A novel membraneless -glucan/O2 enzymatic fuel cell based on -glucosidase (RmBgl3B)/pyranose dehydrogenase (AmPDH) co-immobilized onto buckypaper electrode

A novel membraneless


Introduction
Enzymatic fuel cells (EFCs) are bioelectrochemical devices that utilize enzymatic catalysts to transform the chemical energy of common energy-dense bioorganic fuels such as alcohols or carbohydrates to electrical energy through electrochemical reactions [1][2][3][4][5]. EFCs are emerging systems in the field of green energy sources that take advantage of using a wide variety of biological materials as substrates and operate at physiological conditions, which give them great promise in the ongoing power supply of possibly implantable devices [6,7].
The biological generation of biofuels from the cell wall of plants and fungi as sustainable alternative fuels involves a proper method to transform oligo-and polysaccharides into their respective monosaccharide building blocks, which can be subsequently oxidized by oxidoreductase enzymes [8]. There are only a few works on the design and fabrication of enzymatic biofuel cells using polysaccharides as primary fuels, since only a few biocatalysts are recognized to be able to oxidize long-chain oligosaccharides, whereas most of the studies have been devoted to using monosaccharides as biofuel resources [9,10]. Therefore, a hydrolyzing step is a prerequisite to transform polysaccharides into simple sugars and facilitate their employment as biofuels. An enzymatic cascade system composed of two or more enzymes have been used, which can catalyze a sequential reaction [1,2,11,12]. Cellulose and starch are examples of polysaccharides that are composed of glucose units, and as such can be considered as possible alternative energy source due to their availability, easy biotransformation process and nature-friendly property. Recently, Sony [13] employed cellulose, the major component of the primary cell wall of green plants, in the fabrication of a multi-enzymatic fuel cell. Starch has also been utilized as biofuel by three enzyme cascades including α-amylase, glucoamylase and glucose oxidase [14]. Furthermore, Liu et al. [10] and Cosnier et al. [1] assembled glucoamylase and glucose oxidase in the bioanode for the construction of starch/O 2 biofuel cells. In addition, the disaccharide sucrose has been used as the energy resource in a biofuel cell, capable of generating four electrons per one molecule of sucrose by a three enzyme cascade producing a maximum current density of 344 ± 25 μA/cm 2 in a 0.1 M sucrose solution [15]. The variety of oxidizing biocatalysts can use diverse biological materials as fuel and potential energy sources.
Buckypaper is a thin film-like compact multiwalled carbon nanotube structure  μm) with outstanding electrochemical features and with low toxicity both in vivo and in vitro and has been employed successfully for various direct and mediated electrical wiring approaches of enzymes in biosensors, biofuel cells and implantable biodevices [16][17][18]. It possesses unique features such as high electrical conductivity, high surface area and porosity, low interfacial resistance and allows for efficient mass transport [19]. Its 3D-nano-structure network provides high enzyme loading and efficient electrical communication between the enzyme and mediator molecules irrespective of the orientation of the enzyme inside the conducting platform. Hence, its properties improve the rate of the biocatalytic and bioelectrochemical reactions, which result in higher catalytic currents and power density [20][21][22] and seem prospective for the construction of bioelectrodes for enzymatic fuel cells. The flavoenzyme pyranose dehydrogenase (PDH, EC 1.1.99.29), is an extracellular oxidoreductase, which can be purified from the litter degrading fungi Agaricus meleagris, whose FAD cofactor is covalently attached to the polypeptide chain of the enzyme [23][24][25]. AmPDH is able to catalyze the oxidation of a substrate in pyranose form via monooxidation at C-1, C-2, C-3 or dioxidation at C-1,2, C-2,3 and C-3,4 position [26] and it does not show any anomeric specificity. Deeper oxidation of substrate along with a wide substrate specificity and a larger degree of promiscuity in regioselectivity [27] provide an opportunity for using its beneficial features in the design of biofuel cells. AmPDH is a true dehydrogenase, since it does not exhibit any activity with molecular oxygen as electron acceptor [28]. Another advantageous feature of AmPDH is the ability to operate in physiological conditions. These properties make AmPDH an excellent candidate for the construction of the bioanode in membraneless enzymatic fuel cells [9]. Since no toxic H 2 O 2 is being produced from the reduction of oxygen and consequently the efficiency of the bioanode is preserved in the presence of O 2 .
