Development of an exo-cell assay for quantification of γ-secretase activity in cells. (a) Titration of CHAPSO detergent in the exo-cell assay. CHAPSO detergent was titrated to determine the optimal amount required for stimulating γ-secretase activity. The titration was performed using 10,000 HeLa cells and 1 μM Sb4 substrate. This reaction was incubated for 2.5 hours at 37°C. Supernatent was then collected and analyzed using ruthenylated G2-10* antibody. Activity was quantitated by measuring ECL. For each assay point n = 4, and s.d. is plotted. (b) Titration of the number of HeLa adenocarcinoma cells from which the exo-cell assay can detect γ-secretase activity. The indicated number of HeLa cells were seeded in a 96-well plate and allowed to attach overnight. The next day media was removed and replaced with fresh media containing 0.25% CHAPSO detergent, 1 μM Sb4 substrate, and DMSO or 1 μM Compound E to define background. Values plotted represent the activity quantified for each cell number assay point with GSI-defined background subtracted. For each assay point n = 4, and s.d. is plotted. (c) Dose-dependent inhibition of γ-secretase activity by GSI-34. HeLa cells were seeded 10,000 cells per well of 96-well plate. The cells were treated for 24 hours with the indicated concentration of GSI-34 inhibitor. Cells were then washed once with PBS and then the exo-cell assay was performed using 1 μM Sb4 substrate and 0.25% CHAPSO detergent. (d) IC₅₀ values of distinct GSIs in extended exo-cell assay. IC₅₀ values were obtained for 2 distinct GSI compounds using the extended exo-cell assay. For each data point n = 3, and s.d. is plotted.

Examination of real-time γ-secretase activity in A20 lymphoma and in primary B-CLL patient samples. (a) Correlation between real-time inhibition of γ-secretase activity and GSI-mediated inhibition of A20 mouse lymphoma proliferation. Two 96-well plates were seeded with 50,000 A20 mouse lymphoma cells per well in 100 μl RPMI media. To each of these plates an additional 100 μl of media was added containing DMSO or GSI-34 to indicated final concentration. These plates were incubated for 48 hours at 37°C. Following this incubation, one plate was used in a real-time exo-cell assay to quantitate the real-time inhibition of γ-secretase in A20 cells. Briefly, A20 cells were pelleted and media removed. Fresh media containing 1 μM Sb4 substrate and 0.25% CHAPSO detergent were added and the exo-cell assay was performed. For each assay point n = 4, and s.d. is plotted. Additionally, to the other 96-well plate 2 μCi/ml [³H]thymidine was incubated with the cells for 5 hours. Following 5-hour incubation at 37°C, the amount of tritiated DNA was quantified on a β-counter. For each proliferation assay point n = 10, and s.d. is plotted. (b) Real-time γ-secretase activity in primary B-CLL patient samples. B-CLL cells were seeded in 96-well plate at a concentration of 50,000 cells per well. These were allowed to attach overnight. Subsequently, the media was removed and fresh media was added back that contained either DMSO or 1 μM Compound E inhibitor. This was incubated for 24 hours at 37°C. Cells were then washed once in PBS and exo-cell assay was performed as previously described. For each assay point n = 4, and s.d. is plotted.

Recombinant γ-secretase substrate allows for detection of protease activity directly in cells. (a) Sb4 γ-secretase substrate. Schematic of the truncated Sb4 substrate from the amyloid precursor protein that has an engineered MBP tag as well as Avitag for purification and biotinylation, respectively. A thrombin cleavage site between the MBP tag and avitag allows for the removal of MBP by thrombin treatment following substrate purification. (b) LC-MS analysis identified Sb4 at the expected size and determined that greater than 90% of purified Sb4 is shown to be biotinylated. (c) Development of an exo-cell assay. Utilization of the Sb4 substrate in conjunction with a small amount of CHAPSO detergent allows for real-time examination of γ-secretase activity directly from cells using ECL or homogenous time-resolved fluorescence (HTRF) detection methods in 96-well format.

Potency of γ-secretase inhibitors in various activity assays. The potency of two structurally unique GSIs was assayed in four unique γ-secretase activity assays. IC₅₀ values were determined from the dose response curves using a non-linear regression analysis in the Prism software. An in vitro assay was based on the one previously reported by Li et al. [10], except we utilized Sb4 substrate that eliminated the need for biotinylated antibody. The cell-based activity assay used N2A mouse neuroblastoma cells stably over-expressing APP and a biotinylated 4G8 antibody that binds the C-terminus of the β-amyloid peptide. The exo-cell assay incubated HeLa cells simultaneously with GSI, Sb4 substrate as well as 0.25% CHAPSO detergent prior to detecting substrate cleavage. Finally, the extended exo-cell assay first incubated HeLa cells with GSI for 24 hours. Subsequently, the cells were washed 1× in PBS and then incubated with Sb4 substrate and CHAPSO detergent. The assay was then carried out exactly as described for the original exo-cell method. All assays incorporate ruthenylated G2-10* antibody to detect 40-site cleavage of APP or recombinant Sb4 substrate and quantitated activity by measuring ECL.