Phytomedicine Plus 3 (2023) 100447 Contents lists available at ScienceDirect Phytomedicine Plus journal homepage: www.sciencedirect.com/journal/phytomedicine-plus In silico screening of phytochemicals from Dissotis rotundifolia against Plasmodium falciparum Dihydrofolate Reductase Latif Adams a, Michael Afiadenyo b, Samuel Kojo Kwofie c,d, Michael D. Wilson b, Kwadow Asamoah Kusi e, Dorcas Obiri-Yeboah f, Siobhan Moane a, Michelle McKeon-Bennett a,* a Technological University of Shannon: Midlands Midwest, Midlands campus, Athlone, Ireland b Department of Parasitology, Noguchi Memorial Institute for Medical Research (NMIMR), College of Health Sciences (CHS), University of Ghana, Legon, Accra, Ghana c Department of Biomedical Engineering, School of Engineering Sciences, College of Basic and Applied Sciences, University of Ghana, Legon, Accra, Ghana d West African Center for Cell Biology and Infectious Pathogens, Department of Biochemistry, Cell and Molecular Biology, University of Ghana, Legon, Accra, Ghana e Department of Immunology, Noguchi Memorial Institute for Medical Research (NMIMR), College of Health Sciences (CHS), University of Ghana, Legon, Accra, Ghana f Department of Microbiology and Immunology, School of Medical Sciences, College of Health and Allied Sciences, University of Cape Coast, Cape Coast, Ghana A R T I C L E I N F O A B S T R A C T Keywords: Background: Malaria remains a major health concern in developing countries with high morbidity and mortality, Malaria especially in pregnant women and infants. A major obstacle to the treatment of malaria is a low effectiveness and Antifolate an increase resistance of the parasite to antimalarial drugs. As a result, there is an ongoing demand for new and Dissotis rotundifolia potent antimalarial drugs. Medicinal plants remain a potential source for the development of new antimalarial Phytochemical Plasmodium falciparum Dihydrofolate Reductase drugs. Amongst them is Dissotis rotundifolia is an ethnomedical important plant used in West Africa to treat (PfDHFR) malaria. Molecular docking Purpose: This study aimed at identifying new potential antifolates by virtually screening phytochemicals char- Molecular dynamics simulation acterized from the whole plant methanolic extract of D. rotundifolia against Plasmodium falciparum Dihydrofolate Reductase (PfDHFR). Methods: LC-ESI-Q-TOF-MS analysis was employed to identify the phytochemicals present in the whole plant methanolic extract of D. rotundifolia. These phytochemicals were docked against the catalytic site of PfDHFR. The docking protocol was evaluated using the Area Under the Curve (AUC) of a Receiver Operating Characteristic (ROC) curve. The binding mechanisms and the drug-likeness of the phytochemicals were characterized. A 100 ns Molecular Dynamics (MD) simulation and Molecular Mechanics-Poisson Boltzmann Surface Area (MM-PBSA) calculations were utilized to analyze the stability, the energy decomposition per residue and the binding free energy of the potential leads. Results: Twenty nine phytochemicals were characterized and docked against PfDHFR. Dimethylmatairesinol, flavodic acid, sakuranetin, and sesartemin were identified as potential leads with binding affinities of -8.4, -8.9, -8.6, and -8.9 kcal/mol respectively, greater than a stringent threshold of -8.0 kcal/mol. The potential leads also interacted hydrophobically with critical residue Phe58. A novel critical residue, Leu46 was identified to be crucial in the catalytic activity of PfDHFR. The potential leads were also predicted to be anti-protozoal with a probability of active (Pa) value ranging from 0.319 to 0.537. Conclusion: This study elucidates the potential inhibition of PfDHFR by dimethylmatairesinol, flavodic acid, sakuranetin and sesartemin present in D. rotundifolia. These compounds are druglike, do not violate Lipinski’s rule of five, have a high binding affinity to PfDHFR, and interact with crucial residues involved in the catalytic activity PfDHFR. Dimethylmatairesinol, flavodic acid, sakuranetin and sesartemin could therefore be further investigated and developed as new antifolate drugs for malaria. * Corresponding author at: Technological University of Shannon: Midlands Midwest, Midlands Campus, University Road, Athlone, Co. Westmeath, Ireland. E-mail address: michelle.mckeonbennett@tus.ie (M. McKeon-Bennett). https://doi.org/10.1016/j.phyplu.2023.100447 Available online 1 April 2023 2667-0313/© 2023 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). L. Adams et al. P h y t o m e d i c i n e P lus 3 (2023) 100447 co-factor for the conversion of dUMP to dTMP by providing the methyl Abbreviations group required for this reaction (Figure S1) (Hadni et al., 2021; Iwaloye et al., 2021). Over the years, antifolate drugs were used to inhibit DHFR ADMET Absorption, Distribution, Metabolism, Excretion and hence preventing the production of THF from DHF. This then stops the Toxicity production of the methyl group needed to be used to produce dTMP, dTMP Deoxythymidine monophosphate which is essential for DNA synthesis (Hadni and Elhallaoui, 2017; DHF Dihydrofolate Mharakurwa et al., 2011). Pyrimethamine and cycloguanil are anti- dUMP Deoxyuridine monophosphate folates that have been used as antimalarial drugs (Henriquez and Wil- ITNs Insecticide-Treated Bed Nets liams, 2020). LC-MS Liquid Chromatography Mass Spectrometry DHF interacts with residues Ile14, Ala16, Trp48, Asp54, Phe58, MM-PBSA Molecular Mechanics-Poisson Boltzmann Surface Area Asn108, Leu164 and Thr185 in order to be reduced to THF. Ala16, PASS Prediction of Activity Spectra of Substances Cys50, Ile51, Arg59, Asn108, and Leu164 have mutated leading to drug PfDHFR Plasmodium falciparum Dihydrofolate Reductase resistance (Hadni et al., 2021). All these mutated residues are situated in RMSD Root Mean Square Deviation the active site of PfDHFR. In general, the degree of resistance also in- RMSF Root Mean Square Fluctuation creases with the number of mutations (Henriquez and Williams, 2020). SMILES Simplified Molecular Input Line Entry System Greater degrees of resistance were found in parasites with the SP Sulfadoxine-Pyrimethamine Cys59Arg+Ser108Asn, Asn51Ile+Cys59Arg+Ser108Asn, Asn51I- THF Tetrahydrofolate le+Cys59Arg+Ser108Asn+Ile164Leu, and Cys59Arg+Ser108Asn+Ile WHO World Health Organization 164Leu mutations, compared to parasites with only S108N mutation (Hadni et al., 2021; Kamchonwongpaisan et al., 2004). Ala16- Val+Ser108Thr, a different double mutant, is only connected to cyclo- guanil resistance. Since additional mutations could theoretically occur Introduction to undermine the new antifolates, concerns about the likelihood of developing new antifolates with extended beneficial treatment lifetimes Malaria is one of the leading causes of morbidity and mortality in are greatly heightened by this mutation-based resistance. The parasites’ children and pregnant women (Belete, 2020). According to the World ability to mutate, however, is limited since they require a functioning Health Organization (WHO) Malaria Report, 241 million cases and 627, DHFR (Kamchonwongpaisan et al., 2004). As a result, it might be 000 deaths were recorded in 2020, with 80% of malaria fatalities conceivable to create inhibitors that are resistant to mutations and occurring in children under 5 years old in Africa. Out of 33.8 million would also cause the PfDHFR to stop functioning (Yuvaniyama et al., pregnancies recorded in 2020, 34% were exposed to malaria infection 2003). during pregnancy (WHO, 2021). Medicinal plants remain one of the potential sources for the devel- Malaria in pregnancy is associated with unfavorable outcomes such opment of new effective antimalarial drugs (Uzor et al., 2020). Two as maternal anemia, pre-term delivery, infant mortality, maternal death, antimalarial drugs quinine and artemisinin were extracted and isolated stillbirth, abortion, and low birth weight (El Gaaloul et al., 2022). from the bark of the Cinchona officinalis and Artemisia annua respectively Studies have reported that 20% of all stillbirths in Africa are caused by (Ceravolo et al., 2021). According to WHO, about 80% of low and malaria infection annually (Moore et al., 2017). According to WHO middle-level countries depend on medicinal plants for their primary report, malaria in pregnancy resulted in approximately 819,900 new- healthcare needs (G. Bhat, 2022). These plants are generally easily borns with low birthweight in Africa in 2020 (WHO, 2021). accessible with fewer side effects as compared to orthodox drugs WHO recommends the use of insecticide-treated bed nets (ITNs) and (Mohammadi et