Discovery, Synthesis, and Functional ... - ACS Publications

Subscriber access provided by UNIV OF NEBRASKA - LINCOLN

Article

Discovery, Synthesis, and Functional Characterization of a Novel Neuroprotective Natural Product from the Fruit of Alpinia oxyphylla for use in Parkinson’s Disease Through LC/ MS–Based Multivariate Data Analysis–Guided Fractionation Guohui Li, Zaijun Zhang, Quan Quan, Ren-Wang Jiang, Samuel S.W. Szeto, Shuai Yuan, Wing-Tak Wong, Herman H. C. Lam, Simon Ming Yuen Lee, and Ivan K Chu J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.6b00152 • Publication Date (Web): 01 Jun 2016 Downloaded from http://pubs.acs.org on June 3, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Proteome Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Discovery, Synthesis, and Functional Characterization of a Novel Neuroprotective Natural Product from the Fruit of Alpinia oxyphylla for use in Parkinson’s Disease Through LC/MS–Based Multivariate Data Analysis– Guided Fractionation

by

Guohui Li1,2†, Zaijun Zhang2,3†, Quan Quan1,Renwang Jiang4, Samuel S.W. Szeto1, Shuai Yuan2, Wing-tak Wong5, Herman H. C. Lam1, Simon Ming-Yuen Lee2*, Ivan K. Chu1*

1

Department of Chemistry, The University of Hong Kong, Hong Kong, China; 2State Key Laboratory of Quality Research in Chinese Medicine and Institute of Chinese Medical Sciences, University of Macau, Avenue Padre Tomás Pereira S.J., Taipa, Macao, China; 3Institute of New Drug Research, Guangdong Province Key Laboratory of Pharmacodynamic, Constituents of Traditional Chinese Medicine & New Drug Research, College of Pharmacy, Jinan University, Guangdong, China;4Institute of Traditional Chinese Medicine and Natural Products, College of Pharmacy, Jinan University, Guangzhou 510632, P. R. China; 5Department of Applied Biology & Chemical Technology, The Hong Kong Polytechnic University, Hong Kong, China

*

Address correspondence to: Ivan K. Chu, Department of Chemistry, The University of Hong Kong, Pokfulam, Hong Kong, China. Tel.: (852) 2859 2152, Fax: (852) 2857 1586, E-mail: [email protected] Professor Simon Ming-Yuen Lee, Institute of Chinese Medical Sciences, University of Macau, Av. Padre Tomás Pereira S.J., Taipa, Macao, China. Tel.: (853) 8397 4695, Fax: (853) 2884 1358, E-mail: [email protected]

†

Contributed equally to this study.

1

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract Herein we report the discovery of a novel lead compound, oxyphylla A [(R)-4-(2-hydroxy-5methylphenyl)-5-methylhexanoic acid] (from the fruit of Alpinia oxyphylla), which functions as a neuroprotective agent against Parkinson’s disease. To identify a shortlist of candidates from the extract of A. oxyphylla, we employed an integrated strategy combining liquid chromatography/mass spectrometry, bioactivity-guided fractionation, and chemometric analysis. The neuroprotective effects of the shortlisted candidates were validated prior to scaling up the finalized list of potential neuroprotective constituents for more-detailed chemical and biological characterization. Oxyphylla A has promising neuroprotective effects: (i) it ameliorates in vitro chemical-induced primary neuronal cell damage and (ii) alleviates chemical-induced dopaminergic neuron loss and behavioral impairment in both zebrafish and mice in vivo. Quantitative proteomics analyses of oxyphylla A– treated primary cerebellar granule neurons that had been intoxicated with 1-methyl-4phenylpyridinium revealed that oxyphylla A activates nuclear factor-erythroid 2-related factor 2 (NRF2)—a master redox switch—and triggers a cascade of antioxidative responses. These observations were verified independently through western blot analyses. Our integrated metabolomics, chemometrics, and pharmacological strategy led to the efficient discovery of novel bioactive ingredients from A. oxyphylla while avoiding the non-targeting, labor-intensive steps usually required for identification of bioactive compounds. Our successful development of a synthetic route toward oxyphylla A should lead to its availability on large scale for further functional development and pathological studies. Keywords: natural products · neuroprotective · Parkinson’s disease · proteomics · NRF2

