(M0930-09-51) An Emerging Derivatization Strategy for Minimizing Matrix Effect in Metabolomics Analysis

Purpose: The matrix effect is one of the greatest enemies in biological analysis. Matrix effect is defined as a positive or negative interference produced by a different component from the analyte, affecting the analyte response. Among those matrix effect molecules, lipids play a major role due to the strong ionization capability and the high abundance. The ion suppression or enhancement may cause a series of problems including inaccurate quantitation, unequivocal identification, and low analyte detectability. Recently, more and more novel methodologies were proposed to eliminate the matrix effect, such as changing ion sources, stable isotope labeling, and Solid-phase extraction to remove the matrix. However, none of these methods truly “eliminate” the matrix effect, especially in metabolomics analysis, where those matrix molecules are also the analytes. We believe that rather than removing the matrix molecules, modifying analyte structures to be free of matrix effect is an emerging and promising approach to eliminate matrix effect.

Therefore, this research has two objectives. The first is to introduce a novel derivatization strategy to make the analytes independent of the matrix in metabolomics analysis. The second is to systemically evaluate the matrix effect by post-extraction spiking method for the strategy we proposed.

Methods: Firstly, we established a highly sensitive analytical method for three submetabolome in functional group-based derivatization. To be more specific, we selected three derivatization reagents to target three kinds of the metabolome, with dansyl chloride (Dns-Cl), dansyl piperazine (Dns-PP), and dansyl hydrazine (Dns-HZ) utilized to react with amine-submetabolome, carboxyl-submetabolome and carbonyl-submetabolome, respectively. We established the LC-MS/MS methods for those derivatized metabolites in MRM mode. The reaction conditions are also optimized to obtain the ideal reaction efficiency, including temperature(20, 25, 30, 35, 40, 45, 50, 55, and 60℃) and reaction time (15, 20, 30, 40, 45, 50, 60, 70, 75, 80, 90, 100, 120 minutes).

Secondly, we systemically evaluated the matrix effect in our method with and without the presence of lipids. We applied the lipid removal solid-phase extraction (SPE) sample preparation method to obtain polar metabolites and non-polar lipids subsequently. After the polar metabolome was derivatized with the three derivatization reagents listed above, the resulting solution was reconstituted with 20μL methanol (fraction A) or 20μL of the non-polar lipid matrix (fraction B) obtained from the 2^nd eluent. We applied the post-extraction spiking method to calculate matrix factors and evaluate the matrix effect introduced by lipids, with the equations below to calculate the matrix factor of lipids on derivatized metabolites. (Figure 1)

Results: We established a highly sensitive LC-MS/MS method covering over 200 polar metabolites based on the derivatization method. The optimal reaction temperature and time are optimized with the conditions as the Figure2 shown. Based on this, we evaluated the matrix effect of the metabolites after derivatization. According to FDA, the recommended acceptable range for matrix factor is between 80% and 120%. From the results, we can see among 204 metabolites that we established LC-MS/MS method for by MRM, 110 polar metabolites are detectable after derivatization. Some of the metabolite information was lost partly because of the low concentration in blood or sample lost during SPE adsorption. Among those 110 polar metabolites, we excluded those metabolites which have an SD value over 80% (11 metabolites), indicating sample variance. It turned out that 93 metabolites in 99 metabolites had the Matrix factor between 50% and 150% and 40 metabolites had the matrix factor between 80% and 120% (Figure 3). The result is surprising because most dansyl labeled metabolites had a low/minor matrix effect even eluting with lipids, in contrast to the current research that lipids have a high matrix effect. We assume that due to the dansyl group within this molecular, there’s a large moiety structure and high ionization efficiency within derivatized analytes, thus the derivatized analytes had little dependence on the matrix. In other words, different types of matrices will make a minor effect o the analytical methods, in terms of sensitivity, accuracy, qualification, and quantification.

Conclusion: In this study, we proposed a novel derivatization strategy to reduce the matrix effect in the bioanalytical analysis. Instead of removing the matrix during sample preparation, we modified the analytes structures by derivatization reaction. Currently, some literature supported this theory and we validated the feasibility by using multiple derivatizations. Hopefully, it will be widely applied in the field of bioanalytical methods in metabolomics analysis. Since the matrix effect has been overcome by the derivatization to some extent, this method significantly improves analytical sensitivity and robustness. However, there are still some gaps unsolved here. For the matrix effect, we only evaluated the ion suppression and ion enhancement through the matrix factor. More and more parameters and standards for evaluating matrix effect should be specified, such as the long-term impact on the instrument, the precision and accuracy of this method, etc.

References: [1] XU F, ZOU L, LIU Y, et al. Enhancement of the capabilities of liquid chromatography-mass spectrometry with derivatization: general principles and applications [J]. Mass Spectrom Rev, 2011, 30(6): 1143-72.
[2] FUREY A, MORIARTY M, BANE V. Ion suppression A critical review on causes, evaluation, prevention and applications. [J]. Talanta, 2013, 115: 104-22.
[3] CHEN D, HAN W, SU X, et al. Overcoming Sample Matrix Effect in Quantitative Blood Metabolomics Using Chemical Isotope Labeling Liquid Chromatography Mass Spectrometry [J]. Analytical Chemistry, 2017, 89: 9424-31.

Acknowledgements:Conflict of interest - The authors declare no competing financial interest.

This work was supported by the NSFC (No. 82073812, 81773682, 81773861), Jiangsu Provincial National Science Foundation for Distinguished Young Scholars (No. BK20180027), National Science and Technology Major Project (2017ZX09101001), the Fundamental Research Funds for the Central Universities (2632021PY03), Double First-Class University project, the Program for Jiangsu province Innovative Research Team and a project funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD).


Figure1. workflow of the derivatization strategy based matrix effect study


Figure2. optimization of three derivatization reaction conditions


Figure3. Matrix factor evaluation of derivatized polar metabolites