Comparative analysis of three Callicarpa herbs using high performance liquid chromatography with diode array detector and electrospray ionization-trap mass spectrometry method

Journal of Pharmaceutical and Biomedical Analysis 75 (2013) 239–247

Contents lists available at SciVerse ScienceDirect

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

Comparative analysis of three Callicarpa herbs using high performance liquid chromatography with diode array detector and electrospray ionization-trap mass spectrometry method Yatao Shi a , Chunyong Wu b , Yanhua Chen a , Wenyuan Liu b,∗ , Feng Feng a,∗ , Ning Xie c a

Department of Natural Medicine Chemistry, China Pharmaceutical University, Nanjing, China Department of Pharmaceutical Analysis, China Pharmaceutical University, Nanjing, China c Jiangxi Qingfeng Pharmaceutical Ltd., Ganzhou, China b

a r t i c l e

i n f o

Article history: Received 27 September 2012 Received in revised form 25 November 2012 Accepted 27 November 2012 Available online 7 December 2012 Keywords: Callicarpa nudiflora Hook. et Arn. Callicarpa macrophylla Vahl. Callicarpa kwangtungensis Chun. Phenylpropanoid glycosides and flavonoids HPLC–DAD–ESI-Trap MS

a b s t r a c t Three Callicarpa species, namely Callicarpa nudiflora Hook. et Arn., Callicarpa macrophylla Vahl. and Callicarpa kwangtungensis Chun. are astringency and hemostasis herbs in the traditional Chinese medical systems. Despite their wide use in Chinese medicine, no report on system comparison on their chemical constituents is available so far. High-performance liquid chromatography coupled with diode array detector and electrospray ionization trap mass spectrometry (HPLC–DAD–ESI-Trap MS) technique was used for qualitative and quantitative analyses of the three Callicarpa herbs. Phenylpropanoid glycosides, flavonoids and organic acids were identified by comparing with reference standards or according to their MS/MS fragmentation behaviors. A total of 33 compounds were identified identified or tentatively identified, and 23 of them were reported from these herbs for the first time. Phenylpropanoid glycosides were featured in the three species with their types and contents presenting significant differences. Furthermore, quantitative analysis was conducted by determining four marker phenylpropanoid glycosides (forsythoside B (14), acteoside (15), poliumoside (19), isoacteoside (21)) and two flavonoids (luteolin (30), apigenin (32)). Three flavonoid glucuronides (luteolin-diglucuronide-glucuronide (5), luteolin-diglucuronide (12), apigenin-7-O-␤-glucuronide (24)) were semi-quantified according to their corresponding aglycones. The total contents of the nine major compounds in the three species varied significantly from 8.92 to 40.89 mg/g. © 2012 Elsevier B.V. All rights reserved.

1. Introduction The genus Callicarpa is comprised of 40 more species, about 20 of which have reported ethnomedical uses, and several members of this genus are well known in the traditional medical systems of China and South Asia [1]. Among them, Callicarpa nudiflora Hook. et Arn., Callicarpa macrophylla Vahl. and Callicarpa kwangtungensis Chun. are the common three species in China. All the three Callicarpa species are described as antibiosis, antiphlogosis and hemostasis drugs in Chinese history, which has been confirmed by modern research. Fu et al. [2] reported that C. nudiflora Hook. et Arn. had antimicrobial activities, decreased mice’s abdominal capillary permeability and auricular tumefaction and shorten bleeding and blood-clotting time. C. nudiflora Hook. et

∗ Corresponding authors at: Department of Pharmaceutical Analysis in China Pharmaceutical University, Tongjiaxiang 24, Nanjing 210009, China. Tel.: +86 25 83271038. E-mail addresses: [email protected] (W. Liu), [email protected] (F. Feng). 0731-7085/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jpba.2012.11.038

Arn. is also used in combination with other herbs in preparations to stop internal and external bleeding and to treat diarrhea, dysentery, intestinal worms, skin disorders and dermatitis [3,4]. The increasing application of C. nudiflora Hook. et Arn. has caused the resource shortage in recent years. Other two species, C. macrophylla Vahl. and C. kwangtungensis Chun., are officially recorded in current Chinese Pharmacopoeia (2010 edition) [5,6]. Zhou et al. [7] proved the bacterostasis and hemostasis activities of C. kwangtungensis Chun. extract. Compared with the research data regarding C. nudiflora Hook. et Arn. and C. kwangtungensis Chun., the pharmacological and chemical reports on C. macrophylla Vahl. are relatively fewer. As no system comparison among them was clarified, the three species were often confused in the market. Phytochemical studies on Callicarpa species have been reported [8,9], and the presence of flavonoids, essential oils and terpenoids has been substantiated by detection or isolation of members of these compound classes. Some analytical methods have been established to evaluate the quality of these three Callicarpa species. In Chinese Pharmacopoeia, high-performance liquid chromatography (HPLC) method was used for assay of only verbascoside in C.