2


Page 2 of 33

Page 3 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction Parkinson’s disease (PD) is one of the most common neurodegenerative diseases, affecting up to 1% of those over the age of 65.1 Despite the number of candidate disease-modifying agents that have displayed promising therapeutic effects in preclinical trials, no licensed PD neuroprotective drugs are currently available.2 In recent years, emphasis has being placed on identifying and characterizing potentially active plant-derived pharmaceutical ingredients to fill this persistent unmet demand.3 Historically, natural products have been a main source of compounds displaying medicinal properties and have been used as a platform for drug discovery.4-6 Despite continuing success at providing lead compounds, within the past two decades the use of natural products had diminished, with a focus shifting toward approaches (e.g., combinational chemistry) that might provide a range of molecules for successful lead discovery.7, 8 In practice, however, many of the large screening collections have been unsatisfactory at generating useful targets; instead, it is now recognized that biologically relevant chemical space is better covered by natural products than by synthetic compounds.9,

10

Indeed, natural products remain an enormous untapped resource for continued discovery and development of therapeutic agents for a wide range of disease conditions.11 Therefore, a reemphasis on natural products as the foundation for drug discovery has recently emerged, but this approach is not without significant challenges.7 Natural extracts from plants and microorganisms typically feature a diverse and complex array of metabolites with varying and distinct chemical and physical properties; these compounds can exist over a wide dynamic range of concentrations.12,

13

The

conventional reductionist approach of discovering bioactive compounds from natural products involves bioassay-based screening of concentrated extract samples, followed by sequential rounds of isolation and screening. This strategy is tedious, time-consuming, and often identifies previously characterized compounds;11, 12, 14 it might also be ineffective for identifying compounds present at sub-stoichiometric levels if their activities are masked by less potent, but more abundant, compounds.15 In addition, bioassay-based screening may be confounded by additive or synergistic effects of interacting compounds present in the extracts, thereby complicating downstream purification and identification processes.15, 16 To tackle the complex problem of identifying bioactive compounds from natural products, metabolomics-based approaches are becoming increasingly favored.17, 3


18

Metabolomics is a


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

comprehensive qualitative and quantitative strategy aimed at profiling all of the metabolites in a sample, and correlating their levels to the context of the biological situation under investigation.7 One of the main aspects of metabolomics is the application of multivariate data analysis (MVDA) to simultaneously evaluate a vast number of identified metabolites and determine their contributions to the observed biological activity. MVDA is a powerful technique for the analysis of datasets with a large number of variables (e.g., metabolites), and enables visualization and downstream interpretation of data that correlate with a target variable (e.g., bioactivity).17, 19 To date, however, the metabolomics approach has been focused predominantly on finding system phenomena; its application to the detection of lead compounds remains in an early stage.20 Recently, metabolomics has demonstrated strong potential for use as a component of dereplication strategies, for evaluating the efficacy and variability of natural product extracts, and for the efficient screening of potentially active metabolites by linking the chemical profiles of the examined extracts to their bioactivity data.7, 12, 13, 17, 21

Considering the demonstrated utility of this tool to aid in the discovery of active

compounds from natural products, in this study we designed an integrated purification strategy, incorporating a metabolomics component, for the efficient identification of bioactive metabolites and their isolation from complex natural product extracts while avoiding the untargeted labor-intensive fractionation steps of the classical approach. We chose liquid chromatography/mass spectrometry (LC-MS) as the analytical technique because, combined with electrospray ionization, its sample preparation is simpler than that for gas chromatography/MS; it also features high sensitivity and selectivity, a wide dynamic range, can utilize a combination of different separation phases (i.e., reversed and normal LC phases), and is high-throughput with particular utility toward plant metabolites because of their inherent chemical diversity.22-25 In addition, the MS polarities (positive and negative modes) enable it to cover a more comprehensive range of compound species than other analytical techniques used for chemical profiling.22,

24

This LC-MS approach also has a unique

feature in that it generates a pair-identifier (mass-to-charge ratio/retention time) of the characterized metabolites; this highly useful information can be used to facilitate more efficient downstream isolation of the targeted bioactive compounds.22, 23, 26 Among the various sources of natural products, interest in traditional Chinese medicines from a modern pharmacological perspective has risen steadily, due their long historical use as foods and 4


