Mechanisms of Drug Action

Historical Beginnings

The specificity of drugs for a particular disease has been known since at least the 17th century. The best known example of this is the efficacy of Peruvian bark, the predecessor of quinine, in the treatment of malaria. Sobernheim (1803–1846) first applied the concept of selective affinity to explain the apparent specificity of drugs. He believed, for example, that strychnine had an affinity for spinal cord while digitalis had affinity for the heart. Blake (1814–1893) first demonstrated that inorganic compounds with similar macroscopic crystalline structures exert similar effects when administered intravenously. This triggered a vigorous scientific debate at the turn of the 20th century regarding whether it was the chemical structure or physical properties of drugs that endow them with medicinal properties. This debate was particularly relevant for the theories of actions of general anesthetics because it was believed until recently that their relatively simple and diverse chemical structures precluded the possibility of a specific drug-receptor interaction.

The term receptor was first coined in 1900 by Ehrlich (1854–1915) as a replacement for his original term “side chain” (Seitenkette) that he used to explain the specificity of the antibacterial actions of antitoxins (antibodies). Ehrlich did not originally believe that specific receptors existed for small molecules such as medicinal compounds because they could easily be washed out of the body by solvents. This belief was at odds with the remarkable experimental findings of Langley (1852–1925), who was investigating whether the origin of the automatic activity of the heart resided in the heart muscle itself or was imposed on the heart by the nervous system. He demonstrated that the effect of the plant-derived drug jabonardi—bradycardia—occurred even when innervation was blocked, and that this effect was reversed by applying atropine directly to the heart. He went on to show that the relative abundance of the agonist (jabonardi) over its antagonist (atropine) determined the overall physiologic effect. This observation led Langley to propose that competition of the two drugs for binding to the same substance explained their antagonistic effects on the heart rate. However, the key experiment that led Langley to formulate his receptor concept came 30 years later in 1905, when he showed that the contractile effect of nicotine on skeletal muscle can be antagonized by curare. From the observation that even after application of curare the relaxed muscle contracted following direct application of electric current, he concluded that neither curare nor nicotine acted directly on the contractile machinery. Instead, Langley argued that the drugs interacted with a “receptive substance” that was essential for the initiation of the physiologic actions of the drug.

Modern Development

Langley's concept of “receptive substance” forms the basis of the modern concept of a receptor, but it was not accepted without debate. It took years of work by Clark (1885–1941) and Gaddum (1900–1965), among others, to solidify the receptor concept. Clark demonstrated that the relationship between drug concentration and the physiologic response formed a hyperbolic relationship (the familiar sigmoidal dose-response curve; see later). Clark concluded that the relationship arose from equilibration between the drug and its receptor and argued that the effect was directly proportional to the number of drug-receptor complexes. Ariëns (1918–2002) elaborated on Clark's theory and showed that the affinity of the drug for the receptor is distinct from the ability of the drug-receptor complex to elicit a physiologic response. This distinction was further elaborated by Stephenson (1925–2004), who mathematically defined and quantified efficacy— the propensity of a drug to elicit a response. Through his investigation of sympathomimetic compounds, Ahlquist (1914–1983) found that responses in various tissues occurred with two distinct orders of potency. This led him to propose multiple types of receptors for the same drug (α- and β-adrenergic receptors in this case), and the concept of specificity, which was finally published in 1948 after multiple rejections.

Ahlquist's work is the foundation of modern pharmacology, including the development of the first and still widely used receptor-specific drugs—β-blockers—by Black (1924–2010), who also developed H ₂ -histamine receptor blockers used to diminish stomach acid production in the treatment of peptic ulcer disease. Since Black's fundamental discovery, many receptors have been identified, structures of many drug-receptor complexes have been solved using x-ray crystallography and nuclear magnetic resonance (NMR), and the concept of drug-receptor interactions is now universally considered as the basis of the physiologic actions of drugs.

Drug Binding

Pharmacodynamics is broadly defined as the biochemical and physiologic effects of drugs. Proteins constitute the largest class of drug receptors, but other biomolecules can also be targeted. Proteins and other complex macromolecules can exist in a number of different conformational states. For simplicity, assume just two receptor conformations: physiologically active and inactive. In the case of an ion channel, for instance, the active conformation is an open conformation that allows ion permeation across the membrane, and the inactive conformation is the closed ion channel.

The following equation describes the relationship between the active and inactive states of the receptor (R):

R ^I denotes the inactive (closed) ion channel and R ^A denotes the active (open) ion channel, and k _a and k _i are rate constants for the forward and reverse conformational changes, respectively. In this example, rate k _i is higher than k _a (shown by arrow thickness) to illustrate a situation in which the channel is mostly closed in the absence of drug. The equilibrium relationship between the active and inactive conformations can be written as the ratio of the rate constants:

To initiate its pharmacologic action, a drug (D) first needs to bind its receptor”

In this example, drug D binds an active form of the receptor (R ^A ) to form the complex R ^A *D. As in the first example, this binding reaction has two rate constants, k ₁ and k ₂ , that dictate the rates of drug-receptor complex formation and dissociation, respectively. The equilibrium constant for this binding reaction is therefore the ratio of the rate constants (k ₁ /k ₂ ). The net effect of drug binding is an increase in active receptors as drug selectively binds the active conformation and thus prevents it from converting to the inactive conformation. This type of interaction of drug with a receptor is called agonism (discussed later). Fig. 1.1 shows a more general case, in which one drug (agonist) binds to an active (open) form of the ion channel while another drug (antagonist) selectively binds the closed form.

