Characterization of Two Novel Rumen-Derived Exo-Polygalacturonases: Catalysis and Molecular Simulations

Pectinases are a series of enzymes that degrade pectin and have been used extensively in the food, feed, and textile industries. The ruminant animal microbiome is an excellent source for mining novel pectinases. Two polygalacturonase genes, IDSPga28-4 and IDSPga28-16, from rumen fluid cDNA, were cloned and heterologously expressed. Recombinant IDSPGA28-4 and IDSPGA28-16 were stable from pH 4.0 to 6.0, with activities of 31.2 ± 1.5 and 330.4 ± 12.4 U/mg, respectively, against polygalacturonic acid. Hydrolysis product analysis and molecular dynamics simulation revealed that IDSPGA28-4 was a typical processive exo-polygalacturonase and cleaved galacturonic acid monomers from polygalacturonic acid. IDSPGA28-16 cleaved galacturonic acid only from substrates with a degree of polymerization greater than two, suggesting a unique mode of action. IDSPGA28-4 increased the light transmittance of grape juice from 1.6 to 36.3%, and IDSPGA28-16 increased the light transmittance of apple juice from 1.9 to 60.6%, indicating potential application in the beverage industry, particularly for fruit juice clarification.


Gene Cloning, Expression, and Protein Purification
Two putative GH28 family genes, IDSPga28-4 and IDSPga28-16, were PCR-amplified (Table S1) from the cDNA of Hu sheep rumen fluid using Super pfx DNA polymerase (CWBIO, Beijing, China). The expected fragments were purified and cloned into the truncated pET-30a(+) vector by BamHI + XhoI via homologous recombination using the Trelief™ SoSoo Cloning kit (TsingKe Biotech, Beijing, China). The resulting ligation products were transformed into E. coli BL21 (DE3)-competent cells by heat shock transformation and streaked onto an LB agar plate (5 g/L yeast extract, 10 g/L tryptone, 10 g/L sodium chloride, 20 g/L agar) supplemented with 50 µg/mL of kanamycin, and then incubated at 37 • C for 16 h. Ten colonies of each gene construct were selected for colony-PCR validation. Positive recombinant plasmids were further confirmed by Sanger sequencing (Sangon, Shanghai, China). The recombinant hosts were designated as BL21/pET30a/IDSPga28-4 and BL21/pET30a/IDSPga28-16, respectively. Subsequently, IPTG-induced gene expression, cell pellet collection, and sonication were carried out as described previously [16], and then crude enzyme, resuspended in ice-cold phosphate-buffered saline (PBS, pH 7.4), was loaded onto a HisTrap™ 5 mL column (GE Healthcare BioSciences, Pittsburgh, PA, USA) fitted to an ÄKTA start protein purification system. The bound proteins were eluted with elution buffer, supplemented with a linear gradient of 20-250 mM imidazole. Eluted fractions were analyzed by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and the purified protein fractions were used for enzymatic assays.

Enzyme Activity Assay
Polygalacturonase activity was determined using the 3,5-dinitrosalicylic acid (DNS) method [21] with GalA as the standard. Briefly, purified enzyme solution (15 µL, 0.43 µg protein) and PGA (60 µL, 0.25% w/v) were dissolved in 100 mM citric acid-Na 2 HPO 4 buffer (pH 5.0), unless otherwise stated. After incubation at 40 • C for 10 min, DNS solution (75 µL) was added to terminate the reaction, followed by 10 min in a boiling water bath, and then the absorbance of the reaction mixture was measured at 540 nm, using a Spark multimode microplate reader (Tecan, Männedorf, Switzerland). One enzyme unit (U) was defined as 1 µmol of reducing sugar released per minute. Enzyme concentration was measured by the Bradford method [22] using bovine serum albumin as the standard. To investigate the substrate specificity, purified enzyme was reacted with 0.25% (w/v) PGA, rhamnogalacturonan I, or pectin.

Characterization of Purified Recombinant Enzymes
To determine the optimum pH, purified IDSPGA28-4, or IDSPGA28-16, was incubated with 0.25% (w/v) PGA over a pH range of 3.0-10.0 (pH 3.0-8.0, citric acid-Na 2 HPO 4 buffer; pH 8.0-9.0, Tris-HCl buffer; pH 9.0-10.0, glycine-NaOH buffer) at 40 • C for 10 min, and then the reducing sugar produced was measured as described above. The optimum temperature of each enzyme was determined in buffer at the corresponding optimum pH and temperatures between 30 and 70 • C. The maximum enzyme activity was designated as 100%. All assays were performed in quadruplicate.
The pH stability was determined by measuring the residual activities of IDSPGA28-4 or IDSPGA28-16 under the optimum conditions (40 • C and pH 5.0) after preincubating the enzymes at 25 • C for 1 h in various buffers of pH 3.0-10.0. For the thermostability assay, IDSPGA28-4 or IDSPGA28-16 was pretreated at 30, 40, or 50 • C for 1 h. Aliquots were collected at different time intervals (10,20,30, and 60 min) and used to determine the residual activity. The initial enzyme activity before preincubation was designated as 100%. All assays were performed in quadruplicate.

