Temporal Differentiation of pH-Dependent ... - ACS Publications

Aug 8, 2014 - Removal of Differential Capacitive Interferences in Fast-Scan Cyclic Voltammetry. Justin A. Johnson , Caddy N. Hobbs , and R. Mark Wight...
0 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/ac

Temporal Differentiation of pH-Dependent Capacitive Current from Dopamine Kenji Yoshimi* Department of Neurophysiology, Juntendo University School of Medicine, Hongo 2-1-1, Bunkyo-ku, Tokyo 113-8421, Japan

Adam Weitemier RIKEN BSI, Wako-shi, Saitama, 351-0198, Japan S Supporting Information *

ABSTRACT: Voltammetric recording of dopamine (DA) with fast-scan cyclic voltammetry (FSCV) on carbon fiber microelectrodes have been widely used, because of its high sensitivity to dopamine. However, since an electric double layer on a carbon fiber surface in a physiological ionic solution behaves as a capacitor, fast voltage manipulation in FSCV induces large capacitive current. The faradic current from oxidation/reduction of target chemicals must be extracted from this large background current. It is known that ionic shifts, including H+, influence this capacitance, and pH shift can cause confounding influences on the FSCV recordings within a wide range of voltage. Besides FSCV with a triangular waveform, we have been using rectangular pulse voltammetry (RPV) for dopamine detection in the brain. In this method, the onset of a single pulse causes a large capacitive current, but unlike FSCV, the capacitive current is restricted to a narrow temporal window of just after pulse onset (1 ms) parts of the pulse, it is difficult to measure, both initial and later phase at a proper amplification. Second, with a steep current change, a slight microsecond of temporal error can cause a large variance of recorded current. We introduced a low-pass filter of 2 ms time constant (Figures 1C and 1D). Figure 1C is the waveform of the same recording as Figure 1B, but with a 2-ms low-pass filter. Figure 1D is the same recording as Figure 1C with a longer time scale, including the entire 30 ms pulse. We found that introduction of a lowpass filter and higher gain of amplification improved the measurement of low concentrations of chemicals. FSCV recording on the same carbon fiber microelectrode is shown in Figures 1E−H. Figure 1F shows the pH-dependent changes in the current over the entire FSCV waveform. This

Figure 1. Capacitive current of carbon fiber surface and the effect of pH change, as indicated by the applied potential waveforms and the recorded current change: (A) rectangular applied voltage onset from 0 to 0.2 V; (B and C) current response just after the 0.2 V pulse onset as depicted in panel (A). The carbon fiber was dipped in pH 5.3 (green square) or pH 9.3 (blue diamond) solutions. The low-pass filter was off for panel (B), while the low-pass filter with a time constant of 2 ms was applied for panel (C). The current was recorded at 500 nA/V and digitized by a D/A converter at 10 kHz (every 0.1 ms). (D) Same as that depicted for panel (C) but showing the entire response to a pulse of 30 ms. (E) Triangular applied voltage for FSCV. (F) Current response at FSCV waveform in pH 5.3 (green line) and pH 9.3 (blue dot). The current was recorded at 500 nA/V with a low-pass filter of 0.2 ms time-constant and digitized by D/A converter at 50 kHz (every 0.02 ms). G: Current response at FSCV waveform in PBS pH 7.3, with (magenta) or without (blue) 10 μM dopamine. (H) Differential current of 10 μM dopamine at FSCV. The waveform in PBS pH 7.3 was subtracted as the background.

fact explains why the well-known background-subtracted current−voltage curve of pH response in FSCV has a broad peak. While the pH-induced change was broad, the dopamine reaction had distinctive peaks of oxidation and reduction (see Figures 1G and 1H). Figure 1G is the original measurement, and Figure 1H is the background-subtracted dopamine waveform, after the subtraction of the current in standard buffer (pH 7.3). These results indicate the pH response is 8578

