Flavonoids from Sideritis Species: Human Monoamine Oxidase (hMAO) Inhibitory Activities, Molecular Docking Studies and Crystal Structure of Xanthomicrol

The inhibitory effects of flavonoids on monoamine oxidases (MAOs) have attracted great interest since alterations in monoaminergic transmission are reported to be related to neurodegenerative diseases such as Parkinson’s and Alzheimer’s diseases and psychiatric disorders such as depression and anxiety, thus MAOs may be considered as targets for the treatment of these multi-factorial diseases. In the present study, four Sideritis flavonoids, xanthomicrol (1), isoscutellarein 7-O-[6'''-O-acetyl-β-d-allopyranosyl-(1→2)]-β-d-glucopyranoside (2), isoscutellarein 7-O-[6'''-O-acetyl-β-d-allopyranosyl-(1→2)]-6''-O-acetyl-β-d-glucopyranoside (3) and salvigenin (4) were docked computationally into the active site of the human monoamine oxidase isoforms (hMAO-A and hMAO-B) and were also investigated for their hMAO inhibitory potencies using recombinant hMAO isoenzymes. The flavonoids inhibited hMAO-A selectively and reversibly in a competitive mode. Salvigenin (4) was found to be the most potent hMAO-A inhibitor, while xanthomicrol (1) appeared as the most selective hMAO-A inhibitor. The computationally obtained results were in good agreement with the corresponding experimental values. In addition, the x-ray structure of xanthomicrol (1) has been shown. The current work warrants further preclinical studies to assess the potential of xanthomicrol (1) and salvigenin (4) as new selective and reversible hMAO-A inhibitors for the treatment of depression and anxiety.


Introduction
The genus Sideritis L. (Lamiaceae) comprises more than 150 species mainly distributed in countries of the Mediterranean-Macaronesian region [1]. Sideritis species , which have become very popular recently, are used as herbs either for the preparation of teas, or for their aromatic properties in local cuisines. They are found in a variety of shops, and marketed as mountain tea, dagçayi, adaçayi, yaylaçayi, malotira, té de Puerto, rabo de gato or zaharena in various Mediterranean countries [2][3][4]. Herbal teas prepared form the aerial parts of Sideritis species have also been used in traditional medicine for thousands of years in the treatment of gastrointestinal disorders such as stomach ache, indigestion and flatulence, to alleviate the symptoms of the common cold including fever, flu, sore throath, and bronchitis, and as analgesics, anticonvulsants, sedatives, tonics and diuretics [4][5][6][7]. Moreover, Sideritis species in the form of tea infusions are popularly used for improving memory function and cognitive ability [3]. The species of the genus are known to be rich in essential oils [4,8,9], diterpenes [4,[10][11][12][13] and flavonoids [2,4,5,[11][12][13][14][15][16].
Recently, the inhibitory effects of flavonoids on monoamine oxidases (MAO-A and MAO-B) have attracted great interest among researchers. Monoamine oxidases (MAOs) are flavoenzymes located in the outer membrane of the mitochondria that play an important role in the oxidative catabolism of endogenous and dietary amines [17,18]. MAO contains flavin adenine dinucleotide (FAD) as a co-factor and exists in two isoforms in mammals, named hMAO-A and hMAO-B [19]. MAO-A preferentially deaminates serotonin and norepinephrine, and is selectively inhibited by clorgyline, whereas MAO-B preferentially deaminates phenylethylamine and benzylamine and is selectively inhibited by L-deprenil [20,21]. Inhibitors of MAO-A are clinically used as antidepressants and anxiolytics [22,23], while MAO-B inhibitors are used in the treatment of Parkinson's disease and in the management of symptoms associated with Alzheimer's disease [24].
The studies on the crystal structures of MAOs and the interactions between MAOs and their ligands [24][25][26][27], have led to the design of new possible inhibitors by docking studies. Some currently used MAO inhibitors produce side effects due to a lack of selectivity towards one of the isoforms [28].
Thus, it is necessary to search for more potent, reversible and selective inhibitors of MAO-A and MAO-B. Despite the great amount of studies related to design and chemical synthesis of novel compounds as therapeutic agents or prodrugs, in recent decades identification of new compounds from natural sources or screening the novel biological activities of known natural compounds still attracts great interest.
A wide number of molecules of natural origin belonging to diverse chemical classes such as flavonoids, xanthones, proanthocyanidins, iridoids, alkaloids, curcumins, cannabinoids, etc. have been shown to inhibit human MAOs [29][30][31]. Among them, flavonoids have attracted more interest as they possess a variety of biological activities such as anti-oxidation and cardio-cerebrovascular protection besides their effects on the central nervous system. Their diverse activities give them extra advantages to be the potential multifunctional therapeutic agents for aging-related diseases [29,30]. Flavonoids are present in almost all terrestrial plants, and form part of the human diet via the consumption of vegetables, fruits, beverages and other plant-derived foods. The MAO inhibitory activities of some well-known flavonoids such as apigenin, kaempferol, quercetin, quercitrin, isoquercitrin and rutin have been previously demonstrated by in vitro and/or in vivo studies [29][30][31], however, few docking studies are available on these compounds in order to clarify the interaction of the compounds with the active site of hMAO [32].
In our previous studies, we have shown the anti-inflammatory, anti-nociceptive, anti-ulcerative antioxidant, aldose reductase and glutathione reductase activities of flavonoids isolated from various Sideritis species [6,16,33,34], in addition to the anti-inflammatory, anti-nociceptive, anti-ulcerative activities of their phenylethanoid glycosides and diterpenoids, and the anti-microbial activity of their iridoids and essential oils. Diuretic, anti-inflammatory, anti-spasmodic, antibacterial and antioxidant activities of different extracts obtained from several Sideritis species growing in Turkey were also previously demonstrated by our research group [6]. In the present work, hMAO-A and hMAO-B inhibitory activities of four Sideritis flavonoids, xanthomicrol (1) (4) were determined using recombinant human MAO. The molecular docking studies of these compounds were also performed to gain insights into the intermolecular interactions at the active site 3D space of both the hMAO-A and hMAO-B enzymes. In addition, the crystal structure of xanthomicrol (1), which was found to be the most selective hMAO-A inhibitor among those four flavonoids, has been described. The X-ray diffraction analysis demonstrates its exact structure with a dimeric molecule in the asymmetric unit.

