Glycosylation and methylation in the biosynthesis of isoflavonoids in Pueraria lobata

The pathway for forming isoflavonoid skeletal structure is primarily restricted to the Leguminosae family. Subsequent decorations on the compound backbone by tailoring enzymes would change their biological and medicinal properties. Pueraria lobata is a leguminous plant, and as a traditional Chinese medicine its roots have been ascribed a number of pharmacological activities. Glycosylation and methylation are the main modifying processes in isoflavonoid metabolism in P. lobata roots, resulting in the accumulation of unique glycosylated and methylated end isoflavonoid compounds. For instance, daidzein 8-C-glucoside (i.e., puerarin) and puerarin derivatives are produced only by the Pueraria genus. Puerarin has been established as a clinical drug for curing cardiovascular diseases. To better understand the characteristic isoflavonoid metabolism in P. lobata, this review attempts to summarize the research progress made with understanding the main glycosylation and methylation of isoflavonoids in P. lobata and their biosynthetic enzymes.

Isoflavonoid compounds are considered the main bioactive components of P. lobata (Rong et al., 1998).A significant example is puerarin (i.e., daidzein 8-Cglucoside), an isoflavone that has been used as a prescribed drug in clinical practice for the treatment of cardiovascular diseases (Zhou et al., 2021).The biosynthetic pathway for the formation of isoflavonoid backbone is predominantly conserved in legumes (Falcone Ferreyra et al., 2012).Subsequent modifications on the isoflavonoid skeleton, such as glycosylation and methylation, result in the difference in isoflavonoid composition between different leguminous species.For instance, puerarin and its glycosylated and/or methylated derivatives (e.g.3'-methoxy puerarin, 6''-O-xylosylpuerarin and puerarin 4'-O-glucoside) are produced only by the species within the Pueraria genus (Wang et al., 2020), therefore conferring their unique medicinal value for human.Phytochemical studies revealed glycosylation and methylation as the two major modifications in isoflavonoid metabolism in P. lobata (Wang et al., 2020).As a consequence, pharmacological activities of P. lobata have focused mainly on the specifically glycosylated or/and methylated isoflavonoids, such as puerarin (Wang et al., 2022) and 3'methoxy puerarin (Zhao et al., 2007).There is intense interest in identifying the enzymes responsible for the glycosylation and methylation reactions for isoflavonoid metabolism in P. lobata.This review considers the recent progress in understanding the biochemistry of the glycosylation and methylation for isoflavonoid metabolism in P. lobata.
Identification of UGTS and OMTS acting on isoflavonoids in P. Lobata

