1.   Introduction

Hedyotis corymbosa (Shui Xian Cao, 水线草, HC) is the dried whole plant from the genus Hedyotis belonging to the family Rubiaceae. It was initially recorded in the Illustrated Catalogue of Plants, a Chinese botanical drug monograph. It is bitter, cold, and can enter two channels of the lung and the large intestine in the human body. HC is mainly used for the treatment of scald, but it can also be used to treat malaria when mixed with Dichroae Radix (Chang Shan, 常山) and Verbenae Herba (Ma Bian Cao, 马鞭草) and decocted with water ^[1]. HC is used in the treatment of various cancers in several regions, such as Taiwan and India ^[2]. It is also the main antitumor medicine used in mainland China. However, the quality control standard of HC faces two issues: specificity and systematicness. In mainland China, HC is mixed or incorporated with Hedyotis diffusa (Bai Hua She She Cao, 白花蛇舌草, HD) for the treatment of cancers ^[3], but HC and HD have different chemical substances ^[4] and pharmacological activities ^[5]. Currently, HC is described in the Standard of Traditional Chinese Medicine of Guangdong^[6] and Chinese patent drugs containing HD are described in Chinese Pharmacopoeia ^[7]. However, these two medicinal materials cannot be differentiated. This is because they are congeneric herbs that have similar external shapes. The distinction is, that HD is monangial or has twinflowers from the axil with a short and thick pedicel, and that HC produces the corymb and has the thin pedicels and two to five flowers growing from the same axil. Thus, these two species cannot be differentiated when cut into slices. Further, measurements of oleanolic acid and ursolic acid cannot be used for differentiating these species. HC and HD are ineffectively differentiated in the Standard of Traditional Chinese Medicine of Guangdong, and the report is not systematic and lacks measurements, such as microscopic identification, thin layer chromatography (TLC), and the concentrations of water, ash, heavy metals and pesticide residues. This study primarily focuses on the quality standard of HC, because it is the only source of hedyotiscone A, which exhibits excellent antitumor activity ^[8]. In this study, we (1) qualitatively differentiated HC and HD by TLC, (2) compared the fat-soluble part of the two herbs based on the characteristic spectrum from high-performance liquid chromatography (HPLC), and (3) quantitatively measured hedyotiscone A in HC from different habitats. The microscopic identification method was used to morphologically identify the powder and cross-section of HC. Oven drying, atomic absorption spectrometry and gas chromatography were used to determine the concentrations of water, total ash, acid-insoluble ash, heavy metals and pesticide residues.

2.   Materials and Methods

2.1   Instruments

Inverted biomicroscopy (20022873), MOTIC IMAGES ADVANCED 3.2 processing software; GF₂₅₄ silica gel plates (Qingdao Haiyang Chemical Co. Ltd), ZF-2 ultraviolet instrumentation (Shanghai Anting); Waters 2695-2996 HPLC system with Empower 2 workstations; 202 bench-top drying oven (Beijing Yongguangming Medical Instruments Factory); SX2-10-12 muffle furnace (Shanghai Tanghe); AA-6300C atomic absorption spectrophotometer (Shimadzu Corporation); ASC-6100 automatic sampler (Shimadzu Corporation); GFA-EX7i graphite furnace atomiser (Shimadzu Corporation); Agilent GC-6890N gas chromatography, electron capture detector (ECD); and fine balances, including the AUW analytical balance (Shimadzu Corporation) and the AL-2140 electronic balance (Ohaus).

2.2   Reagents

Methanol was analytically pure; strong acids, such as hydrochloric acid, were the guaranteed reagents (GRs); and distilled water was tri-distilled.

2.3   Reference

Hedyotiscone A was house-fabricated with a purity of 98.45% (refer to figure 1 for its structure); standard solution of lead, arsenic, mercury, cadmium and copper (1 mg/mL, Chinese Research Institute of Metrology); α-666, β-666, γ-666, δ-666 (BHC), o, p'-DDT, p, p'-DDT, p, p'-DDD, p, p'-DDE (DDT), pentachloronitrobenzene (National Institute of Pharmaceutical and Biological Products).

Figure  1.  Structure of hedyotiscone A

下载: 全尺寸图片 幻灯片

2.4   Herbs

Herbs were collected from southern provinces in China. Details are presented in table 1.

