Electrochemical Monitoring of Heterogeneous Peroxygenase Reactions Unravels LPMO Kinetics

Biological conversion of plant biomass depends on peroxygenases and peroxidases acting on insoluble polysaccharides and lignin. Among these are cellulose- and hemicellulose-degrading lytic polysaccharide monooxygenases (LPMOs), which have revolutionized our concept of biomass degradation. Major obstacles limiting mechanistic and functional understanding of these unique peroxygenases are their complex and insoluble substrates and the hard-to-measure H2O2 consumption, resulting in the lack of suitable kinetic assays. We report a versatile and robust electrochemical method for real-time monitoring and kinetic characterization of LPMOs and other H2O2-dependent interfacial enzymes based on a rotating disc electrode for the sensitive and selective quantitation of H2O2 at biologically relevant concentrations. The H2O2 sensor works in suspensions of insoluble substrates as well as in homogeneous solutions. Our characterization of multiple LPMOs provides unprecedented insights into the substrate specificity, kinetics, and stability of these enzymes. High turnover and total turnover numbers demonstrate that LPMOs are fast and durable biocatalysts.


List of Abbreviations
. Preparation of Prussian blue-modified gold electrodes via electrochemical deposition.(A) The Prussian blue layer was deposited on a gold rotating disc electrode (RDE) in a solution containing 1 mM FeCl 3 , 1 mM potassium ferricyanide (K 3 [Fe(CN)₆]), 100 mM KCl and 100 mM HCl by 12 potential sweep cycles between 600 and 900 mV vs. SHE at a scan rate of 20 mV s -1 .The potential window used to deposit Prussian blue was chosen above the redox potential (380 mV.vs SHE) of the catalyst to facilitate a more gentle deposition process, as reported previously by 38,39 .The formation of a small peak at around 650 mV vs. SHE (indicated by an arrow) shows that deposited Prussian blue is being oxidized, i.e. it shows the onset of the anodic peak displayed in panel B. Note that, the obtained current densities during deposition (compared to activation in panel B) are two orders of magnitude lower because the cycling is performed at a much higher potential compared to the redox potential of the catalyst.(B) Activation of the deposited Prussian blue layer by cycling the RDE 20 times (see the Experimental Procedures section for further explanation) between 160 mV and 590 mV vs. SHE at a scan rate of 50 mV s -1 in a solution containing 100 mM KCl and 100 mM HCl.After activation, the H 2 O 2 sensor was dried with N 2 and coated with a Nafion layer.The H 2 O 2 sensor was stored overnight at ambient atmosphere (25 °C, covered with a glass beaker) and then conditioned in the standard assay buffer under the same conditions used during the activation procedure (scan rate 50 mV s -1 , 50 cycles; red voltammograms).

