Betül Kaçar has posted another preprint to bioRxiv describing her work combining molecular paleontology with experimental evolution. I’ve written about Dr. Kaçar’s research, and the Discovery Institute’s bizarre interpretations, before, and I won’t be surprised if the cdesign proponentsists feel compelled to respond again.

The new preprint describes experimental evolution in E. coli bacteria genetically engineered to express an ancient protein in place of its modern counterpart. The gene encoding the protein, Elongation Factor Tu (EF-Tu), exists in two copies in the wild-type E. coli genome. Dr. Kaçar’s team deleted one copy and replaced the other with a gene sequence inferred to be similar to that in E. coli‘s ancestor from 700 million years ago.

Not content to play god by mixing modern and Precambrian genomes, Dr. Kaçar subjected the engineered bacteria to experimental evolution over 2000 generations. Replacement of the modern EF-Tu gene caused an immediate drop in fitness, but over ~1500 generations of evolution, the fitness of the engineered bacteria increased to nearly that of modern E. coli. Surprisingly (to me, anyway), the bacteria adapted not by changes to the EF-Tu gene itself but mostly by changes that affected the expression level of the ancient gene. Five of six replicate populations accumulated mutations in a promoter region upstream of EF-Tu.

Aside from the creative methodology, this work informs a real scientific debate over the relative importance of structural (i.e. protein coding) versus regulatory (i.e. gene expression) changes in phenotypic evolution. Previous bacterial evolution experiments have generally shown a primary role for changes in proteins themselves. In this case, though, few nonsynonymous changes to protein-coding sequences occurred at all, and none in EF-Tu.