Nitric oxide (NO) is produced by endothelial NOS (eNOS) in the intima and by neuronal NOS (nNOS) in the adventitia of cerebral vessels. It dilates the arteries in response to a shear stress, metabolic demands, and chemoregulation. Subarachnoid hemorrhage (SAH) interrupts this regulation; oxyhemoglobin and deoxyhemoglobin, gradually released (p<0.05) from the subarachnoid clot enveloping the conductive arteries, destroy nNOS-containing neurons. This deprives the arteries of NO, leading to initiation of delayed vasospasm (PHASE I). However, the narrowing of the vessel stimulates eNOS through increased shear stress, which normally would lead to an increased production of NO and dilation of arteries. However, this does not happen due to transient eNOS dysfunction evoked by an increase of endogenous competitive NOS inhibitor, asymmetric dimethyl-arginine (ADMA) in CSF in response to the presence of bilirubin-oxidized fragments (BOXes). This eNOS dysfunction sustains vasospasm (PHASE II). ADMA levels are closely correlated with the degree and a time-course of vasospasm in humans (p<0.002) and in primate model (p<0.01) and when ADMA levels decrease vasospasm resolves (PHASE III; p<0.05). Increased levels of ADMA is evoked by its decreased elimination due to decreased presence of ADMA-hydrolyzing enzyme dimethylamine-dimethyl-L-arginine hydrolase (DDAH II; p<0.05) in the arteries in spasm. Thus, inhibition of the L-arginine-methylating enzyme (IPRMT3) or stimulation of the DDAH II may provide new therapeutic venues. We will present 1) experimental data confirming endothelial and neuronal NO synthases dysfunction during vasospasm, 2) experimental and clinical data supporting the hypothesis that BOXes are responsible for eNOS dysfunction during vasospasm and pre-clinical data on 3) probucol influence on ADMA production by endothelial cells exposed to hemoglobin and BOXes in vitro (p<0.05) and 4) in vivo data of a double-blind placebo-controlled primate SAH study of probucol to prevent vasospasm (p=0.09). Despite being negative, the results of this study support the hypothesis that pharmacological lowering of the CSF ADMA levels may prevent development of post-hemorrhagic delayed cerebral vasospasm.
