Introduction

Na+/K+-ATPase, an active transporter of sodium and potassium ions, is responsible for maintaining membrane potentials, the cell volume, and the active transport of other solutes in animal cells1. The therapeutic effect of cardiac glycosides in the treatment of congestive heart failure depends on the reversible inhibition of the Na+/K+-ATPase located in the cell membrane of the human myocardium2,3. Although the inhibition of the Na+/K+-ATPase produces beneficial effects in patients with congestive heart failure, severe side effects and the narrow therapeutic index of cardiac glycosides have evidently limited their clinical applications4.

Many steroid-like compounds found in a variety of Chinese herbs used for promoting blood circulation were demonstrated to be inhibitors of Na+/K+-ATPase, and thus regarded as the active ingredients responsible for their cardiac therapeutic effects via the same molecular mechanism triggered by cardiac glycosides5,6,7,8,9. However, no appreciable level of steroid-like compounds were found in danshen (Salvia miltiorrhiza), a well-known Chinese herb traditionally used for promoting blood circulation10. Instead, lithospermate B (LSB) in complex with Mg2+ was found to be the major soluble ingredient in danshen and shown to be an effective inhibitor of Na+/K+-ATPase, which is presumably responsible for the cardiac therapeutic effect of this herb11. Being non-toxic antioxidants without apparent adverse effects, Mg2+-LSB and LSB may be used as substitutes for cardiac glycosides for the treatment of congestive heart failure12.

To evaluate in vivo pharmacological activities, the metabolic fate of Mg2+-LSB was examined in rats13. Four major metabolites were excreted into bile after the intravenous injection of Mg2+-LSB, and identified via mass spectrometry as meta-O-methylated derivatives of LSB, namely 3-monomethyl-LSB, 3,3′′-dimethyl-LSB, 3,3′′′-dimethyl-LSB, and 3,3′′,3′′′-trimethyl-LSB. These methylated metabolites were found to be potent antioxidants, and thus assumed to be largely responsible for the pharmacological effects of Mg2+-LSB.

In complex with Mg2+, LSB possesses a relatively rigid structure due to the formation of salt bridges between Mg2+ and the four oxygen atoms of the carboxyl groups on the four caffeic acid fragments14. Comparatively, the rigid structure around the salt bridges formed between Mg2+ and carboxyl groups partially mimics the core steroid structure of cardiac glycosides. Recently, we demonstrated that some transition metal ions were able to replace Mg2+ to form stable complexes with LSB15. The in vitro potencies (ie, the inhibition of Na+/K+-ATPase activity) of LSB complexed with Cr3+, Mn2+, Co2+, or Ni2+ increased by approximately 5 times compared with the naturally occurring LSB and Mg2+-LSB. Thus, the transition metal-LSB complexes have the potential to be superior substitutes for cardiac glycosides in the treatment of congestive heart failure. To further explore this potential utilization, we aimed to examine, in this study, the safety and metabolites of transition metal-LSB complexes after intravenous injection in rats.

Materials and methods

Chemicals and reagents

HPLC grade acetonitrile and methanol were purchased from Fisher Scientific (Fair Lawn, NJ, USA). Glacial acetic acid (>99.7%) was obtained from J T Baker Chemical Co (Phillipsburg, NJ, USA). Phosphoric acid (85%) and analytic grade formic acid were bought from Merck Millipore (Gibbstown, NJ, USA). Water was purified by a Millipore clear water purification system (Direct-Q, Millipore, Billerica, MA, USA). Purified LSB was a gift from KO DA Pharmaceutical Co (Taiwan, China). Mg(OH)2 was purchased from Showa Chemical Co (Tokyo, Japan), while NaOH, MnCl2, NiCl2, CrCl3, and CoCl2 were obtained from Sigma-Aldrich Co (St Louis, MO, USA).

Preparation of metal-LSB complexes

Metal-LSB complexes were prepared and characterized as described in our previous study15. Briefly, Mg2+-LSB, and Zn2+-LSB complexes were prepared in 20 mL of H2O by mixing equimolar concentrations of LSB (a final concentration of 50 mmol/L) with Mg(OH)2 and Zn(OH)2, respectively. To prepare Cr3+-LSB, Mn2+-LSB, Co2+-LSB, and Ni2+-LSB complexes, LSB (50 mmol/L) was first precipitated with NaOH (100 mmol/L) in 20 mL of ethanol, and the precipitation was then dissolved by adding CrCl3, MnCl2, CoCl2, and NiCl2 (50 mmol/L) to form metal-LSB complexes, respectively. These metal-LSB complexes were lyophilized at -86 °C and stored at -20 °C prior to usage. The purity of metal-LSB complexes in powder was approximately 85% as estimated by HPLC analysis.

