Chemical characterization with in vitro biological activities of Gypsophila species

Journal of Pharmaceutical and Biomedical Analysis 155 (2018) 56–69

Contents lists available at ScienceDirect

Journal of Pharmaceutical and Biomedical Analysis journal homepage: www.elsevier.com/locate/jpba

Chemical characterization with in vitro biological activities of Gypsophila species Dimitrina Zheleva-Dimitrova a,∗ , Gokhan Zengin b , Vessela Balabanova a , Yulian Voynikov c , Valentin Lozanov d , Irina Lazarova c , Reneta Gevrenova a a

Department of Pharmacognosy, Faculty of Pharmacy, Medical University of Sofia, Bulgaria Department of Biology, Faculty of Science, Selcuk University, Turkey c Department of Chemistry, Faculty of Pharmacy, Medical University of Sofia, Bulgaria d Department of Medical Chemistry and Biochemistry, Faculty of Medicine, Medical University of Sofia, Bulgaria b

a r t i c l e

i n f o

Article history: Received 14 November 2017 Received in revised form 16 March 2018 Accepted 17 March 2018 Available online 21 March 2018 Keywords: Gypsophila Natural products Flavonoids Antioxidant Enzyme inhibitor

a b s t r a c t Methanol-aqueous extracts from the aerial parts of Gypsophila glomerata (GGE), G. trichotoma (GTE) and G. perfoliata (GPE) were investigated for antioxidant potential using different in vitro models, as well as for phenolic and flavonoid contents. The possible anti-cholinesterase, anti-tyrosinase, anti-amylase and anti-glucosidase activities were also tested. The flavonoid variability was analyzed using ultra highperformance liquid chromatography (UHPLC) coupled with hybrid quadrupole-Orbitrap high resolution mass spectrometry (HRMS). Eleven C-glycosyl flavones and 4 O-glycosyl flavonoids, including 2”-Opentosyl-6-C-hexosyl-apigenin/methylluteolin, as well as their mono(di)-acetyl derivatives were found in GGE. Both GGE and GTE shared 2”-pentosyl-6-C-hexosyl-luteolin together with the common saponarin, homoorientin, orientin, isovitexin and vitexin, while di C-glycosyl flavones were evidenced only in GPE. The highest radical scavenging in both ABTS and DPPH assays was noted in GPE, as well as ferric and cupric reducing abilities. However, GTE had the strongest metal chelating activity (17.44 ± 0.51 mg EDTAE/g extract). GPE and GGE were more potent as acetylcholinesterases inhibitors witnessed by 2.09 ± 0.02 mg GALAE/g extract and 1.59 ± 0.09 mgGALAE/g extract, respectively. All flavonoids were found in G. glomerata for the first time. Therefore, further isolation and structural elucidation of newly described acetylated flavonoids are needed in order to determine their relevance in the beneficial properties of the plant. © 2018 Elsevier B.V. All rights reserved.

1. Introduction Gypsophila L. (Caryophyllaceae) species are highly valued and important medicinal herbs widely spread in Asia and Europe [1,2]. Mainland China is one of the most important centers of biodiversity of the genus. Native species in China have been known for their significant ethno-medicinal properties such as G. oldhamiana, G. paniculata and other localized species which have application in traditional Chinese medicine to treat fever, consumptive disease, and infantile malnutrition syndrome [3]. G. oldhamiana roots are recommended in the folk medicine for the treatment of diabetes [3], while G. elegans is used as a traditional medicine for immune disorders and liver diseases [4]. In Turkey both G. bicolor and G.

∗ Corresponding author. E-mail address: [email protected]fia.bg (D. Zheleva-Dimitrova). https://doi.org/10.1016/j.jpba.2018.03.040 0731-7085/© 2018 Elsevier B.V. All rights reserved.

arrostii, var. nebulosa roots are widely used as emulgators in the production of “tahini halvah” [5]. Indeed, triterpenoid saponins from Gypsophila roots are exploited for a variety of purposes including as medicines, detergents and adjuvants [1]. In the last few years our studies on Gypsophila trichotoma native to Bulgaria led to the isolation of a variety of triterpenoid saponin bidesmosides [6,7]. Their synergistic cytotoxicity in combination with type I ribosome-inactivating protein (RIP-I) was evaluated in vitro and used to derive a quantitative structure − activity relationship (QSAR) [8]. To facilitate effective resource utilization, aerial parts of several Gypsophila species including G. elegans, G. trichotoma, G. pacifica, have been investigated for flavonoids [9–11]. Recently, the hepatoprotective and antioxidant properties of C-flavonoids isoorientin-2”-O-␣-l-arabinopyranosyl and isoorientin, isolated from G. elegans, were evidenced in alcohol- and carbon tetrachloride-induced injuries in rats [4,12]. It has been shown that saponarin isolated from G. trichotoma aerial parts is effective

D. Zheleva-Dimitrova et al. / Journal of Pharmaceutical and Biomedical Analysis 155 (2018) 56–69

in attenuating hepatic damage in the paracetamol-, cocaine- and tetrachloride-induced in vitro/in vivo models [10,13,14]. Moreover, the potential of isoorientin-2”-O-␣-l-arabinopyranosyl to induces apoptosis in liver cancer HepG2 cells via mitochondrial-mediated pathway has been reported [15]. In addition, phenylethanoid glycoside verbascosides was identified in G. pilulifera as the main free-radical scavenger present in this species [16]. However, there are only a few studies with respect to the antioxidant potential of Gypsophila species related to C-flavonoid glycosides [4,10,12]. Moreover, little data is available on the enzyme inhibitory activities of the genus Gypsophila [17]. Although a large number of Gypsophila C-flavonoids have been characterized, there are no reports concerning their profiling by LC-ESI/MS. Six C- and O-glycosylflavones from G. paniculata were analyzed by HPLC-DAD-ESI-IT MS and tentative assignment of their structures was proposed [18]. Therefore, it was of interest to extend our assay for the evaluation of antioxidant properties and enzyme inhibitory potential of G. trichotoma and G. glomerata native to Bulgaria, and G. perfoliata originating from Turkey. Based on all above mentioned studies, we aimed at investigating the flavonoid variability in G. trichotoma, G. glomerata and G. perfoliata aerial parts using UHPLC-ESI/HRMS. In addition, our objective was to evaluate their antioxidant potential using different in vitro models and to test the possible enzyme inhibitory activities linked to Alzheimer’s disease and diabetes mellitus. 2. Materials and methods 2.1. Plant material G. trichotoma Wend. aerial parts were collected in August 2004 at the Black Sea coast (Kavarna region) (43◦ 25 N–28◦ 19 E) in Bulgaria, while G. glomerata aerial parts were collected in September, 2010 at Ognyanovo village (Pazardjik region) (42◦ 15 N–24◦ 42 E). The plants were identified by Dr. R. Gevrenova (Faculty of Pharmacy, Medical University-Sofia, Bulgaria). Voucher specimen of G. trichotoma was deposited in the Herbarium of the Faculty of Pharmacy of Nancy, Universite de Lorraine, France (HP101), while voucher specimen of G. glomerata was deposited at Institute of Biodiversity and Ecosystems Research, Bulgarian Academy of Sciences, Sofia, Bulgaria (SOM 171499). G. perfoliata var. perfoliata was collected from Kayseri-Turkey (Ali Mount) during of flowering season (June 2015). Taxonomic identification of the plant material was confirmed by the senior taxonomist Dr. Murad Aydin Sanda, from the Department of Biology, Selcuk University [19,20]. The voucher specimen was deposited at the Herbarium of the Department of Biology, Selcuk University, Konya-Turkey (GZ 1513). Air-dried powdered aerial parts were extracted with 80% methanol (1:25 w/o) (×3) by sonication for 15 min at room temperature to yield the crude extracts (G. glomerata, GGE; G. trichotoma, GTE; G. perfoliata, GPE). 2.2. Chemicals The standards of orientin, homoorientin, vitexin, isovitexin, kaempferol-3-glucoside (astragalin), apigenin, luteolin, diosmetin and kaempferol (≥99% HPLC purity), saponarin and kaempferol3-rutinoside (≥98% HPLC purity), isorhamnetin 3-glucoside (≥95% HPLC purity) were provided from Extrasynthese (Genay, France). 6, 8-diC-glucosyl-apigenin (Vicenin-2) (95% HPLC) was isolated in our laboratory [21]. The stock standard solutions of appropriate concentration were prepared in methanol and were stored at 4 ◦ C in the dark. Acetonitrile (hypergrade for LC–MS), formic acid (HPLCgrade) and methanol (analytical grade) were purchased from Merck (Darmstade, Germany).

57

2.3. UHPLC-ESI/HRMS The LC–MS analyses were performed on a Q Exactive Plus heated electrospray ionization (HESI-II) – high resolution mass spectrometer (HRMS) (ThermoFisher Scientific, Inc., Bremen, Germany) equipped with an ultra high-performance liquid chromatography (UHPLC) system Dionex Ultimate 3000RSLC (ThermoFisher Scientific, Inc.). Operating conditions for the HESI source used in a negative ionization mode were: −2.5 kV spray voltage, 320 ◦ C capillary temperature, 300 ◦ C probe heater temperature, sheath gas flow rate 38 units, auxiliary gas flow 12 units (units refer to arbitrary values set by the Exactive Tune software) and S-Lens RF level 50.00. Nitrogen was used for sample nebulization and collision gas in HCD cell. The LC–MS method was operating in Full scan-dd MS2 /Top 5 with the following settings: 70000 FWHM resolution (at m/z 200), AGC target 1e6, max. IT 50 ms and mass range 100–1000 m/z, while ddMS2 conditions were set to resolution 17500 FWHM (at m/z 200), AGC target 1e3, max. IT 50 ms, isolation window 2.0 m/z and normalased collision energy (NCE) 30. The UHPLC separations were performed on a Kromasil EternityXT C18, 1.8 ␮m, 2.1 × 100 mm (AkzoNobel, Sweeden) with a binary mobile phase consisting of solution A: 0.1% HCOOH and solution B: MeCN (0.1% HCOOH). The following step gradient profile was used: 5% B for 1.0 min, increasing up to 25% B in 14 min, held isocratic at 25% B for 2.0 min, increasing up to 50% B in 1.0 min, held isocratic at 50% B for 2.0 min, increasing up to 95% B in 2.0 min, held isocratic for 2.0 min finally brought back down to 5% B over 0.5 min [22]. Equilibration time was 4.5 min, the flow rate was 0.3 mL/min. The column compartment temperature was set at 40 ◦ C. Data were processed with Xcalibur software ver. 3.0. The calculation of the exact masses and mass measurement errors, prediction of molecular formulas and simulation of monoisotopic profiles were carried out with Xcalibur 3.0 software (ThermoScientific). 2.4. Total phenolics and flavonoids content The total phenolics content was determined by the FolinCiocalteu method [23] with slight modification and expressed as gallic acid equivalents (GAE/g extract), while total flavonoids content was determined using AlCl3 method [24] with slight modification and expressed as rutin equivalents (RE/g extract). 2.5. Biological activities evaluation Antioxidant (DPPH, ABTS radical scavenging, reducing power (CUPRAC and FRAP), phosphomolybdenum and metal chelating (ferrozine method) and enzyme inhibitory assays (cholinesterase (Elmann’s method), tyrosinase (dopachrome method), ␣-amylase (iodine/potassium iodide method) and ␣-glucosidase (chromogenic PNPG method) were reported in our previous papers [24,25]. Antioxidant capacities were expressed as equivalents of Trolox and EDTA (for metal chelating). The enzyme inhibitory activities were evaluated as equivalents of standard inhibitors per gram dry extract (galantamine for AChE and BChE, kojic acid for tyrosinase and acarbose for ␣-amylase and ␣-glucosidase inhibition assays). 3. Results and discussion 3.1. UHPLC-ESI/HRMS analysis of G. glomerata, G. trichotoma and G. perfoliata extracts The methanol extracts from G. glomerata (GGE), G. trichotoma (GTE) and G. perfoliata (GPE) were analyzed by UHPLC gradient elution with Orbitrap-HRMS detection for the first time. According