Page 4 of 33

Page 5 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


medicines with epidemiological evidence of tolerance.27 One of these well-established traditional Chinese medicines, the fruit of Alpinia oxyphylla, has emerged as a potential source of therapeutic compounds because an increasing amount of evidence is indicating its beneficial effects on various neurodegenerative diseases, including Parkinson’s disease (PD).28 Also known as Yi Zhi in Chinese, it has been used historically to treat brain-related disorders (e.g., stroke, amnesia) and many other medical conditions since at least 1328 A.D., as documented in the ancient Chinese medicine book Shi Yi De Xiao Fang.29 In several recent studies, modern molecular pharmacological approaches, using multiple in vitro and in vivo PD neuronal damage models, have demonstrated the protective potential of the ethanolic extract of the fruit of Alpinia oxyphylla (AOE).28, 30-32 Although attempts have been made previously to determine its neuroprotective compounds, substantial profiling of AOE metabolites33-35 has led to protocatechuic acid and chrysin being the only major neuroprotective bioactive ingredients identified.36 In this study we employed a metabolomics-based approach to identify novel neuroprotective compounds from AOE that might be potential leads as PD therapeutics. We used LC-MS to characterize the metabolites present in AOE and correlated their observed bioassay activities using chemometric analyses. Using this approach we identified a novel lead compound, oxyphylla A [(R)4-(2-hydroxy-5-methylphenyl)-5-methylhexanoic acid], that functions as a neuroprotective agent against both in vitro and in vivo PD neuronal damage models. Cellular 1-methyl-4-phenylpyridinium ion (MPP+) intoxication often involves the over-generation of reactive oxygen species (ROS) during oxidative stress.37 The neuroprotective effects elicited by oxyphylla A was hallmarked by a number of the differentially expressed antioxidant enzymes and oxidoreductases. Thus, the protein expression profiles of MPP+–induced cerebellar granule neurons (CGNs) with and without oxyphylla A treatment provide insight on the molecular mechanisms underlying these neuroprotective effects using isobaric tags for relative and absolute quantitation (iTRAQ) tag labeling and multidimensional liquid chromatography/tandem mass spectrometry(LC-MS/MS). We have demonstrated the utility of a multifaceted and integrative strategy involving metabolomics, proteomics, pharmacology and chemometrics for the discovery of A. oxyphylla fruit derived bioactive ingredients.

5



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Experimental Section Chemicals and reagents MCI gel CHP 20P (75–150 µm) was purchased from Mitsubishi Chemical (Tokyo, Japan). Sephadex LH-20 was purchased from GE Healthcare Bio-Sciences AB (Uppsala, Sweden). TLC plates (20 × 20 cm, AlugramSil G/UV 254) were purchased from Macherey-Nagel (Duren, Germany). Silica gel (200–300 mesh) was purchased from Qingdao Marine Chemical Factory (Qingdao, China). ARgrade petroleum ether was obtained from LAB-SCAN (Bangkok, Thailand); AR-grade ethyl acetate was purchased from UNIVAR (NSW, Australia). HPLC-grade MeOH and acetonitrile (ACN) were obtained from Scharlau (Sentmenat, Spain). Jupiter C18 packing material (3 µm particles, 300A° pores) were purchased from Phenomenex (Torrance, CA, USA). Formic acid (≥98%) was purchased from Fluka (St. Louis, MO). The iTRAQ kit was purchased from AB Sciex (Foster City, CA). The Milli-Q system was purchased from Millipore (Bedford, MA). All reagents and materials for cell culturing were purchased from Life Technologies (Carlsbad, CA). Protease inhibitor cocktail was purchased from Roche Applied Science (Mannheim, Germany). Phenylmethylsulfonyl fluoride (PMSF) was purchased from Sigma–Aldrich (St. Louis, MO). Rabbit anti-mouse tyrosine hydroxylase (TH) polyclonal antibody (1:500) was obtained from Millipore (Bedford, MA). Immunol staining primary antibody dilution buffer was obtained from Beyotime (Beijing, China). Diaminobenzidine (DAB) Kit and horseradish peroxidase (HRP)-conjugated goat anti-rabbit secondary antibody were obtained from Gene Company (Shanghai, China). Mouse monoclonal antiTH antibody MAB318 was purchased from Merck Millipore (Darmstadt, Germany). Rabbit monoclonal antibody against erythroid 2-related factor 2 (NRF2) was purchased from Novus Biologicals (Littleton, CO). Rabbit monoclonal antibody against heme oxygenase-1 (HO-1) was purchased from Abcam (Cambridge, MA). Vectastain ABC kit was purchased from Vector Laboratories (Burlingame, DA). RIPA lysis buffer and 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2Htetrazolium bromide (MTT) were purchased from Sigma–Aldrich (St. Louis, MO). MPP+, 1-methyl4-phenyl-1,2,3,6-tetrahydropyridine (MPTP),

DL-dithiothreitol

(DTT), and iodoacetamide (IAA)

were purchased from Sigma–Aldrich (St. Louis, MO). Microcon centrifugal filters were purchased from Merck Millipore (Darmstadt, Germany). Bio-Rad protein assay kit was purchased from Bio6


Page 6 of 33

Page 7 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Rad (Hercules, CA). All other chemicals were purchased from Sigma–Aldrich.

