In vitro evaluation of the effect of C-4 substitution on methylation of 7,8-dihydroxycoumarin: metabolic profile and catalytic kinetics

Daphnetin (7,8-dihydroxycoumarin (7,8-DHC)) and its C-4 derivatives have multiple pharmacological activities, but the poor metabolic stability of these catechols has severely restricted their application in the clinic. Methylation plays important roles in catechol elimination, although thus far the effects of structural modifications on the metabolic selectivity and the catalytic efficacy of human catechol-O-methyltransferase (COMT) remain unclear. This study was aimed at exploring the structure–methylation relationship of daphnetin and its C-4 derivatives, including 4-methyl, 4-phenyl and 4-acetic acid daphnetin. It was achieved by identifying the methylated products generated and by careful characterization of the reaction kinetics. These catechols are selectively metabolized to the corresponding 8-O-methyl conjugates, and this regioselective methylation could be elucidated by flexible docking, in which all the 8-OH groups of these catechols are much closer than the 7-OH groups to catalytic residue LYS144 and methyl donor AdoMet. The results of the kinetic analyses revealed that the Clint values of the compounds could be strongly affected by the C-4 substitutions, which could be partially explained by the electronic effects of the C-4 substituents and the coordination modes of 7,8- dihydroxycoumarins in the active site of COMT. These findings provide helpful guidance for further structural modification of 7,8-DHCs to improve metabolic stability.

Y-LX, 0000-0002-5742-7182 Daphnetin (7,) and its C-4 derivatives have multiple pharmacological activities, but the poor metabolic stability of these catechols has severely restricted their application in the clinic. Methylation plays important roles in catechol elimination, although thus far the effects of structural modifications on the metabolic selectivity and the catalytic efficacy of human catechol-Omethyltransferase (COMT) remain unclear. This study was aimed at exploring the structure-methylation relationship of daphnetin and its C-4 derivatives, including 4-methyl, 4phenyl and 4-acetic acid daphnetin. It was achieved by identifying the methylated products generated and by careful characterization of the reaction kinetics. These catechols are selectively metabolized to the corresponding 8-O-methyl conjugates, and this regioselective methylation could be elucidated by flexible docking, in which all the 8-OH groups of these catechols are much closer than the 7-OH groups to catalytic residue LYS144 and methyl donor AdoMet. The results of the kinetic analyses revealed that the Cl int values of the compounds could be strongly affected by the C-4 substitutions, which could be partially explained by the electronic effects of the C-4 substituents and the coordination modes of 7,8dihydroxycoumarins in the active site of COMT. These findings provide helpful guidance for further structural modification of 7,8-DHCs to improve metabolic stability.
2018 The Authors. Published by the Royal Society under the terms of the Creative Commons Attribution License http://creativecommons.org/licenses/by/4.0/, which permits unrestricted use, provided the original author and source are credited.

