Three New Compounds, Licopyranol A–C, Together with Eighteen Known Compounds Isolated from Glycyrrhiza glabra L. and Their Antitumor Activities

Glycyrrhiza glabra L., known as licorice, is one of the most famous herbs in the world. In this study, we investigated the phytochemical and antitumor activities of G. glabra, especially its anti-colorectal cancer activities. G. glabra was extracted with 70% methanol, and the ethyl acetate layer was separated by silica gel, ODS, LH-20 column chromatography, and semi-preparative HPLC to obtain the compounds. The structures were determined by NMR and MS methods. Three new compounds named licopyranol A–C (1–3), and eighteen known compounds (4–21) were isolated. Compounds with an isoprenyl group or dimethylpyran ring showed better antitumor activities. Licopyranol A (1) and glycyrol (5) both inhibited the proliferation, reduced clone formation and promoted apoptosis of RKO cells. The Western blotting assays showed that glycyrol significantly reduced the expression of E-cadherin, β-catenin, c-Myc, and GSK-3β proteins in RKO cells, suggesting that glycyrol may inhibit the growth of colorectal cancer RKO cells via the Wnt/β-catenin signaling pathway.


Introduction
The genus of Glycyrrhiza, which belongs to the Fabaceae family, consists of about 20 species from all over the world [1]. Among these species, Glycyrrhiza uralensis Fisch, Glycyrrhiza inflata Bat. and Glycyrrhiza glabra L. are used as licorice (Gancao) in traditional Chinese medicine to treat various diseases, such as cough, influenza, diabetes, and cancer, etc. G. glabra is also one of the most commonly used herbs in the world, which possesses hepatoprotective, anti-inflammatory, neuroprotective, antioxidant, and antiviral activities [2]. Phytochemical studies have revealed that G. glabra contains numerous phytochemicals, such as saponins, flavonoids, polysaccharides, and coumarins [3]. The prenylated flavonoids have received extensive attention because of their better biological activities [4]. Licorice is a natural source of prenylated compounds [5]. It has been reported that prenylated flavonoids from G. uralensis induced differentiation of B16-F10 melanoma cells [6]. In addition, glycybridin D (10 mg/kg, ip) decreased tumor mass by 39.7% in a mouse model established by an A549 human lung carcinoma xenograft [2].
The medicinal plants and natural ingredients for cancer treatments have attracted increasing interest [7]. Licorice is currently one of the main ingredients used for inflammatory diseases and cancer in traditional Chinese medicines. Most of the drugs that are used in cancer therapy have some serious side effects. Therefore, Glycyrrhiza glabra, and other natural anticancer agents have been extensively studied due to the fact that these compounds are considered to have better bioactivities and may improve the side effects of cancer therapy [8]. In recent years, biological activities, especially antitumor and cytotoxic

Extraction and Isolation
The ultrasound extraction method has become more popular due to its various features, such as low energy consumption, less extraction time, less active compound degradation, its suitability for thermo-sensitive compounds and high extraction yield [15]. This method has been applied previously for the isolation of secondary metabolites from licorice [16]. Thus, the dried roots and rhizomes (10 kg) were broken to pieces and extracted with 70% methanol 4 times, by soaking overnight and ultrasonic extraction 1 h under 45 kHz at 45 • C; then, the filtrate was combined, concentrated to dryness, and we obtained 966 g of the extract. The extract was diffused in water and successively extracted by petroleum ether, dichloromethane, ethyl acetate and n-butanol, respectively. The ethyl acetate extract (274 g) was applied over a silica gel column, then washed by petroleum ether-ethyl acetate (100:0-0:100, v/v) to acquire fractions A-I.

Cell lines and Culture
Human colorectal cancer cells RKO were purchased from Wuhan Procell Life Sciences Co. Other cells were obtained from the National Infrastructure of Cell Line Resource. All cells were cultured with medium-augmented 10% fetal bovine clear fluid (serum) and 1% double antibodies (penicillin and streptomycin) under a standard incubation temperature at 37 • C in 5% CO 2 . The type of basal medium used for RKO cells was MEM medium; HT-29, HCT-116 cells and gastric cancer MGC-803 cells were cultured in RPMI-1640 medium; lung cancer A549 cells and liver cancer Huh7 cells were cultured in DMEM medium.

