Identification and Characterization of a Putative Kaempferol Glucosyltransferase UGT78G3 in Medicago truncatula

TL;DR

The study identifies and characterizes UGT78G3, a putative enzyme in Medicago truncatula that may glycosylate kaempferol but does not significantly impact its levels in plants.

Contribution

The study identifies UGT78G3 as a putative kaempferol glucosyltransferase and explores its in vitro and in vivo roles.

Findings

UGT78G3 catalyzes the formation of kaempferol 3-O-glucoside in vitro.

Overexpression or mutagenesis of UGT78G3 does not significantly affect kaempferol 3-O-glucoside levels in vivo.

UGT78G3 is a potential target for biocatalyst design for flavonoid glucoside synthesis.

Abstract

UDP-glycosyltransferases (UGTs) represent a large multigene family that play a central role in glycosylating a highly diverse array of natural products, underscoring their critical importance in various biological processes. However, the functional roles of a substantial majority of UGTs remain to be elucidated. In the present study, we characterized the glycosyltransferase UGT78G3, a member of the UGT78 glycosyltransferase family in the model legume Medicago truncatula. Amino-acid sequence analysis revealed a conserved PSPG motif at the C-terminus of UGT78G3. Liquid chromatography-coupled tandem mass spectrometry (LC-MS/MS) analysis demonstrated that UGT78G3 catalyzes the formation of kaempferol 3-O-glucoside in vitro. However, neither UGT78G3 overexpression nor CRISPR/Cas9-mediated mutagenesis resulted in significant changes to the endogenous levels of kaempferol 3-O-glucoside, indicating that UGT78G3 does not play a predominant role in the biosynthesis of kaempferol 3-O-glucoside in vivo. Our findings identify a putative glycosyltransferase in M. truncatula and provide a target for biocatalyst design aimed at synthesizing flavonoid glucosides.

Figures

Full text

1. Introduction

Natural products biosynthesized by plants undergo diverse modifications, including glycosylation, acetylation, methylation, and hydroxylation, that not only diversify their chemical structure, but also underpin their critical roles in industrial and pharmaceutical applications [1]. Among these modifications, glycosylation is an essential and ubiquitous modification step during which sugar moieties are transferred from sugar donors to a wide and diverse range of acceptor substrates, thereby altering their hydrophilicity, stability, chemical properties, and bioactivity [2,3,4,5].

Glycosylation reactions are orchestrated by glycosyltransferases (GTs), which have been categorized into 139 distinct families (GT1~GT139) within the carbohydrate active enzyme database (CAZy; https://www.cazy.org/GlycosylTransferase-family; [6]) (accessed on 1 December 2025). Among these, the GT1 superfamily, frequently designated as uridine diphosphate glycosyltransferases (UDP-glycosyltransferases, UGTs), constitutes the largest group [7,8]. Plant genomes have evolved an extensive repertoire of UGTs. In Arabidopsis thaliana, for instance, 123 UGTs have been identified and categorized into 14 groups and an outgroup [9]. The largest known UGT family has been reported in Eupatorium pubescens, comprising 339 putative UGT genes [10]. A defining characteristic of plant UGTs is the highly conserved plant secondary product glycosyltransferase motif (PSPG), spanning about 44 amino acids, located in the central of the C-terminal domain (CTD). This motif is essential for recognition of the sugar donor [5,11,12]. In contrast, the N-terminal domain, which mediates acceptor recognition and binding, exhibits substantial amino-acid sequence variability; nonetheless, these domains retain a conserved Rossmann fold in their tertiary structures [1,13].

Plant UGTs facilitate the transfer of glycosyl group from UDP-activated donors, most commonly UDP-glucose, to a highly diverse array of natural products, including flavonoids, phenolic acids, sterols, alkaloids, terpenoids, and plant hormones [5,7,14,15,16,17,18]. The broad substrate diversity demonstrated by UGTs establishes them as crucial regulators in various biological processes. These include, but are not limited to, developmental regulation, secondary metabolism, hormone homeostasis, immune defense, responses to environmental stress, and detoxification reactions [19,20,21,22,23,24,25,26].

