Retinoic Acid Induced-Autophagic Flux Inhibits ER-Stress Dependent Apoptosis and Prevents Disruption of Blood-Spinal Cord Barrier after Spinal Cord Injury

Abstract

Spinal cord injury (SCI) induces the disruption of the blood-spinal cord barrier (BSCB) which leads to infiltration of blood cells, an inflammatory response, and neuronal cell death, resulting spinal cord secondary damage. Retinoic acid (RA) has a neuroprotective effect in both ischemic brain injury and SCI, however the relationship between BSCB disruption and RA in SCI is still unclear. In this study, we demonstrated that autophagy and ER stress are involved in the protective effect of RA on the BSCB. RA attenuated BSCB permeability and decreased the loss of tight junction (TJ) molecules such as P120, β-catenin, Occludin and Claudin5 after injury in vivo as well as in Brain Microvascular Endothelial Cells (BMECs). Moreover, RA administration improved functional recovery in the rat model of SCI. RA inhibited the expression of CHOP and caspase-12 by induction of autophagic flux. However, RA had no significant effect on protein expression of GRP78 and PDI. Furthermore, combining RA with the autophagy inhibitor chloroquine (CQ) partially abolished its protective effect on the BSCB via exacerbated ER stress and subsequent loss of tight junctions. Taken together, the neuroprotective role of RA in recovery from SCI is related to prevention of of BSCB disruption via the activation of autophagic flux and the inhibition of ER stress-induced cell apoptosis. These findings lay the groundwork for future translational studies of RA for CNS diseases, especially those related to BSCB disruption.

Keywords: autophagy; blood-spinal cord barrier (BSCB); endocytoplasmic reticulum (ER) stress; retinoic acid (RA); spinal cord injury (SCI).

Figure 1

RA inhibits the increase in BSCB permeability and improves functional recovery after SCI. After SCI, rats were treated with RA and BSCB permeability was measured at 24 h post-SCI by using Evan's Blue dye (n = 5/group). (A) The BBB scores of sham, SCI group and SCI rat treated with RA group. ^* P < 0.05 versus the SCI group, and ^** P < 0.01 versus the SCI group, n = 5. (B) Representative whole spinal cords showing Evan's Blue dye permeabilized into the spinal cord at 24 h post-SCI. (C) Quantification of BSCB permeability data using B by ImageJ software ^**P < 0.01, versus the SCI group, n=5, F=1.848. (D) Quantification of the EB content of the spinal cord (μ g/g), ^**P < 0.01, versus the SCI group, n=5, F=1.482. All data represent mean values ± SEM.

Figure 2

RA induces “Autophagic Flux” in acute SCI. (A) Representative western blots and quantification data of autophagy marker LC3-II in Sham group, SCI rats and treated with RA group. ^**P < 0.01, versus the SCI group, n=5, F=2.122. (B) Representative micrographs showing immunofluorescence with LC3-II. Nuclei are labeled with DAPI (blue) in each group. (C) Representative western blots and quantification data of autophagy marker P62 in Sham group, SCI rats and treated with RA group. ^*P < 0.05, versus the SCI group, n=5, F=1.788.

Figure 3

RA prevents the loss of TJ and AJ proteins after SCI. (A) Representative western blots of AJ proteins β-catenin and P120-catenin in the sham, SCI model and SCI model treated RA groups. (B), (C) Quantification of western blot data from A. (D) Representative western blots of TJ proteins Occludin and Claudin5 in the sham, SCI model and SCI model treated RA groups. (E), (F) Quantification of western blot data from D. All data represent Mean values ± SEM, n =5, ^**P < 0.01, versus the Sham group, ^#P < 0.05, ^##P < 0.01 versus the SCI group, F < 2.749 in all T- tests. (G) Representative micrographs showing double immunofluorescence with Occludin (green) and CD31 (endothelial cell marker, red), nuclei are labeled with DAPI (blue ) in each group, n=5.

