New Diterpenoids from Mesona procumbens with Antiproliferative Activities Modulate Cell Cycle Arrest and Apoptosis in Human Leukemia Cancer Cells

Mesona procumbens is a popular material used in foods and herbal medicines in Asia for clearing heat and resolving toxins. However, phytochemical research on this plant is very rare. In this study, eleven new diterpenoids, mesonols A-K (1–11), comprising seven ent-kauranes, three ent-atisanes, and one sarcopetalane, were isolated from its methanolic extract. Structural elucidation of compounds 1–11 was performed by spectroscopic methods, especially 2D NMR, HRESIMS, and X-ray crystallographic analysis. All isolates were assessed for their antiproliferative activity, and compounds 1–4 showed potential antiproliferative activities against A549, Hep-3B, PC-3, HT29, and U937 cancer cells, with IC50 values ranging from 1.97 to 19.86 µM. The most active compounds, 1 and 2, were selected for further investigation of their effects on cell cycle progression, apoptosis, and ROS generation in U937 human leukemia cancer cells. Interestingly, it was found that compounds 1 and 2 induced antiproliferative effects in U937 cells through different mechanisms. Compound 1 caused cell cycle arrest at the G2/M phase and subsequent cell death in a dose- and time-dependent manner. However, 2-mediated antiproliferation of U937 cells triggered ROS-mediated mitochondrial-dependent apoptosis. These results provide insight into the molecular mechanism involved in the antiproliferative activities of compounds 1 and 2 in U937 cells. Altogether, the study showed that new diterpenoid compounds 1 and 2 from M. procumbens are potent and promising anticancer agents.


Introduction
Natural products (NPs) and NP-like compounds have played an important role in drug discovery. More than 70% of the anticancer drugs that have been approved worldwide over the past seven decades are NPs or are inspired by NP structures [1,2]. NPs have evolved to target multiple proteins and often possess diverse biological activities, leading to a combination of therapeutic effects and toxicity. NP scaffolds can be regarded as "bioactive" or "privileged" scaffolds in chemical space because they have been naturally selected to specifically interact with diverse biological targets [3]. Due to the toxicity of the

