High-Resolution PTP1B Inhibition Profiling Combined with HPLC-HRMS-SPE-NMR for Identification of PTP1B Inhibitors from Miconia albicans

Protein tyrosine phosphatase 1B (PTP1B) is an intracellular enzyme responsible for deactivation of the insulin receptor, and consequently acts as a negative regulator of insulin signal transduction. In recent years, PTP1B has become an important target for controlling insulin resistance and type 2 diabetes. In the present study, the ethyl acetate extract of leaves of Miconia albicans (IC50 = 4.92 µg/mL) was assessed by high-resolution PTP1B inhibition profiling combined with HPLC-HRMS-SPE-NMR for identification of antidiabetic compounds. This disclosed eleven PTP1B inhibitors, including five polyphenolics: 1-O-(E)-caffeoyl-4,6-di-O-galloyl-β-d-glucopyranose (2), myricetin 3-O-α-l-rhamnopyranoside (3), quercetin 3-O-(2″-galloyl)-α-l-rhamnopyranoside (5), mearnsetin 3-O-α-l-rhamnopyranoside (6), and kaempferol 3-O-α-l-arabinopyranoside (8) as well as eight triterpenoids: maslinic acid (13), 3-epi-sumaresinolic acid (14), sumaresinolic acid (15), 3-O-cis-p-coumaroyl maslinic acid (16), 3-O-trans-p-coumaroyl maslinic acid (17), 3-O-trans-p-coumaroyl 2α-hydroxydulcioic acid (18), oleanolic acid (19), and ursolic acid (20). These results support the use of M. albicans as a traditional medicine with antidiabetic properties and its potential as a source of PTP1B inhibitors.


Introduction
Type 2 diabetes is a metabolic syndrome caused by an unbalanced production and/or action of the hormone insulin.Pancreatic β-cells produce insulin as a response to increased blood glucose after digestion of carbohydrates.When insulin binds to the insulin receptor, several downstream signaling pathways become activated, leading to an uptake of glucose from the bloodstream into the cells [1][2][3].For type 2 diabetics, cells from liver and muscles display a decreased response to the insulin stimuli, i.e., they are insulin-resistant, and the resulting decreased glucose uptake leads to hyperglycemia.Insulin resistance is caused by malfunctions in signal transduction such as impaired activation of the insulin receptor tyrosine kinase or down-regulation of the insulin-stimulated phosphatidylinositol-3-OH kinase (PI3K) [4][5][6].Therefore, enzymes involved in this signaling pathway constitute promising targets for managing insulin resistance and for treatment of T2D patients.
Protein tyrosine kinase 1B (PTP1B) is an intracellular enzyme involved in the negative regulation of insulin signaling by dephosphorylating the activated insulin receptor, and consequently PTP1B inhibitors promote a prolonged effect of insulin [6].However, at present, there are no PTP1B inhibitors in clinical use, and PTP1B is therefore an intriguing target in the search for new therapeutic alternatives to the drugs currently available for managing T2D.
Miconia is a genus comprising approximately 1000 species of flowering plants belonging to the family Melastomataceae, and it is native to neo-tropical forests and widely spread in warm areas of the American continent [7].Miconia albicans is a shrub commonly found on the Brazilian savannah and in the Brazilian rain forests.Extracts obtained from various Miconia species have been reported to exhibit antifungal, analgesic, and antiprotozoal activity as well as α-amylase and α-glucosidase inhibitory activity.Although M. albicans is used as a traditional anti-inflammatory medicine in Brazil and for treatment of diabetes in Mexico, the pharmacological properties of this plant species have not yet been investigated in detail.Few studies have reported its antifungal and antibacterial activity as well as its protective effects towards doxorubicin-induced DNA damage, but the individual constituents responsible for these effects have not been identified.The only studies of the phytochemistry of M. albicans led to isolation of the triterpenes α-amyrin and β-amyrin, and the steroids β-sitosterol and stigmasterol [8][9][10][11].
Bioassay-guided fractionation is a widely used method for identification of bioactive constituents in crude plant extracts, but it is usually both laborious and time-consuming.Thus, the combined use of high-resolution inhibition profiling (HR-inhibition profiling) that pinpoints individual bioactive constituents and high-performance liquid chromatography-high-resolution mass spectrometry-solid-phase extraction-and nuclear magnetic resonance spectroscopy (HPLC-HRMS-SPE-NMR) that allows structural identification from analytical-scale HPLC analysis, can accelerate the search for bioactive constituents in complex plant extracts.HR-inhibition profiling/HPLC-HRMS-SPE-NMR have already been used for accelerated identification of α-glucosidase inhibitors [12,13], α-amylase inhibitors [14], PTP1B inhibitors [15], monoamine oxidase inhibitors [16], and antioxidants [17,18] directly from crude extracts of foods and herbal medicine.In this study, we report the PTP1B inhibitory activity of crude defatted ethyl acetate extract of M. albicans as well as the identification of several active polyphenolics and triterpenoids by the use of high-resolution PTP1B inhibition profiling combined with HPLC-HRMS-SPE-NMR.

