High-Throughput Cell Cycle Analysis Using a Microplate Cytometer

The traditional method of determining the cell cycle phase of individual cells is by quantifying their total DNA content using flow cytometry. For screening purposes, however, flow cytometry has several shortcomings, most notably, low throughput, the requirement for a large number of cells, and the inability to analyze adherent cell lines in situ.

With read times enabling over 2000 wells to be analyzed per hour, the rapid whole-well scanning of an Acumen eX3 microplate cytometer (TTP Labtech Ltd., Melbourn Hertfordshire, U.K.) provides true high-throughput cell cycle analysis capability. This is demonstrated through the use of both permeabilized and live cell populations labeled with a range of DNA stains.

Experimental protocol

Figure 1 - Optical configuration of Acumen eX3 microplate cytometer.

For this study, a triple-laser Acumen eX3 (Figure 1) was used for the determination of cell cycle analysis.

DNA dyes

Cell cycle analysis is typically performed on fixed, permeabilized cells using a cell-impermeant nucleic acid stain. However, to analyze live cells, a cell-permeant nucleic acid stain is required. For fixed-cell protocols, the most commonly used DNA dye is propidium iodide (PI).1 While the choices for fixed-cell staining are varied, there are only a few examples of useful cell-permeant nucleic acid stains. The application of microplate cytometry for high-throughput cell cycle analysis using propidium iodide has been demonstrated previously.2 The performance of a range of DNA dyes using all three lasers in an Acumen eX3 was compared.

In this study, five different fluorescent DNA dyes were used, which were suitable for use with one of the Acumen eX3’s lasers. Dyes included Hoechst 34580 (405 nm), Vybrant® DyeCycle™Orange (488 nm) (Molecular Probes [Invitrogen], Eugene, OR), PI (488 nm), TO-PRO-3 (633 nm) (Molecular Probes), and DRAQ5™ (633 nm) (Biostatus Ltd., Leicestershire, U.K.).

Cell cycle arrest protocol

For all assays, HeLa cells were seeded into a 384-well microplate at a density of 2000 cells per well. Cells were incubated at 37 °C/5% CO2 for 18 hr prior to treatment with vehicle or vinblastine for 22 hr at 37 °C/5% CO2.

Fixed-cell staining

For fixed-cell studies, the following fixation protocol was used. The medium was carefully aspirated from each well, and 50 μL of –20 °C 80% ethanol in phosphate-buffered saline (PBS) was added to each well. Cells were incubated at –20 oC for 30 min. Each well was washed twice with 100 μL PBS. After the second wash, the PBS was aspirated off, and 50 μL of a 0.2-mg/mL RNAse solution (DNase free) in PBS was added to each well. The plate was incubated for 1 hr at 37 °C. The RNAse solution was aspirated off the wells, and either 50 μL of a 10-μM PI solution in PBS solution, 50 μL of a 10 μM Hoechst 34580 solution, or 50 μL of a 0.5-μM TO-PRO-3 solution was added to relevant wells. The plate was sealed with a black cover-seal and incubated in the dark for 15 min at room temperature. The plate was loaded into an Acumen eX3 microplate cytometer.

Live cell staining

The live cell stains offered a more simplified protocol as the dyes were added directly to the treated cells. Following incubation with vinblastine for 22 hr as described above, 5 μL of 50 μM DRAQ5 or 5 μL of 50 μM Vybrant DyeCycle Orange was added directly to each well. The plate was sealed with a black cover-seal and incubated in the dark for 30 min at room temperature. Analysis was performed on an Acumen eX3 microplate cytometer using 488-nm or 633-nm excitation.

Plate scanning

The plates were scanned at a sampling resolution of 1 μm in the X direction and 8 μm in the Y direction. The photomultiplier tube (PMT) voltage for each channel was determined for each dye used. Cells were defined as having a width and depth of between 10 and 100 μm. The phases of cell cycle were classified using the Total Intensity object characteristic.

Results and discussion

Figure 2 - Comparison of the fluorescent profiles of a G1 and G2/M nucleus.

Figure 3 - Vinblastine arrest of cells into the G2/M phase of the cell cycle in HeLa cells. The red population indicates cells in G1 phase of the cell cycle; the blue population indicates cells in G2/M phase of the cell cycle.

Figure 4 - Comparison of different DNA stains for cell cycle excited at 405, 488, and 633 nm. a) Propidium iodide, b) Hoechst 34580, c) Vybrant DyeCycle Orange, and d) TO-PRO-3. The red population shows cells in G1; the blue population shows cells in G2/M.