Table  1.  Medicinal material information for Hedyotis corymbosa and Hedyotis diffusa

NO.	Medicinal materials	Collecting locations
1	Hedyotis corymbosa	Chinese University of Hong Kong, Hong Kong
2	Hedyotis corymbosa	Yunan County, Guangdong Province
3	Hedyotis corymbosa	Tianhe District, Guangdong Province
4	Hedyotis corymbosa	Luoding County, Guangdong Province
5	Hedyotis corymbosa	Xinxing County, Guangdong Province
6	Hedyotis corymbosa	University town, Guangdong Province
7	Hedyotis corymbosa	Zhongshan, Guangdong Province
8	Hedyotis corymbosa	Beiliu County, Guangxi Province
9	Hedyotis corymbosa	Guigang County, Guangxi Province
10	Hedyotis corymbosa	Shunde County, Guangdong Province
11	Hedyotis corymbosa	Wulong County, Yunnan Province
12	Hedyotis corymbosa	Fengqing County, Yunnan Province
13	Hedyotis corymbosa	Mengding County, Yunnan Province
14	Hedyotis corymbosa	Shuangjiang autonomous County, Yunnan Province
15	Hedyotis corymbosa	Yongsheng County, Yunnan Province
16	Hedyotis diffusa	Renmin Pharmacy
17	Hedyotis diffusa	Green apple Pharmacy
18	Hedyotis diffusa	Laobaixing Pharmacy
19	Hedyotis diffusa	Yaohaitong Pharmacy
20	Hedyotis diffusa	First Affiliated Hospital of Hunan University of Chinese Medicine

下载: 导出CSV 

| 显示表格

2.5   Characteristic research

2.5.1   TLC differentiation

A reference sample (10 mg hedyotiscone A) and HC powder (5 g) were dissolved in 100 mL methanol in a 100 mL volumetric flask, respectively. The sample was treated with ultrasound for 60 min and filtered. The filtrate was condensed under low pressure to extract the sample. Then, 50 mL water was added and the sample was dispersed by ultrasound.

The filtrate was heated in a 95℃ water bath for 1 h and filtered. The filtrate was then extracted with chloroform. The chloroform layer was transferred and completely evaporated. Methanol was added to bring the sample solution to a total volume of 5 mL. Then, 5 μL of the sample solution and 5 μL of the reference solution was placed on the same silica gel GF₂₅₄ plate, according to the TLC method described in Chinese Pharmacopoeia^[9]. The sample was then extended using an extension system with a petroleum ether-ethyl ratio of 2 : 1, removed, air dried, and placed under an ultraviolet lamp (365 nm) for inspection.

2.5.2   Characteristic HPLC and hedyotiscone A analyses

Sample pre-treatment: 5 mg sample of hedyotiscone A was placed in a 10 mL graduated flask and then 75% methanol was added to obtain a reference solution with a concentration of 0.527 mg/mL. One gram of HC powder (added through a No. 5 sieve) and 25 mL of 75% methanol were placed into a conical flask with a tightly sealed cover. The solution was weighed and treated with ultrasound (power: 160 W; frequency: 59 kHz) for 30 min. The flask was then removed, cooled and weighed again. Methanol (75%) was added to compensate for lost weight. The solution was shaken and filtered, and the filtrate was transferred to be used as the sample solution.

Chromatography conditions: RP C18 column (ZORBOX Eclipse XCB 4.6 × 250 mm, 5 μm) was used with a mobile phase consisting of acetonitrile : water (46 : 54). The wavelength was 345 nm, the sample volume was 7.5 μL, the temperature of the column was 30℃, and the flow rate was 0.7 mL/min.

2.6   Microscopy analyses

2.6.1   Powder microscopy identification

Thirty grams of whole HC were combined with root, stem, leave, flower and fruit samples. Samples were cut into small pieces with lengths ranging from 1~2 cm, ground to a powder (added through a No. 4 sieve). An small amount of the powder was placed on a glass slide and chloral hydrate was added. The powder was then heated to permeabilization using an alcohol burner. Chloral hydrate was added for 1~2 cycles and diluted with glycerol. A cover slip was placed on top and excess liquid was removed after the sample had cooled. The sample was observed and photographed under a high-power microscope (10 × 40)^[10].