General
The methanolic extract of M. procumbens was suspended in H 2 O and then successively partitioned with n-hexane and CH 2 Cl 2 to give two organic layers and an aqueous layer. The CH 2 Cl 2 extract was chromatographed on a C 18 gel flash column and then on a silica gel flash column. The subfractions were further subjected to preparative HPLC using a reversed-phase column to yield eleven new diterpenoids (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11) (Figure 1). All isolated compounds were screened for their antiproliferative activities against five human cancer cell lines, and their molecular mechanism was investigated in U937 human leukemia cancer cells.
A 15,16-dihydroxy-ent-kaurane diterpene skeleton was elucidated for each of the diterpenes 3-7 from the NMR spectra, and all showed the presence of three methines (C-5, C-9, and C-13), three quaternary carbons (C-4, C-8, and C-10), an oxygenated methine (C-15), and an oxygenated quaternary carbon (C-16). According to 1 H-1 H COSY and HMBC correlations (Figure 2A), compound 3 was elucidated as 12,15,16,17-tetrihydroxyent-kaurane. Thus, the planar structure of 3 is a reductive product of 1. The carbonyl carbon (C-15) in 1 is replaced with a hydroxyl group in 3 (δ C 80.3). OH-15 was assigned the β-orientation due to the upfield shift of C-9 (δ C 47.5), which was caused by the γgauche steric compression effect between OH-15 and H-9. Moreover, H-15 (δ H 3.10) showed NOE correlations ( Figure 2B) with H 2 -17, and one proton of H 2 -14 (δ H 0.72) revealed that those protons were on the same side of the cyclopentane ring. From the above findings, the stereochemistry of OH-16 showed an axial-like β-orientation, while the stereochemistry of the 16-hydroxymethyl group showed an equatorial-like α-orientation. Furthermore, the configuration was confirmed by single-crystal X-ray diffraction analysis, and a perspective drawing of 3 is provided in Figure 2C. Finally, the structure of 3 was assigned as 12α,15β,16β,17-tetrahydroxy-ent-kaurane and named mesonol C.
Comparison of the 13 C NMR and 1 H NMR data (Tables 1 and 2) of compound 4 with those of 3 revealed that the only difference was the substitution of a hydroxyl group at C-17 in compound 3 with a chlorine in 4. This finding was confirmed by the chemical shift of C-17 occurring upfield to δ C 53.9 in 4. In addition, the HRESIMS spectrum of 4 showed a pair of sodiated quasimolecular ion peaks at m/z 379.2016 [M + Na] + and 381.1985 [M + Na + 2] + with a ratio of 3:1, suggesting that compound 4 possesses a chlorine atom. The relative configurations of the chiral centers in 4 were found to be the same as those in 3, as determined from a NOESY experiment. Thus, 4 was elucidated as 12α,15β,16βtrihydroxy-17-chloro-ent-kaurane and named mesonol D.
Compound 5 exhibited the same molecular formula (C 20 H 34 O 4 ) and similar NMR data to 3. These two compounds were determined to be stereoisomers because, after detailed analysis of their 2D NMR spectra, their planar structures were identical. The NOESY experiment ( Figure 2B) determined the relative configurations of compound 5, in which correlations were observed from CH 3 -18 to H-5, from H-5 to H-9, from CH 3 -20 to H α -14, from H 2 -7 to H-15, and from H-15 to H β -14. These results indicated that compound 5 also possessed an ent-kaurane skeleton. Moreover, other NOE correlations from H 2 -17 to H-12 and from H 2 -17 to H β -11 were observed in 5, whereas H 2 -17 showed NOE correlations with H-15 and H β -14 in 5. Thus, compound 5 was characterized as 12α,15β,16α,17-tetrahydroxy-ent-kaurane and named mesonol E.
