MOLECULAR CHARACTERIZATION BY MASS-SPECTROMETRY OF THE HUMAN ESTROGEN-RECEPTOR LIGAND-BINDING DOMAIN EXPRESSED IN ESCHERICHIA-COLI

The ligand binding domain of the human estrogen receptor (hER-LBD), encompassing the sequence MDPS(282)AG... V-595 has been expressed at high levels in Escherichia coli from a pET-23d vector, and a purified preparation has been characterized both by mass spectrometry and biochemical methods. Inclusion bodies from the bacterial expression were solubilized by sonication and the hER-LBD was purified to near homogeneity by affinity chromatography over an estradiol-Sepharose column in urea-containing buffer. This material ran as a single peak on reversed-phase HPLC, and sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis showed a band with apparent molecular mass of 31-32 kilodaltons (kDa), somewhat smaller than that expected from the construct (35.6 kDa). Edman degradation revealed a single sequence of MDPSAGDMRA, consistent with an intact N terminus. Further characterization of this material using low resolution matrix-assisted laser desorption ionization mass spectrometry indicated an apparent single protein species of average mass 33,143 daltons (Da). Detailed molecular analysis by electrospray ionization mass spectrometry, however, revealed that this purified hER-LBD preparation was actually composed of a number of both modified and unmodified LED protein fragments between 32,900-33,400 Da. The bulk of this 33-kDa protein mixture consisted of three LED protein fragments with C termini at Ala(571) (70%), Ala(569) (23%), and Ser(575) (4%), with only a trace amount (3%) of the full length expressed LBD (MDPS(282)... V-595; 35, 602 Da). These four protein species also appear to be partially chemically modified by carbamylation. The functional integrity of this hER-LBD preparation was investigated by a ligand-exchange assay, which showed that the hER-LBD retained full estradiol-binding capacity; specific, covalent labeling was also observed using either the electrophilic affinity-labeling ligand tamoxifen aziridine (TAZ) or the photoaffinity-labeling ligand hexestrol diazirine. Thus, this expressed hER-LBD preparation, while appearing nominally pure by conventional biochemical techniques and having the expected ligand-binding, capacity, was shown by sensitive high resolution electrospray ionization mass spectrometry techniques to be largely truncated 20-26 amino acids from the expected C terminus and to have a substantial level of covalent modification arising from the urea. Our findings highlight the power of mass spectrometric analysis to identify and to characterize, at the molecular level, the heterogeneity that may be present in expressed receptor preparations that appear nominally pure by conventional methods of analysis. Such information on the sites and nature of cleavage and modification is important in devising appropriate expression systems and purification protocols and can serve as a specific guide for obtaining preparations of the hER-LBD that are homogenous and potentially better suited for nuclear magnetic resonance and x-ray crystallographic analysis.