Potential implications of Apolipoprotein E in early brain injury after experimental subarachnoid hemorrhage: Involvement in the modulation of blood-brain barrier integrity

Abstract

Apolipoprotein E (Apoe) genetic polymorphisms have been implicated in the long term outcome of subarachnoid haemorrhage (SAH), but little is known about the effect of Apoe on the early brain injury (EBI) after SAH. This study investigated the potential role of APOE in EBI post-SAH. Multiple techniques were used to determine the early BBB disruption in EBI post-SAH in a murine model using wild-type (WT) and Apoe-/- (KO) mice. Progressive BBB disruption (Evans blue extravasation and T2 hyperintensity in magnetic resonance imaging) was observed before the peak of endogenous APOE expression elevation at 48h after SAH. Moreover, Apoe-/- mice exhibited more severe BBB disruption charcteristics after SAH than WT mice, including higher levels of Evans blue and IgG extravasation, T2 hyperintensity in magnetic resonance imaging, tight junction proteins degradation and endothelial cells death. Mechanistically, we found that APOE restores the BBB integrity in the acute stage after SAH via the cyclophilin A (CypA)-NF-κB-proinflammatory cytokines-MMP-9 signalling pathway. Consequently, although early BBB disruption causes neurological dysfunctions after SAH, we capture a different aspect of the effects of APOE on EBI after SAH that previous studies had overlooked and open up the idea of BBB disruption as a target of APOE-based therapy for EBI amelioration research in the future.

Keywords: Apolipoprotein E; Pathology Section; blood-brain barrier; early brain injury; neuroinflamamation; subarachnoid hemorrhage.

Figure 1. SAH grade and overall mortality

Data are expressed as mean ± SEM. A.-B., SAH grade in WT group and Apoe KO group were not significantly different. C. The overall mortality rate within 72h after SAH was 28.57% (34 of 119) in the WT group and 39.02% (16 of 41) in the KO group, with nonstatistical difference.

Figure 2. Vasogenic cerebral edema and neurobehavioral defects after SAH

Data are expressed as mean ± SEM. A. The dotted area in represented the region of interest (ROI) contain the bilateral parietal lobe sensorimotor cortex and hippocampus sensitive to damage. The vasogenic cerebral edema was significantly increased at 24h and peaked at 48h (^**P < 0.01, n = 5) B.-C.. Brain water content (BWC) resulted in the similar change (^**P < 0.01, n = 5) D., RH = right hemisphere, LH = left hemisphere). E.-G.The RR latency and Modified Garcia Score (MGS) decreased dramatically at 6h after SAH relative to their sham-operated counterparts. At later times, the rotorod (RR) latency and MGS performance gradually improved but continuously remained below the sham group levels. Meanwhile, compared with sham-operated mice, body weight loss progressively increased and peaked at 48h after SAH. (**P < 0.01,*P < 0.05 respectively, n = 5).

Figure 3. Endogenous APOE expression after SAH

Data are expressed as mean ± SEM. In sham group, the endogenous APOE expression level was relative low in the astrocytes. After subarachnoid hemorrhage (SAH), APOE in the astrocytes were gradually elevated, up to 24h, the APOE expression was significantly up-regulated and peaked at 48h (^**P < 0.01) A.-C.. The perivascular space APOE positive cells were significantly increased at 24h and 48h, provided an information indicating that APOE function in early brain injury (EBI) is associated with the BBB integrity modulation after SAH (^**P < 0.01) C., D.. (Scale bars = 50μm, Blue lectin positive area indicate microvessels; green GFAP positive area indicate astrocytes; red APOE positive area indicate APOE; n = 5 each group).

Figure 4. Apoe deletion affects mice neurological defects after SAH

Data are expressed as mean ± SEM. The APOE protein expression was totally depleted in Apoe-deficient mice as shown by both western blot analysis A., C. and immunofluorence staining B., D.while the WT mice showed a significant APOE upregulation 48h after SAH (**P < 0.01). (Scale bars = 50μm, Blue lectin positive area indicate microvessels; green GFAP positive area indicate astrocytes; red APOE positive area indicate APOE; n = 5 each group). There was no significant difference in body weight loss (BWL), RR latency, MGS or brain edema between the sham groups (WT-SHAM vs KO-SHAM, @ means p > 0.05, n = 5). During the experimental lesion, Apoe-deficient mice displayed more BWL, shorter RR latency and less MGS compared with WT mice E.-G.. (WT-48H vs KO-48H, **p < 0.01,*p < 0.05 respectively, n = 5).

