3.1 Contents of GRP78 and LC3 increased time dependently post HI
Firstly, ER stress along with autophagy activity in the hippocampus were measured by western blotting at different time points (3hours, 6 hours, 12 hours, and 24 hours) post HI. GRP78 (glucose-regulated protein 78), also referred to as BiP, is a ER resident protein and when under ER stress, it dissociates first from the ER membrane to help destroyed proteins refold and degrade [11]. Stimulation of GRP78 is considered as a biomarker of ER stress [12] and stimulation of autophagy was examined by immunoblotting of LC3B. As indicated in Figure 1, GRP78 expression level was upregulated from 3 h to 24 h post-HI (Figure 1A, B, 3 hours, p < 0.05; 6 hours, p < 0.01; 12 hours, p < 0.05; 24 hours, p < 0.05) and the peak GRP78 level was reported at 6 hours after HI. Increase of LC3BII expression started at 6 hours and attained the peak at 24 hours (Figure 1A, C, 6 hours: p < 0.05; 12 hours: p < 0.05; 24 hours: p < 0.01) after HI treatment. Our data illustrated that ER stress along with autophagy participate in the pathological process of neonatal HIBI. Besides, we selected HI-24 hours model for the successive experiments.
3.2 ER stress triggered autophagy in the neonatal HIBI rats
Next, to assess the association of ER stress with autophagy in the context of neonatal HIBI, the rats were inoculated with tunicamycin (ER stress inducer), 4-PBA (ER stress inhibitor), or 3-MA (autophagy inhibitor). The p62 ubiquitin binding protein participates in the selective autophagy and reduction of p62 is a biomarker for increase of autophagic flux [13]. Results showed that HI insult over-activated ER stress and autophagy as indicated by upregulation of GRP78 as illustrated in Figure 2A, B (p < 0.05 in contrast with the Sham group) and LC3BII (Figure 2A, C, p < 0.01 in contrast with the Sham group), and downregulation of p62 (Figure 2A, D, p < 0.05 in contrast with the Sham group). HI-triggered ER stress and autophagy reinforced by combination with TM (Figure 2, GRP78: p < 0.05 in contrast with HI group; LC3BII: p < 0.05 in contrast with HI group; p62, p < 0.05 in contrast with HI group), but counteracted by 4-PBA administration (Figure 2, GRP78: p < 0.05 in contrast with HI group; LC3BII: p < 0.05 in contrast with HI group; p62, p < 0.05 in contrast with HI group). However, 3-MA was able to restrain autophagy activity (Figure 2, LC3BII: p < 0.05 in contrast with HI group; p62: p < 0.05 in contrast with HI group) without dramatic influences on ER stress (Figure 2, no remarkable difference in GRP78 expression). Altogether, these data illustrated that ER stress occurred upstream of autophagy during neonatal HIBI.
3.3 ER stress-autophagy participates in the cell apoptosis triggered by HIBI
The cell apoptosis in the hippocampus was explored by TUNEL assay. As demonstrated in Figure 3, positive apoptosis cells were seen in HI group (CA1: p < 0.01 in contrast with Sham group; CA3: p < 0.01 in contrast with Sham group), and treatment with TM enhanced HI-induced apoptosis in the hippocampal CA1and CA3 subregions (CA1: p < 0.05 in contrast with HI group; CA3: p < 0.01 in contrast with HI group). Interfering ER stress by administration with 4-PBA restrained HI-induced cell apoptosis (CA1: p < 0.05 in contrast with HI group; CA3: p < 0.05 in contrast with HI group). Similarly, the autophagic inhibitor 3-MA suppressed HI-induced apoptosis via inhibition of autophagy (CA1: p < 0.05 in contrast with HI group; CA3: p < 0.05 in contrast with HI group).
3.4 Sevoflurane post-conditioning alleviated ER stress-mediated autophagy via regulation of the IRE1 cascade
Our previous researches indicated that sevoflurane-conferred neuroprotection against HIBI is linked to repressed excessive autophagy [14,15]. TM was applied in order to elucidate whether inhibition of autophagy by sevoflurane was mediated through suppressing ER stress. Sevoflurane post-conditioning attenuated HI-induced upregulation in GRP78 and LC3BII, and reduction in p62 (Figure 4A, B, HI in contrast with Sham: p < 0.05, p < 0.01, p < 0.01; SPC in contrast with HI: p < 0.05, p < 0.05, p < 0.05; for GRP78, LC3BII, and p62, respectively), but these effects were blocked by TM (Figure 4A, B, SPC + TM in contrast with SPC: all p<0.05 for GRP78, LC3BII, and p62, respectively).
