Identification and structural characterization of a novel acetyl xylan esterase fromAspergillus oryzae

α/β-hydrolase; ferulic acid esterase; crystal structure, ESTHER, Pichia pastoris 7 8 Conflict of interest: The authors declare no conflict of interests. 9 10 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 3 -

1 Acetyl xylan esterase plays a crucial role in the degradation of xylan, the major plant 2 hemicellulose, by liberating acetic acid from the backbone polysaccharides. Acetyl xylan 3 esterase B from Aspergillus oryzae, designated AoAXEB, was biochemically and structurally 4 investigated. The AoAXEB-encoding gene with a native signal peptide was successfully 5 expressed in Pichia pastoris as an active extracellular protein. The purified recombinant 6 protein had pH and temperature optima of 8.0 and 30 °C, respectively, and was stable up to 35 7 °C. The optimal substrate for hydrolysis by purified recombinant AoAXEB among a panel of 8 α-naphthyl esters was α-naphthyl acetate. Recombinant AoAXEB catalyzes the release of 9 acetic acid from wheat arabinoxylan. The release of acetic acid from wheat arabinoxylan 10 increases synergistically with xylanase addition. No activity was detected using the methyl 11 esters of ferulic, p-coumaric, caffeic, or sinapic acids. The crystal structures of AoAXEB in the 12 apo and succinate complexes were determined at resolutions of 1.75 and 1.90 Å, respectively. 13 Although AoAXEB has been classified in the Esterase_phb family in the ESTerases and 14 alpha/beta-Hydrolase Enzymes and Relatives (ESTHER) database, its structural features partly 15 resemble those of ferulic acid esterase in the FaeC family. Phylogenetic analysis also indicated 16 that AoAXEB is located between the clades of the two families. Docking analysis provided a 17 plausible binding mode for xylotriose substrates acetylated at the 2- or 3-hydroxy position. 18 This study expands the repertoire of side chain-degrading enzymes required for complete plant 19 biomass degradation. 20 21 22 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 4 -

1 Xylans are the primary constituents of hemicelluloses and, after cellulose, are the 2 second most abundant renewable polysaccharide in nature. Xylans conventionally contain 3 heterogeneous substituents, such as arabinose, 2- and 3-O-acetyl groups, ferulic (4-hydroxy-3-4 methoxycinnamic), p-coumaric (4-hydroxycinnamic), and 4-O-methylglucuronic acids [1]. 5 Acetyl xylan esterases (AXEs, EC 3.1.1.72) hydrolyze ester linkages to release acetic acid 6 from acetylated xylans [2–4], whereas ferulic acid esterases (FAEs, EC 3.1.1.73) hydrolyze 7 ester-linked ferulic acids. Most of the carboxylesterases adopt the α/β-hydrolase fold as a 8 versatile scaffold and have a Ser-His-Asp (Glu) catalytic triad or a Ser-His catalytic dyad in 9 their active sites for the catalytic mechanism using Ser as the nucleophile [5]. The ESTerases 10 and alpha/beta-Hydrolase Enzymes and Relatives (ESTHER) database is dedicated to the α/β-11 hydrolase fold enzymes and currently classifies 248 subfamilies [6]. AXEs are classified into 12 seven families (Abhydrolase_7, Acetyl-esterase_deacetylase, Acetylxylan_esterase, 13 Antigen85c, Esterase_phb, Cutinase_like, and FaeC families) in the ESTHER database. AXEs 14 belong to nine carbohydrate esterase (CE) families (CE1–CE7, CE12, and CE16) in the 15 Carbohydrate-Active enZYmes (CAZy) database [7]. Most of the characterized enzymes 16 belong to CE1 [8], and further subfamily classifications of fungal CE1 have been proposed [9]. 17 According to the CAZy database classification, the AXEs from Aspergillus spp. belong to CE1 18 [10–13] and CE16 [14]. AXEs from Aspergillus ficuum, Aspergillus luchuensis, Aspergillus 19 oryzae, Podospora anserina, and Parastagonospora nodorum heterologously expressed in 20 Pichia pastoris have been characterized [9–12,15]. AXEs in CE2, CE3, CE6, CE12, and CE16 21 are also classified into the SGNH hydrolase subfamily of the GDSL family [16], which has a 22 conserved GDSL motif around the nucleophilic serine residue instead of the canonical GxSxG 23 motif present in other serine esterases [17]. Notably, as a fungal AXE, the crystal structure of 24 A. luchuensis AXE A (AlAXEA) that is classified into Esterase_phb and CE1 families has been 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 5 - only reported [18], whereas several crystal structures of bacterial SGNH-type AXEs are solved 1 to date. 2 In the present study, we identified a new esterase from A. oryzae that has low 3 sequence similarity to characterized AXEs. Gene AO090005000945 encodes a hypothetical 4 protein in the A. oryzae genomic database (http://www.bio.nite.go.jp/dogan/). Here, we report 5 the identification and characterization of a novel AXE, designated AoAXEB. We determined 6 the crystal structure of AoAXEB to investigate the structural basis of the classification and 7 catalysis of this enzyme. 8 9

