Protection from premature habituation requires functional mushroom bodies in Drosophila

Abstract

Diminished responses to stimuli defined as habituation can serve as a gating mechanism for repetitive environmental cues with little predictive value and importance. We demonstrate that wild-type animals diminish their responses to electric shock stimuli with properties characteristic of short- and long-term habituation. We used spatially restricted abrogation of neurotransmission to identify brain areas involved in this behavioral response. We find that the mushroom bodies and, in particular, the alpha/beta lobes appear to guard against habituating prematurely to repetitive electric shock stimuli. In addition to protection from premature habituation, the mushroom bodies are essential for spontaneous recovery and dishabituation. These results reveal a novel modulatory role of the mushroom bodies on responses to repetitive stimuli in agreement with and complementary to their established roles in olfactory learning and memory.

Figure 1.

Habituation to repeated exposure to electric shock. (A) Experience-dependent attenuation of shock reactivity. Avoidance of a copper grid where 45 V shocks are delivered every 4 sec with 1.2 sec duration each, for the 90-sec test period. 0 denotes avoidance of the electrified grid without prior experience of similar electric shocks (naive), whereas the numbers indicate the number of shocks experienced prior to testing. Each measurement represents the mean performance indices (PI) ± standard error of the mean (SEM) for n ≥ 8 for each wild-type stock; (filled diamonds) w¹¹¹⁸; (open squares) Berlin (B). ANOVA indicated significant differences (F_(11,127) = 39.46; P < 0.0001) among the PIs. Subsequent planned comparisons indicated significant differences (P < 0.001) in shock avoidance of naive animals and the performances after pre-exposure to 11 and 15 shocks. Performance after eight shocks was also different from that of naive animals (P < 0.05). (B) Spontaneous recovery of shock avoidance following exposure to 15 training shocks. PI ± SEMs are shown for n ≥ 8, per time interval for each genotype. ANOVA (F_(7,84) = 12.37; P < 0.001) indicated significant differences. Planned comparisons revealed highly significant differences (P < 0.001) in shock avoidance after 6 and 9 min of rest in comparison to that immediately after training with 15 shocks (0). These animals are distinct from those used in part A, but because the experiments in A and B were performed concurrently, we also compared the performance of naive animals (0 in A) to that that of animals trained with 15 stimuli and allowed to rest for 9 min. These two means were not found to be statistically different (P < 0.1336). (C,D) Animals were trained and tested to the same stimulus strength. The magnitude of shock avoidance decrement is inversely related to stimulus strength. PI ± SEMs are shown for n ≥ 8 per group. Naive avoidance of 45 V shocks was different (planned comparisons, P < 0.001) from those at 60 and 90 V for both w¹¹¹⁸ (C) and B (D). Avoidance of 45 V shocks after exposure to 15 stimuli at 45 V was significantly different from that at 60 or 90 V (planned comparisons, P < 0.001) for w¹¹¹⁸ (C) and B (D). Similarly, avoidance of 60 V shocks after exposure to 15 stimuli at 60 V was significantly different from that at 90 V (planned comparisons, P < 0.001). (E) Reversal of the experience-dependent attenuation of shock avoidance with an olfactory stimulus. PI ± SEMs are shown for n ≥ 8 per group. ANOVA (w¹¹¹⁸), F_(5,57) = 31.26; P < 0.0001; ANOVA (B), F_(5,55) = 24.17; P < 0.0001. Subsequent Dunnett’s tests revealed that compared to the performance of naive animals, exposure to 15 training stimuli significantly (P < 0.001) altered shock avoidance during testing, but two training stimuli did not. Significantly, shock avoidance of animals trained with 15 stimuli, but exposed to 10 sec of benzaldehyde prior to testing (15 × 45 V + BNZ), performed indistinguishably from control animals. The performance of animals shaken after training with 15 stimuli was significantly different (P < 0.001) from that of naive animals.

Figure 2.

Repeated training cycles lead to long-term habituation. (A) The effect of repeated training cycles on habituation measured immediately after training and 30 min post-training. PI ± SEMs are shown for n ≥ 8 per group. ANOVA indicated significant differences: F_(8,61) = 21.09, P < 0.0001. Subsequent Dunnett’s tests revealed highly significant differences (P < 0.0001) in the performance of naive animals compared to that of animals trained with one, two, or three cycles immediately after training or 30 min later. In contrast, the performance of animals tested after three cycles of training followed by a single exposure to benzaldehyde was not significantly different immediately (P = 0.214) or 30 min later (P = 0.101) from that of naive animals. (B) Robust habituation is detected after four training cycles with a 10-min ITI for at least 90 min. PI ± SEMs are shown for n ≥ 7 per group. ANOVA indicated significant effects of ITI and retention interval (P < 0.001). Subsequent planned comparisons revealed that for ITIs of 5 min, performance at all retention intervals except 90 min post-training were significantly different from the performance of naive animals. Similarly, for ITI of 10 min, performances at all retention intervals were significantly different (P < 0.001) from that of naive animals.