240

Y. Shi et al. / Journal of Pharmaceutical and Biomedical Analysis 75 (2013) 239–247

macrophylla Vahl., and both forsythoside B and poliumoside in C. kwangtungensis Chun. Zhang et al. [10] determined the oleanolic acid and ursolic acid in C. nudiflora Hook. et Arn. Hu et al. [11] evaluated the quality of C. nudiflora Hook. et Arn. by determining the quantity of luteolin. Pan et al. [12] established a HPLC method for determination of betulinic acid in C. macrophylla Vahl. Considering the quality assessment, current methods available for analysis of Callicarpa species only focused on one or two compounds, which is difficult to reflect the chemical profiles of herbs. So far, no report on system comparison of their chemical constituents is available, and as a result, no effective method can be applied to simultaneously evaluate the quality of these three Callicarpa species. Thus, the aim of the study is to compare the chemical composition of the three Callicarpa species qualitatively and quantitatively. Based on hyphenated techniques, a new HPLC coupled with diode array detector and electrospray ionization trap mass spectrometry (HPLC–DAD–ESI-Trap MS) method was established and validated. The typical compounds in the three Callicarpa species were characterized and compared, and their fragmentation pathways were also discussed. 2. Experimental 2.1. Chemicals, reagents and herbal materials Reference standards, forsythoside B (14), acteoside (15), poliumoside (19), isoacteoside (21), luteolin (30) and apigenin (32) were purchased from the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). Samioside (16) and rhamnatin (33) (Fig. 1) was isolated from C. nudiflora Hook. et Arn. in our laboratory [13]. All reference standards used in this study showed purities of >97% by HPLC analysis. HPLC-grade methanol, formic acid (J.T. Baker, Phillipsburg, NJ, USA), and ultra-pure water were used for all analyses. The other solvents were from Nanjing Chemical Corporation (Nanjing, China). Herbs of C. nudiflora Hook. et Arn. were collected from authenticated locations in Guangxi Province; C. macrophylla Vahl. and C. kwangtungensis Chun. were collected from Ganzhou, Jiangxi Province in China. The herbal materials were authenticated by the Professor Feng Feng. Specimens were deposited at School of Traditional Chinese Medicine, China Pharmaceutical University. 2.2. Sample preparation Dried raw material was ground and passed through 40 mesh screen. An aliquot of 1.0 g powder was accurately weighed and extracted with 10 mL of 70% methanol by thermal refluxing for 45 min. The extract was cooled at room temperature and the weight loss of solvent during the extract procedure was compensated. Then, the extract was centrifuged at 16,000 rpm for 5 min. The supernatant was filtered through 0.45 ␮m membrane before use. An aliquot of 20 ␮L was injected for analyses. 2.3. HPLC–DAD–ESI-Trap MS system HPLC analysis was performed on a Shimadzu series 2010 HPLC instrument (Kyoto, Japan) equipped with a quaternary pump, a diode array detector, an autosampler and a column compartment. Samples were separated on a Merges C18 column (250 mm × 4.6 mm, 5 ␮m, Hanbon, China). The mobile phase was consisted of methanol (A) and 0.2% formic acid solution (v/v, B). The following gradient elution program was used: 5–15% A for 0–5 min, 15–16% A for 5–8 min, 16–32%A for 8–12 min, 32–52%A for 12–45 min and 52–100% A for 45–65 min. The flow rate was 1.0 mL/min. The column temperature was set to 30 ◦ C. The UV

detector was monitored at 280 nm. The data were processed with Shimadzu 2.0 ChemStation software. HPLC–ESI-Trap MS for qualitative analysis was performed by an Agilent 1100 series LC–MSD Trap SL mass spectrometer with an electrospray interface (ESI) (CA, USA). The mass spectrometer was operated in the negative ion mode with a full scan mass spectra over the m/z range 200–1000, using a cycle time of 1s, a capillary voltage of −3.5 kV, a capillary exit voltage of 165 V, a dry temperature of 300 ◦ C, a high purity nitrogen dry gas of 7.0 L/min, a nitrogen nebulizer pressure of 30.0 psi and a dwell time of 200 ms. The UV spectrum of each compound was collected between 200–400 nm. All the operations, acquisition and analysis of data were controlled by Agilent Mass Hunter Acquisition Software (Ver. A.01.00). 2.4. Quantification and method validation Appropriate amounts of reference standards of forsythoside B (14), acteoside (15), poliumoside (19), isoacteoside (21), luteolin (30) and apigenin (32) were weighed accurately and dissolved separately in methanol to make their stock solutions (1.0 mg/mL). Different volume of each stock solution was transferred to an Eppendorf tube and mixed to make a mixed stock solution. A volume of 100, 80, 50, 30 and 10 ␮L of the mixed stock solution was transferred to an Eppendorf tube, respectively, and diluted to 500 ␮L with 70% methanol to obtain five serial working solutions. An aliquot of 20 ␮L of each working solution was injected into the HPLC instrument for analysis. Five point calibration solutions were determined for construction of calibration curves. Each mixed standard solution was injected in triplicate. Calibration curves were established by plotting the peak area versus concentration (␮g/mL) of each analyte. The LOD and LOQ for each standard were defined at signal-to-noise ratio (S/N) of 3 and 10, respectively. The intra-day variation was determined by analyzing the six replicates on the same day and inter-day variation was determined on three consecutive days. To confirm the repeatability, six independently prepared samples were analyzed. The contents of 6 compounds in herbs were calculated from the corresponding calibration curves. A sample solution was separately tested at 0, 3, 6, 9, 12 and 24 h for appraising stability. The recovery test was determined by standard addition method. A certain amount of herb sample was spiked with the mixed standard solution. The mixture was processed and analyzed under the conditions mentioned above, and six replicates were performed for the analysis. 3. Results and discussion 3.1. Optimization of sample extraction and HPLC conditions In order to extract efficiently the chemical constituents from the herbs, the extraction method (ultrasonic bath, reflux), extraction solvent (30, 50, 70, 100 methanol, v/v), solvent volume (10, 20 and 40 mL) and extraction duration (15, 30, 45 and 60 min) were optimized. By comparing the numbers and areas of characteristic peaks in each chromatogram of different conditions. The optimal extraction conditions were as follow: extracting by thermal refluxing with 10 mL of 70% methanol for 45 min. For HPLC separation, different C18 columns (Agilent Eclipse plus, Agilent SB, Waters Atlantis, Hanbang Merges C18 and YMC ODSA), different mobile phases (acetonitrile-water, methanol–water, acetonitrile–acid aqueous solution, methanol–acid aqueous solution), different detection wavelength (254, 280, 330 and 360 nm) and different column temperatures (25, 30, 35 and 40 ◦ C) were compared. The best chromatographic resolution was achieved by separating samples on a Merges C18 column using methanol (A)

