Rapid Screening and Identification of Diterpenoids in Tinospora sinensis Based on High-Performance Liquid Chromatography Coupled with Linear Ion Trap-Orbitrap Mass Spectrometry

Abstract

Diterpenoids are considered the major active compounds in Tinospora sinensis in virtue of their special structures and activities. Herein, an analytical method was developed for rapid screening and identification of diterpenoids in T. sinensis using high-performmance liquid chromatography coupled with linear ion trap-Orbitrap mass spectrometry (HPLC-LTQ-Orbitrap) in negative ion mode. Two diterpenoid reference standards were first analyzed to obtain their characteristic ESI-MS/MS fragmentation patterns. Then, based on the extracted ion chromatogram (EIC) data-mining method and characteristic fragmentation pathways analysis, diterpenoids in T. sinensis were rapidly screened and identified. After that, an important parameter, Clog P, was adopted to discriminate between the isomers of diterpenoids. As a result, 63 diterpenoids were characterized from the extract of T. sinensis, including 10 diterpenoids and 53 diterpenoid glycosides. Among them, 15 compounds were tentatively identified as new compounds. Finally, target isolation of one diterpenoid glycoside named tinosineside A was performed based on the obtained results, which further confirmed the deduced fragmentation patterns and identified diterpenoid profile in T. sinensis. The results demonstrated that the established method could be a rapid, effective analytical tool for screening and characterization of diterpenoids in the complex systems of natural medicines.

1. Introduction

Tinospora sinensis (Kuanjinteng in Chinese) is commonly known as ‘Gurch’ and belongs to the family Menispermaceae, which is mainly distributed in the tropical parts of the eastern hemisphere. It has been traditionally used as folk medicine for treating debility, dyspepsia, rheumatism, gonorrhea, fever, inflammation, syphilis, ulcer, bronchitis, jaundice, urinary disease, skin disease and liver disease [1,2,3]. Previous phytochemical investigations have discovered that this species contains diterpenoids, alkaloids, flavonoids, lignans, triterpenes, amino acids, and so on. These constituents were reported to exert anti-inflammatory, immunomodulatory, antidiabetic, antitumor, antiadhesive, antiviral, anti-infective, antioxidant, antimutagenic, hematopoietic activities, etc. [4,5,6,7,8]. To our best knowledge, few reports on the systematic analysis of the diterpenoids in T. sinensis are available until now.

Recently, with the development of various data acquisition methods, high-resolution mass spectrometry (HRMS), especially linear ion trap-Orbitrap mass spectrometer (LTQ-Orbitrap), has exhibited excellent performance in the detection of targeted mixture constituents owing to its high speed and detection sensitivity [9,10]. The hybrid linear ion trap-Orbitrap mass spectrometer (LTQ-Orbitrap MS) combines high trapping capacity and MSⁿ scanning function of a linear ion trap together with accurate mass measurements within 3 ppm. The resolution power (up to 100,000) over a wider dynamic range is superior to that of many other mass spectrometers [11,12,13]. In the meantime, the combined application of tandem mass spectrometry for identifying the complicated constituents in traditional Chinese medicines (TCMs) could produce a large quantity of information data, such as molecular weights, elemental compositions, fragmentation patterns of multiple-stage, etc. [10]. Off-line processing of data, such as extracted ion chromatogram (EIC) data-mining method, plays a vital role in component identification [14,15]. These advantages have made LTQ-Orbitrap MS one of the most powerful approaches for the rapid identification and characterization of the multiple constituents found in TCMs [16]. It is well known that studies on flavonoids and alkaloids in T. sinensis by LC-MS have been widely reported, but little attention has been paid to its diterpenoids, mainly due to the complexity and variety of their skeletons, the diversity of their substituents, and a lack of corresponding reference standards. Hence, the characterization of diterpenoids in T. sinensis is of great significance.

In the present study, an HPLC-LTQ-Orbitrap data-acquisition approach combined with the EIC data-mining technique was established for comprehensive identification of the diterpenoids and their derivatives in T. sinensis.

2. Results and Discussion

Owing to the low content of diterpenoids in T. sinensis, a sensitive and reliable HPLC-LTQ-Orbitrap approach was established for determining their accurate masses and molecular formulas. In addition, the characteristic fragmentation pathways of diterpenoids were deduced from two obtained standards, and then used for the rapid identification and characterization of the other diterpenoids in T. sinensis. Furthermore, diterpenoid isomers were differentiated by an important parameter, Clog P, which is the absolute numerical value of the distribution coefficient of substances in a two-phase lipid-water system. Finally, target isolation of one diterpenoid glycoside named tinosineside A was performed based on the obtained results, which further confirmed the deduced fragmentation patterns. As a result, a total of 63 diterpenoids in T. sinensis were screened and divided into diterpenoid aglycones and diterpenoid glycosides (see supplementary materials and Table 1).

Table 1.

Summary of chemical constituents identified in Tinospora sinensis by HPLC-LTQ-Orbitrap.