Cell Viability Assay
The cells were digested with trypsin and resuspended, inoculated into 96-well dishes with a density of (5 × 10 3 per well) and incubated overnight. After the cells were plastered and incubated with complete medium containing various drug concentrations for 48 h, 20 µL of 5 mg/mL MTT solution was injected into each well and cultured in a constant temperature incubator for 4 h at 37 • C; then, the supernatant was disposed of and 150 µL of DMSO was added, agitated for 10 min and at 570 nm, the absorbance was measured by an enzyme marker. Based on the formula below, the percentage of cell viability was calculated: viability percentage of cells (%) = ((OD drug − OD blank )/(OD control − OD blank )) × 100%.

Cell Proliferation Assay
To further investigate the inhibitory effects of the compounds, we designed a multiple concentration-time dependent assay. The RKO cells were incubated with complete medium containing different drug concentrations (10, 20, 40 µM) for 0, 24, 48 and 72 h, respectively. Then, 20 µL of 5 mg/mL MTT solution was injected into all the wells and after 4 h of incubation, the supernatant was removed and 150 µL of DMSO solution was added to all the wells, then the growth curve was plotted, using their absorbance at 570 nm.

Cell Cloning Assay
The RKO cells were digested with trypsin and resuspended; at a density of 1 × 10 3 cells per well, they were inoculated in 6-well dishes and incubated overnight. After the cells were plastered, they were incubated with complete medium containing different drug concentrations for 7 days, and the complete medium was changed once every 3 days. At the end of the culture, the supernatant was disposed of, the cells were washed using PBS, affixed using 4% paraformaldehyde for half an hour, cleaned three times using PBS, then dyed by 0.1% crystalline violet for 10 min, cleaned with PBS to remove the unbound crystalline violet, then photographed and recorded.

Cell Cycle Assay
For cell cycle experiments, RKO cells at the logarithmic growth stage were digested with trypsin and resuspended, at a density of 1 × 10 6 cells per well, then inoculated in 6-well dishes and cultivated overnight. After being plastered, they were incubated separately for 48 h in complete medium containing different concentrations of the drug. After washing with pre-chilled PBS and digestion with trypsin at the end of the culture, the cells were centrifuged at 200× g for 5 min, resuspended with pre-chilled 70% ethanol and fixed at 4 • C overnight. After washing with PBS, 100 µL RNase A solution was incorporated, then the cells were resuspended for 30 min at 37 • C in a water bath; 400 µL PI dying solution was incorporated and mixed, nurtured for 30 min at 4 • C and protected from light for 30 min at 4 • C. The results were detected and analyzed by a CytoFLEX flow cytometer and processed with Modfit LT 5.0.

Cell Apoptosis Assay
The RKO cells were digested with trypsin and resuspended in 6-well dishes at a density of 1 × 10 6 per well and cultured overnight. Then, the cells were plastered and they were incubated in complete medium containing different drug concentrations for 48 h. After the incubation, the supernatant was discarded, the cells were digested with trypsin and collected, washed with PBS, resuspended by adding 1 mL of pre-prepared binding buffer, centrifugated at 300× g for 10 min, the supernatant was removed, 1 mL of binding buffer was added and the cells were resuspended to reach a density of 1 × 10 6 cells/mL. After adding 100 µL of cells to each tube, 5 µL of Annexin V-FITC was incorporated into the tube, mixed gently and nurtured for 10 min without light, then 5 µL of PI solution was added and incubated for 5 min away from light; after completion, we added PBS to 500 µL, mixed gently, assayed and analyzed the results with a CytoFLEX flow cytometry analyzer.

