Fructose Dehydrogenase Electron Transfer ... - Wiley Online Library

DOI: 10.1002/celc.201700861 1 2 3 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

Articles

Fructose Dehydrogenase Electron Transfer Pathway in Bioelectrocatalytic Reactions Michal Kizling[a] and Renata Bilewicz*[b] We present kinetic and mechanistic studies of fructose oxidation by fructose dehydrogenase (FDH) using the electrochemical methods of stationary and rotating disk voltammetry. FDH was physically adsorbed on unmodified multiwalled carbon nanotubes (MWCNT) to study direct electron transfer (DET) parameters, and for comparison an MWCNT with an adsorbed pyrene derivative of naphthoquinone employed as the mediator in mediated electron transfer, MET, was also examined. Kinetic parameters, such as the number of electrons transferred, the turnover number, and the electron transfer rate constant, were

calculated. Comparison of the non-turnover and catalytic behaviour revealed the role of the heme c active site in the electron transfer of FDH. It was also shown that a mediator with a sufficiently low formal potential, such as the naphthoquinone, substitutes for the heme c site in the electron transfer to the electrode. The kinetic parameters of the processes proved that the application of the mediator results in an increase in the rate of fructose catalytic oxidation compared to that of the DET oxidation process.

1. Introduction

D-fructose dehydrogenase is an oxidoreductase which catalyses the oxidation of D-fructose to keto-D-fructose. It contains 3 subunits: subunit I (67 kDa), which contains flavin adenine dinucleotide (FAD), where the catalytic process takes place; subunit II (51 kDa) with three heme c moieties acting as the redox prosthetic groups;[11] and subunit III (20 kDa), which has no cofactors, but most probably plays a significant role in maintaining the stability of the molecule.[12] FDH works as the primary dehydrogenase in the respiratory chain, oxidizes fructose, and transfers the electrons to ubiquinone embedded in the membrane. FDH displays strict substrate specificity to fructose as the electron donor; however, it can transfer electrons to several artificial electron acceptors.[13] It is therefore used as an analytical agent for diagnostic and food analysis.[14] Besides, it is a dioxygen-insensitive enzyme, which makes it very promising for constructing membrane less fructose/ dioxygen enzymatic fuel cells. It has been proved that devices with high power performance and good stability over time can be obtained by using FDH.[15] Previously, it was suggested that the heme c subunit in FDH plays a significant role as an electron transfer site in direct electrochemical communication with the electrode. This hypothesis was based on the observation that the subunit I/II subcomplex (without subunit II) does not show DET-type activity.[16] Also, it was proved that separated subunits can be reconstituted on the electrode surface with full maintenance of enzymatic activity.[17] However, several approaches, such as the use of a mediator[18] or conductive polymer,[19] have been reported to improve the electrocatalytic performance of FDH for mediated electron transfer. On the other hand, high catalytic current density was obtained in the DET-type process, without any mediator, with some porous materials or electrode modified with nanoparticles.[20] Two electrons from the Dfructose molecule are transferred to the FAD catalytic centre of FDH and then successively to the heme c moieties.

One of the most intriguing phenomena studied over the past decades is bioelectrocatalysis, in particular the case of direct electron transfer (DET) between an electrode material and redox-active proteins, protein clusters, such as photosystems[1] or organellas.[2] DET reactions have the benefit that they minimize the loss of thermodynamic efficiency in bioelectrocatalytic reactions. The very first reports on DET with a redox active protein were published as far back as 1977, when Eddowes and Hill[3] and Yeh and Kuwana[4] independently showed that cytochrome c on bipyridyl-modified gold and tindoped indium oxide electrodes, respectively, showed virtually reversible electrochemistry. After that, DET was reported for peroxidases,[5] laccases,[6] and ferredoxins;[7] however, it is worth mentioning that no solid evidence could be provided to support the DET mechanism in the case of the most common enzyme in bioelectrochemistry – native glucose oxidase.[8] Although DET is observed only for a number of redox enzymes, it is one of the most important and interesting subjects in this field. The electrochemistry of redox proteins has been actively studied to establish this electron transfer mechanism due to its envisioned advantages in applications in bioelectrochemical devices, such as biosensors[9] and biological fuel cells.[10]

[a] M. Kizling College of Inter Faculty Individual Studies in Mathematic and Natural Sciences (MISMaP) University of Warsaw Stefana Banacha 2C, 02-097, Warsaw, Poland [b] Prof. R. Bilewicz Faculty of Chemistry University of Warsaw Pasteura 1, 02-094, Warsaw, Poland E-mail: [email protected] Supporting information for this article is available on the WWW under https://doi.org/10.1002/celc.201700861

ChemElectroChem 2018, 5, 166 – 174

Wiley VCH 1801 / 101866

166

2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Freitag, 15.12.2017 [S. 166/174]

