Loading field map
How does a drug molecule, binding fleetingly to a protein on the cell surface, produce a measurable change in the behavior of a whole organism? This question has driven receptor theory and signal transduction for nearly a century. The core tension is between the simplicity of the binding event and the complexity of the cellular response. Early pharmacologists assumed a direct proportionality: more binding meant more effect. But anomalies—partial agonists that never produced a full response, spare receptors that amplified tiny signals, and drugs that blocked without binding to the same site—forced the field to develop increasingly sophisticated models. Each new framework refined, replaced, or coexisted with its predecessors, and several remain in productive tension today.
The first systematic framework, Occupancy Theory, was built on the mass-action law of physical chemistry. A. J. Clark and later J. H. Gaddum proposed that a drug's effect is proportional to the fraction of receptors it occupies. The model assumed a simple equilibrium: drug + receptor ⇌ drug–receptor complex → response. It successfully predicted the shape of dose-response curves and introduced the concept of affinity—how tightly a drug binds to its receptor. But anomalies soon appeared. Partial agonists, such as the early beta-blocker dichloroisoproterenol, occupied receptors fully yet produced only a submaximal response. Some tissues required only a tiny fraction of receptors to be occupied to elicit a maximal response (spare receptors), contradicting the proportionality assumption. These puzzles set the stage for later frameworks that would separate binding from efficacy.
While Occupancy Theory focused on the receptor itself, Earl Sutherland's discovery of cyclic AMP (cAMP) in the 1950s revealed that the receptor is only the first step in a cascade. The Second Messenger Paradigm showed that an extracellular signal (the first messenger) activates a receptor, which then triggers intracellular second messengers—cAMP, calcium ions, inositol trisphosphate—that amplify and diversify the signal. This framework did not replace Occupancy Theory; rather, it absorbed it as the starting point. The receptor's occupancy still mattered, but the response now depended on the state of the entire signaling network. The paradigm also explained how a single drug could produce different effects in different cell types: the same receptor could couple to different intracellular pathways. This insight became infrastructure for all later models, which assume that the receptor is embedded in a dynamic signaling context.
In the 1960s, W. D. M. Paton proposed a radical alternative: the effect of a drug depends not on how many receptors are occupied at equilibrium, but on the rate at which drug molecules associate with and dissociate from receptors. Rate Theory challenged the equilibrium assumptions of Occupancy Theory. It explained why some drugs produced a rapid, transient response that faded even while the drug remained present (tachyphylaxis). However, the theory struggled to account for sustained responses and was mathematically cumbersome. By the early 1970s, most pharmacologists had abandoned it as a standalone model. Yet its kinetic emphasis did not disappear entirely: it influenced later work on receptor desensitization and the time-dependent behavior of signaling pathways, and it remains a minority perspective in some areas of ion-channel pharmacology.
The Two-State Model, developed by Jean-Pierre Changeux and others, proposed that receptors exist in an equilibrium between an inactive state (R) and an active state (R). Agonists stabilize the active state, inverse agonists stabilize the inactive state, and neutral antagonists have equal affinity for both. This model elegantly explained partial agonism: a partial agonist stabilizes R less effectively than a full agonist, so the equilibrium shifts only partway. It also introduced the concept of constitutive activity—receptors spontaneously adopting the active state even without a drug—which was later confirmed experimentally. The Two-State Model did not fully replace Rate Theory; instead, it absorbed the idea that receptors are dynamic, but it framed that dynamics as a conformational equilibrium rather than a kinetic one. It also set the stage for multi-state thinking by showing that even two states could account for phenomena that Occupancy Theory could not.
As molecular biology revealed that many receptors (especially G-protein-coupled receptors, GPCRs) signal through heterotrimeric G proteins, the Ternary Complex Model emerged. It added a third component to the equilibrium: the receptor, the G protein, and the drug form a ternary complex (drug–receptor–G protein). This model explained why guanine nucleotides (which dissociate G proteins) shift agonist binding curves—a phenomenon that had puzzled pharmacologists. It also provided a mechanism for inverse agonism: an inverse agonist stabilizes the receptor in a conformation that has lower affinity for the G protein, reducing basal signaling. The Ternary Complex Model did not discard the Two-State Model; rather, it refined it by specifying that the active state is actually a receptor–G protein complex. The two models coexisted, with the Ternary Complex Model dominating GPCR pharmacology while the Two-State Model remained useful for ligand-gated ion channels.
