Development of Glutamate Sensor for Neurotransmitter Imaging

We report an enzyme-based potentiometric glutamate sensor combined with a semiconductor device for neurotransmitter imaging, the response of which is based on the detection of the interfacial potential of a gold electrode produced by redox and enzyme reactions. First, we confirmed that the interfacial potential increased logarithmically with the increasing concentration ratio of hexacyanoferrate(III) to hexacyanoferrate(II). The proposed sensor could detect the interfacial potential change of the gold electrode with a slope of 66.1 mV/decade. H2O2 and l-glutamate were detected as a function of the change in the interfacial potential of the gold electrode in the range of 0.01–1 mM. Slopes of 55.2 and 41.9 mV/decade were obtained as the sensitivities for H2O2 and l-glutamate, respectively. These estimated values are larger than the theoretical value of 29.5 mV/decade. Although the reason why the difference between the estimated value and the theoretical value occurred could not be revealed, the interfacial potential is expected to depend on some kind of reaction related to H2O2 concentration because the logarithmic response was obtained for H2O2 concentration. Therefore, the results suggested that the sensor can detect the l-glutamate concentration.


Introduction
l-Glutamate is widely known as a major neurotransmitter in mammalian brains. It has been pointed out that the spatiotemporal distribution of extrasynaptic l-glutamate concentrations in brains is an important factor in controlling neural functions. (1) Accordingly, sensing devices that can visualize l-glutamate distribution are required.
Amperometric and potentiometric sensors for detecting H 2 O 2 , as well as those for detecting l-glutamate, have been studied. (2)(3)(4)(5)(6)(7)(8)(9) Among these electrochemical biosensors, potentiometric devices are advantageous for imaging because their sensitivity essentially does not change with decreasing sensing area. In addition, potentiometric image sensors, or pH image sensors, have been developed to visualize the two-dimensional pH distribution in solutions. (10) The sensing of several biomolecules such as adenosine triphosphate (ATP) and acetylcholine (ACh) was demonstrated using pH image sensors combined with enzymatic membranes. (11,12) However, since the sensing of biomolecules relies on the H + generated by enzymatic reactions, the output of the pH image sensor can be affected by the pH variations in a specimen, the ionic strength, and the buffer capacity of the solution. For this reason, development has been focused on biomolecule sensors based on pHindependent potentiometric sensing. (13) In this study, a pH-independent potentiometric glutamate sensor utilizing potential changes of the gold electrode caused by a redox reaction combined with some enzymatic reactions was proposed.

Measurement principle of the concentration of l-glutamate
l-Glutamate sensing is essentially based on the change in interfacial potential between a gold electrode and a solution. A similar study has been reported. (13) Hydrogen peroxide (H 2 O 2 ) as an oxidant is generated by an enzyme reaction involving l-glutamate, as shown in FcMeOH is a metal complex described with the formula C 11 H 12 FeO. The oxidation state of iron of FcMeOH is +2, and that of iron of FcMeOH + is +3. The concentration ratio of FcMeOH to FcMeOH + is changed as a function of the l-glutamate concentration.
In this case, the interfacial potential of the gold electrode is determined by the equilibrium between FcMeOH and FcMeOH + . (13) On the other hand, from Eq. (2), the interfacial potential can also be determined by the equilibrium between H 2 O 2 and H 2 O. Therefore, the interfacial potential is described by Eqs. (3) and (4).
Here, E°′ is the formal potential, R is the universal gas constant, T is the absolute temperature, n is the number of electrons transferred in the redox reaction, and F is the Faraday constant. Ox and Red are the concentrations of an oxidant and a reductant, respectively. Equation (4) reveals that the relationship between the interfacial potential of the gold electrode and the H 2 O 2 concentration becomes 29.5 mV/decade. From Eqs. (1) and (4), the interfacial potential of the gold electrode is expected to increase logarithmically as the l-glutamate concentration increases. The source follower circuit detects the output (V out ) expressed by Here, E is the interfacial potential of the gold electrode, E RE is the potential of the reference electrode to the solution, V ref is the constant voltage applied to the reference electrode, and V gs is the gate-to-source voltage. V gs and E RE are constants. Figure 1 shows a schematic of the measurement setup. It consists of a gold electrode and a source follower circuit with a metal-oxide-semiconductor field-effect transistor (MOSFET). A gold electrode with an area of 1 × 1 mm 2 was deposited on an SiO 2 /Si chip with an adhesion layer of titanium, as shown in Fig. 1. The titanium and gold films were deposited by sputtering, and the thicknesses of the films are about 40 and 300 nm, respectively. The metal films were patterned by a lift-off process. Figure 2 shows a photograph of the gold electrode with the name of each part.

Measurement system
For the source follower circuit, either an n-channel enhancement-mode or an n-channel depletion-mode MOSFET was used. Both have a gate oxide with a thickness of about 100 nm grown by thermal oxidation for 150 min at 1000 °C. In the n-channel enhancement-mode

Measurement in the solutions of hexacyanoferrate(III) and (II)
We first measured the output for different concentration ratios of hexacyanoferrate(III) to hexacyanoferrate(II) to verify the principle of the sensor. The measurements were conducted according to Ref. 14 as follows. The sensor and a reference electrode were immersed in 100 μL of the solutions of hexacyanoferrate(III) and hexacyanoferrate(II) at ratios of 1:9, 2:8, 3:7, 4:6, 5:5, 6:4, 7:3, 8:2, and 9:1. The total concentration of each solution was 10 mM, and hexacyanoferrate(III) and (II) were dissolved in 100 mM HEPES buffer solution. The output was then measured for 5 min.

