Abstract
Highly sensitive and selective monitoring of amino metabolites such as glutamine, arginine, tryptophan and related proteins played significant roles in early diagnosis and warning of lymphoma. But those limited abundance and lacked chromophore group in vivo were bottleneck of multivariate analysis. This work aims to develop a novel UHPLC-Triple-TOF-HRMS method for simultaneous quantitation of 20 kinds of amino metabolites and tracing different proteins based on a new mass spectrometry probe (3-bromopropyl) triphenylphosphonium (3-BMP) with ability of enhance ionization efficiency and targeted labeling amino functional groups. An excellent linearity with R(2) ≥ 0.9995 and inter- and intra-day RSD were 1.43-5.22% and 1.22-5.87%, respectively. Satisfactory recoveries were 87.09-95.82%. Limit of detection (S/N = 3) was 4.0-12.0 fmol. Further, up-regulated haptoglobin, coagulation factor VII and catalase could directly negatively regulate Ala, Lys and Phe, which caused Trp, His, Ser, Asp and Pro expression decreased significantly in lymphoma patients (p < 0.05). Ultimately, a machine learning model was established to predict lymphoma with accuracy rate of 93.68%. Above all, this study would provide multivariate analysis strategy for in-depth explore relationship aminos associated proteins and pathogenesis and helpful for early warning of lymphoma patients under free-disease state.