Early in the history of life, when microscopic single cells were the only game in town, genetic material may have been passed around as easily as baseball cards. This gene swapping, known as horizontal gene transfer, is common among bacteria today, and it may have been occurring before the three domains of life developed.
|Tree of life, divided between major cell types, those with a nucleus (eukaryotes) and without a nucleus (prokaryotes: the bacteria and archaea).
The last universal common ancestor (LUCA) of the bacteria, archaea, and eukaryote domains is estimated to have lived sometime between 3.6 to 4.1 billion years ago, but horizontal gene transfer makes tracing the path of life that far back in time difficult. For example, a phylogenetic tree - a chart that shows evolutionary relationships among species -- could indicate that two unrelated bacteria are closely related if they happened to exchange a gene at some point in their history.
Carl Woese, the microbiologist who used genetics to define the three domains of life, has proposed that the LUCA was actually a large gene pool, a community of organisms with widely shared genetic material. Anthony Poole, a molecular biologist at Stockholm University in Sweden, disagrees with this idea. Instead, Poole thinks horizontal gene transfer may have come into play later in the history of life. In this interview with Astrobiology Magazine's Leslie Mullen, he explains why he thinks horizontal gene transfer took time to develop, and what that means for our understanding of early life.
|Horizontal gene transfer has been a driving force in evolution. All three domains have shared genetic material.
Image Credit: Doolittle, 1999.
Astrobiology Magazine (AM):
Carl Woese said that for the earliest life, horizontal gene transfer was so common that we'll never be able to distinguish individual traits back to the origin of life. But you disagree with this idea?
Anthony Poole (AP): His model essentially is that the rates of horizontal gene transfer very early in the evolution of life were so high that it dominated over normal processes like cell division, or the vertical inheritance from parent to offspring. In his model, the norm would have been high rates of horizontal gene transfer, but there is no evidence whatsoever for or against that theory.
From extremely early in life's history, we can see clear examples of horizontal transfers between the bacteria, archaea, and eukaryotes -- the three domains of life. If, for instance, we find a gene in an archaeon that is widespread in bacteria but otherwise undocumented in archaea, further analysis may indicate that the archaeon acquired that gene from a bacterium through horizontal gene transfer.
While we do see evidence for plenty of gene transfer between the domains, when we get farther back in time, it gets hard to argue that there's been ancient transfers. When we get back to the last universal common ancestor before the three domains separated, we don't have any hard data. Everything is tied together when you get down to the trunk of the tree.
AM:Woese said there was such a free flow of genes back then that everybody was essentially the same organism, a mega-organism.
AP: Yes, he describes a communal ancestor. It was one big gene pool, with everybody swapping genes freely. The problem with that is you will get cheaters emerging -- parasites that are not necessarily contributing to the good of the community. So if there was a completely open system with a free flow of genes, it should suffer from parasite invasion. Of course it's important to test that, because you may find solutions that enable the system to still operate.
|Carl Woese defined the archaea domain of life in 1977 by phylogenetic taxonomy of 16S ribosomal RNA. Woese also was the originator of the RNA world hypothesis, which says that early in the history of life RNA may have once performed functions that today are shared by DNA and proteins.
Credit: University of Illinois at Urbana-Champaign
AM:Wouldn't evolution take care of that? Parasites that are beneficial or not too harmful would survive while those that are destructive would result in the downfall of both the organism and the parasite.
AP: One solution to preventing the crash of an altruistic system due to parasites is to stick it on a two-dimensional surface. But that changes the model. It means there's no longer complete mixing. It provides a structure to the population, where the parasites are not free to move throughout the system. As soon as you have that, you no longer have the model that Woese described.
AM: You're arguing that because an open horizontal gene transfer system is inherently susceptible to parasites, an open system could not last.
AP: Horizontal gene transfer still could be present. It just could not completely dominate in the way Carl Woese has argued.
One of the key problems is: at what level is selection operating? If you have a communal system with genes moving around between individuals freely, is the level of selection on that super-organism, on the cells, or on the genes?
If you look at what happens in horizontal gene transfer in bacteria and archaea, we see indications that it isn't just free transfer. Instead there is a strategy. You have selection on cells -- they're throwing away genes that are not immediately in use. There may be an advantage in keeping the size of the genome small, so carrying around genes that you're not using would be a disadvantage. Either those individuals that have larger genomes will be out-competed, or those genes will be lost from that individual and that individual will live to fight another day.
So there's a selection at the level of the cell that there's an advantage to lose genes that are not being used. There's also selection at the level of the genes. If they're frequently lost, they could go extinct because it may be a one-way street where once you're lost you can never be regained. But genes that are able to develop mechanisms for transfer can find a new host and survive.
