Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.enzyme.chem.msu.ru/ryabov/pdf/JBIC_97.pdf
Дата изменения: Thu May 11 14:38:22 2000
Дата индексирования: Mon Oct 1 21:45:11 2012
Кодировка:

JBIC (1997) 2 : 182190

Q SBIC 1997

ORIGINAL ARTICLE

Alexander D. Ryabov 7 Vasily N. Goral

Steady-state kinetics, micellar effects, and the mechanism of peroxidase-catalyzed oxidation of n -alkylferrocenes by hydrogen peroxide

Received: 15 July 1996 / Accepted: 15 November 1996

Abstract Kinetics of the steady-state oxidation of nalkylferrocenes (alkyl p H, Me, Et, Bu and C5H11) by H2O2 to form the corresponding ferricenium cations catalyzed by horseradish peroxidase has been studied in micellar systems of Triton X-100, CTAB, and SDS, mostly at pH 6.0 and 25 7C. The rate of oxidation of ferrocenes with longer alkyl radicals is too slow to be measured. The reaction obeying the [RFc]:[H2O2] p 2:1 stoichiometry is strictly first-order in both HRP and RFc in a wide concentration range. The corresponding observed second-order rate constants k, which refer to the interaction of the peroxidase compound II (HRPII) with RFc, decrease with the elongation of the alkyl substituent R, and this in turn is accompanied by an increase in the formal redox potentials E7b in the same medium. Increasing the surfactant concentration lowers the rate constants k, the effect being due to the nonproductive binding of RFc to micelles rather than to enzyme inactivation. The micellar effects are accounted for in terms of the Berezin pseudo-phase model of micellar catalysis applied to the interaction of enzyme with organometallic substrates. The oxidation was found to occur primarily in the aqueous pseudo-phase and the calculated intrinsic second-order rate constants kw are (1.9 B 0.5)!10 5, (2.7 B 0.1)!10 4, and (5.9 B 0.6)!10 3 M 1 s 1 for HFc, EtFc, and nBuFc, respectively. The data obtained were used for estimating the self-exchange rate constants for the HRP-II/ HRP couple in terms of the Marcus formalism. Key words Ferrocene 7 Hydrogen peroxide 7 Horseradish peroxidase 7 Micellar effects 7 Kinetics

Abbreviations HRP horseradish peroxidase 7 TOP tobacco peroxidase 7 RFc nalkylferrocene 7 SDS sodium dodecylsulfate 7 CTAB cetyltrimethylammonium bromide 7 ABTS 2,2b-azino-di(3-ethylbenzthiazoline-6-sulfonic acid) 7 cmc critical micelle concentration

Introduction
Peroxidase-catalyzed reactions between peroxides and ferrocene derivatives are important redox processes that comprise a basis for assembling a variety of amperometric biosensors with the mediated electron transfer (for recent review, see [1]). In the case of hydrogen peroxide the reaction is very convenient for the spectrophotometric monitoring of H2O2 in the presence of HRP [2]. The knowledge of basic mechanisms of the peroxidase-catalyzed oxidation of ferrocenes by H2O2, first reported by Epton et al. [3], is thus a challenging goal from both theoretical and applied points of view. A strong impetus of the present investigation arises from our permanent interest in understanding mechanisms of conversion of organometallic compounds in the presence of enzymes in general and oxidoreductases in particular (for reviews, see [4, 5]). It should also be emphasized that ferrocenes are relatively simple molecules which, when oxidized by H2O2 in the presence of HRP, end up in a single-electron oxidation product, viz. ferricenium ions RFc c. The enzymatic oxidation is thus not interfered with by subsequent transformations of intermediate, usually radical, products. Ferrocenes are therefore attractive mechanistic probes for investigating intimate features of the electron transfer in catalysis by such a widespread enzyme as HRP isoenzyme c [68]. Much structural mechanistically relevant information is currently available [912], but there is no indication of a decrease in the number of mechanistic studies with these particular closely related peroxidases (see, for example [1317]). Moreover, numerous intriguing applications of the enzyme raise nov-

A. D. Ryabov (Y) Department of Chemistry, Moscow State University, 119899 Moscow, Russia Fax: c95-939-54-17; e-mail: ryabov@enzyme.chem.msu.su A. D. Ryabov 7 V. N. Goral Division of Chemistry, G. V. Plekhanov Russian Economic Academy, Stremyanny per. 28, 113054 Moscow, Russia

183

el mechanistic questions (see, for example [1820]), one of which arises from the fact that the aqueous solubility of various peroxidase substrates is often increased by surfactants which, in principle, may drastically affect the reactivity [21]. Therefore, effects of surfactants deserve special attention. The ferrocene family shown in Chart 1 is an attractive set of systematically modified compounds with different bulkiness and hydrophobicity and, hence, different affinity toward micelles. Therefore, in this work we report the results of the steadystate kinetic study of the HRP-catalyzed oxidation of ferrocenes by H2O2 and the effects of the nature of both RFc and surfactants (CTAB, SDS and Triton X100) in this enzymatic oxidation. The results obtained, which are rationalized in terms of the pseudo-phase model of micellar catalysis [21, 22], indicate that ferrocene displays "non-enzymatic" kinetics, but nevertheless appears to be a reasonably reactive substrate of HRP. The combination of these two factors allows us to apply the Marcus formalism [23] to the electron transfer involving HRP and calculate the self-exchange rate constant for the HRP compound II/HRP resting state couple.

