The permeability of the blood-brain barrier in mice suffering from fatal lymphocytic choriomeningitis virus infection.

The ultrastructure and the blood-brain-barrier (BBB) permeability were studied in mice suffering from lymphocytic choriomeningitis (LCM). Brains and meninges from mice suffering from LCM virus-induced lymphocytic choriomeningitis were studied by investigating the BBB function and by electron and light microscopy. The cellular exudate in the leptomeninges was located in the subarachnoid space, in arachnoidea and pia, and it was dominated by proliferated pial cells and mononuclear cells, most of which were lymphocytes, while there were only a few neutrophil granulocytes. Many intravascular lymphocytes were seen adhering to as well as penetrating the vessel walls. Many of these lymphocytes were morphologically compatible with T cells. Lymphocytes and larger mononuclear cells were also accumulated in the choroid plexus, and lymphocytes were present in the ventricular system with a tendency to adhere to ependymal epithelial cells. Inspection of the ultrathin sections incubated for horseradish peroxidase (HRP)-activity revealed that the overwhelming part of the peroxidase activity was localized in the extracellular space of the meningeal vessel walls and especially in the abundant intercellular fluid which, like the inflammatory cells, was found in the subarachnoid space in arachnoidea and in pia. In the neuropil, only very small quantities of reaction product were seen intercellularly in the most superficial layers of the cortex. The tight junctions were always intact, but the possibility of a non-demonstrable opening is discussed. Evaluation of the BBB permeability for 2-amino[1-14C]isobutyric acid (AIB) was made by quantitative autoradiography, and it was demonstrated convincingly that AIB concentrations in the subpial and perichorodial tissues were markedly increased in diseased animals as compared to the controls. Our results seem to contradict previous theories on the cause of death resulting from the LCM disease. The findings presented here do not speak in favor of a pronounced brain edema, just as results obtained by us and others do not speak for the possibility of the death being caused by convulsive seizures with subsequent brain anoxia. However, our observations are compatible with the hypothesis that cytotoxic T cells may interact in vivo with virus-infected targets, which are essential for the regulation of the composition of the cerebrospinal fluid. On the other hand, the dysfunction of the BBB demonstrated adds a new element to the pathologic mechanism in a model for the study of virus-induced meningitis.