2.6.2   Observation of cross-sectional tissue

An HC stem was cut into 2-mm-long pieces, fixed in Carnoy's fluid for 1~2 h, and rinsed with water. Transparent xylene was added after 30%, 50%, 70%, 80%, 85% and 100% ethanol dehydration. The transparent material was treated with a mixture of paraffin wax and xylene (1 : 1) and pure paraffin wax for three cycles (20 min for each cycle). The sample was removed after the stem had completely dissolved in the paraffin wax (after 30 min). Paraffin sections were embedded and stained and a transparent section was created with xylene. The slide was removed from the xylene and excess xylene was removed. Gum was added dropwise until the xylene dried and a cover slip was placed on top. The gum was placed under low-power (10 × 5) and high-power (10 × 40) microscopes for observation and to take pictures.

2.7   Water content

HC samples were obtained from 15 different habitats (Table 1). Samples were ground into a powder and filtered through a No. 2 sieve. The oven drying method ^[11] was used to measure water content.

2.8   Ash concentrations

HC samples were obtained from 15 different habitats (Table 1). Samples were ground into a powder and filtered through a No. 2 sieve. Total ash and acid-insoluble ash measurement methods ^[12] were used to determine ash concentrations.

2.9   Heavy metal concentrations

Lead, arsenic, mercury, cadmium and copper standard solutions were prepared, and standard curves were based on the 2015 edition of Chinese Pharmacopoeia^[13]. The samples of HC and lead, arsenic, cadmium, and copper test solutions were prepared based on preparation method B (for lead and mercury) recorded the 2015 edition of Chinese Pharmacopoeia. Atomic absorption spectrometry (AAS) was used for heavy metal measurements ^{[13, 14]}.

2.10   Inspection of pesticide residues

BHC isomers, the DDT standard solution and the test sample solution were prepared as described in the 2015 edition of Chinese Pharmacopoeia^[15]. Gas chromatography (GC) was used to measure pesticide residues ^{[15, 16]}.

3.   Results

3.1   TLC identification

Three characteristic blue fluorescent spots were observed among HC samples, while one blue spot was observed in HD sample. The difference, that the blue fluorescent spots were different in size, was found in different habitats. A significant difference between HC and HD was shown in figure 2.

Figure  2.  TLC chromatograms of HC and HD from different habitats

Note: numbers under the chromatograms represent different habitats.

下载: 全尺寸图片 幻灯片

3.2   Characteristic HPLC and content determination of hedyotiscone A

3.2.1   Characteristic HPLC and selection of the characteristic peak

There were five distinct characteristic peaks for HC, while none were observed for HD. The sharpest peak was observed for hedyotiscone A (Fig. 3).

Figure  3.  HPLC characteristic chromatogram of HC and HD from different origins

下载: 全尺寸图片 幻灯片

3.2.2   Validation of the methodology

In this study, the generated linear equation of hedyotiscone A was Y= 3.15 × 10⁸X+ 312.04 (r² = 1.000) in 0.00005270~0.002108 mg/mL. Precision [relative standard deviation (RSD) 0.92%], repeatability (RSD 0.14%) and recovery (average recovery: 100.42% with RSD of 0.85%) met the requirements for quantitative analysis. The stability test showed hedyotiscone A was stable 20 h after preparation.

3.2.3   Measurement of hedyotiscone A concentrations

Concentrations of hedyotiscone A ranged from 0.00017%~0.00115% with significant variation among different habitats (Table 2). Samples from Yongsheng County in Yunnan Province, Mengding County in Yunnan Province, and Foshan Shunde District of Guangdong Province yielded the highest concentrations of hedyotiscone A, exceeding 0.001%. Samples from Guigang County in Guangxi Province, Shuangjiang County in Yunnan Province, Xinxing County in Guangdong Province, Fengqing County in Yunnan Province, and Yulong County in Yunnan Province yielded the lowest concentrations of hedyotiscone A (< 0.0004%). Concentrations of hedyotiscone A from other habitats were 0.0006% ~ 0.0010%.

