Identification of Î±-Glucosidase Inhibitors in

Article Cite This: J. Nat. Prod. XXXX, XXX, XXX−XXX

pubs.acs.org/jnp

Identification of α‑Glucosidase Inhibitors in Machilus litseifolia by Combined Use of High-Resolution α‑Glucosidase Inhibition Profiling and HPLC-PDA-HRMS-SPE-NMR Tuo Li, Kenneth T. Kongstad, and Dan Staerk* Department of Drug Design and Pharmacology, Faculty of Health and Medical Sciences, University of Copenhagen, Universitetsparken 2, DK-2100 Copenhagen, Denmark

Downloaded via UNIV OF EDINBURGH on January 24, 2019 at 07:12:50 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: Type 2 diabetes is a chronic multifactorial disease affecting more than 425 million people worldwide, and new selective α-glucosidase inhibitors with fewer side effects are urgently needed. In this study, a crude ethyl acetate extract of Machilus litseifolia was fractionated by solid-phase extraction using C18 cartridges to give a fraction enriched in α-glucosidase inhibitors. Subsequent microfractionation and bioassaying of the eluate by high-performance liquid chromatography (HPLC) using a complementary pentafluorophenyl column allowed construction of a high-resolution α-glucosidase inhibition profile (biochromatogram). This was used to target high-performance liquid chromatography−photodiode array detection−highresolution mass spectrometry−solid-phase extraction−nuclear magnetic resonance spectroscopy (HPLC-PDA-HRMS-SPENMR) analysis toward α-glucosidase inhibitors. This led to the identification of 13 dicoumaroylated flavonol rhamnosides, of which seven (8, 10, 12a, 12b, 16, 17, and 18) are reported for the first time, and two lignans, of which one (5) is reported for the first time. IC50 values of isolated compounds toward α-glucosidase range from 5.9 to 35.3 μM, which is 8 to 91 times lower than the IC50 value of 266 μM measured for the reference compound acarbose. currently available α-glucosidase inhibitors such as acarbose have gastrointestinal side effects, such as abdominal bloating, cramping, increased flatulence, or diarrhea.6 Thus, new selective α-glucosidase inhibitory drugs with diminished side effects are urgently needed. Nature is a rich source of bioactive compounds, and 51% of drugs approved from 1981 to 2014 were derived from or inspired by natural products.7 Acarbose and voglibose, two clinically used α-glucosidase inhibitors, are of natural origin,8,9 and several natural products with α-glucosidase inhibitory activity have been identified.10 However, the traditional bioassay-guided fractionation approach used for isolation of bioactive constituents from crude extracts is laborious and time-consuming, and new bioanalytical technologies for accelerated identification of bioactive constituents in crude extracts are needed. One of these new bioanalytical

D

iabetes is a chronic multifactorial disease characterized by elevated and fluctuating blood glucose levels. Diabetic hyperglycemia is caused by a lowered production of insulin by pancreatic β-cells or a lowered response to insulin in the liver and muscle cells. Prolonged hyperglycemia in diabetics causes micro- and macrovascular complications such as diabetic retinopathy, neuropathy, nephropathy, and cardiovascular diseases.1,2 According to the International Diabetes Federation, there are 425 million people with diabetes worldwide, and this number is expected to increase to 629 million in 2045.3 Type 2 diabetes (T2D), caused by a lowered response to insulin in liver and muscle cells, comprises approximately 90% of all diabetes incidences.1 α-Glucosidase is a carbohydrate-metabolizing enzyme occurring in the brush border of the small intestines, where it hydrolyzes terminal nonreducing 1,4-linked α-glucose residues, which are subsequently transferred into the bloodstream. Thus, α-glucosidase inhibitors can lower postprandial blood glucose levels by inhibiting the hydrolysis of dietary carbohydrates in the small intestine.4,5 However, © XXXX American Chemical Society and American Society of Pharmacognosy

Received: July 25, 2018

A

DOI: 10.1021/acs.jnatprod.8b00609 J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products

Article

Figure 1. HPLC chromatogram monitored at 254 nm and the corresponding high-resolution α-glucosidase inhibition profiles of fraction F3 of the crude ethyl acetate extract of Machilus litseifolia.

H]+ ions with m/z 741.1840, 755.2012, 725.1858, and 739.1995, which suggested the molecular formulas C39H32O15, C40H34O15, C39H32O14, and C40H34O14, respectively. This indicated the presence of coumaroylated flavonoid glycosides in fraction F3,23,29,30 a class of compounds of which two previously have been reported for their α-glucosidase inhibitory activity.23 The very lipophilic fraction F5 was a dark green residue with 63% inhibition at 4 μg/mL, but HPLCPDA-HRMS analysis revealed a broad continuum of nonseparable peaks with MS (m/z 593.2756, 621.3058, and 887.5657) and UV (λmax 410, 665, and 730 nm) spectra indicating this fraction to consist of mainly chlorophylls, such as pheophorbide A, α-ethyl pheophorbide, and 3-acetylpheophytin A. Chlorophylls have previously been shown to have strong α-glucosidase inhibitory activity, but generally are otherwise not interesting in drug discovery, and thus fraction F5 was not further investigated. The LC-HRMS analysis of F3, vide supra, indicated the occurrence of several closely related dicoumaroylated flavonoid glycosides. The best separation of these was obtained using a pentafluorophenyl (PFP) column, due to π−π interactions between the pentafluorophenyl moieties on the stationary phase and the aromatic moieties of the analytes. Thus, analytical-scale HPLC using a PFP column and a 76 min gradient of methanol in water provided satisfactory separation of the coumaroylated flavonoid glycosides (see blue trace in Figure 1). The eluate from 1 to 75 min was fractionated into 176 fractions of two 96-well microplates, and the material in each well was assessed for its α-glucosidase inhibitory activity. The results expressed as percent inhibition were plotted at its respective retention time to provide a high-resolution inhibition profile (biochromatogram) with a resolution of 2.4 data points per minute (see red trace in Figure 1). This shows that HPLC peaks 7−14 and 15−19 are correlated with peaks in the biochromatogram, and these peaks as well as wellseparated peaks 1−6 and 15 were subjected to HPLC-PDAHRMS-SPE-NMR analysis. Fraction F3 underwent HPLC-PDA-HRMS-SPE-NMR analysis, with 10 cumulative trappings of peaks 1−19 after 10 successive separations of 0.2 mg per separation (Figure S2, Supporting Information). The amount of material trapped was not sufficient to obtain high-quality NMR spectra of all constituents, and the HPLC-PDA-HRMS-SPE-NMR analysis was therefore repeated with 10 cumulative trappings of peaks 1−19 after 10 successive separations of 0.5 mg per separation (Figure S3, Supporting Information). Compound numbers used in the following correspond to peak numbers; that is, compound 2 corresponds to the material eluted with peak 2,

technologies is high-resolution inhibition profiling combined with hyphenation of high-performance liquid chromatography, photodiode array detection, solid-phase extraction, highresolution mass spectrometry, and nuclear magnetic resonance spectroscopy, i.e., HR-inhibition profiling/HPLC-PDA-SPEHRMS-NMR. High-resolution inhibition profiling provides a biochromatogram that allows pinpointing of HPLC peaks correlated with inhibitory activity and thereby targets subsequent HPLC-PDA-SPE-HRMS-NMR toward bioactive constituents only. This technology has been successfully used for accelerated identification of inhibitors of important T2D targets, i.e., α-glucosidase,11−13 α-amylase,14,15 aldose reductase,16 and PTP1B,15−17 as well as radical scavengers that are involved in the development and progression of T2D.18−20 Machilus litseifolia S. Lee (Lauraceae) is distributed widely in southwest mainland China, and several Machilus species, i.e., M. grijsii, M. leptophylla, M. pauhoi, and M. yunnanensis, have been used in traditional Chinese medicine.21 Flavonoids,22,23 butanolides,24,25 and alkaloids26 are the main classes of compounds in this genus, and various acylated flavonol monoglycosides and proanthocyanidins with α-glucosidase inhibitory activity have been identified in Machilus sp.23,27,28 However, no secondary metabolites have previously been reported from M. litseifolia. Herein, we report the use of HR αglucosidase inhibition profiling/HPLC-PDA-HRMS-SPENMR aiming at discovering α-glucosidase inhibitors from M. litseifolia.

