Here, we show the pressure injection of neuropharmacological substances during single-cell recording in an awake, behaving macaque monkey. This procedure allows pharmacological manipulation in the direct vicinity of a cortical recording site.
The top-down modulation of feed-forward cortical information processing is functionally important for many cognitive processes, including the modulation of sensory information processing by attention. However, little is known about which neurotransmitter systems are involved in such modulations. A practical way to address this question is to combine single-cell recording with local and temporary neuropharmacological manipulation in a suitable animal model. Here we demonstrate a technique combining acute single-cell recordings with the injection of neuropharmacological agents in the direct vicinity of the recording electrode. The video shows the preparation of the pressure injection/recording system, including preparation of the substance to be injected. We show a rhesus monkey performing a visual attention task and the procedure of single-unit recording with block-wise pharmacological manipulations.
In cortical and subcortical areas, neuronal activity is affected by various neuromodulators, for example acetylcholine1. These modulatory effects on neuronal responses have been reported in in vitro studies2, as well as in electrophysiological recordings from anesthetized animals3 and systemic pharmacological manipulations in humans4. Nevertheless, the exact role of different neuromodulators and the involvement of various receptor subtypes are largely unknown. To measure the effects of specific neuromodulators on the activity of single neurons, it is desirable to induce a temporary neuromodulator change as close as possible to the recording electrode. Furthermore, it is important that those manipulations are done in awake animals, as cognitive functions are only present in the absence of anesthesia. Additionally, anesthesia interacts with cholinergic and GABAergic systems5,6 and can lead to changes in neural activity3.
Within the last decades, two main methods of local drug delivery have been developed and refined: iontophoresis and pressure injection. In both methods drugs are delivered through micropipettes made of either glass or steel. With iontophoresis, an electrical current regulates the release of the drug7. Additionally, there is a significant contribution of electro-osmosis to the total amount of ejected molecules8, correlating with the tip diameter9 of the micropipette as well as with the concentration8 of the substance used. Iontophoresis is a powerful tool to quickly and precisely manipulate small volumes of nervous tissue. For iontophoretic injections, multi-barrel micropipettes are usually used10, with one acting as a recording device while the other positions serve as delivery pipettes. A limitation of this method is that only charged molecules can be used, severely limiting the selection of drugs.
Pressure injection uses either air compression or mechanical pressure to eject a substance from a micropipette. Using this method any soluble substance, charged or uncharged, can be used, including large molecules. The method of pressure injection was first described by Reyniers in 1933 and further refined in the 1950s (see Lalley11 for a review). In the 1980s the method was further refined to allow delivery of amounts in the nanoliter range (mainly lidocain12) to a defined brain area13 while simultaneously performing single-cell recording. The ejected volume was usually monitored by observing the movement of a marker, such as the meniscus in the upper part of the pipette13. Pressure injection was first used in the 1990s in awake animals, both extracellularly14 and intracellularly15,16. Based on the cumulative expertise gained in these studies it is now possible to reliably record from different brain structures in combination with pharmacological manipulation (see17 for a comparison of recent pressure injection systems).
An enduring open issue for both drug delivery methods is the difficulty in determining the precise volume injected. This is an even bigger challenge for experiments with awake, behaving rhesus monkeys where the animal performs the experimental task in a separate room. This can be alleviated by the use of a software-controlled system instead of relying on a visual marker to continuously monitor an injection.
The system described here is an extension of a well-established electrophysiological recording system (Mini Matrix System) and combines an injection pipette with multiple parallel-oriented recording electrodes at defined distances in a customizable arrangement. Pharmacological manipulation of the tissue near the recording electrode is possible using only a small amount of substance, ensuring a fast recovery and allowing multiple blocks of injection and control/recovery within the limited time window offered by the behavioral task of the animal.
Animal care and all experimental procedures were conducted in accordance with German laws governing animal care and approved by the district government of Braunschweig, Lower Saxony, Germany.
Note: As the experiment is performed in vivo, it is crucial to maintain the highest possible hygiene standards. Whenever possible, work under sterile conditions.