NO.	tR/min	Identification	Empirical Formula	Proposed Ions	Experimental Mass m/z	Theoretical Mass m/z	Mass Error (×10−6)	MS2 Data (Measured)
1	16.13	(2,3,4,6)-2-(Acetoxymethyl)-6-(((2,4,6,6,10,10)-2-(furan-3-yl)-9,10-dimethoxy-7-(methoxycarbonyl)-6,10-dimethyl-4-oxododecahydro-1H-benzo[f]isochromen-6-yl)oxy)-tetrahydro-2H-pyran-3,4,5-triyl triacetate *	C37H49O17	[M − H]−	765.29578	765.29642	0.836	747(10), 737(48), 721(38), 720(43), 719(100), 697(5), 672(11), 555(18)
2	17.69	Cordioside	C26H33O12	[M − H]−	537.19592	537.19665	1.359	519(4), 491(8), 490(13), 375(100), 357(43), 327(88), 297(4)
3	18.55	(4,6,6,7,9,10,10,10)-2-(Furan-3-yl)-7-hydroxy-10b-methyl-4-oxo-6-(((3,5,6)-3,4,5-trihydroxy-6-(hydroxymethyl)-tetrahydro-2H-pyran-2-yl)-oxy)dodecahydro-1H-benzo-[f]isochromene-9,10-diyl diacetate *	C28H37O14	[M − H]−	597.21661	597.21778	1.959	579(3), 551(60), 476(7), 435(10), 389(96), 359(100)
4	20.51	(2,5,6,8,9,10,12)-15,16-Epoxy-2-hydroxy-6-O-{β-d-xylopyranosyl(1→6)-d-glucopyranosyl}-cleroda-3,13(16),14-trien-17,12-olid-18-oic acid methyl ester *	C32H43O16	[M − H]−	683.25372	683.25456	1.229	665(68), 639(12), 637(16), 615(95), 520(100), 519(79), 370(15), 309(37)
5	21.89	1-Deacetyltinosposide A	C24H33O12	[M − H]−	513.19678	513.19665	−0.253	495(3), 351(12), 333(100), 307(11), 305(5), 271(5)
6	25.01	Cordifoliside D	C26H33O12	[M − H]−	537.19623	537.19665	0.782	519(38), 492(22), 491(100), 490(71), 469(71), 297(38)
7	26.40	Tinospinoside D	C27H35O13	[M − H]−	567.20660	567.20721	1.075	552(1), 521(9), 404(0.5), 359(100), 341(2), 329(2)
8	26.55	(6,6,10,10,10)-Methyl 10-acetoxy-2-(furan-3-yl)-9-hydroxy-10b-methyl-4-oxo-6-(((3,5,6)-3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)-dodecahydro-1H-benzo[f]-isochromene-7-carboxylate *	C28H37O14	[M − H]−	597.21741	597.21778	0.619	565(4), 551(21), 550(25), 535(10), 463(12), 435(12), 389(100), 388(12)
9	26.98	Borapetoside B	C27H35O12	[M − H]−	551.21204	551.21230	0.471	536(5), 533(2), 507(3), 389(100), 388(15), 371(2), 370(7), 329(2)
10	27.65	Amritoside C	C27H35O13	[M − H]−	567.20612	567.20721	1.921	552(6), 529(20), 521(100), 404(10), 341(4)
11	29.43	Tinospinoside B	C27H35O12	[M − H]−	551.21298	551.21230	−1.233	533(9), 532(31), 515(12), 505(100), 389(8), 344(8)
12	29.52	(2,4,6,6,7,7,8,9,9,9)-2-(furan-3-yl)-6a,9b-dimethyl-6-(((2,3,4,5,6)-3,4,5-trihydroxy-6-(hydroxymethyl)-tetrahydro-2H-pyran-2-yl)oxy)decahydro-1H-9,7-(epoxymethano)-oxireno[2′,3′:4,5]benzo[1,2-f]isochromene-4,11(2H)-dione *	C26H31O12	[M − H]−	535.18036	535.18100	1.196	517(0.5), 489(2), 467(2), 373(100), 345(1), 343(2), 313(2)
13	30.82	Tinosposinenside A	C27H35O12	[M − H]−	551.21173	551.21230	1.034	536(5), 533(3), 507(3), 389(100), 374(11), 371(16), 359(2), 341(7)
14	31.03	(6,6,7,9,10,10,10)-2-(Furan-3-yl)-7-hydroxy-10b-methyl-4-oxo-6-(((3,5,6)-3,4,5-trihydroxy-6-(hydroxymethyl)-tetrahydro-2H-pyran-2-yl)oxy)dodecahydro-1H-benzo-[f]isochromene-9,10-diyl diacetate *	C28H37O14	[M − H]−	597.21692	597.21778	1.440	579(4), 561(11), 551(13), 529(19), 515(5), 477(6), 435(12), 389(82), 359(100)
15	31.41	Rumphioside A	C27H35O13	[M − H]−	567.20685	567.20721	0.635	549(14), 548(25), 535(25), 521(34), 405(46), 404(100), 359(56)
16	31.97	Furanoid diterpene glycoside	C26H33O11	[M − H]−	521.20099	521.20173	1.420	506(1), 359(100), 345(3), 344(4), 341(3)
17	32.03	Rumphioside F	C27H35O13	[M − H]−	567.20642	567.20721	1.393	549(22), 521(47), 491(100), 405(11), 404(18), 387(19)
18	33.06	(2,6,6,7,9,10,10,10)-2-(Furan-3-yl)-6,9,10-trihydroxy-10b-methyl-7-(((2,3,4,5,6)-3,4,5-trihydroxy-6-(hydroxyl-methyl)tetrahydro-2H-pyran-2-yl)oxy)decahydro-1H-benzo[f]isochromen-4(2H)-one *	C24H33O12	[M − H]−	513.19635	513.19665	0.585	495(3), 467(1), 351(100), 305(60), 287(7), 161(3)
19	34.12	(2R,5R,6R,8R,9S,10S,12S)-15,16-Epoxy-2-hydroxy-6-O-(β-d-glucopyranosyl)-cleroda-3,13(16),14-trien-17,12-olid-18-oic acid methyl ester	C27H35O12	[M − H]−	551.21185	551.21230	0.816	533(2), 519(3), 505(4), 482(4), 389(100), 327(67)
20	35.07	Tinosineside B	C28H37O14	[M − H]−	597.21716	597.21778	1.038	578(20), 551(100), 529(10), 517(12), 461(25), 179(49)
21	35.26	Palmatoside F	C26H31O12	[M − H]−	535.17987	535.18100	2.111	520(4), 517(2), 488(3), 373(93), 329(100)
22	35.35	(2,6,7,9,10,10,10)-2-(Furan-3-yl)-7,10-dihydroxy-9-methoxy-10b-methyl-6-(((2,3,4,5,6)-3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)deca-hydro-1H-benzo[f]isochromen-4(2H)-one *	C25H35O12	[M − H]−	527.21241	527.21230	−0.209	511(14), 509(3), 494(14), 481(100), 347(6), 300(7)
23	35.80	Amritoside A	C26H35O13	[M − H]−	555.20624	555.20721	1.747	537(5), 513(34), 495(53), 467(19), 393(100), 375(5), 305(74), 287(7)
24	36.87	Rumphioside D	C37H49O17	[M − H]−	765.29431	765.29642	2.757	747(19), 734(20), 721(37), 719(41), 718(100), 697(18), 600(26), 418(4), 393(11)
25	38.12	(4,6,6,9,10,10)-Methyl 9-acetoxy-2-(furan-3-yl)-4a- hydroxy-10b-methyl-4-oxo-6-(((3,5,6)-3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)-dodecahydro-1H-benzo[f]isochromene-7-carboxylate *	C28H37O14	[M − H]−	597.21619	597.21778	2.662	582(73), 579(9), 556(41), 551(28), 550(72), 434(31), 416(33), 389(75), 329(100)
26	38.15	Sagittatayunnanoside D	C26H35O11	[M − H]−	523.21667	523.21738	1.357	508(1), 505(1), 361(100), 347(8), 343(4), 329(4)
27	38.26	Borapetoside H	C33H45O17	[M − H]−	713.26588	713.26512	−1.066	694(12), 668(18), 667(29), 666(100), 645(23), 551(21), 533(6), 389(6)
28	38.38	(5R,6R,8S,9R,10R,12S)-15,16-Epoxy-2-oxo-6-O-(β-d-glucopyranosyl)–cleroda-3,13(16) 14-trien-17,12-olid-18-oic acid methyl ester	C27H33O12	[M − H]−	549.19592	549.19665	1.329	531(0.4), 503(5), 481(3), 387(100), 249(1)
29	38.80	Isocolumbin a	C20H21O6	[M − H]−	357.13321	357.13326	0.140	342(9), 339(2), 313(1), 311(13), 151(100), 135(37)
