Qualitative and Quantitative Analyses of Alkaloids in Uncaria Species by UPLC-ESI-Q-TOF/MS

Abstract

An ultra performance liquid chromatography (UPLC) coupled with quadrupole time-of-flight mass spectrometry (Q-TOF/MS) method has been optimized and established for the rapid analysis of the alkaloids in 22 samples originating from five Uncaria (U.) species. The accurate mass measurement of all the protonated molecules and subsequent fragment ions offers higher quality structural information for the interpretation of fragmentation pathways of the various groups of alkaloids. A total of 19 oxindole alkaloids, 16 indole alkaloids and 1 flavone were identified by co-chromatography of the sample extract with authentic standards, comparison of the retention time, characteristic molecular ions and fragment ions, or were tentatively identified by MS/MS determination. Moreover, the method was validated for the simultaneous quantification of the 24 components within 10.5 min. The potential chemical markers were identified for classification of the U. species samples by principal component analysis (PCA) and orthogonal partial least squared discriminant analysis (OPLS-DA). The results demonstrate the similarity and differences in alkaloids among the five U. species, which is helpful for the standardization and quality control of the medical materials of the U. Ramulus Cum Unics (URCU). Furthermore, with multivariate statistical analysis, the determined markers are more definite and useful for chemotaxonomy of the U. genus.

Some plants of the genus Uncaria (U.) are commonly used for medicinal purposes in Asia.¹⁾ U. rhynchophylla (MIQ.) MIQ. ex HAVIL. (URM), U. macrophylla WALL. (UMW), U. hirsuta HAVIL. (UHH), U. sinensis (OLIV.) HAVIL. (USH), and U. sessilifructus ROXB. (USR) are five commonly used medicinal species in China. Uncariae Ramulus Cum Uncis (URCU), the dried hook-bearing branches of the five plants, is a well known herbal medicine mentioned in Chinese Pharmacopoeias,²⁾ and has been used to treat hypertension, fever, headache, dizziness, stroke and bilious disorders.³⁾ The phytopharmaceuticals containing these herbal plant, such as Tianmagouteng Granules, Yokukansan, and Chotosan, have become better selling herbal medicines for therapy of stroke, hypertension, chronic headache and dizziness in China and Japan.⁴⁾ It was known that oxindole alkaloids and indole alkaloids are the major bioactive components in the U. species, and should be responsible for these aforementioned pharmacology activities.^5,6)

The current study is focusing on the five U. plants because there is a potential issue in relation to the efficacy, quality control and safety of URCU. In Japanese Pharmacopoeias, the URCU was defined as hook or the hook-bearing stem of the three plants, URM, UMW, and USH.⁷⁾ And many other U. species have also been circulated in local herbal market in China as the substitudes. Moreover, the phytopharmaceutical studies demonstrated that bioactivities of the different alkaloids in these plants were significant deviation.⁸⁾ For example, isorhynchophylline (IRN) has strong neuroprotective effect⁹⁾; geissoschizine methyl ether (GME) showed greater effects on vascular responses than hirsutine (HS)¹⁰⁾; strictosidine (S) exhibited potent inhibitory activity on lipopolysaccharide (LPS)-induced nitric oxide (NO) release; only weak inhibitory activities were observed for 11-hydroxy-2′-O-D-glucopyranosyl vincoside lactam (HGVL), dihydrocorynantheine (DC), and corynoxeine-N-oxide (COR-NO).¹¹⁾ Thus, chemical profiling and quantification of alkaloids are necessary for the quality control of the raw materials of URCU as well.

Several holistic chemical profiling methods of the URM have been reported, such as the screening and identification of six oxindole alkaloids and four indole alkaloids in various parts of URM by HPLC,¹²⁾ the quantification of ten oxindole alkaloids and four glycosidic alkaloids using HPLC-UV and LC-MS technologies,¹³⁾ and the profiling of the alkaloids using HPLC-DAD-quadrupole time-of-flight mass spectrometry (Q-TOF/MS) method.¹⁴⁾ However, in these literature, the identification or quantification of the URCU from other four botanical origins were still not fully understood. Therefore, it is important to develop a sensitive and selective method to accurately detect the presence and contents of the bioactive alkaloids to evaluate the biodiversities of the five U. species.

Compared to HPLC, ultra performance liquid chromatography (UPLC) can provide a high peak capacity, greater resolution, increased sensitivity, and higher speed of analysis.^15,16) And UPLC-TOF/MS techniques has also been an effective approach for rapid identification and quantification of multi-components of herbal medicines.^17,18) For example, Zhou et al. simultaneously identified 21 diterpenoids in 12 Salvia species within 10 min by UPLC-Q-TOF/MS.¹⁹⁾ Lu et al. established an UPLC-Q-TOF/MS method for the rapid analysis of isoflavones, saponins and flavones in 16 samples originated from Pueraria lobata and P. thomsonii.²⁰⁾

In the present study, we developed an UPLC-Q-TOF/MS method for the simultaneously qualitative and quantitative analyses of alkaloids in 22 samples collected from various sources for the evaluation of the chemical consistency of the five U. species plants.

Experimental

Chemicals and Reagents

In the present study, a total of 24 compounds which were isolated from leaves and hook-bearing branches of URM in our previous studies^17,18) were used as reference compounds, and the chemical structures were shown in Fig. 1. They included 22-O-β-D-glucopyranosyl isocorynoxeinic acid (GICEA, 1), 18,19-dehydrocorynoxinic acid (DCA, 3), 18,19-dehydrocorynoxinic acid B (DCAB, 7), cadambine (CDB, 8), 3α-dihydrocadambine (DHC, 9), 3β-isodihydrocadambine (IDC, 10), 11-hydroxy-2-O-D-glucopyranosyl vincoside lactam (HGVL, 12), isocorynoxeine (ICO, 14), strictosidine (S, 15), isorhynchophylline (IRN, 16), corynoxeine (COR, 17), corynoxine (CNX, 18), corynoxeine N-oxide (COR-NO, 19), corynoxine B (CNXB, 20), isocorynoxeine N-oxide (ICO-NO, 22), rhynchophylline (RIN, 24), rhynchophylline N-oxide (RIN-NO, 25), isorhynchophylline N-oxide (IRN-NO, 26), dihydrocorynantheine (DC, 28), strictosamide (SCS, 30), geissoschizine methyl ether (GME, 33), hirsuteine (HST, 34), vincoside lactam (VL, 35), and hirsutine (HS, 36). The identity of these compounds were confirmed by melting point, IR, UV, ¹H- and ¹³C-NMR, CD and MS, and their purities evaluated with HPLC-photodiode array detector (PDA) were more than 95%.