Although it might seem at first counterintuitive that even in the absence of agonist a receptor can be found in its active form, modern experimental methods such as single-channel patch clamp recordings can show this directly. An example of such a recording is shown in Fig. 1.2 .

This model of drug stabilizing a receptor in its active conformation nicely explains the actions of gamma-aminobutyric acid (GABA) on GABA _A receptors (see Fig. 1.2 ), but it is a very simplified view in several ways: (1) receptors can have more than two states (e.g., voltage-gated sodium ion [Na ⁺ ] channel). (2) Different conformational states can have different levels of activity rather than the all-or-none view presented here (e.g., nicotinic acetylcholine receptors). (3) Drugs can bind to more than one state of the receptor or at more than one site. However, this model of drug-receptor interactions serves as a foundation for building more sophisticated models. This simplified description is used in the following discussion to derive the basic pharmacologic concepts.

Rearranging the equilibrium expression for drug binding to receptor yields the following expression in which the ratio of the two rate constants is defined as the dissociation constant K _D :

Note that if K _D is small, then k ₁ ≫k ₂ and the complex of drug and its receptor is favored (as illustrated in Eq. 3 ). When the converse is true and K _D is large, the drug-receptor complex is not favored. Thus K _D reflects the propensity of the drug-receptor complex to break down. One can alternatively define affinity as the inverse of K _D , which reflects the propensity of the drug to form a complex with the receptor:

To illustrate the importance of the dissociation constant K _D, the parameter f (fraction of receptor occupied by drug) is first defined as

and then expressed f as a function of drug concentration and the K _D (or affinity) by substituting Eq. 4 :

K _D is a fundamental property of the drug-receptor interaction (given constant conditions such as temperature, pH, and so on) but can be different for different drug-receptor pairs. To illustrate the effect of differences in K _D on the formation of drug-receptor complexes, Eq. 7 is plotted for two drugs characterized by different values of K _D ( Fig. 1.3 ).

From Drug Binding to Physiologic Effect

Clark originally proposed that the number of drug-receptor complexes was directly proportional to the physiologic effect of the drug. Although this is not entirely correct, for simplicity, first assume Clark's theory to derive the basic concentration (dose)-effect (response) curve and illustrate potency. Then Clark's assumption will be relaxed to arrive at the notion of efficacy.

If a physiologic response is directly proportional to the fraction of bound receptors, one should be able to derive the concentration-response relationship simply from the binding curve illustrated in Fig. 1.3 . Indeed, the familiar sigmoid concentration-effect curve shown in Fig. 1.4 results from plotting the same equation as in Fig. 1.3 . The only difference is that drug concentration is plotted on a logarithmic rather than linear scale and the y -axis is labeled as “Fraction of maximal effect.”

Initially, as drug concentration increases, the increase in effect is rather small. In fact, until a certain concentration threshold is reached, no effect is apparent despite increasing drug concentrations. A further increase in drug concentration causes a steep increase in the effect until maximal effect is attained. This sigmoid relationship characterizes actions of many different drugs acting at different receptors. The circles in the plot denote drug concentrations at which half of the maximal effect is attained. This concentration is termed EC ₅₀ (effective concentration for 50% effect). Conceptually this is similar to K _D defined earlier. The major difference is that EC ₅₀ refers to the half maximal effect, w hereas K _D refers to half maximal binding . The smaller the EC ₅₀ , the less drug is required to produce the same effect. This is why EC ₅₀ is commonly used as a measure of drug potency or the ability of the drug to elicit a physiologic response.

The curve in Fig. 1.4 is derived from an abstract notion of equilibrium between bound and free receptors. It is totally independent of the chemical identity of the drug or the receptor—it reflects the general property of drug-receptor interactions and is fundamental to the understanding of the action of any drug.

Efficacy

The concentration-effect curves in Fig. 1.4 depend on the important assumption that the effect of the drug is proportional to the amount of receptor bound by the drug. This hypothesis makes very strong predictions: (1) given high enough concentration, all drugs will give the same maximal effect; and (2) the slope of the curve should be similar for all drugs acting on the same receptor. Indeed, the only difference between the red and blue curves in Fig. 1.4 is that the blue curve is shifted to the right. However, this is not always true, as shown by Stephenson in 1956 in a landmark study ( Fig. 1.5 ).

While investigating the pharmacodynamics of tetramethylammonium (TMA) compounds known to elicit muscle contractions, Stephenson observed that different response curves were not simply shifted versions of each other. Specifically, it appeared that maximal contraction was not always attainable even at the highest concentration of a drug. For instance, even at the highest concentration, octyl-TMA elicited contraction was only 40% of the maximal attainable, whereas 100-fold smaller concentrations of butyl-TMA elicited near maximal contraction (see Fig. 1.5A ). This observation alone does not invalidate Clark's theory. It is possible that octyl-TMA has really low affinity for the receptor and is therefore unable to elicit maximal response in the range of experimental concentrations.