and Discussion 10 Expression of AoaxeB gene and purification of recombinant AoAXEB 11 A comparison of the amino acid sequence of A. oryzae ORF AO090005000945, 12 provisionally named AoaxeB, with the protein database revealed higher sequence identity to 13 hypothetical proteins from Aspergillus flavus, Aspergillus clavatus, and Aspergillus glaucus 14 (Table 1). Low sequence identity was observed with characterized AXEs or FAEs belonging to 15 the FaeC and Esterase_phb families: A. oryzae AXE C (AoAXEC) [15], Talaromyces 16 funiculosus FAE A (TfFAEA; accession number CAC85738.1), and A. oryzae FAE D 17 (AoFaeD) [19] in FaeC, and A. oryzae AXE A (AoAXEA) [11] and AlAXEA [18] in 18 Esterase_phb. Despite its low sequence similarity, AO090005000945 is currently listed as a 19 putative FAE/ poly(3-hydroxybutyrate) depolymerase/AXE in the Esterase_phb family (ID: 20 aspor-q2ur69) in the ESTHER database. The CAZy database does not list this ORF within the 21 CE families. 22 A. oryzae ORF AO090005000945, including the original signal sequence, was 23 successfully engineered for protein expression in the heterologous host P. p a s t o r is. The 24 recombinant protein of AoAXEB was secreted into the medium as an active enzyme and 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 6 - purified using a two-step procedure involving anion exchange and gel filtration 1 chromatography. Purified AoAXEB before and after treatment with endo-β-N-2 acetylglucosaminidase H (Endo-H) migrated using in dodecyl-sulfate polyacrylamide gel 3 electrophoresis (SDS-PAGE) with a molecular mass of approximately 43 kDa and 34 kDa, 4 respectively (Fig. 1), suggesting that the enzyme possessed N-linked oligosaccharides. The 5 NetNglyc 1.0 server predicted two putative N-glycosylation sites (N42 and N48) in the 6 AoAXEB protein (Supplementary Table S1). 7 8 General properties of the purified AoAXEB 9 The optimum pH of purified AoAXEB was 8.0 with α-naphthyl acetate (C2) as the 10 substrate (Fig. 2A), indicating that AoAXEB had a higher optimum pH than AoAXEA and 11 AoAXEC [12,15]. The optimum temperature for AoAXEB activity was 30 °C. Thermal 12 stability studies ranging 25–50 °C were performed in 50 mM sodium phosphate buffer (pH 13 8.0) (Fig. 2B). AoAXEB was stable up to 35 °C. However, the thermal stability decreased to 14 50% of residual activity for 1 h incubation at 40 °C. The thermal stability of AoAXEB was 15 significantly lower than those of AoAXEA and AoAXEC [12,15]. 16 17 Substrate specificity 18 The hydrolytic activity of the purified recombinant AoAXEB was examined using a 19 panel of α-naphthyl esters (C2–C16) as artificial substrates. The optimal substrate for 20 AoAXEB was α-naphthyl acetate (C2), and it showed lower activity toward acyl chain 21 substrates containing three or more carbon atoms (Fig. 2C). The specific activity of AoAXEB 22 toward the C2 substrate (0.033 ± 0.005 units/mg protein) was three-fold higher than that 23 toward α-naphthyl butyrate (C4, 0.011 ± 0.001 units/mg protein). This result suggests that 24 AoAXEB shows a higher specificity for acetic acid esters than AoAXEC [15], although the 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 7 - specific activity of AoAXEB is lower than that of AoAXEC. The Km and kcat values toward the 1 C2 substrate were 0.24 ± 0.12 mM and 0.17 ± 0.03 s-1 for AoAXEB and 1.9 ± 0.4 mM and 4.5 2 ± 0.7 s-1 for AoAXEC, respectively. Therefore, the catalytic efficiency (kcat/Km) for AoAXEB 3 (0.72 s-1･mM-1) was lower than that for AoAXEC (2.4 s-1･mM-1). Methyl ferulate, methyl p-4 coumarate, methyl caffeate, and methyl sinapate were not hydrolyzed, indicating that 5 AoAXEB did not exhibit FAE activity. 6 7 Release of acetic acid from acetylated xylan and synergism with xylanase 8 When AoAXEB was incubated with wheat arabinoxylan at 37 °C for 6 h, the amount 9 of released acetic acid (0.37 ± 0.04 mg/g substrate) increased by 1.85-fold compared to the 10 enzyme-free condition (0.20 ± 0.04 mg/g) (Fig. 2D). A combination of AoAXEB and 11 Thermomyces lanuginosus xylanase (TlXyn) considerably released acetic acid (3.48 ± 0.16 12 mg/g substrate) at 37 °C for 4 h. Thus, the amount of acetic acid released during a shorter 13 incubation with the xylanase was 9.5-fold the amount released during the incubation with 14 AoAXEB alone. These observations suggest that AoAXEB deacetylates xylooligomers instead 15 of xylan polymers. In contrast, AoAXEC acts directly on the acetylated xylan polymer [15]. 16 However, no significant synergistic effect of AoAXEB was observed in the degradation of 17 wheat arabinoxylan by TlXyn (Supplementary Fig. S1). In contrast, the release of acetic acid 18 from xylan increased synergistically upon the addition of xylanase to AoAXEA [12]. 19 20 Crystal structure 21 The crystal structures of recombinant AoAXEB were determined in apo and succinate 22 complex forms at 1.75 and 1.90 Å resolutions, respectively ( Table 2). The crystals of both 23 structures contained one molecule of AoAXEB in the asymmetric unit . The PISA server [20] 24 predicted that the biological assembly in the solution was a monomer. Fig. 3A shows the overall 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 8 - structure of the succinate complex form. A clear electron density map of the succinate molecule 1 was observed at the active site ( Fig. 3B), where the apo structure contained water molecules 2 (Supplementary Fig. S2 ). The succinate molecule may have been derived from trace 3 contamination in the acetate reagent used for crystallography or the fermentation product of P. 4 pastoris used for recombinant protein production because we did not use succinate in the 5 crystallization solution. An acetate molecule was observed on the protein surface because 6 sodium acetate buffer (100 mM) was used in the crystallization solution (Fig. 3A). The apo- and 7 succinate-complex structures are similar, although there is a structural difference far from the 8 active site (Supplementary Figs. S3A and B). The root mean square deviations (RMSD) between 9 the two structures was 0.13 Å for 253 over 298 Cα atoms, and there was no structural difference 10 near the active site . The structure predicted using A lphaFold was close r to the apo form 11 (Supplementary Fig. S3C). 