al., 2020) and serve as an alternative for the develop- preventive treatment using Sulfadoxine-Pyrimethamine (SP) to control ment of new therapeutic agents. and prevent malaria in pregnancy (Anto et al., 2019). SP is administered Dissotis rotundifolia (Sm) Triana is a medicinal plant that belongs to during the second and third trimesters of pregnancy (Al Khaja and the family Melastomataceae. It is commonly referred to as a ‘pink lady’ Sequeira, 2021). SP targets Plasmodium parasite dihydropteroate syn- and is widely distributed in tropical Africa. D. rotundifolia is used to treat thase and dihydrofolate reductase, by decreasing the amount of folic several ailments and diseases such as diarrhea, cough, dysentery, acid the malarial parasite needs to synthesize nucleic acids (Waller and conjunctivitis, bilharzia, peptic ulcer, stomachache, asthma, bronchitis, Sampson, 2018). Intermittent preventive treatment with SP is very tuberculosis, venereal diseases and malaria (Yeboah and Osafo, 2017). effective in decreasing the detrimental effects of malaria during preg- Phytochemical investigation indicates that the whole plant is rich in nancy (Desai et al., 2015). However, these drugs have developed resis- C-glycosylflavones, namely, vitexin, isovitexin, orientin, and isoorientin tance and become less effective (Iwaloye et al., 2021). The emergence of (Rath et al., 2014). Pharmacological studies of this plant have demon- these resistant strains poses a major challenge to malaria control and strated antibacterial (Abere et al., 2010), antiulcer (Adinortey et al., elimination, thus resulting in high morbidity and mortality of malaria 2018), antioxidant (Adinortey et al., 2018), anti-diarrhoeal (Abere et al., (Dhorda et al., 2021), prompting the urgent need for the development of 2010), anti-trypanosomal (Mann et al., 2009) and anti-plasmodial new and effective alternative antimalarial agents. properties (Djehoue et al., 2020). The decoction of the leaves of D. Over the years, antifolate malaria medications have targeted Plas- rotundifolia is ethnomedicinally used in west Africa to treat malaria modium falciparum dihydrofolate reductase (PfDHFR) (Mharakurwa (Lagnika et al., 2016). et al., 2011). Another major target of antimalarial medications is This study aimed at identifying new potential antifolates for malaria PfDHFR)-Thymidylate Synthase (PfDHFR-TS), which is responsible for by virtually screening 29 phytochemicals characterized from a whole DNA synthesis. DHFR is linked metabolically to TS, (DHFR-TS) plant methanolic crude extract of D. rotundifolia against PfDHFR using expressed as a bifunctional protein that plays a crucial role in the folate molecular docking analysis. The binding mechanism of the protein- pathway (Heinberg and Kirkman, 2015; Shamshad et al., 2022). ligand complexes were characterized. Molecular dynamics (MD) simu- DHFR-TS is made up of both the DHFR and the TS domains (Hyde, lations and MolecularMechanics-Poisson Boltzmann Surface Area (MM- 2005). The TS domain converts deoxyuridine monophosphate (dUMP) PBSA) calculations were also undertaken to provide insights into the to deoxythymidine monophosphate (dTMP), which undergoes further binding mechanisms of PfDHFR-ligand complexes. processes to result in DNA synthesis. The DHFR domain, on the other hand, converts dihydrofolate (DHF) to tetrahydrofolate (THF) which then leads to the production of N 5,10 – methylene-THF, which is a 2 L. Adams et al. P h y t o m e d i c i n e P lus 3 (2023) 100447 Materials and methods by GROningen Machine for Chemical Simulation (GROMACS) version 2018 (Abraham et al., 2015). The energy minimization was to ensure Liquid chromatography mass spectrometry (LC-MS) analysis stability during docking. The energy minimized protein and compounds were then subjected to molecular docking. To dock the 29 compounds Sample preparation against PfDHFR, AutoDock Vina v.1.2.0 (Trott and Olson, 2009) was The sample was prepared by measuring 1 mg of dried crude plant employed. Grid box dimensions of 30.1 Å, 32.9 Å and 39.4 Å centered at extract into 1 ml of methanol. The sample was further diluted in a 1:20 41.1 Å, 51.4 Å and 49.8 Å all in the x, y, and z coordinates respectively and filtered into an HPLC autosampler vial through a 0.22 µm PVDF utilized. membrane. Validation of docking protocol High-performance liquid chromatography coupled with electrospray ionization-quadrupole-time of flight-mass spectrometry Pyrimethamine, chloroquine, sulfadoxine, WR99210, and cyclo- guanil were employed as the five potent inhibitors whose Simplified LC-ESI-Q-TOF-MS- qualitative analysis Molecular Input Line Entry System (SMILES) were used to create their Phytochemical profiling of bioactive compounds in D. rotundifolia decoys using the Database of Useful Decoys; Enhanced (DUD;E) whole plant methanolic extract were determined using an Agilent 6520 (Mysinger et al., 2012). Along with those inhibitors, 350 decoys were quadrupole time-of-flight mass spectrometer linked with Agilent 1200 created and docked using AutoDock Vina against PfDHFR. The receiver HPLC system via dual ESI interface (Agilent Technologies, USA). A operating characteristic (ROC) curve was created from the docking data, previously developed method was used for the characterization of and the area under the curve (AUC) was calculated using the simple ROC phytochemicals with some modifications (Shukla et al., 2021). version 1.3.1 (Goksuluk et al., 2016). AUC and ROC were generated D. rotundifolia whole plant crude extract was separated using Agilent using a non-parametric approach of curve fitting (DeLong et al., 2016) Poroshell 120 EC C18 column; 50 mm × 3 mm, 2.7 μm. The eluents were for SE estimation and CI as well as a Type I error of 0.05. water with 0.1% formic acid (A) and acetonitrile (B). The gradient The ligand of the co-crystallized structure of PfDHFR was removed program was followed as 90% (B) for 25 min, then 90% (B) from 25 to and redocked. The co-crystallized ligand and the redocked ligand were 40 min, and then 10% (B) from 40 to 45 min, at a flow rate of 0.5 aligned using the LigAlign algorithm (Heifets and Lilien, 2010) imbed- ml/min. The sample injection volume was 6 μl. ded in PyMOl version 1.3 (Yuan et al., 2017). The Root Mean Square Agilent 6520 QTOF mass spectrometer with electrospray ionization Deviation (RMSD) of the two ligands will then be calculated. The in positive mode was used to conduct the qualitative analysis. As a co-crystallized complex and the redocked complex were then super- nebulizing, collision, and drying gas, nitrogen was utilized. Nebulizer imposed in LigPlot+ version 1.4.5 (Laskowski and Swindells, 2011) to pressure, drying gas flow rate, and capillary temperature were all set to identify the common residues the ligand interacted with in both com- 40 psi, 12 l/min, and 350 ◦C, respectively. Compounds were identified plexes. These approaches aid in accessing AutoDock Vina’s ability to using a mass range of m/z 50–1700 and a resolving power of at least predict the binding pose. 15,000 (FWHM). All of the Ion source’s parameters, including Vcap, the fragmentor, the skimmer, and the octapole radio frequency peak Absorption, distribution, metabolism, excretion, and toxicity prediction voltage, were set to, respectively, 3500 V, 150 V, 65 V, and 750 V. Mass Hunter software version 10 (Agilent Technology) was used for data SwissADME is a web tool that estimates the physicochemical de- analysis. scriptors, pharmacokinetics properties, and ADME parameters of com- pounds (Daina et al., 2017). The SMILES of the compounds were utilized Protein and compounds retrieval to evaluate the ADME parameters of the compounds. Lipinski’s rule, Verber rule, Egan rule, Ghose filter, and Muegge rule were considered The experimentally solved 3D structure of PfDHFR-TS was retrieved for the ADME parameters of the compounds. Toxicity profiling of the from the Protein Data Bank (PDB ID 1J3K) (Burley et al., 2021). The compounds was done with Osiris Datawarrior version 5.5.0 (Sander structure was solved using the X-ray diffraction method and has a res- et al., 2015). Mutagenicity, tumorigenicity, irritability, and reproduc- olution of 2.10 Å. The protein is made up of two chains each of DHFR ible effectiveness were the parameters considered. All compounds that and TS (DHFR, chains A and C, and TS, chains B and D). The DHFR is violated any of the ADMET protocols were eliminated. complexed with WR99210 and NADP while TS is complexed with dUMP (Yuvaniyama et al., 2003). An integrated library composed of two Characterization of binding mechanism antifolate malaria drugs, cycloguanil and pyrimethamine, a PfDHFR inhibitor (WR99201), and 29 compounds characterized from methanol To forecast the nature of interactions between the ligands and the extract of D. rotundifolia whole plant were curated for this study. protein, Ligplot+ version 1.4.5 (Laskowski and Swindells, 2011) was employed to generate 2D protein-ligand interactions. The top hit pose Binding sites identification was preserved in "pdb" file format and then rendered in PyMOL version 2.5.0. The complexes were then used as inputs for Ligplot+ Computed Atlas of Surface Topology of proteins (CASTp) (Tian et al., version.1.4.5. Green dashed lines represent hydrogen bonds and spoked 2018) was employed to evaluate the binding pocket used for this study. arcs extending toward the ligands represent hydrophobic interactions. The pocket was later analyzed in PyMOL version 2.5.0 (Yuan et al., The interaction profiles were created using default parameters. 