■

RESULTS AND DISCUSSION A crude ethyl acetate extract of M. litseifolia showed inhibition of α-glucosidase with an IC50 value of 8.0 ± 0.95 μg/mL. Initial attempts to separate all analytes of the crude extract for high-resolution inhibition profiling were unsuccessful due to the complexity of the extract. The crude extract (2 g) was therefore subjected to reversed-phase solid-phase extraction on LC-18 packing cartridges (60 mL, 10 g) and eluted with 0%, 25%, 50%, 75%, and 100% of water−CH3CN (5:95) added to 0.1% formic acid to yield fractions F1−F5. These five fractions were subjected to HPLC-HRMS, and the total ion current chromatograms are shown in Figure S1, Supporting Information. Fractions F1 and F2 showed broad baseline humps with no distinct peaks, and with only 0.8 and 1.2 mg of material available, no further investigations of the two fractions were performed. Fractions F3 and F4 displayed several distinct peaks, and these fractions were assessed for α-glucosidase inhibitory activity. Fraction F3 exhibited 45% inhibition at a concentration of 25 μg/mL, whereas fraction F4 gave only 24% inhibition at 50 μg/mL. At the same time, preliminary LCHRMS analysis of fraction F3 showed four different main [M + B



Article

−3.5 ppm and C21H26O5Na+, ΔM −2.2 ppm). The 1H NMR spectrum showed characteristic signals for two 1,2,4-trisubstituted benzene systems: δ 6.76 (d, 8.3 Hz, H-5), 7.24 (dd, 8.3 and 2.0 Hz, H-6), and 7.33 (d, 2.0 Hz, H-2), and δ 6.73 (dd, 8.1 and 2.0 Hz, H-6′), 6.87 (d, 8.1 Hz H-5′), and 6.77 (d, 2.0 Hz, H-2′). In addition, three methoxy groups at δ 3.81 (OCH310), 3.81 (OCH3-10′), and 3.78 (OCH3-11′) were positioned at C-3 (δ 148.8), C-4′ (δ 149.0), and C-3′ (δ 149.8), respectively, based on HMBC and ROESY correlations (Figure 3). COSY correlations from CH3-9 (δ 1.11) to CH-8 (δ 3.42), CH-8′ (δ 2.19), CH2-7′ (δ 2.53), and CH3-9′ (δ 0.85) as well as HMBC correlations from H-2, H-9, and H-8′ to C-7 (δ 204.6) established the structure of 5 as 4′-O-methylcinnamophilin, which is a new compound. Comparison of chemical shift and coupling patterns for H-8, H-9, H-7′, H-8′, and H-9′ of 2 and 5 shows that the core part of the two molecules has the same relative configuration, and since naturally occurring cinnamophilin (2) has the 8R,8′S configuration,31 5 was tentatively given an 8R,8′S configuration based on biosynthetic arguments. Full assignments of the 1H and 13C NMR data are given in Table 1, with selected HMBC and ROESY correlations provided in Figure 3 and spectra included in Figures S5−S10, Supporting Information. The material eluted with peaks 8 and 10 showed [M + H]+ ions at m/z 755.1953 (C40H35O15+, ΔM −2.4 ppm) and 755.1988 (C40H35O15+, ΔM 2.3 ppm), respectively, which suggested the molecular formula of C40H34O5 for both compounds. Their 1H NMR spectra showed characteristic signals for a tamarixetin moiety, e.g., δ 7.41 (1H, dd, 8.7, 2.2 Hz, H-6′); 7.34 (1H, d, 2.2 Hz, H-2′), 7.07 (1H, d, 8.7 Hz, H5′), 6.40 (1H, d, 2.1 Hz, H-8), 6.23 (1H, d, 2.1 Hz, H-6), and 3.92 (3H, s, OCH3-7′) for 8. Both compounds also showed signals for a rhamnopyranosyl unit, e.g., δ 5.51 (1H, overlapping signals, H-1″), 5.51 (1H, overlapping signals, H2″), 4.15 (1H, dd, 9.8, 2.9 Hz, H-3″), 4.86 (1H, overlapping signals, H-4″), 3.46 (1H, dd, 9.8, 6.3 Hz, H-5″), and 0.86 (3H, d, 6.3 Hz, H-6″) for 8. However, whereas 8 showed signals for two coumaroyl units both with cis configuration of the double bonds, δ 7.65 (2H, XX′, H-2‴/H-6‴), 6.77 (2H, AA′, H-3‴/ H-5‴), 6.89 (1H, d, 12.8 Hz, H-7‴), 5.75 (1H, d, 12.8 Hz, H8‴) and δ 7.66 (2H, XX′, H-2⁗/H-6⁗), 6.75 (2H, AA′, H3⁗/H-5⁗), 6.90 (1H, d, 12.8 Hz, H-7⁗), 5.86 (1H, d, 12.8 Hz, H-8⁗), 10 showed signals for one coumaroyl unit with cis configuration and one with trans configuration of the double bond: δ 7.50 (2H, XX′, H-2‴/H-6‴), 6.80 (2H, AA′, H-3‴/H5‴), 7.68 (1H, d, 15.9 Hz, H-7‴), 6.42 (1H, d, 15.9 Hz, H-8‴) and 7.67 (2H, XX′, H-2⁗/H-6⁗), 6.75 (2H, AA′, H-3⁗/H5⁗), 6.90 (1H, d, 12.9 Hz, H-7⁗), 5.76 (1H, d, 12.9 Hz, H8⁗). Key HMBC correlations (Figure 3) finally allowed positioning the coumaroyl units at C-2″ and C-4″ of the rhamnopyranosyl moieties, which explained the large downfield shift observed for H-2″ and H-4″. Thus, the structure of 8 was established to be tamarixetin 3-O-(2″,4″-di-Z-p-coumaroyl)-α-L-rhamnopyranoside, and the structure of 10 was established to be tamarixetin 3-O-(2″-E-p-coumaroyl,4″-Z-pcoumaroyl)-α-L-rhamnopyranoside. Both are new compounds, and full assignments of 1H and 13C NMR data are presented in Table 2, selected HMBC and ROESY correlations appear in Figure 3, and spectra are included in Figures S12−S17 and S19−S24, Supporting Information. The material eluted with peak 12 showed a [M + H]+ ion at m/z 755.1934 (C40H35O15+, ΔM −4.8 ppm), which suggested the molecular formula C40H34O15. However, the 1H NMR

compounds 12a and 12b correspond to two compounds coeluting with peak 12, etc. For peaks 16−18, additional preparative-scale isolation was performed to obtain sufficient material for acquisition of 2D homo- and heteronuclear NMR experiments. Comparison of HRMS and 1D and 2D NMR data obtained in the HPLC-PDA-HRMS-SPE-NMR mode with data from the literature allowed identification of 2 as cinnamophilin,31 7a as quercetin 3-O-(2″,4″-di-E-p-coumaroyl)-α-L-rhamnopyranoside,32 9 as kaempferol 3-O-(2″,4″-di-Z-p-coumaroyl)-α-Lrhamnopyranoside,23 11 as kaempferol 3-O-(2″-E-p-coumaroyl,4″-Z-p-coumaroyl)-α-L-rhamnopyranoside,33 13 as kaempferol 3-O-(2″-Z-p-coumaroyl,4″-E-p-coumaroyl)-α-L-rhamnopyranoside,34 14 as kaempferol 3-O-(2″,4″-di-E-p-coumaroyl)-α-L-rhamnopyranoside,23,33 and 19 as 4′-O-methyl-2″,4″di(E)-p-coumaroylafzelin,29 respectively (Figure 2). Retention time, HRMS, and 1H NMR data of 2, 7a, 11, 13, 14, and 19 are given in the Supporting Information (Table S1).