dx.doi.org/10.1021/ac500706m | Anal. Chem. 2014, 86, 8576−8584

Analytical Chemistry

Article

different from electrochemical (faradic) currents derived from oxidation/reduction reactions. The current waveform for each pulse is presented as the temporal change from the pulse onset, instead of conventional cyclic voltammogram. Because there are only two voltage points in rectangular pulses, the abscissa in voltage was not adequate to visualize the waveform. Differential Chemical Responses on the Rectangular Waveform. The differential current induced by pH shift (±0.2) and 1 μM dopamine showed a different delay over the voltage pulse onset (Figure 2A). The change in backgroundsubtracted current after pH shift had the same temporal characteristics as the original capacitive current (Figure 1D), but differed depending on the direction of the shift. This result also shows the current change induced by pH change is mainly derived from the change in the capacitance on the carbon fiber surface. From Figure 2A, we estimated that the differential current, excluding the initial 5 ms, would be less sensitive to pH changes, and current during the initial 5 ms would be selective to pH. Differential current change by other chemicals is presented on Figures 2B and 2C. Ascorbate (10 μM) and DOPAC (10 μM) showed constant electrical current throughout the 30-ms 0.2 V pulse, suggesting low adsorption (Figure 2B). On the other hand, we had unexpected results with adenosine (10 μM) and serotonin (5-HT) (1 μM) (Figure 2C). While 0.2 V was below the oxidation potential of these chemicals, they showed responses similar to basic pH changes. However, the oxidative current of 5-HT and adenosine is observed when the holding and pulse potential were shifted +0.2 V (from 0.2 V to 0.4 V) or +1.2 V (from 1.2 V to 1.4 V), respectively (see Figures 2D and 2E). Note the reductive negative current at the offset (30 ms) for dopamine, DOPAC, and 5-HT (Figures 2A, 2B, and 2D), and its absence for ascorbate and adenosine (Figures 2B and 2E), which are consistent with the reversible and irreversible oxidation of these molecules.33−36 Reactions to the Additional Chemicals. Detection of dopamine and other electrochemically active chemicals was evaluated in vitro at carbon fibers in PBS (Figure 3). Microelectrodes were placed in the flow of PBS, and the flow was switched to test solutions of 0.2 μM dopamine (DA), acidic pH 7.1 PBS (−pH), basic pH 7.5 PBS (+pH), 10 μM ascorbate (AA), 10 μM DOPAC (DC), 10 μM adenosine (Ade), and 0.2 μM 5-HT solutions in PBS. The concentration of each chemical shown in Figure 3 was selected in consideration of the largest changes observed following MFB electrical stimulation-evoked release.15,33−36 Here, the range of dopamine and 5-HT was 0.2 μM, and the other chemicals were set to a larger concentration (×50). The in vivo result in Figure 4 supports the belief that the dopamine concentrations and pH values were within a reasonable range of physiological changes. Test solutions were sequentially applied for 10 s with intervals of 20−40 s (see Figure 3). Three types of voltage application were tested on the same carbon fiber: FSCV, constant-potential amperometry, and RPV. Since the amplitude of differential current of constantpotential amperometry (Figure 3C) and RPV (Figures 3D−3G) is lower than that obtained in FSCV, it is essential to eliminate noise from the power source. It was eliminated as previously described,29,30 by averaging 20 ms of measurement, the whole one-cycle of 50 Hz power supply in our area (i.e., 16.7 ms in 60 Hz area). Therefore, we set three 20-ms-long windows (see Figures 3D and 3E)t0−20, t5−25, and t10−30

Figure 2. Differential current response at rectangular pulses. Background subtracted current and the temporal properties of pH and dopamine reaction at rectangular pulses from 0 to 0.2 V. Low-pass filter was tau = 2 ms. The current while the electrode was in PBS (pH 7.3) immediately before changing the chemical solution was subtracted as the background. (A) Temporal reaction of acidic pH (7.1, −0.2 change), basic pH (7.5, +0.2 change), and dopamine (DA: 1 μM) on the rectangular pulse onset. After dopamine, the current was increased for a longer time duration, compared to pH. (B) Temporal reaction of ascorbate (AA: 10 μM) and DOPAC (DC: 10 μM). (C) Temporal reaction of adenosine (Ade: 10 μM) and serotonin (5-HT: 1 μM). (D) Result of 5-HT at 0.2−0.4 V rectangular pulses (+0.2 V offset). 5HT was oxidized at 0.4 V, while not at 0.2 V. (E) Result of adenosine at 0.8−1.0 V (blue) and 1.2−1.4 V (red) rectangular pulses (+0.8 V and +1.2 V offset). The oxidation current of adenosine was evident at 1.4 V, while only adsorption was observed up to 1.0 V.

