Chemical secondary metabolite profiling of Bauhinia longifolia ethanolic leaves extracts

Industrial Crops & Products 132 (2019) 59–68

Contents lists available at ScienceDirect

Industrial Crops & Products journal homepage: www.elsevier.com/locate/indcrop

Chemical secondary metabolite profiling of Bauhinia longifolia ethanolic leaves extracts

T

Amanda J. Aquinoa, Thayana da C. Alvesb, Regina V. Oliveiraa, Antonio G. Ferreirab, ⁎ Quezia B. Cassa, a b

Núcleo de Pesquisa em Cromatografia (Separare), Chemistry Department of the Federal University of São Carlos, São Carlos, SP, Brazil Laboratory of Nuclear Magnetic Resonance, Chemistry Department of the Federal University of São Carlos, São Carlos, SP, Brazil

A R T I C LE I N FO

A B S T R A C T

Keywords: Bauhinia longifolia Liquid chromatography High resolution mass spectrometry Nuclear magnetic resonance LC-HRMS LC-SPE/NMR

The genus Bauhinia belongs to the family Fabaceae and comprises about 18,000 known species in the world. In Brazil, approximately more than 60 native species of Bauhinia are found. Bauhinia leaves tea infusion or other preparations have been widely used in the Brazilian popular medicine for treatment of several illnesses. Therefore, this work aims to a better understanding of Bauhinia longifolia to provide a valuable database for its quality control, pharmacological and toxicological studies. For the chemical analysis, a liquid chromatography-high resolution quadrupole-time of flight mass spectrometer method was applied to evaluate the chemical profile of Bauhinia longifolia ethanolic leaves extracts, allowing for the identification of 75 compounds comprising chemical classes of phenolic acids, proanthocyanidins, and O-glycosides flavonoids. Most polyphenols were identified for the first time in this specie. In addition, six compounds were isolated and fully identified by liquid chromatography hyphenate to solid phase extraction and nuclear magnetic resonance. Herein, myricitrin, juglanin, afzelin, and bauhiniastatin 2 are reported for the first time for Bauhinia longifolia. Bauhiniastatin 2 has been reported to display anticancer properties towards several cancer cell lines. The chemical profile herein described for the ethanolic leaves extract of Bauhinia longifolia provided essential information of this Brazilian native species.

1. Introduction The genus Bauhinia, known as "pata-de-vaca" (´cow’s paw’) or “unha-de-boi” (‘ox nail’), belongs to the family Fabaceae and comprises about 18,000 known species in the world. Fabaceae is the most common family found in tropical regions of Africa, Asia, Central, and South America, where they are mainly used as folk medicine. In Brazil, approximately more than 60 native species of Bauhinia are found (Silva and da Cechinel Filho (2002); Vaz, 2010). Bauhinia leaves tea infusion or other preparations have been widely used in the Brazilian popular medicine for treatment of several illnesses, especially diabetes (antiglycemiant) (Nogueira and Sabino, 2012). Given its high empirical use for medicinal purposes, added the high interest in research with this plant specie, the Bauhinia genus was included in the RENISUS - National Relation of Medicinal Plants of Interest to the Unified Heatlh System, Brazil, whose purpose is to foster research and development of monographs for quality control of herbal medicine (Brasil - Agência Saúde, 2009).

Phytochemical studies with the genus Bauhinia have identified several secondary metabolites, including terpenes (Duarte-Almeida et al., 2004), quinones (Duarte-Almeida et al., 2004), alkaloids (Mishra et al., 2013), oxepines (Kittakoop et al., 2004), and especially free and glycosylated flavonoids, such as kaempferol and quercetin derivatives (Ferreres et al., 2012; Xu et al., 2012). Moreover, studies have also revealed the potential of the genus Bauhinia as a source of phenols with antioxidant activity (de Sousa et al., 2004). Phenolic secondary metabolites are vital in defense responses and, therefore, they may play a key role in the pathophysiology of many diseases including cancer, diabetes, cardiovascular, and stroke by the management of oxidative stress (Crozier et al., 2009). Despite the chemical and medicinal interest of the genus, little is known about the phytochemistry and pharmacological potential of Bauhinia longifolia when compared to B. forticata and B. variegata (Duarte-Almeida et al., 2004; Farag et al., 2015; Ferreres et al., 2012; Silva et al., 2012). Bauhinia longifolia is native to Brazil and widely distributed in the Cerrado of Minas Gerais and in the Brazilian Atlantic Forest, Amazon,

⁎ Corresponding author at: Núcleo de Pesquisa em Cromatografia (Separare), Chemistry Department of the Federal University of São Carlos, São Carlos, 13565-905, SP, Brazil. E-mail address: [email protected] (Q.B. Cass).

https://doi.org/10.1016/j.indcrop.2019.01.040 Received 31 October 2018; Received in revised form 17 January 2019; Accepted 18 January 2019 0926-6690/ © 2019 Elsevier B.V. All rights reserved.

Industrial Crops & Products 132 (2019) 59–68

A.J. Aquino, et al.

(22.87 °S, 47.07 °W, 22.86 °S and 47.08 °W) by V. A. P. e Carvalho (Carvalho, 2011) and identified by A. S. F. Vaz, Ph.D, of the Botanical Garden Research Institute of Rio de Janeiro and deposited at the Federal University of São Carlos, São Carlos, SP, Brazil herbarium (SPSC) under numbers 8325 and 8330, respectivily. The collected samples have also been registred at SisGen Brazilian plataform of Genetic Heritage and Associated Traditional Knowledge under the number A99450 F. These vegetal samples were dried in a forced air circulation drying oven set at 40 °C for 7 days. For appropriated storage, the dried material was grinded to a fine powder (60 Mesh).

and Caatinga (“Consulta Pública do Herbário Virtual - Bauhinia longifolia,” 2018). The chemical composition of this specie was reported only in three studies, using maceration of either ethanolic or methanolic leaf extracts followed by isolation of their products (dos Santos et al., 2014; Ernestina et al., 2017), or using gas chromatography coupled to mass spectrometry (GC–MS) to analyze the leaves essential oils composition (Duarte-Almeida et al., 2004). In these studies, the compounds identified in the B. longifolia were terpenes (β-caryophyllene, D-germacrene, bicyclogermacrene, isospathulenol, Y-cadinene) and flavonoids (guaijaverin, quercetin, quercitrin, isoquercitrin, hyperin, and avicularin) that exhibited high antioxidant activity similar to those shown for compounds isolated from B. forticata, a specie widely studied and officially recognized as an antidiabetic herbal medicine (de Sousa et al., 2004; Di Stasi et al., 2002; Silva-López and Santos, 2015). Thus, detailed information on the compounds of B. longolilia is essential for further studies regarding pharmacological activity, toxicology, and quality control. In order to quickly identify compounds in complex mixtures, some comprehensive methods have been developed. In the rapidly growing field of plant metabolomics, liquid chromatography coupled to high resolution mass spectrometer (LC-HRMS) provide great alternatives for structural characterization because of the occurrence of fragment ions obtained from different collision energies used together with soft ionization techniques and additional high mass accuracy measurements for elemental composition, which can be combined with online free spectral compound libraries such as MASS BANK (http://www. massbank.eu/) for the identification of compounds (Allard et al., 2017; Demarque et al., 2016; Guan et al., 2011; Wu et al., 2013). Moreover, analysis of samples in both positive and negative ionization modes support the structural elucidation of unknown compounds due to different fragmentation patterns of protonated and deprotonated molecules. Nevertheless, all this information is still not sufficient for complete structural identification of new compounds because of almost identical fragmentation patterns of stereo and positional isomers. Nuclear magnetic resonance (NMR) is an outstanding analytical tool for the confirmation of complex structural molecules. The information obtained from both mass spectrometry (MS) and NMR analysis facilitates structure elucidation of isolated compounds. The hyphenation of liquid chromatography (LC), solid phase extraction (SPE) and nuclear magnetic resonance (LC-SPE/NMR) allows trapping and subsequent analyte-enrichment. The SPE approach enables for solvent exchange from the liquid chromatography mobile phase to deuterated NMR solvents. One (1D) and two dimensional (2D) NMR experiments are then feasible with small sample amounts without the use of traditional timeconsuming phytochemical procedures for sample extraction. To meet this end, the main objective of this work was to evaluate the metabolic chemical profile of B. longifolia leaf ethanolic extracts using advanced, accurate, and precise analytical techniques (LC-HRMS and LC-SPE/NMR) for chemical identification. This paper reports our findings that contribute to a better understanding of the above mentioned medicinal plant and provides a valuable database for quality control, pharmacological studies, and toxicological research.

