Comparison of Anthraquinones, Iridoid Glycosides and Triterpenoids in Morinda officinalis and Morinda citrifolia Using UPLC/Q-TOF-MS and Multivariate Statistical Analysis

Roots of Morinda officinalis and Morinda citrifolia have been interchangeably used in traditional Chinese medicine. However, there is no experimental evidence to support this. In this study, a ultra-performance liquid chromatography quadrupole time-of-flight mass spectrometry (UPLC/Q-TOF-MS)-based approach and a multivariate statistical analysis (MSA) were adopted to compare the difference in the chemical compounds present in the root extract of M. officinalis and M. citrifolia. There were 26 anthraquinones, 15 triterpenes, and 8 iridoid glycosides identified in the root extracts of M. officinalis, 30 anthraquinones, 1 triterpene, and 8 iridoid glycosides in the root extracts of M. citrifolia. Among these, 25 compounds presented in both plants. In addition, a principal component analysis (PCA) showed that these two herbs could be separated clearly. Furthermore, an orthogonal partial least squares-discriminant analysis (OPLS-DA) found 9 components that could be used as chemical markers to discrimination the root extracts of M. officinalis and M. citrifolia. In addition, the results of a Cell Counting Kit 8 (CCK-8) assay and cell colony formation assay indicated that methanol root extracts of M. officinalis and M. citrifolia showed no cell cytotoxicity to normal cells, even promoted the proliferation of normal liver cells. To our knowledge, this is the first time that the differences between the root extracts of M. officinalis and M. citrifolia (Hainan province) have been observed systematically at the chemistry level.


Introduction
Morinda officinalis (M. officinalis) and Morinda citrifolia (M. citrifolia) belong to the family Rubiaceae [1]. The roots of M. officinalis have long been used ass traditional medicine for the strengthening of bones and immunity, and the nourishment of the kidneys [2]. It has also been used to treat impotence, osteoporosis, depression and inflammatory diseases, such as rheumatoid arthritis and dermatitis [2]. Moreover, the clinical effects of M. officinalis have been proved by modern pharmacological studies [3][4][5][6]. The fruit of M. citrifolia is also called Noni in Hawaii and the juice made of it has been used as
Compounds 37, 45, 48, 50, 52-55, 57-63 were determined as triterpenes, and their structures are displayed in Figure 4 [20]. The structure of triterpenes is relatively stable, and it commonly formed ions that were produced by the loss of a unit of H 2 O. To date, approximately 100 compounds have been isolated from M. officinalis, which mainly are polysaccharides, oligosaccharides, anthraquinones, triterpenes, and iridoid glycosides [2]. Approximately, 160 compounds have been isolated and identified from M. citrifolia, which mainly comprise anthraquinones, organic acids, alkaloid, terpenes and volatile compounds [11]. In the present study, 49 compounds from the methanol extract of M. officinalis roots were identified, classified into 26 anthraquinones, 15 triterpenes, and 8 iridoid glycosides. At the same time, 39 compounds were identified from the methanol extract of M. citrifolia roots, including 30 anthraquinones, 1 triterpene, and 8 iridoid glycosides. Among them, 25 compounds were found to be present in both plants. These compounds are classified into 16 anthraquinones, 1 triterpene, and 8 iridoid glycosides. Detailed information on the identified compounds is shown in Table 1.  The compounds with significant differences between groups were called marker compounds, and were plotted at the top right (1) and bottom left (-1). Therefore, nine ions (red box mark) were selected as candidates, to chemically distinguish M. officinalis from M. citrifolia root extracts ( Figure 5B). The variable importance in projection (VIP) plot ( Figure 5C) showed that all the 7 selected potential marker compounds in M. citrifolia root extracts and 2 selected potential marker compounds in M. officinalis root extract have a high VIP value (VIP > 3), suggesting that these marker compounds are largely responsible for the chemical difference between these two samples. Furthermore, the variable average between groups clearly showed content discrepancy of the selected marker compounds in M. citrifolia and M. officinalis root extracts ( Figure 5D). The 9 selected markers were asperulosidic acid (6) Figure 5D). The other seven compounds were relatively higher in M. citrifolia roots. Among them, compounds 11, 15, and 40 are anthraquinones, which are the characteristic compounds of the family Rubiaceae. Asperulosidic acid (6, iridoid glucosides) is one of the most common pharmacologically active ingredients in M. officinalis [2]. Previous reports found that it exhibited anti-inflammatory, anti-renal fibrosis, and antibacterial activities [21,22].

