Electrochemical H 2 O 2 - stat mode as reaction concept to improve the process performance of an unspecific peroxygenase

,


Introduction
The unspecific peroxygenase (UPO) from the basidiomycete fungus Agrocybe aegerita (AaeUPO) (EC 1.11.2.1) was first discovered and documented in 2004 [1].Since then, UPO has attracted a lot of interest due to its ability to selectively introduce oxygen atoms into various organic molecules such as benzene, pyridine and cyclohexane and derivatives thereof [2].UPO catalyzes, among others, epoxidation of alkenes, hydroxylation of alkanes, oxidation of aromatics and N-dealkylations [3].This feature has also drawn the attention of organic chemists, since oxyfunctionalization is one of the most challenging chemical reactions in organic synthesis, especially, the oxyfunctionalization of unactivated C-H bonds [2,4].Until recently, research on cytochrome P450 monooxygenases was mainly in focus for the enzymatic selective introduction of oxygen functionalities [5,6].While P450 monooxygenases are able to incorporate oxygen into organic substrates, these enzymes are relatively unstable, dependent on an expensive cofactor and have low catalytic activity [5][6][7].In comparison, UPOs are fairly stable and require only hydrogen peroxide (H 2 O 2 ), acting simultaneously as the oxygen donor and the electron acceptor [5].
inactivation at an elevated concentration of its co-substrate H 2 O 2 [8].This is one reason why the full technical application of UPOs is still limited [8].There are several established methods already reported to mitigate the inactivating effect of H 2 O 2 .The approaches mainly focus on the adjustment of reaction conditions, especially to keep a constantly low H 2 O 2 concentration.Feeding a diluted H 2 O 2 solution into a reaction medium has been shown to be able to increase the total turnover number (TTN) [9], which is defined as the quotient of moles of the product generated after the enzyme was deactivated and the moles of the used enzyme.However, this approach leads to a volume increase and high local H 2 O 2 concentrations [10].As a result, several in-situ H 2 O 2 generation methods have been investigated.In-situ generation of H 2 O 2 can be accomplished through various approaches, including the utilization of a chemical reductant such as dihydroxyfumaric acid [11], an enzyme e.g., glucose oxidase (GOx) [12], a piezocatalytic method [13], photocatalysis [14] or an electrochemical method [15].
Lately, the application of a specific electrode type called gas diffusion electrode (GDE) has been expanded, particularly in the electrochemical reduction of O 2 to H 2 O 2 [16].The GDE possesses a three-phase boundary consisting of a gas, liquid and solid phase [16].This enables a direct and higher mass transport of O 2 from the atmosphere through the electrode and in contact with the electrolyte [16].Thus, limitations due to low O 2 solubility and diffusivity in the liquid and towards the electrode are avoided [17].The combination of the in-situ generation of H 2 O 2 and the subsequent biocatalytic reaction has been reported.Examples are the oxidation of thioanisole catalyzed by the chloroperoxidase (CPO) from Caldariomyces fumago [10,15], the halogenation of 4-pentenoic acid catalyzed by the vanadium CPO from Curvularia inaequalis [18], and the hydroxylation of ethylbenzene catalyzed by the recombinant AaeUPO (rAaeUPO) [19].The electrochemical in-situ H 2 O 2 generation method does not increase the reaction volume and avoids the formation of by-products (e.g., gluconic acid), which may occur when using diluted H 2 O 2 solution or enzymatic in-situ H 2 O 2 generation with GOx, respectively [10].
Usually, H 2 O 2 is generated in-situ at a constant rate (galvanostatic) [10,15,[18][19][20].However, this approach leads to an accumulation of H 2 O 2 in the medium as the enzyme activity constantly decreases due to H 2 O 2 -dependent enzyme deactivation, the so-called catalase malfunction reaction [8].In turn, accumulation of H 2 O 2 further increases the enzyme deactivation rate.It has been demonstrated that by keeping the H 2 O 2 concentration constant (H 2 O 2 -stat mode), by adjusting the feeding rate of H 2 O 2 to a set H 2 O 2 concentration of 50 µM, the enzyme operational stability could be increased, compared to the continuous addition of H 2 O 2 [9].Moreover, the H 2 O 2 -stat mode was implemented within the bioelectrochemical system, with H 2 O 2 concentration limits set at 0.5 mM and 1.2 mM [21].Nevertheless, due to relative high H 2 O 2 concentration thresholds the enzyme operational stability and the final obtained product concentration were low compared to the galvanostatic mode [21].
In this study, the hydroxylation of 4-ethlybenzoic acid (EBA) catalyzed by rAaeUPO was performed in a GDE system.Two electrogeneration modes were employed to supply the H 2 O 2 in-situ.1) A H 2 O 2stat mode at a concentration limit set between 0.06 mM and 0.28 mM.A custom automation program was developed to regulate the current output of the power supply to the GDE.This program utilized the input from the H 2 O 2 sensor to ensure a constant H 2 O 2 concentration (Fig. 1).2) A galvanostatic mode at a constant current density between 0.8 mA cm − 2 and 6.4 mA cm − 2 , which served as an internal benchmark.The TTN, turnover frequency (TOF) and the productivity were determined and compared.The objective is to find the optimal H 2 O 2 concentration limit under the H 2 O 2 -stat mode, which would enable a high TOF while maintaining a high TTN.

