Skip to main content An official website of the United States government Here's how you know Here's how you know Official websites use .gov A .gov website belongs to an official government organization in the United States. Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites. Search Log in Dashboard Publications Account settings Log out Search… Search NCBI Primary site navigation Search Logged in as: Dashboard Publications Account settings Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide PERMALINK Copy As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice PeerJ . 2026 Apr 9;14:e21066. doi: 10.7717/peerj.21066 Search in PMC Search in PubMed View in NLM Catalog Add to search The C-terminal lectin-like domain modulates the substrate specificity and transglycosylation activity of rice β-galactosidase1 ( Os BGal1) Sunaree Choknud Sunaree Choknud 1 School of Chemistry, Institute of Science, Suranaree University of Technology, Nakhon Ratchasima, Thailand 2 Center for Biomolecular Structure, Function and Application, Suranaree University of Technology, Nakhon Ratchasima, Thailand Find articles by Sunaree Choknud 1, 2 , Thipwarin Rimlumduan Thipwarin Rimlumduan 3 Department of Applied Biology, Faculty of Science and Liberal Arts, Rajamangala University of Technology Isan, Nakhon Ratchasima, Thailand Find articles by Thipwarin Rimlumduan 3 , Surapoj Sanram Surapoj Sanram 4 School of Biomolecular Science and Engineering, Vidyasirimedhi Institute of Science and Technology, Rayong, Thailand Find articles by Surapoj Sanram 4 , Jenis Chiawitthayanan Jenis Chiawitthayanan 1 School of Chemistry, Institute of Science, Suranaree University of Technology, Nakhon Ratchasima, Thailand 2 Center for Biomolecular Structure, Function and Application, Suranaree University of Technology, Nakhon Ratchasima, Thailand Find articles by Jenis Chiawitthayanan 1, 2 , Wipa Suginta Wipa Suginta 4 School of Biomolecular Science and Engineering, Vidyasirimedhi Institute of Science and Technology, Rayong, Thailand Find articles by Wipa Suginta 4 , James R Ketudat Cairns James R Ketudat Cairns 1 School of Chemistry, Institute of Science, Suranaree University of Technology, Nakhon Ratchasima, Thailand 2 Center for Biomolecular Structure, Function and Application, Suranaree University of Technology, Nakhon Ratchasima, Thailand Find articles by James R Ketudat Cairns 1, 2, ✉ Editor: Rogerio Sotelo-Mundo Author information Article notes Copyright and License information 1 School of Chemistry, Institute of Science, Suranaree University of Technology, Nakhon Ratchasima, Thailand 2 Center for Biomolecular Structure, Function and Application, Suranaree University of Technology, Nakhon Ratchasima, Thailand 3 Department of Applied Biology, Faculty of Science and Liberal Arts, Rajamangala University of Technology Isan, Nakhon Ratchasima, Thailand 4 School of Biomolecular Science and Engineering, Vidyasirimedhi Institute of Science and Technology, Rayong, Thailand ✉ Corresponding author. Received 2025 Nov 11; Accepted 2026 Feb 23; Collection date 2026. ©2026 Choknud et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, reproduction and adaptation in any medium and for any purpose provided that it is properly attributed. For attribution, the original author(s), title, publication source (PeerJ) and either DOI or URL of the article must be cited. PMC Copyright notice PMCID: PMC13070324 PMID: 41978877 Abstract Background β-galactosidases (EC 3.2.1.23) are glycosidases that release nonreducing terminal β- d -galactosyl residues from saccharides or glycosides. In plants, most characterized β-galactosidases belong to glycoside hydrolase family 35 (GH35) and function in cell wall remodeling. Over half of plant GH35 enzymes contain a C-terminal lectin-like domain absent in GH35 enzymes from other kingdoms, but its function remains unclear. In this work, we investigate the role of the C-terminal lectin-like domain by the effect of its deletion in rice β-galactosidase 1 ( Os BGal1) on its enzymatic function. Methods We expressed Os BGal1 and a truncated variant lacking the C-terminal domain ( Os BGal1ΔCter) in Pichia pastoris . Both enzymes were purified to homogeneity, with Os BGal1ΔCter exhibiting a higher level of N-glycosylation than the full-length enzyme. Kinetic analysis of p NP-glycosides evaluated substrate preferences. Molecular dynamics (MD) simulations of the predicted structures were used to visualize the effect of the C-terminal lectin-like domain. Results Os BGal1 showed high activity toward p NP-β- d -galactopyranoside and negligible activity toward alternative substrates, except for p NP-β- d -fucopyranoside. In contrast, Os BGal1ΔCter exhibited a 20-fold reduction in catalytic efficiency toward p NP-β- d -galactopyranoside but acquired significant activity toward p NP-β- d -glucopyranoside and higher activity for p NP-β- d -fucopyranoside. Kinetic analysis confirmed this broadened substrate profile, with the truncated enzyme showing higher turnover on glucoside, fucoside, and xyloside substrates. Functional assays demonstrated that Os BGal1 catalyzes transglycosylation, whereas Os BGal1ΔCter lost this activity. The predicted structures of Os BGal1 showed that the C-terminal domain is more than 30 Å from the active site, and MD simulations validated these structures by demonstrating their stability. This study reveals that removal of the C-terminal lectin-like domain of Os BGal1 alters substrate specificity and reduces transglycosylation activity, which may provide a basis for engineering GH35 enzymes with customized catalytic properties. Keywords: Glycoside hydrolase, Accessory domain, Glycosidase, Pichia pastoris , Glycosylation Introduction The β-galactosidases (EC 3.2.1.23) that have been classified by the carbohydrate-active enzymes (CAZy) database to date fall into the amino acid-sequence-based glycoside hydrolase (GH) families 1, 2, 35, 42, 50, and 59 ( Lombard, Henrissat & Garron, 2024 ). β-galactosidases cleave terminal β- d -galactosyl residues from oligosaccharides, polysaccharides, and glycoconjugates. They are widely distributed across bacteria, fungi, plants, and animals, reflecting their fundamental role in carbohydrate metabolism ( Saburi et al., 2023 ; Trainotti et al., 2001 ; Morita et al., 2009 ). Characterized plant β-galactosidases belong to GH family 35 (GH35). Comparative studies have revealed that plants harbor multiple GH35 β-galactosidases, each with distinct isoforms expressed during development ( Ahn et al., 2007 ; Tanthanuch et al., 2008 ). Together, these studies underscore the conserved catalytic core of GH35 enzymes, while highlighting their functional diversification through gene expansion, differential expression, and the acquisition of accessory domains. In apples, strawberries, and tomatoes, specific isoforms correlate with fruit softening and texture ( Smith & Gross, 2000 ; Trainotti et al., 2001 ; Gwanpua et al., 2014 ). Although many studies focus on fruit ripening, β -galactosidase expression has also been correlated with asparagus softening at harvest, as well as with flower opening and senescence, consistent with these enzymes acting in cell wall modifications of agricultural importance. Based on protein sequence comparison of GH35 sequences encoded in the genomes of Arabidopsis, rice ( Oryza sativa ), and the bryophyte moss Physcomitrella patens , plant GH35 enzymes have been split into subfamilies a1, a2, a3, a4, a5, b, c1, c2, and d, ( Ahn et al., 2007 ; Tanthanuch et al., 2008 ). Among these, subfamily d is more closely related to animal β-galactosidase than to the enzymes in the other plant subfamilies. All the Arabidopsis subfamily a1 members ( At BGAL 1, 2, 3, 4, 5, and 12) have been expressed in Pichia pastoris and the proteins characterized and shown to release galactose from galactobiose, galactotriose, and cell wall pectin, but not xyloglucans ( Ahn et al., 2007 ; Gantulga et al., 2008 ; Gantulga et al., 2009 ). Similarly, At BGAL6 (also called MUM2, a subfamily c2 member) was found to modify pectin galactosyl sidechains, and its absence resulted in less seed mucilage extrusion during seed germination ( Dean et al., 2007 ; Macquet et al., 2007 ). In contrast, At BGAL10, a subfamily a5 member, was found to remove galactosyl units from xyloglucan sidechains, and its absence resulted in abnormal xyloglucan repeated units and shorter siliques and sepals ( Sampedro et al., 2012 ). Most of these Arabidopsis GH35 enzymes have been localized to the cell wall, supporting their role in cell wall development ( Chandrasekar & Ra, 2016 ). In rice, fifteen GH35 genes have been identified ( Tanthanuch et al., 2008 ), although grasses like rice have type II cell