Isolation of oleanolic acid from Clematis armandii for differentiation it from adulterants by HPTLC/HPLC

Abstract

Clematis armandii has been used in many Chinese traditional patent medicines all over the world as a form of Caulis Clematidis Armandii. However, it has often been adulterated by Aristolochia manshuriensis, and Iodes vitiginea. A. manshuriensis must not be part of any herbal medicines because it contains aristolochic acid I, a nephrotoxin and potential carcinogen. The current pharmacopoeial methods have had limitation for differentiation of C. armandii from its adulterants. Thus, a specific, comprehensive, sensitive, and reproducible HPTLC/HPLC for quality control of C. armandii was proposed. Oleanolic acid from C. armandii has been isolated by vacuum liquid chromatography and column chromatography. The purified compound has been identified using Ultra Violet spectrophotometry (UV–Vis), Fourier Transform Infrared (FTIR), mass spectrometry (MS) and nuclear magnetic resonance (NMR) spectroscopy. Therefore, oleanolic acid should be detected for positive identification of Caulis Clematis Armandii. Minimum content of oleanolic acid can be evaluated by the validated HPTLC procedure proposed as well.

1. Introduction

Clematis armandii Franch., (Ranunculaceae) is an important medicinal plant, which has been used not only in China as a form of Caulis clematidis armandii (CCA) but also in the United States of America (USA) and the European Union (EU). The plant grows in forests, forest margins, slopes, scrub, along streams at the altitude from 100 to 2400 m. In China, it is distributed in Gansu, Shaanxi, Jiangxi, Hiubei, Hunan, Zhejiang, Fujan, Guangdong, Guangxi, Sichuan, Guizhou, Junan, Xizang [1]. CCA has a bitter taste and a cold property, acting on the heart, small intestine and urinary bladder channels with the actions of promoting urination and relieving stranguria, removing heat from the heart, stimulating menstrual discharge, promoting lactation [2]. At present, the plant is used to replace Caulis aristolochiae manshuriensis (CAM) in 32 Chinese traditional patent medicines. CAM contains aristolochic acid I which is a nephrotoxin and potential carcinogen [3]. There have been reports that overdosage (60 g) of the root of CAM resulted in several acute renal failure and death [4]. Therefore, CAM is no longer included in Pharmacopoeia of the People's Republic of China (PPRC), and not permitted to be sold raw or in products in the United States of America (USA) or the European Union (EU) [5]. Based on a post marketing surveillance programme in Vietnam, CAM Aristolochia manshuriensis, and Caulis Iodes vitiginea (CIV) were often used inappropriately as CCA as adulterants. The challenge to the quality control system becomes significant if CCA is mixed with CAM or CIV, particularly if the material is comminuted. Thus, the differentiation CCA from its adulterants is important for the safe and effective use of this plant species. There were very few reports have been publishes about the differentiation CCA from its adulterants. A loop-mediated isothermal amplification (LAMP) was developed to differentiate CCA from CAM [6]. However, it cannot be applied if the small amounts of adulterant are being mixed with CCA nor a cost-effective method. To the best of our knowledge, none of the chromatographic methods were reported for differentiation CCA from its adulterants. However, several TLC approaches had been used for identification of CCA mentioned in various pharmacopoeias. The current PPRC pattern-oriented method evaluates the authentication on the basis of comparison with reference drug [2], whereas the Hong Kong Chinese Materia Medica Standards (HKCMMS) compound-oriented method evaluates the authentication on the basis of comparison with oleanolic acid (Fig. 1.) as a marker [7], and the European pharmacopoeia (EP) highly pattern-oriented method which evaluates the authentication on the basis of comparison with oleanolic acid and hederagenin as markers with the typical TLC chromatogram provided [7]. Thus, there is a need for a convenient and effective method for quality control of CCA. In this paper, a new HPTLC/HPLC method was developed for differentiation CCA from its adulterants. It is a promising method which is specific, comprehensive, sensitive and practical, and is suitable for the routine quality control of CCA.

