A comprehensive study of pomegranate flowers polyphenols and metabolites in rat biological samples by high-performance liquid chromatography quadrupole time-of-flight mass

Journal of Chromatography A 1604 (2019) 460472

Contents lists available at ScienceDirect

Journal of Chromatography A
journal homepage: www.elsevier.com/locate/chroma

A comprehensive study of pomegranate flowers polyphenols and
metabolites in rat biological samples by high-performance liquid
chromatography quadrupole time-of-flight mass spectrometry
Zainaipuguli Yisimayili a,b,c, Rahima Abdulla a, Qiang Tian c, Yangyang Wang b,c,
Mingcang Chen c, Zhaolin Sun c, Zhixiong Li c, Fang Liu c, Haji Akber Aisa a,b,∗,
Chenggang Huang b,c,∗
a

Key Laboratory of Plant Resources and Chemistry of Arid Zone, Xinjiang Technical Institute of Physics and Chemistry, Chinese Academy of Sciences, Urumqi
830011, China
b
University of Chinese Academy of Sciences, Beijing 100049, China
c
Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai 201203, China

a r t i c l e

i n f o

Article history:
Received 21 May 2019
Revised 14 August 2019
Accepted 21 August 2019
Available online 22 August 2019

Keywords:
Punica granatum L. flowers
Polyphenols
Ellagitannin
Metabolism
HPLC-Q-TOF-MS2

a b s t r a c t
Pomegranate flowers is an ancient medicine that has commonly been used to treat various diseases such
as diabetes. However, no reports are available on the metabolic profile of pomegranate flowers in vivo.
In the present study, with the aid of HPLC-Q-TOF-MS2 , 67 compounds were identified in pomegranate
flowers extract, including 18 ellagitannins, 14 gallic acid and galloyl derivatives, five anthocyanins and
18 flavonoids. Seven compounds were firstly identified. In vivo, 22 absorbed compounds and 35 metabolites were identified in rat biosamples (urine, feces, plasma and tissues) after orally administered with
pomegranate flowers extract. This result showed that not all compounds abundant in pomegranate flowers extract could be absorbed well in plasma and tissues. This finding also suggested a potential correlation between study on metabolic profile of these compounds in vivo and study on strategy of screening
bioactivity of the isolates with in vitro cell systems evaluation. Notably, mono-glucuronide conjugated
metabolite of ellagitannin compound (corilagin) was firstly identified. In addition, this is first report to
identify phase II conjugate metabolites of ellagitannins in vivo after oral administration of ellagitanninsrich extracts (or foods). Thus, characterizing its multiple constitution, absorption and metabolic fate of
these compounds in vivo is helpful to better analyze the active components in pomegranate flowers.
© 2019 The Authors. Published by Elsevier B.V.
This is an open access article under the CC BY license. ( />
1. Introduction
Pomegranate (Punica granatum L.) is widely cultivated for its
widely consumed fruit in the regions of Southeast Asia, the
Mediterranean area and USA. Notably, pomegranate flowers is an
ancient medicine that has commonly been used to treat various diseases such as chronic diarrhea and aphthous stomatitis.
In Unani and Ayurvedic medicine, and in some parts of China,
pomegranate flowers have widely been used to treat diabetes
[1–4].
According to previous studies, the health benefits of
pomegranate flowers have been associated with their polyphenol

content, specifically their anthocyanins, flavonoids and tannins

∗

Corresponding authors.
E-mail addresses: (H.A. Aisa), (C. Huang).

content [5,6]. Anthocyanins, a type of the flavonoids, are the
major pigments responsible for the bright color of pomegranate
flowers [5,7]. Tannins are one group of natural compounds and
the major compounds in pomegranate flowers. Besides, the wide
range of bioactivities of pomegranate flowers have been associated
with the polyphenols isolated from (or present in) pomegranate
flowers such as phenolics, ellagitannins and flavonoids as active
components based on strategy of screening bioactivity of the
isolates with in vitro cell systems evaluation [1,3,5,6,8,9,11]. While
it is difficult to make structure-activity correlation conclusion
among the phytocompounds in the extract or exploring bioactivity
of the isolates with in vitro cell systems evaluation. Because the
ingested compounds, at least part of them, reach the circulatory
system and specific tissues to exert biological effect as a result
of in vivo process of absorption, distribution, metabolism and
excretion [21,25,28]. Thus, further studies such as bioavailability
and metabolism of these compounds in vivo would be required

/>0021-9673/© 2019 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license. ( />

2

Z. Yisimayili, R. Abdulla and Q. Tian et al. / Journal of Chromatography A 1604 (2019) 460472

before exploring their some potential activities. After oral administration of pomegranate flowers extract, the phytocompounds
absorbed as native form and their derived metabolites, at least a
portion of them, may be the functional components responsible
for the bioactivities of pomegranate flowers such as antioxidant,
anti-inflammatory, α -glucosidase inhibitory and hepatoprotective
activities [1,5,6,9–11]. However, no reports are available on the
metabolic profile of pomegranate flowers in vivo. The absence
of scientific evidence for its activities may restrict its further
development including clinical application. Thus, characterizing
its multiple constitution, absorption and metabolic fate of these
compounds in vivo is necessary to better analyze the bioactive
components in pomegranate flowers.
Therefore, in the present study, using rapid and high sensitive, high-performance liquid chromatography quadrupole time-offlight mass spectrometry (HPLC-Q-TOF-MS2 ) method, we characterized the phytochemical profile of pomegranate flowers extract.
Furthermore, the absorbed compounds and their metabolites in
rat plasma, tissues, urine and feces after oral administration of
pomegranate flowers extract were analyzed comprehensively. This
study will provide vital information for finding possible candidates
for the real bioactive compounds in pomegranate flowers and provide a solid basis for further study of biological properties of the
compounds in pomegranate flowers.
2. Materials and methods
2.1. Chemicals and reagents
Reference standards (corilagin, gallic acid, ethyl gallate, ellagic acid, brevifolin, brevifolincarboxylic acid, punicalagin, apigenin, apigenin-7-O-glucoside, kaempferol, luteolin, luteolin-7O-glucoside, isoquercetin, urolithin D, urolithin C, urolithin B,
urolithin A) were used to absolutely identified these compounds
in pomegranate flowers and rat biosamples. These standards
were purchased from the Chengdu MUST Bio-Technology Co. Ltd.
(Chengdu, China). Acetonitrile methanol and formic acid were
bought from Thermo Fisher Scientific Co.Ltd. (Waltham, Massachusetts, USA). Milli-Q System (Millipore, Billerica, MA, USA) was
used to prepare purified water for HPLC. Other chemicals were
analytical-grade and bought from the Sinopharm Chemical Reagent

Co. Ltd. (Shanghai, China).

conditions were as follows: capillary, 40 0 0 V and 350 0 V for positive and negative ionization modes, respectively; nozzle voltage,
500 V; nebulizer, 45 psi; gas temperature and flow rate, 300 °C
and 7 L/min; sheath gas temperature and flow rate, 350 °C and
12 L/min; fragmentor, 100 V; collision energy (CE), 15 eV, 30 eV. The
m/z range of full mass spectra for MS1 was 10 0–170 0. The m/z
range was set from 100 to 1200 for MS2 experiments.
The HPLC-MS system operation and data analysis were carried
out with the Agilent Masshunter Workstation software which contain with Data Acquisition (Version B.05.01) software and Qualitative Analysis (Version B.06.00) software.
2.3. Sample preparation and pretreatment
Dried pomegranate flowers (300 g) were extracted with
ethanol /water (7:3, v/v) three times (solid /liquid ratio was 1:15,
1:15, 1:10, respectively) for (12 h, 6 h, 3 h, respectively) at 60 °C.
The combined extract was concentrated to 1.5 g /mL under vacuum
at 60 °C.
2.4. Animal experiment
Sixty male Sprague-Dawly (SD) rats were purchased from
Shanghai SLAC Laboratory Animal Co. Ltd. (Shanghai, China). Before
the experiment, all the rats were maintained in house with environmentally controlled at a 12 h light-dark cycle and at 22 ± 2 °C
with relative humidity (50 ± 10%) for six days.
Before drug administration, all rats were fasted for 10 h and
they were free to access water. The rats were randomly separated
into eleven groups (n = 5 /group). Groups 1–10 were orally administered with pomegranate flowers extract at a dose of 15 g/kg
and used for collecting blood and tissues samples. After oral administration of pomegranate flowers extract, urine and feces (from
0 to 48 h) were collected from the rats in group 10 which were
kept separately in metabolic cages. The blank biological samples
were collected from the rats in group 11. After oral administration
of pomegranate flowers extract, systematic blood (6–8 mL) from
aorta abdominal and organs (liver, heart, kidney, spleen, and lung)

were collected at 15 min, 0.5 h, 1 h, 2 h, 4 h, 6 h, 8 h, 12 h, 24 h,
48 h (10 time points). The blood biosamples were promptly centrifuged (12,0 0 0 rpm, 10 min). The physiological saline water (0.9%)
was used to homogenize organs.
2.5. Sample preparation

