Magnesium as NMDA receptor blocker in the traditional Chinese medicine Danshen.
Aqueous extracts of the traditional Chinese medicine Danshen, the dried roots of Salvia miltiorrhiza Bunge (Labiatae), blocked N-methyl-D-aspartate (NMDA) evoked currents in cerebrocortical neurons in vitro. The block of the NMDA-evoked currents was voltage dependent and showed the negative slope conductance reminiscent of the effect of [Mg.sup.2+] ions. Atomic absorption spectrophotometry (AAS) revealed that aqueous Danshen extracts contained ~9 mM magnesium. Fractionation of the extracts by high performance liquid chromatography followed by patch clamp recording and AAS indicated that magnesium ions were present in two distinct fractions. One fraction contained ~5 mM magnesium and blocked NMDA-induced currents indicating that it contained mostly free [Mg.sup.2+] ions, while a second fraction did not possess NMDA antagonist activity despite the presence of ~4 mM magnesium suggesting that [Mg.sup.2+] in this fraction was mostly chelated. Following removal of the free [Mg.sup.2+] by ion exchange chromatography, the previously observed block of the NMDA-induced currents was abolished. These data demonstrate that Danshen contains both free and chelated [Mg.sup.2+]. Free [Mg.sup.2+] ions account for the NMDA antagonist activity of Danshen in vitro.
[c] 2004 Published by Elsevier GmbH.
Keywords: Magnesium; Danshen; N-methyl-D-aspartate receptor; Patch clamp; Traditional Chinese medicine
N-methyl-D-aspartate receptor (NMDAR) antagonists are neuroprotective in animal models of stroke; however clinical trials have so far failed because of adverse central nervous system (CNS) effects induced by drugs targeting the NMDAR (Choi, 2002). Although a number of factors contributed to the apparent clinical failure of these drugs (Dichter and Locke, 2003; Legos et al., 2002), the disappointment has been such that some have concluded that the concept of NMDAR antagonism is intrinsically not a valid therapeutic approach (Parsons et al., 1999). Giving up on NMDA antagonists may be premature, however. There is now good evidence that the therapeutic effects of the drug memantine, that has been in use for the treatment of dementia for almost 20 years in Europe, are most likely due to NMDAR antagonist activity (for a review see Parsons et al., 1999). Thus, NMDARs might still prove to be a valid target for neuroprotective therapy provided that the drugs possess the "right" pharmacodynamic profile.
Traditional Chinese medicine (TCM) stroke therapy has a long history and its therapeutic efficacy has been confirmed by recent clinical studies (Gong and Sucher, 1999). Multiple lines of research suggest that TCM preparations commonly used in stroke therapy are neuroprotective with few or no adverse effects (for a review see Gong and Sucher, 1999). To date, however, the molecular basis of their therapeutic actions is largely unknown. We hypothesized that at least part of the neuroprotective effect of TCMs might be due to compounds acting as NMDAR antagonists. If so, then those compounds might be NMDA antagonists with a therapeutically beneficial profile.
Along these lines, we previously identified 22 TCMs that are commonly used in prescriptions for therapy of the acute phase of stroke (Gong and Sucher, 1999) and screened aqueous extracts of these TCMs for NMDAR antagonist activity (Sun et al., 2003). Aqueous extracts of the TCM Danshen, the dried roots of Salvia miltiorrhiza Bunge (Labiatae), blocked NMDA evoked currents in whole cell patch-clamp experiments (Sun et al., 2003). The block of the NMDA-evoked currents was voltage dependent and showed the negative slope conductance reminiscent of the effect of the open channel blockers [Mg.sup.2+] and memantine (Chen et al., 1992). Atomic absorption spectrophotometry revealed the presence of considerable amounts of magnesium in the extracts (Sun et al., 2003). We undertook the present study to determine whether [Mg.sup.2+] or other as yet unidentified compounds accounted for the observed block of NMDA-induced currents by aqueous Danshen extracts.