Y. Shi et al. / Journal of Pharmaceutical and Biomedical Analysis 75 (2013) 239–247

241

OH CH2 HO

O

COOH

OH

COOH CH2

CH2OR1 OO OR3

OH

Glc Api H H Rha H Rha H H

R2 H H H H H H Ac Ac H

HO

Rha Rha Rha Api--Rha Rha Rha Rha Rha Rha

R4 H H H H H CH3 H H CH3

R5 H H H H H H H H CH3

(6) (7) (9) (11) (13) (18)

R2

H H H CH3 CH3 H

Api H Rha Api H H

R3O

R4 OH

H H H H H Ac

R1

(17) OH

O OO ORha

HO

O

R3

O HO

OR1

OR1

OR3

R1

R2

OH

O O

OR2

R3

Cistanoside F (4) OH

CH2OR2 OO ORha

HO

O

R1 (10) (14) (15) (16) (19) (22) (23) (25) (26)

O

Decaffeoylacteoside (3)

Cinnamic acid (2) OR5

O

HO

OH

Citric acid (1)

HO

OH

OH

COOH

R4O

HO

CH2OH OO ORha

CH2OH O ORha OH O OH

OH R1

OR2 OH

R1

R2

(8)

OH

Api

(20) (21)

OCH3 H

H H

H

O R2

R3

H

di-GlcA

H

R4

H

(24)

H

GlcA

H

(28)

H

OGlcA

H

H

(29)

H

OCH3

GlcA

(30) (31) (32) (33)

H CH3 H H

OH H H OCH3

H GlcA H CH3

H H H H OH

Fig. 1. Chemical structures of 33 compounds (20 phenylpropanoid glycosides, 11 flavones and 2 organic acids) identified from the three Callicarpa species. (Glc, glucosyl; Xyl, xylosyl; Api, apiosyl; Rha, rhamnosyl.)

and 0.2% formic aqueous acid (v/v, B) as mobile phase in a gradient elution at 30 ◦ C. The typical chromatograms for three Callicarpa species were shown in Fig. 2.

3.2. Characterization of chemical constituents of three Callicarpa species The chemical constituents of the three Callicarpa species were characterized by the developed HPLC–DAD–ESI-Trap MS method. Because most target compounds showed higher response in negative ion mode than positive mode, negative ion mode was employed in this study, while positive mode was employed when necessary to provide complementary information. A total of 33 compounds, including 20 phenylpropanoid glycosides, 11 flavones and 2 organic acids were identified identified or tentatively identified from three Callicarpa species (Table 1). Their chemical structures are shown in Fig. 1. Among them, eight major compounds (14, 15, 16, 19, 21, 30, 32 and 33) were unambiguously identified by comparing with their reference standards, while the other twenty-five compounds were characterized according to their UV, MS spectra and literatures. Twenty-three of them were reported from the three Callicarpa species for the first time. Phenylpropanoid glycosides featured in Callicarpa species, and 20 phenylethanol glycosides were identified in the three species. The molecular weight of the compounds were determined by the predominant [M−H]− ion in full-scan mass spectra. Further structural information was obtained by analyzing MS spectra and by referring to previous reports on mass spectrometry fragmentation behaviors of phenylethanol glycosides [14]. Based on the

substitution conditions, in this paper they are divided into two types for discussion, named acteoside type (3–4, 10, 14–16, 19, 21, 22–23 and 25) and ␤-substituted acteoside type (6, 7, 8, 9, 11, 13, 18 and 20) phenylethanol glycosides.