Plant material and fractionation of AOE The fruit of A. oxyphylla were purchased from Sam Yao Hong (Macau SAR, China) and authenticated by an experienced pharmacognosist (Dr. Zhixiu Lin, School of Chinese Medicine, The Chinese University of Hong Kong). For the material used in this study, a voucher herbarium specimen was deposited with the Herbarium of the Institute of Chinese Medical Sciences (No. 0011), University of Macau, Macau SAR, China. Following coarse pulverization, 10 kg of the air-dried fruits was obtained. The fruits of A. oxyphylla were extracted three times with 95% aqueous EtOH (90 L) under reflux for 2 h in University o Macau. The combined extracts were concentrated through rotary evaporation at 50 °C, resulting in the AOE (914.55 g).

The fractionation scheme used to separate out the bioactive metabolites in AOE is depicted in Figure S-3. The AOE was re-extracted successively with petroleum ether, ethyl acetate, and EtOH; the ethyl acetate–soluble portion (EA portion) was subjected to further fractionation to yield 11 EA subfractions (Frs. A–K).

Details please refer to the Supporting Information, Experimental

Section.

LC-MS/MS analysis An Agilent 1100 series HPLC system (Agilent Technologies, Palo Alto, CA) coupled to a hybrid triple-quadrupole linear ion trap (LIT) mass spectrometer with a TurboIonSpray ion source (QTRAP, AB Sciex, Concord, Ontario, Canada), set in both positive and negative ion modes, was used for chemical profiling of the metabolites found in fractions (Frs.) A–K: briefly, the crude ethanolic extract of the fruits of Alpinia oxyphylla (AOE) was re-extracted petroleum ether, ethyl acetate, and ethanol successively to yield petroleum ether–soluble portion (PE portion), an ethyl acetate–soluble portion (EA portion), and an ethanol-soluble portion (EtOH portion). The EA portion was subjected to further fractionation to yield 11 EA subfractions (Frs. A–K). Each of the samples (10 µL) from Frs. A–K was injected at 1 mg/mL into an Eclipse XDB C18 column (2.1 × 150 mm, 3.5 µm; Agilent Technologies, Santa Clara, CA) for duplicate runs. The LC and MS data acquisition parameters are 7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

presented in Table S-1. To determine the accurate masses of the top ten bioactive compounds, a platform consisting of an Agilent 1200 series nano pump, equipped with a 100-well plate autosampler (Agilent Technologies, Wilmington, DE), and connected to a triple-quadrupole time-of-fight (TOF) mass spectrometer fitted with a Nanospray III source (TripleTOF 5600; AB Sciex, Concord, Ontario, Canada), was used. The reversed phase (RP) trap column (150 µm i.d.× 50 mm length) and analytical RP nano column (75 µm i.d.× 150 mm length) were packed in-house with Jupiter C18 bulk materials using an ultrahigh-pressure syringe pump with maximum pressure of 6000 psi. Samples were loaded onto an RP trap column and then eluted for LC-MS/MS analysis. The LC and MS data acquisition parameters are presented in Table S-2. High-resolution MS data with isotopic patterns for the ions of interest allowed accurate calculations of the elemental compositions; all of the measured masses were within 6 ppm of the theoretical values for the proposed formulas.

Cell culture Cerebellar granule neurons (CGNs) were isolated and cultured from Sprague–Dawley rats aged eight days, as previously described.38 Briefly, neurons were seeded in a 96-well plate at a density of 1.5 × 105 cells/well containing basal modified Eagle’s medium with 10% fetal bovine serum, 2 mM glutamine, 25 Mm KCl, streptomycin (100 µg/mL), and penicillin (100 units/mL). 10 µM Cytosine arabinoside was added to the culture medium, to inhibit the growth of non-neuronal cells, at 24 h after plating. After implementing this protocol, approximately 95% of the cultured cells observed appeared to be granule neurons. After six days in culture, the CGNs displayed several features of mature neurons.39 The experiments requiring CGNs cells were performed at eight days in vitro (DIV).

MTT assay The MTT assay is a colorimetric assay for testing cell viability based on the principle that intracellular NAD(P)H-dependent oxidoreductases can reduce MTT into insoluble purple formazan, the absorbance of which can be recorded by a microplate reader to reflect the relative number of live cells in a culture medium. CGNs at 8 DIV were pre-treated with serial concentrations of AOE, PE 8


Page 8 of 33

Page 9 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


portion, EA portion, EtOH portion, 11 EA subfractions, oxyphylla A, chrysin, and teuhetenone A, for 2 h. The vehicle group was incubated with 0.1% DMSO. The cells were then exposed to 150 µM MPP+ for 36 h. For assessment of cell viability, MTT solution (5 mg/mL, 15 µL) was added to each well along with the medium (100 µL). The 96-well plate was then placed in the incubator at 37 °C for 4 h. At this point, absolute DMSO (100 µL) was added to each well and then incubation was continued for 10 min. The absorbance at 570 nm was measured using a WallacVictor3™ V microplate reader (PerkinElmer, The Netherlands). The optical density (O.D.) of each treatment group was normalized with respect to that of the untreated control group.