Hydrolysis Product Analysis
Approximately 5 µg of purified enzyme was incubated with 0.25% polygalacturonic acid at pH 5.0 and 37 • C for 3 h. Aliquots sampled at different time intervals (30 min, 1 h, and 3 h) were boiled for 10 min, centrifuged at 25 • C and 10,000× g for 10 min, and then subjected to thin-layer chromatography (TLC) and high-performance anion exchange chromatography (HPAEC). The products formed by treating 0.25% (GalA) n (n = 1, 2, 3) with the purified enzymes for 24 h were analyzed by TLC and HPAEC.
TLC was conducted using a 10 × 10 cm silica gel 60 plate (Merck, Darmstadt, Germany) with the developing solvent, 1-butanol/acetic acid/water = 2:1:1 (v/v/v). The plate was taken out when the solvent front was about 1 cm from the top and, after the solvent had evaporated completely, the plate was sprayed with visualization reagent (sulfuric acid: ethanol = 5:95, v/v). The plate was air-dried, then heated at 105 • C for 10 min, until spots appeared.
HPAEC was performed on an ICS

Molecular Docking and Molecular Dynamics (MD) Simulation
The 3-D structures of (GalA) 2 and (GalA) 3 were downloaded from PubChem (https://pubchem.ncbi.nlm.nih.gov/, accessed on 1 November 2022), and the 3-D structure of IDSPGA28-16 was modeled with SWISS-MODEL (https://www.swissmodel.expasy. org/, accessed on 21 October 2022), using the known structure of Thermotoga maritima exo-polygalacturonase (PDB: 3JUR), which has the highest sequence similarity (41.08%) as the template. Molecular docking between IDSPGA28-16 and (GalA) 2 , or (GalA) 3 , was performed by AutoDock Vina (v1.1, CCSB, Scripps Research Institute, California, USA) [23]. A 3-D box was defined to restrict the conformational sampling space, in which the center point of the catalytic-site Asp residues (D251, D272, and D273) was set as the box center, and the box size was 30 Å, 28 Å, and 26 Å in the x-, y-, and z-dimensions, respectively. Exo-polygalacturonase hydrolyzes the first glycosidic bond from the non-reducing end of PGA, whereas exo-poly-α-digalacturonosidase hydrolyzes the second α-1,4-glycosidic bond from the non-reducing end of the substrate, according to the CAZy database. Therefore, the conformation of the protein-substrate complex, with the nonreducing end inside the catalytic pocket and the highest score from AutoDock Vina, was subjected to molecular dynamics (MD) simulation.
The MD simulations of the complexes were carried out by Gromacs 2019.6 (https://doi.org/10.5281/zenodo.3685922, accessed on 5 November 2022) with the AM-BER ff14SB force field for proteins and GAFF for the sugar molecules, using TIP3P as the water model [24][25][26]. A dodecahedral box was generated, spaced 1 nm away from the periphery of the enzyme-substrate complex, with 69,517 water molecules as the solvent to fill the box for MD simulation of the IDSPGA28-16/(GalA) 2 complex, and 47,197 water molecules for IDSPGA28-16/(GalA) 3 . Because IDSPGA28-16 is negatively charged, 10 of the water molecules were replaced with Na + ions to keep the system electrically neutral. For energy minimization, 50,000 steps of conjugate gradient minimization and a steepest descent energy minimization for every 1000 steps were conducted, with hydrogen bonds constrained by LINCS [27]. The V-rescale method was used for temperature control, the Parrinello-Rahman method for pressure coupling, and the PME method for the calculation of long-range electrostatic interactions. The van der Waals interactions were cut-off at 10 Å. The 100 ns simulation was performed in the isothermal-isobaric NPT ensemble (300 K, 1 atm) with a time step of 2 fs.
Water molecules were removed and the periodicity of the trajectory modified, using the trjconv module of Gromacs, and then the root-mean-square deviation (RMSD) for every 2 ps was calculated by the rms module. The noncovalent interactions (NCIs) of the two complexes were analyzed by Multiwfn 3.8 using the independent gradient model (IGM) and plotted by VMD v1.9.3 (http://www.ks.uiuc.edu/Research/vmd/, accessed on 14 December 2022) [28,29]. The detailed schematic diagrams of the intermolecular interactions, especially hydrogen bonds within 3.5 Å, were plotted by LigPlot + (v2.2, EMBL-EBI, Wellcome Genome Campus, Cambridgeshire, UK).

