This study evaluated the role of
This study evaluated the role of SERT in classical fear conditioning using a SERT knockout rat model. We found a clear disruption of cue-conditioned fear in SERT−/− rats, as measured with the fear-potentiated startle. This disruption was due to a deficit in fear acquisition and was accompanied by development of enhanced contextual conditioned fear. Neither the acquisition nor the glycine transporter of fear-potentiated startle was affected by pharmacological SERT inhibition. Moreover, both the deficit in acquisition of cue-conditioned fear and the development of exacerbated contextual conditioned fear in SERT−/− rats could be reversed by treatment with the CRF1 receptor antagonist CP154,526 during acquisition. The failure to learn the cue-shock contingency resulted in increased contextual fear in SERT−/− rats. This was specifically related to the fear learning deficit, as SERT−/− rats show normal unconditioned anxiety in the light-enhanced startle paradigm (Supplement 1). Furthermore, it is unlikely that the fear learning deficit is a result of increased contextual fear, because results from Experiment 4 show that the prevention of contextual fear by 30mg/kg CP-154,526 did not result in reinstatement of the cued fear expression. In humans, unawareness of the cue-shock contingency also increases contextual fear (Baas et al., 2008), a phenomenon that is specifically associated with maladaptive fear processing in panic disorders patients (Grillon et al., 2007b, Lissek et al., 2009). Interestingly, these patients shown normal startle responding in unconditioned measures of fear and anxiety (Grillon et al., 1994, Melzig et al., 2007). Thus, the current findings in SERT−/− rats fit well with aforementioned human data and resemble both fear- and anxiety-related potentiated startle data in panic disorder patients. Interestingly, several studies indicate altered SERT functioning in panic disorder (Esler et al., 2007, Strug et al., 2010), further strengthening the link between SERT availability and fear learning deficits in anxiety disorders. To the best of our knowledge, this is the first report on the effect of chronic SSRI treatment on the acquisition and expression of fear-potentiated startle in rodents. The lack of effect on expression of cued fear is in line with the single human fear-potentiated startle study available (Grillon et al., 2009). The finding that neither acute nor chronic paroxetine treatment affected the acquisition and expression of cued conditioned fear suggests that the fear learning deficit found in SERT−/− is due to developmental changes and does not result from compromised SERT functioning during fear learning. Serotonin is known to modulate neurodevelopment (Gaspar et al., 2003, Lauder, 1990) and the neural systems regulating anxiety-like behavior are especially sensitive to changes in serotonergic functioning during early development (Gross et al., 2002, Vinkers et al., 2010). It has already been reported that SERT−/− mice show increased spine density and excitatory drive of the BLA (Wellman et al., 2007). Such changes in neuronal structure and excitability may also be responsible for the learning deficit observed in SERT−/− rats. However, the exact mechanism underlying this deficit in SERT−/− rats is subject of further investigation. Interestingly, acute administration of paroxetine during fear acquisition did increase overall startle responding 24h later, which suggests that acute SERT inhibition potentiates the acquisition of contextual conditioned fear. To the best of our knowledge, this is the first time the acute anxiogenic effect of SSRIs in humans, as measured with fear-potentiated startle, is shown in rodents (Burghardt et al., 2004, Grillon et al., 2007a). Together with findings from our concurrent study in humans that SERT S/S carriers show exaggerated contextual fear learning (Heitland et al., 2013), these results suggest that increased extracellular serotonin levels may also be responsible for the increased contextual conditioned fear seen in SERT−/− rats.