1

Articles 1 2 3 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

One of the heme c moieties works as an electron donating site to external electron acceptors in MET-type bioelectrocatalysis (as well as homogeneous enzyme reactions) and to electrodes in DET-type bioelectrocatalysis, although a noncatalytic faradaic signal of FDH could not be identified. The crystal structure of FDH has not yet been solved; however, the size of the FDH molecule was reported to be ca. 7 nm based on atomic force microscopy images.[12] The size of the protein seems to be too large to yield non-catalytic DET signals on a voltammogram; however, a very small, non-catalytic faradaic current density could potentially be amplified by the catalytic turnover of FDH. In the present study, we present a detailed investigation of the reaction of D-fructose dehydrogenase from Gluconobacter sp. adsorbed on different carbon nanotube supports. We describe the role of heme c sites in intrinsic molecular charge transfer. We compared the kinetic and thermodynamic parameters of DET with a mediated system consisting of MWCNT noncovalently modified with a pyrene derivative of naphthoquinone. Ubiquinone (2,3-dimethoxy-5-methyl-6-polyprenyl-1,4benzoquinone) was identified as the physiological electron acceptor of FDH.[21] Therefore, our idea was to utilize a naphthoquinone derivative that could possibly work as an efficient mediating agent due to its relatively fast kinetic and suitable thermodynamic properties, such as a formal potential close to that of FAD. Naphthoquinone immobilized on the electrode surface is a well-known mediator utilized in MET, e. g. in the cases of NAD-dependent glucose dehydrogenase[22] or glucose oxidase.[23] We established the formal potentials of all redox active sites of FDH, and we carefully studied their pH dependence by voltammetry. We measured kinetic parameters: order of the catalysed reaction, kinetic rate constant of electron transfer during catalysis, and the number of electrons exchanged. Based on these data, we were able to assess the role of the heme c group in DET and the advantages gained by introducing the mediating agent and working in the MET regime.

2. Results and Discussion 2.1. Voltammetry of FDH-Modified Electrodes in the Absence of Substrate Figure 1A shows CV of the MWCNT/FDH electrode in Arsaturated pH 5.5 0.1 McIlvaine buffer at 1 mV s1 scan rate. First pair of signals at app. 0.23 V corresponds most probably to a redox process of FAD (peak 1). Wide signal in range from 0.10 to 0.28 correspond to activity of two heme sites (peaks 2 and 3), Figure B clearly shows that observed signal is superposition of two. Finally, at 0.38 V small signal can be distinguished (peak 4), connected probably with activity of buried deeply inside structure third heme site. Significant separation of cathodic and anodic signals (app. 10 mV), especially for first pair of peaks, can be explained with the fact that active site is buried in the apoenzyme structure and its distant from the electrode surface significantly. The peak width at half height (DE 1=2 Þ for FAD was calculated to be 76 mV, which is slightly higher than the theoretical value given by Equation (1): DE 1=2 ¼

90:6 mV n

ð1Þ

However, in case of heme signals, obtained data after deconvolution showed that DE 1=2 was 95.3 and 87.6 mV, respectively (Figure 1B). Both anodic and cathodic peak current densities increased linearly with scan rate in the broad range of scan rates proving that the electrochemical behaviour of FDH has the typical property of surface-controlled electrochemical process, as expected for electroactive species immobilized on the electrode surface. Also, for thin layer voltammetry, integration of CV peaks gives the total amount of charge (Q) passed through the electrode for the reduction or oxidation of electroactive enzyme active site. The surface concentration (G) of FAD and therefore FDH concentration was calculated from Equation (2): Ip ¼

n2 F 2 uAG 4RT

ð2Þ

Where u is a scan rate, A is the surface area and other symbols have their usual meanings. Here, surface concentration

Figure 1. A) Cyclic voltammogram of MWCNT/FDH electrode, 0.1 M McIlvaine buffer pH 5.5; B) observed peaks after background subtraction; numbers refer to observed active sites signals.


Wiley VCH 1801 / 101866

www.chemelectrochem.org

167


Freitag, 15.12.2017 [S. 167/174]

1

Articles 1 2 3 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

of enzyme reached 210 37 pmol cm2, according to FAD and convoluted heme peak charge. To obtain higher resolution and better development of signals relatively close to each other, differential pulse voltammetry (DPV) was utilized for few different electrode system arrangements. Voltammograms of FDH immobilized on MWCNT and MWCN/PNQ at the GCE electrode without fructose are shown in Figure 2A. No redox peaks were observed with bare MWCNT (curve A). However, after FDH adsorption, four welldefined and reversible redox peaks were observed with peak potentials of, respectively, 0.231 V, 0.062 V, 0.054 V, and 0.34 V vs. Ag/AgCl at pH 5.5 (curve B), which clearly correspond to results from CV. To verify the origin of the observed signals suggested above, the experiment was repeated over a wide range of pH. The first peak was shifted with pH, approximately 0.058 V per pH unit. The electrode process of the FAD redox centre is a two electron process coupled to a two proton reaction [Eq (3)]: FDH ðFADÞ þ 2e þ 2Hþ $ FDH ðFADH2 Þ

