The MGG_00097 wild-type (WT) and Y56A mutant genes were cloned into a pHis2 expression vector, which included an N-terminal His6-tag for purification purposes. The constructs were transformed into Escherichia coli BL21 (DE3) cells.

Cultures were induced with 0.2 mM IPTG (isopropyl-β-d-thiogalactopyranoside) at 16°C for 16 h. The bacterial culture was harvested through centrifugation at 2,000 ×g for 15 min at 4°C. The pellets were lysed in 150 mM NaCl, 25 mM Tris-HCl pH 8.0, 0.5 mM TCEP-HCl, and 1 mM PMSF by sonication. The lysate was then centrifuged at 25,000 ×g for 30 min at 4°C. Proteins were initially purified using Ni-NTA affinity chromatography and eluted with 25 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.5 mM TCEP-HCl, and 250 mM imidazole pH 8.0. The eluate of target proteins was digested with TEV protease at 4°C overnight. TEV and His6 were removed by applying the solution to a Hi-Trap Q HP column. Peak fractions were pooled and further purified on a Superdex 200 10/300 GL column equilibrated with 25 mM Tris-HCl pH 8.0, 150 mM NaCl, and 0.5 mM TCEP-HCl. The final protein samples were flash-frozen in liquid nitrogen and stored at −80°C until further use.

Crystals of MGG_00097 (10 mg/ml) were obtained using the hanging-drop vapor diffusion method. The protein solution was mixed with well buffer containing 0.2 M NaCl, 2% ethyl acetate, 1.2 M NaH₂PO₄, and 1.2 M K₂HPO₄. The crystals were soaked in cryoprotectant and flash-cooled in liquid nitrogen for data collection at the Shanghai Synchrotron Radiation Facility (SSRF), which were subsequently processed using HKL2000 (Otwinowski and Minor, 1997) (https://www.hkl-xray.com/). X-ray diffraction data were processed using the PHENIX suite (Jumper et al., 2021; Varadi et al., 2022), and the structure was solved by molecular replacement using the Alphafold prediction model (AF-G4NEI2-F1) as the search model (Adams et al., 2010; Emsley and Cowtan, 2004). All structural representations were prepared with PyMOL software (https://www.pymol.org/). Comprehensive statistics for data collection, structure determination, and refinement are provided in Table 1. Coordinates and the structure factors have been deposited to Protein Data Bank (PDB) with accession codes 8YP9 (MGG_00097-NADP⁺) and 8YP2 (MGG_00097^-Y56A-NADP⁺-Glycerol complex). The NCBI accession code for MGG_00097 is XP_003719043.

Under standard assay conditions, the MGG_00097 (WT, 200 nM; F55A, 200 nM; S172A, 200 nM; Y220A, 200 nM; A254G, 200 nM) activity was spectrophotometrically measured in a reaction mixture containing 0-0.8 mM NADP⁺, 10 mM phosphate (pH 7.6), 150 mM NaCl, and 1% glycerol. The catalytic activity of the WT and mutant enzymes was assessed using a spectrophotometric assay to measure NADPH (abs = 340 nm) consumption. Kinetic parameters (Km and Kcat) were derived from Michaelis-Menten plots and analyzed using GraphPad Prism software (GraphPad Software Inc., San Diego, CA, USA).

Recombinant MGG_00097 was successfully purified to homogeneity, as indicated by a single band at approximately 35.5-kDa on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Fig. 1A). Size-exclusion chromatography confirmed the monomeric state of the protein in solution, consistent with the calculated molecular weight (Fig. 1A). The purified proteins obtained are deemed suitable for subsequent crystallization experiments.

The crystal structure of the holo form of MGG_00097, complexed with NADP⁺, was resolved at 2.3 Å resolution. Only one molecule of MGG_00097 is present in the asymmetry unit. The protein adopts a classical TIM-barrel fold (also known as an (β/α)₈ barrel) (Fig. 1B and C) (Brändén, 1991; Farber and Petsko, 1990; Komoto et al., 2004; Mindnich and Penning, 2009), a hallmark of the AKR family (Liu et al., 2014; Sengupta et al., 2015; Songsiriritthigul et al., 2020), which is mainly composed of eight α-helices (α1-α8) and eight β-strands (β1-β8) (Burton, 2018). In the MGG_00097-NADP⁺ structure, the central region is composed of eight β-strands, while the outer region is surrounded by eight α-helix (Fig. 1C). In addition to the core α-helices and β-strands, the structure of MGG_00097 also includes two secondary structure elements that are common to the AKR superfamily, namely the two auxiliary helices H1 (Pro240-Ser249) and H2 (Glu289-Gly301), pack outside the α-helical barrel (Fig. 1C).