Animal studies

Male Sprague-Dawley rats weighing 250–270 g were purchased from BioLasco, Taiwan Co, Ltd (Taiwan, China). The animals were adapted in a standard controlled environment of 23±2 °C, 60%±10% humidity and a 12-h light/dark cycle, and fed with hard rat chow pellets (Fwusow Ind Corp, Taiwan, China) and purified water ad libitum. The animal experiments were approved by the Institutional Animal Care and Use Committee of the National Chung-Hsing University (IACUC Approval No: 101–107(R)).

Bile collection and preparation

Thirty-three male Sprague-Dawley rats were fasted overnight but had access to water ad libitum. The animals were anesthetized with Zoletil 50® (40 mg/kg, ip; Virbac Laboratories, Carros, France) and remained anesthetized during the surgical operation. LSB or LSB complexes with Mg2+, Zn2+, Ni2+, Mn2+, and Co2+ (50 mg/kg, iv) were dissolved in normal saline and Cr3+-LSB complex was dissolved in 50% poly(ethylene glycol)-400 (v/v) (Fluka Chemie, Buchs, Switzerland); solutions were injected into the right femoral vein. Bile fistulas of the rats were cannulated with PE-20 polyethylene tubing for the collection of bile. The bile was collected into successive tubes on ice at 10 or 30 min intervals for 60 min after a single intravenous dosing. Bile samples of 200 μL were vortex-mixed with two volumes of methanol containing 0.1% H3PO4 for 10 min, and centrifuged at 10 000×g for 20 min at 4 °C. The supernatant was filtered through a 0.22 μm polyvinylidene difluoride (PVDF) membrane filter (PALL Corp, Glen Cove, NY, USA), and used for the following analyses.

Blood sampling and preparation

The left femoral vein was cannulated with PE-50 polyethylene tubing and connected with a 23-gauge needle for blood sampling. After intravenous administration with 100 mg/kg of Zn2+-LSB from the right femoral vein, blood samples of 300 μL were withdrawn in heparinized tubes on ice at 0, 5, 15, 30, and 60 min. Plasma was obtained by centrifugation at 3000×g for 15 min at 4 °C. For analysis, plasma (100 μL) was mixed with methanol (200 μL) containing 0.1% H3PO4 and vortexed for 10 min. After centrifugation at 10 000×g for 20 min at 4 °C, the supernatant was filtered by a 0.22 μm PVDF membrane filter and subjected to HPLC and LC/MS/MS analyses.

HPLC/UV and LC/MS/MS analyses

Bile and plasma samples were analyzed by HPLC coupled to a Waters Corp 600 controller pump with a 2996 photodiode array detector and a 717 autosampler (Milford, MA, USA). The separation was achieved using a Syncronis™ C18 column (250 mm×4.6 mm id, 5 μm) from Thermo Scientific (Waltham, MA, USA). The HPLC mobile phase comprised (A) water with 0.5% acetic acid (v/v) and (B) acetonitrile. The gradient for metabolism analysis started at 95% solvent A and 5% solvent B for 5 min, followed by a linear increase of solvent B to 70% for 20 min, and then a decrease of solvent B to 5% for 5 min. For a continual sample analysis, the column was equilibrated with 5% solvent B for 10 min before the next sample injection. The injection volumes of bile and blood samples were 20 and 150 μL, respectively. In all analyses, the column was kept at room temperature, and the flow rate was 1 mL/min. The metabolites of bile and blood were analyzed by a Thermo Finnigan™ LTQ™ linear ion trap mass spectrometer (Thermo LTQ XL, San Jose, CA, USA) equipped with an electrospray ionization (ESI) interface, and connected to a Thermo Scientific Surveyor LC plus system equipped with a Surveyor MS pump plus and a Surveyor autosampler (Thermo Scientific, San Jose, CA, USA). The mass spectra were obtained in a negative ESI mode. The spray voltage was 3.7 kV and the heated capillary temperature was 300 °C. Sheath gas and auxiliary gas flow rates were 30 and 3 arbitrary units, respectively. The full mass spectra were obtained at a mass-to-charge ratio (m/z) scan rate from 150 to 1500. To obtain the product ion spectra, the relative collision energy of the collision-induced dissociation (CID) was set at 24% for m/z 717 and 23% for m/z 731, 745, and 759. Separations were performed using a Waters Corp Xbridge™ C18 column (100 mm×2.1 mm id, 3.5 μm; Milford, MA, USA) at room temperature, and the injection volume was 5 μL at a flow rate of 0.15 mL/min. The tray and column oven temperature were set at 4 and 30 °C, respectively. The mobile phase comprised (C) water with 0.2% formic acid (v/v) and (D) acetonitrile with 0.2% formic acid (v/v). The program for gradient elution started at 95% solvent C and 5% solvent D for 5 min, followed by a linear increase of solvent D to 70% for 30 min, maintained at 70% solvent D for 8 min, and then a decrease of solvent D to 5% for 2 min. The detection wavelength was set at 288 nm.

