Functional characterization of a lytic polysaccharide monooxygenase from Schizophyllum commune that degrades non-crystalline substrates

Lytic polysaccharide monooxygenases (LPMOs) are mono-copper enzymes that use O2 or H2O2 to oxidatively cleave glycosidic bonds. LPMOs are prevalent in nature, and the functional variation among these enzymes is a topic of great interest. We present the functional characterization of one of the 22 putative AA9-type LPMOs from the fungus Schizophyllum commune, ScLPMO9A. The enzyme, expressed in Escherichia coli, showed C4-oxidative cleavage of amorphous cellulose and soluble cello-oligosaccharides. Activity on xyloglucan, mixed-linkage β-glucan, and glucomannan was also observed, and product profiles differed compared to the well-studied C4-oxidizing NcLPMO9C from Neurospora crassa. While NcLPMO9C is also active on more crystalline forms of cellulose, ScLPMO9A is not. Differences between the two enzymes were also revealed by nuclear magnetic resonance (NMR) titration studies showing that, in contrast to NcLPMO9C, ScLPMO9A has higher affinity for linear substrates compared to branched substrates. Studies of H2O2-fueled degradation of amorphous cellulose showed that ScLPMO9A catalyzes a fast and specific peroxygenase reaction that is at least two orders of magnitude faster than the apparent monooxygenase reaction. Together, these results show that ScLPMO9A is an efficient LPMO with a broad substrate range, which, rather than acting on cellulose, has evolved to act on amorphous and soluble glucans.

www.nature.com/scientificreports/LPMOs can utilize either molecular O 2 or H 2 O 2 as a co-substrate to catalyze the hydroxylation of a carbon in the scissile glycosidic bond (C1 or C4 in cellulose), leading to spontaneous bond cleavage 9,10,18 .Once reduced, a single LPMO molecule acting as a peroxygenase can catalyze multiple turnovers 10,19 .Non-substrate bound reduced LPMOs in solution can react with O 2 to produce H 2 O 2 20,21 , or with H 2 O 2 , generating reactive oxygen species that may lead to damage and autocatalytic inactivation 10 .While the significance of the monooxygenase vs. peroxygenase reaction is still under debate, it is worth noting that LPMO reactions with H 2 O 2 are several orders of magnitude faster than those driven by O 2 4,19,[22][23][24][25][26] .LPMOs currently populate eight of the 17 families in the auxiliary activities (AA) class of the carbohydrateactive enzymes (CAZy) database (http:// www.cazy.org/; 27 ).This class encompasses oxidases, peroxidases, and oxidoreductases in addition to LPMOs 28 .Most characterized fungal LPMOs can be found in the AA9 family, which at the time of writing contains more than 60 functionally characterized LPMOs with activities on insoluble and soluble cellulosic and hemicellulosic substrates.The N-terminal histidine of AA9 LPMOs, which is part of the copper-binding histidine brace, carries a methylation 12 , which helps to protect the enzymes from oxidative damage 29 .Of note, non-methylated variants of these LPMOs have been produced in the yeast Pichia pastoris and are active.Well-studied examples of AA9 LPMOs include NcLPMO9C from Neurospora crassa, LsLPMO9A from Lentinus similis, and CvLPMO9A from Collariella virescens 26,[30][31][32][33][34] , which stand out due to their proven ability to act on soluble substrates, including cello-oligomers, and hemicelluloses such as glucomannan and xyloglucan.
The ability of LPMOs to boost the action of canonical glycoside hydrolases makes them interesting candidates for use in the valorization of recalcitrant polysaccharides in lignocellulosic biomass 2,35,36 .Indeed, modern cellulase cocktails used in lignocellulosic biorefineries contain LPMOs, and their contribution to cellulose saccharification efficiency is evident 3,[37][38][39] .So far, LPMO action in bioprocessing has exclusively been focused on oxidative degradation of cellulose, whereas the potential impact of hemicellulolytic LPMO activities, if present in commercial cocktails, has not been addressed.The continued elucidation of novel LPMOs acting on various lignocellulosic polysaccharides may provide novel tools for biomass processing and may help understand the biological reasons for the large LPMO multiplicity observed in some fungal species.
Basidiomycete wood-decaying filamentous fungi are a rich source of enzymes for the depolymerization of complex plant matter, including LPMOs.The genome of one such basidiomycete fungus, Schizophyllum commune, was first sequenced in 2010, and showed, in addition to genes coding for an extensive array of glycoside hydrolases active on cellulose, xylan, and pectin, the presence of genes encoding 22 putative AA9s 40 .A comparative study of four fungi including S. commune by Zhu and colleagues indicated that the S. commune secretome had significantly higher cellulase and xylanase activities than other white-and brown-rot fungi tested during solid-state fermentation of Jerusalem artichoke stalk.Proteomic analysis of the S. commune secretome revealed the presence of a wide range of cellulolytic and hemicellulolytic enzymes, and eight AA9s, including ScLPMO9A.In addition, a crude secretome-based enzyme cocktail from S. commune outperformed a commercial enzyme blend from Trichoderma longibrachiatum in saccharifications of multiple lignocellulosic substrates, both in conversion of glucan and xylan 41 .A comparative study of S. commune and the closely related Auriculariopsis ampla showed that the gene encoding ScLPMO9A and the orthologous gene in A. ampla are among the most highly upregulated AA9 genes in vegetative mycelium growing on poplar wood 42 .
As an AA9 candidate for in-depth characterization, ScLPMO9A is of particular interest, as it is produced by S. commune under different conditions and when grown on different substrates, hinting at a crucial role of this enzyme during growth and nutrient acquisition.In this study we have cloned, produced, and purified ScLP-MO9A, and performed an in-depth functional characterization of this enzyme.The properties of this singledomain AA9 LPMO, active on soluble substrates, are compared to the properties of the well-studied NcLPMO9C.We show that ScLPMO9A is active on and interacts with a range of soluble and amorphous substrates, whereas its activity on crystalline cellulose substrates is limited, suggesting that this enzyme's natural role is not in saccharification of recalcitrant cellulose.Additionally, we show that ScLPMO9A is a rapid consumer of H 2 O 2 , both in reactions with soluble cellopentaose and in the oxidative depolymerization of amorphous cellulose.

Sequence and structure analysis
A multiple sequence alignment (MSA) was created using T-Coffee Expresso (http:// tcoff ee.crg.cat/ apps/ tcoff ee/ index.html; 43 ) by aligning the sequence of ScLPMO9A with 46 other characterized AA9s, using only the AA9 domains and removing signal peptides.The MSA was edited in AliView 44 , and the resulting MSA was used for phylogenetic analysis using the ProtTest 3.4 software package, by calculating likelihood scores using all included substitution matrices, all improvements, and four categories for rate variation, empirical amino acid frequencies, and a fixed BIONJ JTT tree for base likelihood calculations 45 .A consensus tree was built with all 120 likelihood scores using the Akaike information criterion (AIC).The resulting consensus tree was edited using the iTol v6 online tool (https:// itol.embl.de/; 46 ).Homology modeling of ScLPMO9A using LsLPMO9A bound to cellohexaose (PDB ID 5ACI; 61.1% sequence identity) as a template structure was performed using PHYRE2 47 , and the resulting model was analyzed in PyMOL (The PyMOL Molecular Graphics System, Version 2.0, Schrödinger, LLC).