Fig. 1 Increase in absorbance that accompanies the HRP-catalyzed oxidation of ferrocene into ferricenium ion by H2O2 at pH 6.0 and 25 7C; [HRP] p 1.5!10 8 M, [H2O2] p 1.6!10 4 M, [HFc] p 2.75!10 4 M, [CTAB] p 0.05 M, pH 6; the spectra are recorded at 2 min intervals

function of the total [H2O2] added, the slope being 1.97 B 0.01 in agreement with Eq. 1. It is important to note that the steepness of each step decreases successively because of the consumption of ethylferrocene, and, as we will see below, the slope of each step is in perfect agreement with the steady-state kinetic data. Steady-state Kinetics The recognized problem in the peroxidase catalysis is the lowering of enzymatic activity in an excess of H2O2 [68]. Therefore, the kinetic measurements were preceded by a determination of the optimal hydrogen peroxide concentration for the catalysis. The steady-state

1 RpH (a), Me (b), Et (c), nBu (d), nC5H11 (e), nC6H13 (f)

2

3

4 Chart 1

Results and discussion
Stoichiometry The HRP-catalyzed oxidation of ferrocenes by H2O2 can be followed spectrophotometrically either in the UV or visible spectral regions. It can be seen from Fig. 1 that the sensitivity is an order of magnitude higher in the UV region, but the advantage of using the band at ca. 620 nm is the lack of interference from any other species in the system and, importantly, the fact that the extinction coefficient of the product is independent of the nature and concentration of surfactant used. The reaction stoichiometry given by Eq. 1 was investigated in detail in our previous analytically oriented work [2]. 2 RFc c H2O2
HRP

* 2 RFc

c

c 2OH

P

(1)
Fig. 2 Spectrophotometrically (622 nm) detected formation of ethylferricenium dye as a result of consecutive addition of 0.0004 mmol H 2O 2 to 1.96 mL solution of EtFc: [EtFc] p 5!10 3 M, [HRP] p 10 7 M, [Triton X-100] p 0.04 M, pH 6, 25 7C

However, some data relevant to the present kinetic investigation is worth reconsidering. In particular, an increase in absorbance on consecutive addition of H2O2 to a solution of HRP and EtFc at pH 6 is shown in Fig. 2. The total amount of EtFc c formed is a linear

184

Fig. 3 Steady-state rate of HRP-catalyzed oxidation of HFc vs [H2O2]: [HFc] p 7.2!10 4 ( i) and 4.8!10 4 M (I), [HRP] p 8.7!10 8 M, [Triton X-100] p 0.03 M, pH 6, 25 7C

holds even at a lower concentration of H2O2, viz. 0.5!10 4 M. The same conclusion is reached from the data in Fig. 2, if the slope of each step in the stairway is analyzed. Since each successive slope corresponds to a lower "initial" concentration of EtFc due to its conversion into the ethylferricenium ion, the corresponding "new" [EtFc] could be calculated using p 310 M 1 cm 1 for EtFc c at 622 nm and the steady-state rate at this [EtFc] could be derived. The data calculated in this way is shown in Fig. 4 as open circles, which are very close to the closed squares measured directly as in all other cases. This encouraging agreement demonstrates the absence of peroxidase inactivation by ferricenium cations under the reaction conditions, at least in a matter of 60 min (time scale of the experiment shown in Fig. 2). All ferrocenes in Chart I exhibit a first-order dependence on RFc, and hence the reactions follow the second-order rate law. Rate p k[HRP][RFc] (2) It should be emphasized that Eq. 2, obtained in the steady-state experiment, is untypical of enzymatic processes because the reaction rate does not level off at high substrate concentrations as, for instance, is observed in the HRP-catalyzed oxidation of odianisidine [24], hexacyanoiron(II) [25], guaiacol [26], iodide [26], or ABTS [26, 27]. Ferrocenes thus display "non-enzymatic" reactivity. It is also worth noting that the reaction rate decreases drastically on going from npentylto nhexylferrocene and cannot be measured with good precision in the latter case. The observed second-order rate constants k are summarized in Table 1 together with the formal redox potentials E7b measured under the same conditions. Assuming the accepted mechanism for the peroxidase catalysis shown by Eqs. 35 [68], where HRP-I is an intermediate compound I bearing a ferryl porphyrin radical cation Fe IV p O, and HRP-II is compound II, an intermediate that is one oxidation equivalent above Fe III, we believe HRP c H2O2
k
1

rate of oxidation of ferrocene is shown in Fig. 3 as a function of [H2O2]. As can be seen, there is a broad plateau at an H2O2 concentration around 2!10 4 M which persists at different [HFc] values. A similar dependence was also observed in the case of nBuFc. All other measurements were performed at [H2O2] p 2!10 4 M, i. e. under the most favorable conditions. As would be expected, the steady-state rate is a linear function of the HRP concentration over its range 10160 pM. The representative data demonstrating how the steady-state rate depends on the concentration of different ferrocenes are shown in Fig. 4. The enzymatic oxidation is strictly first order in a wide range of concentrations of all ferrocenes used. The first order