Among the substrates catalyzed by UGTs, flavonoids constitute a class of natural polyphenol substances found across the plant kingdom [5]. Beyond their indispensable roles in plant growth and development, these compounds exhibit potent biological activities, including antimicrobial, antioxidant, antiallergic, anti-inflammatory, and anticancer activity, thereby presenting substantial therapeutic potential for human health [27,28,29,30]. For example, quercetin 3-O-glycoside, synthesized via UGT78D1 in A. thaliana, exhibits therapeutic efficacy against cancer, liver injury, and cardiovascular diseases [31,32,33,34]. Kaempferol 3-O-glucoside synthesized by safflower (Carthamus tinctorius L.) CtUGT4 shows pharmacological activities against liver damage [35,36]. Consequently, biosynthesis, bioactivities, and potential applications of flavonoids have garnered significant research and medical attention.

Medicago truncatula serves as a well characterized model legume, possessing a high-quality reference genome and a variety of genetic resources [37,38]. This species biosynthesizes and accumulates a diverse array of glycosylated natural products, which are predominantly flavonoids (e.g., flavones and anthocyanins), isoflavonoids (including isoflavones and pterocarpans), and triterpene saponins [28,39]. This accumulation suggests the existence of a broad repertoire of UGTs specifically dedicated to the biosynthesis of these compounds. Consistent with this, there are 272 UGTs annotated in M. truncatula genome [12].

However, only a relatively small fraction of UTGs have been functionally investigated. For instance, M. truncatula UGT72L1 has been characterized as a key enzyme in the biosynthesis of epicatechin 3′-O-gulcoside [40]. M. truncatula UGT71G1 is involved in the glycosylation of flavonoids (quercetin and 3-hydroxyflavone) and isoflavonoids (genistein and biochanin A) [41,42]. M. truncatula UGT78G1 exhibits a broad glycosyltransferase activity toward isoflavones (formononetin), flavonols (quercetin, kaempferol, and myricetin), and anthocyanidins (cyanidin and pelargonidin) [1,5]. M. truncatula UGT85H2 demonstrates catalytic activity toward various phenylpropanoid-derived metabolites, including the flavonols, isoflavones, and chalcones [43].

In plants, UGT families are organized into 18 groups. Among these, group F which comprises exclusively the UGT78 family, primarily catalyzes the glycosylation of flavonoids [5,12,44,45,46,47]. In this study, we report the identification of the glycosyltransferase UGT78G3, which belongs to the UGT78 family. An in vitro enzyme assay demonstrated that UGT78G3 catalyzes the glycosylation of kaempferol using UDP-glucose as the glycosyl donor. However, in vivo functional validation via either overexpression or CRISPR/Cas9-mediated mutation of UGT78G3 did not result in significant accumulation of kaempferol 3-O-glucoside in planta, which suggests that the physiological function of UGT78G3 may be negligible or dispensable, likely due to the presence of alternative pathways.

2. Results

2.1. Identification of UGT78G3 as a Candidate Glycosyltransferase in M. truncatula

To elucidate the function of the UGT78 family and identify new candidates involved in the glycosylation of flavonoids, we utilized the model legume M. truncatula, which is characterized by its accumulation of diverse flavonoid glucuronides [22]. Sequence analysis revealed that the M. truncatula genome contains only two annotated members of the UGT78 family: UGT78G1 (Medtr4g128690) and UGT78G3 (Medtr4g128720) (Figure 1A). Notably, UGT78G1 has been characterized as a flavonoid glycosyltransferase, demonstrating broad substrate specificity toward flavonoids and isoflavonoids [5].