Fig. 1. Chemical Structures of Compounds Identified in U. rhynchophylla

Acetonitrile (ACN, HPLC-MS grade) and formic acid (spectroscopy grade) were purchased from Fisher Scientific U.K. (Loughborough, U.K.). Ultrapure water (18.2 MΩ) daily prepared with a Milli-Q water purification system (Millipore, Bedford, MA, U.S.A.). Leucine–enkephalin was obtained from Sigma-Aldrich (St. Louis, MO, U.S.A.). All other chemical reagents were HPLC grade.

Plant Materials

The 22 samples of the five U. species were collected from different provinces of China (as shown in Table 1). The voucher samples were authenticated by Professor Weining Wang (Liaoning Institute for Food and Drug Control, China) and Associate Professor Jiankui Zhang (Liaoning University of Traditional Chinese Medicine, China). They were divided into URM, UMW, UHH, USR, and USH according to the shape of the stem, hook, leaf and peraphyllum, proportion of the epidermis, xylem to pith, and the presence of the trichome. The voucher specimens were kept in the reference library for the medicinal herbs in Shenyang Pharmaceutical University.

Table 1. Collected Herbal Materials of Five Uncaria Species

Sample no.	Species	Location of collection	Parts
1	U. rhynchophylla (MIQ.) MIQ. ex HAVIL.	Guilin Guangxi	stems with hooks
2	〃	Jiujiang Jiangxi	stems with hooks
3	〃	Beijing*	fine powder
4	〃	Jingxian Anhui	stems with hooks
5	〃	Yizhou Guangxi	stems with hooks
6	〃	Yunnan	most stems
7	〃	Laibin Jiangxi	stems with hooks
8	〃	Shaxian Fujian	stems with hooks
9	〃	Tongshan Hubei	stems with hooks
10	〃	Hongjiang Hunan	stems with hooks
11	〃	Guilin Guangxi	stems with hooks
12	〃	Zhejiang	most stems
13	〃	Chongqing	stems with hooks
14	U. macrophylla WALL.	Nanning Guangxi	stems with hooks
15	〃	Nanning Guangxi	stems with hooks
16	U. hirsuta HAVIL.	Nanning Guangxi	stems with hooks
17	〃	Yongjia Zhejiang	stems with hooks
18	U. sessilifructus ROXB.	Nanning Guangxi	stems with hooks
19	〃	Jianhe Guizhou	stems with hooks
20	〃	Mengzi Yunnan	stems with hooks
21	〃	Wuzhou Guangxi	stems with hooks
22	U. sinensis (OLIV.) HAVIL.	Guiyang Guizhou	stems with hooks

*: Purchased from the National Institutes for Food and Drug Control, Beijing, China.

Instrumentation and the Chromatographic Conditions

UPLC analysis was performed using a Waters Acquity UPLC system (Waters, Milford, MA, U.S.A.), equipped with a binary solvent delivery system, an autosampler, and a photodiode array detection system. Separation was performed on a Waters ACQUITY BEH C₁₈ column (2.1×100 mm, 1.7 µm, Waters) at 40°C. The mobile phase consisted of water containing 0.2% formic acid (v/v, A) and acetonitrile containing 0.2% formic acid (v/v, B) with gradient elution (linear gradient from 12% B to 22% B in 8 min, followed by linear gradient from 22% B to 40% B between 8 and 10.5 min). Re-equilibration duration was 2 min between individual runs. The flow rate was 0.6 mL/min and 2 µL samples of standard and sample solution were injected in each run.

Identification of the marker compounds was performed on Waters Q-TOF Xevo G2 instrument (Waters Corp., Manchester, U.K.) equipped with an electrospray ionization (ESI) source, which provides a resolution of 10000 (FWHM) and a mass accuracy error less than 5 ppm. Leucine–enkephalin was used as the lock mass compound to generate an [M+H]⁺ ion (m/z 556.2771) in the LockSpray mode at a concentration of 50 pg/µL at an infusion flow rate of 10 µL/min. The ESI source was operated in positive ionization mode with the capillary voltage of 3.0 kV and the cone voltage was set to 45 V. Source and desolvation temperatures were set at 150 and 450°C, respectively. The cone and nebulization gas flows were 50 and 800 L/h, respectively. All data collected in centroid mode were acquired using Masslynx™ NT 4.1 software (Waters).

The MS^E (E represents collision energy) scanning experiment was carried out as follows: function 1: m/z 100–1000, 0.5 s scan time, 0.02 s inter-scan delay, 6 V collision energy, and function 2: m/z 100–1000, 0.5 s scan time, 0.02 s inter-scan delay, collision energy ramp of 20–50 V. In MS^E scanning, parallel alternate scans are acquired both at low-collision and high-collision energy to obtain the precursor ion information and full-scan mass fragment information in a single analytical run.

Preparation of the Standard Solutions

Twenty-four reference compounds (peaks 1, 3, 7–10, 12, 14–20, 22, 24–26, 28, 30, 33–36) were accurately weighed, and dissolved in methanol to give individual stock solutions with concentrations ranging from 1.0 to 3.5 mg/mL. Series of working standard solutions were prepared by appropriate dilution of stock solutions with methanol in order to prepare calibrators. All solutions were stored at 4°C in a refrigerator before analysis.

Sample Preparation

Powdered herbal materials (0.2 g, passed through a 500 µm mesh sieve) were ultrasonically double-extracted with 10 mL MeOH–H₂O (75 : 25) in an ultrasonic water bath for 30 min. The mixtures were centrifuged at 3000×g for 10 min, and the supernatants were diluted to 25 mL with methanol in a volumetric flask. An aliquot of each 2 µL filtrate filtered through a 0.22 µm PTFE syringe filter was injected into the UPLC instrument for analysis.

Method Validation for Quantification

Among the 35 identified alkaloids, 24 alkaloids were quantified on an UPLC-Q-TOF/MS instrument using quasi-molecular ion chromatograms (EICs, extracted ion chromatograms, with a 0.02 Da window), which peak area was integrated at the expected retention times under full-scan MS conditions

Calibration Curves

Calibration curves (5–8 point) were obtained using external standard calibrations for 24 analytes injecting each solution, and then the calibration curves were constructed by plotting the peak area versus the concentration of each analyte.

