Oxidant stress induced structural damage in estrogen receptor (oxER) identified by mass spectrometry and detected by a novel anti-oxER antibody

4465

Loss of ER DNA-binding function, as observed in a significant proportion of ER-positive human breast cancer samples, can be modeled by oxidant treatment of ER-positive cell lines (e.g. MCF7), recombinant (r) ER, and synthetic ER peptides such as those constituting the two Cys-4 type zinc fingers in the ER DNA-binding domain (DBD). Using Cys oxidizing (e.g. diamide) and arylating (e.g. menadione/K3) reagents and a novel mass spectrometry (MS)/proteomics approach, we have shown that the C-terminal half of the structurally more labile ER second zinc finger (ZnF2B) appears to be the most reactive region associated with loss of ER DNA-binding function, with Cys-240 being more sensitive than Cys-237. To generate a diagnostic antibody (anti-oxER) that would specifically detect thiol reversible disulfides within ER-DBD (e.g. diamide induced) from normal and irreversibly alkylated or arylated (e.g. K3 induced) ER-DBD, we synthesized ZnF2B in three forms: normal linear, linear with Acm-protected Cys residues (to simulate Cys alkylation), and the cyclized/disulfide form. Rabbit antisera to the cyclized/disulfide form was subtracted with the Acm-derivatized peptide, and ELISA titers showed the resulting antisera (anti-oxER) to be specific for cyclized/disulfide ZnF2B. Sensitivity of this immunodiagnostic for oxER was assessed by immunoblotting rER samples treated with increasing diamide exposure (20–200 uM, 10–60 min), confirmed by MS to produce from 15% to >60% ZnF2B disulfide formation, and normalized by immunoblotting for total ER using the C-terminal specific monoclonal, F10. Anti-oxER showed at least an order of magnitude greater affinity for oxER as compared to untreated rER; and this immunoreactivity could be eliminated by oxER disulfide reduction (10mM DTT). ER from MCF7 cells exposed in culture with sufficient diamide or K3 to eliminate ER DNA-binding and then similarly immunobloted with anti-oxER and F10 showed increased anti-oxER reactivity only in the diamide treated sample. ER extracted from 5 cryobanked primary human breast cancer samples and similarly assayed revealed variable oxER content relative to total ER, with oxER/ER ratios ranging from 0.1 to 1.0. These findings indicate that oxidant stress induced ER structural changes defined by MS can be translated into immunodiagnostic assays for use and validation on small clinical samples.