Fast cyclic square-wave voltammetry to enhance neurotransmitter

Oct 25, 2018 - Although fast-scan cyclic voltammetry (FSCV) has been widely used for in vivo neurochemical detection, the sensitivity and selectivity ...
7 downloads 0 Views 1MB Size
Subscriber access provided by Kaohsiung Medical University

Article

Fast cyclic square-wave voltammetry to enhance neurotransmitter selectivity and sensitivity Cheonho Park, Yoonbae Oh, Hojin Shin, Jaekyung Kim, Yu Min Kang, Jeongeun Sim, Hyun U Cho, Han Kyu Lee, Sung Jun Jung, Charles D. Blaha, Kevin E. Bennet, Michael L. Heien, Kendall H. Lee, In Young Kim, and Dong Pyo Jang Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b02920 • Publication Date (Web): 25 Oct 2018 Downloaded from http://pubs.acs.org on October 25, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Fast cyclic square-wave voltammetry to enhance neurotransmitter selectivity and sensitivity Cheonho Park1, Yoonbae Oh2, Hojin Shin1, Jaekyung Kim3, Yumin Kang1, Jeongeun Sim1, Hyun U Cho1, Han Kyu Lee4, Sung Jun Jung4, Charles D. Blaha2, Kevin E. Bennet2,5, Michael L. Heien6, Kendall H. Lee2, In Young Kim1, Dong Pyo Jang*1 1Department

of Biomedical Engineering, Hanyang University, Seoul, Republic of Korea of Neurologic Surgery, Mayo Clinic, Rochester, MN, USA 3Department of Neurology, University of California-San Francisco, San Francisco, CA, USA 4Department of Biomedical Science, Hanyang University, Seoul, Republic of Korea 5Division of Engineering, Mayo Clinic, Rochester, MN 55901, USA 6Department of Chemistry and Biochemistry, University of Arizona, Tucson, Arizona 85721, USA 2Department

ABSTRACT: Although fast-scan cyclic voltammetry (FSCV) has been widely used for in vivo neurochemical detection, the sensitivity and selectivity of the technique can be further improved. In this study, we develop fast cyclic square-wave voltammetry (FCSWV) as a novel voltammetric technique that combines large amplitude cyclic square-wave voltammetry (CSWV) with background subtraction. A large-amplitude, square-shaped potential was applied to induce cycling through multiple redox reactions within a square pulse to increase sensitivity and selectivity when combined with a 2D voltammogram. As a result, FCSWV was significantly more sensitive than FSCV (n=5 electrodes, two-way ANOVA, p=0.0002). In addition, FCSWV could differentiate dopamine from other catecholamines (e.g., epinephrine and norepinephrine) and serotonin better than conventional FSCV. With the confirmation that FCSWV did not influence local neuronal activity, despite the large amplitude of the square waveform, it could monitor electrically induced phasic changes in dopamine release in rat striatum before and after injecting nomifensine, a dopamine reuptake inhibitor.

INTRODUCTION Background subtracted fast scan cyclic voltammetry (FSCV) at carbon fiber microelectrodes (CFM) is a wellrecognized tool in neuroscience research.1-3 It has been widely used to detect neurochemicals in vivo because of its high sensitivity, chemical selectivity, and superior spatial and temporal resolution relative to other in vivo neurochemical monitoring techniques4-5. Efforts to optimize the sensitivity and specificity of the technique for neurochemical detection have applied specific voltammetric waveforms. For example, a triangle shaped waveform consisting of a +1.3 V peak, a holding potential of -0.4 V between scans, and a sweep rate of 400 V/s improved dopamine (DA) detection. Similarly an Nshaped voltammetric waveform has been used to improve serotonin (5-hydroxytryptamine, 5-HT) detection and reduce electrode fouling.6 Meanwhile, a sawhorse waveform with a prolonged switching potential was more selective for adenosine than a conventional triangular waveform because it maximized adenosine oxidation.7 In addition to modifying individual waveforms, we have reported on using two, identical, triangle-shaped waveforms separated by a short interval at a holding potential.8 This scan can discriminate between analytes and mitigates confounding factors on the basis of differences in adsorption properties.