58

D. Zheleva-Dimitrova et al. / Journal of Pharmaceutical and Biomedical Analysis 155 (2018) 56–69

Fig. 1. Total ion chromatograms (TIC) of Gypsophila glomerata (A), G. trichotoma (B) and G. perfoliata (C) extracts in negative ion mode. For compounds number see Table 1.

to the common approach an UHPLC-ESI/HRMS method involving binary solvent system (acetonitrile and 0.1% aqueous formic acid) and gradient elution was applied for the flavonoid profilings of the studied Gypsophila species [22,26]. To achieve the highest sensitivity, the ionization of flavonoids was performed in both negative and positive ion modes. The total ion chromatograms (TIC) of the extracts are shown in Fig. 1. The structures of the assayed compounds are depicted in Fig. 2. Positive and negative ion monitoring of the selected flavonoid standards were compared based on sensitivity (the areas of the chromatographic peaks), and selectivity (purity of the “molecular ion” peak). There is no marked difference in the ionization responces between negative and positive ion modes of ESIHRMS detection. The identification or tentative elucidation of the flavonoids were based on HRMS and MS/MS data, comparison with fragmentation fingerprints observed for the standard flavonoids and literature data [18,26,27]. The fragmentation patterns of the assigned flavonoids, as well as the accurate and exact m/z of the deprotonated [M−H]− and protonated [M+H]+ molecular ions are presented in Table 1. The commonly used nomenclature for flavonoids was adopted to indicate the fragment ions [28]. Cross-ring cleavages of sugar moieties (X) are designated by superscript numbers indicating the bond cleaved [29]. The ions that are formed after the cleavage of the C-ring of the aglycone are denoted i,j A+ /i,j B+ /(in positive ion mode) and i,j A− /i,j B− (in negative ion mode) where the ion A contains A-ring and the ion B − B ring. MS/MS spectral fingerprints of the C-6 (homoorientin and isovitexin) and C-8 (orientin and vitexin) mono glycosyl flavones standards were compared. Thus, when quadrupole-Orbitrap mass spectrometer was employed, the C-8 isomers displayed base peaks [M+H]+ at m/z 449.1078 (orientin) and 433.1129 (vitexin). Even though, C-6 isomers can be distinguished by the higher abundance of the fragment ions at m/z 299.0458 [(M+H)-150]+ (97%) (homoorientin) and 283.0599 (100%) (isovitexin) supported by

Fig. 2. Structures of the assayed compounds.

Table 1 Assignment of flavonoids from Gypsophila trichotoma, G. glomerata and G. perfoliata by UHPLC-HRMS. Peak No.

tRmin

Mono C-glycosyl flavones 5.39 1.

[M−H]- m/z Delta ppm

MS/MS data m/z

[M+H]+ m/z Delta ppm

MS/MS data m/z

Proposed compound

448.1006 C21 H19 O11

447.0930 1.817

447.09 (100), 327.05 (66.41), 357.00 (43.78), 299.05 (12.75), 297.04 (11.51), 285.04 (8.20), 241.05 (0.76), 213.05 (1.42), 201.02 (0.94), 165. 02 (1.40), 151. 00 (0.64), 133. 02 (16.47)

449.1071 −1.643

449.10 (100), 431.09 (37.28), 413.08 (15.00), 383.07 (15.67), 353.06 (34.33), 329.06 (63.12), 299.05 (97.82), 165.01, 287.05 (7.51), 243.53 (1.17), 161.02 (1.98), 137.02 (6.97), 135.04 (4.60), 121.03 (2.14) 449.10 (100), 431.09 (11.05), 413.08 (15.97), 383.07 (6.12), 353.06 (8.94), 329.06 (34.39), 299.05 (24.37), 287.05 (37.60), 161.02 (2.92), 153.01 (18.89), 149.02 (2.74), 137.05 (5.85), 135.04 (3.24) 433.11 (88.28), 415.10 (35.65), 397.09 (18.44), 367.07 (19.29), 337.07 (39.02), 313.07 (64.13), 283.05 (100), 271.05 (7.12), 165.01 (19.63), 163.03 (4.71), 121.02 (11.75) 433.11 (100), 415.10 (11.86), 397.09 (18.01), 367.07 (7.60), 337.07 (6.96), 313.07 (33.46), 283.05 (22.18), 271.05 (7.12), 165.01 (19.63), 163.03 (4.63), 121.03 (11.22)

6-C-glucosylluteolin (Homoorientin)*, a , b , c

2.

5.52

448.1006 C21 H19 O11

447.0932 2.264

447.09 (99.04), 327.05 (100), 357.06 (31.47), 297.04 (13.33), 285.04 (4.49), 255.03 (1.03), 241.05 (0.94), 201. 02 (1.47), 165.02 (1.44), 133.03 (12.38)

449.1073 −1.198

3.

5.97

432.1056 C21 H19 O10

431.0981 1.918

431.09 (100), 341.06 (21.99), 312.05 (6.04), 311.05 (80.25), 283.06 (35.21), 269.04 (4.39), 239.07 (0.97), 117.03 (4.14)

433.1123 −1.347

4.

6.15

432.1056 C21 H19 O10

431.0981 1.918

431.09 (95.60), 341.06 (9.54), 312.05 (11.48), 311.05 (100), 283.06 (32.64), 161.02 (4.49), 269.04 (1.68), 239.07 (1.60), 224.04 (0.61), 197.05 (1.75), 135.04 (3.97), 117.03 (15.61)

433.1125 −1.000

580.1428 C26 H27 O15

579.1359 2.665

579.14 (100), 489.10 (4.74), 459.09 (19.63), 429.09 (13.58), 399.07 (44.77), 369.06 (41.97), 339 (5.47), 133.03 (11.26) 593.15 (100), 503.12 (5.35), 473.11 (19.29), 383.08 (31.32), 353.07 (51.48), 297.08 (18.95) 117.03 (3.59) 579.14 (100), 561.13, 489.10 (6.77), 459.09 (23.31), 429.08 (42.86), 411.07 (3.43), 399.07 (39.32), 369.06 (35.88), 339.05 (6.10), 327.05 (3.45), 311.06 (7.47), 298.05 (43.28), 285.04 (0.36), 255.03 (0.41), 243.03 (0.42), 241.05 (0.99), 239.03 (1.12), 229.05 (0.25), 217.05 (0.37), 213.05 (1.04), 201.02 (1.14), 151.04 (0.97), 135.04 (1.55), 133.03 (9.58), 107.01 (0.43) 563.14 (100), 503.12(5.70) 473.11 (11.06), 443.11 (12.20), 425.09 (4.15), 413.08 (2), 383.08 (27.81), 353.07 (35.34) 337.07 (1.71), 297.08 (13) 149.02 (4.78), 135.04 (3.42), 117.03 (4.78) 593.15 (100), 533.13 (2), 503.12 (8), 473.11 (12), 455.10, 413.09 (37), 395.08 (1.43), 383.08 (46), 323.06 (2), 311.06 (13.8), 283.06 (0.16), 299.06 (3.1), 284.03 (2.9), 256.04 (0.6), 151.00 (0.2)

–

–

(X)C-hexosyl-(X)Cpentosyl-luteolinisomerc

–

–

581.1499 −0.338 C26 H29 O15

581.15 (100), 563.14 (27.76), 545.13 (13.38), 527.12 (8.32), 461.09, 449.05 (6.60), 443.10 (17.66), 425.09 (20.88), 407.08 (20.47), 353.07 (9), 323.06 (13.35), 311.06 (18.16) 203.03 (6.01), 137.02 (12.72), 135.04 (2.02)

6, 8-diC-glucosylapigenin (Vicenin−2)c 6-C-hexosyl−8-Cpentosyl-luteolinc

565.1545 −1.136

565.15 (100), 547.14 (30.77), 529.13 (14), 499.12 (8.37), 481.11, 433.09, 415.08, 427.10 (15), 397.09, 409.09 (21.34), 379.08 (23.75), 337.07 (11.78) 145.03 (1.92), 121.03 (17.24), 119.05 (2.66)

6-C-hexosyl−8-Cpentosyl-apigeninc

595.1655 0.364

595.17 (100), 577.15 (26.05), 559.14 (12.11), 541.13 (10.64), 505.12 (3.77), 463.10(4.7), 439.10 (24.69), 421.09 (29.74), 409.09 (23.19), 393.10 (3.51), 151.04 (5.07)

6-C-hexosyl−8-Cpentosyldiosmetinc

Di C-glycosyl flavones 4.72 5.

6.

4.75

594.1585 C27 H29 O15

593.1522 3.580

7.

4.91

580.1428 C26 H27 O15

579.1355 1.819

8.

5.29

564.1479 C26 H27 O14

563.1411 2.802

9.

