Abstract
Activity of the canonical estrogen receptor (ER) pathway is equivalent to functional activity of the nuclear ER transcription factor. Monoclonal antibodies (MoAbs) that identify nuclear ER in cells and tissue samples are frequently used to assess ER transcriptional activity, however, it remains unclear if this approach is sufficiently predictive of ER pathway activity. This study uses ER-positive breast cancer cell lines (MCF7 and T47D) in which ER transcriptional activity was quantified using an mRNA-based ER pathway activity assay. The relationship between ER activity and nuclear ER staining with ER MoAbs was then investigated. Confirming earlier findings, the results show that while the presence of ER in the cell nucleus is a prerequisite for ER activity, it is not predictive of ER transcriptional activity. There were remarkable differences in the behaviours of the antibodies used in the study. EP1 and 1D5 showed reduced nuclear staining when ER was transcriptionally active, while staining with H4624 was independent of ER activity. To improve discrimination between active and inactive nuclear ER based on ER staining, a method was developed which consists of dual ER MoAb immunofluorescent staining, followed by generation of a digital image with a standard digital pathology scanner. Then a cell nucleus detection algorithm and per cell calculation of the nuclear H4624/EP1 fluorescence intensity ratio was applied, where a high H4624/EP1 ratio predicts an active ER pathway. With this method, the EP1 and 1D5 antibodies are interchangeable. We hypothesize that the transcriptional activation of ER hides the epitope recognized by MoAbs EP1 and 1D5, while H4624 binds an ER epitope that remains accessible during ER pathway activation. The method described in this study should add substantial value to the assessment of ER pathway activity for biomedical research and diagnostics.