Figure 4

RA inhibits ER stress-associated CHOP and caspase-12 expression after SCI. (A) Representative western blots of ER stress markers GRP78, and PDI in the sham, SCI model and SCI model treated RA groups. (B), (C) Quantification of western blot data from A. There were no significant alterations on protein expression of GRP78 and PDI between SCI model and RA treated group ^%P = 0.45, ^&P = 0.70. F_GRP78 = 1.247, F_PDI = 1.026. (D) Representative western blots of CHOP and caspase-12 in the sham, SCI model and SCI model treated RA groups. (E), (F) Quantification of western blot data from D. ^**P < 0.01, versus the SCI group, F_CHOP = 3.569, F_caspase-12 =1.369. All data represent Mean values ± SEM, n =5.

Figure 5

Inhibition of autophagy by CQ abolishes the BSCB protective effect of RA and it's inhibition of apoptotic proteins expression after SCI. (A) Representative whole spinal cords showing Evan's Blue dye permeabilized into spinal cord at 24 h post-SCI. (B) Quantification of BSCB permeability data from B by ImageJ software, ^##P < 0.01 versus SCI group, F =1.653; ^**P < 0.01 versus treated RA groups, n = 4, F = 1.516. (C), (D) Representative western blots and quantification data of TJ proteins Occludin, Claudin5 and β-catenin in each group rats. ^!!P< 0.01, ^@@P< 0.01, ^$$P< 0.01 versus SCI group. ^**P< 0.01, ^##P< 0.01, ^&&P< 0.01 versus treated RA groups, F < 2.139 in all T-tests. (E), (F), (G), (H) Representative western blots and quantification data of ER stress markers GRP78, PDI, CHOP, and Caspase-12 in the sham, SCI model and SCI model treated RA groups, RA compound with CQ, and CQ alone groups. F_PDI =0.723, P = 0.566; F_GRP78 = 2.171, P =0.169. ^&P< 0.05, ^%P< 0.05, versus SCI group. ^**P< 0.01, ^##P< 0.01, versus treated RA groups. All data represent Mean values ± SEM, n = 5, F < 3.034 in all T-tests. (I) Representative micrographs showing double immunofluorescence with NeuN (neuron cell marker, green) and CHOP (red), nuclei are labeled with DAPI (blue ) in each group.

Figure 6

RA protects TG-treated ECs by inducing “Autophagic Flux” in vitro. ECs were treated with TG (10μM) or treated with RA (1 μM) or RA (5 μM), or compound with CQ (100μM) for 6h. (A) MTT results of RA-treated endothelial cells induced by TG. All experiments were repeated four times. All data represent Mean values ± SEM, ^**P< 0.01, versus TG treated cells, F = 1.225. (B), (C), (D) Representative western blots and quantification data of autophagy marker LC3-II and P62 in each group cells. ^**P< 0.01, versus TG treated cells, F_LC3-II = 6.288; F_P62 = 1.067. (E), (F) Representative western blots and quantification data of tight junction proteins Occludin, P120 and β-catenin in each group cells. ^!!P< 0.01, ^#P< 0.05, ^$$P< 0.01 versus TG group. ^**P< 0.01, ^##P< 0.01, ^&&P< 0.01 versus TG groups, F < 2.834 in all T- tests. (G) immunofluorescence staining of Occludin (green) in ECs treated with TG for 6 h, nuclei are labeled with DAPI (blue). (H), (I) Representative western blots and quantification data of ER stress markers GRP78, PDI, CHOP, and Caspase-12 in each group cells, F_PDI =0.678, P = 0.543; F_GRP78 = 0.021, P =0.979. ^%%P< 0.01, ^&&P< 0.01, versus TG group. ^**P< 0.01, ^##P< 0.01, versus treated RA groups. All data represent Mean values ± SEM, n = 4, F < 3.517 in all T-tests.

Figure 7

A model illustrating the BSCB protective effect of RA after SCI. Both autophagy and ER stress activation were involved in SCI, which aim to either repair or further damage the BSCB. Autophagy plays a central role in protecting SCI-induced BSCB disruption, which was induced by RA. ER stress-induced expression of CHOP and caspase-12 activation contributed to ECs death after SCI, which was inhibited by RA.