Technical Comments

Response to Comment on "On the Origin of Interictal Activity in Human Temporal Lobe Epilepsy in Vitro"

See allHide authors and affiliations

Science  25 Jul 2003:
Vol. 301, Issue 5632, pp. 463
DOI: 10.1126/science.1085812

We are glad that our study has stimulated work on human epilepsy and note that Wonzy et al. (1) confirm our finding that a spontaneous interictal activity originates in the subiculum. Their data suggesting that the subiculum of a nonsclerotic temporal lobe generates synchronous activity are intriguing. We wonder whether the diagnosis and site of the epileptic focus was comparable in all of their tissue samples. Our tissue (2) was obtained uniquely from patients with mesial temporal lob epilepsy (MTLE) in which a hippocampal sclerosis was diagnosed by clinical imaging and confirmed by inspection of slices. We therefore could not test the relation between the degree of sclerosis and subicular activity. Rather, we suggested that changes initiated in the subiculum by deafferentation resulting from CA1 cell loss might underlie subicular synchrony.

Did the tissue classified as nonsclerotic in (1) show no CA1 cell loss? We note that hippocampal sclerosis was defined on the Wyler scale (3). Apparently, tissue scored as 1 or 2 on this scale may be “mildly or even moderately sclerotic” with a cell loss of up to 50%. The Wozny et al. data seem to suggest that the probability of observing rhythmic activity was correlated with the degree of sclerosis. If that is the case, perhaps our suggestion should be modified to state that even a loss of less than 50% of CA1 cells may induce changes in the subiculum resulting in an interictal-like activity.

We find the data on neuronal afterhyperpolarizations interesting and agree with Wozny et al. that changes in the intrinsic properties of subicular cells may contribute to an interictal rhythmicity. We would like to clarify one misstatement of our previous work. In tissue from patients with hippocampal sclerosis, we never saw synchrony in the CA3 and CA1 regions or in the dentate area. The data in (1) suggest some differences, especially in the contribution of GABAergic signaling to the rhythmic activity. We would be pleased to try to resolve these differences, which may result from divergent techniques or definitions, by joint experiments in Berlin or in Paris. Several questions may help clarify our differences.

We question how synchronous activity was assessed in (1). We feel that two recordings— either intracellular or extracellular— are essential to define synchrony (4), and we always used two to four extracellular electrodes (5) to explore the subiculum. How large was the region of synchrony in tissue from patients with or without sclerosis in (1)? In (2), rhythmic activity was generated within a region of variable size (approximately 2 × 3 to 5 mm) within, but smaller than, the subiculum. Single cells were always recorded within this zone and all neurons showed intracellular events— depolarizing or hyperpolarizing or mixed— correlated with the interictal-like synchrony. It is possible that the cells in which Wozny et al. detected no rhythmic events were located in subicular regions that did not participate in an interictal-like activity.

We also find it difficult to determine whether the interictal-like synchrony in the tissue obtained from the patients in (1) was suppressed by GABAergic antagonists. Since this was the case for the tissue recorded in (2), we concluded that a defect existed in the GABAergic signaling pathway. We further question how Wozny et al. can be sure that GABAergic signaling in the epileptic subiculum is never depolarizing. Our tests were limited to the subicular cells that fired with rhythmic interictal-like events (2). How many of the cells tested in (1) fired rhythmically? We examined responses to focal stimuli in the presence of glutamatergic receptor antagonists and found depolarizing GABAergic events in 6 out of 10 cells tested. Based on our data and the cells in (1) that displayed a spontaneous rhythmic activity, we estimate that one or two of the cells recorded in sclerotic tissue by Wozny et al. might display depolarizing GABAergic responses. However, this sample size may not be large enough to support a negative conclusion.

In conclusion, the new data presented in (1) suggest that temporal lobe epilepsy is an evolving pathology. Apparently, even a smaller cell loss than the Wyler threshold for sclerosis initiates changes which suffice to construct a pathological subicular network. Presumably, these changes continue to develop during the lifetime of a patient.


Stay Connected to Science

Navigate This Article