Fruit Juice Clarification with Exo-Polygalacturonase
Fresh grape juice, orange juice, and apple juice were extracted from Kyoho grapes, Gannan Navel oranges, and Shaanxi Red Fuji apples (500 g each), respectively, then filtered through eight layers of gauze to remove pulp solids. Ascorbic acid (0.5% w/v) was added to the freshly extracted fruit juice to minimize oxidation. Purified IDSPGA28-4, or IDSPGA28-16, (~500 µg) was incubated with aliquots of juice (5 mL) at 37 • C, without stirring, for 1.5 h. Purified IDSPGA28-4, or IDSPGA28-16, was boiled for 10 min before addition to the juice, as controls. The light transmittance of the treated juice was measured with a UV/vis spectrophotometer (Phoenix, Shanghai, China) at 660 nm (%T 660 ), with the light transmittance of distilled water as 100%. All reactions were performed in triplicate.

Gene Cloning and Sequence Analysis
In the past decade, multi-omics approaches, such as metagenomics and metatranscriptomics, have been adopted widely to improve the understanding of the composition and functionality of the gastrointestinal microbiome in humans and animals [30]. There is great interest in mining novel CAZymes from the ruminant microbiome, because of their promising potential for processing of plant-derived foods and beverages. In this study, two exo-polygalacturonase genes, IDSPga28-4 and IDSPga28-16, were cloned from rumen fluid cDNA. The ORFs of IDSPga28-4 and IDSPga28-16 encoded 541 and 458 amino acids, respectively. Multiple sequence alignment revealed that IDSPGA28-4 and IDSPGA28-16 shared the highest similarity with two GH28 proteins from an unclassified Oscillospiraceae bacterium (96.02% identity with GenBank: MBP3209358) and a Firmicutes bacterium (69.39% identity with GenBank: MBR1735702). However, neither of the latter enzymes have been functionally characterized.
Similar product profiles have been observed previously. Heterologous expression and characterization of five endo-polygalacturonases, BcPGs, from the plant pathogenic fungus, Botrytis cinerea, revealed that BcPG3 and BcPG6 mostly generated GalA and (GalA) 2 from PGA, but trace amounts of (GalA) 3 to (GalA) 6 were also detected during the reaction [9]. The endopolygalacturonase PGD, derived from A. niger, mainly hydrolyzed PGA into GalA and (GalA) 2 as final products, but (GalA) 3 was also observed as an intermediate [38]. However, no intermediate oligogalacturonide products were observed during the timecourse reaction of IDSPGA28-16.

