3.1. GC-MS analysis and total phenolic (TP) and total flavonoid (TF) contents of D. arborea and B. ferruginea
The GC–MS spectrum of D. arborea and B. ferruginea confirmed the presence of various phytocomponents observed at a specific retention time (Fig. 1A, B). The GC–MS chromatograms of these compounds are shown in figures S1 and S2. There were 16 compounds identified from the GC–MS chromatogram of D. arborea and 32 compounds from B. ferruginea (Fig. 1A, S1 and Table S1, S2). Some of these compounds are known as antioxidant agents, according to the literature. For instance, AL-Adhreai et al [40] reported the antioxidant potential of isoxazolidine supported by in vitro, in silico and pharmacodynamic studies. In addition, N-propylacetamide found in B. ferruginea is used as an antiepileptic agent, but it anti-oxidant property is also reported [41]. No previous works on the GC-MS analysis of D. arborea have been found in the literature. The therapeutic potentials of these plants could be attributed to their hight content in phenolics and flavonoids. As shown in table 3, the TP and TF contents in B. ferruginea were significantly (p < 0.001) greater than in D. arborea. There is no previous report on the quantification of TP and TF contents in D. arborea aqueous extract. In this study, the TP content in the aqueous extract of B. ferruginea was 295.25 ± 15.25 mg GAE/g DW, which is greater than the content reported by Afolabi et al. [42] (200.87 ± 1.51 mg GAE/g DW), and Mahomoodally et al. [43] (193.58 ± 0.98 and 187.84 ± 1.88 mg GAE/g DW in the methanolic and aqueous extracts of B. ferruginea stem bark, respectively). The TF content in B. ferruginea (10.42 ± 1.25 mg of QE/g DW) was comparatively lower than that reported by Mahomoodally et al. [43] (24.37–42.31 mg/g) in the aqueous leaf extract of B. ferruginea, but greater than that obtained by Afolabi et al. [42] (0.44 ± 0.01 mg QE/g DW) in the free phenolic extract of the aerial leaves of B. ferruginea. Moreover, Oladejo et al. [44] showed that n-hexane fraction of B. ferruginae contained greater total phenolic (14.14 ± 0.30 mg/GAE g) and flavonoid (7.82 ± 0.10 mg/QUE g) contents than ethyl acetate (13.42 ± 0.41 mg/GAE g; 5.76 ± 0.07 mg/QUE g), butanolic (8.42 ± 0.14 mg/GAE g; 2.60 ± 0.11 mg/QUE g) and residual aqueous fractions (6.57 ± 0.04 mg/GAE g; 3.78 ± 0.06 mg/QUE g). These variations in phenolic and flavonoid contents in B. ferruginea may be ascribed to factors such as climate, place, period of collection and extraction protocols [45, 46]. For instance, in the current work, the stem barks of B. ferruginea were harvested in May 2023 in the village of Bangangte in Cameroon and extracted my macerating in water, while Afolabi et al. [42] collected the aerial leaves of B. ferruginea in Nigeria (Ado-Ekiti, Ekiti State) and extracted using a free phenolic extraction method.
