Exploring structure-activity relationship of S-substituted 2-mercaptoquinazolin-4(3H)-one including 4-ethylbenzenesulfonamides as human carbonic anhydrase inhibitors

Abstract Inhibitory action of newly synthesised 4-(2-(2-substituted-thio-4-oxoquinazolin-3(4H)-yl)ethyl)benzenesulfonamides compounds 2–13 against human carbonic anhydrase (CA, EC 4.2.1.1) (hCA) isoforms I, II, IX, and XII, was evaluated. hCA I was efficiently inhibited by compounds 2–13 with inhibition constants (KIs) ranging from 57.8–740.2 nM. Compounds 2, 3, 4, and 12 showed inhibitory action against hCA II with KIs between 6.4 and 14.2 nM. CA IX exhibited significant sensitivity to inhibition by derivatives 2–13 with KI values ranging from 7.1 to 93.6 nM. Compounds 2, 3, 4, 8, 9, and 12 also exerted potent inhibitory action against hCA XII (KIs ranging from 3.1 to 20.2 nM). Molecular docking studies for the most potent compounds 2 and 3 were conducted to exhibit the binding mode towards hCA isoforms as a promising step for SAR analyses which showed similar interaction with co-crystallized ligands. As such, a subset of these mercaptoquinazolin-4(3H)-one compounds represented interesting leads for developing new efficient and selective carbonic anhydrase inhibitors (CAIs) for the management of a variety of diseases including glaucoma, epilepsy, arthritis and cancer.


Chemistry
Melting points were recorded on a Barnstead 9100 electrothermal melting point apparatus (UK). IR spectra (KBr) were recorded on a FT-IR Perkin-Elmer spectrometer (Perkin Elmer Inc., Waltham, MA). Nuclear magnetic resonance ( 1 H and 13 C NMR) spectra were recorded with Bruker 500 or 700 MHz spectrometers (Zurich, Switzerland) using DMSO-d 6 as the solvent. Micro-analytical data (C, H, and N) were obtained using a Perkin-Elmer 240 analyser (Perkin Elmer Inc., MA) and agreed with the proposed structures within ±0.4% of the theoretical values. Mass spectra were recorded on a Varian TQ 320 GC/MS/MS mass spectrometer (Varian, Palo Alto, CA).

CA inhibition
The hCA I, II, IX, and XII isoenzyme inhibition assay was performed according to the reported method using SX.18 MV-R stopped-flow instrument (Applied Photophysics, Oxford, UK) [43][44][45] . All CA isoforms were recombinant isoforms obtained in-house, as reported earlier 46,47 .

Molecular docking method
Molecular docking was carried out according to the previously reported methods 24

Molecular docking studies
The docking simulations between the hCAs targets, and the most active compounds such as 2 and 3, as well as least active compound such as 6 compared with the prototype 1 lead compound were performed using MOE Suite 53 .

Docking of compounds 2, 3, and 6 with hCA isoenzymes
The most active methyl derivative 2 was docked with the binding pockets of the hCA isoforms II and XII, utilising the different protein crystal structures; 5ULN and 1JD0 downloaded from Protein Data Bank store 55,56 . As shown in Figure 3, and Table 2, the results suggested that the compound 2 displayed similar patterns to the co-crystallized ligands. Firstly, the docking pose of compound 2 in the complex with isoform II formed bidentate chelate with SO 2 NH 2 fragment with zinc metal at a distance range of 2.45-2.01 Å. The sulphonyl part was stabilised by three strong hydrogen bonds through the residues, Thr199 and Leu198 at average bond distance of 2.35-2.22 Å. Moreover, the carbonyl moiety of 4(3H)quinazolinone was stabilised by direct and indirect hydrogen bonding between water and polar residues Gln92 and Asn67 at distance of 2.25 Å as co-crystallized inhibitor S-atom does. In addition, hydrophobic interactions were experienced with methylthioether fragment through Phe131 residue. Secondly, the docking poses of compound 2 in the complex with isoform XII exhibited interactions like the co-crystallized inhibitor especially the SO 2 NH 2 part with Zinc metal and stabilising Thr199 and Thr200 residues. In addition, hydrophobic interactions were experienced with methylthioether fragment through Leu141 residue. The compound 3 was docked with the binding pockets in the hCA isoforms I and IX, utilising the different protein crystal structures; 4WR7 and 5FL4 downloaded from Protein Data Bank store 57,58 . As shown in Figure 4 and Table 2, the results suggested that the compound 3 interacted with both active sites in a similar fashion to the co-crystallized ligands in the pockets. Firstly, the docking pose of compound 3 in the complex with isoform I, consisted of a long, narrow tunnel, leading to a cavity that contained the catalytic Zn 2þ ion chelated with SO 2 NH 2 fragment (2.45-2.01 Å) and a sulphonyl part stabilised by two strong hydrogen bonds through the residues, Thr199 and His200 (2.38-2.41 Å). Moreover, different hydrophobic aromatic interactions were also formed with ethylthioether, phenethyl, and 4(3H)quinazolinone moieties through pockets of Leu131, Leu198, Pro202, Leu141, Trp209, Ala135, and Ala132 residues. Secondly, the docking pose of compound 3 in the complex with isoform IX, revealed that the cavity contained the catalytic Zn 2þ ion forming bidentate chelate with SO 2 NH 2 fragment (2.32-2.12 Å) and the sulphonyl part was stabilised by three hydrogen bonds through the residues, Thr200, His96 and His94 (2.44-1.99 Å). Different hydrophobic aromatic interactions were formed with ethylthioether, phenethyl, and 4(3H)quinazolinone moieties through pockets of Val130, Leu134, and Leu91 residues. The 4(3H)quinazolinone was modulated through molecules of H 2 O in the pocket by polar interactions. In addition, the aryl moiety of benzene sulphonamide formed a H 2 Omediated p-p interactions with certain aromatic amino acids. Thr201 might also play an important role in increasing their binding affinity for the enzyme. In addition, the lease active compound 6 was placed in the hCA I binding cavity ( Figure 5, right panel) and results showed that certain factors affecting the incorrect placement like the insertion of S-bromophenyl ring among polar Gln92 and Asn69 residues and disorienting of planer quinazolinone to Leu131 residue. Moreover, the docking of the least active compound 6 into the hCA-II pocket ( Figure 5, left panel) revealed the intolerance of S side chain bromophenyl moiety into the His119 polar part leading to protrusion out of the pocket and so appeared incompatible with pocket residue that makes it low active.
However, the lead compound 1 was docked into the pockets of hCA II and XII ( Figure 6, Table 2) for comparing its behaviour that showed the loss of SH role in the interactions compared to the potent active 2 and 3 derivatives. These overall docking findings proved that the S-alkylated derivatives exhibited good binding interactions better than the lead compound 1.