Rapid Characterization and Identification of Non-Diterpenoid Constituents in Tinospora sinensis by HPLC-LTQ-Orbitrap MSn

Tinospora sinensis, a kind of Chinese folk medicine, has functions of harmonizing qi and blood, dredging the channels and collaterals, calming and soothing the nerves. In the present study, a method based on high-performance liquid chromatography coupled with linear ion trap-Orbitrap mass spectrometry (HPLC-LTQ-Orbitrap) was developed for the systematical characterization of the non-diterpenoid constituents which possessed remarkable biological activities in T. sinensis, like anti-tumor, anti-inflammatory, hypoglycemic activity and immunomodulatory activity. Based on the accurate mass measurement (<5 ppm), retention times and MS fragmentation ions, 60 non-diterpenoid constituents were unambiguously or tentatively characterized from T. sinensis extract, including 27 alkaloids, 23 phenylpropanoids, seven sesquiterpenoids and three other constituents. Among them, 13 compounds were tentatively identified as new compounds. Finally, three of the non-diterpenoid constituents were purified and identified, which further confirmed the validity of the results. This study demonstrated that the HPLC-LTQ-Orbitrap MSn platform was a useful and efficient analytical tool to screen and identify constituents in natural medicine.


Introduction
Tinospora sinensis is derived from the dried stems of Tinospora sinensis (Lour.) Merr. (family Menispermaceae), which was officially documented in the Chinese Pharmacopoeia (2015) with the name Kuanjinteng [1]. As a kind of folk medicine which is mainly distributed in China, India and Southeast Asia, T. sinensis is commonly employed to treat various diseases. For instance, Tibetan medicine thought that T. sinensis could be used to treat rheumatoid arthritis, while Indian Ayurveda usually employed this ethnic drug in treatment for diabetes [2]. Pharmacological studies and clinical practice have demonstrated that the extracts of T. sinensis possessed various biological activities, including anti-inflammatory, anti-oxidative, anti-radiant, insecticidal and immunosuppressive effects [3][4][5][6].
T. sinensis has complicated chemical composition, including diterpenoids, alkaloids, phenylpropanoids, sesquiterpenes, triterpenoids, sterols, amino acids, and so on. Among them, diterpenoids are considered as the most abundant constituents. In previous work, we have systematic reported of diterpenoids in T. sinensis. A total of 63 diterpenoids were preliminarily identified, including 10 diterpenoid aglycones and 53 diterpenoid glycosides [7]. However, some non-diterpenoid constituents show good pharmacological activities which are probably closely related to its traditional efficacy. For example, two carboxylic acid esters isolated from T. sinensis showed significant PTP1B inhibitory activity in vitro, which represented a novel strategy for the treatment of type II diabetes [8]. Diosgenin isolated from T. sinensis showed good anti-inflammatory activity in carrageenan induced inflammation (paw edema) rodent model [9]. Three isoquinoline alkaloids isolated from T. cordifolia stem, named jatrorrhizine, palmatine and magnoflorine, were proved to possess the potential inhibitory effect on α-glucosidase in vitro and in vivo [10]. Tans-syringin, a typical phenylpropanoid glycoside which was abundant in T. sinensis, showed various activities of anti-tumor, anti-inflammatory, and hypoglycemic activities [11][12][13]. In addition, two sesquiterpenes of tinocordiside and 11-hydroxymustakone isolated from Tinospora cordifolia showed significant immunomodulatory activity [14]. Therefore, in order to comprehensively expound the material foundation of efficacy of T. sinensis, we propose a strategy to screen and identify the non-diterpenoid constituents in this herb.
As a powerful tool with the high resolution and excellent sensitivity to analyze multi-constituents in complex matrices, HPLC-ESI-MS n has been employed for characterization of phytochemical compounds in many areas of food and biological analyses [15,16]. The hybrid linear ion trap-Orbitrap mass spectrometer (LTQ-Orbitrap) was characterized by the higher mass resolution and mass accuracy (within 5 ppm) of orbitrap, MS n scanning function, and high trapping capacity of the linear ion trap [17,18]. Orbitrap allows the potent detection of a great deal of chemical constituents of similar accurate mass with high confidence of compounds identification, especially combined with retention times and the use of mass spectral libraries constructed with authentic standards [19]. These advantages facilitate to rapidly identify and characterize of multiple constituents in TCMs. In general, compounds with the same carbon skeleton usually have similar fragmentation pathway and characteristic product ions in collision-induced dissociation (CID) mode, so that this method also could be employed to identify novel constituents in TCMs [20]. In this study, a method with HPLC-LTQ-Orbitrap was established to comprehensively analyze the non-diterpenoid constituents in T. sinensis.