2017). Biological activity prediction Protein and compounds preparation and virtual docking Prediction of Activity Spectra of Substances (PASS) was used to The spatial data files (SDF) of the compounds were retrieved from predict the activity of the potential leads. PASS is a tool for assessing an PubChem (Sunghwan et al., 2021). The universal force field was used to organic drug-like molecule’s overall bioactivity. Based on the structure minimize the energy of the compounds before they were later converted of organic substances, PASS makes simultaneous predictions of to Auto Dock files (pdbqt) using Open Babel (O’Boyle et al., 2011) in- numerous different forms of biological activity (Filimonov et al., 2014). tegrated into PyRx version 0.8 (Dallakyan and Olson, 2015). The energy The probability of active (Pa) and probability of inactive (Pi) values of the protein was minimized using the CHARM 27 all atoms force field were generated to determine the level of the activities of the compounds. 3 L. Adams et al. P h y t o m e d i c i n e P lus 3 (2023) 100447 An exhaustive search of the literature was also done to identify works in positive ionization mode is represented in Fig. 1. done on the potential leads, their analogs, and derivatives against malaria. Binding pocket identification Molecular dynamics simulation and molecular mechanics Poisson- A binding site is a part of a protein that a ligand binds to with Boltzmann surface area calculation specificity (Stank et al., 2016). The binding pocket used for this work had a surface area of 755.741 Å2 and a volume of 507.282 Å3. This Molecular dynamics (MD) simulations were performed using GRO- predicted pocket was chosen because it contained all the critical residues MACS version 5.1.4. The four potential leads’ complexes, the co- (Table S1) and is large enough to accommodate the ligands to fit through crystallized complex, the cycloguanil complex, and the unbound pro- (Figure S2). Also, previous studies have shown that the mutations at this tein were subjected to a 100 ns MD simulation using the CHARMM27 all- binding pocket are responsible for the resistance of PfDHFR against atoms force field. The ligand topology was generated using SwissParam drugs such as cycloguanil and pyrimethamine as they bind to this pocket (Zoete et al., 2011). Before the simulation, the complexes and the pro- (David et al., 2018; Hadni and Elhallaoui, 2017; Manhas et al., 2019). tein were prepared by initially being solvated in a 1 nm cubic water box This is the site where DHF binds to be reduced to THF (Ibraheem et al., and later neutralized by adding ions. The energy of the complexes and 2022; Yuvaniyama et al., 2003). In addition, this same pocket was used the protein was minimized for relaxation to remove steric clashes or bad in previous studies, where potential PfDHFR inhibitors were identified geometry. Equilibration of temperature (300 K) and density (1020 (Hadni et al., 2021; Iwaloye et al., 2021). kg/m3) was done for steps of 5000. The results of the simulation were then analyzed using root mean square deviation (RMSD), root mean Validation of docking protocol square fluctuation (RMSF), and radius of gyration (Rg). The graphs were then generated via Xmgrace version 5.1.25. A ROC curve was generated after docking five PfDHFR inhibitors and To determine the free binding energy, Molecular Mechanics-Poisson their decoys against the receptor to confirm AutoDock Vina’s capability Boltzmann Surface Area (MM-PBSA) calculations were performed using to differentiate between active and inactive molecules regarding the g_mmpbsa, a tool for high throughput MM-PBSA calculation (Kumari receptor (Fig. 2) (Enninful et al., 2022). To evaluate the docking per- et al., 2014). For every 1 ns step throughout the 100 ns MD, the free formance, the AUC value was calculated. AUC values between 0.5 and energies of the complexes were calculated. The binding free energy 0.7 are regarded as moderate, larger than 0.7 as acceptable, and less contributions of the residues of PfDHFR involved in the binding of each than 0.5 as having poor discrimination ability. The docking model for compound were also estimated via MM-PBSA calculations. The results the receptor has a great discriminating ability if the AUC value is very were plotted using the R programming package (R Core, 2020). near 1 (DeLong et al., 2016). The AutoDock Vina system demonstrated good discriminatory capacity to distinguish between active compounds Results and discussion and decoys with an AUC of 0.733 a type I error of 0.05 and a p-value of 0.01216745. LC-MS analysis The ligands of the co-crystallized structure and the redocked struc- ture were aligned and an RMSD of 0.959 Å was obtained. According to In this present study, the phytochemical profiling of bioactive com- (Alves et al., 2014), an RMSD less than or equal to 2.0 Å indicates a pounds from the methanol extract of D. rotundifolia whole plant were docking tool’s ability to predict the pose of a ligand in a binding pocket analyzed by LC-ESI-QTOF-MS. A total of 29 compounds belonging to (Fig. 3(A) and (B). The co-crystallized and the redocked complexes were different classes such as flavonoids, alkaloids, carboxylic acid, amino later superimposed in LigPlot+. Out of the 11 residues that interacted alcohol, quinoline, lignans, sesquiterpenoid, glycosides, terpenoids, with the ligand, 8 were common to both complexes. There were an lipid, hydroxycoumarin, phenols, azole, quinone, fatty acid, steroid, and overlapping of 2 hydrogen bonds with Asp54 and 5 hydrophobic in- quinoxaline were tentatively characterized from their mass data and MS teractions with Asn108, Phe58, Ile112, Cys54, and Pro113. Ile14 and spectra using Agilent LC-MS Qualitative Software (Mass Hunter) and Leu164 formed hydrophobic interactions with the ligand in the Personal Compound Database and Library (PCDL). Additionally, the co-crystallized structure but in the redocked structure, formed hydrogen following public databases; Pubchem (Sunghwan et al., 2021), Chem- bonds (Fig. 3(C). This corroborates the results from the ligand alignment spider (Pence and Williams, 2010), Phenol-Explorer (Rothwell et al., on AutoDock Vina’s capability to determine the binding pose of ligands. 2013), and Kegg Ligand Database (Kanehisa, 2002) were also used. The base peak chromatogram (BPC) of D. rotundifolia methanol crude extract Fig. 1. The base peak chromatogram (BPC) of Dissotis rotundifolia whole plant methanol extract in positive ionization mode. 4 L. Adams et al. P h y t o m e d i c i n e P lus 3 (2023) 100447 Fig. 2. The ROC curve generated via EasyROC to depict the capability of AutoDock Vina to differentiate between actives and inactives. Molecular docking studies fail as drug candidates (Enninful et al., 2022). Lipinski’s rule (Benet et al., 2016) determines whether a biologically active compound has the Molecular docking primarily studies how two or more molecular chemical and physical properties to be orally bioavailable (Benet et al., structures fit together and the possible interaction existing between the 2016). The number of hydrogen bond donors (≤ 5), hydrogen bond structures (Stanzione et al., 2021). Using a scoring system, docking is acceptors (≤ 10), water-octanol partition coefficient (log P ≤ 5), and utilized to assess the quality of the pose and predict the ligand shape, as molecular weight (≤ 500) are the parameters considered to determine well as its position and orientation within the protein binding site. The the bioavailability of a compound (Benet et al., 2016). Verber’s rule experimental binding mode should ideally be reproducible by the sam- (Veber et al., 2002) also