5.42

594.1585 C27 H29 O15

593.1519 3.074

8-C-glucosylluteolin (Orientin)* , a , b , c

6-C-glucosylapigenin (Isovitexin)* , a , b , c

8-C-glucosylapigenin (Vitexin)* , a , b , c

D. Zheleva-Dimitrova et al. / Journal of Pharmaceutical and Biomedical Analysis 155 (2018) 56–69

Theoret. mass Molecular formula

59

60

Table 1 (Continued) Peak No.

tRmin

Theoret. mass Molecular formula

[M−H]- m/z Delta ppm

C-glycosyl flavones O-glycosylated on the phenolic hydroxyl 10. 593.1514 4.36 594.1585 2.147 C27 H30 O15

5.22

594.1585 C27 H30 O15

593.1511 1.692

C-glycosyl flavones O-glycosylated on the sugar moiety of the C-glycosylation 12. 5.29 580.1428 579.1357 C26 H27 O15 2.165

[M+H]+ m/z Delta ppm

MS/MS data m/z

Proposed compound

593.15 (100), 503.12 (23) 473.11 (66), 341.07 (2), 311.06 (17.09), 293.05 (4.5), 282.05 (62) 269.05 (3.94), 239.03 (1.06), 117.03 (1.55)

595.1658 0.039

6-C−4 -Odiglucosylapigenin (Isosaponarin)c

593.15 (100), 311.05 (39.96), 431.09 (13.49), 297.04 (27.34), 269.04 (4.57), 239.07 (0.40), 117.03 (5.27), 225.05 (0.53), 178.99 (3.46)

595.1671 −2.291

595.17(100), 577.16 (22.43), 475.12 (3.78), 415.10 (5.45), 397.09 (17.08), 379.08 (16.55), 361.07, 337.07 (22.81), 313.07 (64.15), 283.06 (76.87), 271.06(3.59), 121.0282 (4.15) 595.16 (91.34), 577.15 (24.39), 475.12, 433.11 (27.86), 415.10 (24.48), 379.08 (19.45), 337.07 (35.21), 313.07 (66.43), 283.05 (100), 165.01 (12.22), 121.02 (9.05)

579.13 (100), 459.09 (31.24), 447.09, 429.08 (10.56), 369.06 (3.51), 357.06 (14.54), 353.03 (3.79), 327.05 (13.04), 309.04 (19.20), 298.04 (43.28), 297.04 (15.99), 285.04 (7.81), 257.04 (0.17), 255.03 (1.17), 243.03 (0.14), 241.05 (1.39), 239.04 (0.21), 229.05 (0.66), 217.05 (0.15), 213.06 (0.83), 201.02 (1.38), 165.02 (1.52), 151.00 (0.52), 133.03 (8.87), 107.01 (0.21) 593.15 (100), 473.11 (3.7), 413.09 (28.5), 383.08 (2.02), 341.06 (1.98), 311.06 (6.43), 323.06 (2.73), 293.05 (79.68) 269.05 (2.28), 117.03 (8.29)

581.1490 −1.801

581.14 (71.45), 449.10 (100), 431.09 (11.57), 395.07 (15.36), 353.06 (39.70), 329.06 (70.00), 299.05 (89.62), 297.07 (8.73), 283.05 (9.95), 165.01 (13.30), 137.02 (5.46)

2”-O-pentosyl-6-Chexosylluteolina , b , c

595.1649

595.16 (87), 475.12 (1.63), 433.11 (100), 415.10 (14.14), 397.09 (11.10), 379.08 (18.6), 367.08 (12.6), 337.07 (38.67), 313.07 (92.31), 283.06 (92.4), 271.06 (7.67), 229.05 (1.68), 163.04 (2.93), 149.02 (1.97), 145.03 (2.7), 121.03 (7.18), 119.05 (5.7) 565.15 (72.06), 433.11 (100), 415.10 (14.27), 379.08 (17.60), 349.07 (9.93), 337.07 (30.57), 313.07 (70.02), 283.05 (85.38), 267.06 (9.46), 239.07 (1.00), 195.02, 163.03 (3.24), 165.01, 121.02 (0.28) 595.16 (71.07), 577.15, 463.12 (100), 445.11 (13.82), 409.09 (19.94), 367.08 (42.13), 343. 08 (70.00), 313.07 (79.75), 301.06 (5.01), 298.04 (16.33), 195.02, 165.02 (9.87) 607.16 (100), 589.15, 487.12, 475.12 (31.16), 421.09 (14.74), 391.08 (10.91), 379.08 (56.57), 355.08 (88.27), 325.07 (75.10), 310.04 (13.92), 227.03, 151.03 (10.58), 121.02 (7.45)

2”-O-hexosyl−6-Chexosyl-apigeninc

13.

5.68

594.1585 C27 H29 O15

593.1516 2.552

14.

5.88

564.1479 C26 H27 O14

563.1406 1.897

563.14 (100), 413.08 (18.19), 353.06 (2.74), 341.06 (8.26), 311.05 (9.22), 295.06 (3.16), 293.04 (65.93), 283.06 (6.07), 269.04 (4.11), 239.07 (0.17), 135.04 (0.11), 117.03 (5.22)

565.1545 −0.701

15.

6.10

594.1585 C27 H29 O15

593.1511 1.860

593.1512 (100), 443.09 (14.98), 323.05 (43.93), 308.03(12.21), 299.05 (2.06), 285.03 (0.24), 284.03 (1.64), 257.41 (0.22), 256.04 (0.28)

595.1650 −2.209

16.

6.41

606.1585 C28 H29 O15

605.1512 1.824

605.15 (100), 413.11 (17.25), 473.11 (3.75), 353.06 (4.60), 341.06 (4.12), 323.05 (6.62), 311.05 (5.26), 294.10 (5.74), 293.04 (75.77), 269.04 (4.24), 175.00 (7.55), 117.03 (5.35)

607.1649 −1.460

6-C−7-Odiglucosylapigenin (Saponarin)* ,a ,b

2”-O-pentosyl-6-Chexosyl-apigenina , c

2”-O-pentosyl−6C-hexosylmethylluteolina

2”-Oacetylpentosyl-6C-hexosylapigenina

D. Zheleva-Dimitrova et al. / Journal of Pharmaceutical and Biomedical Analysis 155 (2018) 56–69

11.

MS/MS data m/z

Table 1 (Continued) tRmin

Theoret. mass Molecular formula

[M−H]- m/z Delta ppm

MS/MS data m/z

[M+H]+ m/z Delta ppm

MS/MS data m/z

Proposed compound

17.

6.68

636.1690 C29 H31 O16

635.1619 1.950

637.1750 −2.840

637.17 (100), 463.00 (61.68), 445.11 (10.18), 409.09 (16.99), 397.09 (10.55), 367.08 (32.55), 343.08 (70.37), 313.07 (54.45), 301.05, 298.04

2”-O-(acetylpentosyl)−6-Chexosyl-diosmetina

18.

7.11

648.1690 C30 H31 O16

647.1619 1.868

635.16 (100), 443.09 (18.46), 383.07 (4.09), 371 07 (6.21), 353.06 (5.81), 341.06 (6.70), 324.05 (5.83), 323.06 (56.32), 308.03 (15.50), 298.04 (5.39), 297.04 (5.15), 280.03 (7.93), 299.06 (2.66), 284.03 (3.04), 107.00 (0.54) 647.16 (100), 413.08 (17.11), 353.06 (30.11), 323.05 (8.23), 311.05 (4.95), 294.04 (9.07), 293.04 (78.19), 286.06 (3.11), 269.04 (53.23), 117.03 (6.72)

649.1761 −1.093

649.17 (59.02), 433.11 (47.04), 415.00 (6.50), 379.07 (4.65), 337.07 (20.31), 313.07 (50.55), 283.05 (28.16), 217.07 (7.74), 271.05 (5.19)

2”-O-diacetylpentosyl−6-Chexosyl-apigenina

448.1006 C21 H19 O11

447.0930 1.817

447.09 (100), 285.04 (87.46), 284.03 (40.05); 133.02 (5.46), 151.00 (5.03), 107.01 (1.47) 593.15 (100), 285.04 (86.89), 284.03 (50.61), 227.03 (29.91), 257.04 (2.48), 255.02 (43.01), 213.05 (1.47), 151.00 (2.24), 135.00 (1.31), 107.01 (1.44) 447.09 (100), 284.03 (57.68), 255.02 (45.99), 227 (43.57), 285.03 (26.39), 255. 03 (45.99), 243.71 (0.57), 229.05 (1.10), 151.00 (0.87), 107.01 (0.64) 477.10 (100), 315.05 (11.40), 314.04 (52.53), 300.03 (2.94), 285.04 (11.16), 271.02 (24.28), 243.02 (25.68), 227.03 (3.19), 215.03 (2.33), 199.04 (2.51), 151.0026 (1.74)

449.1073 −2.419

7-O-glucosyl-luteolin*, a , b , c

595.1650 −2.176

449.10(12.77), 287.05 (100), 200.00 (2.35), 153.00 (2.99), 137.02 (0.65), 121.02 (2.08) 595.15 (1.38), 449.10 (12.38), 287.05 (100)

449.1073 −2.419

287.05 (100), 265.26 (1.31), 171.84 (1.22)

Kaempferol−3-O-glucoside (Astragalin)*, a

479.1228 2.190

317.06 (100), 302.04 (6.98), 285.03 (2.09)

Isorhamnetin−3-Oglucoside* , a , c

271.0593 −4.864 301.0710 −0.442

271.05 (100), 243.07 (12.15), 195.05 (5.14), 153.01 (10.20), 119.04 (4.72) 301.07 (100), 286.04 (23.87), 258.05 (33.18), 229.04 (1.32)

Apigenin* , a , b , c

O-glycosyl flavonoids 6.26 19.

20.

6.54

594.1585 C27 H29 O15

593.1527 1.945

21.

6.86

448.1006 C21 H19 O11

447.0930 1.817

22.

7.00

478.1111 C22 H21 O12

477.1021 −1.347

Aglycones 23.

10.22

24.

10.51

270.0528 C15 H9 O5 300.0634 C16 H11 O6

269.0455 3.792 299.0561 3.596

* a b c

269.04 (100), 224.10 (1.22), 151.00 (6.14), 149.02 (4.41), 117.03 (16.96 299.05 (100), 284.03 (15.57), 256.04 (15.57), 227.03 (1.28), 177.05 (0.96), 151.00 (4.42) 107.01 (1.22)

Kaempferol−3-O-rutinoside* , a

Diosmetina , b , c

D. Zheleva-Dimitrova et al. / Journal of Pharmaceutical and Biomedical Analysis 155 (2018) 56–69

Peak No.

Compared with standard. G. glomerata. G. trichotoma. G. perfoliata.

