Olfaction, the detection and identification of volatile chemicals, is a primal sense essential for animal survival, reproduction, and social interactions. As such, it is subjected to immense evolutionary pressure toward combining the widest sensory spectrum with the highest possible sensitivity and specificity. To expand their sensory window, mammals dedicate up to 5% of their gene-coding capacity to olfactory receptor (OR) genes, 7-transmembrane G protein–coupled receptors for volatile chemicals [1]. There are ∼1100 functional OR genes in the mouse genome, organized in genomic clusters distributed across most chromosomes. To achieve sensitivity and specificity in odor recognition, mice deploy the ‘one receptor per neuron’ and ‘one receptor per glomerulus’ rules [2]. The former is manifested by the strictly monogenic and monoallelic OR transcription in each mature olfactory sensory neuron (OSN) [3], while the latter by the OR-instructed convergence of all the OSN axons with the same OR protein to the same glomerulus of the olfactory bulb 4, 5. These two principles assure that each OSN is only stimulated by volatile chemicals that activate the OR protein they express, transforming the chemical structure of an odorant into odor-specific patterns of glomerular activation in the brain. Thus, because the OR protein identity determines both odorant detection and axon guidance specificity, OR expression in mature OSNs must be strictly singular.

While singular OR expression is essential for vertebrate olfaction, the rapid evolutionary expansion of the OR gene family from ∼70 OR genes in fish to >1000 OR genes in amphibia and most terrestrial vertebrates [6] poses extreme regulatory challenges. If combinatorial control of gene expression was solely responsible for the singularity and diversity of OR expression, then each newly evolved OR must be controlled by distinct cis-regulatory elements recognized by unique combinations of transcription factors tailored to each OSN subtype [7]. However, both the >1000 OR promoters and >63 intergenic OR enhancers, the Greek Islands 8, 9, 10•• share numerous common transcription factor–binding motifs recognized predominantly by Lin-11, Isl-1, Mec-3-homeobox transcription factor Lhx2 and members of the Ebf family (Ebf1-4) 11, 12, 13, 14. Moreover, single-cell RNA-seq experiments have convincingly shown that the differentiating OSN progenitors do not have the regulatory precision to activate only a single OR, rather they co-transcribe random combinations of 5–15 different OR alleles before switching to a strictly singular OR transcription 15, 16, 17, 18, 19. Consistent with this, single-cell RNA-seq experiments have yet to identify unique combinations of transcription factors correlating with specific OR identities. Thus, while a deterministic process is restricting the OR repertoire that each OSN can choose from [18], additional nondeterministic mechanisms refine this choice and assure that every mature OSN will stably express only one OR allele for the life of the OSN.