Hippocampal ensembles represent sequential relationships among discrete nonspatial events

Abstract

The hippocampus is critical to the temporal organization of our experiences, including the ability to remember past event sequences and predict future ones. Although this fundamental capacity is conserved across modalities and species, its underlying neuronal mechanisms remain poorly understood. Here we recorded hippocampal ensemble activity as rats remembered a sequence of nonspatial events (5 odor presentations unfolding over several seconds), using a task with established parallels in humans. Using novel statistical methods and deep learning techniques, we then identified new forms of sequential organization in hippocampal activity linked with task performance. We discovered that sequential firing fields ("time cells") provided temporal information within and across events in the sequence, and that distinct types of task-critical information (stimulus identity, temporal order, and trial outcome) were also sequentially differentiated within event presentations. Finally, as previously only observed with spatial information, we report that the representations of past, present and future events were sequentially activated within individual event presentations, and that these sequential representations could be compressed within an individual theta cycle. These findings strongly suggest that a fundamental function of the hippocampal network is to encode and preserve the sequential order of experiences, and use these representations to generate predictions to inform decision-making. 2/2415/24

CA1 20-40 Hz oscillatory dynamics reflect trial-specific information processing supporting nonspatial sequence memory

“…One possibility is that 20-40 Hz power reflects a degree of match between the stimulus presented and the stimulus predicted by the animal (based on its knowledge of the sequence). This possibility is well aligned with the hypothesized role of CA1 acting as a comparator between internal representations retrieved from the CA3 and external cues transmitted via the entorhinal cortex (e.g., Hasselmo and Wyble, 1997; Lisman and Grace, 2005) and is consistent with the strong InSeq/OutSeq differentiation observed in spiking activity at the single-cell (Allen et al, 2016) and ensemble (Shahbaba et al, 2019) level. The learning-related power increase we observed is also consistent with this view, as stimulus predictions should improve with learning (resulting in stronger matches on InSeq+ trials).…”

“…In addition to spatial information, accumulating evidence indicates that the hippocampus plays a key role in the processing of temporal information. Consistent with its unique architecture and connectivity (McNaughton and Morris, 1987; Lisman 1999; Foster and Knierim 2012; Buzsáki and Tingley 2018), a growing literature shows the hippocampus is critical for remembering sequences of nonspatial events (Fortin et al 2002; Kesner et al 2002; Allen and Fortin, 2013; Eichenbaum, 2014) and that hippocampal neurons code for temporal relationships among such events (MacDonald et al, 2011; Allen et al 2016; Shahbaba et al, 2019). However, little is known about the oscillatory dynamics associated with this fundamental type of information processing in the hippocampus.…”

Mixing memory and desire: How memory reactivation supports deliberative decision-making.

“…Such investigations will need to pay special attention to the timecourse of retrieval during early learning, which may be dramatically different in dynamics and content from the kind of online reactivation that occurs after many experiences with a task or learning set (Redish, 2016) , and especially when divorced from spatial navigation, the pace of which can confound investigations of the frequency of retrieval of related place field representations. Along these lines, one important recent study examined these dynamics in a non-spatial setting, examining "lookahead" during sequences of odors in well-trained rodents (Shahbaba et al, 2019) . Using a novel combination of decoding methods to identify odor identity representations in dorsal CA1, the authors found that they were able to decode anticipatory sequence reactivations on the scale of a few 100s of milliseconds, consistent with the theta-band rhythms observed in spatial navigation studies.…”

Time as the fourth dimension in the hippocampus

“…In another non-spatial task, a pre-trained sequence of five odors was presented; in the test phase, rats had to report the relative order of two of the odors (Shahbaba et al, 2019). Time cells spanned the presentations of the respective odors (as well as the delays afterwards).…”

“…Conversely, in error trials, inaccurate cue combinations were represented on descending theta phases and erroneous choices were represented on ascending phases. In a task where rats sampled two odors in sequence, and re sponded about their relative order in a test template sequence of five odors, Shahbaba et al (2019) found theta sequencing of past, present and future events.…”