30	38.84	Tinoside	C26H31O11	[M − H]−	519.18542	519.18608	1.271	501(0.1), 473(0.3), 358(0.4), 357(100)
31	39.44	Tinosineside A	C26H35O13	[M − H]−	555.20581	555.20721	2.522	513(100), 495(69), 393(2), 375(8), 333(28), 315(10), 307(13), 305(1)
32	40.03	Tinocapilactone B	C22H25O8	[M − H]−	417.15411	417.15439	0.671	402(42), 371(9), 356(2), 181(100), 166(34), 151(16)
33	40.61	Tinoscorside C	C27H33O12	[M − H]−	549.19586	549.19665	1.438	531(7), 513(2), 503(13), 481(10), 417(5), 387(100)
34	40.79	(1,3,10,10)-9-(2-Methoxy-2,5-dihydrofuran-3-yl)-10a-methyl-4,7-dioxo-3-(((2,3,4,5,6)-3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)tetra-decahydroisobenzofuro[7,1-fg]isochromen-1-yl acetate *	C28H37O14	[M − H]−	597.21655	597.21778	2.060	579(51), 551(100), 528(39), 471(12)
35	41.60	Rumphioside I	C27H35O12	[M − H]−	551.21210	551.21230	0.363	533(9), 519(11), 505(2), 482(3), 339(100), 324(5)
36	42.17	Borapetoside C	C27H35O11	[M − H]−	535.21637	535.21738	1.887	520(1), 517(4), 488(1), 373(100), 359(5), 358(7), 355(2), 341(5)
37	43.22	(1,2,7,8)-1-(2-Ffuran-3-yl)-2-hydroxyethyl)-2-hydroxy-5-(methoxycarbonyl)-1-methyl-7-(((3,4,6)-3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)-1,2,3,4,6,7,8,8-octahydronaphthalene-2-carboxylic acid *	C26H35O13	[M − H]−	555.20630	555.20721	1.639	537(0.2), 513(1), 495(100), 393(0.2), 333(1), 297(2), 178(0.2)
38	43.99	Tinosposinenside B	C28H37O13	[M − H]−	581.22162	581.22286	2.133	563(8), 535(37), 419(2), 373(33), 343(100), 297(6)
39	44.59	Borapetoside A	C26H33O12	[M − H]−	537.19604	537.19665	1.136	519(4), 493(2), 491(62), 375(3), 371(5), 357(1), 341(2), 329(100), 297(1)
40	45.22	8-Hydroxycolumbin	C20H21O7	[M − H]−	373.12759	373.12817	1.554	358(2), 343(45), 325(6), 313(100)
41	45.78	6-Hydroxycolumbin	C20H21O7	[M − H]−	373.12759	373.12817	1.554	358(9), 355(7), 343(79), 329(3), 325(15), 313(100), 305(3), 261(2)
42	46.04	Boropetoside G	C27H37O11	[M − H]−	537.23218	537.23303	1.582	518(1), 375(100), 361(9), 360(2), 357(2), 343(2)
43	46.82	(4,6,6,9,10,10)-Methyl 9-acetoxy-2-(furan-3-yl)-4a-hydroxy-10b-methyl-4-oxo-6-(((3,5,6)-3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)dodec-ahydro-1H-benzo[f]isochromene-7-carboxylate *	C28H37O14	[M − H]−	597.2168	597.21778	1.641	579(23), 565(52), 564(45), 551(27), 529(31), 389(10), 329(100)
44	46.89	Tinocrisposide	C27H35O11	[M − H]−	535.21600	535.21738	2.578	520(2), 516(5), 488(7), 373(100), 359(8), 358(6), 355(3), 341(2)
45	49.44	Cordifolide A	C28H37O12S	[M − H]−	597.19922	597.20002	1.340	579(13), 553(33), 550(53), 533(100), 528(25), 467(43), 466(25), 434(18), 372(32), 359(11)
46	49.89	(2R,5R,6R,8S,9S,10S,12S)-15,16-Epoxy-2-hydroxy-6-O-{β-d-glucopyranosyl-(1→6)-α-d-xylopyranosyl}-cleroda-3,13(16),14-trien-17,12-olid-18-oic acid methyl ester	C32H43O16	[M − H]−	683.25416	683.25456	0.585	665(2), 664(11), 637(2), 615(3), 534(3), 520(100), 502(93), 490(12)
47	50.40	3-((6,6,7,10,10,10)-7-Hydroxy-6a,10b-dimethyl-4,12-dioxo-6-(((3,5,6)-3,4,5-trihydroxy-6-(hydroxymethyl)-tetrahydro-2H-pyran-2-yl)oxy)-2,4,4,5,6,6,7,10,10,10-decahydro-1H-10,7-(epoxymethano)benzo[f]isochromen-2-yl)tetrahydrofuran-2-yl acetate *	C28H37O14	[M − H]−	597.21680	597.21778	1.641	579(1), 561(15), 551(9), 495(6), 487(7), 486(66), 435(4), 433(100), 297(44)
48	51.12	Tinosposinenside C	C26H35O12	[M − H]−	539.21155	539.21230	1.391	521(3), 497(100), 479(86), 478(8), 377(2)
49	52.20	(1,3,10,10)-9-(2-Methoxy-2,5-dihydrofuran-3-yl)-10a-methyl-4,7-dioxo-3a-(((2,3,4,5,6)-3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)tetra-decahydroisobenzofuro[7,1-fg]isochromen-1-yl acetate *	C28H37O14	[M − H]−	597.21655	597.21778	2.060	579(0.4), 551(98), 509(7), 491(100), 481(6), 435(2)
50	59.05	6′-O-Lactoylborapetoside B	C30H39O14	[M − H]−	623.23434	623.23343	−1.460	513(3), 460(100), 458(7), 446(7), 444(4), 297(1), 283(12)
51	59.98	Columbin a	C20H21O6	[M − H]−	357.13297	357.13326	0.812
52	60.37	Tinospinoside E	C26H29O11	[M − H]−	517.16986	517.17043	1.102	499(26), 473(10), 471(20), 381(13), 355(6), 341(11), 162(100), 151(6)
53	61.73	Tinosporaside	C25H31O10	[M − H]−	491.19087	491.19117	0.611	473(28), 472(100), 460(8), 444(0.3), 327(0.4), 312(2)
			C26H33O12	[M − H + HCOOH]−	537.19697	537.19665	−0.596	519(4), 518(11), 491(25), 490(100), 357(3), 343(4), 327(8)
54	62.14	Sagittatayunnanoside B	C33H47O17	[M − H]−	715.28030	715.28077	0.657	670(2), 628(1), 652(1), 552(100), 551(16)
55	63.08	Tinospinoside C	C27H35O12	[M − H]−	551.21155	551.21230	1.361	533(73), 507(90), 505(100), 483(85), 415(84), 389(22), 343(81)
56	63.14	Sagittatayunnanoside C	C32H47O15	[M − H]−	671.28912	671.29094	2.711	653(1), 627(2), 509(50), 347(27), 329(100), 301(70), 241(25)
57	63.83	Tinosponone	C19H21O5	[M − H]−	329.13791	329.13835	1.337	314(100), 311(24), 285(70), 191(18)
58	65.99	Sagittatayunnanoside A	C26H37O10	[M − H]−	509.23792	509.23812	0.393	491(5), 465(5), 347(33), 329(100), 301(89), 257(9), 241(11)
59	69.79	2-O-Lactoylborapetoside B	C30H39O14	[M − H]−	623.23334	623.23343	0.144	608(1), 591(23), 551(7), 486(16), 460(100), 297(34)
60	78.98	Tinotufolin D	C20H25O4	[M − H]−	329.17441	329.17473	0.972	311(36), 285(100), 293(7), 267(10), 249(16)
61	82.48	(2aβ,3α,5aβ,6β,7α,8aα)-6-2-(3-Furanyl)ethyl-2a,3,4,5,5a,6,7,8,8a,8b-decahydro-2a,3-dihydroxy-6,7,8b-trimethyl-2H-naphtho1,8-bcfuran-2-one	C20H27O5	[M − H]−	347.18491	347.18530	1.123	329(63), 303(3), 301(100), 285(3), 187(1)
62	88.91	(3,4,5,8)-Methyl-5-(2-(furan-3-yl)ethyl)-3-hydroxy-5,8a-dimethyl-3,4,4,5,6,7,8,8a-octahydronaphthalene-1-carboxylate *	C20H27O4	[M − H]−	331.19016	331.19038	0.664	313(9), 287(15), 285(100), 283(30), 271(2), 257(5), 243(2), 237(0.4)
63	95.27	Tinotufolin C	C21H31O5	[M − H]−	363.21662	363.21660	−0.055	345(0.1), 334(0.2), 317(69), 295(100)

