Theoretical methods in drug design. The Cambridge Structural Database (CSD) contains more than a quarter of a million small-molecule crystal structures, most of them having biological activity (Allen, 2002). The implications of the CSD in life sciences, especially using theoretical methods related to drug design, have extensively been reported in a recent review (Taylor, 2002). Another popular source of structural information is the Protein Data Bank (PDB) (Berman et al., 2000), that includes more than 45,000 modes of proteins, nucleic acids and other polymers. Even if the amount of data is much lower than the CSD, the PDB has the great advantage in that it also implements co-crystallized structures in the presence of ligands and drugs. In these cases, the information is particularly interesting because the three-dimensional structures of the small co-crystallized compounds can be considered as the bioactive conformations. Moreover, the PDB also includes ligand-receptor complexes obtained by nuclear magnetic resonance (NMR) methods, that in few, but quite interesting cases, provide information about the dynamic equilibrium among bioactive conformations. When drug development is carried out using as a starting structure an experimentally determined complex structure, such as those deposited into the PDB, the term ‘structure-based’ drug design is adopted (Kroemer, 2007). This term is opposed to a more classical paradigm, still in use especially when no structural information is available for the receptor-ligand complex, and known as ‘ligand-based’ drug design (Bacilieri and Moro, 2006). These two terms differ in their relationships with the diseases and are respectively related to the target (protein, receptors or genes) and to reference compounds (known drugs or ligands). The anticancer drug discovery procedure undergoes the same paradigms applied to other drug design projects. The differences between the structure-based and the ligand-based approaches can be summarized in terms of different links with respect to cancer disease (Figure 3.2). It is worth noting that despite the fact that knowledge of the target macromolecular structure is often available, ligand-based drug design represents still an important approach for the discovery of new anticancer agents. Moreover, the opportunity to follow both paradigms in the same drug discovery process usually brings a synergy to the entire development, because the structure-based and the ligand-based approaches are totally different theoretical methodologies. In order to give an idea of the large application of the computational methods available to rational anticancer drug design, we have selected some relevant examples for both paradigms. Target macromolecules (DNA, Enzymes, Proteins) Reference anticancer drugs (Natural or synthetic agents) Structure-based drug design % Fs Ligand-based drug design Figure 3.2 The relationships in cancer disease with target macromolecules and anticancer drugs as function of the drug design paradigm A categorization of them has been applied to those pertinent to the structure-based approach, taking into account the nature of the target involved in the drug discovery process. No categorization has been considered for the ligand-based approach, because the target is not explicitly considered in the computational models. One of the most interesting examples of ligand-based study is related to the inhibitors of the aromatase enzyme. This belongs to the cytochrome P-450 superfamily, whose function is to catalyse the aromatization of androgens, the final step in the biosynthesis of oestrogens (Figure 3.3). Therefore, it is an important factor in sexual development and its inhibition, with the consequent arrest of oestrogen biosynthesis, represents a good therapeutic route for the treatment of oestrogen-dependent breast cancers. Many research groups have identified some potent inhibitors of aromatase that can be collected into two main classes: steroidal analogues of androst-4-ene-3,17-dione, the natural substrate of aromatase, such as compound 1, and the non-steroidal inhibitors, such as compound 2 (Figure 3.4). In absence of any direct information on the tertiary structure of aromatase some molecular modelling studies were carried out with the aim of mapping the active site of the enzyme, performing conformational analysis of the chemical structures of known aromatase inhibitors CHOLESTEROL t TESTOSTERONE AROMATASE cD —-— ESTROGEN estrogen receptor Sx) estrogen receptor BREAST UTERUS AROMATASE INHIBITORS ~—amee> PROLIFERATION emma CANCER Figure 3.3 Schematic representation of oestrogen biosynthesis catalysed by the aromatase with the consequent proliferation in breast and uterus cells 9 § N. N Mt Or 1 2 Figure 3.4 Chemical structures of the steroid (19R)-10-thiiranylestr-4-ene-3,17-dione (1) and the non-steroid compound CGS 16949A (2) Another successful case of ligand-based drug discovery is represented by the poly (ADP-ribose) polymerase-1 (PARP-1) inhibitors. PARP-1 is a nuclear enzyme involved in several cellular processes, such as DNA repair and apoptosis (de Murcia et al., 1994; Dantzer et al., 1999). The enzyme initially recognizes and