Gene transfer could be a strategy that developed for genes to survive in an environment where genes tend to get lost. I think that's why we see horizontal transfer in organisms like bacteria, which are much more affected by their environment. You can find closely related bacteria in extremely different environments, and their adaptability has partly to do with that.
|Archaea are simple cells that don’t have nucei. They resemble bacteria in their cell structure, but the genes of archaea and several of their metabolic pathways are more closely related to those of eukaryotes. Credit: MIT
AM: Horizontal gene transfer may have developed over time as a survival strategy for genes?
AP: Yes. It's difficult for a gene to transfer; it requires a mechanism. It needs to get out of the cell and into another cell. As a biologist, I find it difficult to swallow that gene transfer would be a default state. Viruses can transfer genes, but they need to infect another individual. All of the mechanisms to accomplish that are evolved. We don't know of any default processes by which DNA diffuses out of one cell and into another. That's a type of biology we're unaware of.
AM: Could horizontal gene transfer have evolved from a less complex version?
AP: There are fairly simple mechanisms of transfer. It doesn't require a lot of genes. It could have occurred early, and at significant levels. If there was transfer going on, what does that tell us about the system?
I think what horizontal gene transfer tells us about the system is that there are a combination of selection pressures on genes that are largely cooperating in the cell. Selection on genes may occasionally knock a gene out, and when that happens the only genes that will continue to survive will be those that have a mechanism of transfer. That mechanism could be ridiculously simple, it may not require more than a handful of genes, but it still requires some genes. So given that it requires some mechanism, it could have evolved early owing to its simplicity, but it could not be a default state.
AM: If your hypothesis is correct, how does that change our idea of what early life was like?
AP: It doesn't -- I'm just saying let's be cautious with our speculation about what early life was like. Let's look at the biology of transfer before we use it as a quick remedy for the difficulties in understanding early evolution. Can we explain gene transfer with the tools we have as biologists? My feeling is it's a process that's emerged like any other process, so it's not special in that sense. I think it's a cop out to use the shifting around of genes as a quick way out of dealing with the problems associated with reconstructing very early stages of life.
|A eukaryotic cell is an extremely complex structure.
AM: There are 50 or 60 genes that the three domains of life have in common, and those genes all are related to protein expression. Do you think that is one way of getting a better concept of the last universal common ancestor?
AP: Those 50 or 60 genes are just doing basic things that you find in all cells, like making RNA or protein. They don't do anything that can tell us about the environment of early life, since all cells carry them.
AM: Unless, for instance, the genes couldn't possibly survive in higher temperatures. That would tell you something about the early environment for life.
AP: Yes, but the data doesn't indicate anything like that. We can just say life had machinery to make RNA, it had machinery to make proteins, and maybe a couple of other bits and pieces. We can't say anything about whether life was in this particular temperature, or whether it operated with a particular kind of chemistry. This core of genes just tells us aspects that we already knew: that all cells make proteins. Don't get me wrong, it does give us some useful information. For example, DNA isn't there. It looks like DNA may have evolved more than once, but that's another story.
People have been working on this issue for many years. Initially there was hope you could compare the genes among a number of different genomes, and across the domains of life you would have some core that would be sufficient to run a cell. Then you could perhaps reconstruct that cell, and that would give us some idea of what the earliest cells looked like. That didn't work. What we do see is a clear picture of some of the most conserved components, and among those are things like the sets of genes required for protein synthesis for producing the ribosome, and RNA polymerases that are required for copying RNA.
AM: So in that sense, it doesn't matter who came up with a gene first, because it was so widely shared.
|A bacterial cell is not as complex as a eukaryotic cell, but it still has some very intricate systems and capabilities.
AP: Right. Certain genes do seem to spread very rapidly. But it's a question of the timing. There's no requirement for an early innovation to be shared through horizontal gene transfer. An equally reasonable scenario says one individual produces this great machinery, and it just wipes the floor with the competition.
Another scenario is an innovation turns up very late in evolution, and there is a propensity to transfer it to effectively everybody. If everybody's got it, then we won't necessarily see that it evolved in one lineage and then spread to the others.
AM: In the first scenario the innovator is the last man standing, and all the descendents from then on are like the innovator. In the other scenario, the innovator shares the innovation and everyone changes to become like that organism from that point on. And there's no way to tell which scenario actually happened.
AP: An evolutionary tree just tells us where diversification occurred over time. A tree composed of 1,000 related organisms will always taper down to a primary bifurcation -- one splitting into two. It's the same as when people talk about one Adam and one Eve in human evolution. We can trace genealogies back to a single individual, but it doesn't mean they were alone in the universe, as it were. There were whole populations. Over time individuals don't leave offspring so they disappear from the population. It can take a long time or it can be a short sweep, but the outcome will be that you can trace back to a single individual. That individual didn't necessarily have any special qualities. It's the same with the common ancestor for all archaea. In principle, they all trace back to a single cell, but that doesn't mean they were there in isolation, and it doesn't mean there was anything special about them.