Article pubs.acs.org/ac
Monitoring In Vivo Changes in Tonic Extracellular Dopamine Level by Charge-Balancing Multiple Waveform Fast-Scan Cyclic Voltammetry Yoonbae Oh,† Cheonho Park,† Do Hyoung Kim,⊥ Hojin Shin,† Yu Min Kang,† Mark DeWaele,† Jeyeon Lee,† Hoon-Ki Min,‡,§ Charles D. Blaha,‡ Kevin E. Bennet,‡,∥ In Young Kim,† Kendall H. Lee,*,‡,§ and Dong Pyo Jang*,† †
Department of Biomedical Engineering, Hanyang University, Seoul 04763, Korea Department of Neurologic Surgery, Mayo Clinic, Rochester, Minnesota 55905, United States § Department of Physiology and Biomedical Engineering, Mayo Clinic, Rochester, Minnesota 55905, United States ∥ Division of Engineering, Mayo Clinic, Rochester, Minnesota 55901, United States ⊥ ybrain, Pangyo Digital Center C-dong, 242 Pangyo-ro, Seongnam, Gyeonggi-do 13487, Korea ‡
S Supporting Information *
ABSTRACT: Dopamine (DA) modulates central neuronal activity through both phasic (second to second) and tonic (minutes to hours) terminal release. Conventional fast-scan cyclic voltammetry (FSCV), in combination with carbon fiber microelectrodes, has been used to measure phasic DA release in vivo by adopting a background subtraction procedure to remove background capacitive currents. However, measuring tonic changes in DA concentrations using conventional FSCV has been difficult because background capacitive currents are inherently unstable over long recording periods. To measure tonic changes in DA concentrations over several hours, we applied a novel charge-balancing multiple waveform FSCV (CBM-FSCV), combined with a dual background subtraction technique, to minimize temporal variations in background capacitive currents. Using this method, in vitro, charge variations from a reference time point were nearly zero for 48 h, whereas with conventional background subtraction, charge variations progressively increased. CBM-FSCV also demonstrated a high selectivity against 3,4-dihydroxyphenylacetic acid and ascorbic acid, two major chemical interferents in the brain, yielding a sensitivity of 85.40 ± 14.30 nA/μM and limit of detection of 5.8 ± 0.9 nM for DA while maintaining selectivity. Recorded in vivo by CBM-FSCV, pharmacological inhibition of DA reuptake (nomifensine) resulted in a 235 ± 60 nM increase in tonic extracellular DA concentrations, while inhibition of DA synthesis (αmethyl-DL-tyrosine) resulted in a 72.5 ± 4.8 nM decrease in DA concentrations over a 2 h period. This study showed that CBMFSCV may serve as a unique voltammetric technique to monitor relatively slow changes in tonic extracellular DA concentrations in vivo over a prolonged time period.
D
It is known that adopting higher scan rates can improve the chemical sensitivity of FSCV.10 Although peak current for diffusion-mediated electron transfer varies with the square root of scan rate, peak faradaic oxidation current for DA, which adsorbs to the electrode surface of CFMs, scales proportionally with scan rate. However, at relatively high scan rates (e.g., 300− 400 V/s), the background capacitive charging current also increases in direct proportion to scan rate and thus overwhelms peak DA oxidation currents.11 To maintain the advantage of sensitivity to DA at high scan rates, the background current
opamine (DA) plays an essential role in the central nervous system by modulating a variety of brain circuitries.1 An imbalance in DA extracellular levels is known to be relevant to many disorders such as depression, addiction, schizophrenia, and Parkinson’s disease.2−6 Fast-scan cyclic voltammetry (FSCV) with carbon-fiber microelectrodes (CFMs) has been widely used to measure neurochemicals such as DA because of its high sensitivity, chemical selectivity, and superior spatial and temporal resolution relative to other in vivo neurochemical monitoring techniques, such as microdialysis.7,8 As a result, FSCV measurements of extracellular DA concentrations in the brain have been used to elucidate the neurochemical basis of a number of neurological and psychiatric disorders.1,6,9 © XXXX American Chemical Society
Received: July 9, 2016 Accepted: October 24, 2016 Published: October 24, 2016 A
DOI: 10.1021/acs.analchem.6b02605 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 1. In vitro characterization of holding potential-dependent waveforms and charge-balancing waveforms. (A) Triangular waveforms (scan rate: 400 V/s) with different holding potentials to examine their effects on the capacitive non-Faradaic charging currents. (B) Background cyclic voltammograms for different negative holding potentials. (C) Proposed charge-balancing Mexican hat shaped waveform. (D) Background cyclic voltammogram (10 Hz) of triangular waveform (scan rate: 1200 V/s, solid line) and charge-balancing waveform (dashed line). (E) The response of dopamine (1 μM) at 0.1 Hz repetition times using a triangular waveform (solid line) and a charge-balancing waveform (dashed line). (F) The amount of adsorbed dopamine (1 μM) resulting from different repetition times (0.1 to 10 s). Repetition time of 10 s showed no significant difference in dopamine sensitivity between triangular waveforms and charge-balancing waveforms (n = 5, paired t test, p value = 0.1768; values as the mean ± SEM).
