Gambogic Acid and Epigambogic Acid, C-2 Epimers with Novel Anticancer Effects from Garcinia hanburyi

• Abstract
• Materials and Methods
• Supporting Information
• Acknowledgements
• References

Abstract

Gambogic acid, usually isolated as an inseparable stereomeric mixture of C-2 epimers, was newly separated into two epimers (1 and 2) from the gamboges of Garcinia hanburyi. The stereochemistry at C-2 was clearly defined by extensive spectroscopic analysis and direct comparison of NMR and HPLC data with those of the known R-epimer. Both epimers were examined for their cytotoxicities against human leukemia K562 (K562/S) and doxorubicin-resistant K562 (K562/R) cell lines. Different from doxorubicin (IC₅₀ = 10.78 μM for K562/R and 0.66 μM for K562/S), epimers 1 and 2 exhibited similar activities against both cell lines (IC₅₀ = 1.32 and 0.89 μM for 1, IC₅₀ = 1.11 and 0.86 μM for 2). These results suggested that both epimers were not multidrug resistance (MDR) substrates. Furthermore, epimers 1 and 2 were tested for their inhibitory effects against six human cytochrome P-450 enzymes. Epimers 1 and 2 showed little inhibitory effects toward five of the enzymes except CYP2C9. Interestingly, when tested against CYP2C9, S-epimer 2 had an inhibitory effect 20-fold stronger than that of R-epimer 1.

Gambogic acid (GA, CAS No. 2752 - 65 - 0) is the principal active component of gamboge, the resin from various Garcinia species including G. morella and G. hanburyi [1], [2]. Many modern pharmaceutical studies are focused on its extensive and potent anti-tumor activities [3], [4], [5], [6], [7], [8], [9]. It had been developed as an antitumor drug for clinical testing via intravenous (i. v.) injection in China in the 1970 s. The structure of GA was initially deduced by detailed NMR spectroscopic analyses and chemical synthesis, with the stereochemistry of C-2 undetermined [1], [2]. GA was believed to be an inseparable C-2 epimeric mixture. Continuous efforts were taken to separate and determine these two epimers. The R-configuration had been deduced for C-2 of GA through a series of degradation reactions [10]. A single crystal of R-epimer was obtained from recrystallization of the pyridine salt of GA, and its stereochemistry was then determined by X-ray crystallographic analyses [11]. However, the S-epimer was still unknown. In the current study, GA was isolated from G. hanburyi and further separated into two isomers (1 and 2, in a ratio of 4 : 3) using preparative HPLC. These two isomers were elucidated to be the R-epimer (1) and the S-epimer (2) (Fig. [1]), respectively, by spectral and chromatographic analyses and comparison with those of the known R-epimer crystal. Both epimers were tested for their cytotoxicities against human leukemia K562 (K562/S) and doxorubicin-resistant K562 (K562/R) cell lines and for their inhibitory effects against six human cytochrome P-450 enzymes. This paper presents the isolation, structure elucidation, and bioassay results.

GA, obtained from gamboge by preparative HPLC, presented as one peak on a C₁₈ column (Alltima C₁₈, 5 μ, 4.6 × 250 mm) but as two completely separated peaks on a C₈ column (Alltima C₈, 5 μ, 4.6 × 250 mm). Both peaks showed the same molecular ion peak at m/z = 628 in the HPLC/ESI-MS analysis. They were accordingly isolated to yield compounds 1 and 2. Detailed analysis of their COSY, HMQC, and HMBC spectra indicated that both compounds possessed a basic chemical structure the same as that reported before for GA [1], [2]. NOEs displayed in the ROESY spectra indicated the relative configuration of S, R, S, and R for C-11, C-13, C-14, and C-22 of 1 and 2. The cis-configuration was determined for the olefinic bond between C-27 and C-28 of both 1 and 2, according to the clear NOE between H-27 and H-30. The remaining unknown stereochemistry was that of C-2. Based on the above evidence, compounds 1 and 2 were deduced to be a pair of C-2 epimers. In the previous report, the single crystal of the R-epimer had been obtained from recrystallization of the pyridine salt of GA, and its stereochemistry was then determined by X-ray crystallographic analyses [11]. In the comparison of the ¹H- and ¹³C-NMR data of both epimers with those previously published for the R-epimer, compound 1 was indicated to be the R-epimer. This deduction was confirmed by further chromatographic comparison using HPLC, in which compound 1, the known R-epimer (a crystal sample provided by Dr. Sui Xiong Cai who reported the crystal structure of the pyridine salt of the R-epimer), and their mixture all presented as the same one peak on a C₈ column. Therefore, compound 1 was determined to be the R-epimer, and compound 2 was then elucidated as the S-epimer, named epigambogic acid. In the ¹H-NMR spectra of 1, the known R-epimer [12], and its pyridine salt crystal, all the signals of H-37 and H-32 were completely overlapped near at δ = 5.02 (2H). Different from 1, the proton signals of H-37 and H-32 of 2 were two clear triplets at δ = 5.07 (1H, t, J = 7.0 Hz) and δ = 5.00 (1H, t, J = 7.0 Hz), respectively. In the ¹³C-NMR spectra, the most outstanding difference was that the signal for C-19 was shifted upfield from δ = 27.69 in 1 to δ = 26.96 in 2. These differences might be regarded as the key NMR spectroscopic characteristics of these two epimers.

