Cognitive impairment in epilepsy: the role of network abnormalities

Abstract

The challenges to individuals with epilepsy extend far beyond the seizures. Co-morbidities in epilepsy are very common and are often more problematic to individuals than the seizures themselves. In this review, the pathophysiological mechanisms of cognitive impairment are discussed. While aetiology of the epilepsy has a significant influence on cognition, there is increasing evidence that prolonged or recurrent seizures can cause or exacerbate cognitive impairment. Alterations in signalling pathways and neuronal network function play a major role in both the pathophysiology of epilepsy and the epilepsy comorbidities. However, the biological underpinnings of cognitive impairment can be distinct from the pathophysiological processes that cause seizures.

Keywords: accelerated forgetting; cognition; memory; oscillation; place cells; rate coding; reactivation; replay; temporal coding.

Fig. 1

To assess the effects of interictal spikes on memory rats underwent intra-hippocampal pilocarpine injections to induced SE. Bilateral electrodes were placed in the ventral hippocampus (A). In the Delayed-Match-To-Sample test during the sample step, one of two levers is randomly presented (right or left) and is pressed by the rat (B). Then, in the Delay step, the rat has to poke its nose into a hole in the opposite wall for a random length of time (6–30 seconds). After this time period has elapsed, the first nosepoke into the hole turns off the stimulus light above and extends both levers. Then, in the Match step, the rat has to remember which lever he pressed during the sample phase, and press that same lever again to procure a food reward. During the sampling stage memory is encoded, during the delay phase memory is maintained, and during the match phase memory is retrieved. Performance is recorded for trials without spikes (C, top trace) and trials with spikes (C, bottom). Among trials in which an IIS occurs during the encoding or maintenance epoch of short-term memory, accuracy does not differ from trials without IIS (D). However, IISs during the retrieval phase produces a marked decrease in accuracy. Increasing delays produce decreases in accuracy, regardless of IIS epoch timing. Modified from Kleen et al., (Kleen et al., 2010) with permission. .

Fig. 2

Schematic of SE induced changes that occur over seconds to months. Similar changes can occur with spontaneous seizures.

Fig. 3

Schematic representation of SE induced changes in theta and single cell firing. A. As rat runs around the track several place cells fire. Place cells with partially overlapping place fields (red and blue) will show a different firing relationship on whether the field centers are distant or close to each other. There is a linear relationship between the inter-field distance and the time required to go from the field center to another (running time, T). Because each cell firing is doing phase precession (middle trace), there is also a relationship between the running time (T) and the time interval between APs within the same theta cycle (t). Cells with close fields will fire at short time interval whereas cells with distant fields will fire at long time interval. In the SE rats normal phase precession does not occur and the time interval between APs (t) is variable. As shown in B, the control rat has phase precession whereas the SE rats show phase procession. This abnormality in phase precession is reflected in an abnormal compression of temporal sequences in SE rats. Compared to control pairs of neurons (B1), which show a strong correlation between running-time (T) and theta-time lags (t), SE pairs showed a greater variability as expressed by the total variance about the regression line. Cross correlograms for all control and SE pairs ordered in increasing running time (from -1000 ms to 1000 ms) (B2). Each horizontal line represents the cross-correlogram of an individual pair of place cells with the amplitude represented by a color code. In control pairs, the diagonal band near the zero lag line shifts as running-time lag increases. In contrast, this pattern is much less obvious in SE pairs. C. Rats with SE also have impaired speed-theta frequency correlations. Whereas control rats have an increase in frequency of theta with increasing speed, this is not the case with the SE rats (C1). As shown in the right panel (C2), performance in a spatial memory task is related to the speed-frequency relationship with SE rats with a poor speed-theta frequency showing impaired performance compared to controls. Modified from Lenck-Santini et al. (Lenck-Santini and Holmes, 2008) and Richard et al. (Richard et al., 2013) with permission.

Fig. 4

Example of hippocampal theta modulation of PFC units. Rat is running on a linear track. Two place cells from the hippocampus are recorded as evidenced by increased firing rates of cells as the animal ran through the place fields. APs from the cells are indicated by vertical blue and red lines positioned above the theta rhythm. In the PFC the APs are temporally linked to the APs in the hippocampus although there is a time delay between the APs in the hippocampus and those in the PFC. The modulation of PFC APs by hippocampal theta occurs even in the absence of PFC theta. In a T-maze animals have to make either a forced turn (top figure) or decided whether to go left or right to receive a food award (cocoa puffs). The theta wave can be drawn as both a straight line as well as a circle to represent the 360O of the theta wave. When the animal has a forced turn there is a preferred phase (straight line within the circle) for the APs to occur in both CA1 and PFC. When the animal has to make a decision there is a much stronger phase lock of PFC neurons with the hippocampal theta as evidenced by a thicker line. Figures are based on work by Siapas et al. (Siapas et al., 2005) and Jones et al. (Jones and Wilson, 2005).

Fig. 5

Relationship of coherence with cognition. EEG spectral coherence represents the consistency of the phase difference between two EEG signals when compared over time. In a delayed non-match to sample study in rats with a history seizures electrodes were placed in the PFC and hippocampus. Coherences between the waveforms in the hippocampus and PFC were calculated. Coherence is a measure of synchronization or coupling between two EEG signals and is based mainly on the conformity of phase differences between the EEG signals. The two waves shown from the PFC and hippocampus are of the same amplitude and frequency, but there is a phase shift. Phase differences are typically measured in degrees where a complete cycle is 360 degrees. In this example, the phase difference is approximately 30 degrees. Coherences are dynamically measured when controls and rats with early-life seizures (ELS) are doing the test. Data on left is time-linked to the sample press, and data on the right to the match press, revealing functional differences between control (blue) and ELS (red) rats when the envelopes diverge. ELS rats showed increased theta and gamma coherences following the sample press. The lower panel show dynamic CA1-PFC theta coherence in correct and incorrect trials. In trials with 20-30 second delays, control rats did not show performance-related differences in CA1-PFC coherence between correct (light blue) and incorrect (dark blue). In trials with 20-30 second delays, ELS rats showed increased CA1-PFC coherence in correct trials (light red) relative to incorrect trials (dark red) particularly around the time of the sample press and prior to the match press (arrows). Modified from Kleen et al. (Kleen et al., 2011b) with permission.