LODs and LOQs

The stock solutions containing 24 reference compounds were diluted to a range from 0.01 ng/mL to 0.1 µg/mL and the injection volume was 2 µL. The limits of detection (LOD) and quantification (LOQ) under the present chromatographic conditions were determined at a signal-to-noise ratio (S/N) of 3 and 10, respectively.

Precision, Accuracy and Repeatability

Intra-day and inter-day precision were evaluated by analyzing the known concentrations of the standard stock solutions in six replicates during a single day and the experiments involving triplicates on the successive days. Six different solutions prepared from the same sample were analyzed to confirm the repeatability of the developed assay. Variations were expressed as relative standard deviations (RSD). The recovery was used to evaluate the accuracy of the method. A known amount of standards was added into a certain amount of URM (0.1 g, sample 3). The mixture extracted and analyzed using the method was mentioned above.

Chemometric Data Analysis

The UPLC-MS data of the five U. species samples were analyzed by Markerlynx Applications Manager (V4.1, Waters). The parameters were set as following, retention time (t_R) range 0.3–10.5 min, mass range m/z 150–1000 Da, t_R tolerance 0.1 min, mass tolerance 0.05 Da, width of an average peak at 5% height and peak-to-peak baseline noise were automatically calculated, marker intensity threshold 10.0, noise elimination level 6.0, and isotopic peaks were excluded for analysis.

Results and Discussion

Optimization of Extraction Procedure

In order to obtain a satisfactory extraction efficiency for all analytes, solvents (50%, 75%, and 100% methanol), and time (15 and 30 min) were assessed based on a single factor experiment. The best extraction efficiency was obtained by the sonication extraction using the 75% methanol for 30 min.

Optimization of Chromatographic Separation

A mixed solution of the 24 alkaloid standards and extracts of URCU were used for the optimization of UPLC conditions, the representative UPLC-MS chromatograms of which are presented in Fig. 2. The chromatographic behavior varied with the concentrations of the formic acid (0.0%, 0.1% and 0.2%) and different mobile phase (ACN, MeOH) carried by C₁₈ column. The results obtained showed that the separations on the C₁₈ column using the gradient solvent systems consisting of water (containing 0.2% formic acid, A) and ACN (containing 0.2% formic acid, B), were better than other conditions tested. Using the optimal gradient elution as described in the chromatographic conditions of the Experimental section, the 24 standard compounds were satisfactorily separated within 10.5 min.

Fig. 2. Representative Total Ion Current (TIC) of Standard Solution and Test Solution of URCU for Identification

(A) TIC of ESI-MS of 24 standard solution; (B–F) TIC of ESI-MS from U. rhynchophylla (MIQ.) MIQ. ex HAVIL., U. macrophylla WALL., U. hirsuta HAVIL., U. sinensis (OLIV.) HAVIL. and U. sessilifructus ROXB.

In the UPLC system with mobile phase of acetonitrile/water containing 0.2% formic acid, the t_R of the rhynchophylline-type alkaloids with C₇-S configuration, such as DCA, CNX, ICO, IRN, was shorter than that with C₇-R configuration, such as DCAB, CNXB, COR, RIN. The t_R of the alkaloids with H₃-α configuration, such as SCS and DC, was shorter than that with H₃-β configuration, such as VL and HS. That with the α-configuration of H₂₀, such as IRN and RIN, was shorter than that with the β-configuration of H₂₀, such as CNX and CNXB.

Mass Spectrometry Analysis of the Known U. Alkaloids

The characteristic diagnostic fragmentation ions (CFIs) may be used to rapidly identify the class of alkaloids with similar structures of the U. species. In order to obtain CFIs of these main bioactive compounds, twelve typical alkaloid standards (peaks 3, 7, 14, 16–20, 22, 24, and 25, see Fig. 1) were selected as model compounds, and subjected to fragmentation investigation by UPLC-Q-TOF-MS with MS^E scanning experiments.

As can be seen from Table 2 and Fig. 3, the characteristic fragmentation ions of TMOAs were produced by the indole ring skeleton. In positive MS mode, some CFIs for TMOAs can be delineated on the basis of the mass fragmentation pathways and the intensities of the ion peaks, including the quasi-molecular ions [MH]⁺, [MH−CH₄O]⁺ (C₂₂), [MH−C₂H₈O₂]⁺ (C₁₇), [M−C₅H₈O₃]⁺ (C₁₅), [MH−C₇H₁₀O₃]⁺ (C₂₁, 241.13), [M−C₁₀H₁₄O₃]⁺ (C₃–N₋₄), and [MH−C₁₂H₁₇NO₃]⁺ (C₂–C₅) (as shown in Fig. 3). For instance, ions at m/z 353.19, 321.16, 269.17, 241.13, 160.08 could be used as CFIs for RIN.

