alcohol-containing liquid diet for 6 months followed by a 1-week withdrawal period produces working memory impairment in a T-maze spontaneous alternation task in mice, which persists for at least 6 weeks after the cessation of alcohol intake (31, 47). Moreover, withdrawn mice displaying impaired working memory performance were those that had the lowest pCREB level in the PFC along with a persistent rise of prefrontal corticosterone concentration. Because glucocorticoids in the PFC interact with β-adrenoceptor–cAMP/PKA activity to influence working memory function (92), one route by which elevated glucocorticoid levels may impair PFC-mediated cognitive function long after the cessation of alcohol exposure is by inhibiting the cAMP–PKA cascade. In this context, growing evidence supports a central role for PDE, which is responsible for the breakdown of cAMP, in the regulation of alcohol drinking in rodents [for review, see Ref. (93)]. For example, treatment with various PDE4 inhibitors, including rolipram, produces long-lasting reduction of alcohol intake and preference in C57BL/6J mice (94). Chronic rolipram treatment also results in sustained reduction of alcohol seeking and consumption in alcohol-preferring rats (95, 96). As mentioned earlier, mice subjected to 1- or 6-week alcohol withdrawal from chronic alcohol consumption exhibited working memory impairments accompanied by enhanced