Western Blotting Assay
The RKO cells were digested with trypsin and resuspended in 6-well dishes at a density of 1 × 10 6 per well and cultivated overnight. After the cells were plastered, they were incubated with complete medium containing different drug concentrations for 48 h. After finishing the culture, the supernatant was removed, the cells were dissolved using trypsin and collected by centrifugation at 130× g for 3 min, the supernatant was disposed of and washed three times by PBS; the cells were lysed by adding pre-configured cell lysis solution, containing 1% PMSF and phosphatase inhibitor, in an ice bath for half an hour, centrifugated at 1300× g for 5 min and the supernatant was carefully aspirated for BCA protein quantification. After quantification, we added the load buffer and heated the solution in a metallic bath for 10 min at 100 • C.
Based on the results of BCA protein quantification, 25-30 µg of protein was added to each lane at once and the corresponding marker was added; the voltage was adjusted and SDS-polyacrylamide gel electrophoresis was performed. The gel containing the target and GAPDH protein bands was cut off, then, the gel and PVDF membrane were sandwiched between a cellulose pad and filter paper using a sandwich structure, which was deflated and then placed in a transfer tank and transferred into an ice bath with pre-prepared transfer solution. PVDF membranes were soaked by TBST solution that comprised 5% non-fat milk powder at 25 • C for 2 h. The sealed PVDF membranes were washed 3 times with TBST solution and incubated overnight at 4 • C with a solution containing primary antibodies (diluted 1:1000 with TBST) and the corresponding protein bands in an antibody incubator; after 3 washes with TBST solution, secondary antibodies (diluted 1:10000 with TBST) were incubated for 2 h and washed 3 times with TBST. The above films were developed with the ELC color development system, the X-ray film was pressed, the film was scanned and the film grey values were analyzed with image-pro plus.

Statistical Analysis
GraphPad Prism 9.0.0 was used for the statistical analysis. All experiments were performed in triplicate and the findings were presented as mean ± SD. In order to determine whether there was a significant difference, an unpaired two-tailed Student's t-test was used, and statistical significance was defined as p < 0.05.

Inhibitory Effect of 21 Compounds on 6 Tumor Cell Lines
The antitumor activities of the extracts and isolated compounds were investigated. The extracts of petroleum ether, dichloromethane, and ethyl acetate layers showed better inhibitory effects on RKO and HT-29 cells, and their cell viability values were less than 40% at the concentration of 80 µg/mL for 48 h ( Figure S1).
A total of 21 isolated compounds were tested for their inhibitory effect on human colorectal cancer RKO, HT-29 and HCT-116 cells, lung cancer A549 cells, liver cancer Huh7 cells and gastric cancer MGC-803 cells using MTT assay. As shown in Figure 3, when the tumor cells were processed with different quantities of compound-containing medium (10, 20, 40 µM) for 48 h, compounds 1, 2, 5, 9, and 12 showed better inhibitory effects in a dose-dependent manner. More importantly, some compounds inhibited tumor cells more strongly than 5-fluorouracil when administered at 40 µM, e.g., compounds 1, 2, 5, and 12 better inhibited cells on HT-29, HCT-116 and Huh7, compound 1 better inhibited cells for A549 cells and compounds 1, 5, and 12 better inhibited cells for MGC-803. In addition, compound 9 had a better inhibitory effect on RKO, HT-29 cells and A549 cells in a concentration-dependent manner; compound 15 had a better inhibitory effect on lung cancer A549 cells in a dose-dependent manner, but had less or almost no inhibitory effect on the other five tumor cell lines. the significant inhibitory effect of compound 21 on lung cancer A549 cells in a concentration-dependent manner was also interesting, which was superior to 5-fluorouracil at a low concentration of 10 µM. The IC 50 values of several compounds with better inhibitory effects are listed in Table 2.

