(Being the third part of an occasional series about mutations.)

Here are two pieces of bizarre natural history:

One. The crab spider Thomisus onustus. Some individuals are a most unspiderly color: they are hot pink. This allows them to hide in hot pink flowers — and ambush unwary bees. Other individuals are yellow, and hide in yellow flowers. Even more surprising, their colors are not fixed: move a pink spider to a yellow flower, and she can change her color to match.

Two. In the center of the Atacama desert in northern Chile, it rains less than five millimeters (less than quarter of an inch) a year, yet there are bacteria living inside quartz pebbles. Like plants, these bacteria make energy from sunlight. (The quartz is translucent, so it lets in some light, but at the same time, shields the organisms from the worst of the ultraviolet radiation.)

I mention these oddities because this week, I want to resume my obsession with mutations, and look explicitly at a type that I’ve so far mentioned only in passing: beneficial mutations.

All mutations are accidental changes to DNA. Beneficial mutations are those accidental changes that in some way improve an organism’s chances of surviving and reproducing. Obvious examples from recent decades would be a mutation that confers antibiotic resistance on bacteria that cause disease in humans, or one that confers poison resistance on a rat. Less obvious are mutations that produce hot pink spiders and quartz-dwelling bacteria. But the point is that such mutations allow organisms to evolve to fit their environments better — a process known as adaptation.

Adaptation is the “wow!” factor of nature: when we see something spectacular or exquisite, we are typically looking at an adaptation. And what underpins adaptation is the appearance and spread of beneficial mutations: the process is not possible without them. Yet despite their central role in adaptive evolution, beneficial mutations have — until recently — received surprisingly little attention.

As a result, crucial questions remain open. For instance: is it the case that most beneficial mutations have small effects — and how often do beneficial mutations with large effects happen? And: can we predict which kinds of beneficial mutations will spread through the population by means of natural selection — and which will appear and almost immediately vanish?

The answers hold the keys to understanding how fast organisms can evolve to fit new environments — and thus to practical matters like the appearance of drug resistance in hospitals, as well as a more general view of how likely different populations are to go extinct when the environment changes.

There are several reasons for this neglect of the benign. One — dare I say it — is fashion. In the late 1960s, the geneticist Motoo Kimura proposed the neutral theory of molecular evolution. According to this idea, most mutations are either harmful (and will quickly disappear from the population because those bearing them die) or irrelevant. If this is the case, most genetic variation has no impact on fitness — the technical term for how good an organism is at surviving and reproducing. Kimura’s development of the neutral theory was enormously influential, and prompted a flurry of work investigating whether most genetic variation is irrelevant.

Then it was the turn of deleterious mutations, which became trendy in the late 1980s and early 1990s. Deleterious mutations have been hypothesized to play a central role in a variety of evolutionary phenomena, including (and most prominently) sex. The argument is that organisms with a deleterious mutation rate above a certain threshold must reproduce sexually.

The reason is that sex purges deleterious mutations from the population: sex generates new gene combinations, and thus in each generation it creates some individuals with relatively few deleterious mutations and some with lots. Those with lots die; those with relatively few go on to have large numbers of children. Asexuals, which reproduce by cloning (there are many ways to do this, including laying an egg that doesn’t need to be fertilized, or budding off a piece of yourself), have no way to get rid of deleterious mutations. If the deleterious mutation rate is too high, asexuals will quickly go extinct.

Deleterious mutations are much easier to study than beneficial mutations. First, they are more common. (As I have remarked elsewhere, insofar as a mutation has any effect, the effect is more likely to be harmful than helpful, for the obvious reason that an accidental change is more likely to be a disruption than an improvement.)

Second, whether a mutation is beneficial always — but always — depends on the environment it happens in. If there aren’t any antibiotics around, for example, a mutation that suddenly confers antibiotic resistance isn’t useful, and may even carry a cost. Many harmful mutations, on the other hand, are harmful under any circumstances, such as those that kill you before you are born or make you sterile so you can’t reproduce.

But the times, they are a-changing again. As the genome of one organism after another has been fully sequenced, it’s become clear that, to a much greater extent than anyone expected, many of the mutations that have become established in different species are not neutral, let alone deleterious, but beneficial — adaptive. (Different types of mutation leave different genetic signatures, so analysis of genomes can reveal much about which sorts of mutations have become established in the past.)

And so the study of beneficial mutations has come to the fore. The past few years have seen a burst of mathematical and experimental work designed to get at the questions I mentioned above.

Different generations of Pseudonomas fluorescens. (Photo: Paul B. Rainey/University of Auckland)

For experiments of this kind, bacteria are close to being ideal. For one thing, they grow fast. Bacteria of the species Pseudomonas fluorescens — a beastie that usually lives in the soil — can go from a population of just 500 individuals to 200 million over the course of one night. And because bacterial population sizes are so huge, the odds that some individuals will carry beneficial mutations are reasonably high. It’s like a national lottery: the odds of any particular individual winning are small, but if millions play, the odds are high that somebody will win something.

Another advantage: it’s easy to move bacteria from a comfortable environment, where they have all their nutrients provided, into a tougher environment — one where there’s an antibiotic, or where a crucial nutrient is missing — where most of them will die, and only those with a new beneficial mutation can survive. The individuals with the beneficial mutations can then be collected and grown in different environments to see how they do.

