Bisbenzylisoquinoline alkaloids are among the most important and widely distributed alkaloids in plant resources [1,2]. There are several types of bisbenzylisoquinoline alkaloids, including one, two, or three diphenyl ether linkages and other phenyl ether linkages [2]. Isoliensinine and its analogues liensinine and neferine are three tail-to-head phenolic benzyltetrahydroisoquinoline alkaloid dimers, found to be the major alkaloids in the lotus (Nelumbo nucifera GAERTNER., Nelumbonaceae), especially in its embryo loti “Lien Tze Hsin” (green embryo of mature seed), which is a famous Traditional Chinese Medicine (TCM), primarily used for nervous disorders, insomnia, high fevers with restlessness, and cardiovascular diseases [3,4,5], and also used as an antifebrile, sedative and hemostatic agent [6]. In recent years, these bisbenzylisoquinoline alkaloids were found to have wide biological activities such as antioxidant action, cardiovascular pharmacological effects such as antihypertensive and antiarrhythmic actions [7,8,9], reversing the multidrug resistance (MDR) effect of human carcinomas [10,11], anti-HIV activity [12], and as anti-tuberculosis agents toward multidrug-resistant Mycobacterium tuberculosis [13]. In addition, isoliensinine can inhibit bleomycin-induced pulmonary fibrosis in mice [14] and decrease the overexpression of growth factors PDGF-beta, bFGF, proto-oncogene c-fos, c-myc and hsp70 [15].

In spite of the growing interest in the biological properties of isoliensinine and its analogues, most studies are focused on their isolation and pharmacology, and few studies have investigated their pharmacokinetics [16,17,18] and metabolism [19,20]. In fact, determination and characterization of their metabolites and metabolic pathways are very important for drug discovery and development. Our previous study [21] revealed the fragmentation patterns of isoliensinine and its analogues, and also identified six metabolites of neferine. However, to our knowledge, to date there are no details on the metabolite of isoliensinine.

In this paper, we report three novel metabolites of isoliensinine obtained from dog hepatic microsomal incubations and identified using reversed-phase high performance liquid chromatography (RP-HPLC) and electrospray ionization tandem mass spectrometry (ESI-MS/MS). As a result, these three metabolites were found to be novel desmethyl or didesmethyl products of isoliensinine. The possible metabolic pathways for isoliensinine are proposed. To the best of our knowledge, this is first report of these microsomal metabolites and the metabolic pathways of isoliensinine. It should prove very helpful for evaluation of the drug-like properties of isoliensinine and other bisbenzylisoquinoline alkaloids.

The hepatic microsomes were first prepared according to a reported process [21], and then incubated in vitro with 100 μM isoliensinine. After transformation, an organic solvent was used to extract metabolites which were then analyzed by RP-HPLC. In order to find more bisbenzylisoquinoline metabolites, an on-line DAD detector was set at 200–400 nm to record the UV spectra. As shown in Figure 1, isoliensinine was found to be metabolized into three major metabolites, identified as M1, M2 and M3, and several minor metabolites. Here we focus on characterizing the structures of the three major metabolites.

The on-line DAD spectra (Figure 2) showed the three metabolites have very similar UV spectra to their parent compound isoliensinine and maximum absorbance at 215, 230, and 285 nm, implying that they possess the very similar molecular structures to their parent compounds.

A previous study [21] revealed that demethylation is the major metabolic path in vitro. In order to identify the metabolites, the data-dependent product MS/MS scans were performed for detection of the pre-listed compounds. As shown in Figure 1, metabolites M1 and M2 were found to have the same protonated molecular ion [M+H]⁺ at m/z 597, implying that they are produced by loss of a methyl group from isoliensinine, and metabolite M3 possess a [M+H]⁺ at m/z 583, corresponding to loss of two methyl groups.

Further ESI-MS/MS analyses (Figure 3) showed that these metabolites have typical bisbenzylisoquinoline characteristics. As reported [21], the ESI-MS/MS spectrum of isoliensinine clearly shows the major diagnostic fragment ions at m/z 192, 475, 489, 593, 580 and 568 resulting from the cleavage of the C1'-C9' bond resulting in positive groups CD (Scheme 1), and the loss of 4-ethyl-1-phenol or 4-ethyl-1-methoxybenzene following rearrangements [21,22,23]. In addition, H/D exchange ESI-MS/MS spectra [21] have showed the relative stable fragmentation ions formed by the elimination of H₂O, CH₃NH₂, CH₃OH, and CH₃–N=CH₂. Thus, a possible fragment pattern of isoliensinine as shown in Scheme 1 can be proposed. Clearly, the ions such as at m/z 611, 475 and 192 were key fragment ions to identify its metabolites from dog hepatic microsomes.

