Development and validation of a HILIC Based UPLC-ESI-MS/MS method for the quantification of free carnitine in human plasma

摘要

Carnitine is a naturally occurring amino acid derivative critical for mitochondrial fatty acid metabolism. Because of the extreme hydrophilic nature of the molecule, normal reverse phase column chemistries such as C8, C 18, SB-Aq, and CN are not able to retain the molecule. In the literature, ion pairing reagents such as heptafluorobutyric acid are used to retain carnitine on the column. As an alternative, we developed a rapid chromatographic method using a HILIC column. This technique resulted in a robust, sensitive, high throughput method for the quantification of carnitine with 10ul of plasma. Methods: Plasma (10ul) was spiked with 10ng of D3-carnitine. Proteins were precipitated with the addition of 500 ul of acetonitrile/water (4:1). The supernatant was evaporated and reconstituted in 75ul of methanol. 5ul of the reconstituted sample was analyzed on a tandem quadrupole mass spectrometer (Agilent 6410) coupled with UHPLC (Agilent 1290); separation was achieved using a Agilent ZORBAX rapid resolution high definition HILIC plus column (1.8 μm, 2.1 × 50 mm), using a acetonitrile/water with 0.1% formic acid and 20mM ammonium formate gradient. Standards and samples were run in multiple reaction monitoring (MRM) of the H+ ion with the transitions 162.1 to 103.1 and 165.1 to 103.1. Data: Chromatographic separation of carnitine was attempted using C8, C 18, SB-Aq, Cyano, and HILIC columns. Only the HILIC column was able to retain carnitine. We were able to retain carnitine on a C 18 column using a ion pairing reagent (heptafluorobutyric acid), but the peak shape and width were not acceptable. Using a water/acetonitrile with 0.1% formic acid and 20mM ammonium formate gradient, we were able to develop a 11.5 minute total run (retention time 4.7 minutes) with a peak width of 6 seconds. Run time was minimized by increasing the flow rate during re -equilibration. Linearity, accuracy, and precision were performed. Using D3-carnitine as an internal standard, the method was linear between 25 and 25, 000 ng/ml and sensitive with a limit of detection (LOD) and limit of quantification (LOQ) of 10 and 25ng/ml, respectively. Using a simple acetonitrile/methanol protein precipitation, the average recovery was 88.4 - 95.7%. No significant matrix effect was observed. The method is accurate with quality controls deviating by - 10.1 to - 14.6% and -8.5 to - 13.0% relative to true values within and between days, respectively. The within-run and between -run coefficients of variations were 3.5 - 4.3% and 2.0 - 6.6% respectively, suggesting that the method is precise. After 800 injections, peak shape and resolution were unchanged with a slight drift in the retention time (0.4 minutes). This method was applied to a group of patients given a carnitine supplement.