Human Monoamine Oxidase (hMAO) Inhibitory Activities
The hMAO inhibitory activities of the tested flavonoids ( Figure 1) were investigated using commercially available recombinant hMAO-A and hMAO-B with tyramine as the substrate. Kinetic studies were performed in order to provide data for the interaction of four flavonoids with hMAO isoforms. The kinetic parameters Km (Michaelis-Menten constant) and Vmax (maximum rate) were determined by means of Lineweaver-Burk plots over a substrate range of 0.05-0.5 mM.  Table 1. The results showed that all of the flavonoids tested possessed a degree of selectivity for hMAO-A, with experimental selectivity index (SI) values between 0.007 and 0.086, which suggests that these compounds are highly selective towards hMAO-A. Among them, xanthomicrol (1) appears as remarkably selective, with an experimental SI of 0.007 and selectivity was better than that of moclobemide, a highly potent and selective hMAO-A inhibitor (the experimental SI of moclobemide was determined as 0.010, Table 1). All of the flavonoids tested inhibited hMAO-A in a competitive mode. The kinetic behaviors of xanthomicrol (1) and salvigenin (4), which inhibited hMAO-A potently with experimental Ki values of 0.76 ± 0.03 and 0.54 ± 0.02 μM, respectively, are presented in Figures 2 and 3. Salvigenin (4) appears as the most potent hMAO-A inhibitor among the tested compounds.
MAO inhibitors can be classified as reversible or irreversible, according to their interaction with hMAO isoforms. Although irreversible MAO inhibitors such as selegiline have been extensively used as clinical drugs, this type of inhibition is reported to induce serious cardiovascular toxic effects, mainly provoked by the inhibition of the peripheral MAO-A located in gut, liver and endothelium [35,36]. Besides, the slow and variable enzyme recovery following the withdrawal of irreversible inhibitors is a disadvantage in clinical use, since the turnover rate for MAO biosynthesis in the human brain seems to require about 40 days [37]. Competitive reversible inhibitors have been previously reported to have less influence in the enzyme recovery after withdrawal, and reversible MAO-A inhibitors, such as moclobemide, have no cardiovascular effects, since the ingested tyramine is able to displace the inhibitor from the MAO active site and be metabolized in the normal way by peripheral enzyme in gut and liver [35].  Thus, new selective and reversible inhibitors of MAO-A may be useful therapeutic agents for the treatment of psychiatric diseases and Alzheimer's and Parkinson's diseases devoid of undesirable side-effects, such as cheese effect. All of the flavonoids inhibited hMAO-A reversibly as shown in Table 2. Significant recovery of hMAO-A activity was observed after repeated washing of xanthomicrol (1) and salvigenin (4), indicating that these flavonoids are highly reversible inhibitors of hMAO isoforms.  Figure 4 shows the binding pose (3D and 2D) of xanthomicrol (1) in the active site of hMAO-A isoform. The coumarine ring was inserted between the TYR444 and TYR407 residues, which constitute the hydrophobic cage of the active site, forms two π-π interactions with TYR407. The following five hydrogen bonds form between moieties of the xanthomicrol (1) and the side chains of the active site amino acids of the MAO-A isoform: between the -OH and ASN181, -OCH3 and TYR444 (two), -OH and LYS305 and -OH and GLY66. Analysis of the optimal binding mode for the salvigenin (4) in the hMAO-A active site cavity ( Figure 5) revealed that this compound is docked in the vicinity of the FAD co-factor. Salvigenin forms various electrostatic and van der Waals interactions with the active site residues lining the cavity. A hydrogen bond occurs between the methoxy group of the salvigenin (4) and the hydroxy moiety of TYR444. The other strong electrostatic interactions occur between GLY443 backbone carbonyl and methoxy groups, and backbone carbonyl of ASN181 and methoxy group of the salvigenin (4). The benzene ring of the coumarine moiety of the salvigenin (4) was sandwiched between TYR444 and TYR407 making two π-π interactions.     (4) in the active site of hMAO-B isoform. The major interaction is the hydrogen bond that forms between the carbonyl group of coumarine ring and side chain residue of CYS172. The binding mode adopted by salvigenin (4) fits snugly within a cavity lined with hydrophobic amino acid residues. This hydrophobic pocket includes PHE168, TYR60, ILE199, TYR326, CYS172, ILE198, PHE343, GLN206 and TYR398 amino acids. The potency and selectivity of xanthomicrol (1) and salvigenin (4) for the hMAO enzymes can be evaluated from the above data. The detailed analysis of Figure 5, Figure 7 and Table 1 suggest that salvigenin (4) is a better inhibitor of hMAO-A than of the MAO-B isoform. The reason for that the coumarine core is stabilized by more π-π interactions as well as other various attractions in the hMAO-A active side compared to the hMAO-B isozyme (calculated Ki = 0.867 μM for hMAO-A and calculated Ki = 1.89 μM for hMAO-B). However, xanthomicrol (1) is more selective towards hMAO-A than salvigenin (4) (calculated SIs of xanthomicrol (1) and salvigenin (4) were estimated as 0.029 and 0.239, respectively).