Figure 2. Compounds identified from fraction F3 of the crude ethyl acetate extract of Machilus litseifolia.

The 1H NMR spectrum of the material eluted with peak 7 showed 7a as described above as well as a minor coeluting constituent, 7b. This minor constituent resembled 7a based on the 1H NMR data, but the limited amount obtained prohibited further purification and structural identification of 7b. The 1H NMR spectra of the material eluted with peaks 1, 3, 4, 6, and 15 were also of low quality due to limited amounts of these compounds, but since they were not correlated with peaks in the biochromatogram, no further attempts were made to identify their structure. The material eluted with peak 5 showed a [M + H]+ ion at m/z 359.1840 and a [M + Na]+ at m/z 381.1664, which suggested the molecular formula C21H26O5 (C21H27O5+, ΔM C



Article

Figure 3. Selected HMBC and ROESY correlations of new compounds from fraction F3 of the crude ethyl acetate extract of Machilus litseifolia.

respectively, showed two trans-p-coumaroyl units to be positioned at C-2″ and C-4″ of the rhamnopyranosyl unit. For 12b, HMBC correlations from H-4″ and from H-7⁗ (δ 7.60, 1H, d, 15.9 Hz) to C-9⁗ (δ 168.0) showed the trans-pcoumaroyl unit to be positioned at C-4′, and the downfield shift of H-2″ showed the remaining cis-p-coumaroyl unit to be positioned at C-2″. Thus, 12a was identified as tamarixetin-3O-(2″,4″-di-E-p-coumaroyl)-α-L-rhamnopyranoside and 12b as tamarixetin-3-O-(2″-Z-p-coumaroyl,4″-E-p-coumaroyl)-α-Lrhamnopyranoside. Both 12a and 12b are new compounds, and full assignments of their 1H and 13C NMR data are given in Table 2, with selected HMBC and ROESY correlations presented in Figure 3 and the spectra included in Figures S26− S33, Supporting Information. The material eluted with peak 16 showed a [M + H]+ ion at m/z 739.2017 (C40H35O14+, ΔM 0.6) and a [M + Na]+ ion at

spectrum showed two similar compounds, 12a and 12b, in the ratio 2:1 by integration (Figures S27 and S28, Supporting Information). Both compounds exhibited the same pattern for a tamarixetin moiety with an ABX system for H-2′, H-5′, and H-6′, an AX system for H-6 and H-8, and an aromatic methoxy group as well as signals for a rhamnopyranosyl unit (see Table 2 for full assignments). HMBC correlations from H-1″ (δ 5.70 for 12a and δ 5.62 for 12b) to C-3 (δ 134.6 for 12a and δ 135.0 for 12b) indicated the same glycosidic linkage between C-1″ of the rhamnopyranosyl unit and C-3 of tamarixetin as also observed for 8 and 10. In total, four p-coumaroyl units were observed, three of them with trans configuration of the double bond and one of them with cis configuration of the double bond. For 12a, HMBC correlations from H-2″ and H4″ and from H-7‴ (δ 7.68, 1H, d, 15.9 Hz) and H-7⁗ (δ 7.61, 1H, d, 15.9 Hz) to C-9‴ (δ 167.9) and C-9⁗ (δ 168.2), D



Article

Table 1. 1H NMR (600 MHz) and 13C NMR (151 MHz) Spectroscopic Data of 5a position

δC, type

1 2 3 4 5 6 7 8 9 10 1′ 2′ 3′ 4′ 5′ 6′ 7′

129.6, C 112.1, CH 148.8, C 153.0, C 115.4, CH 124.2, CH 204.6, C 43.6, CH 11.6, CH3 56.3, CH3 134.8, C 114.1, CH 149.8, C 149.0, C 112.7, CH 122.5, CH 41.9, CH2

8′ 9′ 10′ 11′

38.8, 15.4, 56.3, 56.0,

CH CH3 CH3 CH3

a cis-p-coumaroyl and a trans-p-coumaroyl unit. For 17, HMBC correlations from H-2″ and H-7‴ (δ 7.67, 1H, d, 15.9 Hz) to C-9‴ (δ 168.0) and from H-4″ and H-7⁗ (δ 6.90, 1H, d, 12.8 Hz) to C-9⁗ (δ 167.2) demonstrated the trans-p-coumaroyl unit as being positioned at C-2″ and the cis-p-coumaroyl unit to be located at C-4″ of the rhamnopyranosyl unit. The opposite was observed for 18, where HMBC correlations from H-2″ and H-7‴ (δ 6.91, 1H, d, 12.9 Hz) to C-9‴ (δ 166.8) and from H-4″ and H-7⁗ (δ 7.57, 1H, d, 15.9 Hz) to C-9⁗ (δ 168.2) indicated the cis-p-coumaroyl unit as being positioned at C-2″ and the trans-p-coumaroyl unit located at C-4″ of the rhamnopyranosyl unit. Thus, 17 was identified as 4′-O-methyl(2″-E-p-coumaroyl-4″-Z-p-coumaroyl)afzelin and 18 as 4′-Omethyl-(2″-Z-p-coumaroyl-4″-E-p-coumaroyl)afzelin. Both 17 and 18 are new compounds, and full assignments of their 1H and 13C NMR data are given in Table 3, with selected HMBC and ROESY correlations given in Figure 3 and spectra of 17 and 18 included in Figures S42−S47 and S48−S53, Supporting Information, respectively. IC50 values of 10, 11, 13, 14, 16−19, and acarbose toward α-glucosidase from Saccharomyces cerevisiae were determined in triplicate from dose−response curves of nine concentrations. Although compounds 7a, 7b, 8, 9, 12a, and 12b also showed α-glucosidase inhibitory activity according to the highresolution α-glucosidase inhibition profile in Figure 1, the limited amount of 8 and 9 and the inseparable mixtures of 7a/ 7b and 12a/12b did not allow assessment of their IC50 values. The inhibitory activities of all tested compounds are shown in Table 4. The IC50 values observed were in the range 5.9 to 35.3 μM, which is 8 to 91 times more potent than the analogous value of 266.1 μM for the reference compound acarbose. The inhibitory activity of S. cerevisiae α-glucosidase by the compounds isolated in this study is in the same low micromolar range as also reported for two dicoumaroylated kaempferol rhamnopyranosides toward Bacillus stearothermophilus type IV α-glucosidase.23 However, this is the first report of the inhibitory activity of dicoumaroylated flavonol rhamnopyranosides toward α-glucosidase from S. cerevisiae. In general, many flavonoids and flavonoid glycosides show strong α-glucosidase inhibitory activity.35,36 However, a detailed structure−activity relationship (SAR) study of a large number of flavonoid glycosides based on IC 50 determinations using α-glucosidase from the same organism and with the exact same protocol is still lacking. In present study, there seems to be a trend that 4′-methylkaempferol analogues generally show stronger inhibitory activity than quercetin and kaempferol analogues. In this kaempferol series, the presence of one or two Z-p-coumaroyl units favors inhibitory activity compared to E-p-coumaroyl units. However, the lack of all combinations of Z- and E-di-p-coumaroylated kaempferol, quercetin, and 4′-methylquercetin analogues prevents a more detailed SAR discussion. In conclusion, solid-phase extraction with C18 material of a crude ethyl acetate extract of M. litseifolia provided one fraction enriched in α-glucosidase inhibitors, and high-resolution αglucosidase inhibition profiling of this fraction using a complementary PFP column allowed construction of a biochromatogram. Subsequent HPLC-PDA-HRMS-SPENMR analysis led to identification of all major α-glucosidase inhibitors, including seven new dicoumaroylated flavonol rhamnosides. Further investigations of the reported dicoumaroylated flavonoid rhamnosides using mammalian α-glucosidase