Figure S2. Performance characteristics of the H 2 O 2 Sensor. (A)
Current response (blue) and response time (>95% signal transition completed after H 2 O 2 addition, gray) are plotted against the angular velocity of the electrode (s -1 ).The experiment was performed in the presence of 4 g L -1 xyloglucan dissolved in 30 mM sodium acetate buffer, pH 6.0 containing 100 mM KCl.The data points of current response and response time were extracted after titrating 20 µM H 2 O 2 into the system and waiting for approximately 100 s to average the current.As the angular velocity increases, the current response increases and the response time of the sensor decreases.All other experiments were performed at an angular velocity of 50 s -1 (3000 rpm) for optimal sensor performance.(B) Comparison of H 2 O 2 sensor response in four different matrices (substrates resulting in a low or high viscosity solution or suspension).The buffer was 30 mM sodium acetate, pH 6.0, containing 100 mM KCl as supporting, inert electrolyte.xyloglucan, PASC, or CNC were added to test matrix effects.H 2 O 2 was injected 5-times in 20 µM steps reaching a final concentration of 100 µM.These experiments show that the fast response time of the sensor is maintained in different matrices.The equilibrium current is reached within 2 s and recorded for 65 s to obtain the calibration function.The calibration raw data also show that the sensitivity of the electrode may depend on the substrate being studied, so each measurement was calibrated to exclude any influence of a changed sensitivity on the measured initial rates.(C) Currents from panel B and currents derived from three additional identical independent measurements were averaged over 20 s and used to generate calibration cures.These curves show minimal differences in sensitivity between buffer, soluble xyloglucan and the insoluble cellulose substrates (PASC and CNC).From the obtained linear calibration functions, the sensitivity and the limit of detection for H 2 O 2 sensors in different matrices was calculated.Sensitivities were calculated from the slope of the calibration function (nA µM -1 cm -2 ) and the limit of quantitation was calculated by dividing 10x the standard deviation of the baseline (nA cm -1 , 0 µM H 2 O 2 ) with the sensitivity of the electrode.Exemplary calibration data from a measurement series with different xyloglucan, PASC and CNC concentrations is given in Table S2.
Figure S3.Substrate-dependent activity (TN, s -1 ) and stability (Res.Act., %), off-pathway activity and futile H 2 O 2 consumption (n) of the studied LPMOs.All measurements were performed using standard assay conditions as defined in the section "Standard assay conditions" in Materials and Methods and were performed in independent triplicates.For soluble substrates (xyloglucan and glucomannan) the LPMO concentration was 50 nM, for celluloses 200 nM.The rates for the off-pathway activity of LPMOs shown in panel G were determined using 200 nM LPMO.Exemplary raw data are shown in Figures S7A, S11, S13-15.Data obtained from experiments performed under other conditions than in the standard assay are indicated by an asterisk.For these measurements, the initial H 2 O 2 concentration was 50 µM, whereas all other measurements were conducted with 100 µM H 2 O 2 .The turnover numbers and residual activities were calculated as defined in the "Calculation of turnover numbers and residual activities" in the materials and methods section.In short, the observed initial initial rate (v obs ) was corrected with the substrate-dependent background (v sb ) and the rate for the peroxidase-like off-pathway activity (v off-path ).The peroxidase-like activity was only corrected for the v sb measured in the buffer.For the determination of residual activities, once the initial amount of H 2 O 2 added to the reaction was fully consumed, indicated by the H 2 O 2 time trace reaching the baseline (0 µM H 2 O 2 ), 100 µM H 2 O 2 was added a second time.This second time trace was recorded until the baseline was reached again.The exact time of the second addition of H 2 O 2 depended on the progress of the H 2 O 2 time trace.For the calculation of residual activities, the v obs for the first and second H 2 O 2 addition was only corrected with the background for the respective substrate (v sb ) followed by calculating the ratio of the corrected first and second initial rates.Residual activities (%), as shown here, reflect the amount of activity left after consuming 100 µM H 2 O 2 (exceptions are marked with an asterisk; here only 50 µM were consumed).(A-F) Comparison of reaction rates and residual activities of LPMOs acting on different substrates.In the case of studying catalysis in the presence of CNC and/or PASC, GtAA9B, HjAA9B, PcAA9D and NcAA9F showed rapid inactivation at 100 µM H 2 O 2 .Therefore, a lower concentration of H 2 O 2 (50 µM, marked with an asterisk, panels C-F) was used.However, fast inactivation, even at 50 µM H 2 O 2 , did not allow for the determination of reaction rates for PcAA9D (E) and NcAA9F (F) acting on CNC.It is known that both LPMOs are active on crystalline cellulose substrates, but the observed initial rates (v obs , including substrate background and off-pathway activity) were below the limit of the analysis, which was defined as the total observed rate (v obs ) being at least 4-times higher than the standard deviation of the rate of the peroxidase-like activity (v off-path ).Underpinning fast inactivation, the progress of the reaction started to level off well before the H 2 O 2 time trace reached the baseline.Therefore, these LPMOs are marked as inactive on CNC.In the case of NcAA9F acting on PASC (F), v obs was 18-times higher than the standard deviation of the peroxidase-like activity (v off-path ) and the baseline was reached after 200 s, but the rate for the second reaction cycle was equal to the substrate-dependent background (v sb ) measured for PASC, therefore no residual activity is reported.Obviously, both enzymes could not withstand higher H 2 O 2 concentrations or the employed substrates could not provide sufficient protection against inactivation.Similarly, the same criteria were applied to screen LPMO activity (initial H 2 O 2 concentration was 100 µM, data not shown) on the substrates listed in the main text, but only NcAA9C showed activity on substrates besides xyloglucan, PASC and CNC.(G) Off-pathway activity and futile H 2 O 2 turnover of LPMOs (200 nM) determined in buffer using standard assay conditions.v enz was determined by linear regression of the first 10 s of the reactions (R 2 = 0.92 or higher) and converted to turnover numbers.Importantly and expectedly, futile H 2 O 2 turnover leads to fast self-inactivation, as shown by the rapid leveling-off of the H 2 O 2 consumption time trace (see black time trace in Figure 1C of the main manuscript).To illustrate this the average number of H 2 O 2 molecules (n) consumed in this futile reaction prior to enzyme inactivation is also shown (purple data points; secondary y-axis).The number of H 2 O 2 molecules consumed per LPMO molecule was calculated by determining the difference between the starting concentration of H 2 O 2 and the concentration detected after 200 s.At this time point, the time traces in the absence of substrate have levelled off fully, which indicates complete enzyme inactivation.On average, about 120 futile turnovers lead to full inactivation of AA9 LPMOs in the absence of substrate.For comparison, the collective data for productive reactions show that productive consumption of 100 µM H 2 O 2 results in an average 500 (200 nM LPMO) to 2000 (50 nM LPMO) turnovers while maintaining between 23 and 97% of the initial enzyme activity.Clearly, and expectedly, the peroxidase off-pathway reaction, with its risk of causing enzyme inactivation, is drastically inhibited in the presence of substrate.Thus, the correction of the observed LPMO reaction rate, v obs , with the rate of the off-pathway 165 reaction, v off-path , which was applied in all of the experiments described, might lead to slight 166 underestimation of the rate of the peroxygenase reaction, v enz .

