Comparing three methods for lipid identification using mass spectrometry - Data-independent, data-dependent and targeted acquisition

Abstract

Introduction: This study evaluated lipid identification using mass spectrometry.

Background: Three main modes of acquisition are used in mass spectrometry of lipids; data-independent acquisition (DIA), data-dependent acquisition (DDA) and targeted acquisition. DIA is the most including one, yielding a complete but complex data set, while targeted is the least including one, basing acquisition on a predefined list. It has not been established which of these methods is able to identify the most lipids, which is a general goal in lipidomics. Furthermore, a local method for analyzing DIA data does not exist.

Aim(s): The aim of this study was to compare how DIA, DDA and targeted acquisition perform in high-throughput lipid identification and to establish a method for DIA analysis by comparing data-driven and targeted peak picking.

Methods: DIA (MSe), DDA and targeted acquisition mode was used on a Xevo G2 QTOF (Waters, USA) quadrupole time-of-flight mass spectrometer coupled to ultra-high performance liquid chromatography (UHPLC/QTOF-MS). The software MS-DIAL (RIKEN, Japan) and MRMPROBS (RIKEN, Japan) were used for data-driven and targeted peak picking of DIA data, respectively. Standard settings were used for all methods. The DIA method that identified the most lipids was compared to DDA and targeted acquisition. Comparisons were based on the number of lipids identified from a predefined list containing 34 lipids in negative mode and 23 lipids in positive mode, adding up to a total of 35 individual lipids.

Results: Data-driven peak picking of DIA data was able to identify 15 lipids, while targeted peak picking identified 26 lipids. DDA did not yield any usable data, but targeted acquisition produced MS/MS spectra for 17 of the lipids from the list.

Conclusion: When analyzing DIA data, targeted peak picking was more effective than data-driven peak picking, and more lipids were identified with DIA than with targeted acquisition. Further studies are necessary to confirm these findings.

Popular Abstract

How to identify lipids in a sample

Lipids are molecules that do not interact appreciably with water, they are also called fats, and oil is an example of a mixture consisting of lipids. In the body there are thousands of different lipids serving many different functions. One example is as fat depots, but cell membranes, the shell around each cell, also consist of lipids. Lipids are involved in sending signals, both within a cell and throughout the entire body. By determining which lipids are present in a tissue such as blood, you can discover low or high levels that can be the result of a disease or injury. In this study, I have compared three methods to identify lipids in a sample to see which method finds the most lipids.

Lipids can be measured with a technique called mass spectrometry, and quadrupole time-of-flight mass spectrometry was used in this study. It is a procedure in which you insert the sample into a machine which converts the molecules into ions, accelerate them through an electrical field, and measure how long time it takes for the molecules to travel through a chamber. The time is related to the size of the molecule, so the machine is essentially a scale for molecules. In addition to this, if you collide the molecules with a gas, they will fragment before the chamber and you can find out how much separate parts of the molecule weighs. With information about the mass of the entire molecule and its parts, you can figure out which lipid it is.

The three methods compared in this study were one where all lipids in a sample are fragmented, a second in which only the most abundant lipids are fragmented, and a third where a list specifies which lipids to fragment. For the first method, I also tested two different data analysis methods; one where a software does all data processing and gives you a list of lipids, and one where you have to enter a predefined list of lipids and fragments into the software before analysis and the software looks for a match.

The results showed that you can identify the most lipids by measuring with the method that fragment all lipids, and then use a software that identifies lipids based on a predefined list.

• author: Dahlqvist, Sofia
• supervisor: Peter Spégel
• organization: Department of Chemistry
• course: KEML16 20191
• year: 2019
• type: M2 - Bachelor Degree
• subject: Chemistry
• keywords: Analytical chemistry, Data-dependent acquisition, Data-independent acquisition, Lipid identification, Mass spectrometry, Targeted acquisition
• language: English
• id: 8987552
• date added to LUP: 2019-07-04 14:38:17
• date last changed: 2019-07-04 14:38:17