Essays on Science and SocietyEPPENDORF WINNER

Evolution and Revolution in Odor Detection

See allHide authors and affiliations

Science  16 Oct 2009:
Vol. 326, Issue 5951, pp. 382-383
DOI: 10.1126/science.1181998

Molecular neuroscientists have a tendency to seek evolutionarily conserved mechanisms underlying the construction and function of animal brains. This approach unarguably helps to define fundamental principles of neurobiology by integrating insights from diverse model nervous systems. However, while what is true of the brain of a mouse or worm may be relevant to our own, a focus on commonalities overlooks the fact that different animal nervous systems have evolved to operate in distinct ecological contexts. Nowhere is this truer than in the olfactory system, which underlies the detection of myriad volatile chemicals in the environment. Animal olfactory systems display enormous evolutionary capacity, as species acquire and discard olfactory receptor genes, neurons, and behaviors in an everchanging landscape of external chemical stimuli. These modifications often reflect the fact that most relevant odors for a species are themselves derived from evolving organisms such as plant food sources, animal predators, and potential mates.

Insect olfactory systems have been important models since the observations of Fabre in the early 1900s of the potency of a volatile chemical released by a caged female emperor moth in attracting male suitors. Surprisingly, insects lagged well behind vertebrate models in molecular investigations, as odorant receptor (OR) genes were identified in the fruit fly, Drosophila, nearly a decade after the discovery of mammalian ORs. Nevertheless, the small number of Drosophila ORs (62, compared with >1000 in mice) permitted comprehensive functional analysis of the role of these receptors and their circuits in odor perception (1, 2). Many common properties of insect and vertebrate olfactory systems have been appreciated, notably, (i) that individual olfactory sensory neurons (OSNs) express just one type of OR, (ii) that the axons of OSNs expressing the same receptor converge to a defined glomerulus within the primary olfactory center, and (iii) that odors are recognized by specific combinations of ORs to create a spatial “code” of glomerular activation (3).

Eppendorf and Science are pleased to present the prize-winning essay by Richard Benton, the 2009 winner of the Eppendorf and Science Prize for Neurobiology.

Justified by this apparent conservation in olfactory system organization across 500 million years of evolution, my research, with collaborators at The Rockefeller University, aimed to use Drosophila to investigate the mechanisms by which ORs transform odor recognition into neuronal activity. This was a goal, however, that rapidly shifted focus. When initially identified, Drosophila OR proteins were predicted to contain seven transmembrane domains—a structural hallmark of G protein–coupled receptors (GPCRs)—leading to the widespread belief that insect ORs, like their mammalian counterparts, signal via G protein–mediated second-messenger cascades. However, upon bioinformatic reexamination of insect OR sequences, we noted that these receptors exhibited no significant similarity to vertebrate ORs or other GPCRs and, unexpectedly, were predicted to have an inverted membrane topology. We confirmed this provocative computational prediction experimentally, using antibodies against OR epitopes to map their location relative to OSN ciliary membranes. These findings led us to conclude that the long-held dogma of insect ORs being GPCRs was incorrect; rather, they defined a distinct family of transmembrane proteins (46). How insect ORs work remains contentious, although recent analysis by others suggests that they function as odor-gated ion channels (7, 8).

Molecular mechanisms of odor detection in mammals and insects.

(A) Mammalian ORs and other classes of vertebrate chemosensory receptors are members of the seven-transmembrane domain GPCR superfamily, with extracellular N termini and intracellular C termini. (B) (left) By contrast, insect ORs define a distinct family of seven-transmembrane domain receptors with inverse topology to GPCRs and function as heteromeric complexes of an odor-binding OR (red) and a co-receptor OR83b (blue) that likely form an odor-gated ion channel. (Middle) Insect ORs responsible for detecting lipid-derived pheromones additionally require the function of a CD36 protein, a lipid-binding cell surface receptor that acts in immune recognition in other organisms. (Right) A second family of insect olfactory receptors are the IRs, which are homologous to the ionotropic glutamate receptor family of ligand-gated ion channels.

At this point, we were faced with a problem: using Drosophila as a “model” system, but working on receptors that were unique to insects. We decided to fly in the face of this insect-specificity by designing a bioinformatics screen to identify additional factors acting in peripheral odor detection, selecting genes that displayed the same insect-specific orthology as the ORs (by comparative analysis of Drosophila, mosquito, and noninsect genomes) and OSN-specific expression (by expressed sequence tag analysis). These simple criteria were highly effective in recovering all known classes of insect olfactory genes and many uncharacterized molecules (9).

Of the latter group, we first focused on a transmembrane protein, SNMP, that is distantly related to the mammalian lipid-binding protein CD36. We showed that SNMP is expressed specifically in pheromone-sensing OSNs and localizes with ORs in the sensory cilia. Electrophysiological analysis of SNMP mutant OSNs demonstrated that it is essential for pheromoneevoked neuronal activity. Most insect pheromones are lipid-derived, leading us to propose that SNMP acts as a cofactor for pheromonesensing ORs by directly capturing pheromone molecules on the surface of OSN cilia (9, 10). Intriguingly, other CD36-related proteins function as cofactors for Toll-like receptors, permitting detection of bacterially derived lipids to initiate innate immune signaling. Our results therefore highlight a mechanistic parallel between conspecific recognition through pheromonal communication and pathogen recognition through the innate immune system. Molecular homologies have also been noted in pheromone and immune detection in mammals (11); a future challenge is to understand the evolutionary basis of such connections.

Our earlier investigations of ORs also revealed a second problem: that flies lacking the function of all receptors (owing to mutation of the OR co-receptor OR83b) still exhibited behavioral responses to many odors. This observation implied the existence of additional odordetection mechanisms. Our bioinformatics screen identified several uncharacterized ionotropic glutamate receptors (iGluRs)—ligandgated ion channels best studied for their roles in synaptic communication—but these receptors were unusual, bearing divergent ligandbinding domains that lack glutamate-interacting residues. Three lines of evidence supported our hypothesis that these proteins, named ionotropic receptors (IRs), represent a new family of olfactory receptors: First, IRs are expressed in specific combinations in OSNs that do not express ORs; second, the proteins localize to sensory cilia rather than synapses; and finally, misexpression of an IR in an ectopic neuron is sufficient to confer novel odor responses. IRs are therefore likely to define a previously unappreciated “second nose” in insects (12). By contrast to insect ORs, iGluR/IR-like proteins are present across animals, plants, and prokaryotes, which hints that these receptors may represent an ancient mechanism for sensing both intercellular and external chemical cues.

Our studies of the biology of Drosophila odor detection have revealed molecular surprises that invite reconsideration of the basis of the striking similarities in olfactory system organization and function across species (13). Was there a primitive olfactory system in the common ancestor of insects and vertebrates, in which subsequent drastic divergence of the odor-detecting receptors was uncoupled from the maintenance of neuroanatomical and physiological logic? Or, does the common design of olfactory systems across different phyla reflect convergent evolution, indicative of the essential properties of a sensory system responsible for detecting innumerable chemical stimuli? Distinguishing these possibilities is not trivial, but either would yield insight into the mechanisms by which at least this part of the nervous system arose and evolved.


Navigate This Article