3.2. Effects of D. arborea and B. ferruginea on DPPH and ABTS+ free radical scavenging activities
The DPPH assay evaluates the capacity of antioxidants to neutralize DPPH radicals through electron transfer reactions, while the ABTS+ method measures the ability of an antioxidant compound to reduce the 2,2’-azino-bis(3-ethylbenzothiazoline-6-sulfonate) radical cation (ABTS˙+) (dark blue color) into colorless ABTS [47]. There is no report on the ABTS+ free radical scavenging activities of D. arborea and B. ferruginea in the literature. As shown in Fig. 2A and B, the DPPH and ABTS+ free radical scavenging activities of D. arborea, B. ferruginea and ascorbic acids were dose-dependent (except DPPH activity of D. arborea at 2000 µg/ml). At all concentrations, we noticed that the activity of B. ferruginea was significantly (p < 0.05 − 0.001) higher than that of D. arborea. Moreover, B. ferruginea (2000 µg/ml) displayed the highest DPPH activity, compared to D. arborea and ascorbic acid. This strong antioxidant activity of B. ferruginea is consistent with the work of Olarewaju et al. [48] who showed that the DPPH free radical scavenging activity of B. ferruginea aqueous extract is higher than that of ethanolic and ethyl acetate extracts. Moreover, the beta-amyrin acetate fraction in B. ferruginea has been reported as a potent antioxidant agent due to its effective DPPH free radical scavenging potential [49]. In the current work, we observed that the DPPH free radical scavenging activity of D. arborea (7.8–2000 µg/ml) ranges from 20.99–47.50%. Similar, Nwaehujor et al. [50] showed that the methanolic extract of D. arborea (10–400 µg/ml) exhibits a moderate DPPH free radical scavenging activity (from 13% to 50,3%). The yellowish colour observed at high concentration illustrate the ability of the plant extracts to reduce DPPH, probably by donating electron or hydrogen radical, as suggested earlier [50]. The intensity of the yellow color was more pronounce with B. ferruginea, which was correlated with its strong DPPH free radical scavenging activity at high concentrations (Fig. 2A).
3.3. Cupric and ferric reducing antioxidant power of D. arborea and B. ferruginea
The CUPRAC method gauges the electron transfer ability of sample components by assessing their reducing power [51]. On the other hand, the FRAC method test is based on the capacity of a compound to reduce a ferric salt (Fe3+) to ferrous salt (Fe2+) by electron transfer reaction [52]. These reliable methods were used in the present study to estimate the in vitro antioxidant potentials of D. arborea and B. ferruginea. For all compounds, a dose-dependent response was recorded. Indeed, at all concentrations, we noticed that the CUPRAC activity of B. ferruginea was higher than that of ascorbic acid and D. arborea, while ascorbic acid exhibited the strongest FRAC activity. The CUPRAC and FRAC activities of B. ferruginea were higher than that of D. arborea (Fig. 2CD). The low FRAC activity of D. arborea observed in this study corroborates previous reports which showed that the FRAC values of the methanolic extract of D. arborea at 400 µg/ml and 1000 µg/ml were 1.20 mM and 1.34 mM, respectively [50]. The potent FRAC activity of B. ferruginea has been discussed in previous reports. Indeed, Bothon et al. [53] reported that the FRAC activity of the ethanolic extract of Bridelia ferruginea (4.4 ± 0.06 µmol Fe II / g DW) is higher than that of Ceiba pentandra (0.14 ± 0.01 µmol Fe II / g DW). The FRAC activity of various fractions (n-hexane, n-butanol, ethyl acetate and residual aqueous fraction) of B. ferruginea has been previously published [44]. In addition, the powerful FRAC activity of free and bound phenolic-rich extracts of B. ferruginea has been reported [42]. The potent FRAC activity of B. ferruginea observed in the present study is due to its ability to reduce Fe3+ to Fe2+.
3.4. Inhibition of lipid peroxidation by D. arborea and B. ferruginea
The ability of D. arborea, B. ferruginea and ascorbic acid to inhibit lipid peroxidation is shown in Fig. 3. The inhibitory activity of ascorbic acid was significantly (p < 0.001) higher than that of plant extracts. At low doses (7.8–62.5 µg/mL), B. ferruginea was more effective than D. arborea, while the inhibitory trend of both plants was similar at high doses. This result corroborates the work of Oloyede and Babalola [54] who illustrated the capacity of the ethanolic extract of B. Ferruginea to prevent oxidative stress by inhibiting lipid peroxidation. Moreover, the anti-lipid peroxidation effect of B. ferruginea aqueous extract has been reported [55]. The strong antioxidant potential of B. ferruginea could be due to its high phenolic and flavonoid contents, as suggested earlier [54].