2.2. Instrumentations and investigation conditions
The HPLC-Q-TOF-MS2 system (Agilent Technologies, Palo Alto,
CA, USA) which consisted of a HPLC system (1260 Series, coupled to an Agilent Q-TOF mass spectrometer equipped with a Dual
Agilent Jet Stream Electrospray Ionization (ESI) sourse (6530 Series) was used for the identification of components in pomegranate
flowers extract and its metabolites in rat biosamples. The chromatographic separation for pomegranate flowers extract and biological samples were accomplished on an ACE Excel 3 Super C18
column (100 × 2.1 mm, 3.0 μm), (Advanced Chromatography Technologies Ltd. Aberdeen, Scotland). The HPLC flow rate and column
temperature were set at 0.35 mL/min and at 40 °C, respectively. The
optimized mobile phases contain solvent A and solvent B which
were 0.1% formic acid in water and 0.1% formic acid in acetonitrile. An optimized mobile phase gradient elution was as follows:
0–8.0 min, 3.0% B; 8.0–16.0 min, 3.0–8.0% B; 16.0–32.0 min, 8.0% B;
32.0–54.0 min, 8.0–18.0% B; 54.0–60.0 min, 18.0% B; 60.0–65.0 min,
18.0–50.0% B; 65.0–72.0 min, 50.0–80.0% B; 72.0–76.0 min, 80.0–
95.0% B; 76.0–80.0 min, 95% B; 80.1–85.0 min, 3.0% B. In this study,
the mass spectrometric detection for every samples was performed
in both ionization modes. The detection parameters for the MS

Plasma (200 μL at each time point) and 600 μL acetonitrile
were mixed for 5 min and centrifuged (14,0 0 0 rpm, 10 min). The
supernatants were separately evaporated to dryness under vacuum at 40 °C. After removal of the solvent of combined residue
from ten time points, 200 μL of methanol-water (7:3, v/v) was
used to dissolve the residue. The tissues homogenate and urine
biosamples were treated respectively with the same ways as the
plasma. Ground feces were mixed with 10 times of methanol (v/w)
and extracted in ultrasonic bath for two times (for 30 min). The
clear methanol layers were evaporated to dryness under vacuum

at 40 °C. Methanol-water (7:3, v/v) used to dissolve the residue was
200 μL. After centrifuging (14,000 rpm, 10 min) each sample, 10 μL
of sample was used to analysis by HPLC-Q-TOF-MS2 .
3. Results and discussion
In present study, a qualitative analysis of the polyphenols in the
pomegranate flowers extract, absorbed compounds and metabolites in rats orally administered with pomegranate flowers extract
were carried out by using high sensitive HPLC-Q-TOF-MS2 in both
ionization modes.


Z. Yisimayili, R. Abdulla and Q. Tian et al. / Journal of Chromatography A 1604 (2019) 460472

3

Fig. 1. Base peak chromatogram (BPC) of pomegranate flowers extract: (a) positive ion mode, (b) negative ion mode.

3.1. Identification of polyphenols in pomegranate flowers extract
In this study, as shown in Table 1, 67 compounds were
identified in pomegranate flowers extract, including 18 ellagitannins, 14 gallic acid and galloyl derivatives, five anthocyanins
and 18 flavonoids. Seven compounds were firstly identified in
pomegranate flowers. The peak characterization was performed
based on their retention time (tR ), accurate molecular mass (mass
error of less than 5 ppm), major MS/MS fragment ions. Furthermore, the experimental data were compared with commercially
available authentic standards for absolutely identification. The
base peak chromatograms (positive and negative ion mode) of
pomegranate flowers extract were shown in Fig. 1. Analysis of ellagitannins, gallic acid and galloyl derivatives were performed in
the negative ion mode because of stronger response in the MS
spectra. Both positive and negative ion mode were adopted to
identify anthocyanins and flavonoids.
3.1.1. Ellagitannins

Ellagitannins, member of the tannin family, are characterized
as hydrolyzable conjugates containing one or more hexahydroxydiphenoyl (HHDP) group(s) to esterify a sugar, usually glucose
[12]. During their MS/MS fragmentation, it can be observed the
typical losses such as galloyl moiety (152 Da), gallic acid (170 Da),
HHDP (302 Da), galloyl-glucose (332 Da), HHDP glucose (482 Da)
and galloyl-HHDP-glucose (634 Da) residues. Besides, in the negative ESI-IT/Q-TOF-MS2 mode, the characteristic fragment ions observed at m/z 300.99 (which is produced after the spontaneous
lactonization of the HHDP unit into ellagic acid) and m/z 169.01,
indicate the existence of HHDP group and galloyl group in the
molecule, respectively, based on the fragmentation pattern of ellagitannins previously reported in the literatures [12–18].
As shown in Fig. 2, peak 25 showed a protonated molecular ion [M−H]− at m/z 633.0730 (0 ppm) with a molecular formula of C27 H22 O18 . In the MS2 mode, the product ion at m/z
463.0544 [M−H−170 Da]− was occurred via the loss of a gallic acid
from the molecular ion. The typical fragment ion at m/z 300.9995
[M−H−332 Da]– , as a base peak, was formed from a galloyl-glucose

moiety loss from the [M−H]− ion. The typical fragment ion at m/z
169.0130, which was associated with gallic acid, was also observed.
Peak 25 was absolutely identified as corilagin by comparing with
its commercial standard. The proposed fragmentation pattern of
corilagin was shown in Fig. 3.
Peaks 3, 6, 12 and 21 were isomeric compounds. All of the
peaks had a [M−H]− ion at m/z 633.0729 (−0.63 ppm) with molecular formula of C27 H22 O18 . The fragment ion at m/z 463.0544
[M−H−170 Da]– (loss of a gallic acid), the typical ions at m/z
300.9998 and m/z 169.0138 were consistent with those of corilagin.
Thus, peaks 3, 6, 12 and 21 were tentatively identified as galloylHHDP-glucose isomers.
Peak 10 had a protonated molecular ion [M−H]− at m/z
481.0620 (−1.24 ppm) with molecular formula of C20 H18 O14 , and
the MS2 spectrum had fragments at m/z 463.0544 [M−H−18 Da]
and characteristic fragment ion at m/zm/z 300.9995. Thus, peak 10
was tentatively identified as HHDP-glucose.
Peaks 16, 23, 28, 34 and 35 were isomers. All of the peaks had a

[M−H]− ion at m/z 785.0818 (−3.18 ppm) with molecular formula
of C34 H26 O22 . The fragments at m/z 615.0637 [M−H−170 Da]– (loss
of a gallic acid), m/z 463.0544 [M−H−170 Da-152 Da]– (loss of a
gallic acid and a galloyl moiety), the typical ions at m/z 300.9998
and m/z 169.0138 were observed in the MS2 mode. Thus, peaks 16,
23, 28, 34 and 35 were tentatively identified as digalloyl-HHDPglucoside isomers. As an example, the MS/MS spectra and the proposed fragmentation pattern of one digalloyl-HHDP-glucoside were
shown in Figs. 2 and 3.
As shown in Fig. 2, peak 53 had a [M−H]− ion at m/z 937.0910
(−4.58 ppm) with molecular formula of C41 H30 O26 . The product
ions at m/z 767.0748 [M−H−170 Da]– (loss of a gallic acid), m/z
615.0683[M−H−170 Da−152 Da]– (loss of a gallic acid and a galloyl moiety), m/z 465.0754 [M−H−170 Da-302 Da]– (loss of a gallic
acid and HHDP moiety), the typical ions at m/z 300.9998 and m/z
169.0138 were observed in the MS2 mode. Therefore, peak 53 was
tentatively identified trigalloyl-HHDP-glucose.
Peaks 14, 40 and 46 were isomeric compounds. They had a
[M−H]− ion at m/z 951.0719 (−2.73 ppm) with molecular for-


4

Table 1
Summary of the mass spectral data of polyphenols identified in pomegranate flower extract by HPLC-Q-TOF-MS/MS.

1
2b
3
4
5
6
7

8
9
10
11c
12
13
14
15
16
17b
18
19
20b
21
22b , c
23
24
25b
26
27
28
29b , c
30
31
32c

tR (min)
1.46
2.06
2.55

2.92
3.40
4.45
5.01
5.96
9.79
11.98
14.60
14.68
15.26
16.93
18.18
18.22
18.56
19.74
19.89
20.19
21.54
24.25
25.36
26.46
26.50
26.91
27.17
28.13
28.54
30.30
31.29
33.29

Molecular
formula
C13 H16 O10
C7 H6 O5
C27 H22 O18
C20 H20 O14
C13 H16 O10
C27 H22 O18
C13 H16 O10
C20 H20 O14
C21 H10 O13
C20 H18 O14
C15 H11 O7
C27 H22 O18
C20 H20 O14
C41 H28 O27
C27 H31 O16
C34 H26 O22
C48 H28 O30
C27 H31 O15
C20 H20 O14
C13 H8 O8
C27 H22 O18
C9 H10 O5
C34 H26 O22
C34 H24 O22
C27 H22 O18
C21 H21 O11
C21 H21 O10
C34 H26 O22

C12 H8 O6
C21 H10 O13
C27 H24 O18
C14 H10 O8

Calculated
331.0671
169.0142
633.0733
483.0780
331.0671
633.0733
331.0671
483.0780
469.0049
481.0624
465.1028
633.0733
483.0780
951.0745
611.1612
785.0843
1083.0593
595.1663
483.0780
291.0146
633.0733
197.0455
785.0843
783.0686

633.0733
449.1078
433.1135
785.0843
247.0248
469.0049
635.0890
305.0303

Observed
331.0668
169.0140
633.0729
483.0782
331.0668
633.0729
331.0668
483.0782
469.0028
481.0618
465.1023
633.0729
483.0782
951.0719
611.1605
785.0818
1083.0583
595.1660
483.0782
291.0153

