3.1 Synthesis of 2-(2-benzoyl-4-methylphenoxy)quinoline-3-carbaldehyde (5) and preparation of hypothetical compounds (A1-A50) as ligands
The compound 5; 2-(2-benzoyl-4-methylphenoxy)quinoline-3-carbaldehyde was obtained as shown in Scheme 1, and characterized using FTIR, HRMS and 1H-NMR (Figure 3a-c). We investigated drug-likeness for compound 5 which showed mild carcinogenicity (Table 1). In order to identify candidates with better drug-likeness, fifty hypothetical compounds were designed based on structural modifications with substituents in positions 4-, 5-, 6-, 7-, and 8- of the quinoline core, meta position of the methylphenoxy group and para position of the benzoyl group (Figure 4).
3.1.1 Structural elucidation of Compound (5)
White solid; reaction time: 12 h; yield: 60 %; m.pt: 119-121 °C; IR (neat) v max (cm-1) 3057, 2922, 2856, 2739, 1754, 1690, 1612, 1590, 1494, 1461, 1343, 1257, 1199, 1097, 760; 1H NMR (400 MHz, CDCl3) δ 9.75 (s, 1H), 8.32 (s, 1H), 7.97 (d, J = 8.3 Hz, 1H), 7.82 (d, J = 8.2 Hz, 1H), 7.80 (t, J = 8.2 Hz, 1H), 7.74 (t, J = 7.5 Hz, 1H), 7.61 (d, J = 8.2 Hz, 1H), 7.51 (d, 1H), 7.39 (t, 1H), 7.19 (s, 1H), 6.86 (d, 1H), 6.84 (d, 1H), 2.41 (s, 3H). HRMS (ESI): Calc. for [(C24H17NO3)](M+H)+ 368.1281, found 368.1283.
3.2 Toxicity results of compound (5), hypothetical compounds (A1-A50) and ten Antimalarial Reference drugs
All fifty hypothetical compounds (A1-A50) in Figure 4 and ten reference drugs; artesunate, doxycycline, tafenoquine, amodiaquine, artemeter, lumefantrine, primaquine piperaquine, mefloquine and chloroquine were virtually investigated for their toxicity profiles; hepatoxicity, carcinogenicity, immunogenicity, mutagenicity and cytotoxicity as given on Table 1a. Noticeably, while compound 5 and forty-one of the hypothetical derivatives failed one or more of those tests suggesting possible toxic and carcinogenic activities [22], however nine (9) lead compounds A5, A20, A31, A33, A34, A36, A45, A48 and A49 (Figure 5) were fully compliant (Table 1a and b).
Interestingly, results of toxicity for the reference drugs as presented in Table 2a-b returned only mefloquine as compliant, while others had one or more violations in comparison with the nine lead compounds. Mefloquine was therefore selected for further virtual study along with the selected lead compounds.
3.3 In-silico drug-likeness and ADME predictions
Our results in Table 3 shows that the lead compounds have hydrogen bond donors (nitrogen–hydrogen and oxygen–hydrogen bonds (0 – 1), and (<5), while the hydrogen bond acceptors (all nitrogen or oxygen atoms) (4 - 7) also (<10), are in compliance to the rule of five (RO5) [18]. Their molecular weights range (367.40 to 449.40 g/mol) which also fits the (150 to 500 g/mol) rule. The TPSA observed were (56.26 to 104 Å2) which falls within the range (20 to 130Å2) rule. Number of rotatable bonds were not more than 9.
Compounds A20, A45 and A49 were less permeant than others as revealed by their higher negative log Kp values (-5.13 to -5.18) [23]. Unlike others, compounds A31, A36 and A45 showed low gastrointestinal absorption values. This could be attributed to presence of the trifluoromethyl group at position 6 of compound A31, the thiol group at position 8 of compound A36, and both methoxy at position 6 and thiol at position 3 of the methylphenoxyl ring. Interestingly, compound 5 is blood brain barrier permeant unlike all the lead compounds. While compounds A31, A33, A34 and A36 are P-glycoprotein (P-gp) substrates [23, 24].
Six of the nine derivatives in addition to compound 5 exhibited high gastro-intestinal absorption except A31, A36 and A45. Furthermore, unlike all the other lead, compound 5 can penetrate the blood brain barrier (BBB) (Table 3). The nine lead compounds all have good oral bioavailability of 0.55, although with one allowable violation [18]. Inhibition of cytochromes P450 isoenzymes has been linked to the major cause of pharmacokinetics-related drug-drug interactions which could lead to toxic or adverse effects when there is lower clearance, accumulation of drug or metabolites. Table 4 reveals that the lead compounds are CYP2C19 inhibitors, while being substrates to CYP2D6. These cytochromes are involved in the metabolism and elimination of approximately 25% of clinically used drugs involved in the addition or removal of certain functionalities through; hydroxylation, demethylation and dealkylation [25].
3.4 Bioactivity score
The probability of drug leads as potential candidate can be evaluated using their bioactivity scores.