Introduction
Coumarins are widely distributed in food and Chinese medical herbs. With excellent safety, high solubility, good permeability and extensive pharmacological activities, this family has received widespread attention [1]. Daphnetin (7,8-dihydroxycoumarin (7,) is a naturally occurring catechol from the family Thymelaeaceae. Similar to other coumarin derivatives, daphnetin has been reported to have many pharmacological actions including anti-malarial, anti-arthritic, anti-pyretic and anti-cancer properties [2][3][4][5]. 4-Methyl-7,8-DHC was found to exhibit important antioxidant strength, a low cytotoxicity, and could decrease reactive oxygen species production in malignant cell lines [6]. Despite the therapeutic benefits, this family suffers from low oral bioavailability due to poor metabolic stability [7,8]. Pharmacokinetic study after administration in rats shows that daphnetin is rapidly eliminated with quite a short half-life (15 min), due to extensive first-pass metabolism in the liver [7,8].
Owing to the existence of phenolic hydroxyl groups, catechol coumarins could be expected substrates of phase II metabolizing enzymes. In general, hydroxycoumarins could be metabolized by glucuronidation catalysed by uridine-5-diphosphate glucuronosyl transferases or sulfation catalysed by sulfotransferases, as is seen for 7-hydroxycoumarin and 4-methylumbelliferone [9,10]. Taking the catechol moiety into consideration, catechol-O-methyltransferase (COMT) may also participate in the metabolic elimination of catechol coumarins in vivo, such as fraxetin and daphnetin [11,12]. Recent studies demonstrated that daphnetin can be extensively metabolized in human liver to form six conjugated metabolites [13]. Compared with glucuronidation and sulfonation pathways, the methylation of daphnetin had a much higher clearance rate in human liver S9 fractions and contributed to a large amount (37.3%) of the methyl-derived metabolites in human hepatocytes. Reaction phenotyping studies showed a major role of soluble COMT in daphnetin 8-O-methylation. Because of the wide distribution of COMT in human liver, intestine, lung, brain and platelets etc., there is no doubt that methylation is essential in the disposition of these catecholic coumarins [14,15]. Methylation is a crucial pathway for daphnetin clearance and might be involved in pharmacologic actions of daphnetin in humans. However, the structure-methylation relationship of a series of 7,8-DHC analogues remains unclear; understanding of this relationship is necessary for designing structural modifications to improve their metabolic stability.
For the present case study, we selected four 7,8-DHC derivatives, daphnetin, 4-methyl daphnetin (4-MDPN), 4-phenyl daphnetin (4-PDPN) and 4-acetic acid daphnetin (4-ADPN) (figure 1), which contain varied groups substituted in C-4 position as model compounds. The effects of altering the substituent within the C-4 position of 7,8-DHC on the catalytic behaviour of COMT were explored. To the best of our knowledge, no other comparative or systematic studies have yet been published to address these aspects of the structure-methylation relationships of 7,8-DHCs.    then centrifuged at 20 000g, 4°C, for 20 min. Control incubations were performed either without human liver cytosol, or without SAM or without substrate. The supernatants were subjected to ultra-fast liquid chromatography (UFLC) coupled with a diode array detector (DAD) and an electrospray ionization (ESI) mass spectrometer (MS) analyzer. The method also displayed good reproducibility, with the intraday and interday variance both less than 3%. Mass detection was performed on a Shimadzu LC-MS-2010EV instrument with an ESI interface in both positive and negative ion mode from m/z 100 to 800. The detector voltage was set at +1.55 kV and −1.55 kV, for positive and negative ion detection, respectively. The curved desolvation line (CDL) temperature and the block heater temperature were set at 250°C, while the CDL voltage was set at 40 V. Other MS detection conditions were as follows: interface voltage, +4.5 kV and −4.0 kV for positive and negative ion detection, respectively; nebulizing gas (N 2 ) flow was 1.5 l·min −1 and the drying gas (N 2 ) pressure was set at 0.06 MPa. Data processing were performed using the LC/MS SOLUTION v. 3.41 software (Shimadzu, Kyoto, Japan).

Biosynthesis of metabolites and their structural identification
The major methylated conjugate of each substrate including daphnetin, 4-MDPN, 4-PDPN and 4-ADPN was biosynthesized using recombinant human S-COMT and purified for structure elucidation and quantitative analysis. In brief, substrate (1 mM) was incubated with recombinant human S-COMT (0.2 mg protein per ml), 0.1 M Tris-HCl (pH 7.4), 10 mM MgCl 2 , 4 mM DTT and 5 mM SAM in 1 ml of final incubation for 4 h at 37°C. This analytical-scale reaction was scaled up to a volume of 48 ml for each substrate. The stock solution of substrate (100 mM) was prepared in DMSO. The concentration of organic solvent in the final incubation was 1%. The reaction was terminated by the addition of icecold methanol, and then the vessels were transferred to an ice bath and cooled for 20 min. After the removal of protein by centrifugation at 20 000g for 30 min at 4°C, the combined supernatants were loaded on a solid phase extraction (SPE) cartridge (C18PN, 1000 mg, Acchrom Technologies, Beijing, China), which was preconditioned by sequential washing with 6 ml methanol and 6 ml Millipore water. After sample loading, the SPE cartridge was sequentially eluted with 12 ml Millipore water, 12 ml methanol and 12 ml methanol containing 5% formic acid. The entire process was monitored by UFLC-UV, and the methylated conjugate was collected in methanol containing 5% formic acid. After vacuum evaporation, the methylated conjugate of each 7,8-DHC was obtained as a powder and the purity was greater than 96%. The structure of each methylated conjugate was characterized by the NMR technique including 1 H-NMR and 13 C-NMR. All experiments were carried out on a Bruker Avance-400 NMR spectrometer (Bruker, Switzerland). Each methylated conjugate was stored at −80°C before dissolving in DMSO-d 6 (Euriso-Top, Saint-Aubin, France) for NMR analysis. Chemical shifts were given on δ scale and referenced to tetramethylsilane at 0 ppm for 1 H-NMR (400 MHz) and 13 C-NMR (150 MHz).
For the Michaelis-Menten kinetic model, where v is the rate of reaction, [S] is the substrate concentration, V max is the maximum velocity estimate and K m is the apparent affinity constant. All incubations were performed in duplicate. Kinetic constants were obtained using GraphPad PRISM 6 (GraphPad Software, Inc. La Jolla, CA) software.

