Introduction This volume is part of Elsevier’s Learning Trends series. Elsevier Science & Technology Books provides this series of free digital volumes to support and encourage learning and development across the sciences. Titles include content excerpts focused on a central theme to give the reader an introduction to new ideas and information on that topic. This volume in Chemistry Learning Trends introduces readers to a key chapter from the 4th edition of Camille Wermuth’s Practice of Medicinal Chemistry and highlights the interdisciplinary nature of medicinal chemistry. The succeeding articles, from the ScienceDirect Reference Module in Chemistry, Molecular Sciences and Chemical Engineering, will introduce readers to important themes and valuable methods raised in this chapter. Thank you for being a part of the Elsevier community! Table of Contents 1) The Practice of Medicinal Chemistry 4thEdition – Chapter 3 Drug Targets, Target Identification, Validation and Screening by Walter MM Van den Broeck 2) Medicinal and Pharmaceutical Chemistry by Timmerman 3) Perspectives in Drug Discovery by W.T. Comer 4) LIQUID CHROMATOGRAPHY | Affinity Chromatography by D.S. Hage 5) Microarrays by D. Amaratunga H. Göhlmann & P.J. Peeters 6) Systems Biology by L. Coulier, S. Wopereis, C. Rubingh, H. Hendriks, M. Radonjić & R.H. Jellema 7) Comparative Modeling of Drug Target Proteins by B. Webb, N. Eswar, H. Fan, N. Khuri, U. Pieper, G.Q. Dong & A. Sali CHAPTER 3 Drug Targets, Target Identification, Validation, and Screening Walter M.M. Van den Broeck Janssen Infectious Diseases BVBA, Beerse, Belgium OUTLINE I. Introduction 45 C. Haploinsufficiency Profiling in Yeast 58 D. Analysis of Resistant Mutants 59 II. What is a Drug Target? 46 E. siRNA for Target Validation 60 III. The Purpose of Target Identification 47 F. Yeast Three-Hybrid System 61 A. Target-Based Screening. 47 G. DNA Microarrays 63 B. Phenotypic Screening 47 H. Comparative Profiling 64 C. Fast Follower Strategy 50 I. Analysis of the Pathophysiology 65 J. The Study of Existing Drugs 66 IV. Target Options and Treatment Options 51 K. Systems Biology 66 V. Target Deconvolution and Target Discovery 53 L. In Silico Simulation of the Human Patient 67 VI. Methods for Target Identification and VII. Target Validation 68 Validation 54 VIII. Conclusion 68 A. Affinity Chromatography 54 B. Genetic Methods 57 References 68 It doesn’t matter how beautiful your theory is, it doesn’t matter how smart you are or what your name is, if it doesn’t agree with experiment, it’s wrong. Richard P. Feynman (American theoretical physicist 1918–1988) I. INTRODUCTION For ages, humans have been using medicinal substances without tools like DNA microarrays to identify them. Instead, they were guided by theories like the concept of the four humors in Greco-Roman medicine or by spiritual sys- tems like animism. The chances are high that modern medicinal chemists would fully reject these rationales. Today we believe that the essential first step in the discovery of a new cure for a disease is the identification of the protein that is at the basis of that disease. The chances are high that medicinal chemists would fully agree with this rationale, but maybe they shouldn’t. In this chapter, we will see why. The Practice of Medicinal Chemistry. 45 © 2015 Elsevier Ltd. All rights reserved. 46 3. DRUG TARGET IDENTIFICATION AND SCREENING First, we examine why the definition of a drug target is already a bit misleading. Then we explore whether the mantra “first a target, then a drug” is a good guideline. We compare the three most used strategies for drug dis- covery today and assess the role of target identification in these strategies. The next question is what kind of tar- gets we should try to identify. Is the search for the cause of a disease a fruitful road to find new cures? Can we find cures altogether? Finally, after having established the difference between the two meanings of target identifi- cation, we describe briefly the current and most frequently used methods to identify and validate drug targets. II. WHAT IS A DRUG TARGET? In 1891, Paul Ehrlich was experimenting with dyes to stain bacteria. He had already made outstanding contribu- tions in treating infected patients with antitoxins. Together with vaccines, these account for the successful immuno- therapy. Ehrlich saw this immunotherapy as chemical reactions between very complex structures. At that time, the concepts of cells and microorganisms were very new and nobody understood the composition of cells. Maybe a cell was one big molecule, (i.e., a cell-molecule). Ehrlich believed that cell-molecules had side-chains to receive nutrients from outside, and he called these side-chains receptors. He thought that bacteria also had receptors and that the staining of bacteria was a chemical reaction between the dye molecule and the receptors. What if this reac- tion could not only stain the bacteria but also kill