istics. FSCAV provided an estimate of absolute tonic dopamine concentrations in vivo over a 90 min period.29 In the present study, we improved these previous methods by introducing a charge-balancing multiple waveform fast-scan cyclic voltammetry (CBM-FSCV), which allows measurement of relative changes in tonic extracellular DA concentrations for up to several hours by taking advantage of the chemical adsorption characteristics of DA under a manipulated background capacitive current. Here, we have tested two new ideas to minimize time-dependent variance in the capacitive charging current inherent in conventional FSCV. First, we used a chargebalancing waveform with a zero holding potential that takes into account the voltammogram difference among multiple pulses within one series (Mexican hat shaped waveform). When using the conventional triangular FSCV waveform with a −0.4 V holding potential, the background current increases because the double layer near the electrode surface is charged with ions by a negative holding potential.30,31 In contrast, the chargebalancing waveform we designed has a zero holding potential, thus reducing background current variance. Second, we applied dual background subtraction in which after the initial subtraction in the multiple pulses within one hundred waveform train, a single reference point is chosen from which the rest of the time points are subtracted in the same manner as conventional background subtraction. This study shows measurement of stable in vivo extracellular DA concentrations during both pharmacologically induced positive and negative changes in tonic levels, indicating that CBM-FSCV is an efficacious method for monitoring the slow changes in central DA concentrations over a relatively prolonged time period.
must be subtracted in order to quantify phasic changes in DA faradaic current.12 This subtracted cyclic voltammogram provides information about its characteristic redox feature, which distinguishes the oxidizable substances from each other in the extracellular fluid. Although background current subtraction with FSCV has been used extensively to measure in vivo phasic changes in DA,13,14 measurement of slow tonic changes has not been easily accomplished with conventional FSCV because of the inherent instability in background currents over prolonged recording time periods (>90s).14 However, when background drift is accounted for, background signals are stable for 30 min.48 The overall extracellular concentration of DA, which modulates specific behaviors and is implicated in neurological and psychiatric disorders, is influenced by both phasic release over seconds and long-term tonic changes that occur over minutes to hours.15,16 Specifically, tonic DA concentration in the extracellular space, which is generated by low-frequency firing by DA neurons and under local excitatory control,17 critically affects DA transporter function and synaptic plasticity.1,18 In vivo changes in tonic extracellular DA concentrations have been traditionally monitored by microdialysis combined with highperformance liquid chromatography.19,20 In addition, microdialysis has been widely utilized for monitoring other monoamine neurotransmitters in vivo.20,21 In many cases, microdialysis can detect more than one analyte in one sample and this has enabled microdialysis to be used for quantifying basal concentrations and dynamic information regarding neurotransmitters.19,20,22,23 Despite these characteristics, important limitations that include low spatiotemporal resolution and tissue damage remain with microdialysis.24−27 For direct in situ measurement, Atcherley et al.28,29 has recently developed fast-scan controlled-adsorption voltammetry (FSCAV) to measure tonic (denoted as “absolute”) DA extracellular concentrations by exploiting CFM surface adsorption character-
■
MATERIALS AND METHODS Data Acquisition and Analysis. Data were acquired by a commercial electronic interface (NI USB-6251, 16 bit, National B
DOI: 10.1021/acs.analchem.6b02605 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 2. Schematic design of charge-balancing multiple waveform fast-scan cyclic voltammetry (CBM-FSCV). One multiple waveform consists of 100 charge-balancing waveforms and repeats every 10 s (9 s resting gap period; 0.1 Hz).
