Johannes Kepler University, Institute of Analytical and General Chemistry, Altenbergerstraße 69, 4040, Linz, Austria.
Johannes Kepler University, Institute of Analytical and General Chemistry, Altenbergerstraße 69, 4040, Linz, Austria.
Talanta. 2022 Jan 1;236:122849. doi: 10.1016/j.talanta.2021.122849. Epub 2021 Sep 4.
Lipidomics has great potential for the discovery of biomarkers, elucidation of metabolic processes and identifying dysregulations in complex biological systems. Concerning biofluids like plasma or cerebrospinal fluid, several studies for the comparison of lipid extraction solvents have already been conducted. With respect to tissues, which can differ significantly in terms of dry matter content and composition, only few studies are available. The proper selection of an extraction method that covers the complexity and individuality of different tissues is challenging. The goal of this work was to provide a systematic overview on the potential of different extraction methods for a broad applicability. This study covers six different extraction procedures and four different reconstitution solvents applied to ten different porcine tissues. To get an overview of the individual lipid profiles, a workflow was developed for a fast and reliable tentative lipid annotation. Therefore, several machine learning tools were utilized, like the prediction of collision cross sections to support the tentative lipid identification in case of untargeted lipidomics. In terms of data evaluation, unsupervised (e.g. principal component analysis) and supervised (e.g. partial least square - discriminant analysis) methods were applied to visualize and subsequently interpret all generated information. Furthermore, the influence of the tissue composition on the extraction performance was investigated. It could be shown that the ten porcine tissues can be distinguished based on their lipid profile with the applied workflow and that the methyl-tert-butyl ether (MTBE) based extraction method (two-phase) showed the best overall performance for the 16 examined lipid species. With this method the highest number of features (428 in lung tissue) could be annotated. Upcoming one-phase extractions also showed a high potential concerning total number of extracted lipids. Methanol/MTBE/chloroform (MMC) extracted slightly less lipids (393 in lung and liver) than MTBE but turned out to be the best one-phase extraction method. Additionally, the numbers of extracted lipids obtained by isopropanol/water 90:10 (IPA90) (399 in stomach) and by isopropanol/methanol/chloroform (IMC) (395 in stomach) were similar to those of the modified Folch method (402 in stomach). One-phase extractions can therefore clearly be seen as preferable when a high throughput is needed.
脂质组学在生物标志物的发现、代谢过程的阐明以及复杂生物系统中失调的识别方面具有巨大的潜力。关于像血浆或脑脊液这样的生物流体，已经有一些比较脂质提取溶剂的研究。然而，关于组织，由于其干物质含量和组成可能有很大差异，目前只有少数研究可用。选择一种能够涵盖不同组织复杂性和个体性的提取方法具有挑战性。
本工作的目的是提供一个系统的概述，说明不同提取方法在广泛适用性方面的潜力。本研究涵盖了六种不同的提取程序和四种不同的重溶溶剂，应用于十种不同的猪组织。为了获得个体脂质谱的概述，开发了一种快速可靠的暂定脂质注释工作流程。因此，利用了几种机器学习工具，如预测碰撞截面，以支持靶向脂质组学情况下的暂定脂质鉴定。在数据评估方面，应用了无监督（如主成分分析）和有监督（如偏最小二乘-判别分析）方法，以可视化和随后解释所有生成的信息。此外，还研究了组织组成对提取性能的影响。结果表明，在所应用的工作流程下，十种猪组织可以根据其脂质谱进行区分，并且基于甲基叔丁基醚（MTBE）的提取方法（两相）对于 16 种被检测脂质具有最佳的整体性能。使用该方法，可以注释最多的特征（肺组织中的 428 个）。即将到来的单相提取方法在提取总脂质数量方面也显示出很高的潜力。甲醇/MTBE/氯仿（MMC）提取的脂质（肺和肝中的 393 个）略少于 MTBE，但结果是最好的单相提取方法。此外，异丙醇/水 90:10（IPA90）（胃中 399 个）和异丙醇/甲醇/氯仿（IMC）（胃中 395 个）提取的脂质数量与改良 Folch 方法（胃中 402 个）相似。因此，当需要高通量时，单相提取方法显然更可取。
