Isolation and Identification of Twelve Metabolites of Isocorynoxeine in Rat Urine and their Neuroprotective Activities in HT22 Cell Assay

Abstract

Introduction

UH is a well-known herbal medicine mentioned in Chinese and Japanese Pharmacopoeias, which has been used to treat hypertension, headache, and stroke. It is derived botanically from the branches with hooks of five plants of the genus Uncaria (Rubiaceae). RN, IRN, CN, and ICN [(16E,20α)-16,17,18,19-tetradehydro-17-methoxy-2-oxocorynoxan-16-carboxylic acid methyl ester] are the four major tetracyclic oxindole alkaloids found in UH [1] and reported to process many beneficial pharmacological effects on cardiovascular and central nervous system diseases, such as hypertension [2], vasodilatation [3], anti-platelet aggregation [4], sedation [5,6], and neuroprotection [7,8]. Consistently with the pharmacological activities mentioned above, it has been reported that ICN has a protective effect against glutamate-induced neuronal death in cultured cerebellar granule cells by inhibition of Ca²⁺ damage in the rat hippocampus [7], a suppressive effect on 5-HT_2A receptor function [9], and a suppressive activity against lipopolysaccharide-induced NO release in primary cultured rat cortical microglia [8]. It is implied that ICN plays an important role in the biological activities of UH, especially in neuroprotection. Clinically, in China and Japan, the phytochemicals containing UH, such as Tianmagouteng granules, Yokukansan, and Chotosan, have become best-selling herbal medicines for therapy of stroke, hypertension, chronic headache, and others. According to our previous study [1], the average content of ICN in some raw materials of UH was similar to that of IRN and more than RN and CN. Thus, the information on ICN metabolism is helpful for the clinical applications of the phytochemicals containing UH. Despite wide pharmacological studies of ICN, the metabolism and pharmacokinetics of ICN in human and other animals remain obscure. Huang et al. [10] investigated biotransformation in rats orally given ICN, and found four phase II metabolites in rat bile although they were not identified. Also the incubation of the substrates with rat liver microsomes shows that CYP450 plays a key role in their metabolic catalysis. ICN was found by Cai et al. [11] in the plasma of rats which received ICN orally. Wang et al. [12] found that CN was detected in bile, together with its four metabolites, namely, 10 or 11-hydroxy-CN and their 10 or 11-O-glucuronides, after oral administration of CN to rats. In other studies by Wang et al. [13,14], an LC-MS examination detected RN or IRN in plasma, bile, urine, and feces, 10 or 11-O-β-D-glucuronides of RN or IRN in bile, and 10 or 11-hydroxy-RN or IRN in urine and feces, respectively, after oral administration of RN or IRN. Yu et al. [15] proposed that the metabolite of IRN in cat plasma was formed from 16-double-bond reduction after intravenous administration of IRN. The results mentioned above indicate that the metabolism of ICN may be much more complicated than that of RN and IRN due to its vinyl group at C-20.

The liquid chromatography-tandem mass spectrometry (LC-MSⁿ) technique was applied to characterize the urinary, plasma, and/ or biliary metabolites of ICN, IRN, and RN, the major alkaloids of UH [10,13,14]. In these studies, the glucuronide-conjugated position of phase II metabolites and the hydroxylation position of phase I metabolites could not be definitively determined by means of the LC-MSⁿ method alone. Isolation of metabolites and their further structural confirmation on the basis of UV, NMR and MS data are valuable as well. Furthermore, we pay attention to the activity of the metabolites because some phase I metabolites may exert more potent biological activities or side effects than their parent drug [16,17].

The present study describes the isolation and structural elucidation of seven new and six known compounds in urine of rats orally given ICN, and the neuroprotection of ICN and its major metabolites against glutamate-induced HT22 cell death. Moreover, a possible metabolic pathway of ICN was proposed on the basis of the metabolite profile.