Production of his-tagged rAaeUPO
The inoculum seed of Pichia pastoris (X33), which expresses the recombinant protein rAaeUPO-PaDa-I-C6His was prepared as described in [22], in a 50 mL buffered complex glycerol medium (BMGY) containing 25 μg mL − 1 Zeocin.The main fermentation was conducted in a 1 L DASGIP bioreactor system (Eppendorf, Hamburg, Germany) and performed as stated in [22].Modifications to the fermentation process are described in the following.The glycerol batch phase was started by cultivating the inoculum in a 500 mL basal salt medium containing 40 g L − 1 glycerol.Once the initial glycerol was consumed as indicated by the spike of the dissolved oxygen (DO) signal, the glycerol fed-batch phase was started and maintained for 24 h.Afterwards, the glycerol feed was stopped and the methanol fed-batch phase was started to induce the overexpression of rAaeUPO.The DO content and temperature were set at 30% and 30 • C, respectively.To maintain these values, the stirring rate (400-1200 rpm) and aeration rate (30-60 L h − 1 ≙ ca. 1 vvm) were regulated automatically by the system.A 25% v/v ammonia solution was used to maintain the pH at 5. The feeding profiles of glycerol and methanol in the fed-batch phase were set as stated in [23].The biomass was separated from the fermentation broth via centrifugation (Beckmann J2HS, Beckmann Coulter, California, USA) at 5000 rpm for 2 h at 4 • C. The supernatant was sterile-filtered (0.22 µm, DURAPORE, Merck Millipore, Massachusetts, USA) and concentrated by ultrafiltration (10 kDa molecular weight cut off, Minimate TFF Capsule, Pall, New York, USA).rAaeUPOs were dialyzed and concentrated in 0.1 M potassium phosphate (KP i ) buffer, pH 7.

Determination of enzyme activity and concentration
The enzyme activity was quantified using an ABTS assay.The activity assay was conducted spectrophotometrically (Genesys 180, Thermo Scientific, Massachusetts, USA) at 420 nm for 1 min as technical duplicates.The assay consisted of 750 µL 0.1 M Na 2 HPO 4 / 0.1 M citric acid buffer pH 4.4, 100 µL 3 mM ABTS, 50 µL 40 mM H 2 O 2 and 100 µL sample.The sample was added last as it starts the reaction.Directly after adding the sample, the reaction mixture was mixed by pipetting up and down 5 times using the sample pipette tip.The rAaeUPO activity and concentration were calculated as shown below, using equations described previously in [24].
Where v is the rAaeUPO volumetric activity in U mL − 1 , S is the substrate ABTS concentration in the assay in mM, c rAaeUPO is the rAaeUPO concentration in µM, k m is the Michaelis-Menten parameter (50 µM) [5], k cat is the catalytic rate constant (546 s − 1 ) [5] and df is the dilution factor (10, 5 or 1).

