Different sensilla responded to different subsets of stimuli. For example, I9 and I10 responded strongly to theophylline (TPH) but not DEN, whereas I4 and I5 responded
strongly to DEN but not TPH (Figure 1D). Inspection of the response matrix (Figure 3) reveals extensive heterogeneity among the labellar sensilla, and by extension, among the bitter neurons that they contain. The L sensilla exhibited little or no physiological response to our panel of tastants, in agreement with a previous report (Hiroi et al., Panobinostat in vitro 2004). Two of the S sensilla, S4 and S8, also did not respond to any bitter tastants. All other S type sensilla were broadly tuned, responding to 9–15 of the 16 compounds with a spike frequency of ≥10 spikes/s
(Figure 3, Tables S1 and S2). I type sensilla were more narrowly tuned with respect to our panel of tastants, responding to 3–7 compounds. The strongest response was elicited by 10 mM CAF in the S5 sensillum (60.8 ± 3.3 spikes/s; n = 34). A hierarchical clustering analysis identified five functional classes of labellar sensilla: two classes of broadly tuned sensilla (S-a and S-b), two classes of narrowly tuned sensilla (I-a and I-b), and a fifth class that did not display excitatory responses to any of our panel of tastants (L, S-c) (Figures 4A and 4B). The two classes of S sensilla are both broadly tuned, but the S-b sensilla exhibit greater mean responses
to most tastants (Figure 4B). Notably, this class comprises the three sensilla that uniquely exhibited a second Forskolin nmr Isotretinoin high-frequency action potential (Figure 1C). The more narrowly tuned I-a and I-b sensilla respond to complementary subsets of tastants. Maps of the distribution of the sensilla of each class are shown in Figure 4C. The most broadly tuned sensilla (S-a and S-b classes) are located in the medial region of the labellum, while the narrowly tuned sensilla (I-a and I-b classes) are in lateral regions. The three classes of S sensilla are intermingled in the row of medial sensilla, while the I-a and I-b sensilla are restricted to the anterior and posterior portions of the labellum, respectively. We note with interest that among the five bitter compounds that elicited responses >10 spikes/s from the I-a sensilla, three elicited the most aversive behavioral responses (DEN, sparteine sulfate salt [SPS], and (-)- lobeline hydrochloride [LOB]), and one elicited the fifth most aversive response (berberine chloride [BER]) (Figure 2C). The median isoattractive concentration for these five tastants was <0.1 mM; the median concentration for all the others was ∼1 mM. Although gustatory input from other organs such as the legs probably influences this behavior, these results suggest the possibility that different classes of bitter-sensing neurons make different contributions to the behavior of the fly.