Metabolites of Siamenoside I and Their Distributions in Rats

Siamenoside I is the sweetest mogroside that has several kinds of bioactivities, and it is also a constituent of Siraitiae Fructus, a fruit and herb in China. Hitherto the metabolism of siamenoside I in human or animals remains unclear. To reveal its metabolic pathways, a high-performance liquid chromatography-electrospray ionization-ion trap-time of flight-multistage mass spectrometry (HPLC-ESI-IT-TOF-MSn) method was used to profile and identify its metabolites in rats. Altogether, 86 new metabolites were identified or tentatively identified, and 23 of them were also new metabolites of mogrosides. In rats, siamenoside I was found to undergo deglycosylation, hydroxylation, dehydrogenation, deoxygenation, isomerization, and glycosylation reactions. Among them, deoxygenation, pentahydroxylation, and didehydrogenation were novel metabolic reactions of mogrosides. The distributions of siamenoside I and its 86 metabolites in rat organs were firstly reported, and they were mainly distributed to intestine, stomach, kidney, and brain. The most widely distributed metabolite was mogroside IIIE. In addition, eight metabolites were bioactive according to literature. These findings would help to understand the metabolism and effective forms of siamenoside I and other mogrosides in vivo.


Introduction
Mogrosides are a group of cucurbitane-type triterpenoid saponins which have the common aglycone of mogrol [1]. They are responsible for the sweet taste and bioactivities of Siraitiae Fructus (Luo Han Guo in Chinese, the ripe fruits of Siraitia grosvenorii), a traditional Chinese medicine and an edible fruit [2].
Siamenoside I is one of the mogrosides, which is firstly isolated from Siraitia siamensis (a Chinese folk medicine) [3] and then from Siraitia grosvenorii [4]. Its relative sweetness (0.01% solution) to 5% sucrose is determined to be 563, higher than the famous sweetener mogroside V, making it the sweetest cucurbitane glycoside [4].
Besides its intense sweet taste, siamenoside I also has several kinds of bioactivities. It can inhibit the induction of Epstein-Barr virus early antigen (EBV-EA) by 12-O-tetradecanoylphorbol-13-acetate

Profiling the Metabolites of Siamenoside I in Different Biosamples by HPLC-ESI-IT-TOF-MS n
Based on the strategy described in Section 4.7, 86 new metabolites (M1-M86) of siamenoside I were detected altogether in different drug-containing samples by the HPLC-ESI-IT-TOF-MS n technique (Table 1, Table S1, Figure 1, and Figures S1-S25).
Eighty-three metabolites were detected in drug-containing feces; 19 metabolites were found in urine, and only two were detected in plasma.
The other 77 metabolites were tentatively identified by interpretation of their LC-MS n data and by comparison with literature.
These 86 new metabolites of siamenoside I can be classified into 24 classes according to their formative reactions and molecular formulae. , which indicated that their molecular formulae were C60H102O29. The formulae had an additional glucosyl (C6H10O5) than that of siamenoside I (C54H92O24). Hence, they were mogroside V isomers.

Metabolites Formed by Isomerization (M3-M5)
M3-M5 showed [M + HCOOH´H]´at m/z 1169.59, indicating the molecular formula of C 54 H 92 O 24 , which was the same to siamenoside I. Hence, they were mogroside IV isomers, and M3-M4 were further confirmed as mogroside IVA and mogroside IVE by comparison with reference compounds.

Metabolites Formed by Dehydrogenation (M6)
The molecular formula of M6 was predicted to be C 54 H 90 O 24 based on its [M + HCOOH´H]á t m/z 1167.5730, which was formed by loss of two hydrogen atoms from siamenoside I, thus M6 was tentatively identified as dehydrogenated siamenoside I.

Metabolites Formed by Deoxygenation (M7)
The molecular formula of M7 was C 54 H 92 O 23 calculated from its [M + HCOOH´H]´at m/z 1153.5957, which has one less oxygen atom than that of siamenoside I. Accordingly, it was tentatively identified as deoxygenated siamenoside I. In addition, the possible hydroxylation sites of M32 can be deduced by its MS 2 data and one possible structure of M32 is shown in Figure 2a. The nomenclature for the fragmentation pathways and fragment ions of cucurbitanes proposed by the authors [10] were used in this study.

