The optimal final concentration of biotin-GA was determined to be 10 nM for the Hsp90a and Hsp90^ assays and 30 nM for Grp94, because the signal was less robust for Grp94. At the concentrations chosen, we noted excellent signal-to-background ratios (50 to 100). To test the stability of the AlphaScreen signal, the plates were read, incubated another 1.5 hr at room temperature and then read again. We found that the signals were significantly reduced after a second reading. Some of the signals were reduced by as much as 90%. All subsequent experiments were read only once.
To obtain IC50 values, it was important to show that a drop in protein concentration produced a proportional drop in assay signal. The optimal range for Hsp90a was 0.125 to 1 nM, 0.5 to 2 nM for Hsp90^, and 2.5 to 10 nM for Grp94 (Figure 5.3B). The final assay concentrations chosen for the Hsp90 isoforms were 0.25 to 0.5 nM for Hsp90a, 1 nM for Hsp90£, and 3 nM for Grp94. It was notable that the concentration of Hsp90a used in this assay was significantly lower than that used in the HTRF or TRF assays, which represented a dramatic improvement of the detection limits of the biochemical binding assays. Thus, the AlphaScreen assay allowed for the distinction of compounds with potent activities (IC50 < 15 nM) for Hsp90a. Under optimal conditions, the binding potencies of 17AAG and GA for Hsp90a, Hsp90^, and Grp94 were determined. Figure 5.3C demonstrates that 17AAG and GA bind to all three Hsp90 isoforms equipotently (< twofold difference). The presence of DTT along with increased assay sensitivity may explain the lower IC50 values of 17AAG and GA for Hsp90a as determined by AlphaScreen versus TRF assays.
Was this article helpful?