Electrochemical setup
Electrochemical and electroenzymatic experiments were conducted in an undivided reactor.Carbon black GDE (PerOx with PTFE layer, Gaskatel, Kassel, Germany) (A: 12.56 cm 2 , thickness: 250 µm) served as the working electrode and was fixed at the side of the reactor.One side of the GDE faced the liquid phase, while the other side faced the ambient air.A platinum (Pt) wire (Chempur, Karlsruhe, Germany) (99.9%,A: 1.5 cm 2 ) served as the counter electrode.Galvanostatic and dynamic electrical currents were generated by a Keithley 2231a-30-3 DC (Tektronix, Oregon, USA) power supply.Stainless steel crocodile clips were used as connectors.The reactor was equipped with a DULCOTEST PEROX H3 E H 2 O 2 sensor (ProMinent, Heidelberg, Germany), a DULC-OMETER dialog DACb H 2 O 2 sensor module (ProMinent, Heidelberg, Germany) and an NI LabVIEW 2021 SP1 virtual instrumentation program (National Instruments, Texas, USA) (Fig. 1).The H 2 O 2 sensor has a response time of 45 s with a lower and an upper detection limit of 0.006 mM and 0.294 mM, respectively.A constant H 2 O 2 concentration (H 2 O 2 -stat mode) in the medium was maintained by employing an automation program designed in-and controlled by the LabVIEW software (Suppl.Fig. S5).The H 2 O 2 concentration was measured by the H 2 O 2 sensor and the concentration was transmitted to LabVIEW.Lab-VIEW controlled the current output of the power supply and based on the set H 2 O 2 concentration limit, the electrical current sent to the electrode was adjusted to control the H 2 O 2 productivity.For the automation program, the maximum potential, proportional gain and integral time were set to 6 V, 0.01, and 2 min, respectively.The H 2 O 2 concentration limit was set either to 0.06 mM, 0.15 mM, 0.2 mM or 0.28 mM.These values were selected to ensure a relatively balanced distribution across the H 2 O 2 sensor's limit.

Electroenzymatic experiments
The reaction medium contained 200 mL 0.1 M KP i pH 7, 8 mM EBA and 10 nM of rAaeUPO.The medium was stirred at 250 rpm by a magnetic bar (d: 0.5 cm, l: 3 cm).The experiments were conducted at 22 ± 1 • C to minimize thermal deactivation of the enzyme.Galvanostatic experiments were performed at electrical current densities between 0.8 mA cm − 2 and 6.4 mA cm − 2 .In the H 2 O 2 -stat mode, the automation system was engaged.Thus, a dynamic current was applied to the electrodes.Experiments were initiated by either starting the power supply or the automation program.Samples for the quantification of EBA, 4-(1-hydroxyethyl)benzoic acid (HEBA) (20 µL), H 2 O 2 concentration (100-800 µL), and rAaeUPO activity (65-200 µL) were taken periodically from the system.Each experiment was stopped when there was no measurable rAaeUPO activity (slope of the absorbance < 0.01 cm − 1 min − 1 ).Unless otherwise stated, electroenzymatic experiments were performed as duplicates.The TOF refers to the turnover number (TON) per unit time (60 min).The TON is described as the quotient of moles of the product generated at a specific time before the enzyme was deactivated and the moles of the used enzyme.The productivity is defined as the mass of product (derived from the final product concentration) per used reactor volume and time.
The H 2 O 2 productivity was determined in an abiotic environment (without EBA and rAaeUPO) and in galvanostatic mode (0.8 mA cm − 2 -6.4 mA cm − 2 ).The H 2 O 2 concentration was measured periodically over the course of 30 min.Duplicates were performed for each current density.The Faradaic efficiency (F.E.) describes how much energy in form of electrons is consumed for the formation of H 2 O 2 and the formation of side products.The H 2 O 2 F.E. was calculated using the equation given elsewhere [27].

Results and discussion
To determine the optimal H 2 O 2 concentration limit for the rAaeUPOcatalyzed hydroxylation of EBA under the H 2 O 2 -stat mode in the GDE system, several steps were taken.Initially, the electrochemical characterization of the system was conducted to assess the H 2 O 2 productivity.Thereafter, the electroenzymatic hydroxylation of EBA was performed under galvanostatic mode to establish a reference for TOF, TTN, and productivity.Subsequently, electroenzymatic experiments were repeated under H 2 O 2 -stat mode.Finally, the TOF, TTN, and productivity obtained from both H 2 O 2 electrogeneration methods were compared to identify the most efficient approach and eventually the optimal H 2 O 2 concentration limit.