2.3. Sample preparation Dried and ground samples (500 mg) were weighed and extracted with 5 ml of ethanol in conical tubes (50 ml) using an Ultraturrax® homogenizer (IKA®, modelo T18 basic) set at speed 6 for 5 min. After liquid extraction, the samples were centrifuged using a Jouan® centrifuge (model, BR4i) for 10 min at 10.000 x rpm and the supernatant (S1) was obtained. The S1 were evaporated using a Speed-Vac® (Savant®, model SPD131DDA) set at 45 °C overnight. The dried extract obtained from the S1 fraction was identified, weighed, and reconstituted to 50 mg/ml (D1) in (85:15; v/v), MeOH:H2O (v/v). The D1 was fractionated (cleaned up) using SPE to eliminate chlorophylls and other compounds of low polarity (Funari et al., 2012). C18 end-capped (100 mg) SPE cartridges (Varian®) with a volume capacity of 1 ml and a 20-port vacuum manifold system coupled to a Tecnal® vacuum pump (model TE-058) were used. Initially, the SPE cartridge was activated with 2 ml of methanol (MeOH), equilibrated with 2 ml of MeOH:H2O (85:15; v/v), and then loaded with 500 μL of the D1. The eluate was collected, dried in a Speed-Vac® set at 45 °C overnight and resuspended in MeOH:H2O (85:15; v/v) to yield 30 mg/ ml for LC-HRMS and 77 mg/ml LC-SPE/NMR analyses. Prior to analysis, the samples were centrifuged at 9300 x g for 10 min. 2.4. LC-HRMS analysis The ultra-high performance liquid chromatography (UHPLC) system (model Nexera®, Shimadzu) consisted of two quaternary LC-30AD pumps, a DGU-20 A5R degasser, a SIL-30AC autoinjector, a SPD-M30 A diode arrangement detector, a CTO-20 AC furnace, a six-column selector valve, and a CBM 20 A interface. The LabSolutions® workstation software was used to control all modules and for data processing. Fragmentation studies were obtained by chromatographic separation with a Biphenyl Kinetex® column (10 cm x 2.7 μm x 90 Å) Phenomenex® equipped with a guard-column and employing a gradient elution using 100 μM of formic acid in water (solvent A) and methanol (solvent B) as mobile phase at a flow rate of 0.7 ml/min and temperature set at 50 °C. The total run time was 30 min using the following multistep gradient: 0 min, 5% B; 0–20 min, 5–70% B; 20.1–25 min, 100% B for column cleaning and a conditioning cycle time of 5 min with the same initial conditions of 5% B. The injection volume was 0.5 μl (sample concentration: 30 mg/ml). The separated compounds were monitored with a high-resolution mass spectrometer (HRMS) containing a quadrupole time-of-flight mass analyzer (QTOF). The MS analysis was performed using an Impact HD QTOF™ mass spectrometer (Bruker Daltonics, Bremen, Germany) equipped with an electrospray (ESI) interface operating in negative ion mode. A splitter was used for coupling of LC with the MS detector, and the flow entering at the QTOF detector was set at 118 μl/min. The optimal parameters were set as follows: capillary voltage, 4000 V; end plate offset, 500 V; nebulizer, 1 bar; dry heater temperature, 250 °C; dry gas flow, 8 l/min; collision cell energy, 5 eV, and fullMS scan range, m/z 50 – 1500. The mass spectrometer was programmed to perform acquisition in auto MS/MS mode (number of precursors 5) in experiments with different collision energy of 15, 20, 25, 30, 35, 40, 45, or 50 eV for all m/z range analyzed.

2. Materials and methods 2.1. Chemicals Methanol (HPLC grade) was used in all experiments and purchased from J.T. Baker (Philipsburg, USA). Formic acid (p.a. grade) was acquired from Fluka (Buchs, Switzerland). Water was purified in a Milli-Q system (Millipore, São Paulo, Brazil). 2.2. Plant material Leaves of B. longifolia were obtained from the Agronomic Institute of Campinas (IAC), SP, Brazil. The plants were collect in May 2010 60

Industrial Crops & Products 132 (2019) 59–68

A.J. Aquino, et al.

settings: 70 K number of scans; acquisition time 0.47 s; a relaxation delay of 0.1 s; spectral width of 230 ppm, and 32 K data points. The 2D COSY experiment (Bruker cosygpprqf pulse sequence) had the following setting: 16 number of scans; spectral width of 18 ppm in both dimensions; a 90° pulse of 7.25 μs; a relaxation delay of 1 s; 4 K data points in F2 and 256 in F1. The 2D HSQC (Bruker hsqcetgpprsisp2.2 pulse sequence) was obtained in the echo-antiecho phase sensitive mode with the following setting: 32 scans; 18 and 238 ppm spectral widths in the hydrogen and carbon dimensions, respectively, 7.25 μs 90° pulse; 1H-13C coupling constant was 145 Hz with 4 K data points in F2 and 256 in F1. The 2D HMQC had the following setting: 128 scans; 18 and 238 ppm spectral widths in the hydrogen and carbon dimensions, respectively, 7.25 μs 90° pulse with 4 K data points in F2 and 256 in F1, and optimized for nJ 13C-1H of 8 Hz. Data collection and processing were carried out using the Topspin 3.1 software (Bruker Germany). Each 1D and 2D spectra was processed manually adjusting phase, base-line and line broadening factor of 0.3 Hz, if required. For the 2D spectra, 4 K in F1 domain and 1 K in F1 were used. The chemical shift assignments were referenced to deuterated methanol at 3.30 ppm. Structural characterization was established based on NMR spectroscopy analyses, MS/MS fragmentation profile, accurate mass measurements, and comparison with literature data for Bauhinia genus.

In order to perform the LC-SPE experiments using a high-performance liquid chromatography (HPLC) system (topic 2.5), the same chromatographic conditions were used as with the UHPLC-HRMS system. For that, the chromatographic separations were carried out using a Biphenyl Restek® analytical column (150 mm × 4.6 mm i.d., 5 μm) coupled to a guard-column under the same mobile phase used with the Kinetex® collumn, but now at 1 ml/min. The total run time was 80 min using the following multistep gradient: 0 min, 5% B; 0–60 min, 5–80% B; 60–70 min, 100% B for column cleaning, and a conditioning cycle time of 10 min with the same initial conditions of 5% B, using 100 μM of formic acid in water (solvent A) and methanol (solvent B). The injection volume was 1 μl (sample concentration: 77 mg/ml) of the ethanolic extract prepared as described in Section 2.3. A splitter was used for coupling of the LC system with the MS detector, and the flow arriving at the QTOF detector was set at 330 μl/min. The optimal parameters were set as follows: capillary voltage, 4000 V; end plate offset, 500 V; nebulizer, 3 bar; dry heater temperature, 300 °C; dry gas flow, 12 l/min; collision cell energy, 5 eV, and fullMS scan range, m/z 50 – 1200. The mass spectrometer was programmed to perform acquisition in auto MS/MS mode using five precursor ions for fragmentation. The MS and MS/MS data were processed through Data Analysis 4.0 software (Bruker Daltonik), which provided a list of possible elemental formulas by using Bruker Smart Formula. The confirmation of elemental compositions was established with less than 5 ppm mass accuracy. External mass-spectrometer calibration was performed with sodium formate clusters (5 mM sodium formate in water/2-propanol 1/1 (v/v)) in quadratic high-precision calibration (HPC) regression mode. The calibration solution was injected at the end of the analytical run and all the spectra were calibrated prior to compound identification.

