CHAPTER 1 Pharmacodynamics
Mechanisms of Drug Action
The actions of most therapeutic agents are imbued with a certain degree of specificity. In conventional doses, drugs are generally selective in action; that is, they influence a narrow spectrum of biologic events. In addition, the pharmacologic profile of such agents is often markedly dependent on chemical structure; simple molecular modifications may drastically alter drug activity. These attributes of drug action suggest that the tissue components with which drugs interact to cause observable effects are uniquely individualized. Such tissue elements must have highly ordered physicochemical properties to permit particular compounds to combine with them, while prohibiting all others from doing so. They must also be intimately involved with discrete processes of life for drug interactions to exert specific physiologic influences. These “biologic partners” of drug action are given the term receptors.
The existence of receptors for exogenously administered drugs implies that drugs often mimic or inhibit the actions of endogenous ligands for these receptors. Drugs rarely produce novel effects; instead, they modify existing physiologic functions.
For many years after their postulation a century ago, receptors remained an enigma to pharmacologists. Little was known about them other than the probability that they were complex macromolecules possessing a ligand-binding site to interact with specific drugs and an effector site to initiate the pharmacologic response. With the development of biochemical methods for the isolation, solubilization, and characterization of proteins, however, enzymes became available as model systems for the study of drug-receptor interactions. Enzymes exhibit many of the properties that are ascribed to receptors. They are macromolecules having measurable biologic functions and possessing specific reactive sites for selected substrates. The close association between enzymes and receptors was underscored in the early 1940s when it became apparent that some enzymes are drug receptors. The list of drugs that alter known enzymatic activities is extensive and includes angiotensin-converting enzyme inhibitors, allopurinol, anticholinesterases, carbidopa, carbonic anhydrase inhibitors, disulfiram, entacapone, monoamine oxidase inhibitors, protease inhibitors, reverse transcriptase inhibitors, statin cholesterol synthesis inhibitors, sulfonamides, trimethoprim, and various antimetabolites used in cancer chemotherapy.
Besides enzymes (including coenzymes) and other easily solubilized proteins, at least two additional classes of receptors have been identified and are of clinical significance: nucleic acids and membrane-linked proteins. Nucleic acids serve as receptors for a limited number of agents. Certain antibiotics and antineoplastic compounds interfere with replication, transcription, or translation of genetic material by binding, sometimes irreversibly, to the nucleic acids involved. Other drugs, including thyroid hormones, vitamin D analogues, sex steroids, and adrenal corticosteroids, also modify transcription, but here the affected DNA becomes activated or inhibited as a consequence of drug interaction with a separate receptor protein in the cytosol or nucleus of the cell, as described subsequently. The most common receptors of drugs are those located on or within the various membranes of the cell. Their study has been greatly aided in recent years by developments in genomics, proteomics, and informatics. Membrane transporter proteins and metabolic enzymes, described in Chapter 2 for their influence on drug disposition, are themselves targets of drug action. Of greater significance are the many integral membrane proteins that function as receptors for endogenous regulatory ligands, such as neurotransmitters, hormones, and other signaling molecules.
Receptors involved in physiologic regulation can be grouped by molecular structure and functional characteristics into several superfamilies. Most of these receptors have one or more extracellular ligand-binding domains linked by one or more lipophilic membrane-spanning segments to an effector domain often, but not always, located on the cytoplasmic side of the membrane. This arrangement is ideal for the translation of an extracellular signal into an intracellular response. Usually, the endogenous ligand “signal” is hydrophilic and incapable of passive diffusion through the cell membrane. For lipophilic regulatory ligands, such as for thyroid hormone and various steroids, a separate superfamily of intracellular receptors exists. Commonly, drug binding exposes a DNA-binding site on the receptor protein, allowing the receptor to interact with DNA and alter transcription. These major classes of receptors are illustrated in Figure 1-1 and described subsequently.