633.0729
197.0451
785.0818
783.0685
633.0733
449.1085
433.1133
785.0818
247.0244
469.0028
635.0881
305.0300

Ion mode
−

[M−H]
[M−H]−
[M−H]−
[M−H]−
[M−H]−
[M−H]−
[M−H]−
[M−H]−
[M−H]−
[M−H]−
[M]+
[M−H]−
[M−H]−
[M−H]−

[M]+
[M−H]−
[M−H]−
[M]+
[M−H]−
[M−H]−
[M−H]−
[M−H]−
[M−H]−
[M−H]−
[M−H]−
[M]+
[M]+
[M−H]−
[M−H]−
[M−H]−
[M−H]−
[M−H]−

Error (ppm)

MS/MS fragments

−0.90
−1.18
−0.63
0.41
−0.90
−0.63
−0.90

0.41
−4.47
−1.24
−1.07
−0.63
0.41
−2.73
−1.14
−3.18
−0.92
−0.50
0.41
2.40
−0.63
−2.02
−3.18
−0.12
0
1.55
−0.46
−3.18
−1.61
−4.47
−1.42
−0.98

169.0136,
125.0139
463.0535,
313.0560,

169.0136,
463.0535,
169.0136,
313.0560,
425.0149,
463.0510,
303.0486
463.0535,
313.0560,
907.0821,
449.1045,
615.0602,
781.0665,
433.1133,
313.0560,
247.0250,
463.0535,
169.0144,
615.0602,
300.9989
463.0493,
287.0532,
271.0615
615.0602,
219.0305,
425.0149,
465.0677,
273.0061,

Identification

125.0236
300.9998,
169.0137,
125.0236
300.9998,
125.0236
169.0137,
300.9995,
300.9995

169.0138
125.0240

300.9998,
169.0137,
783.0580,
287.0522
463.0499,
621.9980,
271.0606
169.0137,
219.0298,
300.9998,
125.0242
463.0499,

275.0204, 169.0138
125.0240
481.0534, 300.9987

275.0204, 169.0138
125.0240
169.0157, 125.0242

300.9993, 169.0130
300.9997
125.0240
191.0346
275.0204, 169.0138
300.9993, 169.0130

300.9990, 169.0152
153.0162
463.0499,
191.0348
300.9995,
313.0568,
245.0082,

300.9993, 169.0130
169.0157, 125.0242
169.0124
217.0141

Galloyl-glucoside
Gallic acid
Galloyl-HHDP-glucoside
Digalloyl- glucoside
Galloyl-glucoside

Galloyl-HHDP-glucoside
Galloyl-glucoside
Digalloyl-glucoside
Valoneic acid dilactone
HHDP-glucoside
Delphinidin-3-O-glucoside
Galloyl-HHDP-glucoside
Digalloyl- glucoside
HHDP-valoneoyl-glucoside
Cyaniding-3,5-O-diglucoside
Digalloyl-HHDP-glucose
Punicalagin
Pelargonidin-3,5-O-diglucoside
Digalloyl-glucoside
Brevifolincarboxylic acid
Galloyl-HHDP-glucoside
Ethyl gallate
Digalloyl-HHDP-glucoside
Di-HHDP- glucoside
Corilagin
Cyanidin 3-O-glucoside
Pelargonidin 3-O-glucoside
Digalloyl-HHDP-glucoside
Brevifolin
Valoneic acid dilactone
Trigalloyl- glucoside
Methyl brevifolincarboxylate
(continued on next page)

Z. Yisimayili, R. Abdulla and Q. Tian et al. / Journal of Chromatography A 1604 (2019) 460472

No.


Table 1 (continued)
tR (min)

Molecular
formula

Calculated

Observed

Ion mode

Error (ppm)

MS/MS fragments

33c
34
35
36
37a
38
39a , c
40
41c
42b

43
44
45
46
47a , b , c
48a , b
49
50a
51a
52a, b
53
54
55a
56
57a
58a
59a
60a
61

33.89
34.59
37.70
42.12
43.25
43.54
44.15
45.10
46.55
46.98

47.04
47.33
48.36
49.02
49.52
50.23
52.32
53.16
54.51
55.90
56.28
57.53
58.18
58.69
60.73
61.86
62.72
63.76
64.02

C33 H28 O22
C34 H26 O22
C34 H26 O22
C21 H10 O13
C27 H30 O16
C27 H24 O18
C27 H30 O15
C41 H28 O27
C15 H12 O8
C14 H6 O8

C41 H28 O27
C27 H30 O16
C34 H28 O22
C41 H28 O27
C21 H20 O12
C21 H20 O11
C34 H28 O22
C21 H20 O11
C21 H20 O12
C21 H20 O10
C41 H30 O26
C27 H20 O17
C21 H20 O11
C41 H32 O26
C21 H20 O11
C21 H20 O12
C15 H10 O7
C21 H20 O10
C43 H34 O28

775.0999
785.0843
785.0843
469.0049
611.1607 /609.1461
635.0890
595.1657 /593.1512
951.0745
319.0459
300.9990

951.0745
611.1607/609.1461
787.0999
951.0745
465.1028 /463.0882
449.1078 /447.0993
787.0999
449.1078 /447.0993
465.1028
433.1129 /431.0984
937.0953
615.0628
449.1078 /447.0993
939.1109
449.1078 /447.0993
465.1028
303.0499
433.1129 /431.0984
997.1164

775.0937
785.0818
785.0818
469.0028
611.1605
635.0881
595.1660
951.0719
319.0459
300.9998

951.0734
611.1605
787.0972
951.0719
465.1021
449.1085
787.0972
449.1085
465.1021
433.1128
937.0910
615.0621
449.1085
939.1061
449.1085
465.1021
303.0503
433.1128
997.1138

[M−H]−
[M−H]−
[M−H]−
[M−H]−
[M+H]+ /[M−H]−
[M−H]−
[M+H]+ /[M−H]−
[M−H]−
[M−H]−
[M−H]−

[M−H]−
[M+H]+ /[M−H]−
[M−H]−
[M−H]−
[M+H]+ /[M−H]−
[M+H]+ /[M−H]−
[M−H]−
[M+H]+ /[M−H]−
[M+H]+ /[M−H]−
[M+H]+ /[M−H]−
[M−H]−
[M−H]−
[M+H]+ /[M−H]−
[M−H]−
[M+H]+ /[M−H]−
[M+H]+ /[M−H]−
[M+H]+ /[M−H]−
[M+H]+ /[M−H]−
[M−H]−

−7.99
−3.18
−3.18
−4.47
−1.14
−1.42
−0.50
−2.73
0
2.65

−1.15
−1.14
−3.43
−2.73
−1.50
1.55
−3.43
1.55
−1.50
−0.23
−4.58
−1.13
1.55
−5.11
1.55
−1.50
1.31
−0.23
−2.60

757.0854,
615.0602,
615.0602,
425.0149,
449.1045,
465.0677,
433.1133,
907.0821,
273.0036,
283.9967,

933.0629,
449.1045,
617.0745,
907.0821,
303.0484,
287.0573,
617.0745,
287.0573,
303.0499,
271.0612
767.0682,
445.0438,
287.0569,
769.0873,
287.0569,
303.0499,
153.0193
271.0612
953.1216,
169.0149
303.0484,
953.1216,
169.0149
153.0178
153.0173
153.0175
315.0496,

62
63

a

64a , b
65a , b
66a , b
67a
a
b
c

+

−

65.13
65.66

C21 H20 O12
C43 H34 O28

465.1028 /463.0882
997.1164

465.1021
997.1138

[M+H] /[M−H]
[M−H]−

65.70
67.07
67.25
67.31

C15 H10 O6
C15 H10 O5
C15 H10 O6
C17 H14 O7

287.0550
271.0601
287.0550
331.0812

287.0554
271.0603
287.0555
331.0813

[M+H]+ /[M−H]−
[M+H]+ /[M−H]−
[M+H]+ /[M−H]−
[M+H]+ /[M−H]−

/285.0405
/269.0455
/285.0405
/329.0667

−1.50
−2.60
1.39
0.73
1.74
0.30

465.0680,
463.0499,
463.0499,
300.9995,
287.0522,
313.0568,
271.0606,
783.0580,
245.0085,
257.0083,
613.0448,
287.0522,
465.0659,
783.0580,
153.0152
153.0187
465.0659,
153.0187
153.0178

Identification

300.9992,

300.9990,
300.9990,
169.0157,
153.0159
169.0124
153.0179
481.0534,
217.0141
229.0139,
463.0492,
153.0159
313.0582,
481.0534,

169.0139
169.0130
169.0130
125.0242

783.0990, 633.0715, 481.0912, 300.9994,

Ellagitannin
Digalloyl-HHDP-glucoside
Digalloyl-HHDP-glucoside
Valoneic acid dilactone
Luteolin-O-diglucoside
Trigalloyll-glucoside
Apigenin-O-diglucoside
HHDP-valoneoyl-glucoside
Ethyl brevifolincarboxylate

Ellagic acid
Galloyl-HHDP-DHHDP-hexoside
Luteolin-O-diglucoside
Tetragalloyl-glucoside
HHDP-valoneoyl-glucoside
Isoquercetin
Luteolin 7-O-glucoside
Tetragalloyl-glucoside
Luteolin −7-O-glucoside isomer
Tricetin 4 -O-β -glucoside
Apigenin-7-O-glucoside
Trigalloyl-HHDP-glucose
Galloyl-ellagic acid glucoside
Kaempferol-O-glucoside
Pentagalloyl-glucoside
Kaempferol-O-glucoside
Tricetin-4 -O-β -glucoside isomer
Tricetin
Apigenin-7-O-glucoside isomer
Punictannin A

153.0152
783.0990, 633.0715, 481.0912, 300.9994,

Quercetin
Punictannin B

299.0549, 270.0516, 133.1010

Luteolin

Apigenin
Kaempferol
Tricin

300.9987
201.0181
300.9994,
169.0147
300.9987

313.0582, 169.0147

615.0477, 465.0674, 300.9999, 169.0148
300.9999, 169.0132
153.0196
617.0752, 465.0647, 313.0547, 169.0130
153.0196
153.0178

Z. Yisimayili, R. Abdulla and Q. Tian et al. / Journal of Chromatography A 1604 (2019) 460472

No.

