At t = 4 hours, the relative CBF was 228% ± 49% in the triple dosing group and 169% ± 22% in Forskolin molecular weight the intravenous infusion group. Compared with group 1 from experiment B, we found no significant differences (F[2,16] = 0.95, P = 0.41, one-way ANOVA) (Fig. 2B). Based on local clinical recommendations for hypermagnesemia and relevant literature,21 we aimed at achieving a P-Mg > 2 mM after 2 hours. We also aimed at a CSF-Mg level of greater than 20% above baseline after 4 hours. The results of the dose finding study in experiment A demonstrated that groups 2, 3, and 4 achieved a P-Mg greater than 2 mM at t = 2 hours and that group 4 also had a P-Mg

greater than 2 mM at t = 4 hours. Interestingly, we did not find that increased doses of magnesium led to higher CSF-Mg levels at t = 4 hours, even though P-Mg was higher. We concluded that the most appropriate dosing regimen CB-839 was 1.6 mmol/kg MgSO4 intraperitoneally at baseline and 0.8 mmol/kg MgSO4 intraperitoneally after 1 hour, which fulfilled our criteria of acceptable plasma levels and relevant central nervous system bioavailability. Consequently, this dosing regimen was used

in experiment B, in which we, apart from MAP, ICP, and CBF, also studied the potential mechanisms of action of hypermagnesemia. Experiment B showed that induction of hypermagnesemia did not prevent development of intracranial hypertension or cerebral hyperperfusion. We used the well-characterized rat model with PCA and acute hyperammonemia22 and did indeed find cerebral hyperfusion and high ICP. In fact, with the dosing regimen of experiment B, we saw a higher CBF and a tendency toward a higher ICP in rats treated with MgSO4 compared with the corresponding group not receiving MgSO4. Another interesting finding was that

hyperammonemia led to a significant drop in MAP after 1 hour of ammonia infusion and that hypermagnesemia ADAMTS5 in both ammonia and saline infusion animals led to a tendency torward lower MAP; however, it was most pronounced in rats with hyperammonemia. This could indicate that the model itself (PCA + ammonia infusion) induces an initial vasodilatation that is worsened by hypermagnesemia and adds up to the substantial increase in relative CBF that was almost 50% higher in the PCA+NH3+MgSO4 group compared with PCA+NH3+vehicle. To reduce the risk of false-negative results, we also performed experiment C with alternative dosing regimens of MgSO4. The PCA rats appeared to have a lower clearance of MgSO4 than the healthy animals in experiment A, most likely because of the hepatic shunt. We achieved a P-Mg above 2 mM at 2 hours and at the end of the experiment with both a triple-dosing regimen and intravenous infusion. This did not, however, lead to significant positive effects on ICP or CBF. We did observe a tendency toward a slightly lower ICP in the triple dosing group than in the vehicle group.