In Vitro Antibacterial, Antioxidant, Cytotoxicity Activity, and In Silico Molecular Modelling of Compounds Isolated from Roots of Hydnora johannis

The plant Hydnora johannis has been utilized in folk medicine. Analyzing phytochemical composition of dichloromethane/methanol (1 : 1) root part of Hydnora johannis gave oleic acid (1), caffeic acid-2-hydroxynonylester (2), catechin (3), and a pregnane derivative (4). NMR spectroscopy was used to characterize compounds 1–3, while compound 4 was identified through GC-MS analysis and literature comparison. The cytotoxicity of extracts from roots of H. johannis was conducted against MCF-7 cell lines (human breast cancer) by MTT assay. According to the cytotoxicity study, n-hexane extract exhibited a high level of toxicity with 28.9 ± 5.6% cell viability. Antibacterial activity was tested against Escherichia coli, Pseudomonas aeruginosa, Staphylococcus aureus, and Streptococcus pyogen. The highest bacterial growth mean inhibition zone was measured for catechin (3) (13.72 ± 0.05 mm)) against P. aeruginosa at 0.25 mg/mL and acceptable related to standard. Antioxidant activity was studied by the DPPH assay. Based on the data from the antioxidant study, DCM/MeOH extract (70.32%) and catechin (3) showed good antioxidant activity (65.61%) (IC50 0.25 μg/mL) relative to that of the positive control (78.21%, IC50 0.014 μg/mL) at 12.5 μg/mL. In each docking pose, catechin (3) scored higher binding affinity of −7.9, −7.2, and −6.4 kcal/mol towards PqsA, DNA gyraseB, and S. aureus PK, respectively, compared to amoxicillin (−8.1, −6.1, and −6.4 kcal/mol). All five Lipinski rules were obeyed by compounds 1–3, which showed an acceptable drug resemblance. The lipophilicity was computed as less than five (1.47–4.01) indicating a lipophilic property. Catechin (3) obeys Veber's rule implying its good oral bioavailability. Binding affinity scores of catechin (3)-protein interactions are in line with those from in vitro tests, indicating its potential antibacterial effect. The obtained cytotoxicity and antibacterial activity results support the utilization of H. johannis in folk medicine.


Introduction
Genus Hydnora (Hydnoraceae) includes hypogenous parasitic plant species with a massive root system and reduced morphological features [1,2].It is widely distributed in the region of southern Africa, across East Africa to the Arabian Peninsula [3,4].Eight species of genus Hydnora [5] were described.Te plant species of genus Hydnora has been used as medicine for both human and livestock [6,7].Bioactive phytochemical constituents including favonoids, tannins, terpenoids, saponins, and alkaloids have been reported from genus Hydnora [2,8].Hydnora johannis Becc.(synonymous: H. abyssinica A.Br.) is the widely spread species in the African continent.It parasitizes on members of genus Acacia, Albizia, and Kigelia Africana plants [1,9].In Ethiopia, H. johannis occurs dominantly in Shewa, Sidamo, Bale, Arsi, and Hararge, and is known by its local name DechMerech (Afan oro/Amh), Lipti (Som) [10].Te root extracts are used to treat ailments including hemorrhoid, swollen abdomen, cancer, diabetes, dysentery, and kidney and bladder complaints [6].In Sudan, the water extract is given for healing tonsillitis and dysentery [11].Traditional medicine practitioners in Ethiopia recommend the plant as a therapeutic agent for swelled breast [10], painful swelling (gofa, Afan Oromo), hemorrhage [12], cancer, tumors, and infammatory ailments [13].Due to the fact that the plant has prevalent ethnomedicinal uses to treat various diseases, more studies need to be developed about the chemical constituent and biological activities of the plant H. johannis.In this sense, this study seeks to examine cytotoxicity, antibacterial, and antioxidant properties of extracts, and molecular modelling of compounds from CH 2 Cl 2 /CH 3 OH root extract of H. johannis.

Materials and Methods
2.1.Research Procedure.Maceration was used to obtain crude extracts.On precoated silica gel GF254 (Merck) aluminum sheet, analytical thin layer chromatography (TLC) was performed.A UV cabinet with a lamp of 254 and 365 nm was employed to visualize TLC spots, iodine vapor, and vanillin spray.A column was set on silica gel with 60-120 mesh size.Te NMR data were obtained on Bruker Avance 400 MHz spectrometer.Both, GC/MS: coupled GC (7890B, USA) and MS (5977A Network), were obtained from Agilent Technologies.All materials used met the required standard.

Preparation of Plant Materials.
Root part of H. johannis was obtained from Huruta, Arsi Zone, Oromia, Ethiopia, Jan 2020.After being confrmed by the chief technician, Botanist Melaku Masresha, kept at the National Herbarium, AAU, Ethiopia, with voucher code THJ10/17.