over the 30-ms rectangular pulsesand also the average just before the pulse onset (t80−100, Figure S3 in the Supporting Information). Twenty milliseconds (20 ms) from the rectangular pulse onset, t0−t20, includes the pH-sensitive capacitive current. The intermediate (t5−25) is less sensitive to pH but sensitive to dopamine. These characteristics were substantially the same when the background buffer was substituted by a physiological ACSF (see Figure S6 in the Supporting Information). The last t10−30 also detects 8579

dx.doi.org/10.1021/ac500706m | Anal. Chem. 2014, 86, 8576−8584

Analytical Chemistry

Article

As shown in Figure 2B (and also Figure 3D), ascorbate and DOPAC reacted throughout the 0.2 V pulse. A dopamineselective value, t′5−10 = (t5−25) − (t10−30) was calculated, in which the response to ascorbate and DOPAC of t5−25 are suppressed (Figure 3F). As a pH-selective value, we calculated t′0−5 = (t0−20) − (t5−25), which is highly sensitive to pH changes (Figure 3G). Because the amplitude and S/N ratio of t′5−10 was not very high, it may be useful only when transient changes in ascorbate and DOPAC are evident. In addition, the temporal change of ascorbate can be evaluated by the averaged current at 0.0 V (t80−100, just before pulse onset, Figure S3 in the Supporting Information). At the peak potential of dopamine oxidation at FSCV (Figure 3B, representing the white horizontal line in Figure 3A), pH current was as large as dopamine. Note that adenosine made a marked current at the peak (1.3 V) of the triangular waveform. At constant potential amperometry of 0.2 V (Figure 3C), the response to pH was much less than dopamine. The response to ascorbate and DOPAC were evident. At RPV, t5−25 (Figure 3C) and t′0−5 (Figure 3D) were selective to dopamine and pH, respectively. As known previously,10,11 FSCV also successfully discriminated dopamine and pH after post-hoc PCR analysis (see Figures S4 and S5 in the Supporting Information). In short, FSCV current at the dopamine oxidation potential was sensitive to dopamine, pH, DOPAC, and 5-HT, and was relatively less sensitive to ascorbate and adenosine. Note the considerable drift of baseline, which reflects high adsorption of dopamine, adenosine, and 5-HT. In contrast, constant potential amperometry and t5−25 showed negligible response to pH, and the dopamine detection was more significant in t5−25. The baseline was stable throughout the recording and the responses to pH, adenosine, and 5-HT were small, but the response to 10 μM ascorbate and 10 μM DOPAC were larger than that of 0.2 μM dopamine. It appeared that FSCV, compared to RPV, showed a greater delay in recovery of the signal after stimulation. It has been mentioned that adsorption of dopamine on the carbon surface makes the carbon fibers sensitive to dopamine detection, but it also limits the temporal resolution over seconds.34 The temporal distortion by adsorption of dopamine was evaluated as previously reported.12 The adsorption index is defined as the ratio of the remaining amplitude at 10 s after the initiation of PBS washout. When there is no adsorption, the adsorption index would be zero (0). Although the adsorption of 1.0 μM dopamine was negligible upon amperometry (0.01 ± 0.03), it was evident at FSCV (0.28 ± 0.04) and t5−25 (0.23 ± 0.03). Values are expressed in terms of mean ± s.e.m. of four carbon fiber microelectrodes. All four electrodes showed a slightly smaller adsorption at t5−25 than FSCV. At RPV, the higher oxidation current just after the rectangular pulse onset (Figure 2A) strongly suggests the adsorption at 0.0 V duration between pulses. For this reason, constant potential amperometry has been preferred to evaluate fast kinetics, such as studies for reuptake by dopamine transporters. Dopamine and pH Detection In Vivo. The measurement of dopamine and pH change was also confirmed in the brain (see Figure 4). A carbon fiber was placed in the striatum of deeply anesthetized mice (n = 4). Dopamine release was evoked by electric stimulation of ascending monoamine fibers (Medial Forebrain Bundle (MFB)), and brain pH was lowered by CO2 inhalation, as described previously.26,27 While the magnitude of the response to MFB stimulation and CO2