8-C-glycosyltransferase
The 8-C-glycosylation is required for puerarin formation.There is great interest in understanding the biochemical process for the formation of the 8-C-glycosyl group in puerarin (Inoue and Fujita, 1977;Wang et al., 2017;Bao et al., 2022).Some data remain contradictory, particularly regarding the step at which the 8-Cglucosyl group is introduced (Figure 2).Early labeling studies had proposed an upstream intermediate isoliquiritigenin at the chalcone stage, but not daidzein at the isoflavone stage, as an acceptor for the 8-C-glycosylation (Inoue and Fujita, 1977).However, an enzyme assay using the Pueraria root crude protein provided an implication that the C-glucosyl unit in puerarin might be introduced at the isoflavanone stage (He et al., 2011).This assumption is prone to being considered because that a number of flavone C-GTs recognize 2-hydroxyflavanone intermediates as their natural substrates (Brazier-Hicks et al., 2009;Nagatomo et al., 2014;Hirade et al., 2015).Nonetheless, Xi et al. revealed that there are no orthologs of 2-hydroxyflavanone C-GTs in P. lobata (Xi H.T. et al., 2023), indicating that if the C-glycosylation for puerarin biosynthesis occurs at the isoflavanone stage, it is probably catalyzed by a phylogenetically distinct UGT.A very recent labeling study provided evidence that both isoliquiritigenin and daidzein could be incorporated into puerarin in vivo (Adolfo et al., 2022), suggesting that the 8-C-glycosylation can happen at either the chalcone or isoflavone stage, or simultaneously at both levels.
For the first time, a P. lobata C-GT (namely PlUGT43), which directly transfers a glucose group to the C-8 position of daidzein leading to puerarin, was molecularly cloned from the root of P. lobata by Wang et al. (Wang et al., 2017).Through the in vitro assays, PlUGT43 was found to have no or negligible activity with the chalcone intermediate isoliquiritigenin (Wang et al., 2017).However, its isoform (officially named UGT71T5), which shares 99.72% sequence identity with PlUGT43, was recently reported to be capable of catalyzing the Cglycosylation activity against both daidzein and isoliquiritigenin (Adolfo et al., 2022).Incubation of the recombinant PlUGT43 or UGT71T5 with 2-hydroxyisoflavanone did not generate a product matching the 2-hydroxyisoflavanone C-glycoside (Wang et al., 2017;Adolfo et al., 2022), indicating that they had no C-glycosylation activity with 2-hydroxyisoflavanone.The RNAi-mediated down-regulation of UGT71T5 caused a strong reduction in the levels of puerarin in P. lobata hairy roots (Adolfo et al., 2022), confirming that PlUGT43 (UGT71T5) functions as a C-GT at least partially for puerarin biosynthesis in P. lobata.Interestingly, when the 2-HIS (2hydroxyisoflavanone synthase; see its place in the pathway in Figure 2), which is the entry enzyme catalyzing the formation of isoflavonoid backbone (Steele et al., 1999;Jung et al., 2000), was downregulated in P. lobata hairy roots, the potential C-glycosides of isoliquiritigenin and/or liquiritigenin significantly accumulated, when compared to that in the control roots (Adolfo et al., 2022).This data clearly supports that introduction of the C-glucosyl group can take place at the chalcone stage and then the 2-HIS would be able to accommodate the C-glycosides as substrates (Figure 2).Recently, another variant of PlUGT43, designated PlCGT, was identified from P. lobata by Ye et al. (Bao et al., 2022).PlCGT shows 97.01%amino acid identity with PlUGT43, and this variant essentially catalyzes the same C-glycosylation activity as PlUGT43 (Bao et al., 2022).An ortholog of PlUGT43, named PtUGT8, was also isolated from P. thomsonii species (Duan et al., 2022).Despite exhibiting a high sequence identity (96.52%) to PlUGT43, PtUGT8 was shown as primarily having a 7-O-glucosylation activity toward isoflavones whereas not catalyzing the 8-C-glycosylation reaction as does by PlUGT43 (Duan et al., 2022).In view of a previous finding that Oand C-GT can be easily shifted by changing only a few amino acids (Gutmann and Nidetzky, 2012), a subtle difference in the active sites of PtUGT8 and PlUGT43 may plausibly account for this discrepancy.
Taken together, combination of the published data from the in vitro assays using the recombinant UGTs, the in vivo labeling experiments, and the transgenic studies of P. lobata hairy roots strongly supports that during puerarin biosynthesis, the Cglucosylation reaction takes place most likely at either the chalcone or isoflavone stage, or both.
PlUGT2 could also catalyze 4'-or 7-O-glucosylation of puerarin (Wang et al., 2016), which is currently a clinical drug for curing cardiovascular diseases (Zhou et al., 2021).Although puerarin is currently a prescribed drug, its low water solubility is still a serious drawback in clinical applications (Wang et al., 2012;Chen et al., 2021).Glycosylation is an efficient way to increase water solubility (Liu et al., 2016), thus, the identification of PlUGT2 would provide such an opportunity.
One OMT, designated PlOMT4, was cloned (Li et al., 2016b) from P. lobata based on the P. lobata transcriptome database (Wang et al., 2015).Tissue-specific expression analysis revealed that PlOMT4 was expressed most highly in roots, and its transcript was up-regulated by MeJA (Li et al., 2016b).PlOMT4 was found to have the activity of methylating 3'-hydroxy daidzein to form 3'methoxy-daidzein (Li et al., 2016b).PlOMT4 has no activity with the isoflavonoid substrates with free hydroxyl groups at either C7 or C4' (Li et al., 2016b), suggesting the methylation activity of PlOMT4 is region-specific.In addition, PlOMT4 is inactive with 3'-hydroxy puerarin (Li et al., 2016b), indicating that the 8-C-glucosylation of 3'-hydroxy daidzein prevents methylation at the 3'-position, and thereby the 3'-methylation should take place prior to the 8-Cglycosylation during 3'-methoxy-puerarin biosynthesis.PlOMT4 seems to be the only isoflavone specific 3'-O-methyltransferase so far identified from plant species.