Table 2. Compounds Identified in URCU Samples by UPLC-Q-TOF/MS Method

Peak no.	Identified compound	tR (min)	[M+H]+ (m/z)			Molecular formula	Characteristic fragmentation ions (m/z, int.2%)	Comparison with standards
			Detected mass (m/z, int.1%)	Calculated mass (m/z)	Error (ppm)
1	22-O-β-D-Glucopyranosyl isocorynoxeinic acid	0.90	531.2341(4)	531.2343	−0.4	C27H34N2O9	369.18(100)/351.17(17)/319.14(5)/267.15(19)/201.10(6)/160.08(11)	yes
2	Isomer of 22-O-β-D-glucopyranosyl isocorynoxeinic acid	1.80	531.2353(15)	531.2343	1.9	C27H34N2O9	369.18(33)/351.17(4)//267.15(3)/201.10(2)/160.08(3)	no
3	18,19-Dehydrocorynoxinic acid	2.16	369.1808(49)	369.1814	−1.6	C21H24N2O4	369.18(8)/351.17(2)/319.14(6)/267.15(30)/241.13(2)/201.10(21)/160.08(100)	yes
4	Isomer of 11-hydroxy-2-O-D-glucopyranosyl vincoside lactam	2.67	677.2552(4)	677.2558	−0.9	C32H40N2O14	515.21(4)/353.15(4)	no
5	Hyperoside	3.14	465.1049(9)	465.1033	3.4	C21H20O12	465.10/(8)303.05(100)	no
6	Mitraphyllic acid 22-β-D-glucopyranosyl ester	3.37	517.2182(15)	517.2186	−0.8	C26H32N2O9	517.22(15)/201.08(3)	no
7	18,19-Dehydrocorynoxinic acid B	3.58	369.1811(100)	369.1814	−0.8	C21H24N2O4	369.18(6)/351.17(2)/319.14(2)/267.15(7)/241.13(1)/201.10(3)/160.08(100)	yes
8	Cadambine	3.60	545.2139(39)	545.2135	0.7	C27H32N2O10	545.21(15)/383.1632(79)	yes
9	3α-Dihydrocadambine	4.17	547.2295(56)	547.2292	0.5	C27H34N2O10	547.23(13)/385.17(40)/367.17(18)	yes
10	3β-Isodihydro-cadambine	4.36	547.2284(100)	547.2292	−1.5	C27H34N2O10	547.23(13)/385.17(39)/367.17(100)	yes
11	Isomer of isocorynoxeine	4.93	383.1970(100)	383.1971	−0.3	C22H26N2O4	383.20(4)/351.17(2)/319.14(1)/267.15(40)/241.13(100)/224.13(1)/201.08(1)/160.08(53)	no
12	11-Hydroxy-2-O-D-glucopyranosyl vincoside lactam	5.07	677.2563(79)	677.2558	0.7	C32H40N2O14	515.20(6)/353.15(100)	yes
13	Isomer of strictosidine	5.39	531.2354(13)	531.2343	2.1	C27H34N2O9	514.24(3)/369.18(2)/337.15(12)	no
14	Isocorynoxeine	5.46	383.1964(100)	383.1971	−1.8	C22H26N2O4	383.20(10)/351.17(8)/319.14(7)/267.15(37)/241.13(3)/224.13(3)/201.10(30)/160.08(100)	yes
15	Strictosidine	5.68	531.2346(100)	531.2343	0.6	C27H34N2O9	531.23 (17)/514.21(2)/369.18(7)/337.12(3)	yes
16	Isorhynchophylline	6.18	385.2132(100)	385.2127	1.3	C22H28N2O4	385.21(21)/353.19(30)/321.16(8)/269.16(12)/241.13(100)/226.14(3)/160.08(79)	yes
17	Corynoxeine	6.25	383.1968(100)	383.1971	−0.8	C22H26N2O4	383.20(6)/351.17(3)/319.14(2)/267.15(13)/241.13(6)/224.13(4)/201.10(3)/160.08(100)	yes
18	Corynoxine	6.37	385.2132(100)	385.2127	1.3	C22H28N2O4	385.21(19)/353.19(14)/321.16(3)/269.16(6)/241.13(100)/160.08(46)	yes
19	Corynoxeine-NO	6.47	399.1924(100)	399.1920	1.0	C22H26N2O5	399.19(40)/381.18(14)/367.17(4)/355.20(4)/265.13(10)/222.11(30)/206.12(26)/178.12(13)/160.08(100)	yes
20	Corynoxine B	6.55	385.2132(100)	385.2127	1.3	C22H28N2O4	385.21(16)/353.19(8)/321.16(2)/269.17(16)/241.13(21)/226.14(7)/160.08(68)	yes
21	Isocorynoxine B	6.77	385.2130(17)	385.2127	0.8	C22H28N2O4	385.21(29)/353.19(2)/321.16(2)/269.17(5)/160.08(68)	no
22	Isocorynoxeine-NO	6.85	399.1924(100)	399.1920	1.0	C22H26N2O5	399.19(30)/381.18(8)/367.17(1)/355.20(1)/265.13(13)/222.11(30)/206.12(31)/178.12(13)/160.08(56)	yes
23	Mitragynine	6.99	399.2270(100)	399.2284	−3.5	C23H30N2O4	399.23(100)/383.20(15)/367.20(1)/355.20(1)/269.17(5)/201.10(17)/160.08(28)	no
24	Rhynchophylline	7.32	385.2128(100)	385.2127	0.3	C22H28N2O4	385.21(14)/353.19(6)/321.16(1)/269.16(19)/241.13(3)/226.14(6)/160.08(100)	yes
25	Rhynchophylline-NO	7.52	401.2080(100)	401.2076	1.0	C22H28N2O5	401.21(27)/383.19(13)/369.18(5)/357.21(2)/267.15(24)/224.12(33)/208.13(62)/178.09(17)/160.08(90)	yes
26	Isorhynchophylline-NO	7.70	401.2071(100)	401.2076	−1.2	C22H28N2O5	401.21(25)/383.19(13)/369.18(5)/357.21(2)/267.15(11)/224.12(30)/208.13(37)/178.09(10)/160.08(22)	yes
27	Isomer of rhynchophylline	8.06	385.2126(48)	385.2127	−0.3	C22H28N2O4	385.21(8)/353.19(5)/321.16(2)/269.16(9)/241.13(13)/226.14(5)/160.08(52)	no
28	Dihydrocorynantheine	9.27	369.2177(100)	369.2178	−0.3	C22H28N2O3	369.22(3)/337.19(1)/170.10(2)/144.08(100)	yes
29	Raubasine	9.30	353.1861(23)	353.1865	−1.1	C21H24N2O3	353.19(3)/321.16(2)	no
30	Strictosamide	9.32	499.2067(22)	499.2080	−2.6	C26H30N2O8	499.21(2)/337.15(7)/267.11(20)	yes
31	Isomer of vincoside lactam	9.60	499.2081(25)	499.2080	0.2	C26H30N2O8	499.21(3)/337.15(9)/267.11(16)/171.09(73)	no
32	Yohimbine	9.61	355.2021(68)	355.2022	−0.3	C21H26N2O3	355.20(4)/337.20(2)/323.18(1)/225.14(6)/170.10(15)	no
33	Geissoschizine methyl ether	9.71	367.2028(100)	367.2022	1.6	C22H26N2O3	367.20(7)/335.18(1)/209.15(1)/170.10(6)/144.08(100)	yes
34	Hirsuteine	9.97	367.2024(100)	367.2022	0.5	C22H26N2O3	367.20(17)/335.18(3)/209.15(1)/170.10(32)/144.08(100)	yes
35	Vincoside lactam	10.06	499.2083(100)	499.2080	0.6	C26H30N2O8	499.21(2)/337.15(9)/267.11(39)/171.09(100)	yes
36	Hirsutine	10.22	369.2185(100)	369.2178	0.8	C22H28N2O3	369.21(9)/337.19(4)/170.10(21)/144.08(100)	yes

int.₁, The signal intensity data of function 1 in the UPLC/ESI(+)-TOF-MS^E. int.₂, The signal intensity data of function 2 in the UPLC/ESI(+)-TOF-MS^E.