To investigate the evolutionary history of UGT78G3, homologous protein sequences were retrieved from the NCBI database for 17 representative plant species, comprising ten monocotyledons and seven dicotyledons. Subsequently, multiple sequence alignment was performed using CLUSTALW, and an unrooted phylogenetic tree was constructed via the neighbor-joining algorithm. The result revealed that M. truncatula UGT78G3 exhibits high degrees of sequence homology with UGT78G38, UGT78G39, and UGT78G40 in Pisum sativum, UGT78G33 in Arachis hypogaea, UGT78DW2 in Gossypium hirsutum, and UGT78A23, UGT78A24, UGT78A26, and UGT78A27 in Vitis vinifera (Figure 1A), suggesting that these enzymes may act on similar classes of substrates.

Amino-acid sequence alignment delineated a highly conserved PSPG motif within the C-terminus of UGT78G3 and its homologs (Figure 1B), which is essential for the interaction with the sugar donor. Furthermore, a three-dimensional structural model of the MtUGT78G3 protein was generated using the SWISS-MODEL homology-modeling server (https://swissmodel.expasy.org/) (accessed on 1 December 2025). The results revealed that UGT78G3 adopts a conserved GT-B fold characterized by two Rossmann domains (Figure 1C), a structural configuration typical of the glycosyltransferase superfamily. Collectively, these findings suggest that UGT78G3 functions as a potential glycosyltransferase.

2.2. Characterization of the ugt78g3 Mutant in M. truncatula

For functional characterization of UGT78G3, we sought knock-out mutants within the M. truncatula Tnt1 retrotransposon insertion mutant population [48]. Line NF8463 was identified, which harbors a Tnt1 retrotransposon insertion within the UGT78G3 promoter region, located 204 bp upstream of the translational start codon (Figure 2A). Seeds of NF8463 were germinated and genotyped via PCR to identify homozygous individuals for the insertion. RT-PCR analysis was subsequently performed to assess the impact of the Tnt1 insertion on the transcription of UGT78G3. The results demonstrated that the full-length of the UGT78G3 transcript was abolished in these homozygous plants (Figure 2B), and this mutant line was designated ugt78g3. Phenotypic analysis revealed that ugt78g3 mutants exhibit no obvious growth defects; specifically, the plant height, leaf number, and number of internodes were comparable to those of the wild type R108 (WT) (Figure 2C–F).

2.3. Expression Analysis of UGT78G3 and Subcellular Localization of UGT78G3

To examine the temporal–spatial expression profile of UGT78G3, total RNA was isolated from various tissues of the WT (Figure 3A), followed by cDNA synthesis via reverse transcription. Quantitative reverse transcription PCR (qRT-PCR) analysis showed that UGT78G3 is expressed in all examined tissues, including root, stem, leaf, flower, and pod, with a relatively high expression in flower (Figure 3B).

To investigate the subcellular localization of UGT78G3, the 35S:UGT78G3-GFP vector was constructed by ligating GFP to the C-terminus of UGT78G3, under the control of cauliflower mosaic virus (CaMV) 35S promoter. The resulting construct was introduced into tobacco (Nicotiana benthamiana) leaf epidermal cells via Agrobacterium-mediated infiltration. A construct expressing GFP, driven by 35S promoter, was employed as the positive control. Consistent with the results obtained for the GFP control, the UGT78G3-GFP fusion proteins exhibited localization to both the cytoplasm and nucleus (Figure 3C). To further validate this result, the full-length coding sequence of UGT78G3 fused with a N-terminal Myc tag, under the control of 35S promoter, was introduced into the WT. The nuclear and cytoplasmic fractions were isolated from the transgenic 35S:Myc-UGT78G3 seedlings, and Myc-UGT78G3 proteins were detected by immunoblot analysis. The results showed that Myc-UGT78G3 proteins were present within both nuclear and cytoplasmic fractions (Figure 3D), confirming that UGT78G3 is localized to the cytoplasm and nucleus.