Chemical data analysis Markerview software (v. 1.2.1; AB Sciex, Foster City, CA) was used to locate and align the LC-MS peaks.40 Peaks were located with the minimum retention time set to 3 min, a minimum spectral width of 0.3 Da, a subtraction offset of 15 scans, and a subtraction multiplication factor of 1.3. All peaks were subsequently aligned with a retention time tolerance of 0.2 min and a mass tolerance of 0.2 Da across the entire dataset. The noise threshold was set at 1000 cps for the negative mode and 2.0 × 105 cps for the positive mode, according to the background signal abundance in the total ion chromatography profile of the enhanced MS scan. 180 and 644 monoisotopic peaks were detected in the negative and positive modes, respectively. Following peak alignment across the 22 runs (duplicate runs for each of the 11 EA subfractions), the area of a particular peak was normalized to the sum of the corresponding peak areas detected across all the runs. The normalization was performed separately for the datasets acquired in positive and negative modes.41 Detailed lists of the monoisotopic peaks detected in negative and positive modes, and the combined list after normalization are provided in Table S-3.A 22 × 825 (row × column) data matrix, with normalized peak areas and bioactivity values, was inputted for MVDA using SIMICA-P software (v. 13.0; Umetrics, Sweden). Each row represented different samples, and each sample possessed two replicates. Each column represented an LC-MS peak having a unique pair-identifier (m/z/Rt), except for the final column, which represented the bioactivity derived from the MTT assay. A supervised Orthogonal-projections-to-latent-structures (OPLS) model was employed to correlate the LC-MS profiling data with corresponding bioactivity data, according to a Pareto scaling method.13 9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Peakviewer software (v. 2.0; AB Sciex, Foster City, CA) was used to process the HR-MS data. Molecular formulas were generated using the Formula Finder tool.42 Log P values were predicted using ChemBioDraw Ultra 12.0 (Cambridgesoftware, Cambridge, UK).43 Because of space considerations, some experimental methods related to materials, sample preparation, liquid chromatography, mass spectrometry, husbandry and drug treatment conditions, neurobehavioral tests, and data analysis are provided in the Supporting Information.

Purification and structural characterization of oxyphylla A, chrysin, and teuhetenone A (compounds 1–3) The purification scheme used to isolate compounds 1–3 is depicted in Figure S-3. Fr. F was further fractionated through different combination of column chromatographies (e.g. silica gel, Sephadex LH-20, MCI gel) and preparative HPLC to yield purified compound 1-3. UV, IR, NMR and MS spectroscopic techniques were employed for the structure elucidation of compound 1-3. Optical rotation and X-ray diffraction data were also acquired to determine the stereo structure of compound 1. Details please refer to the Supporting Information, Experimental Section.

Total synthesis and X-ray crystallographic of oxyphylla A Oxyphylla A was synthesized from 4-Methylanisole reacting by 9 steps to get the racemic compound i. Then the chiral isomers j and k were obtained by separation on chiral prep-HPLC. The final structure of oxyphylla A was solved by direct methods employing SHELXS97.44 The data for oxyphylla A have been deposited in the Cambridge Crystallographic Data Centre with reference number CCDC 1411040. The detailed crystal data, data collection and refinement parameters please refer to the Supporting Information, Experimental Section.

Mouse and zebrafish husbandry and drug treatment conditions Adult male C57BL/6J mice (8–10 weeks old, 18–22 g) were maintained in a 12-/12-h light/dark cycle with access to water and food ad libitum. They were acclimated for 1 week prior to treatment. 10


Page 10 of 33

Page 11 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


The AB strain of wild-type zebrafish and the general husbandry procedures used in this study were the same as previously described.28 All the animal handling protocols were approved by the Research Ethics Committee, University of Macau. Embryos were collected after natural spawning, staged according to standard criteria, and raised synchronously at 28.5 °C in embryo medium (13.7 mM NaCl, 540 µM KCl, 25 µM Na2HPO4, 44 µM KH2PO4, 300 µM CaCl2, 100 µM MgSO4, 420 µM NaHCO3, pH 7.4). Detailed treatment procedures with AOE and oxyphylla A are presented in the Supporting Information, Experimental Section.

Neurobehavioral tests For experiments involving treatment with AOE, the neurobehavioral changes of the mice were assessed in terms of their performance in Cat Walk and open field tests. For experiments involving the treatment with oxyphylla A, the mice were monitored for neurobehavioral changes using the pole,45 rotarod,46 and footprint tests.47 For the experiments involving zebrafish, the fish behavior was analyzed using a digital video tracking system (Viewpoint ZebraLab system), as previously described.48 Details please refer to the Supporting Information, Experimental Section.

