P12.39.B TUMOR-DRIVEN NEURONAL ALTERATIONS IN GLIOMA PATIENTS: EX VIVO ELECTROPHYSIOLOGICAL INSIGHTS FROM TUMORAL AND PERITUMORAL BRAIN CORTEX

Abstract BACKGROUND Gliomas are primary neuroepithelial tumors mainly confined to the central nervous system, highlighting a critical dependence on the neural microenvironment. However, how neuronal excitability and circuit dynamics vary between the tumoral and peritumoral cortices remains poorly understood. MATERIAL AND METHODS We performed an ex vivo electrophysiological investigation on brain tissues surgically resected from tumoral and peritumoral regions of patients with varying glioma grades. When the surgical approach permitted, tumoral and peritumoral cortices were collected from the same patient, enabling direct comparisons and minimizing the impact of inter-individual variability. Samples were sliced using a vibratome and used for both whole-cell patch-clamp and high-density multielectrode array (HD-MEA) recordings. RESULTS In peritumoral slices, whole-cell recordings showed two distinct neuronal populations based on passive properties, resting membrane potential, and electroresponsiveness: pyramidal neurons (PN) and interneurons (IN). Although passive properties of PNs and INs were comparable between tumoral and peritumoral cortices, functional differences emerged. Interestingly, only 20% (1/5) of INs were spontaneously active in the tumoral cortex versus 71.4% (5/7) in the peritumoral cortex. Moreover, in the tumoral cortex, PNs showed reduced excitability compared to the peritumoral cortex (in response to the same current injection: 14.7 ± 7.5 Hz; n=3 cells, N=3 patients; vs. 2.2 ± 0.9 Hz, n=6, N=4; p=0.04). Additionally, a population of glioma-associated cells (GMCs) was identified only in the tumoral cortex. GMCs exhibited glial-like passive properties and a depolarized resting membrane potential with no spontaneous firing. Upon depolarizing current injections, GMCs generated short, single aberrant spikes significantly different from spikes recorded from PNs in the tumoral sample (overshoot: GMC = 25.5 ± 2.7 mV, n=9, N=3; PN = 80.5 ± 3.7 mV, n=5, N=3; p=4×10-8). This suggests a tumor-induced transformation of glial cells that aberrantly mimic neuronal behavior. Consistent with whole-cell results, HD-MEA recordings revealed rare single-unit spontaneous activity in the tumoral cortex. Interestingly, blocking inhibition triggered widespread local field potentials in one tumoral sample, suggesting an altered inhibitory modulation. Conversely, in the peritumoral cortex, HD-MEA recordings showed sparse single-unit firing that intensified (from 3.2 ± 0.4 Hz, n=61, to 5.8 ± 0.7 Hz, n=54, N=1; p=0.02) and arranged in groups when blocking inhibition (coefficient of variation: from 1.0± 0.1, n=61, N=1, to 3.9 ± 0.2, n=54, N=1; p=3×10-17). CONCLUSION These results provide novel insights into alterations of neuronal excitability and network dynamics driven by the glioma environment, a crucial step toward understanding tumor-brain interactions.