Molecular machines underlie the vast diversity of the living world and are the result of millions of years of selection to optimize them for particular biochemical tasks. If an ancestral gene for a protein is duplicated and the two copies evolve along different paths, they can acquire related but different functions. For some particularly successful families of proteins, this process of duplication and divergence has been repeated many times to create many members of a protein family, each performing a different function. A central question in molecular evolution is how evolution changes the sequence of these family members to produce new molecular functions.
In the labs of Rachelle Gaudet and Andrew Murray, Sriram Srikant sought to answer this question, by combining the expertise of the two labs in protein biochemistry and evolutionary genetics. ATP-binding cassette (ABC) transporters are a particularly interesting case of protein evolution, because they are the largest superfamily of proteins that we know of. As the name suggests these proteins use the chemical energy of ATP hydrolysis to power the transport of substrates across biological membranes. In the last 20 years, biochemistry and structural biology have revealed a conserved molecular mechanism for a large subfamily of these transporters whose members are present in every known species, and have evolved to transport an enormous range of molecules. Given their conserved architecture and mechanism, how does variation in the sequence of these ABC transporters explain their ability to transport such a wide range of substrates, while maintaining the conserved biophysical mechanism?
Sriram took advantage of a set of related ABC exporters that export a class of fungal pheromones that induce the first step in mating. The pheromones in this class are small peptides with a lipid tail attached to their C-terminus. They are produced in the cytoplasm and then exported from cells by dedicated ABC exporters. The peptide sequence of these pheromones varies between different fungal species and the substrate selectivity of the transporters seems to have coevolved with their cognate pheromones. This property allowed the molecular evolution of substrate specificity to be recast as an experimentally tractable problem: which mutations in a pheromone exporter from one species (Yarrowia lipolytica) allow it to transport the pheromone from the brewer’s yeast (Saccharomyces cerevisiae), despite it being 320 million years since the two yeasts shared a common ancestor?
The transporters are known as Ste6 and the substrates as a-factors, so the question becomes which mutations in Y. lipolytica’s Ste6 would allow it to transport a-factor from S. cerevisiae. To report on the export of pheromone, Sriram engineered S. cerevisiae cells to express the a-factor receptor so that the same cell could both secrete and respond to the pheromone by inducing the expression of fluorescent protein. The Y. lipolytica Ste6 can barely export any a-factor from S. cerevisiae, making it possible to screen mutated transporters for versions of the transporter that better export a-factor and thus produce a stronger fluorescent signal.
Sriram found that a number of mutations across the transmembrane domain of Y. lipolytica Ste6 improved its ability to transport a-factor. Interestingly, these mutations are not restricted to a single, small region that would resemble the substrate binding site of an enzyme, but are distributed across the length of the helices that form the large transmembrane cavity. He found that the selected clones have multiple mutations, many of which contribute individually to altering substrate selectivity. The effect of the individual mutations is roughly additive, allowing a set of mutations, each with a detectable individual effect, to produce a large increase in the ability of a transporter to export a substrate that it has not seen for many millions of years. The combination of a large target size for mutations that can alter transport specificity, mutations with large beneficial effects, and mutations that can combine at least additively contribute to the dramatic “evolvability” of ABC transporter substrate selectivity.
Future work would involve mutating transporters whose current function is to transport even more dramatically different substrates to see if the mutational trajectories that change what a transporter exports become more restricted as substrates are more chemically different. Is the space of available mutational trajectories large enough that sequences of functionally orthologous transporters are no closer to each other than distantly diverged transporters? The selection system could also be adapted to study the molecular evolution of other proteins in the mating cascade of fungi, including members of other large protein families like G-protein coupled receptors and MAP kinase cascades.