For instance, how fit are the mutants that survived in the new, tough environment? Are most of them just scraping by, or are they flourishing? And do they grow better or worse in the original, comfortable environment than ordinary, unmutated bacteria do? In other words, do the mutations that improve performance in the tough environment come with a fitness cost in the relaxed one, and if so, how large a cost is it? (In experiments like this, fitness is easy to measure: it is simply how fast bacteria with different mutations grow in a given environment. Measuring fitness in natural populations, in contrast, is exceedingly difficult.)

Experiments of this general type have produced several important results. First, of the beneficial mutations that appear, most have small effects on fitness. This is not surprising: the bigger the effect a mutation has, the bigger the chance that it will be something disruptive, so mutations that have dramatic effects are likelier to be harmful than mutations that produce modest changes. All the same, beneficial mutations with large effects do appear, just much, much less frequently.

However, of the mutations that appear, those with the smallest effects are also the least likely to spread through a population and become established, simply because they don’t confer much of an advantage on their owners. Good mutations with larger effects have a correspondingly larger advantage — and thus their owners are more likely to leave more descendants, and so the mutation will spread more readily. In other words, although mutations of small effect are more common, the actual establishment of adaptive changes will often be due to mutations of medium to large effect, especially on arrival in a tough new place.

Finally, mutations that improve the fit to one environment may dis-improve the fit to another: specializing may come with a cost. Or at least, that’s the result suggested by one set of experiments. Most of the mutations that enabled bacteria to survive in the presence of an antibiotic made them less fit than ordinary bacteria when they were returned to the comfortable environment. But a tiny fraction of the new mutants were super-mutants: their owners fared well in both situations. So far, the details of the genetic basis for this are not understood, nor is it clear whether such individuals would do well in, say, 10 different environments. Nonetheless, the result may have important implications for the development of antibiotic resistance, and the making of bugs and super-bugs.

Together, the study of these different mutational strands — the good, the bad and the irrelevant — is weaving a comprehensive picture of the general distribution of mutations, and thus, the spectrum of genetic variation that appears in nature. When this is complete, we will have a more complete view of how mutations of different types impact different paths and patterns in evolution.

Think of that the next time you meet a pink spider.

**********

NOTES:

For pink spiders (Thomisus onustus), see the plate facing p. 49 of Bristowe, W. S. 1958. “The World of Spiders.” Collins. The ability of this species to change color was confirmed to me by Dr. Marc Théry in an email. For bacteria living inside quartz stones in the Atacama desert, see Warren-Rhodes, K. A. et al. 2006. “Hypolithic cyanobacteria, dry limit of photosynthesis, and microbial ecology in the hyperarid Atacama desert.” Microbial Ecology 52: 389-398.

For a concise description of the process of adaptation and the role of beneficial mutations in it, see Orr, H. A. 2003. “The distribution of fitness effects among beneficial mutations.” Genetics 163: 1519-1526. This paper also discusses some of the open questions, such as how much more common beneficial mutations of small effect are than those of large effect. For a history of the mathematical approaches to beneficial mutations, see Orr, H. A. 2005. “The genetic theory of adaptation: a brief history.” Nature Reviews Genetics 6: 119-127.

For the first statement of the neutral theory, see Kimura, M. 1968. “Evolutionary rate at the molecular level.” Nature 217: 624-626. For a brief review of evolutionary phenomena that may be driven by the deleterious mutation rate, see Kondrashov, A. S. 2001. “Sex and U.” Trends in Genetics 17: 75-77. For deleterious mutation rates and the evolution of sex see, for example, Kondrashov, A. S. 1988. “Deleterious mutations and the evolution of sexual reproduction.” Nature 336: 435-440. (Note that there are two hypotheses for how deleterious mutations can cause the extinction of asexuals, Muller’s ratchet and Kondrashov’s hatchet; here, I am describing Kondrashov’s hatchet. Unlike the ratchet, the hatchet does not depend on population size.)

For a discussion of the extent to which the establishment of beneficial mutations accounts for genetic differences between species, see Eyre-Walker, A. 2006. “The genomic rate of adaptive evolution.” Trends in Ecology and Evolution 21: 569-575.

For an experimental investigation of the raw distribution of fitness effects of beneficial mutations in Pseudomonas fluorescens, see Kassen, R. and Bataillon, T. 2006. “Distribution of fitness effects among beneficial mutations before selection in experimental populations of bacteria.” Nature Genetics 38: 484-488. This paper also reports the discovery of a handful of super-mutants. For experiments showing that mutations of medium to large effects are the mutations that actually become established, see Barrett, R. D. H., MacLean, R. C. and Bell, G. 2006. “Mutations of intermediate effect are responsible for adaptation in evolving Pseudomonas fluorescens populations.” Biology Letters 2: 236-238; see also, Bull, J. J., Badgett, M. R. and Wichman, H. A. 2000. “Big-benefit mutations in a bacteriophage inhibited with heat.” Molecular Biology and Evolution 17: 942-950.

Many thanks to Thomas Lenormand and Sally Otto for general insights on beneficial mutations, to Marc Théry for supplying information on color-changing spiders, and to Thomas Bataillon, Nicholas Judson, Gideon Lichfield, Dmitri Petrov and Jonathan Swire for detailed comments and suggestions.