Structure-based enzyme engineering improves donor-substrate recognition of Arabidopsis thaliana glycosyltransferases

Abstract

Glycosylation of secondary metabolites involves plant UDP-dependent glycosyltransferases (UGTs). UGTs have shown promise as catalysts in the synthesis of glycosides for medical treatment. However, limited understanding at the molecular level due to insufficient biochemical and structural information has hindered potential applications of most of these UGTs. In the absence of experimental crystal structures, we employed advanced molecular modeling and simulations in conjunction with biochemical characterization to design a workflow to study five Group H Arabidopsis thaliana (76E1, 76E2, 76E4, 76E5, 76D1) UGTs. Based on our rational structural manipulation and analysis, we identified key amino acids (P129 in 76D1; D374 in 76E2; K275 in 76E4), which when mutated improved donor substrate recognition than wildtype UGTs. Molecular dynamics simulations and deep learning analysis identified structural differences, which drive substrate preferences. The design of these UGTs with broader substrate specificity may play important role in biotechnological and industrial applications. These findings can also serve as basis to study other plant UGTs and thereby advancing UGT enzyme engineering.

Keywords: deep learning; glycosyltransferases; mass spectrometry; molecular dynamics.

Figure 1.. Product formation catalyzed by UGT 76E1.

(A) MS/MS method to determine product formation using UDP-Glc, UDP-Gal, and UDP-GlcNAc with quercetin catalyzed by UGT 76E1; (B) GAR screen results showing summary of wildtype UGTs 76E4, 76D1, 76E2, 76E1, and 76E5 donor activities where green and red indicate positive and no activity, respectively; (C) structures of donor compounds used in this screening.

Figure 2.. Sequence and structure comparison of group H AtUGTs.

(A) Multiple sequence alignment of AtUGT74F2 with group H AtUGTs reported in this study. Different regions are annotated above the sequences. The positions of mutation are marked in blue (B) Superimposition of 3D structures showing similarity between the models (cyan) and template AtUGT74F2 (red). UDP-sugar is illustrated in CPK.

Figure 3.. A structural comparison of WT and mutant residues investigated in this study.

The proximity of K275 (green) side chain in C1 loop to (A) Glc C2–OH and (B) NAc group in 76E4; (C) P129 on N5 loop in 76D1 is unable to make hydrogen bond with the donor sugar due to the absence of the OH group; (D) Spatial position of G347 on C4 loop in 76D1; (E) Asp or Glu is present in C5 loop at equivalent position of D374 (76E2) in other group H AtUGTs; (F) D374 side chain in 76E2 is unable to make hydrogen bond with the donor sugar. The side chain extended by one carbon atom to E374 enables the formation of a hydrogen bond; (G) The role of T129 in UDP-Gal recognition in 76D1 P129T mutant. T129 interacts with H17 and is involved in the formation of a catalytic triad, which is essential to glycosylation activity; (H) The proximity of C347 side chain to C352 and C364 side chains, where the Cys residues can potentially form a disulfide bond.

Figure 4.. Confirmation of product formation.

Mass spectra—product ion scan confirming glycosylation in mutant 76E4 K275L with (A) UDP-Glc, (B) UDP-Gal and (C) UDP-GlcNAc donor sugars; (D) GAR screen results showing summary of mutant UGTs 76E4 K275L, 76D1 P129T, 76D1 G347C, and 76E2 D374E donor activities where green and red indicate positive and no activity, respectively.

Figure 5.. Structural profiles of UGTs 76E1 and 76E5 against the three sugar substrates.

The 2D t-SNE plots illustrate the latent spaces of (A) 76E1 and (C) 76E5. The catalytic residues are highlighted in the cartoon representation in (B) and (D) taken from the representative 3D structures outlined in boxes in the t-SNE plots. UDP-GlcNAc is shown in stick representation. The enzyme structures are color coded, where UGT-(UDP-Glc) is in green, UGT-(UDP-Gal) in orange, and UGT-(UDP-GlcNAc) in blue.

Figure 6.. Structural profiles of UGT variants based on substrate specificity.

The 2D t-SNE plots illustrate the latent spaces of the enzymes exhibiting positive activity for (A) UDP-Glc, (B) UDP-Gal, and (C) UDP-GlcNAc. The 2D t-SNE plots illustrate the latent spaces of the wildtype UGT 76E4 and the mutant 76E4 K275L against (D) UDP-Glc, (E) UDP-Gal, and (F) UDP-GlcNAc.