1. Preparing the Injection/Recording System
2. Preparing the Substance
3. Daily Preparation of the Injection/Recording System
Note: When mounted in the recording system, the electrodes and micropipette are stored in an enzyme solution (Tergazyme, 1% solution with deionized water) between recordings. The following steps must be performed before every recording.
4. Validation of the Injection System
Note: Although the company calibrates the system, it is recommended to validate the ejected volumes with the materials used in the experimental set-up (tubes, syringes etc.).
5. Acute Recordings
6. Spatial Attention Task
7. Pharmacological Manipulation While Recording
Note: While the monkey is performing the task, inject the substance in a block-wise manner. Three consecutive blocks are defined: control, which acts as a baseline; injection, during which a substance is ejected; and recovery, during which the cells targeted by the injection return to baseline.
8. Post Recording Procedures
Figure 2 depicts the spatial attention task the monkey performed while the injection process was conducted. The monkey was trained to attend to either the stimulus located within the receptive field of the recorded neuron (attend-in), the stimulus located outside of the receptive field (attend-out) or the fixation point (attend-fix). These conditions allow a comparison of neuronal activity in different attentional states.
Figure 3 shows a peri-stimulus time histogram of a sample neuron in an experiment using scopolamine, a muscarinic cholinergic antagonist. The plot demonstrates the response suppression during scopolamine injection versus no injection, when a pattern moving in the cell's preferred direction is presented inside the neuron's receptive field and is attended by the animal. The two first peaks represent the neuron's response to the on- and offset of the spatial cue, which appears inside its receptive field. This is followed by the response to the moving pattern which appears on the screen 500 ms after cue onset. The gray shaded area depicts the analysis period used to calculate the average firing rate for every trial. The green area highlights the suppressive influence of scopolamine injection on the cell's firing rate. The dark green region shows the suppression within the analysis period.
Figure 4A shows the effect of scopolamine on the average firing rate of the sample neuron in each of the three attentional conditions. The neuron's firing rate for the two spatial attention conditions (attention inside or outside of the receptive field of the recording neuron) as well as for the sensory condition (attention at fixation point) dropped shortly after the first injection of the injection block (grey shaded area) and during the recovery block increased after a delay to the same level as before the injection.
Figure 4B shows a control recording from a second sample neuron in which saline (0.9% NaCl) was injected, using the same protocol as for the scopolamine injection. During the injection block no change in the neuron's firing rate was observed compared to the control block.
Figure 1. Set-up used for pharmacological manipulation while recording. (A) Depicts the microinjection pump and the electrophysiological recording system equipped with electrodes and micropipette. The guidetube gap, into which silicon oil is inserted to lubricate the electrodes and micropipette, is shown enlarged. (B) Displays an example micropipette (above) and recording electrode (below). For size comparison, a euro cent (diameter: 16 mm) is placed underneath. Please click here to view a larger version of this figure.
Figure 2. Task design to guide spatial attention. Monkeys were trained to detect a motion direction change in the cued dot pattern. The cue was either placed within the neuron's receptive field (attend-in), as shown in the figure, or outside of it (attend-out). As a sensory control, the monkey was trained to detect a luminance change of the fixation point (attend-fix). Please click here to view a larger version of this figure.
Figure 3. Influence of antagonist scopolamine on firing rate. The peri-stimulus time histogram for a sample neuron is shown for the attend-in condition (attention inside the receptive field of the recorded neuron) during the injection block and during the control block. The x-axis depicts the time in milliseconds after cue onset and the y-axis shows the firing rate in spikes/sec. The grey area depicts the analysis period (300-800 ms after stimulus onset) used to calculate the trial-averaged firing rate. The green shaded area shows the suppression in firing rate across the two conditions. The dark green color highlights the suppression within the analysis period. Please click here to view a larger version of this figure.
