Abstract

Canna edulis Ker-Gawl is a versatile crop that integrates food, energy, and feed in China. The rhizome is a traditional Chinese medicine with a long history. However, there are few studies on chemical components. Crude plant extracts are a complex mixture of several biologically active secondary metabolites. Therefore, rapid and accurate identification and quantification are critical in phytochemical analysis. An efficient approach based on ultrahigh-performance liquid chromatography coupled with Q-Exactive Orbitrap tandem mass spectrometry (UHPLC Q-Exactive Orbitrap MS) was used to analyze the chemical components systematically. We used retention time, accurate molecular weight, parent peaks, fragment peaks, fragmentation characteristics, comparative analysis with the literature, reference standards, and standard databases, including ChemSpider, mzVault, MassBank, and mzCloud. A total of 54 chemical constituents were identified from the methanol extract of the rhizome, including 36 organic acids (with six phenolic acids), three phenols, one sugar (and its derivatives), one coumarin, one triterpenoid, ten N-containing organic compounds (including seven amino acids), and two other compounds. Moreover, three compounds were quantitatively analyzed and a methodological study of the three components was carried out. The established method provided satisfactory precision and accuracy, acceptable recovery, good linearity, and a reasonable detection limit. The UHPLC Q-Exactive Orbitrap MS method was used for the first time to accurately, quickly, and systematically characterize the compounds of the rhizome, revealing the pharmacodynamic substances and providing a theoretical basis for its quality control, in-depth research, and development.

1. Introduction

Canna edulis Ker-Gawl is a perennial herb from the genus Caanna (Cannaceae). Canna edulis is rich in germplasm resources, which can be found in Guizhou province, Yunnan province, Guangxi province, and other places [1]. The rhizome of Canna edulis Ker-Gawl is shown in Figure 1. It is a traditional herb used as medicine and food in China with considerable economic, social, ecological, and medicinal benefits. The herb contains rich trace elements, amino acids, vitamins, and other nutrients [24]. The dry rhizome contains 70–80 g/100 g of starch [5, 6] that is commonly used for functional food and dietary fiber [711] or fermentation of industrial alcohols [1214].

Canna edulis Ker-Gawl was listed as a classical medicinal plant in “Miao medicine” and “Dai medicine,” a branch of traditional Chinese medicine (TCM). It is also called “jiao-yu,” “jiang-yu” (Guangdong Province), “Jiang-ya” (Guangdong, Guizhou Province), “Ba-jiao-ya,” “Xiang-zhu,” “Man-dong” (Dai Medicine), “Hong-jiang-qi” (Miao Medicine), and “Han-ou” in ethnopharmacology. TCM has accumulated information on the use of the herb in ancient manuscripts and recently published books, including “Zi Yuan Zhi,” “Xiang Lan Kao,” “Dian Yao Lu,” “Dai Yao Lu,” “Gui Yao Bian,” and “Zhong Hua Ben Cao.” According to ancient Chinese herbal literature, Canna edulis Ker is sweet and light and has a cold nature. The rhizome clears heat and dampness, protects the liver and gallbladder, cools the blood and detoxifies, and invigorates the spleen and stomach (derived from: http://www.plant.csdb.cn/herb). In China, the herb has been used as a local medicine to treat dysentery, diarrhea, acute icteric hepatitis, metrorrhagia, leukorrhea, irregular menstruation, hemoptysis, and sore swollen poison syndrome.

Modern pharmacological studies have found that Canna edulis Ker -Gawl protects the liver and gallbladder, treats acute jaundice hepatitis, has a gastroprotective effect [15], lowers blood pressure, lowers blood glucose, lowers blood lipids, prevents colon cancer [16], and cardiovascular disease [17, 18] and has antioxidant activities [19, 20]. Previous studies from our group [2125] found that the resistant starch from the rhizome of Canna edulis Ker (Ce-RS3) after 11 weeks of Ce-RS3 intervention showed antidiabetic effects similar to metformin, significantly reduced blood glucose, reduced insulin resistance, increased glucose tolerance, and reduced pathological damage. In addition, Ce-RS3 significantly reduced body weight and dyslipidemia in obese mice.

