Anal. Chem. 2010, 82, 7618–7624
Real-Time in Vivo Simultaneous Measurements of Nitric Oxide and Oxygen Using an Amperometric Dual Microsensor Sarah S. Park,† Minyoung Hong,‡ Cha-Kyong Song,‡ Gil-Ja Jhon,† Youngmi Lee,*,† and Minah Suh*,‡ Department of Chemistry and Nano Science, Ewha Womans University, Seoul, 120-750, Korea, and Department of Biological Sciences, Sungkyunkwan University, Suwon, 440-746, Korea This paper reports a real-time study of the codynamical changes in the release of endogenous nitric oxide (NO) and oxygen (O2) consumption in a rat neocortex in vivo upon electrical stimulation using an amperometric NO/O2 dual microsensor. Electrical stimulation induced transient cerebral hypoxia due to the increased metabolic demands that were not met by the blood volume inside the stimulated cortical region. A NO/ O2 dual microsensor was successfully used to monitor the pair of real-time dynamic changes in the tissue NO and O2 contents. At the onset of electrical stimulation, there was an immediate decrease in the cortical tissue O2 followed by a subsequent increase in the cortical tissue NO content. The averages of the maximum normalized concentration changes induced by the stimulation were a 0.41 ((0.04)-fold decrease in the O2 and a 3.6 ((0.9)-fold increase in the NO concentrations when compared with the corresponding normalized basal levels. The peak increase in NO was always preceded by the peak decrease in O2 in all animals (n ) 11). The delay between the maximum decrease in O2 and the maximum increase in NO varied from 3.1 to 54.8 s. This rather wide variation in the temporal associations was presumably attributed to the sparse distribution of NOS-containing neurons and the individual animal’s differences in brain vasculatures, which suggests that a sensor with fine spatial resolution is needed to measure the location-specific real-time NO and O2 contents. In summary, the developed NO/O2 dual microsensor is effective for measuring the NO and O2 contents in vivo. This study provides direct support for the dynamic role of NO in regulating the cerebral hemodynamics, particularly related to the tissue oxygenation. A gaseous neurotransmitter, nitric oxide (NO), is an important signaling molecule involved in a range of biological/physiological processes, including brain hemodynamics.1-4 Nitric oxide acts * To whom correspondence should be addressed. Fax: + 82-2-3277-2384. E-mail:
[email protected] (Y.L.);
[email protected] (M.S.). † Ewha Womans University. ‡ Sungkyunkwan University. (1) Zhang, F.; Iadecola, C. J. Cereb. Blood Flow Metab. 1994, 14, 574–580.
7618
Analytical Chemistry, Vol. 82, No. 18, September 15, 2010
as a vasodilator modulating the rate and volume of blood flow by controlling the mechanical vasodilation of blood vessels5,6 and increases the bioavailability of many reactive oxygen species (ROS) that are also involved in regulating the cerebral vascular tone.7-9 Brain NO is synthesized endogenously by the NO synthase (NOS) enzyme, of which three isoforms have been identified: neuronal (nNOS), endothelial (eNOS), and inducible NOS (iNOS).10 It was reported that hypoxic stress leads to the overproduction of NO in the cerebral cortex of newborn rats, which is produced from both nNOS and iNOS, probably in an effort to reestablish the normal blood flow.11 Indeed, the mode of NO production is quite complex and finely orchestrated to maintain the optimal cerebral hemodynamics and neuronal cell viability. As the regulatory functions of NO are likely to be concentration-dependent, it is essential to quantify the NO content in vivo to better understand the roles of NO in regulating the cerebral hemodynamics. However, real-time quantitative analysis of the NO content in the brain of an intact animal is challenging due to the rapid diffusion, short half-life, and low basal concentrations of NO.12 Furthermore, NO is generated locally by the activation of NOS within a small confined area and the active NO concentrations are relatively low (a few to hundreds nanomolar), making accurate measurements even more difficult.13,14 Recently, many electrochemical NO single sensors have been developed and used for NO measurements owing to their attractive capabilities, such (2) Buerk, D. G.; Ances, B. M.; Greenberg, J. H.; Detre, J. A. Neuroimage 2003, 18, 1–9. (3) Iadecola, C. Nat. Rev. Neurosci. 