Electroenzymatic hydroxylation of EBA under galvanostatic mode
As a part of the system's electrochemical characterization process in regards to its H 2 O 2 generation capabilities, the H 2 O 2 productivity was determined at various current densities.The electrochemical characterization was conducted in an abiotic environment, without the enzyme and the substrate.
In Fig. 2A, the accumulated H 2 O 2 concentration increased linearly over time for all tested current densities within the 30 min running time.
Based on these results, it can be concluded that there is no indication of O 2 diffusion and mass transfer limitation at the GDE within the tested range.Additionally, as depicted in the Fig. 2B, H ) and the resulting cell potential (y ) show a linear increase with increasing current density.The maximum H 2 O 2 productivity of 5.5 µM min − 1 cm − 2 was achieved at 6.4 mA cm − 2 , the highest tested current density.The highest H 2 O 2 productivity reported here is comparable to those reported in literature for GDE-based systems [18,20,27].It is also observed in Fig. 2B that the F.E. increases from 0.26 at 0.8 mA cm − 2 to 0.50 at 3.2 mA cm − 2 .Upon further increasing the current density, the F.E. shows only minimal improvement and reaches an apparent plateau, with a maximum of 0.55 at 6.4 mA cm − 2 .A similar behavior was reported, where the F.E. increased from 0.60 to 0.78 as the current density was increased from 5 mA cm − 2 to 30 mA cm − 2 [19].A F. E. below 1 means that not all electrons were efficiently used to generate H 2 O 2 , or the resulting concentration of accumulated H 2 O 2 was lower than the theoretical concentration calculated based on the total consumed electrons.Competing reactions such as hydrogen evolution and direct reduction of O 2 to H 2 O are known to reduce the F.E. [25,28].Furthermore, within the electrochemical system the formed H 2 O 2 could be further reduced to H 2 O, oxidized to radicals or decomposed to O 2 and H 2 O, thus reduced the accumulated H 2 O 2 concentration [25,26,28].Surface modification approaches such as thermal oxidation (e.g., using KOH) and coating with carbon nanotubes offer promising ways to enhance the performance of carbon-based electrodes [18,29].These modifications provide a more active surface with O or OH groups and higher current density, respectively [18,29].As a result, H 2 O 2 generation is effectively promoted, leading to an increase in the F.E. [18,29].Additionally, minimizing the contact between the formed H 2 O 2 and counter electrode by placing the counter electrode in a separate compartment is expected also to increase the F.E. of the system.Overall, obtained F.Es. in this study are comparable to the reported values in literature for GDE systems and 3D carbon-based electrodes [10,20,27,30,31].
Following the electrochemical characterization, the electroenzymatic hydroxylation of EBA was performed.The electroenzymatic experiments were conducted initially under the galvanostatic mode by applying various current densities between 0.8 mA cm − 2 and 6.4 mA cm − 2 .The hydroxylation of EBA was catalyzed by rAaeUPO and yielded HEBA as the product.Before starting the experiment, rAaeUPO was added to the reaction mixture.A sample was taken to determine the initial activity using the ABTS assay, which was set as 100% relative activity.Throughout the experiment, enzyme activities were measured relative to the initial activity and expressed as the apparent ABTSactivity due to the coexistence of ABTS and EBA in the sample.Fig. 3A-D show the results of electroenzymatic experiments performed at 0.8 mA cm − 2 , 2.4 mA cm − 2 , 4.0 mA cm − 2 and 5.6 mA cm − 2 , respectively.As the current density is increased, the H 2 O 2 productivity increases correspondingly from 0.37 to 4.6 µM min − 1 cm − 2 .In general, it can be observed that for a period of time the reactions reach an apparent equilibrium in terms of the measured H 2 O 2 concentration, with higher H 2 O 2 concentrations being maintained at increased current densities (Fig. 3E).At the same time, the duration, in which the H 2 O 2 concentration remains constant (termed as apparent equilibrium time) shortens (Fig. 3E).This phenomenon occurred because the relative enzyme activity and the catalytic consumption rate of the H 2 O 2 decreased over the course of the experiment, while the H 2 O 2 productivity remained constant.The apparent equilibrium time was determined using a threshold of 30%, which represents the minimum acceptable deviation from the apparent H 2 O 2 equilibrium concentration.This choice was made considering the generally low H 2 O 2 concentrations observed during the experiment.Opting for a lower threshold, such as 10%, would have resulted in the inability to differentiate deviations from a lower apparent equilibrium H 2 O 2 concentration e.g., 0.13 mM at 2.4 mA cm − 2 , or lower.Consequently, deviations below 30% were considered to be minor fluctuations.At low current density, such as 0.8 mA cm − 2 , the H 2 O 2 generation rate becomes the rate-limiting step of the reaction.As a result, the apparent equilibrium H 2 O 2 concentration is among the lowest compared to other current densities, and the apparent equilibrium time is longer (Fig. 3E) due to higher enzyme stability (71 h).However, at current densities ≥ 2.4 mA cm − 2 , the catalytic consumption rate of H 2 O 2 becomes lower than the H 2 O 2 productivity, making the H 2 O 2 consumption rate the limiting factor of the reaction and leading to a higher apparent equilibrium H 2 O 2 concentration.As more substrate is converted and the enzyme activity gradually decreases, less H 2 O 2 is consumed, resulting in its accumulation in the medium.This accumulation triggers a catalase malfunction reaction, causing even faster enzyme deactivation and resulting in a rapid loss of enzyme activity.Consequently, the apparent equilibrium time decreases with increasing current density (Fig. 3E).In the initial phase of the reaction, the product formation exhibits a linearity for all applied current densities.Nonetheless, the duration of linearity for the product formation differs for each current density.At lower applied current density such as 2.4 mA cm − 2 , the formation rate stays within the linear range for a longer duration (210 min, Fig. 3B), whereas at higher current densities e.g., 5.6 mA cm − 2 , the formation rate deviates from the linear range more quickly (90 min, Fig. 3D) due to higher substrate conversion rate and faster enzyme deactivation.
It is observable in Fig. 4 that the productivity and the TOF are increasing with increasing current density.The highest productivity and TOF obtained under the galvanostatic mode are 10.5 g L − 1 d − 1 and 76.7 s − 1 , respectively.Both are achieved at the highest current density, 6.4 mA cm − 2 .Meanwhile, the TTN reaches its maximum of approximately 650,000 mol mol − 1 at around 2.4 mA cm − 2 and 3.2 mA cm − 2 .An inverse behavior is observed as the current density is increased beyond 3.2 mA cm − 2 .The TTN decreases to 500,000 mol mol − 1 at 6.4 mA cm − 2 .The increasing productivity and TOF could not compensate the faster enzyme deactivation as the current density was increased above 3.2 mA cm − 2 .A faster enzyme deactivation resulted in a reduced final product concentration before all enzyme was deactivated, leading to a decrease in the TTN.A fluctuation in TOF is observed, decreases from 66 s − 1 to 55 s − 1 at 4.8 mA cm − 2 and increases again to 73 s − 1 at 5.6 mA cm − 2 .This observed trend could potentially represent an isolated deviation.Furthermore, other literatures have reported a trend of TOF either remaining stagnant or decreasing with increasing current density, without exhibiting fluctuations [21,26].The maximum TTN obtained under the galvanostatic mode is higher compared to those reported in literatures (400,000 mol mol − 1 ) using a GDE-based system [19,21].A higher TTN obtained here can be explained by a higher enzyme stability due to comparably lower H 2 O 2 productivity.The maximum H 2 O 2 productivity achieved in this study is between 5.8 and 41 times lower [19,21].The relatively small ratio of 0.12 between the counter electrode and the working electrode surface area may restrict the electron flow, potentially diminishing the overall efficiency of the working electrode.This could be an explanation for the observed low H 2 O 2 productivity, especially when considering that other literatures have reported ratios of 0.8 and 1, which could lead to improved performance [19,21].Due to lower H 2 O 2 productivity, the obtained TOF and the productivity are 1.7 and 2.4 times lower, respectively [19,21].