^a Comparison with standards. * Tentatively identified as new compounds.

2.1. The Fragmentation Patterns of Reference Standards

Columbin and isocolumbin yielded their [M − H]⁻ ions at m/z 357.13326 (C₂₀H₂₁O₆). Both of their deprotonated molecular ions generated a serial of fragment ions at m/z 342, m/z 339, m/z 313 and m/z 295, corresponding to [M − H − CH₃]⁻, [M − H − H₂O]⁻, [M − H − CO₂]⁻ and [M − H − H₂O − CO₂]⁻. Columbin produced a fragment ion at m/z 329 through losing carbonyl, and the ion at m/z 329 loss of a CO₂ produced its ESI-MS² base peak ion at m/z 285. The ion at m/z 245 [M − H − 112]⁻ resulted from a Retro-Diels-Alder (RDA) cleavage fragmentation at the 1,4-position of the A-ring. Moreover, the product ion at m/z 245 generated the minor ion at m/z 217 by loss of one molecule of CO.

Meanwhile, isocolumbin produced its ESI-MS² base peak ion at m/z 339 by neutral loss of H₂O. After a RDA cleavage at A-ring and successive fracture of C_9–11 (C₆H₆O, which was D-ring plus –C₂H₂), the prominent ion at m/z 151 was generated (Figure 1). Therefore, the characteristic fragment ions of reference standards were deduced, such as [M − H − 44]⁻ generated by loss of CO₂ from lactone ring, as well as [M − H − 15]⁻, [M − H − 18]⁻, [M − H − 44 − 18]⁻, [M − H − 112]⁻, etc., According to these characteristic fragmentation patterns, diterpenoids in T. sinensis could be rapidly screened and identified.

Figure 1.

Proposed MS/MS mechanistic pathways for columbin and isocolumbin.

2.2. Identification of Diterpenoids Aglycones

The molecular formulas of diterpenoids were predicted by a HRMS database built in-house. Combined with multi-stage mass spectrometry fragmentation information and bibliographic data, two categories of diterpenoids and its derivatives were screened and determined, including 10 diterpenoid aglycones and 53 diterpenoid glycosides. The typical total ion chromatogram (TIC) of T. sinensis in negative ion mode is presented in Figure 2.

Figure 2.

TIC chromatogram of T. sinensis in negative ion mode.

Compounds 29 and 51 generated their [M − H]⁻ ions at m/z 357.13326 (C₂₀H₂₁O₆). They were unambiguously assigned as isocolumbin and columbin, respectively, by comparison of their retention times, literature data and MS fragmentation data with their standards [17,18]. Isocolumbin and columbin constitute a pair of diastereomers, whose retention time and polarity are different under the present chromatographic conditions.

Compound 32 produced its [M − H]⁻ ion at m/z 417.15439 (C₂₂H₂₅O₈). In CID mode, it further generated [M − H − CH₃]⁻, [M − H − H₂O − CO]⁻ and [M − H − H₂O − CO − CH₃]⁻ ion at m/z 402, m/z 371 and m/z 356, respectively. The ESI-MS² base peak ion at m/z 181 (C₁₁H₁₇O₂) was generated by RDA cleavage fragmentation from 1,4-position of the A-ring and fracture of C_9–11 (C₆H₄O₃). The product ion at m/z 181 loss of a methyl generated the prominent ion at m/z 166. Therefore, compound 32 was tentatively deduced as tinocapilactone B [19].

Compounds 40 and 41 produced their [M − H]⁻ ions at m/z 373.12817 (C₂₀H₂₁O₇). Both of their deprotonated molecular ions generated a series of fragment ions at m/z 358 [M − H − CH₃]⁻, m/z 343 [M − H − CO − H₂]⁻, m/z 325 [M − H − 48 amu]⁻ and m/z 313 [M − H − 60 amu]⁻, respectively. In addition, compound 41 generated two minor ions at m/z 305 and m/z 261. By comparison with the MS fragmentation patterns obtained from the reference standards, compounds 40 and 41 were plausibly characterized as 8-hydroxycolumbin and 6-hydroxycolumbin, respectively.

Compound 57 showed the [M − H]⁻ ion at m/z 329.13835 (C₁₉H₂₁O₅). Its ESI-MS² base peak ion at m/z 314 was generated by loss of a methyl. The major fragment ions at m/z 311 and m/z 285 were yielded by neutral loss of H₂O and CO₂, respectively. Moreover, the product ion at m/z 285 generated the predominant product ion at m/z 191 due to the overall fracture of C_9–11 (C₆H₆O). According to the literature data, it was tentatively identified as tinosponone [20].

Compounds 60 and 61 produced their respective [M − H]⁻ ion at m/z 329.17473 (C₂₀H₂₅O₄) and m/z 347.18530 (C₂₀H₂₇O₅). Both of their deprotonated molecular ions generated [M − H − H₂O]⁻, [M − H − CO₂]⁻ and [M − H − H₂O − CO₂]⁻ ions at m/z 311, m/z 329, m/z 285, m/z 303 and m/z 267, m/z 285, respectively. By comparing with the literature data, compounds 60 and 61 were tentatively identified as tinotufolin D and (2aβ,3α,5aβ,6β,7α,8aα)-6-2-(3-furanyl)ethyl-2a,3,4,5,5a,6,7,8,8a,8b- decahydro-2a,3-dihydroxy-6,7,8b-trimethyl-2H-naphtho-1,8-bcfuran-2-one, respectively [21].