Results
The crude defatted extract of M. albicans was found to possess high PTP1B inhibitory activity with an IC 50 value of 4.92 ± 0.31 µg/mL (as determined from the dose-response curve shown in Supplementary Material Figure S1), and it was therefore decided to identify some of the bioactive constituents responsible for this inhibitory activity.

High-Resolution PTP1B Inhibition Profiling and Identification of Active Compounds from Crude Extract of M. albicans
The crude extract was subjected to high-resolution PTP1B inhibition profiling, and the biochromatogram (Figure 1) displayed 12 distinct peaks corresponding to moderate to strong activity eluting between 32 and 62 min.In addition, two large humps with around 100% inhibition were observed in the retention ranges 64-75 min and 75-90 min.Initially, HPLC-HRMS-SPE-NMR analysis of crude extract was performed to identify the material eluted with HPLC peaks 1-12, which are correlated with PTP1B inhibitory activity.HRMS and NMR data are reported in Supplementary Material Table S1.The HRMS data and the inferred molecular formula of the compounds eluting with peaks 7 and 9-12 are given in Table S1 in Supplementary Data, but the amount and purity of these compounds did not allow further structural identification.
Dereplication of the hump from 64 to 75 min correlated with PTP1B inhibitory activity, and based on LC-ESI-HRMS/MS data it showed losses of [M − H − 44] − and [M − H − 62] − , corresponding to a loss of CO2 and CO2 + H2O, which is common for lipids (fatty acids).This suggested that this area contained large amounts of structurally related lipids.Two previous studies have shown that some classes of lipids possess PTP1B inhibitory activity [24,25], which could support the activity observed in the biochromatogram, but the amount and purity of material obtained from this region did not allow structural identification of individual metabolites.
The material eluting with peak 8 showed a molecular ion with m/z 417.The HRMS data and the inferred molecular formula of the compounds eluting with peaks 7 and 9-12 are given in Table S1 in Supplementary Data, but the amount and purity of these compounds did not allow further structural identification.
Dereplication of the hump from 64 to 75 min correlated with PTP1B inhibitory activity, and based on LC-ESI-HRMS/MS data it showed losses of [M − H − 44] − and [M − H − 62] − , corresponding to a loss of CO 2 and CO 2 + H 2 O, which is common for lipids (fatty acids).This suggested that this area contained large amounts of structurally related lipids.Two previous studies have shown that some classes of lipids possess PTP1B inhibitory activity [24,25], which could support the activity observed in the biochromatogram, but the amount and purity of material obtained from this region did not allow structural identification of individual metabolites.
HRMS-based dereplication of peaks eluting from 75 to 90 min suggested the presence of several structurally related pentacyclic triterpenes and saponins.There is a growing interest in naturally occurring triterpenoids because they exhibit a wide spectrum of biological activities including antibacterial, fungicidal, antiviral, cytotoxic, anticancer, and cardiovascular activity.In addition to this, some triterpenoids inhibit PTP1B, a key regulatory enzyme for regulation of blood glucose [26][27][28][29][30].The fraction eluting from 75 to 90 min was therefore isolated by preparative-scale HPLC (hereafter called fraction F1) to allow further analysis using high-resolution PTP1B inhibition profiling combined with HPLC-HRMS-SPE-NMR.