Using high performance liquid chromatography (HPLC) and atomic absorption spectrophotometry (AAS), we found that magnesium ions were present in two distinct fractions. Whole-cell the patch-clamp recordings revealed that only one of the fractions blocked NMDA-induced currents indicating that it contained mostly free [Mg.sup.2+] ions. The second fraction did not possess NMDA antagonist activity despite the presence of ~4 mM magnesium suggesting that the ions were most likely chelated. Following removal of the free [Mg.sup.2+] by ion exchange chromatography, the previously observed block of the NMDA-induced currents was abolished. These experiments show that NMDA antagonist activity in Danshen extracts is due to its significant content of free [Mg.sup.2+].
Materials and methods
Danshen was purchased from the authorized TCM supplier 'The House of Ren Ji' in Shenzhen (China). Acetonitrile (Chrom AR HPLC) was purchased from Mallinckrodt (Mississauga, Ont., Canada). Ion exchange resin (Dowex[R] 50 w X 4-50) was purchased from Aldrich (St. Louis, MO, USA). Neurobasal (NB) medium, serum-free supplement B27 and L-glutamine were purchased from Gibco (Grand Island, NY, USA). Hanks' balanced salt solution (HBSS; with or without [Ca.sup.2+] and [Mg.sup.2+]), sodium pyruvate, glutamate and penicillin/streptomycin, NMDA and glycine were purchased from Sigma (St. Louis, MO, USA). DL-2-amino-5-phosphonopentanoic acid (APV) was obtained from Tocris (Avonmouth, UK).
Extract preparation and fractionation
Dried roots of S. miltiorrhiza (Danshen; 10 g) were washed with double-distilled water (dd[H.sub.2]O) and crashed. The drug was extracted with 100 ml dd[H.sub.2]O under reflux for 5 h using a Soxhlet apparatus. The extract was separated into fractions by preparative HPLC (Waters PrepLC 4000 System, Waters. USA) using a reverse phase column (Alltim C18 100A5u, Alltech. USA). The eluted peaks were monitored at 280 nm using a Waters 2487 Dual absorbance detector (Waters. USA). A gradient solvent system was used for HPLC. Briefly, the eluant solution was a mixture of acetonitrile (A) and water containing 0.1% acetic acid (B). The protocol consisted of elution of the column with 100% solvent B for 10 min, a linear gradient of solvent A (increasing from 20% to 55%) and B for 30 min, followed by a final step of solvent B for 20 min. The flow rate was maintained at 2 ml/min. Fractions were collected at 10 min intervals and immediately freeze-dried under vacuum (Virtis Freeze Dryer, USA). The dried powder was resuspended in dd[H.sub.2]O (at the same volume of the extract applied to the HPLC column in order to maintain the concentration of the individual components unchanged).
Cation exchange and determination of [Mg.sup.2+] concentration
The cation exchange resin was packed into a 30 mm X 100 mm column and washed with 2 M HCl (20 X column volume), followed by [H.sub.2]O (10 X column volume), 1 M NaOH (10 X column volume) and finally 10 volumes of 1 M HCl. HPLC fractions were loaded on the column and eluted with milliQ [H.sub.2]O (ion free, pH 7.0) and immediately freeze-dried under vacuum. The composition of the fractions was monitored by HPLC (Waters analytical HPLC 600). [Mg.sup.2+] content of the fractions was determined using an atomic absorption spectrophotometer (Hitachi, Tokyo, Japan). A calibration curve was generated with standards of known [Mg.sup.2+] concentrations. The absorbance values obtained from fractions were converted into [Mg.sup.2+] concentrations using the standard curve.
Cell culture and whole-cell patch-clamp measurements
Cortical neurons from mouse embryos at day 16 were isolated, cultured and used for whole-cell patch-clamp recording as described in detail previously (Sun et al., 2003).
We used AAS to determine the concentration of magnesium in aqueous Danshen extracts. We found that aqueous extracts prepared from 10 g of Danshen in 100 ml of water contained 8.7 mM magnesium corresponding to ~2 mg of magnesium per gram of Danshen.
Next, we separated the extract into two fractions using preparative HPLC (Fig. 1). The first fraction corresponded to the first 10 min of the HPLC profile, the second fraction to the remainder. Both fractions were tested by patch-clamp assay for their ability to inhibit NMDA-induced currents in mouse cerebrocortical neurons. The patch-clamp data indicated that only the first fraction blocked NMDA-evoked currents.