2004, 5, 347–360. (4) Toda, N.; Ayajiki, K.; Okamura, T. Pharmacol. Rev. 2009, 61, 62–97. (5) Faraci, F. M. Am. J. Physiol. Heart Circ. Physiol. 1990, 259, H1216–H1221. (6) Baumbach, G. L.; Sigmund, C. D.; Faraci, F. M. Circ. Res. 2004, 95, 822– 829. (7) Preckel, M. P.; Leftheriotis, G.; Ferber, C.; Degoute, C. S.; Banssillon, V.; Saumet, J. L. Int. J. Microcirc. Clin. Exp. 1996, 16, 277–283. (8) Yamamoto, Y.; Henrich, M.; Snipes, R. L.; Wolfgang, K. Brain Res. 2003, 961, 1–9. (9) Yamamoto, Y.; Ko ¨nig, P.; Henrich, M.; Dedio, J.; Wolfgang, K. Cell Tissue Res. 2006, 325, 3–11. (10) Methods in Nitric Oxide Research; Feelish, M., Stamler, J., Eds.; John Wiley: Chichester U.K., 1996. (11) Fernandez, A. P.; Alonso, D.; Lisazoain, I.; Serrano, J.; Leza, J. C.; Bentura, M. L.; Lopez, J. C.; Encinas, J. M.; Fernandez-Vizarra, P.; Castro-Blanco, S. Brain Res. Dev. Brain Res. 2003, 142, 177–192. (12) Wang, S.; Paton, J. F. R.; Kasparov, S. Auton. Neurosci. 2006, 126-127, 59–67. (13) Bellamy, T. C.; Garthwaite, J. J. Biol. Chem. 2001, 276, 4287–4292. (14) Wykes, V.; Garthwaite, J. Br. J. Pharmacol. 2004, 141, 1087–1090. 10.1021/ac1013496 2010 American Chemical Society Published on Web 08/17/2010
as real-time, direct analysis with high sensitivity, and a strong potential for clinical application.15,16 Some applications of these sensors for NO analysis in the brain have been reported.12,18-20 The application of either strong sensory stimulation or focal direct electrical stimulation causes a noticeable increase in the neuronal activity, which leads to significant increases in the cerebral metabolic rate of oxygen (CMRO2), ultimately leading to a series of hemodynamic events, such as vasodilation, increase in blood flow. The O2 concentration in the affected brain area decreases when the increased CMRO2 is not met by the corresponding levels of vasodilation, resulting in sustained cerebral hypoxia.3 During cerebral hypoxia, the balance between the NO and O2 levels is disrupted. Regardless of the aforementioned close connection between NO and O2, there are no reports of the real-time simultaneous in vivo analyses for the linked NO and O2 contents in the brain. We recently reported the development of a planar-type dual electrochemical microsensor for the simultaneous measurements of NO and O2.21 This paper reports the in vivo applications of the NO/O2 dual electrochemical microsensor. The real-time codynamic relationships between the NO and O2 concentrations in a rat brain in vivo during cerebral hypoxia were examined using the NO/O2 dual microsensor. The planar configuration and small dimensions of the sensor allowed quantification of the NO and O2 levels at the surface of the intact brain tissue of interest with high vertical and lateral spatial resolution. Cerebral hypoxia was induced experimentally by applying sustained direct cortical electrical stimulation over a 25 s period to the somatosensory cortex of adult male rats. This stimulation induced dramatic changes in brain hemodynamic responses, which behaved similarly with those during epileptic seizures,22 and therefore make possible a clearer verification of our developed sensor’s performance. To best of our knowledge, this is the first report of simultaneous direct in vivo measurements of the NO and O2 contents in the confined area of the brain using electrochemical methods. EXPERIMENTAL SECTION Materials. Sodium nitrite, uric acid, dopamine hydrochloride, and acetaminophen were supplied by Sigma-Aldrich (St. Louis, MO). Phosphate buffered saline (PBS, pH 7.4 at 25 °C) and L-ascorbic acid were purchased from Fisher Scientific (Pittsburgh, PA). Platinizing solution (chloroplatinic acid in water) was obtained from YSI Inc. (Yellow Springs, OH). Lidocaine and isoflurane were purchased from Dai Han Pharm. (Seoul, Korea) and Hana Pharm. (Seoul, Korea), respectively. Pt wire (25 µm diameter, 99.99%) and (15) Bedioui, F.; Villeneuve, N. Electroanalysis 2003, 15, 5–18. (16) Ciszewski, A.; Milczarek, G. Talanta 2003, 61, 11–26. (17) Barbosa, R. M.; Lourenc¸o, C. F.; Santos, R. M.; Pomerleau, F.; Huettl, P.; Gerhardt, G.A.; Laranjinda, J. Methods in Enzymology; Cadenas, E., Packer, L., Eds.; Elsevier: San Diego, CA, 2008; pp 351-367. (18) Ondracek, J. M.; Dec, A.; Hoque, K. E.; Lim, S. A. O.; Rasouli, G.; Indorkar, R. P.; Linardakis, J.; Klika, B.; Mukherji, S. J.; Burnazi, M.; Threlfell, S.; Sammut, S.; West, A. R. Eur. J. Neurosci. 