mRNA结构与功能对筛选损伤时间推断指标的影响研究

Wound age estimation in forensic pathology has been a bottleneck problem not to be solved, and indicators selection for wound age estimation is the key point to solve this problem. In our preceding study of estimating the interval of wound, We found that the expression of different targets RNA was small differences between individual (homogeneity)after injury, but some indicators has shown a greater dispersion of individual trend. It was reported that RNA structure and its stability has a great relationship, but its influence on exact estimating the interval of wound has never been concerned. Especially the AU rich element (ARE) of mRNA3 'untranslated region is closely related to the stability of mRNA, and the stability of mRNA has close relationship with function of Encodes protein? Otherwise the homogeneity in one group between individuals was influenced by the stability of mRNA. Based on the above issues, in this study the differently expressed genes we have screened out will be classify according to presence or absence of ARE structural, and then gene ontology cluster analysis of bioinformatics will be used on the two types of mRNA. After injury animal model was build, the relative expression of the every class represents gene were tested by real-time PCR, through the corresponding data conversion, multi-factor repeated measurement data ANOVA, logistic regression analysis will all be applied to found the relationship between RNA structure, function and homogeneity. We hope it point out the structure and function categories of indicators for estimating the interval of wound, which is a recommendable guideline for selection contusion age indicators in the future.

损伤时间推断一直是法医学亟待解决的瓶颈问题，而筛选用于损伤时间推断的指标是解决这一问题的关键点。本课题组在应用mRNA推断损伤时间研究中发现，不同mRNA指标在相同处理方法下个体间相对表达量的同质性差异很大。研究表明mRNA 3'端非翻译区AU 富集元件(ARE)结构与mRNA稳定性密切相关，同时其稳定性也与编码蛋白质的功能有很大关系，而mRNA稳定性决定了其相对表达量的同质性。因此，本课题拟对二代测序法筛选出的1000多个损伤后差异表达基因进行有无ARE结构分类，进而对两类基因进行功能聚类分析。设计动物损伤模型，高通量real-time PCR检测每类代表基因相对表达量，进行相应数据转化后，应用多因素重复测量设计方差分析、logistic回归分析RNA结构、功能与样本相对表达量同质性之间的关系，希望得出适合推断损伤时间指标的结构和功能类别，为以后遴选损伤时间推断指标提出一个导向性的建议。

骨骼肌损伤时间推断是法医病理学领域中的重点和难点，而作为损伤时间推断最常用的检测指标，基因的筛选却一直缺乏相应指南。有研究指出，mRNA的稳定性通过影响其稳定累积进一步影响其表达同质性。用于检测的基因表达同质性越高，则少数个体越能代表总体的趋势，则推断损伤时间越准确。而mRNA3’UTR区的ARE结构（AU-rich element）通过多种途径影响mRNA的稳定性，因此我们认为无ARE结构的基因表达同质性更高，推断损伤时间更为准确。.在前期测序结果的基础上，我们根据有、无ARE结构及其Gene Ontology分类原则将基因分为两大类、六亚组进行基因筛选，Real-time PCR技术检验基因表达量。通过计算离均差绝对值（DFA）、标准差（STD）、变异系数（CV）、CV’CV、CV评分及相应多元统计分析，对各基因表达量同质性进行评估。.共筛选到32个基因，并检测、计算上述数据。与设想一致，无ARE结构的基因其组内同质性更高，而GO分组中细胞组成亚组(Cellular Component,CC)及分子功能亚组(Molecular Function,MF)基因表现出更佳的同质性。随后对DFA、STD及CV进行聚类分析，并计算聚类组别的CV’CV及CV评分。DFA及STD均得到有显著性差异的三组聚类，且class2均表现出较高的同质性，class2基因构成均以无ARE结构、属于CC及MF基因居多，进一步印证了我们的推测。根据两大类、六亚组分组情况，对各组基因行Fisher判别分析，按正常对照组、损伤后4-24h及损伤后28-48h时分组进行损伤时间推断，得出无ARE结构、属于CC及MF基因交叉验证准确率较高。结合前期研究，对35个基因表达量行PLS-DA，发现当损伤时间分为正常对照组、损伤后4-24h及损伤后28-48h时可实现有效判别，当时间组划为正常对照组、损伤后4-12h、损伤后16-24h及损伤后28-48h时也可实现有效判别。利用步进法Fisher判别分析，三分组、四分组时分别使用13个和14个基因即可达到较高准确率，引入30个未知损伤时间样本时仍能保证较高的判别准确率。这13个和14个基因也以无ARE结构、属于CC及MF基因为主。因此本研究认为无ARE结构的基因和属于细胞组成组或分子功能组的基因表达组内同质性更高，更适用于骨骼肌损伤时间推断。