61

62

D. Zheleva-Dimitrova et al. / Journal of Pharmaceutical and Biomedical Analysis 155 (2018) 56–69

the ions at m/z 329.0652 [(M+H)-120]+ (63%) (homoorientin) and 313.0706 (64%) (isovitexin). Typical ion fragments that characterize C-hexose are 0,1 X+ and 0,2 X+ corresponding to [(M+H)-150]+ and [(M+H)-120]+ , respectively [18,30]. In addition, the loss of water and the formation of m/z 431.0966 (homoorientin) and 415.1020 (isovitexin) was favored for C-6 isomers, which is consistent with Waridel’s LC–MS/MS methodology for the differentiation between C-6 and C-8 glycosyl flavones [30]. In the negative ion mode, C-8 isomers were characterized by the base peak 0,2 X− at m/z 327.0510 [(M−H)-120]− (orientin) and 311.0563 (vitexin). C-6 isomers displayed the base peaks at m/z 447.0930 [M−H]− (homoorientin) and 431.0981 (isovitexin), together with abundant fragment 0,2 X− [(M−H)-120]− (about 60%). In addition, the fragment ion 0,3 X− at m/z 357.0616 [M−H-90]− (homoorientin) and 341.06 (isovitexin) was favorite for C-6 isomers. Indeed, the most important fragmentation pattern for the aglycones luteolin and apigenin of the studied references is the RDA cleavage which generates 0,2 B+ and 1,3 B+ fragment ions. In the spectra of orientin/homoorientin and vitexin/isovitexin the 0,2 B+ at m/z 137.0234 and 121.0288, respectively, had about 10% relative abundance, while 0,2 A+ was also presented, although with less than 1% relative abundance (data not shown). Fragments 0,4 B+ at m/z 179.0336 and (0,4 B+ -H2 O) at m/z 161.0223 (luteolin) had low abundance which is consistent with a previous report [31]. Luteolin and apigenin gave a series of neutral losses of 28 Da (CO), 42 Da (C2 H2 O), 44 Da (CO2 ), 56 Da (2CO) and 72 Da (CO + CO2 ). Consistent with de Rijke, Out, Niessen, Ariese, Gooijer and Udo [31], the studied C-flavonoid standards yielded the fragment ions 1,3 A− and 1,3 B− in (−) ESI/MS [31]. In the mass spectra of orientin and vitexin 1,3 B− ions at m/z 133.0282 and 117.0332, respectively, was the most abundant fragment ion originating from the aglycones (up to 18%), while 1,3 A− at m/z 151.0022, 0,4 A− at m/z 107.0119 and 1,4 B− at m/z 165.0182 (luteolin) was found to be very low (about 1%). Based on characteristic [M+H]+ and [M−H]− ions in the HR ESI/MS analysis, MS/MS fragmentation fingerprints and comparison with the reference standards, compounds 1, 2, 3 and 4 in the studied Gypsophila extracts were identified as homoorientin, orientin, isovitexin and vitexin, respectively (Table 1). Seven isobaric flavonoids 6, 9, 10, 11, 13, 15 and 20 shared the same [М−Н]− at m/z 593.1511 (Table 1, Fig. 1 and 2). The MS/MS spectrum of 6 showed fragments 0,3 X− at m/z 503.1201 [(M−H)90]− and 0,2 X− at m/z 473.1096 [(M−H)-120]− , and abundant ions at m/z 383.0773 [(M−H)-120-90]− and 353.0666 [(M−H)-120120]− (Table 1) indicating the presence of two C-hexosyl moieties as observed for vicenin-2 [32]. The 1,3 B− ion at m/z 117.0332 indicated that ring B of the aglycone is substituted with one hydroxyl group, suggesting apigenin [31]. Considering that in the nature the C-glycosyl moieties appear exclusively at 6 and/or 8 positions of flavones [18], 6 was assigned to vicenin-2, identity further confirmed by comparison with a reference, isolated in our laboratory [21]. Regarding compound 9, typical ions of the di C-glycosyl flavones fragmentation were observed at m/z 503.1204 [(M−H)90]− (0,2 X1 − ) and 473.1096 [(M−H)-120]- (0,2 X0 − ) attributed to C-pentose and C-hexose, respectively, supported by the abundant ions at m/z 413.0884 [(M−H)-60-120]− ([0,2 X0 /0,3 X1 ]− ) and 383.0774 [(M−H)-90-120]− ([0,2 X0 /0,2 X1 ]− ) (Table 1). On the other hand, the relative abundance of the ion resulting from cross ring cleavage of the hexose on the C-glycosyl flavone unit (−120 Da, 0,2 X − ) is related to the position of C-glycosylation, more abun0 dant in 6-C than in 8-C [18]. In the (+) MS/MS spectrum of 9, a loss of 0,2 X0 + was observed together with water molecules providing abundant ions at m/z 439.1021 [(M+H)-120-2H2 O]+ and 421.0925 [(M+H)-120-3H2 O]+ . These data were indicative for 6-C-hexosyl glycosylation, while the low abundant ions 0,2 X1 − at m/z 503.1204 (8%) in (−) MS/MS and 0,2 X1 + 505.1156 [(M+H)-90]+ (3.8%) in (+)

MS/MS implied 8-C-pentosyl glycosylation (Table 1). The fragment ion at m/z 299.0560 [Agl-H]− was indicative for the aglycone. The fragmentation pattern involved the ion at m/z 284.0326 [(Agl-H)CH3 ]− suggesting the presence of methoxy group supported by the ions at m/z 256.0376 [(Agl-H)-CH3 -CO]− and 151.0023 (1,3 A− ) (Table 1). Moreover, 0,2 B+ ion at m/z 151.0389 indicating mass difference of 14 Da with that of luteolin should be attributed to the methoxylation of B-ring [33]. The exact position of the methoxy group at C-4 of the B ring was defined according to the Justesen’s key for deprotoneted methoxylated flavonoids where the loss of (CH3 + CO) and RDA cleavage yielding 1,3 A− allows the differentiation of diosmetin and its isomer [33]. These data suggested the aglycone luteolin 4 -methyl ether (diosmetin) evidenced also by comparison with reference standard. Therefore, 9 was tentatively identified as 6-C-hexosyl-8-C-pentosyl-diosmetin. Concerning compound 7 ([M−H]− at m/z 579.1355), the relative abundance of the ion 0,2 X0 − at m/z 459.0942 [M−H-120]− (19.6%) was indicative for the of 6-C-hexosyl luteolin, while 8-Cpentosyl glycosylation was evidenced by the low abundant ion 0,2 X − at m/z 489.1016 [M−H-90]− (4.7%) (Table 1) [18]. The pre1 cursor ion afforded a series of fragments at m/z 285.0418 [Agl-H]− , 243.0300 [(Agl-H)-C2 H2 O]− , 241.0500 [(Agl-H)-CO2 ]− , 229.0494 [(Agl-H)-2CO]− , 217.0497 [(Agl-H)-C3 O2 ]− , 213.0548 [(Agl-H)-COCO2 ]− , suggesting the presence of 2 hydroxyl groups in a ring B and ␤-hydroxylation in a ring A [31]. These data together with 151.0384 (1,3 A− ) and 133.0282 (1,3 B− ) were consistent with the aglycone luteolin. Thus, 7 was tentatively identified as 6-Chexosyl-8-C-pentosyl-luteolin, while compound 5 was respectively assigned to its isomer (Table 1, Fig. 2). In the same way, compound 8 ([M−H]− at m/z 563.1411) was tentatively identified as 6-Chexosyl-8-C-pentosyl-apigenin witnessed by the aforementioned fragmentation pattern of reference standards (Table 1). The identification of the aglycones luteolin (7) and apigenin (8) was also accomplished by comparison with the fragmentation of commercial standards. In the positive ion mode 11 showed a base peak at 283.0601 [(M+H)-162-150]+ and prominent fragment ion at m/z 433.1135 [(M+H)-162]+ (27%), suggesting the presence of a hexose unit Clinked to the flavonoid skeleton (−150 Da, 0,1 X0 + ) and an O-hexosyl group (−162 Da, Y1 + ) [18] (Fig. 3). Moreover, fragment ions at m/z 415.1010 [(M+H)-162-H2 O]+ , 379.0804 [(M+H)-162-3H2 O]+ , 337.0710 [(M+H)-162-60-2H2 O]+ and 313.0701 [(M+H)-162-120]+ were generated after successive losses of a hexose (−162 Da) together with one or several water molecules and intramolecular breakage of the hexose on the C-glycosyl-flavone skeleton 0,4 X0 + (−60 Da) and 0,2 X0 + (−120 Da). In the (−) MS/MS a loss of a hexose unit and (162 + 120) afforded fragment ions at m/z 431.0972 (Y1 − ) and 311.0562 ([Y1 /0,2 X0 ]− ), respectively (Table 1, Fig. 3). The signal at m/z 269.0450 pointed to the loss of two sugar residues and, at the same time, corresponded to the deprotonated aglycone apigenin. The fragmentation of [apigenin-H]− ion yielded characteristic product ions at m/z 151.0390 (1,3 A− ) and 117.0332 (1,3 B− ), supported by 121.0284 (0,2 B+ ). By comparing retention times with a standard reference together with the assignments of the MS data, compound 11 was ascribed to 6-C-7-O-diglucosyl-apigenin (saponarin). Concerning compound 10, the fragment ions at m/z 503.1197 [(M−H)-90]− (0,3 X0 − ), 473.1085 [(M−H)-120]− (0,2 X0 − ), 341.0661 [(M−H)-90-162]− ([Y1 /0,3 X0 ]− ) and 311.0551 [(M−H)-120-162]− ([Y1 /0,2 X0 ]− ) indicated one C-hexosyl and one O-hexosyl groups [18]. The C-glycosylation at 6 position was evidenced by the higher abundance of the ions at m/z 283.0596 [(M+H)-162-150]+ ([Y1 /0,1 X0 ]+ ) supported by the ion at m/z 313.0706 [(M+H)162-120]+ ([Y1 /0,2 X0 ]+ ). Thus, compound 10 was identified as a 6-C-glycosyl (isovitexin) derivative. In contrast to saponarin, a 17% relative abundance of the ion at m/z 311 (-MS/MS) and the absence

D. Zheleva-Dimitrova et al. / Journal of Pharmaceutical and Biomedical Analysis 155 (2018) 56–69

63

Fig. 3. MS/MS fragmentation of compound 11 (saponarin) in negative ion mode (A) and positive ion mode (B).