High-Resolution PTP1B Inhibition Profiling and Identification of Active Compounds from F1
High-resolution PTP1B inhibition profiling of the triterpenoid-containing fraction F1 obtained by preparative-scale HPLC provided the biochromatogram shown in Figure 3.This displayed several HPLC peaks correlated with moderate to strong PTP1B-inhibitory activity-varying from 33% to 95% inhibition.Initial dereplication of the chromatographic peaks eluted from fraction F1 indicated, based on HRMS and UV data, the presence of several structural isomers of triterpenoids.This fraction was therefore further analyzed by HPLC-HRMS-SPE-NMR to enable identification of PTP1B-inhibitory triterpenoids.Thus, peaks 13-20 were cumulatively trapped (from 10 consecutive separations) on Hysphere GP-resin cartridges based on threshold levels from the UV chromatogram (254 or 280 nm), total ion chromatogram or base peak chromatogram.After drying with nitrogen gas, the trapped metabolites were eluted into 1.7-mm NMR tubes, and 600 MHz spectra were acquired.occurring triterpenoids because they exhibit a wide spectrum of biological activities including antibacterial, fungicidal, antiviral, cytotoxic, anticancer, and cardiovascular activity.In addition to this, some triterpenoids inhibit PTP1B, a key regulatory enzyme for regulation of blood glucose [26][27][28][29][30].The fraction eluting from 75 to 90 min was therefore isolated by preparative-scale HPLC (hereafter called fraction F1) to allow further analysis using high-resolution PTP1B inhibition profiling combined with HPLC-HRMS-SPE-NMR.

