Neuroprotective Autophagic Flux Induced by Hyperbaric Oxygen Preconditioning is Mediated by Cystatin C

A genome editing technique based on transcriptional activator-like effector nucleases (TALEN) was used to produce CysC-knockout rats by targeting exon 1 of CysC [7]. To confirm the efficiency of CysC deletion, sequencing analysis was performed to detect base-pair deletions and western blot to assess the protein levels of CysC.

Rats were randomly assigned to the control or HBO group. The rats in the HBO group were exposed to hyperbaric oxygen (2.5 atmospheres absolute and 100% O₂, 1 h/day) for five consecutive days. The control rats were placed in the hyperbaric chamber with air at normal pressure.

Transient focal cerebral ischemia was achieved using a right middle cerebral artery occlusion (MCAO) model. Briefly, rats were anesthetized intraperitoneally (i.p.) with pentobarbital sodium (50 mg/kg). The right common carotid and external carotid arteries were exposed through a ventral midline neck incision. A monofilament (Beijing Sunbio Biotech Co., Ltd, Beijing, China) with a rounded tip was inserted into the common carotid immediately below the bifurcation. Then, it was advanced into the internal carotid ~ 18 mm–20 mm distal to the bifurcation until a mild resistance was felt, indicating occlusion of the middle cerebral artery. After 2 h, the middle cerebral artery was re-perfused by retracting the intraluminal monofilament. The regional cerebral blood flow of all rats subjected to ischemia/reperfusion (I/R) injury was monitored using a PF 5000 Laser Doppler Perfusion Monitoring Unit (PeriFlux 5000, Perimed AB, Stockholm, Sweden). The intervention was considered successful only if the regional cerebral blood flow sharply decreased to ≤ 30% of the baseline level after MCAO, and increased to ≥ 70% of the baseline level within 10 min of reperfusion.

Twenty microliters of 20 μmol/L CysC siRNA or scrambled siRNA was infused into the right lateral ventricle 3 days before MCAO as reported previously [8]. The rats were anesthetized with pentobarbital sodium and placed in a stereotaxic apparatus. A burr hole was drilled into the skull 1.5 mm lateral and 1.0 mm posterior to bregma over the right hemisphere. A stainless-steel 26-gauge cannula (C315G, Plastic One, Roanoke, VA) was slowly introduced through the burr hole into the right lateral ventricle (3.8 mm beneath the dural surface). Reagents were infused into the ventricle at 0.5 μL/min. The CysC siRNA transfection complex was prepared according to the manufacturer’s instructions. The target sequence 5′-CCCAGACAAATTTGACTAACT -3′ was used.

3-Methyladenine (3-MA) (30 μg in 10 μL artificial cerebrospinal fluid, Sigma-Aldrich) was administered by intracerebroventricular injection 30 min before ischemia [8], while exogenous CysC (40 μg/kg) dissolved in artificial cerebrospinal fluid was delivered into the right lateral ventricle 30 min after reperfusion [7].

Three hours post-reperfusion, the rats were deeply anesthetized before sequential perfusion with ice-cold saline and 4% paraformaldehyde in phosphate-buffered saline (PBS). The brains were removed and dehydrated in a sucrose gradient (20%–30%) in PBS at 4 °C. Coronal sections (10 μm thick) at bregma ± 2 mm were cut on a cryostat and stored at −20 °C until further use. The slices were blocked with 3% normal goat serum in 0.5% Triton X-100 for 30 min at room temperature. Subsequently, the slices were incubated with primary antibodies at 4 °C overnight, followed by incubation with secondary antibodies for 2 h at room temperature. The following antibodies were used: rabbit polyclonal anti-cathepsin B and mouse monoclonal anti-NeuN (both 1:1000, Abcam). DAPI (4’,6-diamidino-2-phenylindole; 1:1000; Sigma-Aldrich) was used to stain nuclei. The following secondary antibodies were used: green-fluorescent Alexa Fluor 488-conjugated donkey anti-mouse and red-fluorescent Alexa Fluor 594-conjugated donkey anti-rabbit (both 1:1000; Abcam). Fluorescent signals in the right penumbra region of cortex (2 mm lateral from the midline) were detected using confocal laser scanning microscopy (FV1000, Olympus, Tokyo, Japan).