M8-M12
In  4 ) generated by s,t D cleavage were observed, which indicated that one hydroxylation site was in c,j A, and the other was in c,j ABC s,t D (Figure 2a).      Further, the possible hydroxylation sites of M53 could be deduced by its MS 2 data, and one possible structure of M53 is shown in Figure 2b.
M53 Besides, the possible dehydrogenation and hydroxylation sites of M74 could be deduced by its MS 2 and MS 3 data, and one possible structure of M74 is shown in Figure 3a.
The characteristic product ions at m/z 479.2977 (C 27 H 43 O 7 , ABCD y E´) produced by y E cleavage, m/z 419.2762 (C 25 H 39 O 5 ) generated by a,c A cleavage, and m/z 195.1347 (C 12 H 19 O 2 ) generated by n,p C cleavage indicated that the C 24 -hydroxyl group of M74 was dehydrogenated, and one of the four tetrahydroxylation sites was at C 2 , one was in n,p CD y E, and the other two were in AB n,p C (Figure 3a).     In addition, the possible dehydrogenation and hydroxylation sites of M85 could be deduced by its MS 2 and MS 3 data, and one possible structure of M85 is shown in Figure 3b.

M78-M84
M85  3 spectra, which indicated that the last of the five hydroxylation sites was in l,p CD w E.

Discussion
The metabolism of siamenoside I in rats was firstly investigated in the present work. In total, 86 new metabolites of siamenoside I were detected in different biological samples from rats, and nine of them were unambiguously identified by comparison with reference compounds, and the others were tentatively identified by careful interpretation of their LC-MS n data.

The Metabolic Pathways of Siamenoside I in Rats
Based on the structures of the metabolites (M1-M86), the metabolic pathways of siamenoside I in rats are proposed and shown in Figure 1. From Figure 1, we can find that the metabolic reactions of siamenoside I include deglycosylation, hydroxylation, dehydrogenation, deoxygenation, isomerization, and glycosylation.
Most of the metabolic reactions of siamenoside I are the same to mogroside V [10]. However, there are also some differences. For example, methylation metabolites were not found in the metabolism of siamenoside I, and deoxygenation is found to be a novel metabolic reaction of mogrosides. Furthermore, pentahydroxylation, didehydrogenation are also found as novel metabolic reactions of mogrosides.
Among These results suggest that siamenoside I has its own metabolism characteristics in comparison with mogroside V.
We also could find the specific metabolites detected in different biosamples. M9 (mogroside IIIE) is detected in all biosamples except muscle, indicating that it is the most widely distributed metabolite.

The Proposed in Vivo Process of Siamenoside I in Rats
From Table 2, we could find that M1-M29 seem to be the main metabolites, and most of the other metabolites are only detected in feces. We think the reasons for this result may be: (1) the first pass effect of these metabolites may be very high, i.e., their hepatic extraction ratios are very high, which leads to their very low contents in plasma, organs, and urine samples. As a result, they are not easily detected in these samples; (2) these metabolites are mainly excreted into bile and then to feces, which make their contents in feces high and detectable.
Based on our research results and general metabolic knowledge, we hypothesize that the in vivo process of siamenoside I in rats may be as follows.
After oral administration, siamenoside I is degraded into its secondary glycosides (e.g., mogroside III, IIIE, IIIA1, IIE, IIA 2 , IA 1 , IE 1 , etc.) and its aglycone (mogrol) or dehydrogenated aglycone (dehydrogenated mogrol) by gastric juice, intestinal juice, intestinal enzymes, or intestinal microflora. Then, mogrol or dehydrogenated mogrol permeate across intestinal mucosa and enter the liver, where they undergo extensive oxidative metabolic reactions to form lots of hydroxylation and/or dehydrogenation metabolites. These polar oxidative metabolites may be largely excreted into the bile and then to the feces, and only a limited amount of them enter general circulation. Besides, some of them may also undergo hepatoenteral circulation and are absorbed into general circulation, and then distributed to different organs, and finally excreted into feces or/and urine.

Animal Experiments and Sample Collection
Eight rats were divided into two groups: two were blank group and the others were test group. Each rat was put into a clean metabolic cage (Suzhou Fengshi Laboratory Animal Equipment Co., Suzhou, Jiangsu, China) and given food and water ad libitum.
Because the research aim is to find the general/average differences between test group rats (drug-containing sample) and blank group rats (blank sample), the individual differences among the same group rats are not taken into consideration. Accordingly, all of the biosamples from each group were combined into one sample which was more representative than individual samples in the following sample collection processes.
The animal experiment lasted six days. The whole urine and feces of days 1-2 were collected as blank urine and feces samples, respectively. On days 3-5, the rats of test group were orally administrated with siamenoside I [50 mg/kg body weight, in normal saline (NS) solution] at 9:00, and all 72-h urine and feces were collected as drug-containing urine and feces samples, respectively. The rats of blank group were orally administrated with the same volume of NS. On day 6 at 9:00, the test and the blank group were treated with siamenoside I and NS again, respectively. After 1 h, blood sample was collected into a vacuum tube with sodium citrate as anticoagulant (Hebei Xinle Technology Co., Ltd., Shijiazhuang, Hebei, China) from rat heart under anesthesia. Then, the organs (heart, liver, spleen, lung, kidneys, stomach, small intestine, brain) and skeletal muscles of rats were collected and washed with NS, separately. All samples were kept at´80˝C before further pretreatment.