Figure 3. In vitro reaction to various chemicals in a flow cell. A carbon fiber microelectrode was placed in a constant flow of PBS (pH 7.3) and the solution was switched to PBS with 0.2 μM dopamine (DA), pH 7.1, pH 7.5, 10 μM ascorbate (AA), 10 μM DOPAC (DC), 10 μM adenosine (Ade), and 0.2 μM serotonin (5-HT). (A) Color plot of background-subtracted FSCV current. The triangular waveform is indicated by the bold blue line on the left. Each sweep was made from the bottom (onset) to the top (10 ms) of the plot. The potential of maximal dopamine oxidation is shown by a white horizontal line. Scale bar on the right indicates −50 nA (dark blue) to +75 nA (green). (B) Differential current at the peak potential of dopamine oxidation during FSCV. (C) Raw current during constant potential amperometry. (D) Color plot of background-subtracted rectangular pulse current. Each pulse was made from the bottom (−20 ms) to the top (60 ms). The timing of the pulse is indicated by a bold blue rectangle on the left side of the color plot. The time window for t5−25 is indicated by the color magenta. Scale bar on the right indicates −5.0 nA (dark blue) to +7.5 nA (green). (E) Current average of t0−20 (green), t5−25 (magenta), and t10−30 (blue). The baselines were vertically shifted by 1 nA. The range of averaging for t5−25 is indicated with magenta line on the color plot (panel (D)). (F) Relative sensitivity of compounds between early (t′5−25) and late (t′10−30) time windows of the DA-sensitive phase of the voltage pulse [t′5−10 = (t5−25) − (t10−30)]. (G) Relative sensitivity of compounds between the pH sensitive phase (t′0−20) and early DA-sensitive phase (t′5−25) of the voltage pulse [t′0−5 = (t0−20) − (t5−25)]. Recordings of FSCV (panel (B)) and t′0−5 (panel (F)) were sensitive to pH changes, while amperometry (panel (C)) and t5−25 (panel (E)) were not. The influence of ascorbate and DOPAC was small in FSCV (panel (B)), while the large reaction was observed in amperometry (panel (C)) and t5−25 (panel (E)).

dopamine, but was less sensitive. The sensitivity of dopamine detection for each method is reviewed in Table S1 and Figure S2 in the Supporting Information. 8580

dx.doi.org/10.1021/ac500706m | Anal. Chem. 2014, 86, 8576−8584

Analytical Chemistry

Article

similar to that of constant potential amperometry. In contrast, t′0−5 was selective to the CO2 response. Table 1 details the relative selectivity of each method and under each condition (MFB stimulation and CO2 inhalation) during in vivo recordings. Nomifensine is a dopamine uptake inhibitor and is known to enhance dopamine overflow.1,12,32 Nomifensine administration enhanced the MFB response but not the CO2 response in all four measurements (dotted magenta line in Figure 4; also see Table 1), confirming that the MFB response was mainly dopamine release. A small MFB response appearing in t′0−5 after nomifensine indicates some contamination of dopamine current at this high value, as expected from Figure 2A. The adsorption index was also calculated for large in vivo MFB responses after nomifensine (see Figure S7 in the Supporting Information). Just like in vitro, the adsorption index was larger for FSCV (0.092 ± 0.015, n = 4) than t5−25 (0.026 ± 0.025), and little for amperometry (0.006 ± 0.006). To support the in vivo detection of dopamine and pH changes (Figure 4), we evaluated the influences of divalent cation of Ca2+. The influence of pH shift and Ca2+ were tested especially carefully, since they can change directly in association with synaptic neurotransmission. Ca2+ in the extracellular space is reported to reduce by 0.1 mM after intense local stimulation in slice.18,21,22 The decrease in Ca2+ concentration from 0.1 to 0.0 mM induced intense current changes at FSCV and t′0−5. However, it was much smaller in the physiological concentration change from 1.1 mM to 1.0 mM (Figure 5) and in ACSF (see Figures S6C and S6D in the Supporting Information, from 1.25 mM to 1.15 mM). The current response to pH was not very influenced by Ca2+ concentrations in RPV (see Figure S6H in the Supporting Information), indicating different mechanisms of interaction with the carbon fiber surface. In order to evaluate the influence of background buffer, PBS was switched to ACSF during a single 120 s of recording (see Figures S6E−H in the Supporting Information). The dopamine response in ACSF was smaller than that in PBS at FSCV (0.73 ± 0.08, n = 3) and t5−25 (0.70 ± 0.01), the result of FSCV is consistent with previous studies.22.23 Unlike dopamine, it was difficult to determine the change in pH sensitivity for FSCV in ACSF, because of the marked change in differential waveform of pH response (S6G). Interestingly, the amplitude of pH response was unchanged at RPV (S6H, ratio of t′0−5 amplitude in ACSF/PBS: 0.99 ± 0.05, n = 3).