ð3Þ

and consequently, the pH affects the potential according to Equation (4): E 0 ¼ E 00 þ

C FDHðFADÞ ðC Hþ Þ2 2:3RT log nF C FDHðFADH2 Þ

ð4Þ

where E0’ = is the formal potential, n is the number of electrons involved in the reaction, T is temperature, and R and F are the gas and Faraday constants, respectively. If the CFDH(FAD) and CFDH(FADH2) are kept constant, the above equation can be simplified to [Eqs. (5), (6)]: E 0 ¼ E 00 þ

2:3RT 2 pH nF

ð5Þ

E 0 ¼ constant þ 0:059pH

ð6Þ

The slope of the obtained E0’ proved the participation of two protons and two electrons in the redox reaction; we therefore assumed that the analysed signal was connected with FAD activity. The observed peak was ascribed to bound FAD and not that released from the protein,[24] since the potential of free FAD appeared at more negative values (Figure 2A). Free FAD gave a peak at ca. 0.315 V[25] and the shift of formal potential with pH was close to 0.059 (curve 6, Figure 2B) The three other signals observed did not show a strict dependence on pH, so they are most probably connected with heme c sites. The E vs. pH relationship displayed a region where the proton coupling to heme oxidation/reduction processes was evident. Between pH 4 and 6.5, the slope of the curves was ca. 30 mV per pH unit. For higher basic values, the heme redox potential became pH independent, but above pH 9 some potential shift with pH was observed again (Figure 2C). The two slopes observed for this relationship suggested that the thermodynamics of the redox activity was controlled by the protonation of two separate acid/base groups. The phenomena


Wiley VCH 1801 / 101866


Figure 2. DPV of: MWCNT (A), MWCNT-FDH (B), MWCNT-PNQ-FDH (C), MWCNT-PNQ (D), and MWCNT-FAD (E); step potential, 2 mV; effective scan rate, 10 mV s1; 0.1 M McIlvaine buffer pH 6.5; Middle) Formal potential of all active sites and bottom) heme sites as function of pH: heme3, heme2 (2), heme1 (3), PNQ (4); bound FAD (5), FAD adsorbed at the electrode surface (6); the buffers used in the presence of 100 mM NaNO3 were 0.1 M McIlvaine buffer (pH 3.5–8), 50 mM Tris/HCl (pH 8–9), 50 mM boric acid (pH 9–10), and CAPS (pH 10–12).

can be explained assuming proton exchange in two His residues ligating the ferrous ion [Eq. (7)]:[26]

168


Freitag, 15.12.2017 [S. 168/174]

1

Articles 1 2 3 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

FerricHemeproteinðHisÞ2 þ e þ 2Hþ þ

$ FerrousHemeproteinðHisH Þ2

ð7Þ

The reaction represented at this point is not purely electrochemical, but rather involves proton binding in a PCET mechanism. Bioenergetically significant redox-linked proton exchange and its influence on the formal potential of metalcontaining active sites have been presented and clearly explained for direct or indirect transmembrane proton pumping[27] or functionalized de novo designed proteins.[28] In this case, it cannot be easily assigned to a particular biological function (especially since the enzyme denatures very quickly at high pH). Thus, while proton exchange mediated by protein structure could determine the redox potential in vivo, it is possible that here proton coupling may be merely incidental. It is, however, probably important for FDH functioning as the primary acceptor in the respiratory chain. Since the formal potentials of heme1 and heme2 (number of heme sites were added according to its appearance from lowest to highest potential) were relatively close to each other, it is possible that they were positioned close to each other in the molecule and were able to contact during electron transfer (Table S1). Heme3’s distant formal potential suggests that it was buried deep in the apoenzyme structure and did not contact the other heme sites electronically. Hibino et al. have created an FDH mutant without heme3, and they proved directly that it does not take part in the electron transfer pathway.[29] Curves C and D on Figure 2A present DPV of electrodes covered with the pyrene derivative of naphthoquinone (PNQ) adsorbed on MWCNT with and without the adsorbed enzyme, respectively. The formal potential of NQ is located closely to FAD and heme1, strictly between them. It can therefore be assumed that the transfer of electrons from the FAD active site to the electrode surface will occur through this mediating agent, due to the decreased energetic cost of electron transfer.