Compound 62 gave a [M − H]⁻ ion at m/z 331.19038 (C₂₀H₂₇O₄). Its MS² spectrum produced the fragment ion at m/z 313, which involves the loss of H₂O. The molecular ion yielded a series of fragment ions at m/z 287 [M − H − CO₂]⁻, m/z 283 [M − H − H₂O − H₂CO]⁻ and m/z 271 [M − H − CH₄ − CO₂]⁻, suggesting the presence of -COOCH₃. The base peak ion at m/z 285 was generated by losing water and carbon monoxide. The [M − H − C₆H₆O]⁻ ion at m/z 237 was generated by the fracture of C_9-11 (C₆H₆O). Moreover, by comparison with the fragmentation patterns obtained from the two reference standards, compound 62 was tentatively identified as a new compound.

Compound 63 generated a [M − H]⁻ ion at m/z 363.21660 (C₂₁H₃₁O₅). It generated a serial of fragment ions at m/z 345 [M − H − H₂O]⁻, m/z 334 [M − H − 29 amu]⁻, m/z 317 [M − H − H₂O − CO]⁻ and m/z 295 [M − H − 2H₂O − OCH₄]⁻. By comparison with the literature data, compound 63 was tentatively identified as tinotufolin C [21].

2.3. Identification of Diterpenoids Glycosides

Compounds 1 and 24 produced their [M – H]⁻ ions at m/z 765.29642 (C₃₇H₄₉O₁₇). After the CID cleavage, both of their further fragmentations resulted in a [M − H − CO₂]⁻ ion at m/z 721 and a [M − H − H₂O − CO]⁻ ion at m/z 719, which were consistent with the characteristic fragmentation pathways of diterpenoids. According to the fragmentation patterns and the values of Clog P, compounds 1 and 24 were plausibly characterized as (2,3,4,6)-2-(acetoxymethyl)-6-(((2,4,6,6,10,10)-2-(furan-3-yl)-9,10-dimethoxy-7-(methoxycarbonyl)-6,10-dimethyl-4-oxododecahydro-1H-benzo[f]isochromen-6-yl)-oxy)tetrahydro- 2H-pyran-3,4,5-triyl triacetate and rumphioside D, respectively.

Compounds 2, 6 and 39 produced their [M − H]⁻ ions at m/z 537.19665 (C₂₆H₃₃O₁₂). They all produced the [M − H − H₂O]⁻ and [M − H − H₂O − CO]⁻ ions at m/z 519 and m/z 491. In addition, compounds 2 and 39 also yielded the [M − H − 162]⁻ ions at m/z 375 due to the overall fracture of dehydrated glucose, which was the ESI-MS² base peak of compound 2. As well as we known, [M − H − 162]⁻ was the characteristic ion of glycosides. Furthermore, compounds 2, 6 and 39 generated the ions at m/z 490 [M − 2H − H₂O − CO]⁻, m/z 357 [M − H − Glc]⁻, m/z 327 [M − H − Glc − 2CH₃]⁻, m/z 297 [M − 2H − CO − Glc − OCH₃]⁻; m/z 492 [M − 2H − CO₂]⁻, m/z 490 [M − 2H − H₂O − CO]⁻, m/z 469 [M − 2H − H₂O − CO − OCH₃]⁻, m/z 297 [M − 2H − CO − Glc − OCH₃]⁻; m/z 493 [M − H − CO₂]⁻, m/z 357 [M − H − Glc]⁻, m/z 341 [M − H − O − Glc]⁻, m/z 329 [M − H − CO − Glc]⁻, m/z 297 [M − 2H − CO − Glc −OCH₃]⁻, respectively. By comparison with the literature data, compounds 2, 6 and 39 were tentatively assigned as cordioside, cordifoliside D and borapetoside A, respectively [22,23].

Compounds 3, 8, 14, 20, 25, 34, 43, 47 and 49 were all observed to possess the same [M − H]⁻ ions at m/z 597.21778 (C₂₈H₃₇O₁₄). Firstly, they were divided into two categories based on whether they produced the [M − H − 112]⁻ ion at m/z 486 by the RDA cleavage from A-ring. Compound 47 generated its ESI-MS² base peak ion at m/z 433 by losing dehydrated glucose. In addition, it also yielded [M − H − 2H₂O]⁻ ion at m/z 561, [M − H − H₂O − CO]⁻ ion at m/z 551, [M − H − 112]⁻ ion at m/z 486, [M − H − dehydrated Glc]⁻ ion at m/z 435, and so on. As far as we knew, there were no related literatures reported. And thus, compound 47 was finally deduced to be a new compound. Compounds 3, 8, 14, 20, 25, 34, 43 and 49 yielded their [M − H − H₂O − CO]⁻ ions at m/z 551, which was the ESI-MS² base peak of compounds 20 and 34, and it also a high intense characteristic fragment ion of compound 49. Compared with the prominent ions of them, there was no difficulty to deduce that compounds 34 and 49 have two lactone rings, which were different to the other compounds. Furthermore, the fragment ions further validated the above deduction. Finally, according to their Clog P values, they were tentatively identified and differentiated. By comparison with the bibliography and MS fragmentation data, compounds 3, 8, 14, 20, 25, 34, 43 and 49 were tentatively identified (Table 1) [24,25].

Compounds 4 and 46 generated their [M − H]⁻ ions at m/z 683.25456 (C₃₂H₄₃O₁₆). Both of their deprotonated molecular ions yielded [M − H − H₂O]⁻ ion at m/z 665, [M − H − H₂O − CO]⁻ ion at m/z 637, [M − H − C₄H₄O]⁻ ion at m/z 615, [M − H − dehydrated Glc]⁻ ion at m/z 520, respectively. Moreover, compound 4 also dissociated into fragment ions [M − H − CO2]⁻ at m/z 639, [M − H − Glc − dehydrated xyl]⁻ at m/z 370 and [M − H − Glc − dehydrated xyl − CO₂ − H₂O]⁻ at m/z 309. According to the literature, compounds 4 and 46 were plausibly characterized as (2R,5R,6R,8S,9S,10S,12S)-15,16-epoxy-2-hydroxy-6-O-{β-d-xylopyranosyl(1→6)–d-glucopyranosyl}-cleroda-3,13(16),14-trien-17,12-olid-18-oic acid methyl ester and (2R,5R,6R,8S,9S,10S,12S)-15,16-epoxy-2-hydroxy-6-O-{β-d-glucopyranosyl-(1→6)-α-d-xylopyran-osyl}-cleroda-3,13(16),14-trien-17,12-olid-18-oic acid methyl ester, respectively [26].