By careful analysis of the mass spectra of the model compounds, it was conformed that ions at m/z 201.08 could be used as CFIs for TMOAs with a vinyl group at C₂₀ (see Table 2) for Mclafferty rearrangement. Opposite to the literature,¹⁴⁾ the ion at m/z 241.12 was found in the mass spectra of eight components and can be used as CFIs for TMOAs, but not for 7S/R configuration and ethyl/vinyl substituent at C₂₀.

Mass Spectrometry Analysis of the 36 Marker Compounds in the Five U. Species Plants

In order to characterize the chemical composition, the mixture solution containing 24 reference alkaloids (see Fig. 1), and the extracts of URM (sample 1) and UMW (sample 14) were subjected to UPLC-ESI-Q-TOF/MS analysis. The identified or tentatively identified compounds were satisfactorily separated under the UPLC-PDA and/or UPLC-MS conditions. Thirty-six specific peaks (labeled peaks 1–36, Fig. 2) in the UPLC chromatograms were characterized by the typical UV absorptions obtained with Waters PDA detector, with absorptions displaying from 209 to 212 nm, 239 to 243 nm, and 278 to 292 nm for oxindole alkaloids, 220 to 225 nm and 276 to 285 nm for indole alkaloids, and 263 to 293 nm and 357 to 362 nm for flavones. By co-chromatography of the sample extracts with the authentic standards, and comparing the retention times, UV spectra, characteristic molecular ions and fragment ions (as shown in Table 2) with those of the authentic standards, the compounds corresponding to the 24 peaks were identified as GICEA (peak 1), DCA (peak 3), DCAB (peak 7), CDB (peak 8), DHC (peak 9), IDC (peak 10), HGVL (peak 12), ICO (peak 14), S (peak 15), IRN (peak 16), COR (peak 17), CNX (peak 18), COR-NO (peak 19), CNXB (peak 20), ICO-NO (peak 22), RIN (peak 24), RIN-NO (peak 25), IRN-NO (peak 26), DC (peak 28), SCS (peak 30), GME (peak 33), HST (peak 34), VL (peak 35), and HS (peak 36).

Due to the absence of reference compounds, the compounds corresponding to the rest 12 peaks were tentatively identified by MS/MS determination, the characteristic UV absorption spectra, and comparison with literature data (see Table 2). The isomer of 22-O-β-D-glucopyranosyl isocorynoxeinic acid (peak 2) are the representative for the structural identification of the oxindole alkaloids and indole alkaloids from U. species samples in the UPLC-Q-TOF-MS determination.

Peak 2 displayed an accurate quasi-molecular ion [M+H]⁺ at m/z 531.2353, corresponding to the molecular formula C₂₇H₃₄N₂O₉, identical to GICEA. The appearance of CFIs, at m/z 369.18, 351.17, 319.14, 267.15, 201.10 and 160.08, was almost the same with GICEA. This was verified by studying the fragmentation pattern of peak 2 and it was compared with GICEA (see Fig. 3). For instance, the MS/MS spectrum of peak 2 contained major fragment ions at m/z 369.18, formed by the loss of C₆H₁₀O₅ (162 Da) [M+H−Glc]⁺. The fragment ion at m/z 351.17 was formed from the precursor ion at m/z 369.18 via the loss of H₂O (18 Da). The fragment ions at m/z 319.14 and 267.15 were from the precursor ion 351.17 via the loss of CH₃OH (32 Da) and C₄H₄O₂ (84 Da). The fragment ion 201.10 was formed from the precursor ion at m/z 267.15 via the Mclafferty rearrangement, loss of C₅H₆ (66 Da). The fragment ion 160.08 was formed from the precursor ion at m/z 319.17 via loss of C₁₀H₉NO (159 Da). Based on the structural data of the alkaloids found from U. spices, the structural differences between these tetracyclic monoterpenoid oxindole alkaloids mainly involved the configurations at C₇ (-R or -S). Under our UPLC conditions, the t_R of configurations at C₇-S (DCA, CNX, ICO, and IRN) is shorter than C₇-R (DCAB, CNXB, COR and RIN), similar to peak 2 and GICEA. According to the structural configuration data and t_R, the compound corresponding to peak 2, together with the CFIs MS/MS data, was tentatively identified as the isomer of 22-O-β-D-glucopyranosyl corynoxeinic acid (R configuration at C₇).

Fig. 3. (A) Partial Fragmentation Pathways and Characteristic Fragmentation Ions of TMOAs; (B) Fragmentation Pathways of GICEA; (C) MS^E Spectra of GICEA; (D) MS^E Spectra of iso-GICEA

Using the similar methods described above, 5 oxindole alkaloids (peaks 2, 6, 11, 21, and 27), 6 indole alkaloids (peaks 4, 13, 23, 29, 31, and 32) and 1 flavone (peak 5) were tentatively identified, as listed in Table 2. However, the position isomerization of the 5 isomers (peaks 4, 11, 13, 27, and 31) cannot be confirmed only based on their MS spectra alone.

Method Validation

All calibration graphs were plotted based on linear regression analysis of the integrated peak areas (y) versus concentrations (x, µg/mL) of the 24 markers in the standard solution at different concentrations. The results are shown in Supplemental Data S1. The sensitivity of the method was evaluated by determining the limits of detection (LODs) and limits of quantitation (LOQs). These parameters were determined by triplicate analysis of a series of decreasing concentrations of standard solution. As shown in Supplemental Data S1, the developed method was very sensitive with most LODs and LOQs of no more than 0.03 ng/mL and 0.1 ng/mL, respectively. Method precision was checked by intra-day and inter-day variability by six consecutive times. The relative standard deviation (RSD) was taken as a measure of the precision and the results are summarized in Supplemental Data S1. From the results obtained, the developed method was found to be precise, with the intra-day variability RSD values between 1.0% and 3.6% and the inter-day variability RSD values between 3.2% and 7.6%. A repeatability test with RSD values between 2.2% and 7.3% suggested that the method was reproducible.

