Figure 2.

The Interneuron PeD12 Plays a Key Role in Behavioral Choice by Activating Withdrawal and Inhibiting Feeding in Response to Touch

(A) The semi-intact head-brain preparation used for recording interneurons and motoneurons of the feeding and withdrawal-response networks. This preparation retains the sensory nerves that carry touch and chemical signals from the lips to the central motor circuits. Paired PeD12 and PlB interneurons (light blue) are located in the pedal ganglia (PeG) and pleural ganglia (PlG), respectively. Feeding motoneurons, B3 and B4 (dark blue), are located in the buccal ganglia (BG). Motoneurons of the whole-body withdrawal network (yellow) are located in several CNS ganglia. The cerebral A cluster is the largest group (6–9 cells), with smaller numbers in the pedal G cluster (3–5 cells) and a single neuron (DLM) in the left parietal ganglion (LPaG). Other CNS ganglia are the right parietal ganglion (RPaG) and the visceral ganglion (VG).

(B) Responses to PeD12 stimulation recorded in the columellar muscle (CM) and the buccal mass (BM). The semi-intact preparation was used for these recordings, but for these experiments, the muscles involved in whole-body withdrawal (CM) and feeding ingestion (BM) were retained, and their contractions were recorded using a force transducer. Sucrose application drives rhythmic feeding movements in the BM until the evoking of a burst of spikes in PeD12 by current injection suppresses feeding despite the continued presence of sucrose. A single large contraction in the CM (^∗) is also caused by PeD12 stimulation (n = 6).

(C1–C3) Electrotonic coupling of PeD12 with motoneurons of the withdrawal-response network. Application of hyperpolarizing square current pulses to PeD12 causes similar but reduced responses in the three corecorded motoneurons (C1). Coupling coefficients recorded in the soma are 0.06 ± 0.01 (n = 5) between PeD12 and Parietal DLM motoneurons, 0.08 ± 0.1 (n = 5) between PeD12 and Pedal G cluster motoneurons, and 0.11 ± 0.02 (n = 12) between PeD12 and Cerebral A cluster motoneurons. Application of touch to the lips (C2) induces bursts of spikes in PeD12 and the three corecorded withdrawal-response motoneurons. A current-induced burst of spikes in PeD12 depolarizes the motoneurons and induces spiking in the motoneurons similar to that produced by touch (C3). All recordings shown in (C1)–(C3) are taken from the same preparation.

(D1–D6) Touch-induced spike activity in PeD12 is both sufficient and necessary for inhibition of feeding. The expanded trace of a B3 fictive feeding burst shows the N1 (protraction), N2 (rasp), and N3 (swallow) phases of the feeding cycle (D1). The inhibition of feeding by touch (D1) is similar to that induced by artificial stimulation of PeD12 (D2), and there is no statistical difference in the two types of data (D3) (n = 6, mean difference scores: touch, −2.4 ± 0.2; PeD12 depolarization, −2.0 ± 0.3; Wilcoxon signed-rank test, W = −8, p = 0.2). Hyperpolarizing PeD12 to suppress spiking (D5) during touch prevents the inhibition of feeding by touch (D4), producing a statistically significant reduction in the difference score (D6) (n = 9, mean difference scores: touch, −2.2 ± 0.2; PeD12 hyperpolarization, −0.9 ± 0.2; Wilcoxon signed-rank test, W = −36, p = 0.014).

In this figure and in the following figures, horizontal bars indicate that either a depolarizing (black) or a hyperpolarizing (gray) square current pulse has been applied. Difference scores in this and other figures are calculated by subtracting the number of feeding bursts in the 20 s before touch from the number of bursts in the 20 s after touch.

Error bars show ±SEM.