3.5. Effects of D. arborea and B. ferruginea on hemolysis
Homolysis assay evaluates the effect of substance on erythrocyte membrane stability using RBCs. The effects of various concentrations (7.8, 15.6, 31.3, 62.5, 125, 250, 500, 1000, 2000 µg/mL) of plant extracts on homolysis after incubation with human RBCs is presented in Fig. 4A, B. According to ASTM F756-00 norms, compounds with hemolysis percentages of more than 5%, between 5 and 2%, and less than 2% are classified as hemolytic, slightly hemolytic and non-hemolytic, respectively [56]. Thus, D. arborea was hemolytic at doses 2000 µg/mL (84.71 ± 8.93%), 1000 µg/mL (63.72 ± 10.20%) and 500 µg/mL (17.58 ± 4.34%), slightly hemolytic at dose 250 µg/mL (2.22 ± 0.46), and non-hemolytic at doses 125 to 7.8 µg/mL. Similarly, Okonkwo and Egeolu [57] reported that the ethanolic extract of D. arborea possesses no anti-hemolytic activity in the RBCs of Wistar albino rats pre-treated with Zidovudine. On the other hand, B. ferruginea was hemolitic at doses 2000 (13.70 ± 8.92%) and 1000 (5.31 ± 3.28%) and non-hemolytic at doses 500 to 7.8 µg/mL, which is consistent with the study of Karou et al. [58]. The hemolitic percentages of the two plant extracts were significantly (p < 0.05 − 0.001) lower than that of Triton X-100, used as a positive control. Overall, the hemolitic effect of B. ferruginea was lower than that of D. arborea (Figs. 4A, B). B. ferruginea (at moderate or low doses) did not cause any significant toxicity to RBC after incubation, probably meaning that it did not affect erythrocyte membrane stability in vitro.
3.6. Cytotoxicity of the aqueous extracts of D. arborea and B. ferruginea in PC3, NIH3T3 and BT474 cells
D. arborea decreased (p < 01.05 − 0.001) the viability of PC3, NIH3T3 and BT474 cells in a dose-dependent manner (Fig. 5A-C). At all doses, the cytotoxicity of D. arborea was most pronounced in BT474 cells. Indeed, the inhibitory values of BT474 cells after 24 hours of incubation with D. arborea at doses 2000, 1000 and 500 µg/ml were 5.56 ± 0.43%, 27.91 ± 2.2% and 29.47 ± 6.95%, respectively (Fig. 5C), indicating a potential candidate for anticancer drug development. The ability of D. arborea to decrease cell growth is in accordance with previous studies on the anticancer properties of this plant. Indeed, it has been proven that the methanolic extract of D. arborea exhibits an anticancer effect in human breast cancer cells (MDA-MB-231) by down-regulating VCAM-1 and EMT proteins through Akt/PKC pathway [59]. On the contrary, the aqueous extract of B. ferruginea had the tendency to feed the cells, since a dose-dependent increase in the viability of NIH3T3 and BT474 cells was observed after incubation (Fig. 5 BC). For instance, at doses 500, 1000 and 2000 µg/ml, the viability of NIH3T3 cells was 69.71 ± 5.1%, 69.88 ± 5.18% and 78.24 ± 10.62%, respectively. On the other hand, high BT474 cells viability was recorded after incubation with B. ferruginea at 500 µg/mL (98.40 ± 10.09%), 1000 µg/mL (98.55 ± 6.27%) and 2000 µg/mL (99.87 ± 4.67%). This trend of increase in the cell viability has been also reported with other plants like Guibourtia tessmannii [60], Calophyllum Brasiliense [61] and Hunteria umbellata [62], meaning that compounds found in the plant extracts could have the capacity to feed cells during incubation. However, the effects on B. ferruginea on cell viability may vary depending on the cells. In fact, the cells viability of PC3 cells after exposure to various concentrations of B. ferruginea was statistically unchanged (P > 0.01) (Fig. 5A). Similarly, Bothon et al. [63] showed that the ethanolic extract of Bridelia ferruginea has no cytotoxic activity on human fibroblast primary culture, as the percentage of cell survival being greater than 60% after exposure to various concentrations (0–25 µg/mL) of plant extract.