walls with less pectin and xyloglucan. Os BGal1 and Os BGal2 were both found to hydrolyze and transglycosylate β-galactosides ( Chantarangsee et al., 2007 ). Os BGal1 also had activity on β-galactobioses, with the highest activity on β-1,3-, followed by β-1,4- and β-1,6- galactobiose, and also released galactose from larch wood arabinogalactan. Given the high expression of Os BGal in rapidly expanding tissues, including 3-day-old seedling shoot, stem apical meristem, initiating panicle and flower ( Chantarangsee et al., 2007 ; Tanthanuch et al., 2008 ), this suggests a role for Os BGal1 in remodeling the cell wall in rapidly growing tissues via its action on arabinogalactans. Another rice isoenzyme, Os BGal13, was also shown to hydrolyze β-1,3-, β-1,6-, and β-1,4-galactobiose disaccharides (in order of activity) and to exhibit strong transglycosylation of β-1,4-galactobiose ( Tanthanuch et al., 2008 ). It hydrolyzed p -nitrophenyl-α- l -arabinoside with 30% of its activity on p -nitrophenyl-β- d -galactoside, suggesting it might also act on both types of monosaccharide residues in arabinogalactan, but no release of arabinose or galactose from larch wood arabinogalactan was detected. More recently, the gene encoding Os BGal9, an isoenzyme more closely related to animal β-galactosidases than other rice isoenzymes, was identified as being regulated by the stress-induced transcription factor SPOTTED-LEAF7, and its expression correlated with resistance to abiotic and biotic stress ( Hoang et al., 2023 ). Os BGal9 was active as a β-galactosidase, but no differences in galactose levels were seen in knockdown and overexpression rice lines. Nonetheless, staining of arabinogalactan in arabinogalactan proteins increased in knockout plants and decreased in overexpression plants, suggesting a role in arabinogalactan remodeling involved in stress resistance. GH35 enzymes share a conserved (β/α) 8 TIM-barrel catalytic core ( Henrissat & Davies, 1997 ), which houses two critical acidic residues that act as the nucleophile and proton donor in the double-displacement mechanism typical of retaining glycosidases ( Sinnott, 1990 ), as well as accessory β-sandwich domains, as shown in Fig. 1 . For example, fungal enzymes such as Penicillium sp. β-galactosidase contain six domains, while human and bacterial homologs typically include two or three ( Rojas et al., 2004 ; Ohto et al., 2012 ). The first structure of a plant GH35 enzyme, tomato β-galactosidase 4 (TBGal4), was solved and found to possess the conserved catalytic barrel and three β-sandwich domains (PDB: 3W5G) ( Eda et al., 2016 ). Despite their conserved catalytic machinery, GH35 enzymes exhibit functional diversity and are adapted to distinct biological contexts. Figure 1. Structural comparison of GH35 β-galactosidases from different kingdoms. Open in a new tab Top: Selected GH35 structures from fungi, human, archaea, and bacteria, including Penicillium sp. (PDB: 1TG7), Homo sapiens (PDB: 3THC), Desulfurococcus amylolyticus (PDB: 8QQH), and Xanthomonas citri (PDB: 7KMN). Bottom: the crystal structure of tomato β-galactosidase TBGal4 (PDB: 3W5G) (left), NMR structure of the isolated C-terminal lectin-like domain of Os BGal1 (PDB: 5IAZ) (center), and predicted full-length model of rice Os BGal1 (right). All the full-length protein structures share a conserved N-terminal (β/α) 8 TIM-barrel catalytic domain (blue/green) but differ in the number and arrangement of β-sandwich domains (orange/red). Of note, none of the experimental full-length GH35 structures contain a C-terminal lectin-like extension, a plant-specific feature highlighted in red in the Os BGal1 model and 5IAZ domain structure. Many plant β-galactosidases carry a unique C-terminal lectin-like domain not found in any of these structures ( Trainotti et al., 2001 ; Ahn et al., 2007 ; Tanthanuch et al., 2008 ). The lack of a complete structure, including the plant-specific C-terminal extension, leaves a gap in our understanding of the structural and functional significance of this unique feature. The C-terminal lectin-like domain structure of rice Os BGal1 has been determined by NMR ( Rimlumduan et al., 2016 ); however, the full-length structure remains unverified. Moreover, although this domain was initially speculated to play a role in carbohydrate binding ( Trainotti et al., 2001 ), no carbohydrate binding activity was detected in the isolated domain ( Rimlumduan et al., 2016 ). Os BGal1, which contains this domain, and Os BGal2, which lacks this domain, were both found to hydrolyze β-galactosides specifically but not β-glucosides ( Chantarangsee et al., 2007 ). Similarly, Arabidopsis GH35 subfamily 1a members At BGAL1 and At BGAL3 contain this C-terminal domain, while At BGAL2, At BGAL4, At BGAL5, and At BGAL12 do not, but they all were found to have similar activity ( Gantulga et al., 2009 ). However, At BGAL1 and 3 had weak activity on alkali-solubilized cell walls, which was not detected in the others. No lectin-like activity could be detected for either At BGAL1 or At BGAL3. At BGAL6 and At BGAL10 also lack the C-terminal domain, so nearly all Arabidopsis GH35 enzymes without the C-terminal domain have been characterized and are found with divergent phylogeny and functions, while only two of the ten enzymes with the domain have been characterized ( Ahn et al., 2007 ). As such, no consistent functional correlation has yet emerged for the C-terminal domain. The functional contribution of the plant β-galactosidase C-terminal lectin-like domain to enzyme function remains poorly understood. Accordingly, in this study, we examine the functional influence of the C-terminal lectin-like domain in Os BGal1. We expressed and purified both full-length Os BGal1 and a variant without the C-terminal domain ( Os BGal1ΔCter) in Pichia pastoris ( P. pastoris ) and used substrate specificity and transglycosylation assays to show how that domain shapes galactoside selectivity and catalytic versatility in GH35 β-galactosidases. Materials & Methods Chemicals The substrates p -nitrophenyl ( p NP)-β- d -galactopyranoside ( p NPGal), p NP-β- d -glucopyranoside ( p NPGlc), p NP-α-L-arabinopyranoside ( p NPAra), p NP-β- d -fucopyranoside ( p NPFuc), p NP-β- d -xylopyranoside ( p NPXyl) were acquired from Sigma-Aldrich (St. Louis, MO, USA). Methanol, ethanol, propanol, isopropanol, butanol, glycerol, and ethylene glycol were purchased from CARLO ERBA Reagents (Barcelona, Spain). Cloning of Os BGal1 and Os BGal1ΔCter A gene encoding OsBGal1 (GenBank: AAM34271.1 , gene locus Os03g0165400 ) optimized for expression in P. pastoris (DDBJ: LC898194 ) was produced by GenScript Corporation (Piscataway, NJ, USA) and provided in the pUC57 vector ( Supplementary Information ). The optimized Os BGal1 cDNA was inserted into the Pst I and Xba I sites in the pPICZαBNH8 plasmid ( Toonkool et al., 2006 ). The cDNA fragment encoding the Os BGal1ΔCter was generated by PCR using pPICZαBNH8/ OsBGal1 as a template with the OsBGal1 Δ Cter Forward primer (5′-CCGCTGCAGTTACTTACGATAAGAAAGC-3′) and Reverse primer (5′GGTCTAGATTATTTAGCCAAAGCAATCTTAG-3′). Amplification was performed by incubating at 95 °C for 5 min, followed by 30 cycles of 95 °C for 30 s, 60 °C for 30 s, and 72 °C for 4 min, then a final extension at 72 °C for 10 min using Pfu polymerase. The PCR product was ligated into the Pst I and Xba I sites in the pPICZαBNH8 plasmid then transformed into competent DH5α cells. The C-terminal domain removed in Os BGal1ΔCter corresponds to amino acid residues 734–851, comprising 118 amino acids. A schematic representation of the expression plasmids is provided in Fig. S1 . The transformants were selected on agar plates containing 25 µg/ml zeocin. The products were confirmed by DNA sequencing (Macrogen, Seoul, Korea). The pPICZαBNH8/ OsBGal1 and pPICZαBNH8/ OsBGal1 Δ Cter plasmids were then digested with Sac I and transformed into P. pastoris strain SMD1168H by electroporation. Clones were selected on agar plates containing 250 µg/ml zeocin as described in the Pichia Manual (Invitrogen, Carlsbad, Calif., USA) and screened for protein expression by a β-galactosidase activity assay, with 20 µL of culture medium, 2 mM p NP-glycosides in 50 mM sodium acetate, pH 5.0, in 140 µL reactions incubated at 30 °C for 20 min. The reactions were stopped by adding 70 µL of 2.0 M Na 2 CO 3 , and the absorbance at 405 nm of the released p -nitrophenolate ( p NP) was measured. Expression and purification of Os BGal1 and Os BGal1ΔCter The Os BGal1 and Os BGal1ΔCter expression clones with the highest β-galactosidase activity were cultured in buffered glycerol-complex medium (BMGY) for 24 hr at 28 °C while shaking at 200 rpm. The cells were collected