THE NATURE OF DRUGS
In the most general sense, a drug may be defined as any substance that brings about a change in biologic function through its chemical actions. In most cases, the drug molecule interacts as an agonist (activator) or antagonist (inhibitor) with a specific mol-ecule in the biologic system that plays a regulatory role. This target molecule is called a receptor. In a very small number of cases, drugs known as chemical antagonists may interact directly with other drugs, whereas a few drugs (osmotic agents) interact almost exclu-sively with water molecules. Drugs may be synthesized within the body (eg, hormones) or may be chemicals not synthesized in the body (ie, xenobiotics, from the Greek xenos, meaning “stranger”). Poisons are drugs that have almost exclusively harmful effects.However, Paracelsus (1493–1541) famously stated that “the dose makes the poison,” meaning that any substance can be harmful if taken in the wrong dosage. Toxins are usually defined as poisons of biologic origin, ie, synthesized by plants or animals, in contrast to inorganic poisons such as lead and arsenic.To interact chemically with its receptor, a drug molecule must have the appropriate size, electrical charge, shape, and atomic composition. Furthermore, a drug is often administered at a loca-tion distant from its intended site of action, eg, a pill given orally to relieve a headache. Therefore, a useful drug must have the nec-essary properties to be transported from its site of administration to its site of action. Finally, a practical drug should be inactivated or excreted from the body at a reasonable rate so that its actions will be of appropriate duration.
Drugs may be solid at room temperature (eg, aspirin, atropine), liquid (eg, nicotine, ethanol), or gaseous (eg, nitrous oxide). These factors often determine the best route of administration. The most common routes of administration are described in Table 3–3. The various classes of organic compounds—carbohydrates, proteins, lipids, and their constituents—are all represented in pharmacol-ogy. As noted above, oligonucleotides, in the form of small seg-ments of RNA, have entered clinical trials and are on the threshold of introduction into therapeutics.
A number of useful or dangerous drugs are inorganic elements, eg, lithium, iron, and heavy metals. Many organic drugs are weak acids or bases. This fact has important implications for the way they are handled by the body, because pH differences in the vari-ous compartments of the body may alter the degree of ionization of such drugs (see text that follows).
The molecular size of drugs varies from very small (lithium ion, MW 7) to very large (eg, alteplase [t-PA], a protein of MW 59,050). However, most drugs have molecular weights between 100 and 1000. The lower limit of this narrow range is probably set by the requirements for specificity of action. To have a good “fit” to only one type of receptor, a drug molecule must be sufficiently unique in shape, charge, and other properties, to prevent its bind-ing to other receptors. To achieve such selective binding, it appears that a molecule should in most cases be at least 100 MW units in size. The upper limit in molecular weight is determined primarily by the requirement that drugs must be able to move within the
body (eg, from the site of administration to the site of action). Drugs much larger than MW 1000 do not diffuse readily between compartments of the body (see Permeation, in following text). Therefore, very large drugs (usually proteins) must often be administered directly into the compartment where they have their effect. In the case of alteplase, a clot-dissolving enzyme, the drug is administered directly into the vascular compartment by intrave-nous or intra-arterial infusion.
Drugs interact with receptors by means of chemical forces or bonds. These are of three major types: covalent, electrostatic, and hydrophobic. Covalent bonds are very strong and in many casesnot reversible under biologic conditions. Thus, the covalent bond formed between the acetyl group of acetylsalicylic acid (aspirin) and cyclooxygenase, its enzyme target in platelets, is not readily broken. The platelet aggregation–blocking effect of aspirin lasts long after free acetylsalicylic acid has disappeared from the blood-stream (about 15 minutes) and is reversed only by the synthesis of new enzyme in new platelets, a process that takes several days. Other examples of highly reactive, covalent bond-forming drugs are the DNA-alkylating agents used in cancer chemotherapy to disrupt cell division in the tumor.Electrostatic bonding is much more common than covalent bonding in drug-receptor interactions. Electrostatic bonds vary from relatively strong linkages between permanently charged ionic molecules to weaker hydrogen bonds and very weak induced dipole interactions such as van der Waals forces and similar phe-nomena. Electrostatic bonds are weaker than covalent bonds.
Hydrophobic bonds are usually quite weak and are probably important in the interactions of highly lipid-soluble drugs with the lipids of cell membranes and perhaps in the interaction of drugs with the internal walls of receptor “pockets.”
