Anti-inflammatory Effects of Punica granatum Linne in Vitro and in Vivo

(1)

Anti-inﬂammatory effects of Punica granatum Linne in vitro and in vivo

Chia-Jung Lee

a

, Lih-Geeng Chen

b

, Wen-Li Liang

a

, Ching-Chiung Wang

a,*

a

School of Pharmacy, College of Pharmacy, Taipei Medical University, 250 Wu-Hsing Street, Taipei 110, Taiwan b

Graduate Institute of Biomedical and Biopharmaceutical Sciences, College of Life Sciences, National Chiayi University, 300 University Road, Chiayi 600, Taiwan

a r t i c l e

i n f o

Article history: Received 1 July 2008

Received in revised form 22 April 2009 Accepted 29 April 2009 Keywords: Punica granatum L. Granatin B Ellagictannin Anti-inﬂammation Nitric oxide

Inducible nitric oxide synthase Cyclooxygenase-2

a b s t r a c t

Inflammation can cause various physical dysfunctions. Punica granatum Linne (pomegranate), a high phe-nolic content fruit, is widely used as an antipyretic analgesic in Chinese culture. Pomegranate has shown potential nitric oxide (NO) inhibition in LPS-induced RAW 264.7 macrophage cells. Moreover, pomegran-ate (100 mg/kg) significantly decreased carrageenan-induced mice paw edema for 1, 3, 4, and 5 h. There-fore, column chromatography combined with in vitro bioassay-guided fractionation was used to isolate the active anti-inflammatory components from the pomegranate. Punicalagin (1), punicalin (2), strictinin A (3), and granatin B (4) were obtained with yields of 0.093%, 0.015%, 0.003%, and 0.013%, respectively. All these hydrolysable tannins inhibited NO production and iNOS expression in RAW 264.7 cells. Among them, 4 showed the strongest iNOS and COX-2 inhibitory effects, and exhibited these effects in the inhi-bition of paw swelling and the PGE2level in carrageenan-induced mice. Taken together, we suggest that 4 could be used as a standard marker for the anti-inflammatory effect of pomegranate.

1. Introduction

Punica granatum Linne, pomegranate (Punicaceae), a common fruit in the Mediterranean and Iran, is widely used for therapeutic formulae, cosmetics, and food seasoning. Pomegranate, also easily acquired from traditional medicine markets, was usually used as an astringent agent (Alper & Acar, 2004), for eliminating parasites (Mudzhiri, 1954; Raj, 1975) and as an antipyretic. The pharmaco-logical functions of pomegranate include antioxidation (Lansky & Newman, 2007), anti-tumour (Khan, Afaq, Kweon, Kim, & Mukhtar, 2007; Lansky & Newman, 2007), anti-hepatotoxicity (Kaur, Jabbar, Athar, & Alam, 2006), anti-lipoperoxidation (Reddy, Gupta, Jacob, Khan, & Ferreira, 2007) and anti-bacteria properties (Menezes, Cordeiro, & Viana, 2006). In hematology, pomegranate could reduce the common carotid intima-medium thickness, thus lower-ing blood pressure and decreaslower-ing low-density lipoprotein (LDL) oxidation and the incidence of heart disease (Aviram et al., 2002, 2004). In our previous studies, we found that extract from the dried peel of the pomegranate could signiﬁcantly inhibit NO production. Hence, we suggested pomegranate contains the anti-inﬂammatory activity components.

In the last few years, many important functions of fresh fruits and vegetables have been reported, and they are now recognised as being good sources of natural antioxidants (Joseph,

Shukitt-Hale, & Lau, 2007), such as grapes, apples, and guavas. The antiox-idants can prevent lipid peroxidation, and DNA and protein dam-age. Polyphenols have been acknowledged to have health-beneficial effects, owing to derived products such as flavonoids, tannins, coumarins, and lignans. According to recent reports, the pomegranate is rich in polyphenols, including mainly ellagitannins, gallotannins (punicalin, punicalagin, pedunculagin, punigluconin, granatin B, and tellimagrandin I) (Satomi et al., 1993) and anthocy-anins (delphinidin, cyanidine and pelargonidin) (Noda, Kaneyuki, Mori, & Packer, 2002). However, the correlation between the phy-tochemicals and the anti-inflammatory properties of the dried peel of the pomegranate has not been investigated. Therefore, this study aimed to clarify the anti-inflammatory activities of pomegranate and its active components.

Inflammation, the first physiological defense system in the hu-man body, can protect against injuries caused by physical wounds, poisons, etc. This defense system, also called short-term inflamma-tion, can destroy infectious microorganisms, eliminate irritants, and maintain normal physiological functions. However, long-term over-inflammation might cause dysfunctions of the regular physi-ology, i.e., asthma and rheumatic arthritis.

We used in vitro and in vivo models to confirm the anti-inflam-matory activity of pomegranate. Lipopolysaccharide (LPS)-induced RAW 264.7 murine macrophages were used in the in vitro study, and carrageenan-induced paw edema in mice served as the in vivo study. LPS can induce several cytokines, such as prostaglan-dins and nitric oxide (NO), which are involved in pro-inflammatory processes. NO can kill bacteria and viruses and is also an important

* Corresponding author. Tel.: +886 2 27361661x6161; fax: +886 2 27329368. E-mail addresses: m303092003@tmu.edu.tw (C.-J. Lee), lgchen@mail.ncyu.e-du.tw(L.-G. Chen),wenlee@tmu.edu.tw(W.-L. Liang),crystal@tmu.edu.tw(C.-C. Wang).

