Authors :
Hongtao Ma, PhD – Weill Cornell Medicine; Peijuan Luo, PhD student – The First Hospital of Jilin University, Department of Neurology; Jing Li, PhD – The First Hospital of Jilin University, Department of Neurology; Fan Yang, PhD – The First Hospital of Jilin University, Department of Neurology; James Niemeyer, PhD – Neurological Surgery – Weill Cornell Medicine; Mingui Zhao, PhD – Neurological Surgery – Weill Cornell Medicine; Dan Li, MD, PhD – Radiology – The First Hospital of Jilin University; Weihong Lin, MD, PhD – The First Hospital of Jilin University, Department of Neurology; Theodore Schwartz, MD – Neurological Surgery – Weill Cornell Medicine
Rationale: Epilepsy can spread slowly across the cortex in a classic Jacksonian march, or rapidly through long-range horizontal projections. The recruitment of distant is a crucial component of the modern conception of epilepsy as a network disease. Whether long-range projections recruit excitatory or inhibitory activity in distant nodes, and this E/I balance impacts the spread of seizures is not well understood.
Methods: Using the Allen Mouse Brain Connectivity Atlas, we injected BMI into the primary sensory cortex (S1), which is known to project to ipsilateral M2 and contralateral S1, and recorded interictal spikes (IIS) with wide-field calcium imaging of excitatory (Thy1-GCaMP6f mice, Jaxlab #024276) and inhibitory neurons (PV-GCaMp6f mice, crossing of Jaxlab #017329 and #028865) across both hemispheres. We quantified the amplitude and participation rate of each of the four visible nodes (iS1, iM2, cS1, cM2) and compared activity to control regions in iV1 and cV2, which are not strongly connected with iS1.
We measured the strength of recruitment and coactivation to investigate the impact and dependence of each node on another.
Results: IIS-associated calcium signal (both Thy-1 and PV) could be recorded from the primary node (iS1) where BMI was injected, as well as from the nodes with monosynaptic connections (iM2 and cS1). Surprisingly, cM2 was also frequently recruited, which was disynaptically connected to iS1. Quantitive analysis showed that both excitatory and inhibitory cells were recruited in all nodes and iS1 was the primary driver of this recruitment. IM2 was most frequently recruited, but all other nodes were activated as well. CS1 displayed more inhibitory cell recruitment, compared with cM2, in which excitatory cells were more frequently recruited.
Conclusions: We have shown how IISs can hijack the known long-range projections between brain areas to create an epileptic network. Inhomogeneities in E/I cell recruitment explain variable node recruitment. Further studies employing cell-specific/node-specific modulation may uncover potential network-based therapies to control intractable epilepsy.
Funding: Not applicable