Neurobiology of Disease (Nov 2023)

Linking epileptic phenotypes and neural extracellular matrix remodeling signatures in mouse models of epilepsy

  • Armand Blondiaux,
  • Shaobo Jia,
  • Anil Annamneedi,
  • Gürsel Çalışkan,
  • Jana Nebel,
  • Carolina Montenegro-Venegas,
  • Robert C. Wykes,
  • Anna Fejtova,
  • Matthew C. Walker,
  • Oliver Stork,
  • Eckart D. Gundelfinger,
  • Alexander Dityatev,
  • Constanze I. Seidenbecher

Journal volume & issue
Vol. 188
p. 106324

Abstract

Read online

Epilepsies are multifaceted neurological disorders characterized by abnormal brain activity, e.g. caused by imbalanced synaptic excitation and inhibition. The neural extracellular matrix (ECM) is dynamically modulated by physiological and pathophysiological activity and critically involved in controlling the brain's excitability. We used different epilepsy models, i.e. mice lacking the presynaptic scaffolding protein Bassoon at excitatory, inhibitory or all synapse types as genetic models for rapidly generalizing early-onset epilepsy, and intra-hippocampal kainate injection, a model for acquired temporal lobe epilepsy, to study the relationship between epileptic seizures and ECM composition. Electroencephalogram recordings revealed Bassoon deletion at excitatory or inhibitory synapses having diverse effects on epilepsy-related phenotypes. While constitutive Bsn mutants and to a lesser extent GABAergic neuron-specific knockouts (BsnDlx5/6cKO) displayed severe epilepsy with more and stronger seizures than kainate-injected animals, mutants lacking Bassoon solely in excitatory forebrain neurons (BsnEmx1cKO) showed only mild impairments. By semiquantitative immunoblotting and immunohistochemistry we show model-specific patterns of neural ECM remodeling, and we also demonstrate significant upregulation of the ECM receptor CD44 in null and BsnDlx5/6cKO mutants. ECM-associated WFA-binding chondroitin sulfates were strongly augmented in seizure models. Strikingly, Brevican, Neurocan, Aggrecan and link proteins Hapln1 and Hapln4 levels reliably predicted seizure properties across models, suggesting a link between ECM state and epileptic phenotype.

Keywords