predicts the bioavailability of drugs by pling process, and it should also be ranked top among all created poses considering the number of rotatable bonds (≤ 10) which is a measure of by the scoring function (Stanzione et al., 2021). The 29 phytochemicals molecular flexibility (Khanna and Ranganathan, 2009) and the Topo- and 3 inhibitors were docked against the predicted binding pocket of logical Polar Surface Area (TPSA ≤ 140 Å2) which is used to analyze the PfDHFR (Fig. 4). -7.0 kcal/mol is a good discriminating threshold for transport of a drug by multi resistance-associated proteins (Fernandes putative protein binders and non-binders (Chang et al., 2007). Inhibition and Gattass, 2009). Ghose filter (Ghose et al., 1999), Egan rule (Egan is not necessarily improved by a more negative binding energy (Li et al., et al., 2000), and Muegge rule (Muegge et al., 2001) also consider the 2014; Pantsar and Poso, 2018). Because approximately 97.7% of in- molecular weight, TPSA, the number of rotatable bonds, hydrogen hibitors have binding energies of less than -7.0 kcal/mol (Chang et al., bonds donors and acceptors, log P and number of atoms to determine the 2007), the threshold of -7.0 kcal/mol filters out approximately 95% of bioavailability of a drug (Table 2). Toxicity testing is a requisite during non-inhibitors (Ahmad et al., 2021). Furthermore, previous studies have drug discovery. Before clinical trials commence, the toxicity of a po- shown that triclosan, an inhibitor of PvDHFR (IC50 = 775 nM) and tential drug must be evaluated (Parasuraman, 2011). Out of the 15 PfDHFR (IC50 > 10 µM) has a binding affinity of -7.557 kcal/mol to compounds, 4 were suitable for further analysis (Table 3). PfDHFR (Bilsland et al., 2018). As such, a stringent threshold of -8.0 kcal/mol was employed. All compounds with binding energies greater than -8.0 kcal/mol were not considered for further analysis. The com- Characterization of protein-ligand interaction pounds identified herein had lower binding energies (higher binding affinity) to PfDHFR than cycloguanil and pyrimethamine, with binding The binding interactions of the hits were elucidated to evaluate the energies of -8.0 kcal/mol and -7.6 kcal/mol, respectively. Out of the important intermolecular bonds involved in the complexes. CID1286 original 29 compounds, 15 had binding energies less than or equal to formed a hydrogen bond with Ala16 with a bond length of 3.11 Å and -8.0 kcal/mol and hence, were considered for downstream analysis hydrophobic interactions with critical residues Phe58, Asp54, and Ile14. (Table 1). CID71944 formed 3 hydrogen bonds with three critical residues, Asp54, Thr185, and Ala16. CID73571 also formed 2 hydrogen bonds with Ala16. Both CID71944 and CID73571 formed hydrophobic interactions Absorption, distribution, metabolism, excretion, and toxicity (ADMET) with Asn108, Cys15, Ser111, Leu40, Val195, Phe58 and Leu46. evaluation CID342737 on the other hand did not form a hydrogen bond but formed hydrophobic interactions with 15 residues (Fig. 5). WR99210 interacted ADMET predictions were performed to remove compounds that may hydrophobically with residues Ile14, Ala16, Cys15, Phe58, Ile112, 5 L. Adams et al. P h y t o m e d i c i n e P lus 3 (2023) 100447 Fig. 3. Overlapping of the Ligands of the co-crystallized structure (green) and the redocked structure (blue) to determine the RMSD (A) and (B). The superimposition of the co-crystallized complex and the redocked complex (C). Leu119, Phe116, Pro113, Leu46, Leu164 and Asn108 while forming two hydrogen bonds with Asp54. Cycloguanil, on the other hand, formed a hydrogen bond with Asn108 and hydrophobic interactions with Phe58, Asp54, Ala16, Cys15, Leu164, Ile14, Tyr170, Leu40, and Leu46 (Figure S3). Interactions with residues Phe58 and Phe116 increase the anti-plasmodial action of naphyl derivatives carrying 1,2,3-triazole compounds significantly (Ibraheem et al., 2022). In addition, Phe58 is a critical residue involved in the catalytic activity of PfDHFR, hence an interaction with Phe58 increases the plausibility of inhibition (Bilsland et al., 2018). Strong binding is indicated by the many hydrogen bonds and the small bond lengths, which could make these ligands intriguing compounds to further study (Enninful et al., 2022). Fig. 4. CID71944 (blue) docked in the binding pocket (brown) of PfDHFR with the rest of the protein shown as ribbon (green). 6 L. Adams et al. P h y t o m e d i c i n e P lus 3 (2023) 100447 Table 1 taxifolin-3-O-α-L-rhamnoside, extracted from Bafodeya benna which has Molecular docking results of the 29 phytochemicals and 4 antifolate drugs added been shown to have antimalarial properties (Xu et al., 2011). Kaemp- as controls with their binding energies. Compounds with binding energy ≤ -8.0 ferol-3-O-rhamnoside, which is also an analog of CID1286 has been kcal/mol were selected for further analysis. proven to possess antimalarial activity (Barliana et al., 2014). CID71944 COMPOUND CID BINDING ENERGY (kcal/mol) has similar side chains as ethylenediamine-N,N’-diacetic acid, an 1. 101,274,424 -9.8 iron-chelating drug that was given to rats injected with P. berghei, and 2. 51,136,360 -9.7 antimalarial activity was observed. Ethylenediamine-N,N’-diacetic acid 3. 5,280,794 -9.7 was later tested on human red blood cells infected with P. falciparum 4. 3,052,765 -9.5 cultures and antiplasmodial activity was observed (Yinnon et al., 1989). 5. 5,280,804 -9.3 6. 121,750 (WR99210) -9.3 CID73571 which is structurally similar to a naringenin derivative iso- 7. 12,309,350 -9.2 vitexin-(I-3,II-3)-naringenin has been identified to possess anti- 8. 85,260,329 -9.2 plasmodial activity (Xu et al., 2011). CID342737 is a benzodioxole and 9. 162,350 -9 studies (Nelson and Hoosseintehrani, 1982) has shown that benzo- 10. 71,944 -8.9 dioxole natural compounds possess antimalarial properties. 11. 342,737 -8.9 12. 73,571 -8.6 13. 46,173,908 -8.4 Molecular dynamics simulations 14. 1286 -8.4 15. 5,281,614 -8.3 The protein-ligand complexes and the unbound protein were sub- 16. 1530 -8.1 17. 9049 (Cycloguanil) -8 jected to 100 ns MD simulations. RMSD, RMSF, and Rg analysis were 18. 5,281,857 -7.9 performed on the MD outputs of all the systems and the graphs were 19. 6,440,940 -7.8 generated. To evaluate the stability of the complexes, the RMSD graph 20. 17,109 -7.7 was analyzed. 21. 4993 (Pyrimethamine) -7.6 The RMSD is used to measure the displacement of the atoms from the 22. 100,067 -7.6 23. 590,929 -7.2 backbone (Raschka, 2017). The main application of RMSD is to compare 24. 5,280,567 -7.1 the differences between the structures that were present during the 25. 276,202 -6.8 simulation period and their reference structure. The RMSD trajectory 26. 3610 -6.7 displays the time-dependent difference between a protein structure and 27. 1923 -6.5 28. 464 -6.4 29. 122,121 -6.2 Table 3 30 8456 -6.2 The toxicity profiling results generated with OSIRIS DataWarrior. Compounds 31. 10,666 -5.9 predicted to be toxic in at least one of the parameters were excluded from the 32 69,421 -5.8 downstream analysis. 33. 37,511 -3.6 Compound Mutagenic Reproductive Tumorigenic Irritant CID effect Prediction of biological activity 101,274,424 None None High None 51,136,360 None None None None Biological activities with a probability of activity (Pa) > probability 5,280,794 None None None None of inactivity (Pi) and Pa > 0.3, are considered potential compounds for 3,052,765 None None High None 1530 High High High None those pharmacological activity investigations (Goel et al., 2010). In this 5,281,614 High None None None study, we considered the antiprotozoal activity of the compounds since 162,350 None None None None the malaria parasite belongs to the phylum protozoa (Dondorp and Von 12,309,350 None None None None Seidlein, 2017). All the compounds were predicted as antiprotozoal 46,173,908 None None None High activities with Pa ranging from 0.319 to 0.537 when Pa Pi. The com- 52,800,804 None None None None > 85,260,329 High None Low None pounds CID324737, CID73571, CID71944, and CID1286 were predicted 1286 None None None None as possessing anti-protozoal activities with Pa values of 0.352, 0.476, 71,944 None None None None 0.426, and 0.537 respectively (Table S4). 