In Figure 6 all the lead compounds generally fall within highly or moderate bioactivity for all the parameters. Specifically, compounds A20 and A48 are highly active for five of the six parameters with bioactivity scores from 0.00 to 0.33. Compounds A5 and A31 have high bioactivity as kinase inhibitor (0.23 and 0.22) with capacity to block cancer cells [26], as nuclear receptor ligands (0.18 and 0.26) for hydrophobic molecules such as fatty acids, cholesterol and lipophilic hormones [27], as glycoprotein receptors GPCR (0.07 and 0.08) which regulate metabolic enzymes, or promoter proteins etc. [28], while A5 is also an enzyme inhibitor with value of 0.1 being able to bind to other available sites on the enzyme [29]. The compound with lowest bioactivity score is A36. All the lead compounds are only moderately active as protease inhibitor with values of (-0.18 to -0.28) implying having the capacity to prevent new HIV cells from becoming mature virus [30]. Only compounds A20 and A48 are highly active as ion channel modulators with values of (0.04 and 0.07) respectively which are above 0.00 [31].
3.5 Molecular docking study
Results from the docking of ligands and reference drugs against HAP is presented in Table 5. The binding energy for compound A31 (-11.3 kcal/mole) and compound A5 (-11.2 kcal/mole) are higher than compound 5 (-10.9 kcal/mol). Additionally, the next six of the lead compounds have binding energies of (-10.8 to -9.8 kcal/mole) all higher than the ten reference drugs investigated, for which the best performing mefloquine had -9.6 kcal/mole while the lowest was chloroquine with -6.0 kcal/mole.
The 3-D structure of P. falciparum histo-aspartic protease (PDB ID: 3QVC) and its hydrogen acceptor/ donor surface interaction with compound A31 is shown in Figure 7a. We observed in the 2-D view, Figure 7b, hydrogen binding interaction of the carbaldehyde oxygen atom of the quinoline core with Phe109 at bond length 3.51 Å. Also strong fluorine bonds of the trifluoromethyl groups at position 7 with Glu86, Arg91, Lys7 and Ala10 (bond lengths ranging from 2.98 to 3.58 Å) may presumably be responsible for the high inhibitory interactions of 2-(2-benzoyl-4-methylphenoxy)-8-methyl-6-(trifluoromethyl)quinoline-3-carbaldehyde (A31) with Plasmodium falciparum (HAP). Other bonds contributing to the interaction as presented in Figure 7b are alkyl, π-π alkyl, π-π stacked and π alkyl at various bond lengths with the residues of the protein.
Hydrophobic interactions (Figure 7c) are highly crucial for the folding of proteins especially in keeping the protein stable and biologically active through decrease in surface area thereby reducing the undesirable interactions with water [32]. Herein, compound A31 exhibits hydrophobic interactions with the P.falciparum HAP amino acid residues especially the 2-benzoyl-4-methylphenoxy side of the molecule interacting with Leu73, Ile80, Tyr112, Phe111, Trp39 and Ile107.
Interaction with the trifluoromethyl side of the quinoline molecule with Glu86, Arg91, Lys7 and Ala10 clearly appears to be hydrophilic due to the electronegative character of fluorine atom as shown in Figure 7c. Solvent accessibility (SAS) is a key feature of proteins for determining the folding and stability of a molecule [33]. In Figure 7c, the solvent accessibility surface, blue region in the 3D interaction is large thereby suggesting better interaction of compound A31 with the binding pocket of the HAP protein.
Similarly, the 3-D structure for compound A5 is shown in Figure 8a with binding energy of -11.2 kcal/mol. The absence of hydrogen bond does not reduce its efficacy as HAP inhibitor due to other interactions such as π-cation (pi electrons of the quinoline core and the amino hydrogen of the side chain of Lys7), alkyl and π-alkyl (ligand and amino acid residues such as Val120, Leu73, Tyr410, Leu73, Ile80, Ile107 and Pro110), π-π stacked and π-π T-shaped (compound A5 and Phe111, Phe109, His32 and Trp39) all contributed to its high binding energy (Figure 8b).
We observed in Figure 8c that compound A5 displays significant hydrophobic interaction with all observed binding residues except Lys7, to suggest good water-lipid interface transport between cell membrane of Pf HAP protein. Furthermore, the quinoline core of compound A5 is exposed to better solvent accessability to imply a more open conformation predisposed to easier interaction with the reactive sites of the target residues [34].
The binding interactions of the best reference drug Mefloquine with P. falciparum histo-aspartic protease residues (Figure 9a) showed that the trifluoro groups attached to position 6 in compound A31 and position 8 in Mefloquine both appear to contribute to their increased activity. However, unlike compound A31, Mefloquine does not have any hydrogen bonding which could have explained its lower binding energy.
Similarly, solvent accessibility surface interaction of compound A31 is higher for Mefloquine suggesting better interaction in the binding pocket of the HAP protein.