Antiangiogenic activity of BAI1 in vivo: implications for gene therapy of human glioblastomas.

Glioblastomas are the most common primary brain tumors in adults. These tumors exhibit a high degree of vascularization, and malignant progression from astrocytoma to glioblastoma is often accompanied by increased angiogenesis and the upregulation of vascular endothelial growth factor and its receptors. In this study, we investigated the in vivo antiangiogenic and antitumor effects of brain-specific angiogenesis inhibitor 1 (BAI1) using human glioblastoma cell lines. Glioblastoma cells were transduced with an adenoviral vector encoding BAI1 (AdBAI1), and Northern and Western blot analyses, respectively, demonstrated BAI1 mRNA and protein expression in the transduced tumor cells. Using an in vivo neovascularization assay, we found that angiogenesis surrounding AdBAI1-transduced glioblastoma cells transplanted into transparent skinfold chambers of SCID mice was significantly impaired compared to control treated cells. Additionally, in vivo inoculation with AdBAI1 of established subcutaneous or intracerebral transplanted tumors significantly impaired tumor growth and promoted increased mouse survival. Morphologically, the tumors exhibited signs of impaired angiogenesis, such as extensive necrosis and reduced intratumoral vascular density. Taken together, these data strongly indicate that BAI1 may be an excellent gene therapy candidate for the treatment of brain tumors, especially human glioblastomas.

Carbon film-based interdigitated array microelectrode used in capillary electrophoresis with electrochemical detection

A carbon film based interdigitated ring-shaped array (IDRA) microelectrode was applied to capillary electrophoresis with electrochemical detection to enhance the detection sensitivity on the basis of the redox cycling of electrochemical reversible species at the IDRA microelectrode. We propose a simple capillary-electrode connection device that consists of an X-Y-Z fiber aligner, an electrochemical cell, and a Nafion tubing joint that will enable the detection capillary to be aligned easily on the IDRA microelectrode and isolate the separation voltage from the electrochemical detection system. We used the off-column amperometric detection of aqueous ferrocene and catecholamines by capillary electrophoresis with an IDRA microelectrode to investigate the effects of the capillary-to-electrode distance and the separation voltage on the response currents in single and dual modes and the collection efficiencies (CE) and redox cycles (Rc) at the IDRA microelectrode. The results show that CE and Rc increase when we increase the distance and lower the separation voltage. The limiting currents also increase as the separation voltage decreases in the dual mode. Under optimum conditions, the CE and Rc of catechol, with good reversibility, reach 83.9% and 3.67, respectively. Our results showed that dual-mode detection with the IDRA microelectrode was capable of achieving lower detection limits than single-mode detection.

Subnanoliter volume wall-jet cells combined with interdigitated microarray electrode and enzyme modified planar microelectrode

Miniaturized wall-jet type flow cells with an active volume of 0.042-15 nL were fabricated for use as highly sensitive electrochemical detectors for capillary electrophoresis/electrochemical detection and small on-line enzyme sensors. The cells consisted of three glass plates and a fused-silica capillary. Two of the plates had microfabricated flow channels and guide trenches for the capillary and working, reference, and counter electrodes. The other plate had a film electrode. When an interdigitated microarray electrode (total area, 66 microm x 64 microm; bandwidth and gap, 2 microm) was installed in the flow cell, the redox cycling enhanced the current at flow rates of less than 100 nL/min even though there were only eight pairs of microbands. A sharp dopamine peak enhanced by the redox cycling was observed when the cell was used for capillary electrophoresis. A square film electrode modified with glutamate oxidase and Os-poly(vinylpyridine) containing HRP was also installed in the flow cell and used to measure neurotransmitter release from cultured nerve cells. When the flow rate was relatively high, the response time of the modified electrode was comparable to that of a cylindrical carbon fiber electrode (33 microm o.d.) modified with the same enzyme and mediator. We observed a transient cathodic current response assigned to the glutamate release with the electrode in the flow cell in a suction mode measurement when we stimulated cultured nerve cells electrically with a dual microelectrode.