3.3. Structures elucidation of acteoside type phenylethanol glycosides Compound 15 with its [M−H]− ion at m/z 623 was identified as acteoside by comparing the retention time and MS/MS data with that of reference substance. The [M−H]− ion lost a caffeoyl group to produce fragment at m/z 461[M−H–caffeoyl]− , lost a H2 O and a CO2 to produce ions of m/z 161 and 135, respectively. The typical fragment at m/z 153 [phenethanol]− is the characteristic for the acteoside type phenylethanol glycosides. This fragmentation pathways are useful for structural elucidation of other compounds. Similar UV absorption spectra and MS/MS fragment profiles were observed in compound 10, 14, 15, 16, 19, 21, 22, 23 and 25, indicating that they possessed similar acteoside type skeleton. Compound 14, 16, 19 and 21 were identified as forsythoside B, samioside, poliumoside and isoacteoside, respectively, by comparing with their reference standards. Compound 10 showed a [M−H]− ion at m/z 785, which is more 162 u than that of 15, suggesting the presence of a glucosyl moiety group. It is characterized as echinacoside based on the literature [14]. Compound 23 produced a [M−H]− ion at m/z 811, which is more 42 u than that of compound 19. Its product ions at m/z 769 [M−CH3 CO]− , 769, 607, 179, 161, 153 and 135, suggesting that 23 is the acetylate of compound 19, named brandioside. Compound 25 was acetylate of 15 (acteoside), named

242

Y. Shi et al. / Journal of Pharmaceutical and Biomedical Analysis 75 (2013) 239–247

Table 1 Identification of 33 compounds from the three Callicarpa species by developed HPLC–DAD–ESI-Trap MS method. tR (min)

UV max (nm)

[M−H]− (m/z)

MS/MS fragments (m/z)

Identification

Callicarpa species

a

1

6.60

210, 286

191 [M−H]−

191 [M−H]− 111 [M−H–2H2 O–CO2 ]−

Citric acid

I, II, III

2a

7.60

220, 276

147 [M−H]−

147 [M−H]− 103 [M−H–CO2 ]−

Cinnamic acid

I, III

3a

15.90

232, 280

461 [M−H]− 923[2M−H]−

461[M−H]− 315[M−H–Rha]− 297[M−H–Rha–H2 O]− 153[M−H–Rha–Glu]− 135[M−H–Rha–Glu–H2 O]−

Decaffeoylacteoside

I, II, III

4a

17.00

242, 330

487 [M−H]− 975 [2M−H]−

487 [M−H]− 179, 161, 135 [caffeoyl]−

Cistanoside F

I, II, III

5a

22.61

250, 268, 334

813 [M−H]−

813 [M−H]− 637 [M−H−176]− 461 [M−H−352]− 351 [2GlcA–H–H2 O]− 285 [M−H−176 × 3]− 193, 175 [GlcA]−

Luteolin-diglucuronideglucuronide

II

6a

23.31

248, 288, 332

771 [M−H]−

771 [M−H]− 753 [M−H–H2 O]− 591 [M−H–H2 O–caffeoyl]− 179, 161, 135 [caffeoyl]− 151 [dehydrophenethanol]− 149 [Apiose]−

␤-OH-forsythoside B

I, II, III

7a

24.28

246, 290, 332

639 [M−H]−

639 [M−H]− 621 [M−H–H2 O]− 459 [M−H–H2 O–caffeoyl]− 179, 161, 135 [caffeoyl]− 151 [dehydrophenethanol]−

Campneoside II

I, III

8a

26.22

248, 288, 332

771 [M−H]−

771 [M−H]− 753 [M−H–H2 O]− 591 [M−H–H2 O–caffeoyl]− 179,161,135[caffeoyl]− 151[dehydrophenethanol]− 149 [Apiose]−

␤-OH-isoforsythoside B

I

9a

26.85

248, 288, 332

785 [M−H]−

785 [M−H]− 767 [M−H–H2 O]− 605 [M−H–H2 O–caffeoyl]− 179, 161, 135 [caffeoyl]− 151 [dehydrophenethanol]− 163, 145 [Rhamnose]−

␤-OH-poliumoside

II

10a

27.61

248, 290, 332

785 [M−H]−

785 [M−H]− 623 [M−H–caffeoyl]− 179, 161, 135 [caffeoyl]− 153 [phenethanol]−

Echinacoside

II

11a

28.48

248, 288, 330

785 [M−H]−

785 [M−H]− 753 [M−CH3 OH]− 591 [M−CH3 OH–caffeoyl]− 179, 161, 135 [caffeoyl]− 151[dehydrophenethanol]− 149 [Apiose]−