Figure S4
. Revealing the activity of 200 nM horseradish peroxidase (HRP) acting on 100 g L -1 Kraft lignin.The experiment was conducted using standard assay conditions (30 mM sodium acetate containing 100 mM KCl, pH 6.0, 30°C, electrode rotation at 50 s -1 ).Kraft lignin was suspended, and partially dissolved, by incubation overnight in the reaction buffer, on an orbital shaker at 30 °C and the pH was adjusted to 6.0 afterwards.(A) Exemplary calibration of the H 2 O 2 sensor in the presence of 100 g L -1 Kraft lignin.Importantly, the stability of the signal and the fast response time of the H 2 O 2 sensor are not affected by the presence of lignin components.The calibration function of this experiment is shown in (B), averaged over 3 independent measurements.(C) H 2 O 2 consumption time trace for 200 nM HRP acting on 100 g L -1 Kraft lignin.The reaction in the absence of substrate (gray time trace), shows no consumption of H 2 O 2 by HRP.The first 50 s of the H 2 O 2 time trace in the presence of substrate were fit using linear regression (fit shown as purple dotted line) to determine the initial rate.Since HRP showed no off-pathway consumption of H 2 O 2 , the obtained enzymatic rate for H 2 O 2 turnover was only corrected with the rate of substrate dependent consumption of H 2 O 2 (0.05 s - 1 ).Three independent measurements showed that HRP converts Kraft lignin with a turnover number of 3.0 ± 0.1 s -1 .Although it is known that HRP can act on lignin components 72 to the best of our knowledge, this is the first time that HRP activity on a real lignin substrate is shown with a continuous method.(D) Image of the electrochemical cell and the H 2 O 2 sensor rotating at 50 s -1 in the presence of 100 g L -1 Kraft lignin.S4. Figure S6.Comparison of NcAA9C activity using either ascorbate or NcCDHIIA for reduction.The dataset was recorded using standard assay conditions.(A) NcCDHIIA was added in different concentrations to find the saturating CDH concentration for NcAA9C acting on 4 g L -1 xyloglucan.The cytochrome domain of NcCDHIIA was fully reduced by incubating the enzyme with 1 mM cellobiose for 30 s, before adding NcAA9C to the reaction.The initial rates obtained for different NcCDHIIA concentrations can be fitted to a hyperbolic function which results in a half saturating concentration of 6.5 ± 0.4 mM and a TN max,app of 40 ± 1 s -1 , which is in good agreement with the TN max,app determined for ascorbate (39 ± 1 s -1 , Table S4).(B) Comparison of LPMO peroxygenase and off-pathway activity in the presence of ascorbate or NcCDHIIA.Due to the oxidase activity of 20 µM NcCDHIIA, i.e. production of H 2 O 2 resulting from oxidation of cellobiose, the off-pathway consumption of H 2 O 2 by NcAA9C in buffer appears "lower" compared to the reaction with ascorbate.However, catalysis of NcAA9C acting on xyloglucan, fueled either by ascorbate or NcCDHIIA used at saturating concentrations results in very similar H 2 O 2 depletion curves, which shows that catalysis is independent of the used electron donor.S7A).The blank sample (red) was retrieved before addition of ascorbate.The next sample (green) was taken 130 s after starting the reaction by subsequent addition of 500 µM ascorbate and the LPMO.Both samples were diluted 50-fold prior to analysis.The 20 mg L -1 standards for cellobiose (DP2), cellotriose (DP3) and cellopentaose (DP5) were applied as a combined standard.The consumption of DP5 was quantified with standards in the range of 5 to 30 mg L -1 .The chromatograms are shown with a 0.3 min time-offset and show that when starting with 100 µM H 2 O 2 and 1 mM substrate, about 10 % of the substrate (i.e., 100 µM) has been converted after 130 s, as expected (Figure S7A).(B) Section of the chromatograms shown in panel A (13-23 min interval) illustrate the decrease in the cellopentaose (DP5) concentration and the concomitant increase in cellotriose (DP3) concentration due to the LPMO catalyzed reaction.Figure S9.Ratio of consumed H 2 O 2 molecules per consumed ascorbate molecule.The dataset was recorded using standard assay conditions (30 mM sodium acetate, pH 6.0, containing 100 mM KCl, 30 °C), with the following exceptions: (A) To determine the stoichiometry, the reaction was run with 200 nM NcAA9C and a very low, 2 µM, ascorbate concentration.The low ascorbate concentration was chosen because it is not sufficient to enable full consumption of the supplied 100 µM H 2 O 2 .After H 2 O 2 consumption was complete (600 s), the remaining H 2 O 2 concentration was used to calculate the ratio of moles of consumed H 2 O 2 to moles of consumed ascorbate, shown in panel (B), with correction for