Results and Discussion

Representative UHPLC profiles showing the rat urinary metabolites are given in Fig. 1. ICN and its metabolites were selectively detected at 245 nm due to their characteristic oxindole nucleus group. Twelve major metabolites together with the parent drug were clearly observed in rat urine compared to the blank control. By means of repeated chromatographic methods on columns of macroporous resin AB-8, Sephadex LH-20, reverse-phase ODS, and preparative HPLC, they were isolated from the rat urine sample, including seven new compounds, M-2, M-3, M-4, M-7, M-10, M-11, and M-12, together with six known alkaloids M-0, M-1, M-5, M-6, M-8, and M-9. Among the seven new compounds, M-10 (20.0 mg), M-3 (18.0 mg), and M-7 (15.0 mg) were three major metabolites, and M-11 (2.7 mg), M-12 (2.5 mg), M-4 (2.4 mg), and M-2 (2.1 mg) four minor ones.

Open in a separate window

Fig. 1

UHPLC-UV chromatograms of blank urine (A), a rat urine sample 0 to 24 h after oral administration of 40 mg/kg of ICN (B), and metabolite standards (C).

Metabolite M-10 was isolated as a white amorphous powder. The molecular formula was determined to be C₂₁H₂₂ N₂O₅ from the [M + H]⁺ quasi molecular ion peak at m/z 383.1607 (calcd. 383.1604) in the ESI-QTOF-MS. This implies that it has 12 degrees of unsaturation. The molecular weight of M-10 was the same as that of ICN (C₂₂H₂₆ N₂O₄), but its molecular composition was different from that of ICN. ¹H-NMR (400 MHz, DMSO-d₆) spectra gave a group of o-substituted benzene ring proton signals (δ_H 7.22, 1H, o, δ_H 7.16, 1H, d, J = 7.0 Hz, δ_H 7.01, 1H, t, J = 7.2 Hz and δ_H 6.88, 1H, d, J = 7.6 Hz), which were in connection with four aromatic carbon signals (δ_C 128.5, 124.2, 121.6, and 109.9) in the HSQC spectra. ¹³C-NMR (100 MHz, DMSO-d₆) spectra gave a lactam carbonyl (δ_C 178.8), two aromatic quaternary carbons (δ_C 130.1 and δ_C 141.7), and a quaternary carbon (δ_C 50.7) signals. These data indicate that M-10 possesses an oxindole structure. Moreover, the resonances for the carbons and protons in the ¹H and ¹³C NMR as well as HSQC spectra included three methylene (δ_H 0.92, 1.75, δ_C 30.7; δ_H 2.47, 2.72, δ_C 40.7; and δ_H 2.63, 3.91, δ_C 44.6), two methines (δ_H 2.63, δ_C 36.5; δ_H 2.63, δ_C 40.8), a methane connected to N or O atom (δ_H 3.68, δ_C 61.7), a vinyl group (δ_H 4.99, 1H, d, J = 9.6 Hz, δ_H 5.02, 1H, d, J = 13.4 Hz, δ_C 116.3; δ_H 5.49, 1H, m, δ_C 138.2), a methoxy (δ_H 3.72, 3H, s, δ_C 61.3), an sp² methine (δ_H 7.23, 1H, s, δ_C 159.8) that should be directly attached to an oxygen, an sp² quaternary carbon (δ_C 110.5), and two carbonyls (δ_C 170.2 and 167.8). HMBC spectra gave some correlations that were determined between H-6 and C-2, C-3, C-5, C-7, and C-8, between H-3 and C-2, C-7, and C-8, between H-21 and C-3, C-15, and C-20 and between H-15 and C-20, respectively. The NMR data of M-10 (Table 1 and 2) suggested that its tetracyclic framework was similar to that of ICN except that C-5 of M-10 was a carbonyl (δ_C 170.2). Moreover, the remaining molecular formula C₆H₈O₃ consisted structurally of two chain substituents, accounting for three degrees of unsaturation and being attached to the cyclic framework separately at C-15 and C-20. The constitution of one of the substituents, C₄H₅O₃, and its location at C-15 were deduced from the HMBC spectrum. HMBC correlations of the substituents were determined between the methoxy protons and C-17 and between H-17 and C-15, C-16, C-22, and the methoxy carbon, respectively. Subsequently, the remaining substituent, a vinyl group (-CH=CH₂), could be attached only to C-20, as evidenced by the appreciable correlation between H-18 and C-20 in the HMBC experiment of M-10. The absolute configuration of M-10 was determined from the CD spectrum and 2D NOESY correlation. As a tetracyclic oxindole alkaloid, its 7S configuration was established according to a negative CE at 285 nm and a positive CE at 220 nm in the CD spectrum [18,19]. Meanwhile, a negative CE at 265 nm indicated that H-3 had an α-orientation [19]. The α-orientation of H-15 was indicated by the strong NOESY correlation of H-3/H-15. The β-orientation of H-20 was indicated by the strong NOESY correlations of H-20/β-H-14 (δ_H 1.75). Thus, M-10 possesses 7S, 3S, 15S, 20R absolute configurations, which are consistent with the absolute configuration (allo A configuration) for ICN [19]. The geometry of the trans C-16–C-17 double bond was confirmed on the basis of the olefinic proton in a downfield shift relative to the corresponding signals of the cis compounds [20]. M-10 was determined to be 5-oxoisocorynoxeinic acid.