* HRP-I

(3)

Table 1 Observed second-order rate constants k for the oxidation of ferrocenes catalyzed by HRP; pH 6, 25 7C, [Triton X100]p0.04 M, [H2O2]p2!10 P4 M RFc HFc MeFc EtFc n-BuFc n-C5H11Fc 1,2,4,1b,2b,4b-Me6Fc FcCOOH FcSOPNa c 3
a b c

k (M

P1

s

P1

)

E7b (mV) 210B 170B 195B 255B 300B 25B 5 5 5 5 5 5

c

Fig. 4 Steady-state rate of HRP-catalyzed oxidation of nalkylferrocenes by H2O2 against their concentrations: [HRP] p 10 7 M, [H2O2] p 2.4!10 4 M, [Triton X100] p 0.04 M, pH 6, 25 7C. Open circles in the case of EtFc were computed from the data in Fig. 2 (for details, see text)

(2.6B0.4)!10 4 (2.0B0.1)!10 4 (3.9B0.2)!10 3 (1.6B0.1)!10 3 398B4 (1.0B0.02)!10 (8.9B0.3)!10 3 771B21 a

3 a,b

Independent of [Triton X-100] pH 7 Extracted from [48]

185

HRP-I c RFc HRP-II c RFc

k

2

* HRP-II c RFc
3

c

(4) (5)

k

* HRP c RFc

c

that the measured rate constants k refer to the ratelimiting interaction of the peroxidase compound II with RFc. In fact, applying the steady-state approximation to HRP-I and HRP-II, with k1 ;10 7 M 1 s 1 [28], and taking into account that k2 is usually higher by a factor of 50100 than k3 [68], it can be demonstrated that HRPII should be a dominant species under the steady-state conditions. This was also confirmed by the UV-visible measurements which showed that HRP-II is the main form of the enzyme in the steady state, consistent with its reduction being rate-limiting. A dozen of reversible transitions HRP a HRP-II were induced by H2O2 and nBuFc in the forward and reverse directions, respectively. The stoichiometric amount of the former converted HRP into HRP-II (HRP-I could not be detected by conventional spectroscopy under the conditions used), as shown by the appearence of the diagnostic bands at 418, 527, and 558 nm [29]. The stoichiometric amount of nBuFc restores the spectrum of the ferri state of HRP. For structurally similar ferrocenes 1, the length of R is crucial in determining the reactivity in micellar systems. Ferrocenes 1 display the expected behavior, viz. the observed rate constants k decrease with increase in their formal redox potentials E7b. The plot of log k against E7b (both measured at 0.04 M Triton X-100) for ferrocenes 1be is a straight line with a slope of (2.7 B 0.5)!10 2 mV 1. Naturally, the electron-rich substrates are characterized by higher rate constants k than those typical of electron-poor molecules, cf. with the Hammett r values of 5.75 and 3.78 reported for substituted anilines and phenols, respectively [30]. Structurally different ferrocenes with substantially lower redox potentials brought about by extensive methylation do not react faster. In particular, 2 is not as reactive as would be expected from its E7b of 25 mV vs SCE. It is oxidized more slowly than methylferrocene 1b by a factor of 20 at pH 6.0, 25 7C and [Triton X100] p 0.04 M. Steric factors seem to play an important, probably decisive, role here. Accordingly, nonsubstituted ferrocene, HFc, has the highest reactivity of all the species studied. It is more reactive than methyl- and ethylferrocene (1b, c), although the observed redox potentials of both are lower by 40 and 15 mV, respectively, than that of HFc (1a).

Fig. 5 Surfactant effects on the observed second-order rate constants for HRP-catalyzed oxidation of ethylferrocene by H2O2: [EtFc] p 4.4!10 4 M, [HRP] p 10 7 M, [H2O2] p 2.4!10 4 M, pH 6, 25 7C. Triton X-100 : I and solid line; CTAB: 1m and broken line. The lines are theoretical curves calculated according to Eq. 8