Identification of the Constituents by HPLC-LTQ-Orbitrap MS n
In order to get adequate structural information of the chemical constituents in T. Sinensis and reveal as many chemical compounds as possible, both positive and negative modes were employed for the comprehensive analysis. For the available standard compounds, these compounds were identified by comparing retention time (t R ) and/or accurate mass. For the standard unavailable compounds, the molecular formula of which were confirmed by compared with the HRMS molecular formula database built in-home, the high-accuracy protonated precursors with an error less than 5 ppm and related literatures. In the present study, a total of 60 compounds (Table 1, Figure 1) were identified or tentatively identified from T. Sinensis extract, including 27 alkaloids, 23 phenylpropanoids, seven sesquiterpenoids and three others. A typical total ion chromatogram (TIC) of T. sinensis in positive and negative ion mode is presented in Figure 2.   suggested the presence of C5-C6 carbon-carbon single bonds, which might due to a more stable π-conjugated system was formed by the losses of two hydrogen. By comparison with reference standard, compound 33 was predicatively deduced as berberine. presence of C9-C10 methoxyl groups. Therefore, combined with bibliography data and fragmentation pathways, these four compounds were tentatively identified as 3-hydroxy-2,9,11-tri-methoxy-5,6-dihydroisoquino[3,2-α]isoquinolinylium, palmaturbine, jatrorrhizine and columbamine, respectively [5,21].  2H] + ion at m/z 334 suggested the presence of C5-C6 carbon-carbon single bonds, which might due to a more stable π-conjugated system was formed by the losses of two hydrogen. By comparison with reference standard, compound 33 was predicatively deduced as berberine. presence of C9-C10 methoxyl groups. Therefore, combined with bibliography data and fragmentation pathways, these four compounds were tentatively identified as 3-hydroxy-2,9,11-tri-methoxy-5,6-dihydroisoquino[3,2-α]isoquinolinylium, palmaturbine, jatrorrhizine and columbamine, respectively [5,21].  [23]. The ion of compound 13 at m/z 340 was generated by loss a CH4 from the quasi-molecular ion. The ions at m/z 192 and m/z 165 were produced by Retro-Diels-Alder (RDA) cleavage fragmentation at 8, 13-position of the C-ring. Moreover, the product ion at m/z 192 generated the minor ion at m/z 190 and m/z 177 by the loss of two hydrogen ions and a methyl group, respectively. Therefore, according to the literature data, compounds 9 and 13 were tentatively deduced as menisperine and tetrahydropamatine [24].  NH] + , which was the characteristic fragment ion of aporphine-type alkaloids [23]. The ion of compound 13 at m/z 340 was generated by loss a CH 4 from the quasi-molecular ion. The ions at m/z 192 and m/z 165 were produced by Retro-Diels-Alder (RDA) cleavage fragmentation at 8, 13-position of the C-ring. Moreover, the product ion at m/z 192 generated the minor ion at m/z 190 and m/z 177 by the loss of two hydrogen ions and a methyl group, respectively. Therefore, according to the literature data, compounds 9 and 13 were tentatively deduced as menisperine and tetrahydropamatine [24].  [23]. The ion at m/z 192 [M + H − 181] + was produced by RDA cleavage fragmentation at 8,13-position of the C-ring. The product ion at m/z 192 generated the minor ion at m/z 177 by the loss of a methyl group. Moreover, the ions at m/z 208 and m/z 165 were produced by α-cleavage fragmentation at B-ring. Therefore, compound 2 was tentatively presumed to be 13-hydroxy-2,3,9,10-tetramethoxy-5,8,13,13a-tetrahydro-6H-isoquino[3,2-α]isoquinolinium.  [23]. However, compound 4 could generated the ESI-MS 2 base peak ion at m/z 297, which involved the loss of a molecule of (CH 3 ) 2 NH, a characteristic fragment ion of aporphine-type alkaloids. The proposed fragmentation pathway of compound 4 is shown in Scheme 2. Combined with bibliography data and fragmentation pathways, these two compounds were tentatively ascertained as magnoflorine and cyclanoline [25].  [23]. However, compound 4 could generated the ESI-MS 2 base peak ion at m/z 297, which involved the loss of a molecule of (CH3)2NH, a characteristic fragment ion of aporphine-type alkaloids. The proposed fragmentation pathway of compound 4 is shown in Scheme 2. Combined with bibliography data and fragmentation pathways, these two compounds were tentatively ascertained as magnoflorine and cyclanoline [25]. . The ion at m/z 369 was produced by the loss of (CH 3 ) 2 NH from the quasi-molecular ion, which suggested the compound 5 might be a kind of aporphine-type alkaloid [23].Moreover, it also generated fragments at m/z 386 [M − H + 2HCOOH − CO] − and m/z 354 [M − H + 2HCOOH − CO − CH 3 OH] − . Therefore, compound 5 was tentatively determined as 2,11-dihydroxy-10-methoxy-6,6-dimethyl-5,6,6a,7-tetra-hydro-4H-dibenzo[de,g]quinolin-6-ium.  Comparing with the literature data and respective fragmentation pathways, compound 29 and 56 was plausibly described as sagitiside A and 2-(4-hydroxy-3-methoxybenzyl)-3-(4-hydroxy-3methoxybenzylidene)butane-1,4-diol. Its ESI-MS 2 base peak ion at m/z 373 was generated by losing dehydrated glucose, and the major product ions at m/z 520 and m/z 505 were produced by the loss of one and two molecules of methyl from the quasi-molecular ion, respectively. In addition, the minor ion at m/z 357 [M − H − Glc − CH4] − generated from the ion at m/z 373 suggested the presence of adjacent methoxyl groups. By comparison with reference standard and literature data, compound 46 was proposed to be tinosposide A [28]. Its ESI-MS 2 base peak ion at m/z 373 was generated by losing dehydrated glucose, and the major product ions at m/z 520 and m/z 505 were produced by the loss of one and two molecules of methyl from the quasi-molecular ion, respectively. In addition, the minor ion at m/z 357 [M − H − Glc − CH 4 ] − generated from the ion at m/z 373 suggested the presence of adjacent methoxyl groups. By comparison with reference standard and literature data, compound 46 was proposed to be tinosposide A [28].