Also, it is noteworthy that direct electron transfer (DET) between the active site of the enzyme and the electrode surface is generally more difficult than mediated electron transfer (MET) [29][30][31][32][33][34], especially for enzymes whose active center is deeply buried in the protein structure The limitation of electron transfer in these enzymes can be solved by using an artificial electronic wire. MET-type enzymatic biofuel cells commonly have higher currents compared to DET-type systems. Os redox polymers constitute one of the most commonly used mediator systems both because they can be "tuned" to encompass a wide voltage range and because their three-dimensional network, fast electrontransfer rate and hydrophilic nature allow the polymeric network to serve as immobilization matrix with a high loading of enzyme molecules [35][36][37], prevent fast denaturation of the enzyme over time and facilitate the electron transfer between the redox center of AmPDH and the conductive support. The used Os redox polymer in this study, [Os(4,4 ′dimethyl-2,2 ′ -bipyridine) 2 poly(N-vinylimidazole) 10 Cl] +2/+ , has a redox potential of 123 mV vs Ag|AgCl (KCl sat ) and thus reveals a high thermodynamic driving force for the electron transfer between the redox center of AmPDH (− 140 mV vs Ag|AgCl 0.1 M KCl, pH 7.4 [26]) and the polymer-bound redox mediator.
The work presented herein demonstrates a bi-catalytic cascade architecture system for the electrochemical oxidation of β-glucan, which is a polysaccharide comprising numerous glucose units with mixed linkages and broadly found in the cell walls of cereals, specified bacteria and fungi. Since no single enzyme is able to directly catalyze the oxidation of β-glucan, a sequential enzymatic process is required breaking down the β-glucan as follows: in the first step the cleavage of the β-1,4 and β-1,3 glycosidic bonds is accomplished by β-glucosidases yielding monomeric D-glucose [38]. Then D-glucose is oxidized by AmPDH to 2-ketoglucose. A bioanode operating through a sequential reaction was fabricated by the co-immobilization of a thermostable β-glucosidase (RmBgl3B) and AmPDH through cross-linking the enzymes and the Os-polymer together with glutaraldehyde on the buckypaper based electrode (Scheme 1) to enable the oxidation of β-glucan. The choice of a thermostable β-glucosidase assures a robust enzyme.
As far as we know, this enzyme cascade and its mediated electrical communication with the buckypaper electrode have not been previously demonstrated for any biofuel cell applications. We believe that this is the first work where β-glucan is employed as the fuel in an EFC. The efficiency of the sequential two-enzyme biocatalytic electrode is evaluated by electrochemical techniques such as cyclic voltammetry and amperometry. Finally, as prove of concept for the conversion of β-glucan to electricity, the RmBgl3B/AmPDH/buckypaper bioanode was connected with the MvBOx/graphite electrode as biocathode (see Scheme 2) to construct a β-glucan/O 2 enzymatic biofuel. Their electrochemical characterization and performance of assembled biofuel cell are described by electromotive force and power density.