Adsorption of a species on the surface of a CFM is a significant factor affecting the voltammetric response in FSCV,9-12 and a higher adsorption strength increases the sensitivity for biogenic catecholamines, such as DA.9, 13-15 Recent studies using multiple waveforms based on CFM surface adsorption properties have been used to measure tonic levels of DA or slow DA changes. For example, Atcherley et al. developed fast-scan controlled-adsorption voltammetry (FSCAV) to measure tonic, extracellular DA concentrations by exploiting CFM surface adsorption.16-17 FSCAV was able to estimate absolute tonic DA concentrations in vivo over a 90 min period.16 In addition, Oh et al. used a multiple waveform FSCV with 100 consecutive charge-balancing waveforms combined with dual background subtraction to minimize temporal variations in background capacitive currents. It was highly selective for DA against 3,4-dihydroxyphenylacetic acid and ascorbic acid, two major chemical interferents in the brain.18 Repeated waveforms have been extended to more complex waveform shapes to improve selectivity with FSCV. For example, Jo et al. recently showed improved identification of the catecholamines DA and norepinephrine (NE) in the brain by combining rectangular and triangular waveforms in voltammetric recordings.19 Two triangular waveforms of

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

different sizes were used to identify hydrogen peroxide (H2O2) in vivo, which is masked by a local pH change.20 Besides these approaches, square-wave voltammetry (SWV) has been developed and considered one of the most powerful electrochemical techniques for analytical applications and electro-kinetic measurements by unifying the advantages of pulse techniques, cyclic voltammetry, and impedance techniques.21 SWV consists of a staircase potential ramp with two equal height and opposite direction square-shaped potential pulses. Cyclic square-wave voltammetry (CSWV), or square wave voltammetry in the fashion of cyclic voltammetry, provides a wealth of electrochemical information, which may consist of multiple voltammetric curves (i.e., upward, downward and net components for both forward and reverse sweeps).22 However, the use of CSWV for in vivo neurochemical detection has not been reported due to its relatively slow scan rate. In this study, we develop fast cyclic square-wave voltammetry (FCSWV) as a novel voltammetric technique combining large amplitude CSWV and background subtraction. A relatively larger amplitude square-shaped potential (> 0.1 V), compared to conventional CSWV, was used in FCSWV to induce analyte oxidation and reduction in a single staircase potential ramp. In addition, a newly designed 2D voltammogram converted from whole current data (voltammogram) after background subtraction was developed to effectively differentiate DA from other analytes. Experimental Section Data acquisition and analysis Data were acquired with a commercial electronic interface (NI USB-6251, 16 bit, National Instruments) with a basestation PC computer and software written in the LabVIEW programming environment (National Instruments, Austin, TX, USA). A current-to-voltage amplifier without an analog filter was used to preserve sharp current responses to the square pulse in FCSWV. The NI USB-6251 was also used to synchronize the applied waveforms and flow injection analysis. After data collection, all waveform parameters and operations, data acquisition and transmission, background subtraction, and signal averaging were performed with custom software control. Data were saved to the base-station computer hard drive as a sequence of unsigned, 2-byte integers for offline processing in MATLAB (MathWork Inc., Natick, MA, USA). GraphPad Prism 6 (GraphPad Software, San Diego, CA, USA) was used to generate figures and statistics. All data are presented as mean ± standard deviation (SD) for “n” electrodes. The same group of electrodes was used for comparisons between FSCV and FCSWV. Root mean square error (RMSE) was calculated from the estimated concentration data and each electrode in the principal component regression (PCR) experiment. Fast scan cyclic voltammetry (FSCV) To compare with FCSWV, a conventional FSCV waveform was used.23 A scan rate of 400V/s and holding potential of 0.4V versus Ag/AgCl was used between scans, the switching potential was 1.3V, and scans were repeated every 200ms to match the repetition time (5Hz) of the proposed waveform.