High-Resolution PTP1B Inhibition Profiling and Identification of Active Compounds from F1
High-resolution PTP1B inhibition profiling of the triterpenoid-containing fraction F1 obtained by preparative-scale HPLC provided the biochromatogram shown in Figure 3.This displayed several HPLC peaks correlated with moderate to strong PTP1B-inhibitory activity-varying from 33% to 95% inhibition.Initial dereplication of the chromatographic peaks eluted from fraction F1 indicated, based on HRMS and UV data, the presence of several structural isomers of triterpenoids.This fraction was therefore further analyzed by HPLC-HRMS-SPE-NMR to enable identification of PTP1B-inhibitory triterpenoids.Thus, peaks 13-20 were cumulatively trapped (from 10 consecutive separations) on Hysphere GP-resin cartridges based on threshold levels from the UV chromatogram (254 or 280 nm), total ion chromatogram or base peak chromatogram.After drying with nitrogen gas, the trapped metabolites were eluted into 1.7-mm NMR tubes, and 600 MHz spectra were acquired.Comparison with 1 H NMR data from literature confirmed 13 to be maslinic acid [31] (Figure 4).
The compound obtained by repeated trapping of peak 15 showed also a molecular ion with m/z 473.3618 [M + H] + suggesting the same molecular formula C 30 H 48 O 4 (∆M = 1.5 ppm).The 1 H NMR spectrum displayed signals similar to those for compound 14, but with H-3 appearing at δ 3.06 as a doublet of doublet with J H3ax,H2ax = 12 Hz and J H3ax,H2eq = 4.1 Hz.Analysis of COSY, NOESY, HSQC and HMBC spectra as well as comparison with 1 H and 13 C NMR data from literature of the related methyl ester [32] allowed identification of 15 as sumaresinolic acid (Figure 4).NMR data from literature of the related methyl ester [32] allowed identification of 14 as 3-episumaresinolic acid (Figure 4).The compound obtained by repeated trapping of peak 15 showed also a molecular ion with m/z 473.3618 [M + H] + suggesting the same molecular formula C30H48O4 (ΔM = 1.5 ppm).The 1 H NMR spectrum displayed signals similar to those for compound 14, but with H-3 appearing at δ 3.06 as a doublet of doublet with JH3ax,H2ax = 12 Hz and JH3ax,H2eq = 4.1 Hz.Analysis of COSY, NOESY, HSQC and HMBC spectra as well as comparison with 1 H and 13 C NMR data from literature of the related methyl ester [32] allowed identification of 15 as sumaresinolic acid (Figure 4).for both 16 and 17 (∆M = 2.7 and 0.7 ppm, respectively).The 1 H NMR spectra resembled that of 13 but revealed additional signals for an AA'XX' aromatic system as well as two cis-positioned olefinic protons (δ 6.88 and δ 5.85, 1H each, 12.8 Hz) for 16 and two trans-positioned olefinic protons (δ 7.64 and δ 6.81, 1H each, 15.9 Hz) for 17.This indicated 16 and 17 to be cisand trans-coumaroyl esters of 13.A clear downfield shift of H-3 from δ 2.91 in 13 to δ 4.63 in 16 and δ 4.64 in 17, shows the coumaroyl unit to be attached at C-3 in both compounds.By comparison of these 1 H NMR data with literature data the structure of 16 and 17 were assigned as 3-O-cis-p-coumaroyl-maslinic acid and 3-O-trans-p-coumaroyl-maslinic acid, respectively [33].The compound obtained by repeated trapping of peak 18 showed also a molecular ion of m/z 619.3974 [M + H] + suggesting the same molecular formula C 39 H 54 O 6 (∆M = 3.1 ppm) as that of 16 and 17.The 1 H NMR spectrum showed signals characteristic of an aromatic AA'XX' system and two trans-coupled olefinic protons (δ 7.63 and 6.38, 1H each, 15.8 Hz) inferring the presence of a trans-coumaroyl unit.Contrary to the seven methyl singlets observed for 13-17, the 1 H NMR spectrum of 18 showed six methyl singlets and a methyl doublet attributed to a methyl group at C-19 (δ 0.95, 3H, d, 6.1 Hz, CH 3 -29).Further analysis based on 1D and 2D NMR and comparison with literature data established that 18 was 3-O-trans-p-coumaroyl-2α-hydroxydulcioic acid [34] (Figure 4).
The compound obtained by repeated trapping of peak 19 showed a molecular ion with m/z 457.3668 [M + H] + suggesting the molecular formula C 30 H 48 O 3 (∆M = 1.8 ppm).Similar to the spectra of 13-15, the 1 H NMR spectrum of 19 displayed signals characteristic of an olean-12-oic acid, and comparison with NMR data from literature allowed identification of 19 as oleanolic acid [35] (Figure 4).The compounds obtained by repeated trapping of peak 20 showed a molecular ion with m/z 457.3672 [M + H] + suggesting the molecular formula C 30 H 48 O 3 (∆M = 0.9 ppm).The 1 HNMR spectrum of 20, showed signals characteristic of an ursan-12-oic acid, and comparison with NMR data from literature allowed identification of 20 as ursolic acid [36].NMR data of 13-20 are given in Supplementary Material (Table S1).
Dose-response curves of 13-20 are shown in Supplementary Material (Figure S2), and the derived IC 50 values are presented in Table 1.Maslinic acid has previously been reported as a bioactive pentacyclic triterpene acid, with anti-tumor, antioxidant, anti-HIV, and antimicrobial activity.In this study 13 was identified as a PTP1B inhibitor with an IC 50 -value of 3.21 ± 0.8 µM, which is in agreement with the previously reported IC 50 -value of 5.93 µM [29].Moreover, studies suggest that PTP1B inhibition of this type of pentacyclic triterpenes is affected by the configuration and substituents at C-2 and C-3 [30].Although the biochromatogram does not give quantitative pharmacological responses, Figure 3 shows that compounds 16 (IC 50 : 0.46 ± 0.07 µM), 17 (IC 50 : 1.08 ± 0.32 µM), and 18 (IC 50 : 1.6 ± 0.50 µM) with a coumaroyl unit at C-3 are correlated with higher PTP1B inhibitory activity as compared to that of 13.According to Qiu and colleagues [30], this structural feature may not only increase the potency but also improve the selectivity of these compounds for PTP1B.PTP1B belongs to the protein tyrosine kinases (PTPs), a family comprising approximately 100 enzymes that are involved in vital processes in the human body, and it has therefore been a challenge to identify selective PT1B inhibitors that might be developed into new medicines [6,30,37].The present work demonstrates that M. albicans constitute a rich source of new potent 3-coumaroyl analogues of maslinic acid, and presents the PTP1B inhibitory activity of 14-18 for the first time.However, further studies are needed to investigate their potential as selective PTP1B inhibitors.