Both epimers were examined for their cytotoxicities against human leukemia K562 (K562/S) and doxorubicin-resistant K562 (K562/R) cell lines. The K562/R subline expresses a high level of multidrug resistance (MDR) transporter on the membrane surface, whereas the parent line (K562/S) does not. The MDR transporter transports actively a wide variety of anticancer agents out of cells and reduces the intracellular drug accumulation to protect the cell. Therefore, when an anticancer agent showed a much higher IC₅₀ to K562/R than to K562/S, this agent would be a substrate of MDR transporter. Doxorubicin (IC₅₀ = 10.78 μM for K562/R and 0.66 μM for K562/S), whose resistant fold (also called the ratio of doxorubicin IC₅₀) between K562/R and K562/S was 16.3, should be a substrate of MDR transporter. It was interesting that unlike doxorubicin, epimers 1 and 2 exhibited similar activities against both cell lines (IC₅₀ = 1.32 μM and 0.89 μM for 1, IC₅₀ = 1.11 and 0.86 μM for 2). Their resistant folds were 1.48 and 1.29, respectively. The results showed that both 1 and 2 had a strong cytotoxicity, similar to that of doxorubicin, but more importantly, they proved to be non-substrates of MDR transporter.

MDR in cancer cells is a significant factor for the failure of chemotherapy in many patients. It is very important to find and develop new anticancer drugs that can overcome MDR of cancer cells, since MDR transporters contributed significantly to the pharmacokinetic disposition of anticancer drugs. Knowledge of substrates, inducers and inhibitors of these transporters is necessary to ensure optimal clinical outcomes [13]. In addition, chemotherapy often requires multidrug combinations, while most anti-cancer drugs are metabolic substrates of cytochrome P450 s. Therefore, drug-drug interactions are important for the combination use of anti-cancer drugs. The likelihood of drug interactions with combination therapy will be very high if these combined drugs are substrates and potent inhibitors or inducers of the cytochrome P450 (CYP) system [14]. We therefore further tested their inhibitory effects toward six major human cytochrome P-450 enzymes: CYP1A2 (phenacetin O-deethylation), CYP2A6 (coumarin 7-hydroxylation), CYP2C9 (diclofenac 4′-hydroxylation), CYP2D6 (dextromethorphan O-demethylation), CYP2E1 (chlorzoxazone 6-hydroxylation) and CYP3A4 (testosterone 6β-hydroxylation). As the results in Table [1] demonstrate, the R-empimer 1 exhibited intermediate inhibition against CYP2C9 in HLM with an IC₅₀ of 9.3 μM, and weak inhibition against CYP1A2, CYP2A6, CYP2D6, and CYP3A4. The S-epimer 2 showed strong inhibition against CYP2C9 with an IC₅₀ of 0.4 μM, and weak inhibition against CYP1A2, CYP2D6, and CYP3A4. Both compounds had no inhibition against CYP2E1. The S-epimer 2 was inactive against CYP2A6, but was much stronger against CYP2C9 than the R-epimer 1. CYP2C9 is one of the most important metabolic enzymes, which is involved in the metabolism of numerous drugs including the anticoagulant drug warfarin and a number of non-steroidal anti-inflammatory drugs [15]. The influence on this enzyme is particularly important if the drug has a narrow therapeutic index, such as warfarin. Compared with the positive control, the results indicated that the S-epimer 2 has the potential for inhibition of CYP2C9, and possibly has drug-drug interactions with the other drugs known to be metabolized by CYP2C9. This considerable difference suggested that some of the biological activities of GA reported previously might be the combined effect of these two epimers which could contribute differentially to the biological activities of GA. Both epimers are certainly worthy of further and separate biological studies.