Isolation of Pure Compounds.
Five hundred grams of airdried and powdered roots of H. johannis has been consecutively extracted by n-hexane, CH 2 Cl 2 /CH 3 OH, and CH 3 OH for 72 h each, three times at normal temperature.Each fltrate was subjected to a rotary evaporator at 40 °C to produce crude extracts weighing 22 g, 35 g, and 30 g, respectively.Ten grams of CH 2 Cl 2 /CH 3 OH (1 : 1) crude extract was loaded to fractionation with rising polarity of n-hexane/ EtOAc and then by DCM/MeOH.Inspecting the TLC profle gave about 120 fractions, each 50 mL.Fractions 31-40 were combined and purifed in a small column with an isocratic mode with 5% EtOAc/n-hexane and yielded compounds 1 (42 mg) and 2 (56 mg).Fractions 67-77 which fractionated with 20% n-hexane in EtOAc, further purifed using a small column with 0.5-2% MeOH in DCM and aforded compound 3 (50 mg).Fractions 19-26 which were collected with 25-30% ethyl acetate in hexane yielded compound 4 (32 mg).

Cytotoxicity Test.
Te antiproliferative efect of n-C 6 H 14 , CH 2 Cl 2 /CH 3 OH, and CH 3 OH extracts from roots of H. johannis was conducted against MCF-7 cell lines (human breast cancer) [14], through MTT assay [15,16].Te MCF-7 cells were acquired from NCCS, Pune, India, grown at 37 °C in DMEM containing 10% FBS and 1% antibiotics (100 U•mL −1 penicillin and 100 μg•mL −1 streptomycin) in a moisten air having 5% CO 2 .Confuent cells were separated using the trypsin-EDTA solution, and 5000 cells per well were used for subculture.At 50% confuence, the media was evacuated, and the cells exposed to 20 μg of the plant extract dissolved in DMSO (20 μL) and set at 37 °C for 24 h in a CO 2 incubator.Finally, samples were incubated with MTT (4 mg•mL −1 ) for 3 h.With the use of a typical microplate reader, absorbance was determined at 540 nm.Percentage cell viability was managed to express cytotoxicity.Te fndings are shown as the average of three replicated experiments.Te equation for calculating cell viability is as follows: Percent cell viability � absorbance of treated absorbance of control × 100%.
(1) Tese strains are among the drug-resistant pathogenic strains of most of the frst-line drugs [17], and any substance toxic to these strains is also accepted to be active on other strains [18].Experiments were carried out at the microbiology laboratory of Adama Science and Technology University.

Antibacterial Activity Test. Te studied compounds
(1-3) were tested for their bacterial growth inhibition.Discdifusion assay based on previously reported protocol was employed [19,20].Inoculums for antibacterial testing were prepared by transferring nearly 2-4 fresh strains with identical morphology into a saline solution.Te turbidity of each bacterial suspension was corrected according to a 0.5 McFarland solution (10 8 CFU/mL) [21].Fresh bacterial suspension of each strain was streaked uniformly onto separate Petri dish holding Mueller-Hinton agar (HiMedia) medium swapping with sterile cotton swap.Stock solutions (5 mg in 5 mL) of samples were made in 4% DMSO, followed by preparation of 0.5 and 0.25 mg/mL of samples (1-3) from corresponding stocks.Sterilized paper discs with 6 mm were 2 Biochemistry Research International prepared and loaded with the sample solution of 100 µL/disc, placed on the surface of the MHA plate, and incubated at 37 °C for 18-24 hr.Standard antibiotic disc of amoxicillin was considered as positive and DMSO as negative controls.Te area of microbial growth inhibition was computed by measuring the diameter (in mm) of the clear area surrounding the paper disc [20,22].Trails were carried out twice, and using SPSS (version 20) statistical software, the outcomes were analyzed and reported as mean ± SD.
2.6.Antioxidant Activity.Te oxidant trapping property of studied samples was performed based on the previously reported DPPH assay [23].Using serial dilution, samples of 200, 100, 50, and 25 g/mL were created, added to a newly made 0.04% DPPH solution in methanol, and incubated at 37 °C, and absorbance at 517 nm was recorded.In conditions similar to those of test samples, we set ascorbic acid as a positive control.Te radical trapping ability in percent for each substance was determined using the following formula [23]: where A control represents absorbance of the DPPH solution and A sample represents absorbance of the test sample.In addition, radical trapping of studied samples was given as IC 50.

