CREB regulation of nucleus accumbens excitability mediates social isolation-induced behavioral deficits

Here, we characterized behavioral abnormalities induced by prolonged social isolation in adult rodents. Social isolation induced both anxiety- and anhedonia-like symptoms and decreased cAMP response element–binding protein (CREB) activity in the nucleus accumbens shell (NAcSh). All of these abnormalities were reversed by chronic, but not acute, antidepressant treatment. However, although the anxiety phenotype and its reversal by antidepressant treatment were CREB-dependent, the anhedonia-like symptoms were not mediated by CREB in NAcSh. We found that decreased CREB activity in NAcSh correlated with increased expression of certain K 1 channels and reduced electrical excitability of NAcSh neurons, which was sufficient to induce anxietylike behaviors and was reversed by chronic antidepressant treatment. Together, our results describe a model that distinguishes anxiety- and depression-like behavioral phenotypes, establish a selective role of decreased CREB activity in NAcSh in anxiety-like behavior, and provide a mechanism by which antidepressant treatment alleviates anxiety symptoms after social isolation. Depression and anxiety are common forms of mental illness in the general population. Although they are classified as distinct syndromes by the Diagnostic and Statistical Manual (American Psychiatric Association), symptoms of depression and anxiety often occur together and to widely varying extents in different subtypes of the illnesses. Despite the importance of these clinical phenomena, very little is known about the distinctions between depression- and anxiety-like symptoms in animal models 1 . Models of ‘active’ stress, such as foot shock, restraint stress, social defeat and learned helplessness, produce depressionand anxiety-like phenotypes; the molecular mechanisms of these models have been extensively studied, but specific molecular mediators of depression versus anxiety symptoms have not yet been described 2‐4 . Even less well studied, however, is a ‘passive’ model of stress and social isolation in adulthood, which, as with active stress, mimics aspects of human depression and anxiety 5,6 . This lack of attention is unfortunate, as social isolation would appear to be particularly relevant to certain subtypes of human depression and anxiety disorders 7,8 . Although social isolation has been studied, most models to date have focused on adulthood behaviors after isolation rearing early in life, either as pups or adolescents, which is a very different model than adulthood social isolation 5 . Reports on adulthood isolation provide