Materials and Methods

Optical rotations were measured with a Jasco P-1010 Polarimeter. ¹H- (400 MHz) and ¹³C- (100 MHz) NMR spectra were recorded on a Brucker DRX-400 spectrometer using TMS as an internal standard. The LC/MS analysis was performed using an Agilent 1100 series combined with MICROMASS Q-TOF-2 spectrometer.

The resin (0.1 g) of Garcinia hanburyi was purchased from National Institute for the Control of Pharmaceutical and Biological Products (NICPBP), P. R. China. A voucher specimen (CMS-0283) is deposited in the Herbarium of the Hong Kong Jockey Club Institute of Chinese Medicine, Hong Kong, China.

The resin (90 mg) was dissolved in 2 mL acetone, and loaded on the preparative HPLC system (Agilent 1100, Alltima C₁₈, 10 μ, 22 × 250 mm) to give GA. The mobile phase was MeOH/0.1 %H₃PO₄ (90 : 10). The flow rate was 1 mL/min. UV detection wavelength was set at UV 360 nm. After the isolation of GA with preparative HPLC, the GA fraction was condensed to remove most of CH₃CN, and the condensed acidified solution was diluted with a large amount of water and loaded on an Sephadex LH-20 CC to remove the acid by eluting with H₂O. The subsequent Me₂CO elution was condensed to dryness, and GA was obtained (35 mg). In the HPLC/ESI-MS analysis, GA presented as one peak (m/z = 628) on a C₁₈ column (Alltima C₁₈, 5 μ, 4.6 × 250 mm) eluted with CH₃CN/0.1 % acetic acid (90 : 10). However, it presented as two completely separated peaks on a C₈ column (Alltima C₈, 5 μ, 4.6 × 250 mm) eluted with CH₃CN/0.1 % acetic acid (75 : 25). These two peaks were isolated by HPLC under the same analytical conditions with each injection of 20 μL acetone solution of GA (35 mg/mL), yielding 1 (12 mg) and 2 (10 mg).

Gambogic acid (1): bright yellow amorphous powder; [α]_D ²⁰: -578° (c 0.201, CHCl₃); UV (MeOH): λ_max = 290 (log ε 4.24), 360 nm (log ε 4.18); IR (KBr): λ_max = 2970, 2928, 1736, 1690, 1632, 1593, 1435, 1331, 1177, 1138, 1049, 810, 671 cm^-1; ESI-MS: m/z = 628 (M⁺); HR-ESI-MS: m/z = 628.3046 [M]⁺, calcd.: 628.3036.

Epigambogic acid (2): bright yellow amorphous powder; [α]_D ²⁰: -486° (c 0.197, CHCl₃); UV (MeOH): λ_max = 290 (log ε 4.24), 360 nm (log ε 4.18); IR (KBr): λ_max = 2971, 2930, 1736, 1691, 1633, 1593, 1435, 1332, 1177, 1138, 1048, 810, 671 cm^-1; ESI-MS: m/z =628 (M⁺); HR-ESI-MS: m/z = 628.3042 [M]⁺, calcd.: 628.3036. Copies of the original spectra are obtainable from the author of correspondence.

