Data_Sheet_5_A Combined Flow Cytometric Semen Analysis and miRNA Profiling as a Tool to Discriminate Between High- and Low-Fertility Bulls.xlsx

Predicting bull fertility is one of the main challenges for the dairy breeding industry and artificial insemination (AI) centers. Semen evaluation performed in the AI center is not fully reliable to determine the level of bull fertility. Spermatozoa are rich in active miRNA. Specific sperm-borne miRNAs can be linked to fertility. The aim of our study is to propose a combined flow cytometric analysis and miRNA profiling of semen bulls with different fertility to identify markers that can be potentially used for the prediction of field fertility. Sperm functions were analyzed in frozen-thawed semen doses (CG: control group) and high-quality sperm (HQS) fraction collected from bulls with different field fertility levels (estimated relative conception rate or ERCR) by using advanced techniques, such as the computer-assisted semen analysis system, flow cytometry, and small RNA-sequencing. Fertility groups differ for total and progressive motility and in the abnormality degree of the chromatin structure (P < 0.05). A backward, stepwise, multiple regression analysis was applied to define a model with high relation between in vivo (e.g., ERCR) and in vitro (i.e., semen quality and DE-miRNA) fertility data. The analysis produced two models that accounted for more than 78% of the variation of ERCR (CG: R2 = 0.88; HQS: R2 = 0.78), identifying a suitable combination of parameters useful to predict bull fertility. The predictive equation on CG samples included eight variables: four kinetic parameters and four DNA integrity indicators. For the HQS fraction, the predictive equation included five variables: three kinetic parameters and two DNA integrity indicators. A significant relationship was observed between real and predicted fertility in CG (R2 = 0.88) and HQS fraction (R2 = 0.82). We identified 15 differentially expressed miRNAs between high- and low-fertility bulls, nine of which are known (miR-2285n, miR-378, miR-423-3p, miR-191, miR-2904, miR-378c, miR-431, miR-486, miR-2478) while the remaining are novel. The multidimensional preference analysis model partially separates bulls according to their fertility, clustering three semen quality variable groups relative to motility, DNA integrity, and viability. A positive association between field fertility, semen quality parameters, and specific miRNAs was revealed. The integrated approach could provide a model for bull selection in AI centers, increasing the reproductive efficiency of livestock.

公牛生育力预测是奶牛育种产业与人工授精（Artificial Insemination，AI）中心面临的核心挑战之一。在人工授精中心开展的精液评估，无法完全可靠地判定公牛的生育力水平。精子富含活性微RNA（microRNA，miRNA），特定的精子源性微RNA可与生育力存在关联。本研究旨在针对不同生育力水平的种公牛精液，联合开展流式细胞术分析与微RNA表达谱分析，以筛选可潜在用于田间生育力预测的生物标志物。本研究借助计算机辅助精液分析系统、流式细胞术、小RNA测序等先进技术，对来自不同田间生育力水平（以估计相对受胎率，estimated relative conception rate，ERCR衡量）的种公牛的冻精剂量（CG：对照组，control group）及分选得到的优质精子组分（high-quality sperm fraction，HQS）开展精子功能分析。不同生育力组的精子总活力、前向运动活力以及染色质结构异常程度均存在显著差异（P < 0.05）。本研究采用逐步向后多元回归分析，构建可有效关联体内（如ERCR）与体外（即精液质量及差异表达微RNA，differentially expressed miRNA，DE-miRNA）生育力数据的预测模型。分析结果得到两个模型，可解释超过78%的ERCR变异（对照组R²=0.88；优质精子组分R²=0.78），筛选得到可用于预测公牛生育力的最优参数组合。对照组样本的预测方程包含8个变量：4项运动动力学参数与4项DNA完整性指标；优质精子组分的预测方程则包含5个变量：3项运动动力学参数与2项DNA完整性指标。对照组（R²=0.88）与优质精子组分（R²=0.82）的实际生育力与预测生育力之间均存在显著关联。本研究在高、低生育力种公牛间筛选得到15个差异表达微RNA（DE-miRNA），其中9个为已知微RNA（miR-2285n、miR-378、miR-423-3p、miR-191、miR-2904、miR-378c、miR-431、miR-486、miR-2478），剩余6个为新发现的微RNA。多维偏好分析模型可依据生育力水平对种公牛进行部分区分，并将精液质量变量聚类为运动特性、DNA完整性与精子活力3个组别。研究揭示了田间生育力、精液质量参数与特定微RNA之间存在正向关联。本研究所采用的整合分析策略可为人工授精中心的种公牛选育提供有效模型，进而提升畜禽的繁殖效率。