73,571 None None None None CID1286 is structurally similar to the natural compound (2R,3R)- 324,737 None None None None Table 2 ADME results generated by SwissADME showing the compounds that violated the various druglikenesss rules of the potential leads. Compound CID Number of violations GI absorption Lipinski’s rule of five Veber rule Egan rule Ghose rule Muegge rule 342,737 0 0 0 0 0 High 71,944 0 0 2 0 0 High 73,571 0 0 0 0 0 High 1286 0 0 0 0 0 High 1530 0 0 0 0 0 High 5,281,614 0 0 0 0 0 High 162,350 1 0 1 1 2 Low 12,309,350 2 0 1 1 3 Low 101,274,424 0 1 1 0 1 Low 51,136,360 2 1 1 0 3 Low 5,280,794 1 0 1 3 2 Low 3,052,765 0 0 0 0 0 High 46,173,908 3 2 1 4 5 Low 52,800,804 2 1 1 1 3 Low 85,260,329 2 1 1 1 4 Low 7 L. Adams et al. P h y t o m e d i c i n e P lus 3 (2023) 100447 Fig. 5. Protein-ligand interaction of (A) PfDHFR-CID 71,944, (B) PfDHFR-CID1286, (C) PfDHFR-CID73571, and (D) PfDHFR-CID342737 complexes. The green dash indicates hydrogen. its reference structure (Enninful et al., 2022). All the complexes and the average of ~0.23 nm. About 0.125 nm displacement was observed be- unbound protein RMSD rose from 0 ns to about 0.225 nm until stability tween the unbound protein and DHFR-WR99210 complex around 60 ns. was reached. The co-crystallized complexed (DHFR-WR99210) rose From 60 ns to the end of the simulation time, an average RMSD value of from 0 to 0.3 nm. From 10 to 20 ns, a high fluctuation was observed for ~0.23 nm was observed. The DHFR-cycloguanil complex was stable the DHFR-WR99210 complex. This depicts a higher displacement of the throughout the simulation time with an average RMSD of ~0.2 nm. This atoms from the backbone. Form 20 to 50 ns, stable fluctuations with an observation was not observed in the DHFR-WR99210 complex. This 8 L. Adams et al. P h y t o m e d i c i n e P lus 3 (2023) 100447 might be due to the WR99210 causing a serious clash with Asn108 and higher RMSF values indicate larger volatility (Dong et al., 2018). All (Yuvaniyama et al., 2003). Throughout the simulation time, the the complexes behaved similarly throughout the simulation time. DHFR-CID1286 complex was steady with an average of 0.25 nm similar Higher fluctuations were observed in DHFR-WR99210 around residue to that of the unbound protein but fluctuated around 14, 25, and 50 ns. 25 with an RMSF of 0.41 nm, DHFR-CID73571 around residue 45 with DHFR-CID1286 was the complex with RMSD closer to the an RMSF around 0.3 nm, DHFR-CID71944 from residues 80 to 100 with DHFR-WR99210. DHFR-CID71944 complex dropped from 0.225 nm to an RMSF around 0.25 nm,DHFR-CID1286 around residue 130 with an 0.175 nm around 8 ns then rose back to 0.225 nm around 20 ns. It then RMSF 0.45 nm and DHFR-CID342737 around residue 190 with an RMSF dropped back to 0.175 nm around 30 ns and stabilized throughout the of 0.15 nm. DHFR-cycloguanil and DHFR-CID71944 complexes fluctu- simulation time with an average RMSD of 0.2 nm while fluctuating ated with an RMSF of 0.35 nm at residue 231 (Fig. 8). around 60 ns. DHFR-CID73571 fluctuates until around 30 ns and then stabilized with an RMSD of ~0.225 nm. DHFR-CID342737 had the Molecular Mechanics- Poisson Boltzmann surface area (MM-PBSA) highest RMSD apart from the DHFR-WR99210 complex, with an average calculation of 0.225 nm until 40 ns then 0.25 nm from 40 ns to the end of the simulation time (Fig. 6). The estimations of the protein-ligand binding affinities produced by Rg is used to evaluate the compactness of proteins (Jiang et al., simulation-based techniques like MM-PBSA are more accurate than 2019). A protein that is folded consistently will probably keep Rg at a those produced by other computational techniques like docking. By very constant value (Lobanov et al., 2008). A protein’s Rg will evolve if examining the binding free energy of protein-ligand complexes, it is it unfolds (Enninful et al., 2022). DHFR-WR2210 complex was stabilized possible to determine which ligands have higher binding affinities to the with an average Rg of ~1.85 nm. DHFR-cycloguanil had an average Rg target. (Wang et al., 2017). of ~1.83 nm. In comparison to the co-crystallized and cycloguanil The binding free energies of the DHFR-CID324737 complex were the complexes, the potential leads complexes are stable indicating the lowest (-84.648 kJ/mol) and corresponded with the docking score of folding of the potential lead complexes. Rg of DHFR-CID71944 stabi- -8.9 kcal/mol. This indicates that the DHFR-CID324737 complex had lized throughout the simulation time. DHFR-CID1286 and the highest binding affinity irrespective of the fact that no hydrogen DHFR-CID73571 behaved similarly. From 80 ns, the Rg of both com- bond was formed. DHFR-CID1286, DHFR-CID71944, and DHFR- plexes dropped but that of DHFR-CID73571 averaged around 1.86 nm, CID73571 had free binding energies of -82.062, -51.988, and 60.157 and that of DHFR-CID1286 with an average of 1.82 nm. kJ/mol with docking scores of -8.4, -8.9 and -8.6 kcal/mol, respectively. DHFR-CID342737 stabilized till 20 ns, then fluctuated until 40 ns where All the complexes had van der Waals forces ranging from -152.495 to a rise was observed. Stability was observed until 70 ns where a sharp rise -114.345 kJ/mol. DHFR-WR99210 complex (redocked complex) had to 1.89 nm was observed, then dropped to 1.84 nm around 85 ns. From binding energy of -9.3 kcal/mol after docking but had a binding free 90 to 100 ns, a stable fluctuation was observed (Fig. 7). energy of -81.474 kJ/mol. DHFR-cycloguanil complexes had a binding To determine which DHFR residues are responsible for the structural energy of -8.0 kcal/mol and a free binding energy of -54.287 kJ/mol. variations, the RMSF trajectories of the DHFR-ligand complexes were This corroborates the docking results where all the potential leads had also assessed. RMSF is connected to crystallographic B-factors and is binding affinity greater than the DHFR-cycloguanil complex. This in- used to assess the flexibility of various protein regions (Sinha and Wang, dicates that the potential leads have a higher chance of inhibiting 2020). These protein domains are crucial in catalysis and ligand binding, PfDHFR. DHFR-WR99210 complex had a higher binding affinity than all Fig. 6. The RMSD plot of the unbound protein (black) and the DHFR-ligand complexes after the 100 ns MD simulation. RMSD plots of DHFR-WR99210, DHFR- Cycloguanil, DHFR-CID1286, DHFR-CID71944, and DHFR-CID73571 DHFR-CID342737, are shown in red, green, blue, yellow, brown and gray respectively. 9 L. Adams et al. P h y t o m e d i c i n e P lus 3 (2023) 100447 Fig. 7. Radius of Gyration plots of the unbound PfDHFR (black) and the PfDHFR-ligand complexes after the 100 ns MD. Rg plots of DHFR-WR99210, DHFR- Cycloguanil, DHFR-CID1286, DHFR-CID71944, and DHFR-CID73571 DHFR-CID342737, are shown in red, green, blue, yellow, brown and gray respectively. the potential leads but from the MM-PBSA calculations, CID1286 and violate Lipinski’s rule, Verber rule, Egan rule, Muegge rule, and Ghose CID342737 had higher binding affinities than thaWR99210. Since it has rule. All the potential leads also have high gastrointestinal absorption been shown that binding energy from simulations is more accurate than and were predicted not to be toxic. computational docking, CID1286 and CID342737 have higher binding These compounds have been predicted to be good binders of PfDHFR affinity to PfDHFR than WR99210. (Table 4). A higher binding affinity by having high binding affinities (-8.4, -8.9, -8.6, and -8.9 kcal/mol, indicates the plausibility of inhibition (Cunha et al., 2016). According to respectively) and interact with the critical residues involved in the ac- previous studies, the binding energy is mostly influenced by electrostatic tivity of PfDHFR in converting DHF to THF. In addition, MD simulation and van der Waal’s forces (Deng et al., 2011). was carried out on the protein-ligand complexes to elucidate the Energy decomposition per residue can be evaluated using MM-PBSA. conformational changes of the complexes and the unbound protein. MM- This entails the breakdown of each residue by taking into account the PBSA calculations identified novel critical residue Leu46, contributing a interactions that each residue is a part of (Gupta et al., 2018). Consid- lot of energy to the catalytic activity of PfDHFR. This depicts the stability ered crucial for protein-ligand binding are residues that contribute en- of the protein-ligand complexes. These compounds can therefore be ergies larger than or equal to 5 kJ/mol or less than or equal to -5 kJ/mol experimentally investigated by in vitro and in vivo techniques to (Dankwa et al., 2022). In the DHFR-WR99210 complex, Asp54 determine their potential efficacy against P. falciparum strains and thus contributed 16.63 kJ/mol, and Phe58 contributed -8.10 kJ/mol be developed into new antifolate drugs for malaria. (Figure S5). Asp54 contributed 10.67 kJ/mol and Phe58 contributed -4.92 kJ/mol in the DHFR-cycloguanil complex (Figure S6). Leu46 