We present the construction of a 3D-printable hyperdrive with eighteen independently adjustable tetrodes. The hyperdrive is designed to record brain activity in freely behaving rats over a period of several weeks.
Monitoring the activity patterns of a large population of neurons over many days in awake animals is a valuable technique in the field of systems neuroscience. One key component of this technique consists of the precise placement of multiple electrodes into desired brain regions and the maintenance of their stability. Here, we describe a protocol for the construction of a 3D-printable hyperdrive, which includes eighteen independently adjustable tetrodes, and is specifically designed for in vivo extracellular neural recording in freely behaving rats. The tetrodes attached to the microdrives can either be individually advanced into multiple brain regions along the track, or can be used to place an array of electrodes into a smaller area. The multiple tetrodes allow for simultaneous examination of action potentials from dozens of individual neurons, as well as local field potentials from populations of neurons in the brain during active behavior. In addition, the design provides for simpler 3D drafting software that can easily be modified for differing experimental needs.
In the field of systems neuroscience, scientists study the neural correlates underlying cognitive processes such as spatial navigation, memory, and decision-making. For these types of studies, it is critical to monitor the activity of many individual neurons during animal behavior. Over the past decades, two important advances have been made to meet the experimental needs for extracellular neural recording in small animals1,2,3. First was the development of the tetrode, a bundle of four microwires used to record neural activity of neurons simultaneously1,2,4. The differential signal amplitudes of activity across the four channels of a tetrode allows for the isolation of individual neuron activity from many simultaneously recorded cells5. In addition, the flexible nature of the microwires allows greater stability of the tetrode minimizing the relative displacement between the tetrode and the target cell population. Tetrodes are now widely used instead of a single electrode for many brain studies in various species, including rodents1,2,6, primates7, and insects8. Second was the development of a hyperdrive carrying multiple independently movable tetrodes, which allows for the simultaneous monitoring of neural activity from larger populations of neurons from multiple recording locations3,9,10,11,12.
The availability of a reliable and affordable multi-tetrode recording device for small animals is limited. The classic hyperdrive, initially developed by Bruce McNaughton13, has been successfully used for neural recordings in freely behaving rats in many labs in the past two decades9,10,14,15. However, for technical reasons, the original components needed to build the McNaughton drive are now very difficult to obtain and are not compatible with recently improved data acquisition interfaces. The other well accepted design of hyperdrive requires the microdrives to be individually handcrafted, which could yield inconsistent results and consume substantial time12. In order to record neural activity from various brain regions in behaving rats, we developed a new hyperdrive using stereolithographic technology. We sought to satisfy the following requirements: (1) the new hyperdrive must allow precise displacement of tetrodes in the brain and provide stable recording from multiple target regions; (2) the new hyperdrive must be compatible with the magnetic quickclip system recently developed to allow easy connection; and (3) the new hyperdrive can be accurately reproduced with materials easily available. Here, we provide a technique for building the 3D-printable hyperdrive containing eighteen independently movable tetrodes, based upon the McNaughton design. In the protocol, we describe the details of the fabrication process of the new hyperdrive, which we have used successfully to record single-neuron action potentials and local field potentials from the postrhinal and medial entorhinal cortices over weeks in a freely behaving rat during natural foraging tasks.
1. Stereolithography of 3D Models
2. Preparation of Accessories (Figure 2).
3. Preparation of the Hyperdrive Components (Figure 3).
4. Assembly of the hyperdrive (Figure 4).
We used a newly built hyperdrive to obtain trial results. The drive was equipped with tetrodes constructed from ø 17 µm (0.0007"), polyimide-coated platinum-iridium (90%-10%) wire. The tips of the tetrodes were plated in platinum black solution to reduce electrode impedances to between 100 and 200 kΩ at 1 kHz. The hyperdrive was implanted 4.6 mm left of the midline and 0.5 mm anterior to the transverse sinus on the skull of a 550 g, male Long-Evans rat. Additional ground wires were connected to skull screws over the cerebellum. All procedures were performed as approved by the Institutional Animal Care and Use Committee (IACUC) of Baylor College of Medicine and were similar to those previously described18. Immediately following the surgical implantation, the tetrodes were advanced 1 mm into the brain. On subsequent days, smaller advanced increments of no more than 80 µm were used. The tetrodes were allowed to stabilize after each advancement for at least 20 h before neural recordings were performed.
To record neural activity, the hyperdrive was connected to a headstage pre-amplifier (Neuralynx, HS-72-QC), and the latter was connected to a data acquisition system with programmable amplifiers (Neuralynx, Digital Lynx SX). Local field potentials were referenced to the ground wire, sampled at 2 kHz, and band-pass filtered at 0.1—500 Hz. Unit activity was referenced to a tetrode with no observable activity located 500 µm from the brain surface, sampled at 32 kHz, and band-pass filtered at 600 Hz-6 kHz. Only spike waveforms above a threshold of 50 µV were recorded.