[FIGURE 1 OMITTED]
In parallel with the electrophysiological studies, we used AAS to measure the [Mg.sup.2+] in the fractions. The AAS data revealed that the concentration of [Mg.sup.2+] in the fraction that blocked NMDA-induced currents was 4.7 mM. Next, we used a cation-exchange column to eliminate [Mg.sup.2+] from this fraction. Subsequent to the ion exchange chromatography, we used AAS to verify that the magnesium had been completely removed. In order to ensure that the procedure did not significantly change the chemical composition of the fraction, we used analytical HPLC to compare the UV absorption profiles of the fraction before and after ion exchange chromatography. AAS demonstrated that the concentration of [Mg.sup.2+] was zero after the ion exchange. At the same time, we did not observe any obvious difference comparing the UV absorption profiles before and after the ion exchange (data not shown).
Whole-cell patch-clamp recordings revealed that removal of [Mg.sup.2+] from the Danshen fraction with NMDA antagonist activity at negative holding potentials led to the total loss of its antagonist activity (Fig. 2) resulting in a linear current/voltage relation (Fig. 3).
The major components in aqueous extracts of the S. miltiorrhiza roots are water-soluble compounds such as tanshinol (danshensu), rosmarinic acid, and lithospermic acid B. Lithospermic acid B has previously been shown to be present in Danshen extracts mostly as [Mg.sup.2+] salt (Tanaka et al., 1989). However, neither lithospermic acid nor any of the other major components of the aqueous extracts showed NMDAR antagonist activity even at millimolar concentration (not illustrated).
The present study demonstrated that aqueous extracts of Danshen contain millimolar amounts of magnesium, which is present in both free and chelated forms. Following removal of the free [Mg.sup.2+] by ion exchange chromatography, the previously observed block of the NMDA-induced currents was abolished.
Magnesium is an essential mineral for humans and thought to be a required co-factor for more than 300 enzymes. Interestingly, the average modern diet contains significantly less than the recommended daily allowance resulting in widespread magnesium deficiency (Marier, 1986). Danshen contains ~2 mg of magnesium per gram. TCM prescriptions such as Huoluo Xiaoling Dan ("Bolus for active energy flow in channels and collaterals"), Danshen Yin ("Red sage drink for removing blood stasis") Dingxian Wan ("Pills for relieving epilepsy") contain 15, 30 and 60 g of Danshen, respectively. Thus, patients might potentially ingest between 30 and 120 mg of (free and chelated) magnesium from Danshen per decoction. Although the fractional absorption of orally administered magnesium is highly variable and it is not known whether or not chelated [Mg.sup.2+]-ions are absorbed, ingestion of tens of milligrams of magnesium might potentially lead to a significant increase of the magnesium concentration in plasma (~25%). Thus, it is conceivable that repeated therapeutic ingestion of TCM decoctions containing Danshen might lead to a sustained increase in plasma magnesium levels that in turn might induce beneficial therapeutic effects.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
Danshen has been reported to possess multiple potentially beneficial activities such as being anti-inflammatory (Lee et al., 1987), reducing platelet aggregation (Onitsuka et al., 1983), protecting against oxidative stress (Weng and Gordon, 1992; Zhang et al., 1990) and protecting the cardiovascular system (Chen et al., 1986; Yagi et al., 1989). Interestingly, at least part of the neuroprotective effects of Danshen may be due to the action of tanshinones, which have been reported recently to reduce stroke size following transient middle cerebral artery occlusion in mice (Lam et al., 2003). Considering the pleiotropic actions of magnesium it remains to be established whether or not any therapeutic effects of Danshen include the action of magnesium as NMDAR channel blocker (Muir, 2002).
We would like to thank Dr. M. Carles for helpful discussions and critical reading of the manuscript. This work was supported by the Hong Kong Research Grants Council and the Area of Excellence Grant AoE/B-15/01 in Neuroscience Research from the University Grants Council of Hong Kong. N.J.S. was recipient of a sponsored research grant from VivoQuest (NY, USA).
Received 10 September 2003; accepted 20 November 2003
Chen, C., Chen, H., Chen, Y., Hsu, H., Hsieh, T., 1986. Isolation of the components of Salviae miltiorrhizae radix and their coronary dilator activities. J. Taiwan Pharmac. Assoc. 38, 226-230.
