Two new flavonoids and anticancer activity of Hymenosporum flavum: in vitro and molecular docking studies

1Chemistry of Natural Compounds Department, National Research Centre, 12622 Giza, Egypt 2Department of Pharmaceutical Medicinal Chemistry, Faculty of Pharmacy, Horus University-Egypt, New Damietta 34518, Egypt 3Industrial Biotechnology Department, Genetic Engineering and Biotechnology Research Institute, University of Sadat City, Sadat City, Egypt 4Department of Molecular Biology, Genetic Engineering and Biotechnology Research Institute, University of Sadat City, Sadat City, Egypt 5Bioinformatic Department, Genetic Engineering and Biotechnology Research Institute, University of Sadat City, Sadat City, Egypt


Introduction
Medicinal plants, since the days of yore, have been a promising source for natural drugs. Their safety aspects, low costs, and accessible approach gave them precedence to be investigated for their bioactive phytochemicals to be applied as natural medicines (1). Phytochemical component, mainly phenolics, have perpetually proved their importance as anticancer drugs (2).
Hepatocellular carcinoma (HCC) affects around one million persons worldwide every year (5). Hepatitis B virus (HBV) infections or mitogens with other tumor cell signaling cascades enhance RAF/MEK/ERK pathway. This pathway has a paramount role in poor prognostication and liver oncogenesis (6). Sorafenib is one of the targeted treatments, recognized as a kinase inhibitor and approved by the Food and Drug Administration (FDA) for HCC treatment (7).
RAF-1 has the leading role in the growth factor signal transduction from the cell membrane to the nucleus. Interestingly, RAF-1 kinase is activated and rapidly phosphorylated through stimulation by many mitogens (8). The mitogen-activated protein kinase (MAPK) pathway is considered the important signal transduction cascade that controls and regulates cell growth. The mechanism of MAPK activation occurs through the binding of growth factors to their cognate receptors. Also, it recruits the small GTPase Ras, which in sequence recruits the serine-threonine kinase RAF-1 (9). MAPK has many pathways that control several cellular processes, differentiation, and drives proliferation and cell survival. One of these pathways is ERK (extracellular-signalregulated kinase) signaling (10).
Numerous researchers established a fruitful combination between sorafenib and other compounds targeting signaling pathways for improving its efficacy. Sorafenib combined with OSU-2S with a synergistic effect to boost the antiproliferative activity on HCC cells, where tumor protein p53 and Protein kinase C-delta (PKCδ) are associated with the regulation of OSU-2S/sorafenibinduced cell death (11). Numerous studies have shown the synergistic effect of sorafenib alongside various active compounds in plant extracts as resveratrol in grapes, peanuts, red wine (12), corosolic acid in Actinidia chinensis (13), and wogonin (5,7-dihydroxy-8-methoxyflavone) in Scutellaria baicalensis (14).
For the time being, new therapeutics discovery or repurposing already existing ones become faster by using several computational methods as molecular docking. Molecular docking is of paramount importance to propose the expected mechanisms of action for the tested compounds and their repurposing (15). Take into consideration the essential role of both RAF-1 and ERK-2 pathways for cancer development. Besides the previously reported effects of anticancer drugs targeting these genes (16), and in continuation to our previous work targeting cancer (17), our goal in this research was to investigate the cytotoxic activity of H. flavum hydroalcoholic leaf extract against human liver tumor cells. We investigated the possible synergistic combinations of the most promising compounds of the leaf extract with sorafenib through RAF/MEK/ERK signaling pathways inhibition, using in silico and in vitro studies. Meanwhile, the fingerprint profile of the leaf extract, the determination of isolated compounds, and their anticancer activities could supply beneficial information on the importance of the plant and derived natural remedies. Collectively, the present study consequences a novel formula of H. flavum hydroalcoholic leaf extract to be used in HCC therapeutic protocols. g), redissolved in 500 mL water for further chemical and biological investigation.

HPLC-ESI-MS/MS analysis Sample preparation
We prepared the sample (120 μg/mL) solution using HPLC analytical grade solvent of MeOH, filtered the sample using a membrane disc-filter (0.2 μm), then subjected it to LC-ESI-MS analysis.