The designated region (penumbra: tissue less severely hypo-perfused than the ischemic core, and where neurons are functionally impaired but not yet irreversibly damaged) in the right cortex (between sections + 2 mm and − 3 mm of bregma) was separated after cold saline perfusion as described previously [9], and whole-cell protein was extracted using a total protein extraction kit (Merck Millipore). An equivalent of 40 µg protein from each sample was resolved on 12% SDS-PAGE. The following primary antibodies were used: anti-CysC (1:1000, Abcam), anti-LC3-I (1:1000, Sigma-Aldrich), anti-LC3-II (1:1000, Sigma-Aldrich), anti-Beclin 1 (1:1000, Cell Signaling), and anti-p62 (1:1000, Abcam). Goat anti-rabbit antibody was used as the secondary antibody (1:5000, Cell Signaling).

Rats were anesthetized with 1% sodium pentobarbital (50 mg/kg i.p.) and perfused transcardially with 150 mL ice-cold 0.9% saline, followed by 500 mL cold mixture of 4% paraformaldehyde, 0.05% glutaraldehyde, and 15% (v/v) saturated picric acid in 0.1 mol/L phosphate buffer (pH 7.4) for 2 h. The brains were excised and fixed in the same fixative at 4 °C for 3 h. Serial coronal sections (50 μm) were cut on a vibratome (VT 1000S, Leica) and incubated overnight with the primary antibody rabbit anti-cathepsin B (1:100 in PBS containing 1% BSA). Subsequently, the sections were washed in PBS and incubated overnight with anti-rabbit IgG conjugated to 1.4 nm gold particles (Nanoprobes) at 1:100. After rinsing, the sections were post-fixed in 2% glutaraldehyde in PBS for 45 min. Silver enhancement was performed in the dark with an HQ Silver Kit (Nanoprobes) for visualization of cathepsin B immunoreactive signals. Before and after the silver enhancement step, the sections were rinsed several times with deionized water and incubated in ABC solution (Sigma-Aldrich) for 4 h, followed by visualization using the glucose oxidase-3,3′-diaminobenzidine method. The immunolabeled sections were fixed in 0.5% osmium tetroxide in 0.1 mol/L phosphate buffer for 1 h, dehydrated in an ethanol gradient and propylene oxide, and embedded in Epon 812 (SPI-CHEM, West Chester, PA). After polymerization, the sections were examined by light microscopy. Three sections containing cathepsin B immunoreactivity were selected from each rat, trimmed under a stereomicroscope, and mounted on new resin stubs. Ultrathin sections were cut on an ultra-microtome (EM UC6, Leica), mounted on mesh grids (6–8 sections/grid) followed by counterstaining with uranyl acetate and lead citrate, and observed in a JEM-1230 electron microscope (JEOL Ltd, Tokyo, Japan). Electron-micrographs were captured by a Gatan digital camera and analyzed with its software (832 SC1000, Gatan, Warrendale, PA).

SPSS 19.0 was used for statistical analysis. All data are presented as mean ± SEM and were analyzed using one-way ANOVA followed by the Dunnett or Bonferroni post hoc test. Differences were considered significant if the P value was < 0.05.

We have shown that the level of CysC is significantly increased after HBO preconditioning in non-ischemic brain [7]. To explore the role of CysC after ischemic injury, we first examined its expression at different time points after cerebral I/R. Compared to control rats without I/R injury, the expression of CysC in the ischemic penumbra was significantly up-regulated as early as 3 h after reperfusion, and the up-regulation was retained until 24 h post-reperfusion (Fig. 1A). Compared with the I/R group, HBO preconditioning further enhanced the expression of CysC at 3 h post-reperfusion [2.62 (0.26) vs 1.72 (0.18), P < 0.01, HBO + I/R vs I/R] (Fig. 1B). These results suggested that both ischemic injury and HBO preconditioning up-regulate CysC in the brain, while HBO preconditioning further enhanced this up-regulation induced by ischemia.

Previous studies have demonstrated that both HBO preconditioning [8] and exogenous CysC promote the induction of autophagy [10]. So, we investigated the role of CysC in mediating autophagy induced by HBO preconditioning. First, we tested LC3-II and Beclin-1, which are reliable markers of autophagy induction. We found that both were significantly up-regulated 3 h after reperfusion [1.43 (0.12) normalized to control, P < 0.05; 1.51 (0.15), P < 0.01] (Fig. 2B, C). Interestingly, HBO-preconditioned rats exhibited a further elevation of these proteins [1.78 (0.13) for LC3-II and 1.96 (0.07) for Beclin-1 in the HBO + I/R group]. On the other hand, as a marker for autophagy inhibition, p62 was not changed after I/R; but HBO preconditioning significantly decreased the p62 level after I/R, indicating that autophagic flux was enhanced (Fig. 2D).

