Abnormal Capillary Vasodynamics Contribute to Ictal Neurodegeneration in Epilepsy

Seizure-driven brain damage in epilepsy accumulates over time, especially in the hippocampus, which can lead to sclerosis, cognitive decline, and death. Excitotoxicity is the prevalent model to explain ictal neurodegeneration. Current labeling technologies cannot distinguish between excitotoxicity and hypoxia, however, because they share common molecular mechanisms. This leaves open the possibility that undetected ischemic hypoxia, due to ictal blood flow restriction, could contribute to neurodegeneration previously ascribed to excitotoxicity. We tested this possibility with Confocal Laser Endomicroscopy (CLE) and novel stereological analyses in several models of epileptic mice. We found a higher number and magnitude of NG2+ mural-cell mediated capillary constrictions in the hippocampus of epileptic mice than in that of normal mice, in addition to spatial coupling between capillary constrictions and oxidative stressed neurons and neurodegeneration. These results reveal a role for hypoxia driven by capillary blood flow restriction in ictal neurodegeneration.

doses of Ketamine-Xylazine were given as needed (one-third to one-half of the original dose) throughout the duration of the recording session (up to eight hours).

Kainate model animals
The non-NMDA receptor agonist kainic acid (Sigma-Aldrich; K0250) or saline control solution was injected subcutaneously with 20-30 mg/kg of kainic acid or saline control solution. Animals were placed in the recording chamber with their head fixed to allow awake intrahippocampal recording of the capillary beds and epidural EEG before kainic acid injection.
To test the kainic acid injection dose, we measured the seizure intensity after different injections for a group of animals. After the injection, we placed the animals in a clear plastic cage and we monitored their locomotor activity and EEG.

Confocal Laser Endomicroscopic imaging methods
We either recorded vasospasms with a single-band Cellvizio (Leica, USA, model FCM-1000) laser scanning microscope targeting a 488 nm laser down the bundle (beveled at the tip to penetrate the tissue) at ≥12 Hz (data in Fig. 1, Supplementary Figs. 1-3 and 8), or a dual-band Cellvizio (Mauna Kea Technologies, Inc, Paris France) targeting both a 488nm and a 660nm laser at ≥12 Hz ( Fig. 4A-R). Each fiber at the beveled surface captured the emitted fluorescence and the photons were descanned into avalanche photodiode detectors 1) The time at which the vasospasm constriction started, i.e. the "onset" when the signal first began to dip below baseline.
2) The time at which the constriction leveled out at its minimum, which typically occurred shortly before the termination of the vasospasm (on average at t = ~80 secs after onset).
3) The time at which the vasospasm terminated (dilation started).
4) The time at which the dilation ends, i.e. when the signal resumed its plateau.
For the onset τ, the data between 1) and 2) above were normalized to the range [0,1] and fitted to a curve of the form: x(t) = exp(-t/ τ) For the termination tau, data between 3) and 4) above were normalized and fitted to: The sphere becomes sequentially larger in 2 μm radius steps, and eventually unmasks the green fluorescein-stained vessel within the stack (at 22 μm in this example, see arrow). The distance of the nearest blood vessel to the central DAPI+ nucleus was determined as the radius of the smallest sphere that contained a vessel. We binned the cell counts by distance as a function of both cohort (KO vs WT), and whether they were AIF+/-. These cell counts created the distance measurements in Figures 3I, 5G-N  or through epilepsy. Epilepsy can further cause a macroscopically hyperemic seizure focus that leads to both cell death from hypoxic apoptosis associated with abnormal vasospasms near ischemic vessels, or from other types of cell death (i.e. excitoxicity) in which cell distances from vessels are spatially random. The combined result is that AIF+ cells tend to be nearer to blood vessels than (randomly distributed) AIF-cells. Scale bars 10 M.

Scientific Reports
Research article Supplementary Figure 4A-H.

Supplementary Movie 2. Blood flow and vasospasm imaging of vessel in
Vasospasm and mural cell recorded (11.7 Hz) in a KO hippocampal capillary. Figure 4I-R.

Supplementary Movie 3. Blood flow and vasospasm imaging of vessel in
Vasospasm, leukocyte blockages, and mural cells recorded (11.7 Hz) in a WT hippocampal capillary.   Figure 3F-H was digitally segmented for nuclei (blue), capillaries (green), and α-SMA antibody labeled mural cells (red). 3D models were created from these data, and the central mural cell from Figure 3F-H was enhanced and magnified. Digital analysis reveals that the mural cell completely surrounds the capillary at the point of constriction, and that the vessel is constricted to an approximate diameter of 1 µm. The mural cell extends down the occluded vessel but the green serum fluorescence is completely cutoff mid-mural cell, indicating that this mural cell completely blocked serum flow during this vasospasm.