Real-time electrochemical imaging using an individually addressable multi-channel electrode.

We developed a real-time electrochemical imaging method that uses a multiple enzyme-modified microelectrode. The method will enable the investigation of the functions of biological materials and cells. To test its effectiveness, we imaged the two-dimensional concentration distribution for hydrogen peroxide and L-glutamate in a standard solution. The multiple electrode consists of an 8 x 8 array of 30 x 30 microm2 carbon micro electrode. Each electrode was connected to a 64-channel potentiostat that could apply a potential to all electrodes at the same time. The multiple electrode was coated with an Os-polyvinylpyridine based polymer (Os-gel) containing horse radish peroxidase (HRP) to detect hydrogen peroxide, which is a very common product of oxidase enzyme. When measuring glutamate, which is a well-known neurotransmitter in the mammalian central nerve system, we modified the electrode with a bilayer of Os-gel-HRP and GluOx. The detection limit of our method was 1 microM and images of the glutamate concentration-distribution changes induced by local injection of glutamate through microcapillary were obtained in real time.

Miniaturized thin-layer radial flow cell with interdigitated ring-shaped microarray electrode used as amperometric detector for capillary electrophoresis.

A chip-type thin-layer radial flow cell was developed as an amperometric detector for capillary electrophoresis. We fabricated a carbon film-based interdigitated ring-shaped array (IDRA) microelectrode with a 2 microm bandwidth and an almost 1 microm gap on a glass plate and used it as a working electrode. A fused-silica capillary was arranged above the IDRA electrode using a guide hole drilled through the acryl plate that formed the flow cell lid. A flow channel for use in connecting the outlet capillary was also fabricated in the acryl plate. We characterized the analytical performance of the IDRA electrode in the microchip flow cell in terms of linear concentration range, sensitivity and concentration detection limit. We achieved a collection efficiency and catechol redox cycle at the IDRA microelectrode of 65% and 1.71, respectively, and thus a high sensitivity and low detection limit of 392.9 pA/microM and 15 nM for dopamine hydrochloride. We examined the reproducibility of the detector and found that the run-to-run and detector-to-detector relative standard deviations were both less than 10%.

[Stimulus effect and jumping of delayed potentials--investigation of human single nerve action potentials].

The jumping phenomenon is defined as an unexpected reduction in latency of evoked nerve responses with augmentation of stimulus intensity. This phenomenon may occur in the presence of segmental demyelination. To determine the severity of neuropathies, we recorded median nerve action potentials at the elbow of 60 patients of neuropathy by means of intraneural microneurography. Submaximal stimuli were delivered at the wrist or at one finger. In 53 of the 60 patients evaluated, delayed single nerve action potentials were recorded by submaximal stimuli that produced less than 10 microV compound nerve action potentials. In three delayed single unit potentials, latencies were shortened unexpectedly by several milliseconds when the intensity of stimulus was gradually increased. With the boundary stimulus intensity, we observed a small amplitude of the corresponding potentials at the delayed and the jumped latencies. Instead, jumping was suppressed with rapid repetitive stimuli. At the stimulus intensity sufficient to provoke jumping, we administered repetitive stimuli at 2Hz to 50Hz. At the rate of 20 to 30Hz, the corresponding potentials again returned to the delayed latencies. Rapid repetitive impulses may evoke cathodal hyperpolarization which raises the excitation threshold of the nerve. Thus, the stimulus effect can be decreased and jumping will be released. In addition, a diminution in amplitude and slight prolongation of the latency of each peak occurred in proportion to the increase in stimulus rate. However, paired stimuli with an interval of 4 to 13 msec was equivalent to a supernormal period and provoked the jumping of a single nerve unit. Conditioning stimuli may have increased the excitability of the corresponding fiber.(ABSTRACT TRUNCATED AT 250 WORDS)

[Autotopagnosia ameliorated by looking at the image reflected in a mirror].