Figure 5A illustrates neural activity recorded from a tetrode located in the postrhinal cortex (2.1 mm below the brain surface), while the animal was freely foraging inside a 1.5 m open box three weeks after implantation. The recording session lasted approximately 30 min and the units recorded remained stable across the entire session (demonstrated by the small variation in spike waveforms). Figure 5B shows local field potentials recorded simultaneously from four different tetrodes located in the medial entorhinal cortex (3.4-3.7 mm deep) while the same animal was actively exploring the open arena seven weeks after implantation. Clear field potential activity in the theta frequency range (6-10 Hz) was present. Individual neuron spike data was isolated using the sorting software MClust (A.D. Redish), and local field potential data was visualized by custom-written Matlab scripts. Examples of low quality tetrode recordings, possibly resulting from a poorly prepared drive, have been shown previously17.
Figure 1: Hyperdrive components created by stereolithographic technology. Image of the 3D-printable hyperdrive components (1¢ coin for size comparison). (A) the hyperdrive core; (B) the protective shield; (C) the protective cap; (D) the central column; (E) the nut; (F) the spacer; (G) the shuttle; (H) the shuttle bolt. Scale bar: 1 cm. These components were created by a UnionTech RSPro450 printer using the plastic material Somos EvoLVe 128. Please click here to view a larger version of this figure.
Figure 2: Custom-designed accessories for hyperdrive construction. These accessories were designed specifically to aid in the preparation of the hyperdrive. Their main components were created by stereolithographic printing. (A) The shuttle bolt holder, which secures the shuttle bolt while tapping the threads. (B) The microdrive assembly station, which guides the cannula insertion into the shuttle. (C) The microdrive rack, which helps to test the assembled microdrives and holds them in place while gluing the cannulas. 1: a microdrive rack base; 2: a microdrive rack with screws fully inserted in the slots; 3: a microdrive rack ready for use. (D) The nut holder, which holds the hyperdrive nut when threading the hole. (E) The hyperdrive core station, which secures the core while hammering the guide rods. (F) The turning tool, which drives the shuttle bolt to rotate in the shuttle. (G) The hyperdrive holder, which helps to place the hyperdrive under a stereoscope. The holder also protects the tetrodes after they have been loaded into the hyperdrive. (H) The rod positioning complex, which helps to position the threaded rods and guide rods in the hyperdrive core. 1: major components of the complex; 2: the upper portion of the complex after assembly; 3: a rod positioning complex in use. Please click here to view a larger version of this figure.
Figure 3: Preparation of the hyperdrive parts before assembly. Images showing the preparation process of the hyperdrive core and the microdrive, as well as other prepared hyperdrive parts. (A) A threaded hyperdrive nut. (B) Preparation of the hyperdrive core. 1: the core with external threads created for the nut; 2: the core placed in a core station with screws fully inserted in the slots; 3: guide rods positioned by the rod positioning complex, ready to be hammered into the core; 4: filling the remaining space in the slots with dilute dental cement; 5: the upper portion of a prepared hyperdrive core. (C) Preparation of the microdrive. 1: a shuttle bolt placed in a shuttle bolt holder base, note the smaller opening is facing away from the experimenter; 2: threading the threads inside the shuttle bolt; 3: insertion of the shuttle bolt into the shuttle; 4: a microdrive placed in the microdrive assembly station base with the cannula guided by the station lid, ready to be inserted; 5: a microdrive with the outer half of the upper cannula tip removed (indicated by arrow); 6: assembled microdrives tested on the microdrive rack. (D) A central column with threaded holes and inserted screw nuts. (E) A hyperdrive cap with four magnets glued in the wells. (F) A 36 mm long guide cannula bundle, with the soldered portion on the left. Please click here to view a larger version of this figure.
Figure 4: Assembly of the hyperdrive. Images showing stages of the hyperdrive assembly. (A) Insertion of the guide cannulas into the core. 1: the guide cannula bundle slid into the silicon tube and the spacer; 2: one guide cannula being placed in its designated hole in the core. Hand writing shows the organization of the guide cannulas; 3: guide cannulas pushed into the core; 4: the core with the guide cannulas inserted and secured by the nut. (B) Assembly of the microdrives to the core. 1: the core with microdrives loaded; 2: the microdrives with polyimide tubes inserted in the cannulas. (C) Insertion of the ground wires into the core. (D) Attachment of the electrode interface board. 1: the hyperdrive with the central column inserted; 2: the hyperdrive with the electrode interface board attached to the central column. (E) Connection of the ground wire to the designated hole in the electrode interface board. (F) A finalized hyperdrive ready for implantation (total weight of 20 g). Please click here to view a larger version of this figure.