3.7. Antioxidant properties of D. arborea and B. ferruginea in PC3, NIH3T3 and BT474 cells
In the present study, PC3, NIH3T3 and BT474 cells were incubated with various concentrations of plant extracts and DPPH and ABTS radical scavenging activity, as well as ferric reducing power (FRP) were measured. In general, these oxidative stress parameters were increased (in a dose-dependent manner) in all cells after incubation with D. arborea and B. ferruginea extracts. In all cells, the DPPH (except in PC3 cells) and ABTS radical scavenging activity, and FRP of B. ferruginea were higher than that of D. arborea (Fig. 6A-I). Indeed, the NIH3T3 cells incubated with B. ferruginea (doses 250–1000 µg/mL), exhibited a significant (P > 0.05) increase in FRP, compared to the cells incubated with D. arborea (Fig. 6F). Moreover, the DPPH radical scavenging activity in PC3 cells after incubation with B. ferruginea (doses 250 µg/ml) was significant (P > 0.05) increased, compared to the cell exposed to D. arborea (Fig. 6G), which indicate the ability of these plants to act as free radical scavengers. The antioxidant activity of B. ferruginea observed in PC3, NIH3T3 and BT474 cells is consistent with the work of Oyebode et al. [63] who reported the free radical scavenging ability of B. ferruginea in vitro. Moreover, the DPPH scavenging activity of the aqueous and ethanolic extracts of B. ferruginea have been reported in FS5 fibroblasts cells [64], justifying the uses of this plant in the treatment of oxidative stress related diseases.
3.8. Drug likeness and toxicity prediction
Drug-Likeness and toxicity prediction of all compounds found in D. arborea and B. ferruginea are summarized in Tables S4 and S5. All compounds were within the acceptable standards, according to Lipinski’s rule which states that a potential drug candidate should not violate no more that one of the following criteria: less than 5 hydrogen bond donors, less than 10 hydrogen bond acceptors, a molecular mass below 500 daltons and less than 5 calculated octanol-water partition coefficient [65–67]. Compounds with high GI absorption satisfying all Lipinski’s rules (5/5) were selected and used for molecular docking against Keap1, GST, PARP10 and p90 RSK enzymes (Figure S3).
3.9. Molecular docking results of selected compounds docked against Keap1 and GST
In the current study, the aqueous extracts of D. arborea and B. ferruginea exhibited a potent antioxidant activity, B. ferruginea, being the most effective. Hence, selected compounds found in these plants were docked against two antioxidant-related enzymes: Keap1 and GST. The importance of Keap1-Nrf2 pathway in the regulation of cellular oxidative stress is well documented. This pathway regulates various antioxidant markers such as heme oxygenase-1, peroxidase-1, superoxide dismutase and glutathione peroxidase [35, 68]. GST is one of the main antioxidant enzymes involved in the prevention of lipid peroxidation and maintenance of cellular homeostasis [69]. Our rationalization for selecting Keap1 and GST as oxidative stress markers and key targets of selected compounds from plant extracts is based on previous studies [35–37].
The binding affinity of selected compounds from D. arborea with Keap1 protein ranges from − 5.72 to -8.17 kcal/mol, which was below the value observed with Nrf2 (-11.82 kcal/mol). Among the compounds with good affinity with Keap1, we found that 2-deoxypentose displayed a strong interaction with the active pocket of Keap1 through three conventional hydrogen bonds with the amino acids Ile-559, Val-604 and Val-463 (Table S6, Fig. 7A). The docking of 1-butanamine, n-methyl or 1-propanamine, N,2-dimethyl with Keap1 indicated the binding interaction with the same amino acid (Ser-602) through two conventional hydrogen bonds, with the binding affinity of -5.90 and − 5.95 kcal/mol, respectively (Table S6, Fig. 7BC). Interestingly, 1,3-Propanediamine N-methyl displayed the best stability with Keap1 protein, compared to all compounds. Indeed, it was linked to Leu-557, Ala-510, Val-463 and Ile-416 residues of protein via six hydrogen bonds with a strong Van-der Walls interactions with nine amino acids, which could justify its strong stability (Table S6, Fig. 7D). Cis-cyclohexane-1,3-diamine showed a binding affinity of -6.73 kcal/mol and interacted with Keap1 through five conventional hydrogen bonds, while Nrf2 was linked to the protein via two hydrogen bonds (Table S6, Fig. 7E). The docking result of selected compounds from B. ferruginea against Keap1 is shown in table S6 and Fig. 8. In brief, the best stability with the protein was observed with 1,3-propanediamine, n-methyl (through 6 hydrogen bonds), followed by N-Methyl-1,3-propanediamine (via 5 hydrogen bonds) and ethen-1,2-diol (through 2 hydrogen bonds) (Table S7 and Fig. 8). However, Nrf2 (a Keap1 inhibitor) displayed the best binding affinity with Keap1.