Fig. 1.

Chemical structure of oleanolic acid.

2. Experimental

Reagents, reference standard and materials.

2.1. Chemicals

Aristolochic acid I (99.0%), lot number: 070M1238V, was purchased from Institute of Drug Quality Control-Ho Chi Minh City, Vietnam. Oleanolic acid (97.0%), lot number: 030520, was purchased from Institute of Drug Quality Control-Ho Chi Minh City, Vietnam. Ethanol, methanol, petroleum ether (60–90 °C), ethyl formate, formic acid, chloroform, toluene, ethyl acetate, acetic acid, cylclohexane, acetone, diethylamine and tin (ii) chloride, phosphoric acid, sulfuric acid of laboratory grade and acetonitrile LC grade, silica gel 60 F₂₅₄ were purchased from Merck, Germany.

2.2. Matrix

Caulis Clematidis Armandii (Powder form) reference drug, lot number: 121409–201402 HH95 9RSS and 121409–201402 Q3YZ-D47K, were purchased from National Institutes for Food and Drug Control, China.

Caulis Clematidis Armandii (Powder form) reference drug, lot number: DST191123-021, was purchased from Chengdu Desite Biological Technology Co., Ltd.

Samples of “Mu Tong” (Stem) were collected at various traditional clinics in Pleiku city, Vietnam.

The samples were identified by molecular DNA barcoding technique indicated Clematis armandii (ID = 98.74%), Aristolochia manshuriensis (ID = 100%), and Iodes vitiginea (ID = 93.0%).

2.3. Instrumentation

HPTLC was performed with a Camag HPTLC instrument including: a CAMAG Linomat 5 sample applicator, CAMAG Automatic Developing Chamber 2, CAMAG TLC Visualizer 2, CAMAG HPTLC Software visionCATS, CAMAG Visualizer Ultimate Package VisionCATS Software. Compounds were separated on HPTLC aluminium plated precoated with silica gel 60 F₂₅₄. The HPLC analyses were performed in a Shimadzu (Japan) chromatography system composed by: quaternary pump (LC-20 AD), photodiode array detector (SPD-M20A), automatic injector (SIL-20AC), column compartment (CTO-20AC) and software controller (LCsolution).

2.4. Extraction and isolation of oleanolic acid

Stems of Clematis armandii were pulverized in mechanical grinder to obtain coarse powder. The powdered materials (8.5 kg) were extracted by maceration at room temperature in methanol 80% for 24 h (3 times), yielding, after evaporation under reduced pressure, a methanolic extract (827 g) which was then fractionated with n-butanol to obtain 218 g of n-butanol extract which was hydrolyzed in methanol in the presence of 7% hydrochloric acid. Cool, a liquid-liquid extraction was performed using petroleum ether (35–60 °C) to obtain an extract (9.4 g), after evaporation under reduced pressure. A portion of this extract (5 g) was subjected to vacuum liquid chromatography [VLC, inner diameter (i.d.) = 40 mm and length (l) = 500 mm] on silica gel 60 (40–63 μm) (100 g), eluting with chloroform–methanol mixtures of increasing polarity (40:0; 40:0.5; 40:1; 40:1.5, v/v). There were 8 major fractions collected and regrouped on the basis of TLC analysis and evaporated under reduced pressure atmosphere. Major fraction III, IV, and V (120.4 mg) were repeatedly subjected to silica gel column chromatography [CC, inner diameter (i.d.) = 20 mm and length (l) = 500 mm] on silica gel 60 (40–63 μm) (25 g), eluting with chloroform–methanol mixtures of increasing polarity(40:0; 40:1.5; 40:3, v/v) to obtain another 6 major fractions. Fractions which showed 1 spot on TLC chromatogram (mobile phase: chloroform–methanol 40:1.5, v/v) were regrouped and evaporated under reduced pressure to give 87.8 mg of whitish solid precipitated which was further recrystallised from methanol to afford 82.5 mg of white crystals.

