The involvement of vagal afferents in the communication pathway from the immune system to the brain was studied. Glutamate was measured in the nucleus tractus solitarius, the brain area receiving glutamatergic vagal afferents, after peripheral injection of lipopolysaccharide or interleukin-1 beta. Intraperitoneal or intravenous saline or intraperitoneal heat-inactivated interleukin-1 beta increased glutamate release, measured by brain microdialysis in freely-moving rats at 20 min (137 +/- 19%, 126 +/- 10% and 133 +/- 6%, respectively), without inducing any other change up to 3 h after injection. Intraperitoneal lipopolysaccharide (10 micrograms/rat) increased glutamate at 20 min (132 +/- 10%) and at 60 min (208 +/- 26%). To compare lipopolysaccharide effectiveness by the two routes, serum levels of interleukin-1 beta, interleukin-6 and tumour necrosis factor-alpha were measured. Intravenous lipopolysaccharide induced each cytokine more rapidly and efficiently than intraperitoneal lipopolysaccharide. Perfusion with tetrodotoxin (1 microM) in the dialysis probe decreased glutamate basal levels by approximately 20% and completely prevented lipopolysaccharide effects, showing the neuronal component of the glutamate measured. Except for the 20-min increase, intravenous lipopolysaccharide (10 micrograms/rat) did not affect glutamate release. Intraperitoneal interleukin-1 beta (4 micrograms/rat) increased glutamate release at 20 min (126 +/- 6%) and at 40 min (150 +/- 18%). These data indicate that vagal glutamatergic system in the nucleus tractus solitarius is activated by intraperitoneal injections of immunoactive compounds.