by centrifugation at 3,000 × g for 20 min at 20 °C, then resuspended in buffered methanol-complex medium (BMMY) at 20 °C while the culture was incubated in a shaking incubator at 200 rpm. Protein expression was induced by adding 1% (v/v) methanol to the culture medium daily for 3 days. The cells were removed from the media by centrifugation, then adjusted to 7.5 with 1.0 M Na 2 PO 3 , and 200 mL of the culture medium was loaded onto a 5-mL immobilized metal-ion affinity chromatography (IMAC) column charged with Co 2+ from cobalt chloride. The column was washed with 3 column volumes (CV) of 20 mM Tris–HCl, pH 8.0, 150 mM NaCl (EQ buffer), and 3 CV of 5 mM imidazole in EQ buffer, then the proteins were eluted with 3 CV of 100 mM imidazole in EQ buffer, collecting 5-ml fractions. The fractions were assayed for β-galactosidase activity and analyzed by SDS-PAGE. Apparently pure fractions containing β-galactosidase activity were pooled and concentrated by centrifugal filtration using a 30 kDa molecular weight cutoff membrane. The proteins were purified by size-exclusion chromatography (SEC) on a Superdex S200 Increase 10/300 GL column in EQ buffer at a flow rate of 0.5 mL/min. The fractions containing β-galactosidase activity were analyzed on SDS-PAGE, and the apparently pure fractions were pooled and concentrated by centrifugal filtration, then kept at −80 °C. Substrate specificity of Os BGal1and Os BGal1ΔCter The substrate specificities were assayed with 5 ng of the enzyme and 1.0 mM p NP-glycosides in 50 mM sodium acetate, pH 5.0, at 30 °C for 20 min. The 140 µL reactions were stopped by adding 70 µL of 2.0 M Na 2 CO 3 , and the absorbance of the released pNP at 405 nm was measured. The concentration was then calculated based on a p NP standard curve. Determination of the pH and temperature optimum and stability of Os BGal1 and Os BGal1ΔCter The pH-activity profiles of Os BGal1 and Os BGal1ΔCter were determined in 140 µL reactions using p NPGal for Os BGal1 and p NPGlc for Os BGal1ΔCter at a substrate concentration of 2.0 mM in 50 mM McIlvaine citrate phosphate buffers ( McIlvaine, 1921 ) with pH values between 2.5 and 8.0 at 0.5-pH-unit increments. The reactions were incubated at 30 °C for 20 min and then stopped with 70 µL of 2 M Na 2 CO 3 . The p NP product was measured by its absorbance at 405 nm. Temperature dependence was assessed at 10−70 °C in reactions conducted in 50 mM sodium acetate, pH 5.0, for 20 min. Then, the reactions were stopped with 70 µL of 2 M Na 2 CO 3 , and the absorbance at 405 nm was measured. Enzyme stability over time was assessed by incubating samples at selected temperatures (10–60 °C) in EQ buffer for varying periods (0.5, 1, 3, 6, 12, and 24 hr). The reactions were then incubated at 30 °C for 20 min in 50 mM sodium acetate, pH 5.0. Then, the reactions were stopped by adding 70 µL of 2 M Na 2 CO 3 , and the absorbance at 405 nm was measured to quantify the p NP released. Determination of Os BGal1 and Os BGal1ΔCter enzyme kinetic parameters The kinetic parameters of Os BGal1 and Os BGal1ΔCter were evaluated from triplicate reactions of hydrolysis of p NPGal, p NPFuc, and p NPGlc. The reactions were carried out in a 140 µL volume with 7.5 ng of the enzyme in 50 mM sodium acetate, pH 5.0, at 30 °C for 20 min. Then, the reactions were stopped by adding 70 µL of 2 M Na 2 CO 3 , and the absorbance at 405 nm was measured to quantify the p NP released. The substrate concentrations used ranged from one-third to three times the apparent K m , or beyond, except for p NPFuc with Os BGal1, for which the K m was too high to reach the upper range. The K m and V max values were determined by non-linear regression analysis of Michaelis–Menten plots with Grafit 5.0 (Erithacus Software, Horley, Surrey, U.K.) and compared to parameters from linear plots of (initial velocity) −1 versus [substrate] −1 (Lineweaver-Burk plots) and [substrate] (initial velocity) −1 versus [substrate] (Hanes-Woolf plots) to verify similar values. Apparent catalytic rate constant ( k cat ) values are calculated by dividing the limiting velocity ( V max ) by the amount of enzyme protein in the reaction. Transglycosylation activity of Os BGal1 and Os BGal1ΔCter The transglycosylation activities were tested using p NPGal, p NPFuc, and p NPGlc as glycosyl donors, and methanol, ethanol, n -propanol, isopropanol, butanol, glycerol, and ethylene glycol as acceptors. The transglycosylation reactions contained 5.0 mM glycosyl donors in a 100 µL volume with 1.0 µg of the enzyme in 50 mM sodium acetate, pH 5.0, at 30 °C for 1 and 12 hr. After stopping the reactions by heating at 90 °C for 5 min, as previously described ( Choknud et al., 2023 ), they were spotted onto silica gel 60 F 254 TLC plates. The TLC plates were developed in a solvent system of butanol:acetic acid:water (2:1:1, v/v/v), then sprayed with 10% sulfuric acid in ethanol and heated at 120 °C until glycoside compounds appeared as dark spots. Statistical analysis We evaluated the significance of differences in activity between Os BGal1ΔCter and Os BGal1 by the t -test implemented in MedCalc Comparison of means (MedCalc Software Ltd. Comparison of means. https://www.medcalc.org/en/calc/comparison_of_means.php (Version 23.4.5; accessed January 3, 2026)). In the cases where activity with Os BGal1 was undetectable, the significance of the Os BGal1ΔCter was calculated with Test for one mean in the MedCalc Software. Molecular dynamics simulation The initial three-dimensional structure of Os BGaI1 was generated with SWISS-MODEL using TBGal4 (3W5G) as a template ( Waterhouse et al., 2018 ; Eda et al., 2016 ). AlphaFold 3 models were also generated, but some active-site amino acid residues were out of position for catalysis, so the SWISS-MODEL structural model was used for further work. A truncated construct lacking the C-terminus ( Os BGaI1ΔCter) was prepared by manually deleting the missing residues to investigate the role of the terminal region. Both the full-length and truncated models were submitted to the CHARMM-GUI web server for system construction with the CHARMM36m force field ( Huang et al., 2017 ; Jo et al., 2008 ). Each protein system was solvated in a TIP3P water box with periodic boundary conditions and neutralized with counter-ions ( Jorgensen et al., 1983 ). Additionally, independent systems were generated by replacing neutralizing ions with 150 mM KCl to assess ion-dependent effects. All simulations were performed using GROMACS 2023 ( Abraham et al., 2015 ). The steepest descent algorithm was applied for energy minimization until a maximum force of less than 1,000 kJ mol − 1 nm − 1 was reached ( Press et al., 2007 ). Positional restraints were applied on heavy atoms with force constants of 400 kJ mol − 1 nm − 2 for the backbone and 40 kJ mol − 1 nm − 2 for side chains ( Huang et al., 2017 ). Subsequent equilibration was carried out in two stages: (i) 125 ps of NVT dynamics with a 1 fs integration step, employing the velocity-rescaling thermostat at 303 K ( Bussi, Donadio & Parrinello, 2007 ); and (ii) 250 ps of NPT equilibration under isotropic pressure coupling with the C-rescale barostat ( τ p = 5 ps, compressibility = 4.5 × 10 − 5 bar − 1 ) at 1 bar ( Parrinello & Rahman, 1981 ). Production simulations were conducted for 50 ns with a 2 fs timestep, using the Verlet cutoff scheme ( Páll & Hess, 2013 ). Long-range electrostatic interactions were treated using the particle-mesh Ewald (PME) method ( Essmann et al., 1995 ), and van der Waals interactions were smoothly switched off between 1.0 and 1.2 nm ( Huang et al., 2017 ). Bond constraints on hydrogen atoms were applied with the LINCS algorithm ( Hess et al., 1997 ). Coordinates were recorded every 100 ps, and trajectory compression was set to 50 ps intervals. Results Predicted structure of Os BGal1 The Os BGal1 model provides a structural framework for visualizing the intact enzyme ( Fig. 1 ), enabling comparison with tomato TBGal4, the only plant GH35 β-galactosidase with a resolved X-ray crystallographic structure ( Eda et al., 2016 ). While the catalytic core is highly conserved, Os BGal1 uniquely carries a C-terminal lectin-like extension (red), which is absent from TBGal4. In the model, this domain lies against the second accessory domain of the GH35 enzyme, comprising amino acid residues S347–T412 and adopting an antiparallel β-sandwich structure composed of seven β-strands, as described in the crystal structure of tomato β-galactosidase TBGal4 ( Eda et al., 2016 ), which interfaces with the catalytic domain, with its nearest residues ∼30 Å away from the catalytic residues in the active site ( Fig. 2 ). Figure 2. Structural superposition of the template tomato