Anal Bioanal Chem (2005) 382: 1374–1379 DOI 10.1007/s00216-005-3309-y
O R I GI N A L P A P E R
Nina Dimcheva Æ Elena Horozova
Horseradish peroxidase-based organic-phase enzyme electrode
Received: 3 February 2005 / Revised: 25 April 2005 / Accepted: 29 April 2005 / Published online: 2 July 2005 Springer-Verlag 2005
Abstract An organic-phase enzyme electrode (OPEE) based on horseradish peroxidase (HRP) immobilized within Nafion on spectroscopic graphite was investigated in acetonitrile. The amperometric electrode response to hydrogen peroxide and cumene hydroperoxide present was found to be the result of the reduction of oxygen, produced upon enzymatic decomposition of both hydroperoxides (i.e., by the catalase-like activity of HRP). The electrode response was found to depend linearly on the hydroperoxide concentration up to 700 lM within the range of potentials from 200 to 400 mV (versus Ag|AgCl). Detection limits of approximately 45 lM for H2O2 and 100 lM for cumene hydroperoxide were determined under the selected experimental conditions. Nernstian dependence (the open circuit voltage of HRPbased electrode versus logarithm of H2O2 concentration) was obtained between 0.2 and 2.0 mM, with a slope of approximately 23 mV per logarithmic unit, suggesting a catalase-like, two-electron disproportionation of the substrate in acetonitrile. Keywords Horseradish peroxidase Æ Organic-phase enzyme electrode Æ Hydroperoxides Æ Catalase-like activity
Introduction Horseradish peroxidase (HRP) is one of the most extensively studied enzymes with large analytical applications in immunoassays and the development of bio-
N. Dimcheva (&) Æ E. Horozova Department of Physical Chemistry, Plovdiv University, 24 Tsar Assen St, 4000 Plovdiv, Bulgaria E-mail:
[email protected]
sensors for hydrogen peroxide monitoring [1]. Its in vivo function is to prevent the oxidative damage of living cells via catalytic oxidation of a variety of electron donors using H2O2 as oxidant. The usual catalytic enzyme cycle comprises three steps: – Oxidation of ferric enzyme with hydrogen peroxide forming the oxidized intermediate (Cpd I) and water: HRP þ H2 O2 ! Cpd I þ H2 O
ð1Þ
– One-electron oxidation of a second substrate S (e.g., phenols, quinones, aromatic amines, etc.) by Cpd I leads to the formation of Cpd II and a radical product: Cpd I þ S ! Cpd II þ S
ð2Þ
– Regeneration of the ferric enzyme by a molecule of substrate with liberation of a water molecule and S•: Cpd II þ S ! HRP þ S þ H2 O
ð3Þ
When immobilized on the electrode surface, HRP catalyzes the reduction of H2O2 according to reaction (1). It is then reduced by the electrode as a source of electrons with simultaneous uptake of two protons from the surrounding medium. The latter reaction is referred to as a direct electron transfer process, and together with reaction (1) represents the direct bioelectroreduction of hydrogen peroxide in buffer solutions. The affect of pH on the heterogeneous electron transfer rate constants for HRP-modified electrodes in hydrogen peroxide electroreduction has been extensively studied [2, 3, 4, 5]. The acceleration of the process at pHs below 7.0 was found to be the result of increased proton concentration, since proton uptake represents the ratelimiting step. The bioelectrochemical reduction of some organic hydroperoxides in aqueous medium proceeds through a similar catalytic cycle, and this was exploited for the development of peroxidase-based amperometric electrodes for the determination of these substances [6, 7].
1375
HRP-based non-aqueous biocatalysis has been reported to be effective in hydrogen peroxide analysis [8, 9], or as a component of a potentiometric cholesterol biosensor [10]. Both direct [11] and mediated [12, 13] bioelectrochemical reduction of hydroperoxides have been reported in water-containing organic media. In water-free aprotic organic solvents, however, bioelectrochemical reduction of hydrogen peroxide on an HRP-modified electrode is implausible due to the absence of protons in the medium. On the other hand, catalase-like activity of HRP could be expected in that medium, provided that hydrogen peroxide or other organic hydroperoxides as the only substrate for the enzyme. It has been demonstrated that HRP exhibits catalase-like activity [14, 15] in aqueous media when hydrogen peroxide is the sole substrate available to the enzyme, i.e., it catalyses the disproportionation of H2O2 leading to the formation of oxygen and water according to the following scheme [16]: Catalase þ H2 O2 ! Cpd I þ H2 O Cpd I þ H2 O2 ! Catalase þ O2 þ H2 O; where Catalase is the ferric enzyme, and Cpd I is catalase Cpd I. The present work deals with characterization of an HRP-based electrode for the direct assay of hydrogen peroxide and cumene hydroperoxide in acetonitrile. The detection principle of the electrode is based on electrochemical reduction of oxygen produced by the catalaselike enzymatic disproportionation of both hydroperoxides.