Statistical analysis

The metabolite intensities (peak areas) were analyzed with Student's t-test performed by SigmaStat (Version 3.5). P values less than 0.05 were considered statistically significant.

Results

Metabolites of metal-LSB complexes in rat bile

To examine the excretion of biliary metabolites, bile samples of three rats were collected after intravenous injection with 50 mg/kg of LSB complexed with Mg2+, Zn2+, Cr3+, Ni2+, Mn2+, or Co2+. Surprisingly, rats injected with Co2+-LSB perished. Three more rats were used to repeat Co2+-LSB intravenous injections; however, all three rats again perished in our experimental conditions. Regardless, similar profiles of metabolites in bile samples were observed when rats were injected with LSB and the rest of metal-LSB complexes except for Mn2+-LSB. For each metal-LSB complex, similar results were observed among the three injected rats; a representative pattern of each metal-LSB was shown to illustrate the metabolite profile (Figure 1).

Figure 1
figure 1

HPLC chromatogram and color of bile collected from rats at basal, 0–30 min and 31–60 min after intravenous administration of LSB (A), Mg2+-LSB (B), Zn2+-LSB (C), Cr3+-LSB (D), Ni2+-LSB (E), or Mn2+-LSB (F). Rats were injected with each sample at 50 mg/kg.

PowerPoint slide

Full size image

Identification of biliary metabolites

Five peaks were consistently observed in the biliary metabolites of rats injected with any of the metal-LSB complexes, as observed in the metabolite profile of Zn2+-LSB in rat bile (Figure 2A). According to previous studies on the metabolites of Mg2+-LSB13,16, the five peaks were putatively identified by LC/MS/MS as LSB and four meta-O-methylated metabolites of LSB, namely 3-monomethyl-LSB (M1), 3,3′′-dimethyl-LSB (M2), 3,3′′′-dimethyl-LSB (M3), and 3,3′′,3′′′-trimethyl-LSB (M4; Figures 2B and 2C). The data displayed the extracted ion chromatograms [M-H]− for LSB at m/z 717 and MS2 ions at m/z 717 and 519 with a 23.3 min retention time, M1 at m/z 731 and MS2 ions at m/z 731 and 533 with a 24.5 min retention time, M2 and M3 at m/z 745 and MS2 ions at m/z 745 and 547 with 26.1 and 25.7 min retention times, respectively, and M4 at m/z 759 and MS2 ions at m/z 759 and 547 with a 24.5 min retention time. ESI-MS showed that M1, M2, M3, and M4 had molecular ion peaks at m/z 731, 745, 745, and 759 [M+CH3-H]−, respectively. The molecular weights of the four metabolites were 14, 28, 28, and 42 mass units higher than that of LSB, as expected of the four methylated metabolites. The same mass spectrometric outcomes were observed for the five equivalent peaks found in the biliary metabolites of rats injected with LSB, Mg2+-LSB, Cr3+-LSB, Ni2+-LSB, or Mn2+-LSB (data not shown).

