Serotonin, Stress, and Conditioning

Abstract

This special issue of Biological Psychiatry focuses on the roles of serotonin (5-HT) in specific brain structures implicated in emotions and stress and their potential relevance to psychological disorders. Historically, clinical observations made in the 1950s from psychotic patients that were administered reserpine, a vesicle storage inhibitor of mono-amines with antihypertensive and sedative properties, established the importance of monoamines in moods (1). Subsequent studies, in particular those examining monoamine metabolites in the brains of suicide victims and findings of anomalous 5-hydroxyindoleacetic acid (the chief metabolite of 5-HT) levels, helped usher the modern view that 5-HT is an important indolamine neurotransmitter/neuromodulator system regulating affects and moods. Recent developments in scientific methods, ranging from molecular biology to neuroanatomy to functional imaging, have enabled the detailed characterization of the serotonergic system in the brain. These advances continue to sculpt the view that understanding the contributions of the serotonergic system to behavior and cognition is imperative for the development of successful treatments for various psychological ailments (e.g., anxiety, depression, fear, and stress disorders) that profoundly impact the quality of human life. In recent decades, stress and fear conditioning in animals (particularly rodents) have been accepted as de facto model systems for gaining insights into various anxiety and fear-related disorders in humans (2,3). In typical stress studies, effects of aversive experiences (such as restraint, anxiogenic drugs) are examined on behavioral (e.g., anxiety, memory) and physiological (e.g., the hypothalamic-pituitary-adrenal [HPA] axis reactivity, hippocampal neurogenesis) measures of interest. In classical or Pavlovian fear-conditioning studies, initially neutral conditioned stimuli (CSs) (such as tone, light, or context) are contingently paired with aversive unconditioned stimuli (USs) (such as electric shock, loud noise) that reflexively evoke unconditioned fear responses (URs). Through CS-US association formation, the CS becomes capable of eliciting conditioned responses (CRs) that resemble innate fear responses. Repeated presentations of a CS (unreinforced) lead to a decrease in CRs (extinction). Although both stress and conditioned fear-evoked behavioral-autonomic-hormonal responses are adaptive (normal) components of the animal’s defense mechanisms, because they share similar characteristics to clinical symptoms of fear- and stress-related disorders, many researchers practice the view that neural mechanisms mediating augmented stress and conditioned fear responses in animals correspond to the root of human disorders. Modern investigations centering on the prefrontal cortex (PFC) and the amygdala as the crucial neural structures involved in the control of stress and fear responses originated from a case study of Phineas Gage in the mid-1800s (4) and the work of Kluver and Bucy on monkeys in 1939 (5). In general, the amygdala is thought to be involved in evoking fear and stress responses, while the PFC exerts (inhibitory) control over the amygdala-mediated defensive behaviors. Consistent with this notion, experimentally induced dysregulations of PFC and amygdalar activities have been shown to alter stress and fear responses in rodents (6). Because amygdala, PFC, and other forebrain structures are innervated by the raphe nuclei (namely the dorsal region) and because 5-HT is implicated in affects and moods (1), serotonergic drugs have garnered substantial interest in stress, fear conditioning, and clinical studies. Although the PFC-amygdalar hypothesis of stress and fear conditioning is considerably supported by animal and human studies, there are unresolved critical issues. A major issue is the precise function of PFC in fear conditioning. Initial studies that examined the ventral medial PFC (vmPFC) (infralimbic and prelimbic cortex) lesion effects on fear conditioning found both a selective impairment of extinction (but not acquisition) of CS-induced fear response (7) and absence of effects on extinction (as well as conditioned inhibition) (8). A recent study further found that the medial PFC (mPFC) is not necessary for acquisition, expression, and extinction of conditioned fear (9). In lieu of these inconsistent results, the recent excitement that the PFC-amygdalar interaction plays a crucial role in fear conditioning/extinction warrants further studies. In this special issue, Bissiere et al. (pages 821–831, in this issue) investigated the role of a specific subregion of the rostral anterior cingulate cortex (rAmy-ACC) that is reciprocally connected with the basolateral amygdala (BLA) (the putative site of associative fear memory formation) in fear conditioning and found that both chemical lesions and reversible inactivations confined to this specific PFC region impaired the acquisition of the conditioned fear to a tone (but not context) CS. In previously fear-conditioned rats, however, pretesting inactivations of rAmy-ACC did not affect the expression of conditioned tone fear. Furthermore, the rAmy-ACC–BLA pathway was confirmed to be glutamatergic, and activation of the rAmy-ACC (via pharmacologically blocking local gamma-aminobutyric [GABA] inhibition) was found to facilitate the acquisition of auditory (but not contextual) fear conditioning. In another study in this issue, Meloni et al. (pages 832–839, in this issue) discovered that the corticotrophin-releasing factor (CRF)-enhanced startle (a behavioral assay of stress) was accompanied by an increased c-Fos expression in dorsal raphe neurons that project mainly to the infralimbic (IL) region of the mPFC, suggesting topographical activation-inactivation coupling of dorsal raphe and IL, respectively. Collectively, these findings highlight the importance of considering specific subregions of ACC and vmPFC exerting differential influences on fear conditioning and stress. Although stress and fear conditioning seem to be useful model systems for gaining insights into human disorders, some caveats must be considered. First, the observations that the symptomatic manifestation of stress and fear-related disorders in humans closely resemble stress and conditioned fear responses in animals do not necessarily indicate that the underlying neuronal mechanisms are the same (post hoc, ergo propter hoc). That is, just as rise in body temperature (fever) has multiple causative factors (cold, stomach aches, etc.), different neural mechanisms could contribute to experimentally enhanced stress/fear responses and anxiety disorders. Closely related to this issue is the significance of the enhanced stress and fear responses. In fear conditioning, for example, stronger conditioned fear (as indexed by response magnitude, resistance to extinction) develops in rats receiving multiple US presentations than those receiving a single US. Therefore, the robustness of conditioned fear after multiple US likely reflects the formation of stronger fear memory. If an experimental manipulation produces augmented conditioned fear (e.g., Bissiere et al. study), then this too may reflect quantitative differences in fear memory. At present, however, there is no convincing evidence that experimentally induced alterations in fear responses (that putatively model human fear disorders) differ from normal fear responses qualitatively at the neurobiological level. A similar concern is applicable to models of stress disorders. Another critical issue is the lack of clear distinction between stress and fear conditioning in many animal studies. Often, the aversive stimuli used in stress experiments produce fear responses. Conversely, because fear conditioning elevates corticosterone and CRF levels, it is also referred to as stress. However, the learned helplessness literature provides compelling evidence that fear conditioning is not sufficient for producing stress effects (10). With these issues in mind, further exploration of the contribution of 5-HT on these models will be critical in clarifying the role of the serotonergic system on the regulation of emotions, moods, and stress, which clearly has broad clinical implications for individual and social well being.