Thioredoxin-interacting protein links endoplasmic reticulum stress to inflammatory brain injury and apoptosis after subarachnoid haemorrhage

Abstract Background Early brain injury (EBI) is considered a major contributor to the high morbidity and mortality associated with subarachnoid haemorrhage (SAH). Both of sterile inflammation and apoptosis are considered the important causes of EBI. Recently, it was confirmed that thioredoxin-interacting protein (TXNIP) not only participates in inflammatory amplification but also stimulates the apoptosis signalling cascade pathway. However, whether the effects of TXNIP influence the pathogenesis of SAH remains unclear. Here, we hypothesize that TXNIP activity induced by endoplasmic reticulum stress (ER stress) may contribute to the pathogenesis of EBI through pro-inflammatory and pro-apoptotic mechanisms. Methods A total of 299 male Sprague–Dawley rats were used to create SAH models. Resveratrol (RES, 60 mg/kg) and two TXNIP small interfering RNA (siRNA) were used to inhibit TXNIP expression. The specific inhibitors of ER stress sensors were used to disrupt the link between TXNIP and ER stress. SAH grade, neurological deficits, brain water content and blood–brain barrier (BBB) permeability were evaluated simultaneously as prognostic indicators. Fluorescent double-labelling was employed to detect the location of TXNIP in cerebral cells. Western blot and TUNEL were performed to study the mechanisms of TXNIP and EBI. Results We found that TXNIP expression significantly increased after SAH, peaking at 48 h (0.48 ± 0.04, up to 3.2-fold) and decreasing at 72 h after surgery. This process was accompanied by the generation of inflammation-associated factors. TXNIP was expressed in the cytoplasm of neurons and was widely co-localized with TUNEL-positive cells in both the hippocampus and the cortex of SAH rats. We discovered for the first time that TXNIP was co-localized in neural immunocytes (microglia and astrocytes). After administration of RES, TXNIP siRNA and ER stress inhibitors, TXNIP expression was significantly reduced and the crosstalk between TXNIP and ER stress was disrupted; this was accompanied by a reduction in inflammatory and apoptotic factors, as well as attenuation of the prognostic indices. Conclusions These results may represent the critical evidence to support the pro-inflammatory and pro-apoptotic effects of TXNIP after SAH. Our data suggest that TXNIP participates in EBI after SAH by mediating inflammation and apoptosis; these pathways may represent a potential therapeutic strategy for SAH treatment.