of the ion at m/z 431 indicated O-linkage of the second hexose moiety at apigenin C 4 position. Attending to these data, 10 was tentatively assigned to isosaponarin. The MS data and chromatographic behavior of 12 ([M−H]− at m/z 579.1357) allowed us to deduce a C-glycosyl flavone Oglycosylated on the sugar unit of the C-glycosylation (Table 1). The fragmentation of [M−H]− yielded the characteristic ion at m/z 459.0930 [(M−H)-120]− (0,2 X0 − ) indicating a C-hexose moiety. The elimination of a pentose unit with a water molecule from O-glycosylation provided the ion at m/z 429.0823 [(M−H)-132H2 O]− (Z1 − ). The loss of 150 Da (132 + H2 O) suggested that this additional pentose could be linked to one of the hydroxyl groups of the primary hexose [18]. The occurrence of the prominent ion at m/z 309.0404 [(M−H)-(132+H2 O)-120]− ([Z1 /0,2 X0 ]− indicated that pentose moiety is not linked to the positions 3”, 4”, or 6” since the fragment (−120 Da) implies these places [18]. On the other hand, the preferential fragmentation leading to [(M−H)-132H2 O]− is characteristic for 2”-O-pentosyl-C-glycosyl derivatives [18,27]. Indeed, the analysis of the [M+H]+ ion at m/z 581.14 showed the fragment ions at m/z 329.0662 [(M+H)-132-120]+ (70%) and 299.0549 [(M+H)-132-150]+ (90%), corresponding to the fragmentation pattern of 6-C-hexosyl-flavone (Table 1). 12 afforded fragment ions at m/z 285.0403 and 133.0281 corresponding to the [aglycone-H]− and 1,3 B− , respectively. The aforementioned fragment ions were in good agreement with the structure of luteolin, corroborated by the fragment 0,2 B+ at m/z 137.0233. Thus, 12 was tentatively assigned as 2”-O-pentosyl-6-C-hexosyl-luteolin and could be structurally related to the previously reported one from G. paniculata L. [18]. In the same way, compound 13 was tentatively assigned as 2”-Ohexosyl-6-C-hexosyl-apigenin. 13 afforded typical abundant ion at

m/z 293.0459 [(M−H)-(162+H2 O)-120] ([Z1 /0,2 X0 ]− suggesting the presence of both O- and C-glycosylation, supported by the ion at m/z 413.0883 [M−H-162-H2 O]− (Z1 − ). The (+) MS/MS spectrum of 13 was similar to that we have just described producing the abundant ions at m/z 313.0703 [(M+H)-162-120]+ and 283.0597 [(M+H)-162150]+ (Table 1). In (+) HRMS analysis, the [М+Н]+ at m/z 565.1545 (14) produced a base peak at m/z 433.1126 [(M+H)-132]+ (Y1 + ), as well as ion at m/z 415.1019 [(M+H)-132-H2 O]+ (Z1 + ), indicating the loss of a pentose unit, as was observed for 6-C-hexosyl-apigenin O-glycosylated on the sugar moiety of the C-glycosylation (Fig. 4). In addition, the MS/MS spectrum showed ions at m/z 313.0704 [(M+H)-132120]+ ([Y1 /0,2 X0 ]+ ) and 283.0598 [(M+H)-132-150]+ ([Y1 /0,1 X0 ]+ ). (Table 1, Fig. 4). This was corroborated by fragmentation features in (−) ESI-MS/MS including loss of both O- and C-linked sugars at m/z 413.0878 [(M−H)-132-H2 O]− (Z1 − ), and 293.0456 [(M−H)(132+H2 O)-120]− ([Z1 /0,2 X0 ]− ). Fragment ions from the cleavage of the aglycone C-ring were consistent with apigenin. Thus, peak 14 was assigned as 2”-O-pentosyl-6-C-hexosyl-apigenin. In the same manner, compound 15 was assigned to 2”-Opentosyl-6-C-hexosyl-methylluteolin. The formation of prominent ions at m/z 343.0810 [(M+H)-132-120]+ ([Y1 /0,2 X0 ]+ ) and 313.0704 [(M+H)-132-150]+ ([Y1 /0,1 X0 ]+ ) was favorite for the presence of 6 C-hexosyl flavones O-glycosylated on the sugar of C-glycosylation (Table 1). (−) MS/MS was afforded significant fragment at 323.0559 [(M−H)-(132+H2 O)-120]− ([Z1 /0,2 X0 ]− ) supporting the 2”-O-pentosyl moiety in 15. Concerning the aglycone, as it has indicated for 9, fragment ions at 299.0555 and 284.0316 in (−) MS/MS and 301.0687 in (+) MS/MS are relevant for the methoxy group (Table 1). The location of the last on the B ring was discerned by the ion at m/z 257.4129 resulting from the loss of C2 H2 O (42 Da) and

64

D. Zheleva-Dimitrova et al. / Journal of Pharmaceutical and Biomedical Analysis 155 (2018) 56–69

Fig. 4. MS/MS fragmentation of compound 14 (2”-O-pentosyl-6-C-hexosyl-apigenin) in negative ion mode (A) and positive ion mode (B).

RDA cleavage of ring B [33]. Based on these data the aglycone was elucidated as metylluteolin (methoxylated either on C-3 or C-4 ). The compound 16 gave [М−Н]− at m/z 605.1512 consistent with molecular formula C28 H28 O15 (calc. for C28 H29 O15 , 1.824 ppm) (Table 1). 16 produced prominent ion at m/z 413.0877 [(М−Н)-132-H2 O-Ac]− (Z1 − ), indicating the loss of an acetylated (Ac) pentose moiety. Moreover, a concomitant loss of 120 Da and (Pent + H2 O + Ac) afforded the abundant ion at m/z 293.0455 ([Z1 /0,2 X0 ]− ) supported by ion at m/z 323.0567 [(М−Н)(132 + H2 O + Ac)-90]− ([Z1 /0,3 X0 ]− ). The aglycone apigenin was evidenced owing to the fact that RDA cleavages 1,3 B− (117) and 0,2 B+ (121) are typical ions of its fragmentation pattern. Thus, peak 16 was tentatively assigned as an acetylated derivative of compound 14 or 2”-O-acetylpentosyl-6-C-hexosyl-apigenin. This hypothesis was further confirmed by the (+) MS/MS spectrum and abundant prominent ions at m/z 379.0814 [(M+H)-132-3H2 O-Ac]+ and 325.0701 [(M+H)-132-H2 O-Ac-90]+ . In the same manner, peak 18 with [М−Н]− at m/z 647.1619 produced a base peak at m/z 293.0457 [(М−Н)-132-H2 O-2Ac-120]− ([Z1 /0,2 X0 ]− ) and a prominent ion at m/z 413.0877 [(М−Н)-132H2 O-2Ac]− (Z1 − ), indicating the presence of additional acetyl group at a pentose moiety with comparison with 16. Indeed, the analysis of [M+H]+ at m/z 649.1761 showed the fragment ions at m/z 433.1132 [(M+H)-132-2Ac]+ (Y1 + ) and 313.0702 [(M+H)132-2Ac-120]+ ([Y1 /0,2 X0 ]+ that corroborate the aforementioned fragmentation pattern (Table 1, Fig. 5). 18 was tentatively assigned as diacetylated derivative of 14 or 2”-O-diacetylpentosyl-6-Chexosyl-apigenin.

The [M−H]− (17) at m/z 635.1619 was consistent with molecular formula C29 H30 O16 indicating 42 mass units more than 15. It generated fragment ions at m/z 593.1498 [(M−H)-Ac]− and 575.1439 [(M−H)-AcOH]− , suggesting the presence of acetyl group (Table 1). In addition, the fragmentation of [M−H]− yielded the characteristic product ions at m/z 443.0983 [(M−H)-132-H2 O-Ac]− (18%), and 323.0561 [(M−H)-132-H2 O-Ac-120]− (56%) indicating acetylated O-pentosyl and C-hexosyl groups as was shown for 16. Fragment ions at m/z 299.0558 [M−H]− , 284.0328 [(M−H)-15]− and 107.0068 (0,4 A− ) were favored for the presence of diosmetin witnessed also by m/z 301.0688 [M+H]+ (Table 1) [33]. (+) MS/MS spectrum gave fragment ions at m/z 343.0809 [(M+H)-132-Ac-120]+ and 313.0704 [(M+H)-132-Ac-150]+ , which is consistent with the structure of 2”-O-(acetylpentosyl)-6-C-hexosyl-diosmetin. In (−) HRMS 20 displayed the fragment ion at m/z 285.0402 [M−H-308]− (87%), while in (+) HRMS this flavonoid gave a base peak at m/z 287.0548 [M+H-308]+ suggesting elimination of a rutinose molecule. Concerning the aglycone, the fragmentation pattern yielded 0,4 A− ion at m/z 107.0124, 0,3 A− at m/z 135.0078 and 1,3 A− at m/z 151.0026. In addition, two abundant fragment ions were also formed by the losses of CH2 O at m/z 255.0200 and (CH2 O + CO) at m/z 227.0346, respectively. The concomitant loss of (CO + CO2 ) resulted in ion at m/z 213.0547 (Table 1). These data were in agreement with the structure of kaempferol [31]. Thus, compound 20 was identified as Kaempferol-3-O-rutinoside by comparing with a reference standard. In the same manner, retention times, fragmentation patterns and monoisotopic profiles of 19, 21, 23 and 24 were in good agreement with those of reference standards Luteolin-7-

D. Zheleva-Dimitrova et al. / Journal of Pharmaceutical and Biomedical Analysis 155 (2018) 56–69

65

Fig. 5. MS/MS fragmentation of compound 18 (2”-O-diacetylpentosyl-6-C-hexosyl-apigenin) in negative ion mode (A) and positive ion mode (B).

O-glucoside, Kaempferol-3-O-glucoside, Apigenin and Diosmetin, respectively. The fragmentation observed for 22 exhibited the loss of hexose moiety at m/z 315.0498 and prominent fragment at m/z 300.0259 [(Agl-CH3 ]− suggesting methoxylated aglycone. Furthermore, the fragmentation pathway involved consequent losses of (H + CO) and CO at m/z 271.0246 (24%) and 243.0296 (26%) together with minor fragments at 227.0343 and 151.0026 corresponding to RDA cleavages of B and A rings, respectively [33]. Based on these data and comparison with reference standard 22 was identified as Isorhamnetin-3-O-glucosde. Besides the above described glycosides, GGE, GTE and GPE also contained a common flavone aglycones apigenin (23) and diosmetin (24) (Table 1). C-glycosyl flavones both O-glycosylated on the phenolic hydroxyl and on the sugar moiety of the Cglycosylation were evidenced in all assayed Gypsophila species (Table 1). However, compounds 14 and 15 together with their acetylated derivatives 16, 17 and 18 were assessed only in GGE, while the di-C-glycosyl flavones 5, 6, 7, 8, and 9 were found in GPE. O-, C-glycosyl flavone 12 could be related to flavonoids 2”O-␣-l-arabinosyl-homoorientin and 2”-O-pentosyl-6-C-hexosylluteolin previously reported in G. elegans and G. paniculata, respectively [11,18]. Saponarin (11) was found in G. elegans and G. trichotoma [10,11]. Compound 15 could be associated with 2”O-pentosyl-6-C-hexosyl-methylluteolin [18], while structure of 13 was consistent with that of 2”-O-glucopyranosyl-isovitexin from G. elegans [11].

