Metabolomic and proteomic profiling of a burn-hemorrhagic shock swine model reveals a metabolomic signature associated with fatal outcomes.

BACKGROUND

Burn-hemorrhagic shock combined injury, a severe condition causing complex stress responses and metabolic disturbances that significantly affect clinical outcomes in both military and civilian settings, was modeled in swine to investigate the associated metabolomic and proteomic changes and identify potential biomarkers for disease prognosis.

METHODS

Eight clean-grade adult male Landrace pigs (4-5 months, average weight 60-70 kg) were used to model burn-hemorrhagic shock combined injury. Serum samples collected at 0 h and 2 h post-injury were analyzed using metabolomic and proteomic measurements. The metabolomic and proteomic data were processed through partial least squares-discriminant analysis (PLS-DA) and the KEGG enrichment etc. Furthermore, the integrate analysis of the metabolomic and proteomic data was generalized by canonical correlation discriminant analysis, and the correlation between metabolites and mortality of the swine model was predicted using a multiple linear regression model by Pearson analysis.

RESULTS

PLS-DA revealed a global shift in each of the metabolomic and proteomic profiles following injury. The levels of 87 signature metabolites including various types of amino acids, fatty acids and acyl-carnitines of different lengths, and many metabolites in the gluconeogenesis, glycolysis, and tricarboxylic acid (TCA) cycle are generally increased (P < 0.05) after injury and can be used as biomarkers. Pathways related to amino acids metabolism and TCA cycle were significantly enriched (P < 0.01). In proteome analysis, we found dramatically altered (P < 0.05) levels of matrix and red blood cell-related proteins, such as type I collagen and hemoglobin. Most importantly, we found that the markedly elevated (P < 0.01) succinic acid, glutaric acid, and malic acid are closely associated (r = 0.863, 0.861, and 0.821, respectively) with injury severity by Pearson analysis, and can predict mortality using a multiple linear regression model.

CONCLUSIONS

The study provides compelling observations that burn-shock swine model undergoes dramatic changes in the acute phase and present a valuable panel for clinical use of prognosis.