Similar to other AKR proteins, the structure of MGG_00097 reveals three large exposed loops, loops A (between β5 and α4; His119-Asn149), B (between β8 and H1; Ser221-Asp239) and C (at the carboxyl-terminus; Arg302-Gln323) (Fig. 2), which are on the top of the barrel and are implicated in the determination of cofactor/substrate-binding site and the substrate specificity (Barski et al., 1996; Ma and Penning, 1999). The lengths and positions in sequence of these loops can vary among AKRs in other families.

The active site of aldo-keto reductases is conserved in both sequence and structure (Jez et al., 1997). Multiple sequence alignment shows that the active site of the enzyme preserves the conserved catalytic tetrad Asp51, Tyr56, Lys86, and His119 (Fig. 3). Interestingly, in all (β/α)₈ barrel proteins, the general position of the active site is maintained, and this folding is thought to be the most common for enzymes (Farber and Petsko, 1990).

In order to understand more fully the structural characteristics of MGG_00097 that determine its substrate specificities, it would be desirable to acquire crystal structures of MGG_00097 bound with a substrate at its active site. In this present study, we initially endeavored to obtain the crystal of WT MGG_00097 binding with substrate, but was not successful. In MGG_00097, Asp51, Tyr56, Lys86, and His119 form the catalytic tetrad, where the tyrosine is the general acid-base involved in proton transfer (Schlegel et al., 1998). Thus, we attempted to obtain a trapped substrate or product/NADP⁺/enzyme complex using a Y56A mutant of MGG_00097. The Y56A mutation enabled the visualization of the open conformation of the cofactor-binding pocket. The binding of glycerol was confirmed within this pocket (Fig. 4A). Structural analysis indicated significant rearrangements in the side chains of Glu29 and Gln226 upon NADP⁺ binding, suggesting these residues act as hinges to facilitate the conformational changes necessary for cofactor binding (Fig. 4B). Aldo-keto reductases carry out a sequential ordered bi-bi catalytic mechanism in which the cofactor binds to the enzyme first and is released last in the reaction cycle (Penning, 2015; Sanli and Blaber, 2001). It has been proposed that the cofactor must bind to the AKR enzyme first in order to be buried at the bottom of the substrate binding chamber, and the substrate molecule then pile up on the nicotinamide ring of the NADP⁺ cofactor (Penning, 2015; Penning et al., 2019). The NADP⁺ molecule with anti-conformation is deep inside a channel connected to the central active site cavity in the ternary complex structure. The nicotinamide and ribose moieties are stabilized by potential hydrogen bonds interactions from the highly conserved residues Asp51, Ser172, and Asn173 (Fig. 4C). The phosphate group and the NADP⁺ adenine moieties are surrounded by Ser221, Lys271, Ser272, Ser273, Thr274, Arg277, and Asn281.

The crystal structures of MGG_00097-NADP⁺ (lightblue) and MGG_00097^Y56A-NADP⁺-glycerol are superimposed (blue) (Fig. 5). Compared to the holo form, the Y56A mutation resulting in the opening of the substrate binding pocket. This conformational change allows for the pocket to be accessible to glycerol for binding (Fig. 5). While the overall conformation of MGG_00097^Y56A is almost identical to the WT, notable conformational changes are observed in the side chains of Glu29 and Gln226. Upon NADP⁺ binding, the Gln226 side chain undergoes a rotation of ~60°, enabling it to tightly fit into the crescent groove of the extended NADP⁺ molecule. In contrast, the Glu29 side chain rotates approximately 30° in proximity to the glycerol molecule (Fig. 5).

The crystal structure of MGG_00097 were determined to investigate the effects of Ala mutations on the catalytic residues Phe55, Ser172, and Tyr220. Additionally, the effects of glycine mutations on Ala254 were also investigated in terms of catalysis (Fig. 6A). The apparent kinetic parameters of the Km-NADP⁺ and Km-glycerol pairs were measured and analyzed using Lineweaver-Burk plots (Dalziel, 1975; Jiang et al., 2016; Schuurink et al., 1990) (Fig. 6B-D). These biochemical analyses aimed to provide a deeper understanding of the changes in enzymatic activity caused by the alanine mutations from a biochemical perspective. The binding of NADP⁺ and glycerol to the enzyme significantly influences their respective Km values (Tables 2 and 3). In Table 3, the Km value of MGG_00097^F55A for glycerol is significantly increased from 343.33 mM (WT) to 3,813 mM, indicating about elevenfold weaken binding affinity. Although Kcat slightly decreased by 1.38-fold compared to that of MGG_00097, Kcat/Km is decreased by 18.5-fold compared to that of the WT, showing that the catalytic efficiency of substrate glycerol reduction. The Km-NADP⁺ of MGG_00097^F55A slightly increased from 0.07 mM (WT) to 0.08 mM. The Kcat-NADP⁺ of MGG_00097^F55A decreased from 87.32/min to 7.51/min, and thus, catalytic efficiency (Kcat/Km) decreased by 12.5-fold. Based on the structure analysis, it is observed that Ser172 potentially interacts with NADP⁺ by hydrogen bonds. Upon mutation of Ser172 to alanine (MGG_00097^S172A), the Km value for NADP⁺ increased from 0.07 mM to 0.14 mM. Additionally, the Kcat value for NADP⁺ decreased significantly from 87.32/min to 17.24/min. The results here shed the catalytic efficiency of the cofactor NADP⁺ is reduced in the MGG_00097^S172A mutant.

