Discrimination of Baphicacanthis Cusiae Rhizoma et radix and its adulterant species and establishment of an assay method for quality control

Optimization of sample preparation

To enhance the efficiency of extraction, a rigorous optimization of the extraction conditions was conducted. Initially, a range of extraction solvents, including chloroform, ethyl acetate, 30%, 50%, 70%, and 100% MeOH, as well as water, were evaluated by assessing the quantity and abundance of peaks in the total ion chromatogram (TIC) associated with each solvent. As a result, 100% MeOH was found to be the most suitable extraction solvent due to its superior extraction efficiency. Subsequently, various extraction methods (ultrasonic extraction, reflux extraction, and Soxhlet extraction), extraction durations (30, 60, and 90 min), and the number of extractions were examined to determine the optimal conditions for achieving maximum extraction efficiency. As a result, the optimal parameters for obtaining sample solutions with enhance extraction efficiency are presented in the section titled “Preparation of standard and sample solutions”.

Establishment of chemical compositions database for NBLG

To achieve a thorough identification of chemical components within NBLG using MS data, it is imperative to establish a comprehensive database that contains all possible compounds. In this study, we constructed a personal database comprising 217 compounds for NBLG identification by referring to the components of the Acanthaceae family listed in the Dictionary of Natural Products (Chapman & Hall/CRC Press, Boca Raton, FL, USA) and conducting a systematic literature review of NBLG and BBLG [3, 15,16,17].

Structural interpretation of constituents based on high-resolution MS and MS/MS data

The “Find by Formula” algorithm in Agilent MassHunter software was employed to comprehensively profile the chemical composition of NBLG. The positive ESI mode was primarily utilized for compound analysis, given that most compounds exhibited higher responses in the positive mode compared to the negative mode. For each compound, the matched score was calculated based on its accurate mass (typically [M + H]+, [M + Na]+ and [M + NH4]+ under the positive scan mode of UHPLC Q-TOF–MS, using theoretical values obtained from our personal database), isotopic abundance, and isotopic spacing. The matched compounds were further characterized based on their MS/MS data, resulting in the tentative characterization of a total of 73 compounds in NBLG. The extracted compound chromatogram (ECC) of the QC sample of NBLG is presented in Fig. 3b, and the MS and MS/MS data of the identified components are summarized in Additional file 1: Table S1. Among the identified compounds, 36 were reported for the first time in NBLG (marked with “#” in Additional file 1: Table S1). The identification of a representative compound from the lignan and alkaloid is shown as follows:

Cpd.99 (tR = 13.45 min) was identified based on an accurate [M + NH4]+ ion at m/z 532.1800 (-2.43 ppm). Its molecular formula was determined as C26H26O11, exhibiting an exclusive match with procumbenoside L in our database. The fragmentation of the molecular ion of Cpd.99 at m/z 532.1800, as illustrated in Fig. 1a, produced a predominant product ion at m/z 378.7631, resulting from the loss of the dimethoxyphenyl group, and another ion at m/z 353.0076 due to the loss of the glucose moiety. Consequently, Cpd.99 was identified as a lignan, namely procumbenoside L. Similarly, Cpd.112 (tR = 14.85 min) was identified based on its[M + H]+ ion at m/z 239.0815 (-0.14 ppm). Its molecular formula was deduced as C14H10N2O2, demonstrating an exclusive correspondence to 3-(2'-hydroxyphenyl)-4(3H)-quinazolinone in our database. Furthermore, Fig. 1b exhibits the fragmentation pattern of the molecular ion of Cpd.112 at m/z 239.0815, revealing characteristic ions at m/z 132.0448 resulting from the loss of the hydroxyphenyl group and the oxygen atom at the carbonyl group of the pyrimidine ring. Successive losses of the carbon atom from the benzene ring led to the formation of ions at m/z 120.0445 and 108.0437. Therefore, Cpd.112 was identified as an alkaloid, namely 3-(2'-hydroxyphenyl)-4(3H)-quinazolinone.

