version date: 1 December 2006 EXERCISEⅢ4 MEDICINAL CHEMISTRY AND MOLECULAR MODELING: AN INTEGRATION FOR THE TEACHING OF DRUG STRUCTURE-ACTIVITY RELATIONSHIP AND THE MOLECULAR BASIS OF DRUG ACTION Ivone Carvalho*, Aurea D. L. borges, and Lilian S.C. Bernardes Faculdade de ciencias Farmaceuticas de ribeirao preto, Universidade de sao paulo, Av. do cafe s/n, Monte Alegre, Ribeirao Preto- SP, 14040-903, Brazil E-mail: carmona@usp.br Abstract: Molecular modeling is described as a tool for understanding fundamental concepts of drug structure-activity relationships in a medicinal chemistry course. The relevant molecular features of antimetabolite drugs are investigated by 3D visualization, their physical properties measured, and the molecular interaction pattern on target macromolecules illustrated by antineoplastic drugs. This approach became a high-quality computing and graphic tool to teach important aspects of biological molecules and drugs and the correlation of their structures and pharmacological actions. The students improve their perception and understanding of the molecular recognition process and may predict molecular properties by handling computer graphics and databases Keywords: molecular modeling; medicinal chemistry; drugs; molecular properties; molecular recognition Modern molecular modeling techniques are a remarkable tool in the search for potential novel therapeutic agents, by helping us to understand and predict the behavior of molecular systems. The powerful modeling components, including molecular graphics, computational chemistry, molecular database, and statistical modeling(QSAr) have assumed an important role in the development and optimization of leading compounds. Moreover, current formation on the proteins 3D structure and functions opens up the possibility of understanding the relevant molecular interactions between a ligand and a target <www.iupac.org/publications/cd/medicinalchemistry
1 EXERCISE III.4 MEDICINAL CHEMISTRY AND MOLECULAR MODELING: AN INTEGRATION FOR THE TEACHING OF DRUG STRUCTURE–ACTIVITY RELATIONSHIP AND THE MOLECULAR BASIS OF DRUG ACTION Ivone Carvalho*, Áurea D. L. Borges, and Lílian S. C. Bernardes Faculdade de Ciências Farmacêuticas de Ribeirão Preto, Universidade de São Paulo, Av. do Café s/n, Monte Alegre, Ribeirão Preto – SP,14040-903, Brazil E-mail: carronal@usp.br Abstract: Molecular modeling is described as a tool for understanding fundamental concepts of drug structure–activity relationships in a medicinal chemistry course. The relevant molecular features of antimetabolite drugs are investigated by 3D visualization, their physical properties measured, and the molecular interaction pattern on target macromolecules illustrated by antineoplastic drugs. This approach became a high-quality computing and graphic tool to teach important aspects of biological molecules and drugs and the correlation of their structures and pharmacological actions. The students improve their perception and understanding of the molecular recognition process and may predict molecular properties by handling computer graphics and databases. Keywords: molecular modeling; medicinal chemistry; drugs; molecular properties; molecular recognition. Modern molecular modeling techniques are a remarkable tool in the search for potential novel therapeutic agents, by helping us to understand and predict the behavior of molecular systems. The powerful modeling components, including molecular graphics, computational chemistry, molecular database, and statistical modeling (QSAR) have assumed an important role in the development and optimization of leading compounds. Moreover, current information on the protein’s 3D structure and functions opens up the possibility of understanding the relevant molecular interactions between a ligand and a target <www.iupac.org/publications/cd/medicinal_chemistry/> version date: 1 December 2006
version date: 1 December 2006 macromolecule. As a consequence, a comprehensive study of drug structure-activity relationships(SARs) can be developed and provide the proper identification of a 3D pharmacophore model for a rational drug design. Although structure-based design is now a medicinal chemistry routine approach, there are still difficulties in teaching fundamental concepts to undergraduate pharmacy students, such as those related to the molecular recognition process [1, 2 An active-learning strategy in medicinal chemistry involves the incorporation of molecular modeling techniques to assist third-year undergraduate students in the understanding of structure-activity principles. This paper focuses on the use