Dehydrogenase-modified carbon-fiber microelectrodes for the

Anal. Chem. 1993, 65, 623-630. 623. Dehydrogenase-Modified Carbon-Fiber Microelectrodes for the. Measurement of Neurotransmitter Dynamics. 2. Covalent...
0 downloads 0 Views 929KB Size
Anal. Chem. 1999, 65, 623-630

02s

Dehydrogenase-Modified Carbon-Fiber Microelectrodes for the Measurement of Neurotransmitter Dynamics. 2. Covalent Modification Utilizing Avidin-Biotin Technology Paul Pantano and Werner G. Kuhr'

Department of Chemistry, University of California, Riverside, California 92521

Dehydrogenase enzymes are immoblllzed onto carbon-fiber mkroelectrode surfaces via avklln-blotln technology and a covahtty linkedhydrophWktether. The avldbbbtln coupkg strategy allows the seiectlvlty of the electrochemkal measurement to be easily changed without reopthizing tho knmobllltatkncomlltlonr. Optlmlzedderlvathatbnconditlonr are demonstrated for dimer, tetramer, and hexamer dehydrogenases. Noneiectroactlvesubdratw (ethanol, ghmse6-phorphate, and glutamate) are quantitated through the detectlon of enzymegenerated NADH by fast scan cyclic staircase voltammetry (100 Vis). The glutamate dehydrogen--7a30mrt h e with a detection limit of 0.5 mM glutamate and a 1-80 mM linear concentration range.

INTRODUCTION The goal of this work is the development of fasbresponding, enzyme-modified microelectrodes for the detection of nonelectroactive species (i.e., amino acid neurotransmitters). Dehydrogenasecatalyzedreactions which oxidize amino acids and simultaneously reduce NAD+ to NADH are very rapid. The NADH generated in this manner acts as an electrontransfer mediator which can be monitored by fast scan cyclic staircase voltammetry (100V/s) at the carbon-fiber surface. Previously, we have demonstrated that fast-scan cyclic staircase voltammetrycan minimize the difficulties associated with the electrochemical measurement of NADH when performed at carbon-fiber microelectrodes.1 In this paper, avidin-biotin coupling is used for the immobilization of the dehydrogenase to the electrode surface. This general strategy allows the selectivity of the measurement to be altered simply by coupling one of 200 possible dehydrogenases to the electrode. While many strategies have been developed for the immobilization of enzymes onto electrode surfaces, few have been devised that afford the small probe size ( ADH > GDH Table 11) seems to indicate that the heterogeneous response is not limited by the catalytic rate of the enzyme but by other physical properties of the modified electrodes. This becomes more evident when one considers the effect of the derivatization on the voltammetry of NADH at the modified electrode. NADH V o l t a m m e t r y at Dehydrogenase-Modified Microelectrodes. The electrochemicalperformance of a carbonfiber surface that acta as both a site of enzyme attachment and as a site of electron transfer is extremely dependent on the chemical and physical properties of the derivatized surface. Since oxidases can utilize very small and electrochemicallyreversibleredox mediators,their electrochemicalperformance was less dependent on the chemical nature of the electrode surface. For dehydrogenases, the procedures used for the derivatization of the electrode surface dramaticallyaffect the rate of electron transfer for NADH. This dependence is demonstrated by the background-subtracted cyclic staircase voltammograms (100 V/s) for the oxidation of 100 pM NADH. before derivatization (Figure 3A), following derivatization with Jeffamine ED-600 (Figure 3B), and following the treatments with sulfo-NHS-biotin and ExtrAvidin (Figure 3C). The electron-transfer properties of the carbon surface degrade as a function of the extent of surface derivatization. The faradaic current recorded between 750 and 10oO mV vs Ag/AgCl decreased by 67 5% relative to the response observed at underivatized carbon fibers following the Jeffamine reaction. Another factor of two was lost following derivatization with sulfo-NHS-biotin and ExtrAvidin. These voltammograms also show that a large overpotential develops for the oxidation of NADH as the electrode is derivatized. The oxidative peak potential is shifted by at least 200 mV positive of that observed at the underivatized carbon-fiber surface. Nonetheless, reproducible and recognizable voltammetry for enzyme-generatedNADH can be obtained from the dehydrogenase-modified microelectrodes (Figure 3D).The top trace represents the background-subtracted cyclic voltammogram (100 V/s) for the oxidation of enzyme-generated NADH following the injection of 6 mM glutamate. The identity of this faradaic peak is confirmed by the bottom trace where the glutamate solution was spiked with 100 pM NADH. Optimization of the analytical performance of these dehydrogenase-modified electrodes can be realized through judicious adjustment of the derivatization conditions and reversibility of NADH oxidation at an enzyme loaded surface. The improved reproducibility of NADH oxidation a t a carbon fiber followingan HC1electrochemical pretreatment has been demonstrated.' The most successful protocol for the production of a GDH-modified electrode to date used a combination of (1) an electrochemical pretreatment of the electrode in HC1; (2) derivatization with the longest available tether molecule (Jeffamine ED-2001); and (3) biotinylation with sulfo-NHS-LC-biotin, Preliminary results from GDHmodified microelectrodes constructed in this manner demonstrated an increase in the success rate (Table I, column 6) as well as an increase in sensitivity. A linear response is observed within a glutamate concentration range of 1-60 mM (r* = 0.981, n = 5) and an approximate limit of detection of 0.5 mM glutamate. A representative time course for the appearance of enzyme-generated NADH at these enzyme electrodes is presented in Figure 4A. The 300-ms rise time is essentially diffusion-limited since injection of NADH produces a response time on the order of 184 f 65 ms.' Variation of the EDC reaction time demonstrates the impact of the derivatization reactions on the voltammetry of NADH (Figure4B,C). The solid lines in each figure represent the NADH voltammetry at the underivatized surface; the crosses represent the voltammetry of NADH after the entire

