(A) Virus was delivered into the ischemic margin. A representative coronal brain section stained with cresyl/violet and X-gal from an animal treated with HSV vector. Viral vector was pre-injected into the ischemic margin (left hemisphere) and the corresponding region in the non-ischemic cortex (right hemisphere), blue color indicates transfected region. (B) Representative brain sections transfected with control vector, p35 and crm A in both the ischemic and non-ischemic cortex are shown. The number of X-gal positive neurons in ischemic cortex (arrows) was much higher in the brain treated with p35 compared with crm A and control vector. (C) Quantification of surviving X-gal positive neurons after infection with viral caspase inhibitors (p35 and crmA) and control vector. The total numbers of X-gal positive neurons in the ischemic cortex were divided by those in the contralateral cortex and expressed as percentages (mean±SEM). Compared to control, p35 gene therapy, but not crmA, significantly increased neuronal survival in the ischemic cortex. (*p<0.05; ANOVA).

(A) Double immunofluorescence staining of β-gal and activated caspase 3 from brains survived 48 hr after stroke. Activated caspase-3 was not detected in the non-ischemic cortex (data not shown). Transfected neurons with control, p35 or crmA vectors were stained with with β-gal (red) and activated caspase 3 (green). Double-labeled neurons are indicated (arrows). Transfected neurons which were not co-labeled with activated caspase 3 were also indicated (astericks). p35 transfection resulted in fewer double-labeled neurons than control or crmA. β-gal, β-galactosidase; DAPI, 4'6-diamindino-2-phenylindole; scale bar, 20μm. (B) Quantification of neurons stained positively for β-gal and activated caspase 3 in brains transfected with viral caspase inhibitors (p35 and crmA) and control vectors. In the ischemic cortex, the number of neurons double-stained with both β-gal and activated caspase 3 were divided by the total number of β-gal positive neurons and expressed as percentages (mean±SEM). p35, but not crmA, delivery inhibited caspase 3 activation (*p<0.05; ANOVA).

(A) Double immunofluorescence staining of β-gal and cytochrome c in ischemic brains 48 hr after stroke. Double-stained neurons were marked with arrows and neurons stained with β-gal only were marked with astericks. An inset was inserted to show a magnified transfected cell with cytochrome c release. The cytochrome c was detected in the ischemic cortex but not in the contralateral cortex (data not shown), suggesting that cytochrome c release from the mitochondria into the cytosol. p35 transfection resulted in a smaller number of double-labeled neurons than control or crmA. Scale bar, 20μm. (B) Percentage of neurons stained positively with cytochrome c which was also positively stained with β-gal in ischemic brains treated with p35 and crmA and control vectors. p35, but not crmA, delivery inhibited cytosolic translocation of cytochrome c (*p<0.05; ANOVA).

(A) Double immunofluorescence staining of β-gal and AIF in ischemic cortex 48 hr after stroke. Positive AIF staining was detected in the ischemic but not non-ischemic cortex (data not shown). After 48 hr of MCAO, most of AIF expression localized at nuclei. Double stained neurons with β-gal and AIF and β-gal only are marked with arrows and astericks, respectively. A transfected cell with nuclear AIF staining was enlarged in an inset. p35 transfection reduced the number of double-labeled neurons than control or crmA. AIF, apoptosis-inducing factor; scale bar, 20μm. (B) Percentage of cells with nuclear AIF translocation among transfected neurons. Quantification of neurons stained positively for β-gal reporter and nuclear AIF after infection of viral caspase inhibitors (p35 and crmA) and control vectors. In ischemic cortex, the counts of β-gal and nuclear AIF double-labeled neurons were divided by those of β-gal single-labeled neurons and expressed as percentages (mean±SEM). p35, but not crmA, inhibited nuclear translocation of AIF compared with control vector (*p<0.05; ANOVA).