2.2. Studies of FDH Electrode Processes in the Presence of Fructose In order to investigate the bioelectrocatalytic activity of FDH towards the oxidation of fructose, linear sweep voltammetry experiments were performed for both films: MWCNT-FDH and MWCNT-PNQ-FDH. Figure 3A and B, respectively, show LSVs of these two systems in Ar-saturated McIlvaine buffer of different pH, ranging from 3.5 to 8, containing 100 mM fructose. For MWCNT-FDH, the shape of the voltammetric curves indicated high resistance, and not a well-developed limiting current plateau. In the case of MWCNT-PNQ-FDH, the voltammograms were different; for all the pH studied a fast response with small attenuation and well-formed limiting current regions could be seen. The current densities were only slightly higher for the electrode with the PNQ mediator, and the highest activity for both systems occurred at a similar pH, close to 5.5. (Figure 3E). The shift of the onset of current with increasing pH for both systems is especially interesting (Figure 3C and D). The shift was ChemElectroChem 2018, 5, 166 – 174

Wiley VCH 1801 / 101866


visibly larger in the case of the MWCNT-PNQ-FDH. In the case of MWCNT-FDH, the current onset potential attained a constant value near pH 6, while for MWCNT-PNQFDH it changed rather regularly over the full range of pH studied. The values obtained were plotted vs pH together with the formal potential of heme1 and PNQ (Figure 4). Importantly, the observed catalytic current onset potential followed the values of the formal potentials of heme1 and PNQ, for the non-mediated and mediated process, respectively. On Figure S1 we showed also plot of heme2 formal potential and catalytic onset to emphasize pronounced distinction between those values in whole pH spectrum. Figure S2 presents first derivative of presented earlier LSV curves, were the maxima of peaks correspond to the inflection points of the catalytic waves. As shown on Figure S2A in whole pH spectrum we observe shift of first peak at pH corresponding to results presented earlier. However, it can be denoted that in all cases we observe superposition of two peaks which imply that formal potential of heme1 indeed control catalytic onset, however heme2 also plays a role in electron transfer process during electrocatalytic reaction. Rotating disc electrodes, RDE, were used to evaluate the electrokinetics of the fructose oxidation reaction by FDH adsorbed on MWCNT supports. The convection in the voltammetry cell allowed faster transport of the reactant from the bulk to the disk electrode. The currents obtained using different disk rotation speeds from 0 to 2000 rpm for representative MWCNTFDH and MWCNT-PNQ-FDH systems are presented in Figures 5A and 5B, respectively. With no rotation, the anodic wave showed fructose oxidation with a steep slope peaking at 0.25 V for MWCNT-PNQ-FDH. The decrease in the disk current indicated mass transport limitation at potentials more positive than 0.25 V. At higher disk rotation speeds, the mass transfer limitation for fructose transport to the electrode surface was minimized, resulting in higher limiting currents. No welldeveloped plateau region was noted for the MWCNT-FDH electrode, probably due to kinetic hindrance, which was eliminated for the other electrode due to the presence of the naphthoquinone moiety. While the rotation speed was increased, the limiting current increased, and the slope of the anodic wave remained close to constant for both systems, indicating stable catalytic activity at all the rotation rates used. The enzymatic electrocatalytic process can be considered as a series of three separate and independent contributions. In the Koutecky-Levich equation [Eq. (8)]: 1 1 1 1 ¼ þ þ Iel ILev Icat IE

ð8Þ

The first term is connected with transport of substrate molecules between the enzyme and bulk solution. This process is addressed by examining the rotation rate dependence of the current, which for a diffusion-controlled reaction is described by the Levich equation [Eq. (9)]:[30] ILev ¼ 0:62nFAcD2=3 u1=6 w1=2

ð9Þ

where n is the number of electrons transferred, F is the Faraday

169


Freitag, 15.12.2017 [S. 169/174]

1

Articles 1 2 3 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

Figure 3. Linear sweep voltammetry of MWCNT-FDH (A) and MWCNT-PNQ-FDH (B) in 100 mM fructose McIlvaine buffer with increasing pH; C), D) Expanded portions of (A) and (B) to see better the shift of onset of the current of both systems with marked pH, respectively; numbers in (C) and (D) correspond to the pH used. E) Activity versus pH plot of MWCNT-FDH (circles) and MWCNT-PNQ-FDH (squares)

constant, A is the electrode surface area, C is the bulk concentration of fructose, D is the diffusion coefficient of fructose (0.661 · 109 m2 s1),[31] u is the kinematic viscosity of the electrolyte (1.793 · 102 cm2 s1),[32] and w is the rotation rate per minute. The second term of eq. 8 describes the intrinsic catalytic properties of the enzyme in its reaction with substrate; it is assumed to be independent of driving force (applied potential) and electrode rotation rate. It can be expressed as the electrochemical form of the Michaelis-Menten expression [Eq. (10)]: Icat ¼

Figure 4. Formal potential values of: heme1 site (filled triangles); PNQ (filled circles); fructose electrooxidation onset for MWCN-FDH (empty triangles); MWCNT-NPQ-FDH (empty circles) vs. pH.