3. Results and discussion 3.1. Chromatographic method development and data processing Chromatographic conditions for the UHPLC tandem MS fragmentation studies were optimized using experimental design to determine the stationary phase, pH, organic modifier, and time of analysis required for efficient separation (Aquino, 2018). In addition, to adjust the ionization conditions, in the negative ionization mode, formic acid at different concentrations (100 μM, 200 μM, 2 mM, 10 mM, and 20 mM) were added to water (A) and methanol (B) mobile phase. The higest MS signal was obtained at 100 μM formic acid (FA). Wu et al., 2004 described similar results at lower concentrations of formic acid (0.1–10 μM) for phenolic compounds in herbal medicines. Fig. 1a shows the chromatographic profile obtained in the optimized LC-HRMS conditions for the ethanolic extract of leaves of B. longifolia in negative ion mode. Fig. 1b display the chromatographic profile with ultraviolet (UV) detection at 254 nm. For the experiments by LC-SPE/NMR, the chromatographic conditions were adjusted to a HPLC system. To meet this end, another biphenyl column (F < 10) was used (Biphenyl Restek®). Moreover, it was observed that both columns [Biphenyl Restek® (150 mm × 4.6 mm i.d., 5 μm) vs. Biphenyl Kinetex® (10 cm x 2.7 μm x 90 Å)] maintained the same elution order for the classes of compounds herein identified, Fig. 2. The identified compounds by LC-HRMS and their proposed fragment ions are shown in Table 1. The chromatographic analysis demonstrated sufficient efficiency to separate isomeric forms of several compounds present in the ethanolic leaf extract. Phenolic acids were detected at the beginning of the chromatogram up to 4.10 min, condensed tannins (proanthocyanidins) and their derivatives eluted from 1.4 to 9.2 min, flavonoids eluted from 6.8 to 12.2 min, and bauhiastatin 2 was detected at 14.6 min (Fig. 1A). The attributions of the inferred compounds (Table 1) were feasible due to comparison of exact masses, MS/MS fragmentations pattern with those spectra found in the literature, or already deposited in spectral libraries online (http://mona.fiehnlab.ucdavis.edu/; http://www. massbank.eu/; https://metlin.scripps.edu) and acquired in equipments with ESI sources and experiments type CID (collision inducted dissociation). All molecular formulas of the identified compounds were calculated using accurate mass considering a mass errors lower than 5 ppm. The identified compounds were searched in the literature to verify if they were present in the Bauhinia ssp and consulted in the

2.5. LC-SPE/NMR analysis The LC-SPE/NMR analysis were carried under the previously described conditions for the Biphenyl Restek® (150 mm × 4.6 mm i.d., 5 μm) analytical column (Section 2.4) but now with an HPLC Agilent system (1200 series, Agilent GmbH) equipped with a quaternary pump (G1311 A), a degasser (G1322 A), a variable wavelength diode array detector (G1315D), an autosampler (G1329 A), and an automatic cartridge exchanger (Bruker Biospin GmbH). The LC system was controlled through HyStar 2.3 software (Bruker). A Knauer (K120 Knauer Smartline Pump Control 100, Bruker Daltonik GmbH©, V01.11) makeup pump diluted the post-column flow with water (1 ml/min) before the peaks were loaded on to the SPE cartridges (Prospekt II SPE unit). 30 consecutive chromatographic runs were performed using injection volumes of 1 μl from a sample of 77 mg/ml and a flow rate of 1.0 ml/min. The compounds were extracted using HySphere Resin GP cartridges (10 mm x 2 mm i.d., 10 μm spherical polydivinylbenzene stationay phase) placed in an automatic cartridge exchange unit. Afterwards, the cartridges were dried using nitrogen for 30 min to remove any residual solvent. Deuterated methanol-d4 (99.8%) (250 μl) was used to elute the retained compounds from the SPE cartridges directly into NMR tubes (Bruker™, 3 mm x 70 mm). The 1D (1H and 13C) and 2D (correlation spectroscopy (COSY), heteronuclear single quantum coherence spectroscopy (HSQC), and heteronuclear multiple quantum coherence spectroscopy (HMBC)) NMR spectra of the isolated compounds were acquired at 299 K employed a Bruker Avance III spectrometer (14.1 T) equipped with a triple resonance inverse of 5 mm (1H/13C/15N) with z-field gradient cryoprobe. The 1D 1H spectrum (Bruker noesypr1d pulse sequence) had the following settings: 16 number of scans; acquisition time 3 s; a relaxation delay of 3 s; a 90° pulse of 7.25 μs; presaturation power of 51.97 dB; spectral width of 18 ppm; mixing time of 0.01 s, and 65 K data points. The 1D 13C spectrum (Bruker zgpg30 pulse sequence) had the following 61

Industrial Crops & Products 132 (2019) 59–68

A.J. Aquino, et al.

Fig. 1. Results of the ethanolic extract of Bauhinia longifolia obtained by UHPLC-HRMS. (A) base peak ion chromatogram in negative ionization mode (-All/MS) and (B) UV chromatogram at 254 nm. Chromatographic conditions: Biphenyl kinetex® column, flow rate at 0.7 ml/min and, temperature set at 50 °C. Gradient elution: 5–80% Methanol.

like gentisic and protocatechuic acid (Kakazu and Horai, 2016; Matsuda et al., 2016a), previously reported for the genus Bauhinia and reported in the Fabaceae family (Compaoré et al., 2012; HMDB0000152; HMDB0001856; Nageshwar et al., 1986). Compounds 3 and 4 (m/z 315) had fragment ions at m/z 152.0114 and 153.0189 due to loss of a hexose moiety [M−H-162.053]−. Thus, these substances were tentatively identified as dihydroxybenzoic acid-hexoside (I) and dihydroxybenzoic acid-hexoside (II). The hydroxybenzoic acid isomers (7, 11) are found in different Bauhinia species (Sashidhara et al., 2013; Shang et al., 2006). Using our experimental conditions, these two compounds eluted at 9.4 and 13.8 min and were characterized due to the presence of a fragment ion at m/z 93.034 (Matsuda et al., 2016b) via loss of a carbon dioxide [M−H-44]− from the carboxylic acid moiety. In addition, the deprotonated molecular ion [M−H]− m/z 299.0771 yielding MS2 fragment ions at m/z 137.0243 (loss of a hexosyl moiety) and m/z 93.0343 (loss of CO2) for the molecular formula (C13H15O8), suggested that compound 9 could be assigned as a hydroxybenzoic acid hexoside. Neutral loss of CO2 [M−H-43.99]− was also observed for peaks 1, 2, 10, 13, 14, and 16. Compound 1 was identified as galloylhexose and showed a precursor ion [M−H]− at m/z 331.0670; compound 2 was assigned as gallic acid and showed a precursor ion [M−H]− at m/z 169.0141; compound 10 was identified as caffeic acid (Matsuda, 2016) due to the precursor ion [M−H]− at m/z 179.0352; compound 13 was assigned as coumaric acid-hexoside with [M−H]− at m/z 325.0927; compound 14 was identified as vanillic acid (Beisken et al., 2014) and presented an ion [M−H]− at m/z 167.0349; compound 16 was assigned as 2-coumaric acid (Matsuda et al., 2016c) and showed an ion [M−H]− at m/z 163.0402. Compounds 1 and 13 showed MS2 with loss of 162.053 typical of hexose and these were characterized as galloyl hexose and coumaric acid-hexoside, respectively. All these compounds have been previously described in Bauhinia ssp (Farag et al., 2015; Liu et al., 2016; Villavicencio et al., 2018).

Fig. 2. Base peak ion chromatogram in negative ionization mode (-All/MS) of the ethanolic extract of Bauhinia longifolia obtained by LC-HRMS when using the HPLC analytical conditions (section 2.4).

HMDB database (http://www.hmdb.ca/) to know if they were present in the plant kingdom and/or in the Fabaceae family. Complete mass spectra of the molecular characteristic are presented in the supplementary material 1. In addition, six compounds (40, 45, 54, 63, 68, and 75) were further isolated by liquid chromatography coupled to solidphase extraction (LC-SPE) and identified by NMR and HRMS data.

3.2. Compounds identified by LC-HRMS 3.2.1. Phenolic acids and derivatives (1-16) In the present work, 16 phenolic acids and phenolic acid derivatives (1–16) were infered (Table 1). Compound 15 was identified as syringic acid with a deprotonated molecule [M−H]− at m/z 197.0455 while compounds 8 and 12 were identified as syringic acid derivatives with [M−H]− at m/z 359.0984 and 359.0978, respectively. The MS2 fragmentation pattern of compounds 8 and 12 showed a fragment ion at m/ z 197 due to the loss of a hexosyl moiety (-162.053 Da; [syringic acidH]−) and a radical anion at m/z 182 by losing a radical methyl group (•CH3) from the precursor ion m/z 197, which coincided with previous report (Kakazu and Horai, 2011). Therefore, it was possible to assign compounds 8 and 12 as syringic acids hexoside. Moreover, these compounds have been already described in Bauhinia blakean (Liu et al., 2016). Compounds 5 and 6 have been proposed as dihydroxybenzoic acid isomers (I and II) because of their matching elemental composition, accurate mass and fragmentation patterns yielding fragment ions at m/z 109 [M−H–44]− (loss of CO2) and m/z 108.0215 [M−H-44-H]− (loss of HCO2) which are characteristic for dihydroxybenzoic acid isomers

3.2.2. Protoanthocyanidins (17-33) The monomeric flavan-3-ols gallocatechin and catechin (19 and 25) were identified by comparing their precursor and fragment ions obtained in negative ionization mode with data available from QTOF spectra (Boettcher, 2016; DB03823, n.d.). Compound 25 (catechin – [M−H]− m/z 289.0720) showed the following fragment ions at m/z 245.0818 [M−H-CO2]− due to loss of one CO2, m/z 221.0818, 203.0711, 205.0505, 151.0402 (origined from Retro-Diels-Alder fragmentation). Gallocatechin (19) was detected with a precursor ion [M−H]− at m/z 305.0670 and generated MS2 fragment ions at m/z 261.0768 [M−H-CO2]−, m/z 237.0769, 221.0452, 219.0662, and 167.0349 (originating from Retro-Diels-Alder fragmentation). Both compounds (catechin and gallocatechin) furnished the m/z 125.024 62

