Abstracts

Rationale:
Tumor-associated seizures (TAS) occur in up to 90% of patients with low-grade glioma (LGG). These events cause increased morbidity and often persist even after resection of the tumor. Multiple groups have shown that TAS arise in the peritumoral cortex, attributed to increased excitation or functional impairment of inhibition. A candidate mechanism is a hypothesized reciprocal relationship between epileptogenicity and gliomagenesis via the mTOR pathway. Studying the inhibition of this pathway and its effects on individual excitatory/inhibitory units in a tumoral microenvironment could be key to optimizing future management of TAS in LGG patients.
Method:
Ex-vivo acute brain slices were obtained from a p53-deleted, PDGF-A-IRES-Cre driven murine model of low-grade glioma in Thy1-GCaMP transgenic mice, in which excitatory pyramidal cells exhibit calcium fluorescence. One cohort was treated with 20 mg/kg of AZD8055 (AZD), an mTOR inhibitor, by oral gavage 5 to 6 hours prior to sacrifice. Slices from this cohort were also bathed in 30 nM AZD in ACSF prior to recording. Recordings were performed using a 96-channel multi-electrode array (400 micron interelectrode spacing, 30 kHz sampling) to acquire electrophysiologic activity for 5 minutes in ACSF and 30 minutes in zero-magnesium solution, with concurrent calcium imaging (50 Hz frame rate). Slices were then fixed and stained for hemagglutinin (tumor) and phospho-s6 (mTOR activity). Thy1-GCaMP positive neurons were identified based on their endogenous fluorescence. The histology was then coregistered to the calcium imaging and microelectrode sites. Standardized methods were applied to filter and cluster spikes semi-automatically based on principal components using a k-means algorithm robust to outliers. Neurons were then sub-classified into putative pyramidal cells and fast-spiking interneurons. Spike-triggered averaging of calcium imaging was performed for each cell type to measure excitatory responses, as represented by transient increases in the GCaMP fluorescence intensity. Statistical significance was measured using a Mann-Whitney U test.
Results:
Across 7 non-AZD and 5 AZD slices, spike sorting revealed a greater number of putative inhibitory units in AZD slices than in non-treated slices (37 of 243 vs. 19 of 466, p < .0001). The inhibitory firing rate was faster in AZD-treated slices than in non-treated ones (.46 vs. .24 spikes/sec, p < .05). Spike triggered averaging revealed transient increases in GCaMP fluorescence temporally correlated with excitatory firing in glioma-infiltrated cortex, for both non-treated and AZD-treated slices, as expected. However, in non-treated slices, a comparable increase in calcium intensity was also observed for inhibitory firing (Fig. 1A). This response was absent or attenuated in AZD-treated slices (Fig. 1B). When statistically comparing the local spike-triggered averages of calcium response between inhibitory and excitatory firing, AZD-treated slices showed a significant difference (p < .0001) and non-treated slices did not (p = .68).