THE ROLE OF THE DENTATE GYRUS AND PERIRHINAL CORTEX IN PATTERN SEPARATION OF NON-SPATIAL MEMORIES (presentación oral)

MIRANDA MAGDALENA

Congreso; Reunión Conjunta de Biociencias; 2017

Successful memory involves not only remembering information over time but also keeping memories distinct and less confusable. The ability to separate the components of memories into distinct memory representations is believed to rely on pattern separation, a computational process by which differences are amplified to disambiguate similar events. By transforming similar experiences into discrete non overlapped-representations, pattern separation enhances the possibility of accurate encoding and retrieval.Despite the importance of this mnemonic function, the molecular mechanisms and signals necessary for the behavioral manifestations of this process remain unknown.Work in the field has suggested that this crucial memory function may be localized to the dentate gyrus (DG) of the hippocampus. However, there is some debate on whether DG granule cells are a domain general pattern separator that allows the differentiation of all kinds of information. Most tasks used to evaluate pattern separation have involved contextual or spatial manipulation, while pattern separation could, in principle, occur during encoding of representations other than spatial, for example for object representations. In this regard, the perirhinal cortex (PRH) is involved in the acquisition and storage of object memories, and it was shown to be crucial for the resolution of tasks with ambiguous features. In this regard, the medial temporal lobe (MTL) ?memory system? that supports episodic memory, consists of several anatomically related structures, including the hippocampus (HC) and perirhinal cortex (PRh). While all these structures function in a mutually interdependent manner, they do not form a functionally homogeneous system, since damage to PRh in rats impairs visual object recognition but not spatial memory, whereas HC damage has the opposite effect. Considering that there could be pattern separation in the spatial domain and in the object domain,these two functions could also depend on different regions of the brain.In this context, our objective is to study the role of the PRH and DG for the behavioral "pattern separation" of object memories and to evaluate the molecular mechanisms involved in this process. Our hypothesis is that non spatial pattern separation might rely more heavily on the PRH than in the DG, but similar molecular mechanisms could be engaged in both structures. We use a modified version of the non-spatial object recognition task to assess the behavioral use of these pattern-separated representations, and manipulated the load of pattern separation during encoding of the task by changing the similarity of the encoded items. To evaluate the requirement of specific plasticity-related proteins in the PRH, antisense oligonucleotides were injected in the PRH of adult male rats with implanted cannulas to inhibit the expression of specific transcripts during or after task acquisition. Through this manipulations we were able to show that the expression of two plasticity-related proteins, the neurotrophic factor BDNF and the immediate early gene Arc, are required for pattern separation of object memories in the PRH and are also required in the DG to solve this task. In addition, we found that Arc is an effector of BDNF for the storage of separate representations of objects in the PRH.Although molecular mechanisms for both hippocampal-dependent and perirhinal-dependent memories have been shown previously, these have almost always been dependent on delay, which suggests a role specifically in persistence but does not indicate a specific effect on behavioral pattern separation or in keeping memories distinct and less confusable. Here we found deficits that are dependent on the load of pattern separation of the task for both DG and PRH. In addition, this raises the question of whether plasticity mechanisms inthe DG are involved in pattern separation of non-spatial information. In this regard, we found that while exogenous BDNF is able to enhance pattern separation when injected into PRH, it is not effective when infused within the DG, indicating that while both structures are necessary, they seem to serve different functions.In the same manner, the AMPA antagonist DNQX, which blocks excitatory transmission, impairs overlapping object memory retrieval when injected in the PRH, but not in the DG. These results could suggest that the DG might not be the structure in which the unique representations for objects are stored, but rather the PRH.Finally, BDNF injected into PRH is able to rescue the memory deficit produce by blockade of Arc in the DG.Moreover, disconnection experiments reveal that the interaction between PRH and DG is necessary for the disambiguation of overlapping object memories. In conclusion, these results reveal a complex interaction between plasticity mechanisms in the PRH and the DG for non spatial pattern separation and posit the PRH as the key structure where unique and non-confusable object representations are stored.Nevertheless, the expression of Arc in the PRH seems to be an essential step for the discrimination of overlapping object memory representations, but not for spatial representations, while the DG seems to play a domain general role in consolidation pattern separated memories.