Chen, H.S., Pellegrini, J.W., Aggarwal, S.K., Lei, S.Z., Warach, S., Jensen, F.E., Lipton, S.A., 1992. Open-channel block of N-methyl-D-aspartate (NMDA) responses by memantine: therapeutic advantage against NMDA receptor-mediated neurotoxicity. J. Neurosci. 12, 4427-4436.
Choi, D.W., 2002. Exploratory clinical testing of neuroscience drugs. Nat. Neurosci. 5 (Suppl.), 1023-1025.
Dichter, M., Locke, R., 2003. Clinical trials in neuroprotection. Xpert Opin. Emerging Drugs 8, 267-271.
Gong, X., Sucher, N.J., 1999. Stroke therapy in traditional Chinese medicine (TCM): prospects for drug discovery and development. Trends Pharmacol. Sci. 20, 191-196.
Lam, B., Lo, A., Sun, X., Luo, H., Chung, S., Sucher, N., 2003. Neuroprotective effects of tanshinones in transient focal cerebral ischemia in mice. Phytomedicine 10, 286-291.
Lee, A., Wu, W., Chang, W., Lin, H., King, M., 1987. Isolation and bioactivity of new tanshinones. J. Nat. Prod. 50, 157-160.
Legos, J.J., Tuma, R.F., Barone, F.C., 2002. Pharmacological interventions for stroke: failures and future. Expert Opin. Investig. Drugs 11, 603-614.
Marier, J., 1986. Magnesium content of the food supply in the modern-day world. Magnesium 5, 1-8.
Muir, K., 2002. Magnesium in stroke treatment. Postgrad Med. J. 78, 641-645.
Onitsuka, M., Fujiu, M., Shinma, N., Maruyama, H.B., 1983. New platelet aggregation inhibition from Tan-shen; Radix of Salvia miltiorrhiza Bunge. Chem. Pharm. Bull. 31, 1670-1675.
Parsons, C.G., Danysz, W., Quack, G., 1999. Memantine is a clinically well tolerated N-methyl-D-aspartate (NMDA) receptor antagonist-a review of preclinical data. Neuropharmacology 38, 735-767.
Sun, X., Chan, L.N., Gong, X., Sucher, N.J., 2003. N-methyl-D-aspartate receptor antagonist activity in traditional Chinese stroke medicines. Neurosignals 12, 31-38.
Tanaka, T., Morimoto, S., Nonaka, G., Nishioka, I., Yokazawa, T., Young, H., Oura, H., 1989. Magnesium and ammonium-potassium lithospermates B, the active principles having a uremia-preventive effect from Salvia miltiorrhiza. Chem. Pharm. Bull. 37, 340-344.
Weng, X., Gordon, M., 1992. Antioxidant activity of quinones extracted from Tanshen (Salvia miltiorrhiza Bunge). J. Agricult. Food Chem. 40, 1331-1336.
Yagi, A., Fujimoto, K., Tanonaka, K., Hirai, K., Takeo, S., 1989. Possible active components of Tanshen (Salvia miltiorrhiza) for protection of the myocardium against ischemia-induced derangements. Planta Med. 55, 51-54.
Zhang, K., Bao, Y., Wu, P., Rosen, R., Ho, C., 1990. Antioxidative components of Tanshen (Salvia miltiorrhiza Bunge). J. Agricult. Food Chem. 38, 1194-1197.
X. Sun (a,b,c), L.N. Chan (a,b), N.J. Sucher (a,b,d,*)
(a) Department of Biology, Hong Kong University of Science & Technology, Clear Water Bay, Kowloon, Hong Kong SAR, China
(b) Molecular Neuroscience Center, Hong Kong University of Science & Technology, Clear Water Bay, Kowloon, Hong Kong SAR, China
(c) Institute of Neuroscience, Nanjing Medical University, Nanjing, Jiangsu, China
(d) Biotechnology Research Institute, Hong Kong University of Science & Technology, Clear Water Bay, Kowloon, Hong Kong SAR, China
*Corresponding author. Department of Neurology/Division of Neuroscience, Children's Hospital and Harvard Medical School, Enders 353, 300 Longwood Avenue, Boston, MA 02115 USA. Tel.: + 1 617 355 3515; Fax: + 1 617 730 0243.
E-mail address: firstname.lastname@example.org (N.J. Sucher).