Chemicals and reagents
The following chemicals and reagents were utilized to construct the bioanode and the biocathode: glycosylated PDH (EC 1.1.99.29; specific activity 32 U mg − 1 [Fc + assay, 20 • C], protein concentration 25 mg mL − 1 [Bradford assay]) from Agaricus meleagris was expressed in Pichia pastoris as previously reported [39]. Rhodothermus marinus β-glucosidase (EC 3.2.1.21) from the glycoside hydrolase family 3 (GH3) (35 U mg − 1 pNP-β-D-Glc) was cloned, produced in Escherichia coli and purified as specifically described below. Myrothecium verrucaria bilirubin oxidase (MvBOx) (EC 1. 3. 3. 5) was supplied with a concentration of 3.61 mg⋅mL − 1 and a specific activity of 370 U⋅mg − 1 in storage buffer containing Tris buffer and 100 mM Na 2 SO 4 (pH 8.0) and employed for construction of the biocathode. Osmium redox polymer, [Os(4,4 ′dimethyl-2,2 ′ -bipyridine) 2 poly(N-vinylimidazole) 10 Cl] +2/+ , with a formal potential (E • ') of 123 mV vs Ag|AgCl (KCl sat ) was prepared in accordance with a well-known procedure [40] and used as mediator in the bioanode to facilitate the electron transfer from the active site of AmPDH to the buckypaper. Glutaraldehyde (GA, Kanto Chemical Co.Inc. Japan) served as a cross-linker. Buckypaper (15-250 µm in thickness) was purchased from NanoTechLabs, Inc. (Buckeye Composites; Yadkinville, NC, USA). Barley β-glucan with low viscosity (Cat. No P-BGBL) was purchased from Megazyme. A stock solution (1 % w/v) of β-glucan was prepared by adding sufficient amount of β-glucan in 0.1 M PBS and heated to boiling along with vigorous stirring then the solution was cooled and stored at room temperature. D-glucose was purchased from ICN Biomedicals Inc. (Aurora, OH, USA). Buffer solution constituents NaH 2 PO 4 , Na 2 HPO 4 and 100 mM KCl utilized for the making of 100 mM phosphate buffer at pH 7. All solutions were prepared with Milli-Q grade water (Millipore, Bedford, MA, USA). Anaerobic and aerobic conditions were established using N 2 or AGA Gas AB (Sundbyberg, Sweden) that were bubbled through the working solutions.

Cloning, production, purification, and enzyme activity of GH 3 β-glucosidase RmBgl3B
Genomic DNA from the thermophilic bacterium Rhodothermus marinus DSM 4253 (grown in DIFCO™ Marin Broth (BD, New Jersey, USA) at 65 • C), was used as template for PCR amplification of gene number 2069 (R. marinus type strain numbering, encoding RmBgl3B) followed by cloning into the vector pET21b (+) (Novagen, Madison, WI) and transformation to E. coli BL21 (DE3) by heat-shock at 42 • C, as described in Ara et al. [29]. Enzyme production was performed in 0.5 L shake flask cultivations at 37 • C with Luria-Bertani (LB) medium containing 100 µg/ ml ampicillin. After reaching an optical density at 620 nm of 0.5, expression was induced by addition of 1 mM IPTG for pET21b and production was continued for 20 h at 30 • C. Cells were harvested by centrifugation at 5000 × g for 15 min at 4 • C and washed twice with 20 mM phosphate buffer pH 7.
Purification of the enzyme was made by immobilized metal ion affinity chromatography (IMAC) using an Ä KTA prime system (Amersham Biosciences, Uppsala, Sweden). Cell pellets were suspended in binding buffer (20 mM sodium phosphate, 500 mM NaCl, 5 mM imidazole, pH 7.4), and lysed by sonication 5 times for 3 min each time at 60 % amplitude and a cycle of 0.5 using a 14-mm titanium probe (UP400 S; Hielscher Ultrasonic GmbH, Teltow, Germany). Cell debris was removed by centrifugation (14000g, 4 • C, 20 min) and the supernatant was applied to a Histrap FF crude column (GE Healthcare) pre-treated with 0.1 M NiSO4. Bound protein was eluted (20 mM sodium phosphate, 500 mM NaCl, 500 mM imidazole, pH 7.4) and fractions of 1 mL were collected.
The activity of the purified RmBgl3B, was first screened and verified using pNP-β-D-Glc at 60 • C. Then the activity was also confirmed using 0.5 % low viscosity β-glucan from barely (Megazyme) as substrate. The reaction was run at two different temperatures (60 • C and 40 • C) using two different pHs 5.6 and 7.4 as described in Ara et al. [41] and in the biofuel cell measurement section described in Material and methods. The activity was measured using the 3,5-dinitrosalicylic acid (DNS) reagent following the increase in the concentration of reducing sugars with time, as described in Aronsson et al. [42]. The enzyme was stored at 4 • C.