Figure 4. Recording in the anesthetized mouse brain (in vivo). A carbon fiber was placed in the middle of the right striatum. MFB stimulation was given to evoke dopamine release in the striatum and CO2 inhalation was given to lower the brain pH. The right MFB was electrically stimulated at time 0 (24 pulses at 30 Hz); 55−60 s later, 18% CO2 was mixed into the anesthetizing gas (1% isoflurane, 30% O2 and 70% N2O). Flow delay from gas-mixing valve to the mouse was estimated as 5.0 s. Blue lines indicate the average of four mice; dotted magenta lines indicate the result of the repeated procedure after dopamine uptake inhibitor (nomifensine, 7 mg/kg, s.c.) administration. (A) FSCV at the peak potential of dopamine oxidation, (B) constant potential amperometry, (C) t5−25, and (D) the t′0−5 value of rectangular pulses of 0−0.2 V. Recordings with FSCV (panel (A)) and t′0−5 (panel (D)) were sensitive to CO2, while amperometry (panel (B)) and t5−25 (panel (C)) were not. Note that nomifensine enhanced MFB but not CO2 reactions. Negative spike in (panel (D)) near time 0 is an artifact by stimulation pulses.

inhalation was identical in FSCV, amperometry showed selective detection of MFB stimulation. At rectangular pulses, t5−25 value was selective to the MFB response, which was Table 1. Detection Sensitivity In Vivoa MFB S/N

CO2 enhancement

FSCV Amp t5−25 t′0−5

64.5 9.5 26.4 4.8

± ± ± ±

10.3 3.1 8.3 1.7

FSCV Amp t5−25 t′0−5

430.0 92.6 145.1 28.5

± ± ± ±

83.8 17.6 32.1 17.1

7.4 9.2 10.1 7.2

± ± ± ±

S/N

Without Nomifensine 41.3 ± 2.2 ± 3.6 ± 34.4 ± After Nomifensine 1.9 41.7 ± 1.7 1.5 ± 3.1 1.2 ± 2.8 23.7 ±

CO2/MFB

13.1 0.6 1.7 15.5

1.34 0.39 0.21 13.43

± ± ± ±

0.47 0.16 0.11 5.55

6.3 0.2 0.3 9.4

0.18 0.02 0.01 3.94

± ± ± ±

0.07 0.01 0.00 2.59

enhancement

1.0 0.8 1.0 0.9

± ± ± ±

0.1 0.3 0.5 0.1

a

S/N ratios of MFB (dopamine) and CO2 (pH) responses and selectivity to CO2 responses against MFB (CO2/MFB). The repeated result after nomifensine administration is indicated in the lower half. Note that nomifensine enhanced MFB responses but not those of CO2. 8581

dx.doi.org/10.1021/ac500706m | Anal. Chem. 2014, 86, 8576−8584

Analytical Chemistry

Article

mimics a part of physiological condition of interstitial environment in the brain, we should remember that the simple use of ACSF still lacks ascorbate,33,34 peptides, lipids, physiological temperature, osmotic pressure, dynamic calcium regulation by non-neuronal components, and so on. It is unrealistic to mimic all of the physiological conditions in vitro, so the only practical way is to test in the real brain (see Figure 4).