Protein expression and purification
A gene fragment containing the signal peptide pelB (MKYLLPTAAAGLLLLAAQPAMA) 48 fused with the gene encoding ScLPMO9A (UniProt ID D8Q364; residues 20-247) was codon-optimized for expression in Escherichia coli and de novo synthesized by GenScript (Piscataway, NJ, USA).Restriction sites for NdeI and NotI were included upstream and downstream of the coding area, respectively.The pelB-ScLPMO9A fragment was isolated from the Genscript vector by digestion with NdeI and NotI, and ligated into the compatible NdeI and NotI sites of the pJB_SP_Sm-vector 49 (replacing the SP_Sm gene fragment), generating pJB_pelB_Sc.The pJB_pelB_Sc www.nature.com/scientificreports/plasmid was transformed into competent E. coli T7 express cells (New England Biolabs, Ipswich, MA, USA) using a heat shock protocol.Plasmid DNA was isolated from the cells using the Wizard® Plus SV Minipreps DNA purification system (Promega, Madison, WI, USA), and the plasmid was verified by full vector sequencing.
Expression of non-labeled and 15 N-isotopically labeled ScLPMO9A was performed as described earlier for other LPMOs 49 .The cell pellet was harvested by centrifugation at 6000 × g and subjected to osmotic shock to prepare a periplasmic extract 50 , which was filtered using a 0.22 µm sterile filter.
Purification of ScLPMO9A was performed by anion-exchange chromatography using an Äkta Purifier system with a 5 mL HiTrap DEAE FF column (GE Healthcare, Uppsala, Sweden), equilibrated with 50 mM Tris-HCl pH 7.5.After loading the sterile-filtered periplasmic extract onto the column, ScLPMO9A was eluted using a 0-500 mM NaCl gradient with 50 mM Tris-HCl pH 7.5 over 90 column volumes.Protein purity was assessed using SDS-PAGE.Fractions containing ScLPMO9A were pooled, buffer exchanged to 50 mM Tris-HCl pH 7.5, and concentrated using a 10-kDa Vivaspin centrifugal tube with PES membrane (Sartorius, Göttingen, Germany) at 10 °C and 6000 × g.For the 15 N-labeled protein used in NMR studies, an additional purification step using size-exclusion chromatography (SEC) was applied.The fractions containing 15 N-labeled ScLPMO9A (identified by SDS-PAGE) were pooled and concentrated to < 5 mL using a 10-kDa Vivaspin centrifugal tube as described above.The concentrated sample was used for further purification by SEC using a HiLoad® 16/600 Superdex® G-75 pg column equilibrated with 25 mM Tris-HCl pH 8.0, 250 mM NaCl with a flow rate of 1 mL/min.Fractions containing 15 N-ScLPMO9A were pooled and buffer exchanged to 25 mM Tris-HCl pH 7.0, 25 mM NaCl, 2.3% glycerol, and concentrated to a final volume of approximately 160 µL as described above.The protein concentration was determined spectrophotometrically at 280 nm using the theoretical molar extinction coefficient (51,005 M −1 •cm −1 ), determined using the ExPASy ProtParam tool 51 .
Expression and purification of NcLPMO9C were performed as described earlier 25 , and copper saturation of both LPMOs was performed as previously described 52 .
The correct amino acid sequence of purified ScLPMO9A was verified by subjecting the purified LPMO to trypsination, as well as extraction and clean-up of the resulting peptides as previously described 53 .Peptides were subsequently analyzed via liquid chromatography-tandem mass spectrometry using a nano UPLC (nanoElute, Bruker Daltonics GmbH, Bremen, Germany) coupled to a trapped ion mobility spectrometry/quadrupole timeof-flight mass spectrometer (timsTOF Pro, Bruker Daltonics GmbH).Peptide separation was achieved using a PepSep Reprosil C18 reverse-phase (1.5 µm, 100 Å) 25 cm × 75 μm analytical column kept at 50 °C coupled to a ZDV Sprayer (Bruker Daltonics GmbH).Prior to sample loading, the column was equilibrated using a pressure of 800 bar.Peptides were separated using an operational flow of 300 nL/min and a 60 min solvent gradient (0-40 min, from 5 to 25% B; 40-45 min, from 25 to 37% B; 45-50 min, from 37 to 95% B; 50-60 min, constant at 95% B).Solvent A consisted of 0.1% (v/v) formic acid in distilled H 2 O (type I, 18.2 MΩ•cm), while the composition of solvent B was 0.1% (v/v) formic acid in acetonitrile.The timsTOF Pro was run in positive ion data-dependent acquisition PASEF mode, using the control softwares Compass Hystar version 5.1.8.1 and timsControl version 1.1.19,and with an acquisition mass range of 100-1700 m/z.The TIMS settings were: 1/ K0 start 0.85 V⋅s/cm 2 and 1/K0 end 1.4 V⋅s/cm 2 , ramp time 100 ms, ramp rate 9.42 Hz, duty cycle 100%.The capillary voltage was set to 1400 V, dry gas to 3.0 L/min, and dry temp to 180 °C.MS/MS settings were as follows: number of PASEF ramps 10, total cycle time 0.53 s, charge range 0-5, scheduling target intensity 20,000, intensity threshold 2500, active exclusion release after 0.4 min, CID collision energy ranging from 27 to 45 eV.The E. coli proteome (UniProt ID UP000002032) was used as reference.We obtained 70% sequence coverage and the obtained sequence, inferred from a combination of MS/MS-based sequencing and peptide masses, was identical to that of UniProt ID D8Q364.
Ascorbic acid (AscA) was used as a reducing agent in all LPMO reactions.Aliquots of a stock solution of 100 mM AscA prepared in TraceSELECT water (Sigma-Aldrich) were prepared and stored at − 20 °C.Aliquots were thawed in the dark immediately prior to use.

Production and consumption of H 2 O 2
An adapted version of the Amplex Red assay 20 was used to quantify H 2 O 2 production by ScLPMO9A and NcLP-MO9C.Reaction mixtures contained 3 µM LPMO, 100 µM Amplex Red (Thermo Fisher Scientific, Waltham, MA, USA), 0.5 U horseradish peroxidase (Sigma-Aldrich), and 50 µM AscA in 50 mM BisTris-HCl pH 6.5, and reactions were initiated by the addition of AscA.The reactions were incubated at 30 °C in a Varioscan LUX plate reader (Thermo Fisher Scientific), and the production of resorufin was measured spectrophotometrically at 563 nm every 22 s over a total time of 6500 s.Control reactions containing 3 µM CuSO 4 in place of the LPMO were performed in parallel.

LPMO reactions with cellulosic and hemicellulosic substates
ScLPMO9A activity was tested with a wide range of cellulosic and hemicellulosic substrates.Reactions with NcLPMO9C were included for comparative purposes.Reactions containing 1 µM ScLPMO9A or NcLPMO9C and individual substrates, or hemicellulosic substrates in combination with PASC, were incubated in 50 mM BisTris-HCl pH 6.5 at 40 °C and 1000 rpm in a Thermomixer (Eppendorf) for 16 h.In reactions with polymeric cellulosic substrates and with hemicellulosic substrates, the substrate concentration was 2 g/L or 4 g/L (with the exception of reactions with sulfite-pulped spruce, which contained 10 g/L substrate).In reactions containing a mixture of PASC and hemicellulosic substrate, the final concentration of both substrates was 2 g/L (total substrate content 4 g/L).In reactions with soluble oligomeric cellulose substrates (cellotetraose, cellopentaose, and cellohexaose), the substrate concentration was either 2 g/L or 1 mM.Reactions were initiated by the addition of 1 mM AscA and stopped by removing insoluble substrates by filtration using a 96-well filter plate (0.45 µm; Merck Millipore, Billerica, MA, USA) operated with a Millipore vacuum manifold system.In the case of soluble substrates, reactions were stopped by boiling for 10 min before filtration.Samples were subsequently stored at -20 °C prior to analysis by high-performance anion exchange chromatography with pulsed amperometric detection (HPAEC-PAD) and/or matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS).All reactions were performed in triplicate, and control reactions without addition of AscA, or only containing relevant substrate(s) and 1 mM AscA, were performed in parallel.