The molecular formula of 6 was inferred by HRESIMS to be C 20  , and H 2 -17 (δ H 3.55 and 3.60) showed HMBC correlations with C-12 (δ C 212.5), and the two COSY fragments of H-9/H-11 and H-13/H-14 supported the connectivity of ring C in 6. Additionally, four hydroxyl groups were attached at C-14, C-15, C-16, and C-17, as demonstrated by the HMBC correlations from H-14 to C-15, H-15 to C-16 and C-17, and H 2 -17 to C-13 and C-16. Furthermore, the chemical shift of C-7 in 6 occurred upfield to 30.5 ppm compared with that in the other ent-kaurane isolates because of the γ-gauche steric compression effects of the 14-OH group. Thus, the fully planar structure of 6 was constructed from the above COSY and HMBC correlations. In the NOESY spectrum ( Figure 2B), the cross peaks of CH 3 -18/H-5/H-9 (β-orientation) and H 3 -19/H 3 -20 (α-orientation) allowed the ent-kaurane skeleton in 6. The key NOESY correlation between H-14 and H 3 -20 revealed that 14-OH was β-oriented. Additionally, H-15 showed NOESY correlations with quasi-equatorial protons H 2 -17 rather than H-9, implying that 15-OH and 16-OH were both β-oriented. Accordingly, 6 was determined to be 14β,15β,16β,17-tetrahydroxy-ent-kaur-12-one and named mesonol F.
Compound 8 was isolated as a white powder and was assigned the formula C 20 H 32 O 4 by HRESIMS. The NMR spectroscopic data (Tables 1 and 2) (Figure 2A), the structure of 8 was elucidated to have similar A, B, and C rings and most likely a structure similar to that of 1-7. In detailed analysis of the HMBC spectrum, a correction from H 2 -17 and H-9 to C-12 was observed, as different from H 2 -17 to C-13 as 1-7, suggesting a migrated C-12/C-16 bond rather than a C-13/C-16 bond in 8. Therefore, compound 8 was proposed to possess a rearranged 16(13→12)-abeo-ent-kaurane skeleton (ent-atisane skeleton) [22]. Additionally, the position of three hydroxyl groups was assigned at C-13, C-16, and C-17, as deduced from their chemical shifts and HMBC correction. The relative configuration of rings A and B in 8 was deduced from similar NMR chemical shifts with ent-atisane-16α-ol [23] and ent-(3β,13S)-3,13-dihydroxyatis-16-en-14-one [24], along with NOESY correlations ( Figure 2B  Comparison of the NMR data of 8 with those of 9 and 10 (Tables 1 and 2) suggests that both structures of 9 and 10 have an ent-atisene skeleton. The major changes of these compounds indicated that the signals for a ketone and an oxygenated methylene in 8 were replaced by two vinyl protons with quaternary carbons in 9 and 10. In addition to the 1 H-1 H COSY and HMBC correlations ( Figure 2A) from H 2 -17 to C-12 and C-15, from H-15 to C-16 and C-8, from H-9 to C-14, and from H-11 to C-8 and C-13, both 9 and 10 likely processed the ent-atisene skeleton. In addition, an exo-methylene group at C-16 and C-17; two hydroxyl groups at C-13 and C-15; and three hydroxyl groups at C-13, C-14, and C-15 were found in 9 and 10, respectively. Accordingly, 9 and 10 were determined to be 13α,15β-trihydroxy-ent-atis-16(17)-ene and 13α,14β,15β-trihydroxyl-ent-atis-16-ene, respectively.