δH (J in Hz) 7.33, d (2.0)

6.76, d (8.3) 7.24, dd (8.4, 2.0) 3.42, m 1.11, d (6.8) 3.81, s 6.77, d (2.0)

6.87, d (8.1) 6.73, dd (8.1, 2.0) A: 2.49, dd (13.4, 6.8) B: 2.57, dd (13.4, 7.8) 2.19, m 0.85, d (6.9) 3.81, s 3.78, s

a

NMR data obtained from HPLC-PDA-HRMS-SPE-NMR analysis with methanol-d4 as elution solvent.

m/z 761.1829 (C40H34O14Na+, ΔM 1.6), which suggested the molecular formula C40H34O14. The 1H NMR spectrum of 16 afforded signals corresponding to a 4′-O-methylkaempferol unit, i.e., an AA′XX′ system for H-2′/H-6′ (δ 7.86) and H-3′/ H-5′ (δ 7.10), an AX system for H-6 (δ 6.22, d, 1.9 Hz) and H-8 (δ 6.39, d, 1.9 Hz), and a methoxy group at C-4′ (δ 3.87, 3H, s). In addition, a rhamnopyranosyl moiety was identified by signals at δ 5.49 (d, 1.7 Hz, H-1″), 5.50 (dd, 3.2, 1.7 Hz, H2″), 4.14 (dd, 9.9, 2.9 Hz, H-3″), 4.86 (overlapping, H-4′), 3.46 (dd, 9.8, 6.2 Hz, H-5″), and 0.83 (d, 6.3 Hz, H-6″), and two p-coumaroyl units with a cis-configuration were identified by signals at δ 6.90 (d, 12.9 Hz, H-7⁗), 5.85 (d, 12.9 Hz, H8⁗), 6.90 (d, 12.9 Hz, H-7‴), and 5.72 (d, 12.9 Hz, H-8‴) in addition to two AA′XX′ spin systems (see Table 3 for resonance assignments). HMBC correlations from H-1″ to C-3 (139.9) and from H-4″ to C-9⁗ (δ 167.1) suggested the rhamnopyranosyl unit is attached to C-3 of the 4′-Omethylkaempferol and one of the cis-p-coumaroyl units to be positioned at C-4″ of the rhamnopyranosyl unit. Although no clear HMBC correlation could be observed from H-2″ to C9‴, the position of the last cis-p-coumaroyl unit was shown to be at C-2″ of the rhamnopyranosyl due to the large downfield shift of H-2″ (δ 167.1). Thus, 16 was identified as 4′-Omethyl-2″,4″-di(Z)-p-coumaroylafzelin, which is a new compound. Full assignments of the 1H and 13C NMR data are given in Table 3, with selected HMBC and ROESY correlations presented in Figure 3 and the spectra provided in Figures S36−S41, Supporting Information. The material eluted with peaks 17 and 18 showed [M + H]+ ions at m/z 739.1988 (C40H35O14+, ΔM −3.1) and 739.1989 (C40H35O14+, ΔM −4.4), suggesting the molecular formula C40H34O14 for both, as also observed for 16 and 19. The 1H NMR spectra of 17 and 18 also resembled those of 16 and 19, but with the differences that 17 and 18 each showed signals for E



Article

Table 2. 1H NMR (600 MHz) and 13C NMR (151 MHz) Spectroscopic Data of 8, 10, 12a, and 12b 8a position

δC, type

2 3 4 5 6 7 8 9 10 1′ 2′ 3′ 4′ 5′ 6′ 7′ Rha 1″ 2″ 3″ 4″ 5″ 6″ Coum-1 1‴ 2‴/6‴ 3‴/5‴ 4‴ 7‴ 8‴ 9‴ Coum-2 1⁗ 2⁗/6⁗ 3⁗/5⁗ 4⁗ 7⁗ 8⁗ 9⁗

158.4, C 135.2, C n.d.b 163.0, C 99.7, CH 163.1, C 94.7, CH 158.5, C 105.5, C 123.4, C 116.4, CH 147.9, C 151.6, C 112.4, CH 122.3, CH 56.4, CH3 99.6, 72.6, 68.3, 74.3, 69.6, 17.5,

CH CH CH CH CH CH3

126.8, 133.7, 115.5, 161.2, 145.7, 115.9, 166.7,

C CH CH C CH CH C

126.9, 133.8, 115.7, 160.0, 145.8, 115.7, 168.2,

C CH CH C CH CH C

δH, (J in Hz)

10a δC, type

7.07, d (8.7) 7.41, dd (8.7, 2.2) 3.92, s

158.3, C 135.1, C n.d.b 163.0, C 99.7, CH 165.8, C 94.6, CH 158.4, C 105.7, C 123.9, C 116.3, CH 147.8, C 151.7, C 112.3, CH 122.3, CH 56.3, CH3

5.51 (overlapping)c 5.51 (overlapping)c 4.15, dd (9.8, 2.9) 4.86 (overlapping)c 3.46, dd (9.8, 6.3) 0.84, d (6.3)

99.5, 73.0, 68.2, 74.2, 69.5, 17.4,

6.23, d (2.1) 6.40, d (2.1)

7.34, d (2.2)

7.65 (XX′) 6.77 (AA′) 6.89, d (12.9) 5.75, d (12.9)

7.66 (XX′) 6.75 (AA′) 6.90, d (12.8) 5.86,d (12.8)

CH CH CH CH CH CH3

127.3, 131.1, 116.6, 160.0, 147.2, 114.4, 168.0,

C CH CH C CH CH C

127.0, 133.7, 115.6, 161.2, 145.6, 115.8, 167.2,

C CH CH C CH CH C

12aa

δH, (J in Hz)

δC, type

7.08, d (8.5) 7.41, dd (8.5, 2.2) 3.95, s


5.60, 5.54, 4.16, 4.94, 3.37, 0.85,

99.0, 72.8, 68.1, 74.4, 69.5, 17.5,

6.22, d (2.1) 6.39, d (2.1)

7.34, d (2.2)

d (1.8) dd (3.4, 1.8) dd (9.8, 3.4) t (9.8) d (9.9, 6.3) d (6.3)

7.50 (XX′) 6.80 (AA′) 7.68, d (15.9) 6.42, d (15.9)

7.67 (XX′) 6.75 (AA′) 6.90, d (12.9) 5.76, d (12.9)

CH CH CH CH CH CH3

127.0, 131.1, 115.5, 160.0, 147.1, 114.4, 167.9,

C CH CH C CH CH C

127.3, 131.3, 116.6, 161.2, 146.7, 114.8, 168.2,

C CH CH C CH CH C

δH, (J in Hz)

12ba δC, type

7.15, d (8.4) 7.42, dd (8.4, 2.3) 3.91, s


5.70, d (1.7) 5.54, dd (3.4,1.7) 4.19, dd (9.8, 3.5) 4.96, t (9.8) 3.32 (overlapping)c 0.85, d (6.3)

