A role for leukocyte-endothelial adhesion mechanisms in epilepsy

Abstract

The mechanisms involved in the pathogenesis of epilepsy, a chronic neurological disorder that affects approximately one percent of the world population, are not well understood. Using a mouse model of epilepsy, we show that seizures induce elevated expression of vascular cell adhesion molecules and enhanced leukocyte rolling and arrest in brain vessels mediated by the leukocyte mucin P-selectin glycoprotein ligand-1 (PSGL-1, encoded by Selplg) and leukocyte integrins alpha(4)beta(1) and alpha(L)beta(2). Inhibition of leukocyte-vascular interactions, either with blocking antibodies or by genetically interfering with PSGL-1 function in mice, markedly reduced seizures. Treatment with blocking antibodies after acute seizures prevented the development of epilepsy. Neutrophil depletion also inhibited acute seizure induction and chronic spontaneous recurrent seizures. Blood-brain barrier (BBB) leakage, which is known to enhance neuronal excitability, was induced by acute seizure activity but was prevented by blockade of leukocyte-vascular adhesion, suggesting a pathogenetic link between leukocyte-vascular interactions, BBB damage and seizure generation. Consistent with the potential leukocyte involvement in epilepsy in humans, leukocytes were more abundant in brains of individuals with epilepsy than in controls. Our results suggest leukocyte-endothelial interaction as a potential target for the prevention and treatment of epilepsy.

Figure 1. Alpha4 integrin, VCAM-1 and PSGL-1 dependence of PMN and T_H1 cell interactions with brain endothelium after SE

a, b. Expression of adhesion molecules in cortical vessels was evaluated by in vivo staining and fluorescence microscopy before (Baseline) (a) and 6 h after the onset of seizure activity induced by pilocarpine (b). Isotype-matched mAb served as a control. Suppression of SE was achieved by administration of 3 mg/Kg Diazepam i.p. 20 min before pilocarpine injection (Diazepam) (b). Scale bar: 200 µm. c, e, f. The frequency of rolling interactions or sticking behavior of fluorescently labeled PMNs and lymphocyte subpopulations in cortical venules was studied before pilocarpine injection (0 h) and at 6 h and 24 h after SE onset by in situ videomicroscopy. The percentage of total cells shown on Y axis was calculated as described in Supplementary methods. The number of venules and animals per group analyzed are provided in Supplementary Table 1b, c. Mean ± SEM are shown. PLNs, freshly isolated peripheral lymph node cells (resting lymphocytes). d. Adherent PMNs 6 h post-SE and T_H1 cells at 24 h post SE are shown in cerebral vessels (arrows). Scale bar: 100 µm. e, f. Anti-adhesion molecule mAbs were used to block leukocyte or endothelial adhesion molecules at 6 h after SE as described at Supplementary methods (Intravital microscopy subsection). PMNs are shown in e and T_H1 cells are shown in f. Groups were compared with control using Kruskall-Wallis test followed by Bonferoni correction of P. **P<0.01; *P<0.001.

Figure 2. Effect of blockade or deficiency in leukocyte adhesion mechanisms on convulsions and seizures

In a-c visual monitoring of mean number of daily spontaneous convulsions (SC) per mouse was performed from day 5–20 after pilocarpine administration. Pilocarpine treated WT mice received no treatment (“epileptic mice”), or were injected i.p. either 1 h after SE-onset with 400 µg α4 integrin-specific mAb or with 150 µg VCAM-1-specific mAb, to model therapy; or 2 h prior to pilocarpine injection with 400 µg α4 integrin-specific mAb (pre-treatment regimen). In both instances treated mice also received antibody (200 µg anti-α4 or control mAb and 150 µg anti-VCAM-1) every other day for 20 days. An isotype-matched control antibody was used as control (a). 10–12 animals/group were monitored for SC for 6 h/day between day 5 and 20 after SE. One representative experiment from a series of 4 (a), 2 (b) and 3 (c) with similar results is shown. In d–f EEGs for each animal were acquired 24 h/day from day 0–20 after pilocarpine administration. Behaviorally, SRS were characterized in our mice by head nodding, forelimb clonus, rearing, and falling. d, days, s, seconds (g) Twenty-sec EEG recordings early after initial seizures and at the end of the second h are shown. Theta activity was occasionally observed in Psgl-1^−/− animals (asterisk, fourth column, second line). (h) Recordings of representative SRS are shown. For d–h data are representative from one experiment with 3 mice/group from two experiments with similar results. ***P<0.001.

Figure 3. Effect of blockade or deficiency in leukocyte adhesion mechanisms on CNS histopathology

Epilepsy-induced structural changes and effects on neuronal cell density alterations were analyzed at 30 days after SE in healthy animals and in mice injected with pilocarpine: control (C57Bl/6, WT) mice, anti-α4 treated animals and Psgl-1^−/− and Fut7^−/− mice. For quantitative evaluation, the number of Nissl stained cell bodies was counted in four sections/animal (3 animals/group) using regions of interest (ROI) sampling. The ROI was delimited by a rectangular frame which was placed within the sampled cortical and hippocampal areas, using upper-left exclusion lines. In the somatosensory (SS) cortex, the rectangular frame was placed vertically throughout the cortical thickness, superficial (II-III) (SS1) and deep (V-VI) layers (SS2) of the cortex. See also the Supplementary Methods for details. Mean ± SEM are shown for stereological counts. Statistical comparison of anti-α4-treated groups or Psgl-1^−/− and Fut7^−/− groups with Control/WT epileptic condition was performed using a two-tail t-test. *P<0.04. CA, corpus ammonis. Scale bar corresponds to 30 µm.

Figure 4. Inhibition or deficiency of leukocyte adhesion pathways maintains the integrity of the BBB

(a) Evans Blue (EB) extravasation was evaluated 18–24 h after pilocarpine injection. In WT (control) mice, EB leakage (asterisks) was observed macroscopically in brains, after removal of the meninges, in the parietal cortex and in particular in the territories relative to the medial cerebral artery (black asterisks) in the superior and lateral views; in the pituitary gland pedunculus (where the BBB is more permeable) in the inferior view; and, finally, in the choroid plexus of the lateral ventricles and in the fourth ventricle as in the sagittal view. WT animals injected only with methylscopolamine did not show any leakage (data not shown). Anti-α4 pretreatment as well as PSGL1- or FucT-VII deficiency (Psgl-1^−/− and Fut7^−/− mice) prevented BBB leakage. Representative brains are shown from one of 2 experiments, 3 mice/condition, with similar results. Scale bar: 4 mm. (b) BBB alterations post-SE were studied in control (C57Bl/6/, WT mice), and anti-α4 treated animals and in Psgl-1^−/− and Fut^−/− mice at 24 h after pilocarpine injection. BBB permeability was studied by the mean of intravenously administered Magnevist® (Nmethylglucamine salt of gadolinium complex of diethylenetriamine pentaacetic acid) a paramagnetic iron oxide contrast agent in MRI. Pseudo-color maps evidencing contrast agent spreading are shown. Representative brains are shown from one of two experiments (3mice/condition) with similar results. Scale bar: 2 mm.