Wiley VCH 1801 / 101866


nFAGk cat C ðC þ K M Þ

ð10Þ

The third term is the contribution due to interfacial electron transfer between the electrode and the primary electron entry or exit active site on the enzyme. In case of FDH, the final exit active sites as shown above are: the heme c or naphthoquinone, for the nonmediated and mediated processes, respec-

170


Freitag, 15.12.2017 [S. 170/174]

1

Articles 1 2 3 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

tively. For the first order reactions, the heterogeneous electron transfer contribution can be described by the Butler-Volmer equation [Eq. (11)] (at potentials distant from Eo’):[33] jIE j ¼ nFAGkE

ð11Þ

The Levich and non-Levich regions of the MWCNT-FDH and MWCNT-PNQ-FDH processes are shown in the plot of Ilim vs. w1/2 (Figure 6B). The number of electrons involved in the catalytic oxidation of fructose was calculated from the Koutecky-Levich equation (16) for a first order reaction [Eq. (13)]:

Where kE is the heterogeneous electron transfer rate constant.

1 1 1 ¼ þ Iel Ikin Bw1=2

The Koutecky-Levich formalism requires the determination of the order of the fructose oxidation reaction, p. Therefore, the current at each potential (Iel) and limiting currents (Ilim) were measured at different rotation rates and inserted into the equation [Eq. (12)]:[34]

Where Iel is the electrode current, Ikin is the kinetic current considering all components that were independent from rotation (i. e. Icat and IE), w is the rotation speed, and B is the Levich slope given by Equation (14):

I logIel ¼ logðIcat þ IE Þ þ plog 1 el Ilim

ð13Þ

B ¼ 0:62nFAcD2=3 u1=6

ð12Þ

where, Icat and IE are the currents from eq. 10 and 11. The values of Ilim and Iel were obtained from the voltammograms recorded for rotation speeds from 25 to 2000 RPM at 0.3 to 0.6 V. No effects from low diffusion control were seen, and the results from the whole range of rotation rates gave a satisfying match to the assumed model. The range of potentials chosen for the calculations ensured that the values were outside the reaction kinetics limited region. The plots of logIel versus log (1-Iel/Ilim) in Figure 6A were linear, with the slope values nearly unity for both systems, showing that the fructose oxidation reaction obeyed first order kinetics. The voltammograms presented in Figure 5 were obtained for wide range of substrate concentration and used to understand the relationship between Ilim and the concentration of fructose as a hydrodynamic function. The relationship can be explained by the presented Koutecky-Levich equation (eq. 8). Due to the difficulty in interpreting the precise value of the limiting current from the experimental data of MWCNT-PNQFDH, the Ilim for 0 RPM was measured at + 0.5 V, corresponding to the current plateau region, and for all other rotation rates it was measured at + 0.6 V.

ð14Þ

To calculate the n value, Koutecky-Levich (Iel1 vs. w1/2) plots for various potentials (from 0.1 to 0.6 for both systems) were plotted for different rotation rates in the kinetic controlled region (Figure 6C). By inserting the slopes values, B, in Eq. 14, the numbers of electrons transferred during the bioelectrochemical oxidation of fructose were calculated to be 1.45 0.10 and 1.92 0.11 for MWCNT-FDH and MWCNT-PNQ-FDH, respectively (Figure 6D). This phenomenon most likely refers to the result from previous section: in the presence of mediator the electron transfer process proceeds through the mediatornaphthoquinone and two electrons are exchanged with FAD. In case of non-mediated system, two heme sites take place in two separate but closely spaced 1e processes. This is in agreement with the results of experiments on stationary electrodes discussed above. Similar behaviour was observed i. e. for fumarate reductase, were intramolecular electron transfer process occurs in different manner: as two separated one electron processes ore collective two electron depending on limiting factor: substrate concentration or enzyme kinetics.[35] Equation (8) predicts the relationship between current, applied potential, substrate concentration, and electrode rotation rate. Combining eq. 8–11 yields a form of the

Figure 5. Rotating disk electrode voltammograms at various disk rotation speeds for A) MWCNT-FDH; B) MWCNT-PNQ-FDH electrodes; McIlvaine buffer pH 5.5 with 3 mM of fructose.


Wiley VCH 1801 / 101866


171


Freitag, 15.12.2017 [S. 171/174]

1

Articles 1 2 3 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

Figure 6. A) Determination of the overall order of the reaction for electrode modified with MWCNT-FDH (squares) and MWCNT-PNQ-FDH (circles), B) Plot showing Levich and non-Levich region of the reaction catalysed by FDH modified electrodes; C) Koutecky-Levich plots for low concentrations of substrates: squares = 50 mM, circles = 0.1 mM, triangles = 0.4 mM. Empty symbols = MWCNT-FDH, filled = MWCNT-PNQ-FDH, D) Number of electrons exchanged in bioelectrocatalytic reaction as a function of electrode potential in region of fructose oxidation

