Adaptation is fast and effective in a fungus

The introduction
Watching adaptation - the process of organisms and populations increasing their fit to the environment - is not easily observed in nature, and when it is, it is often hard to tell what the details of the process are. Crucially, the mutations (and by "mutation" I mean any change to the genome) are mostly unknown, leaving us to guess what kind of genomic changes are responsible for shaping life through evolution.

Fortunately, some organisms are amenable to manipulation and fast evolutionary change: unicellular organisms and viruses have been extensively used with great success to investigate many questions of evolutionary import (though I hurry to caution that such organisms cannot say much about evo-devo and morphological change, as these are areas pertaining to multicellular organisms).

Bacteria (e.g. E. coli), viruses (e.g. φ6), fungi (e.g. yeast), algae (e.g. Chlorella) enlighten us on the mutational basis for evolution, as many experiments have recently shown.

This time a study using the fungus Aspergillus nidulans is manipulated in the lab to attempt to answer an important question in evolutionary theory: During adaptation, how many mutations are needed, and what are their size, i.e. how much do they affect the fitness of the organism?

The short answer is that adaptation is short and fast, requiring only a few mutations. Among those mutations, the first is of greater effect on fitness than the following, which is expected from theory when the organisms are climbing a peak in the fitness landscape, but contrary to the "gradualist view of adaptation dominant since the 1930s."

The experiment
One strain of Aspergillus nidulans is resistant to a fungicide, but resistance is known to be a costly trait, meaning that compared to a fungicide sensitive strain, it grows slower (has lower fitness) when there is no fungicide present. The crazy scientists then let populations of this resistant strain evolve without fungicide, and watch them adapt to the new fungicide-free environment.


Click to go to original figure.

Schoustra et al. placed a few spores (more than 10,000) on a medium (food), and let it grow for five days (about 80 generations) during which the population grew by mitosis (cellular duplication, similar to cell-division of somatic cells in the human body, but not of the gametes (sperm/egg), which is meiosis). After five days either 50,000 or 500 cells were transferred to a fresh medium (designated large and small bottlenecks, respectively). This process was repeated until the populations had gone through 800 generations. Fitness was estimated by measuring the diameter of the colony at generation 0, 80, 160, 240, 320, 480, 640, and 800. These experiments were done 120 times each.

Click to go to original figure.

The results
A fitness of 1 was given to the ancestral strain, and fitness of evolved strains was measured relative to that. In figure 2 the fitness trajectories are plotted as a function of the number of generations into the experiment for large bottlenecks (A) and small bottlenecks (B). Notice how most of the lineages quickly increase their fitness for both large and small bottlenecks.


                      Figure 4. Properties of adaptive walks.
Click to go to original figure.

Without going into the rather complicated details of how the number and effect size of mutations were estimated (it involves a Maximum Likelihood model and comparisons with simulations), we can see from figure 4 that the effect of the first three mutations are not different in the large (triangles) and small (squares) bottleneck lineages, and that the slopes (which are significantly different from zero) are negative, indicating that the first mutations are of the largest effect, followed by the second, and - for the large bottlenecks only - then by the third. (The selection coefficient is a measure of the effect of a mutation, and is given by the fitness of the mutated genotype minus the fitness of the unmutated parent divided by the fitness of the unmutated parent.)

Additionally, the number of mutational steps taken within the 800 generations is very small: for the 50,000 spores bottlenecks the average is 2.39±0.53 standard deviations, and for the 500 spores bottlenecks the average is 2.00±0.74 steps.

The discussion
These results indicate, according to Schoustra et al., that we need to change our expectation of how evolution proceeds:

Taken together, these results imply that the gradualist view of evolution is incorrect; rather, the bulk of adaptation in mutation-limited populations is likely to be achieved by the first few mutational steps.

The authors also found significant differences in genetic variation of the 800 generations evolved lineages, and from this they conclude that

although we do not know in detail the identity of the molecular changes responsible for fitness increases, we can be confident that they were achieved through a variety of genetic routes.

One pleasant surprise is that the different lineages clearly did not evolve the same way. We could have expected that they would all re-evolve sensitivity to the fungicide, thereby increasing their fitness. Rather, from figure 2 we can see that fitness was increased in many different ways, and with differing results. What this indicates is that the fitness landscape in the neighborhood of the ancestral genotype is very rugged (also called epistatic), because many different fitness peaks were climbed. If they hadn't climbed different peaks, we would expect the fitness measurements to be very similar. Instead, there are many different ways to increase fitness, and further, only some include re-evolving fungicide sensitivity (only five lineages re-evolved sensitivity).

I say this is a pleasant surprise because my own work with simulations suggest that fitness landscapes - which is a property of both genotype and environment - are likely to evolve to become epistatic. However, it is bad news for fighting resistant strains in general, because if the strains occupy a rugged area of the fitness landscape, they can increase their growth rate without losing their immunity to the fungicide/antibiotic/other drug.

The reference
Schoustra SE, Bataillon T, Gifford DR, & Kassen R (2009). The properties of adaptive walks in evolving populations of fungus. PLoS biology, 7 (11) PMID:19956798