Studies on Canna edulis Ker-Gawl have focused on cultivation, starch extraction, transformation, and utilization [2628], and there are few studies on its chemical composition. Only about thirty monomers have been obtained by systematic phytochemical enrichment and separation [29]. For example, through the analysis of the chemical components of the ethyl acetate-soluble residue of the methanolic extract derived from dry rhizomes of Canna edulis, Young Sook Yun et al. [30] isolated two phenylpropanoid sucrose esters together with a known phenylpropanoid sucrose ester and four known phenylpropanoids, i.e., caffeic acid, rosmarinic acid, caffeoyl-4′-hydroxyphenyllactic acid, and salvianolic acid B. Through the analysis of the residue of Canna edulis, Zhang et al. [31] isolated 11 compounds from water-soluble extract, i.e., rosmarinic acid, salvianolic acid B, ferulic acid, caffeic acid, 1-caffeoylquinic acid, 3-caffeoylquinic acid, 4-caffeoylquinic acid, 5-caffeoylquinic acid, salicylic acid, and gallic acid. Moreover, by analyzing and summarizing a large number of documents, Zhang et al. [32] reviewed the chemical constituents of Canna edulis and found that the main secondary metabolites of are phenolic compounds.

With the development of the pharmaceutical health and food industry, increasing attention has been paid to the study of Canna edulis Ker -Gawl. UHPLC Q-Exactive Orbitrap MS has excellent resolution, sensitivity, and performance and enables rapid compound identification, structure confirmation, and quantitative target analysis [33, 34]. Therefore, for the first time we used UHPLC Q-Exactive Orbitrap MS analysis technology to perform a comprehensive and systematic chemical composition analysis to provide a basis for further research and development of medicinal substances and quality control.

2. Materials and Methods

2.1. Instruments and Software

UHPLC Q-Exactive Orbitrap MS system: Vanquish Flex UHPLC series Q-Exactive high-resolution mass spectrometer (Themo Fisher Scientific, USA), Ultrasonic apparatus (Shenzhen Guanyijia Technology Co., LTD.), Electronic Balance (Cedis Scientific Instruments (Beijing) Co., Ltd.), Centrifuge (Hunan Xiangyi Laboratory Instrument Development Co., LTD.), Xcalibur 4.4 Workstation (Themo Fisher Scientific, USA), Compound Discoverer3.1 Compound Analysis and Identification Software (Thermo Fisher Scientific, USA), CORTECS UPLC C18 column (2.1 × 100 mm, 1.6 μm) was purchased from Waters.

2.2. Reagents and Materials

Methanol, Acetonitrile, Formic acid were both in mass grade, which were purchased from Thermo Fisher & Fisher Technology Co., Ltd. (China). Distilled Water was obtained from Guangzhou Watsons Food & Beverage Co., Ltd. (China).

Caanna edulis Ker comes from Xingyi medicinal materials planting base, Guizhou Province, acquired in November 2021 and identified as the rhizome of Canna edulis Ker by professor Liu Chunsheng of Beijing University of Chinese Medicine.

Caffeic acid (purity ≥ 98%), ferulic acid (purity ≥ 98%), quinic acid (purity ≥ 98%), citric acid (purity ≥ 98%), and palmitic acid (purity ≥ 98%) were obtained from Shanghai Yuanye Biotechnology Co., Ltd. (China).

2.3. Preparation of Sample and Reference Solutions

20 g of Canna edulis Ker powder was accurately weighed by electronic analytical balance. After 10 mL of methanol was added, an extract was obtained by sonication for 2 h at 25°C. The supernatants were centrifuged at 12 000 r·min−1 for 10 min. Then, using a pipette to remove and filter through a 0.22 μm nylon millipore filter and added to the liquid vial for further analysis.

The reference solutions including caffeic acid, ferulic acid, quinic acid, citric acid, and palmitic acid were dissolved in methanol with appropriate amounts, respectively, as a single control reserve liquid. All the reference solutions were stored at 4°C before analysis.

2.4. Chromatographic Conditions

The UHPLC separation was performed on Vanquish Flex UHPLC System (Thermo Fisher Scientific, San Jose, CA, USA) with a Waters CORTECS UPLC C18 column (2.1 × 100 mm, 1.6 μm) maintained at 35°C. Mobile phases were acetonitrile (A) and 0.1% formic acid aqueous solution (B), with the following gradient elution procedure: 0 min, 10% A; 0 ⟶ 4 min, 10% ⟶ 23% A; 4 ⟶ 15 min, 23% ⟶ 41% A; 15 ⟶ 31 min, 41% ⟶ 90% A; 31 ⟶ 33 min, 90% A; 33-33.1 min, 90% ⟶ 10% A; 33.1 ⟶ 34 min, and 10% A. The flow rate of elution solvent was 0.3 mL·min−1 and injection volume of samples was 5 μL.