the off-pathway consumption of H 2 O 2 (grey curve).The reaction of NcAA9C acting on xyloglucan stopped once ascorbate was completely consumed, as indicated by the flattening of the time course.On the other hand, at saturating ascorbate concentrations (500 µM), the NcAA9C catalyzed reaction consumes all of the added H 2 O 2 , and at a higher rate, as shown in Figure 1D.Panel (B) shows that the ratio of H 2 O 2 molecules per consumed ascorbate decreases at lower substrate concentrations (0.25 g L -1 ), showing that less efficient substrate binding leads to an increased futile H 2 O 2 turnover, which again leads to higher reductant consumption.A comparable ratio of 18 cleavages per ascorbate molecule was previously reported for SmAA10A acting on chitin 73 .Figure S10.Determination of NcAA9C activity on hemicellulose and cellulose substrates.The substrate screening reactions were conducted using standard assay conditions (hemicelluloses and PASC were used at 4 g L -1 , CNC at 100 g L -1 ) and are divided into 3 parts, each consisting of a 5-min time period in which the change in the H 2 O 2 concentration was recorded.First, the change in H 2 O 2 concentration was studied in the presence of the potential substrate (0-5 min) to determine the influence of the substrate on the abiotic H 2 O 2 consumption rate.In the second part, 500 µM ascorbate was added to the electrochemical cell (5-10 min), which in all cases increased H 2 O 2 consumption; this phase provided the substrate background, v sb (µM s -1 ).In the third part, NcAA9C was added to a final concentration of either (A) 50 nM or (B) 200 nM (10-15 min).The slopes obtained for the first two parts are summarized in Table S7, and the turnover numbers derived from the background-corrected reaction rates (v enz ) are summarized in Table S6.Explained in more detail, the obtained rates for the LPMO reactions, v obs , were converted to enzyme rates (v enz ) according to v enz = v obs -v off-path -v sb .The rate for the background reaction with the studied substrate and reductant (v sb ) was obtained from the second part of the progress curves, as outlined above.This is called substrate background correction and its magnitude depended on the studied system.The progress curves illustrate that the backgrounds for some hemicellulosic substrates (e.g.xylan from beechwood) were higher than for cellulosic substrates.The rate for the off-pathway activity of the LPMO (v off-path ) was derived from the third part of the reaction without substrate in the third part.The catalytic activities of all other LPMOs included in this study were assessed using the same experimental conditions, but were only active on xyloglucan, PASC and CNC.Their catalytic performance on these substrates is summarized in Figure S3.xyloglucan; these reactions reveal difference between the LPMOs.NcAA9M, and to some extent GtAA9B, are still able to reach quite high turnover numbers despite the low substrate concentration.This could be explained by their higher tolerance to xylose substitutions on the xyloglucan backbone, which has been previously documented for NcAA9M 57 and a homologue of GtAA9B 56 and which would increase the effective substrate concentration.All experiments were conducted using standard assay conditions, however the LPMO concentration was 1 µM of LPMO and we used a high ascorbate concentration (5 mM) to ensure that the reductant concentration was not limiting.The ascorbate stock was prepared in 30 mM sodium acetate buffer and adjusted to pH 6.0.The co-substrate, H 2 O 2 , was added in 10 µM steps.Each addition was followed by waiting until the H 2 O 2 was completely consumed, after which another 10 µM of H 2 O 2 was added, and this was repeated until only a fraction of the initial LPMO activity remained.The H 2 O 2 titration steps are exemplarily shown by a bundle of blue arrows.Due to the gradual decrease in LPMO activity, the time passing between the injections of 10 µM H 2 O 2 increases over time, as is clearly visible in the right half of the Figures.The H 2 O 2 titration steps are summed up to the total turnover number (TTN) summarized in Table S9.The obtained summed up H 2 O 2 turnovers were corrected for the substrate dependent background consumption of H 2 O 2 over the course of the experiment.The observed turnovers include productive turnovers of the peroxygenase activity, as well as offpathway consumption of H 2 O 2 eventually leading to complete loss in activity.The initial rate obtained after approximately every 5 th titration step was used to estimate the amount of residual activity left, by comparing to the initial rate obtained in the first H 2 O 2 titration step (=100 %).These residual activities are plotted as a function of H 2 O 2 turnover (n) in Figure 3 of the main text.When the residual enzyme activity was close to zero, 1 µM fresh LPMO was added to the reactions, which, in all cases, led to immediate depletion of remaining H 2 O 2 , showing that the decrease in LPMO activity was not caused by substrate or reductant depletion.