Funding contributed -7.32, -5.65, -4.76, and -6.87 kJ/mol in CID1286 (Fig. 9), CID71944, CID73571 and CID342737 complexes’ respectively. This This study was supported by the TUS President’s Doctoral Fellow- indicates that Leu46 may be a critical residue in the activity of PfDHFR. ship, grant number PA01034, the HEA [Higher Education Authority] From previous studies, Asp54 is a critical residue in the catalytic activity and DFHERIS [The Department of Further and Higher Education, of PfDHFR (Singh and Mishra, 2018), this is validated by the results Research, Innovation, and Science, Ireland]. obtained where among all the complexes, Asp54 contributed the highest energy. Availability of data and materials Conclusion All data generated and analyzed during this study are included in this manuscript and the supplementary data Out of 29 bioactive compounds screened against PfDHFR, four po- tential leads, dimethylmatairesinol, flavodic acid, sakuranetin, and sesartemin were identified as potential antifolates after molecular CRediT authorship contribution statement docking. ADMET analysis shows that all the potential leads do not Latif Adams: Conceptualization, Investigation, Methodology, 10 L. Adams et al. P h y t o m e d i c i n e P lus 3 (2023) 100447 Fig. 8. RMSF of the unbound PfDHFR protein (black) and the PfDHFR-ligand complexes. The RMSF plots of DHFR-WR99210, DHFR-Cycloguanil, DHFR-CID1286, DHFR-CID71944, and DHFR-CID73571 DHFR-CID342737, are shown in red, green, blue, yellow, brown and gray respectively. Fig. 9. Energy contribution of the PfDHFR residues involved in the binding in the PfDHFR-CID1286 complex. 11 L. Adams et al. P h y t o m e d i c i n e P lus 3 (2023) 100447 Table 4 Anto, F., Agongo, I.H., Asoala, V., Awini, E., Oduro, A.R., 2019. Intermittent preventive Binding free energies and the other contributing energies of the DHFR-ligand treatment of malaria in pregnancy: assessment of the sulfadoxine-pyrimethamine complexes from MM-PBSA computation Energy values are presented as three-dose policy on birth outcomes in rural northern Ghana. J. Trop. Med. 2019, 1–10. https://doi.org/10.1155/2019/6712685. average standard deviations in kJ/mol. Barliana, M.I., Suradji, E.W., Abdulah, R., Diantini, A., Hatabu, T., Nakajima-Shimada, J., Compound van der Electrostatic Polar SASA Binding Subarnas, A., Koyama, H., 2014. Antiplasmodial properties of kaempferol-3-O- Waals Energy (kJ/ Solvation (kJ/ Energy rhamnoside isolated from the leaves of Schima wallichii against chloroquine- resistant Plasmodium falciparum. Biomed. Reports 2, 579–583. https://doi.org/ Forces mol) Energy mol) (kJ/ 10.3892/br.2014.271. (kJ/mol) (kJ/mol) mol) Belete, T.M., 2020. Recent progress in the development of new antimalarial drugs with CID1286 -152.495 -19.914 ± 107.552 -17.205 -82.062 novel targets. Drug Des. Devel. Ther. 14, 3875–3889. https://doi.org/10.2147/ ± 59.518 14.769 ± 50.624 ± 6.846 ± 41.853 DDDT.S265602. CID342737 -155.890 -25.130 114.059 -17.687 -84.648 Benet, L.Z., Hosey, C.M., Ursu, O., Oprea, T.I., 2016. BDDCS, the Rule of 5 and ± 71.250 15.249 50.339 7.716 64.468 drugability. Adv. Drug Deliv. Rev. https://doi.org/10.1016/j.addr.2016.05.007. ± ± ± ± Bilsland, E., Van Vliet, L., Williams, K., Feltham, J., Carrasco, M.P., Fotoran, W.L., CID71944 -129.803 -53.621 ± 149.483 -18.058 -51.998 Cubillos, E.F.G., Wunderlich, G., Grøtli, M., Hollfelder, F., Jackson, V., King, R.D., ± 19.879 20.936 ± 31.000 ± 2.423 ± 16.282 Oliver, S.G., 2018. Plasmodium dihydrofolate reductase is a second enzyme target for CID73571 -114.345 -24.015 ± 90.103 ± -11.800 -60.157 the antimalarial action of triclosan. Sci. Rep. 8, 1–8. https://doi.org/10.1038/ ± 62.608 17.403 60.616 ± 6.444 ± 44.055 s41598-018-19549-x. WR99210 -144.231 -9.186 ± 88.628 ± -16.684 -81.474 Burley, S., Bhikadiya, C., Bi, C., Bittrich, S., Chen, L., Crichlow, G.V., Christie, C.H., ± 30.563 11.530 26.888 ± 3.242 ± 27.398 Dalenberg, K., Di Costanzo, L., Duarte, J.M., Dutta, S., Feng, Z., Ganesan, S., Cycloguanil -74.333 -11.750 ± 40.638 ± -8.842 -54.287 Goodsell, D.S., Ghosh, S., Green, R.K., Guranović, V., Guzenko, D., Hudson, B.P., ± 53.490 14.079 60.257 ± ± 38.308 Lawson, C.L., Liang, Y., Lowe, R., Namkoong, H., Peisach, E., Persikova, I., 60.174 Randle, C., Rose, A., Rose, Y., Sali, A., Segura, J., Sekharan, M., Shao, C., Tao, Y.-.P., Voigt, M., Westbrook, J.D., Young, J.Y., Zardecki, C., Zhuravleva, M., 2021. RCSB Protein Data Bank: powerful new tools for exploring 3D structures of biological Project administration, Writing – original draft. Michael Afiadenyo: macromolecules for basic and applied research and education in fundamental biology, biomedicine, biotechnology, bioengineering and energy sciences. Nucleic Data curation, Formal analysis, Writing – original draft. Samuel Kojo Acids Res. 49, D437–D451. https://doi.org/10.1093/nar/gkaa1038. Kwofie: Data curation, Formal analysis, Writing – original draft. Ceravolo, I.P., Aguiar, A.C., Adebayo, J.O., Krettli, A.U., 2021. Studies on activities and Michael D. Wilson: Data curation, Formal analysis. Kwadow Asamoah chemical characterization of medicinal plants in search for new antimalarials: a ten year review on ethnopharmacology. Front. Pharmacol. 12, 2478. https://doi.org/ Kusi: Investigation, Methodology, Writing – review & editing. Dorcas 10.3389/fphar.2021.734263/. Obiri-Yeboah: Supervision, Writing – review & editing. Siobhan Chang, M.W., Lindstrom, W., Olson, A.J., Belew, R.K., 2007. Analysis of HIV wild-type Moane: Investigation, Methodology, Supervision, Writing – review & and mutant structures via in silico docking against diverse ligand libraries. J. Chem. Inf. Model. 47, 1258–1262. https://doi.org/10.1021/ci700044s. editing. Michelle McKeon-Bennett: Conceptualization, Funding Cunha, N., Teixeira, G., Martins, T., Souza, A., Oliveira, P., Símaro, G., Rezende, K., acquisition, Project administration, Supervision, Writing – review & Gonçalves, N., Souza, D., Tavares, D., Silva, M., dos Santos, R., 2016. (− )-Hinokinin editing. induces G2/M arrest and contributes to the antiproliferative effects of doxorubicin in breast cancer cells. Planta Med. 82, 530–538. https://doi.org/10.1055/s-0042- 101761. Declaration of Competing Interest Daina, A., Michielin, O., Zoete, V., 2017. SwissADME: a free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules. Sci. Rep. 7, 1–13. https://doi.org/10.1038/srep42717. The authors declare that they have no known competing financial Dallakyan, S., Olson, A.J., 2015. Small-molecule library screening by docking with PyRx. interests or personal relationships that could have appeared to influence Methods in Molecular Biology (Clifton, N.J.), pp. 243–250. https://doi.org/ the work reported in this paper. 10.1007/978-1-4939-2269-7_19. Dankwa, B., Broni, E., Enninful, K.S., Kwofie, S.K., Wilson, M.D., 2022. Consensus docking and MM-PBSA computations identify putative furin protease inhibitors for Acknowledgments developing potential therapeutics against COVID-19. Struct. Chem. https://doi.org/ 10.1007/s11224-022-02056-1. The high-performance computing systems used for biocomputing David, T.I., Adelakun, N.S., Omotuyi, O.I., Metibemu, D.S., Ekun, O.E., Eniafe, G.O., Inyang, O.K., Adewumi, B., Enejoh, O.A., Owolabi, R.T., Oribamise, E.I., 2018. known as Zuputo is located at the WACCBIP, University of Ghana. Molecular docking analysis of phyto-constituents from Cannabis sativa with pfDHFR. Bioinformation 14, 574–579. https://doi.org/10.6026/97320630014574. Supplementary materials DeLong, E.R., Delong, D.M., Clarke-Pearson, D.L., 2016. