Fig 1.
Workflow for cell-based BcrAbl kinase activity assay.
Non-adherent K562 cells are seeded into a 1 mL per well capacity, 96-well filter plate, treated with a BcrAbl kinase inhibitor, and incubated with the cell-penetrating BcrAbl substrate peptide. The substrate is taken up into cells and phosphorylated specifically if BcrAbl kinase is active. Following incubation with the substrate, media is removed by vacuum, lysis buffer is added and incubated with the cells. Following lysis, the soluble cell extract is centrifuged through the filter into a Neutravidin™ capture plate and the BcrAbl substrate is immobilized via its biotin tag. Phosphorylation of the peptide is detected through ELISA using a generic anti-phosphotyrosine antibody.
Fig 2.
Standardization of peptide concentration and cell number BcrAbl kinase activity assay.
Peptide concentration and cell density were standardized for the cell-based BcrAbl kinase activity assay. BcrAbl kinase activity signal reported as relative fluorescent units (RFU, A-C) (A) Cells were exposed to a range of peptide concentrations (2.5–25 μM) to determine the effect on variation and signal to background ratio (n = 8, error bars represent SEM). (B) Cell density was varied using untreated K562 cells to determine the minimum number of cells to perform the assay. (n = 3, error bars present for all data points and represent SEM) (C) Phosphorylated and unphosphorylated BcrAbl substrates were mixed in different ratios, spiked into quenched K562 lysate, and added to the Neutravidin™ coated plate. Each condition contained 30 pmol of phosphorylated substrate and the amount of unphosphorylated substrate was varied to change the ratio (n = 3, error bars present for all data points and represent SEM). (D) K562 cells were exposed to different concentrations of the BcrAbl substrate tagged with Alexa Fluor® 488 and cellular uptake was assessed using flow cytometry. The mean fluorescence intensity is plotted as a function of substrate concentration.
Fig 3.
Characterization of assay reproducibility and validation using well-known BcrAbl inhibitors.
(A) Positive (n = 93) and negative (n = 96) controls (positive control: maximal phosphorylation in the cell-based kinase activity assay; negative control: corresponding maximal inhibition by pre-incubation of cells with imatinib, 10 μM, followed by cell-based kinase activity assay) were performed in parallel using the standardized workflow and the values were plotted to visualize separation of the signal intensities. BcrAbl activity is reported as relative fluorescent units (RFU) (B) BcrAbl kinase activity dose-response curves in K562 cells using well-characterized BcrAbl inhibitors. Cells were treated with 7 concentrations of either imatinib, nilotinib, or dasatinib for 1 hour prior to addition of the BcrAbl peptide and processed for kinase activity analysis. (n = 3, error bars represent SEM).
Fig 4.
Characterization of TKI-resistance using growth inhibition curves.
Dose-response data and fitted curves, and calculated IC50 values from growth inhibition curves. TKI-sensitive K562 and the TKI-resistant K562-IR (top row), K562-NR (middle row), and K562-DR (bottom row) cell lines were exposed to the indicated TKI and assayed for mitochondrial function after 72 hours. Data were fitted to non-linear regression models using either three or four parameters, and the best fit to the data was selected for each set of curves. Curves were compared for logIC50 using the sum-of-squares F-test. IC50 values were derived from these data and used to characterize level of TKI-resistance.
Fig 5.
Dose-response data and fitted curves from the BcrAbl activity assay.
TKI-sensitive K562 (K) and the TKI-resistant K562-IR (I, top row), K562-NR (N, middle row), and K562-DR (D, bottom row) cell lines were exposed to the indicated TKI and BcrAbl activity was assessed as described in the main text. Data were fitted to a three parameter non-linear regression model. Curves were compared for logIC50 using the sum-of-squares F-test. IC50 values were derived from these data and used to characterize involvement of Abl activity in TKI-resistance.
Fig 6.
Alternative presentation of selected data from Figs 4 and 5.
Data for conditions consistent with the maintenance culture conditions for each cell line were re-plotted separately to show status for growth inhibition (A-C) and BcrAbl activity inhibition (D-F) at relevant concentrations of each drug (1 μM IM, 100 nM NL, 10 nM DS). Comparisons between vehicle and inhibitor treated cells for TKI-sensitive K562 (K/parental, left) and the TKI-resistant K562-IR (I, right) are shown in panels A and D; Parental (left) vs. K562-NR (N, right) are shown in B and E; Parental (left) vs. K562-DR (D, right) are shown in C and F. Cell growth is given as normalized absorbance from the XTT assay. BcrAbl activity is given as estimated % total phosphorylated peptide, based on pmol pAbltide detected (calibrated via standard curve) and the total binding capacity of each well (15 pmol). ANOVA with Tukey’s multiple comparisons test was used to compare means, with significant differences indicated by *, **, or ****.
Table 1.
BcrAbl activity and growth inhibition IC50 values for TKI resistant cell lines, shifted relative to parental K562 cells.
Fig 7.
Immunodetection of Abl/BcrAbl, CrkL, and relevant site-specific phosphorylation.
Lysates from parental (K), K562-IR (I), K562-NR (N) and K562-DR (D) cells cultured in the absence (K) or presence (I, N and D) of maintenance concentrations of their respective TKIs were analysed in duplicate using the Wes capillary immunodetection system from ProteinSimple. (A) Individual capillary plots and pseudo-“blot” views (in which signal peaks are represented in “band” form) from chemiluminescence for one replicate from each antibody analysis: total Abl/BcrAbl (αAbl #2862, Cell Signaling Technology), phospho-Abl/BcrAbl (αpY245 ab195839, Abcam), total CrkL (αCrkL 32H4/#3182, Cell Signaling Technology), phospho-CrkL (αpY207 EP270Y/ab52908), and GAPDH (αGAPDH D16H11, Cell Signaling Technology). Peaks are labelled for the respective proteins; note that for GAPDH, one capillary lane (labelled with *) in this replicate exhibited artifactually shifted migration, however immunodetection was consistent with other capillaries. (B) Peaks from total BcrAbl, pYBcrAbl, total CrkL and pYCrkL were normalized to peaks from GAPDH for each duplicate, and plotted to show comparisons between the parental (K) and drug-resistant (I, N, D) cells. Error bars indicate SEM for duplicate runs. (C) GAPDH-normalized signals were used to calculate and compare %phosphorylated/total signal for BcrAbl (top) and CrkL (middle). Estimated % phosphorylated biosensor calibrated from the standard curve and compared with total binding capacity of the well, as described in Fig 6 legend) was re-plotted from data shown in Figs 4 and 6, and shown here (bottom) for comparison to endogenous protein %phospho/total. Note that “blot” images are digital representations of peak intensity, not photographic or scanned images of actual blots, and thus band shapes appear unnatural.