Biological Studies
Studies have shown that many isolated diterpenoid compounds exert antitumor activity against a range of cancer cell types [20]. To evaluate the effects of the isolated diterpene compounds on the growth of human cancer cells, the antiproliferative activities of compounds 1-11 were tested against six different cell lines, including five cancer cell lines, A549, Hep-3B, PC-3, HT29, and U937, and one normal mouse macrophage cell line, RAW 264.7. The IC 50 values are summarized in Table 3. Among the tested compounds, mesonols A-D (1-4) showed potent antiproliferative activities against these cancer cell lines, especially against U937 cells. Of note, mesonols A (1) and B (2) exhibited remarkable antiproliferative activities against U937 cells with IC 50 values of 2.66 and 1.97 µM, respectively, which were more active than the standard drug CPT-11 with an IC 50 value of 4.95 µM. Moreover, mesonols A-D (1-4) were less toxic against normal mouse RAW 264.7 macrophages, implicating the selectivity of mesonols toward cancer cells. Together, these results indicated that the new diterpenoid compounds mesonols A (1) and B (2) from M. procumbens have potency as anticancer agents. >20 >20 >20 >20 >20 >50 6 >20 >20 >20 >20 >20 >50 7 >20 >20 >20 >20 >20 >50 8 (-) a (-) a (-) a (-) a (-) a (-) a 9 >20 >20 >20 >20 >20 >50 10 >20 >20 >20 >20 >20 >50 11 >20 >20 >20 >20 >20 >50 CPT- 11 15 To further investigate the antiproliferative effects of mesonol A or B on U937 cancer cells, the cell cycle profiles in cells treated with various doses of mesonol A (1) or B (2) (0.625, 1.25, 2.5, 5, and 10 µM) for 24 h using flow cytometric analysis were first analyzed. Surprisingly, mesonols A (1) ( Figure 4A) and B (2) ( Figure 4B) showed differential effects on the cell cycle distribution in U937 cells. Mesonol A (1) dose-dependently increased the percentage of the sub-G 1 population (corresponding to apoptotic cells) and reached a plateau at 5 µM. In addition, 10 µM mesonol A (1) not only induced the accumulation of cells at sub-G1 phase but also resulted in a significant increase in the cell population at G2/M phase, implicating that G2/M arrest may be associated with mesonol A (1)-mediated antiproliferation. On the other hand, mesonol B (2) significantly increased the sub-G1 population to 53.86% at 10 µM, compared to 6.33% at 5 µM and 1.58% for the control group. Furthermore, a time course cell cycle analysis to explore the underlying mechanism was also performed. U937 cells were treated with 5 µM mesonol A (1) or 10 µM mesonol B (2) for 3, 6, 9, 12, or 24 h. Mesonol A (1) led to a time-dependent accumulation of cells arrested in the G2/M phase with a concomitant gradual increase cell population of sub-G1 phase after 3 h to 12 h treatment, followed by decreased populations of G2/M and increased population of sub-G1 phase, indicating the correlation between mesonol A (1)-induced G2/M arrest and induction of apoptosis ( Figure 4C). On the other hand, mesonol B (2) did not cause an obvious accumulation of cells in any phase but significantly increased apoptotic cells in the sub-G1 population ( Figure 4D). These results suggested that mesonols A (1) and B (2) may exert their antiproliferative activities against U937 cells through cell cycle arrest at G2/M and induction of apoptosis, respectively. To further confirm the effects of mesonols A (1) and B (2) on cell cycle arrest and apoptosis in U937 cells, the effects of mesonols A (1) and B (2) on critical cell cycle regulators and apoptosis-related proteins, respectively, were next examined. U937 cells were treated with 5 µM mesonol A (1) for 12 h, and the G2/M-associated cell cycle regulators p21, cyclin B1, CDK1, and cdc25c [26] were analyzed by Western blot. As shown in Figure 5A, mesonol A (1) caused a significant increase in p21 protein, which has been implicated in the G2/M checkpoint [27], and decreases in the G2/M progression proteins cdc25c, cyclin B1, and CDK1, confirming that mesonol A induces G2/M arrest in U937 cells. On the other hand, for mesonol B (2), the effects of mesonol B (2) on the status of apoptosis-associated proteins, including Bax, Bcl-2, Bcl-xL, cytochrome c, Apaf-1, caspase-9, caspase-3, and poly(ADP-ribose) polymerase (PARP), in U937 cells were examined ( Figure 5B). Mesonol B (2) induced the activation of caspase-9 and -3, as indicated by decreased pro-caspase-9 and -3 protein levels. PARP, a downstream target of caspase-3, was also significantly cleaved and activated in mesonol B (2)-treated U937 cells. Furthermore, mesonol B (2) induced increases in proapoptosis-related proteins, including Bax, cytochrome c and apaf-1, and decreases in antiapoptotic proteins, including Bcl-2 and Bcl-xL. These results suggest that mesonol B (2) induced intrinsic mitochondrion-dependent apoptosis in U937 cells. Altogether, these results indicated that mesonol A (1) exerts its antiproliferative activities against U937 cells through cell cycle arrest at G2/M, whereas mesonol B (2) exerts its antiproliferative activities through induction of intrinsic apoptosis.
Reactive oxygen species (ROS) play a critical role in mitochondrion-mediated apoptosis. Growing evidence suggests that chemotherapy or radiotherapy could induce apoptosis in cancer cells by increasing intracellular oxidative stress [24,25]. Therefore, whether ROS are involved or not in mesonol B (2)-induced apoptosis and in mesonol A (1)-induced G2/M arrest was further investigated. To this end, whether or not the ROS scavenger Nacetylcysteine (NAC), which is also known to block ROS-mediated apoptosis [28], could inhibit mesonol B (2)-induced apoptosis in U937 cells was examined. As shown in Figure 5C, pretreatment with NAC significantly suppressed mesonol B (2)-induced sub-G1 accumulation compared to mesonol B (2) alone. Western blot analysis also showed that NAC blocked mesonol B (2)-induced apoptosis. Mesonol B (2)-induced activation of cytochrome c, Apaf-1, caspase-9, caspase-3, and PARP was significantly inhibited by NAC ( Figure 5B), suggesting that mesonol B (2) induced apoptosis in a ROS-dependent manner. In contrast, NAC showed little to no effect on the attenuation of mesonol A (1)-induced G2/M arrest ( Figure 5D). Consistently, NAC could not suppress mesonol A (1)-induced activation of p21 or inhibition of cdc25c, cyclin B1, and CDK1 proteins ( Figure 5A), indicating that ROS may not be involved in mesonol A (1)-induced G2/M phase arrest in U937 cells.
Endogenous ROS levels were also measured by DCF-DA, an oxidation-sensitive fluorescent dye, to investigate whether ROS play differential roles in mesonol A (1)-and B (2)-mediated effects. Mesonol A (1) did not induce an increase in intracellular ROS levels compared to those in the control group ( Figure 5E). In contrast, the intracellular ROS levels in mesonol B (2)-treated cells were significantly higher than those in control cells and could be blocked by pretreatment with NAC ( Figure 5F). Together, these results suggest that mesonol B (2)-induced ROS-mediated mitochondrial-dependent apoptosis.