A 52-year-old right-handed man presented progressive dystonia and apraxia of his right hand of five years' duration. He also suffered from parkinsonian features such as rigidity or impaired postural reflexes. Serial investigation of brain MRI revealed progressive cerebral atrophy, which started in the left parietal lobe, and subsequently extended to both hemispheres. He was clinically diagnosed as corticobasal degeneration. He could not point at any part of his own body in response to verbal or visual commands. On the other hand, he could point at every part of the examiner's body or of the illustrated body image. Deep sensations and linguistic functions were not involved. This cognitive impairment was regarded as autotopagnosia. In contrast with inability to recognize any part of the own body in response to the commands, he could name every part of his body as soon as the examiner touched there. Moreover, his symptoms of autotopagnosia were ameliorated by looking at himself in a mirror; he could point at any part of his own body. Disconnection between primary proprioceptive sensory area and the center of body schema was thought to be the mechanism of autotopagnosia in this patient, because the impairment improved with the aid of visual or tactile informations. We speculated the lesion was the left parietal lobe.

[Electrophysiological assessment of median sensory nerve function in HMSN type I by intraneural neurography].

Microneurography (MNG) performed in the forearm segment of the median nerve enabled us to assess compound nerve action potentials reflecting the density of large myelinated sensory fibers around the microelectrode. We examined sensory nerve function of the median nerve in seven patients with hereditary motor and sensory neuropathy type I (HMSN-I) with conventional surface-electrode technique and MNG. In six of seven patients sensory nerve action potentials were not elicited with the surface-electrode method. In contrast, compound nerve action potentials (CNAPs) were evoked in all seven patients with MNG. Although the normal waveform of CNAP is characterized by a large triphasic wave with subsequent small multiphasic waves, the triphasic wave was diminished and prolonged small multiphasic waves were prominent in the patients. Maximal nerve conduction velocity and amplitude of the wave were decreased to 69% and 9% of age-matched control values, respectively. These changes detected in patients with HMSN-I could be interpreted as a result of large myelinated fiber loss and segmental demyelination of sensory nerves. We showed that sensory fiber dysfunction in HMSN-I could be quantitatively evaluated with MNG, even though the sensory action potential was not elicited with the surface-electrode method.

[Midmotoraxonal reexcitation observed in a patient with hereditary motor and sensory neuropathy (HMSN) type IA].

We report a unique late motor unit potential observed in a patient with hereditary motor and sensory neuropathy (HMSN) type IA. The characteristics of the potential were as follows; (1) constant appearance with more than 20 mA of the stimuli, without disappearance with supramaximal stimuli, (2) constant wave-form and amplitude, (3) shorter latency with more proximal stimulation (indirect pathway), (4) remarkable fluctuation of the latency, and (5) disappearance with paired stimuli. The characteristics (1)-(3) correspond to Tomasulo's "peripheral late wave". But our late potential had fluctuation in its latency and was cancelled by paired stimuli, which indicates that this potential originated from the reflection at the midmotoraxonal demyelinating portion. Demyelinating portion is electrically unstable, where the duration of action potential is elongated. Consequently, nerve action potential at that site continues even after the refractory period is over at the adjacent distal portion, which might be the reason that this reflecting potential was evoked.

[An analysis of A-wave with slowed conduction in its afferent and efferent pathways below the branching point].

Motor nerve conduction study was carried out in a 59-year-old woman with 20 years' history of diabetes mellitus. When her right ulnar nerve was stimulated at the wrist, the needle electrode penetrated into the first dorsal interosseous muscle revealed an existence of A-wave. Following characteristics were observed: (1) A-wave appeared with more than 20 mA of the stimuli, in a all-or-none manner at the latency of 54 msec; (2) when stimulated more than 50 mA, A-wave jumped at 10 msec of the latency and were superimposed upon the M-wave; (3) A-wave had a stable latency; (4) when stimulated at the more proximal site, A-wave had longer latency (direct pathway) by strong stimuli, and the potential had shorter latency (indirect pathway) by weak stimuli; and (5) A-wave was not collided by paired stimuli. A-wave conduction velocity was slower than 10 m/sec in the distal part of the reflecting point both in afferent and efferent pathways. This report is the first to show strikingly slowed conduction of A-wave distal to the reflecting point compared to the proximal portion. This fact supports the thesis that A-wave is generated from the branching of immature regenerating fibers. M-wave conduction velocity also decreased around the reflecting point of the A-wave, suggesting that, in addition to moderate diabetic polyneuropathy, entrapment at the cubital tunnel may be involved in generating this A-wave.