2.5. Identification of oleanolic acid by spectroscopic method

2.5.1. UV–Vis

UV–Vis spectrum was recorded on UV-3101PC spectrophotometer of Shimadzu, Japan with solvent as methanol.

2.5.2. FTIR

IR spectrum was measured in KBr disk by Prestige-21 spectrophotometer of Shimadzu, Japan in the range of 4000–600 cm⁻¹.

2.5.3. Mass analysis

ESIMS spectra were gained using 6200 Series TOF and 6500 Series Q-TOF B.06.01 of Agilent, Germany.

2.5.4. NMR analysis

One dimensional NMR spectra were carried out on 500 MHz NMR spectrometer of Avance Bruker AC, Germany in chloroform-d NMR grade.

2.6. Sample and standard preparation

2.6.1. Sample preparation for HPTLC/HPLC

Transfer about 5.0 g of finely powdered CCA to a 250-mL conical flask. Add 100 mL of methanol 80%, sonicate at 60 °C for 60 min. Cool to room temperature, and filter. Repeat with 50 mL of methanol 80%. Combine the extracts, concentrate to dryness at reduced pressure in a rotary evaporator. Dissolve the residue in 5 mL of methanol. Create the precipitate by adding dropwise the obtained methanolic solution in 50 mL of acetone. Filter to obtain the precipitate, evaporate the acetone off and dry the precipitate at 60 °C for 30 min. Dissolve the precipitate in 50 mL of 7% hydrochloric acid solution in methanol, reflux for 3 h. Add to the hydrolysis solution 10 mL of water and extract with five quantities, each of 50 mL, of methylene chloride. Combine the methylene chloride extracts and evaporate to dryness. Dissolve the residue in 5 mL of methanol, shake and filter through a membrane filter (nominal pore size 0.45 μm).

The CAM and CIV Iodes vitiginea were prepared in the same manner as CCA.

2.6.2. Standard preparation

2.6.2.1. Preparation of standard drug solution for HPTLC

The CCA reference drug solution was prepared in the same manner as the test solution.

2.6.2.2. Preparation of standard solution for HPTLC

Dissolve 10 mg of aristolochic acid I reference standard in 200 mL of a mixture of methanol - water (3 : 1). Dilute 2 mL of this solution into 10 mL with a mixture of methanol - water (3 : 1).

Dissolve 5 mg of oleanolic acid reference standard in 50 mL of methanol.

2.6.2.3. Preparation of standard solution for HPLC

Dissolve 5 mg of oleanolic acid reference standard in 100 mL of methanol.

2.7. Thin layer chromatographic conditions

Thin-layer chromatography was carried out by spotting standard and sample preparation of CCA on thin-layer chromatography (TLC) plates of precoated silica-gel 60 F₂₅₄ (2–10 μm) on aluminum sheets (Merck, Germany) using CAMAG Linomat 5 sample applicator. The samples, in the form of bands of length 8 mm were spotted at a constant application rate using nitrogen aspirator. Development of the TLC plates was carried out vertically in a 20 cm × 20 cm twin trough chromatographic tank, pre-saturated for 10 min (with filter paper) with an appropriate developing solvent. The separation was allowed to run a distance of 85 mm from the lower edge of the plate. After developing, the plates were dried at ambient temperature. Mobile phase system used was: chloroform-methanol (40 : 1.5, v/v). Derivatization of the compounds on the TLC plates was made by spraying with a 10% solution of sulfuric acid in ethanol, and heat at 105 °C until the colours of the zones become visible. Detection was made by examining in daylight and under UV light at 254/366 nm.

2.8. Validation of TLC method

Method validation was performed according to AOAC and ICH guideline [8,9]. Thus, for identification purpose, the qualification of sample application, the stability of the analyte which consists of stability of sample in solution; stability of sample on plate; and 2D-separation, specificity, precision which comprises repeatability and intermediate precision and robustness. For assay, linearity and accuracy together with the validation parameter for qualitative method mentioned earlier were performed.