99.2, 72.4, 68.1, 74.4, 69.5, 17.5,

6.22, d (2.1) 6.40, d (2.1)

7.40, d (2.2)

7.50 (XX′) 6.81 (AA′) 7.68, d (15.9) 6.42, d (15.9)

7.54 (XX′) 6.80 (AA′) 7.61, d (15.9) 6.29, d (15.9)

CH CH CH CH CH CH3

127.0, 133.4, 116.7, 160.0, 145.7, 115.7, 168.2,

C CH CH C CH CH C

127.3, 131.1, 115.5, 161.2, 146.7, 114.8, 168.0,

C CH CH C CH CH C

δH, (J in Hz)

6.22, d (2.1) 6.40, d (2.1)

7.39, d (2.3)

7.14, d (8.5) 7.42 (overlapping)c 3.89, s 5.62, d (1.6) 5.52, dd (3.4, 1.8) 4.19, dd (9.7, 3.5) 4.88 (overlapping)c 3.39, dd (9.8, 6.3) 0.83, d (6.2)

7.66 (XX′) 6.77 (AA′) 6.91, d (12.8) 5.86, d (12.8)

7.53 (XX′) 6.84 (AA′) 7.60, d (15.9) 6.29, d (15.9)

a

NMR data obtained from HPLC-PDA-HRMS-SPE-NMR analysis with methanol-d4 as elution solvent. bn.d.: not detected. cSignals overlapping with residual solvent signal of methanol at δ 3.31, water at δ 4.87, or other signals from the molecule. (Maping Group, Gaoqiaoba Village, Enshi, Hubei, People’s Republic of China) in September 2016 and were identified by the botanist Zien Zhao (Wuhan Botanical Garden, Chinese Academy of Sciences, Wuhan, Hubei, China). A voucher specimen (accession number GD138) is deposited at the herbarium of this company. Extraction and Solid-Phase Extraction. Milled plant material (350 g) was extracted with ethyl acetate by sonication for 1 h (three times of 1 L each). The extract was evaporated under reduced pressure to give a crude extract (7 g). Supelclean LC-18 (60 mL, 10 g) SPE cartridges (Bellefonte, PA, USA) were preconditioned by 100 mL of mobile phase B (CH3CN−H2O (95:5) + 0.1% formic acid) followed by conditioning with 100 mL of mobile phase A (CH3CN− H2O (5:95) + 0.1% formic acid). The crude ethyl acetate extract of M. litseifolia (0.3 g) was mixed with 1.5 g of silica gel, loaded onto the SPE cartridges, and eluted with 0% B (3 × 75 mL), 25% B (3 × 75 mL), 50% B (3 × 75 mL), 75% B (3 × 75 mL), and 100% B (3 × 75 mL) to yield fractions F1 (0.8 mg), F2 (1.2 mg), F3 (14 mg), F4 (40 mg), and F5 (120 mg). Fractions F1−F5 were subjected to LCHRMS using the LC-HRMS mode of the HPLC-PDA-HRMS-SPE-

are needed to determine their potential as future T2D drug leads.

■

EXPERIMENTAL SECTION

General Experimental Procedures. α-Glucosidase type I (EC 3.2.1.20, from Saccharomyces cerevisiae, lyophilized powder), pnitrophenyl α-D-glucopyranoside (p-NPG) (≥99.9%), dimethyl sulfoxide (DMSO; ≥99.5%), sodium dihydrogen phosphate dihydrate, disodium hydrogen phosphate, sodium azide, acarbose, methanol-d4, and formic acid were purchased from Sigma-Aldrich (St. Louis, MO, USA). Ethyl acetate, methanol, and acetonitrile were all HPLC grade and obtained from VWR International (Søborg, Denmark). Milli-Q water was purified by 0.22 μm membrane filtration and deionization using a Millipore system (Billerica, MA, USA). The 96-well polypropylene microplates were obtained from Greiner BioOne (Frickenhausen, Germany). Plant Material. The air-dried aerial parts of Machilus litseifolia were purchased from Enshi Dongsheng Plant Development Company F



Article

Table 3. 1H NMR (600 MHz) and 13C NMR (151 MHz) Spectroscopic Data of 16, 17, and 18 16a position

δC

2 3 4 5 6 7 8 9 10 1′ 2′/6′ 3′/5′ 4′ 7′ Rha 1″ 2″ 3″ 4″ 5″ H-6″ Coum-1 1‴ 2‴/6‴ 3‴/5‴ 4‴ 7‴ 8‴ 9‴ Coum-2 1⁗ 2⁗/6⁗ 3⁗/5⁗ 4⁗ 7⁗ 8⁗ 9⁗

158.7, C 139.9, C n.d.c 163.0, C 99.8, CH 166.2, C 94.5, CH 158.4, C 105.5, C 123.4, C 131.4, CH 115.1, CH 163.4, C 55.8, CH3 99.2, 72.5, 68.1, 74.1, 69.5, 17.4,

17a

δH, mult. (J in Hz)

CH CH CH CH CH CH3

126.8, 133.6, 115.6, 160.0, 145.6, 115.6, n.d.c

C CH CH C CH CH

127.3, 133.6, 115.6, 160.0, 145.6, 114.7, 167.1,

C CH CH C CH CH C

δC


3.87, s

158.8, C 135.1, C n.d.c 163.0, C 98.8, CH 165.8, C 93.6, CH 158.3, C 105.8, C 123.4, C 130.6, CH 114.1, CH 163.3, C 55.0, CH3

5.49, d (1.7) 5.50, dd (3.2, 1.7) 4.14, dd (9.9, 2.9) 4.86 (overlapping)d 3.46, dd (9.8, 6.2) 0.83, d (6.3)

98.5, 72.0, 67.3, 73.3, 68.6, 16.7,

6.22, d (1.9) 6.39, d (1.9)

7.86 (XX′) 7.10 (AA′)

7.67 (XX′) 6.76 (AA′) 6.90, d (12.9) 5.85, d (12.9)

7.66 (XX′) 6.77 (AA′) 6.90, d (12.9) 5.72, d (12.9)

CH CH CH CH CH CH3

126.9, 130.1, 115.6, 161.2, 146.2, 113.5, 168.0,

C CH CH C CH CH C

127.2, 132.7, 114.6, 160.0, 144.6, 114.8, 167.2,

C CH CH C CH CH C

18a δC


3.89, s

158.9, C 135.0, C n.d.c 163.2, C 99.7, CH 165.7, C 94.5, CH 158.5, C 105.7, C 123.6, C 131.5, CH 115.1, CH 163.4, C 55.9, CH3

5.61, d (1.6) 5.50, dd (3.4, 1.8) 4.14, dd (9.8, 3.5) 4.94, t (9.8) 3.34 (overlapping)d 0.85, d (6.3)

99.3, 72.4, 68.0, 74.3, 69.5, 17.5,

6.22, d (1.3) 6.40, brs

7.85 (XX′) 7.10 (AA′)

7.50 (XX′) 6.81 (AA′) 7.68, d (15.9) 6.42, d (15.9)

7.68 (XX′) 6.76 (AA′) 6.91, d (12.9) 5.73, d (12.9)

CH CH CH CH CH CH3

127.0, 130.9, 116.5, 161.4, 145.7, 115.5, 166.8,

C CH CH C CH CH C

127.4, 133.5, 115.5, 160.0, 146.6, 114.7, 168.2,

C CH CH C CH CH C

6.23, brs 6.41, brs

7.90 (XX′) 7.19 (AA′) 3.86, s 5.61, d (1.6) 5.50, dd (3.2, 1.7) 4.16, dd (9.9, 2.9) 4.87 (overlapping)d 3.41, m 0.85, d (6.3)