FIGURE 1-1 Examples of four major classes of receptors and signal transduction mechanisms. Arrows denote the receptor ligand-binding sites. A, Intracellular receptors. Lipophilic substances such as steroids can cross the plasma membrane and activate intracellular receptors, which, after translocation to the nucleus, alter gene transcription and, ultimately, synthesis of new protein. B, Ion channel–linked receptors. Drugs such as nicotine can activate ligand-gated ion channels, leading to depolarization (or hyperpolarization) of the plasma membrane. C, G protein–linked receptors. Many drugs can activate G protein–linked receptors, causing release of the α and βγ subunits of associated G proteins. D, Enzyme-linked receptors. Drugs such as insulin promote dimerization of its receptor and activation of the catalytic site on the intracellular end of the receptor.
There are two general classes of ion channels: voltage gated and ligand gated. Voltage-gated ion channels are activated by alterations in membrane voltage. Voltage-gated Na+ channels open when the membrane is depolarized to a threshold potential and contribute to further membrane depolarization by allowing Na+ influx into the cell. As described in Chapter 16, local anesthetics such as lidocaine bind to voltage-gated Na+ channels, leading to blockade of neuronal depolarization. Specific voltage-gated ion channels also exist for K+, Ca++, H+, and Cl−.
In contrast, ligand-gated ion channels are activated in response to the binding of specific ligands or drugs. Many neurotransmitters and drugs and some cytoplasmic ligands activate membrane-bound ligand-gated ion channels, including several types of glutamate receptors and one 5-hydroxytryptamine (5-HT3) receptor promoting Na+, K+, or Ca++ movements, and certain γ-aminobutyric acid and glycine receptors promoting Cl− influx. Depending on the ionic charge and the direction of flow, ligand-gated ion channels can either depolarize or hyperpolarize the cell membrane.
The nicotinic receptor (Figure 1-2), the first membrane-bound drug receptor to be fully characterized,12,22 is an important example of a ligand-gated ion channel. An oligomeric structure, the polypeptide constituents of the nicotinic receptor are arranged concentrically to form a channel through which small ions can traverse the plasma membrane when the receptor is activated by the binding of two acetylcholine (ACh) molecules. As is the case with other ion channels, numerous subtypes of nicotinic receptors exist expressing differing affinities for specific ligands.
FIGURE 1-2 Ribbon structural model of the nicotinic ACh receptor from the electric organ of Torpedo marmorata. Left, View from the synaptic cleft. Five polypeptide units consisting of four different types (α, β, γ, and δ) form a rosette with a hydrophilic pore spanning the center of the oligomer. External regions, which include the ACh binding sites, are highlighted. Arrows indicate the α-subunit tryptophan (W149) that constitutes part of each ligand-binding site. Right, View parallel to the plasma membrane. Each polypeptide subunit includes four α-helical sequences that traverse the plasma membrane; the front two subunits are highlighted. Arrow indicates the same W149 residue. E, External surface (interstitial space); I, internal surface (cytoplasmic space).
(Adapted from Unwin N: Refined structure of the nicotinic acetylcholine receptor at 4 Å resolution, J Mol Biol 346:967-989, 2005.)
G protein–linked receptors, sometimes referred to as metabotropic receptors, constitute the largest superfamily of integral membrane proteins, and collectively serve as targets for approximately half of all nonantimicrobial prescription drugs.9,11 The basic structure of these receptors includes a common seven-membered transmembrane domain. Generally, metabotropic receptors greatly amplify extracellular biologic signals because they activate G proteins, which activate ion channels or, more commonly, other enzymes (e.g., adenylyl cyclase), leading to the introduction or formation of a host of internal second messengers for each extracellular signal molecule detected. This amplification system, which also usually involves an extended duration of activation of G proteins relative to the binding of drug to the receptor, may explain why maximal pharmacologic effects are often observed when only a small proportion of receptors are activated.
G proteins are heterotrimers consisting of α, β, and γ subunits. After receptor activation, guanosine diphosphate attached to the α subunit is replaced by guanosine triphosphate, and the heterotrimer splits into the α monomer and βγ dimer. Many, but not all, of the observed cellular actions are caused by the α subunit (see Figure 5-7). Gαs, the specific α subunit for the G protein associated with β-adrenergic receptors, activates adenylyl cyclase, which catalyzes the synthesis of cyclic adenosine 3′,5′-monophosphate (cAMP).9 cAMP activates protein kinase A, which catalyzes the phosphorylation of serine and threonine residues of certain intracellular proteins, leading to altered cellular function.