(A) NcAA9Ccatalyzed turnover of PASC.(B) NcAA9F-catalyzed turnover of PASC.During this experiment, 500 µL aliquots were collected at ten different time points to determine the total amount of C1-oxidized products formed using HPAEC-PAD analysis.The collected samples were heat inactivated at 90 °C for 10 min.The respective sampling points are indicated by a bundle of red arrows.For clarity, sampling time points after 15 min are not indicated.Progress curves based on H 2 O 2 consumption data and the total amount of off-line detected C1-oxidized products are shown in Figure 3A S9.Residual activities were determined as described in the legend of Figure S16 and are reported in Figure 3 of the main manuscript.Turnover of H 2 O 2 includes productive turnovers of the peroxygenase activity as well as non-productive futile turnovers eventually leading to a loss of activity.4D of the main manuscript.The data set was recorded using standard assay conditions, but the initial H 2 O 2 concentration was varied between 25 and 300 µM.The pHrange was covered by using a 30 mM sodium acetate buffer for pH 5.0 and 6.0, while an imidazole chloride buffer was used for pH 7.0 and pH 8.0.The ascorbate stock solution was pre-buffered to the respective pH of the experiment.In order to only vary one parameter at a time, the ascorbate concentration was kept constant at 3 mM throughout all experiments.This concentration corresponds to 10-times the half saturating concentration for ascorbate at pH 5 (at this pH, the half saturating concentration is at its highest when considering the pH 5 -8 range).This ascorbate concentration should ensure full reduction of the LPMO at all tested pHvalues.For each H 2 O 2 concentration, the residual activity was determined after the initial amount of H 2 O 2 had been consumed.To do so, 25-300 µM H 2 O 2 was added for a second time after reaching the baseline (i.e.0 µM H 2 O 2 ) and the second cycle of H 2 O 2 depletion was followed until the baseline was reached again (where the latter was not always achieved due to enzyme inactivation).Residual activities were calculated as the ratio of the initial reaction rates derived from the second and first reaction cycles of the experiments, as described in Experimental Procedures and the legend of Figure 1 of the main text.The first 15 s of the 1 st and the 2 nd cycle of all measurements were used to determine the initial rate.The obtained kinetic constants are summarized in Table S10.
While the used H 2 O 2 concentrations were almost saturating at pH 5.0 and 6.0, saturation was less pronounced at pH 7.0 and 8.0.Since enzyme inactivation, expectedly, becomes more prominent at higher H 2 O 2 concentrations, and despite the fact that the reported rates are initial rates derived from reasonably straight initial parts of the H 2 O 2 depletion curve, it is not clear whether the observed saturation is due to a true kinetic saturation (e.g. in the form of a half saturating concentration or K M,app. ) or due to overlaying effects related to self-inactivation.(E) To find out if the observed apparent saturation depends on the substrate, we repeated the experiment at pH 6.0 using 100 g L -1 CNC as the substrate for 50 nM NcAA9C.The H 2 O 2 concentration was varied between 25 and 100 µM.The reactions with 75 and 100 µM H 2 O 2 led to full enzyme inactivation after one reaction cycle and for this reason higher H 2 O 2 concentrations were not tested.The dataset shows a small increase in initial reaction rate at increasing H 2 O 2 concentration, with no signs of saturation and early onset of enzyme inactivation.This different behavior with cellulose, compared to xyloglucan, underpins the impact of the substrate on LPMO catalysis and stability.Looking at panels A-E, it is worth noting that the expected low micromolar concentrations of H 2 O 2 in natural ecosystems may lead to stable LPMO reactions with appreciable rate, but are far from saturating.inactivation.These results for GtAA9B underpin the huge impact of the substrate on LPMO performance.While a good soluble, rapidly diffusing substrate like, in this case, xyloglucan leads to fast and stable reactions at almost any H 2 O 2 concentration, enzyme performance in the presence of CNC is very different.

