Artificial theta stimulation impairs encoding of contextual fear memory

Abstract

Several experiments have demonstrated an intimate relationship between hippocampal theta rhythm (4-12 Hz) and memory. Lesioning the medial septum or fimbria-fornix, a fiber track connecting the hippocampus and the medial septum, abolishes the theta rhythm and results in a severe impairment in declarative memory. To assess whether there is a causal relationship between hippocampal theta and memory formation we investigated whether restoration of hippocampal theta by electrical stimulation during the encoding phase also restores fimbria-fornix lesion induced memory deficit in rats in the fear conditioning paradigm. Male Wistar rats underwent sham or fimbria-fornix lesion operation. Stimulation electrodes were implanted in the ventral hippocampal commissure and recording electrodes in the septal hippocampus. Artificial theta stimulation of 8 Hz was delivered during 3-min free exploration of the test cage in half of the rats before aversive conditioning with three foot shocks during 2 min. Memory was assessed by total freezing time in the same environment 24 h and 28 h after fear conditioning, and in an intervening test session in a different context. As expected, fimbria-fornix lesion impaired fear memory and dramatically attenuated hippocampal theta power. Artificial theta stimulation produced continuous theta oscillations that were almost similar to endogenous theta rhythm in amplitude and frequency. However, contrary to our predictions, artificial theta stimulation impaired conditioned fear response in both sham and fimbria-fornix lesioned animals. These data suggest that restoration of theta oscillation per se is not sufficient to support memory encoding after fimbria-fornix lesion and that universal theta oscillation in the hippocampus with a fixed frequency may actually impair memory.

Figure 1. Successive coronal sections through the rat brain illustrating the extent and location of the smallest (black) and largest (dotted) fimbria-fornix lesion.

Numbers show the A-P distance from bregma. The coronal section 0.8 mm from bregma also illustrates the knife lesioning technique used; note that lesioning both in 45° and 70° angles were made bilaterally. VHC, ventral hippocampal commissure.

Figure 2. Schematic illustration of the experiment in the time frame of the entire experiment (A), the test session (B), and initial fear conditioning (C).

After surgical operations and adaptation period the animals underwent testing phase in which the driving current was adjusted. In the test phase, first, in Context A baseline freezing (preshock) was recorded for 3 min with artificial theta stimulation delivery, which was then followed by fear conditioning with three repeated foot shocks. Fear memory was tested by measuring freezing time first 24 h later in the same familiar context (now marked as Context A1), then 26 h later in a totally novel context (B), and finally 28 h later again in the familiar context (A2).

Figure 3. Example of histological slices verifying the effect of fimbria-fornix lesioning on hippocampal cholinergic system.

In the fimbria-fornix lesioned animals (B) a substantial loss of cholinergic innervations was revealed by AChE staining compared to sham operated animals (A). The scale bar represents 1.0 mm and the columns and error-bars mean ± SEM, *** P<0.001, (sham, n = 12; FFX, n = 12).

Figure 4. The EEG analysis revealed the effect of the fimbria-fornix lesioning on theta frequency range (4–12 Hz, marked with a horizontal bar) in the septal CA1 area during exploration (A) and freezing (B).

The lines represent the spectral analysis of EEG of the sham operated (n = 7) and the FFX (n = 6) group. The error-bars represent SEMs, and asterisks denote frequency bands with significant differences at the level P<0.01. (C) Zoomed-in presentation of the same power analysis of sham-operated animals during exploration as in (A). The vertical line denotes the frequency of ATS (8.0 Hz) that was slightly above the mean peak frequency (7.8±0.2 Hz) of sham animals during exploratory behavior. (D) Examples of filtered (2–12 Hz) hippocampal EEG related to exploratory behavior in a sham-operated and a fimbria-fornix lesioned animal during both control situation and artificial theta stimulation (ATS). The occurrence of stimulation pulses is shown on top of the sample EEG sweep.

Figure 5. Freezing behavior expressed as percentage of time spent freezing during 5 minutes in the contexts used in the experiment; preshock, the conditioning chamber (A), the novel chamber (B) and re-exposure to the conditioning chamber (A2).

Dunnett’s two-sided post-hoc test was used to compare each test group with the non-stimulated sham group as a controlThe columns and error-bars represent mean ± SEM, * P<0.05, ** P<0.01, *** P<0.001.

Figure 6. We used the early gene Egr-1 protein product as a molecular marker of the activation of memory consolidation pathways upon re-exposure to the shock context.

The number of Egr-1 positive cells was measured in the left dorsal hippocampal dentate gyrus area (DG), in the lateral amygdala (LA) a. The scale bar in the figures A and B represents 1000 µm and in the inserts 200 µm, * P<0.05.