12 AoAXEB consists of a single domain adopting a typical α/β-hydrolase fold, in which a 13 10-stranded β-sheet is sandwiched by several helices (Fig. 3A). There were three disulfide bonds 14 (C24-C137, C160-C196, and C187-C227). Because we used a protein sample treated with Endo-15 H, a single N-acetylglucosamine was observed at the N-glycosylation sites. The apo-structure 16 was N-glycosylated at N42, N248, and N295 (Supplementary Fig. S3A), whereas the succinate 17 complex contains two N-linked glycans at N42 and N295 (Fig. 3A and Supplementary Fig. S3B). 18 This observation partly contradict s the prediction by NetN glyc ( Supplementary Table S1 ), 19 where N42 and N248 are likely N-glycosylation sites, and N295 and N321 are not. 20 The catalytic triad is located at the center of the molecule and consists of S149, H282, 21 and D213 ( Fig. 3B). The succinate molecule was bound to the active site through numerous 22 hydrophilic interactions. One carboxy group forms direct hydrogen bonds with the side chains 23 of K78, K148, and H282 and a water -mediated hydrogen bond with D292. The other carboxy 24 group forms direct hydrogen bonds with the main chain amides of R76 and N150 and water -25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 9 - mediated hydrogen bonds with D113, Y179, and R224. The interaction with the main chain 1 amides of R76 and N150 corresponds to an oxyanion hole, which plays a key role in the catalysis 2 of α/β-hydrolases (discussed below). 3 4 Structural comparison 5 A Dali structural similarity search revealed that AoAXEB was most similar to 6 Lihuaxuella thermophila poly[(R)-3-hydroxybutyric acid] depolymerase (LtPHBase), which 7 belongs to the Esterase_phb family (Table 3 ) [21]. The second hit was Acremonium 8 alcalophilum FAE D (AaFaeD), which belongs to the FaeC family [22]. The third and fourth 9 hits were Aspergillus sydowii FAE E (AsFaeE) [23] and AlAXEA, respectively [18], both of 10 which belong to Esterase_phb. The fifth hit was a thermostable esterase from Thermotoga 11 maritima (TmEstA) belonging to the 5_AlphaBeta_hydrolase family [24], which showed 12 significantly lower structural similarity than the four proteins listed above. The overall structures 13 of AoAXEB and its five structural homologs are shown in Fig. 4. Interestingly, two of the three 14 disulfide bonds in AoAXEB (C24-C137 and C160-C196) were conserved with AaFaeD in the 15 FaeC family (Figs. 4A and C). Esterase_phb proteins have a conserved disulfide bond near the 16 catalytic site (C40-C75 in AlAXEA, Figs. 4B, D, and E) [18], which is not present in AoAXEB. 17 A disulfide bond corresponding to C160 -C196 in AoAXEB was also found in AsFaeE and 18 AlAXEA but not in the most structurally similar LtPHBase. In summary, AoAXEB has structural 19 features that are intermediate between those of the Esterase_phb and FaeC families. 20 Phylogenetic tree analysis of the amino acid sequences also indicated that AoAXEB was located 21 between two families (Fig. 5). 22 23 Active site features 24 The active sites of AoAXEB were compared with those of FAEs and AXE (Figs. 6A–25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 10 - D). The structures of AaFaeD and AsFaeE complexed with ferulic acid showed an open pocket 1 or surrounding pocket for large aromatic compounds (Figs. 6B and C). In contrast, the putative 2 active site of AoAXEB had a small pocket surrounded by R76, R224, Y179, I216, and T215 3 (Fig. 6A). The pocket size of AoAXEB is similar to that of AlAXEA (Fig. 6D). Our previous 4 study showed that the small pocket of AlAXEA surrounded by the C40-C75 disulfide bond and 5 the large aromatic side chain of W160 accommodated an acetyl group , because site-directed 6 mutants at W160 acquired FAE activity [18]. 7 The oxyanion hole is crucial for the catalytic function of serine and cysteine hydrolases 8 and plays a pivotal role in stabilizing the tetrahedral intermediate during the reaction, typically 9 through hydrogen bonding with the backbone amides of the polypeptides [25]. The oxyanion 10 hole of AoAXEB is formed by the main chain amides of R76 and N150 (Fig. 6E). This is 11 conserved between AaFaeD and AsFaeE (Figs. 6F and G) as hydrogen bond interactions with a 12 carboxylate group in the ligands are similarly formed. 13 We performed automated docking analysis to elucidate the substrate-binding mode of 14 AoAXEB. AXEs liberate acetic acid from their esters with 2- or 3-hydroxy groups on the xylan 15 backbone [4]. The deacetylase activity of AoAXEB on acetylated xylooligosaccharides in wheat 16 arabinoxylan is shown (Fig. 2D). Therefore, for the docking analysis, we used xylotriose models 17 acetylated at the 2- or 3-position of the central sugar unit. As a result, plausible binding modes 18 of the 2-acetylated xylotriose (2AcX3) and 3-acetylated xylotriose (3AcX3) were obtained (Fig. 19 7). The direction of the non-reducing to reducing ends of xylotriose was reversed in 2AcX3 and 20 3AcX3; however, the hydrogen bond interactions with K78, D292, K148, S149, T215, and R223 21 were conserved. Both the 4-hydroxy group at the non-reducing end and the 1-hydroxy group at 22 the reducing end of 2AcX3 and 3AcX3 pointed toward the solvent, suggesting that AoAXEB can 23 bind xylooligosaccharides for longer than four sugar units. The acetyl group in the central sugar 24 units of 2AcX3 and 3AcX3 penetrated the small pocket of the protein , occupying a suitable 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 11 - position under nucleophilic attack by S149 and stabilizing the tetrahedral intermediate by the 1 oxyanion hole. The pocket appears to be able to accommodate an alkyl group with several 2 carbon atoms, but is small enough to favor an acetyl group. Thus, the docking results explained 3 the substrate specificity of AoAXEB (Fig. 2C). 4 5