Experimental Reagents HRP (EC 1.11.1.7) Reanal, Budapest, Hungary) with specific homogeneous activity of the native enzyme of 350–500 U mg1 protein (1 U is the amount of the enzyme which oxidizes 1 lmol of o-dianisidine in the presence of H2O2 per minute at 25C and pH 7.0) and a Reinheitszahl (RZ) of 0.6 were used in the present study, where RZ is the absorbance ratio A403/A275. It is a measure of the hemin content of the peroxidase. Buffer solutions were prepared with analytical grade Na2HPO4Æ12H2O, KOH, H3PO4, and citric acid. Acetonitrile for UV spectroscopy (Fluka) was used as reaction medium; 30% solution of hydrogen peroxide and cumene hydroperoxide, purchased as 80% solution in cumene, both of analytical grade (Fluka) were used without further purification. The polymer (Nafion 117) was obtained from Fluka as a 5% solution in water/alcohols mixture. The polymer solution was neutralized with buffer (pH 7.0) and then diluted with doubly distilled water before use.
Electrode preparation A rod of spectrographic graphite RWI (purchased from Ringsdorff-Werke GmbH, Bonn-Bad Godesberg, Germany) with a diameter of ca. 0.5 cm, fitted into a Teflon cylinder with a platinum current lead, was used as working electrode. The electrode was first polished on emery paper P 400 and then on a filter paper to produce a mirror-like luster. It was then thoroughly rinsed with doubly distilled water and dried at ambient temperature before enzyme immobilization. The enzyme electrode was prepared according to a published procedure [17]. The electrode surface was coated with a 20-lL drop of a mixture of polymer (1.0% Nafion) and enzyme solution (200 lg peroxidase). The coating was allowed to dry at room temperature for 3 h. A similar polymeric coating, but without the enzyme, was used for control experiments. Electrochemical measurements and apparatus The electrochemical workstation was set up from a bipotentiostat type BiPAD (TACUSSEL, Villeurbanne, France) and a digital voltmeter type 1AB105 (ZPU, Pravets, Bulgaria). The electrochemical measurements were carried out in the potentiostatic regime using a three-electrode cell filled with acetonitrile. A silver|AgCl electrode was used as a reference electrode, and a platinum wire as counter electrode. The solution was purged with argon 20–30 min prior to and during the measurements. Tetraethylammonium bromide (Cyanamid, Italy) was used as a supporting electrolyte throughout the measurements. An aliquot of a 1 mM stock solution of the substrate (hydrogen peroxide or cumene hydroperoxide) in acetonitrile was added to 12 mL acetonitrile in the cell in order to obtain the steady state response of the electrode and its dependence on substrate concentration. The current of the electrode was monitored, and when it reached a constant value the next aliquot of substrate stock solution was added. The time required until the current reached a steady state did not exceed 90 s after any of the additions. All the data reported were obtained as the average of at least three independent measurements. A constant temperature in the cell was achieved by means of a thermostat UH (VEB MLW Pru¨fgera¨teWerk, Sitz Freital, Germany); the pH of the buffer for enzyme solution was adjusted using a pH meter OP-208 (Radelkis, Budapest, Hungary).