2.8.1. Qualification of sample application

The performance of automatic sample applicators is tested by applying at least eight times the same sample solution of a standard substance (in different vials) onto a chromatographic plate. After development the chromatogram is evaluated by densitometry and the relative standard deviation(RSD) of the peak areas is calculated. The RSD should not be higher than 2%.

2.8.2. Stability of the analyte

A solution of reference solution of oleanolic acid and reference crude drug solution of C. armandii were prepared and immediately applied on the plate. This solution and the plate are stored in a desiccator under ambient conditions for 3 h. After 3 h, a second solution of those references were freshly prepared and immediately applied on the plate next to the previously applied sample. The set is used to investigate the stability of the analyte on the plate. Next, on the same plate, both solutions of the analyte (the solution stored over 3 h and the freshly prepared solution) are applied. This set evaluates the stability of the analyte in solution. Comparison of the 4 fingerprints are made after detection/derivatization to see if they are not significantly different then the stability of analyte both in solution and on the plate are complied. With a 2-dimensional HPTLC experiment, the solution of the analyte is applied as a spot in the lower left corner of the HPTLC plate. After development, the plate is thoroughly and appropriately dried. Second dimension is performed with turning the plate 90^o using fresh mobile phase in the same condition. If all components line up on the diagonal connecting the line of application with the intersection of the 2 mobile phase fronts, the analyte's stability meet the acceptance criteria.

2.8.3. Specificity

The selectivity was demonstrated by comparing the fingerprint of the sample with the fingerprint of reference crude drug of C. armandii and those of A. manshuriensis, and I. vitiginea as adulterants. A method is specific if the fingerprint of a sample representing the target species is similar with respect to number, position, color, and intensity of zones to the fingerprint on a botanical reference material representing the same species and samples representing a different species produce a different fingerprint.

2.8.4. Precision

Three portions of powdered drug of C. armandii are individually prepared. Onto three 20 × 10 cm plates, three aliquots of each reference solution are applied. The Rf values of three prominent zones across each plate and from plate to plate are evaluated. Visual evaluation was performed. All fingerprints on each plate are identical with respect to number, position, color, and intensity. Across each plate the zones – due to the same compounds – form parallel lines with no disturbance and Rf values differ from expected value not more than: 0.02 (same day); 0.05 (different day).

2.8.5. Robustness

Evaluation of humidity effects to establish the range in which the method performs as expected. The Rf values of three prominent zones must lie within the acceptance criteria of the intermediate precision (Rf ≤ 0.05) [10].

2.8.6. Linearity

The linearity was estimated by regression using the least square method. The slope, intercept and correlation coefficient (r) were calculated and evaluated [9]. There by, samples of the solution at five different concentrations (4.94–246.92 μg mL⁻¹) were spotted twice and the area recovered was calculated.

2.8.7. Accuracy

Recovery was determined by spiking oleanolic acid to the powdered drug before extraction. Nine determinations over 3 concentration levels covering the range of 80%, 100% and 120% the nominal concentration of the procedure. The recovery was calculated by comparing the same samples without spiking.

2.8.8. Limit of detection

The limit of detection (LOD) was calculated on the basis of standard deviation of the response and the slope as shown in Eq. (1).

$$\mathbf{L}\mathbf{O}\mathbf{D}=\frac{\mathbf{3.3}\mathbf{x}\mathbf{\sigma }}{\mathbf{S}}$$ (1)

where σ is the standard deviation of the y-intercept, and S is the slope of the calibration curve.

2.8.9. Statistical analysis

Excel 2016 (Microsoft Office) was used for statistical analysis. A 5% level of significance was selected. HPTLC chromatograms were processed by CAMAG Visualizer Ultimate Package VisionCATS Software.