To confirm the role of CysC, we knocked CysC down by intracerebroventricular administration of CysC siRNA. CysC expression was knocked down by 40% in rats without I/R and HBO preconditioning (Fig. S1). The increase of CysC induced by HBO preconditioning and I/R injury was significantly blocked by CysC siRNA (Fig. 3). Furthermore, knockdown of CysC significantly reduced the up-regulation of LC3-II and Beclin-1 induced by HBO preconditioning and I/R injury (Fig. 4A–C), while the decrease in p62 induced by HBO preconditioning and I/R injury was reversed (Fig. 4D).

To corroborate the role of CysC, we generated CysC^−/− rats using TALEN-based genome-editing to target exon 1 of CysC (Fig. 5A). Sequencing analysis of the targeted CysC locus revealed a 5-bp deletion of the coding sequence (Fig. 5C). Western blot analysis confirmed the gene-targeting efficiency of CysC, demonstrating a significant reduction in CysC expression (Fig. 5B). Compared to the non-ischemic control group, I/R injury increased LC3-II and Beclin-1 in wild-type rats. However, in CysC^−/− rats, the up-regulation of LC3-II and Beclin-1 induced by HBO preconditioning and I/R injury disappeared (Fig. 5D). These results indicated that CysC is an essential mediator in promoting the autophagic flux induced by HBO preconditioning and cerebral ischemia.

Autophagic processing, also described as autophagic flux, includes sequestration of the cytoplasmic cargo into double-membrane vesicles and delivery of the autophagosome to the lysosome for degradation, emphasizing the crucial role of the lysosome during autophagy [11]. As shown in Fig. 6A, in sham-operated wild-type (WT) rats, cathepsin B was observed as fine, granular, peri-nuclear immunoreactivity, indicating intact lysosomal membrane integrity. Three hours post-reperfusion, cathepsin B increased and was diffusely distributed throughout the cytoplasm and nucleus, whereas HBO preconditioning alleviated the diffusion. Meanwhile, we found that in CysC^−/− rats, the improvement of lysosomal membrane integrity induced by HBO preconditioning was totally reversed (Fig. 6B, C). Taken together, these results suggested that the disruption of lysosomal integrity after I/R injury is alleviated by HBO preconditioning in a CysC-dependent manner.

To visualize autophagic flux directly, cortical neuronal ultrastructure was observed with the aid of immuno-electron microscopy. As shown in Fig. 7A, in sham-operated WT rats, cathepsin B was wrapped within the lysosome, where immunogold particles clustered within a single membrane structure. Three hours after reperfusion in the WT rats without HBO preconditioning, a number of autophagosomes could be seen in the neurons of the penumbra, while autolysosomes were rarely detected. Moreover, the immunogold silver-enhanced cathepsin B staining was diffusely located in the cytosol, indicating the disruption of lysosomes. HBO preconditioning significantly attenuated the damage of lysosomes caused by ischemic injury in WT rats, as revealed by fewer immunogold particles outside the lysosome. Interestingly, autolysosomes (lysosomes containing a phagocytic membrane structure cargo) were evident in the HBO preconditioning group. In contrast, HBO preconditioning neither rescued the disruption of lysosomes nor promoted the formation of autolysosomes in CysC^−/− rats (Fig. 7B, C). These results indicated that CysC is essential in preserving lysosomal integrity and forming autolysosomes, thereby promoting the completion of autophagy.

The above results indicated that the elevation of CysC induced by HBO preconditioning enhances autophagy. However, direct evidence linking CysC-autophagy-protection was still missing. Therefore, we used 3-MA to determine whether CysC can still confer the neuroprotection in the absence of autophagy. Exogenous CysC significantly improved neurological function and reduced the infarct volume 72 h after reperfusion compared with the I/R group receiving saline (Fig. 8). The neurobehavioral score of the 3-MA + CysC group was higher than that of the I/R group and lower than that of the CysC group (P < 0.05). The infarct volume in the 3-MA + CysC group was smaller than that in the I/R group and larger than that in the CysC group. These results revealed that the autophagy inhibitor partially reversed the neuroprotection induced by CysC.