Table S1. Summary of published traits of the LPMOs used in this study.
Abbreviations: DP, degree of polymerization; PASC, phosphoric acid swollen cellulose; RAC, regenerated amorphous cellulose; CMC, carboxymethyl cellulose; MCC, microcrystalline cellulose; n.a., not available. 1Note that substrates that are not listed may not have been tested.concentration of 100 µM and for each substrate a newly prepared electrode was used.Depending on the studied system, the number of experiments (n) performed with the same H 2 O 2 sensor varied between 3 and 21.The shown numbers were averaged over the number of measurements performed in the respective system.To evaluate the effect of pH on the performance of the H 2 O 2 sensor, a single sensor was used between pH 4.0 and 8.0 in the presence of 4 g L -1 xyloglucan.To cover the pH-range a 30 mM NaOAc buffer was for pH 4.0 & 5.0, whereas a 30 mM imidazole-chloride buffer was used for pH 6.0-8.0.Importantly, neither different pH values nor changing substrate concentrations affected the quick response time of <2s of the H 2 O 2 sensor.The coefficient of determination refers to the linearity of the obtained calibration function.The limit of quantification was determined by dividing 10 times the standard deviation of the baseline (nA cm -2 ) with the sensitivity (nA µM -1 cm -2 ) of the calibration function.  .Rates were determined at standard assay conditions.All measured rateswere corrected for the substrate dependent background and the off-pathway peroxidase activity of the LPMO, which was recorded in reactions lacking substrate (indicated by 0 g L -1 xyloglucan).For the reaction with xyloglucan and 1 µM NcCDHIIA the rate of this off-pathway activity was excessively low and no correction was needed.The initial rates for the off-pathway activity were fit with R 2 ≥ 0.92, rates for the peroxygenase reaction were fit with an R 2 ≥ 0.98.Some of the underlying raw-data is shown in the main text and in Figure S6.The fully corrected initial rates v enz (µM s -1 ) were converted to turnover numbers (TN, s -1 ) that are independent of the LPMO concentration.The reported TNs were derived from 3 independent experiments and are provided with the standard deviation.Note that the reported TNs on xyloglucan are independent of the enzyme concentration, demonstrating a linear relationship between the enzyme dosage and H 2 O 2 turnover (i.e., a linear relationship between v enz and the enzyme dosage, which translates in equal TNs at all dosages).To ensure an excess of reducing agent, the ascorbate concentration was 5 mM.The ascorbate stock solution was buffered to pH 6.0 to avoid pH variations in the reaction after its addition.The obtained turnover of co-substrate for individual titration steps was summed up to the total turnover numbers and corrected for the substrate dependent background consumption of H 2 O 2 during the course of the experiment.The reported TTNs include productive turnovers of the peroxygenase reaction as well as offpathway consumption of H 2 O 2 eventually leading to self-inactivation indicated by a decrease in residual activity.The residual activity of the LPMOs was calculated from the slope of the last titration step (before terminating the measurement) and compared to the slope of the first titration step (=100 % activity).The initial rates used to calculate residual activities were only corrected for the substrate dependent background consumption of H 2 O 2 as described in the Experimental Procedures and in the main text.The respective TTNs and residual activities were extracted from the titration data-sets of three independent experiments.After stopping the experiments, as a control, the same concentration of LPMO was added again to verify that the observed reduced LPMO rate was not caused by substrate or reductant depletion (which was indeed the case for all reported experiments).Note that the reported TTNs are minimum values, especially in cases where residual activity of the LPMOs was high when the experiment was terminated.The progress of these reactions (residual activity plotted as a function of H 2 O 2 turnover) is shown in the main text, together with data obtained by off-line analysis of the amount of generated C1-oxidized products.Table S10.The effect of pH on the apparent saturation kinetic behavior of the enzymesubstrate, enzyme-reductant, and enzyme-H 2 O 2 interaction and the extrapolated maximal turnover number (TN max,app ) of NcAA9C acting on xyloglucan.Note that, while the interaction of NcAA9C with xyloglucan is described with the apparent steady-state enzyme-substrate complex equilibrium constant (K M,app ), this terminology cannot be used for ascorbate and H 2 O 2 .Since ascorbate is not involved in catalysis (in every catalytic cycle), but only in priming the peroxygenase reaction, defining a half-saturating concentration is more suitable.In the case of H 2 O 2 , the data shown in Figure S22 indicates that inactivation may have a substantial influence on the observed saturation behavior.Therefore, defining a halfsaturating concentration is more appropriate than reporting a true kinetic affinity constant.Two overlapping buffers were used to cover the range between pH 4.0-8.0.Sodium acetate buffer (30 mM) was used between pH 4.0 and 6.0 while 30 mM imidazole chloride buffer was used between pH 6.0 and 8.0.Importantly, the pH-profile in the main text shows that the obtained turnover numbers at pH 6.0 (overlapping pH-profile), are independent of the used buffer system and that imidazole does not affect catalysis of NcAA9C.All experiments were conducted varying only one parameter at the time and otherwise using standard assay conditions (50 nM NcAA9C).The half saturating concentration for ascorbate was determined using 100 µM H 2 O 2 and 4 g L -1 xyloglucan.Based on this data set, the ascorbate concentration was increased to 3 mM (10x half-saturating concentration at pH 5.0) to determine the K M,app for xyloglucan and the half-saturating concentration for H 2 O 2 .The underlying data to obtain half-saturating concentrations and the K M,app are shown in the main text.