6 In this study, we discovered and characterized a novel AXE, AoAXEB, from one of the 7 most important industrial microorganisms, A. oryzae [26]. Although it has been listed as a 8 putative FAE belonging to Esterase_phb in the ESTHER database, AoAXEB exhibited typical 9 substrate specificity for AXE activity and synergy with xylanase in the deacetylation of wheat 10 arabinoxylan. The crystal structure of AoAXEB revealed that it partly had structural features 11 similar to those of FaeC, and phylogenetic analysis indicated that it was situated between the 12 Esterase_phb and FaeC families. Docking analysis elucidated the possible binding mode of 13 acetylated xylooligosaccharides at the active site. Our study suggests that there are still 14 unexplored genes for plant cell wall-degrading enzymes in fungal genomes. 15 16

17 Strains and culture conditions 18 A. oryzae strain RIB40 and P. pastoris strain KM71H were used for the source of the 19 gene cloning and the heterologous expression of the gene, respectively. P. p a s t o r i s 20 transformants were grown at 30 °C in 10 mL of BMGY [1% yeast extract, 2% peptone, 21 1%(v/v) glycerol, 0.00004% biotin, 1.34% yeast nitrogen base with ammonium sulfate, and 22 10%(v/v) 1M potassium phosphate (pH 6.0)] medium in a shaking incubator until the cell 23 density reached an OD600 of 4. Cells were harvested aseptically by centrifugation (3000×g, 10 24 min, 4 °C). The cells were then resuspended in 100 mL of BMMY medium (same composition 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 12 - as BMGY medium, except containing 0.5% (v/v) methanol instead of glycerol) in a 500 mL 1 flask to an OD600 of 1 to start induction. The culture was kept in a shaking incubator at 30 °C 2 for 7 d (180 rpm) with the addition of methanol (0.5 mL) once daily to maintain induction. 3 4 Cloning and expression of AoaxeB gene 5 AoaxeB was annotated as the hypothetical protein AO090005000945 in the A. oryzae 6 genomic database. The protein-coding sequence of AoaxeB was amplified using polymerase 7 chain reaction (PCR) with the forward primer 5’-8 AGGAATTCATGAAGTTTCTCTCAGTAAT-3’ (the EcoRI site is underlined), the reverse 9 primer 5’-GTTCTAGACTATCTCGCCTCGCTCTGGT-3’ (the XbaI site is underlined), and A. 10 oryzae genomic DNA as a template. The PCR products digested with EcoRI and XbaI were 11 cloned into an EcoRI–XbaI-digested pPICZB expression vector (Invitrogen, Waltham, MA, 12 USA). The resulting construct pPICZB-AXEB was used to transform Escherichia coli DH5α, 13 and the positive clone was confirmed using colony PCR. To splice an intron contained within 14 the gene, PCR was performed by inverse PCR using forward primer 5’-15 AAAGAATGGCAAGGAGACCCA-3’ and reverse primer 5’-16 GTTGAGCCCCTGCGGGTACAC-3’ with the KOD-Plus-Mutagenesis Kit (TOYOBO, Otsu, 17 Japan). Gene splicing was verified using DNA sequencing. This procedure yielded an 18 expression plasmid vector containing the AoaxeB gene with its native signal sequence under 19 the control of the alcohol oxidase 1 (aox1) promoter and terminator for expression in P. 20 pastoris as described above. 21 22 Purification of recombinant AoAXEB 23 The recombinant plasmid (pPICZB-AXEB) was linearized with PmeI and 24 subsequently transformed into P. pastoris KH71H using the Pichia EasyComp Transformation 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 13 - Kit (Invitrogen) according to the manufacturer’s protocol. The 7-d cell culture was harvested 1 by centrifugation (5000×g, 15 min). The culture supernatant was used for purification of 2 recombinant AoAXEB. Enzyme purification was performed as previously described [15]. 3 Enzyme purity and molecular mass were evaluated using 12% SDS-PAGE, followed by 4 Coomassie Brilliant Blue staining. The protein concentration was measured using a Micro 5 BCA Protein Assay Kit (Thermo Fisher Scientific Inc., USA). After boiling, the purified 6 AoAXEB protein was treated with 0.5 mU of Endo-H (FUJIFILM Wako Pure Chemical 7 Industries, Osaka, Japan) in 50 mM sodium acetate buffer (pH 5.0) at 37 °C for 18 h. 8 9 Assay of enzyme activity 10 The activity of purified AoAXEB toward wheat arabinoxylan, α-naphthyl acetate 11 (C2), propionate (C3), butyrate (C4), caprylate (C8), laurate (C12), palmitate (C16), methyl 12 ferulate, methyl p-coumarate, methyl caffeate, and methyl sinapate was investigated as 13 described previously [19,27]. The activity as a function of pH for the α-naphthyl acetate 14 substrate was measured using citrate buffer (pH 5.0–6.0), phosphate buffer (pH 6.0–8.0), and 15 Tris-HCl buffer (pH 8.0–9.0) at 30 °C. The activity as a function of temperature in the range 16 30–55 °C, using increments of 5.0 °C, was performed in 50 mM sodium phosphate (pH 8.0). 17 To measure the thermal stability of AoAXEB, the enzyme was incubated at an appropriate 18 temperature for 1 h, and residual activity was determined using α-naphthyl acetate as the 19 substrate. 20 21 Release of acetic acid from acetylated xylan and synergism with xylanase 22 The activity of AoAXEB toward acetylated xylan was assayed using insoluble wheat 23 arabinoxylan (Megazyme, Braw, Wicklow, Ireland), at a final concentration of 2.0%, and in a 24 final volume of 4.0 mL at 37 °C for 6 h in 50mM phosphate buffer (pH 8.0) with purified 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 14 - protein (60 μg). The amount of acetic acid released was determined using an F-kit acetate kit 1 (Roche, Basel, Switzerland). The synergistic effect between AoAXEB and TlXyn (Sigma-2 Aldrich, St. Louis, MO, USA) was investigated by incubation of a 4 mL solution of 2.0% 3 insoluble wheat arabinoxylan (Sigma-Aldrich) in 50 mM phosphate buffer (pH 8.0) for 4 h at 4 37 °C with purified AoAXEB (60 μg) and TlXyn (10 μg). The synergistic effect was 5 determined based on the amount of acetic acid released using an acetic acid assay kit (Roche). 