Apparatus and conditions
Using HPLC-ESI-MS/MS, the aqueous methanolic extract of H. flavum was analyzed. The LC system was Thermo Finnigan (Thermo Electron Corporation, Austin, TX, USA) (18). The reversed-phase column Zorbax Eclipse XDB-C18, with rapid resolution, 4.6×150 mm, 3.5 µm column was used (Agilent, Santa Clara, CA, USA). The mobile phase was water and acetonitrile (ACN) (0.1% formic acid each), and the gradient was employed from 5% to 30% in 60 minutes with a flow rate of 1 mL/min with a 1:1 split before the ESI source. The samples were injected using an autosampler. LCQ-Duo ion trap with a Thermo Quest ESI source was used for MS analysis. We used Xcalibur software (Xcalibur TM 2.0.7, Thermo Scientific, Waltham, MA, USA) to control the system.

Isolation of compounds
Three hundred and fifty ml of water extract was loaded on a polyamide 6 column chromatography. The column was eluted with H 2 O, and then the mixtures of H 2 O-MeOH of decreased polarity to collect ten fractions (1 L, each). The main phenolic fractions obtained were united into six after chromatographic analysis via paper chromatography.
Docking studies Two separate molecular docking studies were performed through MOE 2019.012 suite (24) to propose, quantify, and evaluate the binding scores and interactions of the eight isolated flavonoids of H. flavum leaf extract against RAF-1 and ERK-2 proteins, respectively. The co-crystallized inhibitors of both proteins were involved in our studies as reference standards.

Docking of the isolated extract flavonoids to the binding pockets of RAF-1 and ERK-2
Following the general docking methodology applied before, the two mentioned databases were docked in two separate docking processes (29). Finally, we selected the best pose for each docked compound at each target receptor according to scores, RMSD-refine values, and binding interactions for further studies. Also, we performed a program validation at first for both targets by redocking the co-crystallized inhibitor in each case, and we approved the validity by low RMSD values (<1) between the docked and native forms (10).
In vitro studies Cell line Hepatocellular carcinoma (HepG2 cells) was grown in RPMI media supplemented with 4 mM L-glutamine, 4 mM sodium pyruvate, and 2.5% heat-treated bovine serum albumin (BSA). The normal hepatocyte cells were grown in RPMI media containing 4 mM L-glutamine and 10% BSA. We incubated all cell lines at 37°C under 5% CO 2 condition (30). The imaging of cultured cells was determined by using inverted microscopy with a Zeiss A-Plan 10X.

Cytotoxic concentration 50% (CC 50 )
We tested the purified agent for its cytotoxic effect and calculated the potential CC 50 on HepG2 cells and the normal hepatocytes. Therefore, the cells were cultured in 96-well plates in a density of 10×10 3 cells/well and incubated in a CO 2 incubator at 37ºC. The cells were treated with different concentrations of the purified agent (0-25 mg/mL) followed by overnight incubation. The cell viability rate and the cytotoxic concentration were monitored using an MTT cell growth assay kit (Sigma-Aldrich, Germany). Based on the amount of formazan dye, the CC 50 was measured by measuring the absorbance at 570 nm.

Lactate dehydrogenase (LDH) production
We used the LDH assay kit (Abc-65393) to assess LDH production in the medium collected from culturedtreated cells. According to the manufacturing procedures, 100 µL of lysed cells was incubated with a 100 µL LDH reaction mix for 30 minutes at room temperature. LDH activity was quantified by a plate reader at OD450 nm. The relative LDH production was calculated by dividing the mean values of the treated cells on the mock values, which indicated by fold change (31).

Reverse transcription and quantitative real time-PCR
HepG2 cells were seeded at a density of 2 × 10 5 cells per well in a six-well plate followed by overnight incubation. The cells were treated with 100 μg/mL of each purified flavonoid and (or) SOR followed by incubation at 37°C in a CO 2 incubator for 24 hours. To quantify messenger RNA (mRNA) of indicated genes, we used quantitative reverse transcription polymerase chain reaction (qRT-PCR) to make cDNA construction and amplification in one step via the purified total RNA as a template. Total RNA was extracted 24 hours post-treatment from treated cells and purified using the RNeasy Mini Kit (Qiagen, USA) and TriZol (Invitrogen, USA). The relative expressions of Raf-1 and Erk-2 were detected with the QuantiTect SYBR Green PCR Kit (Qiagen, USA) and oligonucleotides specific for each gene (Table 1). Housekeeping glyceraldehyde 3-phosphate (GAPDH) gene level was used for normalization. The following mixture was prepared for each reaction: 10 μL SYBR Green, 0.5 μL reverse transcriptase (50 U/μL), 0.2 μL RNase inhibitor (20 U/ μL), 1 μL purified total RNA (100 ng/μL), and 1 μL from each primer up to a final volume of 25 μL using RNase free water. According to the manufacturer's protocol, we used the following PCR parameters: 50°C for 30 minutes, 95°C for 3 minutes, 35 cycles (95°C for 30 s, 60°C for 15 s, 72°C for 30 s). We obtained levels of Raf-1 and Erk-2 relative to GAPDH using comparative ΔΔCt equations (16).