2008, 27, 1739–1754. (19) Santos, R. M.; Lourenc¸o, C. F.; Piedade, A. P.; Andrews, R.; Pomerleau, F.; Huettl, P.; Gerhardt, G.A.; Laranjinda, J.; Barbosa, R. M. Biosens. Bioelectron. 2008, 24, 704–709. (20) Brown, F. O.; Finnerty, N. J.; Lowry, J. P. Analyst 2009, 134, 2012–2020. (21) Park, S. S.; Tatum, C. E.; Lee, Y. Electrochem. Commun. 2009, 11, 2040– 2043. (22) Zhao, M.; Suh, M.; Ma, M.; Perry, C.; Geneslaw, A.; Schwartz, T. H. Epilepsia 2007, 48, 2059–2067.
Teflon insulated silver wire (127 µm diameter, 99.99%) were acquired from Good Fellow (Oakdale, PA) and A-M Systems (Sequim, WA), respectively. Theta-type glass capillaries (1.5 mm outer diameter) were products of WPI Inc. (Sarasota, FL). Expanded PTFE membranes (thickness < 19 µm, pore size 0.05 µm) were obtained from W. L. Gore & Associates Inc. (Elkton, MD). Nitric oxide, argon, oxygen, and carbon monoxide gases were obtained from Dong-A Gas Co. (Seoul, Korea). All solvents and chemicals were of analytical-reagent grade and used as received. All aqueous solutions were prepared with deionized water (18 MΩ cm-1). Animal Preparation. Sprague-Dawley male rats (n ) 11, 350-400 g) were used. All procedures involving the animals, their care, and surgery were performed according to the Guide for the Care and Use of Laboratory Animals (National Academy Press, 1996). The animals were placed in a stereotactic frame, and lidocaine (0.5-1.0 mL) was injected around the surgical area for local anesthesia. The animals were anesthetized with isoflurane (2.5-3.0%). The body temperature of the animals was maintained using warm pads, and the level of anesthesia and breathing patterns were closely monitored throughout the experiments. The average breathing rate was ∼38 times/min. The skin was cut and the muscles over the top of the head were removed. The skull over the somatosensory cortex (-1.5 mm posterior to bregma, 2.5 mm lateral to midline) of one hemisphere was then removed. To ensure the health of the cortex, the local field potential (LFP) was recorded with a glass electrode before and during the experiments (Figure 1). Electrical Stimulation and Electrophysiology. A bipolar stimulation electrode (Plastics One) connected to a master 8/SIU-7 was placed in the burr hole located in the somatosensory cortex of one hemisphere. The electrode just touched the surface during the experiment session. Monophasic trains of electrical pulses (pulse-width 200 µs, amplitude 1 mA) were delivered at a frequency of 0.45 Hz for 25 s. A previous study, which used similar but biphasic stimulation parameters, has reported significant changes in the mouse brain hemodynamic signal.23 We expect stronger responses than this previous one because we utilized monophasic stimulations, which deliver stronger impacts on brain tissue than biphasic stimulation. Our stimulation is expected to produce large changes in the hemodynamic signal as similar during epileptic seizures. The signals from the LFP electrode were amplified and band-pass filtered between 0.3 and 300 Hz using an A-M system (WPI Inc.). The signal was then digitized at 1000 Hz using a CED Power 1401 (Cambridge Electronic Design) and recorded on a PC using the Spike 2 program. Each trial was examined for any after-discharge. The trial was discarded if the stimulation evoked an after-discharge. Electrochemical NO/O2 Dual Microsensors. A planar amperometric NO/O2 dual microsensor was fabricated using a previously reported method.21 Briefly, a glass-sealed dual platinized Pt disk working electrode (WE, each Pt diameter ) 25 µm; distance between the two disks ∼ 300 µm) and a coiled Ag/AgCl wire (127 µm diameter) reference electrode (RE) were immersed in an internal solution (30 mM NaCl + 0.22 µM NaOH in distilled water) and covered with an expanded PTFE gas-permeable membrane. (23) Song, C.; Suh, M. J. Korean Phys. Soc. 2008, 54, 1709–1715.
