Albert Lee, Jerome Epsztein, Michael Brecht

Extracellular single unit recording has been the key technique for revealing cellular correlates of behavior in awake and freely moving animals. However, our knowledge about cellular mechanisms underlying behavior could be greatly extended by intracellular recordings. Such recordings measure not only the action potential output of a neuron, but also its subthreshold synaptic inputs and fluctuations, spike threshold, and input resistance. Furthermore, intracellular recording allows manipulation of the cell's membrane potential through direct current injection, and filling of the cell for subsequent anatomical reconstruction of its location and geometry.

To obtain whole-cell recordings in freely moving rats, two major technical problems must be solved. (i) The recording setup must be miniaturized, because standard whole-cell recording equipment is far too large and heavy to be mounted on and carried by a rat. (ii) The recordings must be stabilized, because maintaining intracellular recordings requires a very mechanically stable preparation. Due to these problems, intracellular recording had thus far been limited to in vitro and head-immobilized in vivo experiments.

We have designed a miniature head-mountable whole-cell recording device that can be carried by a small (50-100 g) rat. Mechanical stabilization is achieved by anchoring the recording pipette rigidly in place after the whole-cell configuration has been established. This head-anchored pipette arrangement allows long duration (average 20 min, maximum > 1 hr) recordings in freely moving animals that are remarkably insensitive to mechanical disturbances, followed by anatomical reconstruction of the recorded cells. This technique will allow a wide range of new studies involving detailed measurement and manipulation of the physiological properties of identified cells during behavior.

Figure legend: Whole-Cell Recording of a Hindlimb Motor Cortex Neuron in a Freely Moving Rat.

(A) Reconstruction of the axonal (blue) and dendritic (red) arbors of this layer 3 pyramidal neuron.

(B) Top view of the behavioral arena, showing the trajectory of the rat�s head position for the entire 1 hr duration of this recording (all colors together).

(C) Membrane potential (black, top) over a 5 min period during which the rat moved freely around the arena, and the corresponding speed of head movement (blue, bottom). (D) An example action potential (AP).

(E) Subthreshold membrane potential trace (black, top) over a 1 s period during which

the rat ran at high speed (blue, bottom) along the light blue segment shown in (B).

M1 = primary motor cortex, S1 = primary somatosensory cortex, L1�L6 = cortical layers 1 through 6, WM = white matter, CC = corpus callosum.

(from Lee, Manns, Sakmann, Brecht, Neuron 51: 399-407, 2006).