Effect of surfactants As has already been mentioned, the reaction of Eq. 1 is carried out in the presence of surfactants to increase the solubility of the ferrocenes. Surfactant effects on enzymatic reactivity are well documented in the literature [31-33], and therefore the influence of Triton X100, SDS, and CTAB on the observed second-order

rate constants k was studied in detail. Representative results are shown in Fig. 5. As can be seen, there is a gradual decrease in the rate constants of oxidation of EtFc on increasing the surfactant concentration. Obviously, the retardation may originate from either the enzyme inactivation by surfactants, which has been reported in certain instances [34, 35], or the mass-law retardation due to the non-productive binding of ferrocene molecules to micelles, affording unreactive associates. To rule out the former possibility, the oxidation rate of charged ferrocenes, viz. ferrocenecarboxylic acid (3) and the sodium salt of ferrocenesulfonic acid (4), was measured in the Triton X-100 system. Similar behavior was observed for both of the ferrocenes. The data for the oxidation of FcCOOH at pH 7.0, when the acid is deprotonated (pKa p 4.2 [36]), demonstrates that the rate is unaffected by the surfactant (Fig. 6). Remarkably, the reaction of 3 is of the first order in the absence of surfactant as well, showing that the "nonenzymatic" kinetics are not of micellar origin. Since the affinity of negatively charged species 3 and 4 to micelles of neutral and anionic surfactants should be much lower compared to hydrophobic species 1 (in contrast to 1, charged compounds 3 and 4 are soluble in water), it could be concluded that the micellar effect in Fig. 5 is not due to enzyme inactivation. Rather, it is due to the binding between ferrocenes and micelles. The quantitative treatment of data such as that in Fig. 5 in terms the Berezin "pseudo-phase model" of micellar catalysis [22] is based on the assumption that a bimolecular interaction between reactants may occur both in the "micellar" pseudo-phase and the aqueous phase with the corresponding rate constants km and kw, respectively. However, the model was never, to our

186

kp

km P

HRP

PRFc CVck 1cPRFc CV

w

(7)

Data such as that in Fig. 5 indicate that kw`kmPHRPPRFcCV, as suggested by the decrease in k with increasing surfactant concentration. The fitting of the data in Fig. 5 to Eq. 7 demonstrated that the contribution of the micellar term driven by km into the overall rate is insignificant; Eq. 7 thus transforms into Eq. 8, showing that the oxidation of ferrocenes occurs entirely in the aqueous phase. kp kw 1cPRFc CV (8)

Fig. 6 Triton X-100 effect on the second-order rate constant for the HRP-catalyzed oxidation of FcCOOH by H2O2. [FcCOOH] p 3.2!10 3 M, [HRP] p 10 7 M, 4 [H2O2] p 2.4!10 M, pH 7, 25 7C

knowledge, applied to the second-order reactions between enzymes and inorganic or organometallic compounds. In the framework of the present investigation, the observed second-order rate constant k for the interaction between HRP and RFc in the most general case is given by Eq. 6 [22]. kp km PHRP PRFc CVckw (1PCV) {1c(PHRPP1) CV}{1c(PRFcP1) CV} (6)

where PHRP and PRFc are the partition coefficients for HRP and RFc, respectively, between the micellar and aqueous phases (PA p [A]m/[A]w, A p HRP or RFc), C is the total surfactant concentration without cmc (C p [surfactant]tcmc), and V is the molar volume of ~ micelles. Eq. 6 can be simplified, since PHRP ~ 1 and PRFc k1. In fact, the hydrophilic enzyme molecule is expected to be in the aqueous phase, while hydrophobic, water-insoluble ferrocenes 1 seem to have an increased affinity to the micellar pseudo-phase. With these estimates, and taking into account that one operates at relatively low surfactant concentrations (i.e. CV ~ 1), Eq. 6 transforms into Eq. 7. ~

The best fit values of kw and PRFc in the different micellar media are given in Table 2, and, as anticipated, the rate constants kw measured in micelles of CTAB, Triton X-100, and SDS are similar. In accord with the data in Table 1, the reactivity trend noticed above, viz. the longer the alkyl radical on the ferrocene the lower the rate constants, holds for the intrinsic rate constants kw. The rate decrease on going from R p H to R p nBu, a factor of 32, is significant, suggesting steric retardation. The absolute values of kw indicate that ferrocene is a good substrate of HRP. Since ferrocenes follow first-order kinetics, kw should be compared with the kcat/Km ratio, where kcat and Km are the catalytic and the Michaelis constants for a substrate that obeys Michaelis-Menten kinetics. The available ratios, equal to 0.15!10 5, 1.3!10 5, and 34!10 5 M 1 s 1 for iodide, guaiacol, and ABTS, respectively [26], indicate that the "non-enzymatic" reactivity of ferrocene is higher than that of iodide and comparable with that of guaiacol. The PRFc values for a given ferrocene are similar for different surfactants, suggesting that the hydrophobic interactions govern the binding of uncharged ferrocenes with micelles. Curiously, the PRFc values decrease slightly on going from ferrocene to ethylferrocene. In agreement with this, however, is a qualitative observation that short-chain alkylferrocenes are better and more rapidly soluble in micellar aqueous solutions than HFc. To obtain a value of PRFc independent of kinetics, the binding constants K for HFc and nBuFc with CTAB micelles, Eq. 9, were spectrophotometrically measured.

