, 2009, Lee et al., 2010, Moon et al., 2006 and Moon et al., 2009). Analysis of Gr-GAL4 drivers has shown that Gr5a is expressed in sugar-sensitive neurons in each sensillum, while Gr66a is expressed in a distinct population

of ∼20 neurons that responds to a number of bitter compounds and that mediates aversion ( Chyb et al., 2003, Marella et al., 2006, Thorne et al., 2004 and Wang et al., 2004). Two Gr5a-related genes map to Gr5a-expressing neurons, while a number of other Gr genes appear to be expressed in subsets of Gr66a-expressing neurons ( Dahanukar et al., 2007, Lee et al., 2009, Moon et al., 2009, Thorne and Amrein, 2008, Thorne et al., 2004 and Wang et al., 2004). The sensilla associated with these subsets have not been identified in most

cases, however, and expression of the great majority of Gr genes has not been examined. Historically, a critical question check details in the field has been whether all taste sensilla are functionally equivalent (Hiroi PI3K inhibitor et al., 2002, Marella et al., 2006, Thorne et al., 2004 and Wang et al., 2004). Previous physiological analysis of the labellum revealed that three sensilla, L7, L8, and L9 (Figure 1A), were similar in their responses to all of 50 tested compounds, mostly sugars (Dahanukar et al., 2007). A study of 21 sensilla and four sugars showed that all sensilla responded to all tested sugars, with some quantitative differences among sensilla of different morphology (Hiroi et al., 2002). A survey of a few bitter compounds revealed that none of the longer sensilla on the labellum responded, while all of the shorter hairs that were tested gave indistinguishable responses (Hiroi et al., 2004). An imaging study found that different subpopulations of bitter cells responded to most bitter compounds tested; striking differences in response profiles were not observed (Marella GPX6 et al., 2006). Based on these studies, it has been suggested that bitter-sensitive neurons of the labellum may generally recognize the same bitter

compounds (Cobb et al., 2009 and Marella et al., 2006). A similar model emphasizing functional homogeneity is often cited in mammals, in which multiple bitter receptors are coexpressed and taste receptor cells respond to a broad range of bitter compounds (Adler et al., 2000, Mueller et al., 2005 and Yarmolinsky et al., 2009). However, a systematic analysis of the responses of the labellar taste sensilla to bitter compounds, such as those carried out with Drosophila olfactory sensilla and odorants ( de Bruyne et al., 2001), has not been performed. Because of the limited scope of the extant studies, the basic principles of functional organization that underlie bitter coding in the fly remain unclear. Here we investigate basic principles of bitter coding through a systematic behavioral, physiological, and molecular analysis.