2.9. HPLC chromatographic condition

Chromatographic separation was achieved using a Zorbax SB-C18 column, 250 × 4.6 mm, 5 μm particle size (Agilent, USA), connected with a precolumn (C18, 4.0 × 3.0 mm, 5 μm, Phenomenex, USA). The chromatographic conditions were: injection loop of 20 μL; column compartment at 40 °C; and UV detection at 205 nm. All the mobile phases were filtered in a 0.45-μm filter membrane (Sartorius, Germany) and degassed by an ultrasonic apparatus (TranssonicT700, Fisher Scientific, Germany) for 30 min before use. All volumetric glassware used was previously calibrated. The flow rate was 1.2 mL/min. Programme the chromatographic system as follows.

Time (min)	Acetonitrile (%, v/v)	Methanol–0.1% phosphoric acid solution (50 : 50) (%, v/v)
0–20	60	40
22–23	60 → 100	40 → 0
23–28	100	0
28–30	100 → 60	0 → 40
30–35	60	40

2.10. Validation of HPLC method

2.10.1. Specificity

Sample retention time should match with standard retention time and variation should not be more than 2%. There should not be any interference of blank peak with dilute solvent. The obtained peak purity index of oleanolic acid in sample preparation should confirm the specificity of the method.

2.10.2. Linearity

From the standard solution of oleanolic acid for HPLC, aliquots of 0.1, 0.2, 1.0, 2.0, and 7.0 mL were withdrawn and diluted to 10.0 mL in volumetric flasks, resulting in the following concentrations: 4.87; 9.75; 48.73; 97.46; and 341.11 μg. mL⁻¹. Evaluating linearity was performed by analysis of variance and correlation coefficient determination as well as y-intercept evaluation.

2.10.3. Limit of detection

The limit of detection (LOD) was calculated on the basis of standard deviation of the response and the slope as shown in Eq. (1).

3. Results and discussion

3.1. Spectroscopic identification of oleanolic acid isolated from C. armandii

3.1.1. UV–Vis analysis

The UV–Vis spectrum (MeOH) showed a maximum absorbance wavelength at 205 nm that matched well with literature data [11].

3.1.2. FTIR analysis

The IR spectrum showed a broad absorption at 3456 cm⁻¹, with medium intensity. The strong absorption peak at 2945 cm⁻¹ suggested the presence of a CH₂ symmetric stretching. The band strong signal at 1694 cm⁻¹ indicated the presence of a C O stretch from carboxylic acid. There were bands at 1029 cm⁻¹ showed the presence of C–O stretching and 948 cm⁻¹ confirmed O–H bending of a carboxylic acid. The IR data matched well with literature data [12].

3.1.3. MS analysis

A molecular weight of oleanolic acid was determined by ESIMS (+) with m/z 457.5 [M+H]⁺; 438,9 [M + H–H₂O]⁺; 410,5 [M + H–COOH]⁺; and 392,9 [M + H–H₂O–COOH]⁺ that matched well with literature data [13].

3.1.4. NMR analysis

The ¹³C NMR data matched well with the reference data of oleanolic acid, lot number: 030520, that was purchased from Institute of Drug Quality Control-Ho Chi Minh City, Vietnam.

3.2. Validation of HPTLC procedure

3.2.1. Qualification of sample application

Applying eleven times the same sample solution of oleanolic acid (0.5 mg mL⁻¹) (in different vials) onto a chromatographic plate. After development the chromatogram, the RSD of peak areas is calculated. The RSD was 1.92% (<2%).

3.2.2. Stability of the analyte

Comparision made on the results of freshly prepared samples and stored samples showed that the fingerprints were similar in all detail or the analyte is stable in solution as well as on the plate during the tested time interval. On the other hand, all components line up on the diagonal that connects the application position with the intersection of the 2 mobile phase fronts for 2-dimensional HPTLC experiment.