2.4. UGT78G3 Is Involved in the Glycosylation of Flavonoids

To characterize the substrate specificity and catalytic activity of UGT78G3, we initially evaluated the transcript abundance of genes involved in the flavonoid biosynthesis pathway. The results revealed that the majority of the detected genes exhibited significant changes in expression. Specifically, the expression levels of MtCHS, MtF3’H, MtF3’,5’H, and MtFLS were significantly upregulated, whereas the transcripts for MtCHI and MtF3H were markedly suppressed in the ugt78g3 mutant relative to the WT (Figure 4A–G). Subsequently, the total flavonoids in the ugt78g3 seedlings were quantified relative to the WT. The quantitative results showed that total flavonoids accumulation was significantly decreased in the ugt78g3 mutant compared to WT (Figure 4H). Collectively, these results demonstrated that the flavonoids biosynthesis and metabolism were profoundly altered in ugt78g3, indicating the potential role for UGT78G3 in mediating the glycosylation and metabolic homeostasis of flavonoids.

In addition, flavonoid metabolites from ugt78g3 and the WT seedlings were extracted and quantified by LC-MS/MS using Xevo^TM^ TQ-S triple quadrupole tandem mass spectrometer (Waters, Milford, MA, USA). The quantified metabolites are listed in Supplemental Table S2. Notably, the concentration of kaempferol 3-O-rutinoside was significantly elevated in the ugt78g3 mutant compared to WT (Figure 4I). This accumulation likely results from a metabolic flux shift toward kaempferol 3-O-rutinoside triggered by the loss of UGT78G3 activity.

2.5. UGT78G3 Exhibits Glycosyltransferase Activity Toward Kaempferol In Vitro

To clarify the catalytic activity of UGT78G3 in the glycosylation of kaempferol, we obtained the recombinant UGT78G3 protein (Figure S1) and performed an in vitro enzyme assay. The UDP-glucose (UDP-Glc) was selected as the sugar donor as it represents the most commonly sugar source utilized by plant UGTs [1]. The chromatogram for the kaempferol standard was obtained through LC-MS/MS. Chromatographic separation revealed a retention time of 8.07 min for kaempferol. Subsequent mass spectrometry characterization in the positive ion mode identified the precursor ion at m/z 285.1, which yielded a major product ion at m/z 93.1 under a collision energy of 80 V (Figure 5A).

Following the incubation of kaempferol and UDP-Glc with the enzyme UGT78G3, LC-MS/MS analysis revealed a discrete product peak (Figure 5C). The observed mass spectral characteristics of this peak were consistent with the authentic standard for kaempferol 3-O-glucoside, which exhibited a retention time of 7.45 min, a precursor ion at m/z 449.2, and a major product at m/z 287.2 (Figure 5B,C). These results strongly suggest that kaempferol serves as a substrate of UGT78G3-mediated glycosylation.

2.6. UGT78G3 Overexpression or Mutation Does Not Significantly Affect the Accumulation of Kaempferol 3-O-Glucoside in Planta

To validate the in planta function of UGT78G3, we constructed CRISPR/Cas9-targeted mutated lines of UGT78G3. Two guide RNA, spaced 304 bp apart, were designed to target the first exon of UGT78G3 (Figure 6B,C). These sequences were cloned into the pHSE401-MtU6 vector and introduced into the WT background via Agrobacterium-mediated genetic transformation. Three independent mutated lines were identified: line ugt78g3-03 carries a 4 bp deletion, line ugt78g3-35 contains concurrent 1 bp and 117 bp deletion at distinct site, and line ugt78g3-36 possesses a 38 bp deletion (Figure 6A,C). Each of these alterations results in frameshift mutation, leading to premature translation termination. The phenotype of the mutant lines was comparable to that of the WT (Figure 6D–F). We subsequently analyzed the concentration of kaempferol 3-O-glucoside. The results revealed that its levels in the mutant lines were not significantly different from those in the WT (Figure 6G).