ester hydrochloride (αMPT) were purchased from SigmaAldrich (St. Louis, MO). In Vivo Experiments. Adult male Sprague−Dawley rats weighing 250−350 g were used for the experiments in this studies (n = 6). NIH guidelines were followed for all animal care, and the Hanyang University Institutional Animal Care and Use Committee approved the experimental procedures. Additional details regarding the in vivo experimental protocol is described in the Supporting Information.
Instruments) with a base-station PC computer and software written in the LabVIEW programming environment (National Instruments, Austin, TX). The NI USB-6251 was also used to synchronize the applied waveforms and flow injection analysis control. After data collection, all waveform parameters and operations, data sampling rate acquisition and transmission, background subtraction, signal averaging, and digital filtering were performed by custom software control. Data, in the form of a sequence of unsigned 2-byte integers, were saved to the base-station computer hard drive for offline processing by MATLAB (MathWork Inc., Natick, MA). GraphPad Prism 5 (GraphPad Software, San Diego, CA) was used to generate figures and statistics (one-sample, paired, and unpaired tests, 2way ANOVA, Sionedak’s multiple comparisons test, etc.). All data are presented as mean ± standard error of the mean (SEM) values for n number of electrodes. Fabrication of Carbon Fiber Microelectrodes (CFMs). Electrodes were fabricated as previously described.32 T300 carbon fiber (Cytec Thornel, Woodland Park, NJ) was used for all experiments. The exposed carbon fiber was trimmed to a final length of 50−70 μm using a scalpel.33 Additional details are available in the Supporting Information. Flow Injection Apparatus. A flow-injection analysis system, which consisted of a syringe pump (Harvard Apparatus, Holliston, MA) that directed a buffer solution through a Teflon tube to a 6-port injection valve (Rheodyne, Rohnert Park, CA), was used for in vitro measurements. The injection valve was controlled by a 12 V DC solenoid and was used to transport the analyte from an injection loop to an electrochemical flow cell at a rate of 2 mL/min. A CFM was placed in a flowing stream of buffer, and analyte was injected as a bolus to calibrate each electrode with known concentrations of DA. Chemicals. DA was dissolved in distilled water at a stock concentration of 1 mM diluted with TRIS buffer (15 mM tris(hydroxymethyl)aminomethane, 3.25 mM KCl, 140 mM NaCl, 1.2 mM CaCl2, 1.25 mM NaH2PO4, 1.2 mM MgCl2, and 2.0 mM Na2SO4, with the pH adjusted to 7.4) and preserved in 0.1 M perchloric. From this stock solution, dilutions were made in TRIS buffer for desired final concentrations before starting the flow injection experiments. All pharmacological agents consisting of nomifensine and α-methyl-DL-tyrosine methyl
■
RESULTS AND DISCUSSION Charge-Balancing Waveform. The application of various negative holding potentials (Figure 1A) on the magnitude of voltammogram background capacitance currents was investigated. As shown in Figure 1B, by varying the holding potential from −0.8 V to 0.0 V with the same triangular waveform reversal potential, the magnitude of the background current and corresponding capacitance recorded at CFMs progressively increased (holding potential 0.0 V: 35 μF/cm2, −0.4 V: 51 μF/ cm 2 , −0.8 V: 65 μF/cm 2 ), consistent with previous observations.34 Generally, the reduction process of dopamineo-quinone occurs around −0.2 V in cathodic sweeps at >400 V/ s scan rates. Our previous research has shown that the amount of reduction of dopamine-o-quinone at −0.2 V partially contributes to the overall sensitivity of DA detection at CFMs.32,35 Thus, to retain optimal DA reduction in order to achieve high sensitivity to DA while maintaining relatively low and stabile capacitance at the CFM surface, we designed a charge-balancing