence. Sounds weird? This is the world of selfish replicators, a world that only makes sense from the gene’s-eye view.
Image credit: Wikipedia
Kast, A., Voges, R., Schroth, M., Schaffrath, R., Klassen, R., & Meinhardt, F. (2015). Autoselection of Cytoplasmic Yeast Virus Like Elements Encoding Toxin/Antitoxin Systems Involves a Nuclear Barrier for Immunity Gene Expression PLOS Genetics, 11 (5) DOI: 10.1371/journal.pgen.1005005
Reconstruction of evolutionary trajectories will be a favourite topic on this blog, since it’s a very interesting area that is currently growing rapidly. I already wrote about nice new work showing how heat stability can evolve in thermophiles (heat-loving organisms). Now there’s a new paper looking at how blue color vision arose in the lineage leading to human beings. Both studies asked the same basic and important question: how did evolution get from A to B, given that most of the paths leading from A to B are dead ends? Read the rest of this entry »
In which I propose that creationist rubbish can be turned to gold (and discuss latest creationist “publication”)
Most creationist publishing is rubbish, taking the form of religious argumentation and shamelessly inaccurate propaganda. Creationist writers on both sides of the Atlantic are almost uniformly unworthy of any scientific or scholarly attention, and the only serious question among scientists is whether and how to engage creationist tactics to limit their damaging influence on science education and literacy.
It is quite unusual, then, to see a creationist group publish a scientific paper—containing experimental data—in a peer-reviewed journal. When this happens (about once every two years), it seems worthwhile to read the paper and see whether the creationists are at least doing valid experiments. Read the rest of this entry »
One well-known metaphor for the process of biological evolution is ‘tinkering.’ First proposed by François Jacob in 1977 in a now-famous paper in the journal Science, the idea captures two facets of evolution: the fact that new things must be developed from pre-existing things, and the apparent fact that evolution does not proceed with guidance. The picture is one of an actor mindlessly fiddling with implements, tossing them into the mix to see what happens. A believer might prefer the actor to be mindful, perhaps even goal-driven, but the process shows no evidence of this—Hume’s “stupid mechanic” seems a more apt metaphor to me. But mystical preferences aside, a view of evolution as tinkering brings a set of expectations or predictions to evolutionary thought.
One of those predictions is that innovations in evolution should be rare. More precisely, whenever something “new” appears, we expect it to be built from old stuff, from the components already there. No matter how innovative the new thing looks, we expect it to be a subtle reworking of whatever came before. Read the rest of this entry »
There are many aspects of biology that should inspire wonder in believers and non-believers alike. One of them, in my humble opinion, is the form of genetic recombination known as ‘crossing over.’ You can read a bit more about it elsewhere, but the gist is that in sexually-reproducing creatures, the formation of sperm and eggs involves a mixing-up of genetic material that ensures that the billions of individual sperm or eggs are genetically unique. The process involves the deliberate cutting of DNA, followed by a repair process that often stitches pieces together that weren’t together before. It is an amazing thing, almost audacious in its apparently deliberate drive to create genetic variation, via means that are elsewhere associated with chaos and catastrophe. (Such as cancer, to complete a nice alliterative triplet of horrors.) Read the rest of this entry »
One challenge in writing about basic evolutionary concepts is finding the balance between the creation of jargon-filled geek indulgences and the tedious rehashing of well-known facts and principles. Proteins are a case in point. Many readers will already know that proteins are the workhorses of life, nanomachines with finely-tuned functions and exquisite 3D structures to match. Many, I hope, already know that proteins are manufactured by reference to the famous genetic code, a universal DNA- and RNA-based system. Some perhaps know that proteins must be folded into those exquisite 3D structures, often with help from other proteins, in order to function. And one way to destroy a protein functionally, to cause it to unfold and perhaps even turn inside-out, is to heat it. We’re not talking about extreme heat (say, 500 degrees C as in a self-cleaning oven) that cooks everything down to carbon. We’re talking the kind of heat one finds in the kettle, or in bubbling hot springs in places like Yellowstone National Park in the US. Say, 95 degrees C.
Amazingly, there are life forms (aptly named ‘thermophiles’) that thrive in serious heat. The bacterium Thermus aquaticus is one of the best known—originally discovered in those Yellowstone hot springs, its DNA-copying protein made feasible decades of DNA copying in labs the world over, through a process called the polymerase chain reaction.
This is amazing because proteins in, well, normal organisms like us cannot tolerate 95 C. They unfold, and usually this is permanent. They don’t re-fold when the temperature goes back down. The process is called ‘denaturation.’ Cooking meat is a basic example of this. The heat denatures proteins, and essentially all of the denaturation is irreversible. The meat stays cooked. (Thank heaven. No one wants the haggis coming back to life.)
Somehow, then, the proteins in thermophiles like Thermus aquaticus have become super-stable. They refuse to unfold even when heated to near boiling. We should expect that natural selection, acting on the basic protein structure (the sequence of building blocks, which are amino acids), can hone the nanomachine into a heat-tolerating nanomachine. Read the rest of this entry »