Compounds 1 and 5 Inhibited the Proliferation of RKO Cells
Colorectal cancer (CRC) is a prevalent type of cancer and is the major cause of cancer deaths [33]. To further investigate the inhibitory effects of new compound 1 and the more effective compound 5 on the proliferation of human colorectal cancer RKO cells, we performed cell proliferation and cell cloning, in addition to cell cycle assays.
In cell proliferation experiments, we obtained growth curves for compounds 1 and 5 at different concentrations and for different times, as shown in Figure 4. The results showed that the obvious inhibitory effect of compound 1 on RKO cells occurred at a concentration of 20 µM at around 48 h. However, compound 5 showed a significant inhibitory effect at 10 µM, and the number of cells decreased continuously within 48 h. At around 72 h, the number of cells showed a weak increase, but much less than in the control group, probably because the pro-apoptotic effect of compound 5 on RKO cells changed to an inhibitory effect on their proliferation. In cell proliferation experiments, we obtained growth curves for compounds 1 and 5 at different concentrations and for different times, as shown in Figure 4. The results showed that the obvious inhibitory effect of compound 1 on RKO cells occurred at a concentration of 20 μM at around 48 h. However, compound 5 showed a significant inhibitory effect at 10 μM, and the number of cells decreased continuously within 48 h. At around 72 h, the number of cells showed a weak increase, but much less than in the control group, probably because the pro-apoptotic effect of compound 5 on RKO cells changed to an inhibitory effect on their proliferation.  In cell cloning assays, as in Figure 5, both compounds 1 and 5 inhibited the formation of RKO cell clones and showed a dose-dependent effect. Compounds 1 and 5 inhibited the size and number of RKO cell clones when given at 5 μM, and the inhibition was particularly pronounced at 20 μM, with only some of the smaller clones or almost no clones present. In cell cloning assays, as in Figure 5, both compounds 1 and 5 inhibited the formation of RKO cell clones and showed a dose-dependent effect. Compounds 1 and 5 inhibited the size and number of RKO cell clones when given at 5 µM, and the inhibition was particularly pronounced at 20 µM, with only some of the smaller clones or almost no clones present.

Compounds 1 and 5 Promoted Apoptosis in RKO Cells
As shown in Figure 7, compounds 1 and 5 significantly induced apoptosis in RKO cells, and were potential dose-dependent apoptosis promoters, significantly increasing the rate of apoptosis compared to the control group (4.70 ± 0.40%, early and late apoptosis). Compound 1 increased apoptosis from 8.23 ± 0.29% to 33.42 ± 1.79% after treatment of RKO cells with 10, 20, and 40 μM for 48 h. Under the same conditions, compound 5 had a more significant apoptosis-promoting effect on RKO cells, with the apoptosis rate increasing from 8.14 ± 1.37% to 54.50 ± 0.67%. These results demonstrated that compounds 1 and 5 could promote apoptosis of RKO cells and inhibit RKO cells. (A)

Compounds 1 and 5 Promoted Apoptosis in RKO Cells
As shown in Figure 7, compounds 1 and 5 significantly induced apoptosis in RKO cells, and were potential dose-dependent apoptosis promoters, significantly increasing the rate of apoptosis compared to the control group (4.70 ± 0.40%, early and late apoptosis). Compound 1 increased apoptosis from 8.23 ± 0.29% to 33.42 ± 1.79% after treatment of RKO cells with 10, 20, and 40 µM for 48 h. Under the same conditions, compound 5 had a more significant apoptosis-promoting effect on RKO cells, with the apoptosis rate increasing from 8.14 ± 1.37% to 54.50 ± 0.67%. These results demonstrated that compounds 1 and 5 could promote apoptosis of RKO cells and inhibit RKO cells.

Compounds 1 and 5 Promoted Apoptosis in RKO Cells
As shown in Figure 7, compounds 1 and 5 significantly induced apoptosis in RKO cells, and were potential dose-dependent apoptosis promoters, significantly increasing the rate of apoptosis compared to the control group (4.70 ± 0.40%, early and late apoptosis). Compound 1 increased apoptosis from 8.23 ± 0.29% to 33.42 ± 1.79% after treatment of RKO cells with 10, 20, and 40 μM for 48 h. Under the same conditions, compound 5 had a more significant apoptosis-promoting effect on RKO cells, with the apoptosis rate increasing from 8.14 ± 1.37% to 54.50 ± 0.67%. These results demonstrated that compounds 1 and 5 could promote apoptosis of RKO cells and inhibit RKO cells. (A)