Molecular Modelling.
We performed docking studies to understand the ligand-protein interaction at a molecular level and to reveal the cause of such interactions.Isolated compounds 2 and 3 were separately put into an active binding site of target proteins including PqsA (5OE5), DNA gyraseB (6F86), and S. aureus PK (3T07) [24,25].AutoDock Vina with previously reported protocol [24] applied to perform docking.ChemOfce program (Chem Draw 16.0) was employed to depict the structure of ligand molecules and assigned the correct 2D scheme.Te reduced energy of ligands was monitored by Chemdraw3D.Conformations with reduced energy level were sent into AutoDock Vina to undergo binding simulations.Receptor proteins (6F86, 5OE5, and 3T07) crystal structures were retrieved from the PDB.Using formerly reported standard technique [26], the proteins were designed by eliminating co-existing groups and certain H 2 O. Te selected enzyme fle was then produced by AutoDock version 4.2 (MGL tools 1.5.7)leaving related residue with protein.Initializing the molecule mainly includes the addition of hydrogen atoms and the addition of the Kolman charge to the protein molecule, while, for the ligand molecule, it also includes the addition of the Gasteiger charge, identifying aromatic carbons, and detecting rotatable bonds.In the docking process, a grid box was constructed using 20 × 20 × 20, pointing in x, y, and z directions, respectively, with a grid point spacing of 0.375 A °.  [27] managed to select the one with the minimum binding energy to discuss the interaction.
2.8.Drug-Likeness and ADMET Properties.Te concept of drug resemblance and ADMET properties were designed to disclose important guidelines for initial phases of drug research to enhance the likelihood that a compound will enter clinical trial stages [26].To estimate the drug-likeness and pharmacokinetic parameters, initially, the structure of pure compounds was transformed into their canonical SMILES, then subjected to PreADMET and SwissADME [28].Lipinski's rule [29] and Veber's flter [30] were used to estimate drug-likeness features.
Main ADME parameters [31], such as GIA, skin permeability (logKp) levels, brain access (BBB), permeability glycoprotein (P-gp) substrate/inhibitor, and cytochromes (CYP) inhibition, were computed.CYP P450 inhibition plays a major role in pharmacokinetics-related drug-drug interactions which can result in harmful or related efects attributed to reduced removal/accumulation of drugs or metabolites [28].BOILED-Egg plot was set to evaluate the human intestinal absorption, brain access, and P-gp score for the isolated entity [32].Te toxicity profle including the level of toxicity (LD50, mg/kg)/toxicity class, and other toxicities were predicted using ProTox-II server [33].

Results and Discussion
Purifcation of CH 2 Cl 2 /CH 3 OH (1 : 1) extract of the root of the plant under investigation by chromatography on silica gel resulted from four compounds (Figure 1).NMR spectroscopy and GC/MS data were analyzed by comparing those with the literature data to determine the structure of the studied compounds.
Te GC-MS analysis of fractions 19-26 (25-30% EtOAc in n-hexane as eluent) showed the presence of two peaks as shown in the chromatogram (Figure 2).By comparing the mass spectra (Figure 3) of the composition at the retention time of  which is acceptable as stated by the NCI criteria (estimated IC 50 less than 30 μg/mL) [40].Te cytotoxicity result showed that the evaluated extracts exhibited a promising cytotoxicity efect against the MCF-7 cell line, which support the traditional importance of H. johannis for breast cancer treatment.

Antibacterial Activity.
According to previous study result [41], the extract of dichloromethane/methanol was found to have better activity than n-hexane and methanol extracts towards P. aeruginosa, E. coli, and S. aureus.In this study, the abilities of compounds 1, 2, and 3 to stop bacterial growth, each at 0.25 and 0.50 mg/mL, were evaluated and reported in the zone of inhibition (ZI).Te mean zone inhibition (ZI in mm) results are displayed in Table 5.Samples were reported as mean ± SD (in mm); replicates (N) � 2; concentrations in (mg/mL).

Biochemistry Research International
According to the antibacterial study result, at 0.25 mg/ mL, compound 3 scored the largest ability to stop the growth of all tested strains (Table 5), scoring growth preventing diameter of 13.72 ± 0.05, 12.25 ± 0.1, and 11.62 ± 0.12 mm towards P. aeruginosa, E. coli, and S. pyogen, respectively.Relative to the standard (16.0 ± 0.0 mm, 15.25 ± 0.25 mm, and 16.25 ± 0.25 mm, respectively), this result was found to be good.Cafeic acid-2-hydroxynonylester (2) displayed the IZ value of 10.32 ± 0.12 mm, 9.19 ± 0.06 mm, 8.97 ± 0.22 mm, and 8.2 ± 0.3 mm against P. aeruginosa, E. coli, S. aureus, and S. pyogen, respectively.Oleic acid (1) displayed the moderate zone of inhibition of 7.89 ± 0.05 mm against P. aeruginosa.Te experimental result showed that catechin (3) displayed a considerable activity that goes in line with traditional uses of H. johannis to treat infectious diseases.