[Age-related changes in nerve action potentials of median nerve--an analysis using intraneural neurography].

Age-related decrease in peripheral nerve conduction velocity and the amplitude of sensory nerve action potentials is well established in humans. However, evaluation of the sensory nerve action potentials is subject to technical constraints because of its small size and variability of the values. To elucidate age-related differences, we determined the amplitude of median nerve action potentials in 52 healthy individuals aged 16 to 88 years by means of intraneural neurography. A tungsten microelectrode was inserted into the median nerve trunk at the elbow, and compound nerve action potentials evoked by supramaximal electrical stimulation were recorded. The action potentials consisted mainly of triphasic waves. Maximal conduction velocity (Y, m/sec) declined very mildly with age (X, year), showing the regression line of Y = 67.5 - 0.062X. The peak to peak amplitude of the potentials was 466.6 +/- 20.8 (SE) microV and declined gradually with a decrement of 13 microV per decade in the individuals under the age of 50 years. By contrast, in individuals above the age of 50 years, the amplitude (Y, microV) declined steeply, showing the regression line of Y = 977.1 - 9.86X (r = -0.75, p < 0.001) with age (X, year). A significant correlation was observed between amplitude of the sensory nerve action potentials obtained by the conventional technique and that obtained by the intraneural neurography. Nevertheless, an accelerated decrease was not found in the amplitude of the sensory nerve action potentials with age above 50, probably due to the variability of the values.(ABSTRACT TRUNCATED AT 250 WORDS)

[Detection of subclinical sensory nerve dysfunction in amyotrophic lateral sclerosis--a microneurographic study].

In amyotrophic lateral sclerosis (ALS) sensory function has been considered to be normal on a clinical basis. There are, however, several pathologic reports indicating that peripheral sensory fibers are involved in ALS. To evaluate changes in sensory nerve function quantitatively, we measured compound nerve action potentials (CNAP) of the median nerve by means of intraneural microneurography (MNG) and compared the results between 16 patients with ALS, age-matched 13 patients with Parkinson disease (PD) and 13 healthy controls. A tungsten microelectrode with a tip diameter of 1 micron was inserted percutaneously into the median nerve trunk at the elbow. With supramaximal electric stimulation on the median nerve at the wrist, the largest CNAP was recorded. The configuration of the CNAP was similar among three groups, consisting of a large triphasic wave followed by small multiphaic waves. In ALS patients the peak to peak amplitude (Amp) of the triphasic wave was 189.0 +/- 49.8 microV (mean +/- SD), which was significantly smaller than 273.1 +/- 90.0 microV in PD patients and 325.4 +/- 81.0 microV in normal controls (p < 0.01). The maximal nerve conduction velocity (NCV) in ALS was also reduced when compared with that in PD patients and in normal controls (p < 0.05). The average reduction in Amp in ALS was 58% of that in normal controls, while decrease in NCV was less apparent than Amp and 96% of normal controls. The Amplitude of nerve action potentials is considered to represent the density of large myelinated fibers more than 7 microns around the microelectrode.(ABSTRACT TRUNCATED AT 250 WORDS)

[Intraneural topography of the median nerve at the elbow level--an analysis using intraneural neurography].