Reports document that ER stress can stimulate autophagy through IRE1cascade [16]. The specific IRE1 inhibitor STF-083010 was administrated to further confirm that whether the IRE1-JNK-beclin1 cascade is involved in the neuroprotection of sevoflurane. Proteins in the IRE1 cascade consisting of p-IRE1/IRE1, beclin1, and p-JNK/JNK were also assessed via Western blot. p-IRE1/IRE1, p-JNK/JNK ratio are employed to stand for activation levels of IRE1 and JNK. IRE1 inhibitor reverted the imapcts of TM on the expression of GRP78, LC3BII, and p62 (Figure 4A, B, SPC+TM in contrast with SPC, all p < 0.05; SPC+TM+IRE1 inhibitor in contrast with SPC+TM: all p < 0.05; for GRP78, LC3BII, and p62, respectively), indicating that sevoflurane restrained ER stress-induced autophagy via the IRE1 cascade during HIBI. As presented in Figure 4, obvious increases in rations of p-IRE1/IRE1, p-JNK/JNK, and expression level of beclin1 were observed in ipsilateral hippocampus of rats in contrast with sham group (Figure 4C, D, HI in contrast with Sham: all p < 0.05 for p-IRE1/IRE1, p-JNK/JNK, and beclin1, respectively). Nevertheless, inoculation with sevoflurane dramatically restrained the above changes (Figure 4C, D, SPC in contrast with HI: all p<0.05 for p-IRE1/IRE1, p-JNK/JNK, and beclin1, respectively), which were reversed by combination with TM (Figure 5, SPC+TM in contrast with SPC: all p<0.05 for p-IRE1/IRE1, p-JNK/JNK, and beclin1, respectively). IRE1 inhibitor counteracted the effects of TM (Figure 5, SPC+TM+IRE inhibitor in contrast with SPC+TM: all p<0.05 for p-IRE1/IRE1, p-JNK/JNK, and beclin1, respectively). The above results suggested that sevoflurane post-conditioning suppressed ER stress-autophagy via IRE1-JNK-beclin1 signaling cascade.
3.5 Sevoflurane post-conditioning improved cognitive performance and prevented neuronal loss in rats following HI
To explore the anxiety-like behavior along with the locomotor activity in a novel environment, the rats underwent an open field task at P35. As indicated in Figure 5, no differences were reported among groups in overall travelled distance, time taken in the center region, or the number of feces. These results suggested that treatment factors in this study did not influence anxiety-like behavior or locomotor activity in rats.
From day P35, MWM tasks were carried out to explore spatial learning along with the memory ability. As shown in Figure 5, rats in all groups exhibited a significant downward trend in escape latency. In contrast with the sham group rats, the HI rats exhibited longer escape latency to reach the platform (Figure 5D, day 29 post-HI, p < 0.05; day 30 post-HI, p < 0.001; day 31 post-HI, p < 0.001; day 32 post-HI, p < 0.001; day 33 post-HI, p < 0.001) and fewer platform crossings (Figure 5E, p < 0.001), indicating that rats displayed impaired learning and memory function following HIBI model. Sevoflurane post-conditioning successfully shortened escape latency (Figure 5D, day 30 post-HI, p < 0.001; day 31 post-HI, p < 0.001; day 32 post-HI, p < 0.001; day 33 post-HI, p < 0.001) and increased the number of times of crossing the platform (Figure 5E, p < 0.05), but these effects were reversed by TM (Figure 5 D, E, escape latency, day 30 post-HI, p < 0.01; day 31 post-HI, p < 0.01; day 32 post-HI, p < 0.01; day 33 post-HI, p < 0.01; platform crossing times, p < 0.05). However, treatment with IRE1 inhibitor blocked the negative effects of TM (Figure 5 D, E, escape latency, day 31 post-HI, p < 0.01; day 32 post-HI, p < 0.01; day 33 post-HI, p < 0.01; platform crossing times, p < 0.05).
Nissl staining procedures were conducted to determine neuron status among groups after behavioral tests. There was an irregular scattered organization of the neurons that had increased intracellular space in the hippocampus of the rats in the HI and SPC +TM groups, whereas neurons in the SPC and SPC + IRE1 inhibitor groups preserved a better functional status. Sevoflurane significantly attenuated HI-induced reduction of neuronal density in the CA1 and CA3 hippocampal areas (HI in contrast with Sham: p < 0.001, p < 0.001; SPC in contrast with HI: p < 0.001, p < 0.001; for CA1 and CA3, respectively). Treatment with TM increased neuronal density in the hippocampus (SPC+TM in contrast with SPC: p < 0.001, p < 0.001 for CA1 and CA3, respectively), while IRE1 inhibitor blocked the effects of TM (SPC+TM+IRE1 inhibitor in contrast with SPC+TM: p < 0.001, p < 0.01 for CA1 and CA3, respectively).
These results indicated that sevoflurane post-conditioning alleviated HI-induced cognitive impairments, possibly by regulating ER stress-mediated autophagy via IRE1 signaling cascade.