An in vitro model of maintained drug self-administration

Andrew Norman, Mantana Norman, Vladimir Tsibulsky

University of Cincinnati, Cincinnati, Ohio, USA

The drug self-administration paradigm in animals represents a useful model of addiction in humans. A pharmacokinetic/pharmacodynamic (PK/PD) model was proposed to explain the observed dose-response relationship for cocaine during maintained self-administration in rats (Tsibulsky and Norman, 1999). The core of the in vivo model is equation (1): T = ln(1+Du/Dst)/k, which relates the time between successive drug injections (T) to three parameters: drug unit dose (Du), a minimum maintained drug concentration (Cmin, termed the satiety threshold (Dst) in animals) and the drug first-order elimination rate constant (k). Despite its success, the mechanistic implications of this model are not widely appreciated. We herein provide a simple in vitro model of the PK/PD principles that we hypothesise determine drug self-administration behaviour in animals and, potentially, humans.

The set up of the proposed “thought experiment” consists of three main parts: 1) an isolated tissue that contracts in response to an agonist. A smooth muscle preparation, such as guinea pig ileum or rat vas deferens would suffice. The muscle preparation is maintained in an organ bath in a physiological solution under conditions necessary to maintain the functional integrity of the tissue. The tissue is attached to a lever that indicates the magnitude of the contraction/relaxation of the tissue strip. 2) An actuator, upon contact by the lever, activates a syringe pump that delivers a unit dose of agonist. 3) The agonist is continuously washed out of the organ bath by a constant infusion of physiological solution, which mimics first-order elimination kinetics. When the muscle relaxes to a specified length the lever contacts the actuator, which results in the injection of a dose of agonist into the organ bath. The muscle contracts, raising the lever, disconnecting the actuator. As the agonist concentration declines by washout the tissue relaxes until the lever again contacts the actuator, a dose of agonist is injected and the muscle contracts again. The cycle of agonist injections repeats indefinitely. Competitive receptor antagonists typically increase the rate of agonist self-administration in animals. Antagonists could be added to the physiological solution and the mechanisms underlying the acceleration of agonist self-administration behaviour can also be illustrated by this system.

The system prevents the agonist concentration from falling below a minimum, thereby titrating Cmin, which represents an equipotent agonist concentration. As competitive antagonists increase the equipotent agonist concentration, in the presence of an antagonist, the system will maintain a higher Cmin resulting in a shorter T at a given unit dose even if k is unaltered. Therefore, the antagonist-induced acceleration of agonist self-administration is caused by the more rapid decline in the concentration of agonist at the higher agonist concentrations as dictated by first order elimination kinetics.

This simple model illustrates that an apparently complex animal behaviour can be explained in terms of basic pharmacological principles. The importance of agonist and antagonist concentration in determining behavioural responses is ignored by alternative models of drug self-administration behaviour. Our model provides a rational basis for investigating the mechanisms underlying drug self-administration behaviour and for the development of pharmacotherapeutic strategies for the treatment of drug addiction.

Tsibulsky VL and Norman AB (1999). Brain Res 839: 85-93.