We examined intraneural topography of the median nerve at the elbow level by means of intraneural neurography in 38 individuals. A tungsten microelectrode was inserted into the median nerve trunk at the elbow where largest amplitude of the nerve action potential could be obtained with stimulations over median nerve trunk at the wrist. Each of six areas innervated by the nerve was stimulated supramaximally. The ratio of the amplitude of the nerve action potentials stimulated in each area to that stimulated over median nerve trunk at the wrist was arbitrary defined as territory index. the territory indices of six areas, i.e., thumb, index finger, middle finger, ring finger, little finger and thenar area, were 12%, 14%, 15%, 7%, 0% and 5%, respectively. Based on the investigation of permutations and combinations, the cutaneous fibers in the median nerve trunk at the elbow may be arranged in order of thumb-index-middle-ring fingers from the radial side, and the muscle fibers to the thenar muscles may be located at the palmar side to the index finger area. These results were in accordance with the previous reports obtained by near nerve recordings or anatomical analysis at the wrist. In 42% of the subjects a part of fibers in the median nerve were considered to innervate also little fingers, even though territory index was very small. This method has made it possible to estimate the intraneural topography of the median nerve at the elbow level and to detect Martin-Gruber anastomosis with a small number of sensory fibers.

[Repetitive discharges of proximal origin in a patient with carcinomatous sensory neuropathy].

Repetitive discharges of a motor unit potential were observed during ulnar nerve conduction studies in a patient with pathologically verified carcinomatous sensory neuropathy. The repetitive discharges in this patient had a latency similar to that of the F-wave latency. The first discharge was followed by discharges identical in configuration every 3 to 4 msec. These discharges responded to all of the suprathreshold stimuli, had a constant configuration, and exhibited jitter in their latency. When paired stimuli were applied, the second stimulus collided with the first discharge, leaving the afterdischarges behind at the previous position. Based on these electromyographic findings, the origin of these repetitive discharges is believed to have been the proximal part of a motor axon or a neighboring demyelinating lesion, since degenerative changes have been demonstrated in ventral roots as well as dorsal root ganglia in this disease.

[Stimulus effect of submaximal trains of impulses on nerves].

Although the occurrence of rate-dependent conduction block upon exposure to supramaximal stimuli is a well-known phenomenon, changes in the stimulus effect of submaximal trains of impulses on nerves remain unknown. To investigate changes in the stimulus effect, we evaluated median nerve action potentials using various frequencies of impulse trains. The subjects were 12 healthy controls and 9 patients with diabetic polyneuropathy. A tungsten microelectrode was inserted into the median nerve trunk at the elbow. Weak stimuli that produced 10 microV compound nerve action potentials were repeated. Each value recorded is the average of 20 trials. Repetition of 1- to 5-Hz stimuli yielded the same average wave, but above 7-Hz, the stimuli produced a diminution in amplitude and slight prolongation of the latency of each peak. This was most prominent with 50-Hz or 100-Hz repeated stimuli. The potentials changed amplitude with a waxing and waning pattern, and gradually stabilized at a lower level. The averaged wave corresponded to the record of reduced stimuli at 0.6 mA, maximally. Paired stimuli at an interval of less than 5 msec were equivalent to a relative refractory period, whereas at an interval of 4-18 msec they were equivalent to a supernormal period, and at 12-90 msec, to a subnormal period. The 'jumping' (unexpected shortening of the latency of a potential in response to increased stimulus intensity) of a single nerve unit was provoked or released corresponding to these periods. No differences were found between healthy individuals and patients with diabetic polyneuropathy. As a consequence of electrogenic Na+/K+ pumping, a three-Na+ ion efflux occurred instead of a two-K+ ion influx. Thus, repeated high-frequency impulses induced membrane hyperpolarization that reduced the stimulus effect on nerves. With trains of impulses at a frequency of 100-Hz, which corresponds to a stimulus every 10 msec, the second response was greater than the first, reflecting the supernormal period, but impulse trains provoked hyperpolarization, as mentioned above, and reduced the amplitude of nerve action potentials. The results of this study show that the stimulus effect on nerves decreased at submaximal stimuli greater than 7-Hz, which reduced the amplitude of compound nerve action potentials. Therefore, averaging should be done at a stimulus frequency below 6-Hz.

[Superiority of intrafascicular neurography over conventional nerve conduction studies in evaluating axonal degeneration].