Metabolite M-3 was isolated as a white amorphous powder. The molecular formula was determined to be C₂₇H₃₀O₁₁ from the [M + H]⁺ quasi molecular ion peak at m/z 559.1928 (calcd. 559.1925) in the ESI-QTOF-MS. The [M + H]⁺ ion at m/z 559 and an important fragment ion at m/z 383 originated from the elimination of 176 mass units (glucuronic acid) from [M + H]⁺ ion indicated that M-3 should be a glucuronide conjugate. Its NMR data (Table 1 and 2) suggested that its tetracyclic framework was the same as that of M-10. The constitution and location of both substituents C₄H₄O₃ and C₂H₃ were also deduced from the HMBC spectrum. The signals of an anomeric proton and carbon (δ_H 5.19, 1H, d, J = 7.5 Hz, δ_C 93.8) and a carboxylic group at C-6′ at δ_C 172.0 indicated the presence of a β-D-glucuronic acid moiety [21]. The correlation between an anomeric proton and C-22 in the HMBC experiment of M-3 indicated that a β-D-glucuronic acid moiety could be attached to the C-22 position, as evidenced by the upfield shift of C-22 (−2.5 ppm) relative to the corresponding signals of M-10. The full assignments of proton and carbon signals are summarized in Table 1 and ​and2.2. For M-3, the 7S configuration and α-orientation of H-3 were deduced on the basis of the presence of a negative CE at 285 nm, a positive CE at 220 nm, and a negative CE at 265 nm in the CD data, respectively [18,19]. The α-orientation of H-15 was indicated by the strong NOESY correlation of H-3/H-15. The β-orientation of H-20 was indicated by the strong NOESY correlations of H-20/β-H-14. Thus, M-3 possesses 7S, 3S, 15S, 20R absolute configurations, which are consistent with the allo A configuration for ICN [19]. M-3 was determined to be 5-oxoisocorynoxeinic acid-22-O-β-D-glucuronide.

Metabolite M-7 was isolated as a white amorphous powder. The molecular formula was determined to be C₂₁H₂₄ N₂O₅ from the [M + H]⁺ quasi molecular ion peak at m/z 385.1763 (calcd. 385.1765) in the ESI-QTOF-MS. Its NMR data (Table 1 and ​and2)2) suggested that its tetracyclic framework was the same as that of M-10. The constitution and location of both substituents C₄H₇O₃ and C₂H₃ were also deduced from the HMBC spectrum. For M-7, the 7S configuration and α-orientation of H-3 were deduced on the basis of the presence of a negative CE at 285 nm, a positive CE at 220 nm, and a negative CE at 265 nm in the CD data, respectively [18,19]. The α-orientation of H-15 and H-16 were indicated by the strong NOESY correlations of H-3/H-15 and α-H-14/ H-16. The β-orientation of H-20 was indicated by the strong NO-ESY correlations of H-20/β-H-14. Thus, M-7 possesses 7S, 3S, 15S, 16S, 20R absolute configurations, which are consistent with the allo A configuration for ICN [19]. M-7 was determined to be 17-O-demethyl-16,17-dihydro-5-oxoisocorynoxeine.