Isolated Compounds Identification
The raw spectral analysis data of three compounds, which were purified and identified from T. sinensis, are listed below.

Instrumentation and Condition
HPLC analysis was performed on DIONEX Ultimate 3000 UHPLC system (Thermo Fisher Scientific, Waltham, MA, USA) with a binary pump and an autosampler. Samples were separated on a Sunfire C 18

Conclusions
In this study, an effective and sensitive analytical method by HPLC-LTQ-Orbitrap-MS n was established for systematically characterizing non-diterpenoid constituents and guiding the extraction and isolation in T. sinensis extract. A total of 60 compounds attributed to four categories including 27 alkaloids, 23 phenylpropanoids, seven sesquiterpenoids and three other compounds were identified or tentatively characterized according to the t R and fragmentation pathways. In previous work, we have established a method for the content determination of total alkaloids in T. sinensis, and the results demonstrated that the alkaloid constituents were abundant in this herb, which were in accordance with this study [35]. Besides, 20 compounds were firstly characterized in the genus Tinospora and 13 of them were new compounds. Three natural compounds, including two phenylpropanoids and an alkaloid, were purified and identified from T. sinensis by systemic separation, which also demonstrated that the results of HPLC-LTQ-Orbitrap-MS n was reliable. The results serve well to illustrate the potential fragmentation pathways of non-deterpenoid constituents in T. sinensis, and the HPLC-LTQ-Orbitrap MS n platform was proved as an effective tool for rapid qualitative analysis of constituents. This study not only provides abundant information for better understanding of the chemical compounds in T. sinensis, but also benefits further quality control of this medicine.