Fish are the largest and most diverse group of vertebrates, and like other vertebrates they exhibit complex cognitive abilities such as navigating, socializing, sleeping, hunting, etc. Nevertheless, the neural mechanisms governing fish behavior remain for the most part unknown.
In the past few decades, extracellular recordings from immobilized fish have primarily been implemented to investigate different aspects of the neural basis of behavior1,2. Although this technique is appropriate for some sensory systems, investigation of the full spectrum of the neural basis of behavior is difficult if not impossible in immobilized animals. The first advances involved recording from the Mauthner cells of tethered swimming fish3,4. However, Mauthner cells are disproportionately large and the recorded action potential amplitudes, which can go as high as a few mV, facilitate recording. Later, Canfield et al. described a proof of concept when using a tethered animal to record from the telencephalon of fish5. Another recent technique for recording neural activity from fish is calcium imaging (see reviews by Orger and de Polavieja6, and Vanwalleghem et al.7). This technique was developed for use with zebrafish larvae because the skin and skull are transparent during the larval stage. However, this technique cannot be used to study complex behaviors in later stages of development.
Here, we present a novel technique for recording extracellular neural activity from the brains of freely swimming fish. This is a modified version of the protocol described in Vinepinsky et al.8. The main innovation is the addition of a microdrive that makes it possible to control the position of the electrodes after surgery. The technique is designed for recording from the telencephalon of goldfish using a set of tetrodes that are connected to a neural data logger via a microdrive. The whole setup is wireless and anchored to the fish's skull. The specific weight of the system is equalized to the water-specific weight by adding a small float that allows the fish to swim freely.
The technique is based on the use of a neural data logger that amplifies, digitizes, and stores the signal in an onboard memory device. The logger telemetry system is used to start and stop the recordings, and for synchronization with the video camera. In this protocol, a 16-channel neural logger is used, embedded in a waterproof box together with the microdrive.
The microdrive assembly is fabricated from two main components: the microdrive itself and the microdrive housing (Figure 1A,B). The housing holds the microdrive and the tetrodes, and also acts as the anchor between the skull and the logger box (Figure 1C). The PVC logger box is fabricated using a machine process and is sealed using an O-ring (Figure 1E-G, see also Supplementary Figure 1, Supplementary Figure 2, and Supplementary Figure 3 for a three-dimensional [3D] diagram). At one end, a piece of polystyrene foam is attached to the logger box to compensate for the weight of the implant and provide the fish with a buoyancy-neutral implant. The construction of the microdrive described in the protocol follows the procedure presented by Vandecasteele et al.9 with a modification to attach the microdrive to the housing (Figure 1A). All major steps are presented.
The procedure described in the protocol to prepare the fish skull is similar to the one presented in Vinepinsky et al.8 and is described briefly in the protocol. One day after surgery, the fish are normally fully recovered from the effects of anesthesia and are ready for the behavioral experiments. Note that the tetrode location can be adjusted by turning the microdrive screw. The screw has a spacing of 300 µm per full rotation and an advancement of 75 µm is recommended until the target brain location is reached. An appropriate brain atlas should be consulted to target the specific brain region of interest. It is advisable to test the electrode impedance each time the fish is anesthetized for battery or memory card replacement.