Table  2.  The content of hedyotiscone A in HC

NO.	content (%)	NO.	content (%)	NO.	content (%)
1	0.00017	6	0.00023	11	0.00097
2	0.00076	7	0.00066	12	0.00091
3	0.00085	8	0.00071	13	0.00038
4	0.00021	9	0.00115	14	0.00113
5	0.00083	10	0.00106	15	0.00022

下载: 导出CSV 

| 显示表格

3.3   Microscopic identification

3.3.1   Powder microscopic identification

The powder color was taupe and cork cells and wood fibres were detected. Pores were determined to be paracytic and infinitive. Non-glandular hair and 1~3 cells were detected. A tiny wart was observed on the cell wall. A calcium oxalate needle crystal was also observed. The diameter of the spiral vessel was approximately 10~40 μm. Specific characteristics are shown in figure 4 (a-g). Microscopic identification results were obtained by examining 15 batches of samples.

3.3.2   Inspection of tissue cross-sections

The cuticular layer was composed of one layer of cells with square or rectangular shapes. This layer was covered by a horny layer. Pores and epidermis papilla were observed. The cortex was composed of several layers of parenchyma cells. Calcium oxalate needle crystal bundles were observed, and oil drops were occasionally observed. The endodermis was clear, and the phloem was wide. The cambium zone was not distinct. The xylem was connected to a ring. The medullary portion was extensive and the cell was large. Specific characteristics are shown in figure 4 (h-j).

Figure  4.  Microscopic analysis of the powder and tissue cross-section

Cork fiber (a), wood fiber (20 ×) (b), non-glandular hair (20 ×) (c), spiral vessel (40 ×)(d), stoma (flat shaft) (e), stoma (infinitive) (f), acicular crystal (20 ×) (g), cross-sectional (10 ×) (h), cross-sectional (40 ×, cortex) (i), stem pith (40 ×, acicular crystal) (j)

下载: 全尺寸图片 幻灯片

3.4   Concentrations of water and ash

Results were presented in table 3. Water, ash and acid-insoluble ash were 8.8%~11.2%, 7.0%~15.3% and 0.44%~7.7%, respectively.

Table  3.  The content of water, ash and acid insoluble ash in HC

NO.	Water (%)	Ash (%)	Acid insoluble ash (%)
1	10.72	7.04	1.04
2	8.87	11.89	2.87
3	10.73	13.15	7.67
4	9.62	10.91	2.96
5	10.79	9.17	1.47
6	10.45	12.13	2.07
7	10.47	9.15	2.69
8	10.37	15.30	5.73
9	10.55	9.22	0.44
10	10.48	9.78	2.77
11	10.22	8.41	1.60
12	9.44	10.25	1.51
13	10.50	8.35	1.64
14	10.72	8.42	2.19
15	11.19	8.42	1.50

下载: 导出CSV 

| 显示表格

3.5   Heavy metals

Results are presented in table 4. Concentrations of Pb, As, Hg, Cd, and Cu were 1.60~4.60, 0.650~1.60, 0.0300~0.100, 0.240~0.690 and 10.1~16.6 μg/g, respectively.

Table  4.  Heavy metal content in HC

NO.	Pb (μg/g)	As (μg/g)	Hg (μg/g)	Cd (μg/g)	Cu (μg/g)
1	4.43	1.03	0.07	0.66	16.57
2	2.19	1.17	0.05	0.57	12.34
3	2.23	1.07	0.05	0.58	12.65
4	1.74	0.68	0.07	0.32	11.39
5	2.17	0.96	0.04	0.54	11.94
6	2.53	1.23	0.05	0.66	13.56
7	2.74	1.05	0.07	0.64	12.27
8	4.57	1.09	0.09	0.59	16.57
9	1.63	0.94	0.05	0.37	12.38
10	2.02	1.27	0.05	0.34	12.84
11	3.55	1.13	0.07	0.67	16.98
12	3.64	0.95	0.07	0.64	15.77
13	2.57	0.98	0.07	0.25	10.15
14	3.59	1.51	0.07	0.63	12.25
15	3.46	1.54	0.07	0.68	11.62

下载: 导出CSV 

| 显示表格

3.6   Pesticide residues

Pesticide residue concentrations of all HC samples were less than 0.001 ppm. All heavy metal concentrations were less than 5.0 μg/g, with the exception of copper (10~20 μg/g). Mercury concentrations were less than 0.1 μg/g.

4.   Discussion

Oleanolic acid and ursolic acid were used for the qualitative, and quantitative analysis of HC ^[17] and hedyotiscone A was used for qualitative analysis ^[18]. In this study, the interference ingredient (the pigment that shows red fluorescence at 365 nm) was removed during sample pre-treatment. The most effective developing solvent system was the petroleum ether-ethyl acetate ratio of 2 : 1. Three characteristic blue fluorescent spots were observed in HC samples, while one blue spot was observed in an HD samples. Results showed that TLC was a more effective method for differentiating HC from HD ^[19].

