The hippocampus and pavlovian fear conditioning: Reply to Bast et al.

In a recent article that appeared in Hippocampus, we reviewed findings supporting a mnemonic role for the dorsal hippocampus (DH) in Pavlovian (contextual and tone) fear conditioning (Anagnostaras et al., 2001). We also detailed a view that has emerged over the years from this work that suggests that the hippocampus plays a highly selective role in the acquisition and temporary storage of contextual representations, as opposed to a role in conditional stimulus-unconditional stimulus (CS-US) associations or in permanent storage for which the amygdala has been heavily implicated (Kim and Fanselow, 1992; Phillips and LeDoux, 1992; Young et al., 1994; Maren and Fanselow, 1996; Maren et al., 1996,1997, 1998; Anagnostaras et al., 1999). Because the evidence that DH lesions produce a temporally graded retrograde amnesia selective for contextual fear that accords well with declarative memory deficits in amnesic humans, we have further argued this may be a good model system with which to study the transformation of memory from a hippocampus-dependent to a hippocampus-independent (cortical) state (i.e., consolidation) (Scoville and Milner, 1957; Squire, 1992; Squire and Alvarez, 1995; Hodges and Graham, 1998; Squire et al., 2001; Murre et al., 2001; Frankland et al., 2001). In a letter to Hippocampus regarding our recent article, Bast et al. (2001b) expand on our review to discuss their recent data (and the data of others), focusing in particular on findings from lesions of the ventral hippocampus (VH) and discussing how these are problematic for the view we presented. Although a specific hypothesis on the role of the VH in fear conditioning has not yet been formulated, several interesting findings were reviewed, emphasizing, in particular, the effects of complete or VH lesions on both tone fear conditioning and on remotely acquired fear (e.g., Mumby et al., 1999; Sutherland et al., 2001). These findings are in contrast to DH lesions, in which a severe and selective deficit for recently acquired contextual fear, but not for tone or remotely acquired fear, is typically found (Kim and Fanselow, 1992; Phillips and LeDoux, 1992; Maren et al., 1997; Anagnostaras et al., 1999). In this reply, we address and expand on some of the issues raised by Bast et al. (2001b). Although this can help clarify where agreements lie in some of the empirical findings, we feel there is not yet enough experimental evidence to offer a specific role for the VH in fear conditioning, and it is not yet clear whether these findings challenge the views offered in our review. However, we agree that, as the interface between the DH and amygdala, future work in the area of VH will be essential to our understanding of the neural circuits involved in contextual fear conditioning.