Mexican hat shaped waveform (Figure 1C). In this type of waveform, a negative holding potential (−0.4 V) is added to both sides of the triangular waveform to provide a balance between negative and positive equivalent areas in the waveform. The holding potential was maintained at 0 V between pulses. Application of this waveform resulted in significantly lower background currents (Figure 1D) as shown in Figure 1B. The Mexican hat shaped waveform, which contains reduction processes, resulted in almost equivalent sensitivity to DA in vitro when compared to DA sensitivity using a triangular waveform with a −0.4 V holding potential (Figure 1E).35 C
DOI: 10.1021/acs.analchem.6b02605 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 3. Comparison of multiple waveform background currents between triangular and charge-balancing waveforms. Each waveform shape has the same scan rate (1200 V/s), applied at 100 Hz for 1 s and held for 9 s. (A) Triangular multiple waveform, A1: Background cyclic voltammograms for 100 triangular waveforms, A2: A hundred background voltammogram differences calculated by subtraction of last cyclic voltammogram. (B) Chargebalancing multiple waveform, B1: Background cyclic voltammograms for 100 charge-balancing waveforms, B2: A hundred background voltammogram differences calculated by subtraction of last cyclic voltammogram. (C) The first of the hundred background cyclic voltammograms from A2 (solid line) and B2 (dashed line). (D) The amount of charge difference between the first cyclic voltammogram and the last cyclic voltammogram of a triangular and a charge-balancing multiple waveform. Charge-balancing multiple waveform showed significantly lower charge difference (n = 6, unpaired t test, p value =0.0006, t = 4.980, df = 10). The charge was determined by integrating the current over the total voltammogram (values as the mean ± SEM).
slopes in Mexican hat waveforms are set to 1200 V/s. The negative part of the Mexican hat waveform not only adjusts the balance of electrostatic force but also compensates for the loss of DA adsorption by adopting an interleaved 0 V holding potential.34 The multiple waveform is applied for 1 s at 100 Hz and is repeated every 10 s (Figure 2). Overall, every 10 s, DA accumulates near the CFM for the gap duration (9 s), which is approximately 0.1 Hz. 100 Hz multiple pulses are then applied to limit DA adsorption and consume accumulated DA with oxidation/reduction processes.36 The principal reason we developed the charge-balancing waveform was to decrease the charge difference among multiple pulses by enabling formation of near identical ionic density at the CFM surface while retaining relatively high DA sensitivity that is comparable to the triangular waveform, but with relatively lower capacitance. We investigated how the low capacitance applied to multiple pulses. First, we measured the background current of the triangular multiple waveform (−0.4 V → 1.3 V → -0.4 V, 100 Hz, 100 pulses) (Figure 3A) for comparison with the CBM-FSCV (Figure 3B). Figures 3A1 and 3B1 show background cyclic voltammograms for 100 triangular waveforms, and Figures 3A2 and 3B2 show 100 background voltammogram differences calculated by subtraction of the last cyclic voltammogram. For both waveform shapes, a decaying pattern of background current was detected.37 However, the
