A cranial implant for stabilizing whole-cell patch-clamp recordings in behaving rodents

Abstract

Background: In vivo patch-clamp recording techniques provide access to the sub- and suprathreshold membrane potential dynamics of individual neurons during behavior. However, maintaining recording stability throughout behavior is a significant challenge, and while methods for head restraint are commonly used to enhance stability, behaviorally related brain movement relative to the skull can severely impact the success rate and duration of whole-cell patch-clamp recordings.

New method: We developed a low-cost, biocompatible, and 3D-printable cranial implant capable of locally stabilizing brain movement, while permitting equivalent access to the brain when compared to a conventional craniotomy.

Results: Experiments in head-restrained behaving mice demonstrate that the cranial implant can reliably reduce the amplitude and speed of brain displacements, significantly improving the success rate of recordings across repeated bouts of motor behavior.

Comparison with existing method(s): Our solution offers an improvement on currently available strategies for brain stabilization. Due to its small size, the implant can be retrofitted to most in vivo electrophysiology recording setups, providing a low cost, easily implementable solution for increasing intracellular recording stability in vivo.

Conclusions: By facilitating stable whole-cell patch-clamp recordings in vivo, biocompatible 3D printed implants should accelerate the investigation of single neuron computations underlying behavior.

Fig. 1

Movement-related loss of whole-cell recordings in vivo. (a) Left to right: schematic diagrams depicting approach of patch pipette, whole cell recording configuration and movement-related loss of seal integrity. (b) Top, example motion index trace describing gross movement of the mouse forelimb. Middle, membrane potential recording of a layer 5B pyramidal neuron in the caudal forelimb area of primary motor cortex. Bottom, schematic diagrams depicting forelimb movement during a cued forelimb lever push task. Red dashed line, movement initiation. Vertical scale bars, 1 arbitrary unit (grey, AU) and 20 mV (black), horizontal scale bar, 2 s.

Fig. 2

Implant design. (a) Dimensioned side- and (b) isometric view of the implant design. Measurements are in mm. (c) Four 3D printed implants placed on a UK one pence piece.

Fig. 3

Surgical implantation. (a) Left, Image depicting the outline of a 3 mm glass coverslip, centred above the region of interest. Right, etching the circumference using a 26-gauge needle. 1 – dental cement affixing headplate to skull; 2 – exposed skull; 3 –glass coverslip. Scale bars, 1 mm. (b) Left, Image of the craniotomy outline. Right, a handheld drill with ⊘ 0.6 mm burr used to generate the craniotomy. (c) Left, Image of the craniotomy with central bone area removed (dashed area). Right, fine forceps used to remove bone while maintaining constant irrigation with saline. (d) Left, Image of the durotomy (dashed area). Right, 30-gauge needle and fine forceps are used to remove dura above targeted region. (e) Left, Placement of the implant. Right, implant is aligned to the craniotomy with central well filled with agar. (f) Left, Silicon sealed cranial insert. Right, Coarse forceps are used to press down and align the implant with the surface of the skull. After drying the remaining saline, gel superglue is used to affix the implant in place and silicon is added to the central well.

Fig. 4

Implant reduces speed and amplitude of brain displacements. Two-photon imaging through (a) a conventional craniotomy or (b) an implant. (c) Top, Example 2-photon imaging field-of-view from layer 2/3 in CFA (average time projection of 1 s of raw data, peri-movement initiation; scale bar, 50 µm). Bottom, expanded view of cyan square showing example L2/3 interneuron before and after (average time projection of 4 frames or 100 ms of raw data) the onset of movement. Note xy shift in location (cyan dotted lines). Bottom scale bars, 10 µm. (d) Example trace showing brain displacements from the point of origin across time in the absence (top trace) and presence (bottom trace) of an implant. Horizontal scale bars, 2 s; vertical scale bars, 2 µm. (e) Expanded view of grey dashed rectangle in (d) showing the time (t) and distance (d) of an individual displacement. Speed (s) = distance (d) / time (t). Horizontal scale bar, 200 ms; vertical scale bar, 1 µm. (f-h) Cumulative probability plots showing the distribution of brain displacement amplitudes using a conventional craniotomy with dura removed (teal, N = 3 mice) or implant (orange, N = 3 mice) at 3 different depths from the pial surface (100–140 µm, 300–340 µm, 480–520 µm). (i-k) Cumulative probability plots showing the distribution of brain displacement speeds using a conventional craniotomy with dura removed (teal, N = 3 mice) or an implant (orange, N = 3 mice) at 3 different depths from the pial surface (100–140 µm, 300–340 µm, 480–520 µm).

Fig. 5

Cranial implant improves whole-cell recording stability. (a) Schematic showing patch-clamp recording through a conventional craniotomy. (b) Schematic showing patch-clamp recording through an implant. (c) Violin plots showing median (white circle), mean (thick horizontal line), inter quartile range (thick grey vertical line) and range (thin grey vertical line) of resting membrane potentials (V_rest) using a conventional craniotomy or implant. Dots represent data from individual mice (N = 28 and 59 mice, respectively). (d) Violin plots showing median (white circle), mean (thick horizontal line), inter quartile range (thick grey vertical line) and range (thin grey vertical line) of series resistance measured during whole cell recording using a conventional craniotomy or implant. Dots represent individual recordings from individual mice (N = 28 and 59 mice, respectively). (e) Membrane potential recording from a layer 5B pyramidal neuron using a conventional craniotomy. Green vertical bars represent lever pushes. Black vertical scale bar, 10 mV; horizontal scale bar, 5 s (f) Membrane potential recording from a layer 5B pyramidal neuron using a cranial implant. Green vertical bars represent lever pushes. Black vertical scale bar, 10 mV; horizontal scale bar, 5 s (g) Bar graph showing the probability of maintaining a stable whole-cell recording configuration during behavior when using a conventional craniotomy or implant.