Figure 3.

Impairing neurotransmission in different parts of the brain reveals that the MBs are involved in protection from premature habituation to electric shock. PI ± SEMs are shown for n ≥ 7 per group. The UAS-TNT/+ group was included in this experiment simply to demonstrate that this transgene alone did not precipitate significant behavioral alterations. However, since all strains expressing TNT under the drivers used (signified by >) were accompanied by control strains carrying the driver transgene as a heterozygote in a balanced experimental design. Significant differences in the performance of all strains when naive were not observed (ANOVA: F_(8,89) = 2.37; P < 0.0123). However, significant differences in subsequent shock avoidance were revealed for animals trained with two stimuli (ANOVA: F_(8,82) = 12.41; P < 0.0001). This difference arose from the highly significant difference in avoidance of P247 > UAS-TNT animals compared to their P247/+ controls (P < 0.001, Student’s t-test). No significant differences in the performance of all strains were detected after training with 15 stimuli (ANOVA: F_(9,87) = 1.28; P < 0.0598).

Figure 4.

Premature habituation to shock in animals with structurally or functionally perturbed MBs. (A) Two training stimuli suffice to precipitate significant attenuation in shock avoidance in MB-perturbed animals. PI ± SEMs are shown for n ≥ 7. 0 represents shock avoidance of naive animals. ANOVA indicated significant effects of shock number and genotype (F_(23,176) = 38.61; P < 0.0001). Compared to the performance of naive controls (0 time), significant differences (P < 0.001, Dunnett’s test) in performance arose after one training shock and remained highly significant in MB-perturbed animals. In contrast, 10 training stimuli were required for control strains to exhibit significantly (P < 0.001) attenuated performance compared to naive animals. (B) Recovery of normal (naive) shock avoidance 15 min after training MB-perturbed animals with two stimuli. PI ± SEMs are shown for n ≥ 7. The performance of naive animals of all relevant strains is shown as reference. ANOVA indicated significant effects of the rest interval and genotype (F_(19,154) = 46.61; P < 0.0001). Subsequent Dunnett’s tests with the performance of naive animals of each strain used as control, revealed that following two-stimulus training, the performance of MB-perturbed animals was significantly (P < 0.0001) attenuated after 6 min, but not after 15 min of rest. In contrast, significant differences in shock avoidance of control strains were not detectable immediately, or at later rest intervals. (C) Naive shock avoidance was not restored by an odor pulse following two-stimulus training. PI ± SEMs are shown for n ≥ 8. Compared to the performance of naive flies, there was no significant difference in shock avoidance of B and UAS-TNT/+ after two-stimulus training, or a BNZ pulse after training. In contrast, the performance of naive mbm¹ and P247>UAS-TNT animals remained significantly different from that after two-stimulus training and after the BNZ pulse (P < 0.001). (D) The magnitude of shock avoidance attenuation is inversely related to stimulus strength. PI ± SEMs are shown for mbm¹animals, n ≥ 7. ANOVA indicated significant effects of both stimulus strength and number of stimuli (F_(8,69) = 50.25; P < 0.0001). Subsequently, the performance of animals under 45 V stimulation was compared to those under 60 and 90 V (Hsu’s test). Naive avoidance at 60 and 90 V was significantly different from that at 45 V (P < 0.001). Similarly, shock avoidances at 60 and 90 V were significantly different from those at 45 V after two- or 15-stimulus training (P < 0.001 for both comparisons).

Figure 5.

Protection from premature habituation to shock requires the α/β lobes of the MBs. (A) Structural and functional ablation of the MBs precipitates habituation to shock after two training stimuli. PI ± SEMs are shown for n ≥ 7. The performance of all experimental strains was compared to their relevant heterozygous controls. There were no significant differences in naive avoidances except between 201Y/+ and the experimental 201Y>UAS-TNT (P < 0.001). Two-stimulus training yielded significant differences in the performance of mbm¹, compared to that of B; w¹¹¹⁸-HU, compared to w¹¹¹⁸ alone; P247>UAS-TNT compared to P247/+; 201Y>UAS-TNT, compared to 201Y/+; and c772>UAS-TNT, compared to c772/+ (P < 0.001 for all comparisons, Student’s t-tests). (B) Inhibition of neurotransmission specifically in the α/β lobes leads to premature habituation to shock. PI ± SEMs are shown for n ≥ 7. The performance of all strains after incubation at 32°C (see Materials and Methods) is shown. The performances of animals expressing UAS-shi^ts under P247, 11Y, and 201Y after training with two stimuli were significantly different from those of their respective control heterozygotes (P < 0.001, Student’s t-tests). In contrast, the performances of animals expressing UAS-shi^ts under GH146, H24, 1471, and NP1131 were not significantly different from those of their respective controls.