Multi-Omics and Network-Based Drug Repurposing for Septic Cardiomyopathy

Abstract

Background/objectives: Septic cardiomyopathy (SCM) is a severe cardiac complication of sepsis, characterized by cardiac dysfunction with limited effective treatments. This study aimed to identify repurposable drugs for SCM by integrated multi-omics and network analyses.

Methods: We generated a mouse model of SCM induced by lipopolysaccharide (LPS) and then obtained comprehensive metabolic and genetic data from SCM mouse hearts using ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS) and RNA sequencing (RNA-seq). Using network proximity analysis, we screened for FDA-approved drugs that interact with SCM-associated pathways. Additionally, we tested the cardioprotective effects of two drug candidates in the SCM mouse model and explored their mechanism-of-action in H9c2 cells.

Results: Network analysis identified 129 drugs associated with SCM, which were refined to 14 drug candidates based on strong network predictions, proven anti-infective effects, suitability for ICU use, and minimal side effects. Among them, acetaminophen and pyridoxal phosphate significantly improved cardiac function in SCM moues, as demonstrated by the increased ejection fraction (EF) and fractional shortening (FS), and the reduced levels of cardiac injury biomarkers: B-type natriuretic peptide (BNP) and cardiac troponin I (cTn-I). In vitro assays revealed that acetaminophen inhibited prostaglandin synthesis, reducing inflammation, while pyridoxal phosphate restored amino acid balance, supporting cellular function. These findings suggest that both drugs possess protective effects against SCM.

Conclusions: This study provides a robust platform for drug repurposing in SCM, identifying acetaminophen and pyridoxal phosphate as promising candidates for clinical translation, with the potential to improve treatment outcomes in septic patients with cardiac complications.

Keywords: LC-MS; acetaminophen; metabolomics; network medicine; transcriptomics.

Figure 1

Overall workflow of this study.

Figure 2

Lipopolysaccharide (LPS)-induced septic cardiomyopathy (SCM) in mice. (A) Schematic of SCM mouse model induced by LPS (n = 9). (B) Body weight and heart weight/tibia length ratio (heart weight/TL) in each group. (C) Representative images and data of conventional echocardiography. (D) Serum levels of brain natriuretic peptide (BNP) and cardiac troponin I (cTn-I). (E) Histological examination of mouse hearts with hematoxylin-eosin (H&E) staining (n = 3), blue arrows: inflammatory cell infiltration. Student’s t-test, **** p < 0.0001, ns: no significant difference.

Figure 3

Metabolomics analysis of septic cardiomyopathy (SCM) mouse heart. (A) The intensity variation in five internal standards among all samples. (B) Principal component analysis (PCA) scatter plot shows metabolic differences in mouse hearts between two groups. (C) Volcano plot of metabolites detected in mouse hearts. (D) Normalized intensity of 62 differential metabolites. (E) Pathways enriched based on the differential metabolites.

Figure 4

Transcriptomics analysis of septic cardiomyopathy (SCM) mouse heart. (A) Multidimensional scaling (MDS) scatter plot shows genetic differences in mouse hearts between two groups. (B) Standardized expression of differentially expressed genes (DEGs) in the mouse hearts. (C,D) Gene ontology (GO) enrichment analysis of 453 upregulated (C) and 330 downregulated (D) DEGs. The top 10 (highest odds ratio) enriched GO terms of the biological process category are shown.

Figure 5

Identification of septic cardiomyopathy (SCM)-associated module. (A) A subnetwork of SCM-associated module in the human protein–protein interactome formed by integrating data from metabolomics and transcriptomics analyses. (B) Gene expression of Stat1, Stat2, Stat3, and Eif2ak2 in transcriptomics analysis. False Discovery Rate (FDR) was calculated using moderated t-statistic with the Benjamani–Hochberg procedure. (C) Expression of representative metabolites arachidonic acid, citric acid, glutamate acid, and aspartic acid in metabolomics analysis. p-value was calculated by using the Mann–Whitney U test.

Figure 6

The network connecting 14 FDA-approved drugs and septic cardiomyopathy (SCM).

Figure 7

Acetaminophen (APAP) and pyridoxal phosphate (PLP) protect from cardiac injury in septic cardiomyopathy (SCM) mice. (A) Schematic of SCM mouse model and intervention. (B) Heart weight/tibia length ratio (heart weight/TL) in each group (CON, n = 6; SCM, n = 6; SCM+APAP_L, n = 9; SCM+APAP_H, n = 9; SCM+PLP_L, n = 9; SCM+PLP_H, n = 9). (C) Representative images of conventional echocardiography. (D) Echocardiographic analysis of mouse heart function. EF: ejection fraction; FS: fractional shortening. (E) Serum levels of brain natriuretic peptide (BNP) and cardiac troponin I (cTn-I). (F) Histological examination of mouse hearts with hematoxylin-eosin (H&E) staining (n = 3), blue arrows: inflammatory cell infiltration. (G,H) The immunohistochemistry staining of CD45 (G) and CD68 (H) in mouse hearts, red arrows: CD45-positive cells, green arrows: CD68-positive cells. One-way ANOVA, * p < 0.05, *** p < 0.001, **** p < 0.0001, ns: no significant difference.

Figure 8

Acetaminophen (APAP) and pyridoxal phosphate (PLP) protect the heart against sepsis by regulating inflammation-related pathways and amino acid metabolism pathways, respectively. (A) The workflow of cell experiments. (B,C) Cell viability (B) and lactate dehydrogenase (LDH) activity (C) of H9c2 cells following vehicle, conditioned medium (CM), and APAP treatment (n = 6). (D) Heatmap of normalized expression of genes involved in apoptosis (Bax), cardiac injury (Nppa), and inflammation (Tnfα, Il-1b, Il-6, and Nfkb2) (n = 6). (E) The highlighted subnetwork shows the inferred mechanism-of-action for APAP’s protective effect in septic cardiomyopathy (SCM). (F) Gene expression levels of key targets of APAP in the treatment of SCM (n = 6), CM vs. control: **** p < 0.0001, APAP vs. CM: ^#### p < 0.0001. (G,H) Cell viability (G) and cytotoxicity LDH activity (H) of H9c2 cells following vehicle, CM, and PLP treatment (n = 6). (I) Principal component analysis (PCA) scatter plot based on the expression of genes involved in apoptosis (Bax), cardiac injury (Nppa), and inflammation (Tnfα, Il-1b, Il-6, and Nfkb2) (n = 6). (J) Joint pathway analysis of PLP–SCM interaction network nodes. (K) Normalized concentrations of 18 amino acids in H9c2 (n = 3). One-way ANOVA, ** p < 0.01, *** p < 0.001, **** p < 0.0001.