Measurement of H 2 O 2 solutions of varying concentrations
The gold electrode and a reference electrode were immersed in 45 μL of the enzyme solution. The enzyme solution consisted of 80 mM HEPES buffer solution, 500 μM FcMeOH, and 0.12 units/ μL HRP. FcMeOH was dissolved in methanol before being added to the HEPES buffer solution. During the measurements, 5 μL of H 2 O 2 solution at each concentration diluted with 80 mM HEPES buffer solution was added at 10 min intervals so that the final concentrations of H 2 O 2 became 0.01, 0.05, 0.1, 0.5, and 1 mM. Every H 2 O 2 solution also contained 500 μM FcMeOH.

Measurement of l-glutamate solutions of varying concentrations
The gold electrode and a reference electrode were immersed in 45 μL of the enzyme solution. The enzyme solution consisted of 80 mM HEPES, 500 μM FcMeOH, 0.12 units/μL HRP, and 0.028 units/μL l-glutamate oxidase. FcMeOH was dissolved in methanol before being added to the HEPES solution. During the measurements, 5 μL of l-glutamate solution at each concentration diluted with 80 mM HEPES buffer solution was added at 10 min intervals so that the final concentrations of l-glutamate were 0.01, 0.05, 0.1, 0.5, and 1 mM. Every l-glutamate solution also contained 500 μM FcMeOH. Figure 3 shows the output in response to the concentration ratio of hexacyanoferrate(III) to hexacyanoferrate(II). The output increased as the concentration ratio of hexacyanoferrate(III) to hexacyanoferrate(II) increased. Here, the increase in the output means the increase in the interfacial potential because V gs and E RE are constant. The output was stable for 5 min. Figure 4 shows the relationship between the ratio of hexacyanoferrate(III) to hexacyanoferrate(II) and the output 5 min later. The output change in each ratio of hexacyanoferrate(III) to hexacyanoferrate(II) was plotted after 5 min. The output increased logarithmically with incremental changes in the ratio of hexacyanoferrate(III) to hexacyanoferrate(II) in the range of 1:9−9:1, and the slope of the line is 66.1 mV/decade.  From the Nernst equation, the theoretical value is expected to be 59.1 mV/decade, as shown in Eq. (6). (14) [Fe(CN) 6 ] 3− and [Fe(CN) 6 ] 4− describe the concentrations of hexacyanoferrate(III) and hexacyanoferrate(II), respectively. The results suggest that the proposed sensor can detect the interfacial potential change of the gold electrode with a slope of 66.1 mV/decade.  Figure 6 shows the relationship between the H 2 O 2 concentration and the interfacial potential change of the gold electrode. The interfacial potential change in Fig. 6 was obtained by subtracting 0.0966 V, which corresponds to the potential before H 2 O 2 addition indicated by the dotted line. The time at which the data was employed is described in Fig. 6. The interfacial potential increased  logarithmically as the H 2 O 2 concentration increased, and the slope of 55.2 mV/decade was obtained. Therefore, the interfacial potential is expected to change depending on some kind of reaction related to the H 2 O 2 concentration. However, the estimated value of the slope expressed in Fig. 6 was nearly twice as large as the theoretical value, and the reason why the difference between the two occurred could not be revealed. Further study is necessary to investigate the difference between the two. Figure 7 shows the output in response to the concentration of the l-glutamate solution. The output increased as the final l-glutamate concentration increased and was stable in 8 min after adding each l-glutamate solution for the l-glutamate concentration from 0.01 to 0.1 mM. The output fluctuation in 8 min after adding each l-glutamate solution until adding the next l-glutamate solution was only about 0.5 mV. Figure 8 shows the relationship between the l-glutamate concentration and the interfacial potential change. The interfacial potential change in Fig. 8 was obtained by subtracting 0.0724 V, which corresponds to the potential before l-glutamate addition indicated by the dotted line. The time at which the data was employed is described in Fig. 8. The interfacial potential increased logarithmically as the l-glutamate concentration increased, and the value of the slope was expressed in Fig. 8. Because the slope was close to that for the H 2 O 2 concentration expressed in Fig. 6, the interfacial potential is expected to increase by increasing the H 2 O 2 concentration. However, the slope of 41.9 mV/decade was lower than that for the H 2 O 2 concentration. The enzymatic reaction expressed in Eq. (1) may have proceeded insufficiently.

Conclusions
An enzyme-based potentiometric glutamate sensor was proposed for neurotransmitter imaging. We showed that the proposed sensor can detect the interfacial potential of the gold electrode, which depends on the logarithmic function from the measurement of the solution of hexacyanoferrate(III) and hexacyanoferrate(II). The sensor showed a response to the l-glutamate concentration in the l-glutamate concentration range of 0.01-1 mM with a sensitivity of 41.9 mV/decade. However, the slope differed from the theoretical value expected from the enzymatic reactions. Therefore, further investigation is necessary to reveal the reason why the difference between the estimated value and the theoretical value occurred.