My watch list
my.bionity.com  
Login  

Dynamics of Kinases Involved in Cell Growth Using the xCELLigence System

Jitao Zhang, Angelika Duda, Christian Schmidt, Anja Irsigler, Stefan Wiemann, and Ulrich Tschulena*

Deutsches Krebsforschungszentrum, Division Molecular Genome Analysis, Heidelberg, Germany

*Corresponding author

Introduction

Studies investigating endogenous cellular regulatory mechanisms have substantially pushed the application of high-throughput screens. RNA interference has developed into an extremely powerful tool for mediating the loss-of-function of specific mRNA molecules and preventing their translation into proteins. Numerous studies have helped to uncover novel gene functions in many biological processes, but most have used single endpoint analysis as readout for the respective phenotypes to characterize the outcome of a cascade of regulatory mechanisms.

The xCELLigence System offers a new way to analyze the effect of gene knock downs on living cells, as the system provides continuous, quantitative information about the electrical impedance at the bottom surface of microtiter plates. Thus, any change in cell number, cell morphology or cell attachment can be detected in real-time.

We integrated the Real-Time Cell Analyzer (RTCA) into a high throughput workflow for transfection, analyzed the effect of  human kinases on cell growth in real-time, and monitored the dynamics of the cellular response to decreasing levels of the respective kinases. To this end, we performed a reverse genetic loss-of-function screen with a small interfering RNA (siRNA) library representing 160 kinases. The obtained real-time signatures of the cells provide direct insights into the dynamics of the involved kinases and their impact on cells.

Materials and Methods

E-Plate background measurement and seeding

The background measurement of xCELLigence E-Plates 96 was performed with 100 µl DMEM medium containing penicillin/streptomycin, L-glutamine, and 10% FCS using the standard protocol provided in the software. Following trypsination, cell concentration was determined in a CASY-TT CellCounter and 10,000 HeLa cells were seeded into every well with 100 µl additional DMEM medium.

Automated siRNA-Transfection

Transfection of siRNAs was carried out in a 96-well format. Prior to transfection, DMEM medium was removed from the E-plate’s with the 96-well pipetting head of the liquid handler without touching the cells at the surface of the well bottom. HeLa Cells were washed once with 150 µl Optimem before adding 40 µl Optimem to each well. In parallel, 1.15 µl X-tremeGENE and siRNA were diluted in 20 µl Optimem each, then mixed and incubated for 15 minutes before being added to the E-Plates 96 by the liquid handler’s 96-well pipetting head, leading to a final volume of 80 µl per well and a final siRNA concentration of 60 nM. After 5 hours of incubation in the RTCA MP Station, the transfection mix was removed with the liquid handler; cells were washed with 150 µl of DMEM medium and then filled with 200 µl of DMEM medium before returning the plates to the RTCA MP Station. Cells were then incubated for 90 hours and impedance was measured every 15 minutes for 25 hours. Thereafter measurements were taken at 60-minute interval

Results and Discussion

To identify kinases having an influence on cell growth, we screened a siRNA library consisting of pools of four siRNAs per gene at 60 nM concentration, targeting 160 human kinases. The screen was done in triplicate for all kinases using a RTCA MP Station (Figure 1). As 80 siRNAs and 5 control siRNAs (three replicates on each plate) were screened on one E-Plate 96, all six reading stations of the RTCA MP Station were used in parallel. HeLa cells were transfected in an automated way using a Biomek FXP liquid handling workstation (Figure 2). DMEM growth medium was exchanged with Optimem transfection medium and transfections were carried out using X-tremeGENE siRNA transfection reagent. Five hours later, medium was exchanged again with DMEM growth medium. Impedance was measured every 15 minutes for the first 25 hours following the second medium change, then measurements were continued at 60-minute intervals. To exclude effects on the cells by the medium changes, we normalized values to a timepoint 5 hours after the second medium exchange.


As negative controls, we used mock transfections without siRNA and siAllstar siRNA, which is a negative control siRNA that has no homology to any known mammalian gene. Positive controls were siRNAs targeting WEE1, COPB2 and PLK1, as knock down of these genes is known to induce apoptosis in HeLa cells. These controls were placed in every plate to monitor the reproducibility of the measurements and to permit a plate-to-plate comparison of the data. Different patterns of cell growth were observed for the different controls. Mock-treated and siAll­star-transfected cells showed at first a short reduction in cell index (CI) values after transfection, possibly due to toxic effects of the transfection process, but then a recovery and a steady increase in CI until 90 hours after transfection (Figure 3).