Contents lists available atScienceDirect

Food Chemistry

(2)

Dried peels of Punica granatum, 600g

Homogenization with 70% acetone Filtration to remove the residue 70% acetone extract, 331g

(150.2)

Partition with n-hexane

H2O layer, 308g n-hexane layer, 17g

(149.1) (-) Diaion HP-20 H 20% M 40% M 60% M 70% A DH2O D20M D40M D60M D70 A (164.6) (122.5) (124.0) (132.0) (168.3) Toyopearl HW-40(C) column H 60% M 70% M M:A:H 8 :1: 1 TH2O T60M T70M T712 (146.4) (115.8) (103.1) (177.5) (190.7) M:A:H 7 :1: 2 M:A:H 7 :2: 1 M:A:H 6 :3: 1 70% A T811 T721 T631 T70 (173.9) (144.6) (158.8) ODS column 0.05% TFA : CH3CN = 92 : 8 Punicalagin (1), 556 mg Punicalin (2), 87 mg D, Diaion HP-20 T, Toyopearl HW-40 M, MeOH; A, acetone; H, H2O

TFA, Trifluoroacetic acid

D40M Toyopearl HW-40 (C) column M:A:H 8 :1: 1 M:A:H 7 :2: 1 M:A:H 6 :3: 1 70% M (183.2) 70% A T811 T721 T631 (-) (-) (195.3) 60% M TH2O T60M (174.8) M:A:H 7 :1: 2 H T70M T712 (144.9) (137.2) (189.5) T70 ODS column 0.05% TFA : CH3CN = 85 : 15 ODS column

0.05% TFA : EtOH : EtOAc = 100 : 10 : 5

Granatin B (4), 78 mg Strictinin A (3) 15 mg ODS column 0.05% TFA : CH3CN = 88 : 12 T, Toyopearl HW-40 M, MeOH; A, acetone; H, H2O

TFA, Trifluoroacetic acid

Fig. 1. Isolation ﬂowchart of Punica granatum L. using anti-inﬂammation bioassay-guided fractionation. Numbers in the parentheses were the IC50values (lg/ml) of NO inhibition. All test fractions displayed less than 10% cytotoxicity at 200lg/ml, except D20M, D40M, D20M –TH2O, and D20M – T60M. The IC50values of the test sample exceeded 200lg/ml.

(3)

mediator of vasodilatation; but too much NO might cause hypoten-sion or septicemia (Moncada & Higgs, 2006). Prostaglandins have many physiological activities, such as inducing inﬂammation. Many tissues acutely or chronically generate excess NO and pros-taglandin E2(PGE2) by the overexpression of inducible nitric oxide

synthase (iNOS) and cyclooxygenase-2 (COX-2) in the presence of various inﬂammatory stimulators. RAW 264.7 cells induced by LPS can produce the overexpression of NO and PGE2, and their

reg-ulatory proteins, iNOS and COX-2. Therefore, we used this strategy, combined with chromatography, to isolate the active components, and we used carrageenan-induced paw edema in mice to conﬁrm the in vivo anti-inﬂammatory effects of pomegranate.

2. Materials and methods

2.1. Chemicals

Dimethyl sulphoxide (DMSO), lipopolysaccharide (LPS), 3-(4.5-dimethylthiazol -2-yl) 2.5-diphenyltetrazolium bromide (MTT), indomethacin, N-nitro-L-arginine methyl ester (L-NAME), and the

other chemicals were purchased from Sigma Industries (St. Louis, MO, USA). Dubecco’s modiﬁed Eagle medium (DMEM), fetal bovine serum (FBS), antibiotics, and glutamine were purchased from GIB-CO BRL (Grand Island, NY, USA). Diaion HP-20 gels were bought from Mitsubishi Chemical Industry (Tokyo, Japan). Western blot-ting was performed using an antibody speciﬁc to mouse. iNOS (sc-650), anti-COX-2 (sc-1745), and anti-GAPDH (sc-32233) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA).

2.2. Sample preparation

Test solutions of pomegranate were prepared by dissolving pomegranate in 10% DMSO, which was then stored at 4 °C and used within 1 month. Serial dilutions of test solutions with culture med-ium were prepared before the in vitro assays.

2.3. Determination of total phenols

The total phenol content was determined by the Folin-Ciocalteu method. Pomegranate extract was dissolved in distilled water and mixed with the Folin-Ciocalteu reagent and 7.5% aqueous Na2CO3

solution. After standing for 5 min at 50 °C, the absorbance was measured at 600 nm in an ELISA reader. The amount of total phenol was expressed as gallic acid equivalents (GAE,

l

g gallic acid/mg sample) through the calibration curve of gallic acid. The calibration curve ranged from 7.8 to 250

l

g/ml (R2= 1) (Dudonne´, Vitrac, Cou-tie`re, Woillez, & Me´rillon, 2009).

