Memory Storage Fidelity in the Hippocampal Circuit: The Role of Subregions and Input Statistics

In the last decades a standard model regarding the function of the hippocampus in memory formation has been established and tested computationally. It has been argued that the CA3 region works as an auto-associative memory and that its recurrent fibers are the actual storing place of the memories. Furthermore, to work properly CA3 requires memory patterns that are mutually uncorrelated. It has been suggested that the dentate gyrus orthogonalizes the patterns before storage, a process known as pattern separation. In this study we review the model when random input patterns are presented for storage and investigate whether it is capable of storing patterns of more realistic entorhinal grid cell input. Surprisingly, we find that an auto-associative CA3 net is redundant for random inputs up to moderate noise levels and is only beneficial at high noise levels. When grid cell input is presented, auto-association is even harmful for memory performance at all levels. Furthermore, we find that Hebbian learning in the dentate gyrus does not support its function as a pattern separator. These findings challenge the standard framework and support an alternative view where the simpler EC-CA1-EC network is sufficient for memory storage.Author Summary: It is well known that the hippocampus, a mammalian brain region, has a crucial role in memory formation. Furthermore, it has a remarkable anatomical structure and can be divided into several subregions based on physiological properties. Over the last decades a widely accepted model has evolved suggesting individual roles for each subregion in memory storage. The central idea is that region CA3, with its remarkably many synapses that project into the region itself again, stores the memories within these synapses. In this model a memory is impressed onto the hippocampus by neuronal activation in the hippocampal input region. Recently it has been found that such activations have a certain regularity instead of being random as assumed in the standard model. Here we investigate how well the model performs when storing memories of regular inputs. We find that the proposed function of CA3 actually harms memory performance. Moreover, we show that this function is redundant even in the case of random inputs. These findings call the standard model into question and support an alternative view of how the hippocampus may store memories.