Evolution of symbiosis - in real time!

I'm currently at the 15th Annual International Conference on Microbial Genomics in College Park, Maryland. I just saw a super cool talk by George Church from Harvard Medical School. Here's what they did. They made mutants in E. coli bacteria that were no longer able to make a particular amino acid. One of the mutants was unable to make tryptophan. The other was unable to make tyrosine. So these guys are unable to live, unless they are supplemented with the appropriate amino acid.

Well he asked the question, "can you watch the evolution of symbiosis in action?" He mixed these two cultures, and didn't supplement them. At first, they grew very, very slowly (doubling every 8 hours - much slower than regular E. coli). They were able to supplement the missing amino acids from leakage from their neighbor (i.e. the tryptophan mutant fed the tyrosine mutant and vice versa). After many generations (we're talking hundreds, here) they found that the cells grew more and more quickly. Using next-generation sequencing technology, they found that the increased speed of growth was due to a modification in the pores on the outside of the E. coli cells that allowed them to more easily take up the amino acids.

In a matter of a few months, they can see evolution of a symbiosis between two bacteria that are unable to grow on their own. They're able to watch that symbiosis improve, and molecularly characterize the improvement. How's that for evolution in action? (Though, admittedly, it won't be enough for everybody). Very cool talk (and that was only a small part of his half hour talk!)