Page 2 of 10

Fabrication of carbon fiber microelectrodes (CFMs) Electrodes were fabricated as previously described.24 T300 carbon fiber (Cytec Thornel, Woodland Park, NJ, USA) was used for all experiments. The exposed carbon fiber was trimmed to a final length of 70−120μm with a scalpel.25 The Ag/AgCl reference electrode was fabricated by chloridizing a 31-gauge silver wire.26 Chemicals DA, NE, epinephrine (EP), and 5-HT were dissolved at a stock concentration of 1 mM in 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, pH 7.4) The stock solution was diluted to the final concentrations in Tris buffer before starting the flow injection experiments. All chemicals, including nomifensine, were purchased from Sigma Aldrich (St. Louis, MO, USA). Principal component regression analysis To compare the selectivity of FSCV and FCSWV, PCR analysis was conducted with MATLAB.23, 27 To prepare the PC regression matrix, 0.5μM, 1.0μM, 1.5μM, and 2.0μM of DA, NE, and 5-HT were analyzed at 5Hz. Data for 1.0μM, 2.0μM, 3.0μM and 4.0μM were obtained for EP. Each solution was measured with both FSCV and FCSWV. The number of principal components was chosen to explain 99.5% of the original data. To confirm that the regression matrix was constructed precisely, half of the obtained data were used as a training set, and the other half were used as a test set. To verify the predictive ability of the constructed regression matrix, voltammograms of unknown mixtures were evaluated. The unknown mixtures were 1μM DA and 2μM EP, 1μM DA and 1μM NE, and 1μM DA and 1μM 5-HT. CA2+ imaging protocol to evaluate physiological effects of FCSWV The voltage amplitude of the FCSWV waveform is much larger than the voltage amplitude of the conventional FSCV waveform. Thus, it is important to test whether currents generated by FCSWV voltages induce electrical stimulation of brain tissue adjacent to the CFM. For this purpose, calcium imaging with the below protocol was selected. Mouse brain slices of the striatum were prepared and Ca2+ imaging was performed with IACUC approval. After rapid extraction of the brain, 300-μm thick horizontal striatal slices were prepared with a Vibratome 1000 (The Vibratome Company, Saint Louis, MO, USA). Brain slices were incubated for a period of 60 to 90 minutes with oxygenated artificial cerebrospinal fluid solution (28–30°C) (in mM; NaCl 125, KCl 2.5, CaCl2 2, NaHCO3 26, NaH2PO4 1.25, MgCl2 1, glucose 25, pH 7.4 when bubbled with 95% O2 and 5% CO2) in a submersion-type recording chamber. For Ca2+ imaging, Ca2+ responses of striatal neurons in brain slices were measured with the fluorescent Ca2+ probe (indicator) Fluo-3 AM (10μM, Enzo Life Science, Seoul, Korea) mixed with 1μl pluronic acid (20% in DMSO, Life Technologies, Seoul, Korea) in DMEM medium with 10% FBS (Fetal bovine serum) for 40 min at 37˚C. Fluo-3 AM treated slices on coverslips were mounted on a chamber (12mm Chamlide AC, 500μl total volume, Live Cell Instrument, Seoul, Korea) and placed on an inverted