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Int. Biometrics Soc. 227, 20. https://doi.org/10.3847/0067-0049/227/2/ Supplementary material associated with this article can be found, in 20. the online version, at doi:10.1016/j.phyplu.2023.100447. Deng, N., Zhang, P., Cieplak, P., Lai, L., 2011. Elucidating the energetics of entropically driven protein–ligand association: calculations of absolute binding free energy and entropy. J. Phys. Chem. B 115, 11902–11910. https://doi.org/10.1021/jp204047b. References Desai, M., Gutman, J., L’Lanziva, A., Otieno, K., Juma, E., Kariuki, S., Ouma, P., Were, V., Laserson, K., Katana, A., Williamson, J., Ter Kuile, F.O, 2015. Intermittent screening Abere, T.A., Okoto, P.E., Agoreyo, F.O., 2010. Antidiarrhoea and toxicological evaluation and treatment or intermittent preventive treatment with of the leaf extract of Dissotis rotundifolia triana (Melastomataceae). BMC dihydroartemisinin–piperaquine versus intermittent preventive treatment with Complement. Altern. Med. 10, 71. https://doi.org/10.1186/1472-6882-10-71. sulfadoxine–pyrimethamine for the control of malaria during pregnancy in western Abraham, M.J., Murtola, T., Schulz, R., Páll, S., Smith, J.C., Hess, B., Lindahl, E., 2015. Kenya: an open-lab. Lancet 386, 2507. https://doi.org/10.1016/S0140-6736(15) GROMACS: high performance molecular simulations through multi-level parallelism 00310-4. from laptops to supercomputers. SoftwareX 1 2, 19 25. https://doi.org/10.1016/j. Dhorda, M., Amaratunga, C., Dondorp, A.M., 2021. Artemisinin and multidrug-resistant – – softx.2015.06.001. Plasmodium falciparum - a threat for malaria control and elimination. Curr. Opin. Adinortey, M., Ansah, C., Adinortey, C., McGiboney, J., Nyarko, A., Inhibition, vitro H Infect. Dis. 34, 432–439. https://doi.org/10.1097/qco.0000000000000766. + /K -ATPase, 2018. Antiradical effects of a flavonoid-rich fraction of Dissotis Djehoue, R., Adamou, R., Madjid Amoussa, A.O., Medjigbodo, A.A., Laleye, A., Sanni, A., + rotundifolia, and In silico PASS Prediction of its isolated compounds. J. Nat. Sci. Biol. Djogbenou, L.S., Lagnika, L., Mohamed Elshafei Ali, A., 2020. In vitro assessment of Med. 9, 47. https://doi.org/10.4103/jnsbm.JNSBM_104_17. antiplasmodial activity and acute oral toxicity of Dissotis rotundifolia extracts and Ahmad, S., Waheed, Y., Abro, A., Abbasi, S.W., Ismail, S., 2021. Molecular screening of fractions on Plasmodium falciparum Strains. J. Appl. Life Sci. Int. 23, 8–19. https:// glycyrrhizin-based inhibitors against ACE2 host receptor of SARS-CoV-2. J. Mol. doi.org/10.9734/jalsi/2020/v23i530161. Model. 27, 206. https://doi.org/10.1007/s00894-021-04816-y. Dondorp, A.M., Von Seidlein, L., 2017. Treatment of uncomplicated vivax and ovale Al Khaja, K.A.J., Sequeira, R.P., 2021. Drug treatment and prevention of malaria in malaria, in: infectious Diseases, 2-Volume Set. Elsevier 1014–1025. https://doi.org/ pregnancy: a critical review of the guidelines. Malar. J. 20, 1–13. https://doi.org/ 10.1016/B978-0-7020-6285-8.00117-9 e1. 10.1186/S12936-020-03565-2/TABLES/3. Dong, Y.W., Liao, M.L., Meng, X.L., Somero, G.N., 2018. Structural flexibility and protein Alves, M.J., Froufe, H.J.C., Costa, A.F.T., Santos, A.F., Oliveira, L.G., Osório, S.R.M., adaptation to temperature: molecular dynamics analysis of malate dehydrogenases Abreu, R.M.V., Pintado, M., Ferreira, I.C.F.R., 2014. Docking studies in target of marine molluscs. Proc. Natl. Acad. Sci. USA. 115, 1274–1279. https://doi.org/ proteins involved in antibacterial action mechanisms: extending the knowledge on 10.1073/pnas.1718910115. standard antibiotics to antimicrobial mushroom compounds. Molecules 19, 1672–1684. https://doi.org/10.3390/molecules19021672. 12 L. Adams et al. P h y t o m e d i c i n e P lus 3 (2023) 100447 Egan, W.J., Merz, K.M., Baldwin, J.J., 2000. Prediction of drug absorption using Lobanov, M.Y., Bogatyreva, N.S., Galzitskaya, O.V., 2008. Radius of gyration as an multivariate statistics. J. Med. Chem. 43, 3867–3877. https://doi.org/10.1021/ indicator of protein structure compactness. Mol. Biol. 42, 623–628. https://doi.org/ jm000292e. 10.1134/S0026893308040195. El Gaaloul, M., Tornesi, B., Lebus, F., Reddy, D., Kaszubska, W., 2022. Re-orienting anti- Manhas, A., Lone, M.Y., Jha, P.C., 2019. Multicomplex-based pharmacophore modeling malarial drug development to better serve pregnant women. Malar. J. 21, 1–7. in conjunction with multi-target docking and molecular dynamics simulations for the https://doi.org/10.1186/S12936-022-04137-2/FIGURES/1. identification of PfDHFR inhibitors. J. Biomol. Struct. Dyn. 37, 4181–4199. https:// Enninful, K.S., Kwofie, S.K., Tetteh-Tsifoanya, M., Lamptey, A.N.L., Djameh, G., doi.org/10.1080/07391102.2018.1540362. Nyarko, S., Ghansah, A., Wilson, M.D., 2022. Targeting the Plasmodium falciparum’s Mann, A., Egwim, E.C., Banji, B., Abdukadir, N.-.U., Gbate, M., Ekanem, J.T., 2009. thymidylate monophosphate kinase for the identification of novel antimalarial Efficacy of Dissotis rotundifolia on Trypanosoma brucei infection in rats. African J. natural compounds. Front. Cell. Infect. Microbiol. 12, 1–13. https://doi.org/ Biochem. Res. 3, 005–008. https://doi.org/10.5897/AJBR.9000113. 10.3389/fcimb.2022.868529. Mharakurwa, S., Kumwenda, T., Mkulama, M.A.P., Musapa, M., Chishimba, S., Shiff, C. Fernandes, J., Gattass, C.R., 2009. Topological polar surface area defines substrate J., Sullivan, D.J., Thuma, P.E., Liu, K., Agre, P., 2011. Malaria antifolate resistance transport by multidrug resistance associated protein 1 (MRP1/ABCC1). J. Med. with contrasting Plasmodium falciparum dihydrofolate reductase (DHFR) Chem. 52, 1214–1218. https://doi.org/10.1021/jm801389m. polymorphisms in humans and Anopheles mosquitoes. Proc. Natl. Acad. Sci. USA. Filimonov, D.A., Lagunin, A.A., Gloriozova, T.A., Rudik, A.V., Druzhilovskii, D.S., 108, 18796–18801. https://doi.org/10.1073/pnas.1116162108. Pogodin, P.V., Poroikov, V.V., 2014. Prediction of the biological activity spectra of Mohammadi, S., Jafari, B., Asgharian, P., Martorell, M., Sharifi-Rad, J., 2020. Medicinal organic compounds using the pass online web resource. Chem. Heterocycl. Compd. plants used in the treatment of Malaria: a key emphasis to Artemisia, Cinchona, 50, 444–457. https://doi.org/10.1007/s10593-014-1496-1. Cryptolepis, and Tabebuia genera. Phyther. Res. 34, 1556–1569. https://doi.org/ Bhat, G., 2022. Medicinal plants and Its pharmacological values. Natural Medicinal 10.1002/PTR.6628. Plants. IntechOpen. https://doi.org/10.5772/intechopen.99848. Moore, K.A., Simpson, J.A., Scoullar, M.J.L., McGready, R., Fowkes, F.J.I., 2017. Ghose, A.K., Viswanadhan, V.N., Wendoloski, J.J., 1999. A knowledge-based approach in Quantification of the association between malaria in pregnancy and stillbirth: a designing combinatorial or medicinal chemistry libraries for drug discovery. 1. A systematic review and meta-analysis. Lancet Glob. Heal. 5, e1101–e1112. https:// qualitative and quantitative characterization of known drug databases. J. Comb. doi.org/10.1016/s2214-109x(17)30340-6/attachment/6b66cac6-e176-4981-87a1- Chem. 1, 55–68. https://doi.org/10.1021/cc9800071. 6d88dd931a97/mmc1. Goel, R.K., Singh, D., Lagunin, A., Poroikov, V., 2010. PASS-assisted exploration of new Muegge, I., Heald, S.L., Brittelli, D., 2001. Simple selection criteria for drug-like chemical therapeutic potential of natural products. Med. Chem. Res. 2010 209 (20), matter. J. Med. Chem. 44, 1841–1846. https://doi.org/10.1021/jm015507e. 1509–1514. https://doi.org/10.1007/S00044-010-9398-Y. Mysinger, M.M., Carchia, M., Irwin, J.J., Shoichet, B.K., 2012. Directory of useful decoys, Goksuluk, D., Korkmaz, S., Zararsiz, G., Karaagaoglu, A.E., 2016. EasyROC: an enhanced (DUD-E): better ligands and decoys for better benchmarking. J. Med. interactive web-tool for roc curve analysis using r language environment. R J. 8, Chem. 55, 6582–6594. https://doi.org/10.1021/jm300687e. 213–230. https://doi.org/10.32614/rj-2016-042. Nelson, F.R.S., Hoosseintehrani, B., 1982. Effects of benzylphenol and benzyl-1,3- Gupta, A., Chaudhary, N., Aparoy, P., 2018. MM-PBSA and per-residue decomposition benzodioxole derivatives on fertility and longevity of the yellow fever mosquito energy studies on 7-Phenyl-imidazoquinolin-4(5H)-one derivatives: identification of (Diptera: culicidae). J. Econ. Entomol. 75, 877–878. https://doi.org/10.1093/jee/ crucial site points at microsomal prostaglandin E synthase-1 (mPGES-1) active site. 75.5.877. Int. J. Biol. Macromol. 119, 352–359. https://doi.org/10.1016/j. O’Boyle, N.M., Banck, M., James, C.A., Morley, C., Vandermeersch, T., Hutchison, G.R., ijbiomac.2018.07.050. 2011. Open Babel: an open chemical toolbox. J. Cheminform. 3, 33. https://doi.org/ Hadni, H., Bakhouch, M., Elhallaoui, M., 2021. 3D-QSAR, molecular docking, DFT and 10.1186/1758-2946-3-33. ADMET studies on quinazoline derivatives to explore novel DHFR inhibitors. Pantsar, T., Poso, A., 2018. Binding affinity via docking: fact and Fiction. Molecules 23. J. Biomol. Struct. Dyn. 0, 1–15. https://doi.org/10.1080/07391102.2021.2004233. https://doi.org/10.3390/molecules23081899. Hadni, H., Elhallaoui, M., 2017. Molecular docking and QSAR studies for modeling the Parasuraman, S., 2011. Toxicological screening. J. Pharmacol. Pharmacother. 2, 74–79. antimalarial activity of hybrids 4-anilinoquinoline-triazines derivatives with the https://doi.org/10.4103/0976-500X.81895. wild-type and mutant receptor pf-DHFR. https://doi.org/10.1016/j.heliyon.2019. Pence, H.E., Williams, A., 2010. ChemSpider: an online chemical information resou rce. e02357. J. Chem. Educ. 87, 1123–1124. https://doi.org/10.1021/ed100697w. Heifets, A., Lilien, R.H., 2010. LigAlign: flexible ligand-based active site alignment and R Core, 2020. R: a language and environment for statistical computing. R Foundation for analysis. J. Mol. Graph. Model. 