Integrated Quantitative Targeted Lipidomics and Proteomics Reveal Unique Fingerprints of Multiple Metabolic Conditions

• Published in: Biomolecules (Basel, Switzerland) Vol. 12; no. 10; p. 1439
• Main Authors: Ivanova, Anna A, Rees, Jon C, Parks, Bryan A, Andrews, Michael, Gardner, Michael, Grigorutsa, Eunice, Kuklenyik, Zsuzsanna, Pirkle, James L, Barr, John R
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 01.10.2022; MDPI
• Subjects: Apolipoproteins; artificial neural network classification; Blood lipoproteins; Cardiovascular diseases; Cholesterol; Diabetes mellitus; Diagnosis; Discriminant analysis; dyslipidemias; Ethanol; Glucose; Health aspects; High density lipoprotein; Lipid metabolism; lipidomics; Lipids; Lipoproteins; Metabolic diseases; Metabolic disorders; Metabolic pathways; Metabolism; network analysis; Neural networks; Neurological diseases; Physiological aspects; Plasma; Proteins; Proteolipids; Proteomics; Risk factors; Sodium; Triglycerides; United States; St Louis Missouri; United States > US
• ISSN: 2218-273X; 2218-273X
• DOI: 10.3390/biom12101439

Aberrations in lipid and lipoprotein metabolic pathways can lead to numerous diseases, including cardiovascular disease, diabetes, neurological disorders, and cancer. The integration of quantitative lipid and lipoprotein profiling of human plasma may provide a powerful approach to inform early disease diagnosis and prevention. In this study, we leveraged data-driven quantitative targeted lipidomics and proteomics to identify specific molecular changes associated with different metabolic risk categories, including hyperlipidemic, hypercholesterolemic, hypertriglyceridemic, hyperglycemic, and normolipidemic conditions. Based on the quantitative characterization of serum samples from 146 individuals, we have determined individual lipid species and proteins that were significantly up- or down-regulated relative to the normolipidemic group. Then, we established protein–lipid topological networks for each metabolic category and linked dysregulated proteins and lipids with defined metabolic pathways. To evaluate the differentiating power of integrated lipidomics and proteomics data, we have built an artificial neural network model that simultaneously and accurately categorized the samples from each metabolic risk category based on the determined lipidomics and proteomics profiles. Together, our findings provide new insights into molecular changes associated with metabolic risk conditions, suggest new condition-specific associations between apolipoproteins and lipids, and may inform new biomarker discovery in lipid metabolism-associated disorders.