Figure 2
figure 2

(A) UV and extracted ion chromatograms for the [M-H]− ions of Zn2+-LSB at 717 m/z, the monomethyl-M1 metabolite at 731 m/z, the dimethyl-M2 and M3 metabolites at 745 m/z, and the trimethyl-M4 metabolite at 759 m/z. (B) MS/MS spectra of [M-H]− ions of Zn2+-LSB, the monomethyl-M1, the dimethyl-M2 and M3, and the trimethyl-M4 metabolites. (C) Structures of metal-LSB complexes and four metabolites. Metal represents Mg2+, Zn2+, Cr3+, Ni2+, or Mn2+.

PowerPoint slide

Full size image

In vivo metabolism of Mg2+-LSB, Zn2+-LSB, and Mn2+-LSB

To monitor the metabolism of metal-LSB complexes in detail, bile samples were collected at 10 min intervals for 60 min after rats were intravenously injected with 50 mg/kg of LSB complexed with Mg2+, Zn2+, and Mn2+. Detailed tracking of the three successive bile samples concurrently suggested that in the following metabolism, LSB was first methylated to form M1, which was further methylated to form M2 (relatively fast) and M3 (relatively slow); as a final note, both M2 and M3 were complementarily methylated to form M4 (Figures 3 and 4). It seemed that the methylation of LSB occurred sequentially at three sites, ie, first at position 3, then 3′′, and, finally, the 3′′′ hydroxyl group. Relatively speaking, the metabolism (ie, methylation at the three hydroxyl groups) of Mn2+-LSB was faster than those of Mg2+-LSB and Zn2+-LSB. However, the elimination rate of the M4 methylated from Mn2+-LSB was relatively slow and remained a major metabolite in bile compared with the results from Mg2+-LSB and Zn2+-LSB methylations.

Figure 3
figure 3

HPLC profiles of bile metabolites collected from rats injected with Mg2+-LSB (A), Zn2+-LSB (B), and Mn2+-LSB (C) complexes. Rats were injected with each metal-LSB at 50 mg/kg, and bile samples were collected at 0, 10, 20, 30, 40, 50, and 60 min after injection.

PowerPoint slide

Full size image
Figure 4
figure 4

Proposed metabolic pathway of metal-LSB complex. M1, M2, M3, and M4 stand for 3-monomethyl-LSB, 3,3′′-dimethyl-LSB, 3,3′′′-dimethyl-LSB, and 3,3′′,3′′′-trimethyl-LSB, respectively. Metal represents Mg2+, Zn2+, Cr3+, Ni2+, or Mn2+.

PowerPoint slide

Full size image

Plasma metabolites of Zn2+-LSB

Because similar profiles of metabolites in bile samples were observed when rats were injected with LSB and the metal-LSB complexes, Zn2+-LSB was representatively selected to inspect plasma metabolites of metal-LSB complexes. Zn2+-LSB of 100 mg/kg was used for intravenous injections in rats, and blood samples were collected at 0, 5, 15, 30, and 60 min after injections. In the HPLC profile, M1 and M2 were barely detected in the plasma 5 min after injection (Figure 5A). Nevertheless, monomethyl-LSB (M1), dimethyl-LSB (M2) and trimethyl-LSB (M4) were all detectable in the LC/MS/MS analyses (Figures 5B–5F). In agreement with the HPLC profile, M1 and M2 were detected in the plasma 5 min after injection, while M4 was detected 15 min after injection. However, the relatively minor intermediate metabolite, M3, was undetectable in this analytic condition. Overall, the results suggest that the metabolites of metal-LSB complexes in plasma are fundamentally identical to those in bile.

Figure 5
figure 5

Identification of plasma metabolites after intravenous injection of Zn2+-LSB with a dosage of 100 mg/kg. Zn2+-LSB and its plasma metabolites at 0, 5, 15, 30, and 60 min after injection were analyzed by HPLC (A). Extracted ion chromatograms for the [M-H]− ions of plasma metabolites were detected for the samples of 0 min (B), 5 min (C), 15 min (D), 30 min (E), and 60 min (F) at m/z 717, 731, 745, and 759.