Preparation of Crude Ethyl Acetate Extracts of Leaves of M. albicans
Leaves of Miconia albicans were collected in the National Park of Chapada dos Veadeiros (Alto Paraíso de Goiás, GO, Brazil) in February 2016, identified by botanist Dr. Marcos José da Silva (Universidade Federal de Goiás), and a voucher specimen was deposited at the herbarium of Universidade Federal de Goiás with accession number UFG 7163.The leaves were dried and powdered, and a batch (139.2 g) of this plant material was extracted twice with ethyl acetate (1.0 L) for 2 h under sonication.The extract was filtered and concentrated under reduced pressure to afford a crude extract (9.0 g).Fatty acids and chlorophylls were removed by liquid-liquid partitioning between 90% (v/v) methanol (0.1 L) and petroleum ether (0.9 L), yielding a crude defatted extract (4.0 g) after evaporation of the methanol fraction under reduced pressure.

Microplate-Based PTP1B Inhibition Assay
PTP1B inhibition assaying of the defatted extract, fraction F1 (from preparative HPLC: see Section 3.5) and microfractions (obtained as described below in Section 3.4) were performed according to the microplate-based method previously described [38].In brief, all solutions were prepared in a buffer of 50 mM Tris and 50 mM Bis-Tris containing 100 mM NaCL and adjusted to pH 7. To each well containing microfractions or dilution series of pure compounds were added 18 µL of DMSO and 52 µL of 3.4 mM EDTA in Bis-Tris buffer (to give final well concentrations of 10% DMSO and 1 mM EDTA) as well as 60 µL of a Bis-Tris buffer solution containing 1.5 mM pNPP and 6 mM DTT in phosphate buffer (to give final well concentrations of 0.5 mM pNPP and 2 mM DTT).The microplate was incubated for 10 min at 25 • C, and the reaction was started by addition of 0.001 µg/µL PTP1B stock solution (50 µL) in phosphate buffer (to give a final concentration of 0.05 µg PTP1B per well).The reaction was monitored at 405 nm every 30 s for 10 min to yield the enzyme activity (expressed as cleavage rate AU/s) by using a Thermo Scientific Multiskan FC microplate photometer (Thermo Scientific, Waltham, MA, USA).For each microplate, blank samples (negative controls) containing buffer, enzyme, EDTA, DTT, and pNPP were measured in triplicate, and for the dilution series, RK682 was included as a reference inhibitory compound (i.e., positive control).The percentage inhibition of PTP1B was calculated according to the following equation: The high-resolution PTP1B biochromatograms (high-resolution PTP1B inhibition profiles) were constructed by plotting the inhibition values against the chromatographic retention time.Dose-response curves and IC 50 values for dilution series of the crude extract and pure compounds were obtained using GraFit software, version 5.0.11(Erithacus Software Limited, West Sussex, UK).For PTP1B IC 50 values, the crude extract and isolated compounds were assessed by using the standard assay conditions described above.Nine dilutions were assessed in triplicates and percentage inhibition calculated as mean ± standard deviation in Microsoft Excel according to the above equation.The results were imported and used for assessing dose-response curves and IC 50 values in GraFit software, version 5.0.11(Erithacus Software Limited).Data were fitted to the equation where min is the background, max-min is the y-range, x is the concentration, and slope is the Hill slope.Results are reported as IC 50 values ± standard error.

Isolation of PTP1B Inhibitors from the Extract of M. albicans by Analytical-Scale HPLC
The early-eluting PTP1B inhibitors (observed from 32 to 60 min) were isolated directly from the crude extract using HPLC system 1 operated with the same column, solvent system and flow rate as described in Section 3.4.Twenty successive injections of 20 µL of a solution of the crude extract (equivalent to 400 µg of crude extract each) were separated using the stepwise linear elution gradient: 0 min, 0% B; 50 min, 20% B; 60 min, 30% B; 65 min, 50% B; 85 min, 100% B; 105 min, 100% B. The active compounds observed in the inhibition profile were collected, dried and submitted to analysis by NMR spectroscopy.