Compounds 5 and 18 produced their [M − H]⁻ ions at m/z 513.19665 (C₂₄H₃₃O₁₂). The dominant characteristic fragment ions were presented at m/z 333 [M – H − 180]⁻ and m/z 351 [M − H − 162]⁻ as the ESI-MS² base peak corresponding to the cleavage from glucopyranosyl. After the CID cleavage, both of their further fragmentation resulted in [M − H − H₂O]⁻ at m/z 495 and [M − H − Glc − CO]⁻ at m/z 305. The product ion at m/z 305 of compound 18 further generated the predominant ion at m/z 287 by losing one molecular of water. By comparison with the literature data and MS fragmentation data, compounds 5 and 18 were tentatively identified as 1-deacetyltinosposide A and (2S,6R,6aR,7R,9S,10R,10aR,10bS)-2-(furan-3-yl)-6,9,10–trihydroxy-10b-methyl-7-(((2R,3R,4S,5S,6R)-3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)decahydro-1H-benzo[f]isochromen-4(2H)-one, respectively [4].

Compounds 7, 10, 15 and 17 gave their [M − H]⁻ ions at m/z 567.20721 (C₂₇H₃₅O₁₃). They all produced the [M − H − H₂O − CO]⁻ and [M − H − Glc]⁻ ions at m/z 521 and m/z 405. Besides, compounds 7 and 10 generated m/z 552 by losing a methyl, and the ion at m/z 341 was generated by losing a series of fragment ions of glucose residue, 2H₂O, and CO. Compounds 7 and 15 were also generated [M − H − Glc − CO]⁻ ions at m/z 359, which was the ESI-MS² base peak of compound 7. Meanwhile, compound 7 generated the minor ion at m/z 329. Compound 17 produced its ESI-MS² base peak by losing two molecular of methyl, one molecular of water and carbonyl at m/z 491. Combined with bibliography data and fragmentation pathways, these four compounds were tentatively deduced as tinospinoside D, amritoside C, rumphioside A and rumphioside F, respectively [25].

Compounds 9, 11, 13, 19, 35 and 55 all produced their [M − H]⁻ ions at m/z 551.21230 (C₂₇H₃₅O₁₂). Compounds 9, 13 and 19 produced their ESI-MS² base peak ions at m/z 389 [M − H − dehydrated Glc]⁻, and further generated a series of fragment ions by losing methyl, water, carbon dioxide units at the mean time. According to the literature and the values of Clog P, compounds 9, 13 and 19 were plausibly characterized as borapetoside B, tinosposinenside A and (2R,5R,6R,8R,9S,10S,12S)-15,16-epoxy-2-hydroxy-6-O-(β-d-glucopyranosyl)-cleroda-3,13(16), 14-trien-17,12-olid-18-oic acid methyl ester, respectively [26]. Compounds 11 and 55 generated their ESI-MS² base peak ions at m/z 505 by losing water and carbonyl, and their deprotonated molecular ions loss of glucose residue at m/z 389. According to the fragmentation pathways and the values of Clog P, compounds 11 and 55 were tentatively identified as tinospinoside B and tinospinoside C, respectively. According to the above methods, compound 35 was tentatively defined as rumphioside I.

Compounds 12 and 21 yielded their [M − H]⁻ ions at m/z 535.18100 (C₂₆H₃₁O₁₂). Compound 12 produced its ESI-MS² base peak ion at m/z 373 by loss of a dehydrated glucose and then generated two minor ions at m/z 345 [M − H − dehydrated Glc − CO]⁻ and m/z 313 [M − H − O − dehydrated Glc − CO₂]⁻. While compound 21 generated a series of fragment ions at m/z 520 [M − H − CH₃]⁻, m/z 517 [M − H − H₂O]⁻, m/z 373 [M − H − dehydrated Glc]⁻ and m/z 329 [M − H − dehydrated Glc − CO₂]⁻. According to the fragmentation patterns and the values of Clog P, compounds 12 and 21 were tentatively characterized as (2,4,6,6,7,7,8,9,9,9)-2-(furan-3-yl)-6a,9b-dimethyl-6-(((2,3,4,5,6)-3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)decahydro-1H-9,7-(epoxymethano) oxireno[2′,3′:4,5]benzo[1,2-f]isochromene-4,11(2H)-dione and palmatoside F, respectively.

Compound 16 produced its [M − H]⁻ ion at m/z 521.20173 (C₂₆H₃₃O₁₁). A high intense characteristic fragment ion was present at m/z 359 [M – H − 162]⁻ as its ESI-MS² base peak corresponding to the cleavage of the dehydrated glucose. The product ion at m/z 359 successively generated the predominant product ions at m/z 344 and m/z 341 by losing a methyl and one molecular of water. According to the literature data, compound 16 was tentatively identified as furanoid diterpene glycoside [27].

Compound 22 generated its [M − H]⁻ ion at m/z 527.21230 (C₂₅H₃₅O₁₂). Upon CID mode, its further fragmentation resulted in [M − H − H₂O]⁻ ion at m/z 509, [M − H − H₂O − CH₃]⁻ ion at m/z 494, [M – H − H₂O − CO]⁻ ion at m/z 481, [M − H − Glc]⁻ ion at m/z 347, and [M − H − Glc − CH₂O − CO − CH₄]⁻ ion at m/z 273. As far as we knew, there were no related literatures reported. Thus, compound 22 was tentatively deduced to be a new compound.

Compounds 23, 31 and 37 yielded their [M − H]⁻ ions at m/z 555.20721 (C₂₆H₃₅O₁₃). Compound 31 produced its ESI-MS² base peak ion at m/z 513 [M − Ac]⁻, and the m/z 513 ion generated its predominant ions at m/z 495 [M − H − CH₂CO − H₂O]⁻ and m/z 333 [M − H − CH₂CO − dehydrated Glc]⁻ due to the overall fracture of a hydroxy and glycoside bond. The ion at m/z 333 generated two minor ions at m/z 315 and m/z 305 through losing H₂O and CO, respectively. Besides, its [M − H]⁻ ion also yielded characteristic fragment ion at m/z 393 [M − H − 162]⁻ by losing a dehydrated glucose. By comparison with the literature data, compound 31 was tentatively identified as tinosineside A [24]. Compounds 23 and 37 produced their fragment ions at m/z 537 [M − H − H₂O]⁻, m/z 513 [M − CH₃ − CO]⁻, m/z 495 [M − H − CH₃ − COOH]⁻ and m/z 393 [M − H − dehydrated Glc]⁻. Furthermore, compound 23 generated the fragment ions at m/z 375 [M − H–H₂O–Glc]⁻, m/z 305 [M − H–dehydrated Glc–CH₄ − CO − CO₂]⁻ and m/z 297 [M − H − H₂O − Glc − CO₂ − CH₄]⁻ or [M − H − 2H₂O − Glc − CH₂CO]⁻. According to the fragment ions and the values of Clog P, compounds 23 and 37 were tentatively identified as amritoside A and (1,2,7,8)-1-(2-(furan-3-yl)-2-hydroxyethyl)-2-hydroxy-5-(methoxycarbonyl)-1-methyl-7-(((3,4,6)-3,4,5-trihydroxy-6-(hydroxymethyl)tetrahydro-2H-pyran-2-yl)oxy)-1,2,3,4,6,7,8,8-octahydronaphthalene-2-carboxylic acid, respectively.

