Transcriptome-constrained genome-scale metabolic modeling reveals central carbon and amino acid metabolic...

This study aimed to construct a genome-scale metabolic model (GSMM) for Acinetobacter baumannii and integrate transcriptomic data from colistin, sulbactam, and their combination to delineate the mechanism of synergistic metabolic perturbations. A GSMM of A. baumannii 163560 was reconstructed with CarveMe and curated using literature and Biolog assays. Transcriptomic data under different antibiotic treatments were integrated with the model using RIPTiDe, and flux variability analysis was conducted to identify metabolic perturbations across monotherapies and combination therapies. The curated model (2,103 reactions, 1,492 metabolites, and 1,033 genes) achieved prediction accuracies of 80% for gene essentiality and 80.5% for carbon source utilization. Transcriptomic constraints substantially reduced the active network across all treatments while retaining core pathways such as the TCA cycle and amino acid metabolism. Colistin uniquely activated the glyoxylate cycle, consistent with a stress-adaptive bypass of central carbon metabolism. In contrast, the colistin-sulbactam combination induced distinct shifts toward amino acid, nucleotide, and ion metabolism, with aspartate-related pathways showing characteristic flux increases. Flux variation analysis further demonstrated enhanced TCA cycle activity under colistin and combination therapy, likely reflecting an oxidative stress response. Combination treatment also revealed unique essential reactions involving nucleotide and amino acid biosynthesis. Colistin and sulbactam elicit distinct metabolic rewiring in A. baumannii. Combination therapy drives a shift from central carbon metabolism toward amino acid and ion metabolism. Enzymes such as isocitrate lyase, malate synthase, and aspartate deaminase-absent in mammals-represent potential targets to enhance synergistic antibacterial effects.