The Phenotype Changes of Astrocyte During Different Ischemia Conditions

Abstract

Objectives: Dementia is becoming a major health problem in the world, and chronic brain ischemia is an established important risk factor in predisposing this disease. Astrocytes, as one major part of the blood-brain barrier (BBB), are activated during chronic cerebral blood flow hypoperfusion. Reactive astrocytes have been classified into phenotype pro-inflammatory type A1 or neuroprotective type A2. However, the specific subtype change of astrocyte and the mechanisms of chronic brain ischemia are still unknown.

Methods: In order to depict the phenotype changes and their possible roles during this process, a rat bilateral common carotid artery occlusion model (BCAO) was employed in the present study. Meanwhile, the signaling pathways that possibly regulate these changes were investigated as well.

Results: After four-week occlusion, astrocytes in the cortex of BCAO rats were shown to be the A2 phenotype, identified by the significant up-regulation of S100a10 accompanied by the down-regulation of Connexin 43 (CX43) protein. Next, we established in vitro hypoxia models, which were set up by stimulating primary astrocyte cultures from rat cortex with cobalt chloride, low glucose, or/and fibrinogen. Consistent with in vivo data, the cultured astrocytes also transformed into the A2 phenotype with the up-regulation of S100a10 and the down-regulation of CX43. In order to explore the mechanism of CX43 protein changes, C6 astrocyte cells were handled in both hypoxia and low-glucose stimulus, in which decreased pERK and pJNK expression were found.

Conclusions: In conclusion, our data suggest that in chronic cerebral ischemia conditions, the gradual ischemic insults could promote the transformation of astrocytes into A2 type instead of A1 type, and the phosphorylation of CX43 was negatively regulated by the phosphorylation of ERK and JNK. Also, our data could provide some new evidence of how to leverage the endogenous astrocytes phenotype changes during CNS injury by promoting them to be "protector" and not "culprit".

Keywords: A2 phenotype; Connexin 43; astrocytes; dementia; ischemia.

Figure 1

Cognitive performance tested using a Morris Water Maze in bilateral common carotid artery occlusion (BCAO) rats after 4 weeks of ligation. The rats were tested at a designed time for five consecutive days. (A) Mean escape latency, (B) mean swimming speed, (C) time in target quadrant, (D) number of platform crossover measured. Data were expressed as mean ± SEM, * p < 0.05 versus the control rats; ns, no significance (n = 7).

Figure 2

Reactive astrocyte phenotype in the cortex of BCAO rats after 4 weeks of ligation. (A,B) The protein levels of GFAP were tested using Western blot, and the data were normalized to Gapdh blot. (C) A1 reactive astrocyte phenotypes were tested using Real-time Quantitative PCR. (D) C3 protein concentration was tested using Elisa. (E) A2 reactive astrocyte phenotypes were tested using QPCR. (F) S100a10 protein concentration was tested using Elisa. Data were expressed as mean ± SEM, * p < 0.05 versus the control rats; ns, no significance (n = 6).

Figure 3

Reactive astrocyte function in the cortex of BCAO rats after 4 weeks of ligation. (A) Representative blots of CX43, Aqua4, and ApoE. (B–D) Ratios of CX43, Aqua4, and ApoE to Gapdh or Actin were calculated and compared. Data were expressed as mean ± SEM, * p < 0.05 versus the control rats; ns, no significance (n = 6).

Figure 4

Cobalt chloride and fibrinogen induced toxicity in cultured cortical astrocytes of rats for 24 h. (A,B) Cell viability was determined using a CCK assay. (A) CoCl2 caused a dose-dependent effect on astrocyte viability (n = 3). (B) The effect of fibrinogen was not concentration-dependent or prominent (n = 9). Data were expressed as mean ± SEM, ** p < 0.01 versus the control group; ns, no significance.

Figure 5

Reactive phenotype in cultured cortical astrocytes of rats by different concentrations of CoCl2 and time duration. (A–C) After 24 h, the protein levels of C3 and S100a10 were tested using Western blot, and the data were normalized to Gapdh or Actin blot. (D–F) After 96 h, the protein levels of C3 and S100a10 were tested using Western blot, and the data were normalized to Gapdh or Actin blot. Data are expressed as mean ± SEM, * p < 0.05 versus the control group; ns, no significance (n = 3).

Figure 6

Reactive phenotype in cultured cortical astrocytes of rats by cobalt chloride and low glucose for 24 h. (A–D) The protein levels of C3, S100a10, and GFAP were tested using Western blot and the data were normalized to Gapdh or Actin blot. (E,F) A1 and A2 reactive astrocyte phenotypes were tested using QPCR. Data were expressed as mean ± SEM, * p < 0.05,** p < 0.01 versus the control group; ns, no significance (n = 3).

Figure 7

Reactive phenotype in cultured cortical astrocytes of rats by fibrinogen for 24 h. (A,B) The protein levels of C3, S100a10, and GFAP were tested using Western blot, and the data were normalized to Gapdh or Actin blot. (C,D) A1 and A2 reactive astrocyte phenotypes were tested using QPCR. (E–G) Immunofluorescence staining of GFAP using control (E). Low concentration of fibrinogen and (F) high concentration of fibrinogen (G). (H) Analysis of cell area in immunofluorescence staining of GFAP. Data are expressed as mean ± SEM, * p < 0.05, ** p < 0.01 versus the control group; # p < 0.05, high concentration versus low concentration of fibrinogen group; ns, no significance (n = 3).

Figure 8

Function of reactive cultured cortical astrocytes of rats by cobalt chloride and low glucose for 72 h. (A) Representative blots of CX43, Aqua4, ApoE, and GFAP. (B–E) Ratios of CX43, Aqua4, ApoE, and GFAP to Actin or Gapdh were calculated and compared. Data were expressed as mean ± SEM, * p < 0.05 versus the control group; ns, no significance (n = 3).

Figure 9

Effects of C6 cells by cobalt chloride or low glucose for 24 h. (A) Cell viability was determined using a CCK assay, and CoCl2 caused a dose-dependent effect on C6 cell viability. (B) Representative blots of CX43, Aqua4, and ApoE. (C–E) Ratios of CX43, Aqua4, and ApoE to Gapdh or Actin were calculated and compared. Data were expressed as mean ± SEM, * p < 0.05, ** p < 0.01 versus the control group; ns, no significance (n = 3).

Figure 10

Signal pathway proteins of C6 cells by cobalt chloride or low glucose for 24 h. (A) Representative blots of pERK, pJNK, pAKT. (B–D) Ratios of pERK, pJNK, pAKT to ERK, JNK, and AKT were calculated and compared. Data were expressed as mean ± SEM, * p < 0.05, ** p < 0.01 versus the control group; ns, no significance (n = 3).