Combined targeted and untargeted high-resolution mass spectrometry analyses to investigate metabolic alterations in pompe disease

IntroductionPompe disease is a rare, lysosomal disorder, characterized by intra-lysosomal glycogen accumulation due to an impaired function of alpha-glucosidase enzyme. The laboratory testing for Pompe is usually performed by enzyme activity, genetic test, or urine glucose tetrasaccharide (Glc4) screening by HPLC. Despite being a good preliminary marker, the Glc4 is not specific for Pompe.ObjectiveThe purpose of the present study was to develop a simple methodology using liquid chromatography-high resolution mass spectrometry (LC-HRMS) for targeted quantitative analysis of Glc(4) combined with untargeted metabolic profiling in a single analytical run to search for complementary biomarkers in Pompe disease.MethodsWe collected 21 urine specimens from 13 Pompe disease patients and compared their metabolic signatures with 21 control specimens.ResultsMultivariate statistical analyses on the untargeted profiling data revealed Glc(4), creatine, sorbitol/mannitol, L-phenylalanine, N-acetyl-4-aminobutanal, N-acetyl-L-aspartic acid, and 2-aminobenzoic acid as significantly altered in Pompe disease. This panel of metabolites increased sample class prediction (Pompe disease versus control) compared with a single biomarker.ConclusionThis study has demonstrated the potential of combined acquisition methods in LC-HRMS for Pompe disease investigation, allowing for routine determination of an established biomarker and discovery of complementary candidate biomarkers that may increase diagnostic accuracy, or improve the risk stratification of patients with disparate clinical phenotypes.

Automated summary

This study developed a simple methodology using LC-HRMS for targeted quantitative analysis of Glc(4) combined with untargeted metabolic profiling, aiming to search for complementary biomarkers in Pompe disease. The results revealed several metabolites that were significantly altered in Pompe disease, including Glc(4), creatine, sorbitol/mannitol, L-phenylalanine, N-acetyl-4-aminobutanal, N-acetyl-L-aspartic acid, and 2-aminobenzoic acid. A panel of these metabolites improved sample class prediction compared to a single biomarker.