Error (ppm) and fragment ions taken from the positive ion mode (in case detected in both modes).
Confirmed by using reference standard.
Firstly identified compounds in pomegranate flower.

5

6

Z. Yisimayili, R. Abdulla and Q. Tian et al. / Journal of Chromatography A 1604 (2019) 460472

Fig. 2. The MS/MS spectra of typical ellagitannins and galloyl derivatives in pomegranate flowers.

Fig. 3. Proposed fragmentation pattern of typical ellagitannins and galloyl derivatives in pomegranate flowers.

mula of C41 H28 O27 . These compounds had major fragment ions
at m/z 907.0821[M−H−44 Da]– (loss of a CO2 ), m/z 783.0580, m/z
481.0534, m/z 300.9987 in their MS2 mode. Therefore, peaks 14,
40 and 46 were tentatively identified as HHDP-valoneoyl-glucoside
isomers.
3.1.2. Gallic acid and galloyl derivatives
Peak 2 had a [M−H]− ion at m/z 169.0140 (−1.18 ppm) with
molecular formula of C7 H6 O5 . The product ion at m/z 125.0239
[M−H−44 Da]− was generated via the elimination of CO2 unit
from the molecular ion. Peak 22 showed a [M−H]− ion at m/z
197.0451 (−2.02 ppm) with molecular formula of C9 H10 O5 and the
MS2 spectrum had fragment ions at m/z 169.0144 [M−H−28 Da]–
(loss of a C2 H4 moiety) and m/z 125.0242 [M−H−28 Da-44 Da]–

(loss of a C2 H4 and a CO2 moieties). Peaks 2 and 22 were absolutely identified as gallic acid and ethyl gallate by comparing with
their reference standards, respectively.
Peaks 1, 5 and 7 displayed molecular ion [M−H]− at m/z
331.0668 (−0.90 ppm) with molecular formula of C13 H16 O10 . The
fragment ion at m/z 169.0136 [M−H−162 Da]– (loss of a glucose
moiety), and m/z 125.0236 [M−H−162 Da-44 Da]– were correlated
to gallic acid on the basis of the MS/MS spectra data. Thus, peaks
1, 5 and 7 were tentatively identified as galloyl-glucose isomers.

Peaks 4, 8, 13 and 19 were isomeric compounds. All of them
showed a [M−H]− ion at m/z 483.0782 (0.41 ppm) with molecular formula of C20 H20 O14 . The fragment ion at m/z 313.0560
[M−H−170 Da]– (loss of a gallic acid), the characteristic ions at
m/z 169.0137 and ion at m/z 125.0240 were observed in the MS2


Z. Yisimayili, R. Abdulla and Q. Tian et al. / Journal of Chromatography A 1604 (2019) 460472

mode. Therefore, peaks 4, 8, 13 and 19 were tentatively identified
as digalloyl-glucose isomers.
Peaks 31 and 38 exhibited a [M−H]− ion at m/z 635.0881
(−1.42 ppm) with molecular formula of C27 H24 O18 . The fragment
ions at m/z 465.0689 [M−H−170 Da]– (loss of a gallic acid), m/z
313.0575 [M−H−170 Da-152 Da]– (loss of a gallic acid and a galloyl
moiety), typical fragment ion at m/z 169.0127 and fragment ion at
m/z 125.0254 were observed in the MS2 mode. Therefore, Peaks 31
and 38 were tentatively identified as trigalloyl-glucoside isomers.
As an example, the MS/MS spectra of one trigalloyl-glucoside was
shown in Fig. 2.
Peaks 45 and 49 had a [M−H]− ion at m/z 787.0972
(−3.43 ppm) with molecular formula of C34 H28 O22 . The MS2 spectra of these two peaks exhibited fragments at m/z 617.0772
[M−H−170 Da]– (loss of a gallic acid), m/z 465.0654 [M−H−170 Da152 Da]– (loss of a gallic acid and a galloyl moiety), m/z
313.0531[M−H−170 Da-152 Da-152 Da]– (loss of a gallic acid and
two galloyl moieties), the typical ion at m/z 169.0138. Therefore, peaks 45 and 49 were tentatively identified as tetragalloylglucopyranoside. As an example, the MS/MS spectra and proposed
fragmentation pattern of one tetragalloyl-glucopyranoside were
shown in Figs. 2 and 3.
As shown in Fig. 2, peak 56 exhibited deprotonated molecular
ion [M−H]− at m/z 939.1061 (−5.11 ppm) with molecular formula
of C41 H32 O26 . The fragments at m/z 769.0873 [M−H−170 Da]– , m/z
617.0772 [M−H−170 Da-152 Da]– , m/z 465.0654 [M−H−170 Da152 Da-152 Da]– ,

m/z
313.0531[M−H−170 Da-152 Da-152 Da152 Da]– , the typical fragment ion at m/z 169.0138 and fragment
ion at m/z 125.01 were observed in MS2 experiment. Therefore,
peak 56 was tentatively identified as pentagalloyl-glucoside.
According to the mass fragmentation patterns of galloyl derivatives and ellagitannins, it is suggested that if galloyl derivatives or
ellagitannins have one or more galloyl group(s) to esterify a sugar
(usually glucose), molecular ion firstly remove one molecule gallic acid (C7 H6 O5 , 170 Da), and then continuously lose one or more
galloyl group(s) (C7 H4 O4 , 152 Da) in the mass fragmentation process. The typical losses during their fragmentation are a gallic acid
(170 Da) and a galloyl moiety (152 Da).
3.1.3. Others
Peak 42 had a precursor ion [M−H]− ion at m/z 300.9998
(2.65 ppm) with molecular formula of C14 H6 O8 . The product ions
at m/z 283.9963 and m/z 257.0092 were yielded by the loss of
H2 O (18 Da) unit and CO2 (44 Da) unit from the [M−H]− ion, respectively. Furthermore, the fragments at m/z 229.0133 and m/z
201.0182 were also observed. Thus, peak 42 was absolutely identified as ellagic acid by comparing with its commercial standard.
Peak 29 exhibited a protonated molecular ion [M−H]− at m/z
247.0244 (−1.61 ppm) with molecular formula of C12 H8 O6 . The
fragment ions at m/z 219.0305 and m/z 191.0348 were occurred
via the removal of CO unit and continuing removal of CO unit from
the molecular ion, respectively. Peak 20 had molecular ion [M−H]−
at m/z 291.01 (2.40 ppm) with the molecular formula of C13 H8 O8 .
The fragment ion m/z 247.02 was generated by the loss of CO2
unit from the molecular ion. The fragment ions at m/z 219.02 and
m/z 191.03 were consistent with those of compound 29. Peaks 29
and 20 and were absolutely identified as brevifolin and brevifolincarboxylic acid by comparing with their commercial standards, respectively.
Peak 32 had a [M−H]− ion at m/z 305.0300 (−0.98 ppm) with
molecular formula of C14 H10 O8 . The fragment ions at m/z 273.0061,
m/z 245.0082, m/z 217.0141were observed in the MS2 mode. Peak
41 displayed [M−H]− ion at m/z 319.01 (−1.31 ppm) with molecular formula of C15 H12 O8 . In the MS/MS spectrum, the fragment
ions at m/z 273.0036, m/z 245.0152 and m/z 217.0167 were generated by the removal of C3 H6 O2 unit and continuing removal of CO

7

(28 Da) and two CO (28 Da) unit from the molecular ion, respectively. Moreover, the fragmentation patterns of peaks 32 and 41
were in agreement with fragmentation patterns reported in the literature [13]. Peaks 32 and 41 were tantetively identified as methyl
brevifolincarboxylate and ethyl brevifolincarboxylate.
3.1.4. Anthocyanins
Anthocyanins are naturally occurring plant pigments with
unique chromatographic behavior. Anthocyanins carry an inherent positive charge and can easily donate protons to free radicals
([M]+ ) under (+) ESI condition [19].
Peak 11 had a positively charged molecular ion [M]+ at m/z
465.1023 (−1.07 ppm) with molecular formula of C15 H11 O7 , and
MS/MS fragment ions at m/z 303.0486 [M−162 Da]+ , which this
characteristic matched with the loss of glucose (162 Da) and suggested that the aglycone was delphinidin. Thus, peak 11 was tentatively identified as delphinidin-3-O-glucoside.
Peak 15 had a positively charged molecular ion [M]+ at
m/z 611.1605 with molecular formula of C27 H31 O16 , yielding by
fragment ions at m/z 449.1045 [M−162 Da]+ and m/z 287.0522
[M−162 Da-162 Da]+ , which suggested that the aglycone was
cyanidin. Thus, peak 15 was tentatively identified as cyanidin-3, 5O-diglucoside.
Peak 18 had a positively charged molecular ion [M]+ at m/z
595.1660 (−0.50 ppm) with molecular formula of C27 H31 O15 . In
MS2 mode, the fragment ions at m/z 433.1132 [M−162 Da]+ and
m/z 271.0600 [M−162 Da-162 Da]+ were observed, which suggested
that the aglycone was pelargonidin. Thus, peak 18 was tentatively
identified as pelargonidin-3, 5-diglucoside.
Peak 26 had a positively charged molecular ion [M]+ at m/z
449.1085 with molecular formula of C21 H21 O11 , yielding by fragment ion at m/z 287.0532 [M−162 Da]+ , which suggested that the
aglycone was cyanidin. Thus, peak 26 was tentatively identified as
cyanidin-3-O-glucoside.
Peak 27 had a positively charged molecular ion [M]+ at m/z