Table 2 Best-fit values of the rate constants kw and partition coefficients P (0.07 M phosphate) Ferrocene/ Peroxidase HFc/HRP EtFc/HRP n-BuFc/HRP n-BuFc/TOP
a

RFc

calculated according to Eq. 8. Conditions: 25 7C, pH 6.0 PRFc

kw (M

P1

s

P1

) Triton X-100 SDS
5 4 3 4

CTAB (1.9 (2.6 (5.8 (7.4 B B B B 0.1) 0.7) 0.4) 0.1) !10 !10 !10 !10
5 4 3 4

CTAB
5 4a 3

Triton X-100 458B 254B 395B 366B 64 33 50 18

SDS

a

(1.4 (2.8 (5.5 (6.9

B B B B

0.1) 0.4) 0.4) 0.1)

!10 !10 !10 !10

(2.3B0.4)!10 (3.7B0.6)!10 (6.4B0.2)!10 P

550B 202B 381B 402B

56 15 46 48

392B90 201B19 274B16

a

At 32 7C

187

M c RFc

K
'

'

{M7RFc}

(9)

Here M denotes a micelle. The K values, which are related to PRFc via K p (PRFc1)!V [22], were elucidated from the dependence of the absorbance of HFc and nBuFc on surfactant concentration in the UV region. The highest spectral changes are observed at 220 nm, and the corresponding absorbance versus [CTAB] plot is represented in Fig. 7. A gradual decrease in absorbance allows one to estimate graphically the equilibrium constant K by plotting DA 1 against C 1, where DA is the difference in absorbance of alkylferrocene in the absence (extrapolated value) and in the presence of CTAB, while C is the total CTAB concentration without cmc. An example of the corresponding linear dependence is shown as inset in Fig. 7. Thus, calculated equilibrium constants K and those obtained from the kinetic experiment using K p (PRFc1)!V on the assumption that V ;0.3 dm 3 mol 1 [22] are presented in Table 3. Close correspondence of the two sets of data supports the mechanistic rationale of the micellar effects. A test for the reaction mechanism Since the oxidation occurs in the aqueous phase, the electrostatic effects imposed by charges of micelle and enzyme should be insignificant. To gain evidence for this, TOP, the isoelectric point of which is 3.3 [37], was used. The isoelectric point of HRP isoenzyme c is 8.6, and therefore at pH 6.0 the molecules of HRP and TOP are positively and negatively charged, respectively. If the reaction occurred in the micellar pseudo-phase, the charge of enzyme could play a significant role, especially in the case of ionic micelles. Therefore, the kinetics of TOP-catalyzed oxidation of butylferrocene by H2O2

Table 3 Values of equilibrium constants K (Eq. 9) in the CTAB medium obtained spectrophotometrically and extracted from kinetic data. Conditions: 25 7C, pH 6.0 (0.07 M phosphate) Ferrocene K (M
P1

) From kinetic data 164B17 114B12

Spectrophotometrically measured HFc n-BuFc 164B14 105B 8

were studied in CTAB and Triton X-100 media. Remarkably, the kinetic behavior of TOP appeared to be very similar to that of HRP, including the first-order kinetics in nBuFc and the profiles of k versus surfactant concentrations. The resulting kinetic data summarized in Table 2 indicate close values of PEtFc, in agreement with the micellar effect mechanism, and appreciably higher values of kw in the case of TOP. The latter can presumably be accounted for in terms of the recently observed phenomenon that the reactivity of TOP compound II is at least not lower than that of compound I [38]. Therefore, one can expect higher observed reactivity for TOP than that of HRP in the steady-state kinetic experiment. Self-exchange k couple rate constant for the HRP/HRP-II

22

The Marcus theory of outer-sphere electron transfer [23] has frequently been applied to redox proteins, but seldom to redox enzymes [3941]. The reason seems to be simple, since it is typical of enzymes to bind their low-molecular mass redox partners prior to the innersphere electron transfer. Thus, enzymes appear to be undesirable electron-transfer counterparts, although their tremendous role in redox catalysis encourages a thorough research in this area. Therefore, every example of the outer-sphere behavior of an enzyme provides a unique possibility of treating its reactivity in terms of the Marcus formalism. In this context, the reaction of Eq. 1 is perfect: (a) there are second-order kinetics, (b) much is known of the ferricenium/ferrocene couple, including the redox potentials and the self-exchange rate constants k11, both in non-aqueous [42] and in aqueous media (1.3!10 3 M 1 s 1 at 25 7C and pH 6.5) [43], and (c) the redox potential for the HRP-II/HRP couple, which is relevant to the rate-limiting step, has been determined [44] and recently remeasured [45]. Thus, the self-exchange rate constant k22 for the HRP-II/HRP couple could be evaluated from the Marcus cross-relation [46]: k12 p (k11 k22 K12 f )
1/2

(10)

Fig. 7 Change in absorbance of nbutylferrocene at 220 nm with increasing CTAB concentration at [nBuFc] p 1.44!10 4 M, pH 6, 25 7C. The inset shows a graph for evaluation of the binding constant K (Eq. 9)

on the assumption that f ;1 [23]. In our case, k12 p kw, i. e. the electron transfer occurs in the aqueous phase, and this value is in Table 2. The reaction driving force