629

J

e

E

-3.6

0

a

.-

e

I

0

0

-3.4

5

0

10

fino

15

(SI

0.0

B +++.-A /

L L

I

1000

500

0

PotenUal (mVvs. Ag/AgCl) t

/I

Flgurr 4. Fast scan cyclic staircase voltammetry (100 V/s) of modifled carbon-fiber microelectrodes (electrochsmlcaily pretreated In 1 M HCI). (A) Oxidative peak current for enzymegenerated NADH following an 8-s FIA InJectlon of 3 mM glutamate to a QDKmodM microelectrode. The current between 800 and 1000 mV vs AglAgCi was backgroundcorrectedby subtractbnof the current found between 500 and 700 mV from the same scan. (6)Background-subtracted cyclic staircase voltammograms (12 scans) for the oxidation of 100 pM NADH at a ODH-modified electrode before(line) and after (crosses) the enzyme derlvatizatkm reactlons (EDC reaction time was 1 h). (C) Same as B except that the EDC reactlon time was 14 h.

enzyme modification was completed. The only difference between these two electrodes was a 14-fold increase in the EDC reaction time for the electrode shown in Figure 4C. This demonstrates that the extent of carboxylate activation (directly related to the length of the EDC reaction) changes the coverage of the biotinylated-Jeffamine and ExtrAvidin. This, in turn, reduces the number of sites for attachment of the biotinylated enzyme but increases the availability of electroactive sites for the oxidation of NADH. Of course, a net reduction in surface loading of enzyme may reduce the ultimate sensitivity of the modified microelectrodes, but this must be balanced against the loss of sensitivity due to the effects of the derivatization on the voltammetry of NADH. Thus, it is important to characterize both the protein loading and the electrochemicalproperties of NADH at carbon-fiber microelectrodes derivatized in this fashion.lJ9*20 ACKNOWLEDGMENT This work was supported by the National Institutes of Health (Grant GM44112-01A1) and by a gift from Eli Lilly

650

ANALYTICAL CHEMISTRY, VOL. 65, NO. 5, MARCH 1, 1993

and Company, Inc. We also thank Thomas Hellman Morton for valuable discussionsand Eric W. Kristenaen for the latest revision of his voltammetry software. W.G.K. is the recipient of a Presidential Young Investigator Award (NSF CHE8957394);P.P. was awarded the Edward G. Weston Summer Fellowship of the Electrochemical Society and an ACS

Division of Analytical Chemistry Graduate Fellowehip sponsored by the Eastman Chemical Company.

RECEIVED for review June 2, 1992. Accepted November 4, 1992.