them? What if this dye could do the same in an infected patient? Ehrlich showed that methylene blue was taken up by the malaria parasite and had modest effects in two patients. He was extremely excited by this and coined the term “chemotherapy.” The difference with immunotherapy was that now the antitoxins—which were very complex and difficult to produce and standardize—could be replaced by well-identified chemicals (small molecules) that were easier to produce and handle. We owe the concept that a drug acts by binding to a target molecule to Paul Ehrlich. In his own words:,“Corpora non agunt nisi fixata” or “substances don’t act unless they are bound.” Today this concept is still valid. The Oxford Dictionary of Biochemistry defines a drug target as “a biological entity (usually a protein or gene) that interacts with, and whose activity is modulated by, a particular compound.” Peter Imming [1] uses the following working definition: a molecular structure (chemically definable by at least a molecular mass) that will undergo a specific interaction with chemicals we call drugs because they are administered to treat or diagnose a disease. The interaction has a connection with the clinical effect(s). These definitions could give the misleading impression that a drugÀtarget interaction is a one-to-one relation [2], as if every drug acted by binding to one and only one single specific target. This impression is further strengthened by the ambition of every medicinal chemist, starting with Paul Erhlich himself, to synthesize a “magic bullet,” an ultra-specific compound that would bind only to the target and to nothing else. However, evi- dence is growing that many drugs are successful just because they act on multiple different—not co-located— targets, potentially even hitting several pathways together [3]. Of the 1366 drugs reported in DrugBank 2.0, about 960 have more than one therapeutic target [4], a phenomenon called polypharmacology. As a consequence, searching for a super-selective drug may not always lead to the most active compound. In this perspective, target-based drug screening is not well suited to discover these so called “dirty drugs.” The one-to-one relation also doesn’t fit with drugs that act by binding to a complex of proteins or even a com- plex between proteins and nucleic acids. Many proteins form dimers, trimers, or even more complex constella- tions. In these cases, the drug binding pocket could contain parts of two or more proteins. But the target discovery tools are less well suited to find such targets. Yet another—very obvious—violation of the one-to-one relation is that the same pocket can accommodate many different small molecules. A substantial part of all new drugs is based on this promiscuous behavior of many pockets. The production of close analogues—or, more pejoratively, “me too drugs” —is often seen as a risk averse and profit driven strategy. Nevertheless, these drugs often result in an important incremental progress in activity, side effect profile, or administration facility [5]. A less obvious violation of the one-to-one relation is the fact that a protein can contain multiple pockets. Usually these pockets are all different and could partially overlap, be indirectly connected by allosteric regulation, or be completely separated. The binding to these different pockets could result in different effects. For example, the bind- ing with nucleoside drugs to the active site of a viral polymerase makes it more difficult for the virus to build resis- tance than with nonnucleoside drugs that have their binding site outside the active site of the enzyme. These comments make the picture of a drug target more complex. We could define a drug target as the minimal constellation of molecules that elicit a medically desired effect when bound by a drug. I. GENERAL ASPECTS OF MEDICINAL CHEMISTRY III. THE PURPOSE OF TARGET IDENTIFICATION 47 III. THE PURPOSE OF TARGET IDENTIFICATION Before exploring the plethora of methods to identify drug targets, we should discuss the role and the value of target identification in the drug discovery process. We will describe the role of target identification in the follow- ing three drug discovery strategies for small molecules: À Target-Based Screening Strategy À Phenotypic Screening Strategy À Fast Follower Strategy A. Target-Based Screening Target identification is the cornerstone of target-based screening. The concept underlying this strategy is that at the most fundamental level, most drugs work by binding to a specific target. Therefore, if you want to make a truly new drug, the first thing you have to do is to find a new target. The next step is to find small molecules that bind to this target, preferably as specific as possible. This procedure looks so overwhelmingly self-evident, innovative, and scientific that the complete pharmaceutical research community has been dreaming for decades about realizing this strategy.