Bioactivity data, proteomic sample and proteomics data analysis Biological data are expressed as means ± standard deviation (SD). Statistical significance was determined by using one-way analysis of variation (ANOVA), with the Tukey–Kramer test employed for multiple group comparisons. Significant differences were accepted with p< 0.05.

Proteomic sample& Data analysis: Pooled iTRAQ primary CGNs protein samples (20 µg) were subjected to LC-MS/MS analysis using MD-LC-MS/MS platforms, as previously documented.49-51 MS data were collected on a Triple quadrupole time-of-fight mass spectrometry (Triple TOF 5600, AB SCIEX, Concord, ON, Canada) equipped with a Nanospray III source (AB SCIEX, Concord, ON, Canada). The raw WIFF data were searched against the Rattusnorvegicus Ensembl reference proteome database (Dec. 2013; 29,704 entries: http://www.ensembl.org/Rattus_norvegicus/Info/Index) using ProteinPilot v. 4.5 software (Applied Biosystems, Framingham, MA). The plug-in Proteomics System Performance Evaluation Pipeline (PSPEP), featured in ProteinPilot v.4.5, was used to 11



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 33

determine the false discovery rate (FDR).52 The corresponding unused score with 5% local FDR was used to cut off the number of identified non-redundant proteins and peptides with unique sequence and modification. The iTRAQ quantified proteins were required to have at least 4 spectra with >95% confidence in total from the duplicate runs. The protein ratios were calculated using the weighted average of the natural logarithm of the peptide ratios, as described in the ProteinPilot manual. A protein was deemed as dysregulated when the protein ratio was significantly ≥1.23 or ≤0.81 in the corresponding one-tail t-test (p-value < 0.05).53 The up- and down-regulated proteins were highlighted in red and green, respectively, in Table S-7. Protein pathway and network analysis was performed using the Ingenuity Pathway Analysis (IPA) program (Qigan, Redwood City, CA). Detailed optimized MS acquisition parameters and related data analysis methods are presented in the Experimental Section of the Supporting Information.

Western

blotting,

immunohistochemical

sample

preparation

and

anti-TH

immunostaining After applying the various treatment conditions, the CGNs culture was washed three times with icecold PBS and then the protein was extracted with RIPA buffer (1% Protease Inhibitor and 1% PMSF were added) on ice for 10 min. The cell lysates were centrifuged (12,500×g, 20 min, 4 °C) and then the supernatants were collected and stored at –80 °C until required. The protein contents were quantified using the Bio-Rad protein assay kit. Protein samples (30 µg) were resolved using SDSPAGE and then transferred to PVDF membranes using a semi-dry transfer cell (Bio-Rad). After the membrane had been washed three times with TBST solution, primary antibodies (1:1000) were added and incubated overnight at 4 °C on a rolling mixer. After washing the membranes three times with PBS, horseradish peroxidase-conjugated secondary antibodies (1:2000) were added and incubated at room temperature for another 2 h on a rotating homogenizer. The membranes were then washed three times with PBS and incubated with ECL solution, according to manufacturer’s instructions; and immunoblots were analyzed using a chemiluminescent imaging system. Quantitative assessment of protein blots was performed using Gel DocTM XR (Bio-Rad, Hercules, CA, USA) equipped with QuantityOne software (Bio-Rad). Mice treated with AOE or oxyphylla A 12


Page 13 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


were sacrificed for brain tissue section. The immunohistochemistry method used has been described previously.54 Whole-mount immunostaining of the zebrafish embryos with oxyphylla A treatment was performed as previously described.48 Details please refer to the Supporting Information, Experimental Section.

Results and Discussion The aim of this study was to develop an efficient strategy involving LC-MS–based MVDA-guided fractionation for the identification and isolation of potentially bioactive compounds. This newly developed integrated methodology would then be implemented to identify and characterize novel neuroprotective natural products from AOE, with the goal of isolating potential PD therapeutic lead compounds (Figure 1). In the initial step, to establish the neuroprotective effects of AOE, a crude extract was obtained through an ethanolic hot Soxhlet reflux extraction and examined employing a MPTP–induced PD mouse model. The administration of AOE (i) significantly restored losses of dopaminergic (DA) neurons in the substantia nigra pars compacta (SNpc) (Figure S-1) and (ii) ameliorated MPTP-induced behavioral impairments (Figure S-2), indicating its neuroprotective potential. To characterize the novel neuroprotective constituents of AOE, we subjected it to extractions with three solvents of distinct polarities ranging from low to high, to generate the respective PE portion, EA portion and EtOH portion (Figure S-3). We then tested these three portions, along with AOE, for their abilities to inhibit PD models in PC12 cell and zebrafish28 as well as MPP+-induced cell death of primary CGNs, a commonly used in vitro PD model.39 Consistent with previous findings using other PD experimental models, pretreatment of the CGNs with AOE significantly prevented MPP+-induced damage in a concentration-dependent manner28 with EA portion displaying the highest neuroprotective potency (Figure S-4A). We then further separated EA portion, through column chromatography, into 11 EA subfractions (Frs. A–K) (Figure S-3). Each EA subfraction was subjected to (i) comprehensive chemical profiling analysis through LC–MS/MS in both positive and negative ESI modes (Figure 2) and (ii) the neuroprotective activity assay (Figure 2 and Figures S-4B–D). Next, we applied MVDA to correlate the metabolite profiles of EA subfractions (each species was 13