Koutecky-Levich equation that describes the variation of 1/Ilim with w1/2 for different values of concentration, C [Eq. (15)]: ðC þ K M Þ 1 1 1 ¼ þ þ Iel ð0:62nFAcD2=3 u1=6 w1=2 Þ ðnFAGkcat C Þ nFAGk E

ð15Þ

For further analysis we chose specific experimental conditions under which kinetic factor in the equation could be reduced. By: decreasing C and w, and applying sufficiently large electrochemical driving force j E-Eo’ j , we increased the influence of Levich and enzymatic component in eq. 18, therefore kinetics contribution would be negligible [Eq. (16)]: ðC þ K M Þ 1 1 ¼ þ Iel ð0:62nFAcD2=3 u1=6 w1=2 Þ ðnFAGkcat C Þ

ð16Þ

Iel was measured as a function of w1/2 for a range of low fructose concentrations, and the data were next analysed according to eq. 10.[36] The intercepts on the 1/Iel axis then became 1/Icat. In Figure 7, we have plotted the resulting values of Icat vs. C. As shown, obtained data gave a very good fit to the Michaelis-Menten models in eq. 10 by non-linear regression analysis with kcat 580 54.4 and 710 66.9 for MWCNT-FDH and MWCNT-PNQ-FDH, respectively .The larger value for the latter can be understood by assuming that in the process mediated by naphthoquinone, subunit II does not take part in electron transfer; therefore, the total electron transfer resistance is reduced, which can cause a boost in the reaction rate. Obtained KM values were similar which indicates that enzyme – substrate affinity was not affected by presence of mediator.


Wiley VCH 1801 / 101866


Figure 7. Influence of fructose concentration on Icat value for MWCNT-FDH (squares) and MWCNT-PNQ-FDH (circles) from Koutecky-Levich plots showing non-linear regression fit onto the Michaelis-Menten model.

All kinetic parameters are depicted in Table 1. Finally, all the suggested electron transfer mechanism is shown in Scheme 1.

Table 1. Obtained kinetic parameters for both the direct and mediated electron transfer models. Electron transfer mechanism Direct Electron Transfer Mediated Electron Transfer (system contains PNQ)

172

Obtained kinetic parameters KM [mM] Number kcat [s1] of electrons

kcat/KM [M1 s1]

7.1 0.34 7.8 0.41

8.17 104 9.10 104

1.45 0.10 1.92 0.11

580 54.4 710 66.9


Freitag, 15.12.2017 [S. 172/174]

1

Articles 1 2 3 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

pyrenebutyric acid, naphthoquinone (NQ), dichloromethane, nitric acid, and sulfuric acid were all purchased from Sigma-Aldrich and used without further purification.

Scheme 1. Proposed electron transfer pathway in FDH mediatorless process. Electron movement is shown by arrows.

3. Conclusions This paper presents a detailed electrochemical study of the direct and mediated electron transfer mechanism of fructose dehydrogenase during the bioelectrocatalytic oxidation of fructose. The kinetic parameters of the enzymatic reaction, the catalytic rate constants, Michaelis-Menten constant, turnover number, and the statistical number of electrons involved in electrocatalytic process have been determined based on voltammetry experiments with stationary and rotating disk electrodes. The properties of the mediatorless process were compared with the process mediated by a pyrene derivative of naphthoquinone adsorbed on an MWCNT. Comparison of the calculated electron rate constant of the electrocatalytic process at low substrate concentrations and rotation speeds, as well as the data from non-turnover experiments, prove that electron transfer from FAD to electrode surface directly during electrocatalytic process does not occur and presence of intramolecular (subunit II) or external mediator (naphthoquinone moiety) is essential for reaction occurrence. The comparison of nonturnover experiments results with one taken in presence of substrate lead to the conclusion that in the mediatorless process the heme c site with the lowest formal potential is the exit site for the electron transfer pathway. In the mediated process, the mediator transfers the electrons from FAD to the electrode instead of heme1. Thus, the role of heme1 is played by the mediator: in our case, the naphthoquinone derivative. The calculated numbers of electrons exchanged in the electrocatalytic process suggest that in absence of external mediator we observe two separated one electron transfer process from each low potential heme sites rather than collective two electron reaction. When mediator was present classic two electron was observed.