TBGal4 (purple) with the SWISS-MODEL predicted full-length model of rice Os BGal1(gray) and its C-terminal lectin-like domain of Os BGal1 (red). Open in a new tab The superimposed active-site residues surrounding the galactose molecule from TBGal4 (green sticks). The active-site residues are highly conserved between the two enzymes, except Val117 and Ala119 in TBGal4 correspond to Ile119 and Gly121 in O sBGal1. Domain labels are assigned according to the crystal structure of tomato TBGal4 ( Eda et al., 2016 ). Production of Os BGal1 and Os BGal1ΔCter Recombinant Os BGal1 and Os BGal1ΔCter were successfully expressed in P. patoris and purified by IMAC followed by SEC ( Fig. 3 ). The purified Os BGal1 migrated at an apparent molecular mass of ∼90 kDa ( Fig. 3A , lane 2), whereas Os BGal1ΔCter appeared as a broad smear centered at >100 kDa ( Fig. 3B , lane 2). After IMAC followed by SEC, Os BGal1 and Os BGal1ΔCter yielded approximately 1.2 mg and 1.5 mg of purified protein, respectively, from 1-L cultures, as shown in Table 1 . Treatment with endoglycosidase H had little effect on the full-length enzyme, which migrated at a similar apparent size before and after deglycosylation, indicating a low degree of N-linked glycosylation ( Fig. 3A , lane 3). In contrast, Os BGal1ΔCter displayed a pronounced downward mobility shift, migrating at an apparent molecular mass of ∼50–55 kDa ( Fig. 3B , lane 3), which is substantially lower than its predicted molecular weight. This anomalous migration is consistent with altered SDS–PAGE behavior associated with extensive N-linked glycosylation and truncation. Following deglycosylation, Os BGal1ΔCter gradually lost solubility, precipitating during subsequent handling, and its enzymatic activity declined over time. Os BGal1 is predicted to have two glycosylation sites at Asn392 and Asn461 by the NetNGlyc program ( Gupta & Brunak, 2002 ). Figure 3. Purification of Os BGal1 (A) and Os BGal1ΔCter (B). Open in a new tab Lane 1, protein marker; Lane 2, purified protein by IMAC, followed by S200 SEC; Lane 3, purified protein after deglycosylation with endoglycosidase H. Table 1. Enzyme yields during purification of Os BGal1 and Os BGal1ΔCter. Purification step Yield Specific activity (units mg −1 ) Purification factor (fold) Recovery (%) Protein (mg) Activity (units) Os BGal1 Crude protein 9,182 21,589 2.4 1 100 IMAC 6 268 43 18.3 1.2 SEC 1.2 173 144 61.4 0.8 Os BGal1ΔCter Crude protein 8,818 22,132 2.5 1 100 IMAC 6 170 29 11.4 0.8 SEC 1.5 109 74 29.3 0.5 Open in a new tab Effect of pH and temperature on the enzyme activity of Os BGal1 and Os BGal1ΔCter The effects of pH on the enzymatic activities of Os BGal1 and Os BGal1ΔCter were determined with the p NP-glycoside as the substrate ( Fig. 4A ). Os BGal1 showed the highest activity at pH 4.0, with sharply decreasing activity above pH 5.0, whereas Os BGal1ΔCter had peak activity at approximately pH 5.0–5.5 and retained substantial residual activity across a broader pH window, maintaining more than 50% activity up to pH 8.0. Figure 4. pH and temperature effects on Os BGal1 and Os BGal1ΔCter activity. Open in a new tab (A) Activity versus pH curves for Os BGal1 and Os BGal1ΔCter. (B) Activity versus temperature curves for Os BGal1 and Os BGal1ΔCter. When assayed at varied temperatures, Os BGal1 exhibited the highest activity at 50 °C, whereas Os BGal1ΔCter showed a lower optimum of 40 °C ( Fig. 4B ). The thermal stabilities of Os BGal1 and Os BGal1ΔCter were determined by monitoring residual activity after incubation at different temperatures for various time intervals. Os BGal1 ( Fig. 5A ) maintained nearly full activity at temperatures up to 40 °C for up to 24 hr, but activity declined sharply above 45 °C, with almost complete inactivation within 30 min at 50 °C and higher. In contrast, Os BGal1ΔCter ( Fig. 5B ) retained high activity only up to 30 °C and lost 25% of its activity within one hour and over 60% by 24 hr at 40 °C. Figure 5. Thermal stability of Os BGal1 and Os BGal1Δ Cter. Open in a new tab (A) Residual activity of Os BGal1 after incubation at different temperatures for 0–24 hr. (B) Residual activity of Os BGal1ΔCter under the same conditions. Substrate specificity of Os BGal1 and Os BGal1ΔCter The glycone specificity of Os BGal1 and Os BGal1ΔCter was examined with p NP-glycosides as substrates ( Table 2 ). Os BGal1 displayed the highest activity toward p NPGal, with a specific activity of 133 µmol min −1 mg − 1 . Moderate activity was observed toward p NPFuc (22.0 µmol min −1 mg − 1 ), while only low activities were detected against p NPGlc, p NPAra, and p NPXyl. Table 2. Specific activities of Os BGal1 and Os BGal1ΔCter on p NP-glycosides. Substrate Specific activity ( µmol min −1 mg −1 ) Os BGal1 Os BGal1ΔCter p NP-β- d- galactopyranoside 133 ± 3 18.6 ± 0.5 *** p NP-β- d -glucopyranoside 1.7 ± 0.2 30.1 ± 0.5 *** p NP-β- d -fucopyranoside 22.0 ± 0.5 82.9 ± 0.9 *** p NP-β- d -xylopyranoside ND 0.33 ± 0.10 * p NP-α- l -arabinopyranoside ND 0.75 ± 0.10 ** Open in a new tab Notes. ND not detectable Significance values for Os BGal1ΔCter activities versus wildtype Os BGal1 were determined with the t -test; where wild type activity was undetectable, the test for one mean was used. * p < 0.05. ** p < 0.01. *** p < 0.0001. Os BGal1ΔCter exhibited an altered activity profile. Activity toward p NPGal was reduced 7-fold, whereas activities toward alternative substrates were elevated ( Table 2 ). The truncated enzyme showed specific activity toward p NPFuc, which was 3.8-fold higher than that of wild-type Os BGal1 and fourfold higher than that of Os BGal1ΔCter, toward p NPGal. Similarly, Os BGal1ΔCter activity toward p NPGlc was 17.6-fold higher than that of wild-type Os BGal1 and higher than its activity on p NPGal. Os BGal1ΔCter activities on p NPXyl and p NPAra were also substantially higher than those of Os BGal1, although lower than its activity on p NPGal. Enzyme kinetics of Os BGal1 and Os BGal1ΔCter The kinetic parameters of Os BGal1 and its C-terminal truncated form were determined for the hydrolysis of p NPGal, p NPGlc, and p NPFuc ( Table 3 ). For Os BGal1, kinetic parameters were determined for p NPGal and p NPFuc, whereas very low activity prevented the determination of these parameters for p NPGlc. Os BGal1 hydrolyzed p NPGal with high catalytic efficiency, characterized by a low K m value of 0.20 mM and a k cat of 122 s − 1 , resulting in a specificity constant ( k cat / K m ) of 596 mM − 1 s − 1 . For p NPFuc, Os BGal1 exhibited a K m of 28.5 mM and a k cat of 1,077 s − 1 , corresponding to a lower specificity constant of 37 mM − 1 s − 1 . Table 3. Apparent kinetic parameters for Os BGal1 and Os BGal1ΔCter hydrolysis of p NP-glycoside substrates. K m (mM) k cat (s −1 ) k cat / K m (mM −1 s −1 ) Os BGal1 p NPGal 0.20 ± 0.01 122 ± 11 596 p NPGlc ND ND ND p NPFuc 28.5 ± 2.0 1,077 ± 2 37 Os BGal1ΔCter p NPGal 3.95 ± 0.17 106 ± 0.5 27 p NPGlc 8.56 ± 0.46 286 ± 0.8 33 p NPFuc 5.78 ± 0.28 773 ± 3 133 Open in a new tab Notes. ND not detectable Values are reported with their standard errors, based on all the data. A clear shift in substrate specificity was observed for Os BGal1ΔCter. Although the enzyme retained activity on p NPGal, its K m increased to 3.95 mM, although the k cat was only slightly lower at 106 s − 1 , yielding a specificity constant of 27 mM − 1 s − 1 . For p NPGlc, the enzyme showed a K m of 8.56 mM and a k cat of 286 s − 1 , corresponding to a specificity constant of 33 mM − 1 s − 1 . For p NPFuc, the enzyme exhibited a K m of 5.78 mM and a k cat of 773 s − 1 , resulting in a specificity constant of 133 mM − 1 s − 1 . Transglycosylation activity of Os BGal1 and Os BGal1ΔCter To investigate the transglycosylation activity of Os BGal1 and Os BGal1ΔCter, p NPGal and p NPGlc were tested as sugar donors, and several alcohols were examined as acceptors, including methanol, ethanol, n -propanol, isopropanol, butanol, glycerol, and ethylene glycol. TLC analysis of the Os BGal1 reaction products revealed clear formation of new glycoside spots in the presence of acceptors, corresponding to methyl galactoside (MeGal), ethyl galactoside (EtGal), propyl galactoside (ProGal), isopropyl galactoside ( i -ProGal), butyl galactoside (BuGal), and glycerol galactoside (glycerolGal) were observed ( Fig. 6A ). Ethylene glycol also served as an efficient acceptor to produce its galactoside (ethylene glycolGal) ( Fig. 6B ). Figure 6. Transglycosylation products of Os BGal1 and the Os BGal1ΔCter analyzed by TLC. Open in a new tab (A) Os BGal1 transglycosylation using p NPGal as a donor with methanol (MeOH), ethanol (EtOH), n -propanol (ProOH), isopropanol ( i ProOH), butanol (BuOH) and glycerol as