Mutations in the EPM2A gene encoding a dual-specificity phosphatase (laforin) cause Lafora disease (LD), a progressive and invariably fatal epilepsy with periodic acid-Schiff-positive (PAS+) cytoplasmic inclusions (Lafora bodies) in the central nervous system. To study the pathology of LD and the functions of laforin, we disrupted the Epm2a gene in mice. At two months of age, homozygous null mutants developed widespread degeneration of neurons, most of which occurred in the absence of Lafora bodies. Dying neurons characteristically exhibit swelling in the endoplasmic reticulum, Golgi networks and mitochondria in the absence of apoptotic bodies or fragmentation of DNA. As Lafora bodies become more prominent at 4-12 months, organelles and nuclei are disrupted. The Lafora bodies, present both in neuronal and non-neural tissues, are positive for ubiquitin and advanced glycation end-products only in neurons, suggesting different pathological consequence for Lafora inclusions in neuronal tissues. Neuronal degeneration and Lafora inclusion bodies predate the onset of impaired behavioral responses, ataxia, spontaneous myoclonic seizures and EEG epileptiform activity. Our results suggest that LD is a primary neurodegenerative disorder that may utilize a non-apoptotic mechanism of cell death.

Abstract— Blood‐brain barrier (BBB) transport of choline and certain choline analogs was studied in adult and suckling rats, and additionally compared in the paleocortex and neocortex of adult rats. Saturable uptake was characterized by a single kinetic system in all cases examined, and in adult rat forebrains we determined a Km= 442 ± 60 μM and Vmax= 10.0 ± 0.6 nmol min‐1 g‐1. In 14–15‐day‐old suckling forebrains a similar Km (= 404 ± 88 μM) but higher Vmax (= 12.5 ± 1.5 nmol min‐1 g‐1) was determined. When choline uptake was compared in two regions of the forebrain, similar Michaelis‐Menten constants were determined but a higher uptake velocity was found in the neocortex (i.e. neocortex Km= 310 ± 103 μM and Vmax= 12.6 ± 2.8 nmol min‐1g‐1; paleocortex Km= 217 ± 76 μM and Vmax= 7.2 ± 1.5 nmol min‐1 g‐1). Administration of radiolabelled choline at low (5 μM) and high (100 μM) concentrations, followed by microwave fixation 60 s later and chloroform‐methanol‐water separations of the homogenized brain did not suggest a relationship between concentration and the appearance of label in lipid or aqueous fractions as observed in another in‐vitro study elaborating two‐component kinetics of choline uptake. It was observed that 60s after carotid injection 12–14% of the radiolabel in the ipsilateral cortex was found in the chloroform‐soluble fraction. Hemicholinium‐3 (Ki= 111 μM), dimethylaminoethanol (Ki= 42 μM), tetraethyl ammonium chloride, tetramethyl ammonium chloride, 2‐hydroxyethyl triethylammonium iodide, carnitine, normal rat serum, and to a lesser extent lithium and spermidine all inhibited choline uptake in the BBB. Unsubstituted ammonium chloride and imipramine did not inhibit choline uptake. No difference was observed in blood‐brain barrier choline uptake of unanesthetised, carotid artery‐catheterized animals, and comparable sodium pentobarbital‐anesthetized controls.

No abstract

Immunogold electron microscopy was used to examine human brain resections to localize the GLUT1 glucose transporter. The tissue examined was obtained from a patient undergoing surgery for treatment of seizures, and the capillary profiles examined had characteristics identical to those described previously for active, epileptogenic sites (confirmed by EEG analyses). A rabbit polyclonal antiserum to the full-length human erythrocyte glucose transporter (GLUT1) was labeled with 10-nm gold particle-secondary antibody conjugates and localized immunoreactive GLUT1 molecules in human brain capillary endothelia, with < 0.25% of the particles beyond the capillary profile. Erythrocyte membranes were also highly immunoreactive, whereas macrophage membranes were GLUT1-negative. The number of immunoreactive sites per capillary profile was observed to be 10-fold greater in humans than in previous studies of rat and rabbit brain capillaries. In addition, half of the total number of immunoreactive gold particles were localized to the luminal capillary membrane. We suggest that the blood-brain barrier GLUT1 glucose transporter is up-regulated in seizures, and this elevated transporter activity is characterized by increased GLUT1 transporters, particularly on the luminal capillary membranes. In addition, acute modulation of glucose transporter activity is presumed to involve translocation of GLUT1 from cytoplasmic to luminal membrane sites, demonstrable with quantitative immunogold electron microscopy.