Electroenzymatic hydroxylation of EBA under H 2 O 2 -stat mode
The results from the electroenzymatic hydroxylation of EBA conducted under the galvanostatic mode, discussed in the previous section, served as a reference in this study.Herein, electroenzymatic experiments were conducted once again, this time utilizing the H 2 O 2stat mode at a concentration limit set between 0.06 mM and 0.28 mM, with the intention to increase the enzyme stability and the TTN.
In the H 2 O 2 -stat mode, the H 2 O 2 concentration increases to a predetermined concentration and a steady concentration is maintained throughout the experiment.This is automated via LabVIEW by regulating the electrical current output of the power supply, therefore delivering a dynamic current to the electrodes based on the input from the H 2 O 2 sensor, which measures the H 2 O 2 concentration in the medium.Fig. 5A-D illustrate the results from the hydroxylation of EBA performed under the H 2 O 2 -stat mode with the H 2 O 2 concentration limit set to 0.06 mM, 0.15 mM, 0.2 mM and 0.28 mM, respectively.As soon as the experiment is initiated by engaging the automation system, the power supply increases the current output towards the electrodes to increase the H 2 O 2 productivity and to reach its respective H 2 O 2 concentration limit.
Moreover, in Fig. 6A, the resulting current density, measured H 2 O 2 concentrations, and enzyme relative activity over time obtained from the experiment performed under the H 2 O 2 -stat mode with the H 2 O 2 limit set to 0.15 mM are plotted together.This assessment is performed for the set concentration of 0.15 mM solely for the purpose of exemplifying the automation system and thus, the changes in the current density throughout the experiment, allowing for adjustments of H 2 O 2 productivity.It is apparent from Fig. 6A that the current density is increased to 4 mA cm − 2 within the first 15 min and remains relatively constant up to 60 min.Correspondingly, the H 2 O 2 concentration increases to its limit of 0.15 mM.The measured H 2 O 2 concentration is stable for the whole duration of the experiment.Parallel to the online quantification using the H 2 O 2 sensor, the H 2 O 2 concentrations were also quantified using an offline photometrical method (indicated as H 2 O 2 offline) as a validation of the H 2 O 2 sensor values.In this regard, a maximum deviation of 0.03 mM was observed between the offline and online H 2 O 2 quantification.The observed deviation could be attributed to the use of different calibration systems for each method.The online quantification method, utilizing the H 2 O 2 sensor, employs an internal 2points calibration (set by the manufacturer) with calibration points set at 0 mM and 0.294 mM, which correspond to the theoretical zero value and upper detection limit, respectively.On the other hand, the offline quantification method utilizes a 9-points calibration, with calibration points ranging from 0 mM to 0.1 mM (Suppl.Fig. S3).The observed deviation during the experiment was likely due to reaching the practical lower quantification limit of the online method.This is due to the utilization of a 2-points calibration, which provides fewer reference points.Especially, at lower concentration ranges, resulting in less precise detection of H 2 O 2 .This is reflected in the fact that the highest deviation between the offline and online H 2 O 2 quantification was found in the experiment performed at the H 2 O 2 -stat concentration of 0.06 mM (Fig. 5A).This highlights the importance of performing an offline quantification as a control to an online quantification.After 60 min (Fig. 6A), the current density is steadily decreasing and starts to mimic the declining trend of the relative enzyme activity and the substrate concentration (Fig. 5B).The current output and thus, the current density is reduced to lower the H 2 O 2 productivity since the enzymatic H 2 O 2 consumption is also declining as the enzyme activity decreases.In this way, the amount of generated H 2 O 2 is adjusted to stay equal to the amount of consumed H 2 O 2 keeping the H 2 O 2 concentration constant in the reaction medium.
In general, the final product concentration obtained and the duration of the reaction decrease when the H 2 O 2 -stat concentration limit is increased.By raising the H 2 O 2 concentration limit, the availability of the co-substrate increases, leading to a higher reaction rate (K M, H2O2 : 1.3-1.8mM [5,21,32]).Correspondingly, both TOF and the productivity increase, reaching a maximum of 87.5 s − 1 and 6.9 g L − 1 d − 1 , respectively (Fig. 6B).Additionally, reaching a high TOF and reaction rate at a higher H 2 O 2 -stat concentration limit also increases the possibility of rAaeUPO undergoing catalase and catalase malfunction reactions [8,21].The reason for the aforementioned reactions is that the highly reactive species of rAaeUPO (termed as compound I) formed after binding with the first H 2 O 2 molecule, can react not only with the substrate EBA to yield the product HEBA, but also with a second and third H 2 O 2 molecule [8,21].The reaction of compound I with the second H 2 O 2 molecule yields compound II.Compound II can further react with H 2 O 2 , yielding compound III.The