H 2 O 2 -driven activity on PASC and cellopentaose
To assess the impact of H 2 O 2 on product generation by ScLPMO9A acting on PASC, reactions containing 1 µM LPMO, 2 g/L PASC, 1 mM AscA, and 0, 50, 100, or 250 µM H 2 O 2 in 50 mM Tris-HCl pH 7.5 were prepared.The reactions were initiated by addition of AscA and incubated at 45 °C and 1000 rpm in a Thermomixer (Eppendorf).H 2 O 2 was added to the reactions immediately prior to the AscA.Samples were taken at 3, 6, 9, 30, and 60 min, and remaining insoluble substrate was removed by filtration using a 96-well filter plate (0.45 µm; Merck Millipore) operated with a Millipore vacuum manifold system.Samples were subsequently stored at -20 °C prior to analysis by HPAEC-PAD.All reactions were performed in triplicate, and control reactions without addition of AscA were performed in parallel.
Reactions with cellopentaose contained 1 µM LPMO, 1 mM cellopentaose, 50 µM AscA, and 200 or 400 µM H 2 O 2 in 50 mM sodium acetate pH 5.0.Immediately following the addition of H 2 O 2 , reactions were initiated by addition of AscA and incubated as described above.Samples were taken at various time points and reactions were quenched by addition of NaOH to a final concentration of 100 mM.Samples were subsequently stored at -20 °C prior to analysis of generated native products by HPAEC-PAD.All reactions were performed in triplicate, and control reactions without addition of AscA were performed in parallel.

Synergy with cellulases
Degradation of sulfite-pulped spruce was performed under aerobic conditions in 60 mL screw-cap glass bottles (Wheaton, Millville, NJ, USA) using a working volume of 10 mL.The total enzyme loading was 4 mg protein per g dry matter of substrate, and the substrate content was 10% w/w dry matter.The enzymes added were a 9:1 (based on protein content) mix of Celluclast 1.5 L and Novozym 188, both kindly provided by Novozymes AS (Bagsvaerd, Denmark), and the protein concentrations of these enzyme preparations were determined using the Bio-Rad protein assay (Bio-Rad Laboratories, Hercules, CA, USA), based on the Bradford method 61 , using bovine serum albumin as reference protein.Reactions containing ScLPMO9A contained 3.6 mg of the Celluclast 1.5 L/ Novozym 188 blend and 0.4 mg ScLPMO9A per gram of dry matter (based on protein content).The reactions with sulfite-pulped spruce were initiated by the addition of 1 mM AscA.All reactions were incubated at 50 °C with orbital shaking at 200 rpm in a Minitron Shaker incubator (Infors AG, Bottmingen, Switzerland).Reactions were performed in duplicate, and control reactions without addition of AscA were performed in parallel.Samples of 100 µL were taken at 8, 24, 48, and 72 h, and diluted three times in distilled H2O (type I, 18.2 MΩ•cm) prior to enzyme inactivation by boiling for 15 min before storage at -20 °C.Prior to product quantification, samples were thawed at 4 °C and filtered using a 96-well filter plate (0.45 µm; Merck Millipore) operated with a Millipore vacuum manifold system.Quantification of glucose and cellobiose released during saccharification www.nature.com/scientificreports/ was performed by high-performance liquid chromatography using a Dionex Ultimate 3000 system (Dionex, Sunnyvale, CA, USA) equipped with a Shodex RI-101 refractive index detector (Shodex, Tokyo, Japan).A Rezex ROA-organic acid H + (8%) 300 × 7.8 mm analytical column (Phenomenex, Torrance, CA, USA) was used, operated at 65 °C with 5 mM H 2 SO 4 and an isocratic flow of 0.6 mL/min 38 .Cellobiose levels were below 1 g/L in all samples and are not reported.Glc4gemGlc was quantified by HPAEC-PAD as described below.

Chromatographic analysis of LPMO-derived products by HPAEC-PAD
Products generated in LPMO reactions were analyzed using high-performance anion exchange chromatography with pulsed amperometric detection (HPAEC-PAD) using a Dionex ICS-5000 system (Thermo Fisher Scientific).
The ICS-5000 was equipped with a 3 × 250 mm Dionex CarboPac PA-200 analytical column with a 3 × 50 mm guard column (Thermo Fisher Scientific), and the flow rate was 500 µL/min.Eluents (A: 0.1 M NaOH, B: 0.1 M NaOH containing 1 M NaOAc) were prepared as described previously 62 .All samples were diluted two times in distilled water (type I, 18.2 MΩ•cm) prior to analysis using either a 14-min or a 39-min gradient.The 14-min elution profile used was: 0-5 min, convex upward (Dionex curve 4) from 100% A to 90% A and 10% B; 5-8.5 min, concave upward (Dionex curve 8) from 90% A and 10% B to 100% B; 8.5-8.6 min, linear from 100% B to 100% A; 8.6-14 min, constant at 100% A (reconditioning).The 39-min elution profile has previously been described 63 .Chromeleon version 7.2.9 (Thermo Fisher Scientific) was used for instrument control and analysis.Cellobiose and cellotriose used to prepare standards for quantification of native products generated from cellopentaose by LPMO action were purchased from Megazyme.C4-oxidized standards for quantification of Glc4gemGlc, and for identification of Glc4gemGlc and Glc4gemGlc 2 in product mixtures generated from cellulosic substrates, were produced in-house as previously described 64 .

Product analysis by MALDI-TOF MS
LPMO products from selected samples were analyzed by matrix-assisted laser desorption/ionization time-offlight mass spectrometry (MALDI-TOF MS) using an Ultraflex instrument (Bruker Daltonics GmbH) with a Nitrogen 337 nm laser, as described previously 65 .Sample preparation, data collection, and analysis were performed as previously described 63 .

NMR
Samples of 15 N-labeled copper-free ScLPMO9A (100 -140 µM) were prepared in a Tris-HCl buffer (25 mM Tris-HCl pH 7.0, 25 mM NaCl, 2.3% glycerol) with 10% D 2 O (d-99.9%,Sigma-Aldrich) in 3-mm NMR Essence tubes (Bruker Labscape).NMR spectra were recorded at 25 °C on an 800 MHz Bruker Ascent Advance III HD spectrometer equipped with a 5 mm Z-gradient CP-TCI (H/C/N) cryogenic probe using TopSpin version 3.5 pl7.These analyses were carried out at the NV-NMR center node of the Norwegian NMR platform at NTNU, the Norwegian University of Science and Technology.Two-dimensional 15 N-HSQC (Heteronuclear Single Quantum Coherence) spectra in combination with three-dimensional HNCO and HNCA spectra were recorded to distinguish backbone amide pairs ( 1 H-15 N) from side chains.The NMR data were processed using Bruker TopSpin version 4.1.0and analyzed using CARA version 1.9.1.7.

Titration studies by NMR
Interactions between 15 N-labeled copper-free ScLPMO9A and cellopentaose, cellohexaose, and XG14 were investigated by recording 15 N-HSQC spectra for different substrate concentrations and measuring the chemical shift perturbation (CSP) of the signals compared with 15 N-HSQC spectra recorded in the absence of substrate.Different ligand concentrations had to be used because of differences in the K d values for the different substrates.For cellopentaose, the titration points were 0.05, 0.1, 0.2, 0.4, and 1 mM.For cellohexaose, the titration points were 0.1, 0.3, 0.5, 1.1, and 2.5 mM.For XG14, the titration points were 0.4, 0.8, 1, 1.5, 2, 4, 6, and 8 mM.XG14 is a mixture of xyloglucan oligomers, mainly comprised of oligomers with the sequence XXXGXXXG, where X denotes a glucose with a xylosyl substitution.Based on this, a molecular weight of 2108 g/mol was used in our calculations.Of note, up to three of the X units can be further galactosylated; this was not taken into account in our calculations.
The CSP was calculated using the equation 66 : where Δδ comb is the absolute change in the chemical shift given in hertz (Hz).ΔδH and ΔδN are the changes in the chemical shift of the amide proton (Hz) and amide nitrogen (Hz), respectively.A significant CSP was determined to be Δδ comb ≥ 0.025 ppm for 1 H- 15 N signals.
The affinity of copper-free 15 N-labeled ScLPMO9A towards cellopentaose, cellohexaose, and XG14 was quantified by calculating the dissociation constant (K d ).The K d was determined by plotting the Δδ comb as a function of the substrate concentration for each titration point and fitting the datapoints to the equation: where [P] and [L] are the protein and ligand concentrations, respectively, Δδ comb is the absolute change in chemical shift (in ppm), and Δδ max is Δδ comb obtained when the enzyme is substrate-saturated 67 .The K d was estimated