2.4. Determination of the total ﬂavones content

The total ﬂavanol content was determined by the vanillin assay. Vanillin was dissolved in 80% H2SO4 to prepare the vanillin

re-agent. Pomegranate extract was dissolved in the distilled water and mixed with the vanillin reagent. After standing for 15 min at room temperature, the absorbance was measured at 530 nm in an ELISA reader. The amount of total ﬂavanol was expressed as cat-echin equivalents (CE,

l

g catechin/mg sample) through the calibra-tion curve of catechin. The calibracalibra-tion curve ranged from 7.8 to 250

l

g/ml (R2_{= 1) (}_{He, Liu, & Liu, 2008}_).

2.5. Determination of NO produced from LPS-induced RAW 264.7 cells

RAW 264.7 cells, a murine macrophage cell line, was obtained from American Type Cell Culture (ATCC no. TIB-71; Rockville,

MD, USA). Cells were cultured in DMEM supplemented with 10% FBS, 1% L-glutamine, and 1% penicillin-streptomycin, and

main-tained at 37 °C and 5% CO2. RAW 264.7 cells (4.0 105cells/ml)

were seeded in 96-well plates and then co-treated with LPS (500 ng/ml) and the test samples. After 18 h, the NO production was determined by mixing the culture supernatant with Griess re-agent, and the absorption was detected in an ELISA reader at 530 nm. Anti-inﬂammatory activity was presented in terms of the NO production inhibition percentage. The viability of RAW 264.7 cells was detected by the MTT assay (Tseng, Lee, Chen, Wu, & Wang, 2006).

2.6. Isolation of polyphenols from P. granatum

Dried pomegranate peels (600 g) were purchased from a tradi-tional Chinese medicine store in Taipei, pulverised and ﬁltered through a 20# _{mesh. Pulverised peels were then homogenised}

with 70% aqueous acetone (3 L 4) and the homogenate was then ﬁltered. The ﬁltrate was concentrated by evaporation under reduced pressure (ca. 40 °C) and further freeze-dried to yield the 70% acetone extract (330 g, yield 55%). The 70% acetone extract was dissolved in water and then partitioned with n-hexane to re-move the non-polar compounds. The aqueous layer was chro-matographed over a Diaion HP-20 gel column (9.5 40 cm) with stepwise aqueous MeOH (H2O ? 20% MeOH ? 40%

MeOH ? 60% MeOH ? 70% acetone). Column chromatography was combined with an NO inhibition assay to clarify the bioac-tive fractions. We found that the 20% MeOH (D20M) and 40% MeOH (D40M) fractions of the Diaion column exhibited greater NO inhibition. Furthermore, the D20M fraction was chromato-graphed on a Toyopearl HW-40 (C) (Tosoh Bioscience, Montgom-eryville, PA, USA) gel column (2.5 40 cm) and eluted stepwise with H2O ? 70% MeOH ? MeOH–acetone–H2O (8:1:1, v/v/

v) ? MeOH–acetone–H2O (7:2:1, v/v/v) ? MeOH–acetone–H2O

(6:3:1, v/v/v) ? 70% acetone. The 70% MeOH (D20M –T70M) elu-ate was applied to an ODS column (0.05% triﬂuoroacetic acid:CH3CN, 92:8) and HLBÒ extraction cartridges (Waters,

Mil-ford, MA, USA) to yield punicalagin (556 mg, 1) and punicalin (87 mg, 2). The D40M fraction was chromatographed on a Toyo-pearl HW-40 (C) gel column (2.5 40 cm) and eluted stepwise with H2O ? 70% MeOH ? MeOH–acetone–H2O (8:1:1, v/v/

v) ? MeOH–acetone–H2O (7:2:1, v/v/v) ? MeOH–acetone–H2O

(6:3:1, v/v/v) ? 70% acetone. The 60% MeOH eluate (D40M – T60M) was puriﬁed on an ODS column with 0.05% triﬂuoroacetic acid–CH3CN (88:12) to obtained 15 mg of strictinin A (3). The

Table 1

NO inhibitory effects of four hydrolysable tannins. Compound LPS-induced RAW 264.7

macrophage cells Free radical-scavenging activity NO inhibitiona iNOS activityb,c ONOO scavengingc IC50values (mM) NO inhibition (%) Cytotoxicity (%) SNP clearance (%) 1 69.8 15.7 ± 0.4 35.5 ± 0.7 24.7 ± 6.4 2 78.6 46.7 ± 7.0 67.6 ± 1.4 30.7 ± 1.0 3 63.1 25.4 ± 2.4 45.8 ± 1.1 32.3 ± 0.0 4 33.6 14.8 ± 1.0 41.7 ± 0.7 29.5 ± 0.7 L-NAMEd – 46.4 ± 2.9 7.5 ± 1.6 –

Results are expressed as the mean ± SD of three experiments. a

Co-treatment of LPS and each test compound for 18 h and no cytotoxicites were found in these IC50values.

b

RAW 264.7 cells were stimulated with 1lg/ml LPS and incubated overnight. c _{Each test compound was 100}_l_M.

d

(4)

70% MeOH eluate (D40M – T70M) was puriﬁed on an ODS col-umn with 0.05% triﬂuoroacetic acid–CH3CN (85:15) and 0.05%

trifluoroacetic acid–EtOH–EtOAc (100:10:5, v/v/v) to obtain 78 mg of granatin B (4) (Fig. 1). The structures of these four com-pounds were identified by nuclear magnetic resonance and the purity of each compound (>99.0%) was confirmed by high-perfor-mance liquid chromatography. The formulae and molecular weights of punicalagin (1), punicalin (2), strictinin A (3), and granatin B (4) are C43H28O30 (MW 1048.7), C34H22O22 (MW

782.53), C43H28O30 (MW 634.46), and C41H28O27 (MW 952.66),

respectively.