The G protein system is complex and still incompletely understood. One receptor subtype may activate different G proteins, several receptor subtypes may activate the same G protein, and the ultimate target proteins can exist in tissue-specific isoforms with differing susceptibilities to secondary effector systems. The different G protein pathways can also interact with one another. The complexity of G protein signal transduction provides a sophisticated regulatory system by which cellular responses can vary, depending on the combination of receptors activated and the cell-specific expression of distinct regulatory and target proteins. Several specific membrane-bound G proteins are discussed in Chapter 5.
FIGURE 1-3 Ribbon structural model of the β2-adrenergic receptor. Left, View from the synaptic cleft; right, view parallel to the plasma membrane. For technical reasons, it was not possible to visualize the extracellular N-terminal amino acid chain attached to transmembrane helix 1 (TM1), the intracellular connector between TM5 and TM6, and a major portion of the intracellular C terminal. Binding of a drug ligand (in this case, the β-adrenergic receptor antagonist carazolol) is represented as a stick figure. The rotamer toggle switch tryptophan (W293), which allows TM6 to move in response to drug agonists, is indicated by the space-filling spheres. ECL, Extracellular loop.
(Adapted from Hanson MA, Stevens RC: Discovery of new GPCR biology: one receptor structure at a time, Structure 17:8-14, 2009.)
Enzyme-linked receptors have only one transmembrane domain per protein subunit, with an enzymatic catalytic site on the cytoplasmic side of the receptor. Dimerization of activated receptors usually provides the conformational change required for expression of enzymatic activity. The catalytic sites are commonly protein kinases that phosphorylate tyrosine or, less commonly, serine or threonine residues on target proteins. Autophosphorylation of the receptor also occurs. Some catalytic receptors have guanylyl cyclase or tyrosine phosphatase activity. Insulin, atrial natriuretic peptide, and various growth factors (e.g., epidermal growth factor) activate catalytic receptors. A closely related group of receptors responsible for the action of numerous peptides—including various neurotrophic peptides, growth hormone, and cytokines—lacks enzymatic activity. In such cases, the catalytic site is supplied by a separate nonreceptor protein kinase that interacts with the dimerized receptor.
Many forms of cancer seem to involve mutant variants of enzyme-linked receptors in which the catalytic site or associated nonreceptor protein kinase is continuously activated.4 Approximately half of all oncogenes discovered to date encode for continuously activated protein kinases.
Lipophilic substances capable of crossing the plasma membrane may activate intracellular receptors. Sex steroids, mineralocorticoids, glucocorticoids, thyroid hormones, and vitamin D derivatives all activate specific nuclear receptors that influence DNA transcription.8,20 The typical nuclear receptor is composed of three major subunits: the carboxyl end of the receptor forms the ligand-binding domain, the adjacent segment includes the DNA-binding region, and the amino terminus constitutes the transcription-modulating domain. When a drug (or hormone) binds to the receptor, it folds into the active configuration and dimerizes with a partner receptor. The conformational change results in a dramatic increase in binding to specific DNA sequences. Binding of thyroid hormone to its receptor produces more than a 10-fold increase in receptor affinity for binding to DNA.20 DNA binding of the activated receptor often initiates transcription, leading to increased production of specific proteins. Because this type of signal transduction requires protein synthesis, drugs that activate intracellular receptors typically have a delay of several hours before the onset of their pharmacologic effect. (For this reason, glucocorticoids cannot be used as primary drugs for the management of anaphylaxis.) In some systems, the binding of the drug-receptor complex inhibits transcription. Regardless of the specific mechanism involved, however, the intensity and duration of drug effect is temporally independent of the plasma concentration.
In addition to these intracellular receptors, other enzymes and proteins involved in cellular metabolism and gene expression are receiving increasing scrutiny as potential targets for drug therapy. Nitric oxide, which stimulates guanylyl cyclase directly to form cyclic guanosine 3′,5′-monophosphate (cGMP), and sildenafil, which inhibits the breakdown of cGMP by cGMP-specific phosphodiesterase-5, are two examples of currently available agents acting intracellularly on regulatory enzymes. Finally, structural proteins such as tubulin, which are assembled to form microtubules, are targets for several drugs used in the treatment of cancer, gout, and fungal infections.