Recently, O-glycosyl-C-hexosyl derivatives of methylluteolin have been reported from both G. paniculata and G. pacifica together with the known isovitexin and isoorientin (G. pacifica), and vitexin (G. elegans) [11,18]. 2”-O-rhamnosyl/glucosyl/pentosylisovitexin has been evidenced in G. pacifica, G.elegans and G. paniculata [9,11,18]. Moreover, O-hexosyl-2”-O-pentosyl-6(8)hexosyl-apigenin derivatives have been identified in G. repens [34] and tentatively elucidated in G. paniculata [18]. To the best of our knowledge, only 6”-xylosyl-vitexin, isovitexin and isosaponarin have been reported in G. glomerata [35]. Compounds 5, 6, 7, 8, 9, 16, 17 and 18, together with the known 19, 20, 21, 22, 23 and 24 have not been previously reported in the Gypsophila genus. This is a first report on flavonoids from G. perfoliata. For the first time acetylated C- and O-glycosyl flavones are found in Gypsophila. Flavonoids have proved chemotaxonomically useful at a lower level, within families, genera and species [36]. According to the Turkish plant list, G. perfoliata var. perfoliata is a synonym of G. trichotoma [20]. In line with Bulgarian Flora, flavonoid profilings unambiguously indicated the difference between G. trichotoma var. trichotoma and G. perfoliata var. perfoliata. The presence of a variety of di C-glycosyl flavones in G. perfoliata supports the existence of different species. Within this framework, the results could be useful chemotaxonomic classification of the genus Gypsophila. 3.2. Antioxidant ability Recently, phenolic compounds are gaining interest in food and pharmaceutical areas due to theirs diverse and valued biological

66

D. Zheleva-Dimitrova et al. / Journal of Pharmaceutical and Biomedical Analysis 155 (2018) 56–69

Table 2 Total phenolics and antioxidant properties of the methanolic extracts of G. glomerata, G. trichotoma and G. perfoliata. G. glomerata

G. trichotoma

G. perfoliata

Total bioactive compounds Total phenolic content (mgGAE/g extract) Total flavonoid content (mgRE/g extract)

20.59 ± 0.12a 33.00 ± 0.55

17.07 ± 0.23 35.58 ± 0.39

21.60 ± 0.16 25.07 ± 0.49

Antioxidant assays Phosphomolybdenum (mmol TEs/g extract) DPPH scavenging (mg TEs/g extract) ABTS scavenging (mg TEs/g extract) CUPRAC (mg TEs/g extract) FRAP (mg TEs/g extract) Metal chelating (mg EDTAEs/g extract)

1.13 ± 0.08a 21.65 ± 0.42 49.19 ± 1.11 73.59 ± 1.85 49.26 ± 0.71 12.00 ± 0.39

0.97 ± 0.10 23.45 ± 0.47 40.00 ± 0.32 62.19 ± 1.19 39.37 ± 0.66 17.44 ± 0.51

1.10 ± 0.08 28.53 ± 0.62 72.59 ± 2.17 87.88 ± 1.02 48.99 ± 1.07 15.75 ± 0.07

TE: Trolox equivalent; EDTAE: EDTA equivalent; GAE: Gallic acid equivalent; RE: Rutin equivalent. a Values expressed are means ±S.D. of three parallel measurements.

effects such as antioxidant, antimicrobial, anticancer and antiinflammatory. At this point, total phenolic and flavonoid contents are considered as important parameters for evaluating biological fingerprints. Total phenolic and flavonoid content in the studied Gypsophila extracts were determined by Folin-Ciocalteu and AlCl3 methods, respectively (Table 2). GPE (21.60 mgGAE/g extract) had the higher total phenolic content compared to GGE (20.59 mgGAE/g extract) and GTE (17.07 mgGAE/g extract). This fact was also confirmed by UHLPC-HRMS results. However, total flavonoid content can be ranked as GTE (35.58 mgRE/g extract) > GGE (33.00 mgRE/g extract) > GPE (25.07 mgRE/g extract). Based on our results, the studied Gypsophila species possessed higher level of total phenolic ´ Korkmaz content than that of Gypsophila bitlisensis reported by Is¸yk, and Bursal [37]. Antioxidant abilities of the Gyposphila extracts were tested by different assays including free radical scavenging (DPPH and ABTS assays), reducing power (CUPRAC and FRAP assays), phosphomolybdenum and metal chelating assays. The results are summarized in Table 2. DPPH and ABTS assays are widely used to evaluate free radical scavenging effects of plant extracts or synthetic compounds. In these assays, free radicals are scavenged by antioxidants and then it results in an absorbance decrease. As can be seen in Table 2, GPE exhibited the highest scavenging ability on both DPPH (28.53 mgTE/g extract) and ABTS (72.59 mgTE/g extract). The observed radical scavenging abilities for GPE can be attributed to its richness in phenolic components. Similarly, Chen, Zhang, Chen, Han and Gao [38] reported a linear correlation between total phenolic content and free radical scavenging ability. Interestingly, GGE was more effective on ABTS than GTE. The obtained differences may be explained with the nature of these radicals. This approaches were confirmed by Kim, Lee, Lee and Lee [39], who reported that ABTS is considered as both hydrophilic and lipophilic system but DPPH is hydrophilic. Reducing power is considered as an important indicator of antioxidant properties. Generally, FRAP and CUPRAC assays are used to evaluate reducing abilities of plant extracts or synthetic compounds. In FRAP assay, TPTZ-Fe+3 complex is reduced by antioxidants and it results in an intense blue color complex (TPTZFe+2 complex). Similarly, the transformation of Cu+2 to Cu+ is observed in CUPRAC assay and then an intense red color is developed. The intensity of these colors developed are directly to the reducing abilities of tested samples. As shown in Table 2, GPE and GGE exerted the stronger ferric and cupric reducing abilities compared to GTE. These results may be linked to the presence of phenolic components. In fact, some authors reported that some phenolics exhibited remarkable reducing abilities [40]. Moreover, a positive correlation between total phenolic content values and reducing abilities of different plant extracts was observed in many cases [41]. Consistent with this, GTE (0.97 mmolTE/g extract) had

the lowest antioxidant ability in phoshomolybdenum assay. The assay is considered as one of total antioxidant capacity methods and it is based on the reduction of Mo (VI) to Mo (V) by antioxidants at acidic pH. Transition metals (especially iron and copper) are common catalysts of lipid peroxidation. In this sense, the chelation of these ions is a key element in the antioxidant mechanisms. To investigate the metal chelating effects of Gypsophila extracts, a ferrozine assay was conducted. In contrast to the results of FRAP and CUPRAC, the highest chelating ability was recorded in GTE with 17.44 mgEDTAE/g extract (Table 2). The observed results may be explained with the presence of non-phenolic chelators such as peptides and citric acid in GTE. This assertion is consistent with the fact that the metal chelating ability of phenolic compounds is considered as a secondary antioxidant mechanism [42]. Our findings are in agreement with Wang, Jonsdottir and Ólafsdóttir [43], who reported weak correlation between total phenolics and metal chelating abilities. As far as we know, there is very scarce data on the antioxidant effects of the genus Gypsophila [37]. From this point, the present study could be serve a new scientific basis on this genus. Taken together, the differences observed for the antioxidant properties could be related to the identified compounds or related phytochemicals in these extracts. In this context, it has been found that sugar moieties of flavonoids play important role in their biological activities [44]. For example, C-glyscosyl flavonoids exhibited stronger antioxidant abilities than O-glyscosyl flavonoids as reported by Barreca, Bellocco, Caristi, Leuzzi and Gattuso [45]. For instance, the number and position of sugars and hydroxyl groups could determine the antioxidant abilities of C-glyscosyl flavonoids [46]. In light of the above, GPE contained both monoand di C-glyscosyl flavonoids (especially di C-glycosyl flavonoids) and showed stronger antioxidant potential compared to other extracts. Vicenin-2, detected only in G. perfoliata, has previously been reported as an antioxidant [47]. In addition, mono C-glycosyl flavonoids were identified in all studied extracts which tend to justify the observed biological activities. However, it has been reported that flavonoid glycosides exhibit weaker antioxidant abilities than aglycones, but glycoside bonds tend to increase the bioavailability of these flavonoids [48]. 3.3. Enzyme inhibitory effects In millennia, some factors including modern lifestyle and oxidative stress are triggering to increase the prevalence of global health problems such as Alzheimer’s disease (AD) and diabetes mellitus (DM). For example, 415 million people worldwide were affected by DM in 2015 [49]. Moreover, AD is the main type of dementia and affects about 50 million people and it is expected to reach about 106 million by 2050 [50]. From this perspective,

D. Zheleva-Dimitrova et al. / Journal of Pharmaceutical and Biomedical Analysis 155 (2018) 56–69

67

Table 3 Enzyme inhibitory effects of the methanolic extracts of G. glomerata, G. trichotoma and G. perfoliata. Enzyme Inhibitory assays

G. glomerata

G. trichotoma

G. perfoliata

Acetylcholinesterase (mg GALAE/g extract) Butyrylcholinesterase (mg GALAE/g extract) Tyrosinase (mgKAE/g extract) ␣-amylase (mmol ACAEs/g extract) ␣- glucosidase (mmol ACAEs/g extract)

1.59 ± 0.09a na na 3.94 ± 0.19a 18.01 ± 0.02a

1.43 ± 0.06 na na 3.97 ± 0.09 12.56 ± 0.49

2.09 ± 0.02 na 8.23 ± 0.60a 3.73 ± 0.11 7.10 ± 0.02

GALAE: Galantamine equivalent; KAE: kojic acid equivalent; ACAE: Acarbose equivalent; na: not active. a Values expressed are means ±S.D. of three parallel measurements.

