Biosynthesis of dillapiole/apiole in dill (Anethum graveolens): characterization of regioselective phenylpropene O-methyltransferase.

The phenylpropene volatiles dillapiole and apiole impart one of the characteristic aromas of dill (Anethum graveolens) weeds. However, very few studies have been conducted to investigate the chemical composition of volatile compounds from different developmental stages and plant parts of A. graveolens. In this study, we examined the distribution of volatile phenylpropenes, including dillapiole, in dill plants at various developmental stages. We observed that young dill seedlings accumulate high levels of dillapiole and apiole, whereas a negligible proportion was found in the flowering plants and dry seeds. Based on transcriptomics and co-expression approaches with phenylpropene biosynthesis genes, we identified dill cDNA encoding S-adenosyl-L-methionine-dependent O-methyltransferase 1 (AgOMT1), an enzyme that can convert 6- and 2-hydroxymyristicin to dillapiole and apiole, respectively, via the methylation of the ortho-hydroxy group. The AgOMT1 protein shows an apparent Km value of 3.5 µm for 6-hydroxymyristicin and is 75% identical to the anise (Pimpinella anisum) O-methyltransferase (PaAIMT1) that can convert isoeugenol to methylisoeugenol via methylation of the hydroxy group at the para-position of the benzene ring. AgOMT1 showed a preference for 6-hydroxymyristicin, whereas PaAIMT1 displayed a large preference for isoeugenol. In vitro mutagenesis experiments demonstrated that substituting only a few residues can substantially affect the substrate specificity of these enzymes. Other plants belonging to the Apiaceae family contained homologous O-methyltransferase (OMT) proteins highly similar to AgOMT1, converting 6-hydroxymyristicin to dillapiole. Our results indicate that apiaceous phenylpropene OMTs with ortho-methylating activity evolved independently of phenylpropene OMTs of other plants and the enzymatic function of AgOMT1 and PaAIMT1 diverged recently.

Structural basis for substrate recognition in the Phytolacca americana glycosyltransferase PaGT3.

Capsaicinoids are phenolic compounds that have health benefits. However, the pungency and poor water solubility of these compounds limit their exploitation. Glycosylation is a powerful method to improve water solubility and reduce pungency while preserving bioactivity. PaGT3, a uridine diphosphate glycosyltransferase (UGT) from Phytolacca americana, is known for its ability to glycosylate capsaicinoids and other phenolic compounds. While structural information on several UGTs is available, structures of UGTs that can glycosylate a range of phenolic compounds are rare. To fill this gap, crystal structures of PaGT3 with a sugar-donor analogue (UDP-2-fluoroglucose) and the acceptors capsaicin and kaempferol were determined. PaGT3 adopts a GT-B-fold structure that is highly conserved among UGTs. However, the acceptor-binding pocket in PaGT3 is hydrophobic and large, and is surrounded by longer loops. The larger acceptor-binding pocket in PaGT3 allows the enzyme to bind a range of compounds, while the flexibility of the longer loops possibly plays a role in accommodating the acceptors in the binding pocket according to their shape and size. This structural information provides insights into the acceptor-binding mechanism in UGTs that bind multiple substrates.

An Ambidextrous Polyphenol Glycosyltransferase PaGT2 from Phytolacca americana.

The glycosylation of small hydrophobic compounds is catalyzed by uridine diphosphate glycosyltransferases (UGTs). Because glycosylation is an invaluable tool for improving the stability and water solubility of hydrophobic compounds, UGTs have attracted attention for their application in the food, cosmetics, and pharmaceutical industries. However, the ability of UGTs to accept and glycosylate a wide range of substrates is not clearly understood due to the existence of a large number of UGTs. PaGT2, a UGT from Phytolacca americana, can regioselectively glycosylate piceatannol but has low activity toward other stilbenoids. To elucidate the substrate specificity and catalytic mechanism, we determined the crystal structures of PaGT2 with and without substrates and performed molecular docking studies. The structures have revealed key residues involved in substrate recognition and suggest the presence of a nonconserved catalytic residue (His81) in addition to the highly conserved catalytic histidine in UGTs (His18). The role of the identified residues in substrate recognition and catalysis is elucidated with the mutational assay. Additionally, the structure-guided mutation of Cys142 to other residues, Ala, Phe, and Gln, allows PaGT2 to glycosylate resveratrol with high regioselectivity, which is negligibly glycosylated by the wild-type enzyme. These results provide a basis for tailoring an efficient glycosyltransferase.