Implicit in the interaction of a drug with its receptor is the chemical binding of that drug to one or more specific sites on the receptor molecule. Five basic types of binding may be involved (Figure 1-4).
Covalent bonds arise from the sharing of electrons by a pair of atoms. Although covalent bonds are required for the structural integrity of molecules, they are generally not involved in drug-receptor interactions. Most drugs reversibly associate with their receptors. As described in Chapter 2, the duration of action of these agents is related to how long an effective drug concentration remains in the vicinity of the drug receptors. This time may vary from a few minutes to many days, but usually is on the order of several hours. With bond energies of 250 to 500 kJ/mol, the stabilities of covalent linkages are so great that, when formed, drug-receptor complexes are often irreversible. In these instances, the duration of action is not influenced by the concentration of unbound drug surrounding the receptors. Instead, it may depend on the synthesis of new receptors or on the turnover of the affected cells, processes that often take days to weeks. When the receptors happen to compose or influence the genetic material of a cell, drug effects may be permanent.
Ionic bonds result from the electrostatic attraction between ions of opposite charge. Such associations are relatively weak in an aqueous environment, having bond energies of approximately 20 kJ/mol. Nevertheless, many drugs have a formal charge at physiologic pH, and it is likely that ionic bonds are commonly made with ionized groups located at receptor sites. Because the attraction between ions is inversely proportional to the square of the distance separating them, ionic influences operate over much greater distances than do other interatomic forces. It is reasonable to assume that ionic bonds initiate many drug-receptor combinations.
Although benzene and similar aromatic compounds are hydrophobic solvents, their π electron clouds are capable of interacting with positively charged ions.5 Phenylalanine, tyrosine, and tryptophan—amino acids with aromatic side groups—retain this ability. These amino acids are common constituents at receptor sites for such positively charged drugs as ACh, dopamine, epinephrine, and 5-HT. Individual bond energies are similar to those of hydrogen bonds described subsequently; however, interactions between multiple aromatic amino acids and a single cationic moiety commonly strengthen the overall interaction.
The hydrogen bond represents a special type of interaction between polar molecules. When a hydrogen atom is covalently attached to a strongly electronegative atom such as oxygen or nitrogen, it becomes partially stripped of its electron and takes on some of the characteristics of a bare proton. Strongly electropositive and with an exceedingly small atomic radius, the hydrogen nucleus is able to associate closely with additional electronegative atoms. Hydrogen linkages are generally weaker than ionic bonds (approximately 5 kJ/mol) and are more sensitive to interatomic separation. Functional groups capable of forming hydrogen bonds are common to drugs and receptor sites, however, and if multiple unions occur, the resultant stabilizing force can far outweigh that of a single ionic bond.
Van der Waals forces collectively describe the weak interactions that develop when two atoms are placed in close proximity. The electrostatic attractions that constitute these forces result from reciprocal perturbations in the electron clouds of the atoms involved. These “bonds” are the weakest of the five types described (approximately 0.5 kJ/mol); in addition, they decrease in strength according to the seventh power of the interatomic distance. Paradoxic as it may seem, van der Waals forces are of primary importance in conferring specificity to drug-receptor interactions. Because even electroneutral carbon atoms can participate in such associations, the number of these bonds that connect a drug to its receptor may be large, and the total binding force may be considerable. When minor steric influences prevent an exact fit between a drug and its receptor, the sensitivity of van der Waals forces to interatomic separation forestalls their development, and drug-receptor stability markedly declines.
In addition to the bonding forces already described, hydrophobic interactions between the drug, its receptor, and the aqueous environment can play a major role in stabilizing drug-receptor binding. Water is an unusual liquid with respect to its ability to form hydrogen bonds with itself and with various solutes. The association of a drug with its receptor is enhanced if the drug is hydrophobic or if the surface area of a nonpolar region of the receptor is reduced by drug binding. In either case, stability occurs because of the reduced perturbation of the normal water structure.
