This week’s BEACON Researchers at Work blog post is by Michigan State University postdoc Jeff Morris.
We could probably agree that humans are a lot more complicated than bacteria. At a first glance, we have more moving parts, lots of different encapsulated regions (organs) that do different things, and then there’s that wonderful casserole-sized lump of computer sitting in our heads. If we take a closer look, we find that our genomes are also monstrously more complex. In comparison to bacteria, our chromosomes are full of genetic spam, our genes are regulated by convoluted networks of interacting proteins, and in between transcription and translation there are a battery of modifications that must take place, not to mention active transport of RNA across the nuclear membrane.
So what does all this complexity do for us? One might point to our large, dramatic bodies; but in reality single-celled eukaryotes have genomes just as complex as humans, and sometimes more so. Also, our metabolisms aren’t that much different than bacteria. In fact, bacteria can do lots of reactions no eukaryote can do. Surely, then, our complexity gives us some profound fitness advantage. But the bacteria still comprise the huge majority of Earth’s biomass. They grow faster, are more versatile, more resistant to virtually all stresses, and can occupy environments off-limits to almost all eukaryotes. In this light, our complexity makes us look like Rube Goldberg devices – impressive but inefficient machines comprised of countless bells and whistles of dubious utility.
In his 1995 book Full House, Stephen Jay Gould proposed that complexity should accumulate over the course of evolution by a purely neutral “drunkard’s walk” mechanism if species were just as likely to become more simple as more complex with any given mutational event. Because there is a “wall” of minimum complexity at one end of the spectrum, one would expect to see simple organisms as the dominant “mode” of life, with a long “tail” of increasingly complex (and rare) organisms that gets longer as life goes on (Fig. 1). Of course, this is indeed what one sees; there are orders of magnitude more bacteria than eukaryotes in the world.
Despite Gould’s neutral hypothesis, evolutionary biologists seem determined to prove that natural selection underlies the expansion of complexity in the living world. There have been some very compelling studies by BEACON scientists to this end (for instance, this classic by a 4-piece BEACONite supergroup), but comparatively few that look at selective pressures for simplification. A major focus of my research is on the other side of the coin. We’re all aware of cases of reductive evolution, where simple organisms have more complex ancestors – think tapeworms, which have no digestive tracts. Indeed, reductive evolution is a hallmark of the parasitic lifestyle, driven by the low effective population sizes of parasites and the fact that they are largely isolated from gene flow that could restore lost or damaged genes by recombination. These processes are examples of genetic drift, a random evolutionary process similar to Gould’s drunkard’s walk.
But natural selection can also simplify organisms, and not just parasites. Consider Prochlorococcus (Fig. 2), the world’s most abundant photosynthetic organism. These tiny cyanobacteria are the dominant “algae” in much of the ocean. At ~100,000,000 cells per liter of seawater they are far too populous to be subject to drift. And yet the genomes of Prochlorococcus are roughly half the size of their closest relatives and are missing many important pathways. Some of these missing genes actually make Prochlorococcus dependent on its neighbors for vital functions. For instance, compared to almost all other aerobic bacteria (and most of its cyanobacterial relatives), Prochlorococcus has very little ability to protect itself from hydrogen peroxide. Yet this toxin is continuously produced by sunlight acting on seawater, and would kill off Prochlorococcus if it were living by itself. Fortunately, there are many other bacteria in the ocean that are able to destroy this toxin, and because it is freely membrane-permeable (i.e., it’s a “leaky” function), when they protect themselves they also protect Prochlorococcus and all their other neighbors.
Metabolic pathways like peroxide degradation are expensive, and resources are scarce in the open ocean where Prochlorococcus lives. It is thus reasonable to suspect that the loss of this function gave Prochlorococcus’ ancestor a fitness advantage – but only if its neighbors kept the function. Notably, other common bacteria in the ocean also don’t have the genes necessary to break down peroxide, and these cells also have reduced genomes. Thus, we suspected that there was some common evolutionary process at work here, actively eliminating these genes and making these cells dependent on their neighbors.
We proposed the “Black Queen Hypothesis” (BQH) to explain this phenomenon. The name comes from the card game “Hearts,” where the object is to get the lowest possible score. The queen of spades, however, is worth as much as all the other cards combined, so a primary strategy is to not win her. We propose that there are functions in nature that are like this black queen – they are very costly, so you get a benefit if you aren’t holding her, but nevertheless somebody has to get stuck with her or else you wouldn’t be playing Hearts. For instance, imagine an ocean where no one degraded peroxide – game over, everybody dies. But it doesn’t take everybody working to make the environment safe, so cells can evolve to stop breaking down peroxide up to the point where any further function loss would lead to a level of peroxide that would offset the advantage of the gene loss. This is an example of what evolutionary biologists call negative frequency dependence, and it is striking because it allows the co-existence of very similar organisms. We also suspect that BQH evolution isn’t limited to peroxide and oceanic bacteria, but can be extended to any function that is both leaky and costly.
Nothing in the BQH is truly novel. It’s trivially obvious that complexity can be problematic. For instance, anyone who has written computer code knows that unnecessary lines mean extra clock cycles and slow, sloppy programs. But there hasn’t been much serious systematic thought about what the consequences are for natural selection on genomic and/or functional reduction. With the BQH we show how simple, selfish evolution toward genomic efficiency can create complicated ecological interactions. It can even explain the evolution of mutualistic interactions in well-mixed environments, which was thought to be a virtual impossibility by theorists.
For more information about Jeff’s work, contact him at jmorris at msu dot edu.