To obtain 5-amino-2,4-naphthoquinone (ANQ), to 250 ml of concentrated H2SO4 at 5 8C and under constant stirring, 20 g of NQ was added, and then 70 mg of NaNO3 in 50 ml H2SO4 was introduced in three portions. The mixture was stirred for 25 min at 20 8C and 5 min at 40 8C, cooled, and poured into crushed ice. The precipitate that formed was separated, washed with water, and filtered in dichloromethane through silica gel. 19.3 g (75.1 %) of chromatographically pure 5-nitro-1,4-naphthoquinone (NNQ) was obtained. 50 g of SnCl2 in 50 ml concentrated HCl was then added to 10 g of NNQ in 175 ml of acetic acid at 70 8C. The mixture was heated for 15 min at 100 8C and cooled. The mixture was diluted with 1.5 L of water, and 75 g of FeCl3.6H2O was introduced with stirring. The mixture was neutralized with Na2CO3 to pH 5 and extracted with dichloromethane. The obtained compound (ANQ) was purified on chromatographic silica gel in dichloromethane with a yield of 6.1 g (71.8 %). To a stirred solution of 1-pyrene-4-butyric acid (2 g) in dry DMF (20 mL), HOBt (1.1246 g) and EDC. HCl (1.5956 g) were added under argon. After 30 min at room temperature, a solution of ANQ (2.16 g) in dry DMF (10 mL) was added slowly with a dropping funnel. The reaction was stirred for 16 h at room temperature under Ar. The mixture was then diluted with dichloromethane (120 mL) and washed with aqueous HCl (1 M, 120 mL), aqueous NaOH (1 M, 120 mL), and brine (120 mL). The organic layer was dried over anhydrous MgSO4 and evaporated under vacuum. Recrystallization from MeOH gave the final compound, N-(5,8dioxo-5,8-dihydronaphthalen-2-yl)-4-(1-pyrenyl)butanamide (PNQ), with a yield of 64.3 %. MWCNT films were prepared from a 2.5 mg mL1 suspension in ethanol kept for 20 min under sonication and finally shaken vigorously. The black suspension (20 mL) was deposited onto a GCE, and the solvent was removed under vacuum. In order to obtain electrodes modified with PNQ, the electrodes were incubated in 10 mM PNQ solution in dichloromethane for 1 h and thoroughly rinsed with deionized water, prior to enzyme adsorption. To obtain enzymatic electrodes, the electrodes were incubated in 10 mg ml1 FDH solution at 4 8C for 24 hours. Finally, the electrodes were rinsed with deionized water to remove any loosely bound species and were soaked in McIlvaine buffer of appropriate pH until used. Glassy carbon electrodes (GCE, Ø = 3 mm) were obtained from BASi (US). Prior to modification, the electrodes were polished with 3 mM aluminum oxide powder, rinsed with copious amounts of water and ethanol, and then sonicated for 10 min in ethanol. The experiments were carried out in a three-electrode arrangement employing an Ag/AgCl (KCl sat.) reference electrode, a platinum foil counter electrode, and the GCE as the working electrode. All cyclic voltammetry experiments were carried out using a CH Instruments potentiostat at 22 2 8C.

Acknowledgements Experimental Section Fructose dehydrogenase from Gluconobacter sp. (activity 153 U mg1) was obtained from Sorachim. Citric acid, disodium hydrogen phosphate, ethanol, fructose, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), 3-(cyclohexylamino)-1-propanesulfonic acid (CAPS), boric acid, tris(hydroxymethyl)aminomethane (Tris), dimethylformamide (DMF), hydroxybenzotriazole (HOBt), 1-


Wiley VCH 1801 / 101866


This project was supported by FP7-People-2013-ITN Grant “Bioenergy, Biofuel Cells: From fundamentals to application of bioelectrochemistry” under grant agreement no 607793. Michal Kizling thanks the Polish Ministry of Sciences and Higher Education for support through the project “Diamond Grant” No. 0154 DIA 2013 4 and the Department of Chemistry through the DSM grant No 120000-501/86-DSM-112 700. The authors wish to

173


Freitag, 15.12.2017 [S. 173/174]

1

Articles 1 2 3 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 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

express many thanks to Prof. Magdalena Skompska (University of Warsaw, Faculty of Chemistry) for help with the RDE experiments.

Conflict of Interest The authors declare no conflict of interest. Keywords: electron transfer · fructose dehydrogenase · fructose oxidation · Koutecky-Levich equation · rotating disk electrode

[1] N. Bukhov, R. Carpentier, Photosynth. Res. 2004, 82, 17–33. [2] L. N. Pelster, S. D. Minteer, ACS Catal. 2016, 6, 4495–4999 [3] M. J. Eddowes, H. A. O. Hill, J. Chem. Soc. Chem. Commun. 1977, 21, 771–772. [4] P. Yeh, T. Kuwama, Chem. Lett. 1977, 10, 1145–1148. [5] F. A. Armstrong, M. Lannon, J. Am. Chem. Soc. 1987, 109, 7211–7212. [6] M. Karaskiewicz, D. Majdecka, A. Wieckowska, J. F. Biernat, J. Rogalski, R. Bilewicz, Electrochim. Acta 2014, 126, 132–138. [7] F. A. Armstrong, G. S. Wilson, Electrochim. Acta 2000, 45, 2623–2645. [8] G. S. Wilson, Biosens. Bioelectron. 2016, 82, 7–8. [9] C. Wettstein, K. Kano, D. Schafer, U. Wollenberger, F. Lisdat, Anal. Chem. 2016, 88, 6382–6389. [10] H. Xia, K. So, Y. Kitazumi, O. Shirai, K. Nishikawa, Y. Higuchi, K. Kano, J. Power Sources, 2016, 335, 105–112. [11] M. Ameyama, E. Shinagawa, K. Matsuhita, O. Adachi, J. Bacteriol. 1981, 145, 814–823. [12] M. Tominaga, C. Shirakihara, I. Taniguchi, J. Electroanal. Chem. 2007, 610, 1–8. [13] M. Tominaga, S. Nomura, I. Taniguchi, Biosens. Bioelectron. 2009, 5, 1184–1188. [14] R. Antiochia, G. Vinci, L. Gorton, Food. Chem. 2013, 140, 742–747. [15] a) M. Kizling, M. Dzwonek, B. Olszewski, P. Bacal, L. Tymecki, A. Wieckowska, K. Stolarczyk, R. Bilewicz, Biosens. Bioelectron. 2017, 95, 1– 7; b) H. Funabashi, K. Murata, S. Tsujimura, Electrochemistry 2015, 83, 372–375. [16] S. Kawai, T. Yakushi, K. Matsuhita, Y. Kitazumi, O. Shirai, K. Kano, Electrochem. Commun. 2014, 38, 28–31.