Molecular Docking Studies
Inhibition of MAO by flavonoids appears to be dependent on the phenyl or hydroxyphenyl ring in the 2 position as well as a double bond in the 2 and 3 position of the structure. It has been previously demonstrated that flavonoids isolated from the whole plant and fruits of Cayratia japonica showed MAO inhibitory activity as follows: flavones > flavonols > flavone glycosides > flavanonols. In this series, apigenin inhibited mouse brain MAO-A with an IC50 value of 1.17 μM. It was suggested that MAO inhibitory activity decreased with increasing number of hydroxyl groups on the B ring of the flavones. The presence of electron donating hydrophilic hydroxyl groups in the para position of the B ring has also been reported to favour the MAO-A inhibition [30,31]. Thus, it can be postulated that the p-hydroxyl group in the B ring of xanthomicrol (1) increased the selectivity of this molecule towards hMAO-A.
Monoamine oxidases are responsible for the oxidative deamination of neurotransmitters and dietary amines. Of the two isoenzymes, MAO-A has attracted considerable attention because of its ability to affect the metabolism of serotonin [23]. Depletion of serotonin levels in the central nervous system (CNS) is suggested to be involved in the pathogenesis of depressive disorders [38]. It has been reported that MAO-A activity in the brain is elevated during depression, which causes depletion of serotonin in the brain [39]. Herbal medicines, such as flavonoids, are commonly used for the treatment of depression, along with conventional antidepressants [40]. Thus, in the present study we present four flavonoids isolated from Sideritis species exhibiting highly potent hMAO inhibitory activities. Among them, salvigenin (4) was found as the most potent hMAO-A activity, whereas xanthomicrol (1) appeared as the most selective one towards the hMAO-A isoform.