We then generated transgenic lines overexpressing Myc-UGT78G3 driven by the constitutive 35S promoter (designated as UGT78G3-OE) (Figure 7A). qRT-PCR analysis demonstrated that the UGT78G3 transcript levels in these transgenic plants were over 100-fold higher than those in the WT (Figure 7B). Furthermore, the accumulation of the Myc-UGT78G3 fusion proteins was confirmed by Western blot analysis (Figure 7C). Phenotypic characterization revealed no discernible morphological differences between the overexpression lines and the WT under standard growth conditions (Figure 7D–F). Next, the flavonoid metabolites were extracted from the UGT78G3-OE and WT plants, and the concentration of kaempferol 3-O-glucoside were quantified. Although kaempferol 3-O-glucoside levels were significantly reduced in UGT78G3-OE#13, no significant changes were observed in UGT78G3-OE#14 and UGT78G3-OE#38 plants relative to the WT (Figure 7G).

Taken together, these results demonstrate that neither UGT78G3 overexpression nor mutation significantly affects the accumulation of kaempferol 3-O-glucoside in planta.

3. Discussion

UGTs comprise a diverse group of enzymes that are ubiquitously distributed across various organisms, which mediate the glycosylation of an expansive spectrum of natural products, underscoring their pivotal contributions to secondary metabolite biosynthesis, hormone homeostasis, plant defense, stress response, and detoxification reaction [1,22,24,25]. Flavonoids, which influence plant organ development and exhibit antioxidant and anticancer activities with substantial health benefits for humans, are commonly found in the form of glycosylated derivatives [5,27,29]. This modification, typically orchestrated by UGTs, diversifies their structure and modulates their physicochemical properties, thereby playing a critical role in their bioactivity [13]. Identification of UGTs and functional characterization of UGTs-mediated flavonoid glycosylation not only promises to elucidate the intricate metabolic pathway but also provides a range of available novel flavonoid UGTs for the development of active glycodrugs.

M. truncatula, an annual forage legume, has been employed as a model legume plant due to its simpler genetics and the extensive availability of genetic resources [37,48]. Furthermore, this species accumulates a diverse array of glycosylated flavonoids, making it a valuable model for investigating flavonoid biosynthesis and its associated modifications. While 272 UGTs are annotated in the M. truncatula genome, only a small subset has been functionally investigated [12]. Representative members include UGT71G1, UGT72L1, UGT78G1, UGT84F9, UGT85H2, and so on [5,22,37,40,41,42,43].

In the present study, we identified a putative glycosyltransferase UGT78G3, a member of the UGT78 family (Figure 1A). Primary sequence analysis revealed that UGT78G3 harbors a highly conserved PSPG motif within its C-terminal domain (Figure 1B), which represents a signature identifying plant UGTs. The predicted three-dimensional structural model of the MtUGT78G3 protein showed that it contains two Rossmann domains, each comprising a central β-sheet flanked by α-helices (Figure 1C). These features confirmed that UGT78G3 possesses characteristics typical of plant UGT structures. Furthermore, a marked reduction in total flavonoids content within the retrotransposon Tnt1-insertion mutant ugt78g3, coupled with the dynamic changes in the expression of flavonoid synthesis gene (Figure 4), underscored the requirement of UGT78G3 in maintaining flavonoid homeostasis. A comprehensive examination of flavonoid metabolites from the ugt78g3 mutant compared to the WT highlighted a markedly increased concentration of kaempferol 3-O-rutinoside (Figure 4), which may result from a metabolic flux shift triggered by the loss of UGT78G3 activity. In accordance with this, the in vitro incubation of kaempferol and the sugar donor with the recombinant enzyme UGT78G3 resulted in the production of kaempferol 3-O-glucoside (Figure 5), confirming that UGT78G3 functions as a glycosyltransferase that directly mediates the glycosylation of kaempferol. However, the levels of kaempferol 3-O-glucoside remained largely unaffected in M. truncatula, regardless of whether UGT78G3 was overexpressed or disrupted via CRISPR/Cas9 (Figure 6). The results suggest that while UGT78G3 exhibits enzymatic activity toward kaempferol to generate kaempferol 3-O-rutinoside in vitro, it may not function as a primary glycosyltransferase for this substrate in vivo. This discrepancy between in vitro and in vivo results also suggest that other redundant glycosyltransferases may compensate for the loss of UGT78G3. One such candidate is UGT78G1, which has been shown to catalyze the generation of kaempferol 3-O-glucoside and shares sequence similarity with UGT78G3 (Figure 1A). The generation of a ugt78g3 ugt78g1 double mutant will help elucidate the in vivo function of UGT78G3 and clarify the extent of functional redundancy between these two glycosyltransferases.

