BackgroundTo date, all studies conducted on breast cancer diagnosis have focused on the expression of the full-length 66-kDa estrogen receptor alpha (ERα66). However, much less attention has been paid to a shorter 46-kDa isoform (ERα46), devoid of the N-terminal region containing the transactivation function AF-1. Here, we investigated the expression levels of ERα46 in breast tumors in relation to tumor grade and size, and examined the mechanism of its generation and its specificities of coregulatory binding and its functional activities.MethodsUsing approaches combining immunohistochemistry, Western blotting, and proteomics, antibodies allowing ERα46 detection were identified and the expression levels of ERα46 were quantified in 116 ERα-positive human breast tumors. ERα46 expression upon cellular stress was studied, and coregulator bindings, transcriptional, and proliferative response were determined to both ERα isoforms.ResultsERα46 was expressed in over 70% of breast tumors at variable levels which sometimes were more abundant than ERα66, especially in differentiated, lower-grade, and smaller-sized tumors. We also found that ERα46 can be generated via internal ribosome entry site-mediated translation in the context of endoplasmic reticulum stress. The binding affinities of both unliganded and fully-activated receptors towards co-regulator peptides revealed that the respective potencies of ERα46 and ERα66 differ significantly, contributing to the differential transcriptional activity of target genes to 17β estradiol (E2). Finally, increasing amounts of ERα46 decrease the proliferation rate of MCF7 tumor cells in response to E2.ConclusionsWe found that, besides the full-length ERα66, the overlooked ERα46 isoform is also expressed in a majority of breast tumors. This finding highlights the importance of the choice of antibodies used for the diagnosis of breast cancer, which are able or not to detect the ERα46 isoform. In addition, since the function of both ERα isoforms differs, this work underlines the need to develop new technologies in order to discriminate ERα66 and ERα46 expression in breast cancer diagnosis which could have potential clinical relevance.Electronic supplementary materialThe online version of this article (doi:10.1186/s13058-016-0780-7) contains supplementary material, which is available to authorized users.