Metabolite M-11 was isolated as a white amorphous powder. The molecular formula was determined to be C₂₂H₂₄N₂O₆ from the [M + H]⁺ quasi molecular ion peak at m/z 413.1713 (calcd. 413.1714) in the ESI-QTOF-MS. Its NMR data (Table 1 and ​and2)2) suggested that its tetracyclic framework was similar to that of M-10 except that C-21 of M-11 was a methane connected to N or O atom (δ_H 5.35, 1H, t, J = 3.5 Hz, δ_C 73.7). According to the molecular formula, this implied that a hydroxyl group was attached to C-21 of M-11. The constitution and location of both substituents C₅H₇O₃ and C₂H₃ were also deduced from the HMBC spectrum. For M-11, the 7S configuration and α-orientation of H-3 were deduced on the basis of the presence of a negative CE at 285 nm, a positive CE at 220 nm, and a negative CE at 265 nm in the CD data, respectively [18,19], as evidenced by the substantial NOE correlation of H-9/β-H-14. The α-orientation of H-15 was indicated by the strong NOESY correlations of H-3/α-H-14 and H-15/α-H-14. The β-orientation of H-20 was indicated by the strong NOESY correlations of H-20/β-H-14. The β-orientation of H-21 was confirmed on the basis of the correlations of H-21/β-H-20 in the NOE data. Thus, metabolite M-11 possesses 7S, 3S, 15S, 20R, 21R absolute configurations, which are consistent with the allo A configuration for ICN [19]. M-11 was determined to be 21-hydroxy-5-oxoisocorynoxeine.

Metabolite M-12 was isolated as a white amorphous powder. The molecular formula was determined to be C₂₂H₂₄ N₂O₆ from the [M + H]⁺ quasi molecular ion peak at m/z 413.1713 (calcd. 413.1716) in the ESI-QTOF-MS. Its NMR data (Table 1 and ​and2)2) suggested that its tetracyclic framework was the same as that of M-10. The resonances for the carbons and protons of both substituents C₅H₇O₃ and C₂H₃O in the ¹H and ¹³C NMR as well as HSQC spectra included an epoxide group (δ_H 2.57, 1H, dd, J = 5, 1.5 Hz, δ_C 52.3; δ_H 2.63, 1H, dd, J = 5, 1.5 Hz, δ_H 2.57, 1H, dd, J = 5, 1.5 Hz, δ_C 45.9), a carboxylic carbonyl (δ_C 167.0), two methoxyl groups (δ_H 3.75, 3H, s, δ_C 61.6; δ_H 3.59, 3H, s, δ_C 50.7), and an sp² methine (δ_H 7.30, s, δ_C 160.6) that should be directly attached to the oxygen. The constitution of one of the substituents, C₅H₇O₃, and its location at C-15 were deduced from the HMBC spectrum. The HMBC correlations of the substituents were determined between the one methoxyl proton (δ_H 3.75) and C-17 and between the other methoxyl proton (δ_H 3.59) and C-22 and between H-17 and C-15, C-16, and C-22, respectively. Subsequently, the remaining substituent, an epoxide group, C₂H₃O, could be attached only to C-20, as evidenced by the appreciable correlation between H-18 and C-20 in the HMBC experiment of M-12. For M-12, the 7S configuration and α-orientation of H-3 were deduced on the basis of the presence of a negative CE at 285 nm, a positive CE at 220 nm, and a negative CE at 265 nm in the CD data, respectively [18,19], as evidenced by the substantial NOE correlation of H-9/β-H-14. The α-orientation of H-15 was indicated by the strong NOESY correlations of H-3/α-H-14 and H-15/α-H-14. The β-orientation of H-20 was indicated by the strong NOESY correlations of H-20/β-H-14. Thus, metabolite M-12 possesses 7S, 3S, 15S, 20R absolute configurations, which are consistent with the allo A configuration for ICN [19]. M-12 was determined to be oxireno [18,19]-5-oxoisocorynoxeine.

