S.04. The development of inverse agonists for CNS s103 Today the decision to bring a new drug to clinical development is mainly based on primary pharmacology data from experimental animal models. These animal models are designed to express specific mechanisms, believed to be involved in clinical pain. By modulating these mechanisms pharmacologically, specific targets related to e.g. receptors or enzymes can be identified, hence com- prising a lead target for the development of a new pharmacological entity. However, animal models are approximations and there are difficulties in defining what clinical pain condition the endpoint from these animal models are predictive for. First, our knowledge about which mechanisms, that are involved in specific clinical pain conditions is still limited. Second, we do not know how well mechanisms expressed in the animal models relate to the mechanisms seen in the clinical pain conditions. This lack of linking between experimental pain mechanisms and clinical pain condition represent a major hurdle for the development of new drugs for pain treatment, because it implies that each new compound needs to be tested in clinical pain studies before its value for pain treatment can be known. An increasing number of new compounds are introduced as potential analgesics, but only a limited number of these will turn out to be effective. Clearly, we need to develop new strategies and tools that will increase the predictability of early efficacy data from animals towards the clinical pain condition. One such strategic decision could be to use experimental human volunteer models, where pain related mechanisms can be expressed and measured. Basically, human models of pain related mechanisms are build up in the same way as in the animal models. The concept is to apply a standardised trauma, resulting in reproducible symptoms and signs, reflecting mechanisms that are believed to be involved in clinical pain. Evidence that these mechanisms can be modulated pharmacological by drug classes known to be effective in clinical pain treatment will be supportive. One obvious limitation is that the trauma in humans needs to be temporary, hence only functional changes can be induces to the nerve system. A number of different human pain models have been described in the literature, inclu- sive evidence of the involved mechanism. Experimental animal and human models need to be harmonised in terms of type of injury, mechanisms and endpoints, and if similar pharmacological sensitivity can be obtained across species, this will indicate that the experimental mechanism or target addressed in animals also applies to humans. Even though we still need to link these models to the clinical pain conditions, the use of experimental human models may be a step in the right direction, by reproducing the mechanism- based evidence from animal studies in humans. Information from human volunteers of the lowest dose that is active at the pursued mechanism, onset and duration of action will be valuable for the design of clinical dose-response studies. Furthermore, as a result of the fast advancing gene technology, we may soon be facing a situation where new drugs are developed for specific human targets, without the possibility of testing efficacy in animal models. Well-validated experimental human models will then be neces- sary to address mechanism of action, providing mechanistically proof of the principle. Finally, by having these mechanism based models available, there is an opportunity, via a minor preclinical toxicology pro- gram, to perform single dose experiments with a number of candidate drugs, and in that way providing qualitative input on mechanism, efficacy, side effects and simple DMPK data of value for selecting the best compound for drug development. S.04. The development of inverse agonists for CNS Is.04.011 Receptor models of inverse agonism H.O. Onaran. Ankam University Faculty of Medicine, Dept. Pharmacology and Clinical Pharmacology, Ankara, Turkey Ligands of hormone or neurotransmitter receptors in classical pharmacology have been classified as agonist or antagonist, de- pending on their ability to produce a biological stimulus upon binding to the relevant receptor (i.e depending on their efficacy). Those ligands that initiate a biological signal horn the receptor are named agonist and those that do not as antagonist, assuming that the unligated receptor is silent in terms of its biological activity. However, experimental studies carried out in the last decade, especially on the G protein-coupled receptors, have shown otherwise; receptors, in the absence of ligands, may well possess a biological activity, which can be inhibited by some antagonists. Considering that the inhibition of constitutive receptor activity is a biological signal as well, these ligands have been defined as inverse agonist (or negative antagonist, for their negative efficacy). Here, I briefly reviewed how inverse agonism (or efficacy in general) is formulated in available receptor activation models at the molecular level, and discussed the implication of different formulations. The most macroscopic one of the available receptor models is the ternary complex model (TCM), which has been particularly meant for those receptors that do not possess an intrinsic signaling activity that can be measured experimentally (such as ionic con- ductance or enzymatic activity), but can be read out by a signal transducer such as G protein. According to TCM, receptor and the transducer bind each other with a certain affinity M, which can be modified by ligand binding to the receptor. The affinity M accounts for the basal activity of the receptor, and the degree of change of M upon receptor ligation accounts for the molecular efficacy of ligand. Those ligands that increase, leave unchanged or decrease M, are recognized as agonist, neutral antagonist or inverse agonist, respectively (1). The ability of ligands to modify M is implicitly attributed to a ligand-induced conformational change in the receptor molecule. The next level of microscopicity is the set of allosteric models where the underlying conformational change in the receptor molecule has been made explicit. The simplest form of these models is the two-state model, where receptor is considered in two allosteric conformations (i.e. active; R* and inactive; R) in equilibrium with an equilibrium constant of J (defined as [R*]/[R]), which can be modified by ligand binding. Again, the constant J accounts for the basal activity of receptor, and the degree of change of J by ligation, for the ligand efficacy. Likewise, those ligands that increase, leave unchanged or decrease J, are recognized as agonist, neutral antagonist or inverse agonist, respectively. In the special case of G protein coupled receptors the active form of receptor is considered as the one that can bind G protein (2). Finally, the ultimate level of microscopicity is the multistate stochastic model of receptor action, where the conformational plasticity of the receptor molecule was modeled as a collection of indefinite number of microscopic states. There, the macroscopic state of the receptor is defined as the distribution of microscopic states, which can be modified by ligand binding. Those ligands that induce a “target distribution” in the receptor molecule are considered as full agonist, and all other ligands that induce a distribution deviating from the target one are considered