Lin Chao
e-mail: lchao@biomail.ucsd.edu |
 |
We use a combination of computer modeling and laboratory experimentation to study ecological and evolutionary processes. Ecological and evolutionary studies are often difficult to investigate because they occur slowly over long periods of time. Computer models can speed up time, but we also take advantage of the fact that microbes, bacteria and viruses provide that same advantage because of their short generation time. Furthermore, they can be easily cultured in the laboratory under controlled conditions. However, we also realize that too much control can be dangerous because an experiment that is overly-controlled becomes artificial and not much more than a simulation. And a simulation is always more easily obtained from a computer model. Thus, we are careful to design experiments that teach us more than a computer model. We make sure that the experiments allow us to measure an unknown that could not be determined a priori from a computer model.
We have used microbial systems to study the ecology of host-parasite co-existence and the evolution of sex, mutation rates, transposable elements, group and individual adaptations, and game theory strategies. These projects tested various theoretical models whose biological applicability was constrained by the values of particular parameters. The issue is then whether parameter values realized in biological systems are within the requirement of a model. For example, in studying the advantage of sex in RNA viruses, we tested the reality of Muller’s Ratchet, which is the hypothesis that sex evolved in response to a high deleterious mutation rate. Asexual populations can lose by chance mutation-free genomes and will therefore accumulate deleterious mutations. On the other hand, although sexual populations may also lose their mutation-free genomes, they can recreate them through recombination between two mutated genomes. However, for Muller’s Ratchet to operate, the deleterious mutation rate needs to be sufficiently high to counter the mutation rate and selection intensity for beneficial mutations. Through controlled experiments we were able to show that Muller’s Ratchet can indeed operate in RNA viruses.
A current project in our laboratory tests the viewpoint that evolution by natural selection is more likely to proceed by small steps instead of large steps. The basis of this notion is the geometric model of Fisher (Figure 1), who proposed that small steps are more likely because advantageous mutations of small effect are more common than advantageous mutations of large effect. We tested the model by allowing viral populations to evolve at different sizes and found a positive relationship between step size and population size. In small populations, only the more common advantageous mutations, the ones of small effect, are present and evolution can proceed only by small steps. On the other hand, both large and small advantageous mutations are present in large populations and evolution proceeds by large steps, which also confer a higher selective advantage. These results were surprising because Fisher’s model requires a multi-dimensional phenotypic space (see Figure 1) and it was not known whether the phenotypic space for a virus would be sufficiently complex. Our study provided the first experimental test of Fisher’s model.
Fisher’s model in two dimensions.
Two continuous phenotypes x and y have a fitness optimum at point
o. Fitness drops as x and y move away from the optimum,
and the circle through d and centered around o depicts the phenotypes
that have the same low fitness. If a population is at d,
mutations of small and large effect arising in the population
are represented, respectively, by the small and large circles
centered around d. The dashed portion of the circles corresponds
to mutations that bring the population from d to points closer
to o and are therefore advantageous. The solid portion corresponds
therefore to deleterious mutations, which take the population
away from o. Although the advantageous proportion is approximately
50% of the small circle, it is much less than 50% of the large
circle. Thus, evolution in response to natural selection
should more likely proceed by small steps because advantageous
mutations of small effect will be more common than advantageous
mutations of large effect. This bias against large mutations
becomes stronger as more dimensions are added.