Metabolite M-4 was isolated as a white amorphous powder. The molecular formula was determined to be C₂₂H₂₆ N₂O₅ from the [M + H]⁺ quasi molecular ion peak at m/z 399.1920 (calcd. 399.1922) in the ESI-QTOF-MS which was 16 mass units higher than that of ICN. In the ¹H NMR spectrum, the skeleton proton signals (Table 1) of its parent drug remained except for those of the substituent group on the benzene ring. The linked positions of one hydroxyl group were established by the HMBC spectrum: a significant correlation between a proton signal at δ_H 6.79 (1H, d, J_9,11 = 2.5 Hz, H-9) and a carbon signal at δ_C 152.4 (C-10), or at δ_C 56.4 (C-7), indicated a hydroxyl group attached at C-10 in the aromatic ring. For M-4, the 7S configuration and α-orientation of H-3 were deduced on the basis of the presence of a negative CE at 285 nm, a positive CE at 220 nm, and a negative CE at 265 nm in the CD data, respectively [18,19], as evidenced by the substantial NOE correlation of H-9/β-H-14. The α-orientation of H-15 was indicated by the strong NOESY correlations of H-3/αH-14 and H-15/α-H-14. The β-orientation of H-20 was indicated by the strong NOESY correlations of H-20/β-H-14. Thus, metabolite M-4 possesses 7S, 3S, 15S, 20R absolute configurations. M-4 was identified as 10-hydroxyisocorynoxeine.

Metabolite M-2 was isolated as a white amorphous powder. The molecular formula was determined to be C₂₂H₂₆ N₂O₅ from the [M + H]⁺ quasi molecular ion peak at m/z 399.1920 (calcd. 399.1919) in the ESI-QTOF-MS. This implied that M-2 might be an isomer of M-4, which was further confirmed by NMR data (Table 1). A comparison of the ¹H NMR spectrum of M-2 with that of M-4 showed that their structures were closely related, except for the position of the hydroxyl group on the benzene ring. Then, according to protons signal at δ_H 7.22 (1H, d, J_9,10 = 8.0 Hz, H-9), 6.55 (1H, dd, J_9,10 = 8.0 Hz, J_10,12 = 1.5 Hz, H-10), and 6.69 (1H, d, J_10,12 = 1.5 Hz, H-12), we concluded that a hydroxyl group was attached to C-11 of M-2. For M-2, the absolute configurations of the asymmetric centers at C-3 and C-7 were assigned as S and S, respectively, by comparing its CD spectrum with those of IC and M-4. M-2 was identified as 11-hydroxyisocorynoxeine.

The metabolites M-0, M-1, M-5, M-6, M-8 and M-9 were identified as isocorynoxeine, 18,19-dehydrocorynoxinic acid, 18,19-dehydrocorynoxinic acid B, corynoxeine, isocorynoxeine-N-oxide, and corynoxeine-N-oxide, respectively, by comparing their UV, MS, and t_R data with those of standard compounds. All of them were first isolated as metabolites of UH from rat urine.

The metabolic pathways of indole alkaloids in vivo are related to many reactions such as hydrolysis, O-demethylation [22], glucuronidation [13,14], sulfation [23,24], hydroxylation [25–27], reduction [15], and others [28]. In the present study, eleven phase I metabolites (M-1, M-2, and M-4~M-12), one phase II metabolite (M-3), and ICN (M-0) itself were isolated from the urine of rats given ICN orally and structurally confirmed on the basis of UV, NMR, MS, and CD data. Also the neuroprotection of ICN and its metabolites, M-3, M-6, M-7, and M-10, against 3 mM glutamate-induced HT22 cell death was investigated. According to the metabolite profile, the possible metabolic pathways of ICN in rats are proposed as shown in Fig. 2.

Open in a separate window

Fig. 2

Proposed metabolic pathways of isocorynoxeine.