Galloyl hexose Gallic acid Dihydroxybenzoic acid-hexoside(I) Dihydroxybenzoic acid-hexoside(II) Dihydroxybenzoic acid(I) Dihydroxybenzoic acid (II) Hydroxybenzoic acid (I) Syringic acid-hexoside I Hydroxyc benzoic acid-hexoside Caffeic acid Hydroxybenzoic acid (II) Syringic acid-hexoside (II) Coumaric acid-hexoside Vanillic acid Syringic acid

Coumaric acid (epi)Gallocatechin-(epi)Gallocatechin (epi)Gallocatechin-(epi)Catechin (epi)Gallocatechin

(epi)Catechin-(epi)Gallocatechin (epi)Catechin-(epi)Gallocatechingallic acid (epi)Gallocatechin-(epi)Catechingalloyl (epi)Catechin-(epi)Gallocatechingalloyl (epi)Catechin-(epi)Catechin

Catechin

(epi)Gallocatechin-(epi)Catechingalloyl Gallocatechin-galloyl (epi)Afzelechin-(epi)Gallocatechingalloyl (epi)Catechin-(epi)Catechin-galloyl

(epi)Afzelechin-(epi)Catechin (I)

Catechin gallate (epi)Catechin-(epi)Catechin

(epi)Afzelechin-(epi)Catechin (II)

Kaempferol-hexose-deoxyhexose Kaempferol-deoxyhexosedeoxyhexose-hexose Kaempferol-308 (I) Kaempferol-308 (II) Kaempferol-deoxyhexose Kaempferol-278 (I)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

16 17 18 19

20 21

63

25

26

30

31 32

33

34 35

36 37 38 39

29

27 28

24

23

22

Compound

Peak Number

26.80 27.20 29.20 31.40

21.90 25.60

29.50

23.40 25.70

20.00

19.60

17.60 18.50

17.10

17.10

16.40

16.20

13.00

12.70 12.80

19.10 7.30 11.00 11.20

3.50 3.60 4.60 5.40 5.60 6.20 9.40 10.10 11.40 13.20 13.80 13.60 13.80 14.50 17.80

HPLC Rt(min)a

Table 1 LC-HRMS data of the identified compounds (continued).

8.20 8.40 8.80 9.80

6.80 7.90

9.20

7.00 7.90

5.70

6.00

4.80 5.60

5.00

4.30

4.50

4.70

3.50

3.20 3.40

593.1501 593.1510 447.0926 563.1403

593.1502 739.2085

561.1396

441.0828 577.1346

561.1407

729.1470

457.0775 729.1462

745.1408

289.0720

577.1354

745.1409

745.1413

593.1303 761.1354

163.0402 609.1259 593.1302 305.0670

331.0670 169.0141 315.0721 315.0723 153.0191 153.0191 137.0244 359.0984 299.0771 179.0352 137.0243 359.0978 325.0927 167.0349 197.0455

– 0.60 0.50 – 0.90 – 1.70 1.70 1.70 2.40 – 3.30 2.60 2.80 3.80 4.10 1.40 2.60 2.40

m/z Experimental [M-H]−

UHPLC Rt(min)a

C9H7O3 C30H25O14 C30H25O13 C15H13O7 C30H25O13 C37H29O18

−0.9 −1.5 −0.1 −1.0 −0.4 0.7

C15H13O6

−0.9

C30H25O11 C22H17O10 C30H25O12

−0.9 −0.2 1.0

1.8 0.3 1.5 0.6

1.6 0.8

30 35 35 35 35 35

C27H29O15 C27H29O15 C21H19O11 C26H27O14

35

35 20

20

20

20 35

20

20

C27H29O15 C33H39O19

C30H25O11

C37H29O16

−1.3

1.2

C22H17O11 C37H29O16

0.4 −0.1

C37H29O17

C30H25O12

−0.5

0.2

20

C37H29O17

0.1 20

35

35 20

35 20 20 20

35 20 35 35 35 35 35 25 35 35 35 35 35 20 20

Collision energy (eV)c

C37H29O17

−0.4

C13H15O10 C7H5O5 C13H15O9 C13H15O9 C7H5O4 C7H5O4 C7H5O3 C15H19O10 C13H15O8 C9H7O4 C7H5O3 C15H19O10 C15H17O8 C8H7O4 C9H9O5

Molecular formula [M-H]−

0.3 0.6 0.1 −0.6 1.2 1.4 0.4 0.0 0.4 −1.2 0.6 1.6 0.7 0.3 0.3

Error (ppm)b

284.0325(100.0); 284.0320(100.0); 284.0323(100.0); 284.0321(100.0);

227.0350(40.6); 227.0344(46.3); 227.0349(36.3); 227.0344(42.7);

255.0297(27.6); 285.0375(33.3); 255.0293(25.3); 255.0297(26.6);

(continued on next page)

183.0447(6.0); 255.0291(28.8). 151.0033(1.4);178.9995(0.8). 563.1408(7.3).

745.1443(100.0); 575.1195(48.7); 423.0716(34.7); 457.0778(25.8); 169.0150(20.7); 593.1222(19.3). 305.0668(100.0); 425.0880(60.2); 577.1357(60.0); 289.0718(41.4); 407.0774(29.8); 451.1035(23.6); 125.0244(13.6). 245.0818(91.7); 203.0711(55.5); 125.0244(42.1); 205.0505(39.6); 151.0402(30.6); 221.0818(30.4). 745.1419(100.0); 577.0990 (24.8); 407.0761(13.4); 441.0831(11.3); 559.0909(9.2); 303.0515(8.5); 169.0141(7.5). 169.0145(100.0); 305.0663(11.3); 125.0241(8.3); 161.0248(3.1). 305.0662(100.0); 125.0246(99.4); 161.0237(38.0); 287.0546(35.4); 177.0192(33.4); 169.0138(30.0); 137.0241(20.9); 423.0710(19.4). 729.1469(100.0); 730.1506(39.9); 441.0825(29.3); 577.0998(16.7); 407.0763(13.6); 731.1535(8.6); 559.0872(8.0); 289.0712(7.7). 289.0720(100.0); 561.1407(21.2); 271.0612(13.2); 435.1082(10.2); 407.0774(6.2); 425.0873(4.7). 169.0141(100.0); 125.0243(42.5); 289.0711(27.0); 245.0811(14.3). 289.0721(100.0); 425.0886 (87.9); 577.1368 (61.4); 125.0246 (46.2); 407.0781(39.9); 451.1033(28.4). 289.0707(100.0); 137.0235(23.1); 245.0808(19.6); 125.0246(17.9); 109.0288(15.5); 271.0604(11.3); 164.0104(8.6); 151.0401(7.6); 407.0775(6.5).’ 285.0406(100.0); 447.0919(48.1); 446.0854(41.3); 593.1514(29.4). 284.0323(100.0); 227.0347(88.4); 255.0296(79.7); 285.0382(37.4); 739.2078(19.0).

169.0140 (100.0); 125.0247 (75.2) 125.0245(100.0). 108.0215(100.0); 152.0114(64.3). 109.0288(100.0); 153.0189(71.7). 108.0216(100.0); 109.0258(6.0). 108.0212(100.0); 109.0290(72.3). 93.0348(100.0); 65.0395(39.1); 94.0378(9.8). 197.0455 (100.0); 153.0555 (56.4); 138.0323 (42.1); 182.0221 (34.6). 137.0243(100.0); 93.0343(34.9). 134.0373(100.0); 135.0449(98.4). 93.0346(100.0); 65.0398(63.3). 197.0444(100); 153.0551(59.3). 119.0502(100.0); 163.0397(25.5). 152.0113(100.0); 108.0211(68.4). 121.0293(100.0); 166.9985(95.8); 182.0218(94.9); 123.0087(93.7); 138.0319(31.9); 153.0558(19.4). 119.0502(100.0); 117.0344(17.6). 441.0826(100.0); 609.1259(71.6); 423.0715(64.6); 305.0661(49.3); 483.0906(11.6). 425.0878(100.0); 593.1298(50.4); 289.0714(34.4); 407.0771(28.9); 467.0977(12.5). 125.0243(100.0); 167.0349(44.1); 305.0665(43.1); 219.0662(34.2); 221.0452(17.6); 261.0768(17.1); 237.0769(6.8). 305.0666(100.0); 125.0246(70.3); 423.0718(61.5); 137.0243(25.2); 161.0245(20.0). 761.1368(100.0); 423.0733(38.0); 591.1177(29.8); 443.1926(7.4); 169.0152(7.4); 609.1313(6.8); 303.0537(6.7); 457.0758(6.6). 177.0190(100.0); 125.0240(97.0); 289.0707(95.0); 407.0761(43.9); 285.0406(28.7).

