LAUSR

OBJECTIVE To investigate the protective effect of the hydrogen (H2) inhalation on intestinal injury in severe sepsis mice and its relationship with nuclear factor E2-related factor 2 (Nrf2)/heme oxygenase-1 (HO-1)/high mobility group protein B1 (HMGB1) pathway. METHODS The 280 male wild-type (WT) and Nrf2 knockout (Nrf2-KO) ICR mice (6-8 weeks old, 20-25 g weight) were assigned into 8 groups randomly (WT and Nrf2-KO mice, 4 groups respectively, n = 35): Sham group, H2 inhalation group (Sham+H2 group), sepsis model group which was induced by cecal ligation and puncture (CLP group) and H2 treatment group (CLP+H2 group). In Sham group, except cecal ligation and perforation, the same surgery was used. The mice in Sham group and CLP group inhaled only air for 1 hour and 6 hours, respectively, while the mice in Sham+H2 group and CLP+H2 group inhaled 2% H2 for 60 minutes. After modeling, 20 mice randomly in each group were selected to measure 7-day natural survival rates; the remaining mice(15 mice) were killed in 24 hours after modeling, and the bacterial load was determined by collecting the peritoneal lavage fluid (PLF) of each mouse to measure the colony forming unit (CFU); and the intestinal tissue was taken after perfusion, and the intestinal tissue of 3 mice was used to extract protein for HO-1 and HMGB1 levels by using Western blotting. Intestinal tissue of 3 mice was extracted to prepare histomorphology sections for the identification of HO-1. Intestinal tissue of 3 mice was used to detect the expression of HMGB1 by immunofluorescence. RESULTS The 7-day survival rate of WT and Nrf2-KO mice in CLP group was zero; in WT mice, the 7-day survival rate of CLP+H2 group was significantly higher than that of CLP group (45% vs. 0%, P < 0.05), while no difference of the 7-day survival rate in Nrf2-KO mice was noted between CLP and CLP+H2 groups (0% vs. 0%, P > 0.05). Compared with Sham group, the CFU in the PLF of both mice in CLP group increased significantly. Compared with CLP group, 2% H2 treatment could significantly improve the bacterial clearance rate of WT mice with severe sepsis (CFU, ×103: 34.7±6.3 vs. 74.2±8.1, P < 0.01), but had no significant effect on Nrf2-KO mice with severe sepsis (CFU, ×103: 73.3±7.5 vs. 75.8±8.6, P > 0.05). The results of Western blotting, immunohistochemical and immune-fluorescence showed that the expression of HO-1 and HMGB1 in the CLP group were higher than those in the Sham group. Compared with CLP group, the expression of HO-1 in WT mice of CLP+H2 group was significantly higher [HO-1 protein expression (HO-1/β-actin): 0.716±0.035 vs. 0.460±0.045, HO-1 positive expression (A value): 0.703±0.135 vs. 0.430±0.116, both P < 0.01], while the expression of HMGB1 was significantly lower [HMGB1 protein expression (HMGB1/β-actin): 0.052±0.038 vs. 0.082±0.008, HMGB1 positive expression (A value): 24.530±9.317 vs. 41.830±2.458, both P < 0.01]. In Nrf2-KO mice, compared with CLP group, there was no statistical difference in the above indexes in CLP+H2 group [HO-1 protein expression (HO-1/β-actin): 0.148±0.004 vs. 0.164±0.043, HO-1 positive expression (A value): 0.109±0.020 vs. 0.113±0.025, HMGB1 protein expression (HMGB1/β-actin): 0.109±0.020 vs. 0.113±0.025, HMGB1 positive expression (A value): 39.570±12.660 vs. 42.790±9.680, all P > 0.05]. CONCLUSIONS H2 can inhibit the intestinal injury of mice with severe sepsis through Nrf2/HO-1/HMGB1 pathway, and Nrf2 plays an important role in the treatment of intestinal injury with H2.