The Cytotoxicity of the Methanol Extracts of M. officinalis and M. citrifolia Roots
To further investigate the differences in the methanol extracts of M. officinalis and M. citrifolia roots in terms of cytotoxicity, a CCK-8 assay was used to examine the cell viability of three liver cell lines (normal cells: LO2; cancer cells: HepG2, SMMCH771) after treatment by M. officinalis and M. citrifolia root extracts. As shown in Figure 6A, the two extracts were non-cytotoxic to the cells after a treatment of 24 h at concentrations of up to 100 µg/mL. When treated for 48 and 72 h, lower cell viability was observed in M. citrifolia root extract treatment groups compared to M. officinalis root extract treatment groups. In addition, a colony formation assay was used to evaluate the long-term impact of the M. officinalis and M. citrifolia root extracts on the cells. The results showed that a high dose of M. officinalis and M. citrifolia root extracts promoted the cell proliferation of LO2 cells, and inhibited the cell proliferation of HepG2 and SMMCH771 cells ( Figure 6B). No significant differences were found between the root extracts of M. officinalis and M. citrifolia in terms of long-term impact on HepG2 and SMMCH771 cells. Hence, it is suggested that the methanol extracts of M. officinalis and M. citrifolia roots showed no cell cytotoxicity to normal cells, and even promoted the proliferation of normal liver cells. Meanwhile, compared to M. officinalis root extract, M. citrifolia root extracts exhibited stronger cell cytotoxicity to liver cancer cells. Numerous studies indicated that M. citrifolia extracts exhibited anticancer activity [23]. For instance, the leaf extract of M. citrifolia has an inhibitory effect on lung cancer through strengthening immune response [24]. Its fruit extracts induced apoptosis and suppressed migration in human liver and breast cancer cells [25]. In contrast, few publications report the antitumor activity of M. officinalis. Our results are consistent with previous reports that M. officinalis root extracts show poor cell cytotoxicity toward cancer cells, compared with M. citrifolia root extracts. HPLC-grade methanol, acetonitrile, and water were from Fisher (Pittsburgh, PA, USA). Formic acid (LC/MS grade) was purchased from Merck Millipore (Darmstadt, Germany). ACS grade sodium hydroxide was obtained from Sigma-Aldrich (St. Louis, MO, USA). Distilled water was purchased from Watsons (Guangzhou, China). Lastly, ultrapure water was obtained through Milli-Q purification system (Merck Millipore, Darmstadt, Germany). Other solvents and reagents used were of analytical grade.

Plant Materials
The The plant materials were air-dried and ground for the extraction procedure.

Sample Preparation
The air-dried roots of M. officinalis and M. citrifolia were ground and passed through an 80-mesh sieve. Each powdered sample (1.0 g) was mixed with methanol (50 mL) and then sonicated for 1 h at a temperature of 40 • C. The extract was subjected to centrifugation (12,000 rpm, 10 • C) for 10 min. The supernatant was passed through a 0.2 µm polytetrafluoroethylene (PTFE) syringe filter (VWR Bridgeport, PA, USA) and transferred to a 2 mL transparent vial [19].

Chromatographic Conditions
The separation was conducted on a Waters ACQUITY UPLC system with an ACQUITY UPLC HSS T3 C18 column and 0.1% aqueous formic acid (A) and acetonitrile with 0.1% formic acid (B) was applied as the solvent system. The flow rate was 0.3 mL/min, and each sample solution was injected 1 µL. The column temperature was controlled at 40 • C.