■

DISCUSSION Advanced Discrimination of Capacitive and Faradic Current. We show differential sensitivity to dopamine and pH between the early and late time points in a rectangular voltage pulse. The response to pH appeared just after the onsets of the pulses, while that of dopamine peaked after a delay of a few milliseconds. The temporal pattern of the pH change was the same as the waveform of the capacitive current, suggesting that the current is derived from the changes in electrical capacitance of the electrode surface. The capacitive current is a fundamental distracter in voltammetric recordings with voltage manipulations.2 While the capacitive current influences the recording current throughout the waveform at FSCV, it is concentrated on the initial part of the pulse at RPV. Introduction of adequate low-pass filter (Figure 1C) and precise TH-1 system enabled simultaneous recording of early capacitive current and late faradic current. This not only improved the dopamine sensitivity, but it also enabled differential recording of pH and dopamine. It was previously shown that RPV is suitable for electrochemical detection of dopamine while avoiding interference of pH.9,28 We show that this is simply because the capacitive current at the pulse onset was discarded from analysis. Although the principles of capacitive and faradic current had already been widely known,2,9,19,24 the straight nature of RPV, where it is easy to separate capacitive and faradic current, has been overlooked, and not utilized for voltammetric recordings in the brain. We propose RPV as an alternative detection technique for pH changes, to support the FSCV recordings vulnerable to contamination of capacitive current changes. Importantly, we verified the utility of this approach in vivo through inducing dopamine release and manipulating brain pH (Figure 4). We can additionally recommend the CO2 response shown in this study as a technique to obtain reliable acidic pH training sets for FSCV. Dyanamics of pH and Ca2+. The extracellular space within the brain contains various ionic species needed for neurotransmission and electrochemical equilibrium, including calcium.18−23 We found differential waveforms by calcium and pH shift on FSCV in our own assay, plus largely similar effects between calcium and basic pH shift in RPV. One might predict that pH and other ion changes will occur simultaneously with synaptic activity, so that isolated pH and ionic transients are unlikely to occur under normal conditions. Specifically, Ca2+ entry, and thus loss from the extracellular space near the plasma membrane,21,22 is necessary for dopamine release to occur. This may register as change in capacitive current. However, to date, it is not known to what degree voltammetric methods are able to detect calcium flux in vivo, so this remains an area of future research. Interestingly, we found that the ability of voltammetric recordings to detect Ca2+ decrease was poor in near-physiological background concentrations (Figure 5).

Figure 5. Influence of calcium concentration in HEPES-saline. (A−E) FSCV and (F−J) RPV recordings in the flow of 10 mM HEPES-saline (HS, pH 7.3). Panels (A) and (F) show nonlinear Ca2+ concentration dependency: (A) Current change at the onset of FSCV waveform (0.45 ms from onset) and (F): t′0−5 of RPV. Additional Ca2+ lowered the current at both FSCV (panel (A) and RPV (panel (F)), and the response to additional 0.3 mM Ca2+ was smaller in high Ca2+. Panels (B) and (G) show that the background buffer flow was started with 0.1 mM Ca (Lo) and switched to 1.1 mM Ca2+ (Hi) from 70 s to 160 s, and then back to 0.1 mM. Ca2+ concentration was dropped from 0.1 mM to 0.0 mM (from 10 s to 20 s), and 1.1 mM to 1.0 mM (100 s to 110s). pH was lowered from 7.3 to 7.1 in Lo Ca (from 40 s to 50 s) and Hi Ca (130 s to 140 s). Panels (C)−(E) and (H)−(J) show differential waveforms extracted from the color plot of panels (B) and (G). Note that the peak at 5 ms in FSCV is induced by the influence of HEPES. The inset in panel (C) indicates a 1 mM Ca2+ increase in PBS. Panels (C) and (H) show a 1 mM Ca change from 0.1 mM to 1.1 mM (magenta dotted line: differential from 65 s to 75 s) and 1.1 mM to 0.1 mM (blue solid line: 155 s to 165s). Panels (D) and (I) show a 0.1 mM Ca2+ decrease from 0.1 mM to 0.0 mM (blue solid line) and from 1.1 mM to 1.0 mM (magenta dotted line). Panels (E) and (J) show the acidic pH changes from 7.3 to 7.1 in Lo (light-green solid line) and Hi Ca2+ (dark-green dotted line). 0.1 mM Ca2+ drop was hidden by 1 mM Hi Ca2+ (panels (D) and (I)), while there was no Ca2+ influence on pH responses (panels (E) and (J)). Identical results were repeated on four carbon fibers.

Although in vitro analysis is effective in analyzing the influence of a particular factor, such as calcium, and ACSF 8582

dx.doi.org/10.1021/ac500706m | Anal. Chem. 2014, 86, 8576−8584

Analytical Chemistry

■

Although both H+ and Ca2+ are cations, our results (see Figures S6C and S6D in the Supporting Information) suggest that the interaction of H+ and Ca2+ with the carbon surface is different and independent. An increase of 0.03 μM [H+] (lowered pH from 7.3 to 7.1) sharply increased the capacitive current, while the similar effect was made by decreasing the amount of Ca2+ by 100 μM. In addition, Ca2+ influenced adsorption of dopamine, 5-HT, adenosine, and Ca2+ sensitivity but not on pH shift (see Figure S6 in the Supporting Information). These facts indicate multiple mechanisms of the interaction on carbon fiber surface, which cause capacitive current change. Further studies are awaited to reveal the influences of ionic changes on voltammetric recordings associated with synaptic activities.

