Gene fusions resulting from chromosome translocations are a crucial class of genetic aberrations causing cancer, which have provided treatment targets for a number of cancer types. However, neoplastic gene fusions have not yet been found in the vast majority of estrogen receptor (ER) positive breast cancers. Recently, the cancer genome projects released a large quantity of next-generation sequencing and genomic data for this cancer. Analysis of these data revealed about half of the somatic translocations to be between adjacent or nearly adjacent genomic loci. This finding suggests that these somatic adjacent gene translocations (AGTs) may be far more common than realized, and we speculate that a subset of these AGTs could be pathological events in breast cancer. Indeed, one neoplastic AGT has already been discovered in our preliminary studies, which involves the estrogen receptor gene, ESR1, and a neighboring gene, CCDC170. This fusion endows tumor aggressiveness and endocrine resistance, and appears to be frequent in the most deadly subtype of ER+ breast cancer, luminal B, that is particularly hard to clinically define or manage. We thus hypothesize that ESR1-CCDC170 and other recurrent AGTs may play a crucial role in breast cancer initiation, progression, or therapeutic resistance. The following studies are proposed to test the hypothesis: Aim 1 will further study the role of ESR1-CCDC170 in breast cancer progression, invasion, and endocrine resistance using in vitro knockdown and overexpression models, and will explore the engaged pathways and key effectors by systematic proteomic studies and focused mechanistic studies. Aim 2 will establish the incidence of ESR1-CCDC170 in ER+ breast cancer by interrogating independent patient cohorts, elucidate its clinicopathological relevance, and substantiate its role in tumor formation, metastasis, and endocrine resistance in the in vivo context. Aim 3 will systematically discover new AGTs in breast cancer by integrative analysis of several kinds of genomic profiling and next generation sequencing datasets. We will dissect out true recurrent AGTs driving breast cancer, determine their incidence, and investigate their genetic origins and mechanisms of action. Successful accomplishment of this project may uncover a dark area of breast cancer genetics. Transcripts resulting from AGTs are usually submerged in the overwhelming number of non-genomic transcription-induced chimeras (TICs), whereas conventional cytogenetic assays are insensitive to such close-range translocations. Thus AGTs could have been largely overlooked by the previous gene fusion screens. The new AGTs revealed by this study will not only shed new light on the genetic causes of breast cancer, but will also provide new biomarkers for the prediction of patient outcome and individualized therapy, and attractive drug targets for new and specific breast cancer treatment.