M-3, M-7, and M-10 are three major metabolites. M-10 is believed to be formed through the oxidation at C-5 and hydrolysis of 22-carboxylic methyl ester catalyzed by hepatic cytochrome P450 enzyme [22,29–33]. Moreover, M-10 is also a reactive metabolite that can be successively converted to a glucuronide-conjugate (M-3) via further glucuronidation at 22-carboxylic group catalyzed by UDP-glucuronosyltransferases [22,34]. Most importantly, we found that M-7 is another key metabolite derived from the oxidation at C-5 and reduction of the 16-double bond together with the O-demethylation at C-17 catalyzed by cytochrome P450 [15,24,35]. Wang et al. [13,14] reported that the hydroxylation at C-10 or C-11 of the aryl group was the major metabolic pathway after oral administration of IRN or RN dissolved in DMSO at a single dose of 37.5 mg/kg b.wt. to Wistar rats. In the present study, hydroxylated at C-10 or C-11, derivatives of ICN were only found as minor metabolites in rat urine after oral administration of ICN dissolved in 0.1 M HCl solvent at a single dose of 40.0 mg/kg b.wt. to male Wistar rats. The oxidation at C-5 is the major metabolic pathway for ICN, which is similar to the in vivo metabolism of nicotine [30,33] and (−)-N-{2-[(R)-3-(6,7-di-methoxy-1,2,3,4-tetrahydroisoquinoline-2-carbonyl)piperidino] ethyl}-4-fluorobenzamide [32]. We conducted a comparative experiment in order to explain the difference in major metabolites. The comparative assay showed that peak areas of 10 or 11-hy-droxy-ICN in urine samples of rats orally given ICN dissolved in 0.1 M HCl solvent were lower than those in DMSO, while peak areas of 5-oxo-ICN in urine samples of rats orally given ICN dissolved in 0.1 M HCl solvent were higher than those in DMSO, respectively. It is worth noting that the oral administration of the alkaloid hydrochloride salt to rats is closer to clinical application. Firstly, the alkaloid hydrochloride salt could be formed in the acidic environment of the gastrointestinal tract after oral administration of the alkaloid. Secondly, phytochemicals, such as tablets of Uncaria rhynchophylla total alkaloids, containing chloride salts of rhynchophyllinoid alkaloids have been used for several decades in clinical situations to treat hypertension, headache, and stroke in China. Thus, the oxidation at C-5 should be the major metabolic pathway for tetracyclic oxindole alkaloids, which is consistent with our last study of the metabolic fate of IRN in rats orally given IRN [36].

In addition, we also isolated some minor metabolites except for the parent compound ICN (M-0), all of which were phase I metabolites (M-1, M-2, M-4~M-6, M-8, M-9, M-11, and M-12). It is worth noting that the purity of the dosed ICN was found to be 98.6%, and that these minor metabolites were not impurities of the dosed ICN sample according to its UHPLC analysis. Formation of nine minor phase I metabolites can be explained by three phase I metabolic pathways, isomerization, oxidation and hydrolysis. Some of metabolic pathways mentioned above were firstly reported in the metabolism of tetracyclic oxindole alkaloids found in UH in rats after oral administration. The first is N-oxidation at N-4. M-8 and M-9 were two nitrogen oxides that are believed to be formed through the N-oxidation of N-4 under catalytic action of FMO₃ enzyme [33]. The second is epoxidation. M-12 is a minor phase I metabolite derived from the epoxidation of the vinyl group at C-20 catalyzed by cytochrome P450 [37]. The last is the isomerization. Two pairs of epimers (M-0 and M-6, M-1 and M-5) were obtained. M-6 was the epimer of M-0 that is believed to be formed through the isomerization at C-7 in rats after the oral administration of ICN. IRN is reported to undergo isomerization at the spiro center to form RN in rat urine after oral administration of IRN by our last study [36]. In addition, the isomerization at the spiro centre of rhynchophylline-type oxindole alkaloids in vitro was recognized by Wenkert et al. as early as 1959 and a retro-Mannich ring opening, rotation, and Mannich ring closure was proposed as the mechanism involved. The isomerization at the spiro centre in vivo may occur without enzymatic conversion [38]. M-6 is also a reactive metabolite that can be successively converted to a hydrolyzed metabolite (M-5) via further hydrolysis of 22-carboxylic methyl ester. M-1 was the epimer of M-5 that is believed to be formed through the isomerization at C-20. There is, however, no definitive evidence for isomerization at C-20 of oxindole alkaloids in vitro. Thus, we postulated that isomerization at C-20 may be accomplished by suitable enzymatic processes in vivo.