of computational chemistry and the protein data bank(PDB), accessed from the Web site http:/www.rcsb.org/pdb,tounderstandandpredictthechemicalandmolecularbasis involved in the drug-receptor interactions. a comprehensive study of SArs comprises three approaches. The first one involves comparative analysis of antimetabolite drugs and the corresponding metabolites(named substrates), by representing, visualizing, and superimposing their 3D structures, obtained by minimization processes and molecular alignment techniques. Numerical properties of these molecules are then calculated, the most common being molecular energies and physical constants as partition coefficients, dipolar moment, molecular volume, and interatomic distance. Finally, particular structural feature between substrate and antimetabolite are explored by assessing the electrostatic and geometric patterns required for chemical interaction in the active site of the target molecule, obtained from PDB. Table 1 lists therapeutic targets of interest, describing the enzymes and their corresponding substrates, some PDB files and antimetabolites currently used as antineoplastic, anti-HIV, antibiotic, antihipertensive, anti-inflammatory, cholinergic, and hipolipemic drugs 3 <www.iupac.org/publications/cd/medicinalchemistry/> 2
2 macromolecule. As a consequence, a comprehensive study of drug structure–activity relationships (SARs) can be developed and provide the proper identification of a 3D pharmacophore model for a rational drug design. Although structure-based design is now a medicinal chemistry routine approach, there are still difficulties in teaching fundamental concepts to undergraduate pharmacy students, such as those related to the molecular recognition process [1,2]. An active-learning strategy in medicinal chemistry involves the incorporation of molecular modeling techniques to assist third-year undergraduate students in the understanding of structure–activity principles. This paper focuses on the use of computational chemistry and the protein data bank (PDB), accessed from the Web site http://www.rcsb.org/pdb, to understand and predict the chemical and molecular basis involved in the drug–receptor interactions. A comprehensive study of SARs comprises three approaches. The first one involves comparative analysis of antimetabolite drugs and the corresponding metabolites (named substrates), by representing, visualizing, and superimposing their 3D structures, obtained by minimization processes and molecular alignment techniques. Numerical properties of these molecules are then calculated, the most common being molecular energies and physical constants as partition coefficients, dipolar moment, molecular volume, and interatomic distance. Finally, particular structural features between substrate and antimetabolite are explored by assessing the electrostatic and geometric patterns required for chemical interaction in the active site of the target molecule, obtained from PDB. Table 1 lists therapeutic targets of interest, describing the enzymes and their corresponding substrates, some PDB files and antimetabolites currently used as antineoplastic, anti-HIV, antibiotic, antihipertensive, anti-inflammatory, cholinergic, and hipolipemic drugs [3]. <www.iupac.org/publications/cd/medicinal_chemistry/> version date: 1 December 2006
version date: 1 December 2006 Table 1 Antimetabolites, substrates, and their corresponding therapeutic targets and some PDB files of interest Group Therapeutic class Enzyme Substrate PDB file Drug antimetabolite Antineoplastic Thymidylate deoxyuridylate Itls, Itsn, 5-fluorouracil synthase monophosphate Ihzw, lkzi, trifluorothymidine IcifItsw Anti-HIV Reverse Deoxy I cot I cou. zidovudine pyridyl Idtq, Ifkp, lamivudine monophosphate Hmv, Ihvu stavudine didanosine III Anti-HIV HIV Protease Polyprotein Hvp, Kzk, saquinavir, 2aid, 2bpv, indinavir 2bpw, upj ritonavir Antibiotic Transpeptidase Acil-D-Ala-D- Icef, Iceg, penicillin G, and Carboxy Ala les2 les3 oxacillin carbenicillin cephalexin HMGO-COA IMGO-Coa ldqa, Ihw8, lovastatin, reductase Ihw9, Ihw simvastatin I qax, lqay fluvastatin atorvastidina VI Antihypertensive Dopa L-Dihydroxy- 1js3, 1js6, methyldopa, descarboxilase phenylala Iphh, 1d71 methy lnorepineph rine VIl Antineoplastic Dihydrofolate Dihydrofoli Irb3. lrg7. methotrexate Irh3. lra dfr. 3d trimetrexate VIll Antiinflammatory Cyclooxygenase Arachidonic Idcx, Icqe, ibuprofen Icvu, Iddx, indomethacin Icx2, Pgf nape dIsalicylic IX Choli Acetyl Acetylcholine 2ace, Igqr, neostigmin cholinesterase back, Idx 4, tacrine 2clj, qti pyr deme pi aldoxime <www.iupac.org/publications/cd/medicinalchemistry/>