␤-OCH3 -forsythoside B

III

12a

29.68

254, 266, 334

637 [M−H]−

637 [M−H]− 285 [M−H–352]− 351 [2GlcA–H–H2 O]− 193, 175 [GlcA]−

Luteolin-diglucuronide

II

13a

29.90

236, 290, 332

653 [M−H]−

653 [M−H]− 621 [M−CH3 OH]− 459 [M−CH3 OH–caffeoyl]− 179, 161, 135 [caffeoyl]− 151 [dehydrophenethanol]−

Campneoside I

I, III

14

30.66

236, 245, 290, 332

755 [M−H]−

755 [M−H]− 593 [M−H–caffeoyl]− 179, 161, 135 [caffeoyl]− 153 [phenethanol]− 149 [Apiose]−

Forsythoside B

I, II, III

15

31.81

224, 245, 290, 335

623[M−H]−

623 [M−H]− 461 [M−H–caffeoyl]− 179, 161, 135 [caffeoyl]− 153 [phenethanol]−

Acteoside

I, II, III

No.

Y. Shi et al. / Journal of Pharmaceutical and Biomedical Analysis 75 (2013) 239–247

243

Table 1 (Continued) No.

tR (min)

UV max (nm)

[M−H]− (m/z) −

MS/MS fragments (m/z) −

Identification

Callicarpa species

16

34.09

246, 290, 332

755 [M−H]

755 [M−H] 593 [M−H–162]− 179, 161, 135 [caffeoyl]− 153[phenethanol]−

Samioside

I

17a

34.27

246, 268, 288, 335

621 [M−H]−

621 [M−H]− 351 [2GlcA–H–H2 O]− 269 [M−H–352]− 193,175[GlcA]−

Apigenin-7-O-diglucuronide

III

18a

34.68

246, 290, 330

681 [M−H]−

681 [M−H]− 663 [M−H–H2 O]− 501 [M−H–H2 O–caffeoyl]− 459 [M−H2 O–caffeoyl–CH3 CO]− 179, 161, 135 [caffeoyl]− 151 [dehydrophenethanol]−

2 -acetyl-campneoside II

III

19

35.42

246, 288, 332

769 [M−H]−

769 [M−H]− 607 [M−H–caffeoyl]− 179, 161, 135 [caffeoyl]− 153 [phenethanol]−

Poliumoside

II, III

20a

35.89

246, 290, 330

653 [M−H]−

653 [M−H]− 621 [M−CH3 OH]− 459 [M−CH3 OH–caffeoyl]− 179,161,135 [caffeoyl]− 151 [dehydrophenethanol]−

Isocampneoside I

III

21a

37.34

224, 244, 290, 330

623 [M−H]−

623 [M−H]− 461 [M−H–162]− 179, 161, 135 [caffeoyl]− 153 [phenethanol]−

Isoacteoside

I, III

22a

38.04

246, 290, 332

637 [M−H]−

637 [M−H]− 461 [M−H–feruloyl]− 315 [M−H–feruloyl–Rha]− 193, 175 [feruloyl]−

Eukovoside

III

23a

42.57

250, 290, 332

811 [M−H]−

811 [M−H]− 769 [M−CH3 CO]− 649 [M−H–caffeoyl]− 607 [M−CH3 CO–caffeoyl]− 179, 161, 135 [caffeoyl]− 153 [phenethanol]−

Brandioside

II

24

43.28

246, 268, 336

445 [M−H]−

445 [M−H]− 269 [M−H–176]−

Apigenin-7-O-␤-glucuronide

I

25a

43.52

248, 290, 332

665 [M−H]−

665 [M−H]− 503 [M−H–caffeoyl]− 461 [M–caffeoyl–CH3 CO]− 179, 161, 135 [caffeoyl]− 153 [phenethanol]−

2 -acetylacteoside

III

26a

46.25

246, 290, 332

651 [M−H]−

651 [M−H]− 505 [M−H–Rha]− 475 [M−H–feruloyl]− 193, 175 [feruloyl]−

Martinoside

III

27a

46.76

246, 268, 332

635[M−H]−

635 [M−H]− 351 [2GlcA–H–H2 O]− 283 [M−H–352]− 193, 175 [GlcA]−

Acacetin-diglucuronide

III

28

50.20

250, 268, 288, 332

461 [M−H]−

461 [M−H]− 285 [M−H–176]−

Luteolin-3 -O-␤-glucuronide

I

29a

51.20

248, 288, 332

475 [M−H]−

475 [M−H]− 299[M−H–176]−

Chrysoeriol-7-O-␤-glucuronide

I

30

52.12

254, 268, 342

285 [M−H]−

285 [M−H]− 151 [M−H–134]− 133 [M−H–152]−

Luteolin

I

31

53.93

245, 268, 330

459 [M−H]−

459 [M−H]− 283 [M−H–176]−

Acacetin-7-O-␤-glucuronide

III

32

56.75

254, 268, 292, 336

269 [M−H]−

269 [M−H]− 151 [M−H–118]− 117 [M−H–152]−

Apigenin

I

33

62.82

256, 290, 332

329 [M−H]−

329 [M−H]−

Rhamnazin

I

I, Callicarpa nudiflora Hook. et Arn.; II, Callicarpa kwangtungensis Chun.; III, Callicarpa macrophylla Vahl. a Firstly reported in the three callicarpa species.

