A. The diagram shows the protocols to compare the protective effects of rapid and delayed postconditioning, in which postconditioning was carried out by occluding or releasing the bilateral CCA. Postconditioning was induced by 10 cycles of 10 sec occlusion and 10 sec release of the bilateral CCA (group 2 and 3), or by 3 cycles of 30 sec occlusion and 30 sec release (groups 4 and 5). Rapid postconditioning in groups 2 and 4 was induced immediately after reperfusion; delayed postconditioning in groups 3 and 5 were initiated 3 h after reperfusion. L. left CCA; R, right CCA. B. Representative infarcts stained by TTC from each group. C. Average infarct sizes. Infarct size was measured 2 d after ischemia. Conditions for each group are indicated below the bar. * vs control ischemia (group 1), P<0.05.

The gray region (p) plus the black region (C) represent ischemic injury in a control rat with ischemia alone; C represents infarction in a rat that received ischemia plus postconditioning. The region P spared by postconditioning is defined as the penumbra and the region C is defined as the ischemic core.

Delayed postconditioning (DPC) was conducted with 6 cycles of 15 min occlusion/15 min release in the ipsilateral CCA occlusion 6 h after stroke. Four tests were performed: A. Vibrissae-elicited forelimb placement test. All sham rats showed normal forelimb placing. There was unsuccessful placing of the contralateral forelimb after stroke in control rats; postconditioning attenuated the overall deficit from 1 to 51 d after stroke. The analysis of Two-way repeated measures ANOVA show that there are significant differences between groups: sham vs DPC, P = 0.037; sham vs control ischemia, P<0.001; DPC vs control ischemia, P = 0.003. * vs. before ischemia, P<0.5; #, ##, ###vs. sham, P<0.05, 0.01, 0.001; ††, †††, vs. postconditioning, P<0.01, 0.001. N = 6/each group. B. Postural reflex test. Scores significantly increased in control rats at 1, 2, 7, 10, and 21 d after stroke; postconditioning reduced scores at 1, 2, 7, 10 d after stroke. Two-way repeated measures ANOVA shows that there are significant differences between groups: sham vs DPC, P = 0.058; sham vs control ischemia, P<0.001; DPC vs control ischemia, P<0.001.. *, *** vs. before ischemia, P<0.01, 0.001. ### vs. sham, P<0.001; ††,††† vs. postcon, P<0.01, 0.001; . C.Tail hang test. The number of large right turns induced by tail hanging increased in control rats; postconditioning blocked right turns at 7, 10, and 14 d. Two-way repeated measures ANOVA: Control vs sham, P = 0.022; control vs DPC, P = 0.014; DPC vs sham, P = 0.805. *, vs. before ischemia, P<0.05; ##, ###, vs. sham, P<0.01, 0.001; †, ††, vs. postconditioning, P<0.05, 0.01. D. Home cage forelimb use test. Left-limb-use increased at 1, 2, 7, 10, 14, 21 and 44 d, which was blocked by postconditioning. Two-way repeated measures ANOVA: control ischemia vs sham, P = 0.004; control ischemia vs DPC, P = 0.010; DPC vs sham, P = 0.375. # , ## vs. sham, P<0.05, 0.001; †, ††, vs. postcon, P<0.05, 0.001; ** vs. before ischemia.

Rat brains were harvested at 48 h after stroke for edema measurement. The rat brains were sectioned coronally at 2 mm intervals, generating a total of 6 blocks from rostral to caudal. A. Rostral to caudal slice water content. Water contents in all 6 slices are presented. * vs. sham, P<0.001; # vs. control, P<0.001. B. Mean water content from all slices of the ischemic and non-ischemic hemisphere from rats with sham surgery, control ischemia, and delayed postconditioning. The mean value from the 6 blocks of each hemisphere was calculated, and a mean value from all rats (n = 6/group) is presented. DPC attenuated overall edema after stroke. Ischemic, ischemic hemisphere; non-ischemic, non-ischemic hemisphere; con, control ischemia. * vs. Sham-L, P<0.05; # vs. con, P<0.05; † vs. ischemic hemisphere, P<0.05.

T-PA (2 mg/kg) was injected intravenously for 1h started at 5 h after ischemia of 30 min bCCA occlusion plus permanent dMCA occlusion; postconditioning with 6 cycles of 15 min occlusion/15 release of the left CCA was performed from 6 h after stroke. Rats were killed 48h later for TTC staining and infarct size measurement. T-PA injection significantly worsened ischemic injury compared with control ischemia; delayed postconditioning attenuated its exacerbating effect. N = 6/group. * P<0.05 vs con or t-PA; # P<0.05,vs t-PA (ANOVA on ranks).

