Environmental novelty is signaled by reduction of the hippocampal theta frequency

Abstract

The hippocampal formation (HF) plays a key role in novelty detection, but the mechanisms remain unknown. Novelty detection aids the encoding of new information into memory-a process thought to depend on the HF and to be modulated by the theta rhythm of EEG. We examined EEG recorded in the HF of rats foraging for food within a novel environment, as it became familiar over the next five days, and in two more novel environments unexpectedly experienced in trials interspersed with familiar trials over three further days. We found that environmental novelty produces a sharp reduction in the theta frequency of foraging rats, that this reduction is greater for an unexpected environment than for a completely novel one, and that it slowly disappears with increasing familiarity. These results do not reflect changes in running speed and suggest that the septo-hippocampal system signals unexpected environmental change via a reduction in theta frequency. In addition, they provide evidence in support of a cholinergically mediated mechanism for novelty detection, have important implications for our understanding of oscillatory coding within memory and for the interpretation of event-related potentials, and provide indirect support for the oscillatory interference model of grid cell firing in medial entorhinal cortex.

Figure 1

Effect of novelty on theta frequency. a) Experimental set-up. b) Theta frequency increases with days of experience in environment ‘a’ (days 1-5), and change to an environment within a common setting produces an even greater initial decrease in frequency than first exposure to environment ‘a’ (cf. frequency in environments ‘b’ and ‘c’ on days 1*-3*, black points, to frequency in environment ‘a’ on days 1-3, grey points). The difference in theta frequency from ‘baseline’ (i.e. the mean frequency over days 2-5 for a given rat) is shown for each trial. c) The novelty effect is shown most clearly when each day’s trials within familiar environment ‘a’ (grey points) or novel environments ‘b’ and ‘c’ are averaged (black points). Points show the mean across rats, error bars show s.e.m.

Figure 2

Examples of EEG-electrode recording sites in anterodorsal CA1 (a, b, c) and dorsal subiculum (d), showing photomicrographs for specific rats (see Figures 3a and 4a). Thick arrows point to tissue damage and other indicators of the track made by a given tetrode. Thin arrows point to likely recording locations. The scale bar in A applies to all photomicrographs, and does not take into account tissue shrinkage. a) Location of electrode used in Rat 1 (throughout the experiment; CA1 stratum radiatum). b) Location of electrode used in Rat 2 (CA1 in/around pyramidal layer). c) Location of electrode used in Rat 3 (Days 1-4; CA1 in/around pyramidal layer). d) Location of electrode used in Rat 3 (Days 5, 1*-3*; deep subiculum).

Figure 3

Effect of novelty on theta frequency, no effect of running speed. a) Theta is visibly faster in novel environment ‘b’ (trial 3, black lines) than familiar environment ‘a’ (trial 2, grey lines) on day 1*: individual traces from each rat (left, asterisk marks 10th cycle); power spectra for whole trial (right). b) The effect of novelty on theta frequency does not reflect variations in running speed. Data from each trial were sub-sampled to remove any differences in median speed across trials in the sub-sampled data. Mean theta frequency is shown versus days of experience in familiar environment ‘a’ (grey points) or novel environments ‘b’ and ‘c’ (black points) in sub-sampled data for which median speed is constant across trials. Points show the mean across rats, error bars show s.e.m. See Supplementary Online Material for details.

Figure 4

Theta frequency (a) and median running speed (b) for individual rats and trials. Data from the familiar environment ‘a’ shown in grey, data from the unexpected novel environments ‘b’ and ‘c’ shown in black. See Figure 1 for details of the environments used in each trial.