Analysis of the structure and sequence of ScLPMO9A
Phylogenetic analysis of the ScLPMO9A sequence, shown in Fig. 1, indicated that this enzyme clusters with LsLPMO9A and CvLPMO9A, which are C4-oxidizing LPMOs active on soluble cello-oligosaccharides, mixedlinkage β-glucan, glucomannan, and xyloglucan 32,33 .Of note, a comparative functional characterization study of LsLPMO9A and CvLPMO9A by Simmons et al. indicated that LsLPMO9A potentially has low activity on birchwood xylan, but this specificity was not detected for CvLPMO9A 33 .The third LPMO clustering with ScLP-MO9A (Fig. 1), AnLPMO9A_1602, is the only one in this cluster with a carbohydrate-binding module (belonging to CAZy family 1, CBM1), and has also been reported to cleave cellohexaose 68 .The well-studied C4-oxidizing CBM1-containing NcLPMO9C from Neurospora crassa, shown to be active on cellopentaose, cellohexaose, and to a lesser extent, cellotetraose, in addition to xyloglucan, mixed-linkage β-glucan, and glucomannan 30,65 , appears in the neighboring cluster.
A comparison of the sequences of ScLPMO9A, LsLPMO9A, CvLPMO9A, and NcLPMO9C (Fig. 2) shows that residues that make up the histidine brace (His1 and His81), as well as residues in the second coordination sphere of the copper (His146, Gln161, Tyr163), are conserved in ScLPMO9A.Interestingly, ScLPMO9A has a tryptophan (Trp202) at a solvent-exposed position where other AA9 LPMOs, including LsLPMO9A, NcLPMO9C, and CvLPMO9A, tend to have a tyrosine.Previous studies have shown that this exposed aromatic residue interacts with bound oligomeric substrates 32,69 (Fig. 3).The MSA further shows an alanine residue (Ala78 in ScLPMO9A) shared between ScLPMO9A, LsLPMO9A, and NcLPMO9C, known to be common among C4-oxidizing LPMOs, although this correlation is not absolute (e.g., CvLPMO9A has an Asp in this position 31 ).The conserved Ser residue (Ser80 in ScLPMO9A) adjacent to the second histidine of the histidine brace is prevalent in C4-oxidizing LPMOs 70 .In their study of CvLPMO9A and LsLPMO9A, Simmons et al. noted that (weak) xylan oxidation was only observed for LsLPMO9A.The authors speculated that this may be due to differences in substrate binding residues of the + 2 subsite (Asn28, His66, and Asn67 in LsLPMO9A, compared to Thr28, Arg67, and Val68 in CvLPMO9A) 33 .ScLPMO9A shares two out of three of these residues with LsLPMO9A (Asn28 and His69), but has an Asp70 in place of the Asn.
A model of ScLPMO9A made using LsLPMO9A bound to cellohexaose as a template, shown in Fig. 3, depicts a shallow groove type surface topology, similar to what has been reported for LsLPMO9A 32 .This shallow groove topology differs somewhat from the characteristically flat binding surfaces of LPMOs known for their activity on crystalline substrates 71 .Docking of a hexameric substrate by superposition with the structure of LsLPMO9A bound to cellohexaose showed that the hexamer fits well in the shallow groove of ScLPMO9A, and that binding interactions seen in LsLPMO9A appear to be preserved in ScLPMO9A.Studies of the interactions between ScLPMO9A and oligosaccharides by NMR titration experiments are described below.In order to verify that E. coli-expressed ScLPMO9A was correctly folded and copper-saturated, and to ensure it produced and consumed H 2 O 2 in a manner expected of AA9 LPMOs, we tested ScLPMO9A in assays adapted from Kittl et al. 20 and Breslmayr et al. 57 .The former assay couples H 2 O 2 -production by the LPMO (i.e., oxidase activity) to oxidation of Amplex Red by horseradish peroxidase, which can be monitored spectrophotometrically.The latter assay enables spectrophotometric detection of the formation of coerulignone resulting from    8,25 , indicating that ScLPMO9A was properly folded and contained a coordinated copper in its active site.The redox potential of ScLPMO9A was determined to be 186 ± 10 mV, which is a common, albeit rather low value for AA9 LPMOs.For comparison, using the same method, the redox potential of NcLPMO9C was determined to be 224 ± 3 mV 31 .

Mapping activity on cellulosic substrates
To begin mapping the substrate specificity of ScLPMO9A, three insoluble cellulosic substrates (Avicel, sulfitepulped spruce, and PASC) were tested.Cellopentaose was also included given the activity of ScLPMO9A homologs on cellodextrins.
HPAEC-PAD analysis of product formation after 16 h of incubation (Fig. 4) showed that ScLPMO9A, like LsLPMO9A 26,32 and NcLPMO9C 30 , is a C4-oxidizing cellulose-active LPMO, as evidenced by the reductantdependent accumulation of signals representing the C4-oxidized products Glc4gemGlc and Glc4gemGlc 2 in reactions with cellopentaose and PASC.C4-oxidized oligomers are unstable under the conditions used here and are spontaneously converted to various products including those giving the diagnostic peaks for Glc4gemGlc and Glc4gemGlc 2 and, notably, also including native oligomers lacking the C4-oxidized sugar unit (so, Glc4gemGlc 2 may be converted to Glc 2 ; see also legend to Fig. 4; 72 ).No C1-oxidized reaction products were detected for any of the substrates tested.No activity was detected in reactions with the crystalline model substrate Avicel, in contrast to what has been observed for NcLPMO9C, which is active on PASC, Avicel, and cellulose in steam-exploded spruce 30 .The activity of ScLPMO9A on sulfite-pulped spruce, with an expected crystallinity almost as high as Avicel 73 , was also low, suggesting that ScLPMO9A has a preference for amorphous cellulose, as present in PASC.The main C4-oxidized product generated from cellopentaose was the C4-oxidized dimer, and, accordingly, this reaction mixture showed much more cellotriose than cellobiose, where the latter would be the co-product in a reaction leading to the C4-oxidized trimer.This shows that the pentameric substrate preferentially binds from -3 to + 2, similar to what has been observed for NcLPMO9C.
To further investigate the activity on soluble cello-oligomers and examine possible differences between ScLP-MO9A and NcLPMO9C, reactions with cellotetraose and cellohexaose were analyzed.
Figure 5 shows clear differences between ScLPMO9A and NcLPMO9C.As expected based on previous results 30 , NcLPMO9C showed limited activity on cellotetraose (Glc 4 ), generating only minor amounts of Glc4gemGlc and native cellobiose after 16 h of incubation.ScLPMO9A, on the other hand, completely degraded cellotetraose, producing a mixture of cellobiose and Glc4gemGlc.With cellohexaose (Glc 6 ), NcLPMO9C generated primarily Glc4gemGlc and cellotetraose, and lesser amounts of cellotriose and Glc4gemGlc 2 , indicating a preference for − 4-+ 2 binding.For ScLPMO9A, productive − 3-+ 3 binding was more prominent, as shown by the relatively large amounts of native trimer and Glc4gemGlc 2 .Whether the native dimer and the oxidized dimer generated by this enzyme with cellohexaose result from − 4-+ 2 or -2-+ 4 binding cannot be determined based ).Note that the C4-oxidized products are unstable and that the products labeled as Glc4gemGlc and Glc4gemGlc 2 in fact are diagnostic derivatives of these products, while additional products resulting from tautomerization are visible as minor peaks 30,72 .Also note that the peak heights of oxidized and non-oxidized products cannot be compared due to the instability of the former, and due to different response factors.All reactions were performed with 1 µM LPMO, 2 g/L substrate (or 10 g/L for sulfite-pulped spruce), and 1 mM AscA in 50 mM BisTris-HCl pH 6.