Wiley VCH 1801 / 101866


[17] S. Kawai, Y. Kitazumi, O. Shirai, K. Kano, Electrochim. Acta 2016, 210, 689–694. [18] K. Damar, D. O. Demirkol, Talanta 2011, 87, 67–73. [19] M. Kizling, S. Draminska, K. Stolarczyk, P. Tammela, Z. Wang, L. Nyholm, R. Bilewicz, Bioelectrochem. 2015, 106, 34–40. [20] a) K. Murata, M. Suzuki, K. Kajiya, N. Nakamura, H. Ohno, Electrochem. Commun. 2009, 11, 668–671; b) S. Tsujimura, A. Nishina, Y. Hamano. , K. Kano, S. Shiraishi, Electrochem. Commun. 2010, 12, 446–449. [21] S. Kawai, M. Goda-Tsutsumi, T. Yakushi, K. Kano, K. Matsushita, Appl. Environ. Microbiol. 2013, 79, 1654–1660. [22] S. Abdellaoui, R. D. Milton, T. Quah, S. D. Minteer, Chem. Commun. 2016, 52, 1147–1150. [23] B. Reuillard, A. Le Goff, C. Agnes, M. Holzinger, A. Zebda, C. Gondran, K Eoluarzaki, S. Cosnier, Phys. Chem. Chem. Phys. 2013, 15, 4892–4896. [24] M. Wooten, S. Karra, M. Zhang, W. Gorski, Anal. Chem. 2014, 86, 752– 757. [25] H. Wei, S. Omanovic, Chemi. Biodivers. 2008, 5, 1622–1639. [26] J. E. Baldwin, G. M. Morris, W. G. Richards, Proc R. Soc. B. 1991, 1312, 43– 51. [27] a) P. R. Rich, A. E. Jeal, S. A. Madgwick, A. J. Moody, Biochim. Biophys. Acta 1990, 1018, 29–40; b) P. L. Dutton, C. C. Moser, V. D. Sled, F. Daldal, T. Ohnishi, Biochim. Biophys. Acta 1998, 1364, 245–257. [28] J. M. Shifman, C. C. Moser, W. A. Kalsbeck, D. F. Bocian, P. L. Dutton, Biochem. 1998, 37, 16815–16827. [29] Y. Hibino, S. Kawai, Y. Kitazumi, O. Shirai, K. Kano, Electrochem. Commun. 2017, 77, 112–115. [30] A. C.F Ribeiro, O. Ortona, S. M. N. Simoes, C. I. A. V. Santos, P. M. R. A. Prazeres, A. J. M. Valente, V. M. M. Lobo, H. D. Burrows, J. Chem. Eng. Data 2006, 51, 1836–1840. [31] V. R. N. Telin, J. Telis-Romero, H. B. Mazzotti, A. L. Gabas, Int. J. Food. Prop., 2007, 10, 185–195. [32] H. V. M. Hamelers, A. Ter Hejine, N. Stein, R. A. Rozendal, C. J. N. Buisman, Bioresource Technol. 2011, 102, 381–387. [33] S. L. Gojkovic, S. Gupta, R. F. Savinell, J. Electroanal. Chem. 1999, 462, 63–72. [34] N. S. Parimi, Y. Umasankar, P. Atanassov, R. P. Ramasamy, ACS Catal. 2012, 2, 38–44. [35] H. A. Heering, J. H. Weiner, F. A. Armstrong, J. Am. Chem. Soc. 1997, 119, 11628–11638. [36] K. E. Alcantara, J. F. Rusling, Electrochem. Commun. 2005, 223–226.

Manuscript received: August 16, 2017 Version of record online: November 3, 2017

174


Freitag, 15.12.2017 [S. 174/174]

1