Long-term habituation of the C-start escape response in zebrafish larvae

Abstract

The cellular and molecular basis of long-term memory in vertebrates remains poorly understood. Knowledge regarding long-term memory has been impeded by the enormous complexity of the vertebrate brain, particularly the mammalian brain, as well as by the relative complexity of the behavioral alterations examined in most studies of long-term memory in vertebrates. Here, we demonstrate a long-term form of nonassociative learning-specifically, long-term habituation (LTH)-of a simple reflexive escape response, the C-start, in zebrafish larvae. The C-start is triggered by the activation of one of a pair of giant neurons in the zebrafish's hindbrain, the Mauthner cells. We show that LTH of the C-start requires the activity of NMDA receptors and involves macromolecular synthesis. We further show that the long-term habituated reflex can by rapidly dishabituated by a brief tactile stimulus. Our results set the stage for rigorous, mechanistic investigations of the long-term memory for habituation of a reflexive behavioral response, one that is mediated by a relatively simple, neurobiologically tractable, neural circuit. Moreover, the demonstration of NMDAR and transcriptionally dependent LTH in a translucent vertebrate organism should facilitate the use of optical recording, and optogenetic manipulation, of neuronal activity to elucidate the cellular basis of a long-term vertebrate memory.

Keywords: Habituation; Learning; Memory; NMDAR; Zebrafish.

Figure 1

Habituation of the C-start in larval zebrafish expressed at 1 h posttraining requires NMDA receptor activity. (A) Experimental protocol used to demonstrate 1 h habituation. The Habituation Index = the mean posttest response rate minus mean pretest response rate. See the Materials and methods for additional information. All fish were 5 d post-fertilization (dpf) at the start of the experiments. After the Pretest period, during which the larvae were given 5 auditory pulses separated by 5 min, trained larvae received 6 blocks of habituation training, separated by 20 min. Each block comprised 8 bouts of auditory pulses (120 pulses per bout at 1 Hz, 3 min interbout interval). Following habituation training there was a 1 h period, during which the larvae were unstimulated, followed by the Posttest. The Posttest was identical to the Pretest. Untrained (Test Alone) larvae were treated identically to the Trained larvae, except that they received no stimulation prior to the rest period. See the Materials and methods for additional details. (B) ITH of the C-start. The Trained group exhibited significantly fewer escape responses than did the Test Alone group (asterisk). (C) Effect of NMDA receptor blockade on ITH. The APV-Trained group exhibited significantly less habituation than the Ctrl-Trained group (asterisk). (D) APV exposure did not affect the baseline responsiveness of zebrafish larvae. There was no significant difference between the two groups of untrained larvae (APV-Test Alone and Ctrl-Test Alone) with respect to the number of C-starts elicited.

Figure 2

Habituation of the C-start in zebrafish larvae can persist for ≥ 18 h. (A) Experimental protocol for the fish that received habituation training. Here, posttests occurred at 1 h and 18 h after training. The Test Alone group did not receive habituation training but was otherwise treated identically to the Trained group. After the 1 h Posttest, larvae were left in the dark, unstimulated, for 15 h. Then the room lights were turned on for 1 h after which there was a second posttest. (B) LTH of the C-start. The trained larvae exhibited significantly greater habituation at both the 1 h and 18 h Posttests than did the larvae given the test alone treatment (asterisks).

Figure 3

LTH of the C-start in zebrafish expressed depends on NMDA receptor activity. (A) Experimental protocol for LTH that included only a single posttest at 18 h after training/test alone treatment. (B) Effect of NMDA receptor blockade on LTH. The larvae treated with APV (APV-Trained) exhibited significantly less habituation than larvae treated with the control solution (Ctrl-Trained) 18 h after the end of training (asterisk). (C) APV treatment alone did not alter the responsiveness of larvae. There was no difference in responsiveness between two untrained groups, one given the test alone protocol in control solution and another given the test alone protocol in APV.

Figure 4

LTH of the C-start can be rapidly reversed by a tactile stimulus. (A) Experimental protocol for demonstrating dishabituation of the C-start. Two groups of larvae were trained using the protocol shown in Figure 3A. Animals in the Dishabituation group received a brief touch (arrow) with a hand-held broom bristle 15 min prior to the Posttest. (B) Dishabituation following LTH. The touch dishabituated the C-start, as indicated by the comparison between the Dishabituation group (asterisk) and the group of larvae not given tactile stimulation prior to the Posttest (Habituation group).

Figure 5

LTH of the C-start is disrupted by cold shock. (A) Experimental protocol for testing the effect of cold shock on LTH. LTH training (Fig. 3A) was administered to two groups of larvae. One of the groups (Cold SH-Trained) received cold shock (exposure to control solution at 6–8°C for 2 min) immediately after training whereas the other group (Ctrl-Trained) did not. (B) Effect of posttraining cold shock on LTH. The expression of habituation at 18 h posttraining was blocked in the Cold SH-Trained larvae (asterisk). (C) Cold shock did not affect the baseline responsiveness of the larvae. There was no significant difference in the number of posttest C-starts evoked in untrained larvae that received the cold shock (Cold SH-Test Alone) and in untrained larvae that did not (Ctrl-Test Alone).

Figure 6

The transcriptional inhibitor DRB blocks LTH. (A) Effect of prolonged exposure to DRB on the baseline responsiveness of zebrafish larvae. The number of C-starts evoked in larvae exposed to DRB (75 μM, 150 μM or 300 μM) did not differ significantly from that in larvae exposed simply to the DMSO vehicle solution (0 μM). (B) Experimental protocol for testing DRB’s effect on LTH. The drug was applied to the individual wells (300 μM concentration) immediately after the end of the Pretest, and the Training/Test Alone period commenced 15 min later. The drug remained in the wells throughout the Training/Test Alone period after which it was washed out of the wells with control (E3) medium. The larvae remained in control medium alone for the remainder of the experiment. (C) Effect of inhibiting transcription on habituation expressed at 18 h posttraining. LTH was blocked in the trained fish treated with DRB (asterisk). (D) Effect of exposure to DRB on the baseline responsiveness of larvae. The DRB treatment did not affect the number of C-starts evoked in the Posttest in untrained larvae, as indicated by the lack of a significant difference in the HI at 18 h between the DMSO-Test Alone and DRB-Test Alone groups.