Fragment ions (%)d

A.J. Aquino, et al.

Industrial Crops & Products 132 (2019) 59–68

B8 - Juglanin (kaempferol 3-O-αarabinofuranoside) Kaempferol-200 Kaempferol-278 (II) Kaempferol-214 Kaempferol-182 B9 - Afzelin (Kaempferol-3-O-αrhamnoside) Myricetin-455 Myricetin-hexose Myricetin-pentose (I)

Myricetin-279 Myricetin-pentose-68

Myricetin-pentose(II) Myricetin-309 Myricetin-131 B4 – Myricitrin Myricetin-279 Myricetin-pentose(III) Myricetin-175 Myricetin-217 Quercetin-deoxyheose-hexose-149 Quercetin-309 Quercetin-hexose-deoxyhexose

Quercetin-279 B6 - Avicularin - Quercetin-3-Oarabinofuranoside Quercetin-196

Quecetin-201

Quercetin-279

Quercetin-210 B7 - Quercitrin - Quercetin-3-Orhamnoside Quercetin-pentose Quercetin-desoxyhexose-68 Quercetin-313 Quercetin-217 Isorhamnetin-hexose Isorhamnetin B12 - Bauhiniastatin 2

40

49 50

51 52 53 54 55 56 57 58 59 60 61

62 63

64

65

66

67 68 30.00 30.00 34.20 42.20 30.90 38.50 45.50

29.70 29.90

29.00

28.60

28.30

27.90 28.30

24.20 24.70 24.70 25.00 25.50 26.40 27.00 37.10 23.10 24.00 26.30

23.70 24.10

20.60 22.70 23.60

32.10 33.30 33.60 33.70 33.80

32.00

HPLC Rt(min)a

501.0640 579.1343 510.0881 447.0935 433.0773 515.0807 475.0876 517.0969 477.1040 315.0503 299.0928

– – 9.10 – 9.10 10.70 – 9.50 12.20 14.60

496.0733

– 8.70

579.1346 433.0784

449.0726 625.1413 447.1291 463.0874 595.1288 449.0722 491.0828 533.0927 755.2045 609.1456 609.1449

8.60 8.40

6.90 6.60 – 7.30 7.60 – – – 7.10 7.80 8.00

595.1301 517.0598

771.1987 479.0828 449.0727

– – – 7.00 7.00

485.0693 563.1408 499.0861 467.0743 431.0988

417.0823

m/z Experimental [M-H]−

10.00 10.50 10.80 – 10.60

9.80

UHPLC Rt(min)a

0.8 4.7 1.4 3.6 −0.3 2.2 −0.9

1.5 −0.4 C20H17O11 C24H19O13 C22H19O12 C24H21O13 C22H21O12 C16H11O7 C17H15O5

C21H20NO14 C21H19O11

C26H27O15

C18H17N2O15

−1.1 2.1

C21H14N5O10

2.6

C26H27O15 C20H17O11

C20H17O12 C27H29O17 C22H23O10 C21H19O12 C26H27O16 C20H17O12 C22H19O13 C24H21O14 C33H39O20 C27H29O16 C27H29O16

−0.1 −0.4 1.2 1.6 2.8 0.8 0.7 1.9 −0.6 0.8 2.0 1.6 −1.8

C26H27O16 C23H17O14

0.6 4.9

C33H39O21 C21H19O13 C20H17O12

C18H17N2O14 C26H27O14 C24H19O12 C22H15N2O10 C21H19O10

−1.7 −0.4 4.2 −2.3 −1.1 0.2 0.6 −0.3

C20H17O10

Molecular formula [M-H]−

1.1

Error (ppm)b

35 35 35 35 35 35 35

35 35

35

35

35

35 35

35 35 35 35 35 35 35 35 35 35 35

35 35

35 35 35

35 35 35 35 35

35

Collision energy (eV)c

271.0244(7.7); 287.0198(4.3). 287.0201(21.0); 271.0246(20.1); 625.1409(18.2). 271.0237(6.2); 287.0201(3.8). 271.0242(7.5); 287.0194(4.2). 271.0245(11.8); 287.0196(10.9). 271.0233(5.9); 287.0175(3.4). 271.0237(16.4); 287.0193(12.8). 287.0189(10.1); 271.0237(9.0). 255.0289(25.3); 243.0293(24.6). 243.0297(16.9); 255.0288(10.4). 301.0334((51.1); 243.0289(31.8); 609.1424(16.6); 271.0246(26.0); 301.0314(20.2); 243.0299(14.1); 255.0296(9.6). 301.0338(51.3); 271.0246(14.3); 255.0297(8.1); 302.0371(7.9).

317.0278(28.1); 317.0264(34.4); 317.0270(24.9); 317.0279(35.4); 317.0252(13.7); 317.0258(18.0); 317.0277(28.5); 317.0248(15.4); 271.0243(73.7); 271.0243(31.5); 271.0244(52.9);

317.0260(15.3); 271.0244(12.4); 259.0245(10.2); 287.0190(9.0). 317.0271(42.3); 449.0690(21.6); 359.1493(14.9); 271.0227(13.4);

271.0233(43.2); 259.0244(27.4); 287.0189(26.0). 317.0258(22.3); 271.0240(7.0); 287.0170(5.1). 317.0285(15.9); 271.0221(7.0); 269.0442(5.9); 178.9991(4.9);

300.0265(100.0); 301.0335(100.0); 300.0274(100.0); 300.0261(100.0); 299.0194(100.0); 271.0241(100.0); 225.0556(100.0);

301.0331(31.7); 300.0267(85.4); 301.0325(26.7); 301.0297(17.7); 315.0500(52.6); 300.0270(39.8); 197.0605(83.1);

253.0500(21.0); 447.0924(28.2); 271.0237(13.0); 271.0249(16.2); 300.0249(28.0); 255.0292(38.5); 241.0505(75.3);

271.0235(13.8). 271.0236(28.0). 243.0287(7.2);255.0295(6.2). 243.0303(10.5); 255.0281(6.6). 462.0792(23.5); 271.0251(10.4). 243.0287(24.7). 210.0320(36.1); 196.0525(27.2).

300.0263 (100.0); 301.0342 (91.4); 271.0240 (25.4); 433.0776 (23.9); 243.0287 (18.9); 255.0303 (12.6). 300.0271(100.0); 301.0328(89.2); 302.0378(24.2); 271.0247(23.0); 243.0272(19.3); 433.0728(18.3); 227.0334(8.8); 255.0274(7.1). 300.0268(100.0); 271.0242(21.5); 301.0308(19.6); 243.0295(16.2); 255.0291(8.6); 579.1343(5.6); 227.0340(4.3). 300.0259(100.0); 301.0337(74.6); 447.0928(33.1); 271.0242(23.9); 243.0291(16.0). 300.0274(100.0); 301.0342(66.9); 271.0246(13.3); 255.0293(5.7).

316.0234(100.0); 316.0214(100.0); 316.0221(100.0); 287.0198(4.2). 316.0221(100.0); 316.0215(100.0); 287.0216(12.4). 316.0225(100.0); 316.0217(100.0); 316.0222(100.0); 316.0224(100.0); 316.0218(100.0); 316.0218(100.0); 316.0221(100.0); 316.0221(100.0); 300.0272(100.0); 300.0265(100.0); 300.0271(100.0); 255.0290(16.6). 300.0277(100.0); 300.0275(100.0);

284.0323 (100.0); 285.0385 (40.7); 255.0294 (38.4); 227.0347 (34.7); 183.0445 (12.8). 284.0318(100.0); 285.0392(64.2); 227.0355(47.6); 255.0300(29.5). 284.0323(100.0); 227.0343(35.5); 255.0293(22.2); 285.0369(18.9). 285.0392(100.0); 284.0323(71.0); 227.0352(33.3); 255.0284(31.7). 285.0395(100.0); 227.0342(40.7); 431.0960(34.4); 255.0278(30.6). 284.0329(100.0); 285.0398(69.8); 227.0352(35.6); 255.0301(33.3).

Fragment ions (%)d

a Rt(min) = retention time in minutes; bppm = part per million; ceV = electron volts; dfragment ion (percentage of the ion fragment); m/z = mass-to-charge ratio; HPLC = high performance liquid chromatography; UHPLC = ultra-high performance liquid chromatography.