The specific nature of a particular drug-receptor bond is of less practical importance than the fact that drugs that bind through weak bonds to their receptors are generally more selective than drugs that bind by means of very strong bonds. This is because weak bonds require a very precise fit of the drug to its receptor if an inter-action is to occur. Only a few receptor types are likely to provide such a precise fit for a particular drug structure. Thus, if we wished to design a highly selective short-acting drug for a particular recep-tor, we would avoid highly reactive molecules that form covalent bonds and instead choose a molecule that forms weaker bonds.
A few substances that are almost completely inert in the chemical sense nevertheless have significant pharmacologic effects. For example, xenon, an “inert” gas, has anesthetic effects at ele-vated pressures.
The shape of a drug molecule must be such as to permit binding to its receptor site via the bonds just described. Optimally, the drug’s shape is complementary to that of the receptor site in the same way that a key is complementary to a lock. Furthermore, the phenome-non of chirality (stereoisomerism) is so common in biology that more than half of all useful drugs are chiral molecules; that is, they can exist as enantiomeric pairs. Drugs with two asymmetric centers have four diastereomers, eg, ephedrine, a sympathomimetic drug. In most cases, one of these enantiomers is much more potent than its mirror image enantiomer, reflecting a better fit to the receptor mol-ecule. If one imagines the receptor site to be like a glove into which the drug molecule must fit to bring about its effect, it is clear why a “left-oriented” drug is more effective in binding to a left-hand receptor than its “right-oriented” enantiomer.
The more active enantiomer at one type of receptor site may not be more active at another receptor type, eg, a type that may be responsible for some other effect. For example, carvedilol, a drug that interacts with adrenoceptors, has a single chiral center and thus two enantiomers (Figure 1–2, Table 1–1). One of these enantiomers, the (S)(−) isomer, is a potent β-receptor blocker. The (R)(+) isomer is 100-fold weaker at the β receptor. However, theisomers are approximately equipotent as α-receptor blockers. Ketamine is an intravenous anesthetic. The (+) enantiomer is a more potent anesthetic and is less toxic than the (−) enantiomer. Unfortunately, the drug is still used as the racemic mixture.
Finally, because enzymes are usually stereoselective, one drug enantiomer is often more susceptible than the other to drug-metabolizing enzymes. As a result, the duration of action of one enantiomer may be quite different from that of the other. Similarly, drug transporters may be stereoselective.
Unfortunately, most studies of clinical efficacy and drug elimi-nation in humans have been carried out with racemic mixtures ofdrugs rather than with the separate enantiomers. At present, only a small percentage of the chiral drugs used clinically are marketed as the active isomer—the rest are available only as racemic mixtures. As a result, many patients are receiving drug doses of which 50% is less active, inactive, or actively toxic. Some drugs are currently available in both the racemic and the pure, active isomer forms. Unfortunately, the hope that administration of the pure, active enantiomer would decrease adverse effects relative to those pro-duced by racemic formulations has not been firmly established. However, there is increasing interest at both the scientific and the regulatory levels in making more chiral drugs available as their active enantiomers
Rational design of drugs implies the ability to predict the appropri-ate molecular structure of a drug on the basis of information about its biologic receptor. Until recently, no receptor was known in suf-ficient detail to permit such drug design. Instead, drugs were devel-oped through random testing of chemicals or modification of drugs already known to have some effect . However, the characterization of many receptors during the past three decades has changed this picture. A few drugs now in use were developed through molecular design based on knowledge of the three-dimen-sional structure of the receptor site. Computer programs are now available that can iteratively optimize drug structures to fit known receptors. As more becomes known about receptor structure, ratio-nal drug design will become more common.
The spectacular success of newer, more efficient ways to identify and characterize receptors has resulted in a variety of differing, and sometimes confusing, systems for naming them. This in turn has led to a number of suggestions regarding more rational methods of naming receptors. The interested reader is referred for details to the efforts of the International Union of Pharmacology (IUPHAR) Committee on Receptor Nomenclature andDrug Classification (reported in various issues of Pharmacological Reviews) and to Alexander SPH, Mathie A, Peters JA: Guide toreceptors and channels (GRAC), 4th edition. Br J Pharmacol 2009;158(Suppl 1):S1–S254.