We investigated conventional motor and sensory nerve conduction studies (MCS & SCS) with regard to the sensitivity in detecting axonopathies. Twelve patients with axonal type of polyneuropathy, 2 vincristin neuropathy and 10 cisplatin neuropathy, were examined by MCS & SCS. Their data were compared with those by intrafascicular microneurography (MNG) of the median nerve. Nerve conduction velocities were within normal limits or slightly reduced to 87-99% of the normal. Amplitude of compound muscle action potential (CMAP) by MCS decreased to 4 or 5 mV in vincristin neuropathy, though cisplatin neuropathy presented normal amplitude. Amplitude of sensory nerve action potential (SNAP) by SCS was undetected in one median nerve and in three sural nerves. While, compound nerve action potential (CNAP) by MNG was all recorded, and presented the amplitude value of below 150 microV in seven patients. The waveform was normal or mild neuropathic pattern. No patients presented normal CNAP amplitude and reduced SNAP amplitude. But there were three patients who had normal SNAP amplitude and reduced CNAP amplitude. In SCS we could recognize abnormal only after CNAP amplitude by MNG dropped to below 100 microV. Cisplatin neuropathy demonstrates reduction of CNAP or SNAP amplitude, and vincristin neuropathy further presents reduction of CMAP amplitude. Evaluation of axonopathy is best achieved by nerve action potential amplitude. Conventional surface electrode methods are available for this purpose, but MNG is more sensitive and is capable of quantitative analysis even in severely damaged nerves.

[Comparison of electrophysiological findings between CIDP and HMSN-1].

Chronic inflammatory demyelinating polyneuropathy (CIDP) and hereditary motor sensory neuropathy type 1 (HMSN-1) are representative myelinopathies. In order to differentiate changes in acquired and congenital demyelinating neuropathies, we studied electrophysiologically 9 patients with active phase of CIDP (36.0 +/- 17.6 years old; mean +/- SD) and 7 patients with genetically-proven HMSN-1 A (56.0 +/- 13.6 years old). Motor conduction studies demonstrated longitudinal uniformity in HMSN-1, contrariwise focal conduction block or conduction delay in CIDP. The mean median nerve conduction velocity in the forearm segment and the mean CMAP amplitude stimulated at the wrist were not different between CIDP and HMSN-1 group; 31.8 +/- 7.2 m/sec and 5.6 +/- 2.8 mV in CIDP, and 26.7 +/- 9.8 m/sec and 3.2 +/- 2.6 mV in HMSN-1, respectively. Upper extremity polyneuropathy index (PNI), a mean percentage of normal for 6 indices concerning to the velocity and latency over two nerves obtained by motor conduction studies, was equal and around 50% on the average in each group. Conduction blocks were presented in 7 patients with CIDP and only one patient with HMSN-1. No sensory nerve action potential was recorded in 6 out of 9 patients with CIDP, and in 6 out of 7 patients with HMSN-1. Intrafascicular neurography of the median nerve, stimulated at the wrist and recorded from intrafascicularly inserted microelectrode at the elbow, revealed irregular multiphasic waves which signify severe temporal dispersion. Maximum conduction velocity was similarly reduced to 48 m/sec in CIDP and 44 m/sec in HMSN-1 on the average, but in one patient with HMSN-1 it was maintained to 63 m/sec with conspicuous temporal dispersion of the waveform. Amplitude of the compound nerve action potential (CNAP) decreased more (p < 0.01) in HMSN-1 (26 +/- 11 micro V) than in CIDP (72 +/- 25 micro V). Temporal dispersion of CNAP was prominent in HMSN-1 than in CIDP. In conclusion, electrophysiological changes were more homogeneous in the longitudinal distribution but more heterogeneous in the cross-sectional distribution in HMSN-1 than in CIDP.

[Electrophysiological evaluation of polyneuropathy in juvenile insulin dependent diabetics].