Recovery was carried out by spiking accurate amounts of the 24 standards into sample 3, and then extracting and analyzing them under this proposed method. Each sample was analyzed in triplicate replicates. The total amount of each analysis was calculated from the corresponding calibration curve. The recovery was calculated by the formula: recovery (%)=[(measured amount−original amount)/amount of spiked reference standard]×100. The recovery of the method was in the range of 90.1–108.9%, with RSD less than 7.6%.

Quantification of Oxindole Alkaloids and Indole Alkaloids in the Five U. Species Plants

Chemical profiling and quantification of the oxindole alkaloids and indole alkaloids in the stems with hooks of the five U. species using the optimized UPLC-Q-TOF/MS method were carried out (as shown in Fig. 2 and Table 3). There is a significant difference in the chemical profiling patterns among the five U. species plants. Indole alkaloids, GME (trace–7.68 mg/g), IDC (trace–6.12 mg/g), HS (trace–4.48 mg/g), HST (trace–3.72 mg/g), SCS (trace–2.68 mg/g) and S (0.256–1.01 mg/g), were abundant in the U. species samples. Oxindole alkaloids, RIN (trace–0.739 mg/g), ICO (trace–0.454 mg/g), IRN (trace–0.296 mg/g), and COR (trace–0.232 mg/g) were found in relatively higher amount. N-oxide oxindole alkaloids, COR-NO, ICO-NO, RIN-NO, and IRN-NO were low or trace in U. species samples.

Table 3. Content (mg/g) of Twenty-Four Alkaloids in Materials of Five Uncaria Species

Samplesno.	Analytes
	1	3	7	8	9	10	12	14	15	16	17	18	19	20	22	24	25	26	28	30	33	34	35	36
1	8.45×10−3	2.80×10−3	5.68×10−3	0.110	4.63×10−3	5.13	5.23×10−3	0.227	0.256	0.168	8.20×10−2	nd	2.42×10−3	nd	1.19×10−3	0.185	1.09×10−3	2.05×10−3	2.52×10−3	1.47	0.430	0.98	1.12×10−2	1.37
2	1.42×10−2	1.08×10−3	2.41×10−3	7.02×10−3	nd	0.246	2.20×10−2	0.488	0.409	0.246	0.178	nd	6.02×10−3	nd	3.04×10−3	0.287	2.04×10−3	3.26×10−3	3.57×10−3	0.317	0.177	9.9×10−2	0.107	0.101
3	7.36×10−3	3.55×10−3	1.37×10−2	0.186	1.48×10−2	2.21	1.32×10−2	7.28×10−2	0.757	5.52×10−2	7.30×10−2	3.58×10−3	7.46×10−3	1.42×10−3	4.96×10−3	0.165	4.44×10−3	8.18×10−3	2.65×10−3	2.68	2.44	1.21	4.93×10−2	1.28
4	5.47×10−2	5.27×10−3	4.60×10−3	1.20×10−2	nd	0.452	4.36×10−2	0.375	1.01	0.250	0.232	nd	1.69×10−3	nd	4.92×10−3	0.310	3.23×10−3	7.60×10−4	6.50×10−4	1.67	0.126	0.115	2.39×10−2	0.120
5	2.23×10−3	nd	9.40×10−4	4.62×10−2	9.52×10−3	5.07	nd	1.59×10−2	0.381	2.60×10−2	8.22×10−3	nd	5.30×10−4	nd	3.70×10−4	3.58×10−2	2.30×10−4	nd	2.10×10−4	0.142	1.61	1.90	9.82×10−3	3.65
6	7.20×10−4	5.10×10−4	1.53×10−3	8.19×10−2	4.03×10−2	5.57	7.32×10−3	1.16×10−2	0.367	1.17×10−2	1.34×10−2	7.20×10−4	nd	nd	3.60×10−4	3.99×10−2	nd	nd	3.60×10−3	0.347	3.53	1.23	2.04×10−2	1.59
7	9.08×10−3	6.43×10−3	8.26×10−3	1.63×10−2	nd	0.769	5.22×10−2	0.283	0.544	0.182	9.6×10−2	1.24×10−3	1.08×10−3	1.18×10−3	9.00×10−4	0.175	5.90×10−4	5.50×10−4	9.59×10−3	0.475	0.725	0.748	4.17×10−2	0.898
8	2.50×10−2	1.36×10−2	6.69×10−3	1.99×10−2	6.00×10−4	0.635	1.26×10−2	0.454	0.437	0.264	0.151	nd	4.66×10−3	1.19×10−3	8.49×10−3	0.238	6.30×10−3	4.01×10−3	3.49×10−3	0.334	0.270	4.08×10−2	8.29×10−3	3.65×10−2
9	1.46×10−3	6.90×10−4	1.97×10−3	7.04×10−3	1.06×10−3	0.212	8.45×10−3	0.436	0.573	0.296	0.138	nd	1.62×10−3	nd	3.55×10−3	0.245	2.48×10−3	4.30×10−4	1.87×10−3	0.562	0.305	3.37×10−2	6.72×10−3	3.86×10−2
10	2.31×10−2	4.20×10−4	1.23×10−3	0.130	3.31×10−2	6.12	8.67×10−3	4.19×10−3	0.320	4.49×10−3	1.41×10−3	3.00×10−4	nd	2.60×10−4	nd	4.05×10−3	nd	nd	1.49×10−3	7.98×10−2	2.42	1.37	5.70×10−3	1.61
11	2.60×10−3	2.41×10−3	4.41×10−3	6.76×10−2	1.22×10−3	1.08	7.31×10−3	3.12×10−2	0.503	4.06×10−2	1.14×10−2	1.08×10−3	1.00×10−3	3.03×10−3	1.00×10−3	3.35×10−2	2.12×10−3	4.80×10−4	8.75×10−3	0.554	5.24	3.72	0.187	4.48
12	5.29×10−3	6.01×10−3	1.88×10−2	1.71×10−2	nd	0.552	7.45×10−3	0.155	0.389	6.55×10−2	7.07×10−2	8.70×10−4	2.53×10−3	nd	1.89×10−3	0.114	8.50×10−4	1.19×10−3	1.51×10−3	1.59	2.32×10−2	2.43×10−2	1.92×10−2	2.38×10−2
13	8.20×10−4	nd	1.14×10−3	1.93×10−2	6.40×10−4	0.581	nd	1.78×10−3	0.423	1.93×10−3	1.40×10−4	1.56×10−3	nd	8.30×10−4	nd	4.42×10−3	1.56×10−3	nd	7.33×10−3	0.423	7.68	1.43	1.38×10−2	1.35
14	nd	3.70×10−4	9.90×10−4	nd	6.50×10−4	5.92×10−2	nd	2.13×10−3	0.303	0.124	1.22×10−2	0.394	nd	3.17×10−2	nd	0.739	2.10×10−3	4.35×10−3	8.28×10−3	nd	7.86×10−2	0.120	nd	0.100
15	nd	nd	nd	nd	nd	nd	nd	3.09×10−3	0.292	0.119	1.90×10−2	0.366	6.40×10−4	0.197	5.20×10−4	0.857	6.79×10−2	1.56×10−2	2.84×10−3	6.04×10−2	6.10×10−4	9.9×10−2	nd	5.65×10−3
16	nd	nd	8.80×10−4	nd	nd	nd	nd	1.30×10−4	0.590	2.04×10−3	nd	nd	nd	4.90×10−4	nd	7.77×10−3	nd	nd	nd	0.772	4.00×10−5	2.15×10−3	3.37×10−2	1.01×10−3
17	3.12×10−3	1.41×10−3	2.05×10−3	3.06×10−2	4.80×10−4	0.585	nd	0.163	0.400	0.110	6.58×10−2	nd	1.86×10−3	nd	3.48×10−3	0.136	2.13×10−3	9.50×10−4	9.80×10−4	0.295	0.181	0.638	2.08×10−2	0.549
18	nd	nd	nd	nd	nd	nd	nd	1.17×10−2	0.259	0.246	4.86×10−3	5.30×10−4	nd	5.98×10−3	nd	0.321	1.08×10−3	9.80×10−4	2.59×10−3	nd	1.09×10−2	1.87×10−2	nd	0.102
19	7.40×10−4	4.10×10−4	1.31×10−3	0.202	3.96×10−2	5.40	nd	7.87×10−3	0.362	3.36×10−3	2.83×10−3	6.80×10−4	5.30×10−4	nd	nd	5.50×10−3	1.10×10−4	nd	5.65×10−3	0.656	6.74	1.75	1.99×10−2	1.05
20	nd	nd	1.58×10−3	6.46×10−3	8.20×10−4	0.219	5.51×10−3	9.40×10−4	0.373	7.40×10−4	2.00×10−4	nd	5.20×10−4	nd	3.20×10−4	6.40×10−4	1.40×10−4	nd	5.91×10−3	0.356	3.28	0.566	3.66×10−3	0.520
21	9.00×10−4	5.60×10−4	1.91×10−3	nd	nd	2.85×10−2	nd	1.36×10−2	0.286	5.97×10−3	6.43×10−3	6.90×10−4	7.10×10−4	nd	3.30×10−4	1.02×10−2	nd	nd	nd	0.574	4.53×10−3	5.77×10−3	1.20×10−2	5.62×10−3
22	nd	nd	nd	8.11×10−3	1.15×10−3	0.349	nd	4.60×10−4	0.378	5.80×10−4	2.10×10−4	1.92×10−3	nd	1.22×10−3	5.80×10−4	1.23×10−3	nd	nd	3.00×10−5	0.713	0.603	0.136	6.17×10−3	7.40×10−2