PowerPoint slide

Full size image

Discussion

In the present study, four meta-O-methylated metabolites (M1, M2, M3, and M4) were detected in bile samples of rats after intravenous injections of LSB, Mg2+-LSB, Zn2+-LSB, Cr3+-LSB, Ni2+-LSB, and Mn2+-LSB. These four methylated metabolites were identical to those detected in rat bile after intravenous administrations of LSB and Mg2+-LSB in a previous study16. Presumably, the four methylated LSB metabolites were sequentially formed by a hepatic enzyme, catechol O-methyltransferase (COMT), which catalyzed the transfer of the methyl group from S-adenosyl methionine to the meta-hydroxyl group of the catechol moiety prior to enterohepatic circulation in rats17. The methylation of phenolic compounds tend to result in a lower polarity and a higher metabolic stability by preventing the conjugation of glucuronic acid and sulfate groups, and can thus, be regarded as a route to improve the NQO1-inducing activities of phenolic acids, such as LSB, in danshen18,19. Our results suggest that in the treatment of cardiovascular diseases, for at least a comparable dosage lower than 50 mg/kg, iv in rats, the artificial LSB complexes with transition metals Zn2+, Cr3+, Ni2+, and Mn2+ have potential as safe, even superior therapeutic substitutes for LSB and Mg2+-LSB naturally isolated from danshen.

According to our observations, the metabolic rates of metal-LSB complexes via methylation by COMT were comparable except for that of Mn2+-LSB (Figure 1). M1 and M2 were found to be major metabolites of LSB and of Mg2+-LSB, Zn2+-LSB, Cr3+-LSB, and Ni2+-LSB, while M2 and M4 were the major metabolites of Mn2+-LSB. In contrast with the metabolism of other metal-LSB complexes, Mn2+-LSB-converted M1 was immediately methylated to form M2; however, the elimination rate of its further methylated metabolite, M4, was relatively slow, which resulted in its accumulation as a major metabolite of Mn2+-LSB. It is likely that Mn2+-LSB and its monomethylated metabolite (M1) are better substrates of COMT than other metal-LSB complexes and their monomethylated metabolites, while the trimethylated metabolite (M4) of Mn2+-LSB is relatively resistant to additional metabolism in rats. Whether the different metabolic fate of Mn2+-LSB, in comparison with those of other metal-LSB complexes, leads to a distinctive pharmacological effect on the therapeutic activity of LSB remains to be evaluated.

Unexpectedly, all the rats died within 10 min after intravenous administration with 50 mg/kg of the Co2+-LSB complex. In a preliminary test, rats survived when the intravenous administration of Co2+-LSB was reduced to less than one-fifth (10 mg/kg; data not shown). Cobalt ion is one of the necessary essential elements for humans as suggested by the World Health Organization (WHO), and its daily recommended intake is 5–40 μg/day20. It stimulates the production of erythropoietin and red blood cells for the prevention of anemia. Additionally, cobalt ions raise the blood oxygen-carrying capacity to prevent ischemia and hypoxia. However, an overdose of cobalt ion may be harmful due to its toxic effects on the hematopoietic system21, thyroid22, and lungs23; additionally, its neurotoxicity24, cardiomyopathy25, and carcinogenicity26 have been reported. The adverse effects were consistently observed when the concentration of cobalt ion in human blood exceeded 800 μg/L27. The oral LD50 values of Wistar and Sprague-Dawley rats for single administrations of cobalt ion were reported to be 42, 317, 631, and 3672 mg/kg for cobalt chloride, cobalt carbonate, cobalt sulfate, and tricobalt tetraoxide, respectively28,29,30,31,32. The acute LD50 values of cobalt chloride was 20 mg/kg (equivalent to 9.1 mg/kg cobalt ion) in rats after intravenous injection33. In our experiment, the concentration of cobalt ion in rats injected with 50 mg/kg of Co2+-LSB was equivalent to 3.8 mg/kg of cobalt ion, far below the reported harmful dosage. The reason why rats perished after intravenous injection with 50 mg/kg Co2+-LSB should be clarified in follow-up studies. Meanwhile, whether a lower dosage of Co2+-LSB can be used to develop a new substitute for cardiac glycosides in the treatment of congestive heart failure requires cautious evaluation.

Author contribution

Jason TC TZEN and Tzyy-Rong JINN designed research; Ying-Jie CHEN performed the animal experiments and HPLC analysis; Tse-Yu CHUNG prepared the metal-LSB complexes; Wen-Ying CHEN guided the animal experiments; Chung-Yu CHEN and Maw-Rong LEE performed the LS/MS/MS analysis; Jason TC TZEN and Ying-Jie CHEN wrote the paper.