Preparation of biocathode
For construction of the biocathode, graphite rods (Alfa Aesar, counter-flat top, 3.05 mm diameter, 38.1 mm length, 99.99 %, USA) were polished on wet emery paper (Tufbak, Durite, P1200) and then rinsed thoroughly with deionized water, followed by drying in an airflow. Then, an aliquot of 5 μL of MvBOx [43][44][45] was allowed to dry on the top of the bare graphite and left to dry at room temperature. Then the electrode was gently rinsed with 0.1 M phosphate buffer solution at pH 7.

Preparation of bioanode
Buckypaper pieces, which were used as the electrode material was cut into L-shapes and covered with a small piece of aluminum foil at the connecting ends for electrical connection via a crocodile clip to yield electrodes with a geometric area of 0.25 cm 2 when immersed in solution. The buckypaper electrodes were activated by means of cyclic voltammetry with a voltage range from − 0.8 V to 0.4 V vs Ag|AgCl (KCl sat ) and a scan rate of 10 mVs − 1 in 0.1 M PBS. Subsequently, in order to functionalize the buckypaper and improve the dispersion quality with carboxylic groups [19], a constant potentail of 1.5 V was applied for 90 s [18,46]. The activation procedure was optimized based on the results reported in the literature. The modified electrodes were then rinsed with deionized water and left to dry in air. Afterwards, an aliquot of 10 μL of an Os redox polymer solution (5 mg mL − 1 in water) was spread over the modified buckypaper electrode and allowed to dry at room temperature.
The two-enzyme cascade electrode was fabricated as follows: 5 μL of a glutaraldehyde solution (0.5 % (v/v) in distilled water) was drop-cast on the buckypaper electrode and then immediately appropriate amounts of solutions of AmPDH and RmBgl3B were added onto the buckypaper electrode, gently mixed together and uniformly distributed on the surface of the electrode.
The prepared electrode was allowed to dry in a fridge (4 • C) overnight to yield the biocatalytic anode. The thus-prepared electrode was denoted RmBgl3B-AmPDH/BP. The enzymes were assembled using glutaraldehyde as cross-linker, which provides a strong covalent binding with the amino groups of enzyme and the Os polymer and also increases the enzymatic stability. The AmPDH/BP electrode, which was used for construction of the control experiment was fabricated with the same procedure, except that RmBgl3B was not included.

Electrochemical characterization of bioanode and biocathode
To assess the performances of the prepared bioelectrodes, a conventional three-electrode configuration was used consisting of the working electrodes (RmBgl3B-AmPDH/BP or MvBOx/graphite electrode), an Ag|AgCl (KCl sat ) (Sensortechnik, Meinsberg, Germany) and a Pt wire were used as reference and counter electrode, respectively, for the electrochemical measurement. All potentials were reported versus the Ag|AgCl(KCl sat ) reference electrode. A Palmsens Emstat potentiostat (Palmsens, Utrecht, Netherlands) was used to perform all electrochemical experiments. The electrodes were equilibrated in buffer solution for 5 min before each measurement. For the RmBgl3B-AmPDH/BP bioanode cyclic voltammograms (CVs) were obtained by scanning the range from − 0.15 V to +0.5 V at a scan rate of 10 mVs − 1 . Constant potential amperometry was conducted with consecutive additions of substrate at an applied potential (E app ) of +0.4 V (100 mV more positive than the E • ' of the Os redox polymer) to assure that the electrons are transferred from the FAD of AmPDH to the Os polymer and then to the surface of the buckypaper. All electrochemical measurements were performed in an air-saturated phosphate buffer (50 mM, pH 7.4) containing 0.1 M KCl. In the amperometric experiments, initially the background current was allowed to reach steady state and then successive volumes of the substrate were added into the phosphate buffer solution to give different concentrations of substrates while the calibration current responses were recorded. The apparent Michaelis-Menten constant (K app M ) was calculated by direct fitting the calibration data in the Prism 6 program (GraphPad Software Inc., La Jolla, CA, USA). For the biocathode, MvBOx/ graphite, electrochemical characterization was performed by CV between +0.8 to 0 V at a scan rate of 5 mVs − 1 in air equilibrated, O 2 and N 2 saturated PBS (pH 7.4, 50 mM) solutions.