Activity on hemicellulosic substrates
The ability of ScLPMO9A to degrade hemicellulosic substrates was assessed and compared with that of NcLP-MO9C.Both LPMOs were tested on konjac glucomannan (KGM), mixed-linkage β-glucan (MLBG), tamarind xyloglucan (TXG), and xyloglucan tetradecamer (XG14), alone or in combination with PASC, as various studies have shown that the presence of cellulose may promote LPMO activity on (presumably cellulose-bound) hemicelluloses 8,63,74,75 .Products from reactions containing LPMO, AscA, and hemicellulosic substrate, hemicellulose + PASC, or PASC were analyzed by HPAEC-PAD and, in some cases, MALDI-TOF MS.Since these are single time point measurements and since LPMOs are prone to inactivation, quantitative interpretation of the results presented below requires great care 76 .However, since a suitable substrate protects LPMOs from inactivation, it is safe to assume that major differences in product levels reflect differences in substrate specificity (a better substrate will give more products and the LPMO is less prone to inactivation).Reactions with KGM showed activity of ScLPMO9A and this activity seemed hardly affected by the presence of PASC (Fig. 6A).The product profiles of ScLPMO9A and NcLPMO9C show differences that may indicate differences in substrate-binding preferences and abilities.In particular, ScLPMO9A generates more early-eluting products (5-10 min region).It is also worth noting the substantially higher peak intensities for products generated by ScLPMO9A acting on KGM alone compared to the analogous reaction with NcLPMO9C.The data thus indicate that the two LPMOs have different affinities for glucomannan and/or that they have different cleavage pattern preferences.
Figure 6B shows that ScLPMO9A is clearly more active on MLBG than NcLPMO9C, both in reactions with MLBG alone and in reactions with MLBG and PASC.The activity difference is most pronounced in the reactions with MLBG alone, since the reaction of ScLPMO9A with a mixture of PASC and MLBG yielded less MLBGderived products than the reaction with MLBG alone.The chromatograms for the reactions with ScLPMO9A show a larger variety of products as compared to NcLPMO9C, although this may partly be a false impression due to the general difference in activity.However, one clear and striking difference stands out: when acting on MLBG alone, in contrast to NcLPMO9C, ScLPMO9A generates a relatively high amount of products eluting in the 15-19 min region, which likely are C4-oxidized glucan fragments such as Glc4gemGlc and Glc4gemGlc 2 .This indicates that ScLPMO9A has a greater ability to convert MLBG to small oligomeric products and is, thus, less inhibited by the β-(1,3) bonds in MLBG.
HPAEC-PAD chromatograms for reactions with TXG (Fig. 7A) showed clear activity of ScLPMO9A and NcLPMO9C, both in reactions with PASC/TXG and reactions with TXG alone.ScLPMO9A seemed to generate more products than NcLPMO9C, especially in reactions with only TXG.Overall, the product patterns of the two enzymes look similar and these patterns resemble those generated by previously described LPMOs (including NcLPMO9C) that act on xyloglucan and that are "substitution-sensitive, " where the latter means that they only, or primarily, cleave the glucan chain at a non-substituted glucose [77][78][79] .Still, Fig. 7A shows minor differences in the product spectra of the two LPMOs, and differences were also observed when analyzing reactions with a mixture  www.nature.com/scientificreports/ of xyloglucan tetradecamer, XG14 (Fig. 7B).Thus, the two enzymes do display different cleavage preferences when acting on xyloglucan.MALDI-TOF MS analysis of products generated in the reaction with TXG confirmed that ScLPMO9A is substitution-sensitive, since all abundant products contained a multitude of three pentoses (see Fig. S1 and its legend).For example, the mass spectrum for the various tetrameric products (Fig. S1) showed xyloglucan-derived oxidized species differing by m/z 162 and all containing 3 pentoses (m/z 132), corresponding to the oxidized non-hydrated keto species and the hydrated geminal diol species of xyloglucan fragments GXXX, GXXL, and GXLL (where G is a β-1,4-linked d-glucosyl unit, X is a glucosyl substituted with an α-1,6-linked d-xylosyl, and L corresponds to X but with a further substitution of the xylosyl with a β-1,2-linked d-galactosyl, according to standard xyloglucan nomenclature 80 ).This TXG product pattern resembles what has previously been observed for NcLPMO9C acting on xyloglucan 65,79 .If ScLPMO9A would be able to cleave next to substituted sugars, other products would also have been observed in the spectrum shown in Fig. S1, such as at m/z 951 (4 hexoses, 2 pentoses) and m/z 1539 (6 hexoses, 4 pentoses), as has indeed been observed for TXG-active LPMOs that are less substitution-sensitive 78,79 .
Screening of ScLPMO9A activity on beechwood xylan, birchwood xylan, and wheat arabinoxylan in combination with PASC, using MALDI-TOF MS for product detection, showed products identical to those observed in reactions with only PASC, while reactions with the xylan substrates alone showed no product formation.Despite Reactions contained 1 µM LPMO, 1 mM AscA, and either 2 g/L KGM or MLBG, or, in the reactions containing hemicellulosic substrate and PASC, 2 g/L of each substrate (4 g/L total substrate concentration).Reactions with PASC alone contained 4 g/L PASC.Reactions lacking LPMO contained the aforementioned concentrations of substrate and 1 mM AscA.Reactions were incubated in 50 mM BisTris-HCl pH 6.5 at 40 °C and 1000 rpm for 16 h.Key products derived from LPMO activity on PASC (Glc2, Glc3, Glc4gemGlc, Glc4gemGlc2) are indicated within dashed rectangles (note that there is a slight shift between the PASC + ScLPMO9A and the PASC + NcLPMO9C chromatograms).Control reactions in the absence of AscA did not show any product formation for either LPMO.All reactions were carried out in triplicate and gave similar product profiles.www.nature.com/scientificreports/differing from LsLPMO9A (for which weak xylan activity has been reported) in only one of the three substratebinding residues purported to contribute to xylanolytic activity, and despite this difference being minimal (Asn → Asp), ScLPMO9A did not show reductant-dependent oxidative activity towards xylan.