The effect of repetition time on the sensitivity of DA using both triangular and charge-balancing waveforms was examined (Figure 1F). Since the charge-balancing waveform had a 0 V holding potential that results in a near zero ionic net charge near the CFM, there could be a concern about electrostatic contributions to the sensitivity to the cationic species DA.36 However, by adopting a 10 s repetition time between the 100 waveform trains that results in relatively more adsorption of DA to the CFM, the sensitivity was sufficiently compensated for (Figure 1E). According to Atcherley et al.,28 10 s is sufficient to achieve concentration equilibrium at the electrode surface. Statistically, the Γeq value at 10 s delay time showed no significant difference between triangular and charge-balancing waveforms. However, at lower DA concentration (below 500 nM), CBM-FSCV showed slightly lower sensitivity than conventional FSCV (n = 3, Supporting Information Figure S1). Comparison of the Background Current of a Triangular Multiple Waveform with CBM-FSCV. For the CBM waveforms, a triangular waveform −0.4 V → 1.3 V → −0.4 V was used with a 1200 V/s scan rate and a 100 Hz repetition rate for optimal background current magnitude and sensitivity, as per Atcherley et al.28 On both sides of the waveform, a −0.4 V potential is maintained to balance out the triangle’s positive potential with the waveform potential then returning to a 0 V holding potential (Mexican hat shape). All D
DOI: 10.1021/acs.analchem.6b02605 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 4. Data analysis of dual background subtraction. All voltammograms from one of each series were subtracted from their own last voltammogram (blue) dynamically (① − ② and ④ − ⑤). This process permits the dynamic subtraction voltammogram (③ and ⑥) to be acquired from each series. Once a dynamic subtraction voltammogram has been set as a reference voltammogram (③), each multiple pulses’ dynamic subtraction voltammogram is subtracted from the reference voltammogram (⑥ − ③ = ⑦) in order to observe any changes in DA concentration, such as from DA injection (in vitro) or pharmacologically induced changes (in vivo).
Figure 5. Background stability comparison between conventional background subtraction and dual background subtraction methods in Tris buffer. The charge-balancing multiple waveform at 0.1 Hz was used, and the background subtraction reference point at 10 min after the starting of recording was used for both experiments. (A) Top: Conventional background-subtracted color plot over 2 days. Bottom: cyclic voltammograms determined at hour 1 (Day 0), 24 (Day 1), and 48 (Day 2). (B) Top: 3rd dual background-subtracted color plot over 2 days. Bottom: cyclic voltammograms at hour 1 (Day 0), 24 (Day 1), and 48 (Day 2). (C) The net charge difference near CFM over 2 days. Conventional background subtraction method (black bars, left y axis) showed significant changes in charge amount from hour 2 (2-way ANOVA, p < 0.0001, Sidak’s multiple comparisons test with hour 1). The dual background subtraction method (red bars, right y axis) showed no significant change in charge amount from hour 2 (2-way ANOVA, Sidak’s multiple comparisons test with hour 1). The variation in charge was averaged hourly.
E
DOI: 10.1021/acs.analchem.6b02605 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
Figure 6. CBM-FSCV recordings of DA, DOPAC, ascorbic acid. (A) Top: One hundred of background-subtracted CBM-FSCV responses to DA (100 nM). Bottom: Peak voltammetric currents of dual background-subtracted CBM-FSCV. DA responses from 1st to 66th waveform showed significant differences from the last voltammogram (n = 5, p = 0.0006, one-sample t test). (B) Top: One hundred of the background-subtracted CBM-FSCV responses to DOPAC (20 μM). Bottom: Peak voltammetric current of dual background-subtracted CBM-FSCV. DOPAC responses from 1st to 5th waveform showed significant differences from the last voltammogram (n = 5, p < 0.05, one-sample t test). (C) Top: One hundred of the background-subtracted CBM-FSCV responses to ascorbic acid (200 μM). Bottom: Peak current of dual background-subtracted CBM-FSCV. All ascorbic acid responses showed no significant differences from last voltammogram (n = 5, p = 0.1745, one-sample t test). All values are presented as the mean ± SEM. Horizontal black bars represent significant differences against 100th voltammogram peak current.
