Abstract
ABSTRACT
Advances in molecular characterization have significantly redefined our understanding of breast tumor heterogeneity, enabling improved diagnostics and the development of tailored therapeutic strategies. In particular, integrative omics approaches offer a broader scope for exploring the molecular and biochemical complexity of breast cancer. Notably, metabolomics through high‐resolution magic angle spinning (HR‐MAS) NMR spectroscopy enables nondestructive metabolic profiling of intact tissue samples, preserving the native biochemical context and offering unique potential for refining breast cancer subtyping and gaining deeper insight into disease biology. However, the reproducibility and comparability of HR‐MAS‐based metabolomics across studies are still limited by methodological variations, underscoring the need for standardized and reliable protocols to manage the complexity of breast cancer tissue spectra. In this study, LCModel analysis in combination with the Electronic REference To access In vivo Concentrations (ERETIC2) method as a quantitative reference was implemented for absolute metabolite quantification of breast cancer samples, providing a robust metabolic profiling approach. Generalized linear models were applied to evaluate the associations between quantified metabolites and tumor immunohistochemical classification, and transcriptomic correlation analysis was used to explore gene expression profiles underlying the observed metabolic alterations. A total of 27 metabolites were quantified from breast tumor samples spanning the Luminal A, Luminal B (human epidermal growth factor receptor 2–negative, HER2−), Luminal B (HER2+), and triple‐negative breast cancer (TNBC) subtypes. Strong positive associations were observed between phosphocholine (PCho), betaine (Bet), total choline (tCho), and total creatine (tCr) and immunohistochemical subtype, whereas alanine, glucose, and myo‐inositol showed negative associations. Notably, Luminal B (HER2−) tumors exhibited the highest PCho levels, linked to upregulated expression of choline kinase alpha (CHKA) and other enzymes of the Kennedy pathway. In contrast, Luminal B (HER2+) tumors exhibited elevated levels of alanine (Ala), glucose (Glc), and myo‐inositol (Ins), indicating distinct metabolic reprogramming within the luminal subgroups. Additionally, this study highlights the potential role of betaine (Bet) in interconnecting metabolic and epigenetic pathways in breast cancer, as its concentrations were found to vary across subtypes and correlate with genes involved in fatty acid metabolism, DNA methylation, and estrogen signaling. Robust HR‐MAS NMR profiling captured subtype‐dependent metabolic signatures, and integrative analysis with transcriptomics identified gene expression differences associated with these signatures, thereby advancing the understanding of breast cancer heterogeneity.