Vasodilation by intrathecal lipopolysaccharide of the cerebral arteries after subarachnoid haemorrhage in dogs.

To investigate the influence of inducible nitric oxide synthase on cerebral arteries after subarachnoid haemorrhage (SAH) in vivo, lipopolysaccharide (LPS), a major inducer of inducible nitric oxide synthase, was injected intracisternally into control and SAH model dogs. Intracisternal injection of LPS (0.5 mg) produced a long-lasting, submaximal vasodilation of the basilar artery of control dogs on angiography. This effect became significant at 4 hours after LPS injection and plateaued after 6 hours. This vasodilation was reduced by N(G)-monomethyl-L-arginine. Vasopressin slightly suppressed the vasodilation, while bradykinin increased it. The concentration of L-arginine in CSF decreased after LPS injection, while that of L-citrulline increased. In cytokines, the concentration of tumour necrosis factor-alpha; (TNF-alpha;) in CSF increased transiently at 4 hours after LPS injection, while interleukin-1 beta, interleukin-6, interferon-gamma, did not change. These data suggest that vasodilation by LPS is mainly due to nitric oxide predominantly synthesized by an inducible nitric oxide synthase, proximally induced by TNF-alpha. Our data make it unlikely that SAH itself induces the inducible nitric oxide synthase in vascular tissue, since isolated endothelium-denuded basilar artery from SAH model dogs did not respond to L-arginine. In SAH model dogs, the degree of vasodilation by LPS differed with the severity of vasospasm. Vasodilation was much greater in mild than in severe vasospasm in dogs, and was increased by superoxide dismutase. These findings suggest that the induction of inducible nitric oxide synthase or its activity may be less effective in severe vasospasm.