Astragaloside IV Shows Cardioprotective Metabolic Effects in Diabetic Heart Cells

A new multi-omics study has revealed how astragaloside IV, a saponin drawn from the root of Astragalus membranaceus, a staple of traditional Chinese medicine, appears to shield heart muscle cells from the metabolic wreckage inflicted by high glucose. The research, published in the Journal of Agriculture and Food Research, suggests that the compound’s cardioprotective power may lie in its ability to remodel the tricarboxylic acid cycle and restore mitochondrial energy metabolism in diabetic cardiomyocytes. By combining transcriptomics, targeted metabolomics, machine learning, and molecular docking, the team traced the compound’s effects down to a specific metabolic node: PDHA1, the catalytic subunit that funnels pyruvate into the cell’s energy-producing furnace.

Diabetic cardiomyopathy, the condition at the heart of this work, is a distinct disease entity characterized by ventricular dysfunction that arises without coronary artery disease or hypertension. It is a leading cause of death among people with type 2 diabetes, and its pathogenesis is notoriously multifaceted, involving metabolic disruption, chronic inflammation, oxidative stress, and fibrosis. Under hyperglycemic conditions, myocardial cells burn oxygen abnormally fast, generating excessive reactive oxygen species that damage mitochondrial structure and undermine cardiac contractility. Faulty respiratory chain complexes leak electrons, producing superoxide and driving a self-reinforcing spiral of mitochondrial dysfunction and cell loss through apoptosis and necrosis.

Current clinical strategies for diabetic cardiomyopathy focus on glycemic control, improving myocardial remodeling, easing heart failure, and reducing risk factors, yet the mechanisms of many cardioprotective drugs remain poorly understood, particularly regarding mitochondrial and metabolic dysfunction. This gap has pushed researchers toward natural products with multi-target potential, and astragaloside IV has already shown promise in preclinical models of diabetic nephropathy and cardiomyopathy, with documented antioxidant, anti-inflammatory, and anti-apoptotic activities. What remained unclear was precisely which molecular and metabolic changes the compound triggers to protect cardiomyocytes from diabetic injury.

To find out, the researchers exposed human AC16 cardiomyocytes to a high-glucose environment of 33 millimolar, mimicking the diabetic milieu, and then treated them with astragaloside IV. Viability testing with the CCK-8 assay showed that while the compound itself was toxic at 30 micromolar and above, concentrations of 10 and 20 micromolar significantly rescued the survival of high-glucose-injured cells. Microscopy confirmed that high glucose induced morphological damage, while astragaloside IV restored near-normal cell appearance. Flow cytometry quantified the benefit: cells treated with the compound showed markedly reduced apoptosis compared to untreated high-glucose controls. The team standardized on 10 micromolar for all subsequent experiments.

The transcriptomic arm of the study yielded striking results. RNA sequencing of control, model, and treatment groups, each with six biological replicates, revealed that high glucose dysregulated nearly 3,000 genes, with 2,182 upregulated and 801 downregulated relative to controls. Principal component analysis showed complete separation between control and model clusters, confirming that the damage model was robustly established. After astragaloside IV intervention, the treatment group clustered between the two, and only 596 genes remained differentially expressed compared to controls, a dramatic narrowing indicating that the compound largely reversed the transcriptional chaos wrought by hyperglycemia.

To prioritize the most relevant genes, the researchers intersected their differential expression results with cuproptosis-related gene sets drawn from the OMIM and GeneCards databases and the literature, an approach the authors emphasize was purely an exploratory bioinformatics framework rather than evidence that copper-induced cell death occurred. Nineteen core candidate genes emerged at the intersection of disease, model, and treatment signatures. Seven machine learning algorithms, all achieving perfect classification with an area under the curve of 1.0, were used to rank these genes, and a random forest model narrowed the list to eight. PDHA1, SLC25A3, VEGFA, SERPINE1, and CDKN2A carried the highest feature importance. Functional enrichment placed PDHA1 squarely in alpha-lipoic acid metabolism and the tricarboxylic acid cycle, while other genes mapped to cellular senescence, the Hippo pathway, HIF-1 signaling, and the AGE-RAGE axis implicated in diabetic complications.