CONCLUSIONS Compared to fast-scan cyclic voltammetry (FSCV), our technique using rectangular pulses from 0 to 0.2 V has advantages and disadvantages. First, as discussed above, it can differentiate dopamine from pH clearly without training sets. Second, since the voltage window is relatively low and narrow, rectangular pulses may be made selective for dopamine versus serotonin (5-HT) and adenosine using voltage shifts. By narrowing the potential change to only 0 to 0.2 V, the influence of molecules oxidized at the higher potentials was eliminated from the recording current. On the other hand, FSCV gives stronger electrochemical current to dopamine and is relatively less influenced by ascorbate and 3,4-dihydroxyphenylacetic acid (DOPAC). Since it is easy to switch the waveform from the conventional triangular shape of FSCV to a rectangular pulse, it would be good practice to use both techniques after placing a carbon fiber microelectrode at a particular location in the brain. In conclusion, our technique gives supplemental chemical identification to FSCV. ASSOCIATED CONTENT

S Supporting Information *

Figures S1−S7 and Table S1 are provided as Supporting Information. This material is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

(1) Wightman, R. M.; Amatore, C.; Engstrom, R. C.; Hale, P. D.; Kristensen, E. W.; Kuhr, W. G.; May, L. J. Neuroscience 1988, 25, 513− 523. (2) Kawagoe, K. T.; Zimmerman, J. B.; Wightman, R. M. J. Neurosci. Methods 1993, 48, 225−240. (3) Phillips, P. E.; Stuber, G. D.; Heien, M. L.; Wightman, R. M.; Carelli, R. M. Nature 2003, 422, 614−618. (4) Natori, S.; Yoshimi, K.; Takahashi, T.; Kagohashi, M.; Oyama, G.; Shimo, Y.; Hattori, N.; Kitazawa, S. Neurosci. Res. 2009, 63, 267−272. (5) Badrinarayan, A.; Wescott, S. A.; Vander Weele, C. M.; Saunders, B. T.; Couturier, B. E.; Maren, S.; Aragona, B. J. J. Neurosci. 2012, 32, 15779−15790. (6) Gonon, F.; Buda, M.; Cespuglio, R.; Jouvet, M.; Pujol, J. F. Nature 1980, 286, 902−904. (7) Gerhardt, G. A.; Oke, A. F.; Nagy, G.; Moghaddam, B.; Adams, R. N. Brain Res. 1984, 290, 390−395. (8) Suaud-Chagny, M. F.; Cespuglio, R.; Rivot, J. P.; Buda, M.; Gonon, F. J. Neurosci. Methods 1993, 48 (3), 241−250. (9) Gerhardt, G. A.; Hoffman, A. F. J. Neurosci. Methods. 2001, 109, 13−21. (10) Heien, M. L.; Johnson, M. A.; Wightman, R. M. Anal. Chem. 2004, 76, 5697−5704. (11) Heien, M. L.; Khan, A. S.; Ariansen, J. L.; Cheer, J. F.; Phillips, P. E.; Wassum, K. M.; Wightman, R. M. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 10023−10028. (12) Yoshimi, K.; Naya, Y.; Mitani, N.; Kato, T.; Inoue, M.; Natori, S.; Takahashi, T.; Weitemier, A.; Nishikawa, N.; McHugh, T.; Einaga, Y.; Kitazawa, S. Neurosci. Res. 2011, 71, 49−62. (13) Ariansen, J. L.; Heien, M. L.; Hermans, A.; Phillips, P. E.; Hernadi, I.; Bermudez, M. A.; Schultz, W.; Wightman, R. M. Front. Behav. Neurosci. 2012, 6, 36. (14) Chesler, M. Physiol. Rev. 2003, 83, 1183−1221. (15) Kazemi, H.; Shannon, D. C.; Carvallo-Gil, E. J. Appl. Physiol. 1967, 22, 241−246. (16) Kraig, R. P.; Ferreira-Filho, C. R.; Nicholson, C. J. Neurophysiol. 1983, 49, 831−850. (17) Mutch, W. A.; Hansen, A. J. J. Cereb. Blood Flow Metab. 1984, 4, 17−27. (18) Fedirko, N.; Avshalumov, M.; Rice, M. E.; Chesler, M. J. Neurosci. 2007, 27, 1167−1175. (19) Venton, B. J.; Michael, D. J.; Wightman, R. M. J. Neurochem. 2003, 84, 373−381. (20) Rice, M. E.; Nicholson, C. Anal. Chem. 1989, 61, 1805−1810. (21) Jones, S. R.; Mickelson, G. E.; Collins, L. B.; Kawagoe, K. T.; Wightman, R. M. J. Neurosci. Methods 1994, 52, 1−10. (22) Kume-Kick, J.; Rice, M. E. J. Neurosci. Methods 1998, 84, 55−62. (23) Chen, B. T.; Rice, M. E. Electrochemicals 1999, 11, 344−348. (24) Takmakov, P.; Zachek, M. K.; Keithley, R. B.; Bucher, E. S.; McCarty, G. S.; Wightman, R. M. Anal. Chem. 2010, 82, 9892−9900. (25) Jang, D. P.; Kim, I.; Chang, S. Y.; Min, H. K.; Arora, K.; Marsh, M. P.; Hwang, S. C.; Kimble, C. J.; Bennet, K. E.; Lee, K. H. Analyst 2012, 137, 1428−35. (26) Ziemann, A. E.; Allen, J. E.; Dahdaleh, N. S.; Drebot, I. I.; Coryell, M. W.; Wunsch, A. M.; Lynch, C. M.; Faraci, F. M.; Howard, M. A., III; Welsh, M. J.; Wemmie, J. A. Cell 2009, 139, 1012−1021. (27) Magnotta, V. A.; Heo, H. Y.; Dlouhy, B. J.; Dahdaleh, N. S.; Follmer, R. L.; Thedens, D. R.; Welsh, M. J.; Wemmie, J. A. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 8270−8273. (28) Nakazato, T.; Kagohashi, M.; Yoshimi, K. Biogenic Amines 2006, 20, 23−30. (29) Akiyama, A.; Kato, T.; Ishii, K.; Yasuda, E. Anal. Chem. 1985, 57, 1518−1522. (30) Nakazato, T.; Akiyama, A. J. Neurosci. Methods 1999, 89, 105− 110. (31) Suzuki, A.; Ivandini, T. A.; Yoshimi, K.; Fujishima, A.; Oyama, G.; Nakazato, T.; Hattori, N.; Kitazawa, S.; Einaga, Y. Anal. Chem. 2007, 79, 8608−8615.

