Qiliqiangxin protects against anoxic injury in cardiac microvascular endothelial cells via NRG-1/ErbB-PI3K/Akt/mTOR pathway

Abstract

Cardiac microvascular endothelial cells (CMECs) are important angiogenic components and are injured rapidly after cardiac ischaemia and anoxia. Cardioprotective effects of Qiliqiangxin (QL), a traditional Chinese medicine, have been displayed recently. This study aims to investigate whether QL could protect CMECs against anoxic injury and to explore related signalling mechanisms. CMECs were successfully cultured from Sprague-Dawley rats and exposed to anoxia for 12 hrs in the absence and presence of QL. Cell migration assay and capillary-like tube formation assay on Matrigel were performed, and cell apoptosis was determined by TUNEL assay and caspase-3 activity. Neuregulin-1 (NRG-1) siRNA and LY294002 were administrated to block NRG-1/ErbB and PI3K/Akt signalling, respectively. As a result, anoxia inhibited cell migration, capillary-like tube formation and angiogenesis, and increased cell apoptosis. QL significantly reversed these anoxia-induced injuries and up-regulated expressions of NRG-1, phospho-ErbB2, phospho-ErbB4, phospho-Akt, phospho-mammalian target of rapamycin (mTOR), hypoxia-inducible factor-1α (HIF-1α) and vascular endothelial growth factor (VEGF) in CMECs, while NRG-1 knockdown abolished the protective effects of QL with suppressed NRG-1, phospho-ErbB2, phospho-ErbB4, phospho-Akt, phospho-mTOR, HIF-1α and VEGF expressions. Similarly, LY294002 interrupted the beneficial effects of QL with down-regulated phospho-Akt, phospho-mTOR, HIF-1α and VEGF expressions. However, it had no impact on NRG-1/ErbB signalling. Our data indicated that QL could attenuate anoxia-induced injuries in CMECs via NRG-1/ErbB signalling which was most probably dependent on PI3K/Akt/mTOR pathway.

Keywords: Qiliqiangxin; angiogenesis; anoxia; apoptosis; cardiac microvascular endothelial cell; neuregulin.

Figure 1

QL promoted CMECs migration and capillary‐like tube formation, which were abolished by NRG‐1 siRNA and LY294002 administration. Three random fields from each well were assessed. (A) Representative inverted‐phase contrast microscope images (100×) of cell migration and invasion of the scratch among different groups, as exhibited by percentage of scratch areas, which were quantified using Wimasis WimScratch software from 0 to 12 hrs of anoxia. (B) Tube networks including total number of loops, tubules and branch points were analysed under inverted‐phase contrast microscope (50×) at the end of experiment. Furthermore, a detailed overlay images were provided using Wimasis WimTube software, in which cell‐covered area was in blue, tubules in red, branch points in white and loops by yellow mark. (C) Bar graphs for percentage of scratch area and counts of loops, tubules, branch points among different groups. Percentage of scratch area was counted as scratch area divided by the total area in the same field. Capability of capillary‐like tube formation was described as total number of loops, tubules and branch points. Compared with control group, *P < 0.05, **P < 0.01; compared with anoxia group, ^# P < 0.05, ^{# #} P < 0.01; compared with anoxia+QL group, ^& P < 0.05.

Figure 2

QL significantly reduced the percentage of TUNEL‐positive cells and caspase‐3 activity in CMECs exposed to anoxia. Both NRG‐1 siRNA and LY294002 attenuated such anti‐apoptotic effect. Representative photographs of double staining of TUNEL and DAPI at 12 hrs of anoxia were displayed. Red indicates TUNEL‐positive nuclei; blue indicates DAPI signals. Quantification of CMECs apoptosis was expressed as ratio of TUNEL‐positive nuclei to total nuclei in each randomly chosen field(200×). Caspase‐3 activity was also detected and expressed as fold changes compared with the control value. *P < 0.05, **P < 0.01.

Figure 3

QL up‐regulated expressions of NRG‐1, phospho‐ErbB2, phospho‐ErbB4, phospho‐Akt, phospho‐mTOR, HIF‐1α and VEGF in CMECs exposed to anoxia. NRG‐1 siRNA transfection abolished these up‐regulations. LY294002 also interrupted up‐regulation of phospho‐Akt, phospho‐mTOR, HIF‐1α and VEGF by QL, but did not interfere with NRG‐1 and ErbB expressions. PPARγ expression was not regulated by QL. (A) Western blotting analyses for NRG‐1, phospho‐ErbB2, phospho‐ErbB4, phospho‐Akt, phospho‐mTOR, PPARγ, HIF‐1α and VEGF expressions in CMECs subjected to anoxia, QL treatment and NRG‐1 siRNA transfection. a1–a8 showed corresponding bar graphs. GAPDH and total Akt expressions served as internal control. (B) Western blotting analyses for NRG‐1, phospho‐ErbB2, phospho‐ErbB4, phospho‐Akt, phospho‐mTOR, PPARγ, HIF‐1α and VEGF expressions in CMECs subjected to anoxia, QL treatment and LY294002. b1–b8 showed corresponding bar graphs. GAPDH and total Akt expressions served as internal control. Representative blots from five independent experiments were displayed for (A) and (B). Compared with anoxia group, *P < 0.05, **P < 0.01; compared with anoxia+QL group, ^# P < 0.05, ^{# #} P < 0.01. (C) Real‐time quantitative PCR analyses for NRG‐1 and HIF‐1α mRNA expression normalized to GAPDH. Relative values were calculated for the ratio of NRG‐1 or HIF‐1α to GAPDH and expressed as folds of control. Compared with anoxia group, **P < 0.01; compared with anoxia+QL group, ^{# #} P < 0.01. (D) NRG‐1 concentration in cell supernatants detected by ELISA. QL promoted NRG‐1 secretion into extracellular space, which was abolished by NRG‐1 siRNA, but not by LY294002. **P < 0.01.