Both doxorubicin-resistant (K562/R) and -sensitive K562 (K562/S) cell sublines, purchased from the Tianjin Institute of Hematopathy, the Chinese Academy of Medical Sciences, China, were cultured in RPMI1640 (Gibco, USA) medium and supplemented with 10 volumes of fetal bovine serum at 37 °C in a humidified incubator with 5 % CO₂. The MTT assay was performed using a reported method [16]. Cell lines were seeded into 96-well plates at 6000 viable cells per well. The test chemicals (ADR, DMSO, Th1 and Th2) with different concentrations were loaded in a final volume of 200 μL per well. After 44 h incubation, MTT (5 g/L) was added to each well in a volume of 10 μL and incubated for 4 h. Afterwards, the medium was removed and 200 μL of Me₂SO (37 °C) were added and shaken for 5 min. A 96-well microtiter plate reader was used to determine absorbance values at 570 nm. The mean value of each concentration (n = 3 wells) was obtained. Absorbance of untreated controls was taken as 100 %. The survival rate was calculated as follows: cell survival rate (%) = (T - B)/(U - B) × 100 %, where T (treated) is the absorbance of chemically treated cells, U (untreated) is the absorbance of untreated cells, and B (blank) is the absorbance when neither cells nor chemicals were added. Human liver was obtained from one Chinese autopsy sample (male, aged 37) from Dalian Medical University, with the approval of the ethics committee of Dalian Medical University. HLM were prepared from liver tissue as described in [17]. Protein concentrations of the microsomal fractions were determined by the Lowry method using bovine serum albumin as a standard [18]. The inhibition effects of compounds were characterized using HLM toward six human cytochrome P-450 enzymes based on their probe reaction. Each incubation was performed in a 100 mM phosphate buffer at pH 7.4 containing human microsomal protein, 10 mM glucose 6-phosphate, 1 mM NADP⁺, 4 mM magnesium chloride, 1 unit/mL of glucose 6-phosphate dehydrogenase, and various probe substrates of CYPs and tested compounds (previously dissolved in methanol, whose final concentration was 1 %, v/v) with a range of concentrations in a total volume of 400 μL. And the selective inhibitors of each CYP isoform [furafylline (1A2), 8-methoxypsoralen (2A6), sulphaphenazole (2C9), quinidine (2D6), clomethiazole (2E1), ketoconazole (3A4)] were selected as the positive controls. There was a 3 min preincubation step at 37 °C before the reaction was started by the addition of NADP⁺. After 10 min, the reactions were quenched by adding the same volume of CH₃CN or MeOH and an internal standard. The incubation mixtures were then centrifuged for 10 min at 20,000 × g. An aliquot of the supernatant was analyzed by HPLC. The HPLC system (Shimadzu, Japan) consisted of an SCL-10A system controller, two LC-10AT pumps, a SIL-10A auto injector, a SPD-10AV UV detector or a RF-10A_XL fluorescence detector. The supernatant was analyzed using a Shimadzu C₁₈ column (4.6 × 150 mm, 5 μ) at a flow rate of 1 mL/min. IC₅₀ values (concentration of inhibitor causing 50 % inhibition of original enzyme activity) were calculated by Microsoft Excel software (Microsoft Inc, USA).

Supporting Information

The Supporting Information contains an HPLC chromatogram of GA, the mixture of 1 and the known R-epimer, and 2, ¹H- (400 MHz) and ¹³C- (100 MHz) NMR spectra of 1 and 2 (in CDCl₃, TMS as internal standard), cytotoxic activities of doxorubicin, 1, and 2 (μM), in vitro reaction and detection conditions for CYP-isoform bioassay.

Acknowledgements

We thank Dr. Sui Xiong Cai of Maxim Pharmaceuticals, San Diego, USA for his kind help with a reference crystal sample of the R-epimer 1. This research is funded by the Hong Kong Jockey Club Charities Trust.

