Intracellular neural control of an active feeding structure in Aplysia using a carbon fiber electrode array

Abstract

Background: To study neural control of behavior, intracellular recording and stimulation of many neurons in freely moving animals would be ideal. However, current technologies limit the number of neurons that can be monitored and manipulated. A new technology has become available for intracellular recording and stimulation which we demonstrate in the tractable nervous system of Aplysia.

New method: Carbon fiber electrode arrays (whose tips are coated with platinum-iridium) were used with an in vitro feeding preparation to intracellularly record from and to control the activity of multiple neurons during feeding movements.

Results: In an in vitro feeding preparation, the carbon fiber electrode arrays recorded action potentials and subthreshold synaptic potentials during feeding movements. Depolarizing or hyperpolarizing currents activated or inhibited identified neurons (respectively), manipulating the movements of the feeding apparatus.

Comparison with existing method(s): Standard glass microelectrodes that are commonly used for intracellular recording are stiff, liable to break in response to movement, and require many micromanipulators to be precisely positioned. In contrast, carbon fiber arrays are less sensitive to movement, but are capable of multiple channels of intracellular recording and stimulation.

Conclusions: Carbon fiber arrays are a novel technology for intracellular recording that can be used in moving preparations. They can record both action potentials and synaptic activity in multiple neurons and can be used to stimulate multiple neurons in complex patterns.

Keywords: Aplysia; Carbon fiber electrodes; Intracellular recording; Intracellular stimulation.

Fig. 1.

Animal surgery. (A) An Aplysia californica was anesthetized and pinned dorsal side up. The skin was cut open to reveal the nervous system and the buccal mass (red). Cyan dashed lines indicate the location of the cut. (B) The cerebral–pedal connective and cerebral–pleural connective were severed on both sides, indicated by the cyan dashed lines. The esophagus was also cut. (C) The esophagus was grasped and lifted up. Tissues around the buccal mass, the salivary glands and the buccal artery were cut. (D) The buccal mass was isolated with cerebral ganglia and buccal ganglia.

Fig. 2.

Dish setup, positioning of the buccal mass and cerebral ganglia. (A) The dish had five sections: a back compartment, a front compartment, a narrow platform, two circular holes in the front compartment, and a connecting section between the back and front compartments with a notch in the middle. (B) The buccal mass was treated with a slow injection of 1 mL Aplysia saline to help remove the MgCl₂ from inside the core muscle; care was taken not to inject bubbles. (C) The cerebral ganglia were pulled to the back chamber. The notch was completely sealed with Vaseline. The buccal mass was placed on a Sylgard block, with two silk sutures through the anterodorsal tissue of the jaw and the anteroventral tissue of the jaw.

Fig. 3.

The pinning of the buccal ganglia with its rostral side up. The anterior side is towards the sensory neuron cluster of the buccal ganglia, and the posterior side is towards the motor neuron cluster. (A) Special arch pins were made to keep the buccal ganglia stable during the experiment. (B) The buccal ganglia were pinned to the narrow platform of the dish. The letter “i” stands for ipsilateral side of the buccal mass relative to the nerve attachment; the letter “c” stands for contralateral side relative to the nerve attachment. (C) The procedure of desheathing is shown in sequence. To desheath the buccal ganglia, the dish was rotated 180° so that the incision could be made at the sensory neuron cluster side. After desheathing, the dish was rotated back.

Fig. 4.

The setup of the recording dish. (A) The buccal mass was suspended in the dish. The aerator was placed in the dish. (B) The carbon fiber electrode was inserted into the buccal ganglia with an angle of 60° relative to the solution surface. (C) The full setup of the experiment.

Fig. 5.

Intracellular recording of a biting pattern in a semi-intact preparation. (A) The neural activity of a biting pattern was recorded by a carbon fiber electrode array. Fibers 3, 4, 5, 6, 9 and 13 recorded from six different neurons intracellularly. Fiber 8 recorded a quasi-intracellular action potential from the seventh neuron. Nerve recordings of buccal nerve 2 (BN2) and buccal nerve 3 (BN3) are shown at the bottom (purple). The dashed lines correspond to a series of screenshots of the buccal mass movement at different times. The movements are (1) biting onset, (2) jaw opening and radula protraction, (3) fully open radula (arrow), (4) radula retraction and closure, (5) fully retracted radula and (6) return to rest. (B) Neuron recorded by fiber 5 projects its axon into BN3 as a medium-sized unit (note one-to-one relationship with fixed latency).

Fig. 6.