3.2.3. Specificity

By comparing the major zones of the fingerprint of authenticated CCA, Chuan mu tong CCA reference drug with those of authenticated CAM and CIV Iodes vitiginea as adulterants, difference patterns were obtained in term of position, color, and intensity of zones as illustrated in Fig. 2-a (daylight) and Fig. 2-b (UV 366 nm). This result strengthen and indicate the proposed method's selectivity. The authenticated CCA solution (Sample solution) exhibits three main reddish-violet zones with Rf values of approximately 0.1, 0.5, and 0.6 that correspond in position and color to zones in CCA reference drug solution. Oleanolic acid standard solution 1 exhibits a reddish-violet zone due to oleanolic acid at an Rf of about 0.5. The sample solution exhibits a zone similar in color and Rf value to that due to oleanolic acid in standard solution 1.

Fig. 2.

HPTLC photo-document of C. armandii, A. manshuriensis, and I. vitiginea. Track assignment: 1: blank, 2: C. armandii, 3–4: I. vitiginea, 5–6: A. manshuriensis, 7: C. armandii adulterated with 10% of I. vitiginea, 8: C. armandii adulterated with 10% of A. manshuriensis, 9: oleanolic acid (reference standard), 10: aristolochic acid I (reference standard). (a = daylight, b = UV 366 nm).

3.2.4. Precision

3.2.4.1. Repeatability

For precision, the results for repeatability expressed as the deviation in Rf value from expected value (ΔRf) of three prominent zones in the chromatogram of CCA C. armandii in the same day (ΔRf ≤ 0.02). The method was found to be precise as the repeatability had lower value of 0.02.

3.2.4.2. Intermediate precision

For precision, the results for intermediate precision expressed as the deviation in Rf value from expected value (ΔRf) of three prominent zones in the chromatogram of CCA C. armandii in the different days (ΔRf ≤ 0.05). The method was found to be precise as the intermediate precision had lower value of 0.03.

3.2.4.3. Precision (Quantitative method)

Different amounts of spiked samples were spotted on TLC plate. These spots were analyzed by using the above-described HPTLC method. Precision was expressed as the percent relative standard deviation (%RSD). The method was found to be precise as the intraday precision (n = 3) and the interday precision (n = 5) had lower values of 1.52% and 1.76%, respectively.

3.2.5. Accuracy

Different amounts of spiked samples of oleanolic acid were spotted on TLC plate. These spots were analyzed by using the above-described HPTLC method. Table 3 summarizes the recovery of the analytical method.

Table 3.

Recovery study of oleanolic acid through standard addition (n = 9).

Spike level (%)	Actual amount of oleanolic acid spotted (ng)	Amount of added oleanolic acid (ng)	Area	Observed amount of oleanolic acid (ng)	Recovery (%)	Mean recovery, RSD (%)
80	297.33	238.47	0.01235	545.73	104.16	103.21
	297.33	238.47	0.01227	539.53	101.56
	297.33	238.47	0.01235	545.12	103.91	1.13
	297.33	298.09	0.01301	596.35	100.31	102.87
100	297.33	298.09	0.01313	605.97	103.54
	297.33	298.09	0.01318	609.58	104.75	1.82
	297.33	357.71	0.01396	670.47	104.31	104.16
120	297.33	357.71	0.01398	672.13	104.78
	297.33	357.71	0.01392	667.10	103.37	0.56

3.2.6. Robustness

Variation of Rf values of three prominent zones in the chromatogram of CCA C. armandii with changing relative humidity (RH) of 38% and 58% was studied. The deviations were well within the acceptance criteria of not more than 0.05 in ΔRf.

3.2.7. Linearity

The regression curve of peak areas versus concentrations proved linear with a coefficient of correlation r = 0.9966 and with confidence intervals at P = 0.05. Y = 0.005323 + 0.000129 X. The regression data are given in Table 1, showing good linear relationship in the range of 4.94–246.92 μg mL⁻¹.

Table 1.

Results of ANOVA analysis for calibration curve of oleanolic acid (HPTLC).