29, 93–101. https://doi.org/10.1016/j. Statistical Computing, Vienna, Austria. jmgm.2010.05.005. Raschka, S., 2017. Molecular docking, estimating free energies of binding, and AutoDock Heinberg, A., Kirkman, L., 2015. The molecular basis of antifolate resistance in ’ s semi-empirical force field Introduction molecular docking 1–19. Plasmodium falciparum: looking beyond point mutations. Ann. N. Y. Acad. Sci. 1342, Rath, G., Touré, A., Nianga, M., Wolfender, J.L., Hostettmann, K., 2014. Characterization 10–18. https://doi.org/10.1111/nyas.12662. of C-glycosylflavones from Dissotis rotundifolia by liquid chromatography — UV Henriquez, F.L., Williams, D.R.A.M., 2020. Knowing one’s enemy: the Plasmodium diode array detection — Tandem mass spectrometry. Chromatogr 415 (41), parasite. Antimalar. Agents 49–64. https://doi.org/10.1016/B978-0-08-101210- 332–342. https://doi.org/10.1007/BF02688048, 1995. 9.00002-0. Rothwell, J.A., Perez-Jimenez, J., Neveu, V., Medina-Remon, A., M’Hiri, N., Garcia- Hyde, J.E., 2005. Exploring the folate pathway in Plasmodium falciparum. Acta Trop. Lobato, P., Manach, C., Knox, C., Eisner, R., Wishart, D.S., Scalbert, A., 2013. Phenol- https://doi.org/10.1016/j.actatropica.2005.04.002. Explorer 3.0: a major update of the phenol-explorer database to incorporate data on Ibraheem, W., Makki, A.A., Alzain, A.A., 2022. Phthalide derivatives as dihydrofolate the effects of food processing on polyphenol content. Database 2013, reductase inhibitors for malaria: molecular docking and molecular dynamics studies. bat070–bat070. https://doi.org/10.1093/database/bat070. J. Biomol. Struct. Dyn. 0, 1–11. https://doi.org/10.1080/07391102.2022.2080114. Sander, T., Freyss, J., Von Korff, M., Rufener, C., 2015. DataWarrior: an open-source Iwaloye, O., Elekofehinti, O.O., Kikiowo, B., Fadipe, T.M., Akinjiyan, M.O., Ariyo, E.O., program for chemistry aware data visualization and analysis. J. Chem. Inf. Model. Aiyeku, O.O., Adewumi, N.A., 2021. Discovery of traditional chinese medicine 55, 460–473. https://doi.org/10.1021/ci500588j. derived compounds as wild type and mutant Plasmodium falciparum Dihydrofolate Shamshad, H., Bakri, R., Mirza, A.Z., 2022. Dihydrofolate reductase, thymidylate Reductase Inhibitors: induced Fit Docking and ADME Studies. Curr. Drug Discov. synthase, and serine hydroxy methyltransferase: successful targets against some Technol. 18, 554–569. https://doi.org/10.2174/1570163817999200729122753. infectious diseases. Mol. Biol. Reports 497 (49), 6659–6691. https://doi.org/ Jiang, Z., You, L., Dou, W., Sun, T., Xu, P., 2019. Effects of an electric field on the 10.1007/S11033-022-07266-8, 2022. conformational transition of the protein: a molecular dynamics simulation study. Shukla, V., Singh, P., kumar, D., Konwar, R., Singh, B., Kumar, B., 2021. Phytochemical Polymers (Basel) 11, 1–13. https://doi.org/10.3390/polym11020282. analysis of high value medicinal plant Valeriana jatamansi using LC-MS and it’s in- Kamchonwongpaisan, S., Quarrell, R., Charoensetakul, N., Ponsinet, R., Vilaivan, T., vitro anti-proliferative screening. Phytomedicine Plus 1, 100025. https://doi.org/ Vanichtanankul, J., Tarnchompoo, B., Sirawaraporn, W., Lowe, G., Yuthavong, Y., 10.1016/J.PHYPLU.2021.100025. 2004. Inhibitors of multiple mutants of Plasmodium falciparum dihydrofolate Singh, I.V., Mishra, S., 2018. Molecular docking analysis of pyrimethamine derivatives reductase and their antimalarial activities. J. Med. Chem. 47, 673–680. https://doi. with Plasmodium falciparum dihydrofolate reductase. Bioinformation 14, 232–235. org/10.1021/jm030165t. https://doi.org/10.6026/97320630014232. Kanehisa, M., 2002. The KEGG database. Novartis Found. Symp. 247, 91–101 discussion Sinha, S., Wang, S.M., 2020. Classification of VUS and unclassified variants in BRCA1 101-3, 119–28, 244–52. BRCT repeats by molecular dynamics simulation. Comput. Struct. Biotechnol. J. 18, Khanna, V., Ranganathan, S., 2009. Physiochemical property space distribution among 723–736. https://doi.org/10.1016/j.csbj.2020.03.013. human metabolites, drugs and toxins. BMC Bioinformatics 10 (Suppl 1), S10. Stank, A., Kokh, D.B., Fuller, J.C., Wade, R.C., 2016. Protein binding pocket dynamics. https://doi.org/10.1186/1471-2105-10-S15-S10. Acc. Chem. Res. 49, 809–815. https://doi.org/10.1021/acs.accounts.5b00516. Kumari, R., Kumar, R., Lynn, A., 2014. g_mmpbsa —A GROMACS Tool for high- Stanzione, F., Giangreco, I., Cole, J.C., 2021. Use of Molecular Docking Computational throughput MM-PBSA calculations. J. Chem. Inf. Model. 54, 1951–1962. https://doi. Tools in Drug discovery, 1st ed, Progress in Medicinal Chemistry. Elsevier B.V. org/10.1021/ci500020m. https://doi.org/10.1016/bs.pmch.2021.01.004. Lagnika, L., Djehoue, R., Yedomonhan, H., Sanni, A., 2016. Ethnobotanical survey of Sunghwan, K., Chen, J., Cheng, T., Gindulyte, A., He, J., He, S., Li, Q., Shoemaker, B.A., medicinal plants used in malaria management in South Benin. J. Med. Plants Res. 10, Thiessen, P.A., Yu, B., Zaslavsky, L., Zhang, J., Bolton, E.E., 2021. PubChem in 2021: 748–756. https://doi.org/10.5897/JMPR2016.6219. new data content and improved web interfaces. Nucleic Acids Res 49, Laskowski, R.A., Swindells, M.B., 2011. LigPlot+: multiple ligand–protein interaction D1388–D1395. https://doi.org/10.1093/nar/gkaa971. diagrams for drug discovery. J. Chem. Inf. Model. 51, 2778–2786. https://doi.org/ Tian, W., Chen, C., Lei, X., Zhao, J., Liang, J., 2018. CASTp 3.0: computed atlas of surface 10.1021/ci200227u. topography of proteins. Nucleic Acids Res 46, W363–W367. https://doi.org/ Li, Y., Han, L., Liu, Z., Wang, R., 2014. Comparative assessment of scoring functions on 10.1093/nar/gky473. an updated benchmark: 2. Evaluation methods and general results. J. Chem. Inf. Model. 54, 1717–1736. https://doi.org/10.1021/ci500081m. 13 L. Adams et al. P h y t o m e d i c i n e P lus 3 (2023) 100447 Trott, O., Olson, A.J., 2009. AutoDock Vina: improving the speed and accuracy of characterization by LC-SPE-NMR of some medicinal plant extracts used against docking with a new scoring function, efficient optimization, and multithreading. malaria. Planta Med 77, 1139–1148. https://doi.org/10.1055/s-0030-1270719. J. Comput. Chem. 31 https://doi.org/10.1002/jcc.21334. NA-NA. Yeboah, O., Osafo, N., 2017. Review of the ethno-medical, phytochemical, Uzor, P.F., Prasasty, V.D., Agubata, C.O., 2020. Natural products as sources of pharmacological and toxicological studies on Dissotis rotundifolia (Sm.) Triana. antimalarial drugs. evidence-based complement. Altern. Med. https://doi.org/ J. Complement. Altern. Med. Res. 2, 1–11. https://doi.org/10.9734/JOCAMR/ 10.1155/2020/9385125, 2020. 2017/32212. Veber, D.F., Johnson, S.R., Cheng, H.-.Y., Smith, B.R., Ward, K.W., Kopple, K.D., 2002. Yinnon, A., Theanacho, E., Grady, R., Spira, D., Hershko, C., 1989. Antimalarial effect of Molecular properties that influence the oral bioavailability of drug candidates. HBED and other phenolic and catecholic iron chelators. Blood 74, 2166–2171. J. Med. Chem. 45, 2615–2623. https://doi.org/10.1021/jm020017n. https://doi.org/10.1182/blood.V74.6.2166.2166. Waller, D.G., Sampson, A.P., 2018. Chemotherapy of infections. Med. Pharmacol. Ther. Yuan, S., Chan, H.C.S., Hu, Z., 2017. Using PyMOL as a platform for computational drug 581–629. https://doi.org/10.1016/B978-0-7020-7167-6.00051-8. design. Wiley Interdiscip. Rev. Comput. Mol. Sci. 7 https://doi.org/10.1002/ Wang, C., Nguyen, P.H., Pham, K., Huynh, D., Le, T.N., Wang, H., Ren, P., Luo, R., 2017. wcms.1298. Calculating protein-ligand binding affinities with MMPBSA: method and error Yuvaniyama, J., Chitnumsub, P., Kamchonwongpaisan, S., Vanichtanankul, J., analysis 37, 2436–2446. https://doi.org/10.1002/jcc.24467. Sirawaraporn, W., Taylor, P., Walkinshaw, M.D., Yuthavong, Y., 2003. Insights into WHO, 2021. Word Malaria Report 2021, Word malaria report Geneva: world Health antifolate resistance from malarial DHFR-TS structures. Nat. Struct. Biol. 10, Organization. (2021). 357–365. https://doi.org/10.1038/nsb921. Xu, Y., Capistrano, R., Dhooghe, L., Foubert, K., Lemière, F., Maregesi, S., Baldé, A., Zoete, V., Cuendet, M.A., Grosdidier, A., Michielin, O., 2011. SwissParam: a fast force Apers, S., Pieters, L., 2011. Herbal medicines and infectious diseases: field generation tool for small organic molecules. J. Comput. Chem. 32, 2359–2368. https://doi.org/10.1002/jcc.21816. 14