Multivariate data analysis for the detection of human alpha-acid glycoprotein aberrant glycosylation in pancreatic ductal adenocarcinoma.

Abstract

Relative quantification of human alpha-acid glycoprotein (hAGP) glycan isomers using [12C6]/[13C6]-aniline in combination with multivariate data analysis is proposed as an efficient method for the identification of pancreatic ductal adenocarcinoma (PDAC) glycan biomarkers in serum samples. Intact and desialylated glycans from hAGP, purified from serum samples of patients with PDAC and chronic pancreatitis (ChrP), were labeled with aniline and analyzed by μZIC-HILIC-MS. Afterwards, partial least squares discriminant analysis (PLS-DA) was applied to the relative areas obtained for all glycan isomers in the different samples: pathological (ChrP or PDAC) versus healthy samples. Seven intact glycan isomers with α2-6 linked sialic acids, five of them also fucosylated, were the most meaningful to distinguish between PDAC and ChrP patients. The desialylated glycan isomers also identified by PLS-DA as potential biomarker candidates confirmed that antenna but also core fucosylation could be involved in PDAC. The analysis of intact and desialylated glycan isomers in combination with the multivariate data analysis revealed that the triantennary glycan with two fucoses of hAGP could have in the future a relevant role in the differentiation of patients with PDAC from those with ChrP. SIGNIFICANCE: Multivariate data analysis is currently being used in many omics fields for biomarker discovery. However, to date, no glycomics studies have applied chemometric tools combined with mass spectrometry in a preclinical research. In this work, this methodology has been used to identify altered glycosylation of human alpha-acid glycoprotein in pancreatic ductal adenocarcinoma (PDAC). The obtained results reveal that the triantennary glycan with two fucoses could have a great biomarker potential as it was relevant to differentiate PDAC and chronic pancreatitis (ChrP) patients.