Selected compounds from plant extracts were also docked against GST and piperlongumine (a GST inhibitor) was used for comparison. As expected, piperlongumine exhibited a strong stability with the binding pocket of GST through 3 conventional hydrogen bonds with Cys-101 and Tyr-108 residues (Table S6 and Fig. 7L). Docked interactions of selected compounds from D. arborea against GST are shown in table 9 and Fig. 7. The docking of 2-deoxypentose with GST had showed a binding affinity of -6.63 kcal/mol with a strong interaction through 2 hydrogen bonds (Cys-101 and Asp-306) (Fig. 7G). 2-deoxy-alpha-d-ribopyranose and 1,3-Propanediamine, N-methyl exhibited an affinity of -7.52 and − 5.98 kcal/mol, respectively, and interacted with the active pocket of GST via 2 hydrogen bonds (Table S9 and Fig. 7 HI). In addition, the strong Van-der Walls interactions and the presence of 3 carbon hydrogen bonds (Gln-259, Pro-261, Leu-260) between 2-deoxy-alpha-d-ribopyranose and GST may justify its high binding affinity. Cis-cyclohexane-1,3-diamine displayed good stability with GST compared to Piperlongumine, which could be due to the presence of 5 conventional hydrogen bonds between the ligand and Gln-135, Leu-132 and Ile-143 residues (Table S9 and Fig. 7 KL). Among the compounds found in B. ferruginea, 2-[(methylamino)methyl]cyclohexanol exhibited the highest binding affinity (-8.66 kcal/mol) with GST, which was similar to that of Piperlongumine (-8.85 kcal/mol), used as GST inhibitor in this study (Table S7). Overall, other ligands displayed similar binding affinity with the protein, but their stability was quite different. For instance, 2-[(methylamino)methyl]cyclohexanol was linked to the active pocked of GST via one conventional hydrogen bond (Arg-282), while N-propylacetamide (Arg-221 and Asn-412) and N-Methyl-1,3-propanediamine (Ile-203 and Asn-204) were attached to the protein through two conventional hydrogen bonds (Table S7 and Fig. 8 GHI). In addition to three hydrogen bonds with the residues Cys-101 and Tyr-108 of GST, the highest binding affinity of piperlongumine could be due to the high Van-der Walls interactions with 11 amino acids of the active site of GST (Table S7 and Fig. 8L).
Based on these results, it appears that compounds identified in D. arborea and B. ferruginea have a unique capacity to bind to the active pocket of Keap1and directly inhibit the crucial Keap1–Nrf2 interaction, leading to free Nrf2 release. Moreover, the strong binding affinity of these compounds with GST indicates their antioxidant potential in silico, which corroborates the in vitro antioxidant activity of these plants observed in this study. These findings are consistent with other reports on these plants using animal models. For instance, D. arborea has the ability to boost the antioxidant markers in rats with varicocele [9]. B. ferruginea exhibits a potent antioxidant potential in N-nitrosodiethylamine-treated rats by improving ROS detoxification [70].