The ESI-MS² of the metabolite M1 (Figure 3b) displays characteristic fragmentation ions at m/z 597, 580, 579, 565, 192 and 177. The existence of the diagnostic fragment ion at m/z 580 corresponding to the loss of NH₃[24] indicated M1 is a N-desmethyl metabolite of isoliensinine. There are two possible positions to lose an N-methyl group, one is the 2-N-methyl group, and the other is the 2'-N-methyl group. The prominent fragments ions at m/z 192 and 177 indicated that metabolite M1 has the same CD group as its parent isoliensinine, without the loss of the 2'-N-methyl group. Therefore, the lost group is the 2-N-methyl group, which is further supported by the absence of characteristic ions at m/z 475 and 297, corresponding to the ABFCD or ABE (Scheme 1) moieties, respectively. Due to the loss of the 2-N-methyl group, the ABFCD (Scheme 1) moiety of the metabolite M1 showed a weak capacity to form positive ions at m/z 475. Based on these evidences, the metabolite M1 was identified as 2-N-desmethylisoliensinine.

Metabolite M2 is also an N-demethylation product of isoliensinine because its ESI-MS² spectrum (Figure 3c) showed the prominent diagnostic fragment ion at m/z 580 corresponding to the loss of NH₃ [24]. However, differing from the metabolite M1, the loss of the N-methyl group of M2 occurred on the 2'-N-methyl group of the CD moiety because of the existence of the characteristic fragment at m/z 178. Furthermore, the prominent characteristic fragments ions at m/z 420 and 297, as well as the ion at m/z 475, were supportive of the above elucidation. Its fragmentation patterns were illustrated in Scheme 2. Based on these data, metabolite M2 can be identified as 2'-N-desmethylisoliensinine.

Metabolite M3 showed the protonated ion [M+H]⁺ at m/z 583 (Figure 1), implying that it is probably produced by the loss of two methyl groups. Its ESI-MS/MS spectrum (Figure 3d) showed similar fragment profiles to metabolite M2, implying structural similarity. The diagnostic fragment ion at m/z 566 corresponding to the loss of NH₃, indicated that the metabolite M3 is an N-desmethyl product. The presence of a characteristic fragment at m/z 178 indicates that one lost group occurred on the 2'-N-methyl group of the CD moiety. Furthermore, another demethylation can determined at the position of the 6-O-methyl group of the AB moiety due to the presence of the intensive fragment ions at m/z 461 and 406. Based on these data, the metabolite M3 can be unambiguously identified as 2'-N-6-O-didesmethylisoliensinine.

The occurrence of these new metabolites provides evidential information about the biotransformation of bisbenzylisoquinoline alkaloids. A recent study [25] assumed that in N. nucifera isoliensinine may be biosynthesized from two molecules of (R)-N-methylcoclaurine via C–O oxidative coupling and subsequent O-methylation. In addition, isoliensinine may transform into neferine by further O-methylation [25]. Interestingly, the biotransformation in dog hepatic microsomes is a reversed metabolic process comprising biosynthesis of these bisbenzylisoquinoline alkaloids in N. nucifera. Our previous study indicated that neferine could be metabolized into isoliensinine [21]. The present study indicated that isoliensinine can be further metabolized into three major bisbenzyl-isoquinoline alkaloids M1, M2, M3. The possible metabolic pathway of isoliensinine was summarized in Scheme 3. Clearly, N-demethylation and O-demethylation are two important metabolic pathways of isoliensinine in dog hepatic microsomal incubations, as reported previously for neferine [21].

Demethylation is an important metabolic dealkylation process for compounds with alkyl groups [26]. In the rat hepatic S9 fraction in the presence of an NADPH-generating system, N-demethylation was one of the metabolic pathways of thalicarpine [24] and dauricine [27], which are tail-to-tail benzyltetrahydroisoquinoline alkaloid dimers. A recent study [20] has suggested that both CYP3A and CYP2B are involved in the metabolism of neferine in rat liver microsomes, and CYP3A probably plays a major role. Another study indicated CYP2D6 is also involved in the liver metabolization of neferine to liensinine, isoliensinine and other demethylated metabolites [19]. In the present study, three desmethyl metabolites of isoliensinine are also possibly produced by catalysis by the cytochrome P450 enzyme systems of dog hepatic microsomes.

This work identified three novel metabolites of isoliensinine obtained from dog hepatic microsomes. These metabolites could be formed by demethylation or bisdemethylation of the parent isoliensinine. They showed diagnostic MS fragments such as elimination of benzyl group E or F, and losses of H₂O, CH₃NH₂, CH₃OH, and CH₃–N=CH₂. To the best of our knowledge, this is first report of the microsomal metabolites and the metabolic pathway of isoliensinine. The liver is a very important organ for drug metabolism. Although the current work only focused on the in vitro microsomal metabolism of isoliensinine, the metabolism and the metabolites are very important for further pharmacokinetic and pharmacodynamic studies of isoliensinine and other bisbenzylisoquinoline alkaloids. It is also important for screening and developing new phenolic bisbenzyltetrahydroisoquinoline alkaloids for clinical therapy.