Compound 5 Inhibited RKO Cell Growth via the Wnt/β-Catenin Signaling Pathway
The Cancer Genome Atlas (TCGA) shows that the Wnt signaling pathway i vated in 93% of non-hypermutated CRC and 97% of hypermutated CRC [34]. Ther Wnt/β-catenin signaling is an appropriate drug goal with sufficient potency for the ment of CRC. As shown in Figure 8, compounds 1 and 5 showed inhibitory effects Wnt/β-catenin signaling pathway. There was strong evidence that calmodulin plays in tumor development, invasion and metastasis. Furthermore, E-cadherin is a tumor hallmark molecule that is significant in EMT progression [35]. Compounds 1 and 5 re E-cadherin expression after treatment of RKO cells with 10, 20 and 40 μM for 48 effect of compound 5 was particularly pronounced and dose dependent, probabl result of the synergistic effect of multiple signaling pathways. Over expression catenin is a hallmark driver of the conventional Wnt pathway in colorectal cance Compounds 1 and 5 both reduced the expression of β-catenin in RKO cells, and pound 5 was particularly effective in a dose-dependent manner. The expression of c related c-Myc protein was slightly increased after treatment of RKO cells with comp 1, while it was significantly reduced after treatment with compound 5 in a dose-de ent effect. The GSK-3β protein is a regulatory protein and its aggregation with Ax other proteins will lead to the phosphorylation of β-catenin to p-β-catenin, which wi to its degradation, and thus avoid nucleation and activation of downstream onco Both compounds reduced the expression of GSK-3β protein in RKO cells, and comp 5 had a more significant, dose-dependent effect.

Compound 5 Inhibited RKO Cell Growth via the Wnt/β-Catenin Signaling Pathway
The Cancer Genome Atlas (TCGA) shows that the Wnt signaling pathway is activated in 93% of non-hypermutated CRC and 97% of hypermutated CRC [34]. Therefore, Wnt/βcatenin signaling is an appropriate drug goal with sufficient potency for the treatment of CRC. As shown in Figure 8, compounds 1 and 5 showed inhibitory effects on the Wnt/βcatenin signaling pathway. There was strong evidence that calmodulin plays a role in tumor development, invasion and metastasis. Furthermore, E-cadherin is a tumor main hallmark molecule that is significant in EMT progression [35]. Compounds 1 and 5 reduced E-cadherin expression after treatment of RKO cells with 10, 20 and 40 µM for 48 h. The effect of compound 5 was particularly pronounced and dose dependent, probably as a result of the synergistic effect of multiple signaling pathways. Over expression of β-catenin is a hallmark driver of the conventional Wnt pathway in colorectal cancer [36]. Compounds 1 and 5 both reduced the expression of β-catenin in RKO cells, and compound 5 was particularly effective in a dose-dependent manner. The expression of cancer-related c-Myc protein was slightly increased after treatment of RKO cells with compound 1, while it was significantly reduced after treatment with compound 5 in a dose-dependent effect. The GSK-3β protein is a regulatory protein and its aggregation with Axin and other proteins will lead to the phosphorylation of β-catenin to p-β-catenin, which will lead to its degradation, and thus avoid nucleation and activation of downstream oncogenes. Both compounds reduced the expression of GSK-3β protein in RKO cells, and compound 5 had a more significant, dose-dependent effect.

Discussion
Licorice contains flavonoids, triterpenoid saponins, coumarins and stilbenoids [36], which are considered to have good biological activity [37]. Licorice flavonoids have been found to have anticancer activities with various underlying molecular pathways [7]. Licoricidin suppresses the progress of SW480 human colorectal carcinoma cells via stimulation of cycle arrest, apoptosis and the autophagic pathway [38]. Yan Lin et al. isolated 67 free phenolic compounds from licorice, and 11 of these compounds showed strong cytotoxic activity on 3 human cancer cell types (HepG2, SW480 and MCF7), whereas they demonstrated slight toxicity against the human normal cells of LO2 and HEK293T [39]. Isoprenyl-substituted chalcones had cytotoxic effects on MCF-7, HT-29 and A-2780 cancer cells, while the transformation of chalcones to flavonoids led to reduced anti-proliferative activity [40]. In addition, the isoprenyl flavonoid artonol A displayed cytotoxic activity against human lung cancer cells [41]. Moreover, natural products are considered as main sources of novel structures and discovery of new drugs, with about 48.6% of drugs actually being either natural products or directly derived therefrom [42]. Due to the widespread applications of licorice in traditional medicines, along with their recently reported