ACS Paragon Plus Environment

Page 3 of 10 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

microscope (Olympus IX70, Shinjuku, Tokyo, Japan). Cells were excited with an LED source pE-100 (CoolLED, Andover, Hampshire, UK) at 470nm, and emissions were recorded at 535nm. Fluorescent emissions were recorded every 0.5s. Intracellular Ca2+ concentrations ([Ca2+]i) were measured by digital video microfluorometry with an intensified charge-coupled-device (CCD) camera (QIClick, QImaging, Surrey, Canada) coupled to the microscope and a computer with MetaMorph® NX software (Molecular Devices, San Jose, CA, USA). Ca2+ responses were presented as a pseudo ratio (∆F/F) to compare fluorescence intensities because Fluo-3 AM is a non-ratiometric Ca2+ indicator. Single-wavelength values differ depending on dye uptake in respective cells. The pseudo ratio was calculated with the following formula.28 ∆F/F = (𝐹1 ― 𝐹𝑏𝑎𝑠𝑒) / 𝐹𝑏𝑎𝑠𝑒

In vivo experiments Adult male Sprague Dawley rats weighing 250-350 g were used for this study (n = 3). NIH guidelines were followed for animal care and the Hanyang University Institutional Animal Care and Use Committee approved the experimental procedures. To confirm that FCSWV can measure neurotransmitters in the brain, evoked DA release was measured in rat striatum. To evoke DA release, the medial forebrain bundle (MFB) was stimulated. The evoked DA release in anesthetized rats was measured twice at 10-minute intervals with FCSWV. To confirm that the signal obtained was from DA, a pharmacological test was used. Nomifensine (20 mg/kg, i.p.), a DA reuptake inhibitor, was intraperitoneally injected to change DA extracellular levels. Evoked DA release was measured twice at 10-minute intervals 30 minutes after drug injection.

F1 = measured intensity after stimulation Fbase = measured intensity before stimulation

Figure 1. Schematic fast cyclic square-wave voltammetry (FCSWV) waveform and voltammogram process. (A) FCSWV consists of a staircase potential ramp modified with square-shaped potential pulses. At each step, two, equal height (>0.1V) and oppositely directed potential pulses are imposed. (B) Background subtraction processing in FSCWV. (C) 2D voltammogram reconstruction processing. The voltammogram from a biphasic shape square wave at each step of the ramp is stacked vertically at the ramp potential and displayed with a pseudo-color map.

RESULT AND DISCUSSION Fast cyclic square-wave voltammetry (FCSWV) waveform The terminology used throughout this study was adopted from Helfrick and Bottomley22 to avoid confusion. As shown in Fig. 1A, the FCSWV waveform consists of a square wave oscillation overlaid on a symmetric staircase waveform. Two equal height and oppositely directed square-shaped potential pulses (upward and downward potentials) are superposed on a

single staircase. The staircase starts with a holding potential (EHolding), increases with each stair step (EStaircase) up to a peak potential (EPeak), and then decreases back to EHolding. The magnitude of the upward and downward potentials (ESW) in a square cycle was up to 0.4V, which induces both DA oxidation and reduction during each pulse. In this experiment, the stair steps (EStaircase) and duration of a single stair step (τ) were fixed at 12.5mV and 0.5ms, respectively. When the peak

ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

potential (EPeak) was 0.9V, the FCSWV duration including a forward and reverse sweep was about 88ms. FCSWV was applied every 200ms with 112ms at the holding potential (5Hz repetition frequency). In conventional CSWV, voltammograms were reconstructed with currents sampled at the end of each square pulse to discriminate from the charging current. That is, the end point current responses associated with upward and downward potential pulses compose the forward and backward voltammograms.21 In contrast, in FCSWV, the entire current response to the waveform was sampled at a high sampling rate (1.0 MHz), and background subtraction was used to remove capacitive background current from the whole current response. As shown in Fig. 1B, the whole voltammogram after DA injection was subtracted from the voltammogram before DA injection in the same manner as a conventional background subtraction. The new background subtracted voltammogram could be obtained with multiple redox reactions as shown in Fig. 1B. However, in practice, the background current (capacitive current) could not be completely removed with the background subtraction method because adding DA influences the capacitive current. Thus, background correction processing was used to further minimize residual capacitive current. Detailed background correction methods are described in the supplementary information. For more intuitive FCSWV analysis, a 2D voltammogram representation was designed in this study. When EHolding was -0.2V, EPeak was 0.9V, and EStaircase was 12.5mV, the staircase waveform had 176 steps (=2.2V/0.0125V) with a step duration (τ) of 0.5 ms and sampling rate of 1 MHz, a single step is sampled at 500 points. That is, each voltammogram is sampled by 88, 000 data points, 500 data points for each of 176 steps. In this study, a 1D voltammogram (1 x 88000) was restructured into a 2D voltammogram by graphing each square pulse on the y-axis, the triangle potential on the x-axis, and the current in false color (500 x 176). The current response at a 0.2V step potential, depicted in Fig. 1C, consisted of 250 points for the upward potential (0.55V) and 250 points for the downward potential (-0.25V). The 500 data points were vertically filled at the 0.2V staircase potential column. This process was repeated for each step potential to make a 2D voltammogram. Finally, the reconstructed 2D voltammogram was represented in a pseudo color plot (Fig. 1C). 2D voltammograms from each FCSWV scan were reconstructed at 5Hz and viewed as shown in the supplementary video. FCSWV waveform optimization To investigate the effect of FCSWV parameters on the DA response, three parameters (Amplitude of square wave (ESW), Holding potential (EHolding), and Peak potential (EPeak)) were tested with EStaircase=12.5mV, τ = 0.5ms, and 5 Hz FCSWV repetition frequency. All experiments were conducted with 5 μM DA. The switching potential is considered an important factor in electrochemistry because it relates to chemical sensitivity or unwanted chemical responses. In FCSWV, the maximum potential is defined as the sum of staircase peak potential (EPeak) and square pulse amplitude (Esw). In this experiment, the current response may become unstable or an unwanted response according to the switching potential.29 Therefore, switching potentials between 1.0V to 1.4V were tested at EPeak from 0.6V to 1.0V and a fixed Esw of 0.4V. DA sensitivity slightly increased as the switching potential increased up to 1.4V and reached a maximum sensitivity at 1.4V (Fig. 2A). This result is similar to a previous report.11 However, switching potentials did not differ significantly

Page 4 of 10

(n=5, one-way ANOVA with Tukey’s multiple comparison test). Because the background current becomes unstable when the switching potential exceeds 1.3V we used a 1.3-V switching potential (EPeak=0.9V and Esw=0.4V) in subsequent experiments.29 The holding potential, which is the resting potential between FCSWV waveforms, was evaluated from 0.0V to -0.3V. DA responses did not differ significantly (Fig. 2B, n=5, one-way ANOVA with Tukey’s multiple comparison test). This result was unexpected because, in conventional FSCV, DA sensitivity increased at negative holding potentials due to the positive charge of DA. However, a similar result has been reported previously.30 Two factors may have affected the DA response in FCSWV. First, Esw =0.4V covers the reduction potential (-0.2 V) even when holding at 0.0V. Second, the redox chain reaction at Esw=0.4V in FCSWV dominates DA sensitivity, as shown in Fig. 2C. When Esw increases from 0.1V to 0.4V, the electrochemical response of DA increased, as shown in Fig. 2C. The DA oxidation peak current at Esw=0.4V is about 15 times larger than at Esw=0.1V. A large amplitude Esw, such as 0.4V, covers both DA oxidation and reduction in the upward and downward square pulses. For example, when a square wave with Esw=0.4V is superposed on a 0.2V potential step in a staircase waveform, the square cycle sweeps from 0.6V upward potential (250μsec) to -0.2V downward potential (250μsec). In this case, DA was oxidized to dopamine o-quinone (DOQ) during the upward and downward potentials, and most DOQ reduced back to DA immediately due to the short diffusion time. This redox chain reaction occurred when the magnitude of Esw covered the DA oxidation and reduction potentials within a single square pulse. A higher Esw leads to significantly higher DA sensitivity when compared with a lower Esw as shown in Fig. 2C (n=5, p