摘要目的:观察外源性miR-146a-5p对脂多糖（lipopolysaccharide，LPS）诱导的小鼠胚胎吸收和胎鼠发育不良的改善作用，并初步探讨其作用机制。方法:①将36只成年雌鼠与雄鼠交配，分别于交配前（D0/未孕）、孕第0.5天（day 0.5，D0.5，即见栓当日）、孕第4.5天（day 4.5，D4.5）、孕第7.5天（day 7.5，D7.5）、孕第9.5天（day 9.5，D9.5）和孕第13.5天（day 13.5，D13.5）收集小鼠子宫组织，通过实时荧光定量PCR（quantitative PCR，qPCR）和Western blotting检测不同妊娠期小鼠子宫组织中miR-146a-5p及其靶基因TRAF6蛋白的表达水平；②对孕D7.5小鼠腹腔分别注射生理盐水（对照，记为COL组）、LPS 250 μg/kg（记为LPS250组）、LPS合并尾静脉注射10 nmol miR-146a-5p无关序列（negative control，NC，记为LPS250+NC组）、LPS合并尾静脉注射10 nmol miR-146a-5p激动剂（miR-146a-5p agomir，记为LPS250+miR-146a-5p agomir组），孕第8.5天（day 8.5，D8.5）对小鼠宫内总胚胎数和吸收胚胎数进行计量分析，通过qPCR和Western blotting分别检测子宫组织中肿瘤坏死因子α（tumor necrosis factor α，TNFα）mRNA和TRAF6蛋白表达水平；③将LPS剂量减为50 μg/kg后，将孕D7.5小鼠分为2组，每组3只：一组腹腔注射100 μL LPS（50 μg/kg），同时鼠尾静脉注射10 nmol 的无关序列，记为LPS50+NC组；另一组行腹腔注射100 μL LPS（50 μg/kg），同时鼠尾静脉注射10 nmol 的miR-146a-5p agomir，记为 LPS50+miR-146a-5p agomir组，孕第16.5天（day 16.5，D16.5）对小鼠宫内总胎数/胚胎数、吸收胚胎数、存活胎数及存活胎鼠重量和胎盘重量进行计量分析；④分离培养原代小鼠骨髓源性巨噬细胞（bone marrow-derived macrophages，BMDM），利用LPS刺激诱导其M1极化，再瞬时转染miR-146a-5p模拟物（miR-146a-5p mimics）或其NC片段后，通过qPCR和Western blotting分别检测细胞中 TNFα mRNA和pSTAT1蛋白表达水平。 结果:miR-146a-5p在孕D7.5、D9.5和D13.5小鼠子宫植入部位的表达水平显著高于非植入部位（ P=0.013、 P=0.012、 P=0.003），TRAF6蛋白在D13.5植入部位的表达水平显著低于非植入部位（ P=0.012）。对孕D7.5小鼠腹腔注射250 μg/kg LPS后，孕D8.5时LPS250组的胚胎吸收率为43.13%±3.31%，显著高于COL组（0%， P=0.002），而LPS250+miR-146a-5p agomir组的胚胎吸收率（13.50%±0.87%）显著低于LPS250+NC组（59.33%±4.04%， P=0.001）。当对孕D7.5小鼠腹腔注射低剂量LPS（50 μg/kg）后，D16.5时LPS50+miR-146a-5p agomir组存活胎鼠重量［（0.29±0.09）g］及胎盘重量［（0.06±0.02）g］均显著高于LPS50+NC组［（0.46±0.06）g， P<0.001；（0.07±0.02）g， P=0.021］，两组间吸收胚胎数及胚胎吸收率差异均无统计学意义（均 P>0.05）。与转染NC的BMDM细胞相比，转染miR-146a-5p mimics的BMDM细胞中，pSTAT1蛋白和 TNFα mRNA的表达水平都显著下调（ P=0.012、 P=0.039）。 结论:miR-146a-5p在小鼠胚胎植入后期及胎盘发育期母-胎界面的表达水平显著增高，外源性miR-146a-5p能有效改善LPS诱导的小鼠胚胎吸收和胎鼠发育不良，miR-146a-5p能抑制小鼠巨噬细胞的M1极化活性，提示miR-146a-5p可能通过抑制小鼠母-胎界面巨噬细胞的M1极化而保障妊娠的正常建立和维持。

abstractsObjective:To observe the ameliorative effects of exogenous miR-146a-5p on lipopolysaccharide (LPS)-induced embryonic resorption and fetal mouse dysplasiamice, and to preliminarily investigate its mechanism of action.Methods:1) After 36 healthy adult female mice were mated with male mice, uterine tissues were collected from females on day (D) 0 (D0/not pregnant), D0.5 (the day of embryo observed), D4.5, D7.5, D9.5 and D13.5 of gestation, and the expression levels of miR-146a-5p and its target gene TRAF6 protein in uterine tissues of mice at different gestation periods were detected by real-time fluorescent quantitative PCR (qPCR) and Western blotting. 2) The mice on D7.5 of pregnancy were treated with intraperitoneal injection of saline (control, COL group), intraperitoneal injection of 250 μg/kg LPS (named LPS250 group), LPS combined with tail vein injection of 10 nmol miR-146a-5p unrelated sequence (negative control, NC, named LPS250+NC group), or LPS combined with tail vein injection of 10 nmol miR-146a-5p agonist (miR-146a-5p agomir, named LPS250+miR-146a-5p agomir group). The total number of embryos and the number of absorbed embryos in the uterus of pregnant mice were measured and statistically analyzed on D8.5, and the expression levels of TNFα mRNA and TRAF6 protein in uterine tissues were detected by qPCR and Western blotting. 