Discussion
Licorice contains flavonoids, triterpenoid saponins, coumarins and stilbenoids [36], which are considered to have good biological activity [37]. Licorice flavonoids have been found to have anticancer activities with various underlying molecular pathways [7]. Licoricidin suppresses the progress of SW480 human colorectal carcinoma cells via stimulation of cycle arrest, apoptosis and the autophagic pathway [38]. Yan Lin et al. isolated 67 free phenolic compounds from licorice, and 11 of these compounds showed strong cytotoxic activity on 3 human cancer cell types (HepG2, SW480 and MCF7), whereas they demonstrated slight toxicity against the human normal cells of LO2 and HEK293T [39]. Isoprenyl-substituted chalcones had cytotoxic effects on MCF-7, HT-29 and A-2780 cancer cells, while the transformation of chalcones to flavonoids led to reduced anti-proliferative activity [40]. In addition, the isoprenyl flavonoid artonol A displayed cytotoxic activity against human lung cancer cells [41]. Moreover, natural products are considered as main sources of novel structures and discovery of new drugs, with about 48.6% of drugs actually being either natural products or directly derived therefrom [42]. Due to the widespread applications of licorice in traditional medicines, along with their recently reported bioactivities [43], we aimed to isolate and purify the compounds from the ethyl acetate extract of G. glabra.
In our study, the isolated compounds showed good antitumor activities, of which 10 compounds with isoprenyl or dimethylpyran rings showed better inhibition of tumor cells. In terms of the conformation-activity relationship, compounds 1 and 2 with two dimethylpyran ring substitutions were more cytotoxic to tumor cells than compound 3, with one isoprenyl and one dimethylpyran ring substitution, suggesting that the conversion of the isoprenyl group to the dimethylpyran ring in this class of compounds may be more favorable to its enhanced antitumor activity. In addition, compound 15, after deisoprenylation and methoxy substitution by compound 12, showed a weak inhibitory effect on tumor cells.
Colorectal cancer (CRC) is a serious threat to human health. More than 94% of colorectal cancer situations have mutations in one or more Wnt/β-catenin signaling pathway compositions [44]. Compounds 1 and 5, which were significantly cytotoxic to colorectal cancer RKO cells, were selected for further studies of the anti-cancer mechanism. Our results showed that both compounds 1 and 5 could inhibit the growth of RKO cells by inhibiting cell proliferation, reducing clone formation and promoting their apoptosis, and compound 5 showed a superior effect. It is well known that the Wnt/β-catenin signaling pathway is closely associated with the progression of CRC [45]. Based on Western blotting analysis, compound 5 significantly reduced the expression of E-cadherin, β-catenin, c-Myc and GSK-3β proteins in RKO cells in a dose-dependent manner. It has been demonstrated that glycyrol stimulated cell death related with apoptosis and autophagy in AGS and HCT 116 cells, and inhibited tumor growth in a nude mouse tumor xenograft model bearing HCT 116 cells [46]. The benzofuranyl, isopentenyl and methoxy groups present in glycyrol played an important role in its anti-cancer activity, whereas the furan group led to more improvements [47]. Many natural products play a role in the Wnt/β-catenin signaling pathway [48], such as green tea polyphenols, epigallocatechin-3-gallate and resveratrol [49]. Lonchocarpin flavonoids acted as a Wnt/β-catenin pathway inhibitor, which has both in vitro and in vivo inhibitory effects on cell migration and cell proliferation on HCT116, SW480, and DLD-1 colorectal cancer cell lines [50]. Anticancer bioactive peptide (ACBP) also showed inhibition of proliferation, migration, and cell invasion in three CRC lines (HCT116, RKO, HT29) by suppressing the canonical Wnt signaling pathway [51].
Compound 1 decreased the expression of E-cadherin, β-catenin and GSK-3β proteins, but caused a weak increase in the expression of c-Myc protein. Therefore, the main signaling pathway through which compound 1 exerts its inhibitory effect on RKO cells needs to be further investigated.

Conclusions
In conclusion, we performed phytochemical studies on the roots of G. glabra and isolated 21 compounds, including 3 new compounds, further enriched their chemical composition and screened several compounds with better antitumor activities, which contained isoprenyl and dimethylpyran ring substitutions. Licopyranol A (1) and glycyrol (5) suppressed the growth of RKO cells and inhibited proliferation of RKO cells, reduced clone formation and promoted their apoptosis. The Western blotting results showed that glycyrol could inhibit the growth of RKO cells via the Wnt/β-catenin signaling pathway, suggesting that it had greater potential for antitumor activity, which is expected to become an anticancer seed compound through further studies.