188

DG7 is calculated from E7b p 229 B 2 mV for the HFc c/0 couple found in the micellar media by extrapolation to pure water [47]. The value 210 mV measured for the salt HFc cPF6 without surfactant [48] can also be used, but the adsorptive anodic peak and a large peak separation (130 mV) were observed in the latter case. The calculation gives a k22 of 5.25 M 1 s 1, which is significantly higher than the corresponding estimate based on k12 for the Fe(CN)6 3/Fe(CN)6 4 couple (4.9!10 4 M 1 s 1) [45]. Alternatively, calculation using the previously reported estimate for k22 gives a rate constant for the cross-reaction k12 of 1.8!10 3 M 1 s 1, i.e. too low by two orders of magnitude. It seems reasonable that the ferrocene/ferricenium couple could be more advantageous than the ferri- /ferrocyanide couple, since the work terms should be less important for the uncharged molecule [39], which obeys first-order kinetics. The ferri- /ferrocyanide couple imposes strong charge effects, and this is probably why the values of the self-exchange rate constants are so diverse for related peroxidases. It should be mentioned that the recently reported value of k22 for the Fe IV/Fe III transition in cytochrome c peroxidase is as low as 1.9!10 9 M 1 s 1 [49]. Again, ferrocyanide was used as a redox titrant. General mechanistic considerations It has often been stated in recent publications [5052] that a high-resolution crystal structure of HRP was not available, but, nevertheless, general features of the active-site heme environment were clear from modelbuilding studies [53], NMR data [54], and relevant Xray data for yeast cytochrome c peroxidase [5658]. Moreover, it has very recently been announced that the crystal structure of the recombinant HRP isoenzyme c is also solved [59]. The composition of the active site can be visualized as shown in Fig. 8 [52]. It should also be emphasized that the HRP active site is believed to be polar as opposed to the hydrophobic active site of cytochrome P450 [5658].

In our previous mechanistic study on the oxidation of reduced glucose oxidase by ferricenium ions, we approximated these organometallic substrates by a sphere or ball ca. 5.6 е in diameter [60]. This model seems to be applicable to peroxidase catalysis as well. Thus, there is a hydrophobic, ball-shaped substrate which displays unusual non-enzymatic steady-state first-order kinetics but, nevertheless, is highly reactive. These observations can be rationalized in terms of the following mechanism. The bulky, hydrophobic ferrocene substrate does not seem to bind strongly within the hydrophilic active site region; the two above-mentioned features preclude it for steric reasons and the lack of the hydrophilic versus hydrophobic correspondence. The steric effect is manifested in the rate decrease that accompanies the elongation of the alkyl radical on the ferrocene. Consequently, ferrocene is able to approach only a peripheral site of HRP close to the d-meso and C18H3 heme edge which is ca. 811 е from the iron. It is believed that it is a binding site for aromatic substrates such a guaiacol, and, interestingly, its estimated reactivity is very close to that of ferrocene found in this work. In contrast to guaiacol, ferrocene is more hydrophobic, and this diminishes its affinity to the active site. Consequently, the first-order kinetics are typical of the ferrocene family 1. This implies the outer-sphere electron transfer for the substitutionally inert hydrophobic organometallics and, hence, applicability of the Marcus approach in the peroxidase case. Monoalkylated ferrocenes display normal behavior under the steady-state conditions, and the rate constants decrease with increasing observed redox potentials E7b, which increase as the length of R increases. There should probably be a center on the enzyme surface or close to it that provides, on the one hand, an affinity trap for hydrophobic ferrocene molecules and, on the other, a channel for the electron transfer. Such a channel must exist, because in its absence the HRP enzyme would be unable to exchange electrons directly with an electrode. This phenomenon is now well accepted [61].

Experimental
Reagents Horseradish peroxidase (R/Z p 3) was obtained from Dia-M and used as received. Ferrocene and nbutylferrocene were Aldrich reagents. Ethylferrocene and FcCOOH were purchased from Strem Chemicals and Fluka, respectively. nPentylferrocene was prepared as described elsewhere [48]. MeFc, Me6Fc, and FeSO3H were kindly provided by Dr. M. D. Reshetova. The corresponding ferricenium salts RFc cPF6 were isolated as solids in the case of R p H, Me, Et, nBu and nC5H11 as described, and the corresponding extinction coefficients (lmax, nm) were found to be 260(617), 296(623), 310(622), 242(619), and 128 (623) M 1 cm 1, respectively [60]. The extinction coefficients in the case of all other substituted ferrocenes were obtained by titration against H2O2 in the presence of HRP as recently described [2]. The calculated values of (lmax, nm) are 431(631), 136(631), and 253(745) M 1 cm 1 for FcCOOH, FeSO3H, and Me6Fc, respectively. Surfactants CTAB, SDS, and Triton X-100 were obtained from Merck, Serva,

Fig. 8 Schematic representation of the HRP active side as appeared in [52]

