Alcohol use despite negative consequences is a core phenomenon of alcohol addiction. We recently used alcohol self-administration that is resistant to footshock punishment as a model of this behavior, and found that activity of PKC delta + GABAergic neurons in the central amygdala (CeA) is a determinant of individual susceptibility for punishment resistance. In the present study, we examined whether activation of GABA(B) receptors in CeA can attenuate the activity of PKC delta + neurons in this region, and whether this will result in suppression of punishment- resistant alcohol self-administration in the minority of rats that show this behavior. Systemic administration of the clinically approved GABA(B) agonist baclofen (1 and 3 mg/kg) dose- dependently reduced punishment-resistant alcohol self-administration. Bilateral microinjections of baclofen into CeA (64 ng in 0.3 mu l/side) reduced the activity of PKC delta + neurons, as measured by Fos expression. This manipulation also selectively suppressed punished alcohol self-administration in punishment-resistant rats. Expression analysis indicated that virtually all CeA PKC delta + neurons express the GABA(B) receptor. Using in vitro electrophysiology, we found that baclofen induced hyperpolarization of CeA neurons, reducing their firing rate in response to depolarizing current injections. Together, our findings provide a potential mechanism that contributes to the clinical efficacy of baclofen in alcohol addiction. Therapeutic use of baclofen itself is limited by problems of tolerance and need for dose escalation. Our findings support a mechanistic rationale for developing novel, improved alcohol addiction medications that target GABA(B) receptors, and that lack these limitations, such as e.g., GABA(B) positive allosteric modulators (PAM:s).

For ethical and technical reasons, research in humans has some limitations and requires the support of animal models. Numerous animal models have been developed over the years to study alcohol consumption and model alcohol-related behaviors in several species, including non-human primates, rodents and more recently zebrafish, fruit flies and C. elegans. In this chapter, we provide an overview of the most commonly used animal models of alcohol use disorder (AUD) and discuss their pros and cons. We classify animal models of AUD into two main categories, operant and non-operant paradigms, which covers behavioral procedures developed to model several aspects of human addiction, including primary alcohol reinforcement, physical dependence, loss of control over alcohol intake, progressive choice of alcohol over healthy rewards and relapse. Finally, we will conclude and discuss about other important aspects of human addiction, including interindividual differences, sex differences and social factors, that need to be incorporated into preclinical models of AUD to improve their translational value.