Songlin Xiao, Chunlei Mo, Qixing Liu, Li Xu
Background The role of histidine metabolism in heart failure (HF), particularly its heterogeneous regulation across cardiac cell types, remains unclear. This study aims to define its mechanistic basis and identify key regulatory genes. Methods We integrated multiple transcriptomic datasets, including four conventional and three single-cell RNA-seq (scRNA-seq) cohorts from the GEO database. A multi-step computational biology pipeline was employed, comprising gene set enrichment analysis (e.g., AUCell, ssGSEA), multidimensional machine learning (14 algorithms, e.g., LASSO, random forest), differential expression analysis, and cell-cell communication inference (CellChat, MultiNicheNet) to identify and validate core targets. Results We identified significant activation of the histidine metabolism pathway in the heart failure (HF) microenvironment, particularly within fibroblast and myeloid cells. A high-histidine-metabolism macrophage subpopulation (GPNMB + ) exhibited enriched energy metabolism and upregulated MHC-II signaling, while a corresponding fibroblast subpopulation (Fib_THY1) showed enhanced immune communication. From 23 candidates, a rigorous machine learning strategy pinpointed FRZB and CRYM as core targets. These genes were markedly upregulated in HF, demonstrated reliable diagnostic value (AUC: 0.87–0.98), and correlated strongly with histidine metabolism. Crucially, scRNA-seq revealed distinct cellular localization: CRYM in cardiomyocytes and FRZB in mesenchymal stromal cells. Cell communication analysis suggested a putative pathological positive feedback loop wherein FRZB + stromal cells are predicted to influence metabolic reprogramming in cardiomyocytes. This study proposes a novel model of a “stroma-myocardium” feedback loop. Conclusion This study systematically characterizes the potential role of histidine metabolism dysfunction in HF and supports a model of a putative “stroma-myocardium” positive feedback loop associated with FRZB (in stromal cells) and CRYM (in cardiomyocytes). These computational inferences position FRZB and CRYM as candidate biomarkers, while their regulatory axis provides a novel theoretical framework for potential therapeutic targets in HF.