12-Plex UHPLC-MS/MS analysis of sarcosine in human urine using integrated principle of multiplex tags chemical isotope labeling and selective imprint enriching.

Urinary sarcosine was considered to be a potential biomarker for prostate cancer (Pca). In this work, an integrated strategy of multiplex tags chemical isotope labeling (MTCIL) combined with magnetic dispersive solid phase extraction (MDSPE), was proposed for specific extraction and high-throughput determination of sarcosine by ultra high performance liquid chromatography tandem mass spectrometry (UHPLC-MS/MS). In the past three months, we have developed 8-plex MTCIL reagents with excellent qualitative and quantitative performance. In this work, the multiplexing capacity of MTCIL reagents (MTCIL360/361/362/363/364/365/366/375/376/378/379/381) was increased 1.5-fold from 8-plex to 12-plex. MTCIL359 was prepared and used to label sarcosine standard as internal standard (IS). The structural analogue derivative (MTCIL373-sarcosine) of all targeted MTCIL-sarcosine derivatives was synthesized and used as a novel dummy template to prepare dummy magnetic molecularly imprinted polymers (DMMIPs). The integration of MTCIL and DMMIPs procedures were extremely favorable to excellent chromatographic separation and efficient mass spectrometric detection. The labeling efficiency, chromatographic retention and mass spectrometry responses of MTCIL reagents were consistent for sarcosine. In a single UHPLC-MS/MS run (2.0 min), this method can simultaneously quantify sarcosine in 12-plex urine samples and achieve unbiased concentrations comparison between different urine samples. Analytical parameters including linearity (R 0.989-0.997), detection limits (0.02 nM), precision (2.6-11.5%), accuracy (96.1-107.4%), matrix effect, labeling and extraction efficiency were carefully validated. The proposed method was successfully applied for urinary sarcosine determination of healthy male individuals and Pca patients. It was found that the sarcosine concentrations in these two groups were statistically extremely significantly different (P < 0.001). The developed method was a powerful analytical tool to substantially promote the analysis throughput and large-scale experiments about the potential biomarker research.