3) Then we reduced the dosage of LPS to 50 μg/kg and treated the same groups, named LPS50+NC group, LPS50+miR-146a-5p agomir group, respectively. The total number of fetal mice/embryos, the number of absorbed embryos, the number of surviving fetal mice, the weight of surviving fetal mice and the weight of the placenta were measured and statistically analyzed on D16.5. 4) Primary mouse bone marrow-derived macrophages (BMDM) were isolated and cultured. Mouse BMDM was inducted to M1 polarization by LPS stimulation, and then was transient transfected miR-146a-5p mimics or their NC fragments. The expression levels of TNFα mRNA and pSTAT1 protein were detected by qPCR and Western blotting. Results:The expression level of miR-146a-5p was significantly higher in the implantation sites of D7.5, D9.5 and D13.5 pregnant mice than in the non-implantation sites ( P=0.013, P=0.012, P=0.003), and the protein expression level of TRAF6 was significantly lower in the implantation site of D13.5 pregnant mice than in the non-implantation site ( P=0.012). After intraperitoneal injection of 250 μg/kg of LPS into D7.5 pregnant mice, the embryo absorption rate of the LPS group on D8.5 was 43.13%±3.31%, which was significantly higher than that of COL group (0%, P=0.002), while the embryo absorption rate of the LPS250+miR-146a-5p agomir group (13.50%±0.87%) was significantly lower than that of the LPS250+NC group (59.33%±4.04%, P=0.001). After intraperitoneal injection of 50 μg/kg of LPS combined with tail vein injection of NC or miR-146a-5p agomir to D7.5 pregnant mice, the fetal mouse weight [(0.29±0.09) g] and placental weight [(0.06±0.02) g] of surviving fetal mice in the LPS50+NC group on D16.5 and the LPS50+miR-146a-5p agomir group were statistically significant [(0.46±0.06) g, P<0.001; (0.07±0.02) g, P=0.021], and the differences in the number of absorbed embryos and embryo uptake rate between the two groups were not statistically significant (all P>0.05). The expression levels of both pSTAT1 protein and TNFα mRNA were significantly downregulated in BMDM transfected with miR-146a-5p mimics compared with those transfected with NC ( P=0.012, P=0.039). Conclusion:miR-146a-5p expression levels were significantly increased at the maternal-fetal interface during the late stage of mouse embryo implantation and placental development. Exogenous miR-146a-5p could effectively improve LPS-induced mouse embryo resorption and fetal mouse dysplasia. miR-146a-5p could inhibit the M1 polarization activity of mouse macrophages, suggesting that miR-146a-5p may inhibit the M1 polarization activity of mouse macrophages by suppressing M1 polarization of mouse maternal-fetal interface macrophages to safeguard the normal establishment and maintenance of pregnancy.