Crown-ether-mediated crystal structures of the glycosyltransferase PaGT3 from Phytolacca americana.

Uridine diphosphate glycosyltransferases (UGTs) are ubiquitous enzymes that are involved in the glycosylation of small molecules. As glycosylation improves the water solubility and stability of hydrophobic compounds, interest in the use of UGTs for the synthesis of glycosides of poorly soluble compounds is increasing. While sugar-donor recognition in UGTs is conserved with the presence of a plant secondary product glycosyltransferase (PSPG) motif, the basis of the recognition of the sugar acceptor and the regioselectivity of the products is poorly understood owing to low sequence identity around the acceptor-binding region. PaGT3, a glycosyltransferase from the plant Phytolacca americana, can glycosylate a range of acceptors. To illustrate the structure-function relationship of PaGT3, its crystal structure was determined. The sugar-donor and sugar-acceptor binding pockets in PaGT3 were recognized by comparison of its structure with those of other UGTs. The key feature of PaGT3 was the presence of longer loop regions around the hydrophobic acceptor-binding pocket, which resulted in a flexible and wider acceptor binding pocket. In this study, PaGT3 crystals were grown by co-crystallization with 18-crown-6 ether or 15-crown-5 ether. The crown-ether molecule in the asymmetric unit was observed to form a complex with a metal ion, which was coordinated on two sides by the main-chain O atoms of Glu238 from two molecules of the protein. The crown ether-metal complex resembles a molecular glue that sticks two molecules of PaGT3 together to enhance crystal growth. Thus, this result provides an insight into the substrate-recognition strategy in PaGT3 for the study of glycosyltransferases. Additionally, it is shown that crown ether-metal ion complexes can be used as a molecular glue for the crystallization of proteins.

HmuS from Yersinia pseudotuberculosis is a non-canonical heme-degrading enzyme to acquire iron from heme.

Some Gram-negative pathogens import host heme into the cytoplasm and utilize it as an iron source for their survival. We report here that HmuS, encoded by the heme utilizing system (hmu) locus, cleaves the protoporphyrin ring to release iron from heme. A liquid chromatography/mass spectrometry analysis revealed that the degradation products of this reaction are two biliverdin isomers that result from transformation of a verdoheme intermediate. This oxidative heme degradation by HmuS required molecular oxygen and electrons supplied by either ascorbate or NADPH. Electrons could not be directly transferred from NADPH to heme; instead, ferredoxin-NADP+ reductase (FNR) functioned as a mediator. Although HmuS does not share amino acid sequence homology with heme oxygenase (HO), a well-known heme-degrading enzyme, absorption and resonance Raman spectral analyses suggest that the heme iron is coordinated with an axial histidine residue and a water molecule in both enzymes. The substitution of axial His196 or distal Arg102 with an alanine residue in HmuS almost completely eliminated heme-degradation activity, suggesting that Fe-His coordination and interaction of a distal residue with water molecules in the heme pocket are important for this activity.

Mutagenesis studies of human cystathionine beta-synthase: residues important for heme binding and substrate interactions.

Human cystathionine beta-synthase (CBS) is a pyridoxal 5'-phosphate (PLP) dependent hemoprotein, which catalyzes the condensation of serine and homocysteine. Our mutagenesis studies suggest that Arg-266 is important to sense structural changes in heme-binding site, and that Gln-222 as well as Tyr-223 are involved in interactions with substrates.

Modulation of cystathionine beta-synthase activity by the Arg-51 and Arg-224 mutations.

Human cystathionine beta-synthase (CBS) catalyzes a pyridoxal 5'-phosphate (PLP) dependent beta-replacement reaction to synthesize cystathionine from serine and homocysteine. The enzyme is unique in bearing not only a catalytically important PLP but also heme. In order to study a regulatory process mediated by heme, we performed mutagenesis of Arg-51 and Arg-224, which have hydrogen-bonding interactions with propionate side chains of the prosthetic group. It was found that the arginine mutations decrease CBS activity by approximately 50%. The results indicate that structural changes in the heme vicinity are transmitted to PLP existing 20 A away from heme. A possible explanation of our results is discussed on the basis of CBS structure.