Abstract
AbstractTheta phase precession is thought to confer key computational advantages (e.g. temporal compression suiting spike-timing related plasticity, cognitive relations as phase distances, and population-level coding for directions and sequences). However, direct evidence speaking to: 1) its widely-theorised role in enhancing memorability; 2) its dependence upon sensory input, is lacking. We leveraged the Vector trace cell (VTC) phenomenon to examine these issues. VTCs in subiculum show a simple, unambiguous memory correlate: VTCs remember the distance and direction to a cue after the cue is removed, with a new ‘trace field’ which was not present before the cue was inserted. Regarding memorability, here we show that subsequently-remembered cue fields (those which become trace fields) exhibit higher levels of phase precession than subsequently-forgotten cue fields (those which produce no trace). Thus, phase precession does appear to enhance memorability, consistent with long-established theory. The second issue concerns the extent of phase precession in sensory-elicited vs memory-dependent firing. Phase precession in CA1 is strongly disrupted following deprivation of its Entorhinal, but not CA3, inputs; this could indicate that theta phase precession is largely sensory-driven and absent in memory-dependent fields. Here, however, we show that phase precession is robust in subicular VTC trace fields, i.e. when the cue that originally elicited the new vector field is no longer present. Thus, the much-theorised benefits of phase precession likely apply to memory-dependent fields. These findings have wide implications for oscillatory-based models of memory.
Publisher
Cold Spring Harbor Laboratory