3 Table 1 Antimetabolites, substrates, and their corresponding therapeutic targets and some PDB files of interest. Group Therapeutic class Enzyme Substrate PDB file Drug/ antimetabolite I Antineoplastic Thymidylate synthase deoxyuridylate monophosphate 1tls, 1tsn, 1hzw, 1kzi, 1hvy, 1bq1, 1cif,1tsw 5-fluorouracil, trifluorothymidine II Anti-HIV Reverse transcriptase Deoxythymidylate monophosphate 1cot, 1cou, 1dtq, 1fkp, 1hmv, 1hvu zidovudine, lamivudine, stavudine, didanosine III Anti-HIV HIV Protease Poliprotein 1hvp,1kzk, 2aid,2bpv, 2bpw, 7upj saquinovir, indinavir, ritonavir IV Antibiotic Transpeptidase and Carboxypeptidase Acil-D-Ala-DAla 1cef, 1ceg, 1es2, 1es3 penicillin G, oxacillin, ampicillin, carbenicillin, cephalexin V Hipolipemic HMGO-CoA reductase HMGO-CoA 1dqa, 1hw8, 1hw9, 1hwj, 1qax, 1qay lovastatin, simvastatin, fluvastatin, atorvastidina VI Antihypertensive Dopa descarboxilase L-Dihydroxyphenylalanine 1js3, 1js6, 5pah, 4pah, 1phh, 1d7l methyldopa, methyldopamine, methylnorepineph rine VII Antineoplastic Dihydrofolate reductase Dihydrofolic acid 1rb3, 1rg7, 1rh3, 1ra3, 7dfr, 3dfr methotrexate, piritrexim, trimetrexate VIII Antiinflammatory Cyclooxygenase Arachidonic acid 1dcx, 1cqe, 1cvu, 1ddx, 1cx2, 1pgf ibuprofen, indomethacin, naproxen, acetylsalicylic acid IX Cholinergic Acetylcholinesterase Acetylcholine 2ace, 1gqr, 2ack, 1dx4, 2clj, 1qti neostigmine, tacrine, pyridostigmine, echothiophate, demecarium, pralidoxime <www.iupac.org/publications/cd/medicinal_chemistry/> version date: 1 December 2006
version date: 1 December 2006 3D structure comparisons and overlay sing the Molecular Modeling Pro [4] program, it is possible to construct 3D interactive drug pictures, optimized by reducing the energy of the molecules in some systematic way until a minimum energy conformer is found. Minimization processes can correct unfavorable bond lengths, bond angles, torsion angles, and nonbonded interactions in a starting structure creating a more stable conformation. Mathematical models that perform geometry optimization are divided into classical, mechanical, and quantum mechanical approaches. In computational chemical simulations, the simplified description is a calculated potential energy surface, which applies classical mechanics equations to molecular nuclei, without considering electrons. A set of equations and parameters is called force field, and most molecular modeling programs can choose among several force fields, such as MM2 and AMBER [5] The energy of any atomic arrangement can be stepwise calculated, by assessing how the energy of the system varies as the position of the atoms change. Molecular Modeling Pro can generate and examine automatically many molecular conformations, and their corresponding inter-conversion energy barriers graphically plotted. At the completion of the conformational analysis, the molecule is placed in its low energy conformation. Both procedures can be interactively performed to optimize the geometry of the molecules that now may be compared structurally by overlaying appropriate atoms or functional groups, previously aligned in the atomic coordinates. Rotation and different representation forms(by charge or lipophilicity ) of the model allow a detailed investigation of the conformational and electronic properties of two structures, which could be the substrate and antimetabolite (inhibitor) Two classes of antineoplastic drugs are chosen to exemplify the structure visualization and superposition processes, which involve compounds that act on target enzymes as thymidylate synthase(Ts, Fig. 1)and dihydrofolate reductase(DHFR, Fig. 2). The purpos <www.iupac.org/publications/cd/medicinalchemistry/> 4
4 3D structure comparisons and overlays Using the Molecular Modeling Pro [4] program, it is possible to construct 3D interactive drug pictures, optimized by reducing the energy of the molecules in some systematic way until a minimum energy conformer is found. Minimization processes can correct unfavorable bond lengths, bond angles, torsion angles, and nonbonded interactions in a starting structure, creating a more stable conformation. Mathematical models that perform geometry optimization are divided into classical, mechanical, and quantum mechanical approaches. In computational chemical simulations, the simplified description is a calculated potential energy surface, which applies classical mechanics equations to molecular nuclei, without considering electrons. A set of equations and parameters is called force field, and most molecular modeling programs can choose