Biofuel cell measurements
The β-glucan/O 2 enzymatic fuel cell was assembled in the one compartment electrochemical cell. The RmBgl3B-AmPDH/BP used as the working electrode and the MvBOx/graphite electrode was used as a combined reference and counter electrode in phosphate buffer (50 mM, pH 7.4) solutions that contained 0.5 % (w/v) β-glucan as the biofuel.
The performance of the β-glucan/O 2 enzymatic biofuel cell was evaluated by linear sweep voltammetry (LSV) in the range from 0 V to the open circuit voltage (OCV) at a scan rate of 1 mVs − 1 . The OCV experiment was utilized to record the electromotive force of the membraneless enzymatic fuel cell in buffer solutions that contained 0.5 % β-glucan as the biofuel. All electrochemical measurements were performed at 40 • C, pH 7.4 and under air-saturation. The current density and power density were calculated with respect to the geometrical surface area (just one side) of the corresponding working electrode.

Activity of RmBgl3B
Prior to construction of the bioanode, the activity of RmBgl3B was analysed using the substrate selected for the biofuel cell: the mixed linkage (β-1,4;1,3) glucan from barley, which verified that RmBgl3B was able to release glucose from the substrate ( Table 1). The measurements were made both at the optimum growth temperature of Rhodothermus marinus (60 • C) and at 40 • C (a temperature used for the biofuel cell). Activity verification was important as a previous study had shown that while the enzyme was active on cellobiose, no activity was observed using the longer substrate cellohexaose, indicating a limitation in the length of the substrate. This was hypothesized to be a result of the very deep active site cleft created by the PA14 domain [41] previously proposed as a carbohydrate binding domain [47]. Here, we confirm activity against a mixed linkage β-glucan, thus showing that the 3D-structure of the substrate is important, instead making the enzyme selective for specific glucans (accepting a longer substrate with mixed linkage, but not with β-1,4 linkages only).

Fabrication and characterization of the two enzymes cascade bioanode
The scheme of the sequential enzymatic bioanode electrode is shown in Scheme. 1. Construction of the bioanode electrode involves the coimmobilization of RmBgl3B and AmPDH within the same electrode on the buckypaper electrode while preserving the enzymatic activity. The construction of the bioanode is designed for an intimate connection between RmBgl3B, AmPDH and the Os polymer achieved by GA as crosslinker so that the free glucose monomers produced by the action of RmBgl3B can be directly oxidized by AmPDH and then the electrons relieved from reduced AmPDH are further transferred to the buckypaper electrode through a mediated electron transfer mechanism assured by the Os polymer.
To investigate whether the produced bioelectrocatalytic current is related to oxidation of β-glucan by AmPDH, a control experiment was initially performed by using an AmPDH/BP electrode without RmBgl3B. Fig. 1A shows the CVs obtained with AmPDH/BP operating in a phosphate buffer solution (50 mM, pH 7.4) free of β-glucan (Fig. 1A, dashed line) and in the presence of β-glucan (Fig. 1A solid lines). A pair of redox peaks is observed in solution free of any substrate, which is attributed to the oxidation/reduction of the Os redox polymer. Comparison of CVs recorded for the AmPDH/BP bioelectrode operating in a solution free of β-glucan and in presence of β-glucan at similar conditions shows no bioelectrocatalytic current. An amperometric experiment with successive additions of β-glucan did not exhibit any bioelectrocatalytic current (Fig. 1B). As shown by CV and amperometry, the AmPDH/BP electrode alone is not able to oxidase the β-glucan. Hence, this finding supports that the presence of RmBgl3B used in the design of the sequential enzymatic bioanode is required for complete bioelectrocatalytic oxidation of β-glucan, and is exhibiting a great potential for employment of the presented bioanode for enzymatic biofuel cell development. In order to prove that AmPDH can be compatible together with a second enzyme, RmBgl3B, within the same bioelectrode construction, the RmBgl3B-AmPDH/BP bioanode was constructed and the performance of the immobilized enzymes were evaluated by the same analysis techniques (i. e. CV and amperometry).
