Indoxyl sulfate (IS)-CTSK axis mediates cardiac fibroblast activation in CKD-Related HFpEF

Background

Chronic kidney disease (CKD) is frequently complicated by heart failure with preserved ejection fraction (HFpEF), a phenotype characterized by diastolic dysfunction, myocardial fibrosis, and limited therapeutic options [1]. Uremic toxins, particularly indoxyl sulfate (IS), are increasingly recognized as critical mediators of cardio-renal crosstalk [2]; however, the key downstream effectors and actionable targets remain insufficiently defined.

Objective

To systematically identify IS-driven molecular targets contributing to cardiac remodeling in CKD-related HFpEF and to validate their functional relevance in vivo and in vitro.

Methods

Transcriptomic profiling of human failing and non-failing hearts (GSE141910) was analyzed using differential expression and weighted gene co-expression network analysis (WGCNA) to identify HF-associated modules. Predicted IS-binding proteins were retrieved from ChEMBL, SwissTargetPrediction, and PharmMapper, and intersected with HF-relevant gene sets. A consensus prioritization framework integrating ten machine learning algorithms (including Random Forest, LASSO, SVM, and others) together with protein–protein interaction (PPI) network topological analysis was applied to identify hub targets. Single-cell RNA sequencing (scRNA-seq; GSE298023) was used to determine cell-type–specific expression patterns. Findings were validated in a murine HFpEF model induced by high-fat diet and L-NAME and in primary neonatal mouse cardiac fibroblasts treated with IS, with or without CTSK knockdown.

Results

1. Differential expression and WGCNA identified gene modules strongly correlated with HF progression. Intersection with predicted IS-binding proteins yielded 45 candidate effectors both respond to HFpEF pathology and possess IS-binding potential, suggesting the underlying linking uremic toxicity to HFpEF pathogenesis (Figure A-E).

2. Ensemble machine learning and PPI centrality analysis consistently ranked cathepsin K (CTSK) as the top consensus hub target (Figure F-J).

3. ScRNA-seq analysis revealed that CTSK was predominantly upregulated in cardiac fibroblast clusters. In HFpEF mice, CTSK expression was significantly upregulated in myocardial tissues with co-localizing with activated fibroblasts and accompanied by impaired diastolic function, increased E/A ratio, and preserved ejection fraction. In vitro, IS stimulation enhanced fibroblast activation and myofibroblast differentiation, as evidenced by increased CTSK and POSTN expression and enhanced migratory capacity. SiRNA-mediated CTSK silencing significantly attenuated IS-induced fibroblast activation (Figure K-P).

Conclusion

Our study identifies CTSK as a pivotal mediator of IS-induced cardiac fibroblast activation and myocardial remodeling in CKD-related HFpEF. Targeting the IS–CTSK axis may represent a novel therapeutic strategy to mitigate malignant cardiac remodeling and improve outcomes in patients with CKD-associated HFpEF.Figure