acceptors. Distinct transglycosylation products (MeGal, EtGal, ProGal, i ProGal, BuGal and glycerol-Gal) are observed. (B) Os BGal1 transglycosylation using p NPGal as a donor with ethylene glycol as acceptors. Ethylene glycol supports the formation of clear products (ethylene glycol–Gal). This reaction was first developed in a chloroform:methanol (4:1, v/v) mixture, followed by a butanol:acetic acid:water (2:1:1, v/v/v) mixture for optimal resolution. (C) Os BGal1ΔCter tested with p NPGal and p NPGlc as donors and methanol, ethanol, n -propanol, and isopropanol as acceptors. Transglycosylation products are absent or faint compared to wild-type Os BGal1. In contrast, the Os BGal1ΔCter displayed very little transglycosylation activity ( Fig. 6C ). Os BGal1ΔCter was tested with p NPGal and p NPGlc as donors in the presence of methanol, ethanol, n -propanol, and isopropanol. With both donors, only very weak product bands were detected in the methanol reaction, while no detectable transglycosylation products were observed with other acceptors. Molecular dynamics simulation The structural dynamics of Os BGal1 and Os BGal1ΔCter were examined over 100 ns of molecular dynamics simulations, which showed that both constructs remained folded and compact throughout ( Fig. 7 ). The root-mean-square deviation (RMSD) in both systems equilibrated within the first ∼20 ns. The Os BGal1 (green) stabilized at ∼0.25 nm, while Os BGal1ΔCter (maroon) maintained a similar RMSD (∼0.24 nm), indicating comparable backbone stability and the absence of significant conformational drift ( Fig. 7A ). Per-residue root mean square fluctuation (RMSF) profiles were broadly similar. Most residues fluctuated below 0.25 nm, indicating a rigid backbone in both enzymes. A distinct flexibility peak around residues 739–760 was observed in loop A, located within the C-terminal lectin-like domain, which is absent in the truncated enzyme ( Fig. 7B ). The radius of gyration ( Rg ) indicating global compactness was steady and flat over time for both systems ( Fig. 7C ). The Os BGal1 maintained a larger radius (∼2.99 nm) than Os BGal1ΔCter (∼2.84 nm), a difference of ∼0.15 nm consistent with the added domain, with no evidence of unfolding in either system. Total solvent-accessible surface area (SASA) was correspondingly higher for Os BGal1 (∼350–365 ± 10 nm 2 ) than for Os BGal1ΔCter (∼290–310 nm 2 ) ( Fig. 7D ), as expected for a larger protein. Figure 7. MD analysis of Os BGal1 (green) and Os BGal1ΔCter (maroon) over 100 ns. Open in a new tab (A) The Os BGal1 stabilized at 0.25 nm, while Os BGal1ΔCter maintained a similar RMSD but with slightly lower amplitude. (B) RMSF per residue fluctuated below 0.25 nm; a distinct flexibility peak around residues 739–760 was observed in loop A, located within the C-terminal lectin-like domain. (C) The Rg of Os BGal1 was slightly larger than that of Os BGal1ΔCter by 0.15 nm, and both traces are flat, confirming structural compactness. (D) Total SASA of Os BGal1 was greater than that of Os BGal1ΔCter by 40–55 nm 2 , consistent with the presence of the extra domain. Discussion Glycosylation of Os BGal1 and Os BGal1ΔCter Os BGal1 was expressed in P. pastoris to enable glycosylation, but this also led to differences in glycosylation between the full-length and truncated proteins. The difference in apparent SDS-PAGE patterns between Os BGal1 and Os BGal1ΔCter can be attributed to altered glycosylation. The effect of Endo H digestion on Os BGal1 suggests that glycosylation is limited in the full-length enzyme ( Fig. 3A ). In contrast, the distinct Endo H sensitivity of Os BGal1ΔCter indicates that removal of the C-terminal lectin-like domain exposes the nearby glycosylation site, resulting in hyperglycosylation ( Fig. 3B ), although the protein had an apparent molecular weight of approx. 55 kDa, suggesting it may have been degraded during Endo H digestion. Structural prediction of Os BGal1 by SWISS-MODEL placed the potential N-glycosylation site at Asn392 near the C-terminal lectin-like domain ( Fig. 8 ). In the full-length model, this site appeared partially shielded by the C-terminal domain. In contrast, the Os BGal1ΔCter model lacked the C-terminal domain, leaving Asn392 fully exposed on the protein surface. This suggests that the C-terminal lectin-like domain of Os BGal1 may block access of the glycosylation machinery to Asn392. This structural difference is consistent with the higher apparent glycosylation of Os BGal1ΔCter observed by SDS–PAGE ( Fig. 3 ). Figure 8. Comparison of the Os BGal1 and Os BGal1ΔCter structural models predicted by SWISS-MODEL. Open in a new tab (A) Predicted structure of Os BGal1. (B) Predicted structure of Os BGal1ΔCter, in which the C-terminal domain is absent. The models are colored from blue (N-terminus) to red (C-terminus). Glycosylation sites (magenta) and the predicted active site (blue) are indicated. Ribbon diagrams are shown on the left, and surface representations on the right. Coloring follows the same spectrum from N- to C-terminus. The gradual precipitation and progressive loss of activity observed in Endo H-treated Os BGal1ΔCter suggest that glycosylation contributes significantly to the stability and/or solubility of the truncated enzyme ( Kwon & Yu, 1997 ). In the absence of the C-terminal lectin-like domain, glycans may compensate for reduced structural stability or solubility ( Shental-Bechor & Levy, 2009 ). Once removed, the protein becomes increasingly unstable, eventually precipitating and losing activity. This contrasts with the full-length enzyme, where the C-terminal domain likely shields glycosylation sites and supports stability independent of extensive glycan modification ( Bretthauer & Castellino, 1999 ). On the other hand, the hyperglycosylation caused by the removal of the C-terminal domain could cause some of the effects on catalysis. The large high-mannose glycan of yeast N-glycosylation is different from the smaller, complex glycans found in plants, with only 3 mannose residues substituted with N-acetylglucosamine and xylose, along with fucose on the initial N-acetylglucosamine ( Strasser, 2016 ). Moreover, as noted above, glycosylation of Asn392 may be blocked in the full-length protein, resulting in extensive glycosylation at an unusual site in the truncation. Since the glycosylation sites are on surfaces distant from the active site and are pointed away from it, any effects would likely be indirect. Still, they cannot be ruled out as a cause of the observed change in catalytic behavior. Effect of pH and temperature on the enzyme activity of Os BGal1 and Os BGal1ΔCter The comparison between Os BGal1 and its truncated form Os BGal1ΔCter emphasizes the role of the C-terminal lectin-like domain in shaping the catalytic environment of the enzyme. The full-length enzyme exhibits a sharp activity optimum under acidic pH and elevated temperature, consistent with the physiological conditions of the plant cell wall, where plant GH35 β-galactosidases are expected to operate during cell wall loosening and remodeling ( Smith & Gross, 2000 ; Paniagua et al., 2016 ). In contrast, deletion of the C-terminal domain broadens the pH preference toward a neutral range, increasing the pH optimum by two pH units. This suggests that the C-terminal domain somehow modulates the electrostatic environment of the active site. Since the C-terminal domain contains an equal number of positively and negatively charged residues (10 Arg/Lys and 10 Asp/Glu, pI 7.0), it is unlikely to be due to a long-range electrostatic effect. Removal of the C-terminal domain decreased the temperature optimum, indicating that it contributes to the enzyme’s stability. The full-length enzyme retained activity longer at 40–45 °C, whereas Os BGal1ΔCter was rapidly inactivated under the same conditions. These findings imply that the C-terminal domain not only supports optimal catalysis under the acidic microenvironments of the plant apoplast, where it is thought to be located ( Chantarangsee et al., 2007 ), but also enhances enzyme robustness. Its removal results in a less specialized enzyme that retains activity over a broader pH range but at the expense of efficiency and thermostability. This observation supports the function of the C-terminal domain in stabilizing the catalytic core against heat-induced denaturation and in maintaining the enzyme’s functional conformation during prolonged incubation ( Petrova et al., 2021 ). Substrate specificity of Os BGal1 and Os BGal1ΔCter The C-terminal lectin-like domain of Os BGal1 contributes to substrate specificity despite being quite distant from the active site in the structural model. The full-length enzyme exhibited high activity