James Black and his colleagues developed the Operational Model to address a persistent problem: how to separate the drug's affinity from its efficacy (the ability to produce a response) in a way that is independent of the tissue being studied. The model introduces a transducer function that links receptor occupancy to response, parameterized by a transducer constant (τ, tau). A high τ means the tissue amplifies the signal efficiently; a low τ means the tissue is a poor amplifier. This allowed pharmacologists to compare drugs across different tissues and to predict how a drug would behave in a new system. The Operational Model reconciled insights from both Occupancy Theory (affinity) and Rate Theory (the importance of the tissue context). It coexists with the Ternary Complex Model: the Operational Model describes the overall input-output relationship, while the Ternary Complex Model provides a molecular mechanism for the transducer step. Today, the Operational Model remains the standard tool for quantifying agonism in drug discovery.
By the early 2000s, it became clear that a single receptor can activate multiple signaling pathways (e.g., G-protein-dependent and β-arrestin-dependent pathways) and that different agonists can stabilize different receptor conformations, each favoring a particular pathway. This phenomenon, called biased agonism (or functional selectivity), challenges the assumption that efficacy is a single property of a drug. Instead, a drug can be a full agonist for one pathway and a partial agonist or even an antagonist for another. The dopamine D2 receptor provides a concrete example: some antipsychotic drugs preferentially activate β-arrestin signaling over G-protein signaling, which may explain their therapeutic profile with fewer side effects. Biased agonism extends the Two-State Model into a multi-state conformational ensemble, and it refines the Operational Model by requiring separate efficacy parameters (τ) for each pathway. The relationship between the two frameworks is one of pluralism: the Operational Model provides the mathematical language for quantifying bias, while the conformational ensemble provides the physical picture. Biased agonism is now a major focus of GPCR drug discovery, though predicting bias from chemical structure remains an open challenge.
The most recent framework, Systems Pharmacology, steps back from the receptor to model the entire signaling network. It uses computational methods—ordinary differential equations, network analysis, and machine learning—to simulate how perturbations at the receptor propagate through multiple pathways, feedback loops, and cross-talk with other receptors. This framework does not replace earlier models; rather, it contextualizes them. The Operational Model's τ parameter, for example, becomes a network-level property that depends on the expression levels of many proteins, not just the receptor and its immediate transducer. Systems Pharmacology also addresses phenomena that earlier models could not, such as drug resistance arising from network rewiring or the paradoxical effects of blocking a single node in a robust network. Its distinctive commitment is that the appropriate level of analysis is the system, not the receptor. This has led to productive tension with reductionist approaches: Systems Pharmacology advocates argue that even a perfect receptor model will fail if it ignores network context, while traditional pharmacologists counter that the receptor model is still necessary for understanding the first step of drug action. The two perspectives coexist as complementary levels of explanation.
Today, four frameworks remain actively used: the Second Messenger Paradigm (as infrastructure), the Operational Model of Agonism, Biased Agonism, and Systems Pharmacology. They agree on several points: efficacy is not a fixed property of a drug but depends on the cellular context; receptors are dynamic entities that sample multiple conformations; and a complete understanding requires linking molecular events to tissue-level responses. They disagree on the appropriate level of analysis. The Operational Model and Biased Agonism focus on the receptor and its immediate transducers, providing precise, testable parameters. Systems Pharmacology argues that these parameters are only meaningful within a network model that includes feedback, redundancy, and time delays. The most active debate concerns whether biased agonism can be predicted from receptor structure or must be measured empirically for each drug-pathway pair. This tension drives current research: computational chemists try to design biased ligands in silico, while experimental pharmacologists develop high-throughput assays to measure pathway-specific efficacy. The field has not resolved these questions, but the coexistence of multiple frameworks has made it more powerful than any single model could be.