The bioelectrocatalytic activity of the RmBgl3B-AmPDH/BP electrode with respect to D-glucose oxidation was studied by CV in the range from − 0.15 to 0.5 V. A pair of redox peaks is observed at RmBgl3B-AmPDH/BP electrode in phosphate buffer solution free of any substrate ( Fig. 2A, dashed line), which is attributed to the oxidation/reduction of the Os redox polymer as above. The presence of D-glucose as substrate clearly leads to the generation of a bioelectrocatalytic oxidation current at ca. 0.05 V. The reduction peak current at around +0.15 V decreased in the presence of D-glucose, implying that AmPDH catalyzed the oxidation of D-glucose in the presence of RmBgl3B. In further experiment, we calculated the apparent Michaelis-Menten kinetics for the glucose oxidation at the RmBgl3B-AmPDH/BP electrode by amperometry at a fixed applied potential of +400 mV vs Ag|AgCl(KCl sat ), so that the Os redox polymer is entirely present in its oxidized form. Initially, the background current of the bioanode was allowed to reach steady state and then sequential additions of the D-glucose solution were added into the phosphate buffer solution to give different concentrations of Dglucose (arrows show the points of substrate injection), and its current response was recorded (Fig. 3A). The resulting kinetic plot and its nonlinear regression for glucose are illustrated in Fig. 3B. The K app M value from fitting the amperometry data with the Michaelis − Menten equation for glucose was determined to be 5 ± 1.2 mM and the maximum current density under saturated substrate conditions (j max ) was 736 ± 86 μA cm − 2 . The difference in K app M value for the RmBgl3B-AmPDH/BP electrode compared to the single enzyme bioelectrode, AmPDH/ graphite, reported in previous work [27] for glucose oxidation is probably due to differences in enzyme immobilization conditions, type of electrode material and temperature of the measurements. The response of the RmBgl3B-AmPDH/BP electrode was linear towards glucose concentration in the range between 1 and 850 μM with a correlation coefficient (r) of 0.985 (Fig. 3B).
The efficiency of RmBgl3B-AmPDH/BP bioanode is vital to the performances of the β-glucan/O 2 biofuel cell. To investigate this, we compared the electrocatalysis of the bioanode in absence and presence of β-glucan (0.5 %) by CV. As is shown in Fig. 4 without β-glucan, a pair of redox peaks is seen at ca. 0.15 V and 0.2 V, respectively, reflecting the Os redox polymer. After adding β-glucan into the solution and prior to recording the CV, the RmBgl3B-AmPDH/BP electrode was immersed in the solution for 5 min. This time was considered sufficient to hydrolyze some β-glucan to glucose by RmBgl3B. As shown in Fig. 4 the anodic peak current of the CV at 0.15 V increases meanwhile the reduction current decreases indicating a bioelectrocatalytic oxidation occurring at the electrode surface. This indicates that RmBgl3B catalyzed the hydrolysis of β-glucan to D-glucose, which in turn was electrocatalytically oxidized to 2-ketoglucose by AmPDH with the simultaneous reduction of FAD to FADH 2 as schematically described in Scheme 1. Consequently, the two enzymes are compatible together and in the presence of each other can catalyze the oxidation of β-glucan to 2-ketoglucose. The electrocatalytic current produced in the presence of β-glucan for the RmBgl3B-AmPDH/BP electrode is lower than that in the presence of glucose due to a lower glucose concentration produced in the enzymatic hydrolysis reaction. This is reasonable due to the time required to produce the glucose necessary for electrooxidation at the electrodes. Due to the hydrolysis of β-glucan to glucose by the RmBgl3B and subsequent, oxidation of glucose by the AmPDH at two C-2 and C-3 positions, it can be expected that the combination of the two enzymes in a common electrode can lead to the preparation of efficient bioanode for the construction of the enzymatic fuel cell. Another important advantage of the employment of AmPDH is that it does not show any anomeric specificity. Unlike e.g., glucose oxidase and other glucose oxidizing enzymes that oxidise D-glucose at the C1 position, commonly making use of mutarotase in the architecture of enzymatic biofuel cell, the mutarotase is not necessary when using AmPDH. Moreover, AmPDH is active    also for the first oxidation product, 2-ketoglucose, yielding as second oxidation product 2,3-diketoglucose [48]. Thus at least 4 electrons can be obtained from 1 D-glucose molecule. Depending on the source of PDH, even the second oxidation product can be further oxidized thus yielding a maximum of 6 electrons per 1 D-glucose molecule. Use of this beneficial property simplifies the design of the enzymatic fuel cell system.
