Towards a novel bioelectrocatalytic platform based on “wiring” of pyrroloquinoline quinone-dependent glucose dehydrogenase with an electrospun conductive polymeric fiber architecture

Electrospinning is known as a fabrication technique for electrode architectures that serve as immobilization matrices for biomolecules. The current work demonstrates a novel approach to construct a conductive polymeric platform, capable not only of immobilization, but also of electrical connection of the biomolecule with the electrode. It is produced upon electrospinning from mixtures of three different highly conductive sulfonated polyanilines and polyacrylonitrile on ITO electrodes. The resulting fiber mats are with a well-retained conductivity. After coupling the enzyme pyrroloquinoline quinone-dependent glucose dehydrogenase (PQQ-GDH) to polymeric structures and addition of the substrate glucose an efficient bioelectrocatalysis is demonstrated. Depending on the choice of the sulfonated polyanilline mediatorless bioelectrocatalysis starts at low potentials; no large overpotential is needed to drive the reaction. Thus, the electrospun conductive immobilization matrix acts here as a transducing element, representing a promising strategy to use 3D polymeric scaffolds as wiring agents for active enzymes. In addition, the mild and well reproducible fabrication process and the active role of the polymer film in withdrawing electrons from the reduced PQQ-GDH lead to a system with high stability. This could provide access to a larger group of enzymes for bioelectrochemical applications including biosensors and biofuel cells.


Results and Discussion
Sulfonated polyanilines have been chosen for the electrospinning process because of their conductive properties and since the polymers directly interact with the enzyme PQQ-GDH allowing direct electron transfer between the substrate reduced enzyme and the polymer 42 . Three different sulfonated polyanilines (PABSA [poly(3-aminobenzoic acid-co-3-aminobenzenesulfonic acid)], PAPASA [poly(3-aminobenzeneacetic acid-co-3-aminobenzenesulfonic acid)], PANABMSA [poly(3-aminobenzoic acid-co-2-methoxyaniline-5-sulfonic acid-co-aniline)], Fig. 1) have been used. Apart from sulfonic acid groups, the polymers exhibit carboxylic acid groups, which allow a covalent fixation of the enzyme PQQ-GDH. Differences among the polymers occur due to the presence of different substitutions on the aniline ring (namely methoxy-, ethylcarboxy-and carboxy-groups) or the presence of unsubstituted aniline during copolymerisation.
Characterization of the Copolymers. The synthesized polymers have been characterized by means of gel permeation chromatography (GPC), thermogravimetral analysis (TGA), and IR spectroscopy. The results are presented in Table 1, and the IR spectra are shown in Fig. 2. There, the bands in the range of 3000-3500 cm −1 correspond to stretching vibration N-H bonds for all polymers (not shown). The most prominent absorption band of the carbonyl group of carboxylic acids is found between 1704 cm −1 and 1689 cm −1 . Further bands at 1558-1585 cm −1 and 1488-1506 cm −1 correspond to the ring stretching vibration of the quinoid and benzenoid rings. All polymers show a more intense benzenoid band compared to the quinoid band. The C-N stretching mode gives rise to a band near 1448-1451 cm −1 . The band at 1293-1300 cm −1 is related to the C-N stretching vibration of aromatic amines associated with para-linked aniline units 43 44 . The presence of tetra1,2,4,5 substituted benzene rings has been identified by the appearance of characteristic bands at 851 cm −1 Fabrication and characterization of the 3D electrodes. Attempting to electrospin the polyanilines, we have found that the used polyanilines cannot be processed in electrospinning alone, since electrospraying occurs, probably due to relatively low viscosity and molecular weight of these polymers. Thus, mixtures of sulfonated polyanilines and polyacrylonitrile (PAN) have been used for the electrode fabrication. In order to get a 3D network allowing both electrochemical and optical investigations, the electrospun fiber networks have beenprepared on planar indium tin oxide (ITO) electrodes. After preparation they have been initially characterized by light microscopy in order to check the general quality of polymer networks and to evaluate their thickness (see Supporting Information, Figure S1). Subsequently, the polymer electrodes are characterized by means of scanning electron microscopy. Figure 3 presents light microscopy and SEM images of PAN/PAPASA, PAN/PANAMBSA and PAN/ PABSA composites fabricated from a mixed polymer solution in DMSO. Prior to the spinning process all polymer-containing solutions have been ultrasonicated. Once dispersed, they remain stable in DMSO for at least 3 weeks; neither sedimentation nor aggregation of polymers have been observed. A flow rate of 2 μ l/min is chosen for the electrospinning procedure at almost no dripping. Depending on the chosen polyaniline, the polymer fiber mats show a brownish ( Figure S1: a1 -PAN/PAPASA and c1 -PAN/PABSA) or green-bluish ( Figure S1: b1 -PAN/ PANABMSA) color, which is in agreement with the different redox states (and thus color) of the polyanilines after synthesis. The polymer fiber networks have found to be uniform in their appearance. This is valid for the mats from different polymeric blends and also among multiple electrodes modified with the same polymer mix. Only in some few cases fiber agglomeration could be observed -particularly for fibers made of PAN and PANABMSA, most probably because of the higher viscosity of PANAMBSA compared to other polymers. Figure 4 shows a cut edge of a PAN/PABSA polymer fleece prepared in a 15 min electrospinning process indicating a proper mat thickness in the range of few micrometers. The averaged diameters of the fibers are shown in Table 2. In case of PAN/PANABMSA the mean diameters as well as the standard deviations are slightly larger. Moreover, even when the fibers of all polymer mixtures show a slightly beaded structure, the averaged bead diameters were significantly larger in case of PAN/PANABMSA. As the bead formation is influenced by many parameters, especially concentration and voltage, the relatively high voltage of 25 kV might be the reason for bead formation 45,46 . As for the interaction with water, all mats showed a sponge-like behavior. For instance, when contact angle measurements are conducted, the polymer fiber mats give small angles mostly about 30 °.   in account one can conclude that all three polymer networks provide a suitable interface for electrochemical conversions.

