miR-10b promotes porcine immature Sertoli cell proliferation by targeting the DAZAP1 gene

MicroRNAs(miRNAs) have been widely identified in porcine testicular tissues and implicated as crucial regulators of proliferation, apoptosis, and differentiation in porcine spermatogenesis related cells. However, the function roles of most of the miRNAs that have been identified in Sertoli cells are poorly understood. In the present study, six experiments were conducted to study the regulatory role of miR-10b in porcine immature Sertoli cells. In experiment 1, the results showed that the relative mRNA expression level of miR-10b in porcine testicular tissues decreased quadratically(P<0.001) with increasing age, while the relative mRNA expression level of DAZAP1 gene increased(P<0.001). In addition, the mRNA expression of miR-10b was negatively(P<0.01) correlated with DAZAP1 mRNA expression(r=–0.550). In experiment 2, the results from the bioinformatic analysis and a luciferase reporter assay demonstrated that miR-10b directly targeted the DAZAP1 gene in porcine immature Sertoli cells. DAZAP1 mRNA and protein expressions were both regulated(P<0.05) by miR-10b. In experiments 3 to 5, the over-expression of miR-10b or the siRNA-mediated knockdown of the DAZAP1 gene promoted(P<0.05) porcine immature Sertoli cell proliferation, as determined by the Cell Counting Kit-8(CCK-8) assay and the 5-Ethynyl-2′-deoxyuridine(EdU) assay. However, an annexin V-FITC/PI staining assay and the expression of cell survival-related genes indicated that over-expression of miR-10b or knockdown of DAZAP1 had no effect(P>0.05) on porcine immature Sertoli cell apoptosis. In experiment 6, the co-transfection treatment results showed that miR-10b promoted(P<0.05) porcine immature Sertoli cell proliferation by targeting DAZAP1 gene. Overall, these experiments demonstrated that miR-10b promotes porcine immature Sertoli cell proliferation by targeting the DAZAP1 gene.

miR-34c inhibits proliferation and enhances apoptosis in immature porcine Sertoli cells by targeting the SMAD7 gene

MicroRNAs(miRNAs) are implicated in swine spermatogenesis via their regulations of cell proliferation, apoptosis, and differentiation. Recent studies indicated that miR-34 c is indispensable in the late steps of spermatogenesis. However, whether miR-34 c plays similar important roles in immature porcine Sertoli cells remain unknown. In the present study, we conducted two experiments using a completely randomised design to study the function roles of miR-34 c. The results from experiment I demonstrated that the relative expression level of miR-34 c in swine testicular tissues increased(P=0.0017) quadratically with increasing age, while the relative expression level of SMAD family member 7(SMAD7) decreased(P=0.0009) with curve. Furthermore, miR-34 c expression levels showed a significant negative correlation(P=0.013) with SMAD7 gene expression levels. The results from experiment II indicated that miR-34 c directly targets the SMAD7 gene using a luciferase reporter assay, and suppresses(P<0.05) SMAD7 mRNA and protein expressions in immature porcine Sertoli cells. Overexpression of miR-34 c inhibited(P<0.05) proliferation and enhanced(P<0.05) apoptosis in the immature porcine Sertoli cells, which was supported by the results from the Cell Counting Kit-8(CCK-8) assay, the 5-Ethynyl-2′-deoxyuridine(EdU) assay, and the Annexin V-FITC/PI staining assay. Furthermore, knockdown of SMAD7 via small interfering RNA(siR NA) gave a similar result. It is concluded that miR-34 c inhibits proliferation and enhances apoptosis in immature porcine Sertoli cells by targeting the SMAD7 gene.

基于EEMD-LMD-LSTM-LEC深度学习模型的短时物流需求预测

短时物流需求预测是智慧物流系统的重要组成部分.由于短时物流需求数据具有非平稳性、强随机性、局部突变、非线性等特征,精确预测较为困难.对此,考虑集成经验模态分解(EEMD)、局部均值分解(LMD)、长短期记忆网络(LSTM)以及考虑局部误差校正(LEC),提出用于短时物流需求预测的EEMD-LMD-LSTM-LEC深度学习模型.该预测模型分为两个阶段:第1阶段基于特征分解和特征提取,构建EEMD-LMD-LSTM模型,以降低非线性的原始短时物流需求不平稳及随机变化导致的预测误差;第2阶段构建局部误差校正模型,用于校正第1阶段的预测结果,以减少短时物流需求的局部突变带来的预测误差.实验结果表明,EEMD-LMD-LSTM-LEC短时物流需求预测模型在均方根误差、绝对误差均值、绝对误差百分比和校正决定系数方面,均优于其他11种对比模型,其中包括:数理统计模型—–ARIMA;浅层机器学习模型—–支持向量回归和BP神经网络;深度学习模型—–LSTM和卷积神经网络;组合模型——深度置信网络-LSTM、经验模态分解(EMD)-LSTM、EEMD-LSTM、LMD-LSTM、EMD-LMD-LSTM和EEMD-LMD-LSTM.