Synergy with cellulases
The contribution of LPMOs to the saccharification of cellulose, including cellulose in sulfite-pulped spruce, is well-documented.LPMO-containing cellulase cocktails work better under conditions that promote LPMO activity 36,56 , while addition of LPMOs improves the saccharification power of LPMO-poor cellulase cocktails 38,82 .Interestingly, saccharification reactions with sulfite-pulped spruce, under conditions previously used to reveal the impact of cellulose-active LPMOs, showed that ScLPMO9A did not boost cellulose hydrolysis by an LPMO-poor cellulase cocktail (Fig. S3A).The reaction with AscA and the LPMO did show some LPMO product formation (Fig. S3B), but the glucose production was decreased rather than increased, probably due to the lower cellulase content of this reaction.While higher than in reactions without supplemented LPMO, Glc4gemGlc product levels for the reaction with ScLPMO9A were low compared to what one would expect for a truly cellulose-active LPMO (e.g.Müller et al. 36 ) and decreased over time, which is due to product instability and indicates that LPMO activity had already stopped at the first measuring point, indicative of limited substrate availability.Considering the results described above, indicating that ScLPMO9A only acts on soluble and amorphous substrates, it is likely that these low levels of LPMO products result from action on amorphous subfractions of the material, the degradation of which does not affect overall saccharification efficiency.Reactions contained 1 µM LPMO, 1 mM AscA, and either 2 g/L TXG or XG14, or, in the reactions containing hemicellulosic substrate and PASC, 2 g/L of each substrate (4 g/L total substrate concentration).Reactions with PASC alone contained 4 g/L PASC.Reactions lacking LPMO contained the aforementioned concentrations of substrate and 1 mM AscA.Reactions were incubated in 50 mM BisTris-HCl pH 6.5 at 40 °C and 1000 rpm for 16 h.Key products derived from LPMO activity on PASC (Glc 2 , Glc 3 , Glc4gemGlc, Glc4gemGlc 2 ) are indicated within dashed rectangles (note that there is a slight shift between the PASC + ScLPMO9A and the PASC + NcLPMO9C chromatograms).Control reactions in the absence of AscA did not show any product formation for either LPMO.All reactions were carried out in triplicate and gave similar product profiles.
Vol:.( 1234567890 It is now well-established that LPMOs preferentially utilize H 2 O 2 as a co-substrate to cleave glycosidic bonds and that the resulting peroxygenase reaction is fast 10,23,25 .In fact, a recent study of LsLPMO9A, a close relative of ScLPMO9A, concluded that this enzyme is unable to cleave glycosidic bonds in the absence of H 2 O 2 26 .However, in the absence of substrate, reactions of the reduced LPMO with H 2 O 2 can lead to auto-catalytic oxidation of the enzyme 10,83,84 .To assess the ability of ScLPMO9A to productively use H 2 O 2 , we tested the effect of different initial concentrations of exogenously supplied H 2 O 2 on the activity of ScLPMO9A on PASC (Fig. 8).The formation of oxidized products was assessed by quantifying Glc4gemGlc, which can be quantified with reasonable accuracy 36,85 ; other oxidized products, in particular the other primary soluble product, Glc4gemGlc 2 , and insoluble products (oxidations remaining on the fiber), were not quantified.
Without addition of H 2 O 2 , accumulation of Glc4gemGlc happened at a rate in the order of 0.3 min −1 (estimated from the progress curve in Fig. 8).Assuming that Glc4gemGlc represents about 40% of LPMO cleavages (see below for justification), this means that the LPMO operated at a rate in the order of 0.8 min −1 .Such low rates are commonly observed for LPMOs in AscA-driven reactions 4 .Addition of H 2 O 2 led to a dramatic increase in reaction speed: at the first measuring point, after 3 min, Glc4gemGlc levels amounted to approximately 40% of the added H 2 O 2 for all three levels of inclusion.The progress curves starting at 3 min show slopes quite similar to the curve for the reaction with AscA only.This clearly shows that all H 2 O 2 was consumed after 3 min and that the rest of the reaction was AscA-driven.The fact that the levels of Glc4gemGlc after 3 min amounted to 40% of added H 2 O 2 could be taken to indicate stoichiometric conversion of H 2 O 2 , since other soluble oxidized products and insoluble oxidized products (i.e., oxidized sites remaining in the insoluble substrate) were not monitored and could very well amount to the other 60%.As a rule of thumb, one would expect 50% of oxidized products to remain the insoluble substrate in reactions with an LPMO that does not carry a CBM 86 , although this likely will vary between LPMOs.In any case, the levels of Glc4gemGlc show that a major fraction of the added H 2 O 2 is used productively, and these levels are compatible with productive conversion being close to stoichiometric.
Based on the 3 min time point the reaction with the highest H 2 O 2 concentration, 250 µM, and considering formation of Glc4gemGlc only, the enzyme operated with a rate of at least some 80 min −1 , which is two orders of magnitude higher compared to the reaction with AscA and no added H 2 O 2 .Notably, the progress curve for this reaction shows signs of enzyme inactivation, since the slope of the curve after 3 min is lower compared to the other progress curves.
Rieder et al. have shown that when supplied with H 2 O 2 and a soluble substrate, NcLPMO9C is a very efficient peroxygenase, reaching catalytic rates above 100 s −1 and with the ability to productively use large amounts of H 2 O 2 to stoichiometrically degrade the cello-oligomer substrate 25 .Rates well above 10 s −1 have also been reported for similar reactions with LsLPMO9A 25,26 .Figure 9 shows progress curves for one-minute reactions of ScLPMO9A with cellopentaose at two initial H 2 O 2 concentrations.Conversion of cellopentaose was quantified by monitoring the generation of cellobiose and cellotriose, the amounts of which are expected to be equimolar to the amounts of oxidized trimer and dimer, respectively, although this equimolarity has been questioned in a recent study 26 .When supplemented with 200 µM H 2 O 2 (Fig. 9A), near complete, seemingly almost stoichiometric conversion of H 2 O 2 was achieved within 30 s, with product levels amounting to 190 µM.The actual level of LPMO-catalyzed cleavages may be somewhat lower because, under these conditions, some of the oxidized trimer may be spontaneously converted to the native dimer 72 (this effect is more prominent at higher H 2 O 2 concentrations, as is indeed visible in Fig. 9B).Based on the first 10 s of the experiment, ScLPMO9A reached a rate of at least 11 s −1 .When the H 2 O 2 concentration was increased to 400 µM (Fig. 9B), ScLPMO9A generated slightly less than 300 µM www.nature.com/scientificreports/product in 1 min.Although initial rates appeared higher than when supplemented with 200 µM H 2 O (at least 15 s −1 ), under these conditions full conversion of H 2 O 2 was not observed, and the reaction showed signs of LPMO inactivation and/or reductant depletion.

Binding interactions by NMR
Interactions between copper-free, apo-ScLPMO9A and cellopentaose, cellohexaose, and XG14 were probed by following Chemical Shift Perturbations (CSPs) in 15 N-HSQCs during titrations of the protein with the individual oligosaccharides.Based on the results, the dissociation constants (K d ) for the three substrates were found to have the following order: cellohexaose (in the range of 0.01-0.04mM) < cellopentaose (in the range of 0.03-0.05mM) < XG14 (in the range of 1.63-3.55mM) (Figs. 10 and S4).This trend in binding affinities differs from observations made for the apo-form of NcLPMO9C, which has a higher affinity for the branched XG14 (0.420 ± 0.02 mM) than for 6 (1.1 ± 0.1 mM) 69 .Differences in binding affinities between the two LPMOs can, in part, be explained by variations in their substrate-binding surfaces: the AlphaFold model of ScLPMO9A shows a shallow groove, whereas the surface of NcLPMO9C is flatter (Fig. 11).A groove-like surface is well-compatible with binding of linear substrates, while binding of branched substrates, such as XG14, may be sterically hindered.It is conceivable that such branched substrates would bind better to NcLPMO9C, which has a flatter surface, as has indeed been observed experimentally.
In the 15 N-HSQC spectra, the 1 H-15 N signals of the residues involved in substrate binding are expected to change as the substrate concentration increases during the titration.The 15 N-HSQC spectrum can thus provide quantitative information about the number of residues involved in substrate interaction, even without a resonance assignment linking the 1 H-15 N signal to the primary structure of the protein.The largest changes in CSP values are expected for residues directly involved in substrate binding, while residues farther away from the substrate-binding surface typically display smaller changes or no change 67 .
In response to titration with cellohexaose, 54 1 H-15 N signals showed a CSP ≥ 0.025 ppm, while seven 1 H-15 N signals were significantly perturbed with a CSP ≥ 0.063 ppm, indicating that the corresponding seven residues likely have a direct interaction with cellohexaose (Fig. 10).For the interaction between cellopentaose and ScLP-MO9A, a slightly lower number (47) 1 H-15 N signals showed a CSP ≥ 0.025 ppm, while in this case nine 1 H- 15 N signals displayed a CSP ≥ 0.063 ppm.The changes in CSP showed that the interaction with XG14 involves fewer residues, with 40 1 H-15 N signals having a CSP ≥ 0.025 ppm and only five 1 H- 15 N signals showing a direct (CSP ≥ 0.063 ppm) interaction with XG14.Titration experiments by NMR with NcLPMO9C have previously demonstrated that XG14 binds to a larger area (i.e., interacting with a higher number of residues) compared to cellohexaose 69 , further underpinning differences between this enzyme and ScLPMO9A.
In addition to providing a picture of the number of residues involved in interacting with each substrate, data from 15 N-HSQC spectra also show whether the same or different residues are involved in binding the different substrates (Fig. S5).Comparison of the 1 H- 15 N signals affected by titrations with cellohexaose and cellopentaose shows that some 60% of affected signals are the same, both when looking at CSP ≥ 0.025 ppm and CSP ≥ 0.063 ppm (Fig. S5).The differences between the two substrates may be due to size differences and to the occurrence of different binding modes with similar affinities.Comparison of the with XG14 showed much larger differences, with only about 35% of the affected 1 H-15 N signals being the same when considering CSP ≥ 0.025 ppm.For residues with CSP ≥ 0.063 ppm, the difference between the cello-oligomers and XG14 was even bigger (Fig. S5).Thus, it would seem that binding of XG14 is quite different from binding of cello-oligomers, with, as shown above, the latter being the preferred substrates of ScLPMO9A.Interestingly, a 1 H- 15 N signal with a chemical shift fitting the side chain of a tryptophan residue (chemical shift of H N : 10.29 ppm, N:130.9 ppm) displayed a CSP of 0.041 ppm, 0.040 ppm, and 0.050 ppm when titrated with cellopentaose, cellohexaose, and XG14, respectively (Fig. 12).These CSPs indicate that a tryptophan is a part of the binding surface for all three substrates tested, however, without a chemical shift assignment it is not possible to determine if this is the side chain of Trp202, which, as shown in Fig. 3, likely interacts with the substrate.The favorable CH-π interactions between sugar rings and the side chain of tryptophan 87 could explain the high affinity of ScLPMO9A towards both cellopentaose and cellohexaose compared with NcLPMO9C, which carries a tyrosine at this position 69 (Fig. 12).

