Activation of the septohippocampal system differentiates anxiety from fear in startle paradigms

It has been suggested that different brain areas are involved in the modulation and expression of fear and anxiety. In the present study we investigated these potential differences by using the fear-potentiated-startle (FPS) and lightenhanced-startle (LES) paradigms to differentiate between fear and anxiety, respectively. Male Wistar rats were tested in the FPS and LES paradigm and perfused 1 h after the test session. Fos immunoreactivity (IR) was quantified in 21 brain areas and compared between FPS, LES and four control conditions. Both FPS and LES procedures significantly enhanced the acoustic startle response. A principal component analysis of Fos-IR-data showed that 70% of the changes in Fos-IR could be explained by three independent components: an arousal-component, identifying brain areas known to be activated under conditions of vigilance, arousal and stress, a LES- and an FPS-component. The LES component comprised the septohippocampal system and functionally interrelated areas including nucleus accumbens, anterior cingulate cortex, lateral habenula and supramammillary areas, but not the dorsolateral part of the bed nucleus of the stria terminalis. The central amygdaloid nucleus and the dorsolateral part of the bed nucleus of the stria terminalis loaded exclusively on the FPS component. Analysis of the separate brain areas revealed significantly higher Fos-IR in LES relative to FPS in the anterior cingulate cortex, nucleus accumbens shell, lateral septum, lateral habenula and area postrema. We conclude that the neural circuitry activated during FPS and LES shows clear differences. In anxiety as induced by LES, activation of the septohippocampal system and related areas seems to play a major role. In fear as induced by FPS, the central amygdaloid nucleus and the dorsolateral part of the bed nucleus of the stria terminalis loaded on the same component, but Fos-IR observed in these brain regions did not differentiate between anxiety and fear. Furthermore, principal-component analysis appears a useful tool in detecting and describing correlated changes in patterns of neuronal activity.