Some voluntary C-bends may be Mauthner neuron initiated.

Predatory fish sometimes capture a prey fish first by striking it from the side, allowing the predator to consume the stunned prey head first. The rapid body flexion that the predator uses to stun its prey is similar to the "C" shaped maneuver ("C-bend") that many fish species use when performing a C-start escape response. For most species, one of the two Mauthner neurons initiates the C-start and, together with other reticulospinal neurons, their activity determines the extent of the bend and the ultimate trajectory of the fish. Reported here is initial evidence of previously undescribed behaviors where goldfish strike an object while executing voluntary C-bends that are similar to their C-start escape responses. The overlapping distributions of turn durations, turn angles, and angular velocities suggest that at least some voluntary C-bends are initiated by the Mauthner neuron. This implies that the Mauthner neuron can be activated voluntarily in the absence of predator- or feeding-associated releasing stimuli.

Dry beveling micropipettes using a computer hard drive.

It is sometimes useful in electrophysiological recordings to try various micropipette profiles in order to determine which tip works best in a given experiment. A pipette puller can be used to pull very sharp or blunt pipettes, and to fire polish tips for whole cell patch recordings. Broken tip pipettes can be "bumped" to an acceptable tip diameter under a microscope. However, it may be difficult to rationalize the purchase of a commercial beveling machine simply to test whether beveled pipettes are best for recording intracellularly from the cell types of interest. Presented here are methods that use a surplus computer hard drive to reproducibly dry bevel glass micropipettes. Compared to sharp or broken tip electrodes, pipettes dry beveled with this simple system are superior for making intracellular recordings from cichlid Mauthner neurons. Preliminary data obtained with this inexpensive apparatus may allow investigators to successfully justify the purchase of a commercial beveler.

Functional evidence for visuospatial coding in the Mauthner neuron.

When startled by sound, goldfish make large turns away from a rostral stimulus and small responses away from caudal stimuli, suggesting that rostral startling stimuli recruit larger pools of reticulospinal neurons in the Brainstem Escape Network (BEN) than do caudal stimuli. Consistent with this idea, the zebrafish Mauthner (M-) cell fires when the fish is startled by tail-directed stimuli, but the M-cell homologues (MiD2cm and MiD3cm) are also recruited when the fish is startled by displacing the head. Because vision is known to modulate M-cell activity, a nonstartling, modulatory sensory 'signal' conveyed to the reticular formation may be stronger if the visual sensory image is from a rostral vs. caudal spatial location and could account for a differential neuron pool recruitment and response magnitude. In this study, electrophysiological recordings from cichlid Mauthner neurons showed that visual stimulation of the caudal retina (by a rostral cue) generates a depolarization that is about 1.5 times the amplitude of that generated by stimulation of the rostral retina (by a caudal cue). In behavioral testing, where fish were stimulated visually for 30 ms and then startled by sound, fish startled in the presence of a rostral visual stimulus performed larger amplitude and faster turns than when startled in the presence of a caudal visual stimulus. Thus, M-cell potentials might reflect the strength of visual input to the BEN in general. For a particular visual spatial location, the relative strength of descending visual input appears to contribute to a recruitment of a reticulospinal neuron population that generates a turn magnitude appropriate to the visual cue, and suggests that a retinotopic representation is preserved in the BEN.

Parallel and interrelated neural systems underlying adaptive navigation.

The ability to process in parallel multiple forms of sensory information, and link sensory-sensory associations to behavior, presumably allows for the opportunistic use of the most reliable and predictive sensory modalities in diverse behavioral contexts. Evolutionary considerations indicate that such processing may represent a fundamental operating principle underlying complex sensory associations and sensory-motor integration. Here, we suggest that animal navigation is a particularly useful model of such opportunistic use of sensory and motor information because it is possible to study directly the effects of memory on neural system functions. First, comparative evidence for parallel processing across multiple brain structures during navigation is provided from the literatures on fish and rodent navigation. Then, based on neurophysiological evidence of coordinated, multiregional processing, we provide a neurobiological explanation of learning and memory effects on neural circuitry mediating navigation.

Methods for chronic neural recording in the telencephalon of freely behaving fish.

We have adapted for use in fish several of the procedures employed for recording single neuron activity in freely behaving rodents. Developing a method for single unit chronic recording in freely behaving fish was motivated by a need for a comparison across taxa of telencephalic neural activity evoked during spatial navigation by animals of their environments. However, the procedures outlined here can be modified easily for underwater recording from most aquatic species and from other brain areas. Under anesthesia, bundles of stereotrodes or tetrodes were implanted into the dorsolateral region of the goldfish or cichlid telencephalon. An infrared light emitting diode (LED) was also fixed to the fish's head at the time of surgery. After recovery from anesthesia, fish were allowed to swim freely within a large aquarium. Single unit activity was analyzed and correlated with stimulus conditions, behavior, and the location and movement of the LED recorded by a camera tracking system. The value of this technique is demonstrated by providing the first evidence in fish for navigation-related neural firing, including "place cells" that display location-specific discharge.

Temporal constraints on visually directed C-start responses: behavioral and physiological correlates.

To study the modulatory influences of visual information on Mauthner (M-) initiated C-start responses, short interval visual stimuli were presented to individual cichlid fish prior to being startled with a sound pulse. Because the axon of each Mauthner neuron activates trunk musculature contralateral to the soma, the initial direction of an ensuing startle response provides a behavioral measure of which cell has been driven closer to threshold by visual cues at the time the sound pulse causes one cell to fire. When an LED was illuminated on one side of the fish for 10 ms prior to a startling sound pulse, cichlids reliably turned toward the visual cue. At durations of 15 and 20 ms, fish turned away from the same stimulus. Thus, behavioral evidence suggests that the M-cell contralateral to a visual stimulus appears to be excited first but visual excitation of the M-cell ipsilateral to the visual stimulus follows and predominates. Consistent with the behavioral results, visually evoked excitatory potentials recorded intracellularly in the cichlid M-cells were complex, with initial PSPs showing latencies of about 11.6 ms from contralateral eye stimulation and 15.5 ms from stimulating the ipsilateral eye. PSP latencies in goldfish were longer and more similar for stimuli to the two eyes (about 22 ms). For contrast, sound-evoked PSPs begin within 2 ms. The relative long latencies for visually regulating M-cell function suggest that vision is most adaptive for biasing response direction prior to rather than during a predator's attack.