Discussion
In the present study, eleven new diterpenoid compounds were isolated from the CH 2 Cl 2 layer of a methanolic extract of M. procumbens, including seven ent-kauranes, mesonols A-G (1-7), three ent-atisanes, mesonols H-J (8)(9)(10), and a sarcopetalane, mesonol K (11). Diterpenoid compounds have been reported to exert potent antitumor activity against a range of cancer cell types [20]. For example, the natural compound ent-16β,17αdihydroxykaurane has been shown to exhibit cytotoxicity and apoptotic effects in the human breast cancer cell line MCF-7 [29]. Here, four diterpenoid compounds, mesonols A-B (1-4), were also found, and they showed significant antiproliferative activity against human cancer cell lines, including A549, Hep-3B, PC-3, HT29, and U937. Moreover, the most active mesonols A (1) and B (2) displayed remarkable inhibitory activity against U937 cell lines with IC 50 values of 2.66 and 1.97 µM, respectively, and were even better than the standard drug CPT-11 with an IC 50 value of 4.95 µM. Moreover, both mesonols A (1) and B (2) were less toxic against normal mouse RAW 264.7 macrophages, implicating that these compounds may act selectively against cancer cells and normal cells. These results indicated that mesonols A (1) and B (2) have high potential as anticancer agents, especially against U937 cells. Furthermore, the differential inhibitory effects of mesonols A (1) and B (2) on cell cycle arrest and/or apoptosis in U937 cancer cells were also demonstrated. An earlier study reported that glaucocalyxin A, an ent-kauranoid diterpenoid isolated from Rabdosia japonica var., induced apoptosis in HL-60 cells, as characterized by cell morphology, DNA fragmentation, activations of caspase-3 and -9, and an increased expression ratio of Bax/Bcl-2. The mitochondrial membrane potential loss and cytochrome c release were observed during the induction, and pretreatment of antioxidant NAC could block glaucocalyxin A-induced ROS generation and apoptosis [30]. Another ent-kaurene diterpenoid xerophilusin B has been reported to induce apoptosis and G2/M phase cell cycle arrest in KYSE-150 and KYSE-450 cells. Treatment with xerophilusin B increased the cytochrome c release from mitochondrial to cytosol and activation of caspase-9 and -3, while it downregulated caspase-7 and PARP levels. Moreover, the ratio of Bcl-2/BAX decreased after xerophilusin B treatment [31]. Similarly, our results showed that mesonol B (2) induced the ROS-dependent intrinsic apoptosis pathway and NAC could also block the mesonol B (2)-induced activation of apoptosis in U937 cells; mesonol A (1) caused cell cycle arrest at the G2/M phase and subsequent cell death.
Thus, our bioactive natural product studies on M. procumbens implying mesonols A (1) and B (2) may provide drug candidates for the treatment of cancer.