Compound 26 produced its [M − H]⁻ ion at m/z 523.21738 (C₂₆H₃₅O₁₁). Its ESI-MS² base peak ion at m/z 361 was yielded by loss of a dehydrated glucose. In addition, it generated a series fragment ions, such as m/z 508, m/z 505, m/z 347, m/z 343 and m/z 329, involving the loss of a methyl group, a water molecule, and glucose. By comparing with the fragmentation patterns obtained from two reference standards, compound 26 was tentatively identified as sagittatayunnanoside D. Similarly, based on the literature data, compound 27 was tentatively identified as borapetoside H [28].

Compounds 28 and 33 exhibited their [M − H]⁻ ions at m/z 549.19665 (C₂₇H₃₃O₁₂). Both of their deprotonated molecular ions generated a serial of fragment ions at m/z 531 [M − H − H₂O]⁻, m/z 503 [M − H − H₂O − CO]⁻, m/z 481 [M − H − 2H₂O − CH₃OH]⁻ and m/z 387 [M − H − dehydrated Glc]⁻, respectively. Meanwhile, a product ion at m/z 513 [M − H − 2H₂O]⁻ was observed in the ESI-MS/MS spectra of compound 33, suggesting there was two hydroxy groups at the skeleton. According to the literature data and the values of Clog P, compounds 28 and 33 were tentatively identified as (5R,6R,8S,9R,10R,12S)-15,16-epoxy-2-oxo-6-O-(β-d-glucopyranosyl)–cleroda-3,13(16),14-trien-17,12 -olid-18-oic acid methyl ester and tinoscorside C, respectively [26].

Compound 30 generated its [M − H]⁻ ion at m/z 519.18608 (C₂₆H₃₁O₁₁). The dominant characteristic fragment ion was present at m/z 357 [M − H − 162]⁻ as its ESI-MS² base peak, corresponding to the cleavage from the dehydrated glucose. Meanwhile, the product ions [M − H − H₂O]⁻ (lose of water) at m/z 501 and [M − H − 46]⁻ (lose of water and carbon monoxide) at m/z 473 were monitored, which were in accordance with the diterpenoids glycosides cracking patterns, suggesting this compound might be tinoside.

Compounds 36 and 44 yielded their [M − H]⁻ ions at m/z 535.21738 (C₂₇H₃₅O₁₁). Both of their deprotonated molecular ions generated a serial of fragment ions at m/z 520 [M − H − CH₃]⁻, m/z 488 [M − 2H − CO − H₂O]⁻, m/z 373 [M − H − dehydrated Glc]⁻, m/z 359 [M − H − dehydrated Glc − CH₂]⁻, m/z 358 [M − H − dehydrated Glc − CH₃]⁻, m/z 355 [M − H − Glc]⁻ and m/z 341 [M − H − Glc − CH₂]⁻. According to the literature data and the values of Clog P, compounds 36 and 44 were tentatively identified as borapetoside C and tinocrisposide [29].

Compound 38 gave a [M − H]⁻ ion at m/z 581.22286 (C₂₈H₃₇O₁₃). Its major ions in the MS² spectrum were m/z 563 [M − H − H₂O]⁻, m/z 535 [M − H − H₂O − CO]⁻, m/z 419 [M − H − dehydrated Glc]⁻, m/z 373 [M − H − Glc − CO]⁻ and m/z 343 [M − Glc − CH₂CO₂]⁻, suggesting this compound might contain the glucose and acetyl fragments. Comparison with the parent nucleus and fragmentation patterns, compound 38 was tentatively identified as tinosposinenside B.

Compound 42 exhibited its [M − H]⁻ ion at m/z 537.23303 (C₂₇H₃₇O₁₁). Its ESI-MS² base peak ion at m/z 375 was produced by losing dehydrated glucose, and the product ion at m/z 375 further generated two minor ions at m/z 360 and m/z 357 by losing CH₃ and H2O, respectively. According to the fragmentation pathways, compound 42 was tentatively identified as boropetoside G. Likewise, compound 48 was tentatively defined as tinosposinenside C.

Compound 45 produced its [M − H]⁻ ion at m/z 597.20002 (C₂₈H₃₇O₁₂S), which exhibited the characteristic fragment ion [M − H − C₈H₁₄O₆S]⁻ at m/z 359 suggesting the overall fracture of side chain. S-diterpenoids glycosides were very rare in Tinospora, and there only two compounds were reported in this family so far [30]. By comparison with the fragmentation pathways, compound 45 was plausibly described as cordifolide A.

Compounds 50 and 59 produced their [M − H]⁻ ions at m/z 623.23343 (C₃₀H₃₉O₁₄). According to the respective fragmentation pathways and the values of Clog P, compounds 50 and 59 were tentatively identified as 6′-O-lactoylborapetoside B and 2-O-lactoylborapetoside B, respectively.

Compounds 52–54 produced their respective [M − H]⁻ ion at m/z 517.17043 (C₂₆H₂₉O₁₁), m/z 491.19117 (C₂₅H₃₁O₁₀) and m/z 715.28077 (C₃₃H₄₇O₁₇). None but compound 53 exhibited the [M − H + HCOOH]⁻ adduct ion at m/z 537.19665 (C₂₆H₃₃O₁₂). Take compound 52 for example, its major ions in ESI-MS² spectrum were m/z 499 [M − H − H₂O]⁻, m/z 473 [M − H − CO₂]⁻, m/z 471 [M − H − H₂O − CO]⁻, m/z 455 [M − H − CO₂ − H₂O]⁻, m/z 355 [M − H − dehydrated Glc]⁻, m/z 162 −C₆H₁₀O₅ and m/z 151 –C₅H₁₁O₅, suggesting this compound might contain glucose and undergo a RDA coverage from the A-ring at the 1,4-position. According to the above analysis, compounds 52–54 were tentatively defined as tinospinoside E, tinosporaside and sagittatayunnanoside B, respectively.

Compounds 56 and 58 generated their respective [M − H]⁻ ion at m/z 671.29094 (C₃₂H₄₇O₁₅) and 509.23812 (C₂₆H₃₇O₁₀). The defference between of these molecules was 162 Da, suggesting that compound 56 had one more glucose unit than compound 58. Both of them generated the same ESI-MS² base peak ions at m/z 329 due to the overall fracture of glycoside bond. Furthermore, the product ion at m/z 329 generated a predominant ion at m/z 301 by splitting off CO. Comparing with the parent nucleus and respective fragmentation pathways, compounds 56 and 58 were plausibly identified as sagittatayunnanoside C and sagittatayunnanoside A, respectively.