69 70 71 72 73 74 75

64

46 47 48

41 42 43 44 45

Compound

Peak Number

Table 1 (continued)

A.J. Aquino, et al.

Industrial Crops & Products 132 (2019) 59–68

Industrial Crops & Products 132 (2019) 59–68

A.J. Aquino, et al.

Fig. 3. Fragmentation mechanism of proanthocyanidins in negative mode using QTOF/MS2. CRF: Charge Retention Fragmentation; CRF-RHR Charge Retention Fragmentation - Remote Hydrogen Rearrangement; CMF: Migration Fragmentation - Ɛ-elimination. Afz = Afzelechin; Cat = Catechin; Gal = Gallocatechin.

fragment pattern of myricitin (m/z 287 and 271, respectively), quercetin, and isorhamnetin (m/z 271 and 255, respectively). These two fragment ions also were observed in the experiments with higher collision energies (50 eV) for aglycones-derived compounds. In addition, fragment ions at m/z 178.9989 and 137.0244, or 121.0296 (quercetin and myricetin aglycones, respectively) are ions produced from the retrocyclic reaction of (1,3A0ˉand 1,2A0−), and the fragment ion at m/z 151.0036 was obtained from the elimination of carbon monoxide via charge retention fragmentation (quercetin and myricetin) (Demarque et al., 2016; March et al., 2006; Vukics and Guttman, 2008). The fragmentation mechanism of the O-derived flavonoids (glycosylated or with other losses) used for this characterization is shown in (Fig. 4) (Demarque et al., 2016; March et al., 2006; Vukics and Guttman, 2008). This fragmentation mechanism involved the transfer of protons from the sugar unit to the glycosidic oxygen and heterolytic cleavage of the hemiacetal glycosidic bond or cleavage homolytic activity by charge retention fragmentation (CRF), through radical fragmentation. Thus, losses of one or two hexoside (162), deoxyhexoside (146) and/or pentoside(132) were used to identify the O-glycosylated flavonoids (Pinheiro and Justino, 2012). In addition, for those compounds which fragment ions could not be proposed by fragmentation mechanism (for example compounds 36 and 37), they were them assigned by association of the exact mass, attributed molecular formula, and fragmentation patterns of the aglycone radical [Y0-H]%−. For those compounds, the nomenclature was the name used for the aglycone plus numbers referring to the difference between [Y0-H]%− (or deprotonated molecule ion of the aglycone [Y0H]-) and the exact mass of the deprotonated molecule of the compound [M−H]−).

which were formed after heterocyclic ring fragmentation on ring B or C, and it is a loss of phloroglucinol (Delcambre and Saucier, 2012; Miketova et al., 2000). The identification of proanthocyanidins dimers were based on their fragmentation patterns (Demarque et al., 2016; Li and Deinzer, 2007) illustrated at Fig. 3. For example, (epi)catechin-(epi)catechin (procyanidin B type) exhibited a deprotonated molecule [M−H]− at m/z 577.1351 and fragment ions (20 eV) at m/z 425.0880 (from the retroDiels-Alder rearrangement (RDA) of the heterocyclic ring), at m/z 451.1035 (heterocyclic ring fission – HRF), and m/z 289.0718 ([(epi) catechin-H]−) produced from cleavage through charge migration fragmentation (CMF) by epsilon elimination suggesting carbono-carbon linkage and epi/catechin unit, respectively. Thus, indicating which flavan-3-ol is attached to the upper or lower unit of the dimers, and interflavan linkage at the carbon C-4 with C-8 (Sun and Miller, 2003). 3.2.3. Flavonoids, derivatives and bauhiniastatin (33-75) Using the LC-SPE analytical conditions (Section 2.5), 20 fractions collected by time slice were trapped onto the SPE cartridges and the chemical structures of six compounds (40, 45, 54, 63, 68, and 75) were assigned through analysis of NMR and HRMS data. Among them, the following compounds: myricitrin (54), juglanin (40), afzelin (45), and bauhiniastatin 2 (75) are first time reported for B. longifolia. The discussion of the NMR data will be presented in the Section 3.3. In addition, the mass spectra fragmentation patterns were used to provide the partial structural characterization of flavonoids in relation to the glycosidic, aglycone, allowing the direct analysis of the ethanolic extract. The identification of the glycosylated aglycones was based on homolytic fragmentation (Fig. 4) producing the aglycone radical ion [Y0-H]%−. The kaempferol, quercetin, isorhamnetin, and myricetin aglycones were identified with the fragment ion [Y0-H]%− produced at m/z 285, 300, 315, and 316, respectively. The aglycone was confirmed by the MS2 spectrum in the experiment with fixed collision energy at 50 eV. The isorhamnetin aglycone was also characterized by the loss of a methyl radical (%CH3) producing the fragment ion at m/z 300.0270. Moreover, characteristic fragment ions at m/z 255.0301 [Y0-H−CO]− and another at m/z 227.0352, which refers to [Y0-H-C2O2]−, allowed to confirm the flavonol kaempferol chemical structure. The loss of fragment ions m/z [Y0−CO-2 H]− and [Y0−CO2-2 H]− are characteristic

3.3. Characterization of isolated compounds by LC-SPE/NMR Chromatographic bands eluting at retention time of 25–45 minutes (Fig. 2) were isolated by LC-SPE and analyzed by NMR. Chemical shifts, signals multiplicity and coupling constant (1H NMR and HSQC spectrum) of the identified compounds are presented in Table 2. In the supplementary material 2 are the 1H-NMR spectra, HSQC, HMBC data and the complete discussion of the flavonoids obtained in this experiment. 65

Industrial Crops & Products 132 (2019) 59–68

A.J. Aquino, et al.

Fig. 4. Fragmentation mechanism of O-mono/di or tri-glycosylated flavonoids in negative ion mode using QTOF/MS2. CRF: charge retention fragmentation; CRF-RF: charge retention fragmentation-radical fragmentation; CRF-CME: charge retention fragmentation-carbon monoxide elimination.

The hydrogen experiment conducted for compound 75 (C17H16O5) showed that chemical displacements (Fig. 5) to hydrogens 2, 3 and, 4 were respectively 6.86, 7.06, and 6.59 ppm, characteristic signals for hydrogen of aromatic ring. The multiplicity and constant coupling also suggested an aromatic ring in the chemical structure of this molecule. This was confirmed by COSY experiments, which demonstrated the correlation of hydrogen 3 coupling with hydrogens 2 and 4. For the central ring the hydrogen referring to positions 10 and 11 showed high coupling constant, which is characteristic of vinyl hydrogens. The two methoxyls, hydroxyl, and ethyl groups were verified by the 1H NMR signals. The correlation of the carbon to a bond (J1- HSQC) and HMBC could confirm the correct position of each substituent in the aromatic ring that was further confirmed by comparison to literature data. This molecule was identified as bauhiniastatin 2 which has been reported to display anticancer properties towards several cancer cell lines. (Pettit et al., 2006).

phytochemicals in crude plant extracts. By using this method, a total of 74 compounds were inferred based on MS, MS/MS fragmentation pattern, and comparison with online databases and other relevant bibliographies. In addition, LC-SPE/NMR was successfully applied for the first time in the analysis of Bauhinia longifolia ethanolic leaves extracts. To our knowledge, this research marks the first extensive study of B. longifolia. It has shown that the leaves from this Brazilian species of Bauhinia may well be a good source for bioactive compounds offering new opportunities for pharmaceutical and food industries to develop new products from this plant. In addition, the LC–MS method and the provided phytochemical information can be useful for quality control and/or to explain pharmacology properties. Acknowledments This work was suported by São Paulo State Research FoundationFAPESP with the research grants 2013/01710-1 and 2014/50244-6. The research grants 140469/2013-3 and 308187/2014-8 from the National Council for Scientific and Technological Development (CNPq) are also acknowledged. This work was supported also by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brazil (CAPES) - Finance Code 001. The authors are grateful to Msc. Vinicius

4. Conclusions This work demonstrates that LC-HRMS is a powerful and useful technique for the separation and detection of phenolics, and other

Table 2 Chemical shift (ppm), multiplicity and coupling constant (Hz) of the 1H NMR and HSQC spectrum of the O-derivates flavonoids compounds. 54

63

C/H

1

13

1

6 8 2’ 6’ 3’ 5’ 1’’ 2’’ 3’’ 4’’ 5’’ 6”

6.07 6.19 6.91 6.91 — — 5.28 4.21 3.78 3.35 3.50 0.94

— — 110.8 110.9 — — 104.1 72.7 72.6 74.1 — 18.7

6.06 6.20 7.48 7.45 — 6.86 5.39 4.30 3.90 3.90 3.50 —

H

C

(br s) (br s) (s) (s)

