Technical Note pubs.acs.org/ac
Insertable Fast-Response Amperometric NO/CO Dual Microsensor: Study of Neurovascular Coupling During Acutely Induced Seizures of Rat Brain Cortex Yejin Ha,† Jeongeun Sim,‡ Youngmi Lee,*,† and Minah Suh*,‡,§,∥ †
Department of Chemistry and Nano Science, Ewha Womans University, Seoul, 03760, Republic of Korea Center for Neuroscience Imaging Research (CNIR), Institute for Basic Science (IBS), Suwon, 16419, Republic of Korea § Department of Biomedical Engineering and ∥Samsung Advanced Institute of Health Science and Technology (SAIHST), Sungkyunkwan University, Suwon, 16419, Republic of Korea ‡
S Supporting Information *
ABSTRACT: This paper reports the fabrication of an insertable amperometric dual microsensor and its application for the simultaneous and fast sensing of NO and CO during acutely induced seizures of living rat brain cortex. NO and CO are important signaling mediators, controlling cerebrovascular tone. The dual NO/CO sensor is prepared based on a dual microelectrode having Au-deposited Pt microdisk (WE1, 76 μm diameter) and Pt black-deposited Pt disk (WE2, 50 μm diameter). The different deposited metals for WE1 and WE2 allow the selective anodic detection of CO at WE1 (+0.2 V vs Ag/AgCl) and that of NO at WE2 (+0.75 V vs Ag/AgCl) with sufficient sensitivity. Fluorinated xerogel coating on this dual electrode provides exclusive selectivity over common biological interferents, along with fast response time. The miniaturized size (end plane diameter < 300 μm) and tapered needle-like sensor geometry make the sensor become insertable into biological tissues. The sensor is applied to simultaneously monitor dynamic changes of NO and CO levels in a living rat brain under acute seizure condition induced by 4-aminopyridine in cortical tissue near the area of seizure induction. In-tissue measurement shows clearly defined patterns of NO/CO changes, directly correlated with observed LFP signal. Current study verifies the feasibility of a newly developed NO/CO dual sensor for real-time fast monitoring of intimately connected NO and CO dynamics.
N
urements showed advantages such as high sensitivity and simple visualization. Electrochemical30−35 and electrical36,37 measurements were performed for in vivo and in vitro NO detections providing enhanced time-resolution and efficiency. CO, which has been less studied compared to NO, was also measured through diverse techniques in recent years. Gas chromatography analysis with a detector38 or mass spectrometry39,40 and UV−vis spectroscopy of CO-bound myoglobin 38 were performed as indirect methods. For more direct analysis, laser spectroscopy,41 IR spectroscopy,42 colorimetry,43 fluorescence,44,45 chemiluminescence,46 gravimetry,47 electrochemical,2,48 and electrical49,50 measurements were studied. However, only some of them were applied to biological samples.2,39−41,44−46,48 Although most techniques aim at real-time measurements with high sensitivity for biological NO and CO quantification, electrochemical methods have significant advantages over
itric oxide (NO) and carbon monoxide (CO) share various chemical and biological properties and play various physiological roles as signaling molecules.1,2 NO and CO have quite short half-lives and are generated endogenously at nanomolar to micromolar levels from L-arginine via nitric oxide synthase (NOS)3 and from heme via heme oxygenase (HO),4 respectively. Both of these two gases regulate many diverse biological and physiological processes such as vasodilatation,4−8 immune reaction,4,9−12 neurotransmission,7,12−15 platelet aggregation,4,16,17 and antiapoptosis.4,7,18−20 Because of their biological importance and similarities, large research efforts have been devoted to unravel the complicate linkage between NO and CO. However, the exact relations of NO and CO are still controversial: NO and CO regulate each other inversely12,21 versus CO stimulates NO generation.4,22 Precise and quantitative measurements of their physiological levels have been demanded as a prerequisite to study the roles and correlations between NO and CO, and it is a challenging project. Recently, various direct measurements of NO for biological applications were conducted. For example, chemiluminescence,23−25 fluorescence,26−28 and colorimetric29 meas© XXXX American Chemical Society
Received: November 11, 2015 Accepted: February 8, 2016
A
DOI: 10.1021/acs.analchem.5b04288 Anal. Chem. XXXX, XXX, XXX−XXX
Technical Note
Analytical Chemistry
solution (prepared by mixing 14.5 μL of heptadecafluoro1,1,2,2-tetrahydrodecyl trimethoxysilane, 18 μL of methyltrimethoxysilane, 10 μL of 0.5 M HCl, 160 μL of water, and 727.3 μL of ethanol) to obtain high sensitivity and selectivity to only target gas molecules.52 The fluorinated xerogel solution was loaded to the electrode end surface (∼0.08 μL·mm−2) by a microfluid-dispenser (0.001 s, 3.5 psi, Ultimus, Nordson Co., U.S.A.). It was dried for 2 h at 60 °C in an oven, followed by overnight drying at room temperature. The detailed scheme for the sensor preparation steps and sensor diagram is illustrated in Figure 1. All electrochemical experiments were performed with CHI900B bipotentiostat (CH Instrument Inc., TX).