Bioelectrocatalysis of PQQ-GDH at the polymer fiber network.
To evaluate the effect of the electrospun networks as efficient conductive supports for enzyme immobilization, the resulting electrodes have been modified by the enzyme PQQ-GDH. The enzyme has been chosen since different applications of PQQ-GDH with surface structures have been demonstrated as e.g. in sensing, substance release and bioenergetics [48][49][50] . Here, the enzyme has been covalently bound to the carboxylic acid groups on the aniline rings (Fig. 1). The attractive feature of the sulfonated polyaniline is the conductivity at neutral pH. Moreover, we have already reported on an efficient reaction of PQQ-GDH with these kind of polymers in solution but also on the communication with electrodes modified by similar sulfonated polyaniline copolymers 26 . The approach of covalent fixation of the enzyme in this study has been mainly chosen to ensure a good stability of the enzyme electrodes. Figure 6 summarizes cyclic voltammograms of electrospun PAN/PABSA, PAN/PAPASA and PAN/ PANAMBSA electrodes in the absence and in the presence of the substrate glucose. In the absence of glucose no redox conversion of polymers PABSA and PAPASA can be observed, whereas a clear redox couple with a formal potential about E = + 0.14 V vs. Ag/AgCl can be measured for PANAMBSA, being in good agreement with already reported values for sulfonated polyanilines 26,51,52 . The different voltammetric behavior might be attributed to different initial redox states of copolymers after synthesis (PABSA and PAPASA are in fully oxidized pernigraniline state, whereas PANAMBSA is in a half-oxidized emeraldine salt state) and has also been found in previous studies 8,26,53,54 . Nevertheless, in all three cases significant bioelectrocatalytic currents appear upon addition of glucose. When a bare ITO electrode is incubated in a PQQ-GDH solution and subsequently studied in glucose containing buffer no bioelectrocatalysis could be found, demonstrating the role of the electrospun polymer matrix in the enzymatic reaction.
However, the bioelectrocatalytic signals differ both by their starting potentials and the current intensities (Table 4). If a bioelectrocatalytic current after addition of the substrate is observed at potentials corresponding to the redox conversion of the polymer, mediation can be concluded. In case of ITO-PAN/PANAMBSA electrodes this mechanism might be suggested, since the bioelectrocatalytic current appears at E = + 0.1 V vs. Ag/AgCl indicating that the oxidation process of the polymer may trigger the bioelectrocatalysis. However, for ITO-PAN/ PABSA and ITO-PAN/PAPASA bioelectrocatalysis is observed at lower potentials (Table 4), leading to the conclusion that a direct electron transfer between the suitably oriented enzyme and polymer-modified electrode surface is valid. This is in agreement with already reported starting potentials of PQQ-GDH catalysis at electrodes modified with polymers 13    From the data in Table 4 it is obvious, that depending on used polymers, significantly different intensities of catalytic currents can be observed. ITO-PAN/PAPASA electrodes exhibit lowest bioelectrocatalytic activity. In contrast the intensity of the. glucose-induced catalytic current in the case of PAN/PABSA is found to be 10fold higher than that of ITO-PAN/PAPASA, making ITO-PAN/PABSA electrodes to the most promising candidate for efficient enzyme connection to electrodes and furthermore for potential biosensorial utilization.
In order to optimize the performance of PQQ-GDH-modified 3D polymer mats the electrospinning time has been varied from 15 via 30 to 60 minutes (prior to enzyme fixation). The bioelectrochemical results for these electrodes are summarized in Table 5. It can be clearly concluded that the increase in electrospinning time leads to a slight decrease in catalytic activity of the enzyme electrodes, demonstrating that thinner polymer mats are more suitable for the preparation of electroactive electrodes. Although thicker polymer layers can provide more interaction sites for the immobilization of the enzyme the transport of glucose may limit the overall process. The slight decrease in catalytic current may also be attributed to the large distances for electron transport within the rather thick polymeric mats. Here one has to take into account that a large portion of the fibers is composed of the non-conducting PAN polymer. Thus, the efficient bioelectrocatalysis obtained with these fibers from polymeric blends can be considered a real success. Figure 7 demonstrates a calibration curve obtained for ITO-PAN/PABSA electrodes (n = 3) by evaluating the oxidation current as a function of the glucose concentration at E = + 0.35 V vs Ag/AgCl. For glucose concentrations from 0.0025 to 1 mM a linear response is found. From 2 mM to 8 mM a saturation behavior is observed. An apparent K m value of 0.45 mM glucose can be calculated based on the Michaelis-Menten kinetics. The concentration dependent behavior shows that the system allows a well-defined detection of glucose, but it also demonstrates that that at higher glucose concentration the enzyme electrode cannot follow anymore and the electron withdrawal from the reduced enzyme becomes rate limiting.
Considering the blood glucose level in the reference range between 3.6-5.6 mM one can conclude that the present biosensing system has a potential to be implemented to monitoring the glucose level in human blood. It    has also to be explicitly noted here that glucose detection can be performed at low potentials because bioelectrocatalysis already starts at − 0.1 V vs Ag/AgCl. This is advantageous in complex media, in which other redox-active substances are present. However, limited glucose selectivity of PQQ-GDH is still an issue to be overcome with this respect 60 .
Since long-term stability is an important parameter for the evaluation of the performance of any sensing system, we have traced the stability of our ITO-PAN/PABSA-PQQ-GDH electrodes by testing their activity in glucose solution, after the electrodes were kept at 4 °C in 20 mM MES, pH 6 when not in use. Inset in Fig. 7 shows that the catalytic current density response is maintained over 65% of its initial value after 30 days. The signal decline might be caused by loss in PQQ-GDH activity rather than by leakage of the enzyme from the electrode. Although such kind of bioelectrocatalytic platform is mostly desirable for a single-use application, the improved long-term stability makes this type of architecture promising to be considered for the construction of glucose detection systems.

