Selecting for altered substrate specificity reveals the evolutionary flexibility of ATP-binding cassette transporters

Abstract

<jats:title>Abstract</jats:title><jats:p>ABC transporters are the largest family of ATP-hydrolyzing transporters, with members in every sequenced genome, which transport substrates across membranes. Structural studies and biochemistry highlight the contrast between the global structural similarity of homologous transporters and the enormous diversity of their substrates. How do ABC transporters evolve to carry such diverse molecules and what variations in their amino acid sequence alter their substrate selectivity? We mutagenized the transmembrane domains of a conserved fungal ABC transporter that exports a mating pheromone and selected for mutants that export a non-cognate pheromone. Mutations that alter export selectivity cover a region that is larger than expected for a localized substrate-binding site. Individual selected clones have multiple mutations which have broadly additive contributions to specific transport activity. Our results suggest that multiple positions influence substrate selectivity, leading to alternative evolutionary paths towards selectivity for particular substrates, and explaining the number and diversity of ABC transporters.</jats:p>

Selecting for Altered Substrate Specificity Reveals the Evolutionary Flexibility of ATP-Binding Cassette Transporters

“…In one mating type, these pheromones are small peptides that are C-terminally farnesylated and methyl esterified ( Figures 1A and S1A ); they undergo maturation through a conserved set of enzymes, highlighting the ancestral role of farnesylated pheromones in fungal mating [ 23 ]. We used these common features and a bioinformatic approach to identify and experimentally validate the farnesylated pheromone from Y. lipolytica [ 22 ]. Orthologs of Ste6 (ABCB exporters) [ 24 – 26 ] can be reliably identified by sequence homology ( Figure 1B ).…”

“…We started by testing the ability of Ste6 homologs from different Ascomycetes to rescue S. cerevisiae a-factor export. The Ste6 homolog from Y. lipolytica , which last shared a common ancestor with S. cerevisiae 320 mya, exports S. cerevisiae a-factor poorly but exports the Y. lipolytica a-factor, which lacks any sequence similarity to S. cerevisiae a-factor [ 22 ]. This functional difference in substrate transport allowed us to develop a high-throughput assay for mutations in a transporter that increased its ability to export a-factor from S. cerevisiae .…”

Evolutionary history of ATP‐binding cassette proteins

“…As we recognize the evolutionary history of ABC transporters, we are attempting to place structural and mechanistic knowledge of well-studied members of the family in the appropriate evolutionary context. However, given the extraordinary flexibility of substrate selectivity [95,96,119], substrate binding [37,92], and coupling of nucleotide hydrolysis and conformational changes [40,100,102,103], the limits of functional variation by selection remain unclear. To understand the fundamental sequence and mechanistic constraints placed by evolution on ABC proteins, taking advantage of modern high-throughput experimental approaches and incorporating statistical measures of sequence variation will be essential.…”

“…In fact, the substrate binding site within TMDs may not be at a conserved site or conformational state across homologous transporters [92,93]. Analysis of natural sequence variation [94] and experimental selection to transport a novel substrate [95] suggest that many positions influence the evolution of substrate selectivity. Besides the 'simple' altering of substrate selectivity, evolutionary sequence variation has also modified ABC transporters to function as an ATP-gated Clchannel in CFTR [96] and function as regulatory proteins in the nucleotidegated SUR K + channel complex [97].…”

“…Libraries generated by random mutagenesis of a larger region-of-interest in a protein (>5 residues) produce a small fraction of evolutionary possibilities available. By selecting for a novel function or restoring function by starting from a defective mutation, this approach investigates a fraction of the routes available to evolution [95]. An alternative approach is to generate a complete library of every possible single mutation in a protein, that is, all other 19 amino acids at every position.…”