alternative methods especially for in vivo research due to easy miniaturization, multiple sensing capability, minimized analyte loss, and simple preparation.51 Previously, we reported an electrochemical NO/CO dual sensor for the real-time simultaneous in vivo measurement of NO and CO for the first time.48 However, this sensor had the limitations of a rather slow response time due to the use of a thick poly(tetrafluoroethylene) (PTFE) as a gas permeable membrane; and a relatively large whole sensor body dimension (≥2 mm diameter) due to the use of an outer sensor case to hold an internal solution. Besides, one of the dual electrodes of the sensor responded not only to CO, but also to NO, and therefore, a postcalculation process was needed to obtain separate concentrations of NO and CO from the acquired data that could cause a certain inaccuracy of the quantitative analyses. Several attempts have been reported to reduce the response time and to minimize the sizes of electrochemical NO sensors. Shin et al. reported fluorinated xerogel membrane-derived NO microsensor, which exhibited better response time and selectivity over potent interferents.52 Our group also developed a silanized nanopore sensor, having fast response time, high spatial resolution, and high selectivity.53 In current paper, we present the fabrication and in vivo application of an improved NO/CO dual sensor. This sensor differs from the previous sensors on several points: fast response time, miniaturized size (end plane diameter < 300 μm), and tapered needle-like sensor geometry becoming insertable into biological tissues. More importantly, NO and CO responding signals were completely separated at the dual sensor composed of gold (Au) electroplated and platinized dual microelectrodes, improving the selectivity and quantitative accuracy. The newly developed sensor was applied to simultaneously monitor NO and CO levels for 4-aminopyridine (4-AP) induced seizures in a rat brain. This concomitant recording of endogenous NO and CO during spontaneously heightened neuronal activation may allow disentangling the fast and intimately linked NO and CO dynamics in living brain.
Figure 1. (A) Schematic illustration for the preparation steps of an insertable NO/CO dual microsensor. (B) Cross-sectional views of a dual microelectrode during a course of the NO/CO dual sensor preparation.
WE1 and WE2 were polarized at +0.20 V and +0.75 V versus Ag/AgCl, respectively, in phosphate-buffered saline (PBS, pH 7.4, Fisher Scientific, NJ) solution that can oxidize CO at WE1 and NO at WE2 independently. For obtaining calibration curves, the sensor was fixed in a gastight cell holding 2.5 mL of PBS solution deaerated by Ar gas (Dong-A Gas Co., Korea). The sensor current responses of WE1 and WE2 were monitored, while five aliquots of the NO or CO stock solutions were injected to the deaerated PBS while being stirred. NO and CO stock solutions were prepared by bubbling 10 mL of deaerated PBS solutions with NO and CO gases (Dong-A Gas Co., Korea), respectively.2 Animal Preparation and NO/CO Measurements in Rat Brains. SD (Sprague−Dawley) male rats (250−400 g, n = 8) were anesthetized with 2% isoflurane gas (Hana Pharm., Korea). It was maintained during an experiment. Skin over the head was cut, and the skull was removed. All animal treatments were performed following the Guide for Animal Experiments in Sungkyunkwan University. A local field potential (LFP) electrode that was composed of a glass capillary (WPI Inc., FL), a silver wire, and 0.9% saline to measure brain signals was connected to CED Power 1401/Spike 2 system (Cambridge Electronic Design, U.K.). Another glass capillary filled with 15 mM 4-AP (Sigma, MO) was connected to a nanoinjector (Nanoject II Autonanoliter Injector, Drummond Scientific, PA). The prepared insertable NO/CO microsensor was connected to CHI 1040C multipotentiostat (CH Instrument, TX). The NO/CO sensor, LFP electrode, and 4-AP injecting capillary were placed at corresponding vertices of a triangle on the opened brain with an angle of 45° from the brain surface. They were inserted through the brain tissue to about 1.2 mm
■