244

Y. Shi et al. / Journal of Pharmaceutical and Biomedical Analysis 75 (2013) 239–247

Fig. 2. The typical chromatograms of Callicarpa nudiflora Hook. et Arn. (A), Callicarpa kwangtungensis Chun. (B), Callicarpa macrophylla Vahl. (C) and the mixed reference standards of forsythoside B (14), acteoside (15), poliumoside (19), isoacteoside (21), luteolin (30) and apigenin (32) (D).

2 -acetylacteoside. The two compounds (23 and 25) were found in other Callicarp species [15], which was an aid for their characterization. Compounds 22 and 26 were identified as eukovoside and martinoside, respectively, based on their MS/MS and UV data from literature [16].

3.4. Structures elucidation of ˇ-substituted acteoside type phenylethanol glycosides Compound 7 was identified as campneoside II, a typical ␤hydroxyacteoside compound. It showed the similar fragmentation

Y. Shi et al. / Journal of Pharmaceutical and Biomedical Analysis 75 (2013) 239–247

-H

-H

OH CH2OH OO ORha

HO

O

HO

OH CH2OH OO ORha

OH OH

-H2O

OH

245

HO

O

HO

O

OH

OH

O

Compound 7 m/z 639

m/z 621

-caffeoyl -H OH CH2OH OO ORha

HO O

HO

OH

O

OH

m/z 179

OH

-CO2

m/z 459

-H2O

-H

-H

-H

HO

OH

HO

OH

O OH

HO

O

m/z 161

m/z 135

m/z 151

Fig. 3. The proposed fragmentation pathways of ␤-hydroxyacteoside.

pathways with that of acteoside type phenylethanol glycosides except a characteristic ion at m/z 151, which was formed by losing of a H2 O from the segment of ␤-hydroxyphenethanol. The proposed fragmentation pathways of ␤-hydroxyacteoside were illustrated in Fig. 3. Considering the molecular weight, compound 6 showed more 132 u (apiose), compound 9 showed more 146 u (rhamnose) and compound 18 showed more 42 u (acetyl moiety) than that of compound 7. The compounds 6, 9, 18 were characterized as ␤OH-forsythoside B, ␤-OH-poliumoside and 2 -acetyl-campneoside II, respectively. Compound 8 is an isomer of 6 for their same quasimolecular and product ions. According to their retention time, compound 8 was characterized as ␤-OH-isoforsythoside B. Compounds 13 and 20 were also a pair of isomers. They showed [M−H]− ion at m/z 653, which is more 14 u than that of 7, and product ions at m/z 621 [M−CH3 OH]− , 459 [M−CH3 OH–caffeoyl]− and 151 [dehydrophenethanol]− . Compounds 13 and 20 were identified as campneoside I and isocampneoside I, respectively. Compound 11 showed a molecular weight more 132 u than that of compound 13, which was identified as ␤-methoxyforsythoside B.

3.5. Identification of flavonoids Compounds 30, 32 and 33 were identified as luteolin, apigenin and rhamnazin, respectively, by comparing with the reference substances previously isolated in our laboratory. Compounds 5, 12 and 28 were identified as luteolin derivatives, as they all produced the ion of luteolin aglycone at m/z 285. Compounds 17 and 24 were apigenin derivatives which possessing the ion of apigenin aglycone at m/z 269. Compounds 27 and 31 both showed the ion of acacetin aglycone at m/z 283. Based on their MS/MS

and UV data, these compounds were tentatively identified as a luteolin–diglucuronide (12), apigenin-7-O-diglucuronide (17), apigenin-7-O-␤-glucuronide (24), acacetin–diglucuronide (27), luteolin-3 -O-␤-glucuronide (28) and acacetin-7-O-␤-glucuronide (31). Among them, compound 17, 24, 28 and 31 have been previously reported in Callicarpa species [17,18]. Compound 5 showed quasi-molecular ions at m/z 637 [M−H−176]− , product ions at m/z 461 [M−H−176 × 2]− and 285 [M−H−176 × 3]− , suggesting the presence of three glucuronide units and a luteolin aglycone. It was tentatively identified as a luteolin–diglucuronide–glucuronide [19].

3.6. Method validation for simultaneous quantification of marker compounds Four marker phenylpropanoid glycosides (14, 15, 19, and 21) and two flavonoids (30 and 32) were determined for quality evaluation of the three species. The method was fully validated. As shown in Table 2, all the six compounds showed good linearity (r 2 > 0.9985). The LOQ varied from 2 to 8 ng. The intra- and interday precisions (RSD %) were found to be in the range of 0.11–2.57% and 0.28–2.83%, respectively. The RSD values of repeatability were not more than 3.12% for all analytes. The average recoveries were calculated and all the six analytes showed good recovery rates of 96.3–104.6%, and all the analytes showed good stability within 24 h (RSD < 3.0%). Three flavonoid glucuronides (5, 12 and 24) were semiquantified according to their corresponding aglycones (luteolin and apigenin). Therefore, the established HPLC/UV method was suitable for the simultaneous determination of nine compounds in the three Callicarpa species.