2.7. Determination of NO radical-scavenging activity

Sodium nitroprusside (SNP) is a stable NO donor. We used SNP as an NO donor to evaluate the direct NO radical clearance of sam-ples. The SNP solution was prepared with H2O. Test solution was

added to the sample volume of SNP solution (50 mM). The mixture was incubated at 37 °C, for 5 h. Then, we took the supernatant (1, mixed it with the Griess reagent, and read the absorption on an ELISA reader at 530 nm (Bor, Chen, & Yen, 2006).

2.8. iNOS activity assay in LPS-induced RAW 264.7 cells

Initially, RAW 264.7 cells were pre-stimulated by 1

l

g/ml LPS for 24 h to activate the iNOS. Then, cells were collected and washed twice with PBS to remove the excessive LPS, and the activated RAW 264.7 cells (4.0 105_{cell/ml) were seeded in}

96-well plates. Crude extract and active components were imme-diately added to the cells. The 96-well plate was incubated at 37 °C overnight. The NO detecting method was the same as

pre-viously described. We used N-nitrol-arginine methyl ester (L

-NAME, 400

l

M), a speciﬁc inhibitor of NO synthase enzyme, and as the positive control.

2.9. iNOS and COX-2 expression assay in LPS-induced RAW 264.7 cells

Total cellular protein was extracted with a RIPA solution (radioimmuno-precipitation assay buffer) at 20 °C overnight. We used BSA (bovine serum albumin) as a protein standard to calculate equal total cellular protein amounts. Protein samples (15

l

g) were resolved by denaturing sodium dodecyl sulfate– polyacrylamide gel electrophoresis (SDS–PAGE) using standard methods, and then were transferred to nitrocellulose (Hybond-PVDF) membranes by electroblotting and blocking with 1% BSA. The membranes were probed with the primary antibodies (iNOS, COX-2 and GAPDH) at 4 °C overnight, washed three times with PBST, and incubated for 1 h at 37 °C with alkaline phosphatase-conjugated secondary antibodies. Then, we used an NBT/BCIP commercial kit (Gibco) as the visualising agent. The intensity of the bands was quantiﬁed by computer-assisted image analysis using AlphaInnotech Digital Imaging Systems. Western blotting results are representative of three independent experiments for every data point.

2.10. Model of 1% carrageenan-induced paw edema in ICR mice

Male ICR mice weighing 25 ± 2 g were bought from the National Science Council, Taipei, Taiwan, and maintained at 21 ± 2 °C with food and water ad libitum. They were kept on a 12-h light/12-h dark cycle. All mice used in this experiment were cared for according to

OH HO HO HO OH HO C C O O O O O O O O C O OH OH OH O OH OH C C O O HO HO H OH O

4

O O O O OH O HO HO OH HO OH OH C C O O O O O O OH OH HO HO OH HO HO HO HO OH C O C O

1

O OH HO O OH O O O O O OH OH HO HO OH HO HO HO HO OH C O C O

2

OH HO HO HO OH HO C C O O O OH HO O O O C O OH OH OH

3

(5)

the Ethical Regulations on Animal Research of our university. Edema in the left hind paw of the mice was induced by injecting 50

l

l of 1% (w/v) carrageenan into the subplantar region. The perimeter of the paw was measured 1 h before the carrageenan injection and after 1–6 h, using calipers. One hour before the injection, the pomegran-ate (100 mg/kg), the active component (2.5 and 10 mg/kg), and indomethacin (10 mg/kg, as a positive control) were given orally, while the control group was given distilled water. The blank group was injected with normal saline and given distilled water. Each group consisted of ﬁve animals. After 6 h, mice were sacriﬁced and the serum collected for the PGE2measurement (Tseng et al., 2006).

2.11. Measurement of PGE2production

Serum and cell culture mediums were determined the PGE2

concentrations by the PGE2ELISA kit (Amersham Pharmacia

Bio-tech, Buckinghamshire, UK).

2.12. Statistical analysis

The data are presented as mean and standard deviation (SD). Signiﬁcance was calculated using the Student’s t-test. Differences were considered signiﬁcant for p < 0.05.

3. Results

3.1. Polyphenols isolation from P. granatum L.

The phytochemical components of the 70% acetone extract of pomegranate peels were screened using the Folin-Ciocalteu meth-od and vanillin assay. The screening found that pomegranate was rich in total phenol and ﬂavanol (471.0 ± 32.0

l

g gallic acid equiv-alent/mg in total phenol and 257.0 ± 19.6

l

g catechin equivalent/ mg in total ﬂavanol).

The NO inhibition bioassay-guided fractionation ﬂowchart of pomegranate is shown inFig. 1. These four hydrolysable tannins, punicalagin (1), punicalin (2), strictinin A (3), and granatin B (4), were isolated from D20M and D40M of the pomegranate fractions, with stronger inhibition of NO production; the yields were 0.093%, 0.015%, 0.003% and 0.01%, respectively.