References

• 1 Lin L J, Lin L Z, Pezzuto J M, Cordell G A. Isogambogic acid and isomorellinol from Garcinia hanburyi .  Magn Reson Chem. 1993;  31 340-7
  CrossrefPubMedSearch in Google Scholar
• 2 Asano J, Chiba K, Tada M, Yoshi T. Cytotoxic xanthones from Garcinia hanburyi .  Phytochemistry. 1996;  41 815-20
  CrossrefPubMedSearch in Google Scholar
• 3 Tseng B, Sirisoma N S, Cai S X, Zhang H Z, Kasibhatia S, Ollis K P. et al .Derivatives of gambogic acid and analogs as activators of caspases and inducers of apoptosis. PCT Int Appl 2004: WO 2 004 002 428
  Search in Google Scholar
• 4 Lei Q M, Liu J M. Retrospect and prospect of anti-cancer efficacy of gamboges.  Chin J Canc Prev Treat. 2003;  10 216-9
  PubMedSearch in Google Scholar
• 5 Guo Q L, Zhao L, You Q D, Wu Z Q, Gu H Y. Gambogic aicd inducing apoptosis in human gastric adenocarcinom SGC-7901 cells.  Chin J Nat Med. 2004;  2 106-10
  PubMedSearch in Google Scholar
• 6 Guo Q L, You Q D, Yuan S T, Zhao L. General gambogic acids inhibited growth of human hepatoma SMMC-7721 cells in vitro and in nude mice.  Acta Pharmacol Sin. 2004;  25 769-74
  PubMedSearch in Google Scholar
• 7 Wu Z Q, Guo Q L, You Q D, Zhao L. Growth inhibitory effect of GGAs on experimental tumor in mice and human tumor cell cultured in vitro .  Chin J Nat Med. 2004;  2 99-102
  PubMedSearch in Google Scholar
• 8 Zhao L, Guo Q L, You Q D, Wu Z Q, Gu H Y. Gambogic acid induces apoptosis and regulates expressions of Bax and Bcl-2 protein in human gastric carcinoma MGC-803 cells.  Biol Pharm Bull. 2004;  27 998-1003
  CrossrefPubMedSearch in Google Scholar
• 9 Wu Z Q, Guo Q L, You Q D, Zhao L, Gu H Y. Gambogic acid inhibits proliferation of human lung carcinoma SPC-A1 cells in vivo and in vitro and represses telomerase activity and telomerase reverse transcriptase mRNA expression in the cells.  Biol Pharm Bull. 2004;  27 1769-74
  CrossrefPubMedSearch in Google Scholar
• 10 Cardillo G, Merlini L. Absolute configuration of carbon 2 in the chromene ring of gambogic acid. Tetrahedron Lett 1967: 2529-30
  Search in Google Scholar
• 11 Weakley T JR, Cai S X, Zhang H Z, Keanal J FW. Crystal structure of the pyridine salt of gambogic acid.  J Chem Crystallogr. 2001;  31 501-5
  CrossrefPubMedSearch in Google Scholar
• 12 Zhang H Z, Kasibhatia S, Wang Y, Herich J, Guastella J, Tseng B. et al . Discovery, characterization and SAR of gambogic acid as a potent apoptosis inducer by a HTS assay.  Bioorg Med Chem. 2004;  12 309-17
  CrossrefPubMedSearch in Google Scholar
• 13 Beringer P M, Slaughter R L. Transporters and their impact on drug disposition.  Ann Pharmacother. 2005;  39 1097-108
  PubMedSearch in Google Scholar
• 14 Antoniou T, Tseng A L. Interactions between antiretrovirals and antineoplastic drug therapy.  Clin Pharmacokinet. 2005;  44 111-45
  PubMedSearch in Google Scholar
• 15 Goldstein J A, Morais S M. Biochemistry and molecular biology of the human CYP2C subfamily.  Pharmacogenetics. 1994;  4 285-99
  CrossrefPubMedSearch in Google Scholar
• 16 Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assay.  J Immunol Methods. 1983;  65 5-63
  PubMedSearch in Google Scholar
• 17 Sanderink G J, Bournique B, Stevens J, Petry M, Martinet M. Involvement of human CYP1A isoenzymes in the metabolism and drug interactions of riluzole in vitro .  J Pharm Exper Therap. 1997;  282 1465-72
  PubMedSearch in Google Scholar
• 18 Lowry O H, Roseborough N J, Farr A L, Randall R J. Protein measurement with the folin phenol reagent.  J Biol Chem. 1951;  193 265-75
  PubMedSearch in Google Scholar

Dr. Hongxi Xu

Hong Kong Jockey Club Institute of Chinese Medicine

Shatin

N.T.