Analytical Chemistry, Vol. 82, No. 18, September 15, 2010
7619
Figure 1. Summary of the experimental setup. The illustration shows the stimulation electrode (a star symbol on the right panel), recording the local field potential (LFP) electrode (a circle symbol on the right panel), and the NO/O2 microsensor (a dotted circle on the right panel). The black box on the brain shown on the left, which is centered between lambda and bregma, indicates the exposed brain area for the experiment.
The NO/O2 dual microsensors were calibrated before and after the measurements using standard NO and O2 solutions prepared by purging deoxygenated phosphate-buffered saline (PBS, pH 7.4) with either NO or O2. The sensors were positioned over the brain tissue using a micromanipulator (WPI Inc. Sarasota, FL). Potentials of +0.75 and -0.4 V (vs Ag/AgCl RE) were applied independently to each platinized Pt (WE1 and WE2) of the dual electrode to measure the NO and O2 concentrations, respectively. The currents at the dual electrode were recorded simultaneously using a CHI1000A multipotentiostat (CH Instruments Inc. Austin, TX) with a sampling rate of 200 ms. Data Analysis and Statistical Test. In each animal, the NO and O2 data were normalized, and the temporal profiles of NO and O2 were compared (See the Results and Discussion for the details). The time points of the maximum changes (tO2(or NO),peak, measured as time passed after the electrical stimulation starts), i.e., the peak changes, were defined as the average of the data ranging from 98% to 100% of the maximum value; and the maximum NO and O2 concentrations were obtained as the average of the data ranging tO2(or NO),peak ± 2 s. The average and standard deviation of the peak changes observed in three consecutive trials were calculated for each animal. A paired t test with two-tails was carried out for the statistical evaluation. A p-value < 0.05 was considered significant. RESULTS AND DISCUSSION NO/O2 Dual Microsensor. The performance of the amperometric NO/O2 dual microsensors was good and reliable. Figure 2 shows the dynamic response curves and corresponding calibration curves. The measured currents showed a strong positive correlation with the successive addition of the NO and O2 standard solutions and were linearly proportional to the NO and O2 concentrations. Indeed, the current at WE1 (+0.75 V vs Ag/AgCl) was induced only by the electrochemical oxidation of NO and was directly proportional to the NO concentration. 7620
Analytical Chemistry, Vol. 82, No. 18, September 15, 2010
On the other hand, the current at WE2 (-0.4 V vs Ag/AgCl) was induced by the electrochemical reduction of O2 and NO. For the particular sensor used in Figure 2, the sensitivities for NO at WE1 and at WE2 were 128.8 (r2 ) 0.9967) and 45.2 pA/ µM (r2 ) 0.9841), respectively. On the other hand, the sensor sensitivity for O2, 172.3 pA /µM (r2 ) 0.9999), was observed only at WE2; WE1 did not respond to O2. The currents at the dual sensor can be represented by the following equations:21 IWE1(+0.75 V) ) (SWE1,NOCNO)
(1)
IWE2(-0.4 V) ) (SWE2,NOCNO) + (SWE2,O2CO2)
(2)
where I is the current (amps), S is the sensitivity (amps/molar), C is the concentration (molar), and the subscripts represent the types of the corresponding electrode or gas. The simultaneously measured currents at WE1 and WE2 were converted to the corresponding NO and O2 concentrations by eqs 1 and 2 using the corresponding calibration data obtained instantaneously before the experiments. At a certain time point, the NO concentration could be determined directly from the current measured at WE1 according to eq 1. The O2 concentration could be determined from the current measured at WE2 using the NO concentration (determined from eq 1) according to eq 2. In terms of the sensor selectivity, the dual sensor showed no noticeable responses to the addition of potential interfering agents (e.g., 100 µM nitrite, 100 µM uric acid, 20 µM dopamine, 100 µM ascorbic acid, 100 µM acetaminophen), confirming the selectivity (Figure 3). In addition, the sensor WE1 showed ∼20 times greater sensitivity to NO than to analogous CO, another biological interfering agent (data not shown).21,24 In Vivo Simultaneous Measurements of NO and O2 in Real Time. In a biological system, a range of signals frequently represent the inter-related biological events that usually have very