6 7 Crystallization and structure determination 8 The crystals of AoAXEB were obtained at 20 °C using the sitting-drop vapor 9 diffusion method by mixing equal volumes of protein and reservoir solutions. For apo-form 10 crystals, a protein solution containing 2 mg/mL AoAXEB and 10 mM xylooligosaccharides 11 (FUJIFILM Wako Pure Chemical Industries) and a reservoir solution containing 0.2 M 12 potassium fluoride and 20% PEG3350 (w/v) were used. The electron densities of the 13 xylooligosaccharides were not observed in the resulting crystal structure. For succinate 14 complex crystals, protein solution containing 3 mg/mL AoAXEB and reservoir solution 15 containing 0.1 M sodium acetate buffer (pH 4.5) and 20% (w/v) PEG3000 were used. For data 16 collection, the crystals were cryoprotected using a reservoir solution supplemented with 20% 17 (v/v) PEG200 and flash-cooled by dipping in liquid nitrogen. X-ray diffraction data were 18 collected at 100 K on the beamlines at the Photon Factory of the High Energy Accelerator 19 Research Organization (KEK, Tsukuba, Japan) and the Swiss Light Source (SLS) of the Paul 20 Scherrer Institut (PSI) (Villigen, Switzerland). The preliminary diffraction data were collected 21 at SPring-8 (Hyogo, Japan). The datasets were processed using the XDS [28] and Aimless 22 software [29]. The initial phase was obtained through molecular replacement using MORDA 23 [30] with the LC-Est1 structure (PDB ID: 3WYD, chain A) as a template. Phase improvement 24 and automated model building were achieved using PHENIX (phase and build) [31] and 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 15 - BUCCANEER software [32]. Manual model rebuilding and refinement were performed using 1 Coot (Emsley et al., 2010) and Refmac5 [33]. Polder maps were prepared using PHENIX 2 software [34]. Molecular graphic images were prepared using PyMOL (Schrödinger LLC, 3 New York, NY , USA). 4 5 Phylogenetic and docking analyses 6 Phylogenetic analysis was performed using MEGA 11.0.13 [35]. The protein 7 sequences were aligned using MUSCLE [36]. The docking study was performed using 8 AutoDock Vina 1.2.5 [37]. Using AutoDockTools, polar hydrogen atoms were added to the 9 amino acid residues, and Gasteiger charges were assigned to all atoms of the enzyme. The grid 10 map was prepared with 20 × 20 × 20 points spaced at 1.0 Å distances. The grid box was 11 centered on the carboxy carbon atom of succinate near the catalytic serine of AoAXEB. The 12 exhaustiveness value is 256. The ligand structure was docked at flexible torsion angles, 13 whereas the protein structure was fixed. Docking of 2AcX3 yielded nine binding modes with 14 estimated affinities ranging -6.958–(-6.336) kcal/mol, with the second-best result (-6.893 15 kcal/mol) being selected. Docking of 3AcX3 yielded eight binding modes with estimated 16 affinities ranging -7.299–(-6.338) kcal/mol, with the second-best result (-6.893 kcal/mol) 17 being selected. The first-ranked docking result was not catalytically competent for either 18 ligand. 19 20 Acknowledgments 21 The authors thank Dr. Takatoshi Arkakawa and Dr. Arnaud Chatonnet for their 22 valuable discussions. We also thank the staff of KEK-PF, SLS at PSI, and SPring-8 for the X-23 ray data collection. This research was in part supported by the YU-COE program of Yamagata 24 University (to TKo), JSPS-KAKENHI (19H00929 and 23H00322 to SF and 21K15025 to CY) 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 16 - and the Research Support Project for Life Science and Drug Discovery (Basis for Supporting 1 Innovative Drug Discovery and Life Science Research (BINDS)) from AMED under Grant 2 Number JP22ama121001. 3 4 Author contributions 5 TKo, SF, and YS conceived and supervised the study; CY , TKo, and SF planned 6 experiments; CY , TKa, and TKo performed experiments; TKa and TKo performed the protein 7 production, purification, and biochemical experiments; CY and SF performed protein 8 crystallography; SF performed the phylogenetic and docking analyses; and TKo and SF wrote 9 the manuscript. All authors reviewed the final version of the manuscript. 10 11 Data availability statement 12 Atomic coordinates and structure factors of the crystal structures have been deposited 13 in the Protein Data Bank under accession numbers 9J07 and 9J08. The source data are 14 provided with this paper. 15 16

17 1 Scheller HV & Ulvskov P (2010) Hemicelluloses. Annu Rev Plant Biol 61, 263–289. 18 2 Biely P (2012) Microbial carbohydrate esterases deacetylating plant polysaccharides. 19 Biotechnol Adv 30, 1575–1588. 20 3 Biely P, Westereng B, Puchart V , de Maayer P & Cowan DA (2014) Recent Progress in 21 Understanding the Mode of Action of Acetylxylan Esterases. J Appl Glycosci (1999) 61, 22 35–44. 23 4 Puchart V & Biely P (2023) Microbial xylanolytic carbohydrate esterases. Essays Biochem 67, 24 479–491. 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 17 - 5 Nardini M & Dijkstra BW (1999) α/β Hydrolase fold enzymes: the family keeps growing. 1 Curr Opin Struct Biol 9, 732–737. 2 6 Chatonnet A, Perochon M, Velluet E & Marchot P (2023) The ESTHER database on alpha/beta 3 hydrolase fold proteins - An overview of recent developments. Chem Biol Interact 383, 4 110671. 