Conclusion
In this report the successful fabrication of conducting 3D polymer fiber networks by electrospinning of 3 different, chemically synthesized sulfonated polyanilines together with non-conducting polyacrylonitrile has been demonstrated. Microscopic characterizations of PAN/PABSA, PAN/PAPASA and PAN/PANAMBSA mats have shown that the fibers are uniformly distributed within the network on the ITO surface. The electrochemical characterization of these electrodes in the presence of hexacyanoferrate (II/III) reveals that the polymeric network allows electrochemical conversions to occur. Compared to a bare ITO electrode higher Faradayic currents and faster electron transfer are found.
Moreover, the polymer fibers provide a surface for fixation of active enzyme as exemplified with PQQ-GDH. Significant bioelectrocatalytic currents occur upon addition of glucose to 3D polymer mats with the covalently-bound enzyme. Depending on the choice of the sulfonated polyanilline mediatorless bioelectrocatalysis can be obtained, starting at quite low potentials; i.e., no large overpotential is needed to drive the reaction. A well-defined concentration dependent response is observed for PAN/PABSA. In addition, the mild and well reproducible fabrication process and the active role of the polymer film in withdrawing electrons from the reduced PQQ-GDH lead to a system with high stability. Thus, such electrospun networks developed by making use of a sulfonated polyaniline and its interaction with PQQ-GDH can be considered as a promising candidate for biosensorial purposes.