Since RmBgl3B does not participate in the electrochemical reaction during the hydrolysis of β-glucan, the kinetic behavior of RmBgl3B and maximum electrochemical current density were studied indirectly with respect to the amount of glucose produced. Fig. 5 presents the amperometric response of RmBgl3B-AmPDH/BP electrodes to successive additions of β-glucan at a fixed potential of 400 mV. As can be seen from Fig. 6, the response of the RmBgl3B-AmPDH/BP electrode for additions of β-glucan into the bulk solution was slower than when directly adding D-glucose. Subsequently, we proceeded by calculating the electrontransfer turnover rate and affinity for β-glucan at the RmBgl3B-AmPDH/BP bioelectrode. The resulting kinetic plots are illustrated in Fig. 7.
Comparing the results shown in Fig. 4 with those of Fig. 7 shows that the electrocatalytic current density for β-glucan is lower than for Dglucose under the same conditions. This observation confirms that the rate-limiting step of the overall reaction of β-glucan oxidation is the hydrolysis step, not the enzymatic oxidation reaction, which may be a consequence of the rather low specific activity of RmBgl3B at the conditions used (Table 1). It seems that using a hydrolyzing enzyme with a higher turnover rate would lead to a higher current density for the oxidation of β-glucan at the bioanode. As can be seen from Fig. 7, the intensity of the electrocatalytic current increases with an increase in the concentration of β-glucan and reaches a plateau at a concentration of about 7 μM (0.15 %). The respective calibration curve for β-glucan is depicted in the inset of Fig. 7. The response of the bioelectrode was linear with the concentration of β-glucan in the range between 0.02 and 0.3 μM with a correlation coefficient (r) of 0.9752. The limit of detection (LOD) and sensitivity for β-glucan sensing was calculated to be 0.012 μM (S/N = 3) and 129 AM − 1 cm − 2 . Based on these CV and amperometric results, we propose that the RmBgl3B-AmPDH/BP bioelectrode can be a good candidate for application as bioanode in construction of an enzymatic fuel cell.

Stability test of the RmBgl3B-AmPDH/BP bioelectrode
The operational stability was evaluated by amperometry over time at a certain applied potential viz., the potential corresponding to the voltage of the BFC at its maximal power output.
Three different replicas of the RmBgl3B-AmPDH/BP bioanode were constructed with a similar approach and the operational stability of these bioanodes was evaluated by recording the amperometric response at an applied potential of 0.4 V vs Ag|AgCl (KCl sat ). A continued amperometric measurement for 1 h in the 0.1 M phosphate buffer containing 0.5 % β-glucan (pH 7.4) shows that the electrical output retained almost 85 % of the initial current without a significant decline in the current (Fig. 8). Furthermore, the stability of the prepared electrode was evaluated after five days. Importantly, more than 68 % of the electrode response was maintained even after 5 days storage in a buffer solution at 4 • C, which may, in part, be a consequence of using RmBgl3B, a stable thermophilic enzyme 41 . In addition, the combination of buckypaper with excellent biocompatibility to an enzyme/redox polymer hydrogel   7 film along with GA as the cross-linking agent to covalently bind the enzymes led to immobilization of the enzymes together with the redox polymer without significantly affecting the activity of the enzymes and also preventing enzyme leakage. Therefore, this design renders an improved operational stability for the two-enzyme cascade bioanode.