7.50 (XX′) 6.84 (AA′) 6.91, d (12.9) 5.85, d (12.9)

7.66 (XX′) 6.77 (AA′) 7.57, d (15.9) 6.27, d (15.9)

a

NMR data obtained in methanol-d4 after preparative-scale purification. bNMR data obtained from HPLC-PDA-HRMS-SPE-NMR analysis with methanol-d4 as elution solvent. cn.d.: not detected. dSignals overlapping with residual solvent signal of methanol at δ 3.31, water at δ 4.87, or other signal from the molecule. phase Phenomenex Luna C18(2) column, 150 × 4.6 mm, 3 μm, 100 Å (Phenomenex, Torrance, CA, USA) maintained at 40 °C, and eluted with the following gradient of mobile phase A (CH3CN−H2O (5:95) + 0.1% formic acid) and mobile phase B (CH3CN−H2O (95:5) + 0.1% formic acid) at a flow rate of 0.5 mL/min: 0 min, 0% B; 30 min, 100% B; 50 min, 100% B. High-Resolution α-Glucosidase Inhibition Profiling. Microfractionation for high-resolution inhibition profiles was performed with an Agilent 1200 series instrument consisting of a G1322A degasser, a G1311A quaternary pump, a G1313A high-performance autosampler, a G1316A thermostated column compartment, a G1315B photodiode-array detector, and a G1364C fraction collector, all controlled by Agilent ChemStation vers. B.03.03 software (Agilent, Santa Clara, CA, USA). Separation of 10 μL of a 50 mg/mL solution of F3 was performed on a Phenomenex Kinetix PFP phase, 150 mm × 4.6 mm i.d., 2.6 μm particle size, 100 Å pore size (Phenomenex) with the following elution gradient of mobile phase A (MeOH−H2O (5:95) + 0.1% formic acid) and mobile phase B (MeOH−H2O (95:5) + 0.1% formic acid): 0 min, 60% B; 20 min, 62.5% B; 25 min, 64% B; 35 min, 65% B; 45 min, 65% B; 50 min, 65.5% B; 60 min, 65.5% B; 70

Table 4. IC50 Values of Compounds 10, 11, 13, 14, 16−19, and Acarbose compound

IC50 (μM)a

10 11 13 14 16 17 18 19 acarbose

28.2 ± 0.00 16.1 ± 0.01 35.3 ± 0.02 12.9 ± 0.00 5.9 ± 0.01 6.8 ± 0.00 10.1 ± 0.00 26.9 ± 0.01 266.1 ± 0.01

a

IC50 values measured in triplicate from dose−response curves of nine concentrations. NMR instrument described below. Separations of 10 μL injections of 10 mg/mL solutions of F1−F5 were performed using a reversedG



Article

previous section. In the first HPLC-PDA-HRMS-SPE-NMR analysis with an injection of 10 μL of 20 mg/mL of F3, the materials eluted with peaks 1−5, 7, 8, and 10−19 were trapped, and in the second analysis with injection of 10 μL of 50 mg/mL of F3 the materials eluted with peaks 2−15 and 17−19 were trapped. The SPE cartridges were conditioned with 1000 μL of MeOH (at 6 mL/min) and equilibrated with 500 μL of H2O (at 1 mL/min) prior to trapping. The loaded SPE cartridges were dried with pressurized N2 gas for 45 min each and subsequently eluted with CD3OD into 1.7 mm NMR tubes with a Gilson 215 liquid handler equipped with a 1 mm needle. Chromatographic separation, mass spectrometry, and analyte trapping on SPE cartridges were controlled using Hystar ver. 3.2 software (Bruker Daltonik, Bremen, Germany), whereas the elution process was controlled by Prep Gilson ST ver. 1.2 software (Bruker Biospin, Karlsruhe, Germany). Preparative-Scale Isolation. The crude ethyl acetate extract (4 g) was dissolved in methanol (100 mg/mL) and centrifuged for 5 min at 10 000 rpm before use. Injection volumes of 900 μL were separated on an Agilent 1100 HPLC system equipped with two preparativescale solvent delivery pumps, a multiple-wavelength detector and an autosampler. All separations were performed using a Phenomenex Luna C18 column, 250 mm × 21.2 mm i.d., 5 μm (Phenomenex), operated at room temperature and eluted with the following gradient of mobile phase A (CH3CN−H2O (5:95) + 0.1% formic acid) and mobile phase B (CH3CN−H2O (95:5) + 0.1% formic acid) at a flow rate of 17 mL/min: 0 min, 65% B; 20 min, 82.5% B; 25 min, 95% B; 30 min, 100% B; 55 min, 100% B. After 10 repeated separations, manual collection of the eluate from 2 to 7 min afforded 44.6 mg of material corresponding to fraction F3, as seen by comparing the HRMS data with that for fraction F3 as described in the Extraction and Solid-Phase Extraction subsection. This fraction was separated using the same Agilent 1200 analytical-scale HPLC and method as described in the High-Resolution α-Glucosidase Inhibition Profiling subsection, and a fraction containing peaks 16, 17, 18, and 19 was repurified (repeated injections of 10 μL of a 20 mg/mL solution) using the following gradient at a flow rate of 0.5 mL/min: 0 min, 70% B; 40 min, 72% B; 42 min, 100% B; 47 min, 100% B; 48 min, 70% B. This yielded 0.6 mg of 16, 0.5 mg of 17, 0.5 mg of 18, and 0.6 mg of 19. NMR Experiments. The NMR experiments were recorded in methanol-d4 at 300 K, and 1H and 13C NMR chemical shifts were referenced to the residual solvent signal of methanol-d4 (δ 3.31 and 49.0, respectively). 1H NMR spectra were recorded using 30° pulses and 64 k data points. For the 2D NMR experiments, phase-sensitive DQF-COSY and ROESY spectra were recorded using a gradientbased pulse sequence with a 12 ppm spectral width and 2k × 512 data points (processed with forward linear prediction to 1 k data points); a multiplicity-edited HSQC spectrum was acquired with the following parameters: 1JC,H = 145 Hz, spectral width 12 ppm for 1H and 170 ppm for 13C, 1730 × 256 data points (processed with forward both linear prediction and zero filling to 1k data points) and 1.0 s relaxation delay; HMBC experiment was optimized for nJC,H = 8.0 Hz (longrange), 1JC,H min = 125 Hz, and 1JC,H man = 160 Hz and acquired using a spectral width of 12 ppm for 1H and 240 ppm for 13C, 2k × 256 data points (processed with forward linear prediction to 512 data points and zero filling to 1k data points), and 1.0 s relaxation delay. (8S*,8′S*)-4′-O-Methylcinnamophilin (5): colorless oil; insufficient material was available to obtain optical rotation; HPLC-UV [(MeOH in H2O + 0.05% FA)] λmax 200, 230, 278, 310 nm; 1H and 13 C NMR data, see Table 1; (+) HRESIMS m/z [M + H]+ calcd for C21H27O5+, 359.1853, found 359.1840; [M + Na]+ calcd for C21H26O5Na+, 381.1672, found 381.1664. Tamarixetin 3-O-(2″,4″-di-Z-p-coumaroyl)-α-L-rhamnopyranoside (8): yellowish, amorphous solid; insufficient material was available to obtain optical rotation; HPLC-UV [(MeOH in H2O + 0.05% FA)] λmax 206, 254, 264, 313 nm; 1H and 13C NMR data, see Table 2; (+) HRESIMS m/z [M + H]+ calcd for C40H35O15+, 755.1970, found 755.1953; [M + Na]+ calcd for C40H34O15Na+, 777.1790, found 777.1777.