Statistical analysis
The Microsoft Excel software was used for all histograms and chart preparations. For statistical analysis of the data, we used the Student's two-tailed t test. P values ≤0.05 were considered statistically significant. ΔΔCt analysis was used to determine the mRNA expression fold change detected by qRT-PCR using the following equation: (1) ΔCt = Ct value for gene-Ct value for GAPDH, (2) (ΔΔCt)= 1Ct for value (experimental) -ΔCt for value (control), Expression fold change = (2−ΔΔct) (32).

Results
Isolation and identification of the compounds Eight phenolic compounds were isolated from the hydroalcoholic leaf extract of H. flavum and identified using spectroscopic techniques.  Table 2.   fingerprint for every compound in the extract, despite the similarity of the molecular formula (35). In Table 3, we exhibited the recognition of seventy compounds from the hydroalcoholic extract of H. flavum. Phytochemical constituents were tentatively recognized by contrasting their molecular weights, retention time (Rt), and tandem mass (MS/MS) fragmentation model with formerly reported literature. Figure 2 shows the HPLC-ESI-MS/MS base peak chromatogram of the H. flavum extract. Seventy compounds, up till now, were recognized from the leaf extract of H. flavum organized into hydroxycinnamic acid derivatives, flavonoids, and other miscellaneous compounds.

Docking studies
We performed molecular docking of compounds (1)(2)(3)(4)(5)(6)(7)(8) into the binding pockets of both RAF-1 and ERK-2 including the co-crystallized inhibitor (9) in each case. They got stabilized inside the binding pockets of RAF-1 and ERK-2 by promising scores and bound interactions with the amino acids of both receptors.
By analyzing the binding pockets of RAF-1 and ERK-2 proteins containing the co-crystallized inhibitors, we found that: the co-crystallized inhibitor of RAF-1 was stabilized by forming five H-bonds with Cys424, Lys375, Ser428, and Ile355 amino acids. On the other hand, the co-crystallized inhibitor of ERK-2 was able to compose a covalent bond with Cys164, two H-bonds with Met106 and Ser 151, and a pi-H bond with Ile29.

Cytotoxicity assay
We detected the cytotoxic concentration of 50% (CC 50 ) of the indicated purified H. flavum extract on HepG2 cells and the normal hepatocytes using MTT and LDH production assay kit. Accordingly, the cells were peculated in 96-well/plate in a density of 10 000 cells/well and were left overnight. Then, the cells were treated all night with more than one concentration of H. flavum agent (0-25 mg/mL). Interestingly, the cell viability rate of HepG2 cells was being interrupted at a low concentration of H. flavum treatment (200 ug/mL) and revealed 50% inhibition at the concentration of 600 ug/ml. Meanwhile, the cell viability rate of the normal hepatocytes showed an undetectable toxic effect at the same concentrations of H. flavum treatment (Figures 3A and 3B). The CC 50 of H. flavum agent on the normal cells was almost 1 mg/mL indicating that the plant agent might disturb the cancer cells at a low concentration without any detectable cytotoxic effect on the normal cells. The treated cells were checked for their LDH production, which is considered an indicator of the chemical-mediated cytotoxicity in HepG2 cells   Compounds (1) and (3) against both raf-1 and ERK-2 receptors compared to the docked co-crystallized inhibitor (9) in each case, respectively.

Red dash represents H-bonds and black dash represents H-pi interactions.
and normal cells. There was a significant elevation in the relative LDH production up to threefold in HepG2 cells treated with 200 ug/ml and gradually increased in a dosedependent manner compared to the normal hepatocytes ( Figure 3C). This result further confirms the cytotoxic effect of H. flavum agent on HepG2 cells indicated by the production level of LDH upon treatment compared to the normal cells. Table 6 shows the degree of gene expression of RAF-1 and Erk-2 genes in the HepG2 cell line after treatment with sorafenib, compound 1, compound 3, and their combination. The gene expression levels of the RAF-1 and Erk-2 genes were tested in cell lines prior to the treatment with the 100 µg/mL of sorafenib, compound 1, compound 3, and their combination, along with a comparison to the Error bars indicate standard deviation (SD) of four different replicates. Student two-tailed test was used to determine P values and significance of LDH production level. control using PCR. Figure 4 shows the gene expression patterns for (a) RAF-1 and (b) ERK-2 in hepatocellular carcinoma (HepG2) cell lines, estimated by real-time PCR. The order of gene expression inhibition for the three estimated samples was as follows: sorafenib > compound 1 > compound 3. The inhibitory effect of sorafenib with compound 1 on both RAF-1 and ERK-2 genes was superior to sorafenib alone, which illuminates the strong inhibition of this combination on the RAF/MEK/ERK pathway. The results disclose the underlying molecular mechanisms formerly stated results of the cytotoxic assay. Our results revealed that compound 1 enhanced the cytotoxic effect of sorafenib on HepG2 and decreased cell viability (downregulated the gene expression of both RAF-1 and Erk-2).