General
Optical rotations were determined using a JASCO P-2000 polarimeter. Infrared (IR) spectra were recorded on a Mattson Genesis II spectrometer (Thermo). High-resolution electrospray ionization mass spectrometry (HRESIMS) data were acquired on a Thermo Scientific Q Exactive Focus Orbitrap liquid chromatography-tandem mass spectrometry (LC-MS/MS) instrument equipped with an Ultimate 3000 UHPLC system. Nuclear magnetic resonance (NMR) spectra were recorded on Bruker Avance 400 MHz, Varian Unity Inova 500 MHz, and Varian VNMRS 600 MHz spectrometers. Silica gel 60 (70-230 and 230-400 mesh, Merck) and Sephadex LH-20 (GE) were used for column chromatography. The colorless crystal of 1 was obtained by natural evaporation from MeOH solution. Crystal data were measured by a Brucker D8 VENTURE single-crystal X-ray diffractometer equipped with a dual microfocus X-ray source on Cu radiation (Table S1 in

Mesonol B (2)
The colorless crystal of 2 was obtained by natural evaporation from MeOH solution. Crystal data were measured by a Brucker D8 VENTURE single-crystal X-ray diffractometer equipped with a dual microfocus X-ray source on Mo radiation (Table S2 in

Mesonol C (3)
The colorless crystal of 3 was obtained by natural evaporation from MeOH solution. Crystal data were measured by a Brucker D8 VENTURE single-crystal X-ray diffractometer equipped with a dual microfocus X-ray source on Mo radiation (Table S3 in  The colorless crystal of 7 was obtained by natural evaporation from MeOH solution. Crystal data were measured by a Brucker D8 VENTURE single-crystal X-ray diffractometer equipped with a dual microfocus X-ray source on Mo radiation (Table S4 in

Western Blot Analysis
Cells were lysed in RIPA buffer with phenylmethylsulfonyl fluoride (PMSF; Beyotime Biotechnology, Jiangsu, China). The protein concentration was determined using a Bio-Rad protein assay system (Bio-Rad, Hercules, CA, USA). Equivalent amounts of proteins were separated by SDS-PAGE and then transferred to polyvinylidene difluoride membranes (Bio-Rad). After blocking in TBS containing 5% nonfat milk, the membranes were incubated with primary antibodies (1:1000 dilution) at 4 • C for 12 h and then incubated with a horseradish peroxidase-conjugated secondary antibody (1:5000 dilution). Signals were detected on X-ray film using an ECL detection system (Pierce, Rockford, IL, USA). The relative protein levels were calculated based on β-actin as the loading control.

Cell Cycle Analysis
U937 cells were treated with mesonols A (1) or B (2), dosage range: 0.625-10 µM, for the specified time period, fixed in ice-cold 70% ethanol in PBS, suspended in Krishan's reagent (0.05 mg/mL propidium iodide (PI), 0.1% sodium citrate, 0.02 mg/mL ribonuclease A and 0.3% NP-40), and incubated on ice for 30 min. Data acquisition for 10,000 events was performed with an FACScan system (Becton Dickinson). The distribution of cells in the different phases of the cell cycle was analyzed from the DNA histograms using CELL Quest software. All experiments were performed in triplicate.

Measurement of ROS Generation
U937 cells were seeded onto 96-well microplates at a final concentration of 1 × 10 5 cells per well and treated with mesonol A or B (5 or 10 µM) for 30 min with or without pretreatment with N-acetylcysteine (NAC) (10 mM) for 1 h. The cells were stained with 2 ,7 -dichlorofluorescin diacetate (DCFDA) Cellular Reactive Oxygen Species Detection Assay Kit (Abcam, Cambridge, UK) solution (20 µM) and then incubated at 37 • C for 30 min in the dark. Fluorescence measurements were performed with a fluorescence plate reader (Bio-Rad, Hercules, CA, USA) at Ex/Em = 485/535 nm in end point mode in the presence of compounds, medium, or buffer. All experiments were performed in triplicate.

Statistical Analysis
SPSS (SPSS, Chicago, IL, USA) was used to perform statistical data analysis. All data are presented as the mean ± standard deviation. Groups were compared using one-way analysis of variance (ANOVA) followed by Tukey's test of multiple comparisons; p-values ≤ 0.05 were considered statistically significant.

Conclusions
The results of the present study suggest that new diterpenoid mesonols A (1) and B (2) from M. procumbens have potency as anticancer agents and are more active than the standard drug CPT-11. Mesonols A and B were less toxic against normal mouse RAW 264.7 macrophages, implicating the selectivity of mesonols toward cancer cells. The anticancer mechanism of mesonols A and B was also explored to provide a theoretical basis for the development of targeted anticancer agents.