(br s) (br s) (dd; 3.2; 9.4) (m) (m) (d J = 6.1)

68 13

H (s) (s) (d;2.1) (dd; 2.1, 8.5) (d; 8.5) (br s) (br s) (br s) (br s) (m)

C

— — 116.7 123.0 — 116.7 110.1 84.1 79.7 88.9 62.9 —

40

1

13

H

6.13 6.29 7.31 7.28 — 6.89 5.33 4.20 3.74 3.30 3.50 0.93

(d; 1.6) (d; 1.8) (d; 2) (dd; 2.1; 8.5) (d; 8.3) (d; 1.4) (dd; 1.7; 3.3) (dd; 3.3; 9.4) (m) (m) (d; 6.3)

C

102.1 96.4 116.9 122.9 — 116.5 103.7 72.3 72.2 73.4 72.2 17.7

s = singlet, br s = broad singlet, d = doublet, t = triplet, dd = double of douplet, m = multiplet. 66

45

1

13

H

6.11 6.28 7.93 7.93 6.90 6.90 5.39 4.33 3.89 3.82 3.47 —

(s) (s) (d; 8.8) (d; 8.8) (d; 8.8) (d; 8.8) (br s) (dd; 3.0; 1.0) (dd; 2.9; 5.0) (m) (dd; 1.4; 4.8)

C

— — 132.5 132.5 117.4 117.4 110.1 83.9 79.3 88.8 63.2 —

1

13

H

6.15 6.31 7.74 7.74 6.92 6.92 5.36 4.20 3.69 3.30 3.30 0.91

(d; 2.0) (d; 2.0) (d; 8.7) (d; 8.7) (d; 8.7) (d; 8.7) (d; 1.7) (dd; 1.7; 3.3) (dd; 3.3; 9.4) (m) (m) (d 5.6)

C

101.9 96.5 132.6 132.6 117.5 117.5 104.3 72.9 72.3 73.4 72.3 18.3

Industrial Crops & Products 132 (2019) 59–68

A.J. Aquino, et al.

Fig. 5. 1H NMR, HSQC, COSY, HMBC of the bauhiniastatin 2.

Augusto Perasolo e Carvalho for collecting the samples of Bauhinia longifolia.

antioxidante das folhas de Bauhinia longifolia (Bong.) Steud. Rev. Virtual Química 9. Farag, M.A., Sakna, S.T., El-fiky, N.M., Shabana, M.M., Wessjohann, L.A., 2015. Phytochemical, antioxidant and antidiabetic evaluation of eight Bauhinia L. Species from Egypt using UHPLC–PDA–qTOF-MS and chemometrics. Phytochemistry 119, 41–50. https://doi.org/10.1016/J.PHYTOCHEM.2015.09.004. Ferreres, F., Gil-Izquierdo, A., Vinholes, J., Silva, S.T., Valentão, P., Andrade, P.B., 2012. Bauhinia forficata Link authenticity using flavonoids profile: relation with their biological properties. Food Chem. 134, 894–904. https://doi.org/10.1016/J. FOODCHEM.2012.02.201. Funari, C.S., Eugster, P.J., Martel, S., Carrupt, P.-A., Wolfender, J.-L., Silva, D.H.S., 2012. High resolution ultra high pressure liquid chromatography–time-of-flight mass spectrometry dereplication strategy for the metabolite profiling of Brazilian Lippia species. J. Chromatogr. A 1259, 167–178. https://doi.org/10.1016/J.CHROMA. 2012.03.069. Guan, T.-Y., Liang, Y., Li, C.-Z., Xie, L., Wang, G.-J., Sheng, L.-S., 2011. Recent development in liquid chromatography/mass spectrometry and allied topics for traditional chinese medicine research. Chin. J. Nat. Med. 9, 385–400. https://doi.org/10.3724/ SP.J.1009.2011.00385. HMDB0000152, n.d. Gentisic acid (HMDB0000152). http://www.hmdb.ca/metabolites/ HMDB0000152 (accessed 7.25.18). HMDB0001856, n.d. Protocatechuic acid (HMDB0001856). http://www.hmdb.ca/ metabolites/HMDB0001856 (accessed 7.25.18). Kakazu, Y., Horai, H., 2011. Syringic Acid Mass Spectrum. (accessed 2.20.18). https:// massbank.eu/MassBank/jsp/RecordDisplay.jsp?id=KO001814&dsn=Keio. Kakazu, Y., Horai, H., 2016. Gentisic acid mass spectrum. Mass Bank (accessed 2.20.18). https://massbank.eu/MassBank/jsp/RecordDisplay.jsp?id=KO000577&dsn=Keio. Kittakoop, P., Nopichai, S., Thongon, N., Charoenchai, P., Thebtaranonth, Y., 2004. Bauhinoxepins a and B: new antimycobacterial ibenzo[b,f]oxepins from Bauhinia saccocalyx. Helv. Chim. Acta 87, 175–179. https://doi.org/10.1002/hlca. 200490006. Li, H.-J., Deinzer, M.L., 2007. Tandem mass spectrometry for sequencing proanthocyanidins. Anal. Chem. 79, 1739–1748. https://doi.org/10.1021/ac061823v. Liu, M., Dong, J., Lin, Z., Niu, Y., Zhang, X., Jiang, H., Guo, N., Li, W., Wang, H., Chen, S., 2016. Rapid screening of transferrin-binders in the flowers of Bauhinia blakeana Dunn by on-line high-performance liquid chromatography–diode-array detector–electrospray ionization–ion-trap–time-of-flight–mass spectrometry–transferrin–fluorescence detection system. J. Chromatogr. A 1450, 17–28. https://doi.org/10.1016/J.CHROMA.2016.04.043. March, R.E., Lewars, E.G., Stadey, C.J., Miao, X.-S., Zhao, X., Metcalfe, C.D., 2006. A comparison of flavonoid glycosides by electrospray tandem mass spectrometry. Int. J. Mass Spectrom. 248, 61–85. https://doi.org/10.1016/J.IJMS.2005.09.011. Matsuda, F., 2016. Caffeic Acid Mass Spectrum. (accessed 2.20.18). https://massbank. eu/MassBank/jsp/RecordDisplay.jsp?id=KO000513&dsn=Keio. Matsuda, F., Suzuki, M., Sawada, Y., 2016a. Protocatechuic Acid Mass Spectrum. (accessed 2.20.18). https://massbank.eu/MassBank/jsp/RecordDisplay.jsp?id= PR100601&dsn=RIKEN. Matsuda, F., Suzuki, M., Sawada, Y., 2016b. 4-Hydroxybenzoic Acid Mass Spectrum. (accessed 2.20.18). https://massbank.eu/MassBank/jsp/RecordDisplay.jsp?id= PR100596&dsn=RIKEN. Matsuda, F., Suzuki, M., Sawada, Y., 2016c. 2-Coumaric Acid Mass Spectrum. (accessed 2.20.18). https://massbank.eu/MassBank/jsp/RecordDisplay.jsp?id=KO000442& dsn=Keio. Miketova, P., Schram, K.H., Whitney, J., Li, M., Huang, R., Kerns, E., Valcic, S., Timmermann, B.N., Rourick, R., Klohr, S., 2000. Tandem mass spectrometry studies of green tea catechins. Identification of three minor components in the polyphenolic extract of green tea. J. Mass Spectrom. 35, 860–869. https://doi.org/10.1002/10969888(200007)35:7<860::AID-JMS10>3.0.CO;2-J. Mishra, A., Sharma, A.K., Kumar, S., Saxena, A.K., Pandey, A.K., 2013. Bauhinia variegata leaf extracts exhibit considerable antibacterial, antioxidant, and anticancer activities. Biomed Res. Int., 915436. https://doi.org/10.1155/2013/915436. 2013. Nageshwar, G., Anuradha, S.M.J., Radhakrishnaiah, M., Narayana, L.L., 1986. Distribution pattern of phenolic constituents in species of Bauhinia linn. And its taxonomic significance. Proc. Plant Sci 96, 1–7. https://doi.org/10.1007/ bf03053264. Nogueira, A.C.O., Sabino, C.V.S., 2012. Revisão do gênero Bauhinia abordando aspectos científicos das espécies Bauhinia forficata link e Bauhinia variegata L. De interesse para a indústria farmacêutica. Rev. Fitos 7, 77–84.

Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.indcrop.2019.01.040. References Allard, P.-M., Genta-Jouve, G., Wolfender, J.-L., 2017. Deep metabolome annotation in natural products research: towards a virtuous cycle in metabolite identification. Curr. Opin. Chem. Biol. 36, 40–49. https://doi.org/10.1016/J.CBPA.2016.12.022. Aquino, A.J., 2018. Diferenciação química de quatro espécies de Bauhinia (pata-de-vaca). Federal University of São Carlos - UFSCar. Beisken, S., Earll, M., Baxter, C., Portwood, D., Ament, Z., Kende, A., Hodgman, C., Seymour, G., Smith, R., Fraser, P., Seymour, M., Salek, R.M., Steinbeck, C., 2014. Metabolic differences in ripening of Solanum lycopersicum ‘Ailsa Craig’ and three monogenic mutants. Sci. Data 1, 140029. https://doi.org/10.1038/sdata.2014.29. Boettcher, C., 2016. Catechin Mass Spectrum. URL (accessed 2.27.18). https:// massbank.eu/MassBank/jsp/RecordDisplay.jsp?id=PB002430&dsn=IPB. Brasil - Agência Saúde, 2009. Ministério da Saúde (MS) elabora relação de plantas medicinais de interesse ao SUS. Carvalho, V.A.P., 2011. Caracterização química por cromatografia líquida e análise quimiométrica de espécies vegetais de Bauhinia com aplicação em controle de qualidade de amostras comerciais de pata-de-vaca. Universidade Federal de São Carlos. Compaoré, M., Lamien, C.E., Lamien-Meda, A., Vlase, L., Kiendrebeogo, M., Ionescu, C., Nacoulma, O.G., 2012. Antioxidant, xanthine oxidase and lipoxygenase inhibitory activities and phenolics of Bauhinia rufescens Lam. (Caesalpiniaceae). Nat. Prod. Res. 26, 1069–1074. https://doi.org/10.1080/14786419.2011.559948. Consulta Pública do Herbário Virtual - Bauhinia longifolia, 2018. http://reflora.jbrj.gov. br/reflora/herbarioVirtual/ConsultaPublicoHVUC/ BemVindoConsultaPublicaHVConsultar.do?modoConsulta=LISTAGEM& quantidadeResultado=20&genero=bauhinia&especie=longifolia URL (accessed 11. 2.17). Crozier, A., Jaganath, I.B., Clifford, M.N., 2009. Dietary phenolics: chemistry, bioavailability and effects on health. Nat. Prod. Rep. 26, 1001. https://doi.org/10.1039/ b802662a. DB Epigallocatechin MS/MS Spectrum - ESI-TOF 20V, Negative (DB03823) - DrugBank. 03823, n.d. (accessed 2.27.18). https://www.drugbank.ca/spectra/ms_ms/374093. de Sousa, E., Zanatta, L., Seifriz, I., Creczynski-Pasa, T.B., Pizzolatti, M.G., Szpoganicz, B., Silva, F.R.M.B., 2004. Hypoglycemic Effect and Antioxidant Potential of kaempferol3,7-O-(α)-dirhamnoside From Bauhinia forficata Leaves. https://doi.org/10.1021/ NP030513U. Delcambre, A., Saucier, C., 2012. Identification of new flavan-3-ol monoglycosides by UHPLC-ESI-Q-TOF in grapes and wine. J. Mass Spectrom. 47, 727–736. https://doi. org/10.1002/jms.3007. Demarque, D.P., Crotti, A.E.M., Vessecchi, R., Lopes, J.L.C., Lopes, N.P., 2016. Fragmentation reactions using electrospray ionization mass spectrometry: an important tool for the structural elucidation and characterization of synthetic and natural products. Nat. Prod. Rep. 33, 432–455. https://doi.org/10.1039/ c5np00073d. Di Stasi, L.C., Oliveira, G.P., Carvalhaes, M.A., Queiroz-Junior, M., Tien, O.S., Kakinami, S.H., Reis, M.S., 2002. Medicinal plants popularly used in the brazilian tropical atlantic forest. Fitoterapia 73, 69–91. https://doi.org/10.1016/S0367-326X(01) 00362-8. dos Santos, A.E., Kuster, R.M., Yamamoto, K.A., Salles, T.S., Campos, R., de Meneses, M.D., Soares, M.R., Ferreira, D., 2014. Quercetin and quercetin 3-O-glycosides from Bauhinia longifolia (Bong.) Steud. Show anti-Mayaro virus activity. Parasit. Vectors 7, 130. https://doi.org/10.1186/1756-3305-7-130. Duarte-Almeida, J., Negri, G., Salatino, A., 2004. Volatile oils in leaves of Bauhinia (Fabaceae caesalpinioideae). Biochem. Syst. Ecol. 32, 747–753. https://doi.org/10. 1016/J.BSE.2004.01.003. Ernestina, A., Santos, D., Barreto Junior, C.B., Kuster, R.M., 2017. Flavonoides e atividade

67

Industrial Crops & Products 132 (2019) 59–68

A.J. Aquino, et al.

wine and B-type dehydrodicatechins in an autoxidation mixture of (+)-catechin and (-)-epicatechin. J. Mass Spectrom. 38, 438–446. https://doi.org/10.1002/jms.456. Vaz, A.M.S., da, F., 2010. Checklist of Bauhinia sensu stricto (Caesalpiniaceae) in Brazil. Plant Ecol. Evol. 143, 212–221. https://doi.org/10.5091/plecevo.2010.391. Villavicencio, A.L.C.H., Heleno, S.A., Calhelha, R.C., Santos-Buelga, C., Barros, L., Ferreira, I.C.F.R., 2018. The influence of electron beam radiation in the nutritional value, chemical composition and bioactivities of edible flowers of Bauhinia variegata L. Var. Candida alba Buch.-Ham from Brazil. Food Chem. 241, 163–170. https://doi. org/10.1016/J.FOODCHEM.2017.08.093. Vukics, V., Guttman, A., 2008. Structural characterization of flavonoid glycosides by multi-stage mass spectrometry. Mass Spectrom. Rev. 29. https://doi.org/10.1002/ mas.20212. n/a-n/a. Wu, Z., Gao, W., Phelps, M.A., Wu, D., Miller, D.D., Dalton, J.T., 2004. Favorable effects of weak acids on negative-ion electrospray ionization mass spectrometry. Anal. Chem. 76, 839–847. https://doi.org/10.1021/ac0351670. Wu, H., Guo, J., Chen, S., Liu, X., Zhou, Y., Zhang, X., Xu, X., 2013. Recent developments in qualitative and quantitative analysis of phytochemical constituents and their metabolites using liquid chromatography–mass spectrometry. J. Pharm. Biomed. Anal. 72, 267–291. https://doi.org/10.1016/j.jpba.2012.09.004. Xu, W., Chu, K., Li, H., Zhang, Y., Zheng, H., Chen, R., Chen, L., 2012. Ionic liquid-based microwave-assisted extraction of flavonoids from Bauhinia championii (Benth.) Benth. Molecules 17, 14323–14335. https://doi.org/10.3390/molecules171214323.

Pettit, G.R., Numata, A., Iwamoto, C., Usami, Y., Yamada, T., Ohishi, H., Cragg, G.M., 2006. Antineoplastic Agents. 551. Isolation and structures of bauhiniastatins 1-4 from Bauhinia purpurea. J. Nat. Prod. 69, 323–327. https://doi.org/10.1021/np058075+. Pinheiro, P.F., Justino, G.C., 2012. Structural analysis of flavonoids and R related compounds - A review of spectroscopic applications, in: phytochemicals - A global perspective of their role in nutrition and health. InTech. https://doi.org/10.5772/ 29152. Sashidhara, K.V., Singh, S.P., Srivastava, A., Puri, A., 2013. Main extracts and hypolipidemic effects of the Bauhinia racemosa Lam. Leaf extract in HFD-fed hamsters. Nat. Prod. Res. 27, 1127–1131. https://doi.org/10.1080/14786419.2012.711326. Shang, X., Li, S., Wang, Y., Wang, S., Yang, Y., Shi, J., 2006. Chemical constitutents of Bauhinia aurea. Zhongguo Zhong Yao Za Zhi 31, 1953–1955. Silva, K.L., da Cechinel Filho, V., 2002. Plantas do gênero Bauhinia: composição química e potencial farmacológico. Quim. Nova 25, 449–454. https://doi.org/10.1590/ S0100-40422002000300018. Silva, da, C.M.S., Habermann, G., Marchi, M.R.R., Zocolo, G.J., 2012. The role of matrix effects on the quantification of abscisic acid and its metabolites in the leaves ofBauhinia variegata L. using liquid chromatography combined with tandem mass spectrometry. Brazilian J. Plant Physiol 24, 223–232. https://doi.org/10.1590/ S1677-04202012000300009. Silva-López, R.E., Santos, B.C., 2015. Bauhinia forficata link (Fabaceae). Rev. Fitos 9. https://doi.org/10.5935/2446-4775.20150018. Sun, W., Miller, J.M., 2003. Tandem mass spectrometry of the B-type procyanidins in

68