433.1133 (−0.46 ppm) with molecular formula of C21 H21 O10 , yielding by fragment ion m/z 271.0600 [M−162 Da]+ by the loss of a
glucose moiety, which suggested that the aglycone was pelargonidin. Thus, peak 27 was tentatively identified as pelargonidin-3glucoside.
3.1.5. Flavonoids
Peaks 64 displayed the molecular ion [M+H]+ /[M−H]− at m/z
287.0554 / 285.0409 with molecular formula of C15 H10 O6 , and
main MS/MS fragment ion at m/z 153.0178. Peaks 65 exhibited precursor ion [M+H]+ /[M−H]− at m/z 271.0603 /269.0455
(0.73 ppm) with molecular formula of C15 H10 O5, and main MS/MS
fragment ion at m/z 153.0173. Peaks 66 showed the molecular ion
[M+H]+ /[M−H]− at m/z 287.0555 / 285.0407 with molecular formula of C15 H10 O6 , and main MS/MS fragment ion at m/z 153.0175.
Peaks 64, 65 and 66 were absolutely identified as luteolin, apigenin
and kaempferol by compariing with their authentic standards, respectively.
Peaks 48 showed the molecular ion [M+H]+ /[M−H]− at m/z
449.1085 / 447.0993 with molecular formula of C21 H20 O11 . In the
positive MS2 mode, the fragment ion at m/z 287.0573 (matching with the aglycone of luteolin) appeared after neutral loss of
a glucose (162 Da) moiety from the molecular ion. Peaks 52 had
[M+H]+ /[M−H]− at m/z 433.1139 / 431.0984 with molecular formula of C21 H20 O10 . The positive MS2 spectrum showed the main
product ion at m/z 271.0612 after neutral loss of a glucose (162 Da)
moiety from the molecular ion. Thus, peaks 48 and 52 were
absolutely identified as luteolin-7-O-glucoside and apigenin-7-Oglucoside by comparing with their authentic standards.
Peaks 47 was detected in both ionization modes with molecular ion [M+H]+ /[M−H]− at m/z 465.1028/463.0882 with molecular


8

Z. Yisimayili, R. Abdulla and Q. Tian et al. / Journal of Chromatography A 1604 (2019) 460472

Fig. 4. Extracted ion chromatogram, the MS/MS spectra and chemical structures (from left to right) of pelargonidin 3-O-glucoside (detected only in positive ion mode) and
apigenin-7-O-glucoside (detected in both positive and negative ion mode).

formula of C21 H20 O12 . The main fragment ion at m/z 303.0484 was

observed in the MS2 mode after the loss of a glucose (162 Da) moiety from the molecular ion. Peaks 47 was absolutely identified as
isoquercetin by comparing with its commercial standard.
It is worth to note that several flavonol glycosides and anthocyanin glycosides compounds have the same molecular ions
and mass fragmentation patterns in ESI (+) positive ionization
mode ([M]+ of anthocyanins and [M+H]+ of flavonol glycosides
are the same), occurring as mono- or di-glucosides. For example,
when the aglycone parts of two different species are quercetin
and delphnidin or pelargonidin and apigenin or kaempferol /luteolin and cyanidin. As a representative example, Fig. 4 showed
that pelargonidin 3-glucoside and apigenin-7-O-glucoside share
the same molecular ion, same fragmentation behavior and same
major fragment ions in their positive MS/MS spectrum. However,
these isomeric structures in the positive ion mode could not be
distinguished based on the MS1 and MS2 data. Thus, the similar
fragmentation patterns between these two different species complicate accurate structural elucidation and remain challenging. For
the accurate identification, the three strategies below were taken
into account. First, both positive and negative ionization mode
were used to determine the molecular weight. The identification of
the flavones were carried out based on the observation of the protonated and deprotonated molecules ([M+H]+ and [M−H]− ions),
which have also been described by other authors using IT /Q-TOF
[16,17,21]. Anthocyanins carry an inherent positive charge and can
easily donate protons to free radicals ([M]+ ) under (+) ESI condition. Thus, most reported LC/MS studies of anthocyanins were also
performed in the positive ion mode because of maximum sensitivity [19–24]. Anthocyanins do not ionize at all because of the absence of a free hydroxyl group in the negative ion mode, and deprotonated molecular ion ([M–H]– ) for anthocyanins could not be
detected because of the neutralization of the charge [19]. Thus, the
unique molecular ion [M–2H]– in negative ion mode analysis may
provide additional information for identification of anthocyanins

compounds, but [M–2H]– of anthocyanins and [M–H]– of flavonol
glycosides are also the same. Only a limited number of LC/MS studies of anthocyanins were carried out with the molecular ion [M–
2H]– in negative ion mode, but the full scan MS1 spectrum from
negative ionization mode was complex [22–24].

In this study, the molecular ion [M–2H]– for anthocyanins was
not detected in the extracted ion chromatogram (EIC) when extracting this specific mass from the negative full-scan MS1 dataset,
which there was no mixed peak with the similar retention time
when compared with the positive ion mode. Thus, for the identification of anthocyanins, the positive ion mode was used for
their identification and the negative ionization mode was used
for verification in this study. This characteristic is especially useful for distinguishing anthocyanin glycosides from flavonol glycosides with the same ‘quasi’-molecular ions (M+ = [M+H]+ and
[M–2H]– = [M–H]– ) that co-exist in some plants. It becomes
very simple to distinguish between two different species if negative ionization mode is employed. Thus, it is easy to distinguish rapidly between pelargonidin 3-glucoside and apigenin-7-Oglucoside when comparing the full scan MS spectrum in the negative ionization mode. Secondly, anthocyanins compounds have a
characteristic elution order in reversed phase liquid chromatography (RP-LC), which elute before the flavonol glycosides [25]. The
retention times of the pelargonidin 3-O-glucoside (27.11 min) and
apigenin-7-O-glucoside (55.90 min) on a reverse phase C18 column differed by 28.8 min, which indicate that peak 27 (pelargonidin 3-O-glucoside) must be an anthocyanin with much higher polarity. Thirdly, apigenin-7-O-glucoside were further confirmed by
comparing with the fragmentation pattern and chromatographic
retention time of its authentic standard. Besides, according the
studies previously reported, a photodiode array detector (DAD) to
measure UV/Vis molecular absorbance was used to differentiate
these two different species since anthocyanins have a typical λmax
at ∼330 nm and between 440 and 540 nm [5,18,20,23]. The absorbance maxima for flavonol glycosides were at 250 and 370 nm


Z. Yisimayili, R. Abdulla and Q. Tian et al. / Journal of Chromatography A 1604 (2019) 460472

[21,24]. These strategies are significantly vital considering that MS
and MS/MS date obtained under ESI conditions do not allow identification of isomeric structures for distinguishing anthocyanin glycosides from flavonol glycosides.
3.2. Identification and analysis of absorbed compounds and their
metabolites in rat biosamples
How to find new candidates as bioactive compounds from
medicinal herbs with a higher probability for research and development in drug discovery remain challenging and controversial.
A common strategy for identifying the active constituents from
medicinal herbs in the field of separation science is to screen
bioactivity of the isolates with in vitro cell systems evaluation, although there has been a steadily rise in animal and human studies [3,5,6,8,9,11]. However, sometimes the biological effects of the

screened compounds do not live up to in vivo study. Because the
ingested compounds, at least part of them, reach the circulatory
system and specific tissues to exert biological effect as a result
of in vivo process of absorption, distribution, metabolism and excretion. It is known that the absorbed compounds could be further metabolized by various drug-metabolizing enzymes in vivo
[21,25,28]. For example, the previous studies reported the isolation
of phenolics from pomegranate flowers such as punicatannins A
and B, 1, 6-di-O-galloyl-β -D-glucose and 3, 4, 6-tri-O-galloyl-β -Dglucose. Furthermore, they evaluated the abilities of these isolates
to inhibit α -glucosidase inhibitory activities in vitro for seeking the
bioactive antidiabetic compounds from pomegranate flowers [6,11].
Thus, we were interested in absorption, distribution and metabolic
fate of the compounds after oral administration of pomegranate
flowers extract in vivo, especially those isolates.
In the present study, according to accurate mass and fragmentation pattern generated by the HPLC-Q-TOF-MS2 , 22 absorbed
compounds and their 35 metabolites were absolutely or tentatively identified in rat biosamples after oral administration of
pomegranate flowers extract. The workflow of metabolite identification take three steps, firstly, the probable metabolites were
postulated based on the metabolism rules of compounds. Secondly, the molecular ion [M+H]+ /[M]+ or [M−H]− for probable metabolites were extracted from the full-scan MS1 dataset of
dosed rat biological samples. Thirdly, the peaks detected in the
EIC were further analyzed by the QTOF-MS/MS dataset of dosed
rat biological samples. Among them, 15 absorbed compounds and
metabolites were further confirmed with their authentic standards. Among the metabolites, most of them were found in urine
(19 absorbed compounds and 31 metabolites), feces (21 absorbed
compounds and 25 metabolites) and plasma (15 absorbed compounds and 17 metabolites) samples, only a few of them were
found in tissues, respectively (Table 2). Ellagitannins were abundant in pomegranate flowers extract (18 ellagitannins), but only
corilagin was detected in plasma and tissues. Galloyl derivates
were also abundant in pomegranate flowers extract (14 galloyl
derivates), but none of them was detected in plasma and tissues. Our results indicate that ellagitannins and galloyl derivates
were not well absorbed in plasma and tissues. It is worth to note
that the isolates (punicatannins A and B, 1, 6-di-O-galloyl-β -Dglucose and 3, 4, 6-tri-O-galloyl-β -D-glucose) were not found in
plasma or tissues after oral administration of pomegranate flowers extract. This our in vivo finding have not totally supported
the in vitro findings that these isolates were bioactive antidiabetic