189 and Sigma, respectively. Hydrogen peroxide, buffer components, and other chemicals used were all Reakhim reagents of the highest purity available. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. Acknowledgements The research described in this publication was made possible in part by financial support from the Russian Foundation for Fundamental Research (Grant No 96-03-34328a) and the Russian State Program "Advanced Methods of Bioengineering". We are grateful to the Alexander von Humboldt Foundation for the donation of a Shimadzu UV-160A spectrophotometer. We thank Dr. A. P. Osipov for discussions, Mrs. Yu N. Firsova for the preparation of the ferricenium salts, and Dr. L. M. Lagrimini (Department of Horticulture, Ohio State University, Columbus, Ohio) for kindly providing the TOP. 37. 38. 39. 40. 41. 42. Ryabov AD (1991) Angew Chem Int Ed. Engl 30 : 931941 Ryabov AD (1995) Ross Khim Zh : 137144 Dunford HB (1982) Adv Inorg Biochem 4 : 4180 Everse J, Everse KE, Grisham MB (eds) (1991) Peroxidases in chemistry and biology, vols I and II. CRC, Boca Raton, Fla. Ortiz de Montellano PR (1991) Annu Rev Pharmacol Toxicol 32 : 89107 Sundaramoorthy M, Kishi K, Gold MH, Poulos TL (1994) J Biol Chem 269 : 3275932757 Patterson WR (1995) Biochemistry 34 : 43314341 Poulos TL, Poulos TL, Patterson WB, Sundaramoorthy M (1995) Biochem Soc Trans 23 : 228232 Schuller DJ, Ban N, Huystee RB van, McPherson A, Poulos TL (1996) Structure 4 : 311322 Dunford HB (1995) Xenobiotica 25 : 725733 Veitch NC, Williams JP (1995) Eur J Biochem 229 : 629640 Veitch NC (1995) Biochem Soc Trans 23 : 232240 Smulevich G (1995) Biochem Soc Trans 23 : 240244 Ortiz de Montellano PR, Ozaki SI, Newmyer SL, Miller VP, Hartman C (1995) Biochem Soc Trans 23 : 228232 Uyama H, Kurioka S, Komatsu I, Kobayashi S (1995) Bull Chem Soc Jpn 68 : 32093214 Olson DL, Scheeline A (1995) J Phys Chem 99 : 12041211 Mabrouk PA (1995) J Am Chem Soc 117 : 21412146 Martinek K, Yatsimirsky AK, Levashov AV, Berezin IV (1977) In: Mittal K L (ed) Micellization, solubilization, and microemulsions, vol. 2. Plenum, New York, pp 489507 Berezin IV, Yatsimirsky AK, Martinek K (1973) Usp Khim 42 : 17291760 Marcus RA (1993) Angew Chem Int Ed Engl 32 : 11111121 Lebedeva OV, Ugarova NN, Berezin IV (1977) Biokhimia 42 : 13721379 Ugarova NN, Lebedeva OV, Kurilina TA, Berezin IV (1977) Biokhimia 42 : 15771584 Ozaki S-i, Ortiz de Montellano PR (1995) J Am Chem Soc 117 : 70567064 Childs RE, Bardsley WG (1975) Biochem J 145 : 93103 Jones P, Dunford HB (1977) J Theor Biol 69 : 457461 Chance B (1952) Arch Biochem Biophys 41 : 404408 Dunford HB, Adeniran AJ (1986) Arch Biochem Biophys 251 : 536542 Tanford C (1980) The hydrophobic effect (2nd edn) Wiley, New York Burton SG, Duncun JR (1995) Biotechnol Lett 17 : 627630 Spreti N, Baroletti A, Diprofio P, Germani P, Savelli G (1995) Biotechnol Progr 11 : 107111 Jones MN, Manley P, Wilkinson A (1982) Biochem J 203 : 285291 Bornemann S, Grout DHG, Dalton H, Hutchinson DW (1994) Biocatalysis 11 : 191221 Pendin AL, Leont'evskaya PK, Lvova EI, Nikolsky BP (1969) Dokl. AN SSSR 189 : 115118 Gazarian IG, Lagrimini LM (1996) Phytochemistry 41 : 10291034 Gazarian IG, Lagrimini LM, Ashby GA, Thorneley RNF (1996) Biochem J 313 : 841847 Marcus RA, Sutin N (1985) Biochim Biophys Acta 811 : 265322 Kulys J, Razumas V (1986) Bioamperometrija. Mokslas, Vilnius, pp 160164 Bowler BE, Raphael AL, Gray HB (1990) Progr Inorg Chem 38 : 259322 and references cited therein Yang SE, Chan MS, Whal AC (1980) J Phys Chem 84 : 30943099 Carney MJ, Lesniak JS, Likar MD, Pladziewicz JR (1984) J Am Chem Soc 106 : 25652569 Hayashi Y, Yamazaki I (1979) J Biol Chem 254 : 91019106 Farhangrazi ZS, Fosset ME, Powers LS, Ellis WR (1995) J Biochemistry 34 : 28662871 Canon RD (1980) Electron transfer reactions. Butterworth, London, pp 205206

