Abstracts

Rationale: Hippocampal sclerosis is the main pathological substrate in mesial temporal lobe epilepsy (MTLE), and is characterized by neuronal loss, gliosis and axonal reorganization. mGluR5 is a post-synaptic receptor that modulates glutamatergic neurotransmission. Increased hippocampal mGluR5 protein expression in MTLE has been described in humans and animal models, but PET studies have shown decreased in vivo mGluR5 binding in the epileptogenic hippocampus, suggesting that these receptors might be dysfunctional. Our goal was to analyze the pattern of mGluR5 expression in the hippocampal formation of MTLE patients and to evaluate if mGluR5 expression is associated with changes in neuronal density, mossy fiber sprouting and microtubule-associated protein 2 (MAP2) expression. Methods: We studied 43 hippocampi from refractory MTLE patients who underwent en bloc resection and 10 control hippocampi from necropsy (no pre-mortem neurological disorders). Formalin fixed and paraffin embedded sections were submitted to immunohistochemistry for NeuN, MAP2, mGluR5 and neo-Timm. Hippocampal subfields were defined according to Lorente de No s classification and quantification of neuronal density was performed according to Abercrombie s method. Positive immunoreactive (IR) area and gray value (GV) were estimated with ImageJ software. Statistical analyses were performed using SPSS 11. Results: Neuronal loss was identified in most MTLE hippocampal formation subfields, with sparing of subiculum, parasubiculum and entorhinal cortex. Corrected by neuronal density, mGluR5-IR was increased in all subfields (2.5 to 10-fold higher in MTLE, p<0.02). MTLE hippocampi showed loss of correlation between CA3 and CA1 mGluR5 and neuronal density, which was strong in control hippocampi (R=0.98, p=0.002). Correlation between mGluR5-IR and neuronal density in MTLE was seen only at CA4 (R=0.83, p=0.006). MTLE MAP2-IR was increased in the granular layer and CA2 (p<0.003) and reduced in the other subfields (p<0.03) except for the subiculum and entorhinal cortex, where there were no differences with controls. There was a strong correlation between MAP2 and mGluR5 expression in CA3 in MTLE (R=0.78, p=0.01), but not in control hippocampus. There was a moderate positive correlation between mGluR5-IR in the granular layer and GV of sprouted mossy fibers in the inner molecular layer (R=0.50, p=0.02) in MTLE. Conclusions: mGluR5 upregulation in MTLE hippocampi is present in almost all subfields, despite massive neuronal loss in sectors such as CA1 and prosubiculum. A strong positive correlation between mGluR5 and MAP2 was seen only in CA3, suggesting an increase in postsynaptic expression of the receptor in this sector. Loss of correlation between mGluR5 and neuronal density at sectors with cell loss such as CA1, suggests that non-neuronal or outer neuron perycaria elements may contribute to mGluR5 upregulation seen in MTLE hippocampi.