Concluding remarks
The data presented in this study show that ScLPMO9A is a C4-oxidizing LPMO with activity on amorphous cellulose, soluble cello-oligosaccharides, and various hemicellulose glycans, and with limited ability to contribute to the saccharification of crystalline cellulose.The complete degradation of 1 mM cellotetraose under the conditions tested is of particular interest, since activity on this substrate among cello-oligomer-degrading LPMOs tends to be low 7,30 .Further in-depth analysis of the substrate-binding residues surrounding the active site of ScLPMO9A, preferably based on crystal structures, is needed to explain the structural basis for the observed activity on cellotetraose.
Comparison of ScLPMO9A and NcLPMO9C in the degradation of glucomannan, mixed-linkage β-glucan, and xyloglucan showed that the enzymes share certain properties, such as both being substitution-sensitive in TXG degradation.The product profiles and substrate binding studies did show subtle differences indicating that ScLPMO9A prefers linear over branched substrates.To the best of our knowledge, ScLPMO9A is so far the LPMO with the highest affinity for soluble natural substrates like cellohexaose, with a K d that is an order of magnitude smaller than for NcLPMO9C 69 .During the course of this study, a study containing comparative functional data for eight fungal C4-oxidizing LPMOs, including ScLPMO9A, published 7 .All these LPMOs were expressed in the yeast Pichia pastoris and shown to be active on cello-oligomers and/or hemicellulosic glycans, albeit with seemingly different efficiencies.The conclusions of the aforementioned study regarding the presumed preferred binding mode for cellohexaose are compatible with our observations.However, remarkably, while some of the LPMOs described in this study seemed to show a substrate spectrum similar to the two LPMOs studied above, Frandsen et al. concluded that ScLPMO9A is not active on glucomannan and TXG, nor on cellotetraose.The conclusions of the present study are different.Of note, proteomics analysis (see Materials and methods) showed that the enzyme used in our study was indeed ScLPMO9A (UniProt ID D8Q364).
The discovery that LPMOs may use H 2 O 2 rather than O 2 to break down glycosidic bonds has created some controversy, but has also shown that LPMOs are faster enzymes than originally believed.Although ScLP-MO9A appears to be more sensitive to H 2 O 2 than NcLPMO9C under the conditions tested in the present study,  We also show that ScLPMO9A readily uses H 2 O 2 to degrade PASC, reaching rates in the order of at least several per second rather than about 1 min −1 .To the best of our knowledge, this is the first time that such a high LPMO activity is demonstrated on this much used amorphous cellulosic substrate.All in all, ScLPMO9A seems specifically tailored to work on amorphous and soluble substrates.Given that powerful hydrolytic cellulases co-secreted with LPMOs in natural biomass-degrading ecosystems readily degrade soluble oligosaccharides, it is unlikely that fungi have evolved LPMOs with the specialized purpose of degrading these oligomers.Thus, it is conceivable that enzymes such as ScLPMO9A play hitherto undiscovered roles in lignocellulose conversion, or perhaps even in the conversion of non-lignocellulosic substrates. https://doi.org/10.1038/s41598-023-44278-1

Figure 1 .
Figure 1.Phylogenetic tree of selected AA9 LPMOs.A multiple-sequence alignment of the catalytic domains of 47 LPMOs, including ScLPMO9A and NcLPMO9C, was performed using T-Coffee Expresso and used to create the phylogenetic tree.The colors in the tree represent LPMO regioselectivity on cellulose (blue: C4-oxidizing; red: C1-oxidizing; green: C1/C4-oxidizing; no color: regioselectivity unknown).All four AA9s grouped in the clade containing ScLPMO9A are active on soluble cello-oligosaccharides.

Figure 2 .
Figure 2. Multiple-sequence alignment (MSA) of the catalytic domains of the C4-oxidizing AA9s ScLPMO9A, LsLPMO9A, NcLPMO9C, and CvLPMO9A.Sequence identities between ScLPMO9A and the other AA9s are as follows: LsLPMO9A 61.1%, CvLPMO9A 46.6%, NcLPMO9C 45.6%.Fully conserved residues are indicated by an asterisk (*).Active site histidines are colored blue, and the conserved tyrosine (Tyr163 in ScLPMO9A) that helps shaping the copper site is colored green.Two highly conserved second sphere residues near the copper site (His146 and Gln161) are colored red, whereas a semi-conserved aromatic residue likely involved in substratebinding (Trp202; see main text) is colored purple.Pink arrows and blue rectangles above the amino acid sequences indicate predicted secondary structure elements (strands and helices, respectively) in the model of ScLPMO9A made using the structure of LsLPMO9A as a template (Fig.3).Lines above the sequences represent variable regions in AA9 LPMOs as classified by88 (L2, L3, LS, LC) and89 (Seg.1-5).

Figure 3 .
Figure 3. Structural representation of ScLPMO9A seen from the top (A) and a close-up of the substrate binding surface (B).The model was made with PHYRE2 47 using the structure of LsLPMO9A (5ACI) as a template.The copper ion coordinated in the active site is shown as an orange sphere.Secondary structure elements are shown in light blue (helices), magenta (strands), and light pink (loop regions).The side chains of the active site histidines are colored dark blue, and the side chain of the tyrosine in the proximal axial copper coordination position is colored green.The side chains of His146 and Gln161 are colored red, while the side chain of Trp202 is colored purple.A bound cellohexaose unit, with subsite labelling, is shown in yellow.See main text for more details; additional views comparing the surfaces of ScLPMO9A and NcLPMO9C are provided in the section on the study of binding interactions with NMR below.

