The systemDrive: a Multisite, Multiregion Microdrive with Independent Drive Axis Angling for Chronic Multimodal Systems Neuroscience Recordings in Freely Behaving Animals

Targeting the sleep–wake regulatory network. a, A sagittal brain section with structures implicated in sleep–wake regulation. The structures are distributed across multiple regions of brainstem. Color codes represent the postulated states of vigilance during which specific cell populations within each structure are predominantly active: red is non-REM (NREM) sleep; blue is REM sleep; and green is wake. Brain region names and acronyms are as follows: ventrolateral pre-optic nucleus (VLPO), laterodorsal tegmental nucleus (LDT), pedunculopontine tegmental nucleus (PPT), dorsal raphe (DR), and locus coeruleus (LC). b, Coronal brain sections that highlight SWRS populations DR and LDT underneath sensitive anatomic structures highlighted in orange. These dorsal structures should not be damaged during chronic recording studies, and so safe access to the targets requires independent and angled electrode trajectories. Labels are as follows: vein (v), pineal gland (Pi), cerebral aqueduct (Aq), and fourth ventricle (4V). Depth information is in millimeters with horizontal tics representing 1 mm. Coronal sections were adapted with permission from Paxinos and Watson (2007, their Figs. 95 and 104; copyright Elsevier).

The flexible drive axis. a, Diagram of the FDA. The FDA has a long flexible body tube that is fixed at the top to the microdrive and has a free end with a placement tube. The free implant end is inserted into a skull-fixed guide cannula contralateral to the microdrive fixed point. This creates a second fixed point on the axis. The constrained geometry allows the electrode bundle to move precisely through the tubing and to exit the placement tube at the angle set by the guide cannula. b, Placement of FDAs to multiple targets from multiple implant sites. The independent angling of each FDA provides maximal spatial recording coverage of targets while avoiding sensitive brain structures. During FDA positioning, the placement tube is slid into the guide cannula (diameter, 226 µm). The FDA is positioned once the body tube meets the top of the guide cannula. Depth information is in mm with horizontal tics representing 1 mm. Width-scaled guide cannulas (front section, *; back section, ◊) have been included to better contextualize potential damage to brain from the implant. FDA components on both subpanels are not drawn to scale. Coronal sections were adapted with permission from Paxinos and Watson (2007, their Figs. 95 and 104; copyright Elsevier).

Stereotaxic targeting diagram. Diagram of the FDA stereotaxic targeting method and the stereotaxic unit (STU). a, The ventral end of the FDA can be implanted at a user-defined absolute driving distance Δ away from a target, indicated by the red X, along any arbitrary trajectory. This is achieved by using the STU to carry the guide cannula to the correct coordinates and implant depth. The top of the guide cannula is the same diameter as the body tube (8.9 mil) and thus sets the stopping point of the placement tube (6.3 mil). The difference between the length of the placement tube R and the length from the top of the guide cannula to the tip of the STU α is known by subtracting the two premeasured values. Thus, ventral depth of the FDA is guaranteed by implanting the STU a distance ζ from the target, which is calculated using the Distance to Target equation. b, A series of photographs showing the STU without and with a guide cannula attached. The STU is a surgical stylus that contains a spacer tube (6.3 mil) and a stopper tube (8.9 mil). The guide cannula is slid onto the STU until it reaches the stopper tube. During surgery, the STU is implanted, the guide cannula is fixed to the skull, and then the STU is removed, which leaves just the guide cannula in place. Red scale bar, 2 mm.

Microdrive system details. a, The FDA is made of a tubing structure that encloses a microwire electrode bundle as it is pushed through the drive axis by a drive spring. The shuttle tube is rigidly attached to a bridge on the head of the drive spring and can move through the body tube. As the drive spring is moved up or down with the turning of a drive screw, the shuttle tube and electrodes move with the spring head. The body tube is fixed to the microdrive body with cyanoacrylate. The body tube is very long, as marked by the black break lines. The placement tube is fixed within the bottom of the body tube. Petroleum jelly space fills the ventral end of the placement tube, except at the electrode tips, to prevent fluids and proteins from entering the drive axis and binding the electrode bundle. b, c, Isometric (b) and side (c) 3D-rendered computer-aided design (CAD) model views of the complete microdrive system to scale. Elements of the system include the following: the microdrive body (light gray structure), drive springs and screws, FDAs, an electrode interface board (green printed circuit board) with a 36-position Omnetics connector, an Intan RHD2216 amplifier, and a headmount (dark gray base and cap). The cap of the headmount has been sliced to view the elements inside. d, Top, front, side, and isometric 3D-rendered CAD model views of the microdrive body. The four primary features on the microdrive body are: the drive positions, the screw columns, the base feet, and the spring wedges. The two drive position arrays have tap holes for 000-120 threads and are angled at 20° with respect to the level screw columns. The 20° provides clearance for the screwdriver when driving and allows the long flexible implants to naturally extend over midline, as depicted in the side view of c. The high-density design has 14 usable drive positions to allow for the targeting of implant sites along an extended rostrocaudal range. Each individual drive position is 2.5 mm in width and includes a drive axis hole for fixing the drive axis bundles to the microdrive body. The two screw columns have 00-90 tap holes and are spaced apart by 20 mm. The feet of the microdrive have 00-90 through-holes and are where the microdrive body is secured to the headmount base. The 1 mm tall spring wedge is on the outside of the drive position arrays and is angled at 20°. The microdrive body is 20 mm in length, 16 mm in height, and 18.68 mm in width. e, Isometric 3D-rendered CAD model views of the headmount. The headmount is composed of two components: the base and the cap. The base has a 4-40 tap hole for the cap screws and a smaller 0.66 mm hole that gets enlarged and tapped with 00-90 threads for the foot screws that secure the microdrive body to the headmount base. The base has a centered 15.01 × 11.46 mm opening for access to the cranium, which allows the placement of multiple guide cannulas and extrasensor modalities. The opening has sloped side and back walls to provide extra room for the maneuverability of forceps when placing the microdrive onto an animal. The 3D CAD models for the Omnetics Nano Strip connector and Intan RHD2216 amplifier were downloaded from the Omnetics (part A79026-001, http://www.omnetics.com/products/neuro-connectors/nano-strip-connectors) and Intan (http://intantech.com/downloads.html) web sites.