X-ray Crystallography
The crystal structure of xanthomicrol (4) is shown in Figure 8. The bonding parameters are listed in Table 3. The structure of compound was solved by direct methods using SHELXS-97 [41]. Hydrogen atoms that bonded to carbon and oxygen were positioned geometrically. The asymmetric unit shows two molecules which stand together thanks to intermolecular hydrogen bonding ( Figure 9). Both of hydrogen bonds can be seen in the crystal structure. The intramolecular hydrogen bond, which appears when donor and acceptor atoms are in the same molecule, has a length of 2.562 Å. On the other hand, the distance is 2.796 Å with 119.23° angle for intermolecular hydrogen bonding. Because of the geometrical constraints, it's probably longer in intermolecular interactions [42].
Although synthetic xanthomicrol (1) is commercially available (Apin Chemicals and BOC Sciences), to our knowledge, this is the first report on its crystal structure. The crystal structure of salvigenin (4) has already been reported [43].    The crystal data and structure refinement for xanthomicrol (1) is summarized in Table 4.

Determination of hMAO-A and hMAO-B Activities
The activities of hMAO-A and hMAO-B were determined using p-tyramine as common substrate and calculated as 161.50 ± 8.33 pmol/mg/min (n = 3) and 140.00 ± 7.99 pmol/mg/min (n = 3), respectively. The interactions of the synthesized compounds with hMAO isoforms were determined by a fluorimetric method described and modified previously [45,46]. The production of H2O2 catalyzed by MAO isoforms was detected using Amplex ® -Red reagent, a non-fluorescent probe which reacts with H2O2 in the presence of horse radish peroxidase to produce the fluorescent product resorufin. The reaction was started by adding 200 µM Amplex Red reagent, 1 U/mL horseradish peroxidase (HRP), and p-tyramine (concentration range 0.05-0.5 mM). Control experiments were carried out by replacing the compound and reference inhibitors. The possible capacity of compounds to modify the fluorescence generated in the reaction mixture due to non enzymatic inhibition was determined by adding these compounds to solutions containing only the Amplex Red reagent in a sodium phosphate buffer. No modification was recorded. The possible capacity of compounds to inhibit HRP in the test solution was also determined. The compounds did not inhibit HRP in the experimental medium.

Kinetic Experiments
Synthesized compounds were dissolved in dimethyl sulfoxide, with a maximum concentration of 1% and used in the concentration range of 1.00-100.00 µM for compounds 1 and 4; 5.00-100.00 mM for compounds 2 and 3. Kinetic data corresponding the interaction of hMAO isoforms with the compounds were determined using computer-fit calculation (Prism 4.0, GraphPad SoftwareThe slopes of the Lineweavere-Burk plots were plotted versus the inhibitor concentration and the Ki values were also determined from the x-axis intercept as -Ki. Each Ki value is the representative of single determination where the correlation coefficient (R2) of the replot of the slopes versus the inhibitor concentrations was at least 0.98. SI was calculated as Ki(hMAO-A)/Ki(hMAO-B. The protein was determined according to the Bradford method [47].