Subthreshold synaptic activity was recorded in an actively feeding buccal mass by a CFE array. (A) Synaptic activity impinging on the B3 neuron during several feeding patterns. Fiber 9 recorded intracellular signals from the B3 motor neuron (B3 was identified by the one-to-one match between soma action potential and the largest unit on BN2, indicated by the thicker gray dashed lines in B; three action potentials in B3 occur near the end of the record shown from Fiber 9, and correspond to large units in BN2). The light gray bar at the top indicates the protraction phase, and the dark gray bar indicates the retraction phase. The recording shows two complete feeding patterns in sequence (marked as 1 and 2) and the beginning of the protraction phase of a third feeding pattern (marked as 3). (B) The synaptic interactions between the B3 neuron and the B4/B5 neuron could be observed. The inhibitory synaptic potentials in B3 are matched one-to-one with the B4/B5 action potentials (yellow action potentials, recorded on Fiber 15). The large unit action potentials on the BN3 that are projections from the other B4/B5 neuron also showed a one-to-one inhibitory synaptic activity in B3 neuron (light gray dashed lines). (C) The averaged spikes of B4/B5 action potentials and the averaged signals recorded by fiber 9 are shown superimposed on one another. The strong signal of the B4/B5 action potential was also recorded extracellularly by Fiber 9 (and across other carbon fibers; data not shown), shown as an initial dip in the recording on Fiber 9. The inhibitory effect of B4/B5 on B3 is clear.

Fig. 7.

Stimulation of motor neuron B3 activated the neuron and caused an additional contraction of I3 muscle on the ipsilateral side of the buccal mass during a swallow. (A) Recording of a swallowing pattern without any stimulation. A sequence of screenshots of the buccal mass was recorded at four time points to show (1) the beginning of a swallow, (2) the expansion of the lumen during the protraction phase, (3) the contraction of I1/I3 muscle during the normal retraction phase and (4) return to rest. Two schematic drawings of the buccal mass for states 2 and 3 are shown on top to clarify the expansion of the jaw lumen and its subsequent contraction before and after the retraction phase. (B) Recordings of a swallowing pattern with stimulation of motor neuron B3. The stimulation was sent through fiber 10, and the stimulation period was shown above fiber 4 recording (bar). The stimulation protocol was 100 ms, 200 nA of inhibitory current (light gray bar; indicated by a gray arrow above fiber 4) followed by 2 s, 200 nA excitatory current (dark gray bar). The increase in B3 activity recorded on fiber 9 is indicated by a black arrow pointing to fiber 9, and a black arrow pointing to the projecting action potential on BN2. A sequence of screenshots of the buccal mass was recorded at five time points to show (1) the beginning of the swallow, (2) the expansion of the lumen during the protraction phase, (3) the contraction of the lumen during the retraction phase, (4) the additional contraction after the B3 stimulation and (5) the return to rest. Three schematic drawings of the buccal mass for states 2, 3 and 4 are shown on top to clarify the response of the buccal mass. Comparing A and B, it is clear that an extra contraction (indicated by the arrow in schematic drawing) was observed as a result of the earlier B3 stimulation (indicated by the black arrows in the recordings).

Fig. 8.

Stimulation of a multi-action neuron B4/B5 inhibited the neural activity and released the inhibition on the neuron recorded by Fiber 7. (A) Recording of a feeding pattern without any stimulation. A sequence of screenshots of the buccal mass was taken at five time points to show the following activities: (1) the beginning of a swallow, (2) the maximum expansion of the lumen at the end of the protraction phase, (3) the retraction of the radula with a relaxation of the lumen, (4) the maximum contraction of I1/I3 muscles during the retraction phase, and (5) return to rest. The neuron recorded by fiber 7 burst as the B4/B5 gradually stopped firing (indicated by the arrow under the fiber 7 recording). (B) Recording of a feeding pattern with the inhibition of the multi-action neuron B4/B5. The current was sent through fiber 2, and the stimulation period was shown above the fiber 4 recording (bar). The stimulation protocol was 2 s, - 200 nA current (gray bar). The neuron recorded by fiber 7 burst when B4/B5 was inhibited (indicated by the black arrows in the recordings). A sequence of the screenshots of the buccal mass was shown at the same five time points as A. Compared to A, the firing duration of the B4/B5 neuron was reduced due to the inhibition. The duration between the activities captured by screenshots (3) and (4) was also reduced (compare the length of black bars at the bottom), suggesting that the contraction of the I1/I3 muscles occurred slightly earlier due to the inhibition of B4/B5.