Materials.
If not otherwise stated, all reagents and solvents have been used without further pretreatment.  Copolymer Synthesis PAPASA. was prepared by chemical oxidative polymerization according the following procedure: 1 g (0,006 mol) 3-aminobenzeneacetic acid and 1,14 g (0,006 mol) 3-aminobenzenesulfonic acid were dissolved in 27 ml 0,5 M NaOH. 2,97 g (0,013 mol) ammonium persulfate dissolved in 9 ml water added slowly under continuous stirring for more than 30 min at room temperature. The reaction mixture was stirred for further 24 hrs at room-temperature to complete the reaction. The precipitated product was filtered, washed with aqueous 1.2 M HCl and water and dried under vacuum at 80 °C for 24 h.
PABSA. was synthesized as previously reported 26 . Polymer Characterization. The molecular weights were determined by gel permeation chromatography (GPC) using a Waters apparatus provided with refraction and UV detectors and PL Mixed Gel C Column. Measurements were carried out in DMF using polystyrene standards. IR spectra were recorded with Bio-Rad FTX 3000MX spectrometer. 1 H-NMR measurements were carried out with a Bruker Instrument. Thermogravimetral analysis (TGA) was performed by Perkin Elmer TG 7 apparatus.

Preparation of the polymer solution.
For the preparation of the polymer solution 50 mg/ml polyacrylonitrile and 10 mg/ml of a polyaniline derivative were dissolved in DMSO and ultrasonicated for 11 hours.
Electrospinning. Prior to the preparation of the polymer fibers, the rectangular indium tin oxide (ITO) coated glass slides with surface resistivity 15-25 Ω sq −1 (obtained from Sigma-Aldrich, Taufkirchen, Germany) were cleaned processing DMSO-water-acetone-water-ethanol solvent steps (15 min ultrasonication for each step). The cleaned electrodes were then placed into the cabin of the electrospinning apparatus EC-CLI (IME Technologies, The Netherlands), and the polymer solution was electrospun onto their surfaces from a distance of 35 cm at a voltage of 25 kV with 2 μ l/min spinning rate. The temperature was controlled to be 25 °C and the humidity of 40%. For the spinning process a 27G1" needle (B Braun Melsungen, Germany) was used. The electrospinning procedure was conducted for different times from 5 to 60 min.
Characterization of polymer fiber mats. First the coated electrodes were investigated by means of light microscopy (Olympus DSX500, Olympus Deutschland GmbH, Germany) to confirm the overall structure and uniformity. Additionally the polymer fiber mat coated electrodes where investigated with an electron microscope (JEOL JSM-6510 SEM, JEOL Germany GmbH, Germany) at a voltage of 5 kV. Images where taken at different areas of the electrodes and the diameters of the fibers were measured. Contact angle measurements were conducted with a Dataphysics OCA15EC machine (DataPhysics GmbH, Germany).

Preparation of the enzyme solution.
Soluble glucose dehydrogenase (apoGDH) was gained as a kind gift from Roche. The enzyme was dissolved in 5 mM MES buffer in the presence of 1 mM CaCl 2 , and the pH was adjusted to 5. apoGDH was reconstituted with PQQ in a ratio of 1:1 according to Olsthoorn and Duine 57 . For this purpose, sGDH and PQQ were incubated together for 1 h at room temperature in the dark. Aliquots were stored at − 20 °C. Prior to each measurement, the specific activity of the reconstituted enzyme was determined to be 2200 ± 30 U mg −1 by using 2,6-dichlorophenolindophenol as an electron acceptor.
Enzyme binding. The ITO electrodes on which the electrospun polymer network has been fabricated, have been used to covalently bind PQQ-GDH. In order to fix the enzyme covalently, firstly the polymer-modified electrodes were placed in an EDC/NHS-solution (100 mM/25 mM) in 5 mM MES + 1 mM CaCl 2 , pH 6.5 for 15 minutes and after that washed 3 times with the same buffer before enzyme incubation. For the incubation purpose polymer electrodes have been placed in enzyme solution for 1 h, afterwards washed with buffer prior measurements.
Electrochemistry. The electrodes were characterized electrochemically by means of cyclic voltammetry using a CH Instruments 1230b in a custom made 1 ml measurement cell. A polymer-modified ITO electrodes with the effective electrode area of 0.32 cm 2 ( Figure S2, Supporting Information) were used as working electrodes, together with an Ag/AgCl/1 M KCl reference (Biometra, Germany) and a Pt-wire counter electrodes. Electrochemical characterizations were performed in the presence of K 4 [Fe(CN) 6  Bioelectrochemical measurements were performed at room temperature in the same cell using a 20 mM MES buffer pH 6 as electrolyte. The catalytic response was studied with addition of different glucose concentrations to the solution. Cyclic voltammetric experiments were carried out with a μ Autolab Type II device (Metrohm, The Netherlands). The scan rate was set to 5 mV s −1 . The potential range was chosen as between − 0.4 and + 0.4 V vs. Ag/AgCl. Data analysis was performed using GPES software (General Purpose for Electrochemical System, Eco Chemie, The Netherlands).