Mapping the spatio-temporal pattern of the mammalian target of rapamycin (mTOR) activation in temporal lobe epilepsy

Abstract

Growing evidence from rodent models of temporal lobe epilepsy (TLE) indicates that dysregulation of the mammalian target of rapamycin (mTOR) pathway is involved in seizures and epileptogenesis. However, the role of the mTOR pathway in the epileptogenic process remains poorly understood. Here, we used an animal model of TLE and sclerotic hippocampus from patients with refractory TLE to determine whether cell-type specific activation of mTOR signaling occurs during each stage of epileptogenesis. In the TLE mouse model, we found that hyperactivation of the mTOR pathway is present in distinct hippocampal subfields at three different stages after kainate-induced seizures, and occurs in neurons of the granular and pyramidal cell layers, in reactive astrocytes, and in dispersed granule cells, respectively. In agreement with the findings in TLE mice, upregulated mTOR was observed in the sclerotic hippocampus of TLE patients. All sclerotic hippocampus (n = 13) exhibited widespread reactive astrocytes with overactivated mTOR, some of which invaded the dispersed granular layer. Moreover, two sclerotic hippocampus exhibited mTOR activation in some of the granule cells, which was accompanied by cell body hypertrophy. Taken together, our results indicate that mTOR activation is most prominent in reactive astrocytes in both an animal model of TLE and the sclerotic hippocampus from patients with drug resistant TLE.

Figure 1. Distribution of pS6 immunoreactivity (IR) in the hippocampus of control mice and 6 hours after induction of seizures.

(A-C) In saline-injected control mice (n = 4), the density of pS6 IR varies in the different hippocampal subfields. In dentate gyrus (DG), pS6 IR is weak and only present in some of granule cells (A). In the pyramidal cell layer, pS6 IR is detectable in the entire CA3 subfield (B) and strong in the CA1 region (C). Hippocampus 6 hours post-SE (n = 4) shows significantly increased pS6 IR within the dentate gyrus (D) and CA3 subfields (E) compared with control hippocampus, but pS6 IR is absent in CA1 (F). Insets in D-F show increased expression of pS6 (GFAP-positive, red) in neurons (NeuN-positive, green) of epileptic hippocampus than control hippocampus (A-C). Neuronal death is indicated by nuclear pyknosis. (G) Quantification of pS6 positive neurons in hippocampus at 6 h after seizures (n = 4) versus control mice (n = 4). Error bars are means ± SD., **P<0.01; ***P<0.001 is based on Student’s t-test. ND, not detectable. ML: molecular layer; HS: hippocampal sclerosis; KA, kainic acid. Scale bar in A-F: 200 µm; in insets: 30 µm.

Figure 2. Distribution of pS6 in the hippocampus of control mice and 5 days after induction of seizures.

In CA1 of saline-injected control mice (n = 4), pS6 (green) is present in the cytoplasm of pyramidal neurons (A). Immunofluorescent staining of GFAP (red) shows distribution and morphology of astrocytes under normal physiological conditions (B). There is no colocalization of pS6 and GFAP (C). During the latent phase, increased pS6 expression is observed in the CA1 subfield of KA-injected hippocampus (n = 4) (D), and colocalizes with GFAP, mostly in astrocyte-like cells (arrows in F). Reactive gliosis is indicated by dramatically elevated GFAP (E). pS6 positive glia are present in hippocampus both ipsilateral (ipsi, H) and contralateral (contr, I) to KA injection, although elevation of the former is more significant. Compared with saline-injected control hippocampus (G), dentate gyrus of hippocampus ipsilateral to KA injection (H) contains fewer granule cells with strong pS6 IR (arrow in H), while almost no pS6 positive granule cells are present in the contralateral, non-injected hippocampus (I). (K) Quantification of GFAP positive and pS6/GFAP double positive cells in CA1 subfield of KA-injected mice (n = 4) versus control mice (n = 4) at the second stage. Error bars are means ± SD., **P<0.01; ***P<0.001 is based on Student’s t-test. Scale bar in A-F: 100 µm; in G-I: 200 µm.

Figure 3. pS6 expression in the sclerotic hippocampus (HS) and non-sclerotic controls (non-HS).

Non-sclerotic hippocampus displays weak pS6 IR in glial cells of the hilus (A), while in CA1 and CA3, staining is mainly observed in the cytoplasm of pyramidal cells, and only a few glia are pS6-positive (a-d). In the sclerotic hippocampus, a large number of astrocytes (B) with significantly increased pS6 IR are observed in dentate gyrus (DG) subfields including the hilus (B) and the molecular layer (e). Astrocytes with up-regulated pS6 invade the dispersed granular layer (C). Immunofluorescence image shows colocalization of pS6 (green) with GFAP (red) in astrocytes surrounded by neurons (Inset in C). Glial proliferation and strong pS6 IR are also detected in hilus, CA1 and CA3 of subjects with HS (f-h). (E) Quantification of astrocyte-like pS6 positive cells in CA1, CA3, hilus and DG of HS (n = 13) and non-HS (n = 5). Error bars are means ± SD., *P<0.05; **P<0.01 is based on Student’s t-test. Scale bar in A-B: 400 µm; in C: 50 µm; in D: 20 µm. ML, molecular layer.

Figure 4. mTOR activation in the granule cells of the sclerotic hippocampus is associated with neuronal hypertrophy.

In the dentate gyrus of non-sclerotic hippocampus, granule cells have uniformly sized cell bodies and pS6 expression is below detection level (A). In the sclerotic hippocampus, pS6-positive and negative granule cells are different sizes (B, red and black two-headed arrows indicate pS6 negative and positive neurons, respectively). pS6 positive granule cells in the dentate gyrus have larger cell bodies than do pS6-negative cells (F). A special non-sclerotic hippocampus with recently emerging seizures induced by a glioma also exhibits some granule cells with substantially increased pS6 expression (C), but both pS6-positive and -negative cells have the same cell size (F). KA mice at 2 weeks post-SE were used to study the cell size of pS6 positive and negative neurons, both upper (D) and lower blades (E) of DG were involved in the statistic analysis (F). Error bars are means ± SD., **p<0.01 is based on Student’s t-test. Scale bar in A-C: 50 µm; in D-E: 25 µm.

Figure 5. pS6 IR in hippocampus 2 and 5 weeks after KA-induced seizures.

(A) Dispersion of granule cells is visible at 2 weeks post-SE ipsilateral (ipsi) to the KA injection (n = 4), (C) and is pronounced at 5 weeks (n = 5). pS6 is significantly upregulated in most of the dispersed granule cells at 2 weeks (A), and in the entire dentate gyrus at 5 weeks (B). The morphology and pS6 IR in the corresponding non-injected (contralateral, conr) hippocampus are normal (B, D). Immunofluorescence for pS6 (green) confirms the same observations. (Insets in C, D). Astrocyte-like cells with increased pS6 IR are widespread in the dentate gyrus and CA area of sclerotic hippocampus (Arrows in A, C). Scale bar in A–B: 400 µm.