Pavlov's cockroach: classical conditioning of salivation in an insect

Abstract

Secretion of saliva to aid swallowing and digestion is an important physiological function found in many vertebrates and invertebrates. Pavlov reported classical conditioning of salivation in dogs a century ago. Conditioning of salivation, however, has been so far reported only in dogs and humans, and its underlying neural mechanisms remain elusive because of the complexity of the mammalian brain. We previously reported that, in cockroaches Periplaneta americana, salivary neurons that control salivation exhibited increased responses to an odor after conditioning trials in which the odor was paired with sucrose solution. However, no direct evidence of conditioning of salivation was obtained. In this study, we investigated the effects of conditioning trials on the level of salivation. Untrained cockroaches exhibited salivary responses to sucrose solution applied to the mouth but not to peppermint or vanilla odor applied to an antenna. After differential conditioning trials in which an odor was paired with sucrose solution and another odor was presented without pairing with sucrose solution, sucrose-associated odor induced an increase in the level of salivation, but the odor presented alone did not. The conditioning effect lasted for one day after conditioning trials. This study demonstrates, for the first time, classical conditioning of salivation in species other than dogs and humans, thereby providing the first evidence of sophisticated neural control of autonomic function in insects. The results provide a useful model system for studying cellular basis of conditioning of salivation in the simpler nervous system of insects.

Figure 1. Experimental procedure.

(A), Set-up for measurement of the amount of saliva secretion is illustrated as a lateral view. The distal cut-stump of the salivary duct (SD) was inserted into a small plastic chamber, and the tip of the plastic chamber was covered with white Vaseline (dotted square) to prevent leakage of saliva. Saliva (shaded area) secreted from the SD was collected by a plastic capillary attached to a syringe every 1 min. The amount of secreted fluid was calculated from the length (L) of the fluid column. (B), Stimulus schedules for five P+/V− or P−/V+ forward-pairing trials. One P+/V− or P−/V+ forward-pairing trial consisted of the presentation of peppermint (shaded rectangles) or vanilla odor (hatched rectangles) to an antenna before the presentation of sucrose solution (white squares and white triangles) to the mouth and presentation of vanilla or peppermint odor without pairing with sucrose solution, respectively. (C), Stimulus schedules for five P+/V− backward-pairing trials. One P+/V− backward-pairing trial consisted of the presentation of peppermint odor after the presentation of sucrose solution and subsequent unpaired presentation of vanilla odor. (D), Stimulus schedule for unpaired presentation of peppermint and vanilla odors (CS alone). Peppermint and vanilla odors were alternately presented five times without pairing with sucrose solution. (E), Stimulus schedules for unpaired presentation of sucrose solution (US alone). Sucrose solution was presented five times without pairing with an odor. The duration of olfactory stimulus was 4 sec. The inter-stimulus intervals were 5 min in B–D and 10 min in E. The measurements of salivation responses to peppermint, vanilla and apple odors (Test; black squares) were performed at 30 min (B–E) or 1 day (B) after conditioning or control trials.

Figure 2. Changes in salivation levels upon olfactory (A,B) or gustatory (C,D) stimulations.

In one group of cockroaches, peppermint (solid line), vanilla (broken line) and apple odors (dotted line) were presented for 4 sec to an antenna with intervals of 6 min (A). In another two groups of cockroaches, 4 µl of 500 mM sucrose solution (solid line) or 5 M sodium chloride solution (dotted line) was presented to the mouth (C). The amount of saliva secreted from the salivary duct was measured every 1 min, and values are shown as means±s.e.m. For statistical evaluation, the amount of saliva secreted for a 2-min period after the onset of olfactory (B) or gustatory (D) stimulation (R; black bar) was compared with that before stimulation (R₀; white bar). Asterisks indicate the level of significance (** <0.01, NS>0.05; WCX-test).

Figure 3. Changes in salivation levels in response to odors after P+/V− (A–C) or P−/V+ (D–F) forward-pairing trials.

In one group of cockroaches, peppermint (solid line), vanilla (broken line) and apple odors (dotted line) were presented to an antenna at 30 min after five P+/V− (A) or P−/V+ (D) forward-pairing trials. The odors were presented for 4 sec with intervals of 6 min. The amount of saliva secreted from the salivary duct was measured every 1 min, and values are shown as means±s.e.m. For statistical evaluation, the amount of saliva secreted for a 2-min period after the onset of stimulation (R; black bar) was compared with that before stimulation (R₀; white bar) at 30 min (B,E) or at 1 day (C,F) after five P+/V− or P−/V+ forward-pairing trials. Asterisks indicate the level of significance (** <0.01, NS>0.05; WCX-test).

Figure 4. Changes in the levels of salivation in response to odors after control trials.

In each of three groups of cockroaches, the amount of saliva secreted in response to peppermint, vanilla and apple odors was measured at 30 min after P+/V− backward (A), CS alone (B) or US alone (C) trials. The amounts of saliva were measured every 1 min, and values are shown as means±s.e.m. For statistical evaluation, the amount of saliva secreted for a 2-min period after the onset of olfactory stimulation (R; black bar) was compared with that before stimulation (R₀; white bar). Asterisks indicate the level of significance (** <0.01, NS>0.05; WCX-test).