The sugar donors utilized by UGTs include UDP-glucose, UDP-glucuronic acid, UDP galactose, UDP-rhamnose, among others. Among these, UDP-glucose is the predominant sugar donor for plant UGTs. In M. truncatula, several UGTs have been characterized that are capable of glycosylating flavonoids with UDP-glucose as the sugar donor [22,28]. However, it has also been reported that M. truncatula UGT84F9 can use UDP-glucuronic acid as a sugar donor. Here, we employed UDP-glucose as the sugar donor to investigate the enzymatic activity of M. truncatula UGT78G3. Our in vitro enzyme activity experiments confirmed that UGT78G3 successfully catalyzes the generation of kaempferol 3-O-glucoside using the UDP-glucose as the sugar source. However, the preference and the identification of the optimal sugar donor for UGT78G3 still require further investigation.

Plant UGTs facilitate the attachment of sugar moieties to a diverse range of substrate molecules. While some UGTs exhibit a pronounced substrate selectivity, directing glycosylation reactions toward specific acceptor molecules, others display significant substrate promiscuity, acting on a broad range of chemical groups [1]. For example, M. truncatula UGT78G1 exhibits broad activities toward different types of flavonoids, isoflavonoids, and triterpenoids [5]. In this study, we demonstrated that UGT78G3 shows catalytic activity to flavonol kaempferol in vitro. Further investigation into the substrate specificity of UGT78G3 will be essential to fully elucidate its biological function and its contribution to the M. truncatula metabolic landscape.

4. Materials and Methods

4.1. Plant Materials and Growth Conditions

The M. truncatula genotypes utilized in this study include the wild-type R108 (WT) and the mutant NF8463 (ugt78g3), which was obtained from screening a Tnt1 retrotransposon-inserted mutant population in R108 [48]. The seeds were mechanically scarified using sandpaper, and then imbibed and germinated on moistened filter paper at 4 °C. Following germination, seedlings were transplanted into pots containing a growth substrate consisting of an equal volume blend (1:1 v/v) of soil and vermiculite. Plants were grown in a greenhouse under standardized conditions: a 16/8 h light/dark and a 24 °C day/20 °C night temperature cycle. All primers used are listed in Supplemental Table S1.

4.2. RNA Extraction and Gene Expression Analysis

Total RNA was isolated from different tissues of M. truncatula using TRIzol^®^ Reagent (TAKARA, Kusatsu, Shiga, Japan). Reverse transcription was performed using HiScript^®^ II QRT SuperMix (Vazyme, Nanjing, China) according to the manufacture’s protocol. Quantitative RT-PCR (qRT-PCR) was performed as previously described, using MtActin as an internal control [49]. All primers used are listed in Supplemental Table S1.

4.3. Vector Construction and Plant Transformation

To generate the construct for overexpression of UGT78G3 in M. truncatula, the UGT78G3 coding sequence was amplified from WT and integrated into the Gateway entry vector pCR8. The resulting construct was then utilized in a site-specific recombination reaction (Gateway LR reaction, Invitrogen, Carlsbad, CA, USA) with the plant binary vector pEarleyGate203 to generate the final 35S:Myc-UGT78G3 construct. To facilitate targeted mutagenesis of UGT78G3 using CRISPR/Cas9 system, a specific guide RNA (gRNA) was designed via CRISPR-P web tool (http://crispr.hzau.edu.cn/CRISPR2/help.php) (accessed on 1 June 2024) and inserted into the binary vector pHSE401-MtU6. The constructs were transformed into Agrobacterium tumefaciens strain AGL1 and used for M. truncatula transformation as described [49].

To generate the construct for subcellular localization analysis of UGT78G3, the intermediate construct pCR8-UGT78G3 was recombined with the pMDC83 vector to generate the 35S: UGT78G3-GFP construct.

