Register for an account

X

Enter your name and email address below.

Your email address is used to log in and will not be shared or sold. Read our privacy policy.

X

Website access code

Enter your access code into the form field below.

If you are a Zinio, Nook, Kindle, Apple, or Google Play subscriber, you can enter your website access code to gain subscriber access. Your website access code is located in the upper right corner of the Table of Contents page of your digital edition.

Health

When good mutations swarm?

Gene ExpressionBy Razib KhanMarch 9, 2009 12:00 PM

Newsletter

Sign up for our email newsletter for the latest science news

Evolution is a fact. Lineages change over time, and respond to selection pressures as well as being buffeted by stochastic processes. The arguments about the particular details of the process of evolution can be very vociferous. Scientists are human too, and great evolutionary biologists such as R. A. Fisher stooped to venomous insult when backed into a corner (see his disputes with Sewall Wright during the 1930s). But the superstructure of human foibles and follies rests upon a foundation of genuine scientific dispute, and attempts to refine models which map onto reality. For example there have been long been debates with evolutionary biology about the nature of mutation and adaptation, and a combination of empirical and theoretical work has been very fruitful in elucidating general phenomena. Some of the early disputes between R. A. Fisher and Sewall Wright hinged upon their different views when it came to the role of independent genes of small effect in driving evolution (Fisher tended to give this dynamic more prominence than Wright). Later during the 1960s the Neutral Theory of molecular evolution arose in large part to address data which were poorly explained by the then ascendant models. A few years ago I pointed to recent work which lean toward the proposition that mutations of large effect may be far rarer than mutations of small effect, but that the former are responsible for most evolutionary change because of their larger weights. These are not trivial debates. For evolutionary science to move beyond a purely descriptive exercise the synthesis of model and data must lead to robust theory. Of course easier said than done. One of the pioneers of evolutionary genetics famously observed that "fitness is a bugger!" An interesting paper just came out which I believe sheds light on some of the questions above, even if it is but a sliver of what needs to be known. The Distribution of Fitness Effects of Beneficial Mutations in Pseudomonas aeruginosa:

Adaptation by natural selection depends on the spread of novel beneficial mutations, and one of the most important challenges in our understanding of adaptation is to be able to predict how beneficial mutations impact fitness. Here, we investigate the underlying distribution of fitness effects of beneficial mutations that natural selection acts on during the evolution of antibiotic resistance in the opportunistic human pathogen P. aeruginosa. When the fitness of the wild type is high, most beneficial mutations have small effects. This finding is consistent with existing population genetic models of adaptation based on statistical theory. When the fitness of the wild type is low, most beneficial mutations have large effects. This distribution cannot be explained by population genetic theory, but it can be readily understood by considering the biochemical basis of resistance. This study confirms an important prediction of population genetic theory, and it highlights the need to integrate statistical and biochemical approaches to adaptation in order to understand evolution in stressful environments, such as those provided by antibiotics.

When the fitness of the "wild type" was high, in this experiment that meant that the concentration of the antibiotic was low, most beneficial mutations had small effects and they followed an exponential distribution. This is because out of the set of mutations the beneficial number were being sampled from an extreme tail of the distribution. From the paper:

Adaptation by natural selection ultimately depends on the spread of novel beneficial mutations that increase fitness. Can we predict the fitness effects of beneficial mutations? Gillespie...argued that extreme value theory (EVT) provides a simple answer to this question: the tails of all-Gumbel type distributions (a very flexible type of distribution that includes many familiar distributions, including the normal) are exponential. As such, the fitness effects of beneficial mutations will be exponentially distributed provided that the fitness of the wild-type is high enough so that beneficial mutations are drawn from the extreme tail of the distribution of fitness effects of mutations. It is however unclear how robust this theory is with respect to the fitness of the population prior to selection. As the absolute fitness of the wild-type decreases (for example, because of environmental change), a larger proportion of single mutations will increase fitness. Beneficial mutations will therefore no longer be drawn exclusively from the tail of the distribution, hence the exponential distribution may no longer apply....

When they increased antibiotic concentration the wild type fitness naturally decreased, and the number of beneficial mutations increased. Beneficial being those mutations which increased fitness. You normally assume that most mutations are deleterious; after all, it's easier to tear down than build up. But in an environment which is extremely difficult for the notionally optimal wild type sometimes the tear down is preferable to what you've got. These mutations probably entail costs because of pleiotropy. You modify gene A, and induce changes on genetic pathways 1, 2, 3 ... n. In an environment where the population is reasonably well adapted small changes will result in a shift away from the fitness peak, and selection will then purge these deleterious mutants. In an environment where the population is not well adapted even random walk changes through the fitness landscape may result in positive increments of fitness. Imagine for example that population A is at a fitness peak, but at some point the environment shifts to the point where the fitness landscape warps underneath them. Now the population finds itself at the bottom of a fitness valley. But there is a physiological explanation in this particular experiment at high antibiotic concentrations:

The large effect of beneficial mutations on resistance is consistent with the molecular interactions that occur between rifampicin and RNA polymerase. Structural studies have shown that rifampicin binds to a small, highly conserved pocket of the β-subunit of RNAP and only 12 amino acid residues are involved in direct interactions with rifampicin...Mutations at these residues cause a large increase in resistance...Residues that surround the binding pocket interact only indirectly with rifampicin, and it has been argued that resistance arises at these residues due to amino acid changes that alter the folding of the protein in the binding pocket. We identified only a small number of beneficial mutations...in residues that are involved in indirect Rif-RNAP interactions and mutations at these residues give rise to intermediate levels of rifampicin resistance...This biophysical approach to understanding the effects of beneficial mutations suggests that the data may no longer fit an exponential distribution because of the high specificity of interactions between rifampicin and RNA polymerase: changes to the majority of amino acids that are involved in rifampicin-RNAP interactions results in large increases in resistance and, therefore, large increases in fitness at high concentrations of rifampicin....

This sort of juxtaposition between population genetic theory and physiology takes me back to the Wright-Fisher Controversy over dominance. R. A. Fisher was a theoretical evolutionary biologist, while Sewall Wright was by training an applied geneticist. Their different perspectives likely seeped into their prior commitments. But in today's world where evolutionary genomics offers up a mash-up between math, computation and automated sequencing hopefully there'll be less of this interdisciplinary tension. Nature is One. No need to play population genetics against biophysics.

    2 Free Articles Left

    Want it all? Get unlimited access when you subscribe.

    Subscribe

    Already a subscriber? Register or Log In

    Want unlimited access?

    Subscribe today and save 70%

    Subscribe

    Already a subscriber? Register or Log In