Results and discussion The electrochemical behavior of the peroxidase organicphase enzyme electrode was examined by using polarization curves. The stationary polarization curves of the HRP-based electrode (with subtracted background, i.e.,
1376
a
1.4 1.2
2 1
1.0
-IS /
0.8 0.6 0.4 2'
0.2 0.0 0
200
400
600
b 1.6
2
1.4 1.2 1
1.0 0.8 0.6 0.4 0.2
-0.20
800
CH2O 2 / µ M
-IS / µ A
recorded in the absence of hydroperoxide) in acetonitrile in the presence of hydrogen peroxide and cumene hydroperoxide are depicted at Fig. 1. Both dependencies have similar trends within the potential region studied (from 0 to 600 mV). The current reaches limiting values (plateau) between 100 and 350 mV in both cases and then increases at potentials more negative than 400 mV (most probably as a result of a contribution from non-enzymatic electrochemical reduction of the peroxides, especially at potentials more negative than 500 mV). The concentration dependencies for both hydroperoxides recorded at potentials from 200 to 400 mV showed linear trends up to 700 lM (Fig. 2). The sensitivities determined from the linear portion of these dependencies were found to be practically identical [d IS/d C = (18 ± 3) nA lM1]. Twice as low sensitivity was registered at 100 mV. The control experiments accomplished with an electrode covered with the same polymeric coating but without the enzyme (Fig. 2a and b, curves 2¢) showed that no electrode response was achieved after successive additions of either hydrogen peroxide (Fig. 2a) or cumene hydroperoxide (Fig. 2b) even at 400 mV (i.e., within the specified potential range no electrochemical process takes place with no enzyme present in the cover layer). At potentials more negative than 400 mV, increased noise levels made conditions difficult to work with. Within the potential range from 200 to 400 mV detection limits of 45 lM for hydrogen peroxide and approximately 100 lM for cumene hydroperoxide were determined (at signal to noise ratio 3:1). Our previous studies accomplished with a catalase enzyme electrode in acetonitrile, indicated analogies in the sensitivity of peroxide determination as well as in the linear parts of the concentration depen-
2'
0.0 0
200
400
600
800
C CHP / µ M
I S -I0
A
-0.15
Fig. 2 Concentration dependencies of the electrode response for hydrogen peroxide (a) and cumene hydroperoxide (b) in acetonitrile at 200 mV (1) and 400 mV (2 and 2¢). Curves 2¢ are obtained with the same electrode with no enzyme in the cover layer. Reference electrode Ag|AgCl; temperature 21±1C
-0.10
-0.05
0.00 0
-100
-200
-300
-400
-500
-600
E / mV Fig. 1 Stationary polarization curves (background subtracted) of the HRP-based electrode in acetonitrile in the presence of hydrogen peroxide (open circles) and cumene hydroperoxide (closed circles); reference electrode Ag|AgCl
dencies of the response within the same range of working potentials [18, 19]. Comparative experiments in the presence and absence of oxygen proved that under those working potentials oxygen deliberated in the enzyme layer is reduced electrochemically [17]. In view of these findings we assume that under our experimental conditions the amperometric electrode response is a result of oxygen electrochemical reduction. The latter is formed on enzyme-induced disproportionation of hydroperoxide compounds under investigation, i.e., HRP shows
1377
The possible use of the peroxidase organic-phase enzyme electrode as a potentiometric sensor for hydrogen peroxide was also examined. The open circuit voltage as a function of hydrogen peroxide concentration is linear between 250 and 2,000 lM in semi-logarithmic coordinates (Fig. 3). The measured background open circuit voltage (with no H2O2 present) was 3 mV (versus Ag|AgCl), and its overall change for the whole concentration range studied was 16 mV. Linear regression of the results indicated a change of approximately 23 mV per logarithmic unit, i.e., close to the value of 28 mV, which suggests a two-electron disproportionation of hydrogen peroxide (i.e., catalase-like function). These findings in principle show the possibility of using the OPEE described for potentiometric measurements of hydrogen peroxide in acetonitrile within the concentration range specified. Compared with amperometric electrodes, however, potentiometric sensors have two main disadvantages: – because of the logarithmic relationship between the potential measured and analyte concentration, the latter could be only roughly determined; 5 y = -22.608x - 79.884 R2 = 0.9845
0
E / mV
catalase-like activity in acetonitrile. Moreover, the linear sweep voltammograms of the HPR-based electrode in acetonitrile (not shown) did not indicate bioelectrochemical activity of the enzyme in either the absence or presence of hydrogen peroxide, since no peaks were observed on the voltammograms. Both our preliminary experiments with immobilized HRP enzyme and literature data [14, 15] confirmed its ability to decompose hydrogen peroxide when it is the sole enzyme substrate. Under our experimental conditions, intensive gas (oxygen) release from H2O2 solutions in acetonitrile was observed when immobilized peroxidase was present, whereas in the control sample (i.e., without enzyme), such gas evolution was not observed. The substrate specificity of the HRP-based electrode in acetonitrile was tested at a