5 7 Drula E, Garron ML, Dogan S, Lombard V , Henrissat B & Terrapon N (2022) The 6 carbohydrate-active enzyme database: functions and literature. Nucleic Acids Res 50, 7 D571–D577. 8 8 Adesioye FA, Makhalanyane TP, Biely P & Cowan DA (2016) Phylogeny, classification and 9 metagenomic bioprospecting of microbial acetyl xylan esterases. Enzyme Microb Technol 10 93–94, 79–91. 11 9 Li X, Griffin K, Langeveld S, Frommhagen M, Underlin EN, Kabel MA, de Vries RP & 12 Dilokpimol A (2020) Functional validation of two fungal subfamilies in carbohydrate 13 esterase family 1 by biochemical characterization of esterases from uncharacterized 14 branches. Front Bioeng Biotechnol 8, 542687. 15 10 Chung HJ, Park SM, Kim HR, Yang MS & Kim DH (2002) Cloning the gene encoding acetyl 16 xylan esterase from Aspergillus ficuum and its expression in Pichia pastoris . Enzyme 17 Microb Technol 31, 384–391. 18 11 Koseki T, Furuse S, Iwano K, Sakai H & Matsuzawa H (1997) An Aspergillus awamori 19 acetylesterase: purification of the enzyme, and cloning and sequencing of the gene. 20 Biochemical Journal 326, 485–490. 21 12 Koseki T, Miwa Y , Akao T, Akita O & Hashizume K (2006) An Aspergillus oryzae acetyl 22 xylan esterase: Molecular cloning and characteristics of recombinant enzyme expressed in 23 Pichia pastoris. J Biotechnol 121, 381–389. 24 13 Bauer S, Vasu P, Persson S, Mort AJ & Somerville CR (2006) Development and application 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 18 - of a suite of polysaccharide -degrading enzymes for analyzing plant cell walls. Proc Natl 1 Acad Sci U S A 103, 11417–11422. 2 14 Puchart V , Agger JW, Berrin JG, Várnai A, Westereng B & Biely P (2016) Comparison of 3 fungal carbohydrate esterases of family CE16 on artificial and natural substrates. J 4 Biotechnol 233, 228–236. 5 15 Kato T, Shiono Y & Koseki T (2021) Identification and characterization of an acetyl xylan 6 esterase from Aspergillus oryzae. J Biosci Bioeng 132, 337–342. 7 16 Zhang Y , Ding HT, Jiang WX, Zhang X, Cao HY , Wang JP, Li CY , Huang F, Zhang XY , Chen 8 XL, Zhang YZ & Li PY (2021) Active site architecture of an acetyl xylan esterase indicates 9 a novel cold adaptation strategy. Journal of Biological Chemistry 297, 100841. 10 17 Akoh CC, Lee GC, Liaw YC, Huang TH & Shaw JF (2004) GDSL family of serine 11 esterases/lipases. Prog Lipid Res 43, 534–552. 12 18 Komiya D, Hori A, Ishida T, Igarashi K, Samejima M, Koseki T & Fushinobu S (2017) 13 Crystal structure and substrate specificity modification of acetyl xylan esterase from 14 Aspergillus luchuensis. Appl Environ Microbiol 83. 15 19 Koseki T, Handa H, Watanabe Y ya, Ohtsuka M & Shiono Y (2016) An unusual feruloyl 16 esterase from Aspergillus oryzae: two tryptophan residues play a crucial role for the activity. 17 J Mol Catal B Enzym 133, S560–S568. 18 20 Krissinel E & Henrick K (2007) Inference of macromolecular assemblies from crystalline 19 state. J Mol Biol 372, 774–797. 20 21 Thomas GM, Quirk S, Huard DJE & Lieberman RL (2022) Bioplastic degradation by a 21 polyhydroxybutyrate depolymerase from a thermophilic soil bacterium. Protein Science 31, 22 e4470. 23 22 Phienluphon A, Kondo K, Mikami B, Teo KSK, Saito K, Watanabe T, Nagata T & Katahira 24 M (2024) Structure-Based Characterization and Improvement of an Enzymatic Activity of 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 19 - Acremonium alcalophilum Feruloyl Esterase. ACS Sustain Chem Eng 12, 3831–3840. 1 23 Phienluphon A, Kondo K, Mikami B, Nagata T & Katahira M (2023) Structural insights into 2 the molecular mechanisms of substrate recognition and hydrolysis by feruloyl esterase 3 from Aspergillus sydowii. Int J Biol Macromol 253, 127188. 4 24 Levisson M, Sun L, Hendriks S, Swinkels P, Akveld T, Bultema JB, Barendregt A, van den 5 Heuvel RHH, Dijkstra BW, van der Oost J & Kengen SWM (2009) Crystal structure and 6 biochemical properties of a novel thermostable esterase containing an immunoglobulin -7 like domain. J Mol Biol 385, 949–962. 8 25 Berg JM, Gatto GJ, Hines J, Tymoczko JL & Stryer L (2023) 6. Enzyme Catalytic Strategies. 9 In Biochemistry 10th Edition, pp. 179–209. W. H. Freeman, San Francisco. 10 26 Ichishima E (2016) Development of enzyme technology for Aspergillus oryzae, A. sojae, and 11 A. luchuensis, the national microorganisms of Japan. Biosci Biotechnol Biochem 80, 1681–12 1692. 13 27 Faulds CB & Williamson G (1994) Purification and characterization of a ferulic acid esterase 14 (FAE-III) from Aspergillus niger : Specificity for the phenolic moiety and binding to 15 microcrystalline cellulose. Microbiology (N Y) 140, 779–787. 16 28 Kabsch W (2010) XDS. Acta Crystallogr Sect D-Struct Biol 66, 125–132. 17 29 Evans PR & Murshudov GN (2013) How good are my data and what is the resolution? Acta 18 Crystallogr Sect D-Struct Biol 69, 1204–1214. 19 30 Vagin A & Lebedev A (2015) MoRDa, an automatic molecular replacement pipeline. Acta 20 Crystallogr A Found Adv 71, s19–s19. 21 31 Liebschner D, Afonine P V ., Baker ML, Bunkoczi G, Chen VB, Croll TI, Hintze B, Hung 22 LW, Jain S, McCoy AJ, Moriarty NW, Oeffner RD, Poon BK, Prisant MG, Read RJ, 23 Richardson JS, Richardson DC, Sammito MD, Sobolev O V ., Stockwell DH, Terwilliger 24 TC, Urzhumtsev AG, Videau LL, Williams CJ & Adams PD (2019) Macromolecular 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 20 - structure determination using X -rays, neutrons and electrons: recent developments in 1 Phenix. Acta Crystallogr Sect D-Struct Biol 75, 861–877. 2 32 Cowtan K (2006) The Buccaneer software for automated model building. 