This week’s BEACON Researchers at Work blog post is by University of Washington graduate student Tasneem Pierce.
When I started as an undergraduate in BEACON a year ago, I kept hearing about Avida and how powerful it is to study evolution in action. I decided to teach myself how to use the Avida software, and I quickly discovered that there are no tutorials for biologists interested in the more complex aspects of Avida. Fortunately for me, I was in the heart of BEACON, surrounded by people who were willing to take time to teach me how to use the software. Every single one of the people in the photo below, most of whom are in Dr. Ofria’s Digital Evolution lab, helped me in some aspect of my research with Avida. Now, I am working on creating a tutorial targeted at researchers with no computational background.
Front row (L to R): Charles Ofria, Owen Pierce, and Tasneem Pierce
BEACON’s collaborative atmosphere allowed me to start an Avida project of my own. I started my Avida project when I was doing research in Dr. Lenski’s Experimental Evolution Lab. Dr. Lenski’s long-term evolution experiment studies the genetic changes in twelve populations of Escherichia coli that have evolved for over 50,000 generations. If you are wondering how long that is, I’m 23 years old, and Dr. Lenski’s experiment started just a couple of months before I was born. There have been many cool discoveries in the long-term populations, one being that six of the twelve populations have increased mutation rates. These populations are called mutator lines, as they have damaged methyl-directed DNA mismatch repair systems, which have increased their effective mutation rates by a hundred fold. A high mutation rate alone will typically be maladaptive (more mutations are detrimental than beneficial), but if a mutator causes a rare beneficial mutation, that mutator may hitchhike to fixation meaning that the mutator becomes the dominant organism in the population.
How can a higher mutation rate fix in a population? Again, it is more likely that a higher mutation rate will break something important instead of making something better. What circumstances would lead to the fixation of a higher mutation rate? Using Avida, we can identify conditions under which a population will fix a higher mutation rate if it is easy to knock out a mutation-repair mechanism, but difficult to re-evolve one.
E. coli in nature cycles between the nutrient-rich gut and the external, nutrient-limited environment. In the Lenski lab lines, E. coli starts every 24 hours in fresh media, and by the end of the day, it is in a nutrient-depleted environment. In both of these situations, the E. coli face changing environments. A strategy in a nutrient-rich environment might not be beneficial in a nutrient-limited environment and vice versa. We tried to recreate the changing environments in Avida by rewarding Avidians for task set 1 and punishing them for doing task set 2 during one cycle, and then reversing this by rewarding task set 2 and punishing task set 1 in the next cycle.
Our hypothesis was that a moderate environmental change will select for organisms with a higher mutation rate. We had a variety of environments: a static environment where all of the tasks were consistently rewarded and 6 dynamic environments where there was a toggle between rewarding and punishing traits at different rates (100, 250, 500, 1000, 1500, and 2000 updates).
The starting organisms had a divide instruction with a low mutation rate. Organisms could mutate to have a divide instruction with a higher mutation rate as compared to the starting organism (2x, 3x, 10x higher). The Avidians could evolve a higher mutation rate, but they could not re-evolve a lower mutation rate, similar to how it is much harder to fix a DNA repair mechanism once it is broken.
Our preliminary results indicate the digital organisms can fix a higher mutation rate if they are subjected to a dynamic environment. We see that the ideal environmental change is not too short (not enough time to mutate) or not too long (no incentive to change as the environment is changing slowly). As seen in the graph, populations are more likely to fix a higher mutation rate if the change in the mutation rate is smaller (ex. 2x higher mutation rate is favored over fixation of a 10x higher mutation rate). Our initial runs were 100,000 updates long. When we increased the length of the runs, we found that there is an exponential decay in the number of populations at the lower mutation rate, and gradually most of the populations will fix the higher mutation rate. It is possible that, as long-term line E. coli populations go through more generations, more of the twelve lines will fix a higher mutation rate.
I am currently a first year graduate student in the Kerr lab at University of Washington. My new Avida runs will more closely mimic the Lenski lines as the environmental change will switch to a limited resource system instead of rewards/punishments system and the population size will be limited by requiring resources for successful replication. We hope to find conditions that lead to populations fixing a mutation rate that is a hundred fold higher similar to long term E. coli lines. Stay tuned!
For more information about Neem’s work, you can contact her at neem at uw dot edu.