potential of 400 mV using two other peroxide compounds: meta-chloroperbenzoic acid (m-CPBA, a xenobiotic hydroperoxide) and benzoylperoxide. Although the stationary polarization curves of the enzyme OPEE recorded in the presence of m-CPBA differ from those recorded in background electrolyte (not shown) at a constant potential 400 mV, electrochemical response was not obtained up to concentrations of 700 lM. The most probable explanation for this finding is that no oxygen is produced in this case. The differences in the polarization curves are probably due to HRP deactivation by mCPBA in manner similar to that discussed by Arnao and co-workers [20, 21, 22]. With benzoyl peroxide, no interaction with the enzyme was registered by recording polarization curves and testing electrode response at constant potential with increasing analyte concentration. The temperature dependencies of the electrode response to hydrogen peroxide concentration at working temperatures of 20, 25, 30, and 35C and a potential of 400 mV showed linear trends with minor variation of the electrode sensitivity. In all experiments a detection limit of 45 lM was determined (at signal to noise ratio 3:1), and linear regression analysis of the calibration graphs affords regression coefficients close to unity (Table 1). The decrease of the electrode sensitivity noted with increasing temperature is most probably due to non-catalytic decomposition of hydrogen peroxide in the bulk solution (rate constants for this decomposition in aqueous medium are approximately 5·103 min1 and 8.5·103 min1 at 25 and 30C, respectively [23].
-5
-10
-15
-20 -4
-3
-2
-1
0
log CH2O 2 Fig. 3 Dependence of the open circuit voltage on logarithm of hydrogen peroxide concentration in acetonitrile for the HRP-based OPEE calibration graph; temperature 21±1C; reference electrode Ag|AgCl. The linear regression equation and the regression coefficient are given in the right upper side of the plot area
Table 1 Electrode sensitivity (determined as the slope of the linear portions of the calibration graphs), linear regression equations, and the average regression coefficients for the concentration dependencies of the response at a working potential of 400 mV within the temperature range 20–35C Temperature (C)
Sensitivity dI/dC (nA lM1)
Linear regression equation
20 25 30 35
20 19 17 16
y y y y
= = = =
0.0020 0.0019 0.0017 0.0016
x x0.0177 x0.0446 x0.0869
R2 0.988 0.951 0.997 0.975
1378
– since the open circuit voltage should be measured at equilibrium, the procedure of the measurements is rather time-consuming. The long-term storage stability of the HRP-based electrode was tested over a period of about 400 h. The remaining activity of the OPEE as a function of the storage time is represented at Fig. 4. The remaining activity of the enzyme electrode was estimated as the ratio of the electrode response to that initially registered, at a constant substrate concentration. For a freshly prepared electrode its response sharply decreases to about 60% of the initially registered under equivalent experimental conditions within the first hour of use. Within the next week of storage, the electrode response then smoothly decreases down to about one third of the initially registered value. Since the reaction medium is able to extract water from the immobilized enzyme microenvironment, this drastic decay of activity could be due to water removal (which in turn leads to conformational rigidity of the biocatalyst and consequent loss of its activity) or to biocatalyst deactivation caused by the substrate [15].
Conclusions An enzyme electrode based on HRP immobilized within a film of Nafion on spectrographic graphite was found to be capable of operating in acetonitrile. The amperometric electrode response to hydrogen and cumene hydroperoxides present was found to be the result of the
reduction of oxygen, produced by enzymatic decomposition of both hydroperoxides (i.e., by catalase-like activity of HRP). The optimum range of working potentials was from 200 to 400 mV (versus Ag|AgCl). Within this range calibration plots indicate response linearity up to 700 lM for both hydroperoxides and detection limits of approximately 45 and 100 lM for H2O2 and cumene hydroperoxide, respectively. Similarities with the operational characteristics of a catalase OPEE under the same experimental conditions [18, 19] were noted: sensitivity (17 ± 3) nA lM1; linearity of the calibration graphs up to about 700 lM; detection limits approximately 45 lM for hydrogen peroxide and 100 lM for cumene hydroperoxide within the range of working potentials from 200 to 400 mV. Differences were found, however, in the substrate specificity of the enzymes: the catalase OPEE is applicable to the analysis of m-CPBA or benzoylperoxide [17], whereas the HRP-based enzyme electrode is not. The slope of the Nernstian dependence (the open circuit voltage versus logarithm of hydrogen peroxide concentration) suggests exchange of two electrons. The fair stability of the OPEE developed was assigned to fast enzyme deactivation because of either the influence of the solvent or the biocatalyst damage caused by hydrogen peroxide. Acknowledgments Authors acknowledge kind support of this work from the University of Plovdiv Research Fund and express their gratitude to Ms. D. Georguieva for technical help.