■

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We acknowledge Dr. Yusho Karatsu of Huso Electrochemical Systems, for designing a potentiostat with a variable Butterworth low-pass filter. We thank Dr. Yasuaki Einaga of Keio University, for valuable discussions over pH response on diamond microelectrodes; Dr. Akitane Akiyama of Rainbow Science Lab, for technical suggestion of noise reduction by whole-one cycle averaging; undergraduate students Atsuhiko Shindo, Kotaro Nomura, Akina Miyata, Tomoyuki Sugita, Ryoma Oda, Maiika Fujishiro, and Shiori Kumada, and graduate students Yoshitomo Ueda and Takayuki Jo. This study was supported by Juntendo University Research Institute for Diseases of Old Age, a MEXT-Supported Program for the Strategic Research Foundation at Private Universities. 8583

dx.doi.org/10.1021/ac500706m | Anal. Chem. 2014, 86, 8576−8584

Analytical Chemistry

Article

(32) Oyama, G.; Yoshimi, K.; Natori, S.; Chikaoka, Y.; Ren, Y. R.; Funayama, M.; Shimo, Y.; Takahashi, R.; Nakazato, T.; Kitazawa, S.; Hattori, N. Brain Res. 2010, 1352, 214−222. (33) Rice, M. E. Trends Neurosci. 2000, 23, 209−216. (34) Venton, B. J.; Troyer, K. P.; Wightman, R. M. Anal. Chem. 2002, 74, 539−546. (35) Swamy, B. E.; Venton, B. J. Anal. Chem. 2007, 79, 744−750. (36) Cechova, S.; Venton, B. J. J. Neurochem. 2008, 105, 1253−1263.

8584

dx.doi.org/10.1021/ac500706m | Anal. Chem. 2014, 86, 8576−8584