Parameter		Results
Calibration curve	Slope	Intercept	Correlation coefficient	Range (μg.mL−1)
	0.01288	+0.005323	0.9966	4.94–246.92
Analysis of Variance	Source	Degree of freedom	Sum of Square (SS)	Mean of Square (MS)
	Regression	1	0.001317485	0.001317
	Residual	3	9.99592 × 10−6	1.25 × 10−6
	Total	4	0.001327481

3.2.8. Limit of detection

The LOD for the sample was calculated from calibration plot and the result was 0.29 μg mL⁻¹.

3.3. Validation of HPLC procedure

3.3.1. Specificity

For specificity, the results are shown in Fig. 3. The solvent peak do not interfere with peak of oleanolic acid and peak was baseline resolved. There is no overlap of peaks of standard or sample compared to other from solvents used. It can be observed from the peak purity test that the peak purity index of the peak of oleanolic acid in standard and sample solutions was almost 1.00. This result strengthen and indicate the proposed method's selectivity.

Fig. 3.

Chromatograms obtained for the specificity test. Injections of: A, oleanolic acid standard; B, sample; C, blank. Chromatographic conditions: Zorbax SB- C18 column (250 × 4.6 mm, 5 μm particle size), mobile phase composed by acetonitrile and methanol-0.1% phosphoric acid solution (50: 50) at flow of 1.2 mL min⁻¹ (gradient elution), column compartment at 40 °C, and UV detection at 205 nm.

3.3.2. Linearity

The regression curve of peak areas versus concentrations proved linear with a coefficient of correlation r = 0.9999 and with confidence intervals at P = 0.05. Y = 8308.62 + 12275.89 X. The regression data are given in Table 2, showing good linear relationship in the range of 4.9–341.1 μg mL⁻¹.

Table 2.

Results of ANOVA analysis for calibration curve of oleanolic acid (HPLC).

Parameter		Results
Calibration curve	Slope	Intercept	Correlation coefficient	Range (μg.mL−1)
	12275.89	+8308.62	0.9999	4.9–341.1
Analysis of Variance	Source	Degree of freedom	Sum of Square (SS)	Mean of Square (MS)
	Regression	1	1.17 × 1013	1.17 × 1013
	Residual	3	971000000	324000000
	Total	4	1.17 × 1013

3.3.3. Limits of detection

The LOD for the sample was calculated from calibration plots and the result was 4.84 μg mL⁻¹.

3.4. Sample analysis

The validated HPTLC method was applied for the minimum content test of the marker oleanolic acid in stem of C. armandii, A. manshuriensis, and I. vitiginea. The marker compound of oleanolic acid was found to be present in the stem of C. armandii (0.003%) whereas none of it found in A. manshuriensis nor I. vitiginea.