formation of compound III would eventually lead to a heme-bleaching and irreversible enzyme deactivation [8,21].Moreover, the catalase and catalase malfunction reactions become more pronounced at lower substrate concentrations [15].For EBA, a K M of 2.3 mM was reported [21].In this case, reaching EBA concentrations below its K M leads not only to a reduced reaction rate but also prompting the catalase malfunction reaction due to constant availability of H 2 O 2 in the medium, leading to a faster enzyme deactivation with increasing H 2 O 2 -stat concentration limit (Fig. 6C).This also decreases the obtained final product concentration.The final sampling point for the experiment conducted at the H 2 O 2 -stat limit of 0.06 mM (Fig. 5A) was taken after 24 h.By this time, the enzyme had already been deactivated.Therefore, the enzyme operational lifetime was determined based on the point where the current density was reduced and stabilized (by the automation system) at around 0.16 mA cm − 2 .At this current density, the H 2 O 2 productivity had ceased, effectively preventing its accumulation, due to the absence of H 2 O 2 consumption by the enzyme.Regarding the product HEBA, no product inhibition was observed, at least up to 8 mM.
Overall, the highest analytical yield achieved in this study was 95%.The TTN decreases from the maximum of 710,000 mol mol − 1 at a H 2 O 2stat setting of 0.15 mM to 570,000 mol mol − 1 at 0.28 mM (Fig. 6B).Although the highest TTN is obtained at a set concentration of 0.15 mM, the corresponding TOF (58.0 s − 1 ) and productivity (4.6 g L − 1 d − 1 ) are far from the maximum.As the H 2 O 2 concentration limit is increased from 0.15 mM to 0.2 mM, the TOF increases to 80.3 s − 1 and the productivity increases to 6.1 g L − 1 d − 1 .Nevertheless, further increasing the H 2 O 2 -stat concentration from 0.2 mM to 0.28 mM does not significantly increase the TOF and productivity anymore.Therefore, under these circumstances and in this specific system, it is recommended to set the H 2 O 2 -stat concentration to 0.2 mM as this concentration limit allows not only the achievement of comparably high TOF and productivity, but also a competitive TTN (655,000 mol mol − 1 ), compared to other reported TTNs from comparable reaction systems in a lab-scale (Table 1).
Comparing the key performance indicators from the electroenzymatic experiments conducted under the galvanostatic mode and under H 2 O 2 -stat operation, the maximum TOF achieved using both methods are comparable (Fig. 4, Fig. 6B).However, the highest productivity achieved under the galvanostatic method (10.5 g L − 1 d − 1 ) is higher compared to the one obtained under the H 2 O 2 -stat mode (6.9 g L − 1 d − 1 ).A higher productivity under the galvanostatic method can be explained by a higher H 2 O 2 productivity and a higher accumulation of H 2 O 2 in the medium.However, due to a higher and an everincreasing accumulation of H 2 O 2 under the galvanostatic method, leading to a faster enzyme deactivation, the obtained final product concentration and the TTN decrease.In this regard, the maximum TTN acquired under the H 2 O 2 -stat mode is 10% higher compared to the maximum TTN acquired under the galvanostatic method.Under an optimum condition (H 2 O 2 -stat mode: 0.2 mM, galvanostatic mode: 3.2 mA cm − 2 ), the experiment conducted under H 2 O 2 -stat mode still has a higher TTN (655,000 mol mol − 1 ) and TOF (80.3 s − 1 ), as well as a similar productivity (6.1 g L − 1 d − 1 ).In Table 1, the impact of various H 2 O 2 supply methods on the enzyme stability and thus, also on the TTN for H 2 O 2 -dependent enzymatic reactions are listed.The TTN serves as an important metric to assess the suitability of a biocatalyst for a specific process.It also effectively correlates the yield of the product to the input of the catalyst, providing valuable insights in cost valuation of a reaction system.The highest TTN (900,000 mol mol − 1 ) for rAaeUPO-catalyzed hydroxylation reaction was found in a batch system with manual feeding of H 2 O 2 and immobilized enzyme [33].Compared to the highest TTN reported in literature, the highest TTN in this study is around 20% lower.However, the electrogeneration of H 2 O 2 eliminates the need for a second enzyme or volume increase.Additionally, the GDE system offers key advantages, including easy technical set-up and elimination of O 2 mass transfer limitations.The higher TTN reported in the literature previously can be attributed to enhanced enzyme operational stability resulting from the enzyme immobilization.Enzyme immobilization has been recognized as a significant approach to enhance the performance of bioelectrochemical systems, as also indicated in other literature [20,27].This aspect could serve as an optimization point for the system presented in this study.Furthermore, while the current productivity is low, there is potential for future commercial applications with optimization.An optimization of the productivity under the H 2 O 2 -stat mode could potentially be achieved by employing a fed-batch or continuous process  in order to ensure a constant substrate concentration above the K M value and the application of immobilization technique to increase the enzyme stability.