Docking simulation of 7,8-dihydroxycoumarins into the reported structure of COMT
Flexible docking was performed using DISCOVERY STUDIO (BIOVIA Discovery Studio 2016, Dassault Systèmes, San Diego, USA) [16]. The protein structures of human COMT were taken from the Protein Data Bank [17,18]. Tasks including inserting missing atoms in incomplete residues, modelling missing loop regions, deleting alternate conformations (disorder), standardizing atom names and protonating titratable residues using predicted pK s were performed. Meanwhile, the four compounds were prepared by removing duplicates, enumerating isomers and tautomers, ionization and generating three-dimensional conformations before being used for the docking process. In the ionization process, the four compounds were prepared for treatment at pH 7.5. Herein, the anionic form of 4-ADPN was assumed in the following docking simulation. The CHARMM 40.1 force field was used to represent both the protein and ligand structures. Flexible docking, a fully automated molecular mechanics based induced-fit protein-ligand docking method, includes the following steps: (i) calculates receptor conformations using ChiFlex (a modified version of the ChiRotor sampling algorithm); (ii) creates ligand conformations; (iii) performs ligand docking into each active protein conformation site using LibDock; (iv) clustering to remove similar ligand poses regardless of the protein conformation; (v) refines selected protein side-chains in the presence of the rigid ligand using ChiRotor; and (vi) performs a final ligand refinement using CDOCKER [18,19]. Here in this work, residues included in the active site AC4 of 3BWY (the Protein Data Bank (PDB) code for the structure of human COMT), TRP38, MET40, ASP141, TRP143, LYS144, ASP169, ASN170, LEU198, GLU199, Mg 2+ and AdoMet, are selected to undergo conformation changes during docking. The distances between hydroxyl groups of 7,8-DHCs and the LYS144 residue, AdoMet and Mg 2+ were measured.

Molecular calculations
The pK a values of 8-hydroxyl groups of 7,8-DHCs were calculated using pK a estimation software (MARVINBEANS pK a estimation plug-in, v. 15.6.29, ChemAxon, Budapest, Hungary).