246

Y. Shi et al. / Journal of Pharmaceutical and Biomedical Analysis 75 (2013) 239–247

Table 2 Method validation for quantification of marker compounds in the three Callicarpa species. Compound

Forsythoside B (14) Acteoside (15) Poliumoside (19) Isoacteoside (21) Luteolin (30) Apigenin (32)

Regression equation*

y = 15117x−2227.8 y = 13699x+27676 y = 14847x+4408.8 y = 21368x−1077.1 y = 40668x−3580.4 y = 56638x−887.22

Standard deviation

Slope

Intercept

17.57 18.53 33.52 63.86 50.09 157.27

112.42 669.76 136.51 34.92 108.21 33.75

Linear range (␮g/mL)

r2

LOD (ng/mL)

4.7–470 19.1–1910 4.26–426 2.2–220 0.196–19.6 0.106–10.6

1.0000 1.0000 1.0000 1.0000 0.9998 0.9999

47 63 42 22 20 10

LOQ (ng/mL)

117 191 106 55 49 26

Precision

Recovery (n = 6)

Intra-day (n = 6) RSD (%)

Inter-day (n = 3) RSD (%)

Mean (%)

RSD (%)

0.34 0.11 2.15 1.49 1.64 2.57

1.13 0.28 2.83 1.74 1.82 2.19

98.22 101.50 99.48 99.74 100.80 100.38

1.05 0.98 2.37 1.48 2.73 1.65

In the regression equation y = ax + b, x refers to the concentration (␮g/mL), and y indicates the peak area. r2 refers to the correlation coefficient of regression equations.

Table 3 Contents (mg/g) of 9 marker compounds in the three Callicarpa species (n = 3). Compound

Forsythoside B (14) Acteoside (15) Poliumoside (19) Isoacteoside (21) Luteolin (30) Apigenin (32) Luteolin-diglucuronide-glucuronide (5)a Luteolin-diglucuronide (12)a Apigenin-7-O-␤-glucuronide (24)b

Contents (mg/g) (mean ± sd) C. macrophylla Vahl.

C. nudiflora Hook. et Arn.

C. kwangtungensis Chun.

4.68 ± 0.13 31.84 ± 0.97 0.908 ± 0.037 3.46 ± 0.13 ND ND ND ND ND

0.174 ± 0.007 6.41 ± 0.24 ND 1.984 ± 0.075 0.079 ± 0.003 0.081 ± 0.004 ND ND 0.183 ± 0.008

10.93 ± 0.33 0.862 ± 0.035 6.35 ± 0.19 ND ND ND 0.296 ± 0.009 0.403 ± 0.011 ND

ND, not detected. a Semi-quantified using the calibration curve of luteolin. b Semi-quantified using the calibration curve of apigenin.

3.7. Application Contents of the nine compounds (5, 12, 14, 15, 19, 21, 24, 30 and 32) in the three Callicarpa species (three batches) were shown in Table 3. The chemical constituents and corresponding contents showed significantly different among the three species. Phenylpropanoid glycosides was the major constituents with their contents were 40.7, 8.56 and 18.11 mg/g for C. macrophylla Vahl., C. nudiflora Hook. et Arn. and C. kwangtungensis Chun., respectively. Acteoside (15) was found the major compound of C. macrophylla Vahl and C. nudiflora Hook. et Arn., while forsythoside B (14) and poliumoside (19) were the predominant constituents of C. kwangtungensis Chun. It should be noticed that the content of acteoside (15) was significantly higher in C. macrophylla Vahl than in other two species. Poliumoside (19) was not detected in C. nudiflora Hook. et Arn., and isoacteoside (21) was too low to be quantified in C. kwangtungensis Chun. Meanwhile, total contents of flavones were less than 1.0 mg/g in three species. All five marker flavones were not detected in C. macrophylla Vahl. Luteolin–diglucuronide–glucuronide (5) and luteolin–diglucuronide (12) were only detected in C. kwangtungensis Chun., while luteolin (30), apigenin (32) and apigenin-7-O-␤-glucuronide (24) were just found in C. nudiflora Hook. et Arn. The total content of five marker flavones in C. kwangtungensis Chun. was higher than that in C. nudiflora Hook. et Arn. These characteristics could be used to discriminate and evaluate the quality of the three Callicarpa species.

