Abstract
Olfactory coding relies on primary information from olfactory receptor cells that respond to volatile airborne substances. Despite extensive efforts, our understanding of odor-response profiles across receptors is still poor, because of the vast number of possible ligands (odorants), the high sensitivity even to trace compounds (creating false positive responses), and the diversity of olfactory receptors. Here, we linked chemical purification with a gas chromatograph to single-receptor type recording with transgenic flies using calcium imaging to record olfactory responses to a large panel of chemicals in seven Drosophila ORs: Or10a, Or13a, Or22a, Or42b, Or47a, Or56a, and Or92a. We analyze the data using linear-nonlinear modeling and reveal that most receptors have "simple" response types (mostly positive: Or10a, Or13a, Or22a, Or47a, and Or56a). However, two receptors (Or42b and Or92a) have, in addition to "simple" responses, "complex" response types to some ligands, either positive with a negative second phase or negative with a positive second phase, suggesting the presence of multiple binding sites and/or transduction cascades. We show that some ligands reported in the literature are false positives, because of contaminations in the stimulus. We recorded all stimuli across concentrations, showing that at different concentrations, different substances appear as best ligands. Our data show that studying combinatorial olfactory coding must consider temporal response properties and odorant concentration and, in addition, is strongly influenced by the presence of trace amounts of ligands (contaminations) in the samples. These observations have important repercussions for our thinking about how animals navigate their olfactory environment.