toward p NPGal, consistent with the substrate preferences reported for plant GH35 β-galactosidases ( Ahn et al., 2007 ; Eda et al., 2016 ; Chantarangsee et al., 2007 ; Trainotti et al., 2001 ), with minimal activity toward other glycosides, except for β- d -fucoside. In contrast, Os BGal1ΔCter behaved quite differently. Its activity on p NPGal was much lower, and its specificity constant was nearly 20-fold lower, reflecting a corresponding 20-fold increase in K m . In contrast, it showed significantly higher activity on glucoside, fucoside, and xyloside, with the highest specificity for p NP-β- d -fucoside, followed by p NP-β- d -glucoside. This change means that the C-terminal domain is crucial for maintaining the enzyme’s specificity for galactosides while limiting its activity on other sugars ( Chandrasekar & Ra, 2016 ). These changes suggest that the C-terminal domain is not essential for catalysis itself, as k cat values remained comparable or even higher; however, it is critical for maintaining substrate selectivity and stabilizing galactoside binding. The absence of the C-terminal domain or the extensive glycosylation in Os BGal1ΔCter may expose or alter the substrate-binding environment, thereby reducing the constraints that favor galactose recognition while allowing alternative substrates to be hydrolyzed. Selection for β-galactosides generally relies on two components: selection for the axial 4-hydroxyl group (as opposed to the equatorial 4-hydroxyl in glucose and xylose) and positioning of the side chain in a less activating position to allow the more active galactoside to react but not less active sugars like glucose ( Quirke & Crich, 2020 ). Thus, some structural change in the interactions with the C4 hydroxyl or the C6H 2 OH side chain of the sugar should be necessary to change the glycone specificity. This suggests that either long-range interactions are conveyed through the protein structure to the active site, or the structural model has misplaced at least part of the C-terminal domain, which should be in proximity to the active site in order to change the positions of interacting groups within the glycone-binding site. Notably, both AlphaFold 3 and SWISS-MODEL predicted a similar position for the C-terminal domain. Both GH35 enzymes, with and without the C-terminal domain, have similar substrate specificities, so enzymes lacking the domain have evolved other mechanisms to maintain β-galactoside specificity. This specificity is important, since their roles in plant cell wall remodeling require them to remove only galactose. For instance, At BGAL6 removes only galactosyl side chains from rhamnogalacturonan I, thereby allowing muscilage extrusion ( Macquet et al., 2007 ). If it were to remove other types of side chains, the properties of the mucilage may be affected. Transglycosylation activity of Os BGal1 and Os BGal1ΔCter Our results show that Os BGal1 requires its C-terminal lectin-like domain to catalyze transglycosylation, whereas the Os BGal1ΔCter retained hydrolytic activity but lost nearly all sugar transfer capacity. This finding highlights a domain-dependent mechanism that distinguishes Os BGal1 from other plant GH35 enzymes. For example, tomato TBGal4 β-galactosidase and rice Os BGal2, both of which lack the C-terminal domain, display substrate specificities similar to lectin-like-domain-containing β-galactosidases ( Chantarangsee et al., 2007 ; Eda et al., 2016 ). Thus, the lectin-like extension is not universally required for catalysis within GH35. The role of the C-terminal domain may be context-dependent. In Os BGal1, it may help position acceptor substrates and limit solvent access, creating a microenvironment favorable for glycosidic bond formation. Alternatively, the extensive glycosylation seen in the truncated protein may disrupt this microenvironment. Because transglycosylation requires precise placement of the acceptor and restricted access to water, the full-length enzyme could promote sugar transfer by shielding the active site. Without this domain, the catalytic pocket is likely more solvent-exposed, favoring hydrolysis and permitting broader substrate binding. Thus, the lectin-like domain cloud appears to act as both a specificity filter and a structural shield, enabling efficient hydrolysis and transglycosylation, whereas its absence yields a more flexible but less specialized enzyme. Molecular dynamics simulation Across the simulated timescales, both Os BGal1 and Os BGal1ΔCter remained folded and compact. Os BGal1 exhibited slightly higher global RMSD, along with larger SASA and Rg , attributed to its expanded architecture resulting from the C-terminal lectin-like domain rather than instability of the catalytic barrel. The Os BGal1ΔCter was overall more compact, with lower SASA and Rg . RMSF profiles were broadly similar, consistent with stable cores in both proteins. Taken together, the MD simulations suggest that the full-length and truncated structures predicted by SWISS-MODEL are reasonable and stable, but they cannot readily account for the functional differences between the full-length and truncated forms of Os BGal1. The C-terminal domain may dampen and organize inter-domain motions, while leaving the catalytic barrel intact. Its catalytic influence is probably indirect, unless structure prediction misplaces the C-terminal domain and its actual location is near the mouth of the active site. General discussion The discovery that removing the C-terminal domain could alter the enzyme’s glycone specificity was unexpected, as it was expected to play an auxiliary role, such as helping to bind to cell wall substrates. Thus, the change in activity on small substrates that cannot extend out of the active site during hydrolysis was not anticipated. Initially, we planned to produce the enzyme with and without this domain for structural studies by X-ray crystallography, small-angle X-ray scattering (SAX), or cryogenic electron microscopy (Cryo-EM). It was expected that the enzyme might maintain its basic function, but placement of the C-terminal domain would allow us to better understand the functional implications of this domain. Although we were thus far unable to determine the structure by any of these means, we instead found a remarkable functional shift in glycone specificity and transglycosylation ability. Although we might have hypothesized a change in activity on large, complex natural substrates such as galactans and arabinogalactans, which could have extensive binding across the protein surface, the change in activity on small synthetic glycosides was surprising. The unexpected functional shift opens the possibility of engineering Os BGal1 to meet the needs of applications requiring broader substrate specificity. The Os BGalΔCter can act on additional substrates. It is possible that, instead of removal, the domain sequence could be modified to improve some properties while retaining others. Furthermore, the Os BGalΔCter could be engineered to improve stability and other properties, making it more desirable for industrial applications. Conclusions The C-terminal domain of rice GH35 β-galactosidases1 plays a key role in determining both substrate selectivity and catalytic function. Comparison between full-length and C-terminal truncated enzymes revealed that removal of this domain significantly reduces catalytic efficiency, indicating its importance in stabilizing substrate binding and facilitating turnover. In addition, the truncated enzyme loses preference for β- d -galactoside, demonstrating that the C-terminal domain contributes to selective recognition of galactoside substrates, even though plant GH35 enzymes without this domain, like TBGal4, have similar specificities as full-length Os BGal1. Furthermore, the transglycosylation activity was decreased in the absence of the C-terminal domain, suggesting its involvement in proper substrate positioning for both hydrolysis and transglycosylation reactions. Evidently, either Os BGal1 has evolved to rely on the C-terminal domain for its activity, or the hyperglycosylation that occurs upon its removal affects this activity. Our findings not only clarify the effect of the C-terminal domain on plant GH35 enzymes but also suggest that its removal could alter specificity, enabling the use of this deletion strategy in glycoengineering and plant biotechnology. Supplemental Information Supplemental Information 1. Os BGal1 and Os BGal1ΔCter genes codon-optimized for expression in Pichia pastoris . peerj-14-21066-s001.docx (135.3KB, docx) DOI: 10.7717/peerj.21066/supp-1 Supplemental Information 2. The raw measurement data for enzyme characterization and kinetics. peerj-14-21066-s002.xlsx (116.6KB, xlsx) DOI: 10.7717/peerj.21066/supp-2 Supplemental Information 3. Trajectory of the 100 ns molecular dynamics simulation of Os BGal1. The stability and compact structure of the full-length Os BGal1 during the 100 ns molecular dynamics simulation. Download video file (22MB, mp4) DOI: 10.7717/peerj.21066/supp-3 Supplemental Information 4. Trajectory of the 100 ns molecular dynamics simulation of Os BGal1 ΔCter. The stability and compact structure of Os BGal1ΔCter without the C-terminal lectin-like domain during the 100 ns molecular dynamics simulation. Download video file (21.9MB, mp4) DOI: 10.7717/peerj.21066/supp-4 Acknowledgments The authors acknowledge the use of AI-assisted language tools, including ChatGPT (OpenAI) and Grammarly, for improving English language quality, grammar, and clarity. All scientific content, interpretations, and conclusions were developed by the authors, who take full responsibility for the final manuscript. Funding Statement Funding was provided by Suranaree University of Technology (SUT). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Additional Information and Declarations Competing Interests The authors declare there are no competing interests. Author Contributions Sunaree Choknud conceived and designed the experiments, performed the experiments, analyzed the data, prepared figures and/or tables, authored or reviewed drafts of the article, and approved the final draft. Thipwarin Rimlumduan conceived and designed the experiments, performed the experiments, analyzed the data, authored or reviewed drafts of the article, and approved the final draft. Surapoj Sanram performed the experiments, analyzed the data, authored or reviewed drafts of the article, and approved the final draft. Jenis Chiawitthayanan performed the experiments, analyzed the data, authored or reviewed drafts of the article, and approved the final draft. Wipa Suginta analyzed the data, authored or reviewed drafts of the article, and approved the final draft. James R. Ketudat Cairns conceived and designed the experiments, analyzed the data, prepared figures and/or tables, authored or reviewed drafts of the article, and approved the final draft. DNA Deposition The following information was supplied regarding the deposition of DNA sequences: The sequences are available at DDBJ: LC898194 . Data Availability The following information was supplied regarding data availability: The raw data is available in the Supplemental Files . The molecular dynamics input files and PDB files are available at FigShare: - Choknud, Sunaree; Sanram, Surapoj; R. Ketudat Cairns, James (2025). Molecular dynamics simulation data for GH35 rice β-galactosidases1. figshare. Dataset. https://doi.org/10.6084/m9.figshare.30520754.v1 The sequences are available at DDBJ: LC898194 . References Abraham et al. (2015). Abraham MJ, Murtola T, Schulz R, Páll S, Smith JC, Hess B, Lindahl E. GROMACS: high performance molecular simulations through multi-level parallelism from laptops to supercomputers. SoftwareX. 2015;1–2:19–25. doi: 10.1016/j.softx.2015.06.001. [ DOI ] [ Google Scholar ] Ahn et al. (2007). Ahn YO, Zheng M, Bevan DR, Esen A, Shiu SH, Benson J, Peng HP, Miller JT, Cheng CL, Poulton JE, Shih MC. Functional genomic analysis of Arabidopsis thaliana glycoside hydrolase family 35. Phytochemistry. 2007;68:1510–1520. doi: 10.1016/j.phytochem.2007.03.020. [ DOI ] [ PubMed ] [ Google Scholar ] Bretthauer & Castellino (1999). Bretthauer RK, Castellino FJ. Glycosylation of Pichia pastoris-derived proteins. Biotechnology and Applied Biochemistry. 1999;30:193–200. doi: 10.1111/j.1470-8744.1999.tb00770.x. [ DOI ] [ PubMed ] [ Google Scholar ] Bussi, Donadio & Parrinello (2007). Bussi G, Donadio D, Parrinello M. Canonical sampling through velocity rescaling. The Journal of Chemical Physics. 2007;126:014101. doi: 10.1063/1.2408420. [ DOI ] [ PubMed ] [ Google Scholar ] Chandrasekar & Ra (2016). Chandrasekar B, Ra VDH. Beta galactosidases in Arabidopsis and tomato—a mini review. Biochemical Society Transactions. 2016;44:150–158. doi: 10.1042/BST20150217. [ DOI ] [ PubMed ] [ Google Scholar ] Chantarangsee et al. (2007). Chantarangsee M, Tanthanuch W, Fujimura T, Fry SC, Ketudat Cairns JR. Molecular characterization of β-galactosidases from germinating rice (Oryza sativa) Plant Science. 2007;173(2):118–134. doi: 10.1016/j.plantsci.2007.04.009. [ DOI ] [ Google Scholar ] Choknud et al. (2023). Choknud S, Prawisut A, Gorantla JN, Ketudat Cairns JR. Expression, purification, characterization and glycoside production potential of rice β -d-glucan glucohydrolase I (OsExoI) Process Biochemistry. 2023;127:145–154. doi: 10.1016/j.procbio.2023.02.015. [ DOI ] [ Google Scholar ] Dean et al. (2007). Dean GH, Zheng H, Tewari J, Huang J, Young DS, Hwang YT, Western TL, Carpita NC, McCann MC, Mansfield SD, Haughn GW. The Arabidopsis MUM2 gene encodes a β -galactosidase required for the production of seed coat mucilage with correct hydration properties. The Plant Cell. 2007;19:4007–4021. doi: 10.1105/tpc.107.050609. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Eda et al. (2016). Eda M, Matsumoto T, Ishimaru M, Tada T. Structural and functional analysis of tomato β -galactosidase 4: insight into the substrate specificity of the fruit softening-related enzyme. The Plant Journal. 2016;86(4):300–307. doi: 10.1111/tpj.13160. [ DOI ] [ PubMed ] [ Google Scholar ] Essmann et al. (1995). Essmann U, Perera L, Berkowitz ML, Darden T, Lee H, Pedersen LG. A smooth particle mesh Ewald method. The Journal of Chemical Physics. 1995;103:8577–8593. doi: 10.1063/1.470117. [ DOI ] [ Google Scholar ] Gantulga et al. (2009). Gantulga D, Ahn YO, Zhou C, Battogtokh D, Bevan DR, Winkel BSJ, Esen A. Comparative characterization of the Arabidopsis subfamily a1 β-galactosidases. Phytochemistry. 2009;70:1999–2009. doi: 10.1016/j.phytochem.2009.08.008. [ DOI ] [ PubMed ] [ Google Scholar ] Gantulga et al. (2008). Gantulga D, Turan Y, Bevan DR, Esen A. The Arabidopsis At 1g45130 and At 3g52840 genes encode β-galactosidases with activity toward cell wall polysaccharides. Phytochemistry. 2008;69:1661–1670. doi: 10.1016/j.phytochem.2008.01.023. [ DOI ] [ PubMed ] [ Google Scholar ] Gupta & Brunak (2002). Gupta R, Brunak S. Prediction of glycosylation across the human proteome and the correlation to protein function. Pacific Symposium on Biocomputing. 2002;2002:310–322. doi: 10.1142/9789812799623_0029. [ DOI ] [ PubMed ] [ Google Scholar ] Gwanpua et al. (2014). Gwanpua SG, Buggenhout SV, Verlinden BE, Christiaens S, Shpigelman A, Vicent V, Kermani ZJ, Nicolai BM, Hendrickx M, Geeraerd A. Pectin modifications and the role of pectin-degrading enzymes during postharvest softening of Jonagold apples. Food Chemistry. 2014;158:283–291. doi: 10.1016/j.foodchem.2014.02.138. [ DOI ] [ PubMed ] [ Google Scholar ] Henrissat & Davies (1997). Henrissat B, Davies G. Structural and sequence-based classification of glycoside hydrolases. Current Opinion in Structural Biology. 1997;7(5):637–644. doi: 10.1016/S0959-440X(97)80072-3. [ DOI ] [ PubMed ] [ Google Scholar ] Hess et al. (1997). Hess B, Bekker H, Berendsen HJC, Fraaije JGEM. LINCS: a linear constraint solver for molecular simulations. Journal of Computational Chemistry. 1997;18:1463–1472. doi: 10.1002/(SICI)1096-987X(199709)18:12<1463::AID-JCC4>3.0.CO;2-H. [ DOI ] [ Google Scholar ] Hoang et al. (2023). Hoang TV, Vo KTX, Rahman MM, Zhong R, Lee C, Ketudat Cairns JR, Ye ZH, Jeon JS. SPOTTED-LEAF7 targets the gene encoding β -galactosidase9, which functions in rice growth and stress responses. Plant Physiology. 2023;193:1109–1125. doi: 10.1093/plphys/kiad359. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Huang et al. (2017). Huang J, Rauscher S, Nawrocki G, Ran T, Feig T, Groot BL, Grubmüller H, MacKerell Jr AD. CHARMM36m: an improved force field for folded and intrinsically disordered proteins. Nature Methods. 2017;14:71–73. doi: 10.1038/nmeth.4067. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Jo et al. (2008). Jo S, Kim T, Iyer VG, Impey RW. CHARMM-GUI: a web-based graphical user interface for CHARMM. Journal of Computational Chemistry. 2008;29:1859–1865. doi: 10.1002/jcc.20945. [ DOI ] [ PubMed ] [ Google Scholar ] Jorgensen et al. (1983). Jorgensen WL, Chandrasekhar J, Madura JD, Impey RW, Klein ML. Comparison of simple potential functions for simulating liquid water. The Journal of Chemical Physics. 1983;79:926–935. doi: 10.1063/1.445869. [ DOI ] [ Google Scholar ] Kwon & Yu (1997). Kwon KS, Yu MH. Effect of glycosylation on the stability of alpha (1)-antitrypsin toward urea denaturation and thermal deactivation. Biochimica Et Biophysica Acta/General Subjects. 