Microdrive placement. a, b, Photographs of a fully constructed microdrive with four drive axes. c–i, Chronological photographs from a microdrive placement procedure with the following four nonparallel, separated drive axis targets: VLPO, DR, PPT, and LDT. c, The microdrive is attached to a custom 3D-printed placement tool, which can be manipulated by the stereotaxic device. d, Petroleum-jelly plugs, highlighted by the four black arrows, are first removed from the skull-fixed guide cannulas. Leads of the accessory electrodes, which include cortical screws and hippocampal depth electrodes, are taped down to the side of the headmount base for later riveting. e, f, The microdrive is lowered with the stereotaxic device, and the placement tubes of each drive axis are inserted into the skull-fixed guide cannulas, one of which is highlighted with a black arrow in f. The placement tubes are guided using customized forceps that have small notches near the tips to grab and manipulate the tubes without crushing them. g, All four placement tubes are engaged with their respective guide cannulas. h, The placement tubes are closer to being in complete position. Once the drive axes are in place, the bottom of the body tube, which is highlighted with black marker, will be seated on top of the guide cannula. i, After all the drive axes are seated, the microdrive is secured to the headmount base with the foot screws and the placement tool is removed. Finally, the cortical screw and depth electrode leads are riveted into the EIB.

Targeting accuracy across multiple subjects. Coronal histologic sections taken from a total of four different chronically recorded subjects that demonstrate FDA targeting accuracy and repeatability from multiple brainstem areas. The target structures displayed include the following: bilateral PPT, bilateral LDT, and DR. Implant coordinates and angling for targets are in Table 3. The left column contains histology from three separate subjects, all of whom had other FDA sites. The right column contains histology from a single subject. a–d, NADPH-stained coronal sections that show electrode tracks at bilateral PPT regions (targeting parameters for left column, right column in mm: R = 9.17, 9.28; α = 7.21, 7.41; Δ = 1). e–h, NADPH-stained coronal sections that show electrode tracks at bilateral dorsal and ventral LDT regions (R = 9.20, 9.28; α = 7.16, 7.41; Δ = 2). i–l, Nissl-stained coronal sections that show electrode tracks at the DR nucleus (R = 8.83, 9.28; α = 7.02, 7.41; Δ = 1, 1.25). The red boxes indicate approximate areas from whole-slice images, which are zoomed-in on subsequent images. The yellow dashed lines indicate the estimated boundary of target structures. The red arrows indicate the estimated end of electrode tracks. Scale bars: a, c, e, g, h, i, j, k, l (4× magnified images), 1000 µm; b, f (10× magnified images), 100 µm; d (10× magnified images), 200 µm.

Chronic recording stability. a, b, For an hour-long block of data background, the RMS of the time series from on the microwire in dorsal raphe (a) and one accelerometer channel (b) are computed over 5 s long overlapping windows (overlap = 1 s). c, The RMS background of the microwire channel remains flat with respect to the accelerometer activity during the hour. The maximal range of the background noise (∼7 µV) is much smaller than the amplitude of the action potentials recorded on that channel (c, right).

Simultaneously recorded single units. Example single unit action potentials simultaneously recorded from up to three target SWRS structures. Units across (n = 8) subjects were classified as state dependent by observing changes in firing rates during and outside of characteristic sleep–wake states. The subject numbers correspond to those in Table 4, but the order has been rearranged for the sake of grouping units from the same structures. The y-axis and x-axis of each scale bar represent 50 µV and 1 ms, respectively.

Chronic single-unit stability. Extracellular action potentials recorded simultaneously from PPT, LDT, and DR over a continuous 7 d recording session. Waveforms were averaged over 12 h periods to correspond with light (black traces) and dark (orange traces) cycles, which were separated because of the change in time-of-day dependence of the firing rates. The x-axis is days starting from onset of recordings in a particular recording session. The number subscripts for each region represent the group electrode channel on which each unit was recorded. Data missing from part of day 3 were the result of a computer acquisition failure. Scale: 1 ms.