An innovative strategy for dual inhibitor design and its application in dual inhibition of human thymidylate synthase and dihydrofolate reductase enzymes

Abstract

Due to the diligence of inherent redundancy and robustness in many biological networks and pathways, multitarget inhibitors present a new prospect in the pharmaceutical industry for treatment of complex diseases. Nevertheless, to design multitarget inhibitors is concurrently a great challenge for medicinal chemists. We have developed a novel computational approach by integrating the affinity predictions from structure-based virtual screening with dual ligand-based pharmacophore to discover potential dual inhibitors of human Thymidylate synthase (hTS) and human dihydrofolate reductase (hDHFR). These are the key enzymes in folate metabolic pathway that is necessary for the biosynthesis of RNA, DNA, and protein. Their inhibition has found clinical utility as antitumor, antimicrobial, and antiprotozoal agents. A druglike database was utilized to perform dual-target docking studies. Hits identified through docking experiments were mapped over a dual pharmacophore which was developed from experimentally known dual inhibitors of hTS and hDHFR. Pharmacophore mapping procedure helped us in eliminating the compounds which do not possess basic chemical features necessary for dual inhibition. Finally, three structurally diverse hit compounds that showed key interactions at both active sites, mapped well upon the dual pharmacophore, and exhibited lowest binding energies were regarded as possible dual inhibitors of hTS and hDHFR. Furthermore, optimization studies were performed for final dual hit compound and eight optimized dual hits demonstrating excellent binding features at target systems were also regarded as possible dual inhibitors of hTS and hDHFR. In general, the strategy used in the current study could be a promising computational approach and may be generally applicable to other dual target drug designs.

Figure 1. Conversion of deoxyuridine monophosphate (dUMP) into deoxythymidine monophosphate (dTMP), and dihydrofolate (FH2) to tetrahydrofolate (FH4) by TS and DHFR, respectively.

Figure 2. Training set compounds used in common feature pharmacophore generation.

Figure 3. Strategy of dual inhibitor discovery employed in this study.

Figure 4. Generated pharmacophore model (Dual_Pharma) along with its interfeature distance (A), and its overlay on compound 1 of the training set (B).

Figure 5. Ligand-protein interaction diagram of hDHFR-inhibitor complex (compound 1 in the training set).

The pharmacophore mapping of the same compound is also illustrated. HBA, hydrogen bond acceptor; HBD, hydrogen bond donor; HY_AR, hydrophobic aromatic. The locations of amino acid residues are represented in rectangular boxes, where pink and green colors denote both the hydrogen bond acceptor/donor and nonpolar contacts, respectively.

Figure 6. Ligand-protein interaction diagram of hTS-inhibitor complex (compound 2 in the training set).

The pharmacophore mapping of the same compound is also illustrated. HBA, hydrogen bond acceptor; HBD, hydrogen bond donor; HY_AR, hydrophobic aromatic. The locations of amino acid residues are represented in rectangular boxes, where pink and green colors denote both the hydrogen bond acceptor/donor and nonpolar contacts, respectively.

Figure 7. Chemical structures of identified hits for dual inhibition and their overlay on pharmacophore model (Dual_Pharma).

Figure 8. Binding modes of (A) compound 2 from the training set (B) overlay of all three dual hit compounds in the active site of hTS.

(C) compound 1 from the training set, and (D) overlay of all three dual hit compounds in the active site of hDHFR. Key protein residues and ligands are represented by thick sticks. Hydrogen atoms have been removed for clarity.

Figure 9. Binding modes of all three dual hit compounds in the active sites of hTS and hDHFR.

Key protein residues and hit compounds are represented by thick sticks. Hydrogen atoms have been removed for clarity.

Figure 10. GOLD fitness scores, AutoDock binding energies, and SYLVIA synthetic accessibility scores of top 8 optimized hit compounds.