compounds present in pomegranate flower. According to previous studies, after normal consumption of ellagitannins-rich foods
or extracts, ellagitannins are rarely detected in plasma due to
their low bioavailability. Ellagitannins are metabolized by the intestinal flora to produce ellagic acid and urolithins metabolites
[27–30].

9

Moreover, most of metabolism studies of ellagitannins were
mainly focused on ellagic acid, which is one of the main
hydrolysates of ellagitannins, and urolithins and their derived
metabiolites in plasma, urine and feces in recent years [26–28].
Remarkably, in the present study, ellagitannin corilagin and nine
phase II conjugate metabolites of corilagin were firstly identified
in plasma and tissues after oral administration of pomegranate
flowers extract. As shown in Fig. 5, metabolite C5 (retention
time = 11.96 min) had a molecular ion [M−H]− at m/z 809.1056
(C28 H29 O26 − , 0.24 ppm), which was 176 Da higher than that of corilagin (O5), suggesting that C5 was glucuronide conjugate metabolite. The fragment ion at m/z 633.0676, as a base peak, was produced by natural loss of a glucuronic acid (176 Da) from the
[M−H]− ion. Besides, the fragment ion at m/z 463.0532, the typical fragment ions at m/z 300.9958 and m/z 169.0138 were also
consistent with those of corilagin. Thus, C5 was identified as glucuronidation of corilagin. This is a first report to identify monoglucuronide conjugated metabolite of ellagitannin compound in
vivo. Metabolites C1 and C2 had molecular ion [M−H]− at m/z
713.0298 (C27 H21 O21 S− , −0.56 ppm), which was 80 Da (SO3 ) higher
than that of corilagin. Thus, C1 and C2 were identified as sulfation metabolites of corilagin. Metabolite C3 had a molecular ion
[M−H]− at m/z 647.0894 (C28 H23 O18 − , 0.61 ppm), which was 14 Da
(CH2 ) higher than that of corilagin. Therefore, C3 was identified
as methylation metabolite of corilagin. Metabolite C4 had a molecular ion [M−H]− at m/z 661.1048 (C29 H25 O18 − , 0.30 ppm), which
was 28 Da (2∗ CH2 ) higher than that of corilagin. Therefore, C4 was
identified as Di-methylation metabolite of corilagin. Metabolites C6
and C7 had molecular ion [M−H]− at m/z 823.1208 (C34 H31 O24 − ,
−0.36 ppm), which was 176 Da (a glucuronic acid) and 14 Da (CH2 )
higher than that of corilagin. Therefore, C6 and C7 were identified as glucuronidation and methylation metabolites of corilagin. Metabolite C8 had a molecular ion [M−H]− at m/z 837.1364

(C35 H33 O24 − , −0.67 ppm), which was176 Da (a glucuronic acid)
and 28 Da (2∗ CH2 ) higher than that of corilagin. Therefore, C8 was
identified as glucuronidation and di-methylation metabolite of corilagin. The fragmentation patterns of metabolites C1–C4 and C6–
C8 were similar to that those of metabolites previously reported
for corilagin (Fig. 5) [29]. Furthermore, the binding sites of phase
II conjugate metabolites (C1–C9) of corilagin were determined by
combining with our previous study [29]. The nine phase II conjugate metabolites of corilagin are the methylation, glucuronidation and sulfation conjugated metabolites. This finding raises the
possibility that phase II conjugate metabolites of ellagitannin corilagin may function as biological antioxidant, anti-inflammatory,
α -glucosidase inhibitory and hepatoprotective activities after oral
administration of pomegranate flowers extract.
Besides, 17 metabolites of ellagic acid including urolithins and
their derived metabolites were identified in this study. Ellagic acid
and gallic acid were identified in plasma and tissues. Not only ellagic acid and gallic acid are two main compounds in pomegranate
flowers, as shown in Fig. 6, but also two main the hydrolyzed
metabolite of ellagitannins in rats after oral administration of
pomegranate flowers extract. This may illustrate the wide distribution of ellagic acid and gallic acid in rat biosamples. Metabolites
urolithin D, urolithin C, urolithin A and urolithin B were identified in plasma and some tissues and further confirmed with their
standards. Besides, sulfation, methylation, glucuronidation metabolites of urolithins were found in plasma but not much in different organs, some of which have been identified in previous studies
[30]. The fragmentation patterns of them were consistent with previous study [30]. Flavonoids were also abundant in pomegranate
flowers extract (18 flavonoids), and most of them was detected
in plasma and tissues. The free and mainly glucuronide conjugated flavonoids were identified in plasma and liver. The flavonoids
glucuronide metabolites showed typical loss of a 176 Da (a glu-


10

Table 2
Summary of the mass spectral data and distribution of absorbed compounds and metabolites detected in the rat biological samples orally administrated with pomegranate flower extract.
tR
(min)

Biotransformation

Formula
(neutral)

[M−H]−
Calculated

Observed

MS/MS
fragments

Error
(ppm)

U

F

P

L

S

K

H

Lg

O1
G1
G2
G3
O2
O3
O4
O5a
C1
C2
C3
C4
C5b
C6
C7
C8
C9
O6a
E1
E2
E3
E4
E5
E6
E7
E8
E9a

E10

2.02
8.45
11.24
3.32
5.90
14.60
21.50
26.46
8.04
24.84
43.56
53.47
20.18
22.66
44.91
5.03
33.19
46.95
8.81
9.56
56.36
55.35
57.38
46.43
50.68
47.78
36.16
27.51

Gallic acid
Methylation of gallic acid
Methylation of gallic acid
Di-methylation of gallic acid
Digalloyl-glucoside
Galloyl-HHDP-glucoside
Galloyl-HHDP-glucoside
Corilagin
Sulfation of corilagin
Sulfation of corilagin
Methylation of corilagin
Di-methylation of corilagin
Glucuronidation of corilagin
Glucuronidation and methylation of corilagin
Glucuronidation and methylation of corilagin
Glucuronidation and di-methylation of corilagin
Di-glucuronidation of corilagin
Ellagic acid
Sulfation of ellagic acid
Glycosylation of ellagic acid
Methylation of ellagic acid
Methylation and sulfation of ellagic acid
Methylation and sulfation of ellagic acid
Glucuronidation and methylation of ellagic acid
Glucuronidation and di-methylation of ellagic acid
Methylation and glycosylation of ellagic acid
Urolithin D
Sulfation and di-methylation of Urolithin D

C7 H6 O5
C8 H8 O5
C8 H8 O5
C8 H8 O5
C20 H20 O14
C27 H22 O18
C27 H22 O18
C27 H22 O18
C27 H22 O21 S
C27 H22 O21 S
C28 H24 O18
C29 H26 O18
C28 H30 O26
C34 H32 O24
C34 H32 O24
C35 H34 O24
C39 H38 O30
C14 H6 O8
C14 H6 O11 S
C20 H16 O13
C15 H8 O8
C15 H8 O11 S
C15 H8 O11 S
C21 H16 O14
C22 H18 O14
C21 H18 O13
C13 H8 O6
C15 H12 O9 S

169.0142

183.0299
183.0299
197.0455
483.0780
633.0733
633.0733
633.0733
713.0302
713.0302
647.0890
661.1046
809.1054
823.1211
823.1211
837.1367
985.1375
300.9990
380.9558
463.0518
315.0146
394.9715
394.9715
491.0467
505.0624
477.0675
259.0248
367.0129

169.0141
183.0297

183.0297
197.0452
483.0773
633.0726
633.0726
633.0733
713.0298
713.0298
647.0894
661.1048
809.1056
823.1208
823.1208
837.1364
985.1386
300.9994
380.9555
463.0522
315.0140
394.9718
394.9718
491.0463
505.0621
477.0679
259.0245
367.0129

125.0241
168.0061, 124.0165
168.0067, 124.0165

169.0126, 125.0245
313.0567, 169.0139, 125.0248
463.0514, 300.9992, 169.0144
463.0519, 300.9993, 169.0140
463.0524, 300.9991, 169.0136
633.0713, 463.0522, 300.9993,
633.0713, 463.0522, 300.9993,
463.0490, 300.9990
477.0641, 315.0149, 169.0138
633.0676,463.0532, 300.9958
647.0860, 463.0538, 300.9991
647.0860, 463.0538, 300.9991
661.1035, 477.0670, 315.0153
633.0781, 300.9973
283.9972, 257.0091, 229.0137,
300.9992, 229.0169
300.9995, 201.0220
299.9910
315.0146, 299.9905
315.0146, 299.9905
315.0146, 299.9908, 201.0213
329.0308, 314.0066, 299.9909,
315.0146, 299.9903, 201.0202
241.0152, 231.0299
287.0575, 259.0604