effective therapeutic strategies are in need to control these global problems. The inhibition of key enzymes associated with AD and DM could be targeted to alleviate the symptoms. For example, acetylcholinesterase (AChE) is a target enzyme that catalyzes the hydrolysis of acetylcholine (ACh) in the synaptic gap which results in increased concentration of ACh and thereby improving neurotransmission. Amylase and glucosidase are main enzymes in the catabolism of carbohydrates and inhibition of these enzymes can control blood glucose level in DM patients. In addition, tyrosinase is a key enzyme in the synthesis of melanin, which protects from UV radiation. At this point, tyrosinase is considered as a therapeutic target for preventing skin disorders (SD). Within the framework of the above-mentioned information, several compounds are synthetically produced as enzyme inhibitors (for example, tacrine and donepezil for AD; kojic acid and steroids for SD; acarbose and viglibose for DM). Notwithstanding, these drugs possess unfavorable effects such as gastroinstestinal disturbances and toxicity [51]. The above mentioned facts warrant the need for the discovery of natural and safe enzyme inhibitors. Hence, the enzyme inhibitory properties of the plant extracts were tested against cholinesterases (AChE and BChE), tyrosinase, ␣amylase and ␣-glucosidase. The results are depicted in Table 3. The highest AChE inhibitory effect was noted in GPE (2.09 mgGALAE/g extract), followed by GGE (1.59 mgGALAE/g extract) and GTE (1.43 mgGALAE/g extract). However, the studied extracts were not active on BChE. Again, GGE and GTE did not have any inhibitory effect on tyrosinase and GPE had remarkable anti-tyrosinase effect with 8.23 mgKAE/g extract. Moreover, ␣-amylase inhibition effects of these extracts were close to each other (3.94 mmolACAE/g extract for GGE; 3.94 mmolACAE/g extract for GTE; 3.73 mmolACAE/g extract for GPE). As for ␣-glucosidase, the highest inhibitory effect was observed in GGE (18.01 mmol ACAE/g extract), followed by GTE (12.56 mmolACAE/g extract) and GPE (7.10 mmolACAE/g extract). The observed enzyme inhibitory effects of Gypsophila extracts may be linked to their chemical profiles. With this concept, the detected anti-cholinesterase inhibitory activities for the studied Gyspohila extracts might be caused from the presence of mono-C-glycosyl flavones for example vitexin, isovitexin and orientin [52]. For example, Choi, Islam, Ali, Kim, Kim and Jung [53] reported that C- glycosylated derivatives, including vitexin, exerted potent inhibitory effects on both AChE and BChE. Also, the authors indicated that C-glycosylation at C-8 position exhibited higher AChE inhibitory effect compared to C-6 position. Accordingly, C-glycosylation of luteolin at C-6 position decreased the AChE inhibitory effect [54]. Moreover, 8-C-glycosyl-apigenin (vitexin), isolated from Nelumbo nucifera demonstrated higher cholinesterase activity compared to quercetin-3-O-glycoside [52]. However, it has been reported that 3-O-glycosylation of flavonoids, such as kaempferol or quercetin, assists the inhibition on AChE [44]. Therefore, the cholinesterase inhibition could be attributed to the complex nature of flavonoids and theirs synergistic interactions. In addition, the notable tyrosinase inhibitory activity of GPE could be attributed to di C-glycosyl flavones, detected only in this

extract [55]. It has been reported that 3-O-glycoside derivatives (kaempferol or quercetin derivatives) were not active on tyrosinase. From this point, the hydroxyl group at C-3 position is essential to tyrosinase inhibitory potential. Structure-related activity study of flavons suggested that compounds with a free C-7 hydroxyl group have higher tyrosinase inhibitory activity in comparison with 7-O-glycosides [56]. Regarding to the ␣-glucosidase inhibitory effects, it can be related to the higher concentration of phenolics in GGE [57]. The flavonoids including kaempferol-3-O-rutinoside, astragalin or isorhamnetin-3-O-glucoside could be attributed to observed remarkable glucosidase inhibitory effect of GGE. In according to our approach, these compounds were reported as glucosidase inhibitors in previous studies [58]. In addition, the flavonoid Oand C-glycosides may contribute to the ␣-glucosidase inhibitory activity and some of them were strong anti-glucosidase agents. However, the glycosylation of the flavonoids may increase molecular size and polarity, and lowered the inhibition ability on ␣-glucosidase, although the C-6 glycosylation had relatively less impact than the C-8 glycosylation [44]. Similar fact was also reported for the ␣-amylase inhibition [59]. Moreover triterpenoid saponins isolated from G. oldhamiana showed remarkable ␣-glucosidase inhibitory effects[17]. Taken together, the genus Gypsophila could be considered a potential source of natural enzyme inhibitors for treating diabetes mellitus. To the best of our knowledge, the present study is the first report regarding enzyme inhibitory effects of the studied Gypsophila species. At this point, the presented results could open new insights for developing plant-based drug and nutraceutical formulations. 4. Conclusion In conclusion, the flavonoid profiles of G. glomerata, G. trichotoma and G. perfoliata aerial parts were investigated using UHPLC-ESI/HRMS. A new approach to determine or tentatively elucidate the structures of flavonoids was proposed using Orbitrap mass spectrometer operated in both negative and positive ion modes. Acetylated O, C,-glycosyl flavones and di-C-glycosides were found for the first time in Gypsophila species. The obtained results highlight the studied Gypsophila species’ aerial parts as new sources of bioactive agents including antioxidants, acetylcholinesterase and ␣-glucosidase inhibitors. Further isolation and structural elucidation of the newly described acetylated flavonoids are needed for more thorough understanding of the in vivo activities of G. trichotoma, G. glomerata and G. perfoliata extracts. References [1] S. Böttger, M.F. Melzig, Triterpenoid saponins of the Caryophyllaceae and Illecebraceae family, Phytochem. Lett. 4 (2) (2011) 59–68. [2] M. Henry, Saponins and phylogeny: example of the gypsogenin group saponins, Phytochem. Rev. 4 (2) (2005) 89–94. [3] Q. Chen, J.-G. Luo, L.-Y. Kong, Triterpenoid saponins from Gypsophila altissima L, Chem. Pharm. Bull. 58 (3) (2010) 412–414.

68

D. Zheleva-Dimitrova et al. / Journal of Pharmaceutical and Biomedical Analysis 155 (2018) 56–69

[4] Q.F. Huang, S.J. Zhang, L. Zheng, M. Liao, M. He, R. Huang, L. Zhuo, X. Lin, Protective effect of isoorientin-2 -O-␣-l-arabinopyranosyl isolated from Gypsophila elegans on alcohol induced hepatic fibrosis in rats, Food Chem. Toxicol. 50 (6) (2012) 1992–2001. [5] M. Korkmaz, H. Özc¸elik, Economic importance of Gypsophila L., Ankyropetalum Fenzl and Saponaria L. (Caryophyllaceae) taxa of Turkey, Afr. J. Biotechnol. 10 (47) (2011) 9533–9541. [6] R. Gevrenova, L. Voutquenne-Nazabadioko, D. Harakat, E. Prost, M. Henry, Complete 1H-and 13C NMR assignments of saponins from roots of Gypsophila trichotoma Wend, Magn. Reson. Chem. 44 (7) (2006) 686–691. [7] L. Voutquenne-Nazabadioko, R. Gevrenova, N. Borie, D. Harakat, C. Sayagh, A. Weng, M. Thakur, M. Zaharieva, M. Henry, Triterpenoid saponins from the roots of Gypsophila trichotoma Wender, Phytochemistry 90 (2013) 114–127. [8] R. Gevrenova, A. Weng, L. Voutquenne-Nazabadioko, M. Thakur, I. Doytchinova, Quantitative structure–activity relationship study on saponins as cytotoxicity enhancers, Lett. Drug Des. Discovery 12 (3) (2015) 166–171. [9] W. Nie, J.-G. Luo, L.-Y. Kong, C-Glycosylflavonoids from the aerial part of Gypsophila pacifica, Chin. J. Nat. Med. 8 (04) (2010) 250–252. [10] V. Vitcheva, R. Simeonova, I. Krasteva, M. Yotova, S. Nikolov, M. Mitcheva, Hepatoprotective effects of saponarin, isolated from Gypsophila trichotoma Wend. on cocaine-induced oxidative stress in rats, Redox Rep. 16 (2) (2011) 56–61. [11] F.M. Zhang, Z.G. Tai, L. Cai, Y.B. Yang, F. Li, Z.T. Ding, Flavonoids from Gypsophila elegans and their antioxidant activities, J. Yunnan Unive. (Nat. Sci. Ed.) 33 (2011) 93–95. [12] X. Lin, Y. Chen, S. Lv, S. Tan, S. Zhang, R. Huang, L. Zhuo, S. Liang, Z. Lu, Q. Huang, Gypsophila elegans isoorientin attenuates CCl 4-induced hepatic fibrosis in rats via modulation of NF-␬B and TGF-␤1/Smad signaling pathways, Int. Immunopharmacol. 28 (1) (2015) 305–312. [13] R. Simeonova, M. Kondeva-Burdina, V. Vitcheva, I. Krasteva, V. Manov, M. Mitcheva, Protective effects of the apigenin-O/C-diglucoside saponarin from Gypsophila trichotoma on carbone tetrachloride-induced hepatotoxicity in vitro/in vivo in rats, Phytomedicine 21 (2) (2014) 148–154. [14] R. Simeonova, V. Vitcheva, M. Kondeva-Burdina, I. Krasteva, V. Manov, M. Mitcheva, Hepatoprotective and antioxidant effects of saponarin, isolated from Gypsophila trichotoma wend. on paracetamol-induced liver damage in rats, BioMed Res. Int. 2013 (2013). [15] X. Lin, J. Wei, Y. Chen, P. He, J. Lin, S. Tan, J. Nie, S. Lu, M. He, Z. Lu, Isoorientin from Gypsophila elegans induces apoptosis in liver cancer cells via mitochondrial-mediated pathway, J. Ethnopharmacol. 187 (2016) 187–194. [16] N.K. Chima, L. Nahar, R.R. Majinda, S. Celik, S.D. Sarker, Assessment of free-radical scavenging activity of Gypsophila pilulifera: assay-guided isolation of verbascoside as the main active component, Rev. Bras. Farmacogn. 24 (1) (2014) 38–43. [17] J.-G. Luo, L. Ma, L.-Y. Kong, New triterpenoid saponins with strong ␣-glucosidase inhibitory activity from the roots of Gypsophila oldhamiana, Bioorg. Med. Chem. 16 (6) (2008) 2912–2920. [18] F. Ferreres, A. Gil-Izquierdo, P.B. Andrade, P. Valentão, F. Tomás-Barberán, Characterization of C-glycosyl flavones O-glycosylated by liquid chromatography–tandem mass spectrometry, J. Chromatogr. A 1161 (1) (2007) 214–223. [19] P. Davis, R. Mill, K. Tan (Eds.), Flora of Turkey and the East Aegean Islands, vol. 10, Edinburgh University Press, Edinburgh, 1988. [20] A. Güner, Türkiye bitkileri listesi:(damarlı bitkiler), ANG Vakfı, Istanbul, 2012. [21] R. Gevrenova, D. Zheleva-Dimitrova, S. Ruseva, N. Denkov, S. Konstantinov, V. Lozanov, V. Mitev, Cytotoxic and hepatoprotective effects of Bupleurum flavum flavonoids on hepatocellular carcinoma HEP-G2 cells, Br. J. Pharm. Res. 11 (4) (2016). [22] G. Zengin, A. Aktumsek, R. Ceylan, S. Uysal, A. Mocan, G.O. Guler, M.F. Mahomoodally, J. Glamoclija, A. Ciric, M. Sokovic, Shedding light on the biological and chemical fingerprints of three Achillea species (A. biebersteinii, A. millefolium and A. teretifolia), Food Funct. 8 (3) (2017) 1152–1165. [23] K. Slinkard, V.L. Singleton, Total phenol analysis: automation and comparison with manual methods, Am. J. Enol. Viticult. 28 (1) (1977) 49–55. [24] G. Zengin, C. Sarikurkcu, A. Aktumsek, R. Ceylan, O. Ceylan, A comprehensive study on phytochemical characterization of Haplophyllum myrtifolium Boiss. endemic to Turkey and its inhibitory potential against key enzymes involved in Alzheimer, skin diseases and type II diabetes, Ind. Crops Prod. 53 (2014) 244–251. [25] G. Zengin, A study on in vitro enzyme inhibitory properties of Asphodeline anatolica: new sources of natural inhibitors for public health problems, Ind. Crops Prod. 83 (2016) 39–43. [26] A. Mari, P. Montoro, G. D’Urso, M. Macchia, C. Pizza, S. Piacente, Metabolic profiling of Vitex agnus castus leaves, fruits and sprouts: analysis by LC/ESI/(QqQ) MS and (HR) LC/ESI/(Orbitrap)/MS n, J. Pharm. Biomed. Anal. 102 (2015) 215–221. [27] O.R. Pereira, A.M. Silva, M.R. Domingues, S.M. Cardoso, Identification of phenolic constituents of Cytisus multiflorus, Food Chem. 131 (2) (2012) 652–659. [28] Y. Ma, Q. Li, H. Van den Heuvel, M. Claeys, Characterization of flavone and flavonol aglycones by collision-induced dissociation tandem mass spectrometry, Rapid Commun. Mass Spectrom. 11 (12) (1997) 1357–1364. [29] B. Domon, C.E. Costello, A systematic nomenclature for carbohydrate fragmentations in FAB-MS/MS spectra of glycoconjugates, Glycoconj. J. 5 (4) (1988) 397–409.

