Apart from the handful of true insectophiles among us, most people are totally unaware of the vast majority of insect species. Only a select few species demand more attention because of the behaviors they exhibit that are either beneficial or harmful to humanity. Some species help us by pollinating important food crops; other species feed on those same crops and destroy billions of dollars worth of produce every year. Then, of course, there are the disease vector insects, which include everyone’s favorite flying hypodermic needle—the mosquito. Most of the behaviors that bring these specific insects to our attention, whether in a good or bad way, revolve around their search for and ingestion of food. This is why my lab is particularly interested in the chemosensory systems insects use to find food and in the neurogenetic basis of insect feeding.

Although the goal of our research is to understand and ultimately modify the behavior of pest insects, the tools of modern molecular genetics are not particularly well-developed in mosquitos or any other pest species. Instead, our lab studies the classic model insect Drosophila melanogaster, which is an even better model for the biology of other insect species than it is for human biology. The incredible power we have over the genes of this tiny fly makes it particularly appealing for studying the neurogenetic basis for innate insect behaviors.

One of our major focuses is the olfactory system, which is the main chemosensory system insects use to locate food. Drosophila olfaction is mediated by a divergent family of odorant receptors (ORs), which function as ligand-gated ion channels. The ORs are expressed in olfactory sensory neurons (OSNs) that innervate small hairs called sensory sensilla on the fly’s olfactory organs—the antennae and maxillary palps (Fig. 1). Airborne volatile ligands enter the olfactory sensilla, diffuse through the sensory lymph, and bind to ORs. Odorant binding gates a non-selective cation current through the OR protein, which initiates action potentials in the OSNs. The OSNs project into the brain to the antennal lobes where the pattern of their activity is passed on to projection neurons for further processing.

Most olfactory neurons express at least two OR genes: one subunit that provides ligand-specificity (green) and one chaperoning co-receptor called Orco (blue) (Fig. 2). Most scientists assumed that the insect ORs would be members of the GPCR protein superfamily because they have seven transmembrane domains like the vertebrate odorant receptors. It is now clear, however, that the insect ORs are completely unrelated to all known GPCR subfamilies and that they have an inverted membrane topology with cytoplasmic N-termini. Although there is usually enough homology between insect ORs to recognize them as ORs, they are quite divergent within and between species. We do know, however, that the insect olfactory systems studied thus far function in much the same way because the function of the chaperoning co-receptor Orco has been conserved over at least 250 million years of evolution in species as diverse as moths and mosquitos.

In addition to olfaction, insects use another chemosensory system to guide appropriate feeding behaviors. Flies discriminate food quality with their gustatory system, which is mediated by a large family of gustatory receptors expressed in the mouth parts, legs, and wing margins. Since gustatory ligands are typically soluble, this is a short-range sensory system. One major exception to this rule is in sensing carbon dioxide, which is highly volatile. Most insects are exquisitely sensitive to the concentration of carbon dioxide in the air, but it is the blood-feeding insects that use it for the most nefarious purposes. Mosquitos track the carbon dioxide we release with every breath as one component of a complex and irresistible human odor signature. Since Drosophila also smell carbon dioxide, and use it as a component of a stress pheromone, we sought to identify the molecular receptor for carbon dioxide in the genetically tractable fruit fly instead of the more difficult mosquito. We found that, reminiscent of the heteromeric nature of their ORs, flies detect carbon dioxide using a pair of gustatory receptors (GR21a and GR63a) expressed in the antenna. Co-expression of GR21a and GR63a can confer sensitivity on neurons normally insensitive to carbon dioxide (Fig. 3). Deletion of the Gr63a gene also abolishes sensitivity in the endogenous neurons, establishing the necessity and sufficiency of these receptors in smelling carbon dioxide.

When it comes to carbon dioxide sensitivity in the mosquito, things are just a little more complicated. The malaria mosquito genome contains three clear homologues of the fly carbon dioxide receptors—AgGr22, AgGr23, and AgGr24. These three genes are all co-expressed in the carbon dioxide-sensitive neurons of the mosquito’s maxillary palps, and heterologous expression of these three genes in Drosophila confers carbon dioxide sensitivity. It is still unclear how and why Drosophila species lost one of their carbon dioxide receptors while still maintaining functional sensitivity. One interesting hypothesis, however, is that ORs and GRs do not function as heterodimers at all, but as higher order combinations of multiple subunits. To this end, we are working to characterize the relationships between chemosensory receptor structure and function. We are harvesting functional protein complexes from fly antennae for biochemical experiments. Once these techniques are fully developed, we will be better able to understand the mechanisms by which insect chemosensory receptors function.

In addition, our lab is working to better understand the microenvironment inside insect olfactory neurons that permits OR function. Traditionally, all the odorant receptors that have been studied to any degree (i.e., vertebrate, nematode, and insect) have been difficult to express in cell culture because they seem to have stringent protein trafficking requirements. The recent discovery that insect ORs are actually ligand-gated ion channels was made using a heterologous cell culture systems in spite of poor protein expression levels and even worse cell surface localization. A better understanding of the trafficking requirements of insect chemosensory receptors would be a boon to the olfaction community and would greatly facilitate detailed biochemical and cell biological analyses for these proteins. One of the most interesting ideas regarding a means of enhancing receptor expression in cell culture revolves around the morphology of the neurons in which they are expressed. The ORs of vertebrates, nematodes, and insects all localize to dendrites that extend as part of a specialized non-motile membranous projection called a primary cilium. We are currently exploring the hypothesis that ORs have been difficult to express in cultured cells because they require a specialized coupling to the conserved trafficking machinery that separates the ciliary membrane from the rest of the plasma membrane. If we can identify the accessory factors that couple insect ORs to this machinery, we should be able to dramatically improve surface expression in cultured cells with prominent cilia (Fig. 4).

Ultimately, we hope the establishment of high-throughput systems will allow us to screen chemical libraries for compounds that interfere with specific subsets of insect ORs. Such compounds should allow us to modify the olfactory-guided feeding behaviors of pest insects that cause so much damage around the world.