Fig. 1figure 1

MS/MS spectra and fragmentation pathway of two representative constituents. a Cpd.99 (procumbenoside L) and b Cpd.112 (3-(2-Hydroxyphenyl)-4(3H)-quinazolinone)

Confirmation of constituents using reference standards

To verify the structures of constituents identified in NBLG, eight reference standards, including adenosine (2), 2,4-(1H,3H)-quinazolinedione (31), 2-benzoxazolinone (66), acteoside (77), 1H-indole-3-carbaldehyde (83), tryptanthrine (134), indigo (138), and indirubin (144) were further employed for identification (marked with “*” in Additional file 1: Table S1), and their ECCs are presented in Fig. 3a. For instance, as shown in Fig. 2a (upper panel), Cpd.66 (tR = 10.90 min) was identified with an accurate [M + H]+ ion at m/z 136.0394 (0.73 ppm), and its molecular formula was determined as C7H5NO2. This formula exclusively matched with 2-benzoxazolinone in our database. The fragmentation of the molecular ion of Cpd.66 at m/z 136.0394 yielded predominant product ions at m/z 108.0445 and 80.0503, attributed to the successive loss of CO groups, as well as an ion at m/z 65.0388 resulting from the subsequent loss of an NH group. These findings were further validated through a comparison with the retention time, high resolution MS and MS/MS spectrum of reference standards (lower panel of Fig. 2a). Consequently, Cpd.66 exhibited identical retention time, accurate molecular mass, and fragmentation pathway as 2-benzoxazolinone, confirming its identification as 2-benzoxazolinone.

Fig. 2figure 2

Confirmation of identified constituents by comparing with corresponding reference standards based on retention time, accurate molecular mass, and fragmentation pathway. a Cpd.66 (2-benzoxazolinone) and b Cpd.77 (acteoside)

Likewise, the identification of Cpd.77 is demonstrated in Fig. 2b (upper panel). Cpd.77 (tR = 11.72 min) was identified based on an accurate [M + Na]+ ion at m/z 647.1942 (− 0.62 ppm), and its molecular formula was deduced as C29H36O15. This formula corresponded to acteoside or isoacteoside in our database. Additionally, the fragmentation of the molecular ion of Cpd.77 at m/z 647.1942 yielded predominant product ions at m/z 501.1359, attributed to the loss of the hydroxytyrosol group, and m/z 163.0389, generated from the caffeic acid moiety. These results were further confirmed through a comparison with the reference standard (lower panel of Fig, 2b). Consequently, Cpd.77 displayed identical retention time, accurate molecular mass, and fragmentation pathway as acteoside, confirming its identification as acteoside.

Chemical constituents identified in NBLG

The chemical components identified in NBLG encompass six classes, namely alkaloids, phenylpropanoids, lignans, sesquiterpene lactones, flavonoids, and other compounds. Alkaloids exhibit the highest diversity among the identified components, with a total of 50 alkaloids identified, including 19 indole alkaloids, 13 quinazoline alkaloids, two pyridine alkaloids, one oxazole alkaloid, and other alkaloids. The five most abundant alkaloids found in NBLG were 2-O-β-D-glucopyranosyl-(2H)-1,4-benzoxazin-3(4H)-one (29), 2-O-β-D-glucopyranosyl-4-hydroxy-(2H)-1,4-benzoxazin-3(4H)-one (36), 2-benzoxazolinone (66), 1,2,3,4-tetrahydro-2-oxo-4-quinolinecarboxamide (122), and baphicacanthin A (124). Additionally, ten phenylpropanoids, including acteoside (77) and its isomer (85), were identified in NBLG. Furthermore, NBLG contains four lignans, including 6'-O-(1-hydroxy-4-oxocyclohexylacety) acteoside (133), two sesquiterpene lactones, one flavonoid, and six other types of compounds. This comprehensive chemical profile of NBLG represents the most extensive analysis to date. Nevertheless, further efforts necessitate additional reference compounds for rigorous characterization.