Figure 2. Typical calibration curves for NO and O2 of a dual microsensor at (a) WE1 and (b) WE2. The insets are the corresponding dynamic response curves to the successive increase in the NO and O2 concentrations. +0.75 and -0.4 V (vs Ag/AgCl) were applied to WE1 and WE2, respectively. The calibrations for NO and O2 were carried out separately.
Figure 3. Sensor responses (left column, WE1; right column, WE2) to typical interfering agents: (a) 100 µM nitrite, (b) 100 µM uric acid, (c) 20 µM dopamine, (d) 100 µM ascorbic acid, and (e) 100 µM acetaminophen. Each arrow is marked with an injection of the interfering agents. (Note that the traces are offset for clarity.)
fast temporal latency. Real-time sensing devices that can analyze two different but closely related biological analytes are particularly
advantageous in neurobiological research because they provide spatiotemporally improved information. However, the development Analytical Chemistry, Vol. 82, No. 18, September 15, 2010
7621
Figure 4. Typical O2 and NO concentration changes monitored with the NO/O2 microsensor. Three sequential electrical stimulations (each for 25 s with 1 mA amplitude and 0.45 Hz) were applied at the time region marked with small gray bars under the curves.
of dual sensing devices is difficult because the discrimination of two different analytes is not straightforward. The most common problem is interference between the signals detecting each analyte. Indeed, there are many reports of the development and application of electrochemical NO or O2 sensors, while there are very few reports of simultaneous NO/O2 measurements. Some trials employed separate commercially available NO and O2 single sensors25,26 or a NO/O2 dual sensing microelectrode array system.27 However, the relative positions between the NO and O2 sensing devices and the cross talk between the NO and O2 detection signals as well as sensor selectivity were never clearly stated. More importantly, there are no reports of simultaneous in vivo NO/O2 analysis in the brain in real time. As our developed NO/O2 dual microsensor exhibited negligible NO/O2 detection-signal cross talk with the use of an optimized internal solution composition,21 the NO and O2 levels are simultaneously and quantitatively analyzed at the surface of a rat neocortex using the NO/O2 dual microsensor. To investigate the dynamic relationships between NO and O2, the measurements were carried out with intermittent electrical stimulation on the brain tissue surface. Figure 4 shows the typical NO and O2 concentration traces measured with the dual sensor upon three consecutive electrical stimulations for a single rat. The NO and O2 concentrations were calculated from the currents measured at WE1 and WE2 of the dual sensor using the prior calibration curves from eqs 1 and 2. For statistical analyses and comparison, the NO and O2 concentrations determined at each time point were normalized as follows: CNO(or O2),norm ) CNO(or O2),measured/CNO(or O2),base
(3)
where CNO(or O2),norm is the normalized NO or O2 concentration, CNO(or O2),measured is the measured NO or O2 concentration, and (24) Lee, Y.; Kim, J. Anal. Chem. 2007, 79, 7669–7675. (25) Yomura, Y.; Shoji, Y.; Asai, D.; Murakami, E.; Ueno, S.; Nakashima, H. Life Sci. 2007, 80, 1449–1457. (26) Watanabe, T.; Owada, S.; Kobayashi, H.; Ishiuchi, A.; Nakano, H.; Asakuta, T.; Shimamura, T.; Asano, T.; Koizumi, S.; Jinnouchi, Y.; Katayam, M.; Kamibayasi, M.; Murakami, E.; Otsubo, T. Transplant. Proc. 2009, 39, 3007–3009. (27) Patel, B. A.; Arundell, M.; Quek, R. G.; Harvey, S. L.; Ellis, I. R.; Florence, M. M.; Cass, A. E.; Schor, A. M.; O’Hare, D. Anal. Bioanal. Chem. 2008, 390, 1379–1387.
