PurposeDetailed and exact mechanisms underlying brain arteriovenous malformations (bAVM) are still clinically confusing. Understanding the quantitative changes in proteins and signaling pathways would provide useful information for clinicians to understand the formation and development of bAVM, guiding individualized treatment strategies. This study was performed to establish a large human bAVM proteome database using tandem mass tag labeling and to detect changes of protein expression and pathways in human bAVM. Experimental designThis study used quantitative 6-plex tandem mass tag labeling to profile protein changes in bAVM lesions. Integrated bioinformatics analysis was used to classify and identify the altered proteins and relating signaling pathways. Western blot analyzes were used to validate the proteomic data. ResultsOur work established the first human bAVM proteome databases to date. A total of 1264 proteins were identified, and the expression of 316 proteins was significantly differentially expressed, with 249 upregulated proteins. Bioinformatics analysis demonstrated that the altered proteins had close functional correlations, including integrin cell surface interactions, extracellular matrix organization, and smooth muscle contraction. Three signaling pathways (focal adhesions, tight junctions, and gap junctions), which represent an important arena of cell-cell interactions, were found to be activated in bAVM. The proteomics data are available via ProteomeXchange with identifier PXD003289. Conclusion and clinical relevanceCell-cell interactions, including focal adhesions, tight junctions, and gap junctions, were significantly influenced in human bAVM. Understanding the molecular mechanisms that underlie bAVM would provide useful information for the development of future therapeutic approaches, guiding possible precise and individual treatment strategies.