To generate the construct for prokaryotic expression of UGT78G3, UGT78G3 coding sequence was cloned into the E. coli expression vector pMAL-c2X, downstream of the maltose-binding protein (MBP) tag. This was achieved using EcoRI and SalI restriction enzymes to generate the MBP-UGT78G3 recombinant construct.

All primers used are listed in Supplemental Table S1.

4.4. Subcellular Localization Analysis

The 35S:UGT78G3-GFP construct, the nuclear marker construct 35S:mRFP-AHL22, and the positive control 35S:GFP were independently introduced into A. tumefaciens strain GV3101. Agrobacterium cultures containing the 35S:UGT78G3-GFP or 35S:GFP were mixed with the 35S:mRFP-AHL22 and co-infiltrated into the abaxial epidermis of 4-week-old Nicotiana benthamiana leaves. To minimize gene silencing, the P19 suppressor was included in the infiltration mixture. Fluorescence signals were examined 3 days post-infiltration using a confocal laser scanning microscopy (FV3000, Olympus, Tokyo, Japan).

4.5. Sequence Alignment and Phylogenetic Analysis

Amino-acid sequence alignment was conducted using CLUSTALW (http://www.genome.jp/tools/clustalw/) (accessed on 1 January 2024). A phylogenetic tree was constructed using the neighbor-joining algorithm in MEGA 11.0, with branch support assessed through 1000 bootstrap replicates.

4.6. Flavonoid and Anthocyanin Measurement

Total flavonoid and anthocyanin content were measured by using commercial assay kits G0118F and G0126F (Grace Biotechnology, Suzhou, China) according to the manufacturer’s instructions. Upon completion of the reactions, absorbance was measured at the respective specified wavelengths using a Multiskan Go monochromator-based UV/VIS spectrophotometer (Thermo Scientific, Waltham, MA, USA).

4.7. Enzyme Assays

The MBP-UGT78G3 recombinant construct was transformed into E. coli strain Rosetta and induced with 0.2 mM IPTG. Recombinant MBP-UGT78G3 and MBP control were purified using Amylose Resin (NEB, Ipswich, MA, USA). The purified MBP-UGT78G3 was quantified by the Bio-Rad (Hercules, CA, USA) protein assay reagent.

The glycosylation reaction for MBP-UGT78G3 was performed based on Adiji et al. [22]. Briefly, the enzyme UGT78G3 (0.01–0.1 mg/mL) was mixed with UDP-glucose (1 mM) and acceptor substrate (0.1 mM) in 50 mM Tris-HCl buffer (pH 7.5) to reach a final volume of 100 μL. The reaction mixture was incubated at 30 °C for 12 h to allow for product accumulation. The enzymatic reaction was stopped with 100 μL of methanol, with the resulting products characterized using a UPLC-MS/MS system (UPLC 1290 Infinity II, Agilent Technologies, Santa Clara, CA, USA).

4.8. LC-MS/MS Analysis

LC-MS/MS analysis was conducted with Xevo^TM^ TQ-S triple quadrupole tandem mass spectrometer (Waters, Milford, MA, USA). Chromatographic separation was achieved on an infinityLab poroshell 120 EC-C18 column (2.1 × 150 mm, 2.7 μm, Agilent Technologies, Santa Clara, CA, USA). Mobile phase A and B were composed of 0.1% formic acid and 100 pure acetonitrile, respectively. The flow gradient was 90% A, 10% B, 0–1 min; 90–80% A, 10–20% B, 1–5 min; 80–5% A, 20–95% B, 5–7 min; 5% A, 95% B, 7–8 min; 5–90% A, 95–10% B, 8–8.1 min; and 90% A, 10% B, 8.1–10 min. An amount of 1 μL sample was injected for each analysis. The mass spectra were obtained with the following operating parameters: capillary voltage of 1 kV, source offset of 50 V, source temperature of 150 °C, desolvation temperature of 500 °C. The cone and desolvation gas flows were maintained at 150 L/h and 1000 L/h, respectively.