Mass Spectometry Conditions
A Waters Xevo G2 QTOF equipped with Z-Spray electro spray ionization (ESI) source and MassLynx 4.1 was used for MS data acquisition. Negative ionization mode was applied to acquire data under a mass range of 50−1200 Da with a scan time of 0.2 s and detection time of 22 min; furthermore, both low-energy (function 1) and high-energy (function 2) scan functions were used. The collision energy was 6 V for the low-energy scan function, and 10-45 V for the high-energy scan function. The capillary voltages were set at 2.0 kV, and the sampling cone voltage at 40 V. The source and desolvation temperatures were 100 • C and 400 • C, respectively. The desolvation gas flow rate was 600 L/h, while the cone gas flow rate was 50 L/h.

Mass Data Processing and Analysis
Progenesis QI software was utilized to convert the data of UPLC/Q-TOF-MS, i.e.,; then, the significant differential retention time-exact mass (RT-EM) pairs in the S-plots were picked and exported back into Progenesis QI for compound structural elucidation. The resultant data matrices were subsequently exported to EZinfo 2.0 software (MassLynx v4.1, Waters) for PCA (principal component analysis) and an orthogonal partial least squares-discriminant analysis (OPLS-DA) to examine differences in the samples. All variables obtained from UPLC-MS datasets were mean-centered and scaled to Pareto variance before subjection to PCA and OPLS-DA. The OPLS-DA score plots were depicted by the cross-validation parameter R2Y and Q2, which signify the total explained variation for the X matrix and the predictability of the model, respectively. The sum of squares of the PLS weights was evaluated by the value of VIP (variable importance in projection), showing the relative contribution of each X variable in the model. The variables with VIP >3 were considered influential for the separation of samples in the score plots from the PLS-DA analysis [26,27].

Chemical Composition and Biomarker Identification
Multiple approaches were applied to identify chemical components and biomarkers, which include; comparison with retention times, accurate molecular ions, and characteristic fragment ions of reference compounds; reported data of the same compounds in the literature; online Traditional Chinese Medicine (TCM) Chinese [UNIFI1.7]; and ChemSpider [27].

Cell Viability Assay
Cells were planted into 96-well plates. Various concentrations of the extracts (0.4, 1.2, 3.7, 11.1, 33.3 and 100 µg/mL) were added into the culture media for 24 h, 48 h, and 72 h. At 1 h before each time point, a 10 µL volume of CCK-8 reagent was added to each well. Optical density values at 450 nm were read with a microplate reader (Elx808, Bio Tek, Winooski, VT, USA) and analyzed with GraphPad Prism 7.0 software (GraphPad, San Diego, CA, USA).

Cell Colony Formation Assay
Cells were plated into 12-well plates in a triplicate manner. The root extracts of M. officinalis and M. citrifolia (25,50, and 100 µg/mL) were added to the plates. Culture media were replaced every third day. After 10 days, the culture medium was discarded, and cells were washed by phosphate-buffered saline (PBS). Then, cells were fixed with 4% paraformaldehyde (Beyotime Biotechnology, Shanghai, China) for 30 min, and stained with crystal violet (Beyotime) for 30 min, followed by a wash with distilled water. The colonies were imaged by a HP scanner and counted under an inverted phase contrast microscope.

Data Analysis
Cell experiments were performed in triplicate, and the results are presented as the mean value ± standard error of the mean (SEM). Data were analyzed using GraphPad Prism 7.0 software (GraphPad, San Diego, CA, USA), and p < 0.05 was considered statistically significant.

Conclusions
In the present study, a total of 49 compounds were identified from the methanol extract of the root of M. officinalis and 39 compounds were identified from the methanol extract of the root of M. citrifolia by a UPLC-QTOF-MS/MS analysis. We found 25 ingredients that were present in both plants; they consisted of anthraquinones, triterpene, and iridoid glycosides. Using a QI analysis, we identified 9 components that could be used as chemical markers to distinguish the root extracts of M. officinalis from those of M. citrifolia introduced from Hainan Province. In addition, we found that M. officinalis and M. citrifolia exhibited no cell cytotoxicity towards normal cells, and even promoted the proliferation of normal liver cells. Our study suggests that M. officinalis and M. citrifolia roots have similarities in terms of chemical composition and biological activity. However, further research is needed to explore the feasibility and substitutability of the clinical application of M. officinalis and M. citrifolia.