3.10. Molecular docking results of selected compounds docked against PARP10 and p90 RSK
PARP10 is involved in various aspects of cell function and its activity is associated to cell survival [71] and DNA repairing [72]. p90 RSK is a member of RSK family involved in the regulation of various cellular and metabolic processes, including cell growth [73, 74], cell signaling [75], cellular homeostasis [76] and DNA repair [77]. To predict the cytotoxicity of selected compounds from D. arborea and B. ferruginea, molecular docking on PARP10 and p90 RSK proteins was performed. Veliparib [36] and LJH685 [37] were used as a selective inhibitor of PARP10 and p90 RSK, respectively, according to previous reports.
The binding affinity of selected compounds from D. arborea against PARP10 was similar, but they were matched differently at the active pocket of the protein. Docked interactions of 2,3-butanediol with PARP10 had shown the interactions with Gly-1077 and His-1076 through 3 hydrogen bonds in addition to a strong Van-der Walls interactions with five amino acids (Table S8 and Fig. 9A). 2-deoxypentose interacted with the active site of PARP10 via 4 conventional hydrogen bonds with Arg-1146, Arg-1147, Gly-1145 and Val-1102, but they were only one Van-der Walls interaction with Tyr-1103 and no Alkyl/Pi-Alkyl interactions (Table S8 and Fig. 9B). Among all compounds, the best stable ligand was 2-deoxy-alpha-d-ribopyranose, as it was well-matched in the PARP10 pocket through five hydrogen bonds with Ala-962, Arg-957, Gly-959 and Arg-961, which can also justify its strong binding affinity (Table S8 and Fig. 9C). 1,3-Propanediamine, N-methyl had three hydrogen bonds with the protein via Tyr-1075 and Tyr-1108 residues, while Cis-cyclohexane-1,3-diamine was linked to PARP10 by two conventional hydrogen bonds (Tyr-1108, Tyr-1075) (Table S8 and Fig. 9DE). Veliparib used as control in this experiment exhibited the highest binding affinity (-8.26 kcal/mol) with PARP10, but was not well matched like 2-deoxy-alpha-d-ribopyranose. Thus, its high binding affinity with the protein could be due to the strong Van-der Walls interaction with 10 amino acids (Thr-1079, Asp-1085, Phe-1091, Glu-1106, Arg-1093, Val-1107, Asn-1099, Tyr-1108, His-1076, Thr-1078) as shown in Table S8 and Fig. 9F. Among the compounds isolated from B. ferruginea, we found that 2-[(methylamino)methyl]cyclohexanol exhibited the best binding affinity with PARP10 (-9.80 kcal/mol), which was higher than that of Veliparib (-8.26 kcal/mol) (Table S9). Isobutyramide and N-Methyl-1,3-propanediamine interacted with the active site of PARP10 via 3 conventional hydrogen bonds, while 2-[(methylamino)methyl]cyclohexanol, heptylamine and 4-octen-3-one were attached to the protein through 2 hydrogen bonds (Table S9; Fig. 10A-E). The hydrogen on the pyridine nitrogen of PARP10 formed a conventional hydrogen bond with Asn-1092, while 2 hydrogens of the pyrimidine nitrogen formed 2 hydrogen bonds with the amino acids Cys-1096 and Phe-1095 (Fig. 10F). As reported in previous reports, the main amino acids of the active pocket of PARP10 that interact with compatible ligands are amino acids Tyr-914, Tyr-919, Tyr-932, Ala-911, Ala-921, Gly-888, Ser-927, His-887, Leu-926 and Asn-910 [36, 78, 79]. In the current work, we found that the majority of compounds found in the plant extracts were bound to the active sites of PARP10 via these amino acids. For instance, cyclopropanecarboxamide was linked to Ser-927 via one hydrogen bond (Table S8), while 2-[(methylamino)methyl]cyclohexanol was bound to Tyr-919 through 2 hydrogen bonds (Table S9). Cyclopropanecarboxamide and 4-octen-3-one were linked to two amino acids: (Ser-927; Tyr-919) and (Ala-921, Ser-927), respectively, via two hydrogen bonds (Table S9).