1. Tracing protein 3 chains. Acta Crystallogr Sect D-Biol Crystallogr 62, 1002–1011. 4 33 Murshudov GN, Skubák P, Lebedev AA, Pannu NS, Steiner RA, Nicholls RA, Winn MD, 5 Long F & Vagin AA (2011) REFMAC5 for the refinement of macromolecular crystal 6 structures. Acta Crystallogr D Biol Crystallogr 67, 355–367. 7 34 Liebschner D, Afonine P V ., Moriarty NW, Poon BK, Sobolev O V ., Terwilliger TC & Adams 8 PD (2017) Polder maps: Improving OMIT maps by excluding bulk solvent. Acta 9 Crystallogr D Struct Biol 73, 148–157. 10 35 Tamura K, Stecher G & Kumar S (2021) MEGA11: Molecular evolutionary genetics analysis 11 version 11. Mol Biol Evol 38, 3022–3027. 12 36 Edgar RC (2004) MUSCLE: Multiple sequence alignment with high accuracy and high 13 throughput. Nucleic Acids Res 32, 1792–1797. 14 37 Eberhardt J, Santos -Martins D, Tillack AF & Forli S (2021) AutoDock Vina 1.2.0: New 15 docking methods, expanded force field, and Python bindings. J Chem Inf Model 61, 3891–16 3898. 17 38 Okano H, Hong X, Kanaya E, Angkawidjaja C & Kanaya S (2015) Structural and 18 biochemical characterization of a metagenome -derived esterase with a long N -terminal 19 extension. Protein Science 24, 93–104. 20 39 Kroon PA, Williamson G, Fish NM, Archer DB & Belshaw NJ (2000) A modular esterase 21 from Penicillium funiculosum which releases ferulic acid from plant cell walls and binds 22 crystalline cellulose contains a carbohydrate binding module. Eur J Biochem 267, 6740–23 6752. 24 40 Crepin VF, Faulds CB & Connerton IF (2003) A non-modular type B feruloyl esterase from 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 21 - Neurospora crassa exhibits concentration -dependent substrate inhibition. Biochemical 1 Journal 370, 417–427. 2 41 Gordillo F, Caputo V , Peirano A, Chavez R, Van Beeumen J, Vandenberghe I, Claeyssens M, 3 Bull P, Ravanal MC & Eyzaguirre J (2006) Penicillium purpurogenum produces a family 4 1 acetyl xylan esterase containing a carbohydrate-binding module: characterization of the 5 protein and its gene. Mycol Res 110, 1129–1139. 6 7 8 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 22 - Table 1. Fungal AXEs and FAEs showing similarity to AoAXEB. 1 Strain Accession no. Identity (%) Similarity (%) A. flavus AFL2T_00922 99.6 100 A. clavatus ACLA_065130 73.2 92.9 A. glaucus Aspgl1_0129062 62.1 93.0 A. oryzae (AoAXEC) AO090023000158 24.5 67.4 T. funiculosus (TfFAEA) AJ312296.1 24.1 61.5 A. sydowii (AsFaeE) A0A1L9T9J3 20.4 68.5 A. oryzae (AoFaeD) AO090701000884 19.4 58.2 A. oryzae (AoAXEA) AO090011000745 16.7 64.2 A. luchuensis (AlAXEA) D87681.1 6.7 50.0 Biochemically characterized enzymes are indicated by their abbreviated names in brackets. 2 3 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 23 - Table 2. Crystallographic data collection and refinement statistics. 1 AoAXEB apo AoAXEB + succinate Data collection* Beamline KEK PF BL5A SLS PSI X06DA Wavelength (Å) 1.000 1.000 Space group P212121 P212121 Unit cell (Å) a = 57.614, b = 63.529, c = 83.417 a = 57.027, b = 64.008, c = 83,324 Resolution (Å) 47.41–1.75 47.06–1.90 Total reflections 207,748 (10,264) 133,539 (2,020) Unique reflections 31,605 (1,710) 23,666 (1,011) Rmerge 0.117 (0.859) 0.086 (0.219) Rpim 0.049 (0.379) 0.037 (0.176) Mean I/σ(I) 12.1 (2.1) 13.7 (4.1) CC1/2 0.998 (0.713) 0.994 (0.715) Completeness (%) 100.0 (100.0) 96.0 (65.8) Multiplicity 6.6 (6.0) 5.6 (2.0) Wilson B-factor (Å2) 9.60 11.56 Refinement Resolution (Å) 45.45–1.75 47.10–1.90 Reflections 31,549 23,621 Rwork/ Rfree 0.164/0.199 0.165/0.199 Number of atoms Amino acids 2,318 2,307 Ions 1 (K+) 0 Ligands 52 (1 PGE, 3 NAG) 40 (1ACT, 1SIN, 2NAG) Waters 323 245 Glycosylation N42, N248, N295 N42, N295 Average B-factor (Å2) Protein 17.28 16.3 Ligands 37.25 38.23 Ions 23.5 – Waters 26.1 23.08 Clashscore 1.30 2.86 Molprobity score 0.99 1.39 Ramachandran plot (%) Favored 97.98 95.92 Allowed 1.68 3.06 Outlier 0.34 1.02 RMSD from ideal values Bond lengths (Å) 0.0149 0.0103 Bond angles (°) 1.84 1.62 PDB code 9J07 9J08 *Values in parentheses represent the highest resolution shell. 2 3 4 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 24 - Table 3. Result of DALI structural similarity search 1 Protein ESTHERa CAZyb PDB ID (chain) Z score RMSD (Å) LALIc %IDd LtPHBase Esterase_phb NL 8daj (A) 28.8 2.4 253 21 AaFaeD FaeC NL 8jh9 (A) 27.9 2.3 234 26 AsFaeE Esterase_phb CE1 8iyb (A) 24.4 2.3 226 21 AlAXEA Esterase_phb CE1 5x6s (B) 23.8 2.5 229 20 TmEstA 5_AlphaBeta_ hydrolase NL 3doh (A) 20.8 2.2 206 18 Analyzed using the DALI server (http://ekhidna2.biocenter.helsinki.fi/dali/). 2 aRank 1 family in the ESTHER database (https://bioweb.supagro.inrae.fr/ESTHER/). bFamily 3 in the CAZy database (http://www.cazy.org). NL, not listed;. cNumber of aligned residues. 4 dSequence identity. 5 6 7 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 25 - Figures 1 2 Fig. 1. Coomassie-stained SDS-PAGE of purified AoAXEB. Lane 1, molecular mass markers; 3 lane 2, purified AoAXEB; and lane 3, Endo-H-treated AoAXEB. 4 5 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 26 - 1 Fig. 2. Biochemical properties of AoAXEB. (A) Effects of pH on activity. Enzyme activity was 2 measured in citrate buffer (pH 5 ­6) (closed squares), phosphate buffer (pH 6 ­8) (closed 3 triangles), and Tris-HCl buffer (pH 8­9) (closed diamonds) at 30 °C using α-naphthyl acetate 4 as a substrate. (B) Effects of temperature on activity (closed circles) and stability (open circles). 5 Enzyme activity was measured in phosphate buffer (pH 8) at different temperatures. For the 6 stability assay, aliquots of purified AoAXEB were incubated for 1 h at different temperatures. 7 After cooling on ice, residual enzyme activity was measured at pH 8.0. (C) Substrate specificity 8 measured using different acyl chain α-naphthyl esters (C2 –C16) as chromogenic substrates. 9 Observed maximum activity was set at 100%. (D) Activity toward acetylated xylan. Release of 10 acetic acid from insoluble wheat arabinoxylan was determined as described in the Materials and 11