On the metabolic side, targeted liquid chromatography-mass spectrometry profiling in positive and negative ionization modes identified 60 differential metabolites between control and model cells, and 37 of these were reversed by astragaloside IV. Quality control samples showed coefficients of variation mostly below 20 percent, and supervised OPLS-DA models achieved excellent fit statistics, with R-squared Y and Q-squared values above 0.95 in both modes, confirming the statistical credibility of the group separations. Among the metabolites significantly elevated by the compound were glutathione, oxidized glutathione, S-lactoylglutathione, DL-malic acid, and fumarate, while L-phenylalanine, galactose, and several phosphate intermediates declined. Random forest and support vector machine models independently converged on five key metabolites: 2-prime-deoxyadenosine 5-prime-monophosphate, cytidine 3-prime-monophosphate, epinephrine, oxidized glutathione, and DL-malic acid.

The TCA cycle itself told a subtler story. Six cycle intermediates were measured, including succinic, isocitric, oxaloacetic, malic, aconitic, and fumaric acids, and their changes were heterogeneous rather than uniformly restored. DL-malic acid and fumarate rose significantly after astragaloside IV treatment, while isocitrate and oxaloacetate fell, and succinate and aconitate trended upward without reaching significance. The authors interpret this as a selective reorganization within the cycle rather than a wholesale recovery of the entire metabolic profile. Malate is a pivotal intermediate, regenerating oxaloacetate through mitochondrial malate dehydrogenase and shuttling redox equivalents between cytoplasm and mitochondria via the malate-aspartate shuttle, so its partial restoration hints at rebalanced substrate utilization and redox homeostasis. Spearman correlation analysis added an intriguing twist: most core genes, including PDHA1, showed significant negative correlations with DL-malic acid, suggesting coordinated transcriptional and metabolic remodeling during high-glucose stress and recovery, though the authors are careful to note that correlation does not establish causation.

Canonical correlation analysis strengthened the case for integrated regulation, yielding correlation coefficients of 0.985 and 0.931 for the first two gene-metabolite variate pairs. Pathway enrichment ranked the TCA cycle and pyruvate metabolism as the most significantly affected pathways by topology-based impact values, placing mitochondrial energy metabolism at the center of both the diabetic injury response and the compound’s protective action. Molecular docking then provided computational support for a direct physical interaction: astragaloside IV docked to PDHA1, the E1-alpha subunit of the pyruvate dehydrogenase complex, with a binding energy of minus 8.3 kilocalories per mole, while DL-malic acid bound at minus 6.1, both well within the threshold considered indicative of high affinity. PDHA1 catalyzes the oxidative decarboxylation of pyruvate to acetyl-CoA, the gateway reaction linking glycolysis to the TCA cycle, and its expression rose under sustained high glucose, likely a compensatory response to altered substrate utilization, before being partially restored by treatment.

The authors are candid about the study’s limits. The work rests on an in vitro cell model that cannot capture the systemic metabolic and vascular abnormalities of diabetic cardiomyopathy, and the multi-omics evidence is associative and hypothesis-generating rather than proof of causal mechanism. Docking scores reflect structural compatibility, not biochemical binding or enzymatic regulation. The team plans to validate the PDHA1 interaction using surface plasmon resonance, cellular thermal shift assays, and enzyme activity measurements, to deploy knockdown and overexpression approaches, to conduct targeted metabolic flux analyses, and to extend the work into db/db mice. Still, the convergence of evidence is compelling: astragaloside IV appears to protect cardiomyocytes by rebalancing pyruvate metabolism, reshaping TCA cycle intermediates, and sustaining mitochondrial energy homeostasis. If validated in vivo, a compound derived from a medicinal root long used to strengthen the heart may offer a metabolically informed route to defending it against diabetes.

Subject of Research: Multi-omics analysis of astragaloside IV-mediated TCA cycle and mitochondrial metabolic remodeling in diabetic cardiomyocytes

Article Title: Multi-omics analysis reveals TCA cycle and mitochondrial metabolic remodeling associated with astragaloside IV-mediated cardioprotection

Article References: Chai, J., Wang, Y., Zhang, X., Wang, X., Wang, Z., Guo, S., Cai, Y., Xie, D., Yu, X., Qiu, S., & Zhang, A. (2026). Multi-omics analysis reveals TCA cycle and mitochondrial metabolic remodeling associated with astragaloside IV-mediated cardioprotection. Journal of Agriculture and Food Research, 31, Article 103303. https://doi.org/10.1016/j.jafr.2026.103303

Image Credits: AI Generated

DOI: 10.1016/j.jafr.2026.103303

Keywords: astragaloside IV, diabetic cardiomyopathy, TCA cycle, mitochondrial metabolism, PDHA1, pyruvate metabolism, multi-omics, transcriptomics, metabolomics, molecular docking, DL-malic acid, AC16 cardiomyocytes