Antioxidant Activity.
Te free radical trapping ability of compounds 1, 2, and 3, and extract was evaluated using the DPPH method [23].As per the result, studied extracts and three compounds trap free radicals based on the administered amount (Table 6, Figure 4).At 12.5 μg/mL, the DCM/ MeOH extract exhibited the largest DPPH trapping ability (70.32%) with IC 50 of 1.44 μg/mL, being followed by the MeOH extract (67.44%).At the same lowest concentration, among the tested compounds (1-3), the largest radical trapping ability was shown by catechin (3) (65.61%) at IC 50 of 0.25 μg/mL, relative to the positive control (78.21%) at IC 50 of 0.014 μg/mL.Cafeic acid-2-hydroxynonylester (2) displayed considerable scavenging activity (48.9%), with an IC 50 value of 1.31 μg/mL, whereas oleic acid (1) displayed lower scavenging activity (43.64%) relative to standard (78.21%).Te ability of catechin (3) to trap free radicals suggests the use of this compound as a natural antioxidant (Table 6, Figure 4), supported by the drug-likeness in silico computation study of the compound with higher NHD and NHA values.

Molecular Docking Study.
In each docking pose, catechin (3) scored higher binding energy than compound 2. Te docking results are presented in Tables 7-9.Comparatively, the binding score displayed by catechin (3) towards DNA gyraseB (−7.3) and S. aureus PK (−6.7) was found to be  Biochemistry Research International  higher than that of the drug (−6.1 and −6.4), respectively (Tables 7-9).Te binding score displayed by catechin (3) is in line with its largest in vitro antibacterial activity, which in turn support the previously reported bacterial activity of catechin [32].Accordingly, Figures S4-S6 illustrate how residues interact to stabilize the ligandprotein interaction via H-bonding and other remaining interactions.H-bonding and the interactions are presented in the ball and stick model, and ligands are presented in various colors (Figures 5-7).Overall, the binding afnity of catechin (3) towards PqsA (−7.9 kcal/ mol), despite being lower than that of amoxicillin (−8.3 kcal/mol), is however consistent with its in vitro test against P. aeruginosa.
3.5.Drug-Likeness and ADMET Properties.SwissADME tool was used to calculate physicochemical properties to compute the in silico drug resemblances and ADME parameters to assess molecular pharmacokinetics [28].12 Biochemistry Research International predicted in each case.Te lowest skin permeant was scored by compound 2 (−7.82 cm/s).Compounds 1 and 2 were predicted to be P-gp nonsubstrate (PGP-), having high GIA.Compound 3 was found to be P-gp substrate.No inhibitory efect was predicted by compound 1 towards CYP1A2 and CYP2C9, and compound 3 towards CYP2C9 indicating these isoforms may not be involved in the biotransformation of the corresponding noninhibiting compounds.Compound 2 was found to be noninhibitor of all selected cytochromes.
According to the BOILED-Egg prediction result (Figure 8), all isolated compounds (1-3) fall in the white region indicating the probability of being passively absorbed by the GIT, but none of the isolated compounds permeate the brain (out of the yellow region).Compounds 1 and 2 were detected as being non-PGP− (red spot) and hence not actively pumped up to the gastrointestinal lumen, whereas compound 3 was described as PGP+ (blue dot) and hence actively pumped up from the gastrointestinal lumen.Comparatively, the standard drug amoxicillin, which is located outside white or yellow ellipse, has Biochemistry Research International low probability to permeate through the BBB to access the CNS and low absorption by the gastrointestinal tract.
3.8.Toxicity Prediction.Toxicity profles of studied compounds (1-3) were computed by using the ProTox-II platform [33], and the results are displayed in Table 11.Cafeic acid-2-hydroxynonyl ester (2) and catechin (3) each LD 50 greater than 5000 (class of toxicity less than fve) indicating reduced acute toxicity of each.According to the obtained result, the examined compounds are inactive to induce hepatotoxicity, and in turn unlike to interrupt the normal duty of the liver.Compounds were estimated to be noncarcinogenic, mutagenic, and cytotoxic.Compound 2

Figure 1 :
Figure 1: Compounds' structure obtained up on fractionation of the root of H. johannis.

Figure 3 :
Figure 3: Mass spectrum and structure of compound 4.
Te center grid box is of 29.09 × 22.86 × 30.06A °, 61.68 × 28.33 × 64.29 A °, and −44.36 × −25.35 × −11.00A °for 5OE5, 6F86, and 3T07, respectively.Docking genetic algorithms conditioned on AutoDock Vina and managed to fnd the best docked confguration of a ligand to a protein.Nine conformations for each ligand were taken into account, and PyMOL and Discovery Studio Visualizer