Figure S5 .
Figure S5.Ascorbate concentration-dependency of LPMO catalyzed reactions.Ascorbate concentration-dependent H 2 O 2 turnover by LPMOs was studied using standard assay conditions and either xyloglucan (A) or PASC (B) as substrate.The ascorbate concentration necessary to reach saturating conditions depends on the respective enzyme.In accordance with the saturation curves depicted in this figure, we used 500 µM ascorbate as standard concentration in the measurements for all LPMOs except NcAA9F and HjAA9B (2 mM ascorbate).The ascorbate concentrations required to achieve half saturation for all LPMOs are shown in TableS4.

Figure S7 .
Figure S7.Cellopentaose conversion by NcAA9C.(A) Time courses for H 2 O 2 consumption by NcAA9C in the presence of different cellopentaose concentrations.The experiments were done with 100 nM NcAA9C and standard assay conditions (30 mM sodium acetate buffer, pH 6.0 and 100 mM KCl, 30 °C). (B) The dependence of initial catalytic rates of NcAA9C acting on cellopentaose.The initial linear 15-20 s of the H 2 O 2 time traces were fit with an R 2 of 0.98 or higher and used to calculate turnover numbers (corrected for off-pathway activity and v sb ).The data were fitted to a hyperbolic function (R² = 0.99), which can be used to extrapolate a TN max,app of 75 ± 5 s -1 and a K M,app of 1.2 ± 0.2 mM.(C) Verification of the cellopentaose:H 2 O 2 stoichiometry by comparing the amount of cellopentaose consumed in the reactions of Panel A, as assessed by HPAEC-PAD (dashed bars), with H 2 O 2 consumption, as depicted in panel A. For determination of product formation, reactions were stopped at different time points and the concentration of remaining cellopentaose was quantified.An exemplary chromatogram is displayed in Figure S8.The reactions with 0.1, 0.2 and 0.4 mM cellopentaose were stopped after 200 s by incubation at 90 °C for 10 min.The reaction with 1 mM cellopentaose was stopped after 130 s also by incubation at 90 °C for 10 min.The graph shows that the amount of consumed cellopentaose corresponds well with the amount of H 2 O 2 consumed, confirming the 1:1 stochiometry.Importantly, the shown data for consumption of H 2 O 2 and degradation of cellopentaose are from the same reaction and therefore allow direct comparison of co-substrate utilization and concomitant glycosidic bond cleavage.
Figure S11.H 2 O 2 consumption traces for NcAA9C acting on different concentrations of (A) xyloglucan (50 nM NcAA9C), (B) PASC (100 nM NcAA9C) and (C) CNC (200 nM NcAA9C).The reactions were performed under standard assay conditions (500 µM ascorbate, 30 mM sodium acetate buffer, pH 6.0 with 100 mM KCl at 30 °C).The time traces shown are the average of three independent measurements.The enzyme concentration was varied for the different substrates to obtain comparable reaction rates.The initial rates were obtained by linear fitting the 10 -25 s part of the progress curves and were corrected for the substrate dependent background and the off-pathway activity of NcAA9C as explained in detail in the Materials and methods and the legend of Figure S10.The derived turnover numbers are shown in the main text.All three panels illustrate the importance of high substrate concentrations to maintain NcAA9C peroxygenase activity.Lower substrate concentrations result in enzyme inactivation, as indicated by the slowdown of the enzyme's H 2 O 2 consumption rate before all available H 2 O 2 is consumed.(D) Control experiment performed to validate that the slowdown and eventual cessation of H 2 O 2 consumption is due to irreversible inactivation of NcAA9C.The reaction was performed in standard assay buffer (30 mM sodium acetate, pH 6.0 containing 100 mM KCl) and started by the addition of 500 µM ascorbate followed by the addition of 200 nM NcAA9C to monitor the off-pathway activity of LPMO in the absence of substrate.H 2 O 2 consumption stopped some 50 s after addition of the LPMO.Subsequent addition of 2 g L -1 cellopentaose (DP5) directly into the measurement cell did not result in any change in the H 2 O 2 concentration.
Figure S13.H 2 O 2 time traces for four xyloglucan-active LPMOs, NcAA9M, NcAA9C, GtAA9B, HjAA9B.The reactions were performed under standard assay conditions containing 50 nM LPMO.The time traces shown are an average of three independent measurements.(A) Reactions with 4 g L -1 xyloglucan; similar rates were obtained for NcAA9M, NcAA9C and GtAA9B acting on xyloglucan (TN 30-35 s -1 ).The calculated turnover numbers and the corresponding residual activities are shown in Figure S3.(B) Reactions with 0.25 g L -1xyloglucan; these reactions reveal difference between the LPMOs.NcAA9M, and to some extent GtAA9B, are still able to reach quite high turnover numbers despite the low substrate concentration.This could be explained by their higher tolerance to xylose substitutions on the xyloglucan backbone, which has been previously documented for NcAA9M 57 and a homologue of GtAA9B 56 and which would increase the effective substrate concentration.
Figure S14.H 2 O 2 time traces for C4-oxidizing LPMOs acting on cellulose.Experiments were performed under standard assay conditions with PASC or CNC as substrates (200 nM LPMO, 500 µM ascorbate; except for HjAA9B where 2 mM ascorbate was used to achieve full saturation).(A) NcAA9C and NcAA9M acting on 100 g L -1 CNC, (B) NcAA9C and NcAA9M acting on 4 g L -1 PASC, (C) GtAA9B acting on 4 g L -1 PASC and (D) HjAA9B acting on 4 g L -1 PASC.The derived turnover numbers and the residual activities are summarized in Figure S3.
of the main text and confirm that H 2 O 2 consumption correlates with substrate oxidation.Exemplary chromatograms for offline detection of C1oxidized products of this experiment are shown in Figure S17.(C) NcAA9C-catalyzed turnover of CNC.
Figure S17.HPAEC-PAD analysis of C1-oxidized products generated by NcAA9F acting on 4 g L -1 PASC.The displayed chromatograms (plotted with a 0.1 min offset) show the total amount of oxidized sites generated by NcAA9F.Polymeric and oligomeric material in the reaction mixtures was hydrolyzed in a single reaction by incubating the reaction suspension for 48 h at 40 °C and 1000 rpm with 5 µM Thermobifida fusca Cel6A to hydrolyze soluble products and remaining PASC to native and oxidized dimers (DP2) and trimers (DP3).To obtain time points, the NcAA9F reactions were sampled after which enzyme activity was stopped by heat inactivation (10 min at 90 °C).H 2 O 2 titration and consumption data for this experiment are shown in Figure S16B.The H 2 O 2 titration data and the levels of C1-oxidized products are compared in Figure 3A of the main text.

