| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Marine Biological Laboratory, Woods Hole, MA
2 University of Illinois at Chicago, Chicago, IL
Glutamate is the primary excitatory neurotransmitter in the CNS of vertebrates, activating both ionotropic and metabotropic receptors (1). Glutamate, along with other amino acids, has also been identified as an osmoticum used to regulate the water potential across the plasma membrane of animal cells (2). The molecular mechanisms regulating the release and uptake of glutamate are consequently of considerable physiological interest. The release of glutamate from single cells can be detected with outside-out patches of glutamate receptors pulled from nerve cells (3), but these sensors lack consistency, are technically arduous to construct, and lose sensitivity over time. Fluorescent detection methods have also been used, but are also difficult to make and quantitate (4). Our specific aim is to avoid the difficulties inherent in these other techniques, by developing a noninvasive, real-time micro-biosensor that will quantitatively measure glutamate release and uptake by single cells.
High-sensitivity, real-time detection of glutamate can be achieved using electrochemical detection of H2O2, a byproduct of the enzymatic conversion of L-glutamate to 
-ketoglutarate catalyzed by glutamate oxidase (5). Glutamate oxidase itself has approximately 61- and 325-fold selectivity for glutamate over aspartate and glutamine when immobilized on macroelectrodes (5). H2O2 can be directly detected with a platinum wire electrode polarized to +0.6 V (6). However, this method can be problematic due to (a) electrical drift of the electrode response, which reduces the useful sensitivity and reliability of the measurement, and (b) oxidation of other compounds released by cells, most notably ascorbate. Electrical drift can be eliminated if the electrode is used in a self-referencing format (7). In this mode, measurements are made alternately, near the source of glutamate influx or efflux, and then at a second location a set distance away; the difference in the readings at the two locations is the measurement of the flux. Electrical drift is subtracted out by this process, provided that the signal generated by the electrode occurs more quickly than the rate of electrical drift; if it is too slow, the signal itself will also be subtracted. The oxidation of interfering compounds can be eliminated by using enzyme-coupled detection of H2O2 at a lower electrical potential. This is accomplished by coating the electrode with a redox polymer containing horseradish peroxidase (HRP) and osmium. HRP catalyzes the reduction of H2O2 to water and is itself reduced by osmium(II), which is converted to osmium(III). The electrode then donates electrons to osmium(III), regenerating osmium(II) and, in the process, producing a measurable electrical current (8,9). Our goal in the present work is to determine if such a hydrogel-based glutamate-selective electrode could be miniaturized for use in a self-referencing mode.
Microelectrodes were fabricated in a manner similar to oxygen sensors (10), except that 8-µm carbon fiber, 12-µm gold wire, and 10- and 25-µm platinum wire were used at the reactive surface. Electrodes were dip-coated with Os-gel-HRP redox polymer (BAS, W. Lafayette, IN) and allowed to dry for 10 min. Electrodes were then dip-coated in 50 units/ml glutamate oxidase (Sigma) and allowed to dry for 10–30 min. Glutamate electrodes were polarized to either 0 or +100 mV against a Ag/AgCl reference in physiological saline. The average response of platinum-based electrodes was 0.45 pA ± 0.15 pA/µM of glutamate, whereas gold-based electrodes produced a weaker signal of 0.21 ± 0 pA/µM, and carbon gave the weakest signal, 0.1 ± 0.01 pA/µM. Figure 1A shows the electrical current detected from a platinum gel electrode; it was placed 20 µm from a source pipette containing 25 µM glutamate and moved alternately to a position 50 µm away. The initial period of oscillation was 0.1 Hz, and the electrode completed its translation to each location in 1.5 s. At the arrow, the period of oscillation of the electrode was slowed to approximately 30 s. The record clearly shows that the electrical current induced by glutamate had reached more than 90% of its maximal response within the 10-s time frame of oscillation. While this sensor could be used in a static configuration to detect glutamate, the electrical drift inherent in the electrode, which can limit detection at low concentrations of glutamate, can also be plainly seen as the slow rise in the baseline current (Fig. 1A). Figure 1B shows the differential responses obtained when the electrode was employed in a self-referencing mode to reduce the impact of this drift. The electrode was initially placed 20 µm from a 25 µM glutamate source pipette, and differential recordings were made by subtracting responses obtained at a point 50 µm distant; the rate of oscillation was 0.1 Hz. In this condition, a steady differential signal of approximately 650 fA could be detected. The electrode was then progressively moved to positions more distant from the glutamate source, and differential recordings were obtained in the same fashion. The decline in the differential signal as a function of distance is apparent. Note also the clear, steady, small signal of approximately 50 fA that can be detected with the electrode 100 µm away from the source pipette—a signal that is significantly smaller than the electrical drift and noise depicted in the raw recordings presented in Figure 1A.
|
This work was supported by grants from the National Center for Research Resources (P41 RR01395) and National Science Foundation (009-1240). Special thanks to Robert Lewis, Richard H. Sanger, and Kasia Hammar for their efforts and assistance.
Literature Cited
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