4. Conclusion In conclusion, chemical composition and contents of 9 marker compounds from the three Callicarpa species, C. nudiflora Hook. et Arn., C. macrophylla Vahl. and C. kwangtungensis Chun. were analyzed and compared in the paper. The constituents were

comprehensive characterized by HPLC–DAD–ESI-Trap MS method. A total of 33 compounds were identified or tentatively identified, and 23 of them were reported from the three species for the first time. In addition, 9 marker compounds in the three species were simultaneously determined by the developed method. According to this study, phenylpropanoid glycosides and flavone glycosides were found the major constituents of the three Callicarpa species, but their types and contents were significantly different, which is a measure of the effectiveness and quality assessment of these herb medicines. Acknowledgement This work was supported by the Fundamental Research Funds for the Central Universities (Program No. ZJ10015). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jpba.2012.11.038. References [1] W.P. Jones, A.D. Kinghorn, Biologically active natural products of the genus Callicarpa, Curr Bioact. Compd. 4 (2008) 15–32. [2] J. Fu, S.Y. Kuang, S.X. Wang, Study on the effects of Callicarpa nudiflora tablets in the antibiosis and antiphlogosis and the hemostasis, J. Hainan Univ. (Nat. Sci.) 20 (2002) 154–157. [3] J.P. Cai, L. Dong, W.W. Guan, M.S. Liu, Research progress of Callicarpa nudiflora, Drugs Clin. 27 (2012) 60–63. [4] F.L. Kong, Aciclovir tablets combined with luohuazizhu dispersible tablets for treating herpes zoster in acute exacerbation period, Clin. Pharm. 19 (2010) 50–51. [5] Chinese Pharmacopoeia Commission, Pharmacopoeia of the People’s Republic of China, Vol. 1, 2010th ed., Chinese Medical Science and Technology Press, Beijing, 2010, pp. 18.

Y. Shi et al. / Journal of Pharmaceutical and Biomedical Analysis 75 (2013) 239–247 [6] Chinese Pharmacopoeia Commission, Pharmacopoeia of the People’s Republic of China, Vol. 1, 2010th ed., Chinese Medical Science and Technology Press, Beijing, 2010, pp. 40. [7] B.T. Zhou, X.Z. Li, G.R. Zhong, G.Y. Zheng, Z.Y. Chen, Elementary pharmacodynamical studies of aerial part of Callicarpa kwungtungensis Chun, Chin. J. Modern Med. 16 (2006) 204–206. [8] Z.X. Gu, Y.Q. Liu, D.L. Yan, S. Huang, J. Lu, The research progress of Callicarpa nudiflora Hook.et Arn., Callicarpa macrophylla Vahl. and Callicarpa kwangtungensis, Chun. Chin. Her Med. 8 (2011) 11–12. [9] Y.J. Wang, Y.F. Yang, D. Gao, Advances in studies on chemical constituents in plants of Callicarpa Linn. and their bioactivities, Chin. Trad. Herbal Drugs 39 (2008) 133–138. [10] Y.Q. Zhang, J.B. Hong, W.L. Liu, HPLC determination of oleanolic acid and ursolic acid in folium Callicarpa enudiflorea, J. Hainan Medical Coll. 15 (2009) 5–7. [11] R. Hu, M. Yao, Y.Y. Li, C.Q. Zhu, HPLC content determination of luteolin in Callicarpa nudiflora Hook.et Arn, Trad. Chin. Drug. Res. Clin. Pharmacol. 20 (2009) 271–272. [12] P. Pan, L.Y. Jie, Q.S. Sun, RP-HPLC determination of betulinic acid in Callicarpa macrophylla Vahl, Chin. J. Chin. Mater. Med. 33 (2008) 753–755. [13] J. Zhang, B.Q. Li, F. Feng, Y.P. Tang, W.Y. Liu, Chemical constituents from Callicarpa nudiflora and their hemostatic acitivity, Chin. J. Chin. Mater. Med. 35 (2010) 3297–3301.

247

[14] Y.X. Wang, H.P. Hao, G.J. Wang, P.F. Tu, Y. Jiang, Y. Liang, L. Dai, H. Yang, L. Lai, C.N. Zheng, Q. Wang, N. Cui, Y.T. Liu, An approach to identifying sequential metabolites of a typical phenylethanoid glycoside, echinacoside, based on liquid chromatography–ion trap-time of flight mass spectrometry analysis, Talanta 80 (2009) 572–580. [15] K.A. Koo, S.H. Sung, J.H. Park, S.H. Kim, K.Y. Lee, Y.C. Kim, In vitro neuro protective activities of phenylethanoid glycosides from Callicarpa dichotoma, Planta Med. 71 (2005) 778–780. [16] R. Quirantes-Pine, L. Funes, V. Micol, A. Segura-Carretero, A. FernandezGutierrez, High-performance liquid chromatography with diode array detection coupled to electrospray time-of-flight and ion-trap tandem mass spectrometry to identify phenolic compounds from a lemon verbena extract, J. Chromatogr. A 1216 (2009) 5391–5397. [17] S.S. Subramanian, A.G.R. Nair, T.N.C. Vedantham, Terpenoids and flavones of Callicarpa macrophylla and C. longifolia, Phytochemistry 13 (1974) 306–307. [18] F.P. Gao, H. Wang, W.C. Ye, S.X. Zhao, Chemical constituents form the leaves of Callicarpa nudiflora, J. Chin. Pharm. Univ. 41 (2010) 120–123. [19] M. Schulz, G. Weissenbock, Partial purification and characterization of a luteolin–triglucuronide-specific ␤-glucuronidase from rye primary leaves (Secale cereale), Phytochemistry 26 (1987) 933–937.