among several force fields, such as MM2 and AMBER [5]. The energy of any atomic arrangement can be stepwise calculated, by assessing how the energy of the system varies as the position of the atoms change. Molecular Modeling Pro can generate and examine automatically many molecular conformations, and their corresponding inter-conversion energy barriers graphically plotted. At the completion of the conformational analysis, the molecule is placed in its low energy conformation. Both procedures can be interactively performed to optimize the geometry of the molecules that now may be compared structurally by overlaying appropriate atoms or functional groups, previously aligned in the atomic coordinates. Rotation and different representation forms (by charge or lipophilicity) of the model allow a detailed investigation of the conformational and electronic properties of two structures, which could be the substrate and antimetabolite (inhibitor). Two classes of antineoplastic drugs are chosen to exemplify the structure visualization and superposition processes, which involve compounds that act on target enzymes as thymidylate synthase (TS, Fig. 1) and dihydrofolate reductase (DHFR, Fig. 2). The purpose <www.iupac.org/publications/cd/medicinal_chemistry/> version date: 1 December 2006
version date: 1 December 2006 of this exercise is to provide a geometrical and chemical overview of the substrate and inhibitor of great structural similarity. At the same time, it reinforces the importance of chemical transformations of a key substrate in a biochemical pathway when developing antimetabolite drugs Molecular dimensions and properties Varying the physical and chemical properties of metabolite has often been used in drug design to produce a structure analog such as antimetabolite(metabolite antagonism), required for inhibition of cellular growth and response process. Bioisosterism strategies are frequently used to convert a key substrate into an inhibitor, allowing extra binding interaction to a target enzyme. The substituent modifications can affect various parameters in a drug molecule such as the partition coefficient, electronic density, conformation, bioavailability, and its capacity to establish direct interaction in the receptor domain. Thus, both drug and substrate physical properties can be calculated and compared to give useful information on SARs. Important properties can be estimated in the Molecular Modeling Program, exemplified by partial charges, bond order, dipole moment, ionization potential, electron density distribution solubility, and thermodynamic properties. Some of the physical properties, including volume surface area, density, molecular length, and dipole moment, are affected by geometry, which depends on the minimization process. Additionally, appropriate orientation of the molecule in the x-y plane is required for molecular length comparisons <www.iupac.org/publications/cd/medicinalchemistry/>
5 of this exercise is to provide a geometrical and chemical overview of the substrate and inhibitor of great structural similarity. At the same time, it reinforces the importance of chemical transformations of a key substrate in a biochemical pathway when developing antimetabolite drugs. Molecular dimensions and properties Varying the physical and chemical properties of metabolite has often been used in drug design to produce a structure analog such as antimetabolite (metabolite antagonism), required for inhibition of cellular growth and response process. Bioisosterism strategies are frequently used to convert a key substrate into an inhibitor, allowing extra binding interaction to a target enzyme. The substituent modifications can affect various parameters in a drug molecule such as the partition coefficient, electronic density, conformation, bioavailability, and its capacity to establish direct interaction in the receptor domain. Thus, both drug and substrate physical properties can be calculated and compared to give useful information on SARs. Important properties can be estimated in the Molecular Modeling Program, exemplified by partial charges, bond order, dipole moment, ionization potential, electron density distribution, solubility, and thermodynamic properties. Some of the physical properties, including volume, surface area, density, molecular length, and dipole moment, are affected by geometry, which depends on the minimization process. Additionally, appropriate orientation of the molecule in the x-y plane is required for molecular length comparisons. <www.iupac.org/publications/cd/medicinal_chemistry/> version date: 1 December 2006