3.2. NO inhibitory effects of four hydrolysable tannins

The NO inhibitory effects of 1–4 were measured in LPS-induced RAW 264.7 cells for 18 h. We collected the culture medium to de-tect the NO levels using the Griess reaction. Compounds 1–4 showed less than 10% cytotoxicities in LPS-induced RAW 264.7 cells for 18 h. Compound 4 displayed a more potent NO inhibition

(A) (B) GAPDH COX-2 iNOS 2 (µM) Fold 1.00 9.83 9.71 9.06 10.92 11.70 Fold 1.00 11.97 11.21 8.19 7.30 5.11 B C 12.5 25 50 100 LPS (8h) 1 (µM) iNOS COX-2 GAPDH Fold 1.00 7.72 7.14 7.29 7.64 8.17 Fold 1.00 12.27 13.01 11.45 9.37 7.22 B C 12.5 25 50 100 LPS (18 h) GAPDH COX-2 iNOS 1 (µM) Fold 1.00 10.50 8.74 8.71 6.05 1.94 Fold 1.00 4.47 4.66 4.60 4.96 5.85 B C 12.5 25 50 100 LPS (8h) GAPDH COX-2 iNOS 2 (µM) Fold 1.00 9.80 9.32 9.12 8.94 9.23 Fold 1.00 6.19 4.80 4.60 4.90 4.53 B C 12.5 25 50 100 LPS (18 h) (C) (D) 3 (µM) iNOS COX-2 GAPDH Fold 1.00 15.18 15.23 15.13 14.20 15.00 Fold 1.00 12.30 12.50 12.75 11.48 11.66 B C 12.5 25 50 100 LPS (8h) iNOS COX-2 GAPDH 3 (µM) Fold 1.00 9.68 10.79 10.38 9.66 9.02 Fold 1.00 6.02 5.06 5.06 6.43 2.09 B C 12.5 25 50 100 LPS (18h) 4 (µM) GAPDH COX-2 iNOS Fold 1.00 11.10 9.54 9.86 5.27 2.14 Fold 1.00 15.40 13.09 13.68 3.78 0.59 B C 12.5 25 50 100 LPS (8h) COX-2 iNOS GAPDH 4 (µM) Fold 1.00 12.92 13.46 12.52 13.90 12.93 Fold 1.00 7.77 6.33 6.68 8.44 5.45 B C 12.5 25 50 100 LPS (18h)

Fig. 3. Inducible NOS and COX-2 expression of punicalagin (A), punicalin (B), strictinin A (C) and granatin B (D) induced by LPS (500 ng/ml) in RAW 264.7 cells for 8 and 18 h. (B) Non-LPS-induced group. (C) Solvent control group (H2O). GAPDH was used as an internal control to identify equal amounts of protein loading in each lane. The intensity of the iNOS and COX-2 protein was examined by densitometrus analysis, and expressed as the ratio of iNOS/GAPDH and COX-2/GAPDH. Data from three separate experiments were used, and the picture for one of which is shown.

(6)

in LPS-induced RAW 264.7 cells for 18 h than did the others, and its IC50value was 33.6

l

M (Table 1).

Nevertheless, we would like to clarify whether these hydrolysa-ble tannins had NO radical-scavenging activity or not. SNP was used as an NO radical donor to evaluate the NO-scavenging activity of these four hydrolysable tannins. As shown inTable 1, the four hydrolysable tannins did not signiﬁcantly exhibit SNP clearance at 100

l

M.

3.3. iNOS expression and activities of four hydrolysable tannins in LPS-induced RAW 264.7 cells

Based on the above data, we suggest that the NO inhibition ef-fects of the four hydrolysable tannins could be brought about through the inﬂuence of iNOS protein expression or activity. The

iNOS protein expression in LPS-induced RAW 264.7 cells for 8 and 18 h were observed by Western blot assay. As shown in

Fig. 3, the four hydrolysable tannins displayed dose-dependently inhibitory effects on the iNOS expression at 12.5–100

l

M. In addi-tion, the 8 h group expressed more signiﬁcantly iNOS inhibitory ef-fects than the 18 h group. Therefore, these four hydrolysable tannins could decrease the iNOS protein expressions from the early stage (8 h) to the late stage (18 h). Taken together, 1–4 signiﬁcantly decreased the iNOS expressions in LPS-induced RAW 264.7 cells, and 4 was the strongest among them.

On the other hand, RAW 264.7 cells were pretreated with LPS for 24 h to observe the iNOS activity inhibition of these four hydro-lysable tannins. As seen inTable 1,L-NAME acted as a positive con-trol and displayed signiﬁcant iNOS activity inhibition at 400

l

M. However, these four hydrolysable tannins had no iNOS activity inhibitory effects and even exhibited cytotoxicity at 100

l

M. Hence, these four hydrolysable tannins inhibited NO production through directly decreasing the iNOS protein expressions.