Figure 5: Neural signals recorded by the hyperdrive. Representative recordings showing unit neural activity and local field potential in a behaving rat's brain. (A) Two-dimensional cluster diagrams illustrating individual spikes from simultaneously recorded neurons by a tetrode located in the postrhinal cortex (depth: 2.1 mm). Left: scatter plot showing relationship between peak-to-peak amplitudes of spikes recorded from two electrodes of the tetrode. Each dot corresponds to one spike. Clusters of spikes are likely to originate from the same cell. Four clusters are color coded. Scale bar: 20 µV. Right: spike waveforms (means ± S.D.) of the color-coded cells shown on the left. Note the small variation of the waveforms. Scale bar: 200 µs. (B) Traces of local field potential in the theta frequency range recorded simultaneously from four different tetrodes located in the medial entorhinal cortex (depth: 3.4-3.7 mm) when the rat was freely foraging. Scale bar at bottom left: 500 µV; scale bar at bottom right: 100 ms. Please click here to view a larger version of this figure.
Supplementary Files: The supplementary files include 20 files in the .stl format detailing the hyperdrive components and accessories ready for stereolithographic printing (units in mm), and 1 file in .pdf format which is the blueprint of the turning tool tip ready for machining. The original 3D model files were created with software AutoCAD in .dwg format, which will be available upon request. Please click here to download this file.
Here, we describe the process of constructing a newly developed hyperdrive comprised of eighteen independently movable tetrodes. The drive can be constructed from affordable parts purchased at many available hardware stores, combined with components created by stereolithographic printing. The hyperdrive can be chronically implanted onto a rat's skull using standard surgical procedures and is capable of recording extracellular neural activity while the animal performs various behavioral tasks.
The hyperdrive retains many of the desirable features of the original McNaughton hyperdrive, including the tripod microdrives that are oriented outward by 30 degrees from the drive center13, which provides reliable support for the tetrodes. Once implanted, the hyperdrive affords the execution of small movements of the tetrodes within the brain of an awake animal with considerable precision. One full turn of a shuttle on the threaded rod corresponds to a linear displacement of 317.5 µm. With proper training, an experimenter can advance a shuttle in 1/16 turn steps (20 µm). We designed the hyperdrive for use in adult rats, but the drive could easily be used in any animal with a body size of 350 g or greater (limited by head size). One limitation of the device may be noted in the restricted depth of recording, as the maximum travel distance of the tetrodes along the threaded rods is about 7 mm, which could fall short of deeper structures in some animals' brains.
Stereolithographic printing provides for sufficient resolution to create plastic components in great detail with high fidelity, and has been previously used in hyperdrive fabrication12,19,20. In this case, an industrial printer commonly available through third party production facilities was used. There all the hyperdrive components were printed precisely, including the hyperdrive core, despite its complex geometry, and the small structures such as the ø 0.6 mm through holes and the 0.3 mm thin walls. This precision makes stereolithography an ideal choice for manufacturing hyperdrive components. Based on prior experience, less expensive, desktop 3D printers are less likely to have the precision necessary for reliable reproduction of the hyperdrive components needed. Still, stereolithographic technology has its limitations. First, it has a limited selection of materials. The plastic we chose for the hyperdrive was the most durable of those we have tested, yet it is still not optimal for the manufacture of very small pieces. The shuttles and the shuttle bolts need to be handled with extra caution as they can break during preparation. The plastic components are not autoclavable, as the heat deflection temperature of the material is around 50 °C. In addition, the printing material used is not acetone resistant. These issues could be resolved when new stereolithography materials are developed and tested. Yet, considering the relatively low-cost of stereolithography, the advantages of the technique and cost far exceed the defects. Second, due to the nature of stereolithography, during which photopolymers are photochemically solidified by a UV laser to form a single layer of the desired 3D model21, the objects created by stereolithographic printing are vulnerable to UV light. Consequently, exposing them to strong UV (e.g., direct sunlight) for many hours will irreversibly reduce their physical strength (based on personal communication with the print shop). Considering the environmental UV in the laboratory space (e.g., from the fluorescent lights), it is best to store the stereolithographic components in a dark box when not in use, which will retain components' physical strength for years. Moreover, it is important to use other methods aside from UV light to disinfect the hyperdrive surface before surgery. This test hyperdrive has remained implanted on the rat in good condition in an ordinary laboratory environment over the course of four months, without any indication of a reduction in physical strength or performance.
The 3D printable nature of this hyperdrive also allows rapid modifications and flexible redesign. For example, the hyperdrive can be easily modified to target separated multiple brain regions11. Moreover, this drive could be adjusted to allow simultaneous monitoring of neural activity and local brain manipulation. Incorporation of a microdialysis probe with the array of tetrodes allows for pharmacological activation and deactivation of neurons by the infusion of various drugs during neural recording22. Furthermore, neurons engineered to express light-sensitive channels can be activated or deactivated by the incorporation of an optical fiber in the tetrode bundle and optogenetic technique19. In addition, the drive can be easily rescaled with a fewer number of tetrodes for animals with smaller head sizes, such as mice or juvenile rats.