min, 67% B; 71 min, 100% B, and 76 min 100% B. The eluate from 1 to 76 min was fractionated into 176 wells of two 96-well microplates (omitting the eight wells in the last column for blank controls), leading to a resolution of 2.3 data points per minute. The material in all wells was evaporated to dryness under reduced pressure using a SPD121P Savant SpeedVac equipped with an OFP400 oil-free pump and an RVT400 refrigerated vapor trap (Thermo Scientific, Waltham, MA, USA), and α-glucosidase inhibitory activity was assessed for each well according to the previously described method.37 In short, 10 μL of DMSO was added into each well to dissolve material from microfractionation, and after this 90 μL of 100 mM phosphate buffer (2.65 g of NaH2PO4·2H2O, 4.70 g of Na2HPO4, and 0.10 g of NaN3 in 500 mL of Milli-Q water and adjusting the pH to 7.5) and 80 μL of enzyme solution (2.0 U/mL α-glucosidase enzyme in phosphate buffer) were added and mixed sequentially. After incubation at 28 °C for 10 min, 20 μL of substrate solution (10 mM p-NPG in phosphate buffer) was added to each well, and absorbance was measured at 405 nm every 30 s for 35 min. For each microplate, blank samples containing buffer, enzyme, and substrate were included in triplicate. Incubation and absorbance measurements were performed using a Thermo Scientific Multiskan FC microplate photometer (Thermo Scientific). Percent inhibition of α-glucosidase was calculated using the formula SLOPEsample zy ji zz × 100% %inhibition = jjj1 − z j SLOPE blank z{ k

where SLOPEsample and SLOPEblank are measured enzyme activities expressed as ΔAU/s of analyte-containing samples and blank samples, respectively. Percent inhibition of α-glucosidase for each well was plotted against their respective retention times underneath the HPLC chromatogram to give a high-resolution α-glucosidase inhibition profile (biochromatogram). Inhibitory activities of isolated compounds were assessed by dose− response curves using the same protocol, but adding 10 μL of serial dilutions of the analytes and the reference compound acarbose in DMSO to each well in triplicate. IC50 values were calculated using the below three-parameter equation: y − ymin F(x) = ymin + max x 1 + IC

( ) 50

where ymin is the background, ymax − ymin is the y-range, x is the concentration of the test analyte, and the equation has a standard Hill slope of −1.0 HPLC-PDA-HRMS-SPE-NMR Analysis. HPLC-PDA-HRMSSPE-NMR analyses were performed on a system consisting of an Agilent 1260 chromatograph, a Bruker micrOTOF-Q II mass spectrometer (Bruker Daltonik, Bremen, Germany), a Knauer Smartline 120 pump (Knauer, Berlin, Germany), a Spark Holland Prospekt-2 SPE unit (Spark Holland, Emmen, The Netherlands), a Gilson 215 liquid handler, and a Bruker Avance III 600 MHz NMR spectrometer. The Agilent 1260 system consisted of a G1322A degasser, a G1311A quaternary pump, a G1316A thermostated column compartment, and a G1315A photodiode-array detector. The column eluate was connected to a T-piece splitter directing 1% of the flow to the mass spectrometer and 99% of the flow to automated SPE trapping after postcolumn dilution with a 1 mL/min flow of H2O delivered with the Knauer pump. The micrOTOF-Q II mass spectrometer, equipped with an ESI source, was operated in the positive-ion mode using a drying temperature of 200 °C, a capillary voltage of 4100 V, a nebulizer pressure of 2.0 bar, and a drying gas flow of 7 L/min. A solution of sodium formate clusters was injected automatically at the beginning of each run to enable internal mass calibration. Selected peaks were trapped cumulatively on 10 × 2 mm i.d. Resin GP (general purpose, 5−15 mm, spherical shape, polydivinylbenzene phase) SPE cartridges from Spark Holland after 10 identical separations using UV absorption thresholds at 254 nm to trigger trapping. Separations were performed using the same column, mobile phases, flow rate, and elution gradient as described in the H



Article

Tamarixetin 3-O-(2″-E-p-coumaroyl,4″-Z-p-coumaroyl)-α-Lrhamnopyranoside (10): yellowish, amorphous solid; insufficient material was available to obtain optical rotation; HPLC-UV [(MeOH in H2O + 0.05% FA)] λmax 203, 255, 266, 314 nm; 1H and 13C NMR data, see Table 2; (+) HRESIMS m/z [M + H]+ calcd for C40H35O15+, 755.1970, found 755.1988; [M + Na]+ calcd for C40H34O15Na+, 777.1790, found 777.1766. Tamarixetin 3-O-(2″,4″-di-E-p-coumaroyl)-α-L-rhamnopyranoside (12a): yellowish, amorphous solid; insufficient material was available to obtain optical rotation; HPLC-UV [(MeOH in H2O + 0.05% FA)] λmax 205, 228, 253, 266, 318 nm;1H and 13C NMR data, see Table 2; (+) HRESIMS m/z [M + H]+ calcd for C40H35O15+, 755.1970, found 755.1934; [M + Na]+ calcd for C40H34O15Na+, 777.1790, found 777.1773. Tamarixetin 3-O-(2″-Z-p-coumaroyl,4″-E-p-coumaroyl)-α-Lrhamnopyranoside (12b): yellowish, amorphous solid; insufficient material was available to obtain optical rotation; HPLC-UV [(MeOH in H2O + 0.05% FA)] λmax 205, 228, 253, 266, 318 nm; 1H and 13C NMR data, see Table 2; (+) HRESIMS m/z [M + H]+ calcd for C40H35O15+, 755.1970, found 755.1934; [M + Na]+ calcd for C40H34O15Na+, 777.1790, found 777.1773. 4′-O-Methyl-2″,4″-di-(Z)-p-coumaroylafzelin (16): yellowish, amorphous solid; insufficient material was available to obtain optical rotation; HPLC-UV [(MeOH in H2O + 0.05% FA)] λmax 198, 230, 266, 314 nm; 1H and 13C NMR data, see Table 3; (+) HRESIMS m/z [M + H]+ calcd for C40H35O14+, 739.2021, found 739.2017; [M + Na]+ calcd for C40H34O14Na+, 761.1841, found 761.1829. 4′-O-Methyl-(2″-E-p-coumaroyl,4″-Z-p-coumaroyl)afzelin (17): yellowish, amorphous solid; insufficient material was available to obtain optical rotation; HPLC-UV [(MeOH in H2O + 0.05% FA)] λmax 197, 230, 265, 315 nm; 1H and 13C NMR data, see Table 3; (+) HRESIMS m/z [M + H]+ calcd for C40H35O14+, 739.2021, found 739.1988; [M + Na]+ calcd for C40H34O14Na+, 761.1841, found 761.1846. 4′-O-Methyl-(2″-Z-p-coumaroyl,4″-E-p-coumaroyl)afzelin (18): yellowish, amorphous solid; insufficient material was available to obtain optical rotation; HPLC-UV [(MeOH in H2O + 0.05% FA)] λmax 194, 225, 265, 315 nm; 1H and 13C NMR data, see Table 3; (+) HRESIMS m/z [M + H]+ calcd for C40H35O14+, 739.2021, found 739.1989; [M + Na]+ calcd for C40H34O14Na+, 761.1841, found 761.1804.

■

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS T.L. acknowledges the Chinese Scholarship Council for a scholarship (No. 201508530232). HPLC equipment used for high-resolution bioassay profiles was obtained via a grant from The Carlsberg Foundation, and the 600 MHz HPLC-PDAHRMS-SPE-NMR system was acquired through a grant from “Apotekerfonden af 1991”, The Carlsberg Foundation, and the Danish Agency for Science, Technology and Innovation via the National Research Infrastructure funds. A. Ö nder is thanked for technical assistance.