Discussion
Two new flavonoids (compounds 1 and 2) were isolated from Hymenosporum flavum together with the other six known compounds. Compound 1 was obtained as a yellow powder, with molecular formula C 26 (Table 2) affirmed the presence of 15 flavonol carbon signals, confirming the quercetin aglycone after comparing the NMR data with literature (33). The remaining carbon signals (Table 2) proved the existence of two sugar moieties, which deducted ribose and glucose after comparing the NMR data with literature (33,34). A downfield shift of C-2'' (δ78.3, ribose) also supported the attachment of the glucopyranosyl unit.
Further confirmation was carried out by 13 C NMR where the shift of C-3 up-field (from δ C 135.8 to 134.3,), while that of C-2 downfield (from δ C 146.9 to 157.07) when compared with quercetin (33). Consequently, compound 1 was elucidated as quercetin-3-(glucopyranosyl 1→2 ribopyranoside), where this is the first time isolated in nature.
Compound 2 isolated as yellow needles displayed a dark purple spot on paper chromatogram under UV light, which became yellow upon exposure to ammonia vapor. It has a molecular formula C 26  H-6′) designated a kaempferol flavonol. Two anomeric proton doublets were confirmative for two sugar moieties. The first signal at δ H 5.38 (H-1″) besides its coupling constant (J=3.36 Hz) lower than 4.0 Hz, revealed ribose moiety α-configuration, after the exclusion of arabinose and xylose sugars (33). The other signal at δ H 4.46 with its coupling constant (J=7.72 Hz, H-1′″) evidenced the beta configuration of the glucose sugar.
The 13 C NMR signals (Table 2) affirmed the 15 carbon signals of a flavonol, as compared with literature, confirmed the kaempferol aglycone (33). Other 13 C NMR signals showed two sugar moieties assigned for ribose and glucose after comparison with literature (23,33). A downfield shift of C-2'' (δ78.6, ribose) also evidenced the attachment of glucopyranosyl unit. Further confirmation was carried out by C-3 up-field shift (from δ C 135.5 to 132.7) and C-2 downfield shift (from δ C 146.8 to 155.16), in comparison with kaempferol aglycone (33). Hence, compound 2 was recognized as kaempferol-3-(glucopyranosyl 1→2 ribopyranoside), where is the first time isolated in nature.
The docking simulation results revealed that all the isolated flavonoids (1-8), especially compounds (1) and (3) achieved the best binding scores towards both RAF-1 and ERK-2 receptors, which exceed the bound scores of the docked co-crystallized inhibitors (9) in each case (Table 4). However, compound (3) binding energy on RAF-1 and ERK-2 pockets were -8.85 and -8.17 kcal/mol compared to -7.15 and -6.05 kcal/mol of the docked cocrystallized inhibitor (9) in each case, respectively. The binding scores of compound (1) towards RAF-1 and ERK-2 binding sites were -8.43 and -7.34 kcal/mol compared to -7.15 and -6.05 kcal/mol of the docked co-crystallized inhibitor (9) in each case, respectively. Furthermore, the detailed binding modes of compounds (1) and (3) inside both RAF-1 and ERK-2 receptor pockets were similar to the native co-crystallized inhibitor (10) in binding with nearly the similar crucial amino acids superior to the docked co-crystallized one (9) in each case, respectively (Table 5).
We tried to clarify the role of quercetin glycosides as safe and efficacious anticancer agents. Interestingly, the cell viability rate of HepG2 cells was interrupted at a low concentration of the plant treatment. Meanwhile, the cell viability rate of the normal hepatocytes showed an undetectable toxic effect at the same concentrations of the plant treatment. The relative LDH production was significantly elevated threefold in HepG2 cells treated with 200ug/ml. It gradually increased in a dose-dependent manner when compared to normal hepatocytes. These findings indicate that H. flavum treatment is safe at both low and high doses in normal cells, where the extract provides a potent anticancer treatment.
According to the obtained docking results of the eight isolated flavonoids (1-8) of H. flavum extract compared to the co-crystallized inhibitors of both RAF-1 and ERK-2 receptor pockets, expressed promising idea about their affinities towards these carcinogenic proteins and subsequently expected high efficacies and intrinsic activities in their inhibition as well. As well, this study expected to be hopeful of the isolated flavonoids (1-8) of H. flavum leaf extract against raf-1 and ERK-2 receptors, especially compounds (1) and (3) either alone or in combinations with sorafenib for cancer treatment.
The current study concluded a synergistic combination of sorafenib with H. flavum and a combination of sorafenib with naturally purified quercetin glycosides that sensitize HCC cells towards sorafenib-induced apoptosis. We studied apoptosis-related genes' expression to unravel the underlying molecular mechanisms of the synergistic antitumor effects of H. flavum compounds and sorafenib on hepatocellular carcinoma cell line (HepG2). Therefore, we explored the biological activities of all compounds at a lower dose (100 μg/mL).
Many studies have shown that sorafenib depends on inducing tumor cell apoptosis in multi-cancers. Its mechanism depends on inhibiting the RAF/MEK/ ERK signaling pathway through suppressing tumor cell proliferation and the restriction process of angiogenesis (73). The mechanisms actions of sorafenib are diminishing cell growth, inducing cell cycle arrest in the G0/G1 phase, up regulation of the proapoptotic proteins (caspase 8 and caspase 3), and downregulation of the cell cycle-associated protein cyclin D1 and the anti-apoptotic protein MCL1. Sorafenib decreased the phosphorylation of ERK and MEK (74). Two pathways inhibited ERK in direct and indirect ways. It is better to directly prevent the ERK protein pathway due to its multiple cellular functions as well as organizing the distribution of upstream signals to its nuclear and cytosolic effectors. We succeeded in the extraction of two natural compounds that could inhibit the ERK pathway directly, so that compounds 1 & 3 inhibited ERK pathway alone. Also, both of them improved the efficiency of sorafenib in targeting the ERK pathways. Compounds 1 & 3 also inhibited the ERK pathways indirectly by targeting Raf-1. Raf-1 can activate RAS proteins and indicated that phosphorylation of serine 43 in Raf-1 is responsible for disruption of Ras/Raf association and downstream signaling (8,75). Interestingly, compounds 1 & 3 contributed to overcoming the resistance of HCC. Our findings showed a synergistically inhibitory effect on both Erk-1 and Raf-1 proteins interaction by compounds 1 and 3 in cancer cells.
The safety and inhibitory effects of sorafenib/compound 1 combined with cancer cell indicated that the use of compound 1 should be investigated in the future as a potential complementary to support the application of sorafenib therapy in HCC.