Molecular Dynamics (MD) Simulation
To understand the structural basis of substrate binding and catalysis of IDSPGA28-16, molecular docking and MD simulation analyses were performed on the enzyme structural models. As molecular docking reflects the specific state at a certain moment, the most stable conformation of the complex was selected by molecular docking and then used for MD simulation ( Figure 5). MD simulation of IDSPGA28-16 with (GalA) 2 as a substrate reached equilibrium after 30 ns, with a root-mean-square deviation (RMSD) of 5.11 Å ( Figure 5A). With (GalA) 3 as the substrate, the enzyme-substrate complex reached equilibrium iñ 20 ns, with a much smaller RMSD of 1.89 Å ( Figure 5B). Generally, RMSD is used as a quantitative assessment of similarity between two protein structures or protein-substrate complexes; a relatively low RMSD is preferable for mechanistic simulation [39]. In this case, the conformation with the highest score obtained by molecular docking was used as the initial conformation for MD simulation. The high RMSD of the enzyme-(GalA) 2 complex suggested that (GalA) 2 could bind flexibly, in more than one conformation, whereas (GalA) 3 was relatively tightly constrained.  To further unravel the detailed molecular interactions between key active-site amino acid residues and substrates, the enzyme-(GalA) 2 and enzyme-(GalA) 3 representative structures obtained from MD simulations for 70-100 ns were used to visualize the noncovalent interaction (NCI) iso-surface diagram within 5 Å of the substrate (Figure 5C,D). The (GalA) 2 or (GalA) 3 was positioned at the bottom of the cleft formed by four surrounding loops in IDSPGA28-16 ( Figure S2A,B). It is widely accepted that eight highly conserved residues, including three aspartates, participate in substrate binding and catalysis in endoor exo-polygalacturonases [5,10,11,30]. Seven (N249, D272, D273, H306, G312, R337, and K339) of the eight key resides were located within 5 Å of (GalA) 3 ( Figure 5). More importantly, these residues had strong binding interactions with the substrate; two catalytic aspartates, D272 and D273, formed hydrogen bonds of 2.47 Å and 3.23 Å, respectively, with the +1 subsite of (GalA) 3 ( Figure 5F). Several basic amino acids, including K278, K287, R337, and K339, also interacted with the substrates, contributing to stabilization of the enzyme-substrate complex, in good agreement with previous reports [5,10]. Mutations of the conserved basic amino acids, i.e., arginine (R256N) and lysine (K258N), in the A. niger endopolygalacturonase II, dramatically reduced its catalytic activity, but decreased its K m 10-fold [11]. In the MD stimulation of the enzyme-(GalA) 2 complex, only D272 of the eight key residues was within 5 Å of the substrate ( Figure 5E), consistent with the inability of the enzyme to hydrolyze (GalA) 2 ( Figure 4).
Regarding the different acting modes of endo-and exo-polygalacturonases, the four conserved protein loops, particularly loop 1, appear to contribute strongly to substrate recognition and binding. Loop 1 in endo-polygalacturonases is involved in forming one "wall" of the "substrate path", contributing to a tunnel-like active site, rather than orientating to the active center [4,5,10,11,40]. The substrate tends to lay along the tunnel-like active site in endo-polygalacturonase, which randomly cleaves α-1,4 glycosidic bonds via a single-attack manner, generating pectic oligogalacturonides and monomers [4,9,41,42]. In contrast, loop 1 in exo-polygalacturonase [19,30] and exo-poly-α-digalacturonosidase [5] formed the "back wall" of the substrate-binding cavity, forming to a pocket-like active site (Figures 5 and S2). Interestingly, in addition to several basic amino acid residues, the acidic residue E101, positioned on loop 1, also formed three strong hydrogen bonds with the −1 subsite ( Figure 5F). After PGA, or a pectic oligogalacturonide substrate was bound to the buried basic amino acid residues, its reducing end was blocked by loop 1. Consequently, exo-polygalacturonase could only access and cleave the first α-1,4 glycosidic bond from the reducing end and remove monomers progressively ( Figure 4A,B) [12,19]. Unlike known exo-polygalacturonases, IDSPGA28-16 was incapable of cleaving (GalA) 2 ( Figure 4E,F), probably because of unproductive substrate binding preventing access of the catalytic residues to the α-1,4 glycosidic bond ( Figure 5A,C,E). Taken together, the above stoichiometry and stereochemistry data suggested that IDSPGA28-16 was neither a typical exo-polygalacturonase nor an exo-poly-α-digalacturonosidase. It is an unconventional exo-polygalacturonase.

Effect of Recombinant Enzymes on Fruit Juice Clarification
Many fruits contain high concentrations of pectin, which is responsible for the high viscosity of fruit pulp and the turbidity and often poor yield of freshly pressed fruit juice. Therefore, the treatment of fruit juice, or pulp with pectinases, such as polygalacturonase, pectin lyase, and rhamnogalacturonase, to degrade pectin has been extensively used industrially to increase juice yield, reduce viscosity, and for clarification of the juice, for increased consumer acceptance [43,44]. Although IDSPGA28-4 and IDSPGA28-16 could not degrade the highly esterified (60% methylated) citrus pectin, a previous study [4] reported that exo-polygalacturonase can degrade lightly esterified pectin and clarify grape juice. The effectiveness of IDSPGA28-4 and IDSPGA28-16 for clarification of orange, grape, and apple juice was determined. Treatment with IDSPGA28-4 significantly increased the light transmittance (%T 660 ) of grape juice (pH = 3.6) from 1.6% to 36.3% (p < 0.001) ( Figure 6A), whereas IDSPGA28-16 dramatically increased the %T 660 of apple juice (pH = 3.4) from 1.9% to 60.6% (p < 0.001) and grape juice from 1.1% to 43.8% (p < 0.001) ( Figure 6B). oorganisms 2023, 11, x FOR PEER REVIEW 13 of whereas IDSPGA28-16 dramatically increased the %T660 of apple juice (pH = 3.4) from 1.9 to 60.6% (p < 0.001) and grape juice from 1.1% to 43.8% (p < 0.001) ( Figure 6B). The acidic pH of most fruit juices, such as orange, grape, apple, lemon, and papay requires pectinases with acidic pH optima for industrial application [41]. After pre-inc bation at pH 3.5 for 1 h, IDSPGA28-4 and IDSPGA28-16 retained ~50% of their origin activities ( Figure 3E,F), displaying resilience to acidic pH comparable to many previous reported pectinases [4,45,46]. IDSPGA28-4 and IDSPGA28-16 were less effective for cla fying orange juice (pH = 5.6) than grape and apple juice, even though they had high ca lytic activity from pH 5.0 to 6.0 ( Figure 3E), and the effectiveness in clarifying apple jui was different between two enzymes. The reason may be that the degrees of esterificati of pectin contained in these fruits are different [47] and the two enzymes differ in th preference for the degree of pectin esterification. A cocktail of endo-and exo-acting h drolyses, as well as pectin methylesterases, could further improve the extraction efficien and clarification of fruit juice [4,8,41], which appears to be more attributable to the variab degree of esterification in pectin.