In contrast, siRNAs targeting WEE1, PLK1 and COPB2 all induced a reduction in CI to a value of 0.5 normalized CI and then showed an increase in CI values again (Figure 3). Interestingly, the knockdown of these three genes induced different kinetics for the reduction of CI values, with samples transfected with siRNA targeting COPB2 showing the latest onset of phenotype. This effect might be due to different half-life times of the proteins or to a cascade of reactions leading to a delayed phenotype after the knockdown, which might indicate a more indirect effect. The reproducibility of the controls was extremely high; all replicates showed almost identical curves and had only very little variation with the average coefficient of variance (CV, defined by CV=s/µ) being below 0.05 across all analyzed time points for all controls.

We then analyzed the samples that had been transfected with the siRNA library. 36 of the 160 siRNAs did not induce a significant effect when compared t with mock- or Allstar-control-transfected samples. However, knockdown of several genes induced significant alterations in the growth curve. As expected, knockdown of many (i.e., 84 of 160 kinases induced an inhibitory effect on cell growth). Interestingly, the dynamics of CI changes turned out to be important when comparing the knockdown of different kinases, for example the samples treated with siRNAs targeting CDC2L1, CDC2L2 or CDK10 (Figure 5). The siRNA targeting CDK10 induces a block in growth for the first 35 hours after transfection; however, thereafter the CI increases again, indicating a recovery of proliferation.

In contrast, cells transfected with siRNAs targeting CDC2L1 and CDC2L2 initially showed a normal increase in CI values being indicative of a normal proliferation curve. However, 65 hours after seeding, CDC2L1 siRNA transfected samples show a peak and then a decrease in the CI values, whereas the CDC2L2 siRNA transfected samples showed a peak at 80 hours after seeding before a decline in CI values was apparent. These data are the first indication of different kinetics of the dynamic processes that are involved in RNAi and in the activities of different kinases. Further studies analyzing kinetics of mRNA and protein concentration after siRNA knockdown are needed to elucidate a more detailed picture of kinase dynamics.

These data demonstrate the power of real-time measurements: (a) Differences in the onset and dynamics of different conditions become visible and (b) when comparing to end-point analysis such as WST-1 assays that are commonly performed at a fixed time point, the dynamic response of the cells to the knockdown of the genes would lead to different results and conclusions at any selected time of the end-point analysis (Figure 5). We also detected other examples of inhibitory effects on cell growth. For example, knockdown of Bmx/Etk induces an inhibition of CI increase, when compared with control cells transfected with siAllstar (Figure 4). As it has been described that Etk can have an effect on cell proliferation, this decrease in CI values has been expected [1]. Also, EphB6 knockdown induced a reduction in CI values, indicating an increased proliferation or reduced apoptosis (Figure 4). An influence of EphB6 knockout on T-cell proliferation has already been described [2]. Thus, these effects might also play a role in HeLa cells in our experiments. In contrast, the siRNA targeting EphA4 induced an increase in CI when compared with Allstar-control-transfected samples, indicating an activating effect on cell proliferation (Figure 4).

Conclusions

Our experiments demonstrate that the xCELLigence System can be combined well with reverse-genetic screening experiments using an automated robotic pipeline such as the Biomek FXP liquid handling workstation. The results show a high reproducibility of the system, with the average coefficient of variance being under 0.05. Moreover, the experiments demonstrate the utility of the xCELLigence System to monitor dynamic effects after knockdown experiments and the clear superiority of real-time measurements over end-point analysis. The growth curves generated by transfected cells with different siRNAs showed specific signatures which may lead to a better understanding of the analyzed genes and proteins.

References

1.   Chau CH et al. (2002) Oncogene 21:8817–8829

2.   Luo H et al. (2004) J Clin Inves 114:1762–1773

This article was originally published in Biochemica 2/2009, pages 14-17. ©Springer Medizin Verlag 2009

More about Roche Diagnostics
Your browser is not current. Microsoft Internet Explorer 6.0 does not support some functions on Chemie.DE