section. 12 (A) (C) (B) (D) Fig. 2 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 27 - 1 Fig. 3. Crystal structure of AoAXEB. (A) Overall structure. Polypeptides are shown in 2 rainbow colors. N-Glycans (blue), catalytic triad (magenta), disulfide bonds (sulfur atoms in 3 yellow, indicated with red circles), succinate (yellow), and acetate (yellow) are shown as 4 sticks. (B) Active site. A Polder map (4σ) is shown for the bound succinate. Water atoms and 5 hydrogen bonds are shown as red spheres and yellow dotted lines, respectively. Residues 6 forming direct hydrogen bonds with succinate and other active site residues are shown as 7 green and orange sticks, respectively. 8 9 Fig. 3 N42 N295 N248 acetate C160-C196 C187-C227 C24-C137 (A) (B) S149 H282 K148 D213 Y179 R224D113 R76 K78 D292 I216 T215 N150 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 28 - 1 Fig. 4. Structural comparison with other α/β-hydrolase family enzymes in the top five hits of 2 the structural similarity search. (A) AoAXEB complexed with succinate. (B) L. thermophila 3 poly[(R)-3-hydroxybutyric acid] depolymerase (LtPHBase) complexed with isopropanol (PDB 4 ID: 8DAJ). (C) A. alcalophilum FAE D (AaFaeD) complexed with ferulic acid (PDB ID: 5 8JH9). (D) A. sydowii FAE E (AsFaeE) complexed with ferulic acid (PDB ID: 8IYB). (E) A. 6 luchuensis AXE A (AlAXEA, PDB ID: 5X6S). (F) T. maritima esterase (TmEstA) complexed 7 with diethyl phosphate (PDB ID: 3DOI). The N-terminal Ig-like domain is shown 8 transparently in grey. N-Glycans (blue), catalytic triad (magenta), disulfide bonds (sulfur 9 atoms in yellow and marked with red circles), and bound ligands (yellow) are shown as sticks. 10 LtPHBase (Esterase_PHB) (A) Fig. 4 AoAXEB (Esterase_phb) AsFaeE (Esterase_phb) AlAXEA (Esterase_phb) TmEstA (5_AlphaBeta_hydrolase) AaFaeD (FaeC) (B) (C) (D) (E) (F) C160-C196 C187-C227 C24-C137 C40-75 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 29 - 1 Fig. 5. A phylogenetic tree of esterases in Esterase_phb, FaeC, and 5_AlphaBeta_hydrolase 2 families. Order of structural similarity with AoAXEB is indicated in parentheses. A maximum 3 likelihood tree with the highest log likelihood (-10132.60) is shown. This analysis involved 16 4 amino acid sequences, with a total of 581 positions in the final dataset. Bootstrap values of 200 5 replications are indicated at the branches. The scale indicates branch lengths measures in the 6 number of substitutions per site. The protein names not listed in the text and their ESTHER 7 database IDs are as follows: HoEST, Haliangium ochraceum phospholipase/carboxylesterase 8 (halo1-d0lmj0); LC-Est1, a metagenome-derived esterase (9bact-3WYDseq) [38]; PfFAEB, 9 Penicillium funiculosum FAE B (penfn-faeb) [39]; NcFAE1, Neurospora crassa Fae-1 (neucr-10 faeb) [40]; PpAXEI, Penicillium purpurogenum AXE I (penpu-AXEI) [41]; and AnFaeC, 11 Aspergillus nidulans putative FAE (emeni-faec). 12 13 HoEst TmEstA LC-Est1 AsFaeE PfFAEB NcFAE1 AoAXEA AlAXEA PpAXEI LtPHBase AoAXEB AaFaeD AoAXEC AnFaeC TfFAEA AoFaeD 99 69 100 77 100 66 83 93 58 100 89 98 77 0.50 Esterase_phb FaeC 5_AlphaBeta_ hydrolase Fig. 5 (1) (2) (4) (3) (5) (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 30 - 1 Fig. 6. Active site comparison with FAEs and an AXE. (A–D) Molecular surface 2 presentations. (A) AoAXEB complexed with succinate. (B) AaFaeD complexed with ferulic 3 acid (PDB ID: 8JH9). (C) AsFaeE complexed with ferulic acid (PDB ID: 8IYB). (D) AlAXEA 4 (PDB ID: 5X6S). The catalytic triad residues are indicated with magenta, and other residues 5 forming the active site pocket are indicated with blue (basic residues), yellow (aromatic 6 residues), green (cysteine residues forming a disulfide bond), or orange (other types of 7 residues). (E–G). The catalytic components of AoAXEB (E, green), AaFaeD (F, cyan), and 8 AsFaeE (G, orange). Bound ligands (succinate or ferulic acid in yellow), the catalytic triad 9 (magenta), and oxyanion hole are shown as sticks. Note that the catalytic serine in AaFaeD 10 and AsFaeE takes two alternative conformations, and the catalytic histidine in AsFaeE takes a 11 deviated conformation without forming a hydrogen bond with aspartate. 12 13 (A) Fig. 6 AoAXEB (Esterase_phb) AaFaeD (FaeC) (B) (C) (D) Y179R224 R76 I216 T215 I147 F120 AlAXEA (Esterase_phb) W160 C40- C75 AsFaeE (Esterase_phb) C39- C75 P158 Q163 A166 V205 (E) (F) (G) AoAXEB (Esterase_phb) AsFaeE (Esterase_phb)AaFaeD (FaeC) R76 N150 S149 S119 L50 F120 S119 C39 S120 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint - 31 - 1 Fig. 7. Possible binding modes of acetylated xylooligosaccharides in the active site of 2 AoAXEB. Results of automated docking analysis are shown as the molecular model (left) and 3 surface (right). (A) 2-Acetylated xylotriose (2AcX3, yellow sticks). (B) 3-Acetylated 4 xylotriose (3AcX3, pink sticks). Sugar units from the non-reducing to reducing ends are 5 labeled with –1, 0, and +1. The middle sugar unit labeled “0” is acetylated. 6 7 (A) Fig. 7 S149 K78 T215 R224 R223 R76 K148 0 –1 +1 D292 T215 R224 R76 S149 K78 T215 R224 R223 R76 K148 0 +1 –1 D292 T215 R224 R76 (B) 2AcX3 3AcX3 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted August 23, 2024. ; https://doi.org/10.1101/2024.08.23.609331doi: bioRxiv preprint