2.5. Mass Spectrometry Conditions

MS detection was carried out under the positive and negative ion mode with heated electrospray ionization (HESI) source. The scan mode was full scan/dd MS2. The mass range and the mass resolution were set at 100−1 200 Da and 70 000 FWHM. The operating parameters for positive ionization mode were: spray voltage 3 800 V and capillary temperature 350°C. The flow rate of sheath gas and auxiliary gas were set to 35, 15 (arbitrary units), respectively. As for the negative ionization mode, the operating parameters were as following: spray voltage 3 000 V and capillary temperature 320°C. The flow rate of sheath gas and auxiliary gas were set to 35 and 10 (arbitrary units), respectively.

2.6. Methodological Validation of the Quantitative Analysis

The quantitative analysis established in this study was validated regarding linearity, limit of detection (LOD), and quantification (LOQ), precision and recovery.

2.6.1. Preparation of the Standard Solutions, Linearity, and Sensitivity

We prepared D- (-)-quinic acid, caffeic acid, and ferulic acid stock solutions at 20 mM, respectively. A series calibration standard solutions were prepared by diluted and mixed stock solutions. In order to plot calibration curves, seven standard working solutions at different concentrations were achieved by serially mixing and diluting the stock solutions. The molar concentrations of each component to be measured were quinic acid 0.781 μM, 1.563 μM, 3.125 μM, 6.250 μM, 12.500 μM, 25.000 μM, and 50.000 μM; caffeic acid 7.813 μM, 15.625 μM, 31.250 μM, 62.500 μM, 125.000 μM, 250.000 μM, and 500.000 μM; ferulic acid 0.260 μM, 0.521 μM, 1.042 μM, 2.084 μM, 4.168 μM, 8.335 μM, and 16.667 μM. The calibration curves were constructed by plotting peak area (y) against concentration (x, μM). The LODs and LOQs were determined based on the signal-to-noise (S/N) of 3 and 10, respectively.

2.6.2. Precision

The intraday precision for each compound was determined by analyzing one standard solution samples (n = 6) on the same day (intraday).

2.6.3. Accuracy and Recovery

The standard compound mixtures in three different concentrations (low, middle and high) at known and particular concentrations were added into corresponding untreated samples. The samples were treated and analyzed as described above to determine recoveries of the three components. Three parallel samples were prepared. The RSD% of the areas and contents were calculated. The data were input into GraphPad Prism (version 9.0) for the quantification analysis.

2.7. Data Analysis

Data acquisition and processing were based on the elemental compositions of the precursors. Raw mass spectrum data collected were imported into Compound Discoverer 3.1 (Thermo Fisher Scientific, USA) for data processing. We recorded peak extraction, retention time correction within and between groups, additive ion merging, missing value filling, background peak labeling, and metabolite identification. To identify compounds, we compared the information to databases, including ChemSpider, mzVault, MassBank database (http://www.massbank.jp/Index), and mzCloud (https://www.mzcloud.org/). Then, the exact mass coupled with MS/MS spectrum was matched with the library databases. Finally, the fragmentation patterns of components were illustrated using ChemDraw (version 18.0). For the quantification analyze, the data were input into Graphpad Prism (version 9.0).

3. Results

3.1. Base Peak Ion Chromatogram

The base peak chromatogram of Caanna edulis Ker -Gawl and references are shown in Figure 2.

3.2. Identification Results of Chemical Composition

UHPLC Q-Exactive Orbitrap MS was used for components profiling of the methanol extract of the rhizome of Canna edulis Ker-Gawl. Fifty-four chemical constituents were identified from the methanol extract of the rhizome, with 36 organic acids (including six phenolic acids), three phenols, one sugar (and its derivatives), one coumarin, one triterpenoid, ten N-containing organic compounds (including seven amino acids), and two other compounds. The retention time, molecular ion peak, molecular weight, molecular formula, and proposed compounds are presented in Table 1. Our results from UPLC-LTQ-Orbitrap-MS showed that accurate mass error values were below 5 ppm.

3.3. Identification Process

The identification process of the compounds was as follows.

3.3.1. Organic Acids

We identified 36 organic acid components, including six phenolic acids. Organic acids contain -COOH functional groups that are likely to remove small molecule groups such as H2O, CO2, and -CH3. We took peaks 3, 5, 7, 8, 11, 13, 14, 15, 18, and 20 as examples to explain its lysis rules.

Peak 20, tR 4.650, C10H10O4, was deprotonated in the negative ion mode to produce ion m/z 193.050 51 [M-H], then lost a CH3·group and obtained ion 178.027 33 [M-H-CH3·], or lost a CO2 group and obtained ion 149.024 55 [M-H-CO2]. If both CH3· and CO2 were lost, the fragment of 134.024 52 [M-H-CO2-CH3·]−· was obtained. Compared with the reference, peak 20 was identified as ferulic acid.