Hong Kong

People’s Republic of China

Fax: +852-2603-6263

Email: xuhongxi@hkjcicm.org

References

• 1 Lin L J, Lin L Z, Pezzuto J M, Cordell G A. Isogambogic acid and isomorellinol from Garcinia hanburyi .  Magn Reson Chem. 1993;  31 340-7
  CrossrefPubMedSearch in Google Scholar
• 2 Asano J, Chiba K, Tada M, Yoshi T. Cytotoxic xanthones from Garcinia hanburyi .  Phytochemistry. 1996;  41 815-20
  CrossrefPubMedSearch in Google Scholar
• 3 Tseng B, Sirisoma N S, Cai S X, Zhang H Z, Kasibhatia S, Ollis K P. et al .Derivatives of gambogic acid and analogs as activators of caspases and inducers of apoptosis. PCT Int Appl 2004: WO 2 004 002 428
  Search in Google Scholar
• 4 Lei Q M, Liu J M. Retrospect and prospect of anti-cancer efficacy of gamboges.  Chin J Canc Prev Treat. 2003;  10 216-9
  PubMedSearch in Google Scholar
• 5 Guo Q L, Zhao L, You Q D, Wu Z Q, Gu H Y. Gambogic aicd inducing apoptosis in human gastric adenocarcinom SGC-7901 cells.  Chin J Nat Med. 2004;  2 106-10
  PubMedSearch in Google Scholar
• 6 Guo Q L, You Q D, Yuan S T, Zhao L. General gambogic acids inhibited growth of human hepatoma SMMC-7721 cells in vitro and in nude mice.  Acta Pharmacol Sin. 2004;  25 769-74
  PubMedSearch in Google Scholar
• 7 Wu Z Q, Guo Q L, You Q D, Zhao L. Growth inhibitory effect of GGAs on experimental tumor in mice and human tumor cell cultured in vitro .  Chin J Nat Med. 2004;  2 99-102
  PubMedSearch in Google Scholar
• 8 Zhao L, Guo Q L, You Q D, Wu Z Q, Gu H Y. Gambogic acid induces apoptosis and regulates expressions of Bax and Bcl-2 protein in human gastric carcinoma MGC-803 cells.  Biol Pharm Bull. 2004;  27 998-1003
  CrossrefPubMedSearch in Google Scholar
• 9 Wu Z Q, Guo Q L, You Q D, Zhao L, Gu H Y. Gambogic acid inhibits proliferation of human lung carcinoma SPC-A1 cells in vivo and in vitro and represses telomerase activity and telomerase reverse transcriptase mRNA expression in the cells.  Biol Pharm Bull. 2004;  27 1769-74
  CrossrefPubMedSearch in Google Scholar
• 10 Cardillo G, Merlini L. Absolute configuration of carbon 2 in the chromene ring of gambogic acid. Tetrahedron Lett 1967: 2529-30
  Search in Google Scholar
• 11 Weakley T JR, Cai S X, Zhang H Z, Keanal J FW. Crystal structure of the pyridine salt of gambogic acid.  J Chem Crystallogr. 2001;  31 501-5
  CrossrefPubMedSearch in Google Scholar
• 12 Zhang H Z, Kasibhatia S, Wang Y, Herich J, Guastella J, Tseng B. et al . Discovery, characterization and SAR of gambogic acid as a potent apoptosis inducer by a HTS assay.  Bioorg Med Chem. 2004;  12 309-17
  CrossrefPubMedSearch in Google Scholar
• 13 Beringer P M, Slaughter R L. Transporters and their impact on drug disposition.  Ann Pharmacother. 2005;  39 1097-108
  PubMedSearch in Google Scholar
• 14 Antoniou T, Tseng A L. Interactions between antiretrovirals and antineoplastic drug therapy.  Clin Pharmacokinet. 2005;  44 111-45
  PubMedSearch in Google Scholar
• 15 Goldstein J A, Morais S M. Biochemistry and molecular biology of the human CYP2C subfamily.  Pharmacogenetics. 1994;  4 285-99
  CrossrefPubMedSearch in Google Scholar
• 16 Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assay.  J Immunol Methods. 1983;  65 5-63
  PubMedSearch in Google Scholar
• 17 Sanderink G J, Bournique B, Stevens J, Petry M, Martinet M. Involvement of human CYP1A isoenzymes in the metabolism and drug interactions of riluzole in vitro .  J Pharm Exper Therap. 1997;  282 1465-72
  PubMedSearch in Google Scholar
• 18 Lowry O H, Roseborough N J, Farr A L, Randall R J. Protein measurement with the folin phenol reagent.  J Biol Chem. 1951;  193 265-75
  PubMedSearch in Google Scholar

Dr. Hongxi Xu

Hong Kong Jockey Club Institute of Chinese Medicine

Shatin

N.T.

Hong Kong

People’s Republic of China

Fax: +852-2603-6263

Email: xuhongxi@hkjcicm.org

• Supplementary Material