Figure 4. Effect of scopolamine and saline on firing rate. (A) Antagonist scopolamine injection. The trial-averaged firing rate of the sample cell from Figure 3 over the course of the experiment is shown for the preferred stimulus for all three attentional conditions. The x-axis depicts the trial start time in minutes and the y-axis shows the unit's firing rate in spikes per seconds. Symbols ( attend-in, attend-fix, attend-out) represent the neuron's firing rate within the analysis period in every successfully performed trial, and horizontal lines (solid line: attend-in, dotted line: attend-fix, dashed line: attend-out) show average firing rate for the three different experimental blocks (control, injection, recovery). The grey shaded area shows the injection block, beginning with the first injection and ending 1 min after the last injection. During the injection block 2 nL of 0.1 molar scopolamine were injected every minute with an injection velocity of 2 nl/s. (B) Saline injection. The firing rate of a sample cell over the course of the control experiment is shown for the preferred stimulus for all three attentional conditions. Grey shaded area visualizes the block of saline injection. Please click here to view a larger version of this figure.
Here we have illustrated in detail how to perform reliable and precise injections and high quality single-cell recordings with an "off-the-shelf" pressure injection system. While this method of drug delivery has previously been used in behaving monkeys (reviewed in17), the system presented here has advantages, reviewed below.
As illustrated in Figure 4A, the system described here can provide stable measurement of single neuron activity with and without pharmacological injections in the direct vicinity of the recording site. As shown in Figure 4B, the injection of a control substance, saline, did not lead to a change in firing rate. This control demonstrates that the injection process itself has no measurable influence on the firing properties of the recorded neurons.
The spatial configuration of neuron, recording electrode, and micropipette is of crucial importance in these experiments. Although a precise measurement of their relative positions in the tissue during recording is not possible, we can consider and control for possible sources of variance. First, during volume injection there is a risk that the neuron of interest may be displaced away from the recording electrode, affecting the stability of recorded signals. For that reason it is prudent to compare the firing rate before and after the injection block to verify signal stability. Second, the guide tube configuration of the recording system defines the distance between electrodes and micropipette (e.g., 305 µm in the concentric 3-channel system used in this experiment). As the system provides precise position control for the depth of electrodes and micropipette in the tissue, the distance between them can be minimized by carefully calibrating relative depth before recording (step 3.5), and keeping them at a common depth during recordings.
Potential limitations
In addition to in-house quality control by the manufacturer, the system needs to be validated under lab conditions, as different brands of tubes, syringes etc. can be used and could lead to differences in ejected volumes. Although the system can be used to inject very small volumes as in the experiment shown here, these are below the minimum volume that can be validated due to practical measurement limits in a normal laboratory environment. However, larger injection volumes can be used to infer the relation between the software-defined volume and the volume ejected by the hardware. If transparent tubes are used, an additional visual check of the injection process is possible by measuring the displacement of a visual marker.
Inserting the micropipette into the system is more demanding than electrode insertion, as the diameter of the micropipette is slightly larger and the material is more fragile. In addition, joining the tube to the pin of the micropipette is challenging as it entails a high risk of breaking the upper part of the micropipette. However, the lifetime of a successfully loaded micropipette is several months, even with daily use.
In practice, we have not yet encountered a blockage in the injection system during post-recording cleaning of the system. Nevertheless, no "online" check is possible, and there is a risk that a physical blockage (such as tissue at the micropipette tip) might prevent substance injection. It might therefore be advisable to analyze the data conservatively, such as including only those cells in further analysis that show significant changes in firing rates between control and injection blocks of the experiment.
Despite their small diameter, microelectrodes and pipettes will displace brain tissue and may cause some local tissue damage. This can be minimized by manually positioning the tip of the guide tubes just above the dura mater. The electrodes then penetrate the dura and their intactness is inferred by measuring their impedances online. Afterwards, the micropipette is inserted. When using this approach, regular removal of tissue above the dura is recommended to further reduce the risk of electrode or pipette breakage.