■

(1) American Diabetes Association.. Diabetes Care 2013, 36, S67− S74. (2) World Health Organization. Global Report on Diabetes; 2016; pp 13−14. (3) International Diabetes Federation. IDF Diabetes Atlas, 8th ed.; 2017; pp 7−9; http://www.diabetesatlas.org/. (4) Derosa, G.; Maffioli, P. Arch. Med. Sci. 2012, 8, 899−906. (5) He, Z. X.; Zhou, Z. W.; Yang, Y.; Yang, T.; Pan, S. Y.; Qiu, J. X.; Zhou, S. F. Clin. Exp. Pharmacol. Physiol. 2015, 42, 125−138. (6) Krans, H. M. J. In Side Effects of Drugs Annual; Aronson, J. K., Ed.; Elsevier, 2005; Vol. 28, Chapter 42, pp 509−527. (7) Newman, D. J.; Cragg, G. M. J. Nat. Prod. 2016, 79, 629−661. (8) Truscheit, E.; Frommer, W.; Junge, B.; Muller, L.; Schmidt, D. D.; Wingender, W. Angew. Chem., Int. Ed. Engl. 1981, 20, 744−761. (9) Chen, X. L.; Zheng, Y. G.; Shen, Y. C. Curr. Med. Chem. 2006, 13, 109−116. (10) Kumar, S.; Narwal, S.; Kumar, V.; Prakash, O. Pharmacogn. Rev. 2011, 5, 19−29. (11) Kongstad, K. T.; Ozdemir, C.; Barzak, A.; Wubshet, S. G.; Staerk, D. J. Agric. Food Chem. 2015, 63, 2257−2263. (12) Liu, B.; Kongstad, K. T.; Qinglei, S.; Nyberg, N. T.; Jäger, A. K.; Staerk, D. J. Nat. Prod. 2015, 78, 294−300. (13) Wubshet, S. G.; Moresco, H. H.; Tahtah, Y.; Brighente, I. M. C.; Staerk, D. Phytochemistry 2015, 116, 246−252. (14) Okutan, L.; Kongstad, K. T.; Jäger, A. K.; Staerk, D. J. Agric. Food Chem. 2014, 62, 11465−11471. (15) Zhao, Y.; Kongstad, K. T.; Jäger, A. K.; Nielsen, J.; Staerk, D. J. Chromatogr. A 2018, 1556, 55−63. (16) Tahtah, Y.; Kongstad, K. T.; Wubshet, S. G.; Nyberg, N. T.; Jonsson, L. H.; Jäger, A. K.; Qinglei, S.; Staerk, D. J. Chromatogr. A 2015, 1408, 125−132. (17) Wubshet, S. G.; Tahtah, Y.; Heskes, A. M.; Kongstad, K. T.; Pateraki, I.; Hamberger, B.; Møller, B. L.; Staerk, D. J. Nat. Prod. 2016, 79, 1063−1072. (18) Wubshet, S. G.; Nyberg, N. T.; Tejesvi, M. V.; Pirttilä, A. M.; Kajula, M.; Mattila, S.; Staerk, D. J. Chromatogr. A 2013, 1302, 34− 39. (19) Wiese, S.; Wubshet, S. G.; Nielsen, J.; Staerk, D. Food Chem. 2013, 141, 4010−4018. (20) Wubshet, S. G.; Schmidt, J. S.; Wiese, S.; Staerk, D. J. Agric. Food Chem. 2013, 61, 8616−8623. (21) Jiangsu New Medical College Dictionary of Traditional Chinese Medicine; Shanghai Science and Technology Publishing House: Shanghai, 1977; Vol. 114; pp 1009, 1423. (22) Cheng, W.; Zhu, C. G.; Xu, W. D.; Fan, X. N.; Yang, Y. C.; Li, Y.; Chen, X. G.; Wang, W. J.; Shi, J. G. J. Nat. Prod. 2009, 72, 2145− 2152. (23) Lee, S. S.; Lin, H. C.; Chen, C. K. Phytochemistry 2008, 69, 2347−2353. (24) Giang, P. M.; Otsuka, H.; Son, P. T. Chem. Nat. Compd. 2011, 47, 201−202.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jnatprod.8b00609.

■

REFERENCES

Table S1 with retention times, names, HRMS and 1H NMR data of compounds 2, 5, 8−14, and 16−19 obtained in the HPLC-PDA-HRMS-SPE-NMR mode; Figure S1 showing total ion chromatogram of fractions F1−F5 from solid-phase extraction; Figures S2 and S3 showing trapped peaks from HPLC-PDA-HRMS-SPENMR analysis of fraction F3; and Figures S4−S54 with NMR spectra of known and new compounds identified in this study (PDF)

AUTHOR INFORMATION

Corresponding Author

*Tel: +45 3533 6177. Fax: +45 3533 6041. E-mail: ds@sund. ku.dk (D. Staerk). ORCID

Tuo Li: 0000-0002-5813-9385 Kenneth T. Kongstad: 0000-0003-4487-7886 Dan Staerk: 0000-0003-0074-298X I



Article

(25) Cheng, M. J.; Tsai, I. L.; Lee, S. J.; Jayaprakasam, B.; Chen, I. S. Phytochemistry 2005, 66, 1180−1185. (26) Liu, M. T.; Lin, S.; Wang, Y. H.; He, W. Y.; Li, S.; Wang, S. J.; Yang, Y. C.; Shi, J. G. Org. Lett. 2007, 9, 129−132. (27) Zhao, J.; Ding, H. X.; Song, Q. Y.; Gao, K. Chem. Biodiversity 2011, 8, 1943−1957. (28) Lin, H. C.; Lee, S. S. J. Nat. Prod. 2010, 73, 1375−1380. (29) Huang, H. C.; Yang, C. P.; Wang, S. Y.; Chang, C. I.; Sung, P. J.; Huang, G. J.; Chien, S. C.; Kuo, Y. H. RSC Adv. 2017, 7, 50868− 50874. (30) Ibrahim, M. A.; Mansoor, A. A.; Gross, A.; Ashfaq, M. K.; Jacob, M.; Khan, S. I.; Hamann, M. T. J. Nat. Prod. 2009, 72, 2141− 2144. (31) Wu, T. S.; Leu, Y. L.; Chan, Y. Y.; Yu, S. M.; Teng, C. M.; Su, J. D. Phytochemistry 1994, 36, 785−788. (32) Rao, L. J. M.; Yada, H.; Ono, H.; Yoshida, M. J. Agric. Food Chem. 2002, 50, 3143−3146. (33) Wang, G. J.; Tsai, T. H.; Lin, L. C. Phytochemistry 2007, 68, 2455−2464. (34) Yang, N. Y.; Tao, W. W.; Duan, J. A. Nat. Prod. Res. 2010, 24, 1843−1849. (35) Tadera, K.; Minami, Y.; Takamatsu, K.; Matsuoka, T. J. Nutr. Sci. Vitaminol. 2006, 52, 149−153. (36) Proença, C.; Freitas, M.; Ribeiro, D.; Oliveira, E. F. T.; Sousa, J. L. C.; Tomé, S. M.; Ramos, M. J.; Silva, A. M. S.; Fernandes, P. A.; Fernandes, E. J. Enzyme Inhib. Med. Chem. 2017, 32, 1216−1228. (37) Schmidt, J. S.; Lauridsen, M. B.; Dragsted, L. O.; Nielsen, J.; Staerk, D. Food Chem. 2012, 135, 1692−1699.

J


Identification of Î±-Glucosidase Inhibitors in

Recommend Documents