1997;1335(3):265–272. doi: 10.1016/S0304-4165(96)00143-2. [ DOI ] [ PubMed ] [ Google Scholar ] Lombard, Henrissat & Garron (2024). Lombard V, Henrissat B, Garron M. CAZac: an activity descriptor for carbohydrate-active enzymes. Nucleic Acids Research. 2024;53(1):D625–D633. doi: 10.1093/nar/gkae1045. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Macquet et al. (2007). Macquet A, Ralet MC, Loudet O, Kronenberger J, Mouille G, Marion-Poll A, North HM. A naturally occurring mutation in an Arabidopsis accession affects a beta-d-galactosidase that increases the hydrophilic potential of rhamnogalacturonan I in seed mucilage. The Plant Cell. 2007;19(12):3990–4006. doi: 10.1105/tpc.107.050179. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] McIlvaine (1921). McIlvaine TC. A buffer solution for colorimetric comparison. Journal of Biological Chemistry. 1921;49(1):183–186. doi: 10.1016/S0021-9258(18)86000-8. [ DOI ] [ Google Scholar ] Morita et al. (2009). Morita M, Saito S, Ikeda K, Ohno K, Sugawara K, Suzuki T, Togawa T, Sakuraba H. Structural bases of GM1 gangliosidosis and Morquio B disease. Journal of Human Genetics. 2009;54:510–515. doi: 10.1038/jhg.2009.70. [ DOI ] [ PubMed ] [ Google Scholar ] Ohto et al. (2012). Ohto U, Usui K, Ochi T, Yuki K, Satow Y, Shimizu T. Crystal structure of human β-galactosidase: structural basis of Gm1 gangliosidosis and morquio B diseases. The Journal of Biological Chemistry. 2012;287(3):1801–1812. doi: 10.1074/jbc.M111.293795. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Páll & Hess (2013). Páll S, Hess B. A flexible algorithm for calculating pair interactions on SIMD architectures. Computer Physics Communications. 2013;184:2641–2650. doi: 10.1016/j.cpc.2013.06.003. [ DOI ] [ Google Scholar ] Paniagua et al. (2016). Paniagua C, Blanco-Portales R, Barceló-Muñoz M, García-Gago JA, Waldron KW, Quesada MA, Muñoz-Blanco J, Mercado JA. Antisense down-regulation of the strawberry β-galactosidase gene FaβGal4 increases cell wall galactose levels and reduces fruit softening. Journal of Experimental Botany. 2016;67(3):619–631. doi: 10.1093/jxb/erv462. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Parrinello & Rahman (1981). Parrinello M, Rahman A. Polymorphic transitions in single crystals: a new molecular dynamics method. Journal of Applied Physics. 1981;52:7182–7190. doi: 10.1063/1.328693. [ DOI ] [ Google Scholar ] Petrova et al. (2021). Petrova N, Nazipova A, Gorshkov O, Mokshina N, Patova O, Gorshkova T. Gene expression patterns for proteins with lectin domains in flax stem tissues are related to deposition of distinct cell wall types. Frontiers in Plant Science. 2021;12:634594. doi: 10.3389/fpls.2021.634594. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Press et al. (2007). Press W, Teukolsky SA, Vetterling WT, Flannery BP. Numerical recipes: the art of scientific computing. 3rd edition Cambridge: Cambridge University Press; 2007. [ Google Scholar ] Quirke & Crich (2020). Quirke JCK, Crich D. Glycoside hydrolases restrict the side chain conformation of their substrates to gain additional transition state stabilization. Journal of the American Chemical Society. 2020;142:16965–16973. doi: 10.1021/jacs.0c05592. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Rimlumduan et al. (2016). Rimlumduan T, Hua YL, Tanaka T, Ketudat Cairns JR. Structure of a plant β -galactosidase C-terminal domain. Biochimica Et Biophysica Acta/General Subjects. 2016;1864(10):1411–1418. doi: 10.1016/j.bbapap.2016.07.005. [ DOI ] [ PubMed ] [ Google Scholar ] Rojas et al. (2004). Rojas AL, Nagem RA, Neustroev KN, Arand M, Adamska M, Eneyskaya EV, Kulminskaya AA, Garratt RC, Golubev AM, Polikarpov I. Crystal structures of β -galactosidase from Penicillium sp. and its complex with galactose. Journal of Molecular Biology. 2004;343:1281–1292. doi: 10.1016/j.jmb.2004.09.012. [ DOI ] [ PubMed ] [ Google Scholar ] Saburi et al. (2023). Saburi W, Ota T, Kato K, Tagami T, Yamashita K, Yao M, Mori H. Function and structure of Lacticaseibacillus casei GH35 β-galactosidase LBCZ_0230 with high hydrolytic activity to lacto-N-biose I and galacto-N-biose. Journal of Applied Glycoscience. 2023;70(2):43–52. doi: 10.5458/jag.jag.JAG-2022_0014. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Sampedro et al. (2012). Sampedro J, Gianzo C, Iglesias N, Gutián E, Revilla G, Zarra I. AtBGAL10 is the main xyloglucan β -galactosidase in Arabidopsis, and its absence results in unusual xyloglucan subunits and growth defects. Plant Physiology. 2012;158:1146–1157. doi: 10.1104/pp.111.192195. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Shental-Bechor & Levy (2009). Shental-Bechor D, Levy Y. Folding of glycoproteins: toward understanding the biophysics of the glycosylation code. Current Opinion in Structural Biology. 2009;19(5):524–533. doi: 10.1016/j.sbi.2009.07.002. [ DOI ] [ PubMed ] [ Google Scholar ] Sinnott (1990). Sinnott ML. Catalytic mechanism of enzymic glycosyl transfer. Chemical Reviews. 1990;90(7):1171–1202. doi: 10.1021/cr00105a006. [ DOI ] [ Google Scholar ] Smith & Gross (2000). Smith DL, Gross KC. A family of at least seven β -galactosidase genes is expressed during tomato fruit development. Plant Physiology. 2000;123(3):1173–1183. doi: 10.1104/pp.123.3.1173. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Strasser (2016). Strasser R. Plant protein glycosylation. Glycobiology. 2016;26:926–939. doi: 10.1093/glycob/cww023. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Tanthanuch et al. (2008). Tanthanuch W, Chantarangsee M, Maneesan J, Ketudat Cairns JR. Genomic and expression analysis of glycosyl hydrolase family 35 genes from rice (Oryza sativa L.) BMC Plant Biology. 2008;8:84. doi: 10.1186/1471-2229-8-84. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Toonkool et al. (2006). Toonkool P, Metheenukul P, Sujiwattanarat P, Paiboon P, Tongtubtim N, Ketudat-Cairns M, Ketudat-Cairns J, Svasti J. Expression and purification of dalcochinase, a beta-glucosidase from Dalbergia cochinchinensis Pierre, in yeast and bacterial hosts. Protein Expression and Purification. 2006;48(2):195–204. doi: 10.1016/j.pep.2006.05.011. [ DOI ] [ PubMed ] [ Google Scholar ] Trainotti et al. (2001). Trainotti L, Spinello R, Piovan A, Spolaore S, Casadoro G. β-galactosidases with a lectin-like domain are expressed in strawberry. Journal of Experimental Botany. 2001;52(361):1635–1645. doi: 10.1093/jexbot/52.361.1635. [ DOI ] [ PubMed ] [ Google Scholar ] Waterhouse et al. (2018). Waterhouse A, Bertoni M, Bienert S, Studer G, Tauriello G, Gumienny R, Heer FT, De Beer TAP, Rempfer C, Bordoli L, Lepore R, Schwede T. SWISS-MODEL: homology modelling of protein structures and complexes. Nucleic Acids Research. 2018;46(W1):W296–W303. doi: 10.1093/nar/gky427. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplemental Information 1. Os BGal1 and Os BGal1ΔCter genes codon-optimized for expression in Pichia pastoris . peerj-14-21066-s001.docx (135.3KB, docx) DOI: 10.7717/peerj.21066/supp-1 Supplemental Information 2. The raw measurement data for enzyme characterization and kinetics. peerj-14-21066-s002.xlsx (116.6KB, xlsx) DOI: 10.7717/peerj.21066/supp-2 Supplemental Information 3. Trajectory of the 100 ns molecular dynamics simulation of Os BGal1. The stability and compact structure of the full-length Os BGal1 during the 100 ns molecular dynamics simulation. Download video file (22MB, mp4) DOI: 10.7717/peerj.21066/supp-3 Supplemental Information 4. Trajectory of the 100 ns molecular dynamics simulation of Os BGal1 ΔCter. The stability and compact structure of Os BGal1ΔCter without the C-terminal lectin-like domain during the 100 ns molecular dynamics simulation. Download video file (21.9MB, mp4) DOI: 10.7717/peerj.21066/supp-4 Data Availability Statement The following information was supplied regarding data availability: The raw data is available in the Supplemental Files . The molecular dynamics input files and PDB files are available at FigShare: - Choknud, Sunaree; Sanram, Surapoj; R. Ketudat Cairns, James (2025). Molecular dynamics simulation data for GH35 rice β-galactosidases1. figshare. Dataset. https://doi.org/10.6084/m9.figshare.30520754.v1 The sequences are available at DDBJ: LC898194 . Articles from PeerJ are provided here courtesy of PeerJ, Inc ACTIONS View on publisher site PDF (6.7 MB) Cite Collections Permalink PERMALINK Copy RESOURCES Similar articles Cited by other articles Links to NCBI Databases Cite Copy Download .nbib .nbib Format: AMA APA MLA NLM Add to Collections Create a new collection Add to an existing collection Name your collection * Choose a collection Unable to load your collection due to an error Please try again Add Cancel Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top