4'-O-methyltransferase
In the early 1970s, scientists began a search for the isoflavone 4'-O-methyltransferase (I4'OMT) from plants.At a protein level, a methyltransferase, which catalyses the 4'-O-methylation of the isoflavone daidzein, was purified from Cicer arietinum L (Wengenmayer et al., 1974), indicating that the 4'-O-methylation for biosynthesis of 4'-O-methylated isoflavonoids can take place at the isoflavone stage.However, in alfalfa (Medicago sativa L.) seedlings, radiolabeled daidzein is not incorporated into 4'-Omethylated isoflavonoids (Dewick and Martin, 1979).Paradoxically, biosynthesis of the 4'-O-methylated isoflavonoids in alfalfa suspension cells strongly correlates with the isoflavone 7-O-methyltransferase (I7OMT) activity (Edwards and Dixon, 1991;He et al., 1998), and over-expression of the I7OMT led to enhanced levels of 4'-O-methylated isoflavonoids in the elicited alfalfa leaves (He and Dixon, 2000).In the elicited alfalfa leaves, the operationally soluble I7OMT re-locates to the endoplasmic reticulum where the 2-HIS naturally resides, leading to an interesting hypothesis that the association with other isoflavonoid pathway enzymes may change the region-specificity of I7OMT from the 7-to 4'-position in vivo (Liu and Dixon, 2001).On the other hand, an enzyme assay with the Glycyrrhiza echinata cell-free extract demonstrated that the 4'-O-methylation occurs at the level of 2,7,4'-trihydroxyisoflavanone (Akashi et al., 2000).This is supported by the molecular cloning and characterization of cDNAs encoding the 2,7, 4'-tri-hydroxyisoflavanone 4'-Omethyltransferase (HI4'OMT) from Glycyrrhiza echinata (Akashi et al., 2003), andMedicago truncatula (Deavours et al., 2006).The isoflavone daidzein could not be converted by HI4'OMT, and it only recognizes 2-trihydroxy-isoflavanone as the direct methyl acceptor (Akashi et al., 2003), suggesting that HI4'OMT catalyzes the 4'-O-methylation reaction only at the isoflavanone stage.Therefore, the history leading to the finding of isoflavonoid 4'-O-methyltransferases demonstrates two alternative pathways likely involved: one is the simplest 4'-O-methylation occurring at the stage, and the other is the reaction performed at the level of 2-hydroxyisoflavanones.From P. lobata, Li et al. identified a novel isoflavone 4'-O-methyltransferase (designated PlOMT9) that is capable of directly 4'-O-methylating isoflavones (Li et al., 2016a).Because that PlOMT9 shows the highest degree of amino acid identity with the isoflavone 7-O-methyltransferases (I7OMTs), PlOMT9 was initially presumed as an I7OMT.However, yeast cells expressing PlOMT9 efficiently performed the 4'-Omethylation of daidzein, genistein, prunetin, and isoformononetin (Li et al., 2016a), demonstrating that PlOMT9 functions actually as a I4'OMT.The I4'OMT activity catalyzed by PlOMT9 was further confirmed by in vitro assays using the purified recombinant PlOMT9 (Li et al., 2016a).Moreover, the recombinant PlOMT9 was not active with 2,7,4'-trihydroxy-isoflavanone, which is the natural substrate of HI4'OMT (Akashi et al., 2003).In addition to the main I4'OMT activity, PlOMT9 retains an extremely low 7-Omethylation activity, such as O-methylating daidzein at C7 position to yield trace amounts of isoformononetin.Over-expression of PlOMT9 in Glycine max hairy roots increased the levels of formononetin and ononin (formononetin 7-O-glucoside) by 111.2% and 940.9%, respectively, in comparison with the controls.P. lobata contains a HI4'OMT-like enzyme (Li et al., 2016a), which shares 73% amino acid identity with the HI4'OMT from G. echinata (Akashi et al., 2003), but it is inactive either with 2,7,4'-trihydroxyisoflavanone or the isoflavone daidzein (Li et al., 2016a).

Conclusion and prospects
In summary, utilizing the transcriptomic analysis, in combination with in vitro biochemical analysis of recombinant protein, provides a strong basis for understanding the biosynthetic mechanism of glycosylation and methylation of isoflavonoids in P. lobata (Figure 2).For the O-glycosylation of isoflavonoids in P. lobata, either 7-O-or 4'-O-glycosyltransferase protein is from members of the UGT88E subgroup.For the C-glycosylation of isoflavonoids, PlUGT43 (official designated UGT71T5) is the only isoflavone C-GT identified from plants so far.For the Omethylation of isoflavonoids in P. lobata, both 3'-and 4'-Omethyltransferases perform the methylation reactions at the isoflavone stage, directly utilizing isoflavones as the best acceptors.
Of particular value among the isoflavonoids are puerarin and its derivatives, which are produced exclusively in Pueraria species.Puerarin has been established as a clinical drug to deal with cardiovascular diseases (Wang et al., 2022).By expressing the PlUGT43, in combination with other pathway genes, the production of puerarin directly from glucose could be achieved in yeast at a concentration of 72.8 mg/L (Liu et al., 2021).The poor water solubility of puerarin is still a challenge in narrowing its treatment window in clinical usage (Liu et al., 2016).Considering that glycosylation is the most effective way to increase water solubility of small molecules (Li et al., 2004), the PlUGT2, which is capable of glucosylating puerarin, would provide an excellent template for further designing novel enzymes to increase the water solubility of puerarin.
FIGURE 1The medicinal plant Pueraria lobata and the main glycosylated and methylated isoflavonoids accumulated in it.(A) P. lobata plant.(B) The dried roots of P. lobata used in traditional Chinese medicine.(C) Structures of the main isoflavone glucosides.(D) Structures of the main Omethylated isoflavonoids.