2.4. Target Isolation and Further Verification of Diterpenoids Fragmentation Patterns

Tinosineside A generated its deprotonated molecular ion [M − H]⁻ at m/z 555.20721 (C₂₆H₃₅O₁₃, mass error within 3 ppm). It firstly produced the ESI-MS² base peak ion at m/z 513 [M − Ac]⁻ and further generated the ions at m/z 495 and m/z 375 by losing one molecule of water and dehydrated glucose. The ion at m/z 351 generated minor ions at m/z 333, m/z 315 and m/z 307 through losing H₂O, 2H₂O and CO2, respectively. Besides, its [M − H]⁻ ion also yielded characteristic fragment ion at m/z 393 [M − H − 162]⁻ by losing a dehydrated glucose. The fragmentation pathway was consistent with deduced of compound 31, which further proved the validity of the results (Figure 3).

Figure 3.

Spectra of ion fragments in MSⁿ analysis of tinosineside A in negative ion mode.

3. Materials and Methods

3.1. Chemicals

The reference standards of columbin and isocolumbin (purity over 98%) were procured from Tauto Biotech (Shanghai, China).The structures of these standards were presented in Figure 1. HPLC grade formic acid, acetonitrile and methanol were purchased from Fisher Scientific (Fair Lawn, NJ, USA). Ultrapure water was purchased from Hangzhou Wahaha Group Co., Ltd. (Hangzhou, Zhejiang, China). Material of T. sinensis was purchased from Anguo Linshi Medicinal Materials Co., Ltd. (Anguo, China) and then authenticated by Professor Chun-sheng Liu in Beijing University of Chinese Medicine.

3.2. Sample Preparation

3.2.1. Standard Solutions

The standard solutions of columbin and isocolumbin were prepared in methanol at appropriate concentrations.

3.2.2. Sample Solutions

Powdered dried alcoholic extracts of T. sinensis were weighed accurately (0.13 g) and refluxed with a tenfold excess of ethanol/water (70:30, v/v) three times, and placed in 20 mL of methanol/water (70:30, v/v). The mixture was then extracted in an ultrasonic bath at room temperature for 0.5 h, and the same solvent was added to compensate for the weight lost during the extraction. The extract was filtered and evaporated to near dryness, then placed on a C₁₈ Solid Phase Extraction (SPE) column (J.T.Baker, Phillipsburg, NJ, USA), which was washed with 4 mL distilled water and 4 mL methanol. The methanol eluent was filtered through a 0.22 µm membrane for analysis. All of the solutions were stored at 4 °C and brought to room temperature before analysis.

3.3. Instrumentation and Condition

HPLC analysis was carried out on a DIONEX Ultimate 3000 UHPLC system (Thermo Fisher Scientific, Waltham, MA, USA) with a binary pump and an autosampler. The samples were separated on a Sunfire C18 column (250 mm × 4.6 mm i.d., 5 μm, Waters Corporation, Milford, MA, USA) at room temperature. The mobile phase consisted of acetonitrile (B) and 0.1% (v/v) formic acid in water (A) with the elution gradient set as follows: 0–5 min, 8–12% B; 5–25 min, 12–16% B, 25–45 min, 16–25% B; 45–75 min, 25–46% B; 75–80 min, 46–58% B; 80–95 min, 58–65% B; 95–105 min. The flow rate was set as 1.0 mL/min.

A LTQ-Orbitrap XL mass spectrometer (Thermo Scientific, Bremen, Germany) was connected to the HPLC system via an electrospray ionization (ESI) interface. The effluent was introduced into the ESI source in a post-column splitting ratio of 1:4. Full scan data acquisition was performed from m/z 100 to 1500 in negative ion mode. The important ESI parameters were set as follows: capillary temperature, 350 °C; sheath gas (nitrogen) flow, 30 arb.; auxiliary gas (nitrogen) flow, 10 arb.; electrospray voltage, 3.0 kV; capillary voltage, −35 V; tube lens voltage, −110 V. The resolution of Orbitrap analyzer was set at 30,000 with data-dependent ESI-MS² analysis triggered by the three most abundant ions from one-stage mass spectrometry scanning. Collision-induced dissociation (CID) was performed in LTQ with an activation q of 0.25 and activation time of 30 ms. The isolation width was 2 amu, and the normalized collision energy was set to 35%.

3.4. Peak Selections and Data Processing

A Thermo Xcalibur 2.1 workstation was used for the data acquisition and processing, In order to obtain as many fragment ions of the diterpenoids as possible, the peaks detected with intensity over 30,000 were selected for identification. The chemical formulas for all parent ions of the selected peaks were calculated from the accurate mass using a formula predictor by setting the parameters as follows: C (0–50), H (0–100), O (0–30), S (0–2), N (0–2). Other elements were not considered because they are rarely present in diterpenoids. Furthermore, diterpenoids isomers were differentiated by an important Clog P parameter obtained from Chemdraw to distinguish their polarity. And the exact mass error of all determined compounds was within 3 ppm.

3.5. Extraction and Isolation of Tinosineside A

The air-dried stems (10.0 kg) were extracted three times with tenfold excess of 70% EtOH under reflux for 2 h each at 80 °C. The combined extract was evaporated under reduced pressure to obtain a crude residue. This residue was further dispersed in H₂O, and then successively extracted with CHCl₃, EtOAc, n-BuOH and MeOH. The n-BuOH extract was passed through an AB-8 macroporous resin column and then washed with H₂O, 30% EtOH, 50% EtOH, 70% EtOH and 95% EtOH. The 30% EtOH fraction was further purified by silica gel column chromatography with elution by CHCl₃–MeOH (15:1→4:1, v/v) to give the diterpenoid glycoside named tinosineside A as a white powder.

4. Conclusions

In this study, a sensitive HPLC-LTQ-Orbitrap coupled with EIC data-mining method was established for the rapid characterization of diterpenoids in T. sinensis. According to the characteristic fragmentation pathways of two reference standards, all diterpenoids were divided into two categories, namely diterpenoid aglycones and diterpenoid glycosides. Furthermore, an important parameter Clog P was used to differentiate the isomers of diterpenoids. As a result, 63 diterpenoids were preliminarily identified, including 48 known compounds and 15 new compounds. These compounds included 10 diterpenoid aglycones, 53 diterpenoid glycosides. Finally, target isolation of one diterpenoid glycoside named tinosineside A was performed based on the obtained results, which further confirmed the deduced fragmentation patterns and identified profile of diterpenoids in T. sinensis. This represents the first systematic report of diterpenoids in T. sinensis. The results indicated that the established method could be employed as a rapid, effective technique to screen and identify diterpenoids in T. sinensis. The study also provided significant guidance for the analysis of other herbal medicines or preparations.

Supplementary Materials

The supplementary files are available online.

Author Contributions

Bin Liu and Jia-Yu Zhang designed the experiments; Feng-Xia Guo and Yan-Yan Jiang contributed to the isolation and identification of Tinosineside A; Zi-Jian Wang, Lu-Lu Xu and Jia-Yu Zhang contribute to the data collection and analysis; Sen-Sen Chi, Jia-Yu Zhang and Bin Liu contributed reagents/materials/analysis tools; Lu-Lu Xu, Jia-Yu Zhang and Bin Liu wrote the paper.

Conflicts of Interest

The authors declare no conflict of interest.