In the hospital, blood samples are collected to monitor patients' health states, and thus various protein-based clinical methods have been developed. However, some proteins are found to change in abundances during the process of blood collection and storage. In order to account such pre-analytical effects, we performed liquid chromatography multiple reaction monitoring mass spectrometry (LC-MRM-MS) on 15 selected proteins in plasma samples prepared by varying storage time and temperature of whole blood prior to plasma isolation. Two cytosolic proteins, profilin-1 (PFN1) and thymosin beta-4 (TMSB4X), were absolutely quantified using 15 N-labeled recombinant proteins spiked externally. The other 13 proteins were quantified in a relative way compared with the two reference proteins. Triplicated LC-MRM-MS measurements showed that the median CV of MRM peak areas was 5.7%. The amounts of PFN1 and TMSB4X increased rapidly depending on the storage time between blood collection and plasma preparation. It indicates the leakage of cellular components into the plasma fraction. Relative quantification further revealed that five proteins including PFN1, S10A8, S10A9, S10A11, and TMSB4X showed significant difference (P < 0.05). We further monitored PFN1 and TMSB4X on 40 samples collected for protein diagnostics under a typical clinical study condition. Compared with the plasma samples prepared within a day, the level of both PFN1 and TMSB4X increased in the plasma samples prepared from the blood collected the day before and kept overnight at 4°C (0.51 to 3.11 μg/mL for PFN1 and 0.98 to 5.36 μg/mL for TMSB4X in average). Our result suggests an effort of assuring plasma quality for accurate protein-based diagnosis or biomarker discovery and validation.
Keywords: blood storage; mass spectrometry; multiple reaction monitoring; plasma quality; pre-analytical effect; profilin-1; thymosin beta-4.
© 2018 John Wiley & Sons, Ltd.