Discrimination between NBLG and NBLJ

Due to the scarcity of NBLG resources, NBLJ has been used illegally in the market as a substitute for NBLG. A comprehensive comparison of the chemical components of NBLG and NBLJ can provide pivotal evidence for the reasonable usage of NBLJ. The chemical components in eleven batches of NBLJ, originating from the same whole plant as NBLG, were profiled using the same approach.

A total of 64 compounds were identified in NBLJ, including 55 compounds that were also present in NBLG, while nine compounds (Cpd. 17, 68, 98, 100, 115, 117, 139, 140, 142) were unique in NBLJ. The ECC of the QC sample of NBLJ is shown in Fig. 3c, and the MS and MS/MS data of the identified components are summarized in Additional file 1: Table S1.

Fig. 3figure 3

Photographs and extracted compound chromatogram (ECC) of reference standards and samples. a Eight reference standards; b Baphicacanthis Cusiae Rhizoma et Radix (NBLG); c stem of Baphicacanthus cusia (NBLJ); and d Isatidis Radix (BBLG)

To visualize the overall distribution of chemical constituents in NBLG and NBLJ, the relative abundances of six types of constituents are summarized in Fig. 4a. It can be observed that alkaloids were the most abundant class in both samples, followed by other compounds and phenylpropanoids. Although the chemical components of NBLG and NBLJ are largely similar, some notable differences exist. To identify the chemical markers responsible for distinguishing NBLG from NBLJ, an OPLS-DA analysis was performed. As shown in Fig. 4b, NBLG and NBLJ can be separated into two distinct groups, with an R2X value of 0.427, an R2Y value of 0.973, and a Q2 value of 0.930. A VIP analysis was conducted to identify the components that can differentiate between NBLG and NBLJ, resulting in a total of 34 compounds with VIP > 1 being assigned as markers for the differentiation of NBLG and NBLJ, as demonstrated in Fig. 4c. Notably, all eight of the most significantly different compounds were alkaloids.

Fig. 4figure 4

Discrimination between NBLG and NBLJ based on the chemical profiling of eleven batches of Baphicacanthus cusia (Nees) Bremek. a Comparison of relative percentages of different compound classes within NBLG and NBLJ; b OPLS-DA score plot showing the discrimination between NBLG and NBLJ (R2X = 0.427, R2Y = 0.973, and the Q2 = 0.930); c the 34 chemical markers (VIP > 1) able to differentiate NBLG from NBLJ

In summary, NBLG and NBLJ exhibit high similarity in composition but with significant differences in content. Due to the scarcity of NBLG, NGLJ is sometimes used as a substitute in the market. Therefore, a comprehensive comparison of the chemical constituents of NBLG and NBLJ can provide essential evidence for the appropriate utilization of NBLJ.

Discrimination between NBLG and BBLG

Chemical components in BBLG were profiled using the same strategy as that used for NBLG. Specifically, the ECC of the QC sample of BBLG is depicted in Fig. 3d. The analysis revealed the identification of 84 compounds in BBLG, with 54 compounds being newly identified in BBLG (as denoted by a “^” in Additional file 1: Table S1). Among these identified compounds, only 17 (Cpd. 2, 7, 15, 18, 21, 29, 32, 35, 37, 72, 75, 80, 83, 102, 112, 134, 144) were found to be common with NBLG.

Furthermore, the chart presented in Fig. 5a demonstrates a significant difference in the relative percentage of compound classes between NBLG and BBLG. Notably, the percentage of alkaloids in BBLG was found to be significantly lower than that in NBLG. Moreover, the OPLS-DA score plot (Fig. 5b) displayed a clear separation between NBLG and BBLG, with corresponding R2X, R2Y, and Q2 values of 0.643, 0.983, and 0.964, respectively. Based on these findings, a total of 51 compounds with VIP > 1 were identified as potential chemical markers for differentiating NBLG and BBLG (as shown in Fig. 5c).

