Gene duplication shaped the origin and evolution of the vertebrate olfactory combinatorial code

Abstract

Animals read their chemical environment through olfactory receptors, which form part of the class A GPCRs and are the largest gene family devoted to a single function in animals. Each odorant molecule activates a different combination of receptors, projects distinctly in the brain, and so yields a distinct odour percept. The principles of this coding are well described, and the evolution and regulation of these genes are actively researched. However, how the code was initially assembled and evolved has not been examined. Here we combine phylogenetics, receptor-ligand interactions predicted by paired protein and chemical language models, ancestral sequence reconstruction, and chemical-space analysis. Predicted activation profiles group odorants into six clusters -each enriched for its own set of odour descriptors- so that molecules read by similar receptor combinations tend to smell alike. Reconstructed ancestors place the origin of the code at the first duplication of the family, in the gnathostome ancestor, at the class I/class II split. The first duplicates then diverged asymmetrically into distinct chemical subspaces. One kept a third of the ancestral ligands; the other re-tuned almost completely and took up carboxylic acids -the polar chemistry long associated with class I. Aromatics were retained and enriched on the class II side. Combinatorial coding was not really invented: it arose once, at the first duplication, and was never lost.