3.4. COX2and PGE2inhibitory effects of four hydrolysable tannins

The COX-2 and PGE2inhibitory effects of these four

hydrolysa-ble tannins were evaluated in LPS-induced RAW 264.7 cells for 8 and 18 h. As shown inFig. 3, Compound 4 inhibited COX-2 protein expression in dose-dependent (Fig. 3) and PGE2 productions

(Fig. 4A) more signiﬁcantly than the others after treatment with LPS for 8 h. Moreover, 4 inhibited PGE2 productions in a

dose-dependent manner, and the IC50 value was 66.22 ± 9.4

l

M

(Fig. 4D). However, COX-2 protein expression of LPS-induced RAW 264.7 cells was not inhibited after treatment with the four hydrolysable tannins for 18 h (Fig. 3). Hence, we suggested that 4 could inhibit COX-2 and PGE2production in the early stage (8 h)

because there was less COX-2 expression at that stage, but in the late stage (18 h), COX-2 expression reached a stable state, so 4 had difﬁculty inhibiting its expression.

3.5. Inhibitory effect of pomegranate and 4 on carrageenan-induced paw edema in ICR mice

Carrageenan-induced paw edema in ICR mice was used to eval-uate the in vivo anti-inflammatory model. We used indomethacin, a common non-steroidal anti-inflammatory drug (NSAID) as a po-sitive control. Mice were treated with pomegranate (100 mg/kg) or indomethacin (10 mg/kg) 1 h before carrageenan induction, and had detected paw edema for 6 h. Results showed that carrageenan significantly induced paw edema during the entire experiment in the control group (Table 2). Pomegranate significantly reduced paw edema by more than 50% at 1–5 h after the carrageenan injec-tion, and had even greater potential paw edema inhibitory effects than indomethacin. (A) (B) 0 200 400 600 800 1000 1200 1400 1600 1800 2000 PGE 2 Pr oduction (pg/ml) 4 LPS (500 ng/ml) _- ₊ 1 2 3 + + + + Compound (100 µM) _- -** 0 12.5 25 50 100 0 20 40 60 80 100 PGE 2 inhibition (%) 4 (µM)

Fig. 4. Prostaglandin E2production of 1–4 (A) and a series dose–response curve of 4 (B) induced by LPS-induced RAW 264.7 cells for 8 h. **p < 0.001. n = 3 All data were expressed as the mean ± SD.

Table 2

Effects of pomegranate and granatin B on the paw perimeter of carrageenan-induced mice paw edema. Group Time after injection of carrageenan (h)

1 2 3 4 5 6 Control 1.74 ± 0.30 1.41 ± 0.17 1.41 ± 0.23 1.21 ± 0.13 0.92 ± 0.16 0.88 ± 0.08 Indomethacin (10 mg/kg) 0.70 ± 0.35* _{0.90 ± 0.34}* _{0.79 ± 0.33}* _{0.72 ± 0.15}* _{0.75 ± 0.09}* _{0.62 ± 0.19}* Pomegranate (100 mg/kg) 0.86 ± 0.10* _{0.60 ± 0.07}** _{0.59 ± 0.28}* _{0.42 ± 0.30}* _{0.4 ± 0.36}* _{0.34 ± 0.50} Granatin B (4) 2.5 mg/kg 1.08 ± 0.13* _{0.84 ± 0.20}* _{0.77 ± 0.07}** _{0.75 ± 0.17}* _{0.63 ± 0.03}* _{0.59 ± 0.15}* 10.0 mg/kg 0.98 ± 0.40* _{0.89 ± 0.19}* _{0.68 ± 0.29}* _{0.50 ± 0.41}* _{0.23 ± 0.52}* _{0.48 ± 0.26}*

Values were presented as the mean ± SD of the increasing mice paw edema (mm) of four animals for each group. *_{p < 0.05.}

(7)

Since 4 showed the strongest COX-2 and iNOS inhibitory effects in the in vitro study, we examined the in vivo anti-inflammatory ef-fect of 4. We found that 4 (2.5 and 10 mg/kg) could significantly and dose-dependently reduce paw edema, and the inhibitory abil-ities of the higher dose (10 mg/kg) were the same as that of the indomethacin (Table 2). In addition, 4 had more significant PGE2

inhibitory effects than the indomethacin at 6 h (Fig. 5).

4. Discussion

Pomegranate has been used for centuries as a therapeutic agent for the treatment of inflammatory diseases. According to current reports, pomegranate is a polyphenol-rich fruit, and showed poten-tial as an anti-inflammatory, anti-oxidative and anti-cancer agent in several experimental models (Shukla, Gupta, Rasheed, Khan, & Haqqi, 2008). Chemical analyses have also shown that the phenol compounds of pomegranate contain significantly high levels of hydrolysable tannins, such as punicalin, punicalagin, peduncula-gin, punigluconin (Dudonne´ et al., 2009). Therefore, we used an NO inhibition assay combined with column chromatography to determine which components in pomegranate would have effec-tive anti-inflammatory activity. In this paper, four hydrolysable tannins, punicalagin (1), punicalin (2), strictinin A (3), and granatin B (4), were isolated from pomegranate by bioassay-guided frac-tionation. Each of them displayed a dose-dependently and signifi-cantly inhibitory effect on NO production in LPS-induced RAW 264.7 cells (Table 1). Furthermore, granatin B (4) more strongly inhibited PGE2production and COX-2 expression in LPS-induced

RAW 264.7 cells than the others.

