With the Drosophila melanogaster genome in the bag, the fruit fly has become an important tool for linking the physiology of olfaction with genetics. But although Drosophila has told us much about how insects detect odours, there are (surprisingly) huge gaps in our knowledge.
Until recently we hardly knew any of the chemical compounds that trigger the sensory structures on a fly's antennae. With such basic facts missing it is not surprising that much of the rest of the chain leading from detection to brain is a mystery.
According to Bill Hansson, a chemical ecologist at the Swedish University of Agricultural Sciences in Alnarp, this is because previous work was not focussed on substances relevant to the fly's ecology. `People would just take compounds off the shelf,' he says, `we had no idea what the odours meant for the animals.'
Now his team, with colleagues at the University of Cagliari in Monserrato,Italy have completed a systematic analysis of all Drosophila's olfactory receptor neurons (ORN) (p. 715). The aim was to match each ORN to its ligand. Hansson accepts that a definitive answer is impossible because you can never be sure you haven't missed an important chemical. In vision, for example, the range of the electromagnetic spectrum sets the limits of what the animal can sense. `But in the world of odours you have an unlimited number of structures,' he says.
To narrow down the search the team looked at compounds from six fruits— banana, litchi, mango, papaya, passion fruit and pineapple. They separated each fruit sample into its chemical components using gas chromatography. The stream of chemicals was split in two and directed simultaneously over mounted flies and into a mass spectrometer. That way,recordings of the firing rate of individual ORN's from a tungsten microelectrode inserted into the cells could be tied to particular components of the odour. Refining the method proved to be tough because of the miniature target for the electrode. `It's hell,' says Hansson, `It gave rise to a lot of problems and a lot of swearing in the lab.'
Of the hundreds of chemicals in each sample only a handful brought a response — never more than eight for any one fruit and only 27 in total. So it seems the flies are able to detect key components of useful odours. Also, each ORN typically responded strongly to a single primary ligand, while giving a weaker response to structurally similar chemicals.
In many cases, the ecological relevance of the ligands is clear. For example, acetoin and isoamyl alcohol are microbial breakdown products and ethyl hexanoate and isoamyl acetate are typical fruit volatiles. In behavioural experiments, typically the flies would be indifferent to low concentrations, attracted to intermediate concentrations and repelled by high concentrations. Only one odour (1-hexanol) acted as a repellent across the whole range. This compound is given off by green plant tissue and unripe fruit, so it makes sense that the flies avoid it.
The authors speculate that flies are able to detect potential food sources at a distance by responding to a few key odours with high specificity. He says the next step will be to isolate the genes involved and make comparisons with olfaction in other closely related Drosophila species.