Conclusions
In this study, two exo-polygalacturonases from sheep rumen microbiota, IDSPGA2 4 and IDSPGA28-16, were found to hydrolyze PGA, and the latter was among the mo catalytically active exo-polygalacturonases to date. Distinct from IDSPGA28-4 and prev ously reported exo-polygalacturonases, IDSPGA28-16 was capable of liberating gala turonic acid monomers from substrates only with DP > two, suggesting that it underwe a unique action mode. IDSPGA28-4 and IDSPGA28-16 were effective for fruit juice cla fication, showing potential applications in the food and feed industries.
Supplementary Materials: The following supporting information can be downloaded www.mdpi.com/xxx/s1, Figure S1: Amino acid alignment of bacterial exo-polygalacturonas Three conserved catalytic residues (D251, D272, D273) are highlighted by triangles and five residu thought to participate in substrate-binding are indicated by arrows. Four key loops involved in su The acidic pH of most fruit juices, such as orange, grape, apple, lemon, and papaya, requires pectinases with acidic pH optima for industrial application [41]. After pre-incubation at pH 3.5 for 1 h, IDSPGA28-4 and IDSPGA28-16 retained~50% of their original activities ( Figure 3E,F), displaying resilience to acidic pH comparable to many previously reported pectinases [4,45,46]. IDSPGA28-4 and IDSPGA28-16 were less effective for clarifying orange juice (pH = 5.6) than grape and apple juice, even though they had high catalytic activity from pH 5.0 to 6.0 ( Figure 3E), and the effectiveness in clarifying apple juice was different between two enzymes. The reason may be that the degrees of esterification of pectin contained in these fruits are different [47] and the two enzymes differ in their preference for the degree of pectin esterification. A cocktail of endo-and exo-acting hydrolyses, as well as pectin methylesterases, could further improve the extraction efficiency and clarification of fruit juice [4,8,41], which appears to be more attributable to the variable degree of esterification in pectin.

Conclusions
In this study, two exo-polygalacturonases from sheep rumen microbiota, IDSPGA28-4 and IDSPGA28-16, were found to hydrolyze PGA, and the latter was among the most catalytically active exo-polygalacturonases to date. Distinct from IDSPGA28-4 and previously reported exo-polygalacturonases, IDSPGA28-16 was capable of liberating galacturonic acid monomers from substrates only with DP > two, suggesting that it underwent a unique action mode. IDSPGA28-4 and IDSPGA28-16 were effective for fruit juice clarification, showing potential applications in the food and feed industries.
Supplementary Materials: The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/microorganisms11030760/s1, Figure S1: Amino acid alignment of bacterial exo-polygalacturonases. Three conserved catalytic residues (D251, D272, D273) are highlighted by triangles and five residues thought to participate in substrate-binding are indicated by arrows. Four key loops involved in substrate binding and catalysis are underlined in blue (IDSPGA28-16) or purple (IDSPGA28-4). Figure S2: Predicted three-dimensional structure and hydrophobic surface of IDSPGA28-4 (A-B) and IDSPGA28-16 (C-D). Exo-polygalacturonase from Thermotoga maritima; (PDB: 3JUR) served as the template for homology modeling. Eight conserved key residues are depicted as sticks. Four key loops involved in substrate binding and catalysis are highlighted in azure (IDSPGA28-16) or violet (IDSPGA28-4). The proposed catalytic center is indicated in red. Table S1: Oligonucleotides used in this study. Data Availability Statement: The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.