Peak 11, tR 1.595, C6H6O2, produced ion m/z 109.029 49[M-H] in negative mode. Further cleavage produced the secondary fragment ion m/z 91.018 82 [M-H-H2O], 81.034 61 [M-H-CO], 65.014 47 [M-H-CO2]. Combined with the standards online database and the information in the MassBank database and literature, peak 11 was identified as catechol.

Peak 13, tR 2.007, and peak 14, tR 2.167, C4H6O5, were easily deprotonated in the negative ion mode to produce ion m/z 133.014 22 [M-H]. The main fragment ion peaks in the secondary mass spectrometry were m/z 115.003 65 [M-H-H2O], 89.024 48 [M-H-CO2], 87.008 74, 72.993 16, and 71.013 90 [M-H-H2O-CO2]. Combining the information with the standards online database, the MassBank database, and the literature, peaks 13 and 14 were identified as malic acid (i.e., they are enantiomers).

Peak 3, tR 0.951, showed ion at [M-H]-m/z 191.056 12 in negative ion mode. Further cleavage produced the secondary fragment ion m/z 173.045 41 [M-H-H2O]-, 127.039 98 [M-H-H2O-CH2O2], 111.008 74 [M-H-2H2O-CO2], 87.008 74, and 85.029 48 [M-H-2H2O-CO2-C2H2]. Combining the retention time, fragment information, and characteristic with reference, peak 3 was identified as D- (-) -quinic acid. The secondary mass spectral profile of peak 3 is shown in Figure 3.

Peak 5 was found at 1.000 min, possessing the quasi-molecular ion [M-H] at m/z 191.019 74. Further cleavage produced the secondary fragment ion m/z 173.046 17 [M-H-H2O], 146.938 60 [M-H-CO2], 129.019 18 [M-H-H2O-CO2], 111.008 75 [M-H-2H2O-CO2], 102.948 79, 87.008 76, and 85.029 50 [M-H-H2O-2CO2]. Combining the retention time, fragment information, and characteristics with reference, peak 5 was identified as citric acid. The secondary mass spectral profile of peak 5 is shown in Figure 4.

Peak 7 was found at 1.126 min and yielded parent ion [M-H]m/z 117.019 33 in negative ion mode. Further lysis of secondary fragment ion m/z 99.008 74 [M-H-H2O], 73.029 46 [M-H-CO2], compared with the standard database, information from MassBank database and references [35], peak 10 was identified as succinic acid. The secondary mass spectrogram of peak 7 is shown in Figure 5.

Peak 8, tR 1.191, had the deprotonated ions at m/z 169.014 14 [M-H] in negative ion mode. The daughter ion at m/z 125.024 35 is attributed to the loss of CO2. Combined with the standard database, the MassBank database, and literature information [36], peak 8 was identified as gallic acid. The secondary mass spectral profile of peak 8 is shown in Figure 6.

Peak 15 was found at 2.228 min and yielded parent ion [M-H] at m/z 153.019 32. Further cleavage generated the daughter ion m/z 123.045 17, 109.029 45 [M-H-CO2], and 108.021 45 [M-H-CO2-H·]−·. Combined with the standard database and the information in the MassBank database and literature [37], peak 15 was identified as a gentisic acid. The secondary mass spectral profile of peak 15 is shown in Figure 7.

Peak 18 (tR 2.859) gave [M-H] at m/z 179.034 96 in negative ion mode. The daughter ion at m/z 135.045 18[M-H-CO2] was attributed to the loss of CO2. In addition, compared with the information with reference, peak 18 was identified as caffeic acid. The secondary mass spectral map of peak 18 is shown in Figure 8.

3.3.2. Amino Acids

Seven amino acids were identified in the samples. Amino acids often contain -NH2 and -COOH functional groups. Generally, amino acids lose H2O or -COOH, and some amino acids lose NH3. Peaks 4 and 50 are used as examples to explain the lysis rules.

Peak 4 was found at 0.978 min and yielded the parent ion [M-H] at m/z 131.045 94. Further cleavage produced the daughter ion m/z 114.019 65 [M-H-NH2], 113.035 70 [M-H-H2O], 111.020 13, 87.056 59[M-H-CO2], 72.009 07, and 70.029 80 [M-H-CO2-NH3]. Retention time, fragmentation information, and feature peaks matched the information in a local database (mzVault) and the HMDB database. Thus, peak 4 was identified as asparagine. The secondary mass spectral profile of peak 4 is shown in Figure 9.