Figure 4 .
Figure 4. HPAEC-PAD chromatograms of products generated in reactions of ScLPMO9A with cellopentaose (Glc 5 ; green), PASC (orange), sulfite-pulped spruce (blue), and Avicel (yellow).Substrate names are provided directly above chromatograms in colors corresponding to the chromatogram.Key products derived from LPMO activity (Glc 2 , Glc 3 , Glc4gemGlc, Glc4gemGlc 2 ), and Glc 5 are indicated within dashed rectangles.The black chromatogram shows a standard containing Glc4gemGlc and Glc4gemGlc 2 (cellopentaose treated with NcLPMO9C38 ).Note that the C4-oxidized products are unstable and that the products labeled as Glc4gemGlc and Glc4gemGlc 2 in fact are diagnostic derivatives of these products, while additional products resulting from tautomerization are visible as minor peaks30,72 .Also note that the peak heights of oxidized and non-oxidized products cannot be compared due to the instability of the former, and due to different response factors.All reactions were performed with 1 µM LPMO, 2 g/L substrate (or 10 g/L for sulfite-pulped spruce), and 1 mM AscA in 50 mM BisTris-HCl pH 6.5, and were incubated at 40 °C and 1000 rpm for 16 h.Control reactions lacking AscA did not show any formation of native or C4-oxidized products.All reactions were carried out in triplicate and gave identical product profiles.

Figure 6 .
Figure 6.HPAEC-PAD chromatograms for reactions with ScLPMO9A or NcLPMO9C and PASC, konjac glucomannan (KGM), mixed-linkage (β-1,3;1,4) glucan (MLBG), or mixtures of PASC and KGM or MLBG.Panel A shows PASC and KGM, while Panel B shows reactions with PASC and MLBG.Reaction set-ups are indicated directly above chromatograms in colors corresponding to the relevant chromatogram.Reactions contained 1 µM LPMO, 1 mM AscA, and either 2 g/L KGM or MLBG, or, in the reactions containing hemicellulosic substrate and PASC, 2 g/L of each substrate (4 g/L total substrate concentration).Reactions with PASC alone contained 4 g/L PASC.Reactions lacking LPMO contained the aforementioned concentrations of substrate and 1 mM AscA.Reactions were incubated in 50 mM BisTris-HCl pH 6.5 at 40 °C and 1000 rpm for 16 h.Key products derived from LPMO activity on PASC (Glc2, Glc3, Glc4gemGlc, Glc4gemGlc2) are indicated within dashed rectangles (note that there is a slight shift between the PASC + ScLPMO9A and the PASC + NcLPMO9C chromatograms).Control reactions in the absence of AscA did not show any product formation for either LPMO.All reactions were carried out in triplicate and gave similar product profiles. https://doi.org/10.1038/s41598-023-44278-1

Figure 7 .
Figure 7. HPAEC-PAD chromatograms for reactions with ScLPMO9A or NcLPMO9C and PASC, tamarind xyloglucan (TXG), xyloglucan tetradecamer (XG14), or mixtures of PASC and TXG or XG14.Panel A shows reactions with PASC and TXG, while Panel B shows reactions with PASC and XG14.Reaction setups are indicated directly above the chromatograms in corresponding to the relevant chromatogram.Reactions contained 1 µM LPMO, 1 mM AscA, and either 2 g/L TXG or XG14, or, in the reactions containing hemicellulosic substrate and PASC, 2 g/L of each substrate (4 g/L total substrate concentration).Reactions with PASC alone contained 4 g/L PASC.Reactions lacking LPMO contained the aforementioned concentrations of substrate and 1 mM AscA.Reactions were incubated in 50 mM BisTris-HCl pH 6.5 at 40 °C and 1000 rpm for 16 h.Key products derived from LPMO activity on PASC (Glc 2 , Glc 3 , Glc4gemGlc, Glc4gemGlc 2 ) are indicated within dashed rectangles (note that there is a slight shift between the PASC + ScLPMO9A and the PASC + NcLPMO9C chromatograms).Control reactions in the absence of AscA did not show any product formation for either LPMO.All reactions were carried out in triplicate and gave similar product profiles.

Figure 8 .
Figure 8.Effect of H 2 O 2 on Glc4gemGlc production by ScLPMO9A in reactions with PASC.The figure shows the production of Glc4gemGlc by 1 µM ScLPMO9A in reactions containing 2 g/L PASC, 1 mM AscA, and different initial concentrations of supplemented H 2 O 2 (0, 50, 100, or 250 µM).Reactions were performed in 50 mM Tris-HCl pH 7.5 at 45 °C and 1000 rpm.Control reactions lacking AscA did not show any formation of Glc4gemGlc.Error bars represent standard deviation between triplicates.

BFigure 9 .
Figure 9. Peroxygenase activity of ScLPMO9A acting on cellopentaose.The figure shows time courses for product formation in reactions containing 1 µM ScLPMO9A, 50 µM AscA, 1 mM cellopentaose, and 200 µM (Panel A) or 400 µM (Panel B) H 2 O 2 .Reactions were performed in 50 mM sodium acetate pH 5.0 and were incubated at 40 °C and 500 rpm.Note that cleavage of cellopentaose leads either to a dimeric or a trimeric product; for example, in panel A, at 30 s, approximately 190 µM of cellopentaose has been converted, resulting in 70 µM of dimer and 120 µM of trimer.Samples were taken at 5, 10, 30, and 60 s.Error bars represent standard deviation between triplicates.

Figure 10 .
Figure 10.Binding interactions between ScLPMO9A and cellopentaose (Glc 5 ), cellohexaose (Glc 6 ), and xyloglucan tetradecamer (XG14).Binding interactions between the substrates and the LPMO were measured by the number of backbone 1 H-15 N signals showing significant chemical shift perturbations (CSP ≥ 0.025 ppm), and perturbations corresponding to direct interactions (CSP ≥ 0.063 ppm) when titrated with the substrates.The dissociation constant (K d ) of the interactions was calculated by plotting the combined CSP (Δδ comb in ppm) against the substrate concentration (mM).Schematic presentations of Glc 5 , Glc 6 , and XG14 are placed above the plots.β-1,4-linked glucose monomers are shown as blue circles, α-1,6-linked xylosyl substitutions are shown as orange stars, and β-1,2-linked galactosyl substitutions are shown as yellow circles.Note that XG14 is a mixture of oligomers and that the number and positions of xylosyl and galactosyl substitutions are unknown.

Figure 11 .
Figure 11.Substrate-binding surface topology of ScLPMO9A and NcLPMO9C.The figure shows the substratebinding surface topology of an AlphaFold model of ScLPMO9A (Panel A) and an X-ray crystal structure of NcLPMO9C (PDB ID 4D7U; Panel B).The side chains of the copper-binding histidines and a conserved surface-exposed aromatic residue are shown as colored sticks.The black dashed line illustrates how a linear polysaccharide chain could interact with the substrate-binding surface of the LPMOs.

Figure 12 .
Figure 12.Changes in the chemical shift of a tryptophan residue (H N : 10.294, N = 130.890) upon titration with the three substrates cellopentaose (Glc 5 ), cellohexaose (Glc 6 ), and xyloglucan tetradecamer (XG14).Panel A: Overlay of 15 N-HSQC spectra recorded when ScLPMO9A was titrated with 0.05, 0.1, and 1.0 mM cellopentaose.Panel B: Overlay of 15 N-HSQC spectra recorded when ScLPMO9A was titrated with 0.1, 0.5, and 2.6 mM cellohexaose.Panel C: Overlay of 15 N-HSQC spectra recorded when ScLPMO9A was titrated with 0.4, 2.0, and 8.0 mM XG14.Arrows indicate the direction of the change in chemical shift in all three panels.
57n assay adapted from Breslmayr et al.57was used to measure H 2 O 2 consumption by the LPMOs.Reaction mixtures contained 3 µM LPMO, 1 mM 2,6-dimethoxyphenol (Sigma-Aldrich), and 100 µM H 2 O 2 in 50 mM BisTris-HCl pH 6.5, and reactions were initiated by addition of the LPMO.Reactions were incubated at 30 °C in a Varioscan LUX plate reader (Thermo Fisher Scientific), and the absorbance at 469 nm was measured every 30 s over a total time of 600 s.Control reactions containing 3 µM CuSO 4 in place of the LPMO were performed in parallel.
15ing the signals of four 1 H-15N pairs in the titration experiments with cellohexaose and XG14, and of two 1 H-15N pairs for titrations with cellopentaose, to provide a K d range.