Three different extraction methods (reflux extraction, ultrasound extraction and Soxhlet extraction) were used to extract hedyotiscone A from the powder sample after the pre-treatment. Using the integral value of the peak area as the inspection index, the three extraction methods showed a similar peak area for hedyotiscone A. Thus, the ultrasound extraction method was selected to simplify the process. By comparing eight different extraction solvents, including methanol, 75% methanol, 50% methanol, 95% ethanol, 75% ethanol, 50% ethanol, ethyl acetate and chloroform, the largest peak area of hedyotiscone A was obtained with 75% methanol. Therefore, 75% methanol was determined to be the most effective extraction solvent. A comparison of different ultrasound extraction durations showed that the peak area of hedyotiscone A did not increase after 30 min. Thus, an ultrasound extraction of 30 min was determined to be the best extraction duration. When comparing the solid-to-liquid ratio, the largest peak area was observed when the ratio was 1 g : 25 mL. Thus, 25 mL was determined to be the most effective solvent volume.

Two different mobile phases (46 : 54 acetonitrile-0.5% acetate acid glacial and 46 : 54 acetonitrile-water), three different flow rates (1.0, 0.8 and 0.75 mL/min), three different injection volumes (10, 8 and 7 μL), three different mobile phase ratios (47 : 53, 46 : 54, 45 : 55), and three different inspection wavelengths (325, 345 and 365 nm) were compared using the purity of the hedyotiscone A peak as the index (with the purity angle less than the purity threshold). The largest pure characteristic peak was observed when the wavelength was 345 nm, the mobile phase was acetonitrile-water (46 : 54), the flow rate was 0.75 mL/min, and the injection volume was 7 μL.

According to the literature on HD, the effective antitumor components of HD consist of low polar components, such as triterpene ^{[20, 21]} and anthraquinone ^[20], which are extracted by petroleum ether and chloroform, rather than large polarity components such as flavonoids and iridoid glycoside ^[22]. Results showed that hedyotiscone A was the most representative characteristic peak in low polar components, so hedyotiscone A in HC may be antitumor. Fewer studies have examined HC than HD, so the fingerprint of the two herbs was analyzed with HPLC.

Low heavy metal concentrations may be closely related to the source of the medicinal drugs. All HC samples were obtained from wild picking in an industrial area without artificial cultivation, which prevented contamination by heavy metals and pesticide residues.

5.   Conclusion

Calcium oxalate needle crystal bundles can be observed and oil drops can be observed occasionally in cross-sectional tissue. Three characteristic blue fluorescent spots were observed in HC samples by TLC. Results showed five distinct characteristic peaks for HC that were not observed for HD. The sharpest peak was hedyotiscone A. The linear range of hedyotiscone A was 0.00005270~0.002108 mg/mL. The concentration of hedyotiscone A ranged from 0.00017% to 0.00113% across different habitats. Moisture, ash and acid-insoluble ash concentrations were 8.80%~11.2%, 7.00%~15.3% and 0.440%~7.70%, respectively. Concentrations of Pb, As, Hg, Cd and Cu were 1.60~4.60, 0.650~1.60, 0.0300~0.100, 0.240~0.690 and 10.1~16.6 μg/g, respectively. Pesticide residues were less than 0.00100 ppm. Specific TLC and HPLC characteristic spectra were highly established and can differentiate HD and HC. A comprehensive quality standard system was also developed, including microscopic identification and measuring the concentrations of water content, ash, heavy metals and pesticide residues. This comprehensive and adaptive characterization system can effectively assure the quality and authenticity of HC, and provide a valuable reference for clinical application.

Acknowledgements

We thank for the funding support from the National Science foundation of China (No. 81503041), the Science Research Projects of Chinese Pharmacopoeia (No. Z18), the Science Research Projects of the Hunan Provincial Department of Education (No. 17C1213), the Science Research Projects of the Hunan Provincial Department of Education (No. 14C0860), and the Key Projects of the Changsha Municipal Science and Technology Bureau (No. K1406030-31).

Competing Interests

The authors declare no conflict of interest.