voltammogram. Therefore, it is necessary to subtract the difference voltammogram (the amount of charge difference) from the pre-DA injection reference point for the detection of DA (dual background subtraction). The ionic density near the CFM has been shown to determine the shape of the cyclic voltammogram.37,38 A voltammogram, collected as a reference point, is under limited ionic density by a 0 V holding potential. However, in the case of the charge-balancing multiple waveform, during a 0 V holding potential there are little to no electrostatic forces present to attract ions to the CFM surface, only mass transport, which is determined by the ionic concentration of the buffer medium.39 In addition, the difference voltammogram within one series is more consistent than the original background current because the decay pattern for background currents to a new equilibrium state is time-dependent and markedly affected by the form of the applied voltage waveform (see Figure 3A2,B2). In Figure 5, we examined the stability of dual background subtraction against conventional background subtraction. The backgroundsubtracted voltammogram (Figure 5A) was calculated by using only the initial voltammograms from each series, whereas the dual background-subtracted voltammogram (Figure 5B) was calculated by subtracting the 100th voltammogram from the 3rd voltammogram. The recording continued over a 2 day period in an isothermal/humidity condition (enclosed sealed beaker) in Tris buffer to minimize changes in ionic concentrations by evaporation. As depicted in Figure 5A, the conventional background subtraction method exhibited marked changes in relative background currents over the 2 day test period. Conversely, the dual background subtraction method showed comparatively minimal changes in relative background currents for the same test period. In Figure 5C, the background charge
charge-balancing multiple waveform exhibited a significantly smaller background current difference over all pulses derived from the newly designed waveform compared to the conventional triangular waveform. Figure 3C represents the current difference between the 1st and 100th voltammogram of the triangular (Figure 3A2) and the charge-balancing multiple waveforms (Figure 3B2). In this representative example, the multiple triangular waveform showed approximately 300 nA of peak-to-peak current difference. In contrast, the chargebalancing multiple waveform showed 175 nA of peak-to-peak current difference (Figure 3C). Consequently, the chargebalancing multiple waveform resulted in a significantly lower charge difference (106.5 ± 8.99 pC), compared to the conventional triangular multiple waveform (267 ± 31.08 pC) among pulses (Figure 3D, n = 6). Dual Background Subtraction in CBM-FSCV. Because the charge-balancing multiple waveform diminished charge differences among multiple pulses under restricted ionic conditions, we used the background subtraction method to remove the rest of the background differences. In Figure 4, all voltammograms in each series are subtracted from their own last voltammogram (dynamic subtraction), and then a single reference point is chosen from which to subtract the rest of the time points, in the same manner as a conventional background subtraction method. This method is termed “dual background subtraction” due to the fact that there are two reference voltammograms: one within each train series; and one initial reference used for original background subtraction. As shown in the voltammograms in Figure 4, after DA injection, it is possible to detect the DA signature with dynamic subtraction alone because the adsorption kinetics are changed by varying repetition time.28,36 However, there is still a minor background current difference existing in the dynamic subtracted F
DOI: 10.1021/acs.analchem.6b02605 Anal. Chem. XXXX, XXX, XXX−XXX
Article
Analytical Chemistry
does not adsorb onto CFM,36 and this feature gives AA its relatively low sensitivity and fast decaying pattern (Figure 6C, top and bottom). In the case of AA, there were no significant differences between the first and last voltammogram peak currents (n = 5). Overall, the selectivity of CBM-FSCV over DA can be achieved by taking the sixth or later voltammogram. For practical applications, the present data suggest that the sixth dual background-subtracted voltammogram can serve as the boundary signal in order to significantly minimize the interference from electroactive interferents, such as DOPAC and AA, while achieving maximal sensitivity to DA using CBMFSCV. As a result, delaying analysis until the sixth CBM-FSCV dual background-subtracted response to DA yielded a sensitivity for DA of 85.40 ± 14.30 nA/μM and a limit of detection of 5.8 ± 0.9 nM (3 times the RMS noise in a cyclic voltammogram, n = 3) at CFMs in vitro which is sufficient for quantifying DA in vivo in relevant physiological ranges16 (Figure 7).
differences from the reference time point were averaged hourly from both data sets and compared from the beginning of data collection (hour 1). As shown in Figure 5C, significant timedependent changes in charge amount were observed using the conventional background subtraction method (2-way ANOVA, p < 0.0001, Sidak’s multiple comparisons test), whereas in marked contrast, no significant changes in charge amount were observed using the dual background subtraction method (2-way ANOVA, Sidak’s multiple comparisons test). CBM-FSCV of Dopamine and Other Electroactive Interferents. It is well-known that DA highly adsorbs to CFMs at relatively low concentrations (
