The hippocampus is known to play a critical role in processing information about the temporal context in which our experiences occur. However, it remains unclear how hippocampal oscillations are associated with this type of information processing, and how their functional organization is influenced by connectivity gradients. Here we investigate these issues by examining local field potential activity recorded across the proximodistal axis of CA1 as rats performed a complex odor sequence memory task. More specifically, we took advantage of the paradigm’s precise time-locking and diverse cognitive demands to evaluate the correspondence between specific patterns of oscillatory activity and specific forms of information processing. Our first main finding is that spectral content differed between behavioral states (odor sequence processing epochs vs running epochs) and that, for each behavioral state, the power of recruited oscillations showed significantly distinct gradients along the proximodistal axis. Odor sequence processing epochs were characterized by increased power in the 4-8 Hz and 20-40 Hz range, with 20-40 Hz oscillations showing a power gradient increasing toward proximal CA1. In contrast, running epochs were characterized by increased power in the 8-12 Hz range and in a broad but modest increase in power across higher frequency ranges (>24 Hz), with power gradients increasing toward proximal and distal CA1, respectively. Our second main finding is the discovery that 20-40 Hz oscillations were specifically and significantly modulated by sequence memory performance. We found that 20-40 Hz power increased with knowledge of the sequence and carried trial-type-specific information. These results link 20-40 Hz oscillations with trial-specific processing of nonspatial information critical for order memory judgments, which is consistent with the demonstrated role of the hippocampus in processing temporal relationships among events. The prominence of such oscillations in proximal CA1 is also consistent with evidence that medial entorhinal input is important for the processing of temporal information in CA1.

Data was initially collected using a Plexon MAP system and then processed in Matlab. Data (behavioral and neural) from the original .plx files was organized in a format used by the Fortin lab and collaborators and stored as a .mat file. Behavioral and neural data are stored in separate files with a common organization. For each session the data was arranged in an NxM matrix where rows indicate a timestamp and columns represent different events and/or data collected at that time. Further description of the data organization can be found in the ReadMe and at https://github.com/FortinLab/statMatrixNeuralAnalysis.

Associated code located at https://github.com/FortinLab/Gattas_et_al_2020. A readme (Filenames README.txt) is included to indicate the naming convention for the associated files.