Thirty patients with juvenile insulin dependent diabetes mellitus (IDDM) were electrophysiologically evaluated. In addition to the conventional motor and sensory nerve conduction studies, intrafascicular microneurography was performed in the median nerve. In this method a tungsten microelectrode was inserted into the median nerve trunk at the elbow, and a compound nerve action potential (CNAP) was recorded with supramaximal electrical stimulation at the wrist. The subjects' age ranged from 8 to 31 years with an average (SD) of 15.4 (6.2) years; the disease duration varied from 1 to 23 years with an average (SD) of 8.3 (5.8) years. Polyneuropathy index (PNI), expressed as a mean percentage of the normal for twelve indices over the four nerves obtained by motor conduction studies, was 93.9% on the average in patients with IDDM. The mean amplitude of CNAP obtained by intrafascicular microneurography was 417 microV. These results indicate that neuropathy in IDDM is milder than that in adult non-insulin dependent diabetes mellitus (NIDDM). The mean value of PNI decreased at a rate of 0.56% per year; the mean glycosylated hemoglobin (A1c) level was as high as 8.2 +/- 0.9%, findings consistent with those of the previous analysis of adult patients with NIDDM. The PNI value had a significant negative correlation with the duration of diabetes mellitus (p < 0.001) and with mean glycosylated hemoglobin (A1c) level (p < 0.01). CNAP amplitude had a tendency to correlate with duration of diabetes mellitus (p < 0.1). In patients with IDDM we can tell exactly when the disease occurred. Progression of neuropathy in juvenile IDDM was identical to that of adult NIDDM. Careful management of diabetes mellitus is of importance to prevent the progression of neuropathy.

Small-volume on-line sensor for continuous measurement of gamma-aminobutyric acid.

We report the first on-line electrochemical sensor for the continuous measurement of gamma-aminobutyric acid (GABA), which is a well-known inhibitory neurotransmitter in the nervous system. The sensor is composed of a glutamate oxidase (GluOx) and catalase immobilized small-volume enzymatic reactor and a glassy carbon (GC) electrode modified with a top layer film consisting of gabase and GluOx coimmobilized bovine serum albumin and an Ospoly(vinylpyrridine) bottom layer film containing horseradish peroxidase. The response of the sensor depends on the alpha-ketoglutarate concentration and is almost saturated when its concentration is 100 times higher than GABA. The sensor exhibits a sensitivity of 1.56 nA/microM for GABA under optimized conditions and shows almost no response when 10 microM glutamate is continuously injected. A detection limit of 0.1 microM is obtained with a linear range of 0.1-10 microM. GABA can be measured in the absence of alpha-ketoglutarate when there is L-glutamate in the sample solution, which is a typical condition for the extracellular measurement of cultured nerve cells.

On-line electrochemical sensor for selective continuous measurement of acetylcholine in cultured brain tissue.

An on-line acetylcholine (ACh) sensor was developed in order to determine extracellular ACh concentrations without interference from choline (Ch). The sensor is composed of a small-volume enzymatic prereactor (22-microL inner volume) in which choline oxidase and catalase are immobilized in series. Carbon electrodes were modified with an acetylcholine esterase (AChE), choline oxidase (ChOx), and osmium poly(vinylpyridine)-based redox polymer containing horseradish peroxidase (Os-gel-HRP). The sensor sensitivity was 43.7 nA/microM (+/- 0.15, n = 3) for ACh under optimized conditions. Almost no response was seen when 100 microM Ch was continuously injected. The detection limit for ACh with the sensor was comparable to that obtained using liquid chromatography with electrochemical detection combined with an enzymatic reactor. The electrical stimulation of cultured rat hippocampal tissue resulted in an extracellular ACh increase of 20 nM (+/- 11 nM, n = 3). This increase was observed continuously with our online sensor combined with a microcapillary sampling probe located very close to the tissue. The continuous measurement of ACh and Ch using a split disk carbon film dual electrode in which one electrode surface was modified with ChOx/Os-gel-HRP and the other with AChE-ChOx/Os-gel-HRP bilayer film was also demonstrated to improve the response time by eliminating the prereactor.