Fig. 5figure 5

Discrimination between NBLG and BBLG based on the chemical profiling of eleven batches of NBLG and ten batches of BBLG. a Comparison of relative percentages of different compound classes within NBLG and BBLG; b the OPLS-DA score plot illustrating discrimination between NBLG and BBLG (R2X = 0.643, R2Y = 0.983, and the Q2 = 0.964); c the 51 chemical markers (VIP > 1) able to differentiate NBLG from BBLG

Collectively, our results demonstrate significant differences between the chemical profiles of NBLG and BBLG, providing chemical evidence for the independent recognition of NBLG and BBLG in the CHP. Importantly, the marked divergence between NBLG and BBLG suggests that BBLG cannot be used as a substitute for NBLG in clinical or manufacturing applications.

Method validation for HPLC fingerprint analysis

Seven peaks, which collectively accounted for over 90% of the total chromatographic peak area, were selected as characteristic peaks for method validation in the HPLC fingerprint analysis. Among them, peak 4 was identified as the reference peak due to its appropriate retention time and high abundance. To assess the precision, reproducibility, and stability of the HPLC fingerprint method, the RSDs of the relative retention times (RRTs) of the seven characteristic peaks were determined. The RSD values of all the RRTs were found to be below 0.8%, indicating that the established HPLC fingerprint method was repeatable and reliable.

HPLC fingerprint analysis of NBLG

The HPLC fingerprints of eleven batches of NBLG were established by selecting and normalizing the retention times of common peaks in the chromatograms, as shown in Fig. 6a. Common peaks with good stability and resolution in all batches were identified, leading to the recognition of seven characteristic peaks in the HPLC fingerprint of NBLG, as depicted in Fig. 6b. The identity of peaks 1 to 7 was confirmed through the employment of commercial standards (Fig. 6c) and UHPLC-Q-TOF–MS analysis as discussed above. The peaks were successively identified as 2-O-β-D-glucopyranosyl-(2H)-1,4-benzoxazin-3(4H)-one, 2-O-β-D-glucopyranosyl-4-hydroxy-(2H)-1,4-benzoxazin-3(4H)-one, 2-benzoxazolinone, acteoside, isoacteoside, indigo, and indirubin. The RRTs and acceptable ranges (± 5%) of the seven characteristic common peaks of NBLG were provided. The chemical structures of these peaks are demonstrated in Fig. 6d.

Fig. 6figure 6

HPLC fingerprint of NBLG. a HPLC fingerprint of eleven batches of NBLG; b reference chromatogram of eleven batches of NBLG; c HPLC chromatogram of two reference standards; d chemical structures of the seven characteristic compounds identified in NBLG

HPLC fingerprinting analysis is a widely used and valuable technique for characterizing Traditional Chinese Medicine (TCM) [18]. However, informative HPLC fingerprints for NBLG are limited, with only a few peaks observed in the chromatogram or the identification of limited components [19]. For example, the Hong Kong Chinese Materia Medica Standards (HKCMMS) displayed an NBLG HPLC fingerprint with only three characteristic peaks, including indigo, indirubin, and an unidentified component. Previous literature has reported that the bioactivities of NBLG are derived from various components. Therefore, in this study, we established an HPLC fingerprint of NBLG that included seven components from two distinct classes of components (alkaloids and phenylpropanoids). This comprehensive fingerprint can be more reasonably applied in the quality assessment of NBLG.

Selection of chemical markers for the quality control of NBLG

The current edition of the CHP does not provide a specification for the content of marker compounds in NBLG, which is crucial for ensuring the quality control of Chinese medicines. The CHP guidelines suggest that ideal marker compounds should be abundant, easily obtainable, exhibit bioactivity, and have a specific presence. However, for NBLG, previously proposed marker compounds such as indigo, indirubin, tryptanthrin, adenosine, and 4(3H)-quinazolione have relatively low content in NBLG (less than 1/10000) [20]. Moreover, indigo, one of the marker compounds, is quite unstable and needs to be detected within 8 h. As shown in our previous chemical study [21] and the profiling conducted in the current study, the chemical components of NBLG are complex and diverse, including alkaloids, flavonoids, organic acids, glycosides, pentacyclic triterpenes, amino acids, anthraquinones, carbohydrate, and sterols. Therefore, selecting a suitable marker compound that consistently exists in NBLG with relatively high abundance is essential for the establishment of an assay method.