Comparison to alternative methods
The system used here shows clear advantages compared to other pressure injection systems. One strong advantage is the diameter of the micropipette (approximately 100 µm), which is half the size of other available probes17 and therefore minimizes neural tissue damage. In contrast to previous designs, the current system employs spatially separated recording electrode and micropipette. Although other systems provide a smaller distance between electrode and pipette, the system described here allows independent depth changes of electrodes and pipette, thus permitting variable relative distances within a recording session. Importantly, no compromise regarding recording quality needs to be made, as the injection system is an extension of an established recording device. While only one micropipette and thus one substance is used in this protocol, it is possible to inject several substances within one experimental procedure. To achieve this, several micropipettes can be threaded into separate guide tubes and connected to syringes mounted in individual injection pumps. Finally, controlling the system is easy, as only one computer program is needed to advance the electrodes and micropipette, and to perform the pressure injection during the experiment.
Comparing pressure injection to iontophoresis, there are relative advantages and disadvantages. For example, pressure injection requires greater volumes to be introduced into the tissue than iontophoresis, thus increasing the risk of neuronal displacement. The current protocol used volumes in the nL range, and we rarely experienced noticeable changes in a recorded cell's signal quality. The system also allows larger volumes to be injected, which is potentially useful for behavioral manipulations but could impact stability of neuronal recording. A clear advantage of pressure injection over iontophoresis is the larger variety of useable substances as there is no requirement to use charged substances. However, pH values should be checked and compared between experimental and control substances (e.g., saline).
The question might arise why to use the long-established method of pressure injection instead of newer techniques such as optogenetics for manipulating neural activity. Although well established in rodents, optogenetics is not yet reliably established in rhesus monkeys. In particular, it does not yet allow the local manipulation of cells selective for a particular neurotransmitter type. In the longer run, we see great potential for the combination of the advantages of pharmacological manipulations with the advantages of optogentic manipulations in elucidating the neural basis of cognitive functions.
Here we have shown how pressure injection can be used to pharmacologically manipulate a locally restricted area in the brain of awake, behaving rhesus monkeys. We hope that this method inspires other scientists to investigate neuromodulatory contributions to the dynamics of neuronal activity.
The authors have nothing to disclose.
This work was supported by grants of the Deutsche Forschungsgemeinschaft through the Collaborative Research Center 889 "Cellular Mechanisms of Sensory Processing" to S.T. (Project C04). We thank Sina Plümer, Leonore Burchardt, Dirk Prüsse, Klaus Heisig and Ralf Brockhausen for technical and animal-related support and our collaborators in the Stem Cell Unit of the German Primate Center, Dr. Katharina Debowski and Anna Magerhans, for technical assistance in the filtration process.
(-)-Scopolamine hydrochloride | Sigma-Aldrich | 55-16-3 | Mr 339.81 g/mol |
NaCl 0.9% | B. Braun Melsungen AG | 3079870 | 5ml |
Terg-a-zyme | Sigma-Aldrich | Z273287 | enzyme detergen |
Hydrogen peroxide | Roth | Used in 3% solution with deionized water | |
Ethanol | Chemie-Vertieb Hannover | 104642 | 70% |
Deionized water | |||
Injekt 40 Duo | B. Braun Melsungen AG | 9166432V | Syringe and needle |
Eppendorf Safe-Lock microcentrifuge tubes, amber | Eppendorf | 0030 120.191 | 1,5ml |
Quarzglass micropipette | Thomas Recording | ||
Recording electrode | Thomas Recording | quartz/platinum-tungsten fiber electrode; impedance value 1-2 MΩ and 0.3-0.5 MΩ | |
PharmedBPT-Schlauch | Saint-Gobain Performance Plastics | 3702003 | Size: 0,25 x 2,05 mm (Wd: 0,9mm) |
Loctite 401 | Henkel | 233641 | Superglue |
Silicon oil | Thomas Recording | M-1000 | |
Minisart RC15 | Sartorius | 17761———-R | Syringe filter |
Multichannel Micro Injection System | Thomas Recording | multichannel microelectrode manipulator “System Eckhorn” equipped with microelectrodes and micropipettes and a precision multichannel microinjection pump | |
McLab | custom | internal lab software to control stimulus presentation |