−0.59
−1.09
−1.09
−1.52

−1.44
−1.10
−1.10
0
−0.56
0.56
0.61
0.30
0.24
−0.36
−0.36
−0.67
−0.35
1.32
−0.78
0.86
−1.90
0.75
0.75
−0.81
−0.59
0.83
−1.15
0

+
+
+
+
–

+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+

+
+
+
–
+
–

+
+
+
+
+
+
–
+
+
–
–
+
+
+
–
+
+
+
+
+
+
+

+
+
+
–
–
–
–

+
–
–
+
+
+
+
+
+
–
+
–
–
+
+
–
–
–
+
+
+

+
–
–
–
–
–
–
+

–
–
+
–
–
+
+
+
–
+
+
–
+
+
+
+
–
+
–
–

+
–
–
–
–
–
–
+
–

–
–
–
–
–
–
–
–
+
–
–
+
–
–
–
–
+
–
–

+
–
–
–
–
–
–
+
–
–

–
–
–
+
+
+
–
+
–
–
–
–
–
–
–
–
–
+

+
–
–
–
–
–
–
+
–
–
–

–
–
+
+
–
–
+
–
–
–
–
–
–
–
+
–
–

–
–
–
–
–
–
–
–
–
–
–
–

–
–
–
–
–
+
–
+
–
–
+
+
–
–
–
–

169.0144
169.0144

201.0192

201.0234

(continued on next page)

Z. Yisimayili, R. Abdulla and Q. Tian et al. / Journal of Chromatography A 1604 (2019) 460472

NO

Table 2 (continued)

E11
E12a
E13
E14
E15a
E16
E17a
O7a
O8a
O9a
O10
O11
O12
O13
O14a
O15a
A1
O16
O17a
O18a
L1
L2
O19a
O20
O21
T1
T2

T3
O22
a
b

tR
(min)

Biotransformation

11.96
11.14
13.77
9.37
58.98
35.17
67.47
20.23
24.32
29.12
32.34
46.64
34.66
18.83
67.11
56.44
52.79
43.25
65.75
50.29

54.99
41.62
67.29
56.40
58.65
61.29
49.28
51.81
67.30

Methylation and glucuronidation of Urolithin D
Urolithin C
Glucuronidation of urolithin C
Di-glucuronidation of urolithin C
Urolithin A
Sulfation and glucuronidation of Urolithin A
Urolithin B
Brevifolincarboxylic acid
Ethyl gallate
Brevifolin
Methyl brevifolincarboxylate
Ethyl brevifolincarboxylate
Digalloyl-HHDP-glucoside
Pelargonidin 3,5-diglucoside
Apigenin
Apigenin-7-O-glucoside
Glucuronidation of apigenin
Apigenin-di-glucoside
Luteolin
Luteolin 7-O-glucoside

Glucuronidation of luteolin
Di-glucuronidation of luteolin
Kaempferol
Kaempferol-3-O- glucoside
Tricetin
Glucuronidation and di-methylation of tricetin
Di-lucuronidation of tricetin
Di-glucuronidation and methylation of tricetin
Tricin

Formula
(neutral)

[M−H]−
Calculated

Observed

MS/MS
fragments

C21 H22 O11
C13 H8 O5
C19 H16 O11
C25 H24 O17
C13 H8 O4
C19 H16 O13 S
C13 H8 O3
C13 H8 O8
C9 H10 O5

C12 H8 O6
C14 H10 O8
C15 H12 O8
C34 H26 O22
C27 H31 O15
C15 H10 O5
C21 H20 O10
C21 H18 O11
C27 H31 O15
C15 H10 O6
C21 H20 O11
C21 H18 O12
C27 H26 O18
C15 H10 O6
C21 H20 O11
C15 H10 O7
C23 H22 O13
C27 H26 O19
C28 H28 O19
C17 H14 O7

449.0725
243.0299
419.0620
595.0941
227.0350
483.0239
211.0401
291.0146
197.0455

247.0248
305.0303
319.0459
785.0843
595.1663
271.0601
433.1129
447.0922
595.1663
287.0550
449.1078
463.0871
639.1192
287.0550
449.1078
303.0499
507.1133
655.1141
669.1298
331.0812

449.0731
243.0294
419.0617
595.0936
227.0348
483.0234
211.0402
291.0146
197.0450

247.0242
305.0300
319.0455
785.0838
595.1669
271.0605
443.1133
447.0926
595.1668
287.0555
449.1085
463.0877
639.1189
287.0558
449.1088
303.0505
507.1139
655.1136
669.1292
331.0818

273.0409,
215.0349,
243.0254
419.0614,
198.0320,
403.0688,
167.0502,
247.0252,
169.0144,

219.0305,
273.0061,
273.0036,
615.0602,
433.1133,
153.0173
271.0612
271.0612
433.1133,
153.0178
287.0573,
287.0573,
463.0871,
153.0175
287.0573,
153.0193
331.0825,
479.0829,
493.0984,
315.0496,

259.0606
187.0401
243.0245
182.0371
227.0359
139.0552
219.0296,
125.0242
191.0348

245.0082,
245.0085,
463.0499,
271.0606

191.0348

217.0141
217.0141
300.9990, 169.0130

271.0606
153.0187
153.0187
287.0573, 153.0187
153.0187
303.0499
303.0499
317.0567, 303.0513
299.0549, 133.1010

Error
(ppm)

U

F

P

L

S

K

H

Lg

1.33
−2.05
−0.71
−0.84
−0.88
−1.03
0.28
0
−2.53
−2.42
−0.98
−1.25
−0.63
1.00
1.47
0.92
0.89
0.84
1.74
1.55

1.29
−0.46
2.78
2.22
1.96
1.18
−0.76
−0.89
1.81

+
+
–
–
+
–
–
+
–
+
+
+
–
+
+
+
+
+
+
+

+
+
+
+
+
+
+
+
+

+
+
+
–
–
–
+
+
+
+
+
+
+
+
+
+
+
+
+
+

+
+
+
+
+
+
–
–
+

–
+
–
–
+
+
+
+
–
+
+
+
–
+
+
+
+
–
+
+

+
–
+
–
+
–
–
–
+

+
–
–
+
–
+
+
+
–
+
+
+
–
–
+
+
+
–
+
+

+
–
+
–
+
–
–
–
+

–
–
–
–
–
–
–
+
–
+
+
–
–
–
+
–
+
–
–
–

+
–
–
–
–
–
–
–
–

–
+
+
–
–
–
–
–
–
+
+
–
–
–
+
–
–
–
–
–

–
–
–
–
–
–
–
–
–

–
–
–
+
–
–
–
+
–
+
–
–
–
–
–
–
–
–
–
–

–
–
–
–
–
–
–
–
–

–
+
–
–
–
–
–
–
–
+
–
–
–
–
–
–
–
–
–
–

–
–
–
–
+
–
–
–
–

Z. Yisimayili, R. Abdulla and Q. Tian et al. / Journal of Chromatography A 1604 (2019) 460472

NO

Confirmed by using reference standard.
Firstly identified metabolites in vivo.

11


12

Z. Yisimayili, R. Abdulla and Q. Tian et al. / Journal of Chromatography A 1604 (2019) 460472

Fig. 5. Extracted ion chromatogram and the MS/MS spectra of phase II conjugate metabolites of corilagin in rats after oral administration of pomegranate flowers extract.

Fig. 6. The possible metabolic pathways of corilagin (ellagitannins to urolithins) in rats orally administered with pomegranate flowers extract. (Me: Methylation; GluA:
Glucuronidation; Glu: Glycosylation; SO3 : Sulfation). Solid and dotted box represent the confirmed and possible determination of binding sites of respective metabolites.

curonic acid), which were traditionally considered to be monoglucuronides. However, except absorbed compound pelargonidin3, 5-O-diglucoside, there was not detected any metabolites of

the four anthocyanins in plasma and tissues. Interestingly, brevifolin and its three derivates were all detected in plasma and

liver. This result indicates that brevifolin and its derivates were
well absorbed in plasma. Thus, characterizing its multiple constitution, absorption and metabolic fate of these compounds in vivo
is helpful to better analyze the active components in pomegranate
flowers.