Reversibility Experiments
Reversibility of the MAO inhibition with newly synthesized compounds was determined by centrifugation-ultrafiltration method previously described [48]. Recombinant enzymes (hMAO-A or B) were incubated with a single concentration of the synthesized compounds or the reference inhibitors in a sodium phosphate buffer (0.05 M, pH 7.4) for one hour at 37 °C. An aliquot was stored at 4 °C and used for the measurement of MAO-A and -B activities. The remaining incubated sample was placed in an Ultrafree-0.5 centrifugal tube with a 30 kDa Biomax membrane and centrifuged at 9000× g for 20 min at 4 °C. The enzyme retained in the 30 kDa membrane was re-suspended in a sodium phosphate buffer at 4 °C and centrifuged two more times. After the third centrifugation, the enzyme retained in the membrane was re-suspended in sodium phosphate buffer (300 mL) and an aliquot of this suspension was used for MAO-A and -B activity determination. Control experiments were performed simultaneously to determine the 100% MAO activity by replacing the compounds and standards with appropriate dilutions of the vehicles. The corresponding values of percent (%) MAO isoform inhibition was separately calculated for samples with and without repeated washing.

Molecular Docking Studies
Molecular models of the inhibitors (1,4) were built and optimized using SPARTAN 10.0 at PM3 level on iMac 3,4 GHz Intel Core i7 processor. The crystal structures of MAO-A, and MAO-B were extracted from the Protein Data Bank (PDB) [49] Their PDB codes are 2Z5X and 2V5Z for MAO-A and MAO-B respectively. Each structure was cleaned of all water molecules and inhibitors as well as all non-interacting ions were removed before being used in the docking studies. For MAO-A and MAO-B, one of the two subunits was taken as the target structure. For each protein all hydrogens were added and minimized using "Clean Geometry" toolkit, then for more complete optimization final geometry was submitted to "Prepare Macromolecule" protocol of Discovery Studio (Accelerys) assigning CHARMM force field. Missing hydrogen atoms were added based on the protonation state of the titratable residues at a pH of 7.4. Ionic strength was set to 0.145 and the dielectric constant was set to 10. Autodock Tool (ADT) interface was used for the preparation of input files of macromolecules and ligands for docking calculations. Docking was achieved using AutoDock 4.2.6 coding scripts [50] using HPxw8600-Work-Station. The detailed docking procedures are given elsewhere [51,52]. Acelrys Visualization 4.0 program was used for rendering the 2D and 3D pictures. To render visible the detailed interactions of the docked poses of the inhibitors (1,4) Acelrys Visualization 4.0 program was used.

X-ray Analysis
Single crystal X-ray diffraction patterns of xanthomicrol were collected at 293 K with Bruker APEX-II CCD diffractometer from a K780 X-ray generator applying 50 kV and 40 mA. All data were accumulated using a graphite monochromated Mo Kα (λ = 0.71073 Å) radiation. The least-squares methods was applied on the basis of all reflections with F 2 > 2σ(F 2 ) to determine the lattice parameters.

Conclusions
The present study provides an evidence for the isolated flavonoids from Sideritis species as new potent and selective hMAO-inhibitors. The results suggest two compounds, xanthomicrol and salvigenin as potential natural starting molecules for developing novel selective MAO-A inhibitors, for prevention and treatment of psychiatric disorders such as depression and anxiety, and also cognitive impairments in Alzheimer's and Parkinson's Diseases. Current study gives a clear insight in the inhibitory ability of these flavonoids towards hMAO; however, more studies are needed to clarify their efficacy in vivo. In addition, in the present study, crystal structure of xanthomicrol, which is the most selective MAO-A inhibitor tested, has been introduced for the first time.

Supplementary Data
Additional material available from Cambridge Crystallographic Data Center as Deposition No. CCDC 1051489 comprises H-atom coordinates, thermal parameters and remaining bond lengths and angles. These data can be obtained free of charge via http://www.ccdc.cam.ac.uk/conts/retrieving.html, or from the Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; E-Mail: deposit@ccdc.cam.ac.uk.