Structural elucidation of metabolites is one of the most challenging tasks in drug metabolism studies. In recent years, comparisons of ESI-MSⁿ data and HPLC retention times with synthetic standards have been commonly used to identify the structures of metabolites. However, when the standards are difficult to synthesize, some metabolite structures deduced only from LC/MSⁿ data may not be correct, especially if isomeric metabolites are present. In our study, four pairs of isomers (M-0 and M-6, M-1 and M-5, M-2 and M-4, and M-8 and M-9) were obtained, which had identical MS data in the LC-MSⁿ determination. Therefore, their exact structures could not be identified using LC/MSⁿ data alone. In these cases, preparation of metabolites and further identification on the basis of NMR data are needed. Of course, direct isolation of the metabolites from urine, bile, or feces of humans or animals is difficult, but it is the most reliable method for the identification of metabolites. We have determined the definitive structures of 12 metabolites and the parent drug by examination of MS, CD, and NMR spectra. These results are important for better understanding of its in vivo metabolic fate and disposition in rats.

Glutamate is a mammalian central nervous system neurotransmitter in normal neuronal cells. However, high concentrations of glutamate produce oxidative stress, resulting in neurodegeneration such as in stroke, trauma, Alzheimer’s disease, and Parkinson’s disease [39–41]. HT22 cells, an immortalized mouse hippocampal cell line, have been widely used in recent years as an in vitro model for studying the mechanism of glutamate-induced neurotoxicity [42,43]. The neuroprotective effects of some metabolites of ICN, together with ICN, against 3 mM glutamate-induced HT22 cell death were investigated. Cell viability was determined in HT22 cells by the calcein AM cell viability assay [42]. As shown in Fig. 3, ICN exhibited a significant neuroprotective effect against glutamate-induced cell death at the maximum concentration (100 μM). In a parallel experiment, ZYC-26, an estrogen-like compound, produced a significant neuroprotective effect against glutamate-induced cell death [44]. However, little or weak neuroprotective activities were observed at the maximum concentration (100 μM) for M-3, M-6 (CN), M-7, and M-10. The C-7 configuration of ICN and CN seemed to affect their neuroprotective activities, which is consistent with the report that antagonistic activity of alkaloids from UH on 5-HT_2A receptors in the brain is closely related to the 7S configuration of the oxindole moiety. It was apparent that ICN is initially administered to the rat in an active form and then becomes converted to its inactive forms excreted in urine through some in vivo biotransformations such as the oxidation at C-5, the reduction of the 16-double bond, O-demethylation at C-17, hydrolysis at 22-carboxylic methyl ester, and/or the isomerization at C-7. It is usual for the alkaloid to be eventually eliminated in urine in its inactive (or less than fully active) form in vivo.

Open in a separate window

Fig. 3

Effects of isocorynoxeine and some metabolites against 3 mM glutamate-induced HT22 cell death. HT22 cells were treated simultaneously with 3 mM glutamate. Cell viability was determined by calcein AM assay after 24 h exposure to the various samples. All data were normalized to percentage survival of control. Data are represented as mean ± SEM for n = 8; * p < 0.05, ** p < 0.01, and *** p < 0.001 versus glutamate-treated group alone. CN: corynoxeine; G: glutamate; ICN: isocorynoxeine; ZYC-26, 2-(1-adamantyl)-4-methylestrone.

In summary, seven new together with six known compounds were isolated from the urine of rats given ICN orally and structurally confirmed by UV, NMR, MS, and CD techniques. It was evidenced for the first time that the major metabolic pathways for ICN should be the oxidation at C-5 and hydrolysis at the 22-carboxylic methyl ester. The formed metabolites were excreted either free or conjugated.

Materials and Methods

Supplementary Material

Acknowledgments

Abbreviations

Footnotes

References