Technical Development of Molecular Profiling Technologies

临床分子分析中心（CMPC）的技术开发工作旨在扩大可分析的临床样本数量，并改进下一代测序技术。尽管临床研究人员有最好的意图，但适当的生物标本的积累仍然是实施CMPC精准医学使命的最具挑战性的方面。出于这个原因，我们已经把精力集中在分析福尔马林固定石蜡包埋（FFPE）标本的问题上。使用FFPE的能力非常有吸引力，因为这种标本类型适合常规病理实验室实践。我们对癌症中DNA甲基化的研究提供了这个领域最近项目的一个例子。使用一种新的基于微阵列的平台，我们已经确定可以像在冷冻标本中一样准确地分析FFPE标本中的DNA甲基化位点。这将为这类研究打开大量的组织标本档案。为了补充这种肿瘤标本类型，我们经常使用唾液标本来提供种系DNA，以更好地识别癌症标本中的体细胞突变。正如预期的那样，基于DNA的分析相对稳定，然而，RNA是一个更不稳定的模板。我们已经探索了从档案材料（如FFPE）中获得mRNA签名的可能性。这是很困难的，因为尽管平台技术在本质上没有限制，但在这些受损样品中发现的片段RNA在稳定之前（热缺血时间，处理时间，处理化学等）和稳定后的不同程度的时间和存储条件下受到许多变量的影响。除此之外，我们还在Stevenson等人（2015）关于稳定提取的RNA以供下游微阵列使用的文献中发表了“通过表达微阵列评估RNAstable中总RNA的长期稳定性”；如果在实验室中实现，它可以节省-80℃的存储成本。为了确保可重复性和提供高质量的结果，我们制定了从临床标本中提取核酸的完善的标准操作程序。我们已经成功地从FFPE样品中提取DNA用于甲基化分析；值得注意的是，这些样品非常适合用于比较基因组杂交和DNA测序。我们已经开发了一种从细胞学幻灯片中提取DNA的方案，并且已经能够从这种材料中产生非常高质量的DNA拷贝数和突变谱。几十年来，苯酚和氯仿等有机溶剂一直用于从血液和组织中纯化核酸。然而，这些化学品的使用和产生的废物造成了健康危害问题和处置问题。因此，我们已经研究并验证了在不使用有机溶剂的情况下从研究和临床标本中提取DNA、RNA和microRNA的新方案。这些努力说明了我们致力于将基因组分析技术的效用扩展到实际可获得的临床样本。一个重要的新技术发展领域是下一代测序技术的应用和方案的制定，特别是用于临床标本的文库建设。与微阵列相比，这些新技术提供了以更深入、更稳健的方式生成肿瘤标本基因组图谱数据的可能性。例如，有可能描绘出大量的药物靶点的突变，这些突变可能会促进肿瘤的生长，这些数据可以纳入未来的临床试验设计。目前正在努力使商业协议适应人类标本检测的现实；在其中一篇（Chaisaingmongkol et al., 2017）中，我们发现改变阻滞剂对覆盖深度有深远的影响。

The Clinical Molecular Profiling Core's (CMPC) technology development efforts are directed at expanding the number of clinical samples which can be analyzed and improving next-generation sequencing technologies. Despite the best intentions of clinical researchers, accrual of appropriate biospecimen remains the most challenging aspect of implementing the CMPC's precision medicine mission. For this reason, we have directed efforts to the problem of analyzing formalin fixed paraffin embedded (FFPE) specimens. The ability to use FFPE is extremely attractive since this specimen type fits into routine pathology laboratory practices. An example of a recent project in this area is provided by our study of DNA methylation in cancer. Using a novel microarray based platform, we have established that it is possible to profile sites of DNA methylation in FFPE specimens as accurately as in frozen specimens. This will open up large archives of tissue specimens to this type of research. To complement this tumor specimen type, we regularly use saliva specimens to provide germline DNA to better identify somatic mutations in the cancer specimens. As expected, DNA based assays are relatively robust, however, RNA is a much more labile template. We have explored the possibility of obtaining mRNA signatures from archival material such as FFPE. This is difficult because although the platform technology is not intrinsically limiting, the fragmented RNA found in such compromised samples are subject to many variables in sample processing prior to stabilization (warm ischemia time, processing time, processing chemistry etc.) and after stabilization to varying degrees of time and conditions of storage. In addition to this, we have published in the literature on stabilizing extracted RNA for downstream microarray use in Stevenson et al. (2015), "Long-term stability of total RNA in RNAstable as evaluated by expression microarray"; if implemented in the lab, it could save on -80C storage costs. To help ensure reproducibility and provide for quality results, we have developed refined standard operating procedures for extracting nucleic acid from clinical specimens. We have successfully extracted DNA from FFPE samples for use in methylation assays; significantly, these samples have been very suitable for comparative genomic hybridization and DNA sequencing. We have developed a protocol for extracting DNA from cytology slides and have been able to generate remarkably high quality DNA copy number and mutation profiles from this material. Organic solvents such as phenol and chloroform have been used for decades to purify nucleic acids from blood and tissues. However, the use and waste produced with these chemicals creates health hazard issues and problems of disposal. Therefore, we have investigated and validated new protocols for extraction of DNA, RNA, and microRNA from both research and clinical specimens without the use of organic solvents. These efforts are illustrative of our commitment to extend the utility of genome profiling technologies to realistically obtainable clinical samples. An important new area of technology development is in the application of next-generation sequencing technologies and development of protocols especially in library construction for use with clinical specimens. These new technologies offer the possibility of generating genomic profiling data on tumor specimens in a much deeper and more robust way than has been possible with microarrays. For example, it may become possible to profile large numbers of drug targets for mutations which may promote tumor growth, data which could be incorporated into future clinical trials design. Efforts are ongoing to adapt commercial protocols to the reality of testing human specimens; in one (Chaisaingmongkol et al., 2017) we discovered that changing the blockers had a profound effect on the depth of coverage.