References
Remaining activity / %
100 80 60 40 20 0 0
100
200
300
400
Storage time / h Fig. 4 Remaining activity (%) as a function of storage time for HRP-based OPEE; working potential 200 mV (versus Ag|AgCl); substrate concentration 350 lM, temperature 21±1C. The electrode remaining activity was determined as the ratio of the measured electrode response (subtracted background) to its initially registered value at a constant substrate concentration
1. Gorton L, Lindgren A, Larsson T, Munteanu FD, Ruzgas T, Gazarian I (1999) Anal Chim Acta 400:91–108 2. Ferapontova E, Gorton L (2001) Electrochem Commun 3:767– 774 3. Christenson A, Dimcheva N, Ferapontova E, Gorton L, Ruzgas T, Stoica L, Shleev S, Yaropolov A, Haltrich D, Thorneley R, Aust S (2004) Electroanalysis 16:1074–1092 4. Ferapontova E, Puganova E (2002) J Electroanal Chem 518:20–26 5. Ferapontova E, Gorton L (2002) Bioelectrochemistry 55:83–87 6. Wollenberger U, Wang J, Ozsoz M, Scheller F, Gonzalez-Romero E (1991) Bioelectrochem Bioenerg 26:278–296 7. Ossina MA, Bogdanovskaya VA, Tarassevich MR, Ussipovich AI (2004) Russ J Electrochem 40:923–933 8. Iwuoha EI, Saenz de Villaverde D, Garcia NP, Smyth MR, Pingarron JM (1997) Biosens Bioelectron 12(8):749–761 9. Wang J, Lin Y, Chen L (1993) Analyst 118:277–280 10. Ciucu A, Ciucu C, Baldwin RP (2002) Roum Biotechnol Lett 7:625–630 11. Popescu IC, Cso¨regi E, Gorton L (1996) Electroanalysis 8(11):1014–1019 12. Schubert F, Saini S, Turner APF (1991) Anal Chim Acta 245:133–138 13. Schubert F, Saini S., Turner APF, Scheller F (1992) Sens Actuators B 7:408–411 14. Hiner ANP, Hernandez-Ruiz J, Williams GA, Arnao MB, Garcia-Canovas F, Acosta M (2001) Arch Biochem Biophys 392:295–302 15. Hernandez-Ruiz J, Arnao MB, Hiner ANP, Garcia-Canovas F, Acosta M (2001) Biochem J 354:107–114 16. Dunford HB (1999) Heme peroxidases. Wiley, New York
1379 17. Horozova E, Dimcheva N, Jordanova Z (2002) Bioelectrochemistry 58:181–187 18. Dimcheva N, Horozova E, Jordanova Z (2004) Bulgarian Chem Ind 75:1–7 19. Dimcheva N, Horozova E, Shougova T (2005) Monatshefte fur Chemie 136:147–152 20. Hiner ANP, Ruiz JH, Arnao MB, Lopez JNR, Canovas FG, Acosta M (2002) J Enzyme Inhib Med Chem 17:287–291
21. Arnao MB, Hernandez-Ruiz J, Varon R, Garcia-Canovas F, Acosta M (1995) J Mol Catal A Chem 104:179–191 22. Rodriguez-Lopez JN, Hernandez-Ruiz J, Garcia-Canovas F, Thorneley RNF, Acosta M, Arnao MB (1997) J Biol Chem 272:5469–5476 23. Dimcheva N (2001) PhD Thesis, Bulgarian Academy of Sciences, Sofia