Analytes: 1 GICEA, 3 DCA, 7 DCAB, 8 CDB, 9 DHC, 10 IDC, 12 HGVL, 14 ICO, 15 S, 16 IRN, 17 COR, 18 CNX, 19 CORNO, 20 CNXB, 22 ICONO, 24 RIN, 25 RINNO, 26 IRNNO, 28 DC, 30 SCS, 33 GME, 34 HST, 35 VL, 36 HS. nd, not detectable (<limit of detection or minimum of the linear range), as described in Experimental section of Method Validation for Quantification.

GME, IDC, HS, HST, and ICO were abundant in URM (2.32×10⁻²–7.68 mg/g for GME, 0.212–6.12 mg/g for IDC, trace–4.48 mg/g for HS, trace–3.72 mg/g for HST and trace–0.454 mg/g for ICO) than other four U. species (trace–6.74 mg/g for GME, trace–5.40 mg/g for IDC, trace–1.35 mg/g for HS, and trace–0.163 mg/g for ICO). RIN, CNX, CNXB were also riched in UMW (0.739–0.857 mg/g for RIN, 0.366–0.394 mg/g for CNX, trace–0.197 mg/g for CNXB) than other four U. species (trace–0.321 mg/g for RIN, trace for CNX, trace for CNXB).

Several studies on HPLC analysis of the URM have been reported,^12–14) and most of them focused on chemical profiling of alkaloids. However, there were large variations in the content of these bioactive alkaloids in the U. species samples according to the species and different geographic locations, which may be responsible for the different therapeutic efficacies of these herbal plants. The present study demonstrated the difference of chemical composition between the five U. species by UPLC-Q-TOF/MS analysis. The established UPLC-Q-TOF/MS profiles working in 10.5 min may be more practical for the chemical characterization of the URCM samples than the HPLC methods.

Multivariate Statistical Analysis

To further visualize the difference between the UPLC-Q-TOF-MS profiles obtained from the five U. species samples, unsupervised principal component analysis and supervised orthogonal partial least squared discriminant analysis were performed to process data and figure out the potential chemical markers for their difference.

The score plot obtained by all observations using 4781 Pareto-scaled variables is displayed in Fig. 4A. A clear separation can be seen between the five U. species samples. In the scatter-plot and S-plot, each point represents an ion t_R-m/z pair. The t_R-m/z pair points at the ends of score plot represent characteristic markers with the most confidence to each group (see Fig. 4B). Six ions, a (t_R 6.17 min, m/z 385.2120), b (t_R 7.31 min, m/z 385.2121), c (t_R 5.42 min, m/z 383.1966), d (t_R 6.27 min, m/z 383.1963), e (t_R 6.35 min, m/z 385.2135), f (t_R 8.05 min, m/z 385.2122) at the top left corner and three ions, g (t_R 9.70 min, m/z 367.2015), h (t_R 10.21 min, m/z 369.2173), i (t_R 9.96 min, m/z 367.2018) at the top right corner are the marker compounds of the U. species samples which contribute most to difference between the five sources, respectively.