The results in Fig. 1.5B definitively rule out this possibility. If the small contraction size elicited by octyl-TMA is due to its low affinity for the receptor, then some receptors will remain unoccupied even at maximal octyl-TMA concentration. If so, additional butyl-TMA should bind these available receptors and make the contraction maximal. Yet, in contrast to this prediction, addition of butyl-TMA did very little to the contraction elicited by octyl-TMA alone. Thus weak affinity of octyl-TMA for the receptor cannot explain submaximal response and Clark's theory therefore must be incomplete.

To explain these observations, Stephenson generalized Clark's theory by proposing that the response R is not directly proportional to the fraction of receptor bound by drug, but instead is some function F of the variable he referred to as stimulus S:

$R=F\left(S\right)$

where S is a product of the efficacy (e) and the fraction of the receptors occupied f :

$S=ef$

In the case of muscle contraction, F can be conceptualized as excitation-contraction coupling and efficacy as the ability of the drug-receptor complex to produce excitation. By substituting Eq. 7 into Eq. 9 , affinity A, drug concentration D, and efficacy e can be combined in the same equation:

$S=e\left(\frac{A\ast \left[D\right]}{A\ast \left[D\right]+1}\right)$

For conditions where the fraction of the occupied receptors is small, this simplifies to

$S=eAD$

Accordingly, even when the fraction of the occupied receptors is small, the observed physiologic effect can be quite large if the efficacy is high. Conversely, even if affinity is high but the efficacy is low, the overall effect can be quite low. Therefore the overall drug potency for a given system is a function of two variables that characterize drug-receptor interactions: affinity and efficacy .

Full Agonists, Partial Agonists, and Inverse Agonists

Drugs can be classified based on the features of their concentration-effect relationships. This section focuses on different kinds of agonists and the following section discusses different forms of antagonism. First, features of drug-receptor interactions that make a particular drug an agonist are defined. In schema Eq. 3 , it is assumed that drug binds only the active conformation of the receptor (see also Fig. 1.1 ). In a more general case (schema Eq. 12 ), drug can bind both active and inactive receptor conformations with different affinities.

The higher the affinity for the active conformation, the more equilibrium will be driven to the active receptor conformation until essentially all receptors are activated. This is called a full agonist. If the affinities for both active and inactive conformations of the receptor are comparable, the drug will be unable to convert a significant fraction of the receptor to the active conformation, even at high concentrations (reviewed for glutamate receptors). This drug is called a partial agonist.

This is a microscopic level description of the basis of drug agonism, but in most cases, there is no detailed understanding of the molecular events. It is difficult to measure experimentally the differences in affinity for different conformational states of a receptor. Usually this problem is solved by performing molecular dynamics simulations.

There is, however, a way to discover differences between agonists by characterizing their concentration-effect curves on a macroscopic level. Recall that the overall effect of a drug depends on two factors: affinity and efficacy. According to Eq. 10 , efficacy determines the maximal effect attainable at the limit of high drug concentration, and affinity determines the range of drug concentrations at which the steep portion of the concentration-effect curve occurs. Therefore the effect of drug affinity can be isolated by scaling the y -axis of the concentration-response curve to the maximal effect attainable for that drug, and differences in efficacy can be characterized by comparing maximal attainable effects.

Fig. 1.6A shows two drugs that are distinguished by their affinity (higher for the red drug) scaled relative to the maximal effect attainable for each drug. When the effect of each drug is plotted relative to the absolute maximal effect (see Fig. 1.6B ), it becomes evident that although the red drug has higher affinity, it has lower efficacy with a maximal response of one-third of that attainable by the blue drug. Therefore the red drug is a partial agonist, whereas the blue drug is a full agonist. Although the shapes of the plots in Figs. 1.6A and B appear quite different, in fact the relationship between their EC ₅₀ values stays exactly the same regardless of how the data are plotted.

Up until this point we made an implicit assumption that concentration-response curves start at zero. In other words, in the absence of drug there is no effect. In Fig. 1.6C this assumption is relaxed. The plot in Fig. 1.6C shows the behavior of a system exhibiting intrinsic receptor activity, even in the absence of drug. This occurs in systems where even in the absence of drug a significant number of receptors exist in their active conformations. The blue curve shows a full agonist and the red curve shows a partial agonist. When the black drug is added, it appears that the intrinsic activity of the receptor is diminished. This can occur if the drug has a higher affinity for the inactive conformation of the receptor. This drug-receptor interaction is called inverse agonism; an inverse agonist is a drug that has a negative efficacy . If the inverse agonist was added after adding the full agonist, the overall effect would be diminished, suggesting that the inverse agonist is an antagonist (see later). In fact, the distinction between an antagonist and an inverse agonist can be subtle and is often evident only in genetically modified systems that express constitutively active receptors. For instance, the commonly used β-blockers such as propranolol are in fact inverse agonists at β-adrenergic receptors.