Evidence for mTOR pathway activation in a spectrum of epilepsy-associated pathologies

Abstract

Introduction: Activation of the mTOR pathway has been linked to the cytopathology and epileptogenicity of malformations, specifically Focal Cortical Dysplasia (FCD) and Tuberous Sclerosis (TSC). Experimental and clinical trials have shown than mTOR inhibitors have anti-epileptogenic effects in TS. Dysmorphic neurones and balloon cells are hallmarks of FCDIIb and TSC, but similar cells are also occasionally observed in other acquired epileptogenic pathologies, including hippocampal sclerosis (HS) and Rasmussen's encephalitis (RE). Our aim was to explore mTOR pathway activation in a range of epilepsy-associated pathologies and in lesion-negative cases.

Results: 50 epilepsy surgical pathologies were selected including HS ILAE type 1 with (5) and without dysmorphic neurones (4), FCDIIa (1), FCDIIb (5), FCDIIIa (5), FCDIIIb (3), FCDIIId (3), RE (5) and cortex adjacent to cavernoma (1). We also included pathology-negative epilepsy cases; temporal cortex (7), frontal cortex (2), paired frontal cortical samples with different ictal activity according to intracranial EEG recordings (4), cortex with acute injuries from electrode tracks (5) and additionally non-epilepsy surgical controls (3). Immunohistochemistry for phospho-S6 (pS6) ser240/244 and ser235/236 and double-labelling for Iba1, neurofilament, GFAP, GFAPdelta, doublecortin, and nestin were performed. Predominant neuronal labelling was observed with pS6 ser240/244 and glial labelling with pS6 ser235/236 in all pathology types but with evidence for co-expression in a proportion of cells in all pathologies. Intense labelling of dysmorphic neurones and balloon cells was observed in FCDIIb, but dysmorphic neurones were also labelled in RE and HS. There was no difference in pS6 labelling in paired samples according to ictal activity. Double-labelling immunofluorescent studies further demonstrated the co-localisation of pS6 with nestin, doublecortin, GFAPdelta in populations of small, immature neuroglial cells in a range of epilepsy pathologies.

Conclusions: Although mTOR activation has been more studied in the FCDIIb and TSC, our observations suggest this pathway is activated in a variety of epilepsy-associated pathologies, and in varied cell types including dysmorphic neurones, microglia and immature cell types. There was no definite evidence from our studies to suggest that pS6 expression is directly related to disease activity.

Figure 1

pS6 in HS ILAE type 1 with (A-J) and without (K-L) dysmorphic changes. HS ILAE type 1 with dysmorphic changes (A-J) (A) The dentate gyrus in a case with ILAE HS type 1 and additional glassy balloon-like astroglial cells on H&E which show membranous positivity for CD34 (inset). (B) pS6 235/236 labelling of hypertrophic CA4 neurones and (C) in the granule cell layer was noted in HS cases with dysplasia features; inset shows co-localisation of labelling in a proportion of cells with the two pS6 antibodies. (D) pS6 235/236 also labelled small immature cells with bipolar or multipolar processes including in the basal layer of the dentate gyrus (D) as well as through the dentate gyrus (E). Co-localisation between doublecortin (DCX) and pS6 was noted in some of these small cells in the dentate gyrus (F) as well as with nestin (G); in addition both pS6 markers co-labelled a proportion of small multipolar cells in HS. (I) With pS6 240/244 prominent labelling of horizontal cells in the stratum moleculare of the hippocampus, in addition to more distinct labelling of pyramidal cells through hippocampal subfields was noted. HS ILAE type 1: Intense labelling of CA4 neurones but the the granule cells were more variably negative (K) or positive (L) with pS6 240/244. White arrowheads in all images indicate double-labelled cells. Bar in A, C, D, E, F, G, H and J equivalent to approximately 35 microns and in B, I, K and L approximately 50 microns.