Preparation and standardization of solutions Solutions of alkylferrocenes in an aqueous micellar medium were prepared as follows. Ferrocene (0.0232 g, 0.12 mmol) was added to 50 mL of Triton X-100 solution (0.04 M) in 0.07 M phosphate (pH 6) and stirred for 2 h to achieve homogeneity, with [HFc] p 0.0025 M. Solutions of other alkylferrocenes were prepared in a similar way, but 20 min stirring was needed to give [RFc] p 0.008 M. Prolonged stirring was not needed to dissolve FcCOOH and FeSO3H in buffered aqueous solution even in the absence of surfactants. Solutions of H2O2 were standardized by UV spectroscopy using the extinction coefficient of 72.8 M 1 cm 1 at 230 nm [62]. Solutions of HRP were quantified by measuring the absorbance at 403 nm for the native enzyme ( p107!10 3 M 1 cm 1) and at 418 nm ( p95!10 3 M 1 cm 1) [29] for HRP-II after addition of the corresponding amount of H2O2. Good agreement between the results of the two measurements was observed.

Kinetic and equilibrium measurements All kinetic and spectrophotometric measurements were carried out on a Shimadzu UV-160A spectrophotometer equipped with a CPS-240A cell positioner/temperature controller. The reactions were initiated by the addition of 10 mL of the HRP solution (2.05!10 5 M) to the reaction mixture containing 1.0 mL of ferrocene solution (2.5!10 3 M), 30 mL of H2O2 solution (0.013 M), and 970 mL of 0.07 M phosphate buffer in a 1-cm quartz cell. Typical concentration ranges used in this work are as shown in Figs. 4 and 5. The development of absorbance at the wavelength of maximum absorbance for a particular ferricenium cation was monitored until ca. 60% of the total amount of ferrocene was oxidized. Practically linear steady-state portions of kinetic curves were observed up to the 55% conversion, and the corresponding slopes of absorbance vs time plots were taken as steady-state rates. The equilibrium binding of HFc and nBuFc with micelles of CTAB was measured as follows. nButylferrocene (0.0139 g, 0.057 mmol) was added to 60 mL of CTAB solution (0.0142 M) at pH 6 (phosphate buffer) and stirred for 20 min to dissolve nBuFc. The stock solution of [CTAB] p 0.1 M was used for creating the necessary concentration of CTAB. The absorbance of 0.14 mM nBuFc was measured at 220 nm in the CTAB concentration range 2.157 mM. The binding of HFc to the CTAB micelles was measured in the same manner.

References
1. Gorton L (1995) Electroanalysis 7 : 2345 2. Goral VN, Nelen' MI, Ryabov AD (1995) Anal Lett 28 : 21392148 3. Epton R, Hobson ME, Marr G (1978) J Organomet Chem 149 : 231

43. 44. 45. 46.

190 47. Ryabov AD, Amon A, Gorbatova RK, Ryabova ES, Gnedenko BB (1995) J Phys Chem 99 : 1407214077 48. Bazhenova MA, Bogoush SS, Herbst AG, Demeschik TV, Komarovskaya Yu G, Kurova VS, Reshetova MD, Ryabov AD, Ryabova ES, Firsova Yu N (1996) Izv RAN, ser Khim: 25752581 49. Cheung E, English AM (1995) Can J Chem 73 : 11811186 50. Henriksen A, Gajhede M, Baker P, Smith AT, Burke JF (1995) Acta Cryst D 51 : 121123 51. Ortiz de Montellano PR, Ozaki SI, Newmayer SL, Miller VP, Hartmann C (1995) Biochem Soc Trans 23 : 223227 52. Rodriguez-Lopez JN, Smith AT, Thorneley RNF (1996) J Inorg Biol Chem 1 : 136142 53. Smith AT, Du P, Loew GH (1994) In: La Mar GN (ed) Nuclear magnetic resonance of paramagnetic macromolecules. Kluwer, Dordrecht, pp 7593 54. Loew GH, Du P, Smith AT (1995) Biochem Soc Trans 23 : 250256 55. La Mar GN, Hernandez G, De Ropp JS (1992) Biochemistry 31 : 91589168 56. Finzel BC, Poulos TL, Kraut J (1984) J Biol Chem 259 : 1302713036 57. Poulos TL, Finzel BC, Howard AJ (1987) J Mol Biol 195 : 687700 58. Edwards SL, Poulos TL, Kraut J (1984) J Biol Chem 259 : 1298412988 59. Gajhede M, Henriksen A, Schuller DJ, Poulos TL, Smith AT (1996) Plant Peroxidase. Biochemistry and Physiology. IV International Symposium, July 610. Vienna, Austria. Book of Abstracts O-2 60. Ryabov AD, Firsova Yu N, Nelen' MI (1996) Appl Biochem Biotechnol 61 : (in press) 61. Popescu IC, Zetterberg G, Gorton L (1995) Biosensors and Bioelectronics 10 : 443461 and references cited therein 62. George P (1953) Biochem J 54 : 267271