Abstract: Objective The aim of this study was to systematically analyze the molecular characteristics of breast cancer metastasis from the primary focus to lymph node based on single-cell transcriptomic technology,and screen potential biomarkers that could be used to assess the risk of lymph node metastasis of breast cancer and predict survival prognosis. Methods A single-cell RNA sequencing dataset of breast cancer(GSE161529)was integrated to construct a single-cell transcriptomic map by normal breast tissues,primary lesion and paired lymph node metastasis samples.Through CNV analysis,and combining cell trajectory analysis,functional enrichment,and differential expression analyses,epithelial cell subsets and key genes closely related to tumor progression were screened,and survival analysis was conducted based on the TCGA breast cancer cohort to verify the clinical prognostic value of key genes. Results A total of 86,825 high-quality cells were obtained,and 7 main cell types were identified.The epithelial cells in both the primary lesion and metastatic lymph nodes had malignant characteristics,and there were two main differentiation pathways:NT→PT and NT→PT→MLN.Epithelial cells were further divided into 8 subclusters(C1-C8).Among them,the C4 subgroup was mainly associated with tumor proliferation,whereas the C8 subgroup was closely related to metastasis.The high expression of core genes PCLAF and ERO1A was significantly correlated with poor prognosis in patients(P＜0.001). Conclusion This study reveals the cellular differentiation lineage of breast cancer from primary to metastasis,and identifies PCLAF and ERO1A as key genes driving the progress of breast cancer,which can be used as potential biomarkers to predict lymph node metastasis of breast cancer and evaluate the prognosis of patients.

Key words: single cell transcriptome sequencing, breast cancer, proliferating cell nuclear antigen clamp-associated factor, endoplasmic reticulum oxidoreductase 1 alpha, prognosis