binds to DNA damage sites and catalyses the synthesis of linear and branched ADP-ribose polymers from the substrate nicotinamide adenine dinucleotide (NAD+), with concomitant release of nicotinamide. Thus, PARP-1 inhibitors can be useful in the repair of DNA damage induced by radiation during cancer therapy. Early studies identified various analogues of 3-aminobenzamide, such as compound 4 (Figure 3.6) as inhibitors of PARP-1; however, these compounds lack potency and specificity, and some are poorly water soluble. Subsequent design of constrained analogues locking the carboxamide group in its probable bioactive conformation (compounds 5, 6 and 7, Figure 3.6) led to compounds with poor potency; however, compounds 6 and 7 displayed the desired biological effect at high concentrations. Since compound 7 can form an intramolecular hydrogen bond between the carboxamide and the nitrogen of the heterocycle ring (Figure 3.6), a restricted scaffold such as compound 8 (Figure 3.7) has been suggested. Derivatives based on compound 8 were confirmed to act as resistance-modifying agents for radiotherapy, but they present potential problems in both synthesis and stability. Another series of tricyclic compounds is based on the cyclo-homologation of the lactam ring corresponding to the 2-phenyl-1H-benzimidazoles (Figure 3.8, 9). Derivatives based on 9 appeared to be promising candidates, able to mimic well the reference benzimidazole-4-carboxamide molecule 7, which has been adopted as a benchmark. Some examples of the derivatives of compound 9 were synthesized and led to the discovery of potent inhibitors compounds 10, 11 and 12 (Figure 3.9). Structure-activity relationships studies indicate that the carboxamide hydrogen is essential; in fact, its replacement by a methyl group causes the complete loss of binding affinity. Compound 12 shows a 25-fold increase in efficacy when compared to the reference molecule 7 and presents the best biological profile in cellular assays. The design concept was validated by X-ray crystallographic analyses of complexes with the PARP-1 target. In Figure 3.10 the representative inhibitor 2-(3-methoxyphenyl)benzimidazole-4-carboxamide, analogue to compound 7, bound at the PARP active site is reported (PDB code 1EFY). Its carbonyl oxygen accepts hydrogen bonds from the side chain OH of Ser904 and from the Gly863 backbone amide. The ligand amide donates a hydrogen bond to the Gly863 backbone carbonyl oxygen. The conformation observed for the carboxamide group of the co-crystallized ligand was also observed in the solid state (i.e., measured for the protein-free compound) of other benzimidazole analogues, confirming the essential role of molecular modelling data in the drug discovery process (White et al., 2000).
Key Takeaways
- Structure-based and ligand-based drug design are two important approaches in anticancer research.
- The Cambridge Structural Database (CSD) and Protein Data Bank (PDB) provide valuable structural information for drug discovery.
- Molecular modeling can help identify potential binding modes of inhibitors to target enzymes.
Practical Tips
- Utilize computational methods like molecular modeling to predict the binding affinity of potential drugs before extensive synthesis.
- Consider both structure-based and ligand-based approaches in drug design to enhance the chances of success.
- Validate your drug candidates through X-ray crystallography or other structural biology techniques.
Warnings & Risks
- Be cautious with compounds that lack water solubility, as they may not be effective in vivo.
- Compounds with high binding affinity do not always translate to therapeutic efficacy; further testing is necessary.
- The synthesis and stability of constrained scaffolds can present significant challenges.
Modern Application
While the techniques described in this chapter are rooted in historical methods, they still hold relevance for modern survival preparedness. Understanding these approaches can help in developing effective treatments for various diseases, including cancer. The advancements in computational tools have made it easier to predict and optimize drug candidates, reducing the time and cost of drug development.
Frequently Asked Questions
Q: What are some examples of ligand-based inhibitors mentioned in this chapter?
The chapter mentions two main classes of aromatase enzyme inhibitors: steroidal analogues like compound 1 (androst-4-ene-3,17-dione) and non-steroidal inhibitors such as compound 2 (CGS 16949A).
Q: How does the structure-based drug design approach differ from ligand-based drug design?
Structure-based drug design uses experimentally determined complex structures, while ligand-based drug design relies on reference compounds and computational models. Structure-based approaches are more targeted to specific targets but may require structural information.
Q: What is the significance of molecular modeling in drug discovery according to this chapter?
Molecular modeling helps in mapping the active site of enzymes, performing conformational analysis of chemical structures, and identifying potential binding modes for inhibitors. This can guide the design of more effective drugs.