Discussion
In    and in recombinant S-COMT. Considering the catalytic mechanism, COMT-mediated O-methylation is reported to be a two-step S N 2 displacement reaction, which begins with proton transfer from the premethylated catechol-OH group to the amino group of Lys144, and then the ionized catechol-O-attacks the methyl group of AdoMet to yield the methylation product [14]. Consequently, a decrease in pK a value of pre-methylated catechol-OH group would facilitate the first step [22]. In this case, the pK a values of the 7-and 8-hydroxyl groups of 7,8-DHCs were determined using pK a estimation software, and a relatively lower pK a value of 8-hydroxyl of 4-PDPN may partially explain the higher catalytic efficiency of this compound compared with the other three. It could be assumed that the C-4 substitutes may influence the electronic cloud density of the 8-OH group of 7,8-DHCs and further impact the deprotonation process. Moreover, from the perspective of the enzymatic environment, a slightly more tight coordination of the 8-O atom of 4-PDPN to the metal centre, indicated by the shorter distance between the 8-O atom of 4-PDPN and Mg 2+ compared with the other three compounds, may also account for the relatively higher Cl int of 4-PDPN [21]. A more favourable interaction between 4-PDPN and the active site of human COMT could also be observed in our flexible docking simulation. For 4-PDPN, 14 of the 111 docked poses could form the desired octahedral coordination required for the methyl-transfer reaction to occur with the oxygen atoms in the catecholic hydroxyl group coordinating to the Mg 2+ rigidly with average distances of 2.10-2.30 Å. However, the amount of octahedral coordination formed in the case of the other three compounds is much less. For 4-ADPN, only one out of 57 docked poses are ligated to the Mg 2+ with average distances of 2.00-2.60 Å. It seems that an orientation effect provided by the C-4 phenyl group which is beneficial to the binding process of 7,8-DHC towards COMT and a relatively high efficiency of 'correct' binding would be beneficial to the subsequent methyl-transfer step. Hence, the catalytic efficiency is the comprehensive result jointly induced by both the physiochemical properties of the substrate and the interactions between substrate and enzyme. Further work should be conducted to better understand the mechanism accounting for the difference among 7,8-DHCs concerning their COMT-mediated catalysis.
The level of COMT activity in human tissues has been used as a biochemical index to assess individual differences in normal and in pathological conditions, and the availability of a sensitive assay could be useful in the development of new COMT inhibitors [23,24]. However, the applicability of currently available substrates for the measurement of COMT activity in human tissues could be relatively limited owing to the atypical kinetics and time-consuming chromatographic separation of the two methylated products [25,26]. Taken together, the formation of a mono-methylated product and the Michaelis-Menten kinetics of methylation implies that 7,8-DHC derivatives can serve as practical probes for the assessment of COMT activity in complex biological samples in vitro. In fact, 4-MDPN has been developed to be a probe for sensing the real activities of COMT in cell and tissue preparations, and this study could provide useful information on the design of COMT probes with high affinity, good reactivity and ideal kinetic behaviours for sensitive quantitative determination of COMT activity [27].
It has been reported that the dihydroxyl moiety of 7,8-DHCs is essential to their biological activities, such as anti-oxidant and anti-cancer activities [28][29][30][31][32]. Accordingly, the retention of catechol phenols is crucial for designing better daphnetin derivatives as therapeutic agents. During the past half century, many coumarin derivatives including a range of C-4 derivatives were synthesized and reported, as these derivatives can be readily synthesized and various substitutes can be easily introduced to the C-4 position by total synthesis or semi-synthesis [33]. Therefore, C-4 modification, such as C-4 acetic acid substitution, is a potential strategy to balance the activity and the metabolic stability, as a relatively low intrinsic clearance was observed for 4-ADPN.
In this work, we carried out the methylation of 7,8-DHCs by COMT using human liver cytosol and recombinant S-COMT. Only 8-O-methylated 7,8-DHCs were detected under the experimental conditions applied. A computing model of the enzyme was constructed, and the docking of substrate molecules to the enzyme was performed to verify this regioselective methylation. All of the four 7,8-DHCs were relatively good COMT substrates, while 4-PDPN showed the lowest K m value. These findings are in agreement with the docking results using a human S-COMT crystal structure as template. In fact, this constitutes the first report of the systemic structure activity study between COMT and 7,8-DHCs. The results will give us more information about the metabolism of 7,8-DHCs in humans and enhance the knowledge of regioselectivity of methylation by human COMT, which would be very helpful to guide the further structural modification of 7,8-DHCs with improved metabolic stability.

Conclusion
Our findings demonstrated that the C-8 hydroxyl group is the only metabolic site of 7,8-DHCs metabolized by human COMT, while the C-7 hydroxyl group could hardly be methylated. Furthermore, we also found that the electronic effects of the C-4 substituents on 7,8-DHCs and the coordination modes of 7,8-DHCs in the active site of COMT could strongly affect the catalytic efficiency. All of these findings are very helpful for guiding the further structural modification of 7,8-DHCs with improved metabolic stability.
Data accessibility. The dataset supporting this article has been uploaded as electronic supplementary material.