Normalization matters: tracking the best strategy for sperm miRNA quantification
Normalization matters: tracking the best strategy for sperm miRNA quantification
复制标题
DOI：
10.1093/molehr/gaw072
复制
发表时间：
2017-01-01
影响因子：
4
通讯作者：
Anton, Ester
中科院分区：
文献类型：
--
作者：
Corral-Vazquez, Celia;Blanco, Joan;Anton, Ester
研究问题：精子microRNA（miRNA）定量逆转录聚合酶链反应（qRT-PCR）使用单重检测的最可靠标准化策略是什么？简要答复：使用hsa-miR-100- 5 p和hsa-miR-30 a-5 p的平均表达作为精子miRNA qRT-PCR数据标准化器被认为是最佳策略。已知内容：平均居中方法是miRNA高通量表达分析的最可靠的标准化策略。研究设计，大小持续时间：使用四种方法将先前发表的精子miRNA表达谱的循环阈值（Ct）值标准化：（i）平均居中限制（MCR）方法（作为参考策略）;（ii）小核RNA RNU 6 B的表达;（iii）通过一致性相关限制（CCR）算法选择的四种miRNA的表达：hsa-miR-1005 p、hsa-miR-146 b-5 p、hsa-miR-92 a-3 p和hsa-miR-30 a-5 p;（iv）这些miRNA中的两种的组合实现了与MCR的最高接近度。来自736个精子miRNAs的表达谱数据来自先前发表的在生育供体（n = 10）和不育患者（n = 10）中进行的研究。38）。对于每个测试的标准化分子，评估不同样品和群体中的表达普遍性和均匀性作为被认为是有效候选物的不可或缺的要求。不同的标准化策略的可靠性进行了比较，MCR的基础上的一组差异表达的miRNA（DE-miRNA）检测人群之间，相应的预测目标和相关的丰富的生物processing.Main结果和机会的作用：所有测试的normalizers被发现是普遍存在的和非差异表达的人群之间。RNU 6 B是样本中表达最不均匀的候选者。与MCR输出相比，通过RNU 6 B进行的数据归一化导致了严重误导的结果，对靶基因和丰富的生物过程进行了零预测。Hsa-miR-146 b-5 p和hsa-miR-92 a-3 p比RNU 6 B更均匀地表达，但其结果仍然显示与参考方法的接近性很小。hsa-miR-100- 5 p和hsa-miR-30 a-5 p与MCR的相似性最高。针对两种miRNA的组合的归一化达到关于检测到的DE-miRNA的最佳接近等级（曲线下面积= 0.8）。该组合在预测的靶基因（72.3%的真阳性）及其相应的富集生物学过程（70.4%的真阳性）方面也表现出最佳性能。大规模数据：不适用。限制，排除原因：本研究集中于精子miRNA qRT-PCR分析。在其他细胞类型或组织中使用所选的标准化物仍需要确认。研究结果的更广泛意义：使用高通量检测方法寻找基于精子miRNA表达的新生育力生物标志物是生殖遗传学领域即将面临的挑战之一。在这种情况下，必须使用单重测定法验证结果。在这项研究中提出的标准化策略将在这方面提供一个通用的选项，允许标准化的验证数据，而不会导致有意义的变化的结果。相反，在精子-miRNA表达研究中，RNU 6 B的qRT-PCR数据标准化应被丢弃。
STUDY QUESTION: What is the most reliable normalization strategy for sperm microRNA (miRNA) quantitative Reverse Transcription Polymerase Chain Reactions (qRT-PCR) using singleplex assays?SUMMARY ANSWER: The use of the average expression of hsa-miR-100-5p and hsa-miR-30a-5p as sperm miRNA qRT-PCR data normalizer is suggested as an optimal strategy.WHAT IS KNOWN ALREADY: Mean-centering methods are the most reliable normalization strategies for miRNA high-throughput expression analyses. Nevertheless, specific trustworthy reference controls must be established in singleplex sperm miRNA qRT-PCRs.STUDY DESIGN, SIZE DURATION: Cycle threshold (Ct) values from previously published sperm miRNA expression profiles were normalized using four approaches: (i) Mean-Centering Restricted (MCR) method (taken as the reference strategy); (ii) expression of the small nuclear RNA RNU6B; (iii) expression of four miRNAs selected by the Concordance Correlation Restricted (CCR) algorithm: hsa-miR-1005p, hsa-miR-146b-5p, hsa-miR-92a-3p and hsa-miR-30a-5p; (iv) the combination of two of these miRNAs that achieved the highest proximity to MCR.PARTICIPANTS/MATERIALS, SETTING, METHODS: Expression profile data from 736 sperm miRNAs were taken from previously published studies performed in fertile donors (n = 10) and infertile patients (n = 38). For each tested normalizer molecule, expression ubiquity and uniformity across the different samples and populations were assessed as indispensable requirements for being considered as valid candidates. The reliability of the different normalizing strategies was compared to MCR based on the set of differentially expressed miRNAs (DE-miRNAs) detected between populations, the corresponding predicted targets and the associated enriched biological processes.MAIN RESULTS AND THE ROLE OF CHANCE: All tested normalizers were found to be ubiquitous and non-differentially expressed between populations. RNU6B was the least uniformly expressed candidate across samples. Data normalization through RNU6B led to dramatically misguided results when compared to MCR outputs, with a null prediction of target genes and enriched biological processes. Hsa-miR-146b-5p and hsa-miR-92a-3p were more uniformly expressed than RNU6B, but their results still showed scant proximity to the reference method. The highest resemblance to MCR was achieved by hsa-miR-100-5p and hsa-miR-30a-5p. Normalization against the combination of both miRNAs reached the best proximity rank regarding the detected DE-miRNAs (Area Under the Curve = 0.8). This combination also exhibited the best performance in terms of the target genes predicted (72.3% of True Positives) and their corresponding enriched biological processes (70.4% of True Positives).LARGE SCALE DATA: Not applicable.LIMITATIONS, REASONS FOR CAUTION: This study is focused on sperm miRNA qRT-PCR analysis. The use of the selected normalizers in other cell types or tissues would still require confirmation.WIDER IMPLICATIONS OF THE FINDINGS: The search for new fertility biomarkers based on sperm miRNA expression using high-throughput assays is one of the upcoming challenges in the field of reproductive genetics. In this context, validation of the results using singleplex assays would be mandatory. The normalizer strategy suggested in this study would provide a universal option in this area, allowing for normalization of the validated data without causing meaningful variations of the results. Instead, qRT-PCR data normalization by RNU6B should be discarded in sperm-miRNA expression studies.