To calibrate the performance of the RmBgl3B-AmPDH/BP at different temperatures, we measured the oxidation current between 20 • C and 50 • C. The oxidation current using the RmBgl3B-AmPDH/BP electrode at 0.22 V was 66.3 μA/cm 2 at 20 • C, 252.5 μA/cm 2 at 30 • C, 781.2 μA/cm 2 at 40 • C, and 623.4 μA/cm 2 at 50 • C. Increasing the temperature of the solution up to 40 • C enhanced the performance because the activity of RmBgl3B and AmPDH depends on temperature. A further decrease in the catalytic current was observed by raising the temperature to 50 • C, which is likely due to denaturation of AmPDH as RmBgl3B has an optimum for activity at 90 • C in solution [41,42].

Bilirubin oxidase-based biocathode
The performance of the biocathode was assessed by cyclic voltammetry in 0.1 M phosphate buffer (pH 7.4) containing 0.5 % β-glucan. The MvBOx/graphite biocathode has an open circuit voltage (OCV) value of 530 mV in equilibrium with air ( Fig. 9), which is in accordance with the reported value of 480 mV (vs NHE, pH 5.3) [43,44]. As can be seen in Fig. 3, CVs of the biocathode obtained in solutions free of O 2 show no catalytic current, while in the O 2 saturated buffer solution, and somewhat at conditions in equilibrium with air, a pronounced cathodic current started at about 0.52 V. These observations confirm that the MvBOx/graphite electrode has a satisfactory performance for use as biocathode in the enzymatic fuel cell.

Evaluation of a membraneless β-glucan/O 2 biofuel cell
As illustrated above, CV analysis of the RmBgl3B-AmPDH/BP electrode demonstrates that the used two enzymes are adaptable. In addition, amperometric experiments and the obtained result from the Michaelis-Menten model propose that this bi-enzymatic architecture could oxidize the β-glucan (glucose) and recommend it as a favorable candidate in the construction of an enzymatic fuel cell.
To assess the practical performance of the RmBgl3B-AmPDH/BP, the bioanode was coupled to the MvBOx/graphite biocathode and evaluated in a membrane-less β-glucan/O 2 biofuel cell configuration. The two bioelectrodes were immersed in air-saturated phosphate buffer containing 0.5 % β-glucan at 40 • C and equilibrated for 30 min, this time was considered to produce a sufficient glucose concentration for a reasonable oxidation rate at the bioanode.
The OCV of the biofuel cell was recorded versus time and it reached steady state after approximately 30 min (Fig. 10). Since the anodic and cathodic open circuit potentials were − 507 and 63 mV vs Ag|AgCl (KCl sat ), respectively, it is expected that the OCV of the prepared cell in 0.5 % β-glucan was about 570 mV. However, it was 550 mV, which is slightly lower than the expected value due to the resistance on the electrode surface.
Linear sweep voltammograms were recording the catalytic current exactly from above the steady-state OCV value (current = 0) to 0 mV with a scan rate of 0.1 mV/s during the biofuel cell operation and was then utilized to acquire the polarization curve. The power density of the biofuel cell was obtained by multiplying the values of voltage and current density. Fig. 11 represents the current density and calculated power density versus the voltage of the assembled biofuel cell (j-V and P D -V curve) operated at 40 • C. The anodic current generated by the RmBgl3B-AmPDH/BP electrode, corresponding to D-glucose oxidation, was obtained at potentials more positive than 0.1 V vs Ag|AgCl (KCl sat ), whereas the cathodic current for the reduction of oxygen was generated