Figure S20 :
Figure S20: The pH-dependency of the second-order reoxidation rate of NcAA9C by H 2 O 2 in absence of substrate.Reoxidation rates were recorded in 50 mM sodium acetate (pH 5.0 & 6.0), imidazole-HCl (6.0 -8.0), or MOPS-HCl (7.0) buffer at 30 °C.The second-order rate for re-oxidation of NcAA9C-Cu(I) by H 2 O 2 shows no pH-dependency, fluctuating at about 100,000 M -1 s -1 between pH 5.0 and 8.0.While this may seem in contrast with the pH-activity profiles described in Figure 4 of the main text, the low pH-dependency of the rate of reoxidation by H 2 O 2 cannot be extrapolated to what happens during productive catalysis, since binding of the substrate has a massive effect on H 2 O 2 turnover, increasing it by as much as 50fold 28 .

Figure S23 .
Figure S23.H 2 O 2 , substrate and inactivation-dependent kinetics of 50 nM NcAA9M and GtAA9B acting on 4 g L -1 xyloglucan or 100 g L -1 CNC.The data set was recorded using standard assay conditions (30 mM sodium acetate containing 100 mM KCl, pH 6.0, 500 µM ascorbate, 30 °C), but the initial H 2 O 2 concentration was varied between 25 and 300 µM.Residual activities after one reaction cycle were determined as described in the legend of Figure S22.(A-B) H 2 O 2 and substrate (xyloglucan vs. CNC) dependent kinetics of NcAA9M.Turnover numbers for NcAA9M acting on xyloglucan increase almost linearly between 25 and 300 µM H 2 O 2 and residual activities only decrease to half of the initial activity at the highest H 2 O 2 concentration.No apparent saturation could be observed; therefore, no apparent catalytic rate constants were derived from this dataset.For CNC the situation is very different, i.e. lower rates and more inactivation, much alike what was observed for NcAA9C (Figure S22B, E) (C-D) The same experiment was conducted using GtAA9B and for this LPMO the difference between xyloglucan and CNC was even bigger.Activity on xyloglucan increased almost linearly between 25 and 300 µM H 2 O 2 , with remarkably little enzyme inactivation, whereas the apparent activity on CNC was low at all tested H 2 O 2 concentrations, with rapid enzyme

Table S2 . H 2 O 2 sensor performance in buffer and different substrate-containing solutions or suspensions.
The data shown were obtained from experiments using NcAA9C acting under standard assay conditions on different xyloglucan, PASC or CNC concentrations in 30 mM sodium acetate buffer pH 6.0 containing 100 mM KCl.The calibration data was obtained from experiments with a final H 2 O 2

Table S3 . Rates of catalysis for NcAA9C acting on various (hemi-)cellulosic substrates (xyloglucan, PASC, CNC and MCC).
Either ascorbate or NcCDHIIA was used as reductant, as indicated in the Table

Table S5 . Correlation of H 2 O 2 consumed by NcAA9F acting on 4 g L -1 PASC and concomitant formation of C1-oxidized products in the same reaction according to HPAEC-PAD analysis.
The quantified C1-oxidized dimers and trimers derived from samples obtained during the TTN experiment were summed up to obtain the total amount of oxidized products (soluble and insoluble).The data displayed here are also shown in the main text.The data obtained during the H 2 O 2 titration experiment was corrected with the substrate dependent consumption of H 2 O 2 .Therefore, oxidized products according to the H 2 O 2 sensor include productive turnovers of the peroxygenase reaction and non-productive turnovers eventually leading to inactivation of the enzyme.The difference between oxidized products according to HPAEC-PAD analysis and the H 2 O 2 sensor at later sampling points may be explained by an increasing off-pathway consumption of H 2 O 2 .Importantly, both measurements show good agreement at earlier time points (until 10 min).

Table S6 . Ranked activity of NcAA9C acting on hemicelluloses and cellulosic substrates.
All measurements were conducted using standard assay conditions (500 µM ascorbate, 100 µM H 2 O 2 , 30 mM sodium acetate buffer, pH 6.0, 100 mM KCl, at 30 °C).The substrate concentration was 4 g L -1 for all substrates, except for CNC and MCC (100 g L -1 ).For hemicelluloses and cellopentaose, 50 nM NcAA9C was used, while 200 nM NcAA9C was used for CNC and MCC.The hemicellulosic substrates and cellopentaose are ranked according to the obtained turnover numbers, followed by amorphous and crystalline cellulose substrates.All other studied LPMOs were also tested in reactions with all the listed substrates, but were inactive on all substrates except xyloglucan, PASC and CNC.

Table S7 . Abiotic substrate-dependent change in H 2 O 2 concentration.
The experiment was performed using standard assay conditions at pH 6.0.In the first part of the experiment (0-5 min), the change in H 2 O 2 concentration in the presence of substrate was recorded, followed by recording the change in H 2 O 2 concentration in the presence of 500 µM ascorbate (5-10 min).This is termed the substrate dependent background (v sb ), which was used to correct the observed initial rates (v obs ), as described in the Experimental Procedures.The slopes were determined by linear regression.The raw data is displayed in FigureS10.

Table S8 .
Comparison of the apparent dissociation constant of the enzyme-substrate complex (K S,app ) in the presence of saturating amounts of reductant, and extrapolated maximal turnover numbers (TN max,app ) for different LPMOs acting on different substrates.The K M,app for cellopentaose reported for NcAA9C is also given in millimolar concentration in parentheses.The underlying datasets are shown in Figure2Bof the main text and in FigureS7A.

Table S9 . Total turnover numbers (TTN) of LPMOs acting on various substrates.
The substrates used were xyloglucan (50 µM H 2 O 2 titration steps, except NcAA9M which was also tested with 25 µM H 2 O 2 titrations steps), PASC and CNC (10 µM H 2 O 2 titration steps).The experiments were conducted under standard assay conditions using 50 nM LPMO for xyloglucan or 1 µM LPMO for PASC and CNC.