Figure 5. Apoe deficiency aggravates vasogenic brain edema and BBB permeability

Data are expressed as mean ± SEM. A.-C. The Vasogenic cerebral edema and brain water content were not significantly deferent from each other in the sham groups (WT-SHAM vs KO-SHAM, @ represents that p > 0.05, n = 5). However, after SAH, T₂-weight image (T2WI) MRI detected vasogenic cerebral edema in the prior dotted area and brain water content were dramatically deferent from each other with a higher edema level in the Apoe-deficient mice. The BBB permeability was not significantly different in the sham group detected by Evans blue (EB) extravasation D. and plasma immunoglobulin G immunofluorence staining E.-F.. However, both the brain EB content and extravascular IgG deposition were significantly elevated with higher levels far from the wild-type mice in Apoe-deficient mice (^*P < 0.05, ^**P < 0.01, n = 5 respectively, RH = right hemisphere, LH = left hemisphere, Blue lectin positive area indicate microvessels; green area indicate perivascular deposition of IgG; n = 5 each group).

Figure 6. Apoe deficiency aggravates BBB structure damage

Data are expressed as mean ± SEM. A.-D. The three major TJs were simultaneously decreased at 48h post-SAH. Furthermore, the decrease of TJs was more obvious in the KO-48H group (@ represents that p > 0.05, ^*P < 0.05, ^**P < 0.01 respectively, n = 5). E. TEM scanning showed TJs disruption and vacuolations between endothelial cells as pointed by the black arrows. In the Apoe-deficient mice group, more TJs disruption and bigger vacuolations between endothelial cells could be observed accompanied with microvessels lumen deformation (subjacent figures represent the dotted area in the corresponding ones above).

Figure 7. Apoe deficiency aggravates endothelial apoptosis

Data are expressed as mean ± SEM. A.-D. In comparison with wild-type mice, the apoptotic related protein BAX, Cleaved Caspase-3 expressions were higher in the Apoe-deficient mice while the antiapoptosis protein BCL2 was significantly lower (@ represents that p > 0.05, *p < 0.05, **p < 0.01 respectively, n = 5). E.-F.. Apoe-deficient mice displayed more endothelial cells death detected by Lectin and TUNEL double stainning (Red lectin positive area indicate microvessels; blue area indicate TUNEL and blue DAPI area indicate nucleus; @ represents that p > 0.05, ^*p < 0.05, ^**p < 0.01 respectively, n = 5).

Figure 8. Apoe deficiency aggravates MMP-9 upregulation after SAH

Data are expressed as mean ± SEM. A.-B.The protein expression analysed using western blotting demonstrated a greater levels of MMP-9 after SAH in Apoe-deficient mice when compared with WT ones. C.-D. The MMP-9 enzymatic activity was also higher in Apoe-deficient mice at 48h after SAH. E.The immunofluoresence stain showed that the perivascular MMP-9 location was abundantly increased after SAH, perivascular MMP-9 positive area was more obvious in the Apoe-deficient mice. (Blue lectin positive area indicate microvessels; red area indicate perivascular location of MMP-9; @ represents that p > 0.05, ^**p < 0.01 respectively, n = 5).

Figure 9. Apoe deficiency aggravates the inflammatory response

Data are expressed as mean ± SEM. Although the main mediators of MMP-9 like CypA and phosphorylated nuclear factor kappa B subunit p65 (p-P65) were slighty different in the sham group, the discrepancy has greater statistical significance at 48h after SAH (^*P < 0.05, ^**P < 0.01 respectively, n = 5) A.-C.. The enzyme-linked immuno sorbent assay (ELISA) measurement of TNF-α, IL-6 and IL-1β, another three inflammatory mediators of MMP-9 downstream the CypA- NF-κB related pathway, showed no significant difference of these inflammatory cytokines. Nevertheless, the discrepancies of these cytokines have greater levels at 48h after SAH in Apoe-deficient mice compared with WT mice (@ represents that p > 0.05, ^**P < 0.01 respectively, n = 5) D.-F..