Peak 50 (tR 0.987) was found in positive ion mode, generating [M + H]+ at m/z 166.086 03. Further lysis produced m/z 131.049 27, 120.080 99, 103.054 60, 93.070 36. Combined with the reference and MassBank database, peak 50 was identified as L-phenylalanine. The secondary mass spectral profile of peak 50 is shown in Figure 10.

3.3.3. Fatty Acids

Two fatty acids were identified (palmitic and oleic acid). Palmitic acid was identified by combining with reference and database and was found at 31.420 min in negative ion mode, yielding parent ion [M-H] at m/z 255.233 17. Further lysis produced secondary fragment m/z 236.902 79 [M-H-H2O]. Peak 43 had the corresponding character when compared with references and databases. Therefore, peak 43 was identified as palmitic acid. Its secondary mass spectrometry is shown in Figure 11. Oleic acid was identified by combining the standard database, MassBank database, and related literature [38]. Peak 44 was found at 31.830 min, produced [M-H] peak at m/z 281.248 75, and matched the database and literature information. Thus, peak 44 was identified as oleic acid.

3.4. Method Validation for Quantitation of Three Compounds

Regression equations were obtained by plotting corresponding peak areas versus different concentrations. All the regression equations exhibited excellent linearity, the values of linear ranges (r2) from analytical curves were >0.999 and linearity equations were listed in Table 2. RSDs of the precision test ranged from 0.9% to 1.7%. In addition, the accuracy of the proposed method was assessed. The results indicated that the UHPLC Q-Exactive Orbitrap method possessed good accuracy with recoveries ranging from 103.0% to 104.8%, while all RSDs were less than 3%. The optimized and validated UHPLC Q-Exactive-Orbitrap method was used for quantitative analysis. The contents of the three compounds are listed in Table 2.

4. Discussion

UHPLC Q-Exactive Orbitrap MS technology is characterized by high resolution, fast analysis speed, and good sensitivity. It is a necessary means to analyze small molecule compounds. Using UHPLC Q-Exactive Orbitrap MS technology for the first time, this study systematically characterized the chemical profile of Canna edulis Ker-Gawl. After optimizing the method, we selected pure methanol as the extraction solvent and acetonitrile 0.1% formic acid water as the mobile phase. In this study, we identified 54 compounds from methanol extracts and carried out quantitative analysis and methodological validation on three of them. This study identified more peaks and compounds in negative ion mode. Forty-five chemical components (mainly organic acid components) were identified in negative ion mode, and nine chemical components (mainly amino acid components) were identified in positive ion mode. Most of the identified compounds have good biological activity, such as, caffeic acid, gallic acid, gentisic acid, ferulic acid, and other phenolic compounds. Moderate intake of these compounds can present promising effects in the prevention of diseases [17, 3941] such as diabetes, obesity, Parkinson’s, cardiovascular disease, and others. In view of antioxidant activity, these compounds can be developed as natural food additives [31, 42, 43].

Plants of the homologous family contain parent nuclei of similar compositions. According to the literature, Canna edulis Ker is a cultivarietas of Canna indica L. Studies have found that Canna indica Linn. (Cannaceae) contains flavonoids, polyphenols, essential oils, and anthocyanins [4449]. These chemical components were not identified in this study. The possible reasons are changes in chemical composition due to processing, low content, and difficulty ionizing in this ion mode. To characterize the components of Canna edulis Ker-Gawl, molecular networks with the GNPS database and other technical methods can be used.

5. Conclusion

In this study, the UHPLC Q-Exactive Orbitrap MS method was used for the first time to accurately, quickly, and systematically characterize the compounds of the rhizome of Canna edulis Ker-Gawl. 54 chemical constituents were identified from the methanol extract of the rhizome, including 36 organic acids (with six phenolic acids), three phenols, one sugar (and its derivatives), one coumarin, one triterpenoid, ten N-containing organic compounds (including seven amino acids), and two other compounds. Moreover, three compounds were quantitatively analyzed and a methodological study was carried out. The established method provided satisfactory precision and accuracy, acceptable recovery, a good linearity and a reasonable detection limit. We preliminarily revealed the pharmacodynamic substance basis of this rhizome and provided a basis for quality control, in-depth research, and development.

Data Availability

The data used to support the findings of this study are included within the article.

Conflicts of Interest

The authors declare that there are no conflicts of interest regarding this study.

Acknowledgments

This study was funded by the Special Investigation of Science and Technology Basic Resources (Grant no. 2018FY100704) and Science and Technology Program Project of Guizhou Province (Grant no. [2020]4Y074).