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

assigned an m/z-and-retention-time data pair as an identifier) with their respective neuroprotective activities to reveal any potential bioactive constituents in these complex metabolite profiles. Specifically, we employed an orthogonal-projections-to-latent-structures (OPLS) model to reveal the key potential bioactive metabolites in the fruit of A. oxyphylla. In the OPLS model, the inter-class variance in X was modeled by the predictive component tp while the intra-class variance orthogonal to Y was modeled by the orthogonal component to; this model provided a means to remove the systematic variation from descriptor variables X that were not correlated to response variables Y.55, 56 As illustrated in Figure 3A, in the score plot along the direction of the first predictive component tp1, all 11 EA subfractions were classified into two groups according to the potency of their neuroprotective activities. Frs. A–C, J, and K were classified in the low bioactivity group, while Frs. D–I were placed in the high neuroprotective activity group, which could be divided unambiguously into two subgroups along the first orthogonal component to1, demonstrating intra-class variation. In this manner, it was possible to provide a more robust and definitive interpretation for the grouping displayed in the score plot. Accordingly, this observation confirms that the first predictive component tp1, in the OPLS model, was associated with the bioactivity. Metabolites exhibited the greatest influence on the discrimination in the score plot were identified in the loading plot as potential biochemically significant metabolites (Figure 3B). The data points corresponding to 1, 2, and 3 carried the first three highest weights (PP1) along the tp1direction, indicating that they made the greatest contribution to the discrimination in score plots according to bioactivity (Figure 3B). Moreover, the significance of the variables in the OPLS model can also be presented in an S-plot, in which both the covariance p and correlation p (corr) of the variables projected to the predictive component can be visualized (Figure 3C). This plot, therefore, provides more information to assess the reliability of the variables’ contribution; by eliminating the usual suspects and false positives, we could identify the variables having biological significance relating to bioactivity.55, 57 By also taking the confidence interval of each variable into account, we compiled a shortlist comprising the top 10 ranked MS peaks based on both the magnitude of their contributions and their reliability to the model with cutoff values of p1 > 0.07 and p (corr)1 > 0.6 (Table S-4, MVDA result against single concentration please refer to Table S-5). In addition, all 10 metabolites were statistically significant based on their jackknifed confidence intervals (Figure 3D). Compound 1 was positioned on the wing 14


Page 14 of 33

Page 15 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


of the S-plot with the highest p1value, indicating the greatest contribution to class separation and the highest p (corr)1 value (indicative of the greatest degree of reliability; Figure 3C). Meanwhile, compound 2 had the second-highest p1 value and also a high p (corr) 1 value (Figure 3C). MVDA indicated that 1, 2, and 3, with features of 235.4/20.7, 253.4/21.2, and 195.2/15.7 (m/z ratio/retention time), had the greatest influence on the neuroprotective activity; therefore, we tentatively regarded them as the bioactive metabolites having the highest importance with respect to their characterization. High-resolution MS data with isotopic patterns for the ions of interest allowed accurate calculations of the elemental compositions; all of the measured masses were within 6 ppm of the theoretical values for the proposed formulas (Table S-4). The top three ranked compounds, 1– 3, that exerted the greatest influence on neuroprotective activity in the OPLS model were scaled-up, isolated chromatographically, and further characterized (Figure S-3 and Table S-4). Based on semiquantitative analysis through LC-MS, we determined that Fr. F contained the most abundant quantities of 1, 2, and 3 among all the other fractions. Accordingly, we subjected Fr. F to further fractionation, using three comprehensive chromatographic separation methods, to isolate the individual compounds. The first approach, involving silica gel chromatography (separation mechanism according to molecular polarity) coupled with Sephadex LH-20 chromatography, led to the isolation of 1. We combined MCI gel chromatography (separation based on hydrophobicity) and Sephadex LH-20 chromatography (acting as a molecular sieve) with preparative reversed-phase HPLC to provide pure forms of compounds 2 and 3. Using the multiple bioassays against MPP+induced PC12 cell toxicity (data not shown), CGN toxicity, and dopaminergic neuron loss and swimming behavior impairments in PD zebrafish larvae model, we confirmed the neuroprotective activities of these three purified compounds (Figure 4A, Figure S-5 and Figure S-18). We identified compounds 2 and 3 as chrysin and teuhetenone A, respectively; their MS/MS spectra and LC retention times were virtually identical to those of authentic standard samples under similar experimental conditions, with