In summary, we solved the crystal structures of MGG_00097 in cofactor bound forms (MGG_00097-NADP⁺ complex), as well as the MGG_00097^Y56A-NADP⁺-glycerol ternary complex. The binary structure of MGG_00097-NADP⁺ shares the typical TIM fold and the NADP⁺ molecular with anti-conformation is deep inside a channel connected to the central active site cavity within the (β/α)₈ barrel. Compared with previous studies of other AKRs, all other AKRs have the conserved catalytic tetrad residues, including AKR2E4 (D53, Y58, K87, and H120), AKR11C1 (D46, Y51, K76, and H119), and AKR11B4 (D54, Y59, K84, and H130) (Marquardt et al., 2005; Richter et al., 2010; Yamamoto and Wilson, 2013). In the MGG_00097 amino acid sequence, the conserved catalytic tetrad D51, Y56, K86, and H119 surrounded the substrate glycerol. The phosphate group forms a charged pair with D51. The histidine is believed to play a role in substrate orientation and transition state stabilization (Ehrensberger and Wilson, 2004; Khurana et al., 1998). A consensus among most AKRs is their preference for the coenzyme, NADPH over NADH, based on MGG_00097^Y56A-NADP⁺ structural analysis, the residues K271, S272, S273, T274, and R277 contribute to a hydrogen bond network encompassing the 2’monophosphate of adenosine monophosphate (Abraham et al., 2022; Jez et al., 1997). These amino acids likely act as key contributing elements for MGG_00097’s NADP⁺ preference.

Structural comparison between MGG_00097-NADP⁺ and MGG_00097^Y56A-NADP⁺-glycerol identified a conformational alteration of Glu29 and Gln226. The Y56A mutation causes the phenol group on tyrosine to be replaced by hydrogen ions, causes a conformational change in the protein and opens the substrate binding pocket to be accessible to glycerol. The F55A (amino acid lost π bond of benzene ring after mutation) and Y220A (amino acid changed from hydrophilic amino acid to hydrophobic amino acid after mutation) were located at both ends of glycerol molecule, and Km value was significantly increased after mutation, indicating that the binding affinity between protein and substrate glycerol was significantly weakened after mutation. These results indicate that F55 and Y220 are the key amino acid sites of substrate-binding pocket. Ser172 is the key residue in cofactor-binding pocket.

The appressoria of M. grisea generate enormous turgor pressure by accumulating glycerol at concentrations as high as 3M, enabling the appressorium to breach the plant cuticle. The synthesis of glycerol in eukaryotic cells can occur via several routes. In yeast, glycerol production from carbohydrates involves the enzyme GPD, which converts dihydroxyacetone phosphate to glycerol-3-phosphate in an NADH-dependent manner (Ansell et al., 1997). Glycerol can also be synthesized from dihydroxyacetone and glyceraldehyde by an NADPH-dependent reductases. In the filamentous fungus Aspergillus nidulans, these reactions are carried out by a single enzyme known as NADPH-dependent GD (Redkar et al., 1995). In the mycelium of M. grisea, the activity of GPD was low and did not show significant changes in response to hyperosmotic stress. However, the activities of dihydroxyacetone reductase and glyceraldehyde reductase were detected in the mycelium of M. grisea and increased approximately threefold under hyperosmotic stress (Thines et al., 2000). In developing appressoria, the activities of both enzymes were present and remained active even 48 h after the appressoria were fully mature (Thines et al., 2000).

In this study, we demonstrate that MGG_00097 exhibits NADPH-dependent oxidoreductase activity, specifically oxidizing glycerol to glyceraldehyde. NADPH-dependent oxidoreductases catalyze the reduction or oxidation of a substrate coupled to the oxidation or reduction, respectively. Thus, MGG_00097 may play a key role in supporting the normal life processes of M. grisea or in facilitating plant host infection as glyceraldehyde reductase. MGG_00097 crystal structures provided mechanistic details of cofactor-dependent AKR activity. AKRs encoding NADP(H)-dependent oxidoreductases exist in nearly all phyla and function in the phase 1 metabolism of endogenous substrates and xenobiotics (Jez et al., 1997). MGG_00097 might be a good candidate for the development of stress-tolerant rice and could play a vital role in agricultural technology.