Structure–activity relationships (SAR) of natural products have been found to influence the various pharmacological functions, such as antioxidant and anti-inflammation activities. Up to the present, over 3000 kinds of tannins have been identified, chiefly as secondary metabolites in green plants. Tannin is involved in a large proportion of phenolic derivatives in plants and is divided into two types: hydrolysable and condensed tannins. Many possi-ble permutations are offered by substitution and conjugation, and this could explain why so many tannin derivatives occur naturally (Fylaktakidou, Hadjipavlou-Litina, Litinas, & Nicolaides, 2004). Based on the structural similarities of these four hydrolysable tan-nins, we divided them into two groups. One was punicalagin (1) and punicalin (2) and the other was strictinin A (3) and granatin B (4) (Fig. 2). Among them, 1 and 2 were previously recognised as inhibitors of the pro-inflammatory for anti-edematogenic

activ-ity on carrageenan-induced paw edema in rats (10 mg/kg) (Lin, Hsu, & Lin, 1999). However, there is no scientific evidence that 3 and 4 have anti-inflammatory activities. In this study, we first found that both 3 and 4 had potential NO inhibitory effects in LPS-induced RAW264.7, and that 4 was the better one (Table 1). Moreover, 4 showed the strongest COX-2 and iNOS inhibitory ef-fects in an in vitro assay, and significantly reduced paw edema and PGE2inhibitory effects in an in vivo assay.

Dehydrohexahydroxydiphenoyl (DHHDP), a substitution in C-2 and C-3 of 4, was the difference between 4 and 3, and probably played an important role in anti-inflammation. In related investi-gations, tannins with DHHDP units invariably had strong anti-inflammatory and anti-oxidant activity. (Feldman, 2005) Geraniin, a well-documented hydrolysable tannin, has been reported to have excellent NO radical-scavenging and iNOS inhibitory activities (Kumaran & Karunakaran, 2006). Mallotusinic acid and euphorbin E also had potent scavenging effects on DPPH free radicals. (Okuda, 2005) Hence, we concluded once the hydrolysable tannins had the DHHDP group, they would appear to have better anti-inflamma-tory activities.

Many studies have demonstrated that the massive production of NO and PGE2via the pro-inﬂammatory proteins iNOS and COX-2

played an important physiological role in inflammation. Evidence has shown that NO production was initiated by treating LPS for 8 h and reached a stable state for 18 h, and that COX-2 expression began at 6–8 h and reached the maximal level at 16–24 h. (Caughey, Cleland, Penglis, Gamble, & James, 2001; Reher, Harris, Whiteman, Haiand, & Meghji, 2002). Therefore, we used two treatment times (8 and 18 h) to measure the anti-inflammatory effects of these hydrolysable tannins, because of the different mechanism and pro-duction time of these inflammation mediators (COX-2, iNOS, etc.). On the basis of the above results, we found that 4 could significantly decrease NO and PGE2 production through inhibiting iNOS and

COX-2 expression. Little NO production was found in LPS-induced RAW 264.7 for 8 h. NO was released from the process of converting

L-arginine toL-citruline by iNOS, while NO was initially produced at

8 h and reached the maximum level at 18 h; therefore, NO produc-tion from LPS-induced RAW 264.7 was not detected at 8 h. Unlike NO production, activated COX-2 could convert arachidonic acid to PGE2in only 30 min. Hence, the COX-2 and PGE2inhibitory effects

of 4 from LPS-induced RAW 264.7 at 8 h were stronger than those at 18 h. In summary, we suggest that 4 is an effective anti-inﬂamma-tory compound and has dual roles in anti-inﬂammation, by decreas-ing PGE2 production in the early stage and decreasing NO

production in the late stage.

In conclusion, 4 not only displayed the best NO inhibitory abil-ities in LPS-induced RAW 264.7, but also had the strongest PGE2

inhibitory effects in the in vitro and in vivo assays. Taken together, 4 could be used as a standard marker compound to determine the potential anti-inﬂammatory effect of pomegranate.

References

Alper, N., & Acar, J. (2004). Removal of phenolic compounds in pomegranate juices using ultraﬁltration and laccase-ultraﬁltration combinations. Die Nahrung, 48(3), 184–187.

Aviram, M., Dornfeld, L., Kaplan, M., Coleman, R., Gaitini, D., Nitecki, S., et al. (2002). Pomegranate juice ﬂavonoids inhibit low-density lipoprotein oxidation and cardiovascular diseases: Studies in atherosclerotic mice and in humans. Drugs under Experimental and Clinical Research, 28(2–3), 49–62.

Aviram, M., Rosenblat, M., Gaitini, D., Nitecki, S., Hoffman, A., Dornfeld, L., et al. (2004). Pomegranate juice consumption for 3 years by patients with carotid artery stenosis reduces common carotid intima-media thickness, blood pressure and LDL oxidation. Clinical Nutrition, 23(3), 423–433.

Bor, J. Y., Chen, H. Y., & Yen, G. C. (2006). Evaluation of antioxidant activity and inhibitory effect on nitric oxide production of some common vegetables. Journal of Agricultural and Food Chemistry, 54(5), 1680–1686.

Caughey, G. E., Cleland, L. G., Penglis, P. S., Gamble, J. R., & James, M. J. (2001). Roles of cyclooxygenase (COX)-1 and COX-2 in prostanoid production by human

0 10 20 30 40 50 60 PGE 2 inhibition (%) Control Indo 2.5 10 4 (µM) a b

Fig. 5. Prostaglandin E2concentrations of 4 on carrageenan-induced mice paw edema for 6 h. Indomethacin served as a positive control and 10 mg/kg dosage used. a

p < 0.01. The p value was calculated by comparison with the control group. b_{p < 0.05. The p value was calculated by comparison with the indomethacin group} (n = 3). All data were expressed as mean ± SD.