1

H and13C NMR spectra providing unambiguous structural

confirmation (Figures S-6–S-11).58,

59

Our identification of chrysin, a known and characterized

neuroprotective compound,36 confirmed the effectiveness of this MVDA-based approach. Because no authentic standard for 1 was available, we performed an extensive structural and stereochemical investigation using UV, IR, 1H NMR,13C NMR COSY, HSQC, and HMBC spectroscopies and 15



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

optical rotation, resulting in the identification of 1 as the novel compound (R)-4-(2-hydroxy-5methylphenyl)-5-methylhexanoic acid (Figure 5, Figures S-12–S-16; Table S-6).60 We have named this compound “oxyphylla A,” following the conventional nomenclature for metabolites from the fruit of A. oxyphylla.35 Although chrysin (2) has potential therapeutic benefits, it displays low oral bioavailability;61 this significant drawback, from a pharmacological perspective, has been demonstrated in vivo to result from its extensive cellular metabolism and efflux of metabolites back into the intestines for hydrolysis and elimination.61 Our present study is the first to report the potential neuroprotective effects of teuhetenone A (3); unfortunately, it has a low probability of traversing the blood–brain barrier (BBB), with a predicted octanol/water partition coefficient (log P, a measure of lipophilicity) of 1.4. It has been suggested that a value of log P of at least 1.5 is required for satisfactory penetration of the BBB, with the optimal range being between 2 and 3.62, 63 Oxyphylla A, however, has a predicted log P value of 3.2, suggesting great potential for crossing the BBB; indeed, i.g. administration of oxyphylla A (20 mg/kg dosage) to rats led to its subsequent presence at 10 µM in the cerebrospinal fluid (Figure S-17). Thus, from the top three compounds, we selected oxyphylla A alone for further in vivo pharmacodynamics studies. First, we synthesized, purified, and chirally separated it in gram quantities; the synthetic procedure is detailed in the Supporting

Information—Experimental

Section and in a patent application.64 X-ray

crystallography confirmed the absolute configuration of oxyphylla A to be R, consistent with the optical rotation reported for (–)-sesquichamaenol (Ref. #: CCDC 1411040, Supporting Information—Experimental Section).65 To assess the neuroprotective effects of oxyphylla A in vivo, initially we used a zebrafish PD model. Oxyphylla A moderated MPTP-induced DA neuron loss and alleviated zebrafish larvae swimming behavioral impairments in a dose-dependent manner (Figure S-18). Next, we further examined oxyphylla A using an MPTP-treated PD mouse model; evidence exists for the selective MPTPinduced DA neuron loss inhibiting mitochondrial function in primates, and in mice under oxidative stress.66 Treatment with oxyphylla A significantly moderated the TH–positive DA neuron loss of the MPTP-treated mouse model in a concentration-dependent manner—comparable with the performance of the positive control (rasagiline, a monoamine oxidase-B inhibitor and an FDAapproved drug for the treatment of PD).66 At a dosage of 20 mg/kg, the efficacy of oxyphylla A was 16


Page 16 of 33

Page 17 of 33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


similar to that of rasagiline, with 85% survival of TH-positive DA neurons (Figures 4B-C and S-19). The motor phenotypes of the PD mice were also accessed using pole, footprint, and rotarod tests to mimic human neurological diagnosis. Oxyphylla A significantly ameliorated movement abnormalities in a dose-dependent manner—decreasing the duration of the pole and footprint tests, prolonging the time on the rod, and improving the stride length (Figure 4D and Figure S-19)—with the maximum effect at a concentration of 20 mg/kg (comparable with that of rasagiline). Most notably, the efficacy of oxyphylla A was considerably higher than that of rasagiline for mice subjected to the rotarod test, one of the main and most sensitive tests for assessing motor function in mice (Figure 4D). The protein expression profiles of MPP+-induced CGNs in the presence and absence of oxyphylla A provided insight into the molecular mechanisms underlying these neuroprotective effects: a total of 2649 non-redundant proteins were confidently identified (local FDR

Discovery, Synthesis, and Functional ... - ACS Publications

Recommend Documents