LJH685 used as a selective p90 RSK inhibitor had the best binding affinity (-9.61 kcal/mol) with the protein, but was not well-fitted into the RSK pocket, compared to other ligands from plant extracts (Table S8). For instance, 2-deoxypentose, 2-deoxy-alpha-d-ribopyranose and 1,3-Propanediamine, N-methyl- formed 3 hydrogen bonds with p90 RSK, while LJH685 interacted with the protein only through one conventional hydrogen bond (Table S8; Fig. 9H-J). Cyclopropanecarboxamide possessed a binding energy of -6.96 kcal/mol and interacted with the active site of p90 RSK via one hydrogen bond (with Lys-142), but exhibited the highest Alkyl/Pi-Alkyl interactions (Leu-70, Met-99, Val-101, Phe-83) among all ligands found in D. arborea (Table S8; Fig. 10).
The docking of compounds found in B. ferruginea showed that 2-[(methylamino)methyl]cyclohexanol displayed the highest binding affinity with p90 RSK (-10.60 kcal/mol) and was well-matched to the active pocket via 3 conventional hydrogen bonds with amino acids Leu-285, Ala-283 and Arg-300 (Table S9; Fig. 10H). The additional Alkyl/Pi-Alkyl interactions (with 2 amino acids: Met-287 and Arg-305) and Van-der Walls interactions (with 6 amino acids: Leu-299, Lys-284, Ile-280, Phe-268, Pro-237 and Lys-304) may justify its high binding affinity with p90 RSK. N-Methyl-1,3-propanediamine and 1,3-propanediamine, n-methyl- also possessed 3 hydrogen bonds with p90 RSK, but displayed lower binding affinity (-5.64 kcal/mol and − 5.64 kcal/mol, respectively) (Table S9; Fig. 10IJ). The docking of 4-methoxy-2-butyn-1-ol and 4-octen-3-one with p90 RSK had shown interaction with one amino acid through one hydrogen bond, with the binding energies of -6.45 kcal/mol and − 8.07 kcal/mol, respectively. However, 4-octen-3-one displayed additional Alkyl/Pi-Alkyl interactions with seven amino acids (Val-82, Lys-100, Ala-98, Leu-147, Val-131, Leu-150, Leu-200) which could justify its high binding affinity (Table S9; Fig. 10JK).
In the present study, we found that B. ferruginea exhibited a potent antioxidant activity, while D. arborea decreased the proliferation of PC3, NIH3T3 and BT474 cells. The proposed mechanism of action of these plants is summarized in Fig. 11. Based on the results obtained, it could be proposed that B. ferruginea could exhibit its potent antioxidant activity by inhibiting the peroxidation of cellular lipids (i), which could promote fatty acid oxidation at the level of mitochondria (ii), resulting in the reduction of ROS and finally leading to the reduction of oxidative stress (iii). In addition, this plant could directly increase the production of antioxidant enzyme [63, 64] and Keap1(iv), as also reported by previous authors. The good binding interactions of selected compounds against Keap1 and GST supports this probable mechanism. Bax, cytochrome C and caspases 3/9 are the main proapoptotic factors involved in the intrinsic pathway of apoptosis. D. arborea could activate Bax (i), which could enter mitochondria and induce release of mitochondrial cytochrome C (ii). This proapoptotic factor could then activate caspase 3 and 9 (iii), which could promote apoptosis and decrease cell proliferation. Under stress or pathological conditions, the activation of ERK1/2 induces apoptosis and cell death via p90 RSK and other various signalling cascades [80]. In the current study, the phytocompounds found in D. arborea exhibited a strong interaction with the active pocket of p90 RSK. For instance, 2-deoxy-alpha-d-ribopyranose was the most stable compounds, as it was attached to the active pocket of the protein via 5 conventional hydrogen bonds. This ligand could activate ERK1/2 and phosphorylate p90 RSK (iv), which in turn could increase apoptosis and decrease cell proliferation. Another partway could be via PARP10 phosphorylation. It is well known that PARP10 modulates apoptosis, DNA repair and cell viability [81–83]. This ligand could directly modulate PARP10 phosphorylation (v), which may promote apoptosis and decrease cell proliferation. Further mechanistic studies are highly needed.