7622
Analytical Chemistry, Vol. 82, No. 18, September 15, 2010
Figure 5. Representative patterns of the NO and O2 concentration change profiles obtained for three different rats as a function of time upon an electrical stimulation.
CNO(or O2),base is the basal NO or O2 concentration, which was obtained by averaging the measured corresponding concentrations for 50 s immediately before electrical stimulation. A focal direct cortical electrical stimulation elicited a transient increase in the tissue O2 demand resulting in an abrupt decrease in tissue oxygenation, followed by a subsequent increase in tissue NO release. Then, the released NO induced vasodilation in order to compensate for the increased O2 consumption of the brain tissue. Therefore the O2 concentration was consequently restored to the initial one after reaching tO2,peak, and then NO also returned to its initial level. As seen in Figure 4, for the same single animal, NO and O2 concentration changes responding to three sequential electrical stimulations were fairly similar to one another in terms of the degree of NO increase and O2 decrease as well as the times at maximum NO or O2 level changes. The average basal NO and O2 concentrations measured at the surface of the brain tissue of 11 rats were 0.5 ± 0.4 µM (mean ± SD) and 230 ± 20 µM (mean ± SD), respectively. After the in vivo measurements for ∼3 h, the NO and O2 sensitivities observed at both WE1 and WE2 of the dual sensor were typically maintained within ∼± 10% of the initial sensitivities obtained immediately before the measurements. Dynamic Relationship between NO and O2 Concentrations upon Electrical Stimulation-Induced Cerebral Hypoxia. Figure 5 illustrates the representative trends distinctive in the shape of the NO and O2 concentration change profiles as a function
Figure 6. (a) The observed time points of the peak O2 concentration changes (tO2,peak, white bars) and the corresponding peak NO concentration changes (tNO,peak, black bars) for all the animals examined (n ) 11). (b) The maximum normalized O2 concentration changes (∆CO2,norm, 0) and corresponding normalized NO concentration changes (∆CNO,norm, 9). The data points were presented intentionally in the order of NO concentration changes from the greatest to the smallest in order to show the close relationship between the changed amounts of NO and O2 concentrations clearly.
of time upon electrical stimulation measured for 3 different rats out of 11. All of the observations showed rather sharp changes in the NO and O2 concentrations before tO2(or NO),peak than after tO2(or NO),peak. The observed individual transient patterns were, however, different depending on the animal. The changing rates of the NO concentration are strongly affected by the ones of the O2 consumption. In fact, the more rapid decrease in the O2 concentration tended to be correlated with the more rapid increase in the NO concentration before tO2(or NO),peak; and the more rapid recovery of the O2 concentration corresponded to the more rapid return of NO concentration to the initial levels after tO2(or NO),peak, respectively. Furthermore, the rates of increase in the NO concentration decreased only after the O2 consumption reached its peak in all animals, even though there was a large variation in the delays between the tO2,peak and tNO,peak (Figures 5 and 6a). In the dual sensor, NO and O2 gases were detected amperometrically at the surfaces of WE1 and WE2 after passing through the PTFE gas permeable membrane. Indeed, the response times of the dual sensor for NO and O2 were slightly different (