The binding of a drug to its receptor is generally not related to a particular attractive force but results from the conjoint action of ionic, cation-π, hydrogen, van der Waals, and (rarely) covalent linkages, often in synchrony with hydrophobic interactions. Each type of association contributes differently to the drug-receptor complex. When random movement causes a drug molecule to approach or collide with the receptor surface, ionic attractions, closely followed by cation-π interactions, are the first to develop. Unable to convey specificity or stability to a drug-receptor union by themselves, these forces nevertheless serve to draw in and partially orient the drug to its receptor. As the intermolecular separation diminishes, hydrophobic influences, hydrogen bonding, and subsequently van der Waals forces become prominent. In concert, these interactions provide for the specificity of drug action; without an exact fit, binding is impaired, and the drug cannot adhere well enough to influence receptor function. Covalent linkage confers a high degree of permanency to the drug-receptor complex. Fortunately, though, irreversible binding is uncommon in therapeutics. Many agents are used to produce a single, temporary effect; covalent attachment would preclude such use. In many instances, covalent bonding would make drug regimens more difficult to administer and adverse reactions more troublesome to treat.
Examination of structure-activity relationships (SARs) is a time-honored method of studying drug-receptor interactions. In SAR investigations, specific features of the structure of a drug molecule are identified and then altered systematically to determine their influence on pharmacologic activity. The chemical features that are most often involved in these considerations are the presence and type of ionic charge; the effect of neighboring groups on the degree of ionization; hydrogen-bonding capability; and steric factors such as the size of alkyl side chains, the distance between reactive groups, and the three-dimensional configuration of such groups. SAR studies of closely related agents (congeners) have led to an understanding of the chemical prerequisites for pharmacologic activity and, on a practical level, made possible the molecular modification of drugs to provide enhanced or even novel therapeutic effects, while reducing the incidence and severity of toxic reactions. In addition, SAR studies serve to illustrate how the combined action of the various binding forces described earlier are necessary for maximal drug activity, which yields certain clues concerning the physicochemical properties of the receptor sites involved that are of value to investigators seeking to unravel the exact structure of these sites.
A recent example of SARs is provided by the study of the binding of norepinephrine and related drugs to the β2-adrenergic receptor (Figure 1-5).2 The norepinephrine molecule is composed of a catechol residue (a benzene ring with two hydroxyl groups in the meta and para positions) connected by a two-carbon intermediate chain to a nitrogen terminus that is positively charged at physiologic pH. The presence of a cationic nitrogen locus is essential for full activity; loss of the ionic charge by removing the nitrogen moiety or replacing it with a nonionic carbon group virtually eliminates drug action, as does replacement of the receptor’s aspartate residue (D113) with a neutral amino acid. Hydrogen bonds involving both ring hydroxyl groups with corresponding serine residues (S203, S204, and S207) greatly increase potency (by 25-fold, 33-fold, and 39-fold, respectively) but preclude entry of the drug into the central nervous system (CNS). Replacement of a hydroxyl with a larger group generally eliminates agonist activity at β receptors but may result in antagonistic effects. Another hydrogen bond between the β-hydroxyl group and its asparagine counterpart (N293) increases binding affinity 38 times. The distance separating the catechol and nitrogen moieties of the molecule is likewise crucial for full activity. Electrostatic interactions involving the benzene ring and aromatic amino acid residues of the receptor protein (e.g., F290) also contribute to the binding of norepinephrine.
FIGURE 1-5 Ribbon model of the interaction of norepinephrine (NE) with the β2-adrenergic receptor. The transmembrane segments involved in agonist binding (TM3, TM5, TM6, TM7) are shown along with their serine and asparagine residues (S203, S204, S207, and N293), involved in hydrogen binding (dashed lines), and the aspartate residue (D113) that forms an ionic bond. Other amino acids involved in agonist binding are also identified.
(Adapted from Bhattacharya S, Hall SE, Li H, et al: Ligand-stabilized conformational states of human β2 adrenergic receptor: insight into G-protein-coupled receptor activation, Biophys J 94:2027-2042, 2008.)
Although norepinephrine can fully activate the β2 receptor in vitro, it requires higher concentrations than those achieved physiologically. Epinephrine, the natural β2 receptor ligand, has a nitrogen-bound methyl group that increases β2 activity by 10 to 50 times. Increasing the size of the alkyl moiety on the nitrogen further increases β-adrenergic selectivity. Because alkyl moieties do not form hydrogen or ionic bonds, this finding implies that van der Waals forces or hydrophobic interactions, or both, contribute significantly to the binding of epinephrine and congeners with selective β2-adrenergic properties.