Conclusion
Two newly identified flavonols; quercetin-3-O-(glucopyranosyl 1→2 ribopyranoside) (1) and kaempferol-3-O-(glucopyranosyl 1→2 ribopyranoside) (2), along with other six flavonoids, were isolated from the leaf extract of H. flavum. Moreover, we identified seventy compounds from the HPLC-PDA/MS/MS of the hydroalcoholic extract, for the first time. The cytotoxic activity of the plant extract confirmed its potential action on HepG2 cells indicated by the production level of LDH upon treatment compared with the normal cells. Furthermore, compounds 1 and 3, which showed the best results in silico were further examined in vitro using qRT-PCR. They exhibited promising inhibitory activities against both RAF-1 and ERK-2 gene expressions. Also, both of them improved the efficiency of sorafenib in targeting both RAF-1 and ERK-2 pathways indicating synergistic combinations confirmed by the in vitro results of the MTT assay and PCR. The results revealed that compounds 1 and 3 could regulate the division of the tumor cells by restoring the sustained RAS/RAF/ERK signaling pathway and managing the programmed cell death. This mechanism showed that the isolated glycosides are safe in cancer cells' treatment with a noticeable effect on cell proliferation and angiogenesis. Our findings could be promising for further preclinical and clinical studies on the studied compounds, especially for compounds 1 and 3, either alone or in combinations with sorafenib for cancer treatment.

Authors' contribution
RFT and WAE designed the study, conducted the phytochemical analysis, and wrote the manuscript. EAK supervised the work, and critical revision of the article.
On the other hand, AIA, HK and AH implemented the animal experiment, biochemical and gene expression analysis. AAA conducted the molecular docking studies and contributed to the manuscript writing. All authors read and approved the final version and agreed to publish it.