Z. Yisimayili, R. Abdulla and Q. Tian et al. / Journal of Chromatography A 1604 (2019) 460472

4. Conclusion
In summary, the present study applied HPLC-Q-TOF-MS2 to
characterize the polyphenols composition of pomegranate flowers
and then their appearance as native form, including their metabolites in rat urine, feces, plasma and tissues after oral administration of pomegranate flowers extract. The 67 compounds identified
in pomegranate flowers, but only 22 compounds detected in rat
biosamples. This result showed that not all compounds abundant
in pomegranate flowers extract could be absorbed well in plasma
and tissues. This finding also suggested a potential correlation between study on metabolic profile of these compounds in vivo and
study on strategy of screening bioactivity of the isolates with in
vitro cell systems evaluation. To the best of our knowledge, this
is the first time to analyze the metabolic profile of pomegranate
flowers in vivo. This study was expected to provide significant information to find possible candidates for the real bioactive compounds in pomegranate flowers and provide a solid basis for the
study of quality control of pomegranate flowers.
Declaration of Competing Interest
None.
Acknowledgment
This present study was financially supported by The Science and Technology Service Network Initiative of the Chinese
Academy of Sciences (KFJ-STS-QYZD-066). The authors thank professor Chenggang Huang for supporting this study.
Reference
[1] J. Xu, Y. Zhao, H.A. Aisa, Anti-inflammatory effect of pomegranate flowers in

lipopolysaccharide (LPS)-stimulated RAW264.7 macrophages, Pharm. Biol. 55
(2017) 2095–2101.
[2] A. Aslani, B. Zolfaghari, F. Davoodvandi, Design, formulation and evaluation of an oral gel from Punica granatum flowers extract for the treatment of recurrent aphthous stomatitis, Adv. Pharm. Biol. 6 (2016) 391–
398.
[3] A. Kam, K.M. Li., V. Razmovski-Naumovski, S. Nammi, J. Shi, K. Chan, G.Q. Li,
A comparative study on the inhibitory effects of different parts and chemical
constituents of pomegranate on a-amylase and a-glucosidase, Phytother. Res.
27 (2013) 1614–1620.
[4] M. Ali, N. Sharma, Phytochemical investigation of the flowerss of Punica granatum, Indian J. Chem. 45B (2006) 1681–1685.
[5] L. Zhang, Q. Fu, Y. Zhang, Composition of anthocyanins in pomegranate
flowerss and their antioxidant activity, Food Chem. 127 (2011) 1444–
1449.
[6] T. Yuan, Y. Ding, C. Wan, L. Li, J. Xu, K. Liu, A. Slitt, D. Ferreira, I.A. Khan,
N.P. Seeram, Antidiabetic ellagitannins from pomegranate flowerss: inhibition of α -glucosidase and lipogenic gene expression, Org. Lett. 14 (2012)
5358–5361.
[7] Z. Ben-Simhon, S. Judeinstein, T. Trainin, R. Harel-Beja, I. Bar-Ya’akov, H. Borochov-Neori, D. Holland, A "White" anthocyanin-less pomegranate (Punica
granatum L.) caused by an insertion in the coding region of the leucoanthocyanidin dioxygenase (LDOX; ANS) gene, PLOS ONE. 10 (2015) 1–21.
[8] M. Russo, C. Fanali1, G. Tripodo, P. Dugo, R. Muleo, L. Dugo, L.D. Gara, L. Mondello, Analysis of phenolic compounds in different parts of pomegranate
(Punica granatum) fruit by HPLC-PDA-ESI/MS and evaluation of their antioxidant activity: application to different Italian varieties, Anal. Bioanal. Chem. 410
(2018) 3507–3520.
[9] R.S.E. Dine, Q. Ma, Z.A. Kandil, A.M. El-Halawany, Triterpenes as uncompetitive
inhibitors of α glucosidase from flowerss of Punica granatum L, Nat. Prod. Res.
28 (2014) 2191–2194.

13

[10] G. Kaur, Z. Jabbar, M. Athar, M.S. Alam, Punica granatum (pomegranate) flowers extract possesses potent antioxidant activity and abrogates FE-NTA induced
hepatotoxicity in mice, Food Chem. Toxicol. 44 (2006) 984–993.
[11] T. Yuan, C. Wan, H. Ma, N.P. Seeram, New phenolics from the flowerss of Punica
granatum and their in vitro α -glucosidase inhibitory activities, Planta Med. 79

(2013) 1674–1679.
[12] R. Garcia-Villalba, J.C. Espin, K. Aaby, C. Alasalvar, M. Heinonen, G. Jacobs,
S. Voorspoels, T. Koivumaki, P.A. Kroon, E. Pelvan, S. Saha, F.A. Tomas-Barberan,
Validated method for the characterization and quantification of extractable and
nonextractable ellagitannins after acid hydrolysis in pomegranate fruits, juices,
and extracts, J. Agric. Food Chem. 63 (2015) 6555–6566.
[13] M. Zhu, X. Dong, M. Guo, Phenolic profiling of duchesnea indica combining macroporous resin chromatography (MRC) with HPLC-ESI-MS/MS and ESI-IT-MS, Molecules 20 (2015) 22463–22475.
[14] Y. Liu, N.P. Seeram, Liquid chromatography coupled with time-of-flight tandem mass spectrometry for comprehensive phenolic characterization of
pomegranate fruit and flowers extracts used as ingredients in botanical dietary
supplements, J. Sep. Sci. 41 (2018) 3022–3033.
[15] A. Taamalli, I. Iswaldi, D. Arraez-Roman, A. Segura-Carretero, A. Fernandez-Gutierrez, M. Zarrouk, UPLC-QTOF/MS for a rapid characterisation of phenolic compounds from leaves of Myrtus communis L, Phytochem. Anal. 25
(2014) 89–96.
[16] R. Abdulla, S. Mansur, H. Lai, A. Ubul, G. Sun, G. Huang, H.A. Aisa, Qualitative
analysis of polyphenols in macroporous resin pretreated pomegranate husk extract by HPLC-QTOF-MS, Phytochem. Anal. 28 (2017) 465–473.
[17] A. Mahfoudhi, F.P. Prencipe, Z. Mighri, F. Pellati, Metabolite profiling of
polyphenols in the Tunisian plant Tamarix aphylla (L.) Karst, J. Pharm. Biomed.
Anal. 99 (2014) 97–105.
[18] V. Brighenti, S.F. Groothuis, F.P. Prencipe, R. Amir, S. Benvenuti, F. Pellati,
Metabolite fingerprinting of Punica granatum L. (pomegranate) polyphenols by
means of high-performance liquid chromatography with diode array and electrospray ionization-mass spectrometry detection, J. Chromatogr. A 1480 (2017)
20–31.
[19] N. Acero, A. Gradillas, M. Beltran, A. Garcia, D.M. Mingarro, Comparison of phenolic compounds profile and antioxidant properties of different sweet cherry
(Prunus avium L.) varieties, Food Chem. 279 (2019) 260–271.
[20] M. Lee, J.S. Park, D.S. Choi, M.Y. Jung, Characterization and quantitation of anthocyanins in purple-fleshed sweet potatoes cultivated in Korea by HPLC-DAD
and HPLC-ESI-QTOF-MS/MS, J. Agric. Food Chem. 61 (2013) 3148–3158.
[21] X. Ying, M. Liu, Q. Liang, M. Jiang, Y. Wang, F. Huang, Y.I. Xie, J. Shao, G. Bai,
G. Luo, Identification and analysis of absorbed components and their metabolites in rat plasma and tissues after oral administration of ‘Ershiwuwei shanhu’
pill extracts by UPLC-DAD/Q-TOF-MS, J. Ethnopharmacol. 150 (2013) 324–338.
[22] Q. Tian, R.M. Aziz, G.D. Stoner, S.J. Schwartz, Anthocyanin determination in
black raspberry (Rubus occidentalis) and biological specimens using liquid

chromatography-electrospray ionization tandem mass spectrometry, J. Food
Sci. 70 (2005) C43–C48.
[23] H. Olsen, K. Aaby, G.I.A. Borge, Characterization, quantification, and yearly variation of the naturally occurring polyphenols in a common red variety of curly
kale (Brassica oleracea L. convar. acephala var. sabellica cv. ’Redbor’), J. Agric.
Food Chem. 58 (2010) 11346–11354.
[24] J. Sun, L.Z. Lin, P. Chen, Study of the mass spectrometric behaviors of anthocyanins in negative ionization mode and its applications for characterization of
anthocyanins and non-anthocyanin polyphenols, Rapid Commun. Mass Spectrom. 26 (2012) 1123–1133.
[25] X. Zhang, A. Sandhu, I. Edirisinghe, B. Burton-Freeman, An exploratory study of
red raspberry (Rubus idaeus L.) (poly)phenols/metabolites in human biological
samples, Food Funct. 9 (2018) 806–819.
[26] C. Garcia-Nozi, F. Vaillant, Metabolic fate of ellagitannins: implications for
health, and research perspectives for innovative functional foods, Crit. Rev.
FoodSci. Nutr. 54 (2014) 1584–1598.
[27] R. Gonzalez-Barrio, P. Truchado, H. Ito, J.C. Espín, F.A. Tomas-Barberan, UV and
MS identification of urolithins and nasutins, the bioavailable metabolites of
ellagitannins and ellagic acid in different mammals, J. Agric. Food Chem. 59
(2011) 1152–1162.
[28] M. Larrosa, M.T. García-Conesa, J.C. Espín, F.A. Tomás-Barberán, Ellagitannins,
ellagic acid and vascular health, Mol. Asp. Med. 31 (2010) 513–539.
[29] Z. Yisimayili, X. Guo, H. Liu, Z. Xu, R. Abdullaa, H.A. Aisa, C. Huang, Metabolic
profiling analysis of corilagin in vivo and in vitro using high-performance liquid chromatography quadrupole time-of-flight mass spectrometry, J. Pharm.
Biomed. Anal. 165 (2019) 251–260.
[30] J.C. Espin, R. Gonzalez-Barrio, B. Cerda, C. Lopez -Bote, A.I. Rey, F.A. Tomas-Barberan, Iberian pig as a model to clarify obscure points in the bioavailability
and metabolism of ellagitannins in humans, J. Agric. Food Chem. 55 (2007)
10476–10485.