Date Published: January 18, 2018
Author(s): Nathan T. Rodeberg, Stefan G. Sandberg, Justin A Johnson, Paul E. M. Phillips, R. Mark Wightman.
Fast-scan cyclic voltammetry (FSCV) has been used for over 20 years to study rapid neurotransmission in awake and behaving animals. These experiments were first carried out with carbon-fiber microelectrodes (CFMs) encased in borosilicate glass, which can be inserted into the brain through micromanipulators and guide cannulas. More recently, chronically implantable CFMs constructed with small diameter fused-silica have been introduced. These electrodes can be affixed in the brain with minimal tissue response, which permits longitudinal measurements of neurotransmission in single recording locations during behavior. Both electrode designs have been used to make novel discoveries in the fields of neurobiology, behavioral neuroscience, and psychopharmacology. The purpose of this Review is to address important considerations for the use of FSCV to study neurotransmitters in awake and behaving animals, with a focus on measurements of striatal dopamine. Common issues concerning experimental design, data collection, and calibration are addressed. When necessary, differences between the two methodologies (acute vs chronic recordings) are discussed. The topics raised in this Review are particularly important as the field moves beyond dopamine toward new neurochemicals and brain regions.
The monitoring of molecules in the brain has undergone significant advances in the past four decades. One of the earliest techniques for measuring neurotransmitter release was push–pull perfusion, a method that uses a cannula for sample collection prior to downstream analysis.1,2 However, the direct interface of the perfusate with brain tissue raised concerns with sample contamination and flow-induced damage to the surrounding environment. To address these issues, this procedure was later adapted to incorporate a dialysis membrane, creating the technique known as microdialysis.3–6 Microdialysis restricts flow to the probe, which minimizes brain damage and maintains sample purity. Equilibration of analytes across the membrane according to their concentration gradients results in concentration changes in the dialysate reflective of fluctuations in the brain. Microdialysis is highly versatile, with its sensitivity, selectivity, and number of analytes that can be monitored simultaneously dependent on the detection method employed. Its main limitation is spatiotemporal resolution, as microdialysis probes are typically at least 200 μm in diameter, and samples are historically collected approximately every 5–20 min to allow sufficient sample volume accumulation at low flow rates.5–8 Recent improvements, largely due to reduction in the minimum volume needed for sample analysis, have permitted microdialysis measurements on a subminute time scale.9–12
Measurements using FSCV with CFMs were originally conducted in anesthetized animals.25,26,40,41 However, these studies could not reveal direct information about neurotransmission during behavior. The first FSCV measurements in freely moving animals detected dopamine release in terminal regions, evoked by electrical stimulation of afferent axonal pathways in rats. These experiments used acutely implanted glass-encased CFMs lowered into the brain using head-mounted microdrives.42–44 Later, behavioral evoked dopamine was detected by this approach,45 and these types of recordings became routine, primarily due to improved sensitivity obtained by increasing the anodic limit of the waveform46,47 to maintain oxygen-containing moieties on the electrode surface which enhance adsorption of positively charged analytes (such as dopamine).27 FSCV has been adapted for multimodal recordings with simultaneous extracellular electrophysiological recordings48–51 and iontophoresis51–54 at the same probe.
Extracellular dopamine is detected via its oxidation and reduction at the carbon-fiber surface, producing a voltammetric current proportional to its local concentration. However, how to obtain this concentration has been a matter of considerable debate and development within the field. Original calibrations of in vivo voltammetric data directly converted the voltammetric current at the peak oxidation potential for dopamine into a concentration using an externally obtained calibration factor. However, various electroactive substances can interfere at the oxidative peak for dopamine, including ascorbic acid,26,35,101,102 dopamine metabolites,17,26,102,103 pH,80,104–107 and other ions.104,107 Because this method is univariate (i.e., only uses a single measurement point to predict concentration), it cannot separate out these interferences.108,109 While anatomical and pharmacological criteria can increase confidence in the identity of the measured signal, univariate analysis will fail if interfering analytes significantly contribute.
The authors of this Review are in general agreement that, when appropriate caution is observed, both acute and chronic CFMs can be used for detection of behaviorally evoked dopamine release in regions of the striatum using FSCV. In support of the reliability of these measurements, there is generally high concordance between results from FSCV of dopamine concentration fluctuations in the striatum with either acute or chronic electrodes, and electrophysiological recordings of dopamine neurons in the midbrain, with many key findings reproduced across approaches. These replications include the characterization of reward prediction-error signals62,70,140 that convey quantitative information.63,141 They include demonstrations that dopamine signals to reward-related cues are sensitive to factors that influence subjective value such as delayed reward delivery (temporal discounting),62,142,143 or subjective risk preference,144–146 and concur that there is stronger encoding of reward size than effort-based response cost by dopamine signals.147,148 An uncertainty-like signal following presentation of a Pavlovian stimulus predicting probabilistic reward has been identified and replicated across methodologies149,150 as have observations of partial generalization between sensory stimuli that are associated with different economic values,62,151 which can come in the form of a presumed sensory signal, temporally separated from a value signal.63,152,153 The success of chronic electrodes is notable, as it has long been held that chronically implanted electrodes are prone to failure. A recent review of glucose biosensors documents the importance of chronic sensors for monitoring in diabetes. The chief problem to their use is the foreign body response that impairs sensor performance.154 It may be that the finding that very small electrodes remain functional will be very useful to other health related fields involving biosensors.