Deciphering Metabolomic Signatures Associated with Sickle Cell Disease Using Mouse Models

• Published in: Blood 2023-11, Vol.142 (Supplement 1), p.2484-2484
• Main Authors: Montllor-Albalate, Claudia, Potdar, Alka A, Chen, Yu-Wei, DeGuzman, Francis, Yu, G. Karen
• Format: Article
• Language: English
• ISSN: 0006-4971; 1528-0020
• DOI: 10.1182/blood-2023-178709

Introduction: Sickle cell disease (SCD) is debilitating and affects millions of people worldwide. Although SCD is caused by a single-point mutation of the β-globin chain of hemoglobin, the clinical manifestations are heterogeneous, and there is a lack of well-characterized predictive biomarkers of disease progression. Metabolomics is a promising tool to interrogate disease pathophysiology and underlying molecular mechanisms. Although prior metabolomic studies have been performed on human and murine red blood cells (RBCs) in the context of SCD, no systematic investigation across commonly used mouse models of SCD has been published. In the present study, we examined the blood cell metabolome of 2 widely used SCD mouse models to study and compare the metabolomic changes in disease models. Methods: Blood samples were collected from 5 groups of mice (10-14 weeks old, n=6-7 per group): wild-type (WT) C57BL/6J mice; Townes AA, AS, and SS mice; and Berkeley SCD mice. Direct injection high-resolution mass spectrometry was used to generate untargeted metabolomics data. Three matrices (RBC, plasma, and whole blood) were isolated from each mouse group. The RBC fraction, containing peripheral blood mononuclear cells, was separated from plasma via centrifugation of whole blood. The current study focuses on the RBC dataset.Raw profile data were centroided, merged, and recalibrated using methods previously described (Fuhrer et al. Anal Chem 2011). Putative annotations were generated based on compounds contained in the Human Metabolome Database, KEGG, and ChEBI databases using accurate mass per charge (tolerance 0.001 m/z) and isotopic correlation patterns. Z-score normalization was applied across all samples for unsupervised clustering using principal component analysis (PCA) and to identify metabolites with significant changes between groups. Metabolites were mapped to central RBC metabolic pathways using MetaboAnalyst (Xia et al. Nat Protoc 2011) and manual curation. Results: Unsupervised PCA that used all detectable RBC metabolites resulted in clear separation of mouse groups by disease status (Figure 1), suggesting robust metabolomic changes in mouse models of SCD. Focusing on the top divergent metabolites across the mouse groups, several enriched pathways were identified, such as amino acid metabolism, pyrimidine and purine metabolism, and aminoacyl-tRNA biosynthesis, all potentially attributable to the buildup of RBC precursors. Metabolites characteristic of immatur