Conclusion
It is clear that the mode of H 2 O 2 electrogeneration impacts the enzyme's operational stability and the overall productivity.The presented results demonstrate that each mode has its own advantages and disadvantages.On the one hand, galvanostatic mode offers a higher productivity at a higher current density but suffers from a faster enzyme deactivation due to a continuously increasing concentration of H 2 O 2 and, therefore, excess of H 2 O 2 .As a result, the final product concentration and the TTN are reduced.On the other hand, operation under H 2 O 2 -stat condition provides the possibility to achieve high TOF and TTN, albeit at a lower productivity.The advantage of the H 2 O 2 -stat mode lies in its ability to adapt to the changes of the H 2 O 2 consumption rate over time, in accordance to the progress of the reaction.Therefore, an excess of H 2 O 2 is prevented protecting the enzyme from rapid deactivation.The key performance indicators such as productivity and TOF obtained under the H 2 O 2 -stat mode are comparable to those reported in literature.Notably, the TTN obtained in this study is higher than all reported values for rAaeUPO-catalyzed reaction in bioelectrochemical systems, to the best of our knowledge.Furthermore, in order to gain a comprehensive understanding of the inactivation mechanism in different H 2 O 2 -dependent enzymes, the H 2 O 2 -stat method introduced here can be applied in the study of these enzymes.