HPLC-HRMS-SPE-NMR Analyses of Triterpenoids from F1
HPLC-HRMS-SPE-NMR analyses of fraction F1 were performed on a system comprising an Agilent 1260 HPLC (Santa Clara, CA, USA), a Bruker micrOTOF-Q II mass spectrometer (Bruker Daltonik, Bremen, Germany), a Knauer Smartline K120 pump (Knauer, Berlin, Germany), a Spark Holland Prospekt-2 SPE unit (Spark Holland, Emmen, The Netherlands), a Gilson 215 liquid handler (Gilson, Middleton, WI, USA), and a Bruker Avance III 600 MHz NMR spectrometer (Bruker Biospin, Rheinstetten, Germany).The Agilent HPLC system consisted of a G1329B autosampler, a G1311B quaternary pump with built-in degasser, a G1316A thermostatted column compartment, equipped with Phenomenex (Phenomenex Inc., Torrance, CA, USA) Luna C 18 (2) column (150 mm × 4.6 mm i.d., 3 µm particle size, 100 Å pore size), and a G1315D photodiode array detector, and the column, elution solvents, temperature, and flow rate were the same as described in Section 3.4.A T-piece after the HPLC column directed approximately 1% of the eluate to the micrOTOF-Q II mass spectrometer equipped with an electrospray ionization source.Spectra were acquired in positive ionization 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.To enable internal mass calibration, a solution of sodium formate clusters was automatically injected at the beginning of each run.The remaining 99% of the eluate was directed to the photodiode array detector and subsequently diluted with water at a flow rate of 1 mL/min delivered by the Knauer pump.Hysphere GP-resin SPE cartridges (10 × 2 mm i.d., Spark Holland, Emmen, The Netherlands) were preconditioned with 500 µL of acetonitrile and equilibrated with 500 µL of water.Selected chromatographic peaks were trapped from 10 consecutive chromatographic separations using threshold levels from the UV chromatogram (200 nm and 254 nm), the total ion chromatograms (TIC), or base peak chromatograms (BPC) to trigger trapping.The loaded SPE cartridges were dried with pressurized N 2 gas for 30 min and subsequently eluted with methanol-d 4 (30 µL) into 1.7-mm o.d.NMR tubes (Bruker Biospin, Rheinstetten, Germany) using the Gilson liquid handler equipped with a 1-mm needle.HPLC separations, mass spectrometry and analyte trapping on SPE cartridges were controlled by Hystar version 3.2 software (Bruker Daltonik, Bremen, Germany), whereas the elution processes were controlled by Prep Gilson ST version 1.2 software (Bruker Biospin, Rheinstetten, Germany).

NMR Experiments
NMR experiments were recorded at 300 K in methanol-d 4 and acquired on a Bruker Avance III 600 MHz NMR spectrometer ( 1 H operating frequency of 600.13 MHz) equipped with a Bruker SampleJet sample changer and a 1.7-mm cryogenically cooled inverse triple-resonance TCI probe (Bruker Biospin, Rheinstetten, Germany).The one-dimensional 1 H spectra were acquired with 30 • pulses with a 20 ppm spectral width, 2.72 s acquisition time, and 1.0 s relaxation delay, collecting 256 or 512 FIDS, each consisting of 64 k data points, and Fourier transformed to 256 k data points with a line broadening of 0.1 Hz.Two-dimensional homonuclear DQF-COSY and NOESY experiments were recorded using a gradient-based pulse sequence with a 20 ppm spectral width and 2 k × 512 data points (processed with forward linear prediction to 1 k data points).Multiplicity-edited HSQC spectra were acquired with 12 ppm spectral width for 1 H and 170 ppm for 13 C, 2 k × 256 data points, and 1.0 s relaxation delay.HMBC experiments were recorded with a 12 ppm spectral width for 1 H and 210 ppm for 13 C, 2 k × 192 data points and 1.0 s relaxation delay.

Figure 1 .
Figure 1.High-resolution PTP1B inhibition profiling of the crude ethyl acetate extract of M. albicans.Black trace: UV absorption at 254 nm; Green trace: PTP1B inhibition profiling.