4. Discussion

Although a number of reports have been published about the determination of oleanolic acid in CCA using various techniques [[14], [15], [16], [17]]. But to the best of our knowledge, none of the reports on isolation of oleanolic acid from CCA were published. Our isolation of oleanolic acid helped reconfirm the presence of it in CCA. The use of oleanolic acid as a marker for TLC identification of CCA as per earlier version of PPRC [18,19], current HKCMMS [7], and current EP [20] has been in line with our finding. Therefore, oleanolic acid should be detected for positive identification of CCA. However, Liu et al. had not found oleanolic acid in 9 batch samples of CCA and suggested to reinvestigate the index component of Clematis spp after mentioning that it used to be an index component in PPRC for the identification of CCA [21]. Afterwards, Bai et al. had suggested that CCA reference drug be used instead of β-sitosterol to improve the TLC identification [14]. That might be why from PPRC 2010 onwards [2,22], oleanolic acid had no longer been a marker but CCA reference drug was used instead. The TLC procedure in current PPRC did not help differentiate CCA from CAM, and CIV because the sample preparation was not able to obtain oleanolic acid. Regarding the tracks of CCA, CAM, and CIV on the TLC chromatogram, they were similar in term of the number of zones, Rf values, and intensities. In respect of the TLC method in EP, the hydrolysis was omitted resulting in the absence of oleanolic acid in the sample solution. Although the sequence of zones present in the chromatograms obtained with the reference solution and the test solution described, it cannot be applied due to the hydrolysis significantly is significant disadvantage of the sample preparation [20]. Although the TLC method in HKCMMS required oleanolic acid as a marker, it need to be improved since the use of 12 M hydrochloric acid for hydrolysis may decompose or carbonise the residue obtained. In addition, the mobile phase and the derivatization could not help distinguish CCA from its adulterants. Therefore, the proposed HPTLC method allows the sample of CCA mixed with CIV to be identified by examining the plate under ultraviolet light at 365 nm, the chromatogram showed a purplish-brown fluorescent zone (Rf 0.84). The CCA mixed with CAM could be identified by examining the plate under ultraviolet light at 254 nm, the chromatogram showed a zone due to aristolochic acid I (Rf 0.07). In regarding CCA, after derivatization, it could be differentiated from CAM and CIV by examining the plate under daylight and under ultraviolet light at 365 nm. The chromatogram showed a purple, yellowish-orange fluorescent zone (Rf 0.10), light pink, light pink fluorescent zone due to oleanolic acid (Rf 0.48), and light purple, light orange fluorescent zone (Rf 0.62). Furthermore, the minimum content of oleanolic acid by HPTLC could be realized due to the fully validated method. The validated HPLC method for identification of oleanolic helped confirm the specificity of the overall chromatographic identification as per ISO/DIS 19609-2 international standard [23]. The HPLC assay method in EP describes the sample preparation with the adding of 6 M hydrochloric acid but without the heat to catalyse the hydrolysis process [20]. Our experiment had confirmed that oleanolic acid was not in the test solution prepared by EP assay method resulting in an HPLC chromatogram without oleanolic acid peak. In our proposed method, reflux the test solution in 7% hydrochloric acid solution in methanol for 3 h was performed to obtain oleanolic acid. The minimum content of 0.30% of oleanolic acid (dried drug) required by EP should also be considered since Qing et al. reported an HPLC procedure for determination of oleanolic acid in CCA with the use of hydrochloric acid and reflux and found that the results of oleanolic acid in CCA vary from 0.009 to 0.0124%. However, the mobile phase included triethylamine and acetic acid to improve the peak shape of oleanolic acid [15]. In our proposed method, the mobile phase did not include acetic acid and triethylamine to avoid any import-export nuisance of these chemicals but still ensuring the similar symmetry of oleanolic acid peak at about 1.05. Yang et al. reported the determination on oleanolic acid in CCA by microwave-assisted extraction -HPLC/MS [16]. Obviously, it is not suitable for routine quality control of CCA.

5. Conclusion

Oleanolic acid was isolated from Caulis Clematis Armandii for the first time. A new HPTLC/HPLC method for differentiation Caulis Clematis Armandii from Caulis Aristolochia Manshuriensis and Caulis Iodes Vitiginea was developed and validated in accordance to the validation parameters required by the ICH Q2 (R1) guidelines. This new method can identify not only Caulis Clematis Armandii itself but also in case Caulis Clematis Armandii mixed with Caulis Iodes Vitiginea or Caulis Aristolochia Manshuriensis in a specific, comprehensive, and practical manner. Oleanolic acid should be detected for positive identification of Caulis Clematis Armandii. The minimum content of oleanolic acid in Caulis Clematis Armandii can also be performed by the validated HPTLC as well. This method of analysis can be useful for laboratories and regulatory agencies including pharmacopoeia commissions in quality control or developing monograph.

Author contribution statement

Ha Minh Hien: Conceived and designed the experiments; Performed the experiments; Analyzed and interpreted the data; Wrote the paper.

Huynh Ngoc Thuy Tram: Performed the experiments; Analyzed and interpreted data; Contributed reagents, materials, analysis tools or data.

Tran Viet Hung: Conceived and designed the experiments; Contributed reagents, materials, analysis tools or data; Wrote the paper.

Data availability statement

Data will be made available on request.

Declaration of interest's statement

The authors declare no conflict of interest.