LDHA-mediated CREB1 Lactylation Promotes Breast Cancer Brain Metastasis by Transcriptionally Activating Cytoskeletal Remodeling

Background Brain metastasis (BM) is a leading cause of mortality in breast cancer patients. This study utilizes single-cell RNA sequencing (scRNA-seq) to identify high-risk malignant sub-clusters and uncover potential therapeutic targets driving BM. Methods Five pairs of scRNA-seq data of primary breast tumors (PT) and brain metastases (BM) were retrieved from the Gene Expression Omnibus (GEO) database. Following unsupervised clustering and cell-type annotation, specialized bioinformatic pipelines—including inferCNV and pseudotime trajectory analysis—were implemented to delineate malignant subpopulations and evolutionary states associated with high brain metastatic potential. Molecular mechanisms were investigated using ChIP-qPCR, nascent RNA labeling assay, and dual-luciferase assays. Clinical significance was evaluated in a 124 paired-patient cohort using propensity score matching (PSM), while in vivo metastatic potential was assessed via a murine carotid artery injection model. Results We identified a specific malignant epithelial sub-cluster (Cluster 3) characterized by high lactate levels and superior BM potential, and subsequently establishing CREB1 lactylation as a biomarker of this cluster. Mechanistically, LDHA-mediated lactylation of CREB1 at Lysine 136 (K136) enhances its transcriptional activity, upregulating cytoskeletal genes including CALML5, CNN2, and PDLIM1. This axis facilitates pseudopodia formation, cellular migration. In vivo, LDHA knockdown significantly reduced intracranial tumor burden. Clinically, high lactylation scores and high CALML5 expression are independent predictors of brain metastasis and overall survival. Conclusions The LDHA-lactylation-CREB1-cytoskeleton axis is a novel driver of brain metastasis, serving as a promising therapeutic target and prognostic biomarker for breast cancer.