EXPERIMENTAL SECTION Preparation of Insertable Amperometric NO/CO Sensors. The sensor fabrication process was similar to our previous works, but with some modifications.2,54 First, a thetatype glass capillary (1.5 mm in diameter, WPI Inc., FL) was pulled by a pipet puller (PC-10, Narishige, Japan) to form a needle-sharp sensor body. Two platinum (Pt) wires (76 and 50 μm in diameter, Sigma, MO, and Good Fellow, Oakdale, PA, respectively) were thermally sealed in the pulled theta-type glass capillary, and the end part was polished vertically to expose the 76 and 50 μm Pt microdisks as working electrode 1 (WE1) and working electrode 2 (WE2), respectively. WE1 and WE2 were etched in 1.2 M CaCl2 solution (water/acetone = 2:1 (v/v), Sigma, MO) for 10 and 2 s, respectively, at 3 V ac voltage (60 Hz) to form the recessed micropores. The pores were filled with Au or Pt electrodeposited layers independently as follows. For WE1, the porous Au layer was electroplated in an aqueous solution of 5 wt % gold(III) chloride hydrate (HAuCl4, Sigma, MO) and 0.3 M NaCl (Sigma, MO) by amperometry at −0.1 V versus Ag/AgCl for 30 s. For WE2, Pt black was electrodeposited in 3% platinizing solution (YSI Inc., OH) by amperometry at −0.1 V versus Ag/AgCl for 5 s. The electrode was rinsed with water thoroughly, and then the entire electrode end surface was coated with a fluorinated xerogel B
DOI: 10.1021/acs.analchem.5b04288 Anal. Chem. XXXX, XXX, XXX−XXX
Technical Note
Analytical Chemistry
electrode: Formation of an oxidized Pt film inhibits adsorption of CO to the electrode surface, which is a necessary initial step in the electrocatalytic oxidation of CO on Pt.57 The direct coating of an electrode surface with a fluorinated xerogel solution (i.e., absence of an internal solution) and the experiments carried out at physiological pH 7.4 are presumably the reasons for the selective NO oxidation at WE2 of the current sensor. Figure 3 shows representative dynamic current responses of a dual sensor to NO and CO concentration changes and the corresponding calibration curves. The current of WE1 increased linearly proportional to CO concentration increase, but negligible current change responding to NO concentration change was observed over the range of tested concentrations. On the other hand, WE2 generated an anodic current increasing proportionally to NO concentration increase without noticeable CO oxidation. This confirms the selective anodic detection of CO at WE1 and that of NO at WE2, in a good agreement to LSV result (Figure S1). Therefore, CO or NO concentrations can be measured independently by converting the WE1 or WE2 currents using the corresponding calibration curves, which improves the sensor quantitative accuracy. Au deposited WE1 exhibited a relatively low sensitivity to CO even though larger Pt wire (76 μm diameter) was used as a substrate electrode for WE1 compared to WE2 (50 μm diameter). In fact, CO sensitivity of WE1 (26 ± 14 pA μM−1, n = 6) was about one-seventh of NO sensitivity of WE2 (180 ± 46 pA μM−1, n = 6). However, the absence of noticeable NO oxidation current at WE1 was advantageous as it allowed the selective detection of CO. Fortunately, physiological CO levels are known to be much greater than NO,48 and therefore, the relatively low CO sensitivity of the sensor is seemingly not a critical problem for the biological applications. The dual sensor showed a great linearity between the measured currents and NO or CO concentrations for the tested concentration range (0.18−9.0 μM for CO at WE1 and 0.020−2.0 μM for NO at WE2, R2 > 0.999, n = 5). The detection limits were determined to be ∼180 nM of CO at WE1 and ∼6.0 nM of NO at WE2 (S/ N = 3). As prepared sensor showed effective gas permeability, that is, fast response time, by employing thin-layer coating of fluorinated xerogel. The response times (t90%, time to reach 90% of the steady-state amperometric signal) were estimated as 4.5 ± 1.3 s for WE1 and 3.1 ± 0.2 s for WE2 (n = 5) using typical sensor dynamic current responses similar to Figure 3. These current response data were obtained in a test solution under magnetic stirring to homogenize the solution into which an aliquot of standard NO or CO solution was added successively. This procedure makes a certain time is required to attain an intended homogeneous constant concentration in the solution after the injection the standard solution. Thus, the practical t90% of the sensor in a pre-equilibrated system would be shorter than the estimated one. In fact, the response times of the current sensor were 3.7× and 4.8× faster at WE1 and WE2, respectively, compared to the previously reported one.2 This is attributed to the use of fluorinated xerogel coating for the gas selectivity instead of PTFE gas-permeable membrane, resistive diffusional barrier for NO or CO transport. Fast response time, that is, high time resolution, is an important sensor characteristic for studying fast dynamics such as brain seizure. Sensor selectivity was verified by measuring the sensor currents responding to a variety of biological interfering species. Both WE1 and WE2 currents were not changed upon the
depth, and eventually, their three ends were positioned closely inside of the brain (estimated distance among the three ends