In this study, an HPLC fingerprint of NBLG was established, revealing seven common peaks that were identified as characteristic components of NBLG and proposed as ideal chemical markers. Four of these components, namely 2-O-β-D-glucopyranosyl-4-hydroxy-(2H)-1,4-benzoxazin-3(4H)-one, 2-benzoxazolinone, acteoside, and indigo, were not detected in BBLG. Furthermore, significant differences were observed in their content between NBLG and NBLJ (Fig. 7), suggesting their potential as effective markers for distinguishing NBLG from NBLJ and BBLG. Nevertheless, the instability of indigo and the commercial unavailability of 2-O-β-D-glucopyranosyl-4-hydroxy-(2H)-1,4-benzoxazin-3(4H)-one render them less ideal markers. In contrast, 2-benzoxazolinone and acteoside are considered preferable due to their commercial availability and relative abundance. Additionally, 2-benzoxazolinone has been found to exhibit anti-influenza activity [22], while acteoside has displayed antioxidant and anti-inflammatory activity [23]. These observations suggest that these two compounds could potentially serve as active constituents of NBLG, making them suitable candidates as ideal chemical markers for NBLG.

Fig. 7figure 7

Relative content (%) of the seven characteristic compounds within NBLG, NBLJ, and BBLG. Statistical significance indicated by *: P-value < 0.05, **: P-value < 0.01, and ***: P-value < 0.001

Method validation for HPLC quantitative analysis

Validation was performed for the simultaneous determination of 2-benzoxazolinone and acteoside in NBLG. The calibration curves for both compounds exhibited excellent linearity across the test range, with R2 values exceeding 0.9999, as shown in Table 2. The LOD and LOQ for acteoside were determined to be 0.42 µg/mL and 0.85 µg/mL, respectively, while those for 2-benzoxazolinone were 0.43 µg/mL and 0.88 µg/mL, respectively. The precision, repeatability, and stability of both compounds exhibited RSD values below 2%. To assess the recovery, the found amounts of 2-benzoxazolinone and acteoside were in the range of 447.18 to 458.05 µg and 754.48 to 770.86 µg, respectively, resulting in mean recoveries of 102.7% and 97.2%, respectively, with both RSD values below 1%. All parameters met the acceptance criteria, indicating the reliability and accuracy of the method for determining 2-benzoxazolinone and acteoside in NBLG.

Table 2 Regression equation, linear range, limit of detection (LOD), limit of quantification (LOQ), precision, repeatability, stability, and recovery of 2-benzoxazolinone and acteoside in NBLGQuantitative determination of the two chemical markers

The developed HPLC method was employed to quantify the content of 2-benzoxazolinone and acteoside in eleven batches of NBLG. As shown in Table 3, the content of 2-benzoxazolinone in NBLG ranged from 61.90 to 548.70 μg/g, with a corresponding content range in dry weight of 0.007–0.062%, and a mean value of 0.023%. For acteoside, its content in NBLG varied from 47.20 to 2219.30 μg/g, with a corresponding content range in dry weight of 0.005–0.546% and a mean value of 0.102%. It is important to note that our findings revealed significant fluctuations in the content of 2-benzoxazolinone and acteoside among different batches of NBLG. Despite standardized processing and storage conditions for all samples, the wide variations among batches are likely attributed to the diverse growth environments and harvesting seasons of the NBLG samples [24]. These influential factors necessitate further investigation in future studies.

Table 3 Contents of 2-benzoxazolinone and acteoside in eleven batches of NBLG (n = 2)

Based on these results, we propose a new quality control standard for NBLG, which specifies that the limit of content in dry weight for both 2-benzoxazolinone and acteoside should not be less than 0.010%. Currently, there are no explicit regulation regarding the component content of NBLG, and the content of 50% ethanol extract remains the primary indicator of NBLG’s quality [16]. Therefore, our proposed standard would provide a crucial benchmark for assessing the quality of NBLG.

留言 (0)

沒有登入
gif