Blood Samples
The blood samples were centrifuged at 3000 rpm for 20 min at 4˝C, and the supernatant plasma samples were collected. Afterward, 8 mL methanol was added to 2 mL of plasma sample, and then was mixed and centrifuged at 9000 rpm for 30 min. The supernatant was collected and evaporated to dryness at 55˝C by a rotatory evaporator. The 15 mg residue was reconstituted in 1.00 mL methanol, filtrated through 0.45 µm filter membrane, and stored at 4˝C before analysis.

Urine Samples
Urine samples were filtered and then evaporated to dryness under vacuum at 55˝C by a rotatory evaporator. Subsequently, the residue was ultrasonically extracted with 10 mL methanol for 30 min, and the extract was then centrifuged at 9000 rpm for 30 min. The supernatant was transferred to another tube and evaporated to dryness at 55˝C by a rotatory evaporator. Next, a 100 mg residue was dissolved in 1.00 mL methanol, filtered through 0.45 µm membranes, and stored at 4˝C before analysis.

Feces Samples
Feces samples were dried at 55˝C and pulverized. Subsequently, the 10 g powder was ultrasonically extracted with 50 mL methanol for 30 min for three times, and the three supernatants were combined and evaporated to dryness. Next, the residue was mixed with 20 mL methanol and centrifuged at 9000 rpm for 30 min. The supernatant was evaporated to dryness again at 55˝C and the residue was collected. Then, 15 mg of residue was dissolved in 1.00 mL methanol, 0.45 µm membranes, and stored at 4˝C before analysis.

Organ and Skeletal Muscle Samples
Each organ was weighed, minced, and homogenized in 4 times volume (mL/g) of 4˝C NS by a homogenizer (Ultra-Turrax T8, Ika-Werke, Gmbh & Co. KG, Staufen, Germany). Then, 10 mL homogenate was mixed with 9 times volume (mL/mL) of acetonitrile, ultrasonically treated for 30 min, and centrifuged at 12,000 rpm for 30 min at 4˝C. Afterward, the supernatant was collected and evaporated to dryness at 50˝C. The residue was dissolved in 1.00 mL methanol, filtered through 0.45 µm filter membrane, and stored at 4˝C before analysis. The skeletal muscle samples were treated by the same method.

Strategy for Profiling and Identification of the Metabolites of Siamenoside I in Biosamples
The strategy [14] previously proposed by the authors was used to find and identify the metabolites of siamenoside I in the present study. In short, the base peak chromatograms (BPCs) of drug-containing samples and blank samples were thoroughly analyzed and compared. Meanwhile, the possible metabolites predicted by general metabolism rules were also screened and confirmed by comparing their corresponding extracted ion chromatograms (EICs). The metabolites were identified by comparison of their retention times and MS n data with those of reference compounds, or tentatively identified by interpretation of their MS n data.

Preliminary Evaluation of the Relative Contents of Siamenoside I and Its Metabolites in Biosamples
To preliminarily estimate the relative contents of siamenoside I and its metabolites in biosamples, the peak area of each metabolite calculated from its NI EIC was used. In the present study, only the data of different organ samples were comparable, since they were prepared and analyzed by the same method.

Conclusions
The metabolism of siamenoside I in rats was studied for the first time. In total, 86 new metabolites were detected. Nine of them were unambiguously identified by comparison with reference compounds, and the other 77 were tentatively identified by HPLC-ESI-IT-TOF-MS n technique. The metabolic pathways of siamenoside I in rats were proposed based on the structures of metabolites. The metabolic reactions of siamenoside I were found to be deglycosylation, hydroxylation, dehydrogenation, deoxygenation, isomerization, and glycosylation, among which deoxygenation, pentahydroxylation, and didehydrogenation were novel metabolic reactions of mogrosides. In addition, 23 metabolites were new metabolites of mogrosides. The distributions of siamenoside I and its 86 metabolites in rat organs were firstly reported, and they were mainly distributed to intestines, the stomach, kidneys, and the brain. Mogroside IIIE was the most widely distributed metabolite. Eight metabolites had bioactivities, indicating that they might contribute to the bioactivities of siamenoside I. These findings not only provide valuable information on the metabolism and disposition of siamenoside I and mogrosides in rats but also provide useful information on the chemical basis of the pharmacological effects of siamenoside I and mogrosides in vivo.