(8)

endothelial cells: Selective up-regulation of prostacyclin synthesis by COX-2. Journal of Immunology, 167(5), 2831–2838.

Dudonne´, S., Vitrac, X., Coutie`re, P., Woillez, M., & Me´rillon, J. M. (2009). Comparative study of antioxidant properties and total phenolic content of 30 plant extracts of industrial interest using DPPH, ABTS, FRAP, SOD, and ORAC Assays. Journal of Agricultural and Food Chemistry.

Feldman, K. S. (2005). Recent progress in ellagitannin chemistry. Phytochemistry, 66, 1984–2000.

Fylaktakidou, K. C., Hadjipavlou-Litina, D. J., Litinas, K. E., & Nicolaides, D. N. (2004). Natural and synthetic coumarin derivatives with anti-inﬂammatory/ antioxidant activities. Current Pharmaceutical Design, 10(30), 3813–3833. He, X., Liu, D., & Liu, R. H. (2008). Sodium borohydride/chloranil-based assay for

quantifying total ﬂavonoids. Journal of Agricultural and Food Chemistry, 56(20), 9337–9344.

Joseph, J. A., Shukitt-Hale, B., & Lau, F. C. (2007). Fruit polyphenols and their effects on neuronal signaling and behavior in senescence. Annals of the New York Academy of Sciences, 1100, 470–485.

Kaur, G., Jabbar, Z., Athar, M., & Alam, M. S. (2006). Punica granatum (pomegranate) ﬂower extract possesses potent antioxidant activity and abrogates Fe-NTA induced hepatotoxicity in mice. Food and Chemical Toxicology, 44(7), 984–993. Khan, N., Afaq, F., Kweon, M. H., Kim, K., & Mukhtar, H. (2007). Oral consumption of

pomegranate fruit extract inhibits growth and progression of primary lung tumors in mice. Cancer Research, 67(7), 3475–3482.

Kumaran, A., & Karunakaran, R. J. (2006). Nitric oxide radical scavenging active components from Phyllanthus emblica L.. Plant Foods for Human Nutrition, 61(1), 1–5.

Lansky, E. P., & Newman, R. A. (2007). Punica granatum (pomegranate) and its potential for prevention and treatment of inﬂammation and cancer. Journal of Ethnopharmacology, 109(2), 177–206.

Lin, C. C., Hsu, Y. F., & Lin, T. C. (1999). Effects of punicalagin and punicalin on carrageenan-induced inﬂammation in rats. The American Journal of Chinese Medicine, 27(3–4), 371–376.

Menezes, S. M., Cordeiro, L. N., & Viana, G. S. (2006). Punica granatum (pomegranate) extract is active against dental plaque. Journal of Herbal Pharmacotherapy, 6(2), 79–92.

Moncada, S., & Higgs, E. A. (2006). The discovery of nitric oxide and its role in vascular biology. British Journal of Pharmacology, 147(Suppl. 1), S193–S201. Mudzhiri, M. S. (1954). Punica granatum bark extract as a therapeutic agent in

tapeworm infection. Meditsinskaia Parazitologiia i Parazitarnye Bolezni, 4, 311–313.

Noda, Y., Kaneyuki, T., Mori, A., & Packer, L. (2002). Antioxidant activities of pomegranate fruit extract and its anthocyanidins: Delphinidin, cyanidin, and pelargonidin. Journal of Agricultural and Food Chemistry, 50(1), 166–171. Okuda, T. (2005). Systematics and health effects of chemically distinct tannins in

medicinal plants. Phytochemistry, 66, 2012–2031.

Raj, R. K. (1975). Screening of indigenous plants for anthelmintic action against human Ascaris lumbricoides: Part-II. Indian Journal of Physiology and Pharmacology, 19(1), 47–49.

Reddy, M. K., Gupta, S. K., Jacob, M. R., Khan, S. I., & Ferreira, D. (2007). Antioxidant, antimalarial and antimicrobial activities of tannin-rich fractions, ellagitannins and phenolic acids from Punica granatum L. Planta Medica, 73(5), 461–467. Reher, P., Harris, M., Whiteman, M., Haiand, H. K., & Meghji, S. (2002). Ultrasound

stimulates nitric oxide and prostaglandin e2 production by human osteoblasts. Bone, 31(1), 236–241.

Satomi, H., Umemura, K., Ueno, A., Hatano, T., Okuda, T., & Noro, T. (1993). Carbonic anhydrase inhibitors from the pericarps of Punica granatum L.. Biological and Pharmaceutical Bulletin, 16(8), 787–790.

Shukla, M., Gupta, K., Rasheed, Z., Khan, K. A., & Haqqi, T. M. (2008). Consumption of hydrolyzable tannins-rich pomegranate extract suppresses inﬂammation and joint damage in rheumatoid arthritis. Nutrition, 24(7–8), 733–743.

Tseng, S. H., Lee, H. H., Chen, L. G., Wu, C. H., & Wang, C. C. (2006). Effects of three purgative decoctions on inﬂammatory mediators. Journal of Ethnopharmacology, 105(1–2), 118–124.