To further find the most suitable chemical markers for the discrimination between URM and the other sources, the OPLS-DA was performed to generate score plot and S-plot. The VIP (variable importance in the projection) value ensured the significance of the potential markers. The marker ions, e, b, f, h, c, and d (see Fig. 4C) contribute the VIP value to difference URM and UMW are 34.33, 28.17, 22.64, 12.58, 12.24, 8.26, respectively. The marker ions, c and g (see Fig. 4D), b, h, and i (see Fig. 4E), c and h (see Fig. 4F), contribute higher value to difference URM and UHH, USH, USR, respectively.

Fig. 4. (A) PCA Scores Plot; (B) Scater-Plot of the Five Uncaria Species Samples; (C–F) PCA S-Plots of U. rhynchophylla (MIQ.) MIQ. ex HAVIL. and the Other Sources, U. macrophylla WALL., U. hirsuta HAVIL., U. Sinensis (OLIV.) HAVIL. and U. sessilifructus ROXB., Respectively

The points a–i in the Scater-plot (B) and S-plots (C–F) represent the potential chemical markers of the URCU samples.

The top nine leading markers mentioned above were structurally identified as IRN (a, 16), RIN (b, 24), ICO (c, 14), COR (d, 17), CNX (e, 18), iso-rhynchophylline (f, 27), GME (g, 33), HS (h, 27), and HST (i, 34), respectively. The results of the multivariate statistical analysis supported the quantification results in our study. Furthmore, a reliable fingerprint of plant herbs, and multivariate statistical analysis should be established on the basis of analysis of large amount of samples in the future, especially in the USH species, to comprehensively discuss the difference between the five U. species because chemical profiling in plants is subjected to influence of geographic locations as the authors described.

Conclusion

The present study is the first report on an UPLC-Q-TOF/MS method for the rapid analyses of the bioactive alkaloids in the URCU, the dried hook-bearing branches of the five U. species, URM, UMW, UHH, USH, and USR. Some fragmentation pathways of the alkaloids were proposed to identify the observed components tentatively. Moreover, a validated method was successfully applied for the simultaneous quantification of the oxindole alkaloids and indole alkaloids with good accuracy and precision. The results demonstrate the similarity and differences in alkaloids among the five U. species, which is helpful for the standardization and quality control of the medical materials of URCU. Furthermore, with multivariate statistical analysis, the determined markers are more of representative.

Acknowledgment

This research was supported by the National Natural Sciences Foundation of China (No. 81173544), and the Specialized Research Fund for the Doctoral Program of Higher Education of China (SRFDP) (20112134120010), and the Distinguished Professor Fund of Liaoning Province of 2011.

References

• 1) Heitzman M. E., Neto C. C., Winiarz E., Vaisberg A. J., Hammond G. B., Phytochemistry, 66, 5–29 (2005).
• 2) Chinese Pharmacopoeia Commission, “Pharmacopoeia of the People’s Republic of China,” Vol. 1, People’s Medical Publishing House, Beijing, 2010.
• 3) Ndagijimana A., Wang X. M., Pan G. X., Zhang F., Feng H., Olaleye O., Fitoterapia, 86, 35–47 (2013).
• 4) Sekiya N., Shimada Y., Shibahara N., Takagi S., Yokoyama K., Kasahara Y., Sakakibara I., Terasawa K., Phytomedicine, 9, 636–640 (2002).
• 5) Haginiwa J., Sakai S., Aimi N., Yamanaka E., Shinma N., Yakugaku Zasshi, 93, 448–452 (1973).
• 6) Gerhard L., Phytother. Res., 18, 259–274 (2004).
• 7) Japanese Pharmacopoeia Editorial Committee, “Japanese Pharmacopoeia,” 16th ed., Hirokawa Press, Tokyo, 2013.
• 8) Kang T. H., Murakami Y., Takayama H., Kitajima M., Aimi N., Watanabe H., Matsumoto K., Life Sci., 76, 331–343 (2004).
• 9) Shimada Y., Goto H., Itoh T., Sakakibara I., Kubo M., Sasaki H., Terasawa K., J. Pharm. Pharmacol., 51, 715–722 (1999).
• 10) Aisaka K., Hattori Y., Kihara T., Ishihara T., Endo K., Hikino H., Planta Med., 51, 424–427 (1985).
• 11) Ma B., Wu C. F., Yang J. Y., Wang R., Kano Y., Yuan D., Helv. Chim. Acta, 92, 1575–1585 (2009).
• 12) Yamanaka E., Kimizuka Y., Aimi N., Sakai S., Haginiwa J., Yakugaku Zasshi, 103, 1028–1033 (1983).
• 13) Qu J., Gong T., Ma B., Zhang L., Kano Y., Yuan D., Chem. Pharm. Bull., 60, 23–30 (2012).
• 14) Xie S. L., Shi Y. Y., Wang Y. X., Wu C. Y., Liu W. Y., Feng F., Xie N., J. Pharm. Biomed. Anal., 81–82, 56–64 (2013).
• 15) Zhou Y., Huang S. X., Pu J. X., Li J. R., Ding L. S., Chen D. F., Sun H. D., Xu H. X., J. Pharm. Biomed. Anal., 56, 916–927 (2011).
• 16) Zhou Y., Xu G., Choi F. F., Ding L. S., Han Q. B., Song J. Z., Qiao C. F., Zhao Q. S., Xu H. X., J. Chromatogr. A, 1216, 4847–4858 (2009).
• 17) Wu L., Hao H. P., Wang G. J., Curr. Drug Metab., 13, 1251–1265 (2012).
• 18) Kuang Y., Qiu F., Kong W. J., Luo J. Y., Cheng H. Y., Yang M. H., J. Chromatogr. B Analyt. Technol. Biomed. Life Sci., 939, 98–107 (2013).
• 19) Zhou Y., Lee S., Choi F. F., Xu G., Liu X., Song J. Z., Li S. L., Qiao C. F., Xu H. X., Anal. Chim. Acta, 678, 96–107 (2010).
• 20) Lu J., Xie Y. Y., Tan Y., Qu J. L., Matsuda H., Yoshikawa M., Yuan D., Chem. Pharm. Bull., 61, 941–951 (2013).