Figure 2

pS6 in FCD subtypes. FCD II: (A) Intense labelling was confirmed in dysmorphic neurones in FCDIIa and FCD IIb (B) as anticipated, highlighting the tigroid appearance of the cytoplasm (B). Dysmorphic neurones were also intensely positive with pS6 235/6 (C), although the cytoplasm of balloon cells appeared somewhat weaker. (D) Double labelling of pS6 with DCX confirmed the balloon cells as pS6 positive, with some balloon cells co-labelling with DCX (arrowhead, inset); small DCX bipolar cells were not always pS6 labelled (arrow) and wrapping of a DCX-positive cell processes around a pS6-positive balloon cell (inset) was also noted. FCD IIIa: (E) FCD type IIIa (adjacent to HS) with laminar neuronal loss demonstrated with NeuN labelling and clusters of neurones in layer II accompanied by mark superficial cortical gliosis (F); in these cases labelling of the residual neurones in layer II with pS6 ser 235/6 (G) and pS6 ser 240/244 (H) was noted. Double labelling studies confirmed co-localisation of pS6 240/244 with GFAPdelta isoform in the cortex (I) and white matter (J) and between nestin and pS6 235/6 in small glial cells particularly in perivascular regions (K), as well as focally between DCX-positive small cells and pS6 235/236 (L). FCD IIIb: pS6 highlighted intense labelling of scattered cortical neurones in adjacent dyslaminar cortex, as well as neurones trapped within the tumour (inset), but negative labelling of the small tumour cells (M). FCDIIId: Disrupted cortex adjacent to an old perinatal infarct in this case showed a prominent ‘tramline’ labelling pattern of pyramidal cells (N); inset shows prominent labelling of astroglial cells in the region of chronic cortical scarring in an FCD IIId case. Bar in A, B, C, D, I, J, K and L equivalent to approximately 35 microns, in e,f,g,h and m to 50 microns and in n to 100 microns.

Figure 3

Rasmussen’s encephalitis. (A) Areas of active encephalitis and cortical scarring highlighted with increased numbers of HLADR-positive microglia/macrophages, as well as interstitial rod-like CD163-positive microglia cells (inset). (B) On adjacent sections, labelling with pS6 235/236 showed a similar distribution of cellular labelling as well as with pS6 240/244 (C). (D) pS6 240/244 highlighted many of the enlarged dysmorphic neurones in RE cases, as confirmed with co-labelling with neurofilaments (inset). (E) pS6 236/26 demonstrated labelling of smaller glial cells as well as (F) bipolar rod shape cells. Double labelling studies in RE cases confirmed the majority of pS6 labelled cells were not GFAP-positive astroglia (G) but co-labelled with populations of iba1-positive cells (microglial marker) (H), nestin and doublecortin (inset) (I) positive cells. Bar in A, B, C equivalent to approximately 100 microns, in d approximately 50 microns and E, F, G, H, I approximately 30 microns.

Figure 4

Non-lesional epilepsy and control groups. (A) Intense labelling of morphologically different cells types, including macrophages and multipolar cells (inset), around an organising intracranial electrode track mark. (B and C) Paired samples from regions of different ictal activity, based on intracranial monitoring: (B) is from case 34, sample 1 (ictal onset zone) where less pS6 labelling was seen with both markers (illustrated here with pS6 240/244) compared to (C) sample 2 which represented less active/seizure spreading zone where more pS6 labelling of small cells was observed as well as neuronal cells (inset). (D) Temporal lobe cortex with no specific pathology, adjacent to HS, where prominent ‘tramline’ labelling of cortical neurones was observed with pS6 240/244. (E) Another pathologically normal temporal lobe in epilepsy with a prominent perivascular labelling pattern of neuronal cells with pS6 235/236 (arrowhead) and (F) labelling of small glial like cells around vessels in the white matter and inset in the subpial layer (inset). Bar in A, B, C, E, F equivalent to approximately 50 microns and in d to 100 microns.