A. Comparative timelines for delayed postconditioning, in which postconditioning was conducted by occluding or releasing the ipsilateral CCA alone. Delayed postconditioning was performed by 6 cycles of occluding or releasing the left CCA alone; each occlusion or release lasted for 15 min. Delayed postconditioning was initiated at 3 h (group 2), 6 h (group 3), or 12 h (group 4) after reperfusion. Rats in group 1 received 1–2% isoflurane for 3h starting from 6h after reperfusion; this group serves as a control for postconditioning. L. left CCA; R, right CCA. B. Representative infarcts stained with TTC from each group. C. Average infarct size in rats treated with delayed postconditioning. Conditions for each group are indicated below the bar. *, vs. ischemic control (Group 1), P<0.05; # vs group 2, P<0.05.

Representative MRI images at 5 h (top two rows) and 24 h (middle two rows) from rats subjected to control ischemia and delayed postconditioning after stroke are shown. Control ischemia (con) was induced by 30 min bCCA occlusion combined with permanent dMCA occlusion; delayed postconditioning (DPC) was conducted from 6 h to 9 h by 6 cycles of 15 min release/15 min occlusion of the ipsilateral CCA. TTC staining showed infarction (bottom two rows) corroborated T2 weighted MRI. The rats were euthanized after MRI. Rat brains were sliced into 5 coronal slices and stained with TTC. The infarct region revealed by TTC staining is consistent with that revealed by MRI.

A. Representative PET imaging for FDG uptake and corresponding TTC staining. PET imaging was conducted 11 h after stroke; delayed postconditioning (DPC) was 6 cycles of 15 min occlusion/15 min release in the ipsilateral CCA. In the ischemic core, no apparent FDG uptake was detected (* indicated by the dark region defined as the ischemic core); in the penumbra, the signal was weaker than that in the contralateral hemisphere. There is a non-specific signal caused by the skin incision outside the brain (arrow). Rats were euthanized 2 d after stroke for TTC staining. The infarct region corroborates with the region with reduced FDG uptake. B. Delayed postconditioning (DPC) did not affect ischemic core sizes measured from PET imaging. PET images from a total of 10 levels, with a 1 mm distance between adjacent levels, were selected. The core area was measured and normalized to the whole contralateral hemisphere and expressed as a percentage; an average value from all measured levels is presented. There was no difference in the core size between rats receiving control ischemia and delayed postconditioning, suggesting that delayed postconditioning did not change FDG uptake in the ischemic core. C. DPC improved FDG uptake. The optical densities of Region of interest (ROI) in the selected 10 coronal slices from rostral (level 1) to caudal (level 10) were measured; ROIs are indicated on the inserted diagram (right corner). Optical densities in three ROIs (1 mm diameter circle) in the cortex from the ipsilateral and contraleral cortex were measured. The average density in the ipsilateral hemisphere was divided by that in the contralateral hemisphere and expressed as a percentage. FDG uptake is significantly higher in the caudal slices (slices 5–10) vs. rostral slices (slices 1–4) in rats receiving DPC; it is also significantly higher in slice 10 than in other slices in rats receiving control ischemia. DPC improved FDG uptake in slices 5 to 10 compared with corresponding slices in control rats. *, *** vs. corresponding time points in control rats, P<0.05, 0.001, respectively. † vs slice 2,3,4,5, P<0.05;. # vs. slice 1,2,3,4, P<0.01; ##, vs. slice 1, 2, 3, 4, 5, P<0.05 (Anova on Ranks followed with Dunn's test). *, *** P<0.05, 0.001, vs corresponding slices in the DPC groups. D. Square means of overall values from all 10 slices. DPC significantly improved overall FDG uptake. *** vs. control ischemia, P<0.001.

The ischemic core and penumbra and the non-ischemic hemisphere were dissected for Evan's blue detection. A. Time course of BBB leakage after stroke. The amount of Evan's blue in the ischemic and non-ischemic hemisphere was detected at 6, 24, and 48 h after stroke. Evan's blue penetrated into the ischemic brain tissue as early as 6 h, and persisted up to 48 h. More leakage occurred in the core than in the penumbra. Pen, penumbra; contra, contralateral hemisphere. ** vs. corresponding contralateral hemisphere, P<0.01; ### vs. corresponding penumbra and contralateral hemisphere, P<0.001; & vs. 6 h and 24 h in the penumbra, P<0.05 (Rank sum test). B. The effect of DPC on BBB was detected at 24 h and 48 h after stroke. DPC reduced BBB leakage at 48 h but not at 24 h after stroke in the penumbra; it had no effect on BBB leakage in the ischemic core. N = 5–6/group. C, control ischemia; P, delayed postconditioning. * vs. corresponding contralateral hemisphere, P<0.05; *** vs. corresponding contralateral hemisphere or penumbra, P<0.001. # vs. 48 h-control penumbra, P<0.05.