Declaration of Generative AI and AI-assisted technologies in the writing process
During the preparation of this work the author(s) used ChatGPT in order to improve the readability.After using this tool, the author(s) reviewed and edited the content as needed and take(s) full responsibility for the content of the publication.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 1 .
Fig. 1.Schematic representation of the complete electroenzymatic reaction system, including the H 2 O 2 sensor, H 2 O 2 sensor module, power supply and Lab VIEW control unit.The control unit is used to regulate the current output sent to the electrodes and to maintain a constant H 2 O 2 concentration.GDE: gas diffusion electrode (working electrode), Pt: platinum (counter electrode).Dashed line: working cycle of the automation system.Solid line: electric circuit between the power supply and the electrodes.
2 O 2 productivities, resulting cell potentials and the F.Es. are shown as a function of the current density.The H 2 O 2 productivity (y [µM min − 1 cm −

Fig. 3 .
Fig. 3. Hydroxylation of EBA catalyzed by rAaeUPO in a GDE system with in-situ H 2 O 2 generation at A) 0.8 mA cm − 2 , B) at 2.4 mA cm − 2 , C) at 4.0 mA cm − 2 and D) at 5.6 mA cm − 2 .Reaction conditions: 200 mL 0.1 M KP i pH 7, 8 mM EBA, 10 nM rAaeUPO, 250 rpm, temperature: 22 ± 1 • C. EBA: 4-ethylbenzoic acid, HEBA: 4-(1hydroxyethyl)benzoic acid.See Fig. S4.A-D for the full data set.E) Apparent equilibrium H 2 O 2 concentration and apparent equilibrium time as a function of current density.The apparent equilibrium time describes the duration, in which the H 2 O 2 concentration remains relatively constant during the experiment.The apparent equilibrium time is the duration until the H 2 O 2 concentration deviates from the apparent equilibrium H 2 O 2 concentration by 30%.The threshold of 30% was chosen due to overall low H 2 O 2 concentrations during the experiment.Deviations below 30% were interpreted as minor fluctuations.Data shown are average from technical duplicates.

Fig. 4 .
Fig. 4. Corresponding TTN, TOF and productivity as a function of current density.Data shown are average from technical duplicates.

Fig. 6 .
Fig. 6.A) Time-dependent relative enzyme activity, current density and H 2 O 2 concentrations for the hydroxylation of EBA operated under H 2 O 2 -stat mode with the H 2 O 2 limit set to 0.15 mM.H 2 O 2 offline: an offline H 2 O 2 quantification via a photometrical method serves as a control for the online quantification.B) Corresponding TTN, TOF and productivity as a function of the set H 2 O 2 -stat concentration.C) Half-life time (t 1/2 ) and deactivation constant (k deact ) of rAaeUPO at different H 2 O 2stat concentrations.The half-life time of rAaeUPO was determined by dividing the actual enzyme operational lifetime observed during the experiment (experiments shown in Fig. 5. A-D) by two.Deactivation constant was determined from the half-life time (t 1/2 ).k deact = ln (2) t1/2 .Duplicates were performed.

Table 1
Comparison of the impact of different H 2 O 2 supply method on the total turnover number (TTN) of H 2 O 2 -dependent enzymatic reactions.