It is possible to generate three-dimensional models of fluorescent objects, including nuclei. This permits visual determination of the number of nuclei present in each cell. The multiple sampling also means that integration of fluorescence intensity is not affected by the presence of multiple nuclei, leading to well-defined peaks for cells in G1 and G2/M phases, as shown in Figure 2. From a histogram showing total dye intensity in the nucleus, the relative amounts of DNA in each nucleus can be determined. Using this parameter, all of the single cells in the well were classified into the different phases of the cell cycle. Figure 3 shows a defined shift of the cells from G1 into G2/M phase upon cell cycle arrest by 10 μM vinblastine treatment for 22 hr. By calculating the total cell count across the whole area of the well, mathematical models were then used to calculate the percentage of cells occupying the different phases of the cell cycle from G1, S, and G2/M.

By using this method of calculating dye intensity and relating it to DNA content, the use of different fluorescent dyes on cell cycle analysis was examined. In addition to PI, cells were also stained with Hoechst 34580, Vybrant DyeCycle Orange, and TO-PRO-3. The analysis shows good correlation between all four dyes, giving typical DNA histograms under control conditions as shown in Figure 4.

Live cell stains

Live cell dyes are less frequently used in cell cycle analysis. The main advantages of using live cell dyes is that the assay protocol is relatively much simpler, predominantly since it does not require a fixation step or, in the case of PI, an RNAse treatment step. DRAQ5 was used as a 633-nm excitable DNA stain. DRAQ5 has a far-red emission and has previously been used in DNA-specific, stoichiometric cell cycle analyses. Because it is a far-red dye, it has no overlap with green fluorescent protein/fluorescein isothiocyanate (GFP/FITC) and there is no need for it to be washed out, making it well suited for a homogeneous one-step assay.

Figure 5 - Comparison of the 633-nm excitable dye, DRAQ5, to other DNA stains.

The concentration dependence of vinblastine against the percentage of cells in G1 and G2/M phase of the cell cycle was determined. As can be seen in Figure 5, it was found that DRAQ5 gave comparable responses to those obtained with the more established dyes, PI and Hoechst 34580. In addition to the concentration response curves, the effect of using nocozadole to induce G2/M arrest was also investigated, and the results demonstrate that the effects were similar to those obtained using vinblastine.


Summary

The utility of microplate cytometry for cell cycle compound profiling has been demonstrated using as standard agents vinblastine and nocozadole, which arrested HeLa cells in the expected phase of the cell cycle. The ability to use any of the Acumen eX3’s lasers for cell cycling illustrates that there is a large degree of flexibility to use the remaining lasers for multiplexing cell cycle studies with a secondary assay such as mitotic index.

The results demonstrate that there was no real difference in using either the cell-impermeant nucleic acid stains, PI and TO-PRO-3, or the permanent nucleic acidic stains, Hoechst 34580 and DRAQ5. For assays determining cell cycle analysis only, where the protocol is much easier than fixation steps, there is no negative impact when using live cell stains.

For screening purposes, the throughput of a microplate cytometer for cell cycle analysis is unparalleled, since it is able to analyze in a few hours what normally takes a week on a flow cytometer.3 Because all the cell processing is performed with microplates, it is also more amenable to automation. The novel design features of the Acumen eX3 microplate cytometer used in these studies permit multiplex, whole-well analysis at high read times compatible with primary screening campaigns, with daily throughputs of 30,000 compounds per day being reported for some assays. For cell cycle analysis, automated throughput of 384 samples in 10 min has been achieved.

In summary, microplate cytometry offers a high-content, high-throughput approach to cell cycle analysis that can eliminate current bottlenecks in drug discovery screening campaigns.

References

  1. Crissman, H.A.; Steinkamp, J.A. Rapid, simultaneous measurement of DNA, protein, and cell volume in single cells from large mammalian cell populations. J. Cell Biol. 1973, 59, 766–71.
  2.  Bowen, W.P.; Wylie, P.G. Application of laser-scanning fluorescence microplate cytometry in high content screening. Assay and Drug Development Technol.2006, 4(2), 209–21.
  3. Kittler, R.; Pelletier, L.; Heninger, A.-K.; Slabicki, M.; Theis, M.; Miroslaw, L.; Poser, I.; Lawo, S.; Grabner, H.; Kozak, K.; Wagner, J.; Surendranath, V.; Richter, C.; Bowen, W.; Jackson, A.L.; Habermann, B.; Hyman, A.A.; Buchholz, F. Genome-scale RNAi profiling of cell division in human tissue culture cells. Nature Cell Biol. 2007, 9(12), 1401–12.

    Dr. Wylie is Product Manager, TTP Labtech Ltd., Melbourn Science Park, Melbourn Hertfordshire SG8 6EE, U.K.; tel.: +44 1763 262626; fax: +44 1763 261964; e-mail: [email protected].