[30] P. Waridel, J.-L. Wolfender, K. Ndjoko, K.R. Hobby, H.J. Major, K. Hostettmann, Evaluation of quadrupole time-of-flight tandem mass spectrometry and ion-trap multiple-stage mass spectrometry for the differentiation of C-glycosidic flavonoid isomers, J. Chromatogr. A 926 (1) (2001) 29–41. [31] E. de Rijke, P. Out, W.M. Niessen, F. Ariese, C. Gooijer, A.T. Udo, Analytical separation and detection methods for flavonoids, J. Chromatogr. A 1112 (1–2) (2006) 31–63. [32] M.J. Simirgiotis, J. Bórquez, G. Schmeda-Hirschmann, Antioxidant capacity, polyphenolic content and tandem HPLC–DAD–ESI/MS profiling of phenolic compounds from the South American berries Luma apiculata and L. Chequón, Food Chem. 139 (1) (2013) 289–299. [33] U. Justesen, Collision-induced fragmentation of deprotonated methoxylated flavonoids, obtained by electrospray ionization mass spectrometry, J. Mass Spectrom. 36 (2) (2001) 169–178. [34] M. Elbandy, T. Miyamoto, M.-A. Lacaille-Dubois, Sulfated lupane triterpene derivatives and a flavone C-glycoside from Gypsophila repens, Chem. Pharm. Bull. 55 (5) (2007) 808–811. [35] V.N. Darmograi, P.E. Krivenchuk, V.I. Litvinenko, Flavonoids of (Gypsophila paniculata L.), Farmatsiia 18 (2) (1969) 30–32. [36] R.J. Grayer, M.W. Chase, M.S. Simmonds, A comparison between chemical and molecular characters for the determination of phylogenetic relationships among plant families: an appreciation of Hegnauer’s Chemotaxonomie der Pflanzen, Biochem. Syst. Ecol. 27 (4) (1999) 369–393. [37] M. Is¸ık, M. Korkmaz, E. Bursal, Determination of antioxidant properties of Gypsophila bitlisensis, Int. J. Pharmacol. 11 (2015) 366–371. [38] G.-L. Chen, X. Zhang, S.-G. Chen, M.-D. Han, Y.-Q. Gao, Antioxidant activities and contents of free, esterified and insoluble-bound phenolics in 14 subtropical fruit leaves collected from the south of China, J. Funct. Foods 30 (2017) 290–302. [39] D.-O. Kim, K.W. Lee, H.J. Lee, C.Y. Lee, C. Vitamin, equivalent antioxidant capacity (VCEAC) of phenolic phytochemicals, J. Agric. Food Chem. 50 (13) (2002) 3713–3717. [40] K. Sevgi, B. Tepe, C. Sarikurkcu, Antioxidant and DNA damage protection potentials of selected phenolic acids, Food Chem. Toxicol. 77 (2015) 12–21. [41] T.K. Koley, C. Kaur, S. Nagal, S. Walia, S. Jaggi, Antioxidant activity and phenolic content in genotypes of Indian jujube (Zizyphus mauritiana Lamk.), Arab. J. Chem. 9 (2016) 1044–1052. [42] C.A. Rice-Evans, N.J. Miller, G. Paganga, Structure-antioxidant activity relationships of flavonoids and phenolic acids, Free Radic. Biol. Med. 20 (7) (1996) 933–956. [43] T. Wang, R. Jonsdottir, G. Ólafsdóttir, Total phenolic compounds, radical scavenging and metal chelation of extracts from Icelandic seaweeds, Food Chem. 116 (1) (2009) 240–248. [44] J. Xiao, Dietary flavonoid aglycones and their glycosides: which show better biological significance? Crit. Rev. Food Sci. Nutr. 57 (9) (2017) 1874–1905. [45] D. Barreca, E. Bellocco, C. Caristi, U. Leuzzi, G. Gattuso, Distribution of C-and O-glycosyl flavonoids, (3-hydroxy-3-methylglutaryl) glycosyl flavanones and furocoumarins in Citrus aurantium L. juice, Food Chem. 124 (2) (2011) 576–582. [46] P. Zeng, Y. Zhang, C. Pan, Q. Jia, F. Guo, Y. Li, W. Zhu, K. Chen, Advances in studying of the pharmacological activities and structure–activity relationships of natural C-glycosylflavonoids, Acta Pharm. Sin. B 3 (3) (2013) 154–162. [47] M. Satyamitra, S. Mantena, C. Nair, S. Chandna, B. Dwarakanath, U. Devi, The antioxidant flavonoids, orientin and vicenin enhance repair of radiation-induced damage, SAJ Pharm. Pharmacol. 1 (1) (2014) 1–9. [48] S. Kumar, A.K. Pandey, Chemistry and biological activities of flavonoids: an overview, Sci. World J. 2013 (2013) 1–16. [49] IDF, IDF Diabetes Atlas, seventh edition, 2015. [50] M.J. Prince, World Alzheimer Report 2015: the global impact of dementia: an analysis of prevalence, incidence, cost and trends, 2015. [51] N. Martins, I.C. Ferreira, Neurocognitive Improvement Through Plant Food Bioactives: A Particular Approach to Alzheimer’s Disease, Food Bioactives, Springer, 2017, pp. 267–298. [52] H.A. Jung, S. Karki, J.H. Kim, J.S. Choi, BACE1 and cholinesterase inhibitory activities of Nelumbo nucifera embryos, Arch. Pharmacal Res. 38 (6) (2015) 1178–1187. [53] J.S. Choi, M.N. Islam, M.Y. Ali, E.J. Kim, Y.M. Kim, H.A. Jung, Effects of C-glycosylation on anti-diabetic, anti-Alzheimer’s disease and anti-inflammatory potential of apigenin, Food Chem. Toxicol. 64 (2014) 27–33. [54] J.S. Choi, M.N. Islam, M.Y. Ali, Y.M. Kim, H.J. Park, H.S. Sohn, H.A. Jung, The effects of C-glycosylation of luteolin on its antioxidant, anti-Alzheimer’s disease, anti-diabetic, and anti-inflammatory activities, Arch. Pharmacal Res. 37 (10) (2014) 1354–1363. [55] A. Morgan, M.N. Jeon, M.H. Jeong, S.Y. Yang, Y.H. Kim, Chemical components from the stems of Pueraria lobata and their tyrosinase inhibitory activity, Nat. Prod. Sci. 22 (2) (2016) 111–116. [56] Y.-J. Kim, H. Uyama, Tyrosinase inhibitors from natural and synthetic sources: structure, inhibition mechanism and perspective for the future, Cell. Mol. Life Sci. 62 (15) (2005) 1707–1723. [57] M.A. Alam, I. Zaidul, K. Ghafoor, F. Sahena, M. Hakim, M. Rafii, H. Abir, M. Bostanudin, V. Perumal, A. Khatib, In vitro antioxidant and, ␣-glucosidase inhibitory activities and comprehensive metabolite profiling of methanol

D. Zheleva-Dimitrova et al. / Journal of Pharmaceutical and Biomedical Analysis 155 (2018) 56–69 extract and its fractions from Clinacanthus nutans, BMC Complement. Altern. Med. 17 (1) (2017) 181. [58] J. Xiao, G. Kai, K. Yamamoto, X. Chen, Advance in dietary polyphenols as (-glucosidases inhibitors: a review on structure-activity relationship aspect, Crit. Rev. Food Sci. Nutr. 53 (8) (2013) 818–836.

[59] Y. Xiao-Ping, S. Chun-Qing, Y. Ping, M. Ren-Gang, ␣-Glucosidase and ␣-amylase inhibitory activity of common constituents from traditional Chinese medicine used for diabetes mellitus, Chin. J. Nat. Med. 8 (5) (2010) 349–352.

69