Fig 1.
Overview of the xCELLigence impedance-based technology.
(A) Multi-well plates have gold electrodes embedded on the bottom that measure the conductivity of the medium through a small current. Adherent cells cover the surface of the electrodes and impede the flow of the current. (B) Typical cell proliferation curve measured through the standardized impedance parameter (Cell Index) shows the different phases of initial cell adhesion, exponential growth, confluency and cell death. The Cell Index without cells is 0. This is based on [20].
Fig 2.
Impedance measurement of OP9 stromal cell adhesion and displacement by BCP-ALL cells.
(A) Impedance was measured over 7 days in wells containing OP9 alone (10,000 cells/well), US7 alone (50,000 cells/well) and US7 plated on OP9 feeder cells as indicated (n = 2). When added to the well, OP9 stromal cells adhere to the electrodes and increase impedance (Cell Index) over time. PDX-derived BCP-ALL cells (US7) do not attach to the electrodes and do not affect the measured impedance; however, when co-cultured with OP9, they cause a reduction in impedance. (B) Impedance in co-culture of OP9 cells with different amounts of US7 cells as indicated below the figure (n = 2, triangle = no cells, diamond = 50,000 cells, square = 5,000 cells, circle = 500,000 cells). The reduction in impedance is proportional to the number of added US7 cells. (C) Bright field images of OP9 alone (10,000 cells/well) or OP9 and US7 BCP-ALL cells (50,000 cells/well) 7 days after leukemia cell addition. Dark areas are generated by electrodes blocking the light path in the inverted microscope. Cells are visible in the clear space between electrodes. Scale bar = 100 μm.
Fig 3.
Different BCP-ALL cells exhibit unique patterns of impedance signal displacement.
Primary (non-cultured) BCP-ALL samples Ph+ BM61, Ph+ BM13 and BM47 were plated (100,000 cells/well) one day after OP9 feeder cells (10,000 cells/well) and impedance was measured over more than 7 days (n = 2). BM13 and BM61 were run on the same plate, whereas BM47 was from a different experiment.
Fig 4.
OP9 displacement can be used for monitoring drug cytotoxicity on BCP-ALL cells.
(A) US7 cells treated with 100 nM vincristine as indicated (n = 2). (B) Cell Index profile of dose-response to vincristine treatment of OP9+US7. Dotted line indicates the time point used for calculating the IC50 (n = 3 except 1.25nm and 315pm n = 2). (C) Dose-response curve and IC50 calculated from the Cell Index profile in panel B. The IC50 value for vincristine treatment of US7 was 1.87 nM. Statistical analysis was done using one-way ANOVA followed by Dunnett’s multiple comparison test. Error bars represent standard deviation. p values relative to control: ***<0.001; ****<0.0001 (D) ICN13 cells treated with 20 nM vincristine (n = 2). (E) Images of cells indicated in the figure inside the 16-well plate devices, captured after 7 days of treatment with 10 nM (US7) or 20 nM (ICN13) vincristine. Scale bar = 100 μm.
Fig 5.
Comparison of end-point tissue culture with OP9 displacement.
US7 cells were plated on 96-well tissue culture plates (TC = tissue culture), HTS 96 well Transwell plates (TW = transwell), and E-Plate 16 devices (XC = xCELLigence) as indicted in the figure (all coated with fibronectin) and exposed to either DMSO or 1 nM vincristine (vcr). On day 9, cells were harvested and examined for (A) live cell numbers using Trypan blue exclusion (B) apoptotic cells by FACS using 7-AAD and Annexin V (C) living cell percentages based on FACS and negativity for 7-AAD. Statistical analysis was done using two-way ANOVA, followed by Bonferroni posttests. Error bars represent standard deviation. p values: *<0.05, **<0.01; ***<0.001; ****<0.0001. (D) xCELLigence continuous monitoring of parallel cultures under the indicated drug treatments (n = 3). Dotted lines in inset indicate time points of early differentiation visible between the controls and treatment wells. (E) Correlation of cell index values with conventional cell viability measurements. Error bars represent standard deviation. p values of correlation coefficient: *<0.05, **<0.01; ***<0.001.
Fig 6.
Different drug treatments generate unique patterns of impedance signal displacement in BCP-ALL cell co-cultures.
Ph-positive BCP-ALL cells include BLQ1 (A) and BLQ5 (B) cells. Top graphs: different concentrations of ponatinib (pon). Dotted lines show the time point used for the % displacement calculation (also see S3 Fig). Bottom graphs: Percentage impedance signal displacement from treatment of BLQ1 and BLQ5 with the indicated concentrations of four drugs at the time points indicated in S3 Fig. Statistical significance was determined using one-way ANOVA, followed by followed by Dunnett’s multiple comparison test. Error bars represent standard deviation (n = 3). p values relative to control: *<0.05; **<0.01; ***<0.001; ****<0.0001.
Fig 7.
Treatment of BLQ5 co-cultured on human MSC with VX-680.
(A) BLQ5 cells were plated on HS5 stromal cells, resulting in a reduction of impedance (n = 3). (B) Treatment of BLQ5 cells with different concentrations of VX-680 as indicated (n = 3). Dotted lines show the d5 and d9 time points at which cells were harvested for (C) live cell numbers determined by Trypan blue exclusion. Statistical significance was determined using one-way ANOVA, followed by followed by Dunnett’s multiple comparison test. Error bars represent standard deviation. p values relative to DMSO control: ****<0.0001. or (D) Percentage apoptotic cells characterized by FACS using 7-AAD and Annexin V.
Fig 8.
OP9 displacement can be utilized to test new drugs for BCP-ALL treatments.
(A) Impedance profile of three BCP-ALL cell lines treated with the JAK2 inhibitor ruxolitinib (rux) (n = 3). Dotted lines show the time points used for the calculation of % displacement in panel B (also see S4 Fig) (B) Percentage impedance signal displacement from treatments with the indicated drug concentrations (n = 3). Statistical analysis was done using one-way ANOVA followed by Dunnett’s multiple comparison test to compare treatment samples versus the OP9+ALL+DMSO control wells. Error bars represent standard deviation. p values relative to control: *<0.05; **<0.01; ***<0.001; ****<0.0001.