In summary, the easy mutability coupled with the simpler, more affordable method of constructing an effective neural recording implant that can be reliably and accurately reproduced, makes this hyperdrive a powerful tool in the field.
The authors have nothing to disclose.
We thank the Moser Lab at the Kavli Institute for Systems Neuroscience and Centre for Neural Computation, Norwegian University of Science and Technology, for the chronic neural recording procedures in rats. This work was supported by NIH grant R21 NS098146, and Human Frontier Science Program Long-Term Fellowship LT000211/2016-L to L. Lu.
Welding rod | Blue Demon | ER308L-035-01T | Stainless steel, 0.035" in diameter |
Screw | McMaster | 91771A060 | Stainless steel, flat head, 0-80 thread, 5/8" in length |
Screw | McMaster | 91772A051 | Stainless steel, pan head, 0-80 thread, 5/32" in length |
Screw | McMaster | 92196A056 | Stainless steel, socket head, 0-80 thread, 5/16" in length |
Screw | McMaster | 92196A055 | Stainless steel, socket head, 0-80 thread, 1/4" in length |
Screw | McMaster | 95868A131 | Nylon, socket head, 2-56 thread, 3/16" in length, black |
Screw nut | McMaster | 90730A001 | Stainless steel, narrow hex, 0-80 thread |
Shoulder screw | McMaster | 90298A213 | Stainless steel, 8-32 thread, 3/16" in diameter, 1/4" in length |
Cup screw | McMaster | 92313A105 | Stainless steel, 4-40 thread, 3/16" in length |
Thumb screw | McMaster | 94323A592 | Nylon, 8-32 thread, 3/8" in length, black |
Magnet | Apex | M3X1MMDI | Neodymium, 3 mm X 1 mm disc |
Metal tubing | Small Parts | B00137QHNS | Stainless steel, 23 gauge, 0.0253" OD, 0.013" ID, 0.006" wall |
Metal tubing | New England Small Tube | Custom-made | Stainless steel, 30 gauge, 0.012/0.0125" OD, 0.007/0.008" ID, full hard |
Heat-shrink tubing | McMaster | 7856K72 | 0.09" ID before shrinking, blue |
Silicone tubing | A-M Systems | 807300 | 0.040" ID, 0.085" OD |
Polyimide tubing | A-M Systems | 823400 | 0.0045" ID, 0.0005" wall |
Ground wire | A-M Systems | 791500 | 0.005" bare, 0.008" coated, half hard |
Tetrode wire | California Fine Wire | Custom-made | 0.0007" in diameter, platinum-iridium (90%-10%), HML and VG coating |
EIB | Neuralynx | EIB-72-QC-Large | |
Gold pins | Neuralynx | large EIB pins | |
Tap | Balax | 01302-000 | M1.2 thread size |
Tap | McMaster | 2522A811 | 0-80 thread size, bottoming |
Tap | McMaster | 2522A771 | 0-80 thread size, plug |
Tap | McMaster | 26955A94 | 3/8"-24 thread size, bottoming |
Tap | McMaster | 2522A713 | 2-56 thread size |
Tap | McMaster | 2522A715 | 4-40 thread size |
Tap | McMaster | 2522A718 | 8-32 thread size |
Die | McMaster | 2576A457 | 3/8"-24 thread size, 1" OD |
Drill bit | McMaster | 30585A82 | Wire gauge 65, 0.035" in diameter |
Drill bit | McMaster | 30585A83 | Wire gauge 66, 0.033" in diameter |
Drill bit | McMaster | 30585A87 | Wire gauge 70, 0.028" in diameter |
Drill bit | McMaster | 30585A88 | Wire gauge 71, 0.026" in diameter |
Drill bit | McMaster | 30585A91 | Wire gauge 73, 0.024" in diameter |
Drill bit | McMaster | 8870A23 | 3/16" in diameter |
Dremel disc | Wagner | 31M | Diamond coated, 22 mm in diameter, 0.17 mm in thickness |
Steel wire | Precision Brand | 21212 | 0.012" in diameter, full hard |
Steel wire | Precision Brand | 21007 | 0.007" in